diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,428535 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 61215,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3445475101470947,
+      "learning_rate": 0.0005999999996049288,
+      "loss": 10.4731,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.277515411376953,
+      "learning_rate": 0.0005999999984197153,
+      "loss": 9.9074,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.00109601020813,
+      "learning_rate": 0.0005999999964443598,
+      "loss": 9.3624,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.896841287612915,
+      "learning_rate": 0.0005999999936788619,
+      "loss": 8.9861,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8488935232162476,
+      "learning_rate": 0.0005999999901232217,
+      "loss": 8.6725,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8929357528686523,
+      "learning_rate": 0.0005999999857774393,
+      "loss": 8.2981,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8532694578170776,
+      "learning_rate": 0.0005999999806415148,
+      "loss": 8.0649,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7290151119232178,
+      "learning_rate": 0.0005999999747154479,
+      "loss": 7.7909,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5064324140548706,
+      "learning_rate": 0.0005999999679992389,
+      "loss": 7.5804,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3336801528930664,
+      "learning_rate": 0.0005999999604928875,
+      "loss": 7.4451,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1641117334365845,
+      "learning_rate": 0.0005999999521963943,
+      "loss": 7.2529,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0922843217849731,
+      "learning_rate": 0.0005999999431097587,
+      "loss": 6.9598,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9296258091926575,
+      "learning_rate": 0.0005999999332329811,
+      "loss": 6.7569,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8542349934577942,
+      "learning_rate": 0.0005999999225660615,
+      "loss": 6.7468,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8458271622657776,
+      "learning_rate": 0.0005999999111089997,
+      "loss": 6.6628,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6604408621788025,
+      "learning_rate": 0.0005999998988617959,
+      "loss": 6.7246,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6733983755111694,
+      "learning_rate": 0.00059999988582445,
+      "loss": 6.5789,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5909510850906372,
+      "learning_rate": 0.0005999998719969623,
+      "loss": 6.5809,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7529872059822083,
+      "learning_rate": 0.0005999998573793326,
+      "loss": 6.4471,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5217795968055725,
+      "learning_rate": 0.000599999841971561,
+      "loss": 6.5395,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6023449897766113,
+      "learning_rate": 0.0005999998257736476,
+      "loss": 6.5889,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.570929229259491,
+      "learning_rate": 0.0005999998087855924,
+      "loss": 6.3548,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62796950340271,
+      "learning_rate": 0.0005999997910073954,
+      "loss": 6.3005,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5817890167236328,
+      "learning_rate": 0.0005999997724390567,
+      "loss": 6.4104,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5530886650085449,
+      "learning_rate": 0.0005999997530805763,
+      "loss": 6.349,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.643932580947876,
+      "learning_rate": 0.0005999997329319543,
+      "loss": 6.3305,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5512265563011169,
+      "learning_rate": 0.0005999997119931908,
+      "loss": 6.3231,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6873916387557983,
+      "learning_rate": 0.0005999996902642858,
+      "loss": 6.1347,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7107514142990112,
+      "learning_rate": 0.0005999996677452392,
+      "loss": 6.3309,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6492154002189636,
+      "learning_rate": 0.0005999996444360514,
+      "loss": 6.1696,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5806317925453186,
+      "learning_rate": 0.0005999996203367222,
+      "loss": 6.1405,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6508877873420715,
+      "learning_rate": 0.0005999995954472518,
+      "loss": 6.1023,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7064071893692017,
+      "learning_rate": 0.00059999956976764,
+      "loss": 6.0525,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6533799767494202,
+      "learning_rate": 0.0005999995432978872,
+      "loss": 6.2293,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5126307606697083,
+      "learning_rate": 0.0005999995160379933,
+      "loss": 6.0087,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6564686298370361,
+      "learning_rate": 0.0005999994879879586,
+      "loss": 6.1167,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6836735606193542,
+      "learning_rate": 0.0005999994591477828,
+      "loss": 5.8947,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6649318337440491,
+      "learning_rate": 0.0005999994295174661,
+      "loss": 6.0403,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6308470368385315,
+      "learning_rate": 0.0005999993990970087,
+      "loss": 6.0636,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5832207202911377,
+      "learning_rate": 0.0005999993678864107,
+      "loss": 6.0658,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5933493971824646,
+      "learning_rate": 0.000599999335885672,
+      "loss": 6.07,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6838207244873047,
+      "learning_rate": 0.0005999993030947929,
+      "loss": 5.8559,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6562783718109131,
+      "learning_rate": 0.0005999992695137733,
+      "loss": 5.7866,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6013045310974121,
+      "learning_rate": 0.0005999992351426134,
+      "loss": 5.631,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.579399824142456,
+      "learning_rate": 0.0005999991999813132,
+      "loss": 6.0032,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6205735206604004,
+      "learning_rate": 0.0005999991640298728,
+      "loss": 5.4523,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5519394874572754,
+      "learning_rate": 0.0005999991272882925,
+      "loss": 5.6271,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5361047983169556,
+      "learning_rate": 0.0005999990897565721,
+      "loss": 6.0044,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6790111064910889,
+      "learning_rate": 0.0005999990514347119,
+      "loss": 5.683,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6192212700843811,
+      "learning_rate": 0.0005999990123227119,
+      "loss": 5.8079,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6024751663208008,
+      "learning_rate": 0.0005999989724205721,
+      "loss": 5.7658,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7180526852607727,
+      "learning_rate": 0.000599998931728293,
+      "loss": 5.7054,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7811439037322998,
+      "learning_rate": 0.0005999988902458743,
+      "loss": 5.7338,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.678833544254303,
+      "learning_rate": 0.0005999988479733163,
+      "loss": 5.7092,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.633277952671051,
+      "learning_rate": 0.0005999988049106191,
+      "loss": 5.6259,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4287935495376587,
+      "learning_rate": 0.0005999987610577827,
+      "loss": 5.6371,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7538682818412781,
+      "learning_rate": 0.0005999987164148074,
+      "loss": 5.7429,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5953795909881592,
+      "learning_rate": 0.0005999986709816932,
+      "loss": 5.7226,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7172217965126038,
+      "learning_rate": 0.0005999986247584401,
+      "loss": 5.5155,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7222607135772705,
+      "learning_rate": 0.0005999985777450485,
+      "loss": 5.4241,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7364643216133118,
+      "learning_rate": 0.0005999985299415183,
+      "loss": 5.5181,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7325772047042847,
+      "learning_rate": 0.0005999984813478497,
+      "loss": 5.6243,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6013771891593933,
+      "learning_rate": 0.000599998431964043,
+      "loss": 5.5167,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5635715126991272,
+      "learning_rate": 0.000599998381790098,
+      "loss": 5.7081,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6563093662261963,
+      "learning_rate": 0.000599998330826015,
+      "loss": 5.4648,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5509313941001892,
+      "learning_rate": 0.0005999982790717942,
+      "loss": 5.4311,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7320454716682434,
+      "learning_rate": 0.0005999982265274357,
+      "loss": 5.5097,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6863231062889099,
+      "learning_rate": 0.0005999981731929395,
+      "loss": 5.5005,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.543449342250824,
+      "learning_rate": 0.0005999981190683059,
+      "loss": 5.5048,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5858511328697205,
+      "learning_rate": 0.000599998064153535,
+      "loss": 5.5506,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7430124878883362,
+      "learning_rate": 0.0005999980084486269,
+      "loss": 5.2905,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5860936045646667,
+      "learning_rate": 0.0005999979519535819,
+      "loss": 5.5213,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5965762138366699,
+      "learning_rate": 0.0005999978946683999,
+      "loss": 5.1871,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6979895234107971,
+      "learning_rate": 0.0005999978365930813,
+      "loss": 5.3346,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7160101532936096,
+      "learning_rate": 0.000599997777727626,
+      "loss": 5.1066,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5626764297485352,
+      "learning_rate": 0.0005999977180720344,
+      "loss": 5.459,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6080682277679443,
+      "learning_rate": 0.0005999976576263064,
+      "loss": 5.6085,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6767348647117615,
+      "learning_rate": 0.0005999975963904425,
+      "loss": 5.4253,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5835095047950745,
+      "learning_rate": 0.0005999975343644425,
+      "loss": 5.3682,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5829843282699585,
+      "learning_rate": 0.0005999974715483068,
+      "loss": 5.2841,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6556689739227295,
+      "learning_rate": 0.0005999974079420355,
+      "loss": 5.3828,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5656118392944336,
+      "learning_rate": 0.0005999973435456287,
+      "loss": 5.4779,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.553550660610199,
+      "learning_rate": 0.0005999972783590868,
+      "loss": 5.4759,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5812600255012512,
+      "learning_rate": 0.0005999972123824096,
+      "loss": 5.3517,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6190153956413269,
+      "learning_rate": 0.0005999971456155976,
+      "loss": 5.4019,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5636712312698364,
+      "learning_rate": 0.0005999970780586508,
+      "loss": 5.2974,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.546588659286499,
+      "learning_rate": 0.0005999970097115694,
+      "loss": 5.4874,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6141658425331116,
+      "learning_rate": 0.0005999969405743537,
+      "loss": 5.431,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5903357267379761,
+      "learning_rate": 0.0005999968706470037,
+      "loss": 5.4673,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5161350965499878,
+      "learning_rate": 0.0005999967999295199,
+      "loss": 5.2426,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6005253791809082,
+      "learning_rate": 0.000599996728421902,
+      "loss": 5.4007,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5921294093132019,
+      "learning_rate": 0.0005999966561241506,
+      "loss": 5.3612,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5736501812934875,
+      "learning_rate": 0.0005999965830362657,
+      "loss": 5.327,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5713847279548645,
+      "learning_rate": 0.0005999965091582475,
+      "loss": 5.2216,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.583519697189331,
+      "learning_rate": 0.0005999964344900962,
+      "loss": 5.101,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6061059236526489,
+      "learning_rate": 0.0005999963590318122,
+      "loss": 5.3313,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.605904221534729,
+      "learning_rate": 0.0005999962827833954,
+      "loss": 5.3037,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5677119493484497,
+      "learning_rate": 0.0005999962057448462,
+      "loss": 5.2484,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6556852459907532,
+      "learning_rate": 0.0005999961279161646,
+      "loss": 5.3307,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6364889144897461,
+      "learning_rate": 0.000599996049297351,
+      "loss": 5.3118,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6421851515769958,
+      "learning_rate": 0.0005999959698884056,
+      "loss": 5.3212,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5902762413024902,
+      "learning_rate": 0.0005999958896893285,
+      "loss": 5.2373,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5422645211219788,
+      "learning_rate": 0.00059999580870012,
+      "loss": 5.3965,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6656525731086731,
+      "learning_rate": 0.0005999957269207802,
+      "loss": 5.3206,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6076167821884155,
+      "learning_rate": 0.0005999956443513094,
+      "loss": 5.2297,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5958263278007507,
+      "learning_rate": 0.0005999955609917078,
+      "loss": 5.2308,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6043804287910461,
+      "learning_rate": 0.0005999954768419757,
+      "loss": 5.3056,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5793024301528931,
+      "learning_rate": 0.0005999953919021132,
+      "loss": 5.1175,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6292415857315063,
+      "learning_rate": 0.0005999953061721206,
+      "loss": 5.2404,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.539836585521698,
+      "learning_rate": 0.0005999952196519981,
+      "loss": 5.3884,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5717549324035645,
+      "learning_rate": 0.0005999951323417458,
+      "loss": 5.172,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5996467471122742,
+      "learning_rate": 0.0005999950442413642,
+      "loss": 5.1867,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6268661022186279,
+      "learning_rate": 0.0005999949553508534,
+      "loss": 5.2663,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6602498888969421,
+      "learning_rate": 0.0005999948656702136,
+      "loss": 5.0668,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5790102481842041,
+      "learning_rate": 0.0005999947751994451,
+      "loss": 5.1133,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6533108949661255,
+      "learning_rate": 0.000599994683938548,
+      "loss": 5.1798,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6445215344429016,
+      "learning_rate": 0.0005999945918875227,
+      "loss": 5.1873,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5656430125236511,
+      "learning_rate": 0.0005999944990463693,
+      "loss": 5.0912,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5833689570426941,
+      "learning_rate": 0.0005999944054150882,
+      "loss": 5.1466,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6031045317649841,
+      "learning_rate": 0.0005999943109936796,
+      "loss": 5.1367,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6581068634986877,
+      "learning_rate": 0.0005999942157821436,
+      "loss": 5.1474,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5878651738166809,
+      "learning_rate": 0.0005999941197804806,
+      "loss": 5.116,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6598708629608154,
+      "learning_rate": 0.0005999940229886909,
+      "loss": 5.2586,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6158697605133057,
+      "learning_rate": 0.0005999939254067746,
+      "loss": 5.2573,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.624995231628418,
+      "learning_rate": 0.0005999938270347321,
+      "loss": 5.2498,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5899373292922974,
+      "learning_rate": 0.0005999937278725636,
+      "loss": 5.2714,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5740645527839661,
+      "learning_rate": 0.0005999936279202693,
+      "loss": 5.1358,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.609225332736969,
+      "learning_rate": 0.0005999935271778495,
+      "loss": 5.2404,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6659890413284302,
+      "learning_rate": 0.0005999934256453046,
+      "loss": 4.9877,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5626158714294434,
+      "learning_rate": 0.0005999933233226347,
+      "loss": 5.2386,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5483737587928772,
+      "learning_rate": 0.0005999932202098401,
+      "loss": 5.2031,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6274495124816895,
+      "learning_rate": 0.0005999931163069211,
+      "loss": 5.1119,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6077089309692383,
+      "learning_rate": 0.000599993011613878,
+      "loss": 4.9585,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5846816301345825,
+      "learning_rate": 0.000599992906130711,
+      "loss": 5.2099,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6238222718238831,
+      "learning_rate": 0.0005999927998574204,
+      "loss": 5.0893,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.720881462097168,
+      "learning_rate": 0.0005999926927940066,
+      "loss": 5.0845,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4975273013114929,
+      "learning_rate": 0.0005999925849404696,
+      "loss": 5.0894,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6107990741729736,
+      "learning_rate": 0.00059999247629681,
+      "loss": 5.0879,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6482533812522888,
+      "learning_rate": 0.000599992366863028,
+      "loss": 5.1011,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5770416855812073,
+      "learning_rate": 0.0005999922566391237,
+      "loss": 4.9711,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5987694263458252,
+      "learning_rate": 0.0005999921456250976,
+      "loss": 5.2741,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5902721285820007,
+      "learning_rate": 0.00059999203382095,
+      "loss": 4.9697,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5815165042877197,
+      "learning_rate": 0.000599991921226681,
+      "loss": 5.0085,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6174842119216919,
+      "learning_rate": 0.000599991807842291,
+      "loss": 5.1251,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5789567232131958,
+      "learning_rate": 0.0005999916936677804,
+      "loss": 5.0212,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5186668634414673,
+      "learning_rate": 0.0005999915787031493,
+      "loss": 5.0934,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5524886250495911,
+      "learning_rate": 0.0005999914629483983,
+      "loss": 5.0444,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6883447766304016,
+      "learning_rate": 0.0005999913464035273,
+      "loss": 5.0623,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5598011016845703,
+      "learning_rate": 0.000599991229068537,
+      "loss": 5.1861,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6614800691604614,
+      "learning_rate": 0.0005999911109434274,
+      "loss": 4.9011,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5990419983863831,
+      "learning_rate": 0.0005999909920281991,
+      "loss": 4.9618,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5454047322273254,
+      "learning_rate": 0.0005999908723228522,
+      "loss": 5.1169,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6134856939315796,
+      "learning_rate": 0.000599990751827387,
+      "loss": 5.0239,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6012156009674072,
+      "learning_rate": 0.000599990630541804,
+      "loss": 5.24,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5873432755470276,
+      "learning_rate": 0.0005999905084661033,
+      "loss": 5.1557,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6696448922157288,
+      "learning_rate": 0.0005999903856002855,
+      "loss": 5.0485,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5300459265708923,
+      "learning_rate": 0.0005999902619443507,
+      "loss": 5.0126,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5879744291305542,
+      "learning_rate": 0.0005999901374982992,
+      "loss": 4.8251,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6002203822135925,
+      "learning_rate": 0.0005999900122621314,
+      "loss": 4.8551,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.584845244884491,
+      "learning_rate": 0.0005999898862358478,
+      "loss": 5.0479,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7027909755706787,
+      "learning_rate": 0.0005999897594194485,
+      "loss": 5.1492,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6529715061187744,
+      "learning_rate": 0.0005999896318129339,
+      "loss": 5.0112,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7082669734954834,
+      "learning_rate": 0.0005999895034163043,
+      "loss": 4.8184,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6252927184104919,
+      "learning_rate": 0.0005999893742295602,
+      "loss": 4.8673,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6685312390327454,
+      "learning_rate": 0.0005999892442527017,
+      "loss": 4.9796,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6606252193450928,
+      "learning_rate": 0.0005999891134857293,
+      "loss": 5.0516,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6550689339637756,
+      "learning_rate": 0.0005999889819286433,
+      "loss": 4.9988,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6406249403953552,
+      "learning_rate": 0.000599988849581444,
+      "loss": 5.2205,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5724087953567505,
+      "learning_rate": 0.0005999887164441319,
+      "loss": 4.8884,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6238783597946167,
+      "learning_rate": 0.0005999885825167073,
+      "loss": 5.0775,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6989482641220093,
+      "learning_rate": 0.0005999884477991703,
+      "loss": 4.9027,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5641810894012451,
+      "learning_rate": 0.0005999883122915216,
+      "loss": 5.0401,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5841800570487976,
+      "learning_rate": 0.0005999881759937614,
+      "loss": 4.9143,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6667760610580444,
+      "learning_rate": 0.0005999880389058902,
+      "loss": 4.7318,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5549502372741699,
+      "learning_rate": 0.0005999879010279082,
+      "loss": 4.9401,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.562578022480011,
+      "learning_rate": 0.0005999877623598156,
+      "loss": 4.8758,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6487370729446411,
+      "learning_rate": 0.0005999876229016132,
+      "loss": 4.8428,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5677328705787659,
+      "learning_rate": 0.000599987482653301,
+      "loss": 4.94,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5713781714439392,
+      "learning_rate": 0.0005999873416148795,
+      "loss": 4.9779,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6296773552894592,
+      "learning_rate": 0.0005999871997863491,
+      "loss": 5.0549,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.609196126461029,
+      "learning_rate": 0.0005999870571677102,
+      "loss": 4.8211,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5543152689933777,
+      "learning_rate": 0.0005999869137589631,
+      "loss": 4.9272,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6066011190414429,
+      "learning_rate": 0.0005999867695601082,
+      "loss": 4.9566,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5656384229660034,
+      "learning_rate": 0.0005999866245711458,
+      "loss": 4.9814,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6074258685112,
+      "learning_rate": 0.0005999864787920765,
+      "loss": 4.723,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5915899872779846,
+      "learning_rate": 0.0005999863322229005,
+      "loss": 4.6722,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5853182673454285,
+      "learning_rate": 0.0005999861848636183,
+      "loss": 4.8963,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5973837375640869,
+      "learning_rate": 0.00059998603671423,
+      "loss": 4.9082,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5674445033073425,
+      "learning_rate": 0.0005999858877747365,
+      "loss": 4.8848,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6249287128448486,
+      "learning_rate": 0.0005999857380451376,
+      "loss": 4.7804,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5839101076126099,
+      "learning_rate": 0.0005999855875254343,
+      "loss": 4.9382,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6714131236076355,
+      "learning_rate": 0.0005999854362156265,
+      "loss": 4.9891,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6222569942474365,
+      "learning_rate": 0.0005999852841157148,
+      "loss": 4.8221,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.671981692314148,
+      "learning_rate": 0.0005999851312256997,
+      "loss": 4.7026,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5763186812400818,
+      "learning_rate": 0.0005999849775455815,
+      "loss": 4.816,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5959866642951965,
+      "learning_rate": 0.0005999848230753605,
+      "loss": 4.9962,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6573556661605835,
+      "learning_rate": 0.0005999846678150372,
+      "loss": 4.8953,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6091251373291016,
+      "learning_rate": 0.0005999845117646122,
+      "loss": 4.9322,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6092210412025452,
+      "learning_rate": 0.0005999843549240856,
+      "loss": 5.0708,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6042549014091492,
+      "learning_rate": 0.000599984197293458,
+      "loss": 4.8596,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6179243326187134,
+      "learning_rate": 0.0005999840388727296,
+      "loss": 4.9343,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.558975875377655,
+      "learning_rate": 0.0005999838796619012,
+      "loss": 4.736,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6428976058959961,
+      "learning_rate": 0.0005999837196609727,
+      "loss": 4.7473,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6300376057624817,
+      "learning_rate": 0.0005999835588699451,
+      "loss": 4.5575,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6382180452346802,
+      "learning_rate": 0.0005999833972888184,
+      "loss": 4.649,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6212637424468994,
+      "learning_rate": 0.0005999832349175932,
+      "loss": 4.9914,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5995153188705444,
+      "learning_rate": 0.0005999830717562699,
+      "loss": 4.7988,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6228866577148438,
+      "learning_rate": 0.0005999829078048489,
+      "loss": 4.7471,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5948235988616943,
+      "learning_rate": 0.0005999827430633307,
+      "loss": 4.7403,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5553930401802063,
+      "learning_rate": 0.0005999825775317156,
+      "loss": 4.7738,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58103346824646,
+      "learning_rate": 0.0005999824112100042,
+      "loss": 4.8885,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6065656542778015,
+      "learning_rate": 0.0005999822440981968,
+      "loss": 4.7514,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6187020540237427,
+      "learning_rate": 0.000599982076196294,
+      "loss": 4.7373,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6189316511154175,
+      "learning_rate": 0.0005999819075042961,
+      "loss": 4.7754,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6775714159011841,
+      "learning_rate": 0.0005999817380222035,
+      "loss": 4.7958,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6709480285644531,
+      "learning_rate": 0.0005999815677500168,
+      "loss": 4.6884,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5949779152870178,
+      "learning_rate": 0.0005999813966877363,
+      "loss": 4.7673,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6341105699539185,
+      "learning_rate": 0.0005999812248353626,
+      "loss": 4.9962,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7519387006759644,
+      "learning_rate": 0.0005999810521928962,
+      "loss": 4.685,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6779834628105164,
+      "learning_rate": 0.0005999808787603373,
+      "loss": 4.8129,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6366468667984009,
+      "learning_rate": 0.0005999807045376866,
+      "loss": 4.7325,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6042128205299377,
+      "learning_rate": 0.0005999805295249444,
+      "loss": 4.6506,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7438672780990601,
+      "learning_rate": 0.0005999803537221112,
+      "loss": 4.741,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6630218029022217,
+      "learning_rate": 0.0005999801771291876,
+      "loss": 4.7997,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5924032330513,
+      "learning_rate": 0.0005999799997461738,
+      "loss": 4.6237,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6375287175178528,
+      "learning_rate": 0.0005999798215730705,
+      "loss": 4.6495,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6222159266471863,
+      "learning_rate": 0.000599979642609878,
+      "loss": 4.7176,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6462298631668091,
+      "learning_rate": 0.0005999794628565969,
+      "loss": 4.79,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5806784629821777,
+      "learning_rate": 0.0005999792823132276,
+      "loss": 4.7,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6010195016860962,
+      "learning_rate": 0.0005999791009797706,
+      "loss": 4.7524,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.594200849533081,
+      "learning_rate": 0.0005999789188562264,
+      "loss": 4.8351,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.647185206413269,
+      "learning_rate": 0.0005999787359425954,
+      "loss": 4.7011,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6233699321746826,
+      "learning_rate": 0.0005999785522388783,
+      "loss": 4.8662,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6353042721748352,
+      "learning_rate": 0.0005999783677450753,
+      "loss": 4.5278,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6044039130210876,
+      "learning_rate": 0.000599978182461187,
+      "loss": 4.7688,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5654269456863403,
+      "learning_rate": 0.000599977996387214,
+      "loss": 4.7677,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6043969988822937,
+      "learning_rate": 0.0005999778095231566,
+      "loss": 4.6497,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6131507754325867,
+      "learning_rate": 0.0005999776218690154,
+      "loss": 4.4584,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6189717650413513,
+      "learning_rate": 0.0005999774334247909,
+      "loss": 4.7738,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6453130841255188,
+      "learning_rate": 0.0005999772441904836,
+      "loss": 4.7744,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5762782096862793,
+      "learning_rate": 0.0005999770541660939,
+      "loss": 4.6736,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6359663009643555,
+      "learning_rate": 0.0005999768633516224,
+      "loss": 4.6294,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5880213379859924,
+      "learning_rate": 0.0005999766717470696,
+      "loss": 4.7633,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5674964189529419,
+      "learning_rate": 0.0005999764793524359,
+      "loss": 4.9202,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6202216744422913,
+      "learning_rate": 0.000599976286167722,
+      "loss": 4.8709,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6854368448257446,
+      "learning_rate": 0.0005999760921929283,
+      "loss": 4.5826,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6487476825714111,
+      "learning_rate": 0.0005999758974280551,
+      "loss": 4.6156,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.650581955909729,
+      "learning_rate": 0.0005999757018731034,
+      "loss": 4.6199,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6073902249336243,
+      "learning_rate": 0.0005999755055280732,
+      "loss": 4.6756,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8026038408279419,
+      "learning_rate": 0.0005999753083929653,
+      "loss": 4.6487,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5802706480026245,
+      "learning_rate": 0.0005999751104677803,
+      "loss": 4.8061,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6185906529426575,
+      "learning_rate": 0.0005999749117525185,
+      "loss": 4.5397,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6637834310531616,
+      "learning_rate": 0.0005999747122471805,
+      "loss": 4.7092,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6466782093048096,
+      "learning_rate": 0.0005999745119517669,
+      "loss": 4.8699,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7292912602424622,
+      "learning_rate": 0.0005999743108662781,
+      "loss": 4.6157,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6685052514076233,
+      "learning_rate": 0.0005999741089907146,
+      "loss": 4.4988,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6441407799720764,
+      "learning_rate": 0.0005999739063250772,
+      "loss": 4.7335,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7168517708778381,
+      "learning_rate": 0.0005999737028693662,
+      "loss": 4.7464,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5829117298126221,
+      "learning_rate": 0.0005999734986235822,
+      "loss": 4.6709,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6130709052085876,
+      "learning_rate": 0.0005999732935877258,
+      "loss": 4.942,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5860216617584229,
+      "learning_rate": 0.0005999730877617974,
+      "loss": 4.8475,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6693971157073975,
+      "learning_rate": 0.0005999728811457975,
+      "loss": 4.6322,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6591095924377441,
+      "learning_rate": 0.0005999726737397269,
+      "loss": 4.5757,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6148874759674072,
+      "learning_rate": 0.0005999724655435862,
+      "loss": 4.6693,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6252702474594116,
+      "learning_rate": 0.0005999722565573754,
+      "loss": 4.8816,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6648889780044556,
+      "learning_rate": 0.0005999720467810956,
+      "loss": 4.8358,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6808665990829468,
+      "learning_rate": 0.0005999718362147471,
+      "loss": 4.7751,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6691820621490479,
+      "learning_rate": 0.0005999716248583306,
+      "loss": 4.7164,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7447707653045654,
+      "learning_rate": 0.0005999714127118465,
+      "loss": 4.466,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6744222640991211,
+      "learning_rate": 0.0005999711997752954,
+      "loss": 4.3235,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7479273676872253,
+      "learning_rate": 0.000599970986048678,
+      "loss": 4.63,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7194980382919312,
+      "learning_rate": 0.0005999707715319947,
+      "loss": 4.4956,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6701153516769409,
+      "learning_rate": 0.000599970556225246,
+      "loss": 4.5292,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.653418779373169,
+      "learning_rate": 0.0005999703401284328,
+      "loss": 4.734,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6312591433525085,
+      "learning_rate": 0.0005999701232415552,
+      "loss": 4.6344,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6036155223846436,
+      "learning_rate": 0.0005999699055646143,
+      "loss": 4.5935,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6546796560287476,
+      "learning_rate": 0.0005999696870976103,
+      "loss": 4.4745,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6474147439002991,
+      "learning_rate": 0.0005999694678405438,
+      "loss": 4.3296,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6968685984611511,
+      "learning_rate": 0.0005999692477934153,
+      "loss": 4.6786,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6009911894798279,
+      "learning_rate": 0.0005999690269562259,
+      "loss": 4.6661,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6383833289146423,
+      "learning_rate": 0.0005999688053289757,
+      "loss": 4.7877,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6396824717521667,
+      "learning_rate": 0.0005999685829116652,
+      "loss": 4.6966,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6602899432182312,
+      "learning_rate": 0.0005999683597042953,
+      "loss": 4.4193,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6245065927505493,
+      "learning_rate": 0.0005999681357068665,
+      "loss": 4.4753,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6217121481895447,
+      "learning_rate": 0.0005999679109193794,
+      "loss": 4.3305,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5806468725204468,
+      "learning_rate": 0.0005999676853418345,
+      "loss": 4.57,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7057567238807678,
+      "learning_rate": 0.0005999674589742325,
+      "loss": 4.7142,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6089333891868591,
+      "learning_rate": 0.0005999672318165738,
+      "loss": 4.5676,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6079134941101074,
+      "learning_rate": 0.0005999670038688592,
+      "loss": 4.7218,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.611993670463562,
+      "learning_rate": 0.0005999667751310893,
+      "loss": 4.5366,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5751462578773499,
+      "learning_rate": 0.0005999665456032647,
+      "loss": 4.5681,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6958464980125427,
+      "learning_rate": 0.0005999663152853858,
+      "loss": 4.714,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7553857564926147,
+      "learning_rate": 0.0005999660841774534,
+      "loss": 4.6257,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6197561621665955,
+      "learning_rate": 0.0005999658522794681,
+      "loss": 4.6357,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5696096420288086,
+      "learning_rate": 0.0005999656195914305,
+      "loss": 4.7116,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6267894506454468,
+      "learning_rate": 0.0005999653861133411,
+      "loss": 4.6124,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6374806761741638,
+      "learning_rate": 0.0005999651518452006,
+      "loss": 4.4663,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.619802713394165,
+      "learning_rate": 0.0005999649167870097,
+      "loss": 4.6616,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5879377722740173,
+      "learning_rate": 0.0005999646809387689,
+      "loss": 4.5237,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6228669881820679,
+      "learning_rate": 0.0005999644443004789,
+      "loss": 4.7327,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0745501518249512,
+      "learning_rate": 0.0005999642068721403,
+      "loss": 4.742,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6069528460502625,
+      "learning_rate": 0.0005999639686537537,
+      "loss": 4.5265,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6730210185050964,
+      "learning_rate": 0.0005999637296453195,
+      "loss": 4.7896,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.665211021900177,
+      "learning_rate": 0.0005999634898468389,
+      "loss": 4.4585,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6540265679359436,
+      "learning_rate": 0.0005999632492583121,
+      "loss": 4.2711,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6752997040748596,
+      "learning_rate": 0.0005999630078797397,
+      "loss": 4.6462,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6478506326675415,
+      "learning_rate": 0.0005999627657111227,
+      "loss": 4.4006,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6323385238647461,
+      "learning_rate": 0.0005999625227524613,
+      "loss": 4.4891,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6497457027435303,
+      "learning_rate": 0.0005999622790037563,
+      "loss": 4.3089,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6490249037742615,
+      "learning_rate": 0.0005999620344650085,
+      "loss": 4.6512,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6243768930435181,
+      "learning_rate": 0.0005999617891362185,
+      "loss": 4.8247,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6056337356567383,
+      "learning_rate": 0.0005999615430173868,
+      "loss": 4.6136,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.59369295835495,
+      "learning_rate": 0.0005999612961085141,
+      "loss": 4.5969,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.627199113368988,
+      "learning_rate": 0.0005999610484096011,
+      "loss": 4.4371,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6465263962745667,
+      "learning_rate": 0.0005999607999206484,
+      "loss": 4.492,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6473185420036316,
+      "learning_rate": 0.0005999605506416567,
+      "loss": 4.4302,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6702262759208679,
+      "learning_rate": 0.0005999603005726267,
+      "loss": 4.6095,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6607844829559326,
+      "learning_rate": 0.0005999600497135589,
+      "loss": 4.6373,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6345941424369812,
+      "learning_rate": 0.0005999597980644541,
+      "loss": 4.602,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8155503869056702,
+      "learning_rate": 0.0005999595456253129,
+      "loss": 4.7077,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5928239822387695,
+      "learning_rate": 0.000599959292396136,
+      "loss": 4.5808,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6200684309005737,
+      "learning_rate": 0.000599959038376924,
+      "loss": 4.4643,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6029160618782043,
+      "learning_rate": 0.0005999587835676778,
+      "loss": 4.6415,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6697453260421753,
+      "learning_rate": 0.0005999585279683977,
+      "loss": 4.706,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6388944983482361,
+      "learning_rate": 0.0005999582715790847,
+      "loss": 4.4938,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6525525450706482,
+      "learning_rate": 0.0005999580143997393,
+      "loss": 4.4293,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6061077117919922,
+      "learning_rate": 0.0005999577564303621,
+      "loss": 4.4593,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6646733283996582,
+      "learning_rate": 0.000599957497670954,
+      "loss": 4.5408,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6451191306114197,
+      "learning_rate": 0.0005999572381215155,
+      "loss": 4.6175,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6172031164169312,
+      "learning_rate": 0.0005999569777820475,
+      "loss": 4.3329,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7403270602226257,
+      "learning_rate": 0.0005999567166525505,
+      "loss": 4.4766,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7167713642120361,
+      "learning_rate": 0.0005999564547330251,
+      "loss": 4.6084,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.605120062828064,
+      "learning_rate": 0.0005999561920234722,
+      "loss": 4.6095,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6738837957382202,
+      "learning_rate": 0.0005999559285238925,
+      "loss": 4.443,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6576931476593018,
+      "learning_rate": 0.0005999556642342866,
+      "loss": 4.4609,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6313718557357788,
+      "learning_rate": 0.0005999553991546552,
+      "loss": 4.4452,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5991811156272888,
+      "learning_rate": 0.0005999551332849989,
+      "loss": 4.5898,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6302996277809143,
+      "learning_rate": 0.0005999548666253186,
+      "loss": 4.3871,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.637596607208252,
+      "learning_rate": 0.0005999545991756149,
+      "loss": 4.563,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6226162910461426,
+      "learning_rate": 0.0005999543309358886,
+      "loss": 4.5605,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6292939186096191,
+      "learning_rate": 0.0005999540619061402,
+      "loss": 4.3628,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6156478524208069,
+      "learning_rate": 0.0005999537920863705,
+      "loss": 4.3611,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6901237368583679,
+      "learning_rate": 0.0005999535214765803,
+      "loss": 4.4794,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5876708626747131,
+      "learning_rate": 0.0005999532500767703,
+      "loss": 4.5018,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6149948239326477,
+      "learning_rate": 0.0005999529778869411,
+      "loss": 4.423,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6238973140716553,
+      "learning_rate": 0.0005999527049070935,
+      "loss": 4.4507,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6290735602378845,
+      "learning_rate": 0.0005999524311372282,
+      "loss": 4.6397,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5849318504333496,
+      "learning_rate": 0.0005999521565773459,
+      "loss": 4.4881,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6476309299468994,
+      "learning_rate": 0.0005999518812274474,
+      "loss": 4.4975,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6004956364631653,
+      "learning_rate": 0.0005999516050875334,
+      "loss": 4.6363,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.640617311000824,
+      "learning_rate": 0.0005999513281576045,
+      "loss": 4.5486,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6288664937019348,
+      "learning_rate": 0.0005999510504376616,
+      "loss": 4.523,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6244643330574036,
+      "learning_rate": 0.0005999507719277054,
+      "loss": 4.4213,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6235512495040894,
+      "learning_rate": 0.0005999504926277365,
+      "loss": 4.4436,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6489076614379883,
+      "learning_rate": 0.0005999502125377557,
+      "loss": 4.3302,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7294739484786987,
+      "learning_rate": 0.0005999499316577639,
+      "loss": 4.3603,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6268585920333862,
+      "learning_rate": 0.0005999496499877615,
+      "loss": 4.5018,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6179900765419006,
+      "learning_rate": 0.0005999493675277497,
+      "loss": 4.4187,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6882840991020203,
+      "learning_rate": 0.0005999490842777288,
+      "loss": 4.3276,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6455632448196411,
+      "learning_rate": 0.0005999488002376999,
+      "loss": 4.6731,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6357992887496948,
+      "learning_rate": 0.0005999485154076634,
+      "loss": 4.3314,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.641894519329071,
+      "learning_rate": 0.0005999482297876204,
+      "loss": 4.5446,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6699628829956055,
+      "learning_rate": 0.0005999479433775714,
+      "loss": 4.411,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6083547472953796,
+      "learning_rate": 0.0005999476561775173,
+      "loss": 4.4048,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6556709408760071,
+      "learning_rate": 0.0005999473681874587,
+      "loss": 4.6152,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6575928330421448,
+      "learning_rate": 0.0005999470794073965,
+      "loss": 4.407,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6290593147277832,
+      "learning_rate": 0.0005999467898373314,
+      "loss": 4.4726,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6279957890510559,
+      "learning_rate": 0.0005999464994772643,
+      "loss": 4.5635,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6792528033256531,
+      "learning_rate": 0.0005999462083271957,
+      "loss": 4.564,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6177847981452942,
+      "learning_rate": 0.0005999459163871266,
+      "loss": 4.3569,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6043108701705933,
+      "learning_rate": 0.0005999456236570577,
+      "loss": 4.5328,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7578732371330261,
+      "learning_rate": 0.0005999453301369897,
+      "loss": 4.3748,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6486274600028992,
+      "learning_rate": 0.0005999450358269234,
+      "loss": 4.4287,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6370539665222168,
+      "learning_rate": 0.0005999447407268597,
+      "loss": 4.2936,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5845693945884705,
+      "learning_rate": 0.0005999444448367993,
+      "loss": 4.2919,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6829562783241272,
+      "learning_rate": 0.0005999441481567428,
+      "loss": 4.6738,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6204245090484619,
+      "learning_rate": 0.0005999438506866911,
+      "loss": 4.5717,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6141250133514404,
+      "learning_rate": 0.0005999435524266453,
+      "loss": 4.5029,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6364489197731018,
+      "learning_rate": 0.0005999432533766056,
+      "loss": 4.2914,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5723608732223511,
+      "learning_rate": 0.0005999429535365734,
+      "loss": 4.4488,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6832333207130432,
+      "learning_rate": 0.0005999426529065489,
+      "loss": 4.3861,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6163421869277954,
+      "learning_rate": 0.0005999423514865334,
+      "loss": 4.8537,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6012544631958008,
+      "learning_rate": 0.0005999420492765273,
+      "loss": 4.525,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6326162815093994,
+      "learning_rate": 0.0005999417462765318,
+      "loss": 4.4894,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7766426801681519,
+      "learning_rate": 0.0005999414424865472,
+      "loss": 4.5683,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5986698269844055,
+      "learning_rate": 0.0005999411379065746,
+      "loss": 4.5938,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6302661895751953,
+      "learning_rate": 0.0005999408325366149,
+      "loss": 4.243,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6291135549545288,
+      "learning_rate": 0.0005999405263766688,
+      "loss": 4.6281,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6244451403617859,
+      "learning_rate": 0.000599940219426737,
+      "loss": 4.3295,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6079283952713013,
+      "learning_rate": 0.0005999399116868203,
+      "loss": 4.3864,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6124751567840576,
+      "learning_rate": 0.0005999396031569197,
+      "loss": 4.6859,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6195314526557922,
+      "learning_rate": 0.0005999392938370358,
+      "loss": 4.6973,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7498607039451599,
+      "learning_rate": 0.0005999389837271697,
+      "loss": 4.4272,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7137424349784851,
+      "learning_rate": 0.0005999386728273219,
+      "loss": 4.5234,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5930689573287964,
+      "learning_rate": 0.0005999383611374934,
+      "loss": 4.5479,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6761220097541809,
+      "learning_rate": 0.000599938048657685,
+      "loss": 4.4618,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6029164791107178,
+      "learning_rate": 0.0005999377353878975,
+      "loss": 4.362,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6475263237953186,
+      "learning_rate": 0.0005999374213281318,
+      "loss": 4.5006,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.647782564163208,
+      "learning_rate": 0.0005999371064783885,
+      "loss": 4.2719,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5890044569969177,
+      "learning_rate": 0.0005999367908386688,
+      "loss": 4.5949,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.627471923828125,
+      "learning_rate": 0.0005999364744089731,
+      "loss": 4.4189,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5897189378738403,
+      "learning_rate": 0.0005999361571893026,
+      "loss": 4.4445,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5901340246200562,
+      "learning_rate": 0.0005999358391796578,
+      "loss": 4.3554,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6106919646263123,
+      "learning_rate": 0.00059993552038004,
+      "loss": 4.3254,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7342495918273926,
+      "learning_rate": 0.0005999352007904495,
+      "loss": 4.3228,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5682928562164307,
+      "learning_rate": 0.0005999348804108875,
+      "loss": 4.4674,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6223228573799133,
+      "learning_rate": 0.0005999345592413548,
+      "loss": 4.4122,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.632500946521759,
+      "learning_rate": 0.0005999342372818523,
+      "loss": 4.5875,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.629377543926239,
+      "learning_rate": 0.0005999339145323805,
+      "loss": 4.271,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6136090755462646,
+      "learning_rate": 0.0005999335909929405,
+      "loss": 4.4486,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6139892339706421,
+      "learning_rate": 0.0005999332666635332,
+      "loss": 4.51,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8123599290847778,
+      "learning_rate": 0.0005999329415441595,
+      "loss": 4.2626,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6359814405441284,
+      "learning_rate": 0.0005999326156348201,
+      "loss": 4.5825,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.702951967716217,
+      "learning_rate": 0.000599932288935516,
+      "loss": 4.5257,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6103016138076782,
+      "learning_rate": 0.0005999319614462479,
+      "loss": 4.4541,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6839025616645813,
+      "learning_rate": 0.0005999316331670166,
+      "loss": 4.4537,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6513589024543762,
+      "learning_rate": 0.0005999313040978233,
+      "loss": 4.2952,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6502857804298401,
+      "learning_rate": 0.0005999309742386685,
+      "loss": 4.3789,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6159502267837524,
+      "learning_rate": 0.0005999306435895533,
+      "loss": 4.484,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5981889963150024,
+      "learning_rate": 0.0005999303121504786,
+      "loss": 4.3018,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6399751901626587,
+      "learning_rate": 0.0005999299799214451,
+      "loss": 4.3074,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6635521054267883,
+      "learning_rate": 0.0005999296469024537,
+      "loss": 4.399,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5930505990982056,
+      "learning_rate": 0.0005999293130935054,
+      "loss": 4.4843,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6015510559082031,
+      "learning_rate": 0.0005999289784946011,
+      "loss": 4.4935,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.690433919429779,
+      "learning_rate": 0.0005999286431057415,
+      "loss": 4.4129,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7033190727233887,
+      "learning_rate": 0.0005999283069269276,
+      "loss": 4.4088,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5974972248077393,
+      "learning_rate": 0.0005999279699581602,
+      "loss": 4.2984,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6253489255905151,
+      "learning_rate": 0.0005999276321994402,
+      "loss": 4.2365,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6363905668258667,
+      "learning_rate": 0.0005999272936507687,
+      "loss": 4.4612,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6299953460693359,
+      "learning_rate": 0.0005999269543121463,
+      "loss": 4.2972,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7025599479675293,
+      "learning_rate": 0.0005999266141835741,
+      "loss": 4.168,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7058206796646118,
+      "learning_rate": 0.000599926273265053,
+      "loss": 4.2049,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.639482319355011,
+      "learning_rate": 0.0005999259315565837,
+      "loss": 4.2935,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5956512093544006,
+      "learning_rate": 0.0005999255890581672,
+      "loss": 4.2922,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6834009289741516,
+      "learning_rate": 0.0005999252457698045,
+      "loss": 4.1426,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6566224694252014,
+      "learning_rate": 0.0005999249016914964,
+      "loss": 4.4575,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6373875737190247,
+      "learning_rate": 0.0005999245568232438,
+      "loss": 4.634,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.637948751449585,
+      "learning_rate": 0.0005999242111650476,
+      "loss": 4.4227,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.713834285736084,
+      "learning_rate": 0.0005999238647169089,
+      "loss": 4.2999,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.606334924697876,
+      "learning_rate": 0.0005999235174788283,
+      "loss": 4.3889,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6799372434616089,
+      "learning_rate": 0.0005999231694508069,
+      "loss": 4.3517,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6416102051734924,
+      "learning_rate": 0.0005999228206328457,
+      "loss": 4.5535,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6423821449279785,
+      "learning_rate": 0.0005999224710249454,
+      "loss": 4.0969,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6177202463150024,
+      "learning_rate": 0.0005999221206271071,
+      "loss": 4.3001,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5873739719390869,
+      "learning_rate": 0.0005999217694393317,
+      "loss": 4.3533,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6107127666473389,
+      "learning_rate": 0.00059992141746162,
+      "loss": 4.2398,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6472650170326233,
+      "learning_rate": 0.0005999210646939731,
+      "loss": 4.3991,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.611621081829071,
+      "learning_rate": 0.0005999207111363916,
+      "loss": 4.4359,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6676717400550842,
+      "learning_rate": 0.000599920356788877,
+      "loss": 4.3993,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6014063954353333,
+      "learning_rate": 0.0005999200016514296,
+      "loss": 4.5267,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6274306774139404,
+      "learning_rate": 0.0005999196457240508,
+      "loss": 4.2282,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6827574372291565,
+      "learning_rate": 0.0005999192890067413,
+      "loss": 4.3729,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7055410742759705,
+      "learning_rate": 0.0005999189314995022,
+      "loss": 4.2916,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.654010534286499,
+      "learning_rate": 0.0005999185732023343,
+      "loss": 4.2391,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5943931937217712,
+      "learning_rate": 0.0005999182141152385,
+      "loss": 4.2879,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6318141222000122,
+      "learning_rate": 0.000599917854238216,
+      "loss": 4.2995,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6332406401634216,
+      "learning_rate": 0.0005999174935712676,
+      "loss": 4.2975,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.654248058795929,
+      "learning_rate": 0.0005999171321143941,
+      "loss": 4.3066,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6300703287124634,
+      "learning_rate": 0.0005999167698675967,
+      "loss": 4.4163,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6364467740058899,
+      "learning_rate": 0.0005999164068308762,
+      "loss": 4.4665,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.650804340839386,
+      "learning_rate": 0.0005999160430042337,
+      "loss": 4.2126,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6396269202232361,
+      "learning_rate": 0.00059991567838767,
+      "loss": 4.4289,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7045847177505493,
+      "learning_rate": 0.0005999153129811861,
+      "loss": 4.1509,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.895271360874176,
+      "learning_rate": 0.0005999149467847831,
+      "loss": 4.3426,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6522945165634155,
+      "learning_rate": 0.0005999145797984617,
+      "loss": 4.4832,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6630837321281433,
+      "learning_rate": 0.000599914212022223,
+      "loss": 4.4587,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6792482733726501,
+      "learning_rate": 0.0005999138434560681,
+      "loss": 4.2637,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7289097309112549,
+      "learning_rate": 0.0005999134740999979,
+      "loss": 4.4558,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6192370653152466,
+      "learning_rate": 0.0005999131039540131,
+      "loss": 4.3225,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6023765206336975,
+      "learning_rate": 0.0005999127330181151,
+      "loss": 4.4778,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6397039294242859,
+      "learning_rate": 0.0005999123612923046,
+      "loss": 4.4022,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5905510187149048,
+      "learning_rate": 0.0005999119887765827,
+      "loss": 4.468,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7561242580413818,
+      "learning_rate": 0.0005999116154709504,
+      "loss": 4.2759,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6308673024177551,
+      "learning_rate": 0.0005999112413754084,
+      "loss": 4.2214,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6405158638954163,
+      "learning_rate": 0.0005999108664899582,
+      "loss": 4.3654,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6524988412857056,
+      "learning_rate": 0.0005999104908146003,
+      "loss": 4.5213,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6339336037635803,
+      "learning_rate": 0.0005999101143493359,
+      "loss": 4.5757,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6563646793365479,
+      "learning_rate": 0.000599909737094166,
+      "loss": 4.4423,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6299603581428528,
+      "learning_rate": 0.0005999093590490917,
+      "loss": 4.1991,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8304985761642456,
+      "learning_rate": 0.0005999089802141137,
+      "loss": 4.5737,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6237910389900208,
+      "learning_rate": 0.0005999086005892332,
+      "loss": 4.298,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6513141989707947,
+      "learning_rate": 0.0005999082201744512,
+      "loss": 3.9653,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6443023085594177,
+      "learning_rate": 0.0005999078389697687,
+      "loss": 4.3509,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6092896461486816,
+      "learning_rate": 0.0005999074569751865,
+      "loss": 4.4119,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6280472874641418,
+      "learning_rate": 0.000599907074190706,
+      "loss": 4.1763,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6605474352836609,
+      "learning_rate": 0.0005999066906163279,
+      "loss": 4.1999,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.633695125579834,
+      "learning_rate": 0.0005999063062520532,
+      "loss": 4.5811,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6050964593887329,
+      "learning_rate": 0.0005999059210978832,
+      "loss": 4.2844,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6477183699607849,
+      "learning_rate": 0.0005999055351538186,
+      "loss": 4.315,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6902852654457092,
+      "learning_rate": 0.0005999051484198606,
+      "loss": 4.4602,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9078108668327332,
+      "learning_rate": 0.00059990476089601,
+      "loss": 4.6133,
+      "step": 491
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6101062893867493,
+      "learning_rate": 0.000599904372582268,
+      "loss": 4.5548,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6342147588729858,
+      "learning_rate": 0.0005999039834786357,
+      "loss": 4.232,
+      "step": 493
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6037482619285583,
+      "learning_rate": 0.0005999035935851142,
+      "loss": 4.4302,
+      "step": 494
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6121346950531006,
+      "learning_rate": 0.0005999032029017041,
+      "loss": 4.2391,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6090756058692932,
+      "learning_rate": 0.0005999028114284067,
+      "loss": 4.4575,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.569670557975769,
+      "learning_rate": 0.0005999024191652231,
+      "loss": 4.2884,
+      "step": 497
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5985832810401917,
+      "learning_rate": 0.0005999020261121541,
+      "loss": 4.3026,
+      "step": 498
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6532287001609802,
+      "learning_rate": 0.000599901632269201,
+      "loss": 4.3152,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5899796485900879,
+      "learning_rate": 0.0005999012376363647,
+      "loss": 4.1943,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6288335919380188,
+      "learning_rate": 0.0005999008422136463,
+      "loss": 4.4126,
+      "step": 501
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6783986687660217,
+      "learning_rate": 0.0005999004460010467,
+      "loss": 4.1275,
+      "step": 502
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.612891435623169,
+      "learning_rate": 0.0005999000489985671,
+      "loss": 4.3033,
+      "step": 503
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6028085947036743,
+      "learning_rate": 0.0005998996512062085,
+      "loss": 4.2218,
+      "step": 504
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5980657935142517,
+      "learning_rate": 0.0005998992526239719,
+      "loss": 4.2868,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6469926238059998,
+      "learning_rate": 0.0005998988532518584,
+      "loss": 4.4216,
+      "step": 506
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6261191368103027,
+      "learning_rate": 0.0005998984530898691,
+      "loss": 4.1823,
+      "step": 507
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5999051928520203,
+      "learning_rate": 0.0005998980521380048,
+      "loss": 4.0664,
+      "step": 508
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6515395641326904,
+      "learning_rate": 0.000599897650396267,
+      "loss": 4.5476,
+      "step": 509
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5981988906860352,
+      "learning_rate": 0.0005998972478646564,
+      "loss": 4.2978,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6036791205406189,
+      "learning_rate": 0.0005998968445431742,
+      "loss": 4.3236,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6372940540313721,
+      "learning_rate": 0.0005998964404318213,
+      "loss": 4.2261,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.612691342830658,
+      "learning_rate": 0.0005998960355305989,
+      "loss": 4.2686,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6639612913131714,
+      "learning_rate": 0.0005998956298395082,
+      "loss": 4.4662,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6196693181991577,
+      "learning_rate": 0.00059989522335855,
+      "loss": 4.244,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6042441725730896,
+      "learning_rate": 0.0005998948160877256,
+      "loss": 4.4187,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.584118127822876,
+      "learning_rate": 0.0005998944080270359,
+      "loss": 4.2727,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6688718795776367,
+      "learning_rate": 0.0005998939991764821,
+      "loss": 4.4981,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6350945234298706,
+      "learning_rate": 0.0005998935895360651,
+      "loss": 4.0629,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6718504428863525,
+      "learning_rate": 0.0005998931791057863,
+      "loss": 4.3385,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.666595995426178,
+      "learning_rate": 0.0005998927678856464,
+      "loss": 4.1755,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6215299963951111,
+      "learning_rate": 0.0005998923558756467,
+      "loss": 4.4248,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6000812649726868,
+      "learning_rate": 0.0005998919430757883,
+      "loss": 4.3693,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.643562912940979,
+      "learning_rate": 0.0005998915294860722,
+      "loss": 3.9877,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6312112808227539,
+      "learning_rate": 0.0005998911151064996,
+      "loss": 4.3255,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6362340450286865,
+      "learning_rate": 0.0005998906999370716,
+      "loss": 4.2147,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6441465616226196,
+      "learning_rate": 0.0005998902839777889,
+      "loss": 4.2851,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6550605893135071,
+      "learning_rate": 0.0005998898672286532,
+      "loss": 4.167,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6101124286651611,
+      "learning_rate": 0.0005998894496896651,
+      "loss": 4.5298,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6387701630592346,
+      "learning_rate": 0.0005998890313608261,
+      "loss": 4.1748,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6303190588951111,
+      "learning_rate": 0.0005998886122421369,
+      "loss": 4.2512,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7182632088661194,
+      "learning_rate": 0.0005998881923335989,
+      "loss": 4.3773,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5540059208869934,
+      "learning_rate": 0.0005998877716352132,
+      "loss": 4.3919,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6034284234046936,
+      "learning_rate": 0.0005998873501469808,
+      "loss": 4.1638,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6099483966827393,
+      "learning_rate": 0.0005998869278689028,
+      "loss": 4.3032,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6097630262374878,
+      "learning_rate": 0.0005998865048009803,
+      "loss": 4.1468,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5770888328552246,
+      "learning_rate": 0.0005998860809432145,
+      "loss": 4.2715,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6351972818374634,
+      "learning_rate": 0.0005998856562956064,
+      "loss": 4.2979,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6256553530693054,
+      "learning_rate": 0.0005998852308581573,
+      "loss": 4.1972,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.625809907913208,
+      "learning_rate": 0.0005998848046308682,
+      "loss": 3.9932,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6050224900245667,
+      "learning_rate": 0.0005998843776137402,
+      "loss": 4.2899,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.617949366569519,
+      "learning_rate": 0.0005998839498067745,
+      "loss": 4.069,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7680637836456299,
+      "learning_rate": 0.0005998835212099722,
+      "loss": 4.264,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6679553389549255,
+      "learning_rate": 0.0005998830918233344,
+      "loss": 4.2505,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0453165769577026,
+      "learning_rate": 0.0005998826616468622,
+      "loss": 4.356,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6193425059318542,
+      "learning_rate": 0.0005998822306805568,
+      "loss": 4.4011,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6391511559486389,
+      "learning_rate": 0.0005998817989244194,
+      "loss": 4.08,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6139493584632874,
+      "learning_rate": 0.000599881366378451,
+      "loss": 4.206,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6752720475196838,
+      "learning_rate": 0.0005998809330426528,
+      "loss": 4.1488,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6094870567321777,
+      "learning_rate": 0.0005998804989170259,
+      "loss": 4.1779,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5940812230110168,
+      "learning_rate": 0.0005998800640015715,
+      "loss": 4.3568,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6211913228034973,
+      "learning_rate": 0.0005998796282962907,
+      "loss": 4.4739,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5913729071617126,
+      "learning_rate": 0.0005998791918011847,
+      "loss": 4.2086,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6634578704833984,
+      "learning_rate": 0.0005998787545162547,
+      "loss": 4.2911,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.636097252368927,
+      "learning_rate": 0.0005998783164415017,
+      "loss": 4.1991,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6963991522789001,
+      "learning_rate": 0.0005998778775769269,
+      "loss": 4.0183,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6141919493675232,
+      "learning_rate": 0.0005998774379225315,
+      "loss": 4.5183,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5955811142921448,
+      "learning_rate": 0.0005998769974783167,
+      "loss": 4.3453,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6223891377449036,
+      "learning_rate": 0.0005998765562442835,
+      "loss": 4.3894,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6813943386077881,
+      "learning_rate": 0.0005998761142204333,
+      "loss": 4.2083,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6669309735298157,
+      "learning_rate": 0.000599875671406767,
+      "loss": 4.566,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6055184602737427,
+      "learning_rate": 0.0005998752278032859,
+      "loss": 4.0433,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6754607558250427,
+      "learning_rate": 0.0005998747834099912,
+      "loss": 4.0311,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6373229622840881,
+      "learning_rate": 0.000599874338226884,
+      "loss": 4.33,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5642352104187012,
+      "learning_rate": 0.0005998738922539656,
+      "loss": 4.2819,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5712973475456238,
+      "learning_rate": 0.000599873445491237,
+      "loss": 4.3986,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6137987375259399,
+      "learning_rate": 0.0005998729979386994,
+      "loss": 4.4378,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6173729300498962,
+      "learning_rate": 0.0005998725495963542,
+      "loss": 4.3934,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6119434833526611,
+      "learning_rate": 0.0005998721004642024,
+      "loss": 4.3429,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6111935377120972,
+      "learning_rate": 0.000599871650542245,
+      "loss": 4.2851,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7171507477760315,
+      "learning_rate": 0.0005998711998304835,
+      "loss": 4.4319,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5924344062805176,
+      "learning_rate": 0.000599870748328919,
+      "loss": 4.4248,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6047422289848328,
+      "learning_rate": 0.0005998702960375526,
+      "loss": 4.3181,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.61612468957901,
+      "learning_rate": 0.0005998698429563856,
+      "loss": 4.2176,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6250529289245605,
+      "learning_rate": 0.0005998693890854192,
+      "loss": 4.3786,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6662085652351379,
+      "learning_rate": 0.0005998689344246544,
+      "loss": 4.4409,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6052320599555969,
+      "learning_rate": 0.0005998684789740926,
+      "loss": 4.1686,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6721617579460144,
+      "learning_rate": 0.0005998680227337351,
+      "loss": 4.1626,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6326162815093994,
+      "learning_rate": 0.0005998675657035827,
+      "loss": 4.4329,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6244572401046753,
+      "learning_rate": 0.0005998671078836369,
+      "loss": 4.3406,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6222200989723206,
+      "learning_rate": 0.0005998666492738989,
+      "loss": 4.3156,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6773638725280762,
+      "learning_rate": 0.0005998661898743698,
+      "loss": 4.2542,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7391511797904968,
+      "learning_rate": 0.0005998657296850509,
+      "loss": 4.293,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6230476498603821,
+      "learning_rate": 0.0005998652687059434,
+      "loss": 4.1702,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6014927625656128,
+      "learning_rate": 0.0005998648069370485,
+      "loss": 4.0259,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6436325311660767,
+      "learning_rate": 0.0005998643443783674,
+      "loss": 4.1263,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6496565937995911,
+      "learning_rate": 0.0005998638810299013,
+      "loss": 4.0934,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5946733951568604,
+      "learning_rate": 0.0005998634168916515,
+      "loss": 4.3685,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5709477066993713,
+      "learning_rate": 0.0005998629519636191,
+      "loss": 4.0456,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6664632558822632,
+      "learning_rate": 0.0005998624862458054,
+      "loss": 4.2432,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6627117395401001,
+      "learning_rate": 0.0005998620197382117,
+      "loss": 4.492,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6205899715423584,
+      "learning_rate": 0.0005998615524408391,
+      "loss": 4.3008,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6240425109863281,
+      "learning_rate": 0.0005998610843536888,
+      "loss": 4.4099,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6774874329566956,
+      "learning_rate": 0.0005998606154767621,
+      "loss": 4.2297,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6011747121810913,
+      "learning_rate": 0.0005998601458100603,
+      "loss": 4.2218,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6347803473472595,
+      "learning_rate": 0.0005998596753535847,
+      "loss": 4.2884,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6915126442909241,
+      "learning_rate": 0.0005998592041073364,
+      "loss": 4.3879,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6727957129478455,
+      "learning_rate": 0.0005998587320713165,
+      "loss": 4.0906,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6125076413154602,
+      "learning_rate": 0.0005998582592455266,
+      "loss": 4.2448,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6534925699234009,
+      "learning_rate": 0.0005998577856299677,
+      "loss": 4.1813,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6901077628135681,
+      "learning_rate": 0.0005998573112246411,
+      "loss": 4.0296,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6255195140838623,
+      "learning_rate": 0.000599856836029548,
+      "loss": 4.22,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6073751449584961,
+      "learning_rate": 0.0005998563600446898,
+      "loss": 4.3775,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6304476857185364,
+      "learning_rate": 0.0005998558832700675,
+      "loss": 4.5508,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6237664222717285,
+      "learning_rate": 0.0005998554057056825,
+      "loss": 4.3309,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6360412836074829,
+      "learning_rate": 0.0005998549273515362,
+      "loss": 4.2754,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6329114437103271,
+      "learning_rate": 0.0005998544482076297,
+      "loss": 4.4678,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6190249919891357,
+      "learning_rate": 0.0005998539682739643,
+      "loss": 4.0623,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7699590921401978,
+      "learning_rate": 0.0005998534875505413,
+      "loss": 4.368,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6379511952400208,
+      "learning_rate": 0.0005998530060373618,
+      "loss": 4.0845,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6318610310554504,
+      "learning_rate": 0.0005998525237344272,
+      "loss": 4.1786,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.654819905757904,
+      "learning_rate": 0.0005998520406417388,
+      "loss": 4.219,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6683518886566162,
+      "learning_rate": 0.0005998515567592979,
+      "loss": 4.333,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6329825520515442,
+      "learning_rate": 0.0005998510720871057,
+      "loss": 4.2552,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6308939456939697,
+      "learning_rate": 0.0005998505866251635,
+      "loss": 4.142,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.557669460773468,
+      "learning_rate": 0.0005998501003734724,
+      "loss": 4.2415,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6747656464576721,
+      "learning_rate": 0.0005998496133320339,
+      "loss": 4.3223,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6284106373786926,
+      "learning_rate": 0.0005998491255008494,
+      "loss": 4.3184,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6386296153068542,
+      "learning_rate": 0.0005998486368799197,
+      "loss": 4.0359,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6140933632850647,
+      "learning_rate": 0.0005998481474692466,
+      "loss": 4.0157,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6379926204681396,
+      "learning_rate": 0.0005998476572688312,
+      "loss": 4.1319,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5984822511672974,
+      "learning_rate": 0.0005998471662786747,
+      "loss": 4.157,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5686042904853821,
+      "learning_rate": 0.0005998466744987786,
+      "loss": 4.2332,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6098901629447937,
+      "learning_rate": 0.0005998461819291439,
+      "loss": 4.0068,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7193558216094971,
+      "learning_rate": 0.0005998456885697722,
+      "loss": 4.1709,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6442568898200989,
+      "learning_rate": 0.0005998451944206646,
+      "loss": 4.2862,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.610493004322052,
+      "learning_rate": 0.0005998446994818225,
+      "loss": 4.1749,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6829988360404968,
+      "learning_rate": 0.0005998442037532471,
+      "loss": 4.2972,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6680061221122742,
+      "learning_rate": 0.0005998437072349397,
+      "loss": 4.1303,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6559966206550598,
+      "learning_rate": 0.0005998432099269019,
+      "loss": 4.1799,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.610496461391449,
+      "learning_rate": 0.0005998427118291347,
+      "loss": 4.2051,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5574377775192261,
+      "learning_rate": 0.0005998422129416393,
+      "loss": 4.2476,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6652995347976685,
+      "learning_rate": 0.0005998417132644175,
+      "loss": 4.1745,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.609061598777771,
+      "learning_rate": 0.0005998412127974702,
+      "loss": 4.3128,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6088485717773438,
+      "learning_rate": 0.0005998407115407989,
+      "loss": 4.3749,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6192953586578369,
+      "learning_rate": 0.0005998402094944048,
+      "loss": 3.9438,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6353934407234192,
+      "learning_rate": 0.0005998397066582894,
+      "loss": 4.1745,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0005871057510376,
+      "learning_rate": 0.0005998392030324539,
+      "loss": 4.5231,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5929376482963562,
+      "learning_rate": 0.0005998386986168996,
+      "loss": 4.0178,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6462288498878479,
+      "learning_rate": 0.0005998381934116279,
+      "loss": 4.1872,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6652570366859436,
+      "learning_rate": 0.0005998376874166401,
+      "loss": 4.4516,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6204379200935364,
+      "learning_rate": 0.0005998371806319375,
+      "loss": 4.2446,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6597891449928284,
+      "learning_rate": 0.0005998366730575216,
+      "loss": 4.1447,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.708060622215271,
+      "learning_rate": 0.0005998361646933934,
+      "loss": 4.3571,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5907878279685974,
+      "learning_rate": 0.0005998356555395546,
+      "loss": 4.0995,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5822071433067322,
+      "learning_rate": 0.0005998351455960062,
+      "loss": 4.3305,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6510571837425232,
+      "learning_rate": 0.0005998346348627499,
+      "loss": 4.2952,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.624159574508667,
+      "learning_rate": 0.0005998341233397869,
+      "loss": 4.1151,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5877444744110107,
+      "learning_rate": 0.0005998336110271185,
+      "loss": 4.1762,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.664666473865509,
+      "learning_rate": 0.0005998330979247459,
+      "loss": 4.0597,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5993242263793945,
+      "learning_rate": 0.0005998325840326708,
+      "loss": 4.3319,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.603900671005249,
+      "learning_rate": 0.0005998320693508944,
+      "loss": 4.1354,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5658103227615356,
+      "learning_rate": 0.000599831553879418,
+      "loss": 4.3008,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.588505208492279,
+      "learning_rate": 0.0005998310376182429,
+      "loss": 3.9547,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6646924018859863,
+      "learning_rate": 0.0005998305205673707,
+      "loss": 4.3278,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5992798805236816,
+      "learning_rate": 0.0005998300027268024,
+      "loss": 4.0916,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6579504013061523,
+      "learning_rate": 0.0005998294840965397,
+      "loss": 4.2061,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6638225317001343,
+      "learning_rate": 0.0005998289646765839,
+      "loss": 4.3004,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6212108135223389,
+      "learning_rate": 0.0005998284444669363,
+      "loss": 4.25,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7648120522499084,
+      "learning_rate": 0.0005998279234675982,
+      "loss": 3.9266,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6046731472015381,
+      "learning_rate": 0.0005998274016785711,
+      "loss": 4.0652,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6333780884742737,
+      "learning_rate": 0.0005998268790998563,
+      "loss": 4.2736,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6316869854927063,
+      "learning_rate": 0.0005998263557314553,
+      "loss": 4.2631,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.593485951423645,
+      "learning_rate": 0.0005998258315733692,
+      "loss": 4.3236,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.633284330368042,
+      "learning_rate": 0.0005998253066255997,
+      "loss": 4.0417,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5952419638633728,
+      "learning_rate": 0.000599824780888148,
+      "loss": 4.3424,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6344154477119446,
+      "learning_rate": 0.0005998242543610155,
+      "loss": 4.1042,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6327022910118103,
+      "learning_rate": 0.0005998237270442037,
+      "loss": 4.174,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7157790660858154,
+      "learning_rate": 0.0005998231989377139,
+      "loss": 4.0206,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6288419365882874,
+      "learning_rate": 0.0005998226700415474,
+      "loss": 4.3456,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6706027984619141,
+      "learning_rate": 0.0005998221403557058,
+      "loss": 3.9778,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6601995825767517,
+      "learning_rate": 0.0005998216098801904,
+      "loss": 4.2968,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6394633650779724,
+      "learning_rate": 0.0005998210786150024,
+      "loss": 4.3944,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7500145435333252,
+      "learning_rate": 0.0005998205465601435,
+      "loss": 4.0035,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6336989998817444,
+      "learning_rate": 0.000599820013715615,
+      "loss": 4.241,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6461735367774963,
+      "learning_rate": 0.0005998194800814182,
+      "loss": 4.223,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6548101902008057,
+      "learning_rate": 0.0005998189456575547,
+      "loss": 4.1025,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6322773694992065,
+      "learning_rate": 0.0005998184104440257,
+      "loss": 4.2219,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6030046939849854,
+      "learning_rate": 0.0005998178744408328,
+      "loss": 4.1642,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6334202289581299,
+      "learning_rate": 0.0005998173376479773,
+      "loss": 4.2389,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6108695864677429,
+      "learning_rate": 0.0005998168000654606,
+      "loss": 4.0175,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6236847639083862,
+      "learning_rate": 0.0005998162616932841,
+      "loss": 4.3389,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6061368584632874,
+      "learning_rate": 0.0005998157225314493,
+      "loss": 4.2202,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6532244682312012,
+      "learning_rate": 0.0005998151825799576,
+      "loss": 4.0348,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5872693061828613,
+      "learning_rate": 0.0005998146418388105,
+      "loss": 4.2631,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6573641300201416,
+      "learning_rate": 0.0005998141003080092,
+      "loss": 4.1371,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6546458601951599,
+      "learning_rate": 0.0005998135579875554,
+      "loss": 4.1695,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6271075010299683,
+      "learning_rate": 0.0005998130148774502,
+      "loss": 4.2471,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5936674475669861,
+      "learning_rate": 0.0005998124709776953,
+      "loss": 4.1321,
+      "step": 689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5933693051338196,
+      "learning_rate": 0.0005998119262882921,
+      "loss": 4.1484,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6198687553405762,
+      "learning_rate": 0.000599811380809242,
+      "loss": 4.4228,
+      "step": 691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6083481907844543,
+      "learning_rate": 0.0005998108345405465,
+      "loss": 4.266,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6238859295845032,
+      "learning_rate": 0.0005998102874822068,
+      "loss": 4.3885,
+      "step": 693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6317378878593445,
+      "learning_rate": 0.0005998097396342245,
+      "loss": 4.2074,
+      "step": 694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6172046661376953,
+      "learning_rate": 0.0005998091909966011,
+      "loss": 4.0928,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6623038649559021,
+      "learning_rate": 0.000599808641569338,
+      "loss": 4.0324,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6745285987854004,
+      "learning_rate": 0.0005998080913524365,
+      "loss": 4.3255,
+      "step": 697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5575293302536011,
+      "learning_rate": 0.0005998075403458984,
+      "loss": 4.2219,
+      "step": 698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6386272311210632,
+      "learning_rate": 0.0005998069885497249,
+      "loss": 4.236,
+      "step": 699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7148878574371338,
+      "learning_rate": 0.0005998064359639173,
+      "loss": 4.089,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6956779360771179,
+      "learning_rate": 0.0005998058825884775,
+      "loss": 4.1711,
+      "step": 701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.610063910484314,
+      "learning_rate": 0.0005998053284234067,
+      "loss": 4.3074,
+      "step": 702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7103130221366882,
+      "learning_rate": 0.0005998047734687062,
+      "loss": 4.1535,
+      "step": 703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6960937976837158,
+      "learning_rate": 0.0005998042177243776,
+      "loss": 3.9969,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7523322105407715,
+      "learning_rate": 0.0005998036611904225,
+      "loss": 3.8138,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6804961562156677,
+      "learning_rate": 0.0005998031038668422,
+      "loss": 4.1109,
+      "step": 706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6763580441474915,
+      "learning_rate": 0.0005998025457536382,
+      "loss": 3.9702,
+      "step": 707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8153755068778992,
+      "learning_rate": 0.0005998019868508121,
+      "loss": 4.0995,
+      "step": 708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6804799437522888,
+      "learning_rate": 0.0005998014271583652,
+      "loss": 4.2456,
+      "step": 709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6983351707458496,
+      "learning_rate": 0.0005998008666762989,
+      "loss": 3.9431,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6080770492553711,
+      "learning_rate": 0.0005998003054046151,
+      "loss": 4.277,
+      "step": 711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7080684900283813,
+      "learning_rate": 0.0005997997433433148,
+      "loss": 4.0764,
+      "step": 712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6443161964416504,
+      "learning_rate": 0.0005997991804923997,
+      "loss": 4.0913,
+      "step": 713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6476138234138489,
+      "learning_rate": 0.0005997986168518713,
+      "loss": 4.2866,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.603281557559967,
+      "learning_rate": 0.000599798052421731,
+      "loss": 4.1755,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6660357713699341,
+      "learning_rate": 0.0005997974872019804,
+      "loss": 4.0975,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6493331789970398,
+      "learning_rate": 0.0005997969211926208,
+      "loss": 4.2041,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6349976062774658,
+      "learning_rate": 0.000599796354393654,
+      "loss": 4.0962,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6662265062332153,
+      "learning_rate": 0.000599795786805081,
+      "loss": 4.0407,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5812216997146606,
+      "learning_rate": 0.0005997952184269038,
+      "loss": 4.2963,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5997583866119385,
+      "learning_rate": 0.0005997946492591237,
+      "loss": 4.0069,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6585355997085571,
+      "learning_rate": 0.0005997940793017422,
+      "loss": 4.0195,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6478659510612488,
+      "learning_rate": 0.0005997935085547606,
+      "loss": 4.1715,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6393293142318726,
+      "learning_rate": 0.0005997929370181809,
+      "loss": 4.086,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6129328608512878,
+      "learning_rate": 0.0005997923646920041,
+      "loss": 4.2473,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6000876426696777,
+      "learning_rate": 0.0005997917915762319,
+      "loss": 3.8759,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6004976034164429,
+      "learning_rate": 0.0005997912176708657,
+      "loss": 4.2669,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.698418140411377,
+      "learning_rate": 0.0005997906429759074,
+      "loss": 4.2933,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6390034556388855,
+      "learning_rate": 0.0005997900674913581,
+      "loss": 4.2472,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6226664781570435,
+      "learning_rate": 0.0005997894912172196,
+      "loss": 4.0674,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6232365965843201,
+      "learning_rate": 0.0005997889141534931,
+      "loss": 4.3033,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6623662114143372,
+      "learning_rate": 0.0005997883363001803,
+      "loss": 4.1758,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9481949806213379,
+      "learning_rate": 0.0005997877576572828,
+      "loss": 4.5139,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6264766454696655,
+      "learning_rate": 0.000599787178224802,
+      "loss": 4.109,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6743102073669434,
+      "learning_rate": 0.0005997865980027395,
+      "loss": 4.422,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.688208281993866,
+      "learning_rate": 0.0005997860169910968,
+      "loss": 4.1469,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5970245003700256,
+      "learning_rate": 0.0005997854351898754,
+      "loss": 4.2083,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6264445781707764,
+      "learning_rate": 0.0005997848525990769,
+      "loss": 4.251,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6513589024543762,
+      "learning_rate": 0.0005997842692187027,
+      "loss": 3.8674,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6829968690872192,
+      "learning_rate": 0.0005997836850487545,
+      "loss": 3.9729,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6139025688171387,
+      "learning_rate": 0.0005997831000892338,
+      "loss": 4.1896,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6279214024543762,
+      "learning_rate": 0.0005997825143401421,
+      "loss": 4.1086,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6188284158706665,
+      "learning_rate": 0.0005997819278014808,
+      "loss": 4.0587,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6335148811340332,
+      "learning_rate": 0.0005997813404732517,
+      "loss": 3.9916,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.643291711807251,
+      "learning_rate": 0.0005997807523554563,
+      "loss": 4.2194,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6920807957649231,
+      "learning_rate": 0.000599780163448096,
+      "loss": 3.9606,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6325848698616028,
+      "learning_rate": 0.0005997795737511725,
+      "loss": 4.0133,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6075696349143982,
+      "learning_rate": 0.0005997789832646874,
+      "loss": 4.1146,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.658141016960144,
+      "learning_rate": 0.000599778391988642,
+      "loss": 4.1188,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5975100994110107,
+      "learning_rate": 0.000599777799923038,
+      "loss": 4.1244,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6382097601890564,
+      "learning_rate": 0.0005997772070678771,
+      "loss": 4.0732,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.623789370059967,
+      "learning_rate": 0.0005997766134231606,
+      "loss": 4.1322,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6385670900344849,
+      "learning_rate": 0.0005997760189888902,
+      "loss": 4.1762,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.640335738658905,
+      "learning_rate": 0.0005997754237650675,
+      "loss": 4.0456,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6698756814002991,
+      "learning_rate": 0.0005997748277516941,
+      "loss": 4.3383,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6713323593139648,
+      "learning_rate": 0.0005997742309487714,
+      "loss": 4.0747,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.61420738697052,
+      "learning_rate": 0.0005997736333563011,
+      "loss": 4.1113,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6245611310005188,
+      "learning_rate": 0.0005997730349742847,
+      "loss": 4.3387,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5801929235458374,
+      "learning_rate": 0.0005997724358027238,
+      "loss": 4.0914,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6528918743133545,
+      "learning_rate": 0.0005997718358416201,
+      "loss": 4.148,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5901684761047363,
+      "learning_rate": 0.000599771235090975,
+      "loss": 4.0853,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6030728220939636,
+      "learning_rate": 0.0005997706335507902,
+      "loss": 4.071,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.624138355255127,
+      "learning_rate": 0.0005997700312210672,
+      "loss": 4.134,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.647909939289093,
+      "learning_rate": 0.0005997694281018077,
+      "loss": 4.2178,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6185994148254395,
+      "learning_rate": 0.0005997688241930131,
+      "loss": 4.2237,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6649905443191528,
+      "learning_rate": 0.0005997682194946852,
+      "loss": 4.4995,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6107615828514099,
+      "learning_rate": 0.0005997676140068255,
+      "loss": 4.1487,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6697407364845276,
+      "learning_rate": 0.0005997670077294355,
+      "loss": 4.0481,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6354249715805054,
+      "learning_rate": 0.000599766400662517,
+      "loss": 4.0766,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.63181072473526,
+      "learning_rate": 0.0005997657928060715,
+      "loss": 3.8411,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6216081976890564,
+      "learning_rate": 0.0005997651841601006,
+      "loss": 4.1454,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6391425728797913,
+      "learning_rate": 0.0005997645747246058,
+      "loss": 4.0997,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6952412724494934,
+      "learning_rate": 0.0005997639644995888,
+      "loss": 4.2179,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6759364008903503,
+      "learning_rate": 0.0005997633534850514,
+      "loss": 4.1654,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6045799851417542,
+      "learning_rate": 0.0005997627416809948,
+      "loss": 4.2034,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6865895390510559,
+      "learning_rate": 0.0005997621290874209,
+      "loss": 4.0764,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6113295555114746,
+      "learning_rate": 0.0005997615157043312,
+      "loss": 4.2651,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.687916100025177,
+      "learning_rate": 0.0005997609015317275,
+      "loss": 4.2646,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6994914412498474,
+      "learning_rate": 0.0005997602865696111,
+      "loss": 3.9585,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6061134338378906,
+      "learning_rate": 0.0005997596708179839,
+      "loss": 4.0875,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.606025218963623,
+      "learning_rate": 0.0005997590542768475,
+      "loss": 4.321,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6358824968338013,
+      "learning_rate": 0.0005997584369462033,
+      "loss": 4.1621,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6684304475784302,
+      "learning_rate": 0.0005997578188260531,
+      "loss": 4.2266,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5966183543205261,
+      "learning_rate": 0.0005997571999163985,
+      "loss": 4.0322,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5628737211227417,
+      "learning_rate": 0.0005997565802172411,
+      "loss": 3.8574,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7218655347824097,
+      "learning_rate": 0.0005997559597285825,
+      "loss": 3.857,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6253189444541931,
+      "learning_rate": 0.0005997553384504246,
+      "loss": 4.1211,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7218075394630432,
+      "learning_rate": 0.0005997547163827686,
+      "loss": 3.6799,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5922057628631592,
+      "learning_rate": 0.0005997540935256165,
+      "loss": 4.3404,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6513999700546265,
+      "learning_rate": 0.0005997534698789698,
+      "loss": 4.0325,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6125958561897278,
+      "learning_rate": 0.0005997528454428301,
+      "loss": 3.9919,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6877921223640442,
+      "learning_rate": 0.0005997522202171991,
+      "loss": 4.1427,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6257963180541992,
+      "learning_rate": 0.0005997515942020785,
+      "loss": 3.9095,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6871378421783447,
+      "learning_rate": 0.0005997509673974699,
+      "loss": 4.2673,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7322933077812195,
+      "learning_rate": 0.0005997503398033748,
+      "loss": 4.0279,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6184875965118408,
+      "learning_rate": 0.0005997497114197952,
+      "loss": 4.0317,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.620513916015625,
+      "learning_rate": 0.0005997490822467323,
+      "loss": 4.1609,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6280710101127625,
+      "learning_rate": 0.0005997484522841882,
+      "loss": 3.9997,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6425851583480835,
+      "learning_rate": 0.0005997478215321642,
+      "loss": 4.3098,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6624174118041992,
+      "learning_rate": 0.0005997471899906623,
+      "loss": 3.9097,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6578688621520996,
+      "learning_rate": 0.0005997465576596839,
+      "loss": 4.2133,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6733409762382507,
+      "learning_rate": 0.0005997459245392308,
+      "loss": 4.0564,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6345073580741882,
+      "learning_rate": 0.0005997452906293045,
+      "loss": 4.1252,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6336762309074402,
+      "learning_rate": 0.0005997446559299069,
+      "loss": 4.1081,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6436794996261597,
+      "learning_rate": 0.0005997440204410395,
+      "loss": 4.0327,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0349910259246826,
+      "learning_rate": 0.0005997433841627042,
+      "loss": 4.0259,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6200885772705078,
+      "learning_rate": 0.0005997427470949023,
+      "loss": 4.1537,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6146049499511719,
+      "learning_rate": 0.0005997421092376358,
+      "loss": 4.2592,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6383852958679199,
+      "learning_rate": 0.0005997414705909062,
+      "loss": 3.9864,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6226063966751099,
+      "learning_rate": 0.0005997408311547153,
+      "loss": 4.0929,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6670148968696594,
+      "learning_rate": 0.0005997401909290648,
+      "loss": 4.1746,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6614567637443542,
+      "learning_rate": 0.0005997395499139562,
+      "loss": 3.9935,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5947785973548889,
+      "learning_rate": 0.0005997389081093914,
+      "loss": 4.0051,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6563360691070557,
+      "learning_rate": 0.000599738265515372,
+      "loss": 4.1408,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.626630961894989,
+      "learning_rate": 0.0005997376221318996,
+      "loss": 4.2392,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5788909792900085,
+      "learning_rate": 0.0005997369779589762,
+      "loss": 4.0375,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5952694416046143,
+      "learning_rate": 0.0005997363329966031,
+      "loss": 4.0138,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6254683136940002,
+      "learning_rate": 0.0005997356872447822,
+      "loss": 4.0292,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5747913718223572,
+      "learning_rate": 0.0005997350407035153,
+      "loss": 4.1415,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.638823926448822,
+      "learning_rate": 0.0005997343933728038,
+      "loss": 4.1323,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6529001593589783,
+      "learning_rate": 0.0005997337452526498,
+      "loss": 4.0443,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.62080317735672,
+      "learning_rate": 0.0005997330963430547,
+      "loss": 4.1244,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6070016622543335,
+      "learning_rate": 0.0005997324466440202,
+      "loss": 4.1765,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7610416412353516,
+      "learning_rate": 0.0005997317961555483,
+      "loss": 4.0887,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6447186470031738,
+      "learning_rate": 0.0005997311448776404,
+      "loss": 4.2312,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6255171298980713,
+      "learning_rate": 0.0005997304928102985,
+      "loss": 3.8122,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6879158616065979,
+      "learning_rate": 0.0005997298399535241,
+      "loss": 4.0435,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6047009229660034,
+      "learning_rate": 0.000599729186307319,
+      "loss": 4.0208,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5998377203941345,
+      "learning_rate": 0.0005997285318716847,
+      "loss": 4.055,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5977486371994019,
+      "learning_rate": 0.0005997278766466233,
+      "loss": 4.238,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5870335698127747,
+      "learning_rate": 0.0005997272206321365,
+      "loss": 3.7925,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6094736456871033,
+      "learning_rate": 0.0005997265638282256,
+      "loss": 4.2518,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6728547215461731,
+      "learning_rate": 0.0005997259062348929,
+      "loss": 4.1802,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5765520334243774,
+      "learning_rate": 0.0005997252478521396,
+      "loss": 3.8716,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.604479193687439,
+      "learning_rate": 0.0005997245886799679,
+      "loss": 4.1638,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7261841297149658,
+      "learning_rate": 0.0005997239287183791,
+      "loss": 4.0187,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6174228191375732,
+      "learning_rate": 0.0005997232679673752,
+      "loss": 4.0628,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5952420234680176,
+      "learning_rate": 0.0005997226064269579,
+      "loss": 4.1506,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5986089110374451,
+      "learning_rate": 0.0005997219440971291,
+      "loss": 4.0894,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5981305241584778,
+      "learning_rate": 0.0005997212809778903,
+      "loss": 3.893,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.747908353805542,
+      "learning_rate": 0.0005997206170692432,
+      "loss": 4.2173,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6060889959335327,
+      "learning_rate": 0.0005997199523711899,
+      "loss": 4.1177,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6315199732780457,
+      "learning_rate": 0.0005997192868837317,
+      "loss": 4.1163,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6924545764923096,
+      "learning_rate": 0.0005997186206068708,
+      "loss": 3.9744,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6298468112945557,
+      "learning_rate": 0.0005997179535406086,
+      "loss": 4.0924,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5951611399650574,
+      "learning_rate": 0.000599717285684947,
+      "loss": 3.8841,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6348850727081299,
+      "learning_rate": 0.0005997166170398876,
+      "loss": 4.0741,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6831207275390625,
+      "learning_rate": 0.0005997159476054326,
+      "loss": 3.9413,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6047587394714355,
+      "learning_rate": 0.0005997152773815834,
+      "loss": 4.0532,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6082019805908203,
+      "learning_rate": 0.0005997146063683417,
+      "loss": 4.0897,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6189398169517517,
+      "learning_rate": 0.0005997139345657095,
+      "loss": 4.1492,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.624375581741333,
+      "learning_rate": 0.0005997132619736885,
+      "loss": 4.3931,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6344230771064758,
+      "learning_rate": 0.0005997125885922805,
+      "loss": 3.903,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6171305179595947,
+      "learning_rate": 0.0005997119144214872,
+      "loss": 4.1388,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6373316645622253,
+      "learning_rate": 0.0005997112394613102,
+      "loss": 4.3171,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5979803204536438,
+      "learning_rate": 0.0005997105637117516,
+      "loss": 4.0232,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6416492462158203,
+      "learning_rate": 0.0005997098871728131,
+      "loss": 4.4098,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6283490061759949,
+      "learning_rate": 0.0005997092098444964,
+      "loss": 3.9947,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5568273663520813,
+      "learning_rate": 0.0005997085317268033,
+      "loss": 4.0547,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6507592797279358,
+      "learning_rate": 0.0005997078528197356,
+      "loss": 3.9341,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.635384202003479,
+      "learning_rate": 0.0005997071731232952,
+      "loss": 4.0064,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5979812145233154,
+      "learning_rate": 0.0005997064926374837,
+      "loss": 3.8342,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6876376867294312,
+      "learning_rate": 0.000599705811362303,
+      "loss": 4.1938,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6582483649253845,
+      "learning_rate": 0.0005997051292977549,
+      "loss": 4.0655,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6212353706359863,
+      "learning_rate": 0.0005997044464438413,
+      "loss": 4.1384,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5523200631141663,
+      "learning_rate": 0.0005997037628005636,
+      "loss": 4.1858,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6378701329231262,
+      "learning_rate": 0.0005997030783679241,
+      "loss": 4.112,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5988345742225647,
+      "learning_rate": 0.0005997023931459242,
+      "loss": 4.095,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6165454983711243,
+      "learning_rate": 0.0005997017071345659,
+      "loss": 3.8559,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6014971733093262,
+      "learning_rate": 0.000599701020333851,
+      "loss": 3.8783,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6640937924385071,
+      "learning_rate": 0.0005997003327437812,
+      "loss": 4.1482,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6299766302108765,
+      "learning_rate": 0.0005996996443643585,
+      "loss": 4.0896,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6197918057441711,
+      "learning_rate": 0.0005996989551955847,
+      "loss": 3.9987,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6115261316299438,
+      "learning_rate": 0.0005996982652374614,
+      "loss": 4.0155,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6456928253173828,
+      "learning_rate": 0.0005996975744899906,
+      "loss": 4.2857,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6480864882469177,
+      "learning_rate": 0.000599696882953174,
+      "loss": 3.983,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6337156295776367,
+      "learning_rate": 0.0005996961906270136,
+      "loss": 4.0381,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.610868513584137,
+      "learning_rate": 0.0005996954975115111,
+      "loss": 4.123,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6607982516288757,
+      "learning_rate": 0.0005996948036066683,
+      "loss": 4.2524,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6953883171081543,
+      "learning_rate": 0.000599694108912487,
+      "loss": 3.9404,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6293426156044006,
+      "learning_rate": 0.0005996934134289692,
+      "loss": 3.9467,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6299402117729187,
+      "learning_rate": 0.0005996927171561166,
+      "loss": 4.2004,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7695729732513428,
+      "learning_rate": 0.0005996920200939309,
+      "loss": 3.9446,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.662330687046051,
+      "learning_rate": 0.0005996913222424144,
+      "loss": 4.1542,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6112087965011597,
+      "learning_rate": 0.0005996906236015684,
+      "loss": 4.1722,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6271827220916748,
+      "learning_rate": 0.000599689924171395,
+      "loss": 3.9089,
+      "step": 886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6005352735519409,
+      "learning_rate": 0.0005996892239518961,
+      "loss": 3.9845,
+      "step": 887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6041455864906311,
+      "learning_rate": 0.0005996885229430734,
+      "loss": 4.1323,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5910174250602722,
+      "learning_rate": 0.0005996878211449289,
+      "loss": 4.2648,
+      "step": 889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6130709648132324,
+      "learning_rate": 0.0005996871185574643,
+      "loss": 4.0298,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6232227087020874,
+      "learning_rate": 0.0005996864151806815,
+      "loss": 4.2897,
+      "step": 891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6949803233146667,
+      "learning_rate": 0.0005996857110145823,
+      "loss": 4.0198,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5966178774833679,
+      "learning_rate": 0.0005996850060591687,
+      "loss": 4.1471,
+      "step": 893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6501827836036682,
+      "learning_rate": 0.0005996843003144424,
+      "loss": 4.0682,
+      "step": 894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6076254844665527,
+      "learning_rate": 0.0005996835937804054,
+      "loss": 4.1748,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5926339030265808,
+      "learning_rate": 0.0005996828864570594,
+      "loss": 4.019,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6619225144386292,
+      "learning_rate": 0.0005996821783444064,
+      "loss": 3.9084,
+      "step": 897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6074878573417664,
+      "learning_rate": 0.0005996814694424483,
+      "loss": 4.0513,
+      "step": 898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7126258015632629,
+      "learning_rate": 0.0005996807597511868,
+      "loss": 4.0084,
+      "step": 899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6678495407104492,
+      "learning_rate": 0.0005996800492706237,
+      "loss": 3.9053,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6096392869949341,
+      "learning_rate": 0.0005996793380007613,
+      "loss": 3.9624,
+      "step": 901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6041682362556458,
+      "learning_rate": 0.000599678625941601,
+      "loss": 4.0106,
+      "step": 902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6434935331344604,
+      "learning_rate": 0.000599677913093145,
+      "loss": 4.0789,
+      "step": 903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7096984386444092,
+      "learning_rate": 0.000599677199455395,
+      "loss": 4.1337,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5979908108711243,
+      "learning_rate": 0.0005996764850283529,
+      "loss": 4.1843,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6335223913192749,
+      "learning_rate": 0.0005996757698120206,
+      "loss": 4.2159,
+      "step": 906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.685057520866394,
+      "learning_rate": 0.0005996750538064001,
+      "loss": 3.8512,
+      "step": 907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6344968676567078,
+      "learning_rate": 0.0005996743370114932,
+      "loss": 4.1279,
+      "step": 908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6115414500236511,
+      "learning_rate": 0.0005996736194273017,
+      "loss": 4.1364,
+      "step": 909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6755195260047913,
+      "learning_rate": 0.0005996729010538275,
+      "loss": 4.0321,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5959694981575012,
+      "learning_rate": 0.0005996721818910725,
+      "loss": 4.2033,
+      "step": 911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6436964869499207,
+      "learning_rate": 0.0005996714619390389,
+      "loss": 4.0886,
+      "step": 912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6065239310264587,
+      "learning_rate": 0.0005996707411977281,
+      "loss": 4.1504,
+      "step": 913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5962178707122803,
+      "learning_rate": 0.0005996700196671423,
+      "loss": 3.8577,
+      "step": 914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6235368847846985,
+      "learning_rate": 0.0005996692973472835,
+      "loss": 4.1173,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6299294233322144,
+      "learning_rate": 0.0005996685742381532,
+      "loss": 3.9826,
+      "step": 916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6154347062110901,
+      "learning_rate": 0.0005996678503397537,
+      "loss": 4.0741,
+      "step": 917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6413784623146057,
+      "learning_rate": 0.0005996671256520867,
+      "loss": 3.9614,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.631839394569397,
+      "learning_rate": 0.0005996664001751542,
+      "loss": 4.0662,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6405590772628784,
+      "learning_rate": 0.0005996656739089581,
+      "loss": 4.0365,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6293720006942749,
+      "learning_rate": 0.0005996649468535002,
+      "loss": 4.2496,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6363210082054138,
+      "learning_rate": 0.0005996642190087825,
+      "loss": 3.9949,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6418921947479248,
+      "learning_rate": 0.000599663490374807,
+      "loss": 4.0203,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6134077906608582,
+      "learning_rate": 0.0005996627609515755,
+      "loss": 3.9338,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6744232177734375,
+      "learning_rate": 0.00059966203073909,
+      "loss": 3.9921,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6086677312850952,
+      "learning_rate": 0.0005996612997373524,
+      "loss": 3.9985,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6225132942199707,
+      "learning_rate": 0.0005996605679463644,
+      "loss": 4.061,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6579686999320984,
+      "learning_rate": 0.0005996598353661284,
+      "loss": 4.1547,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6591225862503052,
+      "learning_rate": 0.0005996591019966459,
+      "loss": 4.1103,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.601435661315918,
+      "learning_rate": 0.0005996583678379191,
+      "loss": 3.9549,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6121599674224854,
+      "learning_rate": 0.0005996576328899497,
+      "loss": 4.0187,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.62762850522995,
+      "learning_rate": 0.00059965689715274,
+      "loss": 4.0643,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.638881266117096,
+      "learning_rate": 0.0005996561606262914,
+      "loss": 4.0128,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5881949067115784,
+      "learning_rate": 0.0005996554233106063,
+      "loss": 3.9358,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6519078612327576,
+      "learning_rate": 0.0005996546852056865,
+      "loss": 3.9074,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6900700926780701,
+      "learning_rate": 0.0005996539463115339,
+      "loss": 4.0253,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6109591126441956,
+      "learning_rate": 0.0005996532066281505,
+      "loss": 4.1243,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6539261341094971,
+      "learning_rate": 0.0005996524661555382,
+      "loss": 4.1093,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6261901259422302,
+      "learning_rate": 0.000599651724893699,
+      "loss": 4.0016,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6098498702049255,
+      "learning_rate": 0.0005996509828426347,
+      "loss": 4.0939,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6256742477416992,
+      "learning_rate": 0.0005996502400023475,
+      "loss": 4.1468,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6049453616142273,
+      "learning_rate": 0.0005996494963728392,
+      "loss": 4.1078,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.593472957611084,
+      "learning_rate": 0.0005996487519541119,
+      "loss": 4.2345,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6599992513656616,
+      "learning_rate": 0.0005996480067461673,
+      "loss": 4.0287,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6049973964691162,
+      "learning_rate": 0.0005996472607490076,
+      "loss": 3.9218,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6687982678413391,
+      "learning_rate": 0.0005996465139626346,
+      "loss": 4.0952,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6398851275444031,
+      "learning_rate": 0.0005996457663870505,
+      "loss": 3.7504,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6496148109436035,
+      "learning_rate": 0.0005996450180222571,
+      "loss": 3.9623,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6317514777183533,
+      "learning_rate": 0.0005996442688682562,
+      "loss": 4.1198,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6482225656509399,
+      "learning_rate": 0.0005996435189250501,
+      "loss": 3.883,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.607512891292572,
+      "learning_rate": 0.0005996427681926406,
+      "loss": 4.0248,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6879410147666931,
+      "learning_rate": 0.0005996420166710297,
+      "loss": 4.0689,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6295230984687805,
+      "learning_rate": 0.0005996412643602196,
+      "loss": 4.2013,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6885456442832947,
+      "learning_rate": 0.0005996405112602119,
+      "loss": 4.1025,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5858647227287292,
+      "learning_rate": 0.0005996397573710087,
+      "loss": 4.0533,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6196497082710266,
+      "learning_rate": 0.0005996390026926121,
+      "loss": 4.094,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6471155285835266,
+      "learning_rate": 0.0005996382472250241,
+      "loss": 3.8297,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6373564004898071,
+      "learning_rate": 0.0005996374909682465,
+      "loss": 4.142,
+      "step": 958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6640136241912842,
+      "learning_rate": 0.0005996367339222815,
+      "loss": 4.2055,
+      "step": 959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6296126842498779,
+      "learning_rate": 0.0005996359760871311,
+      "loss": 3.9527,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6657034754753113,
+      "learning_rate": 0.0005996352174627971,
+      "loss": 4.1435,
+      "step": 961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.632871687412262,
+      "learning_rate": 0.0005996344580492817,
+      "loss": 3.8951,
+      "step": 962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6474413275718689,
+      "learning_rate": 0.0005996336978465866,
+      "loss": 4.1062,
+      "step": 963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5804777145385742,
+      "learning_rate": 0.0005996329368547141,
+      "loss": 3.9748,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6162649989128113,
+      "learning_rate": 0.0005996321750736662,
+      "loss": 3.8642,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5436090230941772,
+      "learning_rate": 0.0005996314125034447,
+      "loss": 4.0974,
+      "step": 966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.648061990737915,
+      "learning_rate": 0.0005996306491440517,
+      "loss": 3.9705,
+      "step": 967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6477342844009399,
+      "learning_rate": 0.0005996298849954894,
+      "loss": 4.1531,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6069393157958984,
+      "learning_rate": 0.0005996291200577596,
+      "loss": 4.0453,
+      "step": 969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6199972629547119,
+      "learning_rate": 0.0005996283543308643,
+      "loss": 4.2185,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6699131727218628,
+      "learning_rate": 0.0005996275878148055,
+      "loss": 4.0178,
+      "step": 971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7701675295829773,
+      "learning_rate": 0.0005996268205095855,
+      "loss": 3.9939,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5983409285545349,
+      "learning_rate": 0.000599626052415206,
+      "loss": 3.9987,
+      "step": 973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6925124526023865,
+      "learning_rate": 0.0005996252835316692,
+      "loss": 3.917,
+      "step": 974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6102199554443359,
+      "learning_rate": 0.000599624513858977,
+      "loss": 3.9685,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6447969079017639,
+      "learning_rate": 0.0005996237433971315,
+      "loss": 4.3062,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.606885552406311,
+      "learning_rate": 0.0005996229721461347,
+      "loss": 4.2991,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6305561065673828,
+      "learning_rate": 0.0005996222001059887,
+      "loss": 4.1132,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6146825551986694,
+      "learning_rate": 0.0005996214272766955,
+      "loss": 3.8875,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6634001731872559,
+      "learning_rate": 0.0005996206536582571,
+      "loss": 3.9914,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.623626708984375,
+      "learning_rate": 0.0005996198792506755,
+      "loss": 4.1724,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6683825850486755,
+      "learning_rate": 0.000599619104053953,
+      "loss": 4.0585,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5538197159767151,
+      "learning_rate": 0.0005996183280680914,
+      "loss": 3.9352,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6251077055931091,
+      "learning_rate": 0.0005996175512930927,
+      "loss": 4.3014,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.593470573425293,
+      "learning_rate": 0.0005996167737289589,
+      "loss": 4.0789,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6242483258247375,
+      "learning_rate": 0.0005996159953756923,
+      "loss": 4.0246,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6253484487533569,
+      "learning_rate": 0.0005996152162332949,
+      "loss": 3.9758,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6085364818572998,
+      "learning_rate": 0.0005996144363017686,
+      "loss": 3.8767,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6171862483024597,
+      "learning_rate": 0.0005996136555811156,
+      "loss": 4.01,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5950732231140137,
+      "learning_rate": 0.0005996128740713379,
+      "loss": 3.7366,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6203235983848572,
+      "learning_rate": 0.0005996120917724374,
+      "loss": 4.0769,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.608241856098175,
+      "learning_rate": 0.0005996113086844165,
+      "loss": 3.8565,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.612503707408905,
+      "learning_rate": 0.0005996105248072769,
+      "loss": 4.1548,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5867249369621277,
+      "learning_rate": 0.000599609740141021,
+      "loss": 3.8754,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6069759726524353,
+      "learning_rate": 0.0005996089546856507,
+      "loss": 4.0749,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5676235556602478,
+      "learning_rate": 0.000599608168441168,
+      "loss": 3.8812,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6222755908966064,
+      "learning_rate": 0.0005996073814075751,
+      "loss": 3.8984,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5938562750816345,
+      "learning_rate": 0.0005996065935848739,
+      "loss": 3.961,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5748042464256287,
+      "learning_rate": 0.0005996058049730666,
+      "loss": 3.823,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6213597059249878,
+      "learning_rate": 0.0005996050155721553,
+      "loss": 3.9996,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.622829258441925,
+      "learning_rate": 0.0005996042253821422,
+      "loss": 3.9153,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5811119675636292,
+      "learning_rate": 0.000599603434403029,
+      "loss": 4.0641,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6401235461235046,
+      "learning_rate": 0.0005996026426348183,
+      "loss": 4.0072,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6247691512107849,
+      "learning_rate": 0.0005996018500775117,
+      "loss": 4.0948,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8084251284599304,
+      "learning_rate": 0.0005996010567311115,
+      "loss": 4.232,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.611668586730957,
+      "learning_rate": 0.0005996002625956198,
+      "loss": 3.9149,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6284043788909912,
+      "learning_rate": 0.0005995994676710386,
+      "loss": 4.1114,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6413992047309875,
+      "learning_rate": 0.0005995986719573701,
+      "loss": 4.0627,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.769960880279541,
+      "learning_rate": 0.0005995978754546165,
+      "loss": 4.0868,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6613034605979919,
+      "learning_rate": 0.0005995970781627795,
+      "loss": 3.9804,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6815891265869141,
+      "learning_rate": 0.0005995962800818617,
+      "loss": 3.9572,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6836370825767517,
+      "learning_rate": 0.0005995954812118648,
+      "loss": 3.9664,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6569324731826782,
+      "learning_rate": 0.000599594681552791,
+      "loss": 4.1149,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6019830703735352,
+      "learning_rate": 0.0005995938811046426,
+      "loss": 3.9853,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.612848699092865,
+      "learning_rate": 0.0005995930798674216,
+      "loss": 4.155,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5986717343330383,
+      "learning_rate": 0.00059959227784113,
+      "loss": 4.0356,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6132051348686218,
+      "learning_rate": 0.0005995914750257701,
+      "loss": 3.7446,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6531481742858887,
+      "learning_rate": 0.0005995906714213439,
+      "loss": 3.7868,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6747217178344727,
+      "learning_rate": 0.0005995898670278535,
+      "loss": 4.0415,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6562297344207764,
+      "learning_rate": 0.0005995890618453009,
+      "loss": 3.7626,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8349602222442627,
+      "learning_rate": 0.0005995882558736885,
+      "loss": 3.96,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7195594906806946,
+      "learning_rate": 0.0005995874491130183,
+      "loss": 3.9682,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5842819809913635,
+      "learning_rate": 0.0005995866415632924,
+      "loss": 4.1458,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6021990776062012,
+      "learning_rate": 0.0005995858332245129,
+      "loss": 3.7984,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6178818941116333,
+      "learning_rate": 0.000599585024096682,
+      "loss": 3.95,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7198466062545776,
+      "learning_rate": 0.0005995842141798018,
+      "loss": 3.782,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6762264370918274,
+      "learning_rate": 0.0005995834034738744,
+      "loss": 4.1151,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6355637311935425,
+      "learning_rate": 0.0005995825919789021,
+      "loss": 4.1855,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6325914859771729,
+      "learning_rate": 0.0005995817796948869,
+      "loss": 3.8144,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6539778709411621,
+      "learning_rate": 0.0005995809666218308,
+      "loss": 4.0136,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.701380729675293,
+      "learning_rate": 0.0005995801527597362,
+      "loss": 4.0071,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.617536723613739,
+      "learning_rate": 0.0005995793381086051,
+      "loss": 3.9345,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5975783467292786,
+      "learning_rate": 0.0005995785226684397,
+      "loss": 3.9294,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5805318355560303,
+      "learning_rate": 0.0005995777064392421,
+      "loss": 4.0045,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6723288893699646,
+      "learning_rate": 0.0005995768894210144,
+      "loss": 3.9988,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7065977454185486,
+      "learning_rate": 0.000599576071613759,
+      "loss": 3.8351,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7041149139404297,
+      "learning_rate": 0.0005995752530174777,
+      "loss": 3.907,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6054835319519043,
+      "learning_rate": 0.0005995744336321729,
+      "loss": 3.8096,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6375568509101868,
+      "learning_rate": 0.0005995736134578468,
+      "loss": 3.9835,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6124348640441895,
+      "learning_rate": 0.0005995727924945014,
+      "loss": 3.9451,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.787834644317627,
+      "learning_rate": 0.0005995719707421388,
+      "loss": 3.9742,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6277994513511658,
+      "learning_rate": 0.0005995711482007615,
+      "loss": 3.9475,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6354180574417114,
+      "learning_rate": 0.0005995703248703713,
+      "loss": 4.2671,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5767564177513123,
+      "learning_rate": 0.0005995695007509705,
+      "loss": 4.0588,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6037499904632568,
+      "learning_rate": 0.0005995686758425614,
+      "loss": 3.8606,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7180562019348145,
+      "learning_rate": 0.000599567850145146,
+      "loss": 3.6251,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6226548552513123,
+      "learning_rate": 0.0005995670236587265,
+      "loss": 3.8062,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6276943683624268,
+      "learning_rate": 0.0005995661963833051,
+      "loss": 4.17,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6406780481338501,
+      "learning_rate": 0.0005995653683188841,
+      "loss": 4.0168,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.660855770111084,
+      "learning_rate": 0.0005995645394654655,
+      "loss": 3.7325,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7150101661682129,
+      "learning_rate": 0.0005995637098230516,
+      "loss": 4.2903,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6027021408081055,
+      "learning_rate": 0.0005995628793916445,
+      "loss": 3.9363,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6147125959396362,
+      "learning_rate": 0.0005995620481712464,
+      "loss": 4.002,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6540363430976868,
+      "learning_rate": 0.0005995612161618595,
+      "loss": 4.1231,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7382659316062927,
+      "learning_rate": 0.0005995603833634862,
+      "loss": 3.7431,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6010420918464661,
+      "learning_rate": 0.0005995595497761283,
+      "loss": 4.0042,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6575014591217041,
+      "learning_rate": 0.0005995587153997882,
+      "loss": 3.9075,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6506360769271851,
+      "learning_rate": 0.0005995578802344682,
+      "loss": 4.0998,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.670628547668457,
+      "learning_rate": 0.0005995570442801703,
+      "loss": 3.9776,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6330320239067078,
+      "learning_rate": 0.0005995562075368969,
+      "loss": 3.9504,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.604716956615448,
+      "learning_rate": 0.0005995553700046501,
+      "loss": 3.8521,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6562084555625916,
+      "learning_rate": 0.000599554531683432,
+      "loss": 4.1304,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6587114334106445,
+      "learning_rate": 0.000599553692573245,
+      "loss": 3.8732,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6262410283088684,
+      "learning_rate": 0.0005995528526740911,
+      "loss": 3.9372,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6062853336334229,
+      "learning_rate": 0.0005995520119859727,
+      "loss": 4.017,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6965979337692261,
+      "learning_rate": 0.000599551170508892,
+      "loss": 3.9419,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6158636212348938,
+      "learning_rate": 0.0005995503282428512,
+      "loss": 3.903,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6285214424133301,
+      "learning_rate": 0.0005995494851878524,
+      "loss": 4.057,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7115497589111328,
+      "learning_rate": 0.0005995486413438979,
+      "loss": 3.8139,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5884736776351929,
+      "learning_rate": 0.00059954779671099,
+      "loss": 3.9513,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6723683476448059,
+      "learning_rate": 0.0005995469512891308,
+      "loss": 3.9108,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.605546236038208,
+      "learning_rate": 0.0005995461050783226,
+      "loss": 4.0745,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5845876336097717,
+      "learning_rate": 0.0005995452580785676,
+      "loss": 4.0279,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6146058440208435,
+      "learning_rate": 0.000599544410289868,
+      "loss": 4.0084,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6121573448181152,
+      "learning_rate": 0.0005995435617122261,
+      "loss": 3.9437,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6386142373085022,
+      "learning_rate": 0.0005995427123456441,
+      "loss": 4.1899,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.607537567615509,
+      "learning_rate": 0.0005995418621901243,
+      "loss": 3.7694,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8437908291816711,
+      "learning_rate": 0.0005995410112456688,
+      "loss": 3.7674,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6184353232383728,
+      "learning_rate": 0.00059954015951228,
+      "loss": 3.8997,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6318781971931458,
+      "learning_rate": 0.0005995393069899601,
+      "loss": 3.9576,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6466130614280701,
+      "learning_rate": 0.0005995384536787112,
+      "loss": 4.0041,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.668156087398529,
+      "learning_rate": 0.0005995375995785357,
+      "loss": 4.0425,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5986151695251465,
+      "learning_rate": 0.0005995367446894358,
+      "loss": 3.7717,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6351150870323181,
+      "learning_rate": 0.0005995358890114137,
+      "loss": 4.0434,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6070703864097595,
+      "learning_rate": 0.0005995350325444719,
+      "loss": 4.0091,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7131597399711609,
+      "learning_rate": 0.0005995341752886124,
+      "loss": 3.9062,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6343297958374023,
+      "learning_rate": 0.0005995333172438374,
+      "loss": 4.2335,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6075549721717834,
+      "learning_rate": 0.0005995324584101494,
+      "loss": 3.9601,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5800171494483948,
+      "learning_rate": 0.0005995315987875505,
+      "loss": 4.0497,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5841231346130371,
+      "learning_rate": 0.0005995307383760431,
+      "loss": 3.9826,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6168744564056396,
+      "learning_rate": 0.0005995298771756293,
+      "loss": 4.2034,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.660862922668457,
+      "learning_rate": 0.0005995290151863114,
+      "loss": 4.1355,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7201418280601501,
+      "learning_rate": 0.0005995281524080919,
+      "loss": 3.7802,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6478432416915894,
+      "learning_rate": 0.0005995272888409728,
+      "loss": 3.8778,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6431102156639099,
+      "learning_rate": 0.0005995264244849564,
+      "loss": 3.8396,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6155719757080078,
+      "learning_rate": 0.0005995255593400452,
+      "loss": 3.9207,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6385898590087891,
+      "learning_rate": 0.0005995246934062412,
+      "loss": 3.9631,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6196228265762329,
+      "learning_rate": 0.0005995238266835468,
+      "loss": 3.9081,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6495975255966187,
+      "learning_rate": 0.0005995229591719643,
+      "loss": 4.0299,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6047267913818359,
+      "learning_rate": 0.000599522090871496,
+      "loss": 4.1038,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5929258465766907,
+      "learning_rate": 0.0005995212217821441,
+      "loss": 4.0514,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6673069000244141,
+      "learning_rate": 0.000599520351903911,
+      "loss": 3.8946,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.597668468952179,
+      "learning_rate": 0.000599519481236799,
+      "loss": 3.8061,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6228914260864258,
+      "learning_rate": 0.0005995186097808102,
+      "loss": 4.0257,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6423768997192383,
+      "learning_rate": 0.0005995177375359471,
+      "loss": 4.0893,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6223159432411194,
+      "learning_rate": 0.0005995168645022118,
+      "loss": 4.0435,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6253622174263,
+      "learning_rate": 0.0005995159906796068,
+      "loss": 3.755,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6040705442428589,
+      "learning_rate": 0.0005995151160681344,
+      "loss": 4.1872,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5690031051635742,
+      "learning_rate": 0.0005995142406677967,
+      "loss": 3.9943,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6597207188606262,
+      "learning_rate": 0.0005995133644785963,
+      "loss": 3.9716,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6696934103965759,
+      "learning_rate": 0.0005995124875005352,
+      "loss": 3.759,
+      "step": 1111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6375954747200012,
+      "learning_rate": 0.000599511609733616,
+      "loss": 3.8252,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.636753499507904,
+      "learning_rate": 0.0005995107311778406,
+      "loss": 3.9468,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5825153589248657,
+      "learning_rate": 0.0005995098518332117,
+      "loss": 3.8568,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6210393309593201,
+      "learning_rate": 0.0005995089716997316,
+      "loss": 4.1555,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6214787364006042,
+      "learning_rate": 0.0005995080907774024,
+      "loss": 3.9536,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.678767740726471,
+      "learning_rate": 0.0005995072090662267,
+      "loss": 3.835,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6073644161224365,
+      "learning_rate": 0.0005995063265662063,
+      "loss": 4.2358,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6090550422668457,
+      "learning_rate": 0.0005995054432773442,
+      "loss": 3.9443,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.715369462966919,
+      "learning_rate": 0.0005995045591996423,
+      "loss": 4.0607,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5823421478271484,
+      "learning_rate": 0.0005995036743331031,
+      "loss": 4.0344,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6796244382858276,
+      "learning_rate": 0.0005995027886777287,
+      "loss": 3.9048,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5604933500289917,
+      "learning_rate": 0.0005995019022335217,
+      "loss": 4.0935,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6292743682861328,
+      "learning_rate": 0.0005995010150004844,
+      "loss": 3.9694,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6105656623840332,
+      "learning_rate": 0.0005995001269786189,
+      "loss": 3.8372,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6465719938278198,
+      "learning_rate": 0.0005994992381679277,
+      "loss": 3.7912,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.704645574092865,
+      "learning_rate": 0.0005994983485684133,
+      "loss": 3.8925,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6321631073951721,
+      "learning_rate": 0.0005994974581800779,
+      "loss": 4.1433,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6182036399841309,
+      "learning_rate": 0.0005994965670029237,
+      "loss": 3.9114,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6689995527267456,
+      "learning_rate": 0.0005994956750369532,
+      "loss": 4.0865,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6217949986457825,
+      "learning_rate": 0.0005994947822821688,
+      "loss": 4.1664,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.608690083026886,
+      "learning_rate": 0.0005994938887385727,
+      "loss": 4.0256,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6573523283004761,
+      "learning_rate": 0.0005994929944061675,
+      "loss": 3.9891,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6570764780044556,
+      "learning_rate": 0.0005994920992849552,
+      "loss": 3.9587,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5835046172142029,
+      "learning_rate": 0.0005994912033749385,
+      "loss": 4.1042,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6346054077148438,
+      "learning_rate": 0.0005994903066761196,
+      "loss": 4.0866,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6046596169471741,
+      "learning_rate": 0.0005994894091885007,
+      "loss": 3.9774,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6285713911056519,
+      "learning_rate": 0.0005994885109120845,
+      "loss": 3.9324,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5951323509216309,
+      "learning_rate": 0.0005994876118468731,
+      "loss": 3.8939,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6478222012519836,
+      "learning_rate": 0.0005994867119928689,
+      "loss": 3.9917,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6328380703926086,
+      "learning_rate": 0.0005994858113500746,
+      "loss": 3.8191,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6400364637374878,
+      "learning_rate": 0.0005994849099184921,
+      "loss": 4.0963,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5783833861351013,
+      "learning_rate": 0.000599484007698124,
+      "loss": 3.9709,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6741927862167358,
+      "learning_rate": 0.0005994831046889727,
+      "loss": 3.8906,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6302924156188965,
+      "learning_rate": 0.0005994822008910406,
+      "loss": 3.7855,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5713663101196289,
+      "learning_rate": 0.0005994812963043299,
+      "loss": 3.963,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5913187265396118,
+      "learning_rate": 0.0005994803909288431,
+      "loss": 3.9689,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6074280738830566,
+      "learning_rate": 0.0005994794847645827,
+      "loss": 3.9716,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.626340925693512,
+      "learning_rate": 0.000599478577811551,
+      "loss": 4.1273,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5834325551986694,
+      "learning_rate": 0.0005994776700697503,
+      "loss": 3.8578,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.621330738067627,
+      "learning_rate": 0.000599476761539183,
+      "loss": 3.8196,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6001361608505249,
+      "learning_rate": 0.0005994758522198516,
+      "loss": 3.9403,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6255820393562317,
+      "learning_rate": 0.0005994749421117584,
+      "loss": 4.1259,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.631493330001831,
+      "learning_rate": 0.0005994740312149058,
+      "loss": 4.104,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6368297934532166,
+      "learning_rate": 0.0005994731195292964,
+      "loss": 4.0128,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6331319808959961,
+      "learning_rate": 0.0005994722070549323,
+      "loss": 3.8825,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5888481736183167,
+      "learning_rate": 0.000599471293791816,
+      "loss": 4.1243,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5803425312042236,
+      "learning_rate": 0.00059947037973995,
+      "loss": 3.9828,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5989425182342529,
+      "learning_rate": 0.0005994694648993368,
+      "loss": 3.7897,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6215832233428955,
+      "learning_rate": 0.0005994685492699785,
+      "loss": 3.6121,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6019163131713867,
+      "learning_rate": 0.0005994676328518778,
+      "loss": 4.244,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5976054668426514,
+      "learning_rate": 0.0005994667156450368,
+      "loss": 4.1499,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6580859422683716,
+      "learning_rate": 0.0005994657976494583,
+      "loss": 3.814,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6277784705162048,
+      "learning_rate": 0.0005994648788651445,
+      "loss": 4.0219,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6280736327171326,
+      "learning_rate": 0.0005994639592920976,
+      "loss": 4.0019,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7152631878852844,
+      "learning_rate": 0.0005994630389303205,
+      "loss": 3.9115,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6251851916313171,
+      "learning_rate": 0.0005994621177798153,
+      "loss": 4.104,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6360085606575012,
+      "learning_rate": 0.0005994611958405846,
+      "loss": 3.9413,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5834187865257263,
+      "learning_rate": 0.0005994602731126306,
+      "loss": 3.8453,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6263565421104431,
+      "learning_rate": 0.000599459349595956,
+      "loss": 3.9509,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6582068204879761,
+      "learning_rate": 0.000599458425290563,
+      "loss": 3.7823,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6431542038917542,
+      "learning_rate": 0.0005994575001964542,
+      "loss": 3.9593,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6388793587684631,
+      "learning_rate": 0.0005994565743136321,
+      "loss": 3.8094,
+      "step": 1173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6203798055648804,
+      "learning_rate": 0.0005994556476420988,
+      "loss": 3.8218,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6218488812446594,
+      "learning_rate": 0.0005994547201818571,
+      "loss": 3.8657,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6742687225341797,
+      "learning_rate": 0.0005994537919329092,
+      "loss": 3.8143,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6174846887588501,
+      "learning_rate": 0.0005994528628952577,
+      "loss": 3.8962,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7459059357643127,
+      "learning_rate": 0.000599451933068905,
+      "loss": 4.0231,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.602178692817688,
+      "learning_rate": 0.0005994510024538535,
+      "loss": 4.0269,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6028525233268738,
+      "learning_rate": 0.0005994500710501058,
+      "loss": 4.0828,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6567595601081848,
+      "learning_rate": 0.0005994491388576641,
+      "loss": 4.0124,
+      "step": 1181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6130238771438599,
+      "learning_rate": 0.000599448205876531,
+      "loss": 4.0882,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5951813459396362,
+      "learning_rate": 0.000599447272106709,
+      "loss": 3.8659,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6704297065734863,
+      "learning_rate": 0.0005994463375482006,
+      "loss": 4.0565,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6474865674972534,
+      "learning_rate": 0.0005994454022010081,
+      "loss": 4.048,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5812511444091797,
+      "learning_rate": 0.0005994444660651339,
+      "loss": 3.8526,
+      "step": 1186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5992949604988098,
+      "learning_rate": 0.0005994435291405808,
+      "loss": 3.9835,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6317020654678345,
+      "learning_rate": 0.0005994425914273509,
+      "loss": 3.8099,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6197634339332581,
+      "learning_rate": 0.000599441652925447,
+      "loss": 3.8242,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6547470092773438,
+      "learning_rate": 0.0005994407136348715,
+      "loss": 3.9396,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.594266951084137,
+      "learning_rate": 0.0005994397735556267,
+      "loss": 3.9765,
+      "step": 1191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.647244930267334,
+      "learning_rate": 0.000599438832687715,
+      "loss": 3.8405,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6598258018493652,
+      "learning_rate": 0.0005994378910311393,
+      "loss": 3.7971,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6706022620201111,
+      "learning_rate": 0.0005994369485859016,
+      "loss": 3.9114,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5980822443962097,
+      "learning_rate": 0.0005994360053520047,
+      "loss": 3.8791,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6569104194641113,
+      "learning_rate": 0.000599435061329451,
+      "loss": 3.9084,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6087574362754822,
+      "learning_rate": 0.000599434116518243,
+      "loss": 3.7479,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6148558259010315,
+      "learning_rate": 0.0005994331709183831,
+      "loss": 3.9402,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6539002656936646,
+      "learning_rate": 0.0005994322245298739,
+      "loss": 3.9145,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6310211420059204,
+      "learning_rate": 0.0005994312773527178,
+      "loss": 3.8744,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.630675196647644,
+      "learning_rate": 0.0005994303293869175,
+      "loss": 3.8331,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5651322603225708,
+      "learning_rate": 0.000599429380632475,
+      "loss": 4.0296,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5851694941520691,
+      "learning_rate": 0.0005994284310893934,
+      "loss": 3.6882,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6342992186546326,
+      "learning_rate": 0.0005994274807576749,
+      "loss": 3.8372,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5836645364761353,
+      "learning_rate": 0.0005994265296373221,
+      "loss": 4.0055,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.628150999546051,
+      "learning_rate": 0.0005994255777283374,
+      "loss": 3.7181,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6232638955116272,
+      "learning_rate": 0.0005994246250307232,
+      "loss": 3.8581,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6497608423233032,
+      "learning_rate": 0.0005994236715444823,
+      "loss": 3.6943,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6445043087005615,
+      "learning_rate": 0.0005994227172696171,
+      "loss": 3.918,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5761111378669739,
+      "learning_rate": 0.0005994217622061301,
+      "loss": 3.9378,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5777541995048523,
+      "learning_rate": 0.0005994208063540237,
+      "loss": 3.8161,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6353874206542969,
+      "learning_rate": 0.0005994198497133006,
+      "loss": 3.8907,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6581094264984131,
+      "learning_rate": 0.0005994188922839633,
+      "loss": 3.8506,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.69733065366745,
+      "learning_rate": 0.0005994179340660141,
+      "loss": 4.0021,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6014714241027832,
+      "learning_rate": 0.0005994169750594558,
+      "loss": 3.829,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6537860035896301,
+      "learning_rate": 0.0005994160152642908,
+      "loss": 3.9849,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7314767837524414,
+      "learning_rate": 0.0005994150546805216,
+      "loss": 3.9777,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6319352984428406,
+      "learning_rate": 0.0005994140933081507,
+      "loss": 3.7624,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5953729152679443,
+      "learning_rate": 0.0005994131311471808,
+      "loss": 3.8365,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.609235405921936,
+      "learning_rate": 0.0005994121681976144,
+      "loss": 4.1573,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6277957558631897,
+      "learning_rate": 0.0005994112044594538,
+      "loss": 4.0577,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5987586379051208,
+      "learning_rate": 0.0005994102399327019,
+      "loss": 4.0041,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6414321660995483,
+      "learning_rate": 0.0005994092746173609,
+      "loss": 3.9939,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6084062457084656,
+      "learning_rate": 0.0005994083085134336,
+      "loss": 4.004,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6363257169723511,
+      "learning_rate": 0.0005994073416209225,
+      "loss": 3.7369,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6323297023773193,
+      "learning_rate": 0.0005994063739398299,
+      "loss": 3.8556,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7020496726036072,
+      "learning_rate": 0.0005994054054701585,
+      "loss": 3.9934,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6019614338874817,
+      "learning_rate": 0.000599404436211911,
+      "loss": 3.985,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6175892353057861,
+      "learning_rate": 0.0005994034661650899,
+      "loss": 3.9954,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.611221969127655,
+      "learning_rate": 0.0005994024953296976,
+      "loss": 3.852,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5849113464355469,
+      "learning_rate": 0.0005994015237057368,
+      "loss": 3.7631,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6553965210914612,
+      "learning_rate": 0.0005994005512932099,
+      "loss": 4.1467,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6709908843040466,
+      "learning_rate": 0.0005993995780921197,
+      "loss": 4.0218,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6183760762214661,
+      "learning_rate": 0.0005993986041024686,
+      "loss": 3.8164,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5950647592544556,
+      "learning_rate": 0.0005993976293242591,
+      "loss": 4.1307,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6528424620628357,
+      "learning_rate": 0.000599396653757494,
+      "loss": 3.8645,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6041496396064758,
+      "learning_rate": 0.0005993956774021757,
+      "loss": 4.0503,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6346700191497803,
+      "learning_rate": 0.0005993947002583067,
+      "loss": 3.978,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6335063576698303,
+      "learning_rate": 0.0005993937223258898,
+      "loss": 3.8234,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6392975449562073,
+      "learning_rate": 0.0005993927436049273,
+      "loss": 3.9517,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5925444960594177,
+      "learning_rate": 0.0005993917640954221,
+      "loss": 3.798,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6346644163131714,
+      "learning_rate": 0.0005993907837973766,
+      "loss": 3.9064,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5913667678833008,
+      "learning_rate": 0.0005993898027107933,
+      "loss": 3.9169,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7143539190292358,
+      "learning_rate": 0.000599388820835675,
+      "loss": 3.7224,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6183956265449524,
+      "learning_rate": 0.000599387838172024,
+      "loss": 3.7269,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.614376962184906,
+      "learning_rate": 0.0005993868547198432,
+      "loss": 3.9984,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5987918376922607,
+      "learning_rate": 0.000599385870479135,
+      "loss": 4.0514,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5989696383476257,
+      "learning_rate": 0.000599384885449902,
+      "loss": 3.9283,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6367683410644531,
+      "learning_rate": 0.0005993838996321468,
+      "loss": 4.1072,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6802504658699036,
+      "learning_rate": 0.0005993829130258721,
+      "loss": 3.7503,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6775780320167542,
+      "learning_rate": 0.0005993819256310804,
+      "loss": 4.0616,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5724057555198669,
+      "learning_rate": 0.0005993809374477744,
+      "loss": 3.735,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6689765453338623,
+      "learning_rate": 0.0005993799484759566,
+      "loss": 3.8647,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6421110033988953,
+      "learning_rate": 0.0005993789587156295,
+      "loss": 3.9385,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.696961522102356,
+      "learning_rate": 0.000599377968166796,
+      "loss": 3.847,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6425986886024475,
+      "learning_rate": 0.0005993769768294583,
+      "loss": 3.857,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6519110798835754,
+      "learning_rate": 0.0005993759847036195,
+      "loss": 3.846,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5740974545478821,
+      "learning_rate": 0.000599374991789282,
+      "loss": 3.8013,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6745760440826416,
+      "learning_rate": 0.0005993739980864483,
+      "loss": 3.9241,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6441271305084229,
+      "learning_rate": 0.000599373003595121,
+      "loss": 3.7695,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6193081140518188,
+      "learning_rate": 0.0005993720083153029,
+      "loss": 3.985,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6134259700775146,
+      "learning_rate": 0.0005993710122469966,
+      "loss": 3.7961,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6477553248405457,
+      "learning_rate": 0.0005993700153902046,
+      "loss": 3.9387,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6375543475151062,
+      "learning_rate": 0.0005993690177449298,
+      "loss": 4.016,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.615598738193512,
+      "learning_rate": 0.0005993680193111744,
+      "loss": 3.823,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6295831203460693,
+      "learning_rate": 0.0005993670200889412,
+      "loss": 3.7599,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6454052329063416,
+      "learning_rate": 0.0005993660200782331,
+      "loss": 3.9033,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6237940192222595,
+      "learning_rate": 0.0005993650192790525,
+      "loss": 3.7442,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6554262042045593,
+      "learning_rate": 0.0005993640176914019,
+      "loss": 3.8714,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.591227114200592,
+      "learning_rate": 0.0005993630153152841,
+      "loss": 3.6987,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5855788588523865,
+      "learning_rate": 0.0005993620121507018,
+      "loss": 3.872,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6025470495223999,
+      "learning_rate": 0.0005993610081976575,
+      "loss": 3.8734,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6828283071517944,
+      "learning_rate": 0.0005993600034561539,
+      "loss": 3.8746,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6492578983306885,
+      "learning_rate": 0.0005993589979261939,
+      "loss": 3.7428,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6107031106948853,
+      "learning_rate": 0.0005993579916077796,
+      "loss": 3.7983,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6975229382514954,
+      "learning_rate": 0.0005993569845009142,
+      "loss": 3.6364,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6226152777671814,
+      "learning_rate": 0.0005993559766056,
+      "loss": 3.9707,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6262789368629456,
+      "learning_rate": 0.0005993549679218398,
+      "loss": 3.9105,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5812487006187439,
+      "learning_rate": 0.0005993539584496362,
+      "loss": 3.5877,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5997166037559509,
+      "learning_rate": 0.000599352948188992,
+      "loss": 3.8289,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5774847269058228,
+      "learning_rate": 0.0005993519371399096,
+      "loss": 4.0577,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6224757432937622,
+      "learning_rate": 0.0005993509253023919,
+      "loss": 3.7593,
+      "step": 1282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5884726047515869,
+      "learning_rate": 0.0005993499126764415,
+      "loss": 4.0062,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6184850931167603,
+      "learning_rate": 0.0005993488992620608,
+      "loss": 3.7726,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6030580997467041,
+      "learning_rate": 0.000599347885059253,
+      "loss": 4.0764,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6678992509841919,
+      "learning_rate": 0.0005993468700680204,
+      "loss": 4.0444,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6198200583457947,
+      "learning_rate": 0.0005993458542883657,
+      "loss": 3.8496,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5855886936187744,
+      "learning_rate": 0.0005993448377202916,
+      "loss": 3.8943,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.684818685054779,
+      "learning_rate": 0.0005993438203638009,
+      "loss": 3.8199,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5692946910858154,
+      "learning_rate": 0.0005993428022188961,
+      "loss": 3.9652,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6946492791175842,
+      "learning_rate": 0.00059934178328558,
+      "loss": 4.1611,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5940146446228027,
+      "learning_rate": 0.0005993407635638553,
+      "loss": 3.9545,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5907729864120483,
+      "learning_rate": 0.0005993397430537246,
+      "loss": 3.9311,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6589469909667969,
+      "learning_rate": 0.0005993387217551907,
+      "loss": 3.795,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5783873200416565,
+      "learning_rate": 0.0005993376996682561,
+      "loss": 3.9994,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6160930395126343,
+      "learning_rate": 0.0005993366767929237,
+      "loss": 3.9999,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5932180881500244,
+      "learning_rate": 0.000599335653129196,
+      "loss": 4.0102,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.635309100151062,
+      "learning_rate": 0.0005993346286770759,
+      "loss": 3.7957,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6134257912635803,
+      "learning_rate": 0.000599333603436566,
+      "loss": 3.8315,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5721672773361206,
+      "learning_rate": 0.0005993325774076688,
+      "loss": 4.0402,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5775968432426453,
+      "learning_rate": 0.0005993315505903874,
+      "loss": 3.8583,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.618675172328949,
+      "learning_rate": 0.0005993305229847242,
+      "loss": 3.8447,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6464390754699707,
+      "learning_rate": 0.0005993294945906821,
+      "loss": 4.1014,
+      "step": 1303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.660879373550415,
+      "learning_rate": 0.0005993284654082638,
+      "loss": 3.9139,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5997679233551025,
+      "learning_rate": 0.0005993274354374718,
+      "loss": 3.9929,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6345100998878479,
+      "learning_rate": 0.0005993264046783089,
+      "loss": 4.0084,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5924948453903198,
+      "learning_rate": 0.000599325373130778,
+      "loss": 3.8919,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6048806309700012,
+      "learning_rate": 0.0005993243407948815,
+      "loss": 3.8444,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5796457529067993,
+      "learning_rate": 0.0005993233076706224,
+      "loss": 3.8349,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6050766706466675,
+      "learning_rate": 0.0005993222737580033,
+      "loss": 3.9042,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6207561492919922,
+      "learning_rate": 0.0005993212390570269,
+      "loss": 4.2432,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.722430408000946,
+      "learning_rate": 0.0005993202035676959,
+      "loss": 3.7475,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6052895784378052,
+      "learning_rate": 0.0005993191672900132,
+      "loss": 3.9763,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6674761772155762,
+      "learning_rate": 0.0005993181302239815,
+      "loss": 4.0753,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6250360012054443,
+      "learning_rate": 0.0005993170923696033,
+      "loss": 3.8684,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6339815855026245,
+      "learning_rate": 0.0005993160537268815,
+      "loss": 3.8393,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6206067204475403,
+      "learning_rate": 0.0005993150142958189,
+      "loss": 3.7968,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5794596672058105,
+      "learning_rate": 0.000599313974076418,
+      "loss": 3.9016,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.706591784954071,
+      "learning_rate": 0.0005993129330686818,
+      "loss": 3.796,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6444628834724426,
+      "learning_rate": 0.000599311891272613,
+      "loss": 3.8181,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5789697766304016,
+      "learning_rate": 0.0005993108486882142,
+      "loss": 3.9068,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6501109004020691,
+      "learning_rate": 0.0005993098053154882,
+      "loss": 3.836,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6823614239692688,
+      "learning_rate": 0.0005993087611544379,
+      "loss": 3.8886,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5772250890731812,
+      "learning_rate": 0.0005993077162050658,
+      "loss": 3.7892,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.674119770526886,
+      "learning_rate": 0.0005993066704673748,
+      "loss": 4.0057,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.541368305683136,
+      "learning_rate": 0.0005993056239413677,
+      "loss": 3.9643,
+      "step": 1326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6814593076705933,
+      "learning_rate": 0.0005993045766270472,
+      "loss": 3.8573,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.607441782951355,
+      "learning_rate": 0.000599303528524416,
+      "loss": 3.8976,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5948410630226135,
+      "learning_rate": 0.0005993024796334768,
+      "loss": 3.7664,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6131972074508667,
+      "learning_rate": 0.0005993014299542327,
+      "loss": 4.0518,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6085745096206665,
+      "learning_rate": 0.0005993003794866861,
+      "loss": 3.8089,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6138327717781067,
+      "learning_rate": 0.0005992993282308398,
+      "loss": 3.8027,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6015024781227112,
+      "learning_rate": 0.0005992982761866968,
+      "loss": 3.7218,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6075776815414429,
+      "learning_rate": 0.0005992972233542597,
+      "loss": 3.8605,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.603941798210144,
+      "learning_rate": 0.0005992961697335314,
+      "loss": 3.692,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6697930097579956,
+      "learning_rate": 0.0005992951153245146,
+      "loss": 3.7488,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6258661150932312,
+      "learning_rate": 0.000599294060127212,
+      "loss": 4.0558,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5765604972839355,
+      "learning_rate": 0.0005992930041416264,
+      "loss": 4.1127,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5883685946464539,
+      "learning_rate": 0.0005992919473677607,
+      "loss": 4.1005,
+      "step": 1339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6068031191825867,
+      "learning_rate": 0.0005992908898056176,
+      "loss": 4.0732,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6263118386268616,
+      "learning_rate": 0.0005992898314551999,
+      "loss": 4.0635,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.630369246006012,
+      "learning_rate": 0.0005992887723165102,
+      "loss": 3.9827,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5971052646636963,
+      "learning_rate": 0.0005992877123895516,
+      "loss": 3.8703,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6135636568069458,
+      "learning_rate": 0.0005992866516743269,
+      "loss": 3.8379,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.63181471824646,
+      "learning_rate": 0.0005992855901708387,
+      "loss": 3.7176,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6445341110229492,
+      "learning_rate": 0.0005992845278790898,
+      "loss": 3.7877,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6358599662780762,
+      "learning_rate": 0.000599283464799083,
+      "loss": 3.6308,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6615734100341797,
+      "learning_rate": 0.0005992824009308212,
+      "loss": 3.8782,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6317094564437866,
+      "learning_rate": 0.0005992813362743072,
+      "loss": 3.935,
+      "step": 1349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6740353107452393,
+      "learning_rate": 0.0005992802708295438,
+      "loss": 3.6289,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.604249894618988,
+      "learning_rate": 0.0005992792045965336,
+      "loss": 3.8375,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6756212115287781,
+      "learning_rate": 0.0005992781375752797,
+      "loss": 3.8007,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5996575355529785,
+      "learning_rate": 0.0005992770697657848,
+      "loss": 3.8507,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6114326119422913,
+      "learning_rate": 0.0005992760011680516,
+      "loss": 4.0967,
+      "step": 1354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6802350878715515,
+      "learning_rate": 0.000599274931782083,
+      "loss": 3.5996,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6232761740684509,
+      "learning_rate": 0.0005992738616078819,
+      "loss": 3.8647,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5817407369613647,
+      "learning_rate": 0.0005992727906454511,
+      "loss": 4.0256,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6215899586677551,
+      "learning_rate": 0.0005992717188947933,
+      "loss": 4.0016,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6303964257240295,
+      "learning_rate": 0.0005992706463559113,
+      "loss": 3.675,
+      "step": 1359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6754851937294006,
+      "learning_rate": 0.0005992695730288081,
+      "loss": 3.977,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5957140922546387,
+      "learning_rate": 0.0005992684989134864,
+      "loss": 3.8991,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6352207660675049,
+      "learning_rate": 0.0005992674240099491,
+      "loss": 4.0731,
+      "step": 1362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.634380042552948,
+      "learning_rate": 0.000599266348318199,
+      "loss": 3.8853,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6464090347290039,
+      "learning_rate": 0.000599265271838239,
+      "loss": 3.8244,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6003564596176147,
+      "learning_rate": 0.0005992641945700717,
+      "loss": 3.7622,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.674795389175415,
+      "learning_rate": 0.0005992631165137002,
+      "loss": 4.0823,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6135823726654053,
+      "learning_rate": 0.0005992620376691273,
+      "loss": 3.7498,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6516311764717102,
+      "learning_rate": 0.0005992609580363558,
+      "loss": 3.8798,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6563093066215515,
+      "learning_rate": 0.0005992598776153883,
+      "loss": 3.9035,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5973907709121704,
+      "learning_rate": 0.0005992587964062282,
+      "loss": 3.9296,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6241722702980042,
+      "learning_rate": 0.0005992577144088778,
+      "loss": 3.9025,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5976197123527527,
+      "learning_rate": 0.0005992566316233403,
+      "loss": 3.9788,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5596520304679871,
+      "learning_rate": 0.0005992555480496183,
+      "loss": 4.1813,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7279554009437561,
+      "learning_rate": 0.0005992544636877149,
+      "loss": 3.7869,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6230193376541138,
+      "learning_rate": 0.0005992533785376328,
+      "loss": 3.7344,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6615080833435059,
+      "learning_rate": 0.0005992522925993748,
+      "loss": 3.793,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6379623413085938,
+      "learning_rate": 0.000599251205872944,
+      "loss": 4.0211,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6381189823150635,
+      "learning_rate": 0.000599250118358343,
+      "loss": 3.9455,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6189027428627014,
+      "learning_rate": 0.0005992490300555748,
+      "loss": 3.8361,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5973135232925415,
+      "learning_rate": 0.0005992479409646424,
+      "loss": 4.1191,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6134343147277832,
+      "learning_rate": 0.0005992468510855483,
+      "loss": 3.5492,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5866838693618774,
+      "learning_rate": 0.0005992457604182957,
+      "loss": 3.7574,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6180101633071899,
+      "learning_rate": 0.0005992446689628873,
+      "loss": 3.9139,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6157495379447937,
+      "learning_rate": 0.0005992435767193261,
+      "loss": 3.9326,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6040421724319458,
+      "learning_rate": 0.0005992424836876149,
+      "loss": 3.9882,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.577000081539154,
+      "learning_rate": 0.0005992413898677565,
+      "loss": 4.0152,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6416754722595215,
+      "learning_rate": 0.0005992402952597539,
+      "loss": 3.7716,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6492377519607544,
+      "learning_rate": 0.00059923919986361,
+      "loss": 3.7164,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6138007640838623,
+      "learning_rate": 0.0005992381036793277,
+      "loss": 3.7265,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6215953230857849,
+      "learning_rate": 0.0005992370067069097,
+      "loss": 3.9165,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5956389307975769,
+      "learning_rate": 0.0005992359089463591,
+      "loss": 3.9124,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5724150538444519,
+      "learning_rate": 0.0005992348103976786,
+      "loss": 3.8611,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6579985022544861,
+      "learning_rate": 0.0005992337110608713,
+      "loss": 3.4344,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5854646563529968,
+      "learning_rate": 0.00059923261093594,
+      "loss": 3.828,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6103180646896362,
+      "learning_rate": 0.0005992315100228875,
+      "loss": 3.8028,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5735791325569153,
+      "learning_rate": 0.0005992304083217168,
+      "loss": 3.9679,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5912772417068481,
+      "learning_rate": 0.0005992293058324308,
+      "loss": 3.7386,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6024033427238464,
+      "learning_rate": 0.0005992282025550325,
+      "loss": 3.8819,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5784563422203064,
+      "learning_rate": 0.0005992270984895246,
+      "loss": 3.7262,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6742603182792664,
+      "learning_rate": 0.0005992259936359102,
+      "loss": 3.7894,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5925095677375793,
+      "learning_rate": 0.0005992248879941919,
+      "loss": 3.9177,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5972510576248169,
+      "learning_rate": 0.0005992237815643732,
+      "loss": 3.8498,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6998504400253296,
+      "learning_rate": 0.0005992226743464563,
+      "loss": 3.9739,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6386610269546509,
+      "learning_rate": 0.0005992215663404446,
+      "loss": 3.9983,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6281149983406067,
+      "learning_rate": 0.0005992204575463409,
+      "loss": 3.82,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6512899994850159,
+      "learning_rate": 0.0005992193479641481,
+      "loss": 3.5674,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.613243043422699,
+      "learning_rate": 0.0005992182375938692,
+      "loss": 3.8449,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6193988919258118,
+      "learning_rate": 0.0005992171264355069,
+      "loss": 3.8452,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6317744851112366,
+      "learning_rate": 0.0005992160144890643,
+      "loss": 3.7912,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.604690432548523,
+      "learning_rate": 0.0005992149017545443,
+      "loss": 3.8883,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5913760662078857,
+      "learning_rate": 0.0005992137882319498,
+      "loss": 3.9683,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6362264156341553,
+      "learning_rate": 0.0005992126739212839,
+      "loss": 3.9218,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5891861319541931,
+      "learning_rate": 0.0005992115588225493,
+      "loss": 3.8159,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6369192004203796,
+      "learning_rate": 0.0005992104429357491,
+      "loss": 3.9504,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6242348551750183,
+      "learning_rate": 0.0005992093262608861,
+      "loss": 3.7089,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5739800333976746,
+      "learning_rate": 0.0005992082087979635,
+      "loss": 3.9268,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7472577095031738,
+      "learning_rate": 0.0005992070905469839,
+      "loss": 3.9839,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5870568156242371,
+      "learning_rate": 0.0005992059715079504,
+      "loss": 3.7954,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6301756501197815,
+      "learning_rate": 0.000599204851680866,
+      "loss": 3.9817,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6080438494682312,
+      "learning_rate": 0.0005992037310657336,
+      "loss": 3.7202,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6272107362747192,
+      "learning_rate": 0.0005992026096625561,
+      "loss": 3.6994,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6598101258277893,
+      "learning_rate": 0.0005992014874713366,
+      "loss": 3.9831,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6518803834915161,
+      "learning_rate": 0.0005992003644920779,
+      "loss": 3.6686,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5766271948814392,
+      "learning_rate": 0.0005991992407247831,
+      "loss": 3.6382,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6695455312728882,
+      "learning_rate": 0.0005991981161694549,
+      "loss": 3.7335,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5740464329719543,
+      "learning_rate": 0.0005991969908260966,
+      "loss": 3.9502,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6571002006530762,
+      "learning_rate": 0.0005991958646947111,
+      "loss": 3.7527,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6101444959640503,
+      "learning_rate": 0.0005991947377753011,
+      "loss": 3.9642,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6039525270462036,
+      "learning_rate": 0.0005991936100678697,
+      "loss": 3.6638,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7623158097267151,
+      "learning_rate": 0.0005991924815724201,
+      "loss": 3.9647,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5585444569587708,
+      "learning_rate": 0.000599191352288955,
+      "loss": 3.7619,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6341544985771179,
+      "learning_rate": 0.0005991902222174774,
+      "loss": 3.9703,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6054025888442993,
+      "learning_rate": 0.0005991890913579904,
+      "loss": 3.9349,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6357590556144714,
+      "learning_rate": 0.0005991879597104969,
+      "loss": 3.6478,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6057144403457642,
+      "learning_rate": 0.0005991868272749998,
+      "loss": 3.7992,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6412066221237183,
+      "learning_rate": 0.0005991856940515024,
+      "loss": 3.8659,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6738225817680359,
+      "learning_rate": 0.0005991845600400072,
+      "loss": 4.0249,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6093243360519409,
+      "learning_rate": 0.0005991834252405177,
+      "loss": 3.6715,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6469516158103943,
+      "learning_rate": 0.0005991822896530364,
+      "loss": 4.0306,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6203641891479492,
+      "learning_rate": 0.0005991811532775667,
+      "loss": 3.9434,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6038147807121277,
+      "learning_rate": 0.0005991800161141114,
+      "loss": 3.8991,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6124120950698853,
+      "learning_rate": 0.0005991788781626735,
+      "loss": 3.8865,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6234179735183716,
+      "learning_rate": 0.0005991777394232559,
+      "loss": 4.0951,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.622256875038147,
+      "learning_rate": 0.0005991765998958618,
+      "loss": 3.8243,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5846306085586548,
+      "learning_rate": 0.0005991754595804942,
+      "loss": 3.9512,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6416991949081421,
+      "learning_rate": 0.0005991743184771558,
+      "loss": 3.8896,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6234089732170105,
+      "learning_rate": 0.00059917317658585,
+      "loss": 3.8233,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6189374327659607,
+      "learning_rate": 0.0005991720339065796,
+      "loss": 3.6576,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5880756378173828,
+      "learning_rate": 0.0005991708904393476,
+      "loss": 3.8074,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6507730484008789,
+      "learning_rate": 0.0005991697461841571,
+      "loss": 4.0946,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5725138783454895,
+      "learning_rate": 0.0005991686011410111,
+      "loss": 3.7917,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6646773219108582,
+      "learning_rate": 0.0005991674553099126,
+      "loss": 3.728,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5751292109489441,
+      "learning_rate": 0.0005991663086908644,
+      "loss": 3.7794,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6166512966156006,
+      "learning_rate": 0.0005991651612838698,
+      "loss": 3.9293,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6255661845207214,
+      "learning_rate": 0.0005991640130889319,
+      "loss": 3.8986,
+      "step": 1455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5917665958404541,
+      "learning_rate": 0.0005991628641060534,
+      "loss": 3.8005,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5992826223373413,
+      "learning_rate": 0.0005991617143352375,
+      "loss": 3.8569,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6259722113609314,
+      "learning_rate": 0.0005991605637764872,
+      "loss": 3.9048,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6164818406105042,
+      "learning_rate": 0.0005991594124298056,
+      "loss": 3.9696,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5641024708747864,
+      "learning_rate": 0.0005991582602951958,
+      "loss": 3.7273,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6404457688331604,
+      "learning_rate": 0.0005991571073726605,
+      "loss": 3.7909,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6667171120643616,
+      "learning_rate": 0.0005991559536622031,
+      "loss": 3.7106,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5996845364570618,
+      "learning_rate": 0.0005991547991638264,
+      "loss": 3.8542,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6826181411743164,
+      "learning_rate": 0.0005991536438775337,
+      "loss": 4.0089,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6228737235069275,
+      "learning_rate": 0.0005991524878033277,
+      "loss": 3.7689,
+      "step": 1465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6522650718688965,
+      "learning_rate": 0.0005991513309412116,
+      "loss": 3.9532,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6846997141838074,
+      "learning_rate": 0.0005991501732911886,
+      "loss": 3.8377,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6278219223022461,
+      "learning_rate": 0.0005991490148532616,
+      "loss": 3.7568,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6812443733215332,
+      "learning_rate": 0.0005991478556274336,
+      "loss": 3.8747,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7166856527328491,
+      "learning_rate": 0.0005991466956137077,
+      "loss": 3.9553,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5980615615844727,
+      "learning_rate": 0.0005991455348120871,
+      "loss": 4.0037,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6599940061569214,
+      "learning_rate": 0.0005991443732225746,
+      "loss": 3.7815,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6811275482177734,
+      "learning_rate": 0.0005991432108451735,
+      "loss": 3.9142,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.59503173828125,
+      "learning_rate": 0.0005991420476798867,
+      "loss": 3.6994,
+      "step": 1474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.597121000289917,
+      "learning_rate": 0.0005991408837267173,
+      "loss": 3.911,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5894047021865845,
+      "learning_rate": 0.0005991397189856684,
+      "loss": 3.7512,
+      "step": 1476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6419175267219543,
+      "learning_rate": 0.000599138553456743,
+      "loss": 3.892,
+      "step": 1477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6146329641342163,
+      "learning_rate": 0.0005991373871399444,
+      "loss": 3.7919,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6020705103874207,
+      "learning_rate": 0.0005991362200352754,
+      "loss": 3.749,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6040729284286499,
+      "learning_rate": 0.0005991350521427391,
+      "loss": 3.8971,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6802822947502136,
+      "learning_rate": 0.0005991338834623386,
+      "loss": 3.9495,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7739499807357788,
+      "learning_rate": 0.0005991327139940771,
+      "loss": 4.0075,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7212753295898438,
+      "learning_rate": 0.0005991315437379576,
+      "loss": 3.7131,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5949774980545044,
+      "learning_rate": 0.0005991303726939832,
+      "loss": 3.9272,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6401901841163635,
+      "learning_rate": 0.0005991292008621569,
+      "loss": 3.9621,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6232290863990784,
+      "learning_rate": 0.0005991280282424819,
+      "loss": 3.8078,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.644420862197876,
+      "learning_rate": 0.0005991268548349613,
+      "loss": 3.8307,
+      "step": 1487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6507530808448792,
+      "learning_rate": 0.000599125680639598,
+      "loss": 3.85,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6406475305557251,
+      "learning_rate": 0.0005991245056563954,
+      "loss": 3.74,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6073602437973022,
+      "learning_rate": 0.0005991233298853562,
+      "loss": 3.7604,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6374882459640503,
+      "learning_rate": 0.0005991221533264839,
+      "loss": 3.842,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6282268762588501,
+      "learning_rate": 0.0005991209759797813,
+      "loss": 3.845,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6874997615814209,
+      "learning_rate": 0.0005991197978452517,
+      "loss": 4.1104,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.663975715637207,
+      "learning_rate": 0.0005991186189228981,
+      "loss": 3.9287,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6389418840408325,
+      "learning_rate": 0.0005991174392127236,
+      "loss": 3.6839,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5825007557868958,
+      "learning_rate": 0.0005991162587147314,
+      "loss": 4.0187,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5879276394844055,
+      "learning_rate": 0.0005991150774289245,
+      "loss": 3.8135,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6391376852989197,
+      "learning_rate": 0.0005991138953553059,
+      "loss": 4.0896,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5957635045051575,
+      "learning_rate": 0.0005991127124938791,
+      "loss": 3.896,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6745945811271667,
+      "learning_rate": 0.0005991115288446468,
+      "loss": 3.4926,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6167849898338318,
+      "learning_rate": 0.0005991103444076124,
+      "loss": 3.8955,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6245574355125427,
+      "learning_rate": 0.000599109159182779,
+      "loss": 3.7455,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5934446454048157,
+      "learning_rate": 0.0005991079731701495,
+      "loss": 3.659,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5914957523345947,
+      "learning_rate": 0.0005991067863697271,
+      "loss": 3.7625,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6563999652862549,
+      "learning_rate": 0.0005991055987815151,
+      "loss": 4.0097,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6522777676582336,
+      "learning_rate": 0.0005991044104055166,
+      "loss": 3.9068,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6069871187210083,
+      "learning_rate": 0.0005991032212417345,
+      "loss": 3.9719,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.630017876625061,
+      "learning_rate": 0.0005991020312901721,
+      "loss": 3.7736,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.631232738494873,
+      "learning_rate": 0.0005991008405508326,
+      "loss": 3.6984,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6594641208648682,
+      "learning_rate": 0.000599099649023719,
+      "loss": 3.8675,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6070505976676941,
+      "learning_rate": 0.0005990984567088344,
+      "loss": 3.959,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6136875748634338,
+      "learning_rate": 0.0005990972636061821,
+      "loss": 3.8672,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7074840664863586,
+      "learning_rate": 0.0005990960697157652,
+      "loss": 3.8252,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6983718276023865,
+      "learning_rate": 0.0005990948750375868,
+      "loss": 3.9131,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6260075569152832,
+      "learning_rate": 0.0005990936795716501,
+      "loss": 3.6873,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6146523356437683,
+      "learning_rate": 0.0005990924833179582,
+      "loss": 3.8511,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.586942732334137,
+      "learning_rate": 0.0005990912862765143,
+      "loss": 3.8194,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6059939861297607,
+      "learning_rate": 0.0005990900884473213,
+      "loss": 3.8406,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.609937310218811,
+      "learning_rate": 0.0005990888898303827,
+      "loss": 3.4061,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.59391850233078,
+      "learning_rate": 0.0005990876904257015,
+      "loss": 3.8884,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6493670344352722,
+      "learning_rate": 0.000599086490233281,
+      "loss": 3.9793,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6219658851623535,
+      "learning_rate": 0.000599085289253124,
+      "loss": 3.8025,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6138971447944641,
+      "learning_rate": 0.0005990840874852342,
+      "loss": 3.8796,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6053948998451233,
+      "learning_rate": 0.0005990828849296143,
+      "loss": 4.0149,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.651043713092804,
+      "learning_rate": 0.0005990816815862678,
+      "loss": 3.9163,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6234416365623474,
+      "learning_rate": 0.0005990804774551975,
+      "loss": 3.7939,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5939651727676392,
+      "learning_rate": 0.0005990792725364068,
+      "loss": 3.8339,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6041796207427979,
+      "learning_rate": 0.000599078066829899,
+      "loss": 3.7584,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.636833906173706,
+      "learning_rate": 0.0005990768603356771,
+      "loss": 3.8717,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6864349246025085,
+      "learning_rate": 0.0005990756530537442,
+      "loss": 3.786,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5589094161987305,
+      "learning_rate": 0.0005990744449841037,
+      "loss": 3.7716,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6430098414421082,
+      "learning_rate": 0.0005990732361267585,
+      "loss": 3.7062,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6288703083992004,
+      "learning_rate": 0.0005990720264817122,
+      "loss": 3.8552,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5796893835067749,
+      "learning_rate": 0.0005990708160489676,
+      "loss": 3.8646,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5982096791267395,
+      "learning_rate": 0.000599069604828528,
+      "loss": 3.9246,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5953184366226196,
+      "learning_rate": 0.0005990683928203967,
+      "loss": 3.7108,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.621066153049469,
+      "learning_rate": 0.0005990671800245768,
+      "loss": 3.9115,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6250355839729309,
+      "learning_rate": 0.0005990659664410715,
+      "loss": 3.7263,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6018726229667664,
+      "learning_rate": 0.0005990647520698839,
+      "loss": 4.0379,
+      "step": 1539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5801904201507568,
+      "learning_rate": 0.0005990635369110174,
+      "loss": 3.8927,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.61872398853302,
+      "learning_rate": 0.0005990623209644751,
+      "loss": 3.806,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5771560072898865,
+      "learning_rate": 0.0005990611042302602,
+      "loss": 3.8253,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5889159440994263,
+      "learning_rate": 0.0005990598867083759,
+      "loss": 3.827,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5858890414237976,
+      "learning_rate": 0.0005990586683988254,
+      "loss": 3.9579,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6480501890182495,
+      "learning_rate": 0.0005990574493016119,
+      "loss": 3.844,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6195836067199707,
+      "learning_rate": 0.0005990562294167387,
+      "loss": 3.8677,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6144270300865173,
+      "learning_rate": 0.0005990550087442088,
+      "loss": 3.7589,
+      "step": 1547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6432852745056152,
+      "learning_rate": 0.0005990537872840257,
+      "loss": 3.8853,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6419529914855957,
+      "learning_rate": 0.0005990525650361925,
+      "loss": 3.9961,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6389473080635071,
+      "learning_rate": 0.0005990513420007122,
+      "loss": 3.7316,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6231613159179688,
+      "learning_rate": 0.0005990501181775885,
+      "loss": 3.8515,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5940372347831726,
+      "learning_rate": 0.0005990488935668242,
+      "loss": 3.9313,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6202555298805237,
+      "learning_rate": 0.0005990476681684226,
+      "loss": 3.9028,
+      "step": 1553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5787298083305359,
+      "learning_rate": 0.0005990464419823871,
+      "loss": 3.8886,
+      "step": 1554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.639350950717926,
+      "learning_rate": 0.0005990452150087208,
+      "loss": 3.8559,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5683527588844299,
+      "learning_rate": 0.000599043987247427,
+      "loss": 3.9704,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6336866617202759,
+      "learning_rate": 0.0005990427586985088,
+      "loss": 3.8659,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6109759211540222,
+      "learning_rate": 0.0005990415293619695,
+      "loss": 4.0727,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6308028101921082,
+      "learning_rate": 0.0005990402992378125,
+      "loss": 3.7325,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6601607799530029,
+      "learning_rate": 0.0005990390683260408,
+      "loss": 3.8482,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6257616877555847,
+      "learning_rate": 0.0005990378366266577,
+      "loss": 3.9858,
+      "step": 1561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6197351813316345,
+      "learning_rate": 0.0005990366041396666,
+      "loss": 3.7637,
+      "step": 1562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5823824405670166,
+      "learning_rate": 0.0005990353708650706,
+      "loss": 3.8847,
+      "step": 1563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6095857620239258,
+      "learning_rate": 0.000599034136802873,
+      "loss": 3.9395,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6098863482475281,
+      "learning_rate": 0.000599032901953077,
+      "loss": 3.7885,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5995204448699951,
+      "learning_rate": 0.0005990316663156859,
+      "loss": 3.8914,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5945135354995728,
+      "learning_rate": 0.0005990304298907029,
+      "loss": 3.9679,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5825366973876953,
+      "learning_rate": 0.0005990291926781313,
+      "loss": 3.8133,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6503740549087524,
+      "learning_rate": 0.0005990279546779745,
+      "loss": 3.8859,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6176950335502625,
+      "learning_rate": 0.0005990267158902354,
+      "loss": 3.8342,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6380782127380371,
+      "learning_rate": 0.0005990254763149176,
+      "loss": 3.5816,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6276407837867737,
+      "learning_rate": 0.0005990242359520243,
+      "loss": 3.9316,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5958492159843445,
+      "learning_rate": 0.0005990229948015587,
+      "loss": 3.6879,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5630125403404236,
+      "learning_rate": 0.000599021752863524,
+      "loss": 3.7691,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6117847561836243,
+      "learning_rate": 0.0005990205101379236,
+      "loss": 3.73,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6020464301109314,
+      "learning_rate": 0.0005990192666247607,
+      "loss": 3.8803,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5830046534538269,
+      "learning_rate": 0.0005990180223240385,
+      "loss": 3.6786,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6638304591178894,
+      "learning_rate": 0.0005990167772357605,
+      "loss": 4.1394,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5913933515548706,
+      "learning_rate": 0.0005990155313599298,
+      "loss": 3.9041,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5918366312980652,
+      "learning_rate": 0.0005990142846965498,
+      "loss": 3.675,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6066710948944092,
+      "learning_rate": 0.0005990130372456237,
+      "loss": 3.7399,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5532450675964355,
+      "learning_rate": 0.0005990117890071548,
+      "loss": 3.6508,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5913922786712646,
+      "learning_rate": 0.0005990105399811464,
+      "loss": 3.7544,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5768578052520752,
+      "learning_rate": 0.0005990092901676018,
+      "loss": 3.748,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6337912678718567,
+      "learning_rate": 0.0005990080395665242,
+      "loss": 3.8081,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5935748815536499,
+      "learning_rate": 0.000599006788177917,
+      "loss": 3.8911,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6132523417472839,
+      "learning_rate": 0.0005990055360017835,
+      "loss": 3.9166,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6079237461090088,
+      "learning_rate": 0.000599004283038127,
+      "loss": 3.7937,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6035066843032837,
+      "learning_rate": 0.0005990030292869507,
+      "loss": 3.8474,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.620436429977417,
+      "learning_rate": 0.0005990017747482579,
+      "loss": 3.6852,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6006262302398682,
+      "learning_rate": 0.0005990005194220521,
+      "loss": 3.7319,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6008548140525818,
+      "learning_rate": 0.0005989992633083364,
+      "loss": 4.134,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5870209336280823,
+      "learning_rate": 0.0005989980064071142,
+      "loss": 4.0057,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.595600962638855,
+      "learning_rate": 0.0005989967487183888,
+      "loss": 3.8465,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5959302186965942,
+      "learning_rate": 0.0005989954902421634,
+      "loss": 3.701,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5945109724998474,
+      "learning_rate": 0.0005989942309784416,
+      "loss": 3.9228,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.614081084728241,
+      "learning_rate": 0.0005989929709272264,
+      "loss": 3.7123,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.59315425157547,
+      "learning_rate": 0.0005989917100885214,
+      "loss": 3.7997,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5761898756027222,
+      "learning_rate": 0.0005989904484623296,
+      "loss": 3.6235,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6658716201782227,
+      "learning_rate": 0.0005989891860486546,
+      "loss": 3.8065,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5968477129936218,
+      "learning_rate": 0.0005989879228474997,
+      "loss": 3.6979,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.630476713180542,
+      "learning_rate": 0.0005989866588588681,
+      "loss": 3.6387,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6172245144844055,
+      "learning_rate": 0.0005989853940827631,
+      "loss": 3.687,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6125102043151855,
+      "learning_rate": 0.0005989841285191881,
+      "loss": 3.7127,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6071942448616028,
+      "learning_rate": 0.0005989828621681466,
+      "loss": 3.7638,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6062251329421997,
+      "learning_rate": 0.0005989815950296417,
+      "loss": 3.706,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5893798470497131,
+      "learning_rate": 0.0005989803271036769,
+      "loss": 3.8942,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5914480090141296,
+      "learning_rate": 0.0005989790583902553,
+      "loss": 3.7829,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6149440407752991,
+      "learning_rate": 0.0005989777888893805,
+      "loss": 3.6808,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8986303210258484,
+      "learning_rate": 0.0005989765186010557,
+      "loss": 3.543,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5829508900642395,
+      "learning_rate": 0.0005989752475252843,
+      "loss": 3.7894,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6794427633285522,
+      "learning_rate": 0.0005989739756620697,
+      "loss": 3.7064,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7055894732475281,
+      "learning_rate": 0.0005989727030114151,
+      "loss": 3.8842,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5715697407722473,
+      "learning_rate": 0.000598971429573324,
+      "loss": 3.9117,
+      "step": 1614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6290931701660156,
+      "learning_rate": 0.0005989701553477996,
+      "loss": 3.5982,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5757427215576172,
+      "learning_rate": 0.0005989688803348455,
+      "loss": 3.6704,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6020493507385254,
+      "learning_rate": 0.0005989676045344648,
+      "loss": 3.5587,
+      "step": 1617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5768998265266418,
+      "learning_rate": 0.000598966327946661,
+      "loss": 3.5811,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6348850131034851,
+      "learning_rate": 0.0005989650505714374,
+      "loss": 3.9682,
+      "step": 1619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6411147713661194,
+      "learning_rate": 0.0005989637724087973,
+      "loss": 3.7195,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6555300354957581,
+      "learning_rate": 0.0005989624934587443,
+      "loss": 3.647,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6385052800178528,
+      "learning_rate": 0.0005989612137212816,
+      "loss": 3.8637,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6256349086761475,
+      "learning_rate": 0.0005989599331964127,
+      "loss": 3.765,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6104692220687866,
+      "learning_rate": 0.0005989586518841408,
+      "loss": 3.721,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5835363864898682,
+      "learning_rate": 0.0005989573697844692,
+      "loss": 3.9361,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6491838693618774,
+      "learning_rate": 0.0005989560868974015,
+      "loss": 3.8728,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.602606475353241,
+      "learning_rate": 0.0005989548032229411,
+      "loss": 3.7802,
+      "step": 1627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5809963941574097,
+      "learning_rate": 0.0005989535187610912,
+      "loss": 3.5699,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5906050801277161,
+      "learning_rate": 0.0005989522335118553,
+      "loss": 3.7924,
+      "step": 1629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6175136566162109,
+      "learning_rate": 0.0005989509474752367,
+      "loss": 3.8836,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5688444972038269,
+      "learning_rate": 0.0005989496606512389,
+      "loss": 3.9392,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6032138466835022,
+      "learning_rate": 0.0005989483730398652,
+      "loss": 4.0405,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6493321657180786,
+      "learning_rate": 0.000598947084641119,
+      "loss": 3.824,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5763911008834839,
+      "learning_rate": 0.0005989457954550038,
+      "loss": 3.7681,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5691786408424377,
+      "learning_rate": 0.0005989445054815229,
+      "loss": 3.9562,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6753783226013184,
+      "learning_rate": 0.0005989432147206796,
+      "loss": 3.7078,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6098892092704773,
+      "learning_rate": 0.0005989419231724775,
+      "loss": 3.478,
+      "step": 1637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6037442684173584,
+      "learning_rate": 0.0005989406308369199,
+      "loss": 3.73,
+      "step": 1638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6473791003227234,
+      "learning_rate": 0.0005989393377140101,
+      "loss": 3.7052,
+      "step": 1639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6919148564338684,
+      "learning_rate": 0.0005989380438037517,
+      "loss": 3.7036,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6580774188041687,
+      "learning_rate": 0.000598936749106148,
+      "loss": 3.8456,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5730836987495422,
+      "learning_rate": 0.0005989354536212024,
+      "loss": 3.5748,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5901459455490112,
+      "learning_rate": 0.0005989341573489185,
+      "loss": 3.7192,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6304981708526611,
+      "learning_rate": 0.0005989328602892994,
+      "loss": 3.9909,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6005657911300659,
+      "learning_rate": 0.0005989315624423487,
+      "loss": 3.6228,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.61674565076828,
+      "learning_rate": 0.0005989302638080698,
+      "loss": 3.7769,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.598010778427124,
+      "learning_rate": 0.0005989289643864662,
+      "loss": 3.7453,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6126073598861694,
+      "learning_rate": 0.0005989276641775412,
+      "loss": 3.6865,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6504325270652771,
+      "learning_rate": 0.0005989263631812982,
+      "loss": 3.6676,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6758439540863037,
+      "learning_rate": 0.0005989250613977408,
+      "loss": 3.8634,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5557368993759155,
+      "learning_rate": 0.0005989237588268722,
+      "loss": 3.7872,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6107842922210693,
+      "learning_rate": 0.0005989224554686961,
+      "loss": 3.6353,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5767660140991211,
+      "learning_rate": 0.0005989211513232157,
+      "loss": 3.9362,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6302205324172974,
+      "learning_rate": 0.0005989198463904344,
+      "loss": 3.8184,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6364771723747253,
+      "learning_rate": 0.0005989185406703559,
+      "loss": 3.93,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6427229046821594,
+      "learning_rate": 0.0005989172341629834,
+      "loss": 3.7459,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.604668378829956,
+      "learning_rate": 0.0005989159268683205,
+      "loss": 3.7998,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6303444504737854,
+      "learning_rate": 0.0005989146187863705,
+      "loss": 3.8282,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5857745409011841,
+      "learning_rate": 0.0005989133099171371,
+      "loss": 3.8037,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5540629625320435,
+      "learning_rate": 0.0005989120002606233,
+      "loss": 3.9089,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.732204794883728,
+      "learning_rate": 0.000598910689816833,
+      "loss": 3.7783,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6092149615287781,
+      "learning_rate": 0.0005989093785857694,
+      "loss": 3.9062,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5495967864990234,
+      "learning_rate": 0.0005989080665674361,
+      "loss": 3.8753,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6210874319076538,
+      "learning_rate": 0.0005989067537618364,
+      "loss": 3.8944,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6496333479881287,
+      "learning_rate": 0.0005989054401689739,
+      "loss": 3.9576,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6235400438308716,
+      "learning_rate": 0.0005989041257888519,
+      "loss": 3.8495,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5927395224571228,
+      "learning_rate": 0.000598902810621474,
+      "loss": 3.6989,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5796555280685425,
+      "learning_rate": 0.0005989014946668437,
+      "loss": 3.6428,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6373754739761353,
+      "learning_rate": 0.0005989001779249643,
+      "loss": 3.6226,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5925307273864746,
+      "learning_rate": 0.0005988988603958395,
+      "loss": 3.735,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6357890963554382,
+      "learning_rate": 0.0005988975420794724,
+      "loss": 3.7731,
+      "step": 1671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6149181127548218,
+      "learning_rate": 0.0005988962229758668,
+      "loss": 3.844,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6464425921440125,
+      "learning_rate": 0.0005988949030850261,
+      "loss": 3.7181,
+      "step": 1673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6317658424377441,
+      "learning_rate": 0.0005988935824069538,
+      "loss": 3.5902,
+      "step": 1674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5768055319786072,
+      "learning_rate": 0.0005988922609416532,
+      "loss": 3.7552,
+      "step": 1675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6145294904708862,
+      "learning_rate": 0.000598890938689128,
+      "loss": 4.0868,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6101502180099487,
+      "learning_rate": 0.0005988896156493815,
+      "loss": 3.8068,
+      "step": 1677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6006686091423035,
+      "learning_rate": 0.0005988882918224175,
+      "loss": 3.8113,
+      "step": 1678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6369596719741821,
+      "learning_rate": 0.0005988869672082389,
+      "loss": 3.5701,
+      "step": 1679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6296815276145935,
+      "learning_rate": 0.0005988856418068498,
+      "loss": 3.7768,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6762238144874573,
+      "learning_rate": 0.0005988843156182534,
+      "loss": 3.77,
+      "step": 1681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5993692278862,
+      "learning_rate": 0.0005988829886424532,
+      "loss": 3.8124,
+      "step": 1682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6622273921966553,
+      "learning_rate": 0.0005988816608794527,
+      "loss": 3.7658,
+      "step": 1683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5899795293807983,
+      "learning_rate": 0.0005988803323292556,
+      "loss": 3.7341,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6564462780952454,
+      "learning_rate": 0.0005988790029918651,
+      "loss": 3.6619,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6012856364250183,
+      "learning_rate": 0.0005988776728672847,
+      "loss": 3.7838,
+      "step": 1686
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5971396565437317,
+      "learning_rate": 0.0005988763419555182,
+      "loss": 3.7886,
+      "step": 1687
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5735385417938232,
+      "learning_rate": 0.0005988750102565689,
+      "loss": 3.8873,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5730529427528381,
+      "learning_rate": 0.0005988736777704403,
+      "loss": 3.7951,
+      "step": 1689
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5634326338768005,
+      "learning_rate": 0.0005988723444971361,
+      "loss": 3.9817,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5796641707420349,
+      "learning_rate": 0.0005988710104366595,
+      "loss": 3.6427,
+      "step": 1691
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5546978116035461,
+      "learning_rate": 0.0005988696755890142,
+      "loss": 3.839,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6504260897636414,
+      "learning_rate": 0.0005988683399542037,
+      "loss": 3.6866,
+      "step": 1693
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6330404281616211,
+      "learning_rate": 0.0005988670035322316,
+      "loss": 3.6307,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.686377227306366,
+      "learning_rate": 0.0005988656663231012,
+      "loss": 3.5898,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6027424335479736,
+      "learning_rate": 0.0005988643283268162,
+      "loss": 3.6807,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6752616763114929,
+      "learning_rate": 0.0005988629895433801,
+      "loss": 3.5909,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5653634071350098,
+      "learning_rate": 0.0005988616499727965,
+      "loss": 3.9474,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6292624473571777,
+      "learning_rate": 0.0005988603096150686,
+      "loss": 3.7446,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6524072289466858,
+      "learning_rate": 0.0005988589684702004,
+      "loss": 3.8772,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6119723320007324,
+      "learning_rate": 0.0005988576265381951,
+      "loss": 3.5762,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6504294872283936,
+      "learning_rate": 0.0005988562838190562,
+      "loss": 3.6652,
+      "step": 1702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6257925033569336,
+      "learning_rate": 0.0005988549403127875,
+      "loss": 3.6179,
+      "step": 1703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6477129459381104,
+      "learning_rate": 0.0005988535960193924,
+      "loss": 3.7043,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6320105791091919,
+      "learning_rate": 0.0005988522509388744,
+      "loss": 3.7492,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5984609127044678,
+      "learning_rate": 0.0005988509050712371,
+      "loss": 3.8237,
+      "step": 1706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5916309952735901,
+      "learning_rate": 0.0005988495584164841,
+      "loss": 3.5611,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5889941453933716,
+      "learning_rate": 0.0005988482109746188,
+      "loss": 3.6234,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5852148532867432,
+      "learning_rate": 0.0005988468627456448,
+      "loss": 3.5888,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6306596398353577,
+      "learning_rate": 0.0005988455137295657,
+      "loss": 3.9278,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.63728266954422,
+      "learning_rate": 0.0005988441639263851,
+      "loss": 3.6708,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5717198848724365,
+      "learning_rate": 0.0005988428133361064,
+      "loss": 3.8049,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5758329033851624,
+      "learning_rate": 0.0005988414619587333,
+      "loss": 3.7912,
+      "step": 1713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6172877550125122,
+      "learning_rate": 0.0005988401097942693,
+      "loss": 3.6147,
+      "step": 1714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6377825140953064,
+      "learning_rate": 0.0005988387568427179,
+      "loss": 3.7007,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5383907556533813,
+      "learning_rate": 0.0005988374031040828,
+      "loss": 3.6447,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6975224018096924,
+      "learning_rate": 0.0005988360485783674,
+      "loss": 3.6552,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5704695582389832,
+      "learning_rate": 0.0005988346932655755,
+      "loss": 3.9312,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.583870530128479,
+      "learning_rate": 0.0005988333371657104,
+      "loss": 3.8014,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5745240449905396,
+      "learning_rate": 0.0005988319802787759,
+      "loss": 3.9566,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5921694040298462,
+      "learning_rate": 0.0005988306226047754,
+      "loss": 3.7942,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.586078405380249,
+      "learning_rate": 0.0005988292641437126,
+      "loss": 3.8475,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6041373610496521,
+      "learning_rate": 0.000598827904895591,
+      "loss": 3.7767,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6381456255912781,
+      "learning_rate": 0.0005988265448604143,
+      "loss": 3.7119,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6016688942909241,
+      "learning_rate": 0.0005988251840381858,
+      "loss": 3.9402,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6326636672019958,
+      "learning_rate": 0.0005988238224289094,
+      "loss": 3.89,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5598689317703247,
+      "learning_rate": 0.0005988224600325886,
+      "loss": 3.7806,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6054954528808594,
+      "learning_rate": 0.0005988210968492268,
+      "loss": 3.8775,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6688248515129089,
+      "learning_rate": 0.0005988197328788279,
+      "loss": 3.8513,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6302428841590881,
+      "learning_rate": 0.0005988183681213953,
+      "loss": 3.7177,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6412402391433716,
+      "learning_rate": 0.0005988170025769326,
+      "loss": 3.6202,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6302105784416199,
+      "learning_rate": 0.0005988156362454434,
+      "loss": 3.7965,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5867551565170288,
+      "learning_rate": 0.0005988142691269314,
+      "loss": 3.5873,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.670638382434845,
+      "learning_rate": 0.0005988129012214,
+      "loss": 3.755,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6355201601982117,
+      "learning_rate": 0.000598811532528853,
+      "loss": 3.9034,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6599445939064026,
+      "learning_rate": 0.000598810163049294,
+      "loss": 4.0809,
+      "step": 1736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5947389602661133,
+      "learning_rate": 0.0005988087927827264,
+      "loss": 3.8277,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6837261319160461,
+      "learning_rate": 0.000598807421729154,
+      "loss": 3.8107,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6405160427093506,
+      "learning_rate": 0.0005988060498885804,
+      "loss": 3.678,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6684130430221558,
+      "learning_rate": 0.0005988046772610091,
+      "loss": 3.8897,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6249805688858032,
+      "learning_rate": 0.0005988033038464438,
+      "loss": 3.9512,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.591820478439331,
+      "learning_rate": 0.0005988019296448882,
+      "loss": 3.6929,
+      "step": 1742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6458953022956848,
+      "learning_rate": 0.0005988005546563457,
+      "loss": 3.7441,
+      "step": 1743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6687983870506287,
+      "learning_rate": 0.0005987991788808201,
+      "loss": 3.7556,
+      "step": 1744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6560901999473572,
+      "learning_rate": 0.000598797802318315,
+      "loss": 3.7612,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6090633273124695,
+      "learning_rate": 0.000598796424968834,
+      "loss": 3.8793,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.647350549697876,
+      "learning_rate": 0.0005987950468323806,
+      "loss": 3.8172,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6341491937637329,
+      "learning_rate": 0.0005987936679089586,
+      "loss": 3.9446,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5984036326408386,
+      "learning_rate": 0.0005987922881985717,
+      "loss": 3.6023,
+      "step": 1749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6060883402824402,
+      "learning_rate": 0.0005987909077012233,
+      "loss": 3.8801,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6400834918022156,
+      "learning_rate": 0.0005987895264169172,
+      "loss": 3.8829,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6266987919807434,
+      "learning_rate": 0.0005987881443456569,
+      "loss": 3.6213,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6001605987548828,
+      "learning_rate": 0.0005987867614874463,
+      "loss": 3.7277,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6193922758102417,
+      "learning_rate": 0.0005987853778422887,
+      "loss": 3.8871,
+      "step": 1754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6383598446846008,
+      "learning_rate": 0.000598783993410188,
+      "loss": 4.0488,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5988031029701233,
+      "learning_rate": 0.0005987826081911478,
+      "loss": 3.8313,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6383107304573059,
+      "learning_rate": 0.0005987812221851716,
+      "loss": 3.7355,
+      "step": 1757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5527706146240234,
+      "learning_rate": 0.0005987798353922632,
+      "loss": 3.9147,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5697839856147766,
+      "learning_rate": 0.0005987784478124263,
+      "loss": 3.7764,
+      "step": 1759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6100751757621765,
+      "learning_rate": 0.0005987770594456644,
+      "loss": 3.7297,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5448328256607056,
+      "learning_rate": 0.0005987756702919813,
+      "loss": 3.7812,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6359660625457764,
+      "learning_rate": 0.0005987742803513805,
+      "loss": 3.583,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5905768275260925,
+      "learning_rate": 0.0005987728896238657,
+      "loss": 3.7894,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6355024576187134,
+      "learning_rate": 0.0005987714981094407,
+      "loss": 3.7716,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5862598419189453,
+      "learning_rate": 0.0005987701058081091,
+      "loss": 3.7433,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6261117458343506,
+      "learning_rate": 0.0005987687127198745,
+      "loss": 3.8265,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5825438499450684,
+      "learning_rate": 0.0005987673188447406,
+      "loss": 3.598,
+      "step": 1767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6495029926300049,
+      "learning_rate": 0.0005987659241827111,
+      "loss": 3.4557,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5948395729064941,
+      "learning_rate": 0.0005987645287337896,
+      "loss": 3.6985,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6291310787200928,
+      "learning_rate": 0.0005987631324979799,
+      "loss": 3.7719,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6354256868362427,
+      "learning_rate": 0.0005987617354752856,
+      "loss": 3.7825,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5604045391082764,
+      "learning_rate": 0.0005987603376657103,
+      "loss": 3.9863,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6706216931343079,
+      "learning_rate": 0.0005987589390692578,
+      "loss": 3.7109,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6647695302963257,
+      "learning_rate": 0.0005987575396859318,
+      "loss": 3.6884,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6067073941230774,
+      "learning_rate": 0.000598756139515736,
+      "loss": 3.7603,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6076330542564392,
+      "learning_rate": 0.0005987547385586739,
+      "loss": 3.887,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6393039226531982,
+      "learning_rate": 0.0005987533368147493,
+      "loss": 3.9307,
+      "step": 1777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5665969848632812,
+      "learning_rate": 0.0005987519342839661,
+      "loss": 3.8139,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5718681812286377,
+      "learning_rate": 0.0005987505309663275,
+      "loss": 3.8209,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6644183993339539,
+      "learning_rate": 0.0005987491268618377,
+      "loss": 3.6834,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6793142557144165,
+      "learning_rate": 0.0005987477219705002,
+      "loss": 3.6879,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8465895652770996,
+      "learning_rate": 0.0005987463162923187,
+      "loss": 3.7265,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6196627020835876,
+      "learning_rate": 0.0005987449098272968,
+      "loss": 4.0914,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5777091383934021,
+      "learning_rate": 0.0005987435025754384,
+      "loss": 3.7122,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5573776960372925,
+      "learning_rate": 0.000598742094536747,
+      "loss": 4.0411,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6396626830101013,
+      "learning_rate": 0.0005987406857112266,
+      "loss": 3.6454,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5943922996520996,
+      "learning_rate": 0.0005987392760988805,
+      "loss": 3.7657,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6125734448432922,
+      "learning_rate": 0.0005987378656997128,
+      "loss": 3.7891,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6283426880836487,
+      "learning_rate": 0.000598736454513727,
+      "loss": 3.7106,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6084054708480835,
+      "learning_rate": 0.0005987350425409268,
+      "loss": 3.9321,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5988473296165466,
+      "learning_rate": 0.0005987336297813161,
+      "loss": 4.0916,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.650684118270874,
+      "learning_rate": 0.0005987322162348985,
+      "loss": 4.0522,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5723590850830078,
+      "learning_rate": 0.0005987308019016778,
+      "loss": 3.8183,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5800668001174927,
+      "learning_rate": 0.0005987293867816575,
+      "loss": 3.7359,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5976575016975403,
+      "learning_rate": 0.0005987279708748416,
+      "loss": 3.9004,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6423118114471436,
+      "learning_rate": 0.0005987265541812336,
+      "loss": 3.5266,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5797125697135925,
+      "learning_rate": 0.0005987251367008374,
+      "loss": 3.9024,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5562116503715515,
+      "learning_rate": 0.0005987237184336568,
+      "loss": 3.6905,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6045980453491211,
+      "learning_rate": 0.0005987222993796953,
+      "loss": 3.6083,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.696354866027832,
+      "learning_rate": 0.0005987208795389567,
+      "loss": 3.8496,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6290586590766907,
+      "learning_rate": 0.0005987194589114449,
+      "loss": 3.5814,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5684269070625305,
+      "learning_rate": 0.0005987180374971635,
+      "loss": 3.8016,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5839782953262329,
+      "learning_rate": 0.0005987166152961163,
+      "loss": 3.5221,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5923112630844116,
+      "learning_rate": 0.000598715192308307,
+      "loss": 3.5328,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5886315703392029,
+      "learning_rate": 0.0005987137685337393,
+      "loss": 4.0307,
+      "step": 1805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6204468607902527,
+      "learning_rate": 0.0005987123439724172,
+      "loss": 3.8404,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5841030478477478,
+      "learning_rate": 0.0005987109186243441,
+      "loss": 3.3384,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5855621099472046,
+      "learning_rate": 0.0005987094924895241,
+      "loss": 3.7011,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6314960718154907,
+      "learning_rate": 0.0005987080655679607,
+      "loss": 3.7732,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5664706826210022,
+      "learning_rate": 0.0005987066378596577,
+      "loss": 3.8319,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6282777786254883,
+      "learning_rate": 0.0005987052093646189,
+      "loss": 3.6498,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5864231586456299,
+      "learning_rate": 0.0005987037800828482,
+      "loss": 3.8712,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6137101054191589,
+      "learning_rate": 0.0005987023500143491,
+      "loss": 3.6108,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6736273169517517,
+      "learning_rate": 0.0005987009191591254,
+      "loss": 3.6828,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6778322458267212,
+      "learning_rate": 0.0005986994875171811,
+      "loss": 3.6604,
+      "step": 1815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6052943468093872,
+      "learning_rate": 0.0005986980550885198,
+      "loss": 3.9966,
+      "step": 1816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6130955815315247,
+      "learning_rate": 0.0005986966218731454,
+      "loss": 3.6581,
+      "step": 1817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.61712247133255,
+      "learning_rate": 0.0005986951878710616,
+      "loss": 3.704,
+      "step": 1818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6492055654525757,
+      "learning_rate": 0.0005986937530822721,
+      "loss": 3.7532,
+      "step": 1819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6408119797706604,
+      "learning_rate": 0.0005986923175067806,
+      "loss": 3.8313,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.577404797077179,
+      "learning_rate": 0.0005986908811445911,
+      "loss": 4.156,
+      "step": 1821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6154677271842957,
+      "learning_rate": 0.0005986894439957074,
+      "loss": 3.8513,
+      "step": 1822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5824177265167236,
+      "learning_rate": 0.0005986880060601332,
+      "loss": 3.7039,
+      "step": 1823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6346336603164673,
+      "learning_rate": 0.0005986865673378722,
+      "loss": 3.801,
+      "step": 1824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6315222978591919,
+      "learning_rate": 0.0005986851278289283,
+      "loss": 3.5617,
+      "step": 1825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.628288209438324,
+      "learning_rate": 0.0005986836875333052,
+      "loss": 3.6983,
+      "step": 1826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6149936318397522,
+      "learning_rate": 0.0005986822464510067,
+      "loss": 3.8962,
+      "step": 1827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5607312917709351,
+      "learning_rate": 0.0005986808045820367,
+      "loss": 3.8976,
+      "step": 1828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6367286443710327,
+      "learning_rate": 0.000598679361926399,
+      "loss": 3.7893,
+      "step": 1829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6114875078201294,
+      "learning_rate": 0.0005986779184840973,
+      "loss": 3.5419,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6330634355545044,
+      "learning_rate": 0.0005986764742551355,
+      "loss": 3.9017,
+      "step": 1831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5696288347244263,
+      "learning_rate": 0.0005986750292395173,
+      "loss": 3.6879,
+      "step": 1832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6309328079223633,
+      "learning_rate": 0.0005986735834372465,
+      "loss": 3.8622,
+      "step": 1833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5509409308433533,
+      "learning_rate": 0.000598672136848327,
+      "loss": 3.8122,
+      "step": 1834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6046048402786255,
+      "learning_rate": 0.0005986706894727627,
+      "loss": 3.7887,
+      "step": 1835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6264524459838867,
+      "learning_rate": 0.0005986692413105571,
+      "loss": 3.9682,
+      "step": 1836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6554989218711853,
+      "learning_rate": 0.0005986677923617142,
+      "loss": 4.0354,
+      "step": 1837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5879736542701721,
+      "learning_rate": 0.0005986663426262379,
+      "loss": 3.8353,
+      "step": 1838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6094560623168945,
+      "learning_rate": 0.0005986648921041319,
+      "loss": 3.7207,
+      "step": 1839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5921733975410461,
+      "learning_rate": 0.0005986634407954001,
+      "loss": 3.7837,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5805111527442932,
+      "learning_rate": 0.0005986619887000463,
+      "loss": 3.7929,
+      "step": 1841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6623033285140991,
+      "learning_rate": 0.0005986605358180743,
+      "loss": 3.669,
+      "step": 1842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5905426144599915,
+      "learning_rate": 0.0005986590821494878,
+      "loss": 3.8428,
+      "step": 1843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6057484149932861,
+      "learning_rate": 0.0005986576276942908,
+      "loss": 3.5711,
+      "step": 1844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5728998184204102,
+      "learning_rate": 0.0005986561724524871,
+      "loss": 3.8205,
+      "step": 1845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6299675703048706,
+      "learning_rate": 0.0005986547164240805,
+      "loss": 3.7557,
+      "step": 1846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6935710906982422,
+      "learning_rate": 0.000598653259609075,
+      "loss": 3.6427,
+      "step": 1847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6175175309181213,
+      "learning_rate": 0.0005986518020074742,
+      "loss": 3.7941,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5965023636817932,
+      "learning_rate": 0.000598650343619282,
+      "loss": 3.6033,
+      "step": 1849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6782153844833374,
+      "learning_rate": 0.0005986488844445022,
+      "loss": 3.6862,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6123353838920593,
+      "learning_rate": 0.0005986474244831389,
+      "loss": 3.8051,
+      "step": 1851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6483194828033447,
+      "learning_rate": 0.0005986459637351957,
+      "loss": 3.9932,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5647997856140137,
+      "learning_rate": 0.0005986445022006764,
+      "loss": 3.6815,
+      "step": 1853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6573106646537781,
+      "learning_rate": 0.0005986430398795849,
+      "loss": 3.9537,
+      "step": 1854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5951365828514099,
+      "learning_rate": 0.0005986415767719254,
+      "loss": 3.7697,
+      "step": 1855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6256296038627625,
+      "learning_rate": 0.0005986401128777012,
+      "loss": 3.7458,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7118420600891113,
+      "learning_rate": 0.0005986386481969165,
+      "loss": 3.4327,
+      "step": 1857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5771075487136841,
+      "learning_rate": 0.0005986371827295751,
+      "loss": 3.6589,
+      "step": 1858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6408962607383728,
+      "learning_rate": 0.0005986357164756809,
+      "loss": 3.9137,
+      "step": 1859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6218885183334351,
+      "learning_rate": 0.0005986342494352376,
+      "loss": 3.7506,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6180911064147949,
+      "learning_rate": 0.0005986327816082491,
+      "loss": 3.6067,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6280100345611572,
+      "learning_rate": 0.0005986313129947194,
+      "loss": 3.9251,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5840473175048828,
+      "learning_rate": 0.0005986298435946523,
+      "loss": 3.654,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5487592220306396,
+      "learning_rate": 0.0005986283734080517,
+      "loss": 3.7433,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6708935499191284,
+      "learning_rate": 0.0005986269024349213,
+      "loss": 3.6582,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5990443825721741,
+      "learning_rate": 0.0005986254306752653,
+      "loss": 3.7431,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6027263402938843,
+      "learning_rate": 0.0005986239581290873,
+      "loss": 3.6634,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6863844394683838,
+      "learning_rate": 0.0005986224847963912,
+      "loss": 3.727,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6209395527839661,
+      "learning_rate": 0.0005986210106771811,
+      "loss": 3.853,
+      "step": 1869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6714705228805542,
+      "learning_rate": 0.0005986195357714605,
+      "loss": 3.7967,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5347830057144165,
+      "learning_rate": 0.0005986180600792337,
+      "loss": 3.8287,
+      "step": 1871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6196743249893188,
+      "learning_rate": 0.0005986165836005044,
+      "loss": 3.8296,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5981053113937378,
+      "learning_rate": 0.0005986151063352763,
+      "loss": 3.7604,
+      "step": 1873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5930126905441284,
+      "learning_rate": 0.0005986136282835537,
+      "loss": 3.7541,
+      "step": 1874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.604667067527771,
+      "learning_rate": 0.0005986121494453401,
+      "loss": 3.658,
+      "step": 1875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6272248029708862,
+      "learning_rate": 0.0005986106698206398,
+      "loss": 3.6897,
+      "step": 1876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.572104811668396,
+      "learning_rate": 0.0005986091894094563,
+      "loss": 3.752,
+      "step": 1877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6113089323043823,
+      "learning_rate": 0.0005986077082117936,
+      "loss": 3.6702,
+      "step": 1878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6149545311927795,
+      "learning_rate": 0.0005986062262276559,
+      "loss": 3.8337,
+      "step": 1879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5942882299423218,
+      "learning_rate": 0.0005986047434570466,
+      "loss": 3.546,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6164019703865051,
+      "learning_rate": 0.00059860325989997,
+      "loss": 3.713,
+      "step": 1881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6244223713874817,
+      "learning_rate": 0.0005986017755564299,
+      "loss": 3.5671,
+      "step": 1882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6063665747642517,
+      "learning_rate": 0.0005986002904264301,
+      "loss": 3.8345,
+      "step": 1883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5902014374732971,
+      "learning_rate": 0.0005985988045099747,
+      "loss": 3.7012,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5616203546524048,
+      "learning_rate": 0.0005985973178070675,
+      "loss": 3.5898,
+      "step": 1885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.610362708568573,
+      "learning_rate": 0.0005985958303177124,
+      "loss": 3.7553,
+      "step": 1886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5817036032676697,
+      "learning_rate": 0.0005985943420419134,
+      "loss": 3.6755,
+      "step": 1887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6104779839515686,
+      "learning_rate": 0.0005985928529796743,
+      "loss": 3.6263,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5836775898933411,
+      "learning_rate": 0.000598591363130999,
+      "loss": 3.797,
+      "step": 1889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.60204017162323,
+      "learning_rate": 0.0005985898724958917,
+      "loss": 3.6424,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5760531425476074,
+      "learning_rate": 0.000598588381074356,
+      "loss": 3.6849,
+      "step": 1891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6347927451133728,
+      "learning_rate": 0.0005985868888663961,
+      "loss": 3.6574,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6068776845932007,
+      "learning_rate": 0.0005985853958720157,
+      "loss": 4.0513,
+      "step": 1893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6126516461372375,
+      "learning_rate": 0.0005985839020912189,
+      "loss": 3.9252,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5863696932792664,
+      "learning_rate": 0.0005985824075240096,
+      "loss": 3.7798,
+      "step": 1895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6498505473136902,
+      "learning_rate": 0.0005985809121703916,
+      "loss": 3.8625,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5926463007926941,
+      "learning_rate": 0.000598579416030369,
+      "loss": 3.5835,
+      "step": 1897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6369209885597229,
+      "learning_rate": 0.0005985779191039457,
+      "loss": 3.7343,
+      "step": 1898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5705779194831848,
+      "learning_rate": 0.0005985764213911255,
+      "loss": 3.7468,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5890957713127136,
+      "learning_rate": 0.0005985749228919126,
+      "loss": 3.7234,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6325515508651733,
+      "learning_rate": 0.0005985734236063108,
+      "loss": 3.9433,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6154495477676392,
+      "learning_rate": 0.000598571923534324,
+      "loss": 3.8898,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5961557626724243,
+      "learning_rate": 0.0005985704226759563,
+      "loss": 3.7667,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5520904064178467,
+      "learning_rate": 0.0005985689210312114,
+      "loss": 3.959,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5621737241744995,
+      "learning_rate": 0.0005985674186000935,
+      "loss": 3.7459,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6092246174812317,
+      "learning_rate": 0.0005985659153826065,
+      "loss": 3.7063,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6204909086227417,
+      "learning_rate": 0.0005985644113787544,
+      "loss": 3.74,
+      "step": 1907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.618402361869812,
+      "learning_rate": 0.0005985629065885412,
+      "loss": 3.6501,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5384846329689026,
+      "learning_rate": 0.0005985614010119705,
+      "loss": 3.7754,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5484468340873718,
+      "learning_rate": 0.0005985598946490467,
+      "loss": 3.6136,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5607725977897644,
+      "learning_rate": 0.0005985583874997736,
+      "loss": 3.765,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5928760170936584,
+      "learning_rate": 0.0005985568795641551,
+      "loss": 3.6978,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6223922967910767,
+      "learning_rate": 0.0005985553708421951,
+      "loss": 3.8356,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6009296178817749,
+      "learning_rate": 0.0005985538613338979,
+      "loss": 3.8473,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5862912535667419,
+      "learning_rate": 0.0005985523510392673,
+      "loss": 3.8365,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5991998910903931,
+      "learning_rate": 0.0005985508399583072,
+      "loss": 3.6978,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6147690415382385,
+      "learning_rate": 0.0005985493280910217,
+      "loss": 3.631,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6772075295448303,
+      "learning_rate": 0.0005985478154374147,
+      "loss": 3.7764,
+      "step": 1918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6254436373710632,
+      "learning_rate": 0.0005985463019974901,
+      "loss": 3.6895,
+      "step": 1919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5893893837928772,
+      "learning_rate": 0.0005985447877712521,
+      "loss": 3.9327,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6510739922523499,
+      "learning_rate": 0.0005985432727587045,
+      "loss": 3.7668,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6356686949729919,
+      "learning_rate": 0.0005985417569598515,
+      "loss": 3.6375,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7121893763542175,
+      "learning_rate": 0.0005985402403746969,
+      "loss": 3.8785,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6303704977035522,
+      "learning_rate": 0.0005985387230032449,
+      "loss": 3.7847,
+      "step": 1924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6012903451919556,
+      "learning_rate": 0.0005985372048454992,
+      "loss": 3.6605,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.631798267364502,
+      "learning_rate": 0.000598535685901464,
+      "loss": 3.4997,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6432697176933289,
+      "learning_rate": 0.0005985341661711432,
+      "loss": 3.6908,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6546198129653931,
+      "learning_rate": 0.0005985326456545409,
+      "loss": 3.5806,
+      "step": 1928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6076374053955078,
+      "learning_rate": 0.0005985311243516611,
+      "loss": 3.6955,
+      "step": 1929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.673281192779541,
+      "learning_rate": 0.0005985296022625078,
+      "loss": 3.8208,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.614531397819519,
+      "learning_rate": 0.000598528079387085,
+      "loss": 3.7834,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6124190092086792,
+      "learning_rate": 0.0005985265557253965,
+      "loss": 4.1048,
+      "step": 1932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5843042135238647,
+      "learning_rate": 0.0005985250312774466,
+      "loss": 3.6868,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5952600836753845,
+      "learning_rate": 0.0005985235060432393,
+      "loss": 3.7835,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5896327495574951,
+      "learning_rate": 0.0005985219800227785,
+      "loss": 3.6072,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5987282395362854,
+      "learning_rate": 0.0005985204532160683,
+      "loss": 3.8684,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.661604642868042,
+      "learning_rate": 0.0005985189256231125,
+      "loss": 3.6414,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6821339726448059,
+      "learning_rate": 0.0005985173972439154,
+      "loss": 3.9158,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5843617916107178,
+      "learning_rate": 0.000598515868078481,
+      "loss": 3.8082,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5790466070175171,
+      "learning_rate": 0.0005985143381268132,
+      "loss": 3.8143,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.639920175075531,
+      "learning_rate": 0.0005985128073889161,
+      "loss": 3.7693,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.615778386592865,
+      "learning_rate": 0.0005985112758647937,
+      "loss": 3.6194,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6140244007110596,
+      "learning_rate": 0.00059850974355445,
+      "loss": 3.5417,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6662708520889282,
+      "learning_rate": 0.0005985082104578892,
+      "loss": 3.5064,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6774072051048279,
+      "learning_rate": 0.0005985066765751151,
+      "loss": 3.8021,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6490411758422852,
+      "learning_rate": 0.000598505141906132,
+      "loss": 3.5669,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6608024835586548,
+      "learning_rate": 0.0005985036064509437,
+      "loss": 3.6388,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5928901433944702,
+      "learning_rate": 0.0005985020702095543,
+      "loss": 3.6946,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6092929840087891,
+      "learning_rate": 0.000598500533181968,
+      "loss": 3.5945,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6284611821174622,
+      "learning_rate": 0.0005984989953681887,
+      "loss": 3.932,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5998930335044861,
+      "learning_rate": 0.0005984974567682205,
+      "loss": 3.8786,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6137663125991821,
+      "learning_rate": 0.0005984959173820674,
+      "loss": 3.6663,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6030510663986206,
+      "learning_rate": 0.0005984943772097337,
+      "loss": 3.7237,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5598939061164856,
+      "learning_rate": 0.0005984928362512231,
+      "loss": 3.652,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6012648940086365,
+      "learning_rate": 0.0005984912945065397,
+      "loss": 3.7399,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5890420079231262,
+      "learning_rate": 0.0005984897519756879,
+      "loss": 3.4773,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6079639792442322,
+      "learning_rate": 0.0005984882086586714,
+      "loss": 3.8788,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6101313233375549,
+      "learning_rate": 0.0005984866645554945,
+      "loss": 3.924,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5866898894309998,
+      "learning_rate": 0.000598485119666161,
+      "loss": 3.6906,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7030449509620667,
+      "learning_rate": 0.0005984835739906753,
+      "loss": 3.7014,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6568990349769592,
+      "learning_rate": 0.0005984820275290413,
+      "loss": 3.7585,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5689438581466675,
+      "learning_rate": 0.000598480480281263,
+      "loss": 3.6166,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6452824473381042,
+      "learning_rate": 0.0005984789322473446,
+      "loss": 3.7331,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5881951451301575,
+      "learning_rate": 0.0005984773834272902,
+      "loss": 3.7973,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5840572714805603,
+      "learning_rate": 0.0005984758338211037,
+      "loss": 3.763,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5961487889289856,
+      "learning_rate": 0.0005984742834287894,
+      "loss": 3.6024,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5628345608711243,
+      "learning_rate": 0.0005984727322503512,
+      "loss": 3.8128,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5680345296859741,
+      "learning_rate": 0.0005984711802857933,
+      "loss": 3.8158,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5904361605644226,
+      "learning_rate": 0.0005984696275351198,
+      "loss": 3.6442,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6462759971618652,
+      "learning_rate": 0.0005984680739983346,
+      "loss": 3.826,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5882953405380249,
+      "learning_rate": 0.0005984665196754421,
+      "loss": 3.7568,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5906187891960144,
+      "learning_rate": 0.0005984649645664461,
+      "loss": 3.7037,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5672358870506287,
+      "learning_rate": 0.0005984634086713509,
+      "loss": 3.8733,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6111389994621277,
+      "learning_rate": 0.0005984618519901605,
+      "loss": 3.617,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6110372543334961,
+      "learning_rate": 0.000598460294522879,
+      "loss": 3.7275,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5851089954376221,
+      "learning_rate": 0.0005984587362695105,
+      "loss": 3.6567,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6242319345474243,
+      "learning_rate": 0.0005984571772300592,
+      "loss": 3.698,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5997954607009888,
+      "learning_rate": 0.0005984556174045292,
+      "loss": 3.5633,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6077648401260376,
+      "learning_rate": 0.0005984540567929244,
+      "loss": 3.4955,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6089391112327576,
+      "learning_rate": 0.0005984524953952492,
+      "loss": 3.8301,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6587269902229309,
+      "learning_rate": 0.0005984509332115074,
+      "loss": 3.6826,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6356779336929321,
+      "learning_rate": 0.0005984493702417033,
+      "loss": 3.7191,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6114301085472107,
+      "learning_rate": 0.0005984478064858411,
+      "loss": 3.8376,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6225729584693909,
+      "learning_rate": 0.0005984462419439248,
+      "loss": 3.5604,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5687344074249268,
+      "learning_rate": 0.0005984446766159585,
+      "loss": 3.8095,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6089506149291992,
+      "learning_rate": 0.0005984431105019463,
+      "loss": 3.5823,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6365668177604675,
+      "learning_rate": 0.0005984415436018925,
+      "loss": 3.8444,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6325536370277405,
+      "learning_rate": 0.000598439975915801,
+      "loss": 3.8402,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6012512445449829,
+      "learning_rate": 0.0005984384074436761,
+      "loss": 3.7274,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5962508916854858,
+      "learning_rate": 0.0005984368381855219,
+      "loss": 3.6733,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.618022084236145,
+      "learning_rate": 0.0005984352681413424,
+      "loss": 3.7132,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5967841148376465,
+      "learning_rate": 0.0005984336973111419,
+      "loss": 3.9452,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.590883195400238,
+      "learning_rate": 0.0005984321256949245,
+      "loss": 3.5217,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5896095633506775,
+      "learning_rate": 0.0005984305532926943,
+      "loss": 3.7342,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5817958116531372,
+      "learning_rate": 0.0005984289801044554,
+      "loss": 3.6573,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.586776077747345,
+      "learning_rate": 0.000598427406130212,
+      "loss": 3.7158,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.61875981092453,
+      "learning_rate": 0.0005984258313699684,
+      "loss": 3.688,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5944945216178894,
+      "learning_rate": 0.0005984242558237285,
+      "loss": 3.8244,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.62650465965271,
+      "learning_rate": 0.0005984226794914965,
+      "loss": 3.7835,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6252263188362122,
+      "learning_rate": 0.0005984211023732767,
+      "loss": 3.6683,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6127182841300964,
+      "learning_rate": 0.000598419524469073,
+      "loss": 3.7427,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5782539248466492,
+      "learning_rate": 0.0005984179457788897,
+      "loss": 3.529,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6279388666152954,
+      "learning_rate": 0.0005984163663027311,
+      "loss": 3.4904,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6365188360214233,
+      "learning_rate": 0.0005984147860406012,
+      "loss": 3.75,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6217809319496155,
+      "learning_rate": 0.0005984132049925041,
+      "loss": 3.6439,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6590561270713806,
+      "learning_rate": 0.0005984116231584441,
+      "loss": 3.7347,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5828232169151306,
+      "learning_rate": 0.0005984100405384253,
+      "loss": 3.7325,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5944784879684448,
+      "learning_rate": 0.0005984084571324519,
+      "loss": 3.7656,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6109013557434082,
+      "learning_rate": 0.0005984068729405281,
+      "loss": 4.0064,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6408974528312683,
+      "learning_rate": 0.000598405287962658,
+      "loss": 3.7111,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6209164261817932,
+      "learning_rate": 0.0005984037021988458,
+      "loss": 3.8332,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6082508563995361,
+      "learning_rate": 0.0005984021156490956,
+      "loss": 3.6786,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6087662577629089,
+      "learning_rate": 0.0005984005283134117,
+      "loss": 3.8417,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6204620599746704,
+      "learning_rate": 0.0005983989401917982,
+      "loss": 3.5133,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5919209122657776,
+      "learning_rate": 0.0005983973512842595,
+      "loss": 3.7543,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6268076300621033,
+      "learning_rate": 0.0005983957615907995,
+      "loss": 3.7173,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6411733031272888,
+      "learning_rate": 0.0005983941711114224,
+      "loss": 3.8086,
+      "step": 2017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5614747405052185,
+      "learning_rate": 0.0005983925798461325,
+      "loss": 3.7973,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6606065630912781,
+      "learning_rate": 0.000598390987794934,
+      "loss": 3.6554,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5416567325592041,
+      "learning_rate": 0.0005983893949578311,
+      "loss": 3.9936,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7004711627960205,
+      "learning_rate": 0.000598387801334828,
+      "loss": 3.8626,
+      "step": 2021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.572654664516449,
+      "learning_rate": 0.0005983862069259288,
+      "loss": 3.517,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6083114147186279,
+      "learning_rate": 0.0005983846117311377,
+      "loss": 3.6627,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.639316976070404,
+      "learning_rate": 0.0005983830157504591,
+      "loss": 3.6865,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.715901792049408,
+      "learning_rate": 0.0005983814189838969,
+      "loss": 4.052,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.581177830696106,
+      "learning_rate": 0.0005983798214314555,
+      "loss": 3.903,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5649012327194214,
+      "learning_rate": 0.0005983782230931391,
+      "loss": 3.772,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6346585154533386,
+      "learning_rate": 0.000598376623968952,
+      "loss": 3.5737,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.70558762550354,
+      "learning_rate": 0.0005983750240588982,
+      "loss": 3.7745,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6112974286079407,
+      "learning_rate": 0.000598373423362982,
+      "loss": 3.7495,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5762995481491089,
+      "learning_rate": 0.0005983718218812075,
+      "loss": 3.8097,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5994111895561218,
+      "learning_rate": 0.0005983702196135793,
+      "loss": 3.5908,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6250690221786499,
+      "learning_rate": 0.0005983686165601012,
+      "loss": 3.8148,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5601446628570557,
+      "learning_rate": 0.0005983670127207777,
+      "loss": 3.8132,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.619858980178833,
+      "learning_rate": 0.0005983654080956128,
+      "loss": 3.698,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6175944805145264,
+      "learning_rate": 0.0005983638026846108,
+      "loss": 3.7716,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6550002098083496,
+      "learning_rate": 0.0005983621964877761,
+      "loss": 3.9279,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5929962396621704,
+      "learning_rate": 0.0005983605895051127,
+      "loss": 3.7389,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6178693771362305,
+      "learning_rate": 0.0005983589817366249,
+      "loss": 3.6407,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6257408857345581,
+      "learning_rate": 0.0005983573731823172,
+      "loss": 3.8194,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6745170950889587,
+      "learning_rate": 0.0005983557638421933,
+      "loss": 3.7114,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6017802357673645,
+      "learning_rate": 0.000598354153716258,
+      "loss": 3.6747,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5711172223091125,
+      "learning_rate": 0.000598352542804515,
+      "loss": 3.7881,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6025456786155701,
+      "learning_rate": 0.000598350931106969,
+      "loss": 3.7451,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6247069239616394,
+      "learning_rate": 0.0005983493186236241,
+      "loss": 3.5347,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6139021515846252,
+      "learning_rate": 0.0005983477053544843,
+      "loss": 3.7963,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5898500084877014,
+      "learning_rate": 0.0005983460912995542,
+      "loss": 3.6405,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5590971112251282,
+      "learning_rate": 0.000598344476458838,
+      "loss": 3.8969,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.613969624042511,
+      "learning_rate": 0.0005983428608323397,
+      "loss": 3.7783,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6687971353530884,
+      "learning_rate": 0.0005983412444200639,
+      "loss": 3.541,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5752463340759277,
+      "learning_rate": 0.0005983396272220146,
+      "loss": 3.7437,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6484838724136353,
+      "learning_rate": 0.000598338009238196,
+      "loss": 3.5128,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6085900664329529,
+      "learning_rate": 0.0005983363904686127,
+      "loss": 3.8193,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6057000756263733,
+      "learning_rate": 0.0005983347709132686,
+      "loss": 3.567,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.580085039138794,
+      "learning_rate": 0.0005983331505721683,
+      "loss": 3.9294,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5920203924179077,
+      "learning_rate": 0.0005983315294453158,
+      "loss": 3.8553,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5826650857925415,
+      "learning_rate": 0.0005983299075327155,
+      "loss": 3.6751,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6206904053688049,
+      "learning_rate": 0.0005983282848343717,
+      "loss": 3.8432,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5721469521522522,
+      "learning_rate": 0.0005983266613502885,
+      "loss": 3.9363,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.545362114906311,
+      "learning_rate": 0.0005983250370804702,
+      "loss": 3.6203,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5554171800613403,
+      "learning_rate": 0.0005983234120249213,
+      "loss": 3.9277,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.606959879398346,
+      "learning_rate": 0.0005983217861836459,
+      "loss": 3.5465,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5941494703292847,
+      "learning_rate": 0.0005983201595566484,
+      "loss": 3.9184,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6147699952125549,
+      "learning_rate": 0.000598318532143933,
+      "loss": 3.7999,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.570624589920044,
+      "learning_rate": 0.000598316903945504,
+      "loss": 3.6566,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5827869772911072,
+      "learning_rate": 0.0005983152749613656,
+      "loss": 3.7424,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6157439351081848,
+      "learning_rate": 0.0005983136451915222,
+      "loss": 3.569,
+      "step": 2067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6273912787437439,
+      "learning_rate": 0.0005983120146359781,
+      "loss": 3.8541,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.589424192905426,
+      "learning_rate": 0.0005983103832947376,
+      "loss": 3.7674,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5925363898277283,
+      "learning_rate": 0.0005983087511678049,
+      "loss": 3.7414,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6019349098205566,
+      "learning_rate": 0.0005983071182551845,
+      "loss": 3.8578,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5667245984077454,
+      "learning_rate": 0.0005983054845568803,
+      "loss": 3.7414,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5880571007728577,
+      "learning_rate": 0.0005983038500728971,
+      "loss": 3.8056,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6357612013816833,
+      "learning_rate": 0.0005983022148032389,
+      "loss": 3.624,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6150004267692566,
+      "learning_rate": 0.00059830057874791,
+      "loss": 3.6649,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5906575918197632,
+      "learning_rate": 0.0005982989419069148,
+      "loss": 3.6441,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6579732298851013,
+      "learning_rate": 0.0005982973042802578,
+      "loss": 3.7872,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5626339912414551,
+      "learning_rate": 0.0005982956658679429,
+      "loss": 3.6208,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6239688396453857,
+      "learning_rate": 0.0005982940266699747,
+      "loss": 3.5179,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5556694269180298,
+      "learning_rate": 0.0005982923866863574,
+      "loss": 3.732,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6616612672805786,
+      "learning_rate": 0.0005982907459170954,
+      "loss": 3.6723,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.592753529548645,
+      "learning_rate": 0.0005982891043621929,
+      "loss": 3.8336,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6064931750297546,
+      "learning_rate": 0.0005982874620216543,
+      "loss": 3.6037,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6325314044952393,
+      "learning_rate": 0.000598285818895484,
+      "loss": 3.5522,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6174896955490112,
+      "learning_rate": 0.0005982841749836863,
+      "loss": 3.8025,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5929061770439148,
+      "learning_rate": 0.0005982825302862654,
+      "loss": 3.7384,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.597075879573822,
+      "learning_rate": 0.0005982808848032258,
+      "loss": 3.8178,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5691017508506775,
+      "learning_rate": 0.0005982792385345717,
+      "loss": 3.6708,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6304391026496887,
+      "learning_rate": 0.0005982775914803075,
+      "loss": 3.8078,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6006346344947815,
+      "learning_rate": 0.0005982759436404376,
+      "loss": 3.7204,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6283980011940002,
+      "learning_rate": 0.0005982742950149661,
+      "loss": 3.4926,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6308605074882507,
+      "learning_rate": 0.0005982726456038977,
+      "loss": 3.7156,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6353732943534851,
+      "learning_rate": 0.0005982709954072365,
+      "loss": 3.6275,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7076538801193237,
+      "learning_rate": 0.0005982693444249868,
+      "loss": 3.7353,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5855659246444702,
+      "learning_rate": 0.0005982676926571532,
+      "loss": 3.555,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6150306463241577,
+      "learning_rate": 0.0005982660401037398,
+      "loss": 3.6312,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6198219656944275,
+      "learning_rate": 0.000598264386764751,
+      "loss": 3.7313,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6115104556083679,
+      "learning_rate": 0.0005982627326401914,
+      "loss": 3.8263,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6037588119506836,
+      "learning_rate": 0.000598261077730065,
+      "loss": 3.6085,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6247969269752502,
+      "learning_rate": 0.0005982594220343764,
+      "loss": 3.7709,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6451380848884583,
+      "learning_rate": 0.0005982577655531298,
+      "loss": 3.8026,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6023620963096619,
+      "learning_rate": 0.0005982561082863298,
+      "loss": 3.5155,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6898672580718994,
+      "learning_rate": 0.0005982544502339805,
+      "loss": 3.7734,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.620043933391571,
+      "learning_rate": 0.0005982527913960863,
+      "loss": 3.7091,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6267081499099731,
+      "learning_rate": 0.0005982511317726518,
+      "loss": 3.7821,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5801389813423157,
+      "learning_rate": 0.0005982494713636812,
+      "loss": 3.5639,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5749569535255432,
+      "learning_rate": 0.0005982478101691788,
+      "loss": 3.7611,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6347259283065796,
+      "learning_rate": 0.0005982461481891491,
+      "loss": 3.6399,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6066511869430542,
+      "learning_rate": 0.0005982444854235964,
+      "loss": 3.6464,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6020014882087708,
+      "learning_rate": 0.0005982428218725252,
+      "loss": 3.5554,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6045134663581848,
+      "learning_rate": 0.0005982411575359398,
+      "loss": 3.835,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6153691411018372,
+      "learning_rate": 0.0005982394924138446,
+      "loss": 3.2436,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.608252227306366,
+      "learning_rate": 0.0005982378265062439,
+      "loss": 3.7666,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5522031784057617,
+      "learning_rate": 0.0005982361598131422,
+      "loss": 3.5962,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6099326014518738,
+      "learning_rate": 0.0005982344923345439,
+      "loss": 3.5498,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6369093060493469,
+      "learning_rate": 0.0005982328240704533,
+      "loss": 3.6785,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5921229720115662,
+      "learning_rate": 0.0005982311550208749,
+      "loss": 3.7165,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6032543778419495,
+      "learning_rate": 0.000598229485185813,
+      "loss": 3.6678,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.616646409034729,
+      "learning_rate": 0.000598227814565272,
+      "loss": 3.6858,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.610694944858551,
+      "learning_rate": 0.0005982261431592564,
+      "loss": 3.5347,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5749186873435974,
+      "learning_rate": 0.0005982244709677704,
+      "loss": 3.8255,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5923564434051514,
+      "learning_rate": 0.0005982227979908186,
+      "loss": 3.4269,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6011732816696167,
+      "learning_rate": 0.0005982211242284054,
+      "loss": 3.4419,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5927369594573975,
+      "learning_rate": 0.0005982194496805351,
+      "loss": 3.6693,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.575406551361084,
+      "learning_rate": 0.0005982177743472122,
+      "loss": 3.688,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6836567521095276,
+      "learning_rate": 0.000598216098228441,
+      "loss": 3.7712,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5466444492340088,
+      "learning_rate": 0.0005982144213242261,
+      "loss": 3.7642,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6006033420562744,
+      "learning_rate": 0.0005982127436345718,
+      "loss": 3.6812,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6072657704353333,
+      "learning_rate": 0.0005982110651594824,
+      "loss": 3.7972,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6421571373939514,
+      "learning_rate": 0.0005982093858989625,
+      "loss": 3.7497,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5948945879936218,
+      "learning_rate": 0.0005982077058530165,
+      "loss": 3.8954,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6586947441101074,
+      "learning_rate": 0.0005982060250216488,
+      "loss": 3.7188,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6014887094497681,
+      "learning_rate": 0.0005982043434048638,
+      "loss": 3.7739,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6018311977386475,
+      "learning_rate": 0.000598202661002666,
+      "loss": 3.6456,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6336399912834167,
+      "learning_rate": 0.0005982009778150596,
+      "loss": 3.6948,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.586958646774292,
+      "learning_rate": 0.0005981992938420493,
+      "loss": 3.7109,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.641817033290863,
+      "learning_rate": 0.0005981976090836396,
+      "loss": 3.9746,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5889933705329895,
+      "learning_rate": 0.0005981959235398347,
+      "loss": 3.5553,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5783056616783142,
+      "learning_rate": 0.0005981942372106391,
+      "loss": 3.7265,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6405118107795715,
+      "learning_rate": 0.0005981925500960574,
+      "loss": 3.8398,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.585486114025116,
+      "learning_rate": 0.0005981908621960937,
+      "loss": 3.7428,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6322093605995178,
+      "learning_rate": 0.0005981891735107528,
+      "loss": 3.7375,
+      "step": 2142
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6491268873214722,
+      "learning_rate": 0.0005981874840400389,
+      "loss": 3.7564,
+      "step": 2143
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5832539796829224,
+      "learning_rate": 0.0005981857937839566,
+      "loss": 3.7131,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5929149389266968,
+      "learning_rate": 0.0005981841027425102,
+      "loss": 3.6175,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5840476155281067,
+      "learning_rate": 0.0005981824109157044,
+      "loss": 3.5056,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5910953879356384,
+      "learning_rate": 0.0005981807183035436,
+      "loss": 3.7863,
+      "step": 2147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6023166179656982,
+      "learning_rate": 0.000598179024906032,
+      "loss": 3.4796,
+      "step": 2148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6427032351493835,
+      "learning_rate": 0.0005981773307231743,
+      "loss": 3.5623,
+      "step": 2149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5569535493850708,
+      "learning_rate": 0.0005981756357549749,
+      "loss": 3.7547,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6145084500312805,
+      "learning_rate": 0.0005981739400014383,
+      "loss": 3.8746,
+      "step": 2151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6303276419639587,
+      "learning_rate": 0.0005981722434625689,
+      "loss": 3.8413,
+      "step": 2152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6206766963005066,
+      "learning_rate": 0.0005981705461383712,
+      "loss": 3.7907,
+      "step": 2153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6251280903816223,
+      "learning_rate": 0.0005981688480288496,
+      "loss": 3.8965,
+      "step": 2154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6077283620834351,
+      "learning_rate": 0.0005981671491340087,
+      "loss": 3.6284,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6264916658401489,
+      "learning_rate": 0.0005981654494538528,
+      "loss": 3.7238,
+      "step": 2156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5991299152374268,
+      "learning_rate": 0.0005981637489883866,
+      "loss": 3.7317,
+      "step": 2157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6194969415664673,
+      "learning_rate": 0.0005981620477376144,
+      "loss": 3.7232,
+      "step": 2158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5819993019104004,
+      "learning_rate": 0.0005981603457015409,
+      "loss": 3.7031,
+      "step": 2159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6135178804397583,
+      "learning_rate": 0.0005981586428801703,
+      "loss": 3.7341,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5985665917396545,
+      "learning_rate": 0.0005981569392735072,
+      "loss": 3.774,
+      "step": 2161
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5931420922279358,
+      "learning_rate": 0.0005981552348815562,
+      "loss": 3.6447,
+      "step": 2162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7057386636734009,
+      "learning_rate": 0.0005981535297043216,
+      "loss": 3.6028,
+      "step": 2163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5932669043540955,
+      "learning_rate": 0.0005981518237418081,
+      "loss": 3.8519,
+      "step": 2164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6052893996238708,
+      "learning_rate": 0.0005981501169940199,
+      "loss": 3.8123,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3449329137802124,
+      "learning_rate": 0.0005981484094609618,
+      "loss": 3.8314,
+      "step": 2166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5868152976036072,
+      "learning_rate": 0.0005981467011426381,
+      "loss": 3.7069,
+      "step": 2167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6063858866691589,
+      "learning_rate": 0.0005981449920390534,
+      "loss": 3.7754,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6096360087394714,
+      "learning_rate": 0.0005981432821502122,
+      "loss": 3.6768,
+      "step": 2169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6433746814727783,
+      "learning_rate": 0.000598141571476119,
+      "loss": 3.781,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5999099016189575,
+      "learning_rate": 0.0005981398600167782,
+      "loss": 3.9962,
+      "step": 2171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6435213685035706,
+      "learning_rate": 0.0005981381477721944,
+      "loss": 3.5761,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5516665577888489,
+      "learning_rate": 0.0005981364347423722,
+      "loss": 3.6282,
+      "step": 2173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5798248052597046,
+      "learning_rate": 0.000598134720927316,
+      "loss": 3.54,
+      "step": 2174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6035856604576111,
+      "learning_rate": 0.0005981330063270302,
+      "loss": 3.4801,
+      "step": 2175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6492979526519775,
+      "learning_rate": 0.0005981312909415195,
+      "loss": 3.7736,
+      "step": 2176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6277186870574951,
+      "learning_rate": 0.0005981295747707882,
+      "loss": 3.9117,
+      "step": 2177
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5910218954086304,
+      "learning_rate": 0.0005981278578148412,
+      "loss": 3.8235,
+      "step": 2178
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7807388305664062,
+      "learning_rate": 0.0005981261400736827,
+      "loss": 3.6833,
+      "step": 2179
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5943345427513123,
+      "learning_rate": 0.0005981244215473174,
+      "loss": 3.7929,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.578851580619812,
+      "learning_rate": 0.0005981227022357497,
+      "loss": 3.7879,
+      "step": 2181
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6276848316192627,
+      "learning_rate": 0.0005981209821389841,
+      "loss": 3.4234,
+      "step": 2182
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6022424697875977,
+      "learning_rate": 0.0005981192612570253,
+      "loss": 3.7131,
+      "step": 2183
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.669609546661377,
+      "learning_rate": 0.0005981175395898777,
+      "loss": 3.4337,
+      "step": 2184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6255220174789429,
+      "learning_rate": 0.0005981158171375459,
+      "loss": 3.7007,
+      "step": 2185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5757836699485779,
+      "learning_rate": 0.0005981140939000344,
+      "loss": 3.6742,
+      "step": 2186
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.552218496799469,
+      "learning_rate": 0.0005981123698773478,
+      "loss": 3.7315,
+      "step": 2187
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5861135125160217,
+      "learning_rate": 0.0005981106450694904,
+      "loss": 3.842,
+      "step": 2188
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6071463823318481,
+      "learning_rate": 0.0005981089194764672,
+      "loss": 3.7262,
+      "step": 2189
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6482838988304138,
+      "learning_rate": 0.0005981071930982823,
+      "loss": 3.4363,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5629037618637085,
+      "learning_rate": 0.0005981054659349405,
+      "loss": 3.7168,
+      "step": 2191
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6590587496757507,
+      "learning_rate": 0.0005981037379864463,
+      "loss": 3.7393,
+      "step": 2192
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6444178819656372,
+      "learning_rate": 0.0005981020092528041,
+      "loss": 3.6663,
+      "step": 2193
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5685077905654907,
+      "learning_rate": 0.0005981002797340187,
+      "loss": 3.8345,
+      "step": 2194
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5644817352294922,
+      "learning_rate": 0.0005980985494300946,
+      "loss": 3.632,
+      "step": 2195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6295031309127808,
+      "learning_rate": 0.000598096818341036,
+      "loss": 3.5742,
+      "step": 2196
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5638957023620605,
+      "learning_rate": 0.000598095086466848,
+      "loss": 3.7152,
+      "step": 2197
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5969396829605103,
+      "learning_rate": 0.0005980933538075349,
+      "loss": 3.5886,
+      "step": 2198
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5802797675132751,
+      "learning_rate": 0.0005980916203631011,
+      "loss": 3.6731,
+      "step": 2199
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5977463722229004,
+      "learning_rate": 0.0005980898861335515,
+      "loss": 3.6158,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5731961727142334,
+      "learning_rate": 0.0005980881511188904,
+      "loss": 3.8086,
+      "step": 2201
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5767385363578796,
+      "learning_rate": 0.0005980864153191226,
+      "loss": 3.5526,
+      "step": 2202
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6023370623588562,
+      "learning_rate": 0.0005980846787342524,
+      "loss": 3.4739,
+      "step": 2203
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5843697786331177,
+      "learning_rate": 0.0005980829413642847,
+      "loss": 3.6017,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5621604323387146,
+      "learning_rate": 0.0005980812032092238,
+      "loss": 3.8069,
+      "step": 2205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6262988448143005,
+      "learning_rate": 0.0005980794642690744,
+      "loss": 3.8661,
+      "step": 2206
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.594237744808197,
+      "learning_rate": 0.0005980777245438411,
+      "loss": 3.6751,
+      "step": 2207
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5916548371315002,
+      "learning_rate": 0.0005980759840335284,
+      "loss": 3.7278,
+      "step": 2208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5857176184654236,
+      "learning_rate": 0.0005980742427381409,
+      "loss": 3.646,
+      "step": 2209
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6896742582321167,
+      "learning_rate": 0.0005980725006576832,
+      "loss": 3.5892,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.678264856338501,
+      "learning_rate": 0.0005980707577921599,
+      "loss": 3.8561,
+      "step": 2211
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5566852688789368,
+      "learning_rate": 0.0005980690141415756,
+      "loss": 3.6284,
+      "step": 2212
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5891258716583252,
+      "learning_rate": 0.000598067269705935,
+      "loss": 3.4689,
+      "step": 2213
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6085823178291321,
+      "learning_rate": 0.0005980655244852424,
+      "loss": 3.8165,
+      "step": 2214
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6259714365005493,
+      "learning_rate": 0.0005980637784795027,
+      "loss": 3.454,
+      "step": 2215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6584763526916504,
+      "learning_rate": 0.0005980620316887203,
+      "loss": 3.6297,
+      "step": 2216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5733041167259216,
+      "learning_rate": 0.0005980602841128998,
+      "loss": 3.8796,
+      "step": 2217
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6462345123291016,
+      "learning_rate": 0.000598058535752046,
+      "loss": 3.6018,
+      "step": 2218
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6512056589126587,
+      "learning_rate": 0.0005980567866061634,
+      "loss": 3.7185,
+      "step": 2219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6104543209075928,
+      "learning_rate": 0.0005980550366752565,
+      "loss": 3.313,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5999908447265625,
+      "learning_rate": 0.00059805328595933,
+      "loss": 3.7027,
+      "step": 2221
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5871838331222534,
+      "learning_rate": 0.0005980515344583886,
+      "loss": 3.8634,
+      "step": 2222
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5965851545333862,
+      "learning_rate": 0.0005980497821724366,
+      "loss": 3.6761,
+      "step": 2223
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6039659380912781,
+      "learning_rate": 0.000598048029101479,
+      "loss": 3.769,
+      "step": 2224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6120340824127197,
+      "learning_rate": 0.0005980462752455204,
+      "loss": 3.5583,
+      "step": 2225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.605404257774353,
+      "learning_rate": 0.0005980445206045649,
+      "loss": 3.8279,
+      "step": 2226
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5787206888198853,
+      "learning_rate": 0.0005980427651786179,
+      "loss": 3.8282,
+      "step": 2227
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5788346529006958,
+      "learning_rate": 0.0005980410089676833,
+      "loss": 3.9292,
+      "step": 2228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6174167394638062,
+      "learning_rate": 0.0005980392519717661,
+      "loss": 3.3578,
+      "step": 2229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5495995283126831,
+      "learning_rate": 0.000598037494190871,
+      "loss": 3.7939,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6039677262306213,
+      "learning_rate": 0.0005980357356250023,
+      "loss": 3.7216,
+      "step": 2231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.598181962966919,
+      "learning_rate": 0.0005980339762741651,
+      "loss": 3.7851,
+      "step": 2232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5918916463851929,
+      "learning_rate": 0.0005980322161383636,
+      "loss": 3.6826,
+      "step": 2233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5864900350570679,
+      "learning_rate": 0.0005980304552176026,
+      "loss": 3.8629,
+      "step": 2234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5640453696250916,
+      "learning_rate": 0.0005980286935118868,
+      "loss": 3.4874,
+      "step": 2235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6437229514122009,
+      "learning_rate": 0.0005980269310212207,
+      "loss": 3.7286,
+      "step": 2236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5866195559501648,
+      "learning_rate": 0.000598025167745609,
+      "loss": 3.6045,
+      "step": 2237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6194109320640564,
+      "learning_rate": 0.0005980234036850565,
+      "loss": 3.7216,
+      "step": 2238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6149755120277405,
+      "learning_rate": 0.0005980216388395676,
+      "loss": 3.7214,
+      "step": 2239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5646103024482727,
+      "learning_rate": 0.0005980198732091471,
+      "loss": 3.6628,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6370061635971069,
+      "learning_rate": 0.0005980181067937996,
+      "loss": 3.7355,
+      "step": 2241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6375163793563843,
+      "learning_rate": 0.0005980163395935297,
+      "loss": 3.6579,
+      "step": 2242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.600969135761261,
+      "learning_rate": 0.0005980145716083423,
+      "loss": 3.8107,
+      "step": 2243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5838377475738525,
+      "learning_rate": 0.0005980128028382416,
+      "loss": 3.8588,
+      "step": 2244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5946401357650757,
+      "learning_rate": 0.0005980110332832328,
+      "loss": 3.7246,
+      "step": 2245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6003056764602661,
+      "learning_rate": 0.0005980092629433202,
+      "loss": 3.4673,
+      "step": 2246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6012457609176636,
+      "learning_rate": 0.0005980074918185083,
+      "loss": 3.7586,
+      "step": 2247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5687946677207947,
+      "learning_rate": 0.0005980057199088024,
+      "loss": 3.4091,
+      "step": 2248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6248691082000732,
+      "learning_rate": 0.0005980039472142065,
+      "loss": 3.7712,
+      "step": 2249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5778698325157166,
+      "learning_rate": 0.0005980021737347258,
+      "loss": 3.629,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.24909245967865,
+      "learning_rate": 0.0005980003994703647,
+      "loss": 4.0102,
+      "step": 2251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6096078157424927,
+      "learning_rate": 0.0005979986244211277,
+      "loss": 3.7681,
+      "step": 2252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6022841334342957,
+      "learning_rate": 0.0005979968485870198,
+      "loss": 3.6394,
+      "step": 2253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6294955015182495,
+      "learning_rate": 0.0005979950719680455,
+      "loss": 3.8566,
+      "step": 2254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.641569972038269,
+      "learning_rate": 0.0005979932945642096,
+      "loss": 3.5181,
+      "step": 2255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5697000622749329,
+      "learning_rate": 0.0005979915163755168,
+      "loss": 3.6019,
+      "step": 2256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6431049704551697,
+      "learning_rate": 0.0005979897374019715,
+      "loss": 3.6606,
+      "step": 2257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6189910769462585,
+      "learning_rate": 0.0005979879576435786,
+      "loss": 3.6255,
+      "step": 2258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6022100448608398,
+      "learning_rate": 0.0005979861771003429,
+      "loss": 3.7233,
+      "step": 2259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6038528084754944,
+      "learning_rate": 0.0005979843957722688,
+      "loss": 3.8155,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5769616961479187,
+      "learning_rate": 0.0005979826136593612,
+      "loss": 3.4622,
+      "step": 2261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5804604291915894,
+      "learning_rate": 0.0005979808307616248,
+      "loss": 3.7446,
+      "step": 2262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5664138793945312,
+      "learning_rate": 0.0005979790470790642,
+      "loss": 3.9118,
+      "step": 2263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6211099028587341,
+      "learning_rate": 0.0005979772626116841,
+      "loss": 3.7406,
+      "step": 2264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.624521791934967,
+      "learning_rate": 0.0005979754773594894,
+      "loss": 3.7997,
+      "step": 2265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5502910614013672,
+      "learning_rate": 0.0005979736913224844,
+      "loss": 3.7596,
+      "step": 2266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5801634788513184,
+      "learning_rate": 0.0005979719045006743,
+      "loss": 3.5966,
+      "step": 2267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6068748831748962,
+      "learning_rate": 0.0005979701168940633,
+      "loss": 3.5563,
+      "step": 2268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6216042041778564,
+      "learning_rate": 0.0005979683285026565,
+      "loss": 3.4848,
+      "step": 2269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5503404140472412,
+      "learning_rate": 0.0005979665393264585,
+      "loss": 3.7462,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6132727265357971,
+      "learning_rate": 0.000597964749365474,
+      "loss": 3.8308,
+      "step": 2271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6045131087303162,
+      "learning_rate": 0.0005979629586197075,
+      "loss": 3.877,
+      "step": 2272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6383094191551208,
+      "learning_rate": 0.0005979611670891641,
+      "loss": 3.8716,
+      "step": 2273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6388508677482605,
+      "learning_rate": 0.0005979593747738483,
+      "loss": 3.7067,
+      "step": 2274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.630363941192627,
+      "learning_rate": 0.0005979575816737648,
+      "loss": 3.6567,
+      "step": 2275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6004341244697571,
+      "learning_rate": 0.0005979557877889184,
+      "loss": 3.6867,
+      "step": 2276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5606918931007385,
+      "learning_rate": 0.0005979539931193137,
+      "loss": 3.6417,
+      "step": 2277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5711063146591187,
+      "learning_rate": 0.0005979521976649556,
+      "loss": 3.607,
+      "step": 2278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6434407830238342,
+      "learning_rate": 0.0005979504014258488,
+      "loss": 3.7508,
+      "step": 2279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6066713929176331,
+      "learning_rate": 0.0005979486044019979,
+      "loss": 3.6702,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.566117525100708,
+      "learning_rate": 0.0005979468065934078,
+      "loss": 3.7053,
+      "step": 2281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6023035645484924,
+      "learning_rate": 0.000597945008000083,
+      "loss": 3.7924,
+      "step": 2282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.649734377861023,
+      "learning_rate": 0.0005979432086220287,
+      "loss": 3.6891,
+      "step": 2283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6085857152938843,
+      "learning_rate": 0.000597941408459249,
+      "loss": 3.6721,
+      "step": 2284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.594468355178833,
+      "learning_rate": 0.0005979396075117492,
+      "loss": 3.7176,
+      "step": 2285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5784687995910645,
+      "learning_rate": 0.0005979378057795337,
+      "loss": 3.6966,
+      "step": 2286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.586226224899292,
+      "learning_rate": 0.0005979360032626073,
+      "loss": 3.8213,
+      "step": 2287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5555378198623657,
+      "learning_rate": 0.0005979341999609749,
+      "loss": 3.7659,
+      "step": 2288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6186184287071228,
+      "learning_rate": 0.0005979323958746411,
+      "loss": 3.7084,
+      "step": 2289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5890335440635681,
+      "learning_rate": 0.0005979305910036108,
+      "loss": 3.6129,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.586846113204956,
+      "learning_rate": 0.0005979287853478886,
+      "loss": 3.7499,
+      "step": 2291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6445232033729553,
+      "learning_rate": 0.0005979269789074793,
+      "loss": 3.5936,
+      "step": 2292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5615684390068054,
+      "learning_rate": 0.0005979251716823877,
+      "loss": 3.8463,
+      "step": 2293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6355283856391907,
+      "learning_rate": 0.0005979233636726186,
+      "loss": 3.5838,
+      "step": 2294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5905806422233582,
+      "learning_rate": 0.0005979215548781766,
+      "loss": 3.5728,
+      "step": 2295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6034395694732666,
+      "learning_rate": 0.0005979197452990665,
+      "loss": 3.3597,
+      "step": 2296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5971759557723999,
+      "learning_rate": 0.0005979179349352932,
+      "loss": 3.7747,
+      "step": 2297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5553016662597656,
+      "learning_rate": 0.0005979161237868615,
+      "loss": 3.7177,
+      "step": 2298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6574896574020386,
+      "learning_rate": 0.000597914311853776,
+      "loss": 3.8484,
+      "step": 2299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5912606120109558,
+      "learning_rate": 0.0005979124991360414,
+      "loss": 3.713,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5491510629653931,
+      "learning_rate": 0.0005979106856336628,
+      "loss": 3.7381,
+      "step": 2301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5802510976791382,
+      "learning_rate": 0.0005979088713466447,
+      "loss": 3.7492,
+      "step": 2302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5774723291397095,
+      "learning_rate": 0.000597907056274992,
+      "loss": 3.7597,
+      "step": 2303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6084253191947937,
+      "learning_rate": 0.0005979052404187094,
+      "loss": 3.7054,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5598316788673401,
+      "learning_rate": 0.0005979034237778018,
+      "loss": 3.7911,
+      "step": 2305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6073153614997864,
+      "learning_rate": 0.0005979016063522738,
+      "loss": 3.6144,
+      "step": 2306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7126467227935791,
+      "learning_rate": 0.0005978997881421304,
+      "loss": 3.7347,
+      "step": 2307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6775062084197998,
+      "learning_rate": 0.0005978979691473763,
+      "loss": 3.632,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6118178367614746,
+      "learning_rate": 0.0005978961493680162,
+      "loss": 3.5525,
+      "step": 2309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6124061346054077,
+      "learning_rate": 0.000597894328804055,
+      "loss": 3.6992,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6024648547172546,
+      "learning_rate": 0.0005978925074554975,
+      "loss": 3.8113,
+      "step": 2311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6617346405982971,
+      "learning_rate": 0.0005978906853223485,
+      "loss": 3.7497,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5842245221138,
+      "learning_rate": 0.0005978888624046127,
+      "loss": 3.5113,
+      "step": 2313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6647869348526001,
+      "learning_rate": 0.0005978870387022949,
+      "loss": 3.7863,
+      "step": 2314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6231033205986023,
+      "learning_rate": 0.0005978852142154001,
+      "loss": 3.8737,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5999862551689148,
+      "learning_rate": 0.000597883388943933,
+      "loss": 3.774,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6005386114120483,
+      "learning_rate": 0.0005978815628878982,
+      "loss": 3.832,
+      "step": 2317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6149207353591919,
+      "learning_rate": 0.0005978797360473009,
+      "loss": 3.7553,
+      "step": 2318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5701582431793213,
+      "learning_rate": 0.0005978779084221456,
+      "loss": 3.6775,
+      "step": 2319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5574246644973755,
+      "learning_rate": 0.0005978760800124372,
+      "loss": 3.6211,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6085734963417053,
+      "learning_rate": 0.0005978742508181805,
+      "loss": 3.6918,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6652593612670898,
+      "learning_rate": 0.0005978724208393804,
+      "loss": 3.8293,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5725169777870178,
+      "learning_rate": 0.0005978705900760418,
+      "loss": 3.7319,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6140974164009094,
+      "learning_rate": 0.0005978687585281692,
+      "loss": 3.7711,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5903205871582031,
+      "learning_rate": 0.0005978669261957676,
+      "loss": 3.7894,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.553845226764679,
+      "learning_rate": 0.000597865093078842,
+      "loss": 3.6955,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5786354541778564,
+      "learning_rate": 0.0005978632591773969,
+      "loss": 3.8862,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5850010514259338,
+      "learning_rate": 0.0005978614244914375,
+      "loss": 3.5009,
+      "step": 2328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5797577500343323,
+      "learning_rate": 0.0005978595890209683,
+      "loss": 3.7438,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5699589848518372,
+      "learning_rate": 0.0005978577527659943,
+      "loss": 3.7208,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6231520771980286,
+      "learning_rate": 0.0005978559157265203,
+      "loss": 3.7847,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5472413897514343,
+      "learning_rate": 0.0005978540779025511,
+      "loss": 3.8139,
+      "step": 2332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6323192119598389,
+      "learning_rate": 0.0005978522392940917,
+      "loss": 3.6501,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8912398219108582,
+      "learning_rate": 0.0005978503999011467,
+      "loss": 3.9693,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5911180377006531,
+      "learning_rate": 0.000597848559723721,
+      "loss": 3.5229,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5412778258323669,
+      "learning_rate": 0.0005978467187618198,
+      "loss": 3.7783,
+      "step": 2336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5602182745933533,
+      "learning_rate": 0.0005978448770154474,
+      "loss": 3.6555,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6185539364814758,
+      "learning_rate": 0.000597843034484609,
+      "loss": 3.5993,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6103714108467102,
+      "learning_rate": 0.0005978411911693094,
+      "loss": 3.7237,
+      "step": 2339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6211607456207275,
+      "learning_rate": 0.0005978393470695534,
+      "loss": 3.5363,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5573241710662842,
+      "learning_rate": 0.0005978375021853459,
+      "loss": 3.9038,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6002277731895447,
+      "learning_rate": 0.0005978356565166917,
+      "loss": 3.7787,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6228034496307373,
+      "learning_rate": 0.0005978338100635958,
+      "loss": 3.5716,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5964246988296509,
+      "learning_rate": 0.0005978319628260629,
+      "loss": 3.5264,
+      "step": 2344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6132970452308655,
+      "learning_rate": 0.0005978301148040978,
+      "loss": 3.6864,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6156790852546692,
+      "learning_rate": 0.0005978282659977058,
+      "loss": 3.5236,
+      "step": 2346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5750837326049805,
+      "learning_rate": 0.0005978264164068912,
+      "loss": 3.7859,
+      "step": 2347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6026135087013245,
+      "learning_rate": 0.0005978245660316592,
+      "loss": 3.6397,
+      "step": 2348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5814940929412842,
+      "learning_rate": 0.0005978227148720146,
+      "loss": 3.8765,
+      "step": 2349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6358333826065063,
+      "learning_rate": 0.0005978208629279623,
+      "loss": 3.6133,
+      "step": 2350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6395803689956665,
+      "learning_rate": 0.0005978190101995071,
+      "loss": 3.7048,
+      "step": 2351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5972474813461304,
+      "learning_rate": 0.000597817156686654,
+      "loss": 3.7461,
+      "step": 2352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5645350217819214,
+      "learning_rate": 0.0005978153023894079,
+      "loss": 3.6762,
+      "step": 2353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6386092305183411,
+      "learning_rate": 0.0005978134473077736,
+      "loss": 3.5883,
+      "step": 2354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.578127384185791,
+      "learning_rate": 0.0005978115914417559,
+      "loss": 3.7026,
+      "step": 2355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5714936256408691,
+      "learning_rate": 0.0005978097347913598,
+      "loss": 3.6753,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6167704463005066,
+      "learning_rate": 0.0005978078773565903,
+      "loss": 3.7017,
+      "step": 2357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6012189984321594,
+      "learning_rate": 0.000597806019137452,
+      "loss": 3.6736,
+      "step": 2358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6270298361778259,
+      "learning_rate": 0.00059780416013395,
+      "loss": 3.5279,
+      "step": 2359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6216274499893188,
+      "learning_rate": 0.0005978023003460893,
+      "loss": 3.4673,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6030069589614868,
+      "learning_rate": 0.0005978004397738744,
+      "loss": 3.6831,
+      "step": 2361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5585080981254578,
+      "learning_rate": 0.0005977985784173107,
+      "loss": 3.7218,
+      "step": 2362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6055890321731567,
+      "learning_rate": 0.0005977967162764027,
+      "loss": 3.8694,
+      "step": 2363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58843994140625,
+      "learning_rate": 0.0005977948533511555,
+      "loss": 3.5457,
+      "step": 2364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6120928525924683,
+      "learning_rate": 0.0005977929896415741,
+      "loss": 3.7752,
+      "step": 2365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6189221739768982,
+      "learning_rate": 0.000597791125147663,
+      "loss": 3.6428,
+      "step": 2366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6048324108123779,
+      "learning_rate": 0.0005977892598694276,
+      "loss": 3.5921,
+      "step": 2367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5761814117431641,
+      "learning_rate": 0.0005977873938068725,
+      "loss": 3.5946,
+      "step": 2368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6603354811668396,
+      "learning_rate": 0.0005977855269600027,
+      "loss": 3.6483,
+      "step": 2369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5804461240768433,
+      "learning_rate": 0.0005977836593288233,
+      "loss": 3.5695,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5955277681350708,
+      "learning_rate": 0.0005977817909133389,
+      "loss": 3.6131,
+      "step": 2371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6651073098182678,
+      "learning_rate": 0.0005977799217135547,
+      "loss": 3.7153,
+      "step": 2372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5861976146697998,
+      "learning_rate": 0.0005977780517294754,
+      "loss": 3.7365,
+      "step": 2373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5709624290466309,
+      "learning_rate": 0.000597776180961106,
+      "loss": 3.6133,
+      "step": 2374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6309444904327393,
+      "learning_rate": 0.0005977743094084514,
+      "loss": 3.681,
+      "step": 2375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5742753148078918,
+      "learning_rate": 0.0005977724370715167,
+      "loss": 3.7932,
+      "step": 2376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5696350336074829,
+      "learning_rate": 0.0005977705639503067,
+      "loss": 3.4466,
+      "step": 2377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6605788469314575,
+      "learning_rate": 0.0005977686900448262,
+      "loss": 3.7482,
+      "step": 2378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5416749715805054,
+      "learning_rate": 0.0005977668153550804,
+      "loss": 3.7162,
+      "step": 2379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6018760204315186,
+      "learning_rate": 0.0005977649398810741,
+      "loss": 3.5618,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6329287886619568,
+      "learning_rate": 0.0005977630636228123,
+      "loss": 3.4429,
+      "step": 2381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6521981358528137,
+      "learning_rate": 0.0005977611865802999,
+      "loss": 3.7848,
+      "step": 2382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.585591733455658,
+      "learning_rate": 0.0005977593087535417,
+      "loss": 3.6851,
+      "step": 2383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5864081382751465,
+      "learning_rate": 0.0005977574301425429,
+      "loss": 3.6888,
+      "step": 2384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6254026889801025,
+      "learning_rate": 0.0005977555507473083,
+      "loss": 3.7411,
+      "step": 2385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6233348250389099,
+      "learning_rate": 0.0005977536705678429,
+      "loss": 3.9431,
+      "step": 2386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.583896279335022,
+      "learning_rate": 0.0005977517896041516,
+      "loss": 3.7437,
+      "step": 2387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5994085073471069,
+      "learning_rate": 0.0005977499078562394,
+      "loss": 3.7577,
+      "step": 2388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6116136908531189,
+      "learning_rate": 0.0005977480253241112,
+      "loss": 3.527,
+      "step": 2389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6351898908615112,
+      "learning_rate": 0.0005977461420077721,
+      "loss": 3.627,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6024461984634399,
+      "learning_rate": 0.0005977442579072269,
+      "loss": 3.7272,
+      "step": 2391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.608726978302002,
+      "learning_rate": 0.0005977423730224807,
+      "loss": 3.7063,
+      "step": 2392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7524157166481018,
+      "learning_rate": 0.0005977404873535383,
+      "loss": 3.5555,
+      "step": 2393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6578160524368286,
+      "learning_rate": 0.0005977386009004048,
+      "loss": 3.7474,
+      "step": 2394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6224360466003418,
+      "learning_rate": 0.0005977367136630852,
+      "loss": 3.6999,
+      "step": 2395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6000403165817261,
+      "learning_rate": 0.0005977348256415843,
+      "loss": 3.6645,
+      "step": 2396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5816790461540222,
+      "learning_rate": 0.0005977329368359072,
+      "loss": 3.4327,
+      "step": 2397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6410402059555054,
+      "learning_rate": 0.0005977310472460588,
+      "loss": 3.4193,
+      "step": 2398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.607941210269928,
+      "learning_rate": 0.0005977291568720442,
+      "loss": 3.5642,
+      "step": 2399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6291766166687012,
+      "learning_rate": 0.0005977272657138683,
+      "loss": 3.5734,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6615599989891052,
+      "learning_rate": 0.000597725373771536,
+      "loss": 3.7853,
+      "step": 2401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5869166254997253,
+      "learning_rate": 0.0005977234810450524,
+      "loss": 3.6767,
+      "step": 2402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.647530198097229,
+      "learning_rate": 0.0005977215875344224,
+      "loss": 3.3796,
+      "step": 2403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5894125699996948,
+      "learning_rate": 0.000597719693239651,
+      "loss": 3.3358,
+      "step": 2404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5925880074501038,
+      "learning_rate": 0.0005977177981607434,
+      "loss": 3.5175,
+      "step": 2405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6181460022926331,
+      "learning_rate": 0.0005977159022977043,
+      "loss": 3.7263,
+      "step": 2406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5811930298805237,
+      "learning_rate": 0.0005977140056505389,
+      "loss": 3.8623,
+      "step": 2407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6250795125961304,
+      "learning_rate": 0.000597712108219252,
+      "loss": 3.7118,
+      "step": 2408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5673991441726685,
+      "learning_rate": 0.0005977102100038488,
+      "loss": 3.7807,
+      "step": 2409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.60816890001297,
+      "learning_rate": 0.0005977083110043341,
+      "loss": 3.5926,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5878978371620178,
+      "learning_rate": 0.0005977064112207131,
+      "loss": 3.7433,
+      "step": 2411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5771352648735046,
+      "learning_rate": 0.0005977045106529906,
+      "loss": 3.7076,
+      "step": 2412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5752608776092529,
+      "learning_rate": 0.0005977026093011719,
+      "loss": 3.6284,
+      "step": 2413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5990986227989197,
+      "learning_rate": 0.0005977007071652616,
+      "loss": 3.7138,
+      "step": 2414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5749576091766357,
+      "learning_rate": 0.0005976988042452651,
+      "loss": 3.5255,
+      "step": 2415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5812985897064209,
+      "learning_rate": 0.0005976969005411871,
+      "loss": 3.8895,
+      "step": 2416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5838669538497925,
+      "learning_rate": 0.0005976949960530329,
+      "loss": 3.604,
+      "step": 2417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5880911946296692,
+      "learning_rate": 0.0005976930907808073,
+      "loss": 3.5379,
+      "step": 2418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6260443925857544,
+      "learning_rate": 0.0005976911847245153,
+      "loss": 3.5423,
+      "step": 2419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6164513826370239,
+      "learning_rate": 0.0005976892778841622,
+      "loss": 3.9167,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6131654977798462,
+      "learning_rate": 0.0005976873702597527,
+      "loss": 3.7101,
+      "step": 2421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6063238978385925,
+      "learning_rate": 0.000597685461851292,
+      "loss": 3.6194,
+      "step": 2422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5958883762359619,
+      "learning_rate": 0.0005976835526587851,
+      "loss": 3.8709,
+      "step": 2423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.623001217842102,
+      "learning_rate": 0.0005976816426822369,
+      "loss": 3.7071,
+      "step": 2424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5648288726806641,
+      "learning_rate": 0.0005976797319216527,
+      "loss": 3.8756,
+      "step": 2425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5784841179847717,
+      "learning_rate": 0.0005976778203770373,
+      "loss": 3.9232,
+      "step": 2426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6105582118034363,
+      "learning_rate": 0.0005976759080483958,
+      "loss": 3.7534,
+      "step": 2427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5793781280517578,
+      "learning_rate": 0.0005976739949357333,
+      "loss": 3.7578,
+      "step": 2428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5981903672218323,
+      "learning_rate": 0.0005976720810390547,
+      "loss": 3.6772,
+      "step": 2429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.585236132144928,
+      "learning_rate": 0.0005976701663583652,
+      "loss": 3.5096,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6657644510269165,
+      "learning_rate": 0.0005976682508936699,
+      "loss": 3.6456,
+      "step": 2431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5640720129013062,
+      "learning_rate": 0.0005976663346449735,
+      "loss": 3.8045,
+      "step": 2432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6080378293991089,
+      "learning_rate": 0.0005976644176122813,
+      "loss": 3.732,
+      "step": 2433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5977432727813721,
+      "learning_rate": 0.0005976624997955984,
+      "loss": 3.7909,
+      "step": 2434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6109653115272522,
+      "learning_rate": 0.0005976605811949296,
+      "loss": 3.6774,
+      "step": 2435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6155614256858826,
+      "learning_rate": 0.0005976586618102802,
+      "loss": 3.7383,
+      "step": 2436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5923848748207092,
+      "learning_rate": 0.0005976567416416552,
+      "loss": 3.74,
+      "step": 2437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6466518640518188,
+      "learning_rate": 0.0005976548206890597,
+      "loss": 3.8091,
+      "step": 2438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6154281497001648,
+      "learning_rate": 0.0005976528989524985,
+      "loss": 3.5068,
+      "step": 2439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5379983186721802,
+      "learning_rate": 0.000597650976431977,
+      "loss": 3.8493,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5665551424026489,
+      "learning_rate": 0.0005976490531275,
+      "loss": 3.6155,
+      "step": 2441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5697283744812012,
+      "learning_rate": 0.0005976471290390727,
+      "loss": 3.8031,
+      "step": 2442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6457461714744568,
+      "learning_rate": 0.0005976452041667002,
+      "loss": 3.6258,
+      "step": 2443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5855603218078613,
+      "learning_rate": 0.0005976432785103875,
+      "loss": 3.6725,
+      "step": 2444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.642481803894043,
+      "learning_rate": 0.0005976413520701397,
+      "loss": 3.5866,
+      "step": 2445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5984083414077759,
+      "learning_rate": 0.0005976394248459619,
+      "loss": 3.6047,
+      "step": 2446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6179513931274414,
+      "learning_rate": 0.0005976374968378591,
+      "loss": 3.7578,
+      "step": 2447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5873163938522339,
+      "learning_rate": 0.0005976355680458364,
+      "loss": 3.4911,
+      "step": 2448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6037702560424805,
+      "learning_rate": 0.0005976336384698989,
+      "loss": 3.556,
+      "step": 2449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6157534718513489,
+      "learning_rate": 0.0005976317081100517,
+      "loss": 3.5765,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5667474865913391,
+      "learning_rate": 0.0005976297769662997,
+      "loss": 3.6285,
+      "step": 2451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5748934149742126,
+      "learning_rate": 0.0005976278450386483,
+      "loss": 3.6365,
+      "step": 2452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5666347146034241,
+      "learning_rate": 0.0005976259123271025,
+      "loss": 3.5536,
+      "step": 2453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6123989224433899,
+      "learning_rate": 0.0005976239788316671,
+      "loss": 3.6866,
+      "step": 2454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5808203220367432,
+      "learning_rate": 0.0005976220445523476,
+      "loss": 3.6159,
+      "step": 2455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6113956570625305,
+      "learning_rate": 0.0005976201094891489,
+      "loss": 3.6135,
+      "step": 2456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5960624814033508,
+      "learning_rate": 0.0005976181736420762,
+      "loss": 3.7485,
+      "step": 2457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6104657649993896,
+      "learning_rate": 0.0005976162370111343,
+      "loss": 3.8493,
+      "step": 2458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5828893780708313,
+      "learning_rate": 0.0005976142995963285,
+      "loss": 3.6571,
+      "step": 2459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6023302674293518,
+      "learning_rate": 0.000597612361397664,
+      "loss": 3.8401,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5271471738815308,
+      "learning_rate": 0.000597610422415146,
+      "loss": 3.8007,
+      "step": 2461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6546827554702759,
+      "learning_rate": 0.0005976084826487792,
+      "loss": 3.4448,
+      "step": 2462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.610604465007782,
+      "learning_rate": 0.0005976065420985689,
+      "loss": 3.6248,
+      "step": 2463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5978257060050964,
+      "learning_rate": 0.0005976046007645203,
+      "loss": 3.578,
+      "step": 2464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.599612832069397,
+      "learning_rate": 0.0005976026586466386,
+      "loss": 3.7923,
+      "step": 2465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6239912509918213,
+      "learning_rate": 0.0005976007157449286,
+      "loss": 3.4729,
+      "step": 2466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6227158904075623,
+      "learning_rate": 0.0005975987720593957,
+      "loss": 3.6351,
+      "step": 2467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5865234732627869,
+      "learning_rate": 0.0005975968275900448,
+      "loss": 3.6103,
+      "step": 2468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5831554532051086,
+      "learning_rate": 0.0005975948823368812,
+      "loss": 3.758,
+      "step": 2469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6110011339187622,
+      "learning_rate": 0.0005975929362999099,
+      "loss": 3.6188,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6303837895393372,
+      "learning_rate": 0.000597590989479136,
+      "loss": 3.7157,
+      "step": 2471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.602101743221283,
+      "learning_rate": 0.000597589041874565,
+      "loss": 3.6808,
+      "step": 2472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6086721420288086,
+      "learning_rate": 0.0005975870934862015,
+      "loss": 3.7579,
+      "step": 2473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5447176694869995,
+      "learning_rate": 0.0005975851443140509,
+      "loss": 3.8872,
+      "step": 2474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6065565347671509,
+      "learning_rate": 0.0005975831943581184,
+      "loss": 3.7509,
+      "step": 2475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.582368791103363,
+      "learning_rate": 0.0005975812436184089,
+      "loss": 3.6838,
+      "step": 2476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5834957957267761,
+      "learning_rate": 0.0005975792920949278,
+      "loss": 3.7677,
+      "step": 2477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5708243250846863,
+      "learning_rate": 0.00059757733978768,
+      "loss": 3.6435,
+      "step": 2478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5927959084510803,
+      "learning_rate": 0.0005975753866966708,
+      "loss": 3.8947,
+      "step": 2479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5893408060073853,
+      "learning_rate": 0.0005975734328219054,
+      "loss": 3.4956,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5960896611213684,
+      "learning_rate": 0.0005975714781633887,
+      "loss": 3.7718,
+      "step": 2481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6134823560714722,
+      "learning_rate": 0.0005975695227211261,
+      "loss": 3.9268,
+      "step": 2482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5897719264030457,
+      "learning_rate": 0.0005975675664951225,
+      "loss": 3.7412,
+      "step": 2483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5636825561523438,
+      "learning_rate": 0.0005975656094853834,
+      "loss": 3.6555,
+      "step": 2484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6075690984725952,
+      "learning_rate": 0.0005975636516919136,
+      "loss": 3.4279,
+      "step": 2485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6178411245346069,
+      "learning_rate": 0.0005975616931147185,
+      "loss": 3.6293,
+      "step": 2486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6706565022468567,
+      "learning_rate": 0.0005975597337538032,
+      "loss": 3.2851,
+      "step": 2487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5771182775497437,
+      "learning_rate": 0.0005975577736091727,
+      "loss": 3.4126,
+      "step": 2488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5679176449775696,
+      "learning_rate": 0.0005975558126808323,
+      "loss": 3.7656,
+      "step": 2489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5970303416252136,
+      "learning_rate": 0.0005975538509687872,
+      "loss": 3.5579,
+      "step": 2490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.557059645652771,
+      "learning_rate": 0.0005975518884730425,
+      "loss": 3.693,
+      "step": 2491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.713829755783081,
+      "learning_rate": 0.0005975499251936034,
+      "loss": 3.7477,
+      "step": 2492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5846167206764221,
+      "learning_rate": 0.000597547961130475,
+      "loss": 3.8181,
+      "step": 2493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5585032105445862,
+      "learning_rate": 0.0005975459962836627,
+      "loss": 3.7385,
+      "step": 2494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6127833724021912,
+      "learning_rate": 0.0005975440306531715,
+      "loss": 3.6375,
+      "step": 2495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6165521144866943,
+      "learning_rate": 0.0005975420642390064,
+      "loss": 3.6615,
+      "step": 2496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5974895358085632,
+      "learning_rate": 0.0005975400970411729,
+      "loss": 3.6218,
+      "step": 2497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6264663338661194,
+      "learning_rate": 0.0005975381290596759,
+      "loss": 3.5796,
+      "step": 2498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5844413638114929,
+      "learning_rate": 0.0005975361602945209,
+      "loss": 3.5147,
+      "step": 2499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6054651141166687,
+      "learning_rate": 0.0005975341907457129,
+      "loss": 3.6625,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6386021971702576,
+      "learning_rate": 0.0005975322204132571,
+      "loss": 3.6716,
+      "step": 2501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6154173612594604,
+      "learning_rate": 0.0005975302492971586,
+      "loss": 3.6322,
+      "step": 2502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5977914929389954,
+      "learning_rate": 0.0005975282773974228,
+      "loss": 3.5976,
+      "step": 2503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6002846360206604,
+      "learning_rate": 0.0005975263047140547,
+      "loss": 3.5668,
+      "step": 2504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5692142844200134,
+      "learning_rate": 0.0005975243312470596,
+      "loss": 3.6982,
+      "step": 2505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6204922795295715,
+      "learning_rate": 0.0005975223569964427,
+      "loss": 3.4332,
+      "step": 2506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6070844531059265,
+      "learning_rate": 0.0005975203819622091,
+      "loss": 3.5652,
+      "step": 2507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6402035355567932,
+      "learning_rate": 0.0005975184061443641,
+      "loss": 3.9418,
+      "step": 2508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5677720308303833,
+      "learning_rate": 0.0005975164295429129,
+      "loss": 3.5175,
+      "step": 2509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6889002323150635,
+      "learning_rate": 0.0005975144521578607,
+      "loss": 3.6094,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6181936264038086,
+      "learning_rate": 0.0005975124739892126,
+      "loss": 3.6502,
+      "step": 2511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6118708848953247,
+      "learning_rate": 0.000597510495036974,
+      "loss": 3.7122,
+      "step": 2512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5670859217643738,
+      "learning_rate": 0.0005975085153011499,
+      "loss": 3.677,
+      "step": 2513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5545445680618286,
+      "learning_rate": 0.0005975065347817457,
+      "loss": 3.8839,
+      "step": 2514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6169227361679077,
+      "learning_rate": 0.0005975045534787666,
+      "loss": 3.7485,
+      "step": 2515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6185595393180847,
+      "learning_rate": 0.0005975025713922177,
+      "loss": 3.763,
+      "step": 2516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5830788612365723,
+      "learning_rate": 0.0005975005885221044,
+      "loss": 3.66,
+      "step": 2517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5783771276473999,
+      "learning_rate": 0.0005974986048684315,
+      "loss": 3.6327,
+      "step": 2518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6497078537940979,
+      "learning_rate": 0.0005974966204312048,
+      "loss": 3.4842,
+      "step": 2519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5857980251312256,
+      "learning_rate": 0.0005974946352104292,
+      "loss": 3.6608,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5564964413642883,
+      "learning_rate": 0.00059749264920611,
+      "loss": 3.7486,
+      "step": 2521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6064032912254333,
+      "learning_rate": 0.0005974906624182523,
+      "loss": 3.7437,
+      "step": 2522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5508198142051697,
+      "learning_rate": 0.0005974886748468616,
+      "loss": 3.4906,
+      "step": 2523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.588590145111084,
+      "learning_rate": 0.0005974866864919429,
+      "loss": 3.5258,
+      "step": 2524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6488778591156006,
+      "learning_rate": 0.0005974846973535015,
+      "loss": 3.5921,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.634125292301178,
+      "learning_rate": 0.0005974827074315427,
+      "loss": 3.5026,
+      "step": 2526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6007183790206909,
+      "learning_rate": 0.0005974807167260717,
+      "loss": 3.5792,
+      "step": 2527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5504261255264282,
+      "learning_rate": 0.0005974787252370937,
+      "loss": 3.6759,
+      "step": 2528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6158027052879333,
+      "learning_rate": 0.000597476732964614,
+      "loss": 3.564,
+      "step": 2529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6696975827217102,
+      "learning_rate": 0.0005974747399086379,
+      "loss": 3.4829,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6032002568244934,
+      "learning_rate": 0.0005974727460691706,
+      "loss": 3.7942,
+      "step": 2531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5709253549575806,
+      "learning_rate": 0.0005974707514462173,
+      "loss": 3.8751,
+      "step": 2532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6123468279838562,
+      "learning_rate": 0.0005974687560397832,
+      "loss": 3.5406,
+      "step": 2533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7552351355552673,
+      "learning_rate": 0.0005974667598498737,
+      "loss": 3.6732,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6883453130722046,
+      "learning_rate": 0.0005974647628764941,
+      "loss": 3.5411,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.614341676235199,
+      "learning_rate": 0.0005974627651196495,
+      "loss": 3.5317,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5992317795753479,
+      "learning_rate": 0.0005974607665793452,
+      "loss": 3.694,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5962361693382263,
+      "learning_rate": 0.0005974587672555865,
+      "loss": 3.5415,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6169871687889099,
+      "learning_rate": 0.0005974567671483786,
+      "loss": 3.5371,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.575014591217041,
+      "learning_rate": 0.0005974547662577268,
+      "loss": 3.692,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.66556316614151,
+      "learning_rate": 0.0005974527645836365,
+      "loss": 3.4761,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5992640852928162,
+      "learning_rate": 0.0005974507621261129,
+      "loss": 3.7176,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6112333536148071,
+      "learning_rate": 0.0005974487588851611,
+      "loss": 3.4594,
+      "step": 2543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6187512874603271,
+      "learning_rate": 0.0005974467548607866,
+      "loss": 3.6286,
+      "step": 2544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5970472693443298,
+      "learning_rate": 0.0005974447500529946,
+      "loss": 3.5783,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.622491180896759,
+      "learning_rate": 0.0005974427444617903,
+      "loss": 3.935,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5881618857383728,
+      "learning_rate": 0.000597440738087179,
+      "loss": 3.7599,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5989762544631958,
+      "learning_rate": 0.0005974387309291662,
+      "loss": 3.5662,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5770282745361328,
+      "learning_rate": 0.000597436722987757,
+      "loss": 3.5526,
+      "step": 2549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5764257907867432,
+      "learning_rate": 0.0005974347142629567,
+      "loss": 3.6311,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5779116749763489,
+      "learning_rate": 0.0005974327047547706,
+      "loss": 3.5682,
+      "step": 2551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5993980169296265,
+      "learning_rate": 0.000597430694463204,
+      "loss": 3.9816,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6707353591918945,
+      "learning_rate": 0.0005974286833882621,
+      "loss": 3.7115,
+      "step": 2553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5597850680351257,
+      "learning_rate": 0.0005974266715299504,
+      "loss": 3.5479,
+      "step": 2554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5796374678611755,
+      "learning_rate": 0.0005974246588882741,
+      "loss": 3.6176,
+      "step": 2555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6326395869255066,
+      "learning_rate": 0.0005974226454632384,
+      "loss": 3.728,
+      "step": 2556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5709221363067627,
+      "learning_rate": 0.0005974206312548487,
+      "loss": 3.7504,
+      "step": 2557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6459200978279114,
+      "learning_rate": 0.0005974186162631104,
+      "loss": 3.8128,
+      "step": 2558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6132066249847412,
+      "learning_rate": 0.0005974166004880286,
+      "loss": 3.6669,
+      "step": 2559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.593331515789032,
+      "learning_rate": 0.0005974145839296088,
+      "loss": 3.7232,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5764754414558411,
+      "learning_rate": 0.0005974125665878561,
+      "loss": 3.4763,
+      "step": 2561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5679061412811279,
+      "learning_rate": 0.000597410548462776,
+      "loss": 3.3199,
+      "step": 2562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6383253931999207,
+      "learning_rate": 0.0005974085295543739,
+      "loss": 3.6122,
+      "step": 2563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6798160672187805,
+      "learning_rate": 0.0005974065098626548,
+      "loss": 3.8301,
+      "step": 2564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5811006426811218,
+      "learning_rate": 0.0005974044893876241,
+      "loss": 3.7606,
+      "step": 2565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6266860961914062,
+      "learning_rate": 0.0005974024681292874,
+      "loss": 3.6762,
+      "step": 2566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5711492896080017,
+      "learning_rate": 0.0005974004460876498,
+      "loss": 3.6717,
+      "step": 2567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5752375721931458,
+      "learning_rate": 0.0005973984232627166,
+      "loss": 3.5786,
+      "step": 2568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6051990985870361,
+      "learning_rate": 0.0005973963996544932,
+      "loss": 3.7287,
+      "step": 2569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6400178074836731,
+      "learning_rate": 0.000597394375262985,
+      "loss": 3.4631,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6259292364120483,
+      "learning_rate": 0.0005973923500881971,
+      "loss": 3.5266,
+      "step": 2571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5960250496864319,
+      "learning_rate": 0.0005973903241301352,
+      "loss": 3.5943,
+      "step": 2572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6018271446228027,
+      "learning_rate": 0.0005973882973888042,
+      "loss": 3.6866,
+      "step": 2573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5785349011421204,
+      "learning_rate": 0.0005973862698642097,
+      "loss": 3.7706,
+      "step": 2574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7196159362792969,
+      "learning_rate": 0.0005973842415563572,
+      "loss": 3.6392,
+      "step": 2575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6078731417655945,
+      "learning_rate": 0.0005973822124652516,
+      "loss": 3.8852,
+      "step": 2576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6295037865638733,
+      "learning_rate": 0.0005973801825908987,
+      "loss": 3.5333,
+      "step": 2577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5880255103111267,
+      "learning_rate": 0.0005973781519333034,
+      "loss": 3.6906,
+      "step": 2578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5796219110488892,
+      "learning_rate": 0.0005973761204924714,
+      "loss": 3.6351,
+      "step": 2579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5962153077125549,
+      "learning_rate": 0.0005973740882684079,
+      "loss": 3.7197,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5779426097869873,
+      "learning_rate": 0.0005973720552611183,
+      "loss": 3.7243,
+      "step": 2581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6200467348098755,
+      "learning_rate": 0.000597370021470608,
+      "loss": 3.7223,
+      "step": 2582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5809329152107239,
+      "learning_rate": 0.0005973679868968822,
+      "loss": 3.4354,
+      "step": 2583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.617591142654419,
+      "learning_rate": 0.0005973659515399463,
+      "loss": 3.667,
+      "step": 2584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6338459253311157,
+      "learning_rate": 0.0005973639153998058,
+      "loss": 3.6246,
+      "step": 2585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5794420838356018,
+      "learning_rate": 0.000597361878476466,
+      "loss": 3.6043,
+      "step": 2586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6019072532653809,
+      "learning_rate": 0.0005973598407699322,
+      "loss": 3.6836,
+      "step": 2587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5721585154533386,
+      "learning_rate": 0.0005973578022802098,
+      "loss": 3.7401,
+      "step": 2588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6371548771858215,
+      "learning_rate": 0.0005973557630073042,
+      "loss": 3.5781,
+      "step": 2589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5836586356163025,
+      "learning_rate": 0.0005973537229512208,
+      "loss": 3.7384,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5855005383491516,
+      "learning_rate": 0.0005973516821119649,
+      "loss": 3.6999,
+      "step": 2591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5802505016326904,
+      "learning_rate": 0.0005973496404895419,
+      "loss": 3.7894,
+      "step": 2592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6217300295829773,
+      "learning_rate": 0.0005973475980839571,
+      "loss": 3.7012,
+      "step": 2593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.657211422920227,
+      "learning_rate": 0.0005973455548952159,
+      "loss": 3.6203,
+      "step": 2594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5941809415817261,
+      "learning_rate": 0.0005973435109233238,
+      "loss": 3.7269,
+      "step": 2595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5752348303794861,
+      "learning_rate": 0.0005973414661682862,
+      "loss": 3.7513,
+      "step": 2596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59751957654953,
+      "learning_rate": 0.0005973394206301083,
+      "loss": 3.5195,
+      "step": 2597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5890262722969055,
+      "learning_rate": 0.0005973373743087956,
+      "loss": 3.4186,
+      "step": 2598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5875232815742493,
+      "learning_rate": 0.0005973353272043535,
+      "loss": 3.5819,
+      "step": 2599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5776680111885071,
+      "learning_rate": 0.0005973332793167874,
+      "loss": 3.6035,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.652600884437561,
+      "learning_rate": 0.0005973312306461027,
+      "loss": 3.5718,
+      "step": 2601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6072568893432617,
+      "learning_rate": 0.0005973291811923046,
+      "loss": 3.6524,
+      "step": 2602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5894418954849243,
+      "learning_rate": 0.0005973271309553988,
+      "loss": 3.7459,
+      "step": 2603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6451753377914429,
+      "learning_rate": 0.0005973250799353906,
+      "loss": 3.6723,
+      "step": 2604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5642520785331726,
+      "learning_rate": 0.0005973230281322852,
+      "loss": 3.6376,
+      "step": 2605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6151884198188782,
+      "learning_rate": 0.0005973209755460883,
+      "loss": 3.6376,
+      "step": 2606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5698564052581787,
+      "learning_rate": 0.0005973189221768051,
+      "loss": 3.5028,
+      "step": 2607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5662623047828674,
+      "learning_rate": 0.000597316868024441,
+      "loss": 3.7416,
+      "step": 2608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6105310916900635,
+      "learning_rate": 0.0005973148130890017,
+      "loss": 3.506,
+      "step": 2609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5945753455162048,
+      "learning_rate": 0.0005973127573704923,
+      "loss": 3.7686,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5580887794494629,
+      "learning_rate": 0.0005973107008689182,
+      "loss": 3.4169,
+      "step": 2611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5982559323310852,
+      "learning_rate": 0.0005973086435842851,
+      "loss": 3.8084,
+      "step": 2612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.545141875743866,
+      "learning_rate": 0.0005973065855165981,
+      "loss": 3.7542,
+      "step": 2613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5577916502952576,
+      "learning_rate": 0.0005973045266658629,
+      "loss": 3.7296,
+      "step": 2614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6226997971534729,
+      "learning_rate": 0.0005973024670320848,
+      "loss": 3.61,
+      "step": 2615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5903077721595764,
+      "learning_rate": 0.0005973004066152692,
+      "loss": 3.5793,
+      "step": 2616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6483430862426758,
+      "learning_rate": 0.0005972983454154215,
+      "loss": 3.7537,
+      "step": 2617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5781821608543396,
+      "learning_rate": 0.0005972962834325472,
+      "loss": 3.7462,
+      "step": 2618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6237555742263794,
+      "learning_rate": 0.0005972942206666517,
+      "loss": 3.709,
+      "step": 2619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.538802444934845,
+      "learning_rate": 0.0005972921571177405,
+      "loss": 3.8094,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6104321479797363,
+      "learning_rate": 0.0005972900927858188,
+      "loss": 3.5672,
+      "step": 2621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5900411605834961,
+      "learning_rate": 0.0005972880276708925,
+      "loss": 3.6144,
+      "step": 2622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6186203360557556,
+      "learning_rate": 0.0005972859617729665,
+      "loss": 3.6618,
+      "step": 2623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5977591872215271,
+      "learning_rate": 0.0005972838950920465,
+      "loss": 3.5381,
+      "step": 2624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5627738237380981,
+      "learning_rate": 0.000597281827628138,
+      "loss": 3.5466,
+      "step": 2625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5746119022369385,
+      "learning_rate": 0.0005972797593812464,
+      "loss": 3.4968,
+      "step": 2626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6021314859390259,
+      "learning_rate": 0.0005972776903513771,
+      "loss": 3.6362,
+      "step": 2627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5764266848564148,
+      "learning_rate": 0.0005972756205385356,
+      "loss": 3.5602,
+      "step": 2628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.560896098613739,
+      "learning_rate": 0.0005972735499427273,
+      "loss": 3.9325,
+      "step": 2629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5999066829681396,
+      "learning_rate": 0.0005972714785639576,
+      "loss": 3.5947,
+      "step": 2630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6318387389183044,
+      "learning_rate": 0.0005972694064022322,
+      "loss": 3.4485,
+      "step": 2631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6000919938087463,
+      "learning_rate": 0.0005972673334575562,
+      "loss": 3.6263,
+      "step": 2632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6186161041259766,
+      "learning_rate": 0.0005972652597299354,
+      "loss": 3.4619,
+      "step": 2633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6438592076301575,
+      "learning_rate": 0.0005972631852193751,
+      "loss": 3.8048,
+      "step": 2634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6455599069595337,
+      "learning_rate": 0.0005972611099258808,
+      "loss": 3.8894,
+      "step": 2635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6064257025718689,
+      "learning_rate": 0.0005972590338494579,
+      "loss": 3.6067,
+      "step": 2636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5977962613105774,
+      "learning_rate": 0.0005972569569901119,
+      "loss": 3.8438,
+      "step": 2637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5873711109161377,
+      "learning_rate": 0.0005972548793478483,
+      "loss": 3.4951,
+      "step": 2638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5406238436698914,
+      "learning_rate": 0.0005972528009226725,
+      "loss": 3.5039,
+      "step": 2639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6180803179740906,
+      "learning_rate": 0.0005972507217145901,
+      "loss": 3.816,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5730354189872742,
+      "learning_rate": 0.0005972486417236065,
+      "loss": 3.5936,
+      "step": 2641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5667976140975952,
+      "learning_rate": 0.0005972465609497273,
+      "loss": 3.7131,
+      "step": 2642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5629318356513977,
+      "learning_rate": 0.0005972444793929577,
+      "loss": 3.7713,
+      "step": 2643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6278749704360962,
+      "learning_rate": 0.0005972423970533033,
+      "loss": 3.6363,
+      "step": 2644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.637045681476593,
+      "learning_rate": 0.0005972403139307698,
+      "loss": 3.6883,
+      "step": 2645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5682446956634521,
+      "learning_rate": 0.0005972382300253625,
+      "loss": 3.7383,
+      "step": 2646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5715309977531433,
+      "learning_rate": 0.0005972361453370868,
+      "loss": 3.4358,
+      "step": 2647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5956022143363953,
+      "learning_rate": 0.0005972340598659483,
+      "loss": 3.5463,
+      "step": 2648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5942005515098572,
+      "learning_rate": 0.0005972319736119526,
+      "loss": 3.667,
+      "step": 2649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6336799263954163,
+      "learning_rate": 0.000597229886575105,
+      "loss": 3.5427,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5908926129341125,
+      "learning_rate": 0.0005972277987554111,
+      "loss": 3.6074,
+      "step": 2651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6374375224113464,
+      "learning_rate": 0.0005972257101528763,
+      "loss": 3.4208,
+      "step": 2652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5974894762039185,
+      "learning_rate": 0.0005972236207675063,
+      "loss": 3.6029,
+      "step": 2653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5751729011535645,
+      "learning_rate": 0.0005972215305993065,
+      "loss": 3.6742,
+      "step": 2654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5705915689468384,
+      "learning_rate": 0.0005972194396482823,
+      "loss": 3.8229,
+      "step": 2655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6375955939292908,
+      "learning_rate": 0.0005972173479144394,
+      "loss": 3.5791,
+      "step": 2656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6054398417472839,
+      "learning_rate": 0.0005972152553977831,
+      "loss": 3.4664,
+      "step": 2657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5961582064628601,
+      "learning_rate": 0.0005972131620983191,
+      "loss": 3.7149,
+      "step": 2658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5579370856285095,
+      "learning_rate": 0.0005972110680160528,
+      "loss": 3.7323,
+      "step": 2659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6265712380409241,
+      "learning_rate": 0.0005972089731509897,
+      "loss": 3.494,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5627157688140869,
+      "learning_rate": 0.0005972068775031353,
+      "loss": 3.8952,
+      "step": 2661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6081935167312622,
+      "learning_rate": 0.0005972047810724954,
+      "loss": 3.6359,
+      "step": 2662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6013399958610535,
+      "learning_rate": 0.0005972026838590753,
+      "loss": 3.7558,
+      "step": 2663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7497528791427612,
+      "learning_rate": 0.0005972005858628803,
+      "loss": 3.5532,
+      "step": 2664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5855658650398254,
+      "learning_rate": 0.0005971984870839162,
+      "loss": 3.8398,
+      "step": 2665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5891215205192566,
+      "learning_rate": 0.0005971963875221886,
+      "loss": 3.7379,
+      "step": 2666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5891420841217041,
+      "learning_rate": 0.000597194287177703,
+      "loss": 3.7002,
+      "step": 2667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5906747579574585,
+      "learning_rate": 0.0005971921860504646,
+      "loss": 3.6601,
+      "step": 2668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5525054931640625,
+      "learning_rate": 0.0005971900841404794,
+      "loss": 4.0302,
+      "step": 2669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6058871746063232,
+      "learning_rate": 0.0005971879814477525,
+      "loss": 3.795,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5821624398231506,
+      "learning_rate": 0.0005971858779722898,
+      "loss": 3.6976,
+      "step": 2671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5979516506195068,
+      "learning_rate": 0.0005971837737140966,
+      "loss": 3.5227,
+      "step": 2672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.558814287185669,
+      "learning_rate": 0.0005971816686731786,
+      "loss": 3.5519,
+      "step": 2673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5667821168899536,
+      "learning_rate": 0.0005971795628495413,
+      "loss": 3.7112,
+      "step": 2674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5700370669364929,
+      "learning_rate": 0.0005971774562431901,
+      "loss": 3.528,
+      "step": 2675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5918428301811218,
+      "learning_rate": 0.0005971753488541307,
+      "loss": 3.6761,
+      "step": 2676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6164225935935974,
+      "learning_rate": 0.0005971732406823688,
+      "loss": 3.6645,
+      "step": 2677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5859614610671997,
+      "learning_rate": 0.0005971711317279096,
+      "loss": 3.698,
+      "step": 2678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5908759236335754,
+      "learning_rate": 0.0005971690219907589,
+      "loss": 3.7241,
+      "step": 2679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6182529330253601,
+      "learning_rate": 0.0005971669114709222,
+      "loss": 3.5972,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.576896071434021,
+      "learning_rate": 0.000597164800168405,
+      "loss": 3.679,
+      "step": 2681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5724910497665405,
+      "learning_rate": 0.000597162688083213,
+      "loss": 3.7296,
+      "step": 2682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6576551198959351,
+      "learning_rate": 0.0005971605752153516,
+      "loss": 3.6643,
+      "step": 2683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6664922833442688,
+      "learning_rate": 0.0005971584615648264,
+      "loss": 3.5636,
+      "step": 2684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.639898955821991,
+      "learning_rate": 0.000597156347131643,
+      "loss": 3.6449,
+      "step": 2685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.680196225643158,
+      "learning_rate": 0.000597154231915807,
+      "loss": 3.4298,
+      "step": 2686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6258304119110107,
+      "learning_rate": 0.0005971521159173239,
+      "loss": 3.8802,
+      "step": 2687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5518138408660889,
+      "learning_rate": 0.0005971499991361994,
+      "loss": 3.5749,
+      "step": 2688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5616171956062317,
+      "learning_rate": 0.000597147881572439,
+      "loss": 3.7806,
+      "step": 2689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5623195171356201,
+      "learning_rate": 0.0005971457632260482,
+      "loss": 3.4841,
+      "step": 2690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6076975464820862,
+      "learning_rate": 0.0005971436440970326,
+      "loss": 3.7146,
+      "step": 2691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5725387334823608,
+      "learning_rate": 0.0005971415241853979,
+      "loss": 3.622,
+      "step": 2692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5814027786254883,
+      "learning_rate": 0.0005971394034911497,
+      "loss": 3.5529,
+      "step": 2693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5468875169754028,
+      "learning_rate": 0.0005971372820142934,
+      "loss": 3.4851,
+      "step": 2694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6255273222923279,
+      "learning_rate": 0.0005971351597548346,
+      "loss": 3.641,
+      "step": 2695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6269606351852417,
+      "learning_rate": 0.000597133036712779,
+      "loss": 3.7258,
+      "step": 2696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6093981266021729,
+      "learning_rate": 0.0005971309128881324,
+      "loss": 3.3623,
+      "step": 2697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5752262473106384,
+      "learning_rate": 0.0005971287882808999,
+      "loss": 3.6917,
+      "step": 2698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6046422123908997,
+      "learning_rate": 0.0005971266628910874,
+      "loss": 3.7567,
+      "step": 2699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6624324917793274,
+      "learning_rate": 0.0005971245367187005,
+      "loss": 3.6769,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5971564054489136,
+      "learning_rate": 0.0005971224097637447,
+      "loss": 3.7973,
+      "step": 2701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5755377411842346,
+      "learning_rate": 0.0005971202820262258,
+      "loss": 3.5409,
+      "step": 2702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6195607781410217,
+      "learning_rate": 0.0005971181535061491,
+      "loss": 3.6003,
+      "step": 2703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6000617742538452,
+      "learning_rate": 0.0005971160242035204,
+      "loss": 3.6304,
+      "step": 2704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5995765328407288,
+      "learning_rate": 0.0005971138941183454,
+      "loss": 3.4578,
+      "step": 2705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5783835053443909,
+      "learning_rate": 0.0005971117632506295,
+      "loss": 3.609,
+      "step": 2706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5949788093566895,
+      "learning_rate": 0.0005971096316003783,
+      "loss": 3.7868,
+      "step": 2707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5998488068580627,
+      "learning_rate": 0.0005971074991675976,
+      "loss": 3.599,
+      "step": 2708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5909786820411682,
+      "learning_rate": 0.0005971053659522929,
+      "loss": 3.5773,
+      "step": 2709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5821101665496826,
+      "learning_rate": 0.0005971032319544698,
+      "loss": 3.5233,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.587090253829956,
+      "learning_rate": 0.000597101097174134,
+      "loss": 3.7505,
+      "step": 2711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.622183620929718,
+      "learning_rate": 0.0005970989616112913,
+      "loss": 3.3969,
+      "step": 2712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6129473447799683,
+      "learning_rate": 0.0005970968252659468,
+      "loss": 3.6862,
+      "step": 2713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.659932017326355,
+      "learning_rate": 0.0005970946881381066,
+      "loss": 3.5053,
+      "step": 2714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6454612016677856,
+      "learning_rate": 0.0005970925502277761,
+      "loss": 3.4388,
+      "step": 2715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7548837065696716,
+      "learning_rate": 0.0005970904115349612,
+      "loss": 3.51,
+      "step": 2716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6142184138298035,
+      "learning_rate": 0.000597088272059667,
+      "loss": 3.6681,
+      "step": 2717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7135729193687439,
+      "learning_rate": 0.0005970861318018997,
+      "loss": 3.6722,
+      "step": 2718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6415162682533264,
+      "learning_rate": 0.0005970839907616646,
+      "loss": 3.5688,
+      "step": 2719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5897626280784607,
+      "learning_rate": 0.0005970818489389675,
+      "loss": 3.8595,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6751065850257874,
+      "learning_rate": 0.000597079706333814,
+      "loss": 3.435,
+      "step": 2721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6195039749145508,
+      "learning_rate": 0.0005970775629462096,
+      "loss": 3.6634,
+      "step": 2722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6431128978729248,
+      "learning_rate": 0.0005970754187761602,
+      "loss": 3.6171,
+      "step": 2723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5993673205375671,
+      "learning_rate": 0.0005970732738236713,
+      "loss": 3.5535,
+      "step": 2724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6270466446876526,
+      "learning_rate": 0.0005970711280887485,
+      "loss": 3.5313,
+      "step": 2725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5823235511779785,
+      "learning_rate": 0.0005970689815713976,
+      "loss": 3.6458,
+      "step": 2726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5822427272796631,
+      "learning_rate": 0.0005970668342716242,
+      "loss": 3.8305,
+      "step": 2727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5477180480957031,
+      "learning_rate": 0.0005970646861894338,
+      "loss": 3.6481,
+      "step": 2728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.578444242477417,
+      "learning_rate": 0.0005970625373248323,
+      "loss": 3.5803,
+      "step": 2729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6037389636039734,
+      "learning_rate": 0.0005970603876778252,
+      "loss": 3.5236,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6162851452827454,
+      "learning_rate": 0.0005970582372484182,
+      "loss": 3.6997,
+      "step": 2731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6191104650497437,
+      "learning_rate": 0.0005970560860366171,
+      "loss": 3.6575,
+      "step": 2732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6048828959465027,
+      "learning_rate": 0.0005970539340424273,
+      "loss": 3.3794,
+      "step": 2733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6663912534713745,
+      "learning_rate": 0.0005970517812658547,
+      "loss": 3.3146,
+      "step": 2734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5750559568405151,
+      "learning_rate": 0.0005970496277069048,
+      "loss": 3.5452,
+      "step": 2735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5543981790542603,
+      "learning_rate": 0.0005970474733655833,
+      "loss": 3.8258,
+      "step": 2736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5796617865562439,
+      "learning_rate": 0.0005970453182418961,
+      "loss": 3.7764,
+      "step": 2737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5608425140380859,
+      "learning_rate": 0.0005970431623358485,
+      "loss": 3.4656,
+      "step": 2738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6683704853057861,
+      "learning_rate": 0.0005970410056474466,
+      "loss": 3.7797,
+      "step": 2739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6158978343009949,
+      "learning_rate": 0.0005970388481766957,
+      "loss": 3.7061,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5848861932754517,
+      "learning_rate": 0.0005970366899236015,
+      "loss": 3.5359,
+      "step": 2741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5714642405509949,
+      "learning_rate": 0.0005970345308881701,
+      "loss": 3.4968,
+      "step": 2742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5503971576690674,
+      "learning_rate": 0.0005970323710704068,
+      "loss": 3.5294,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6631685495376587,
+      "learning_rate": 0.0005970302104703174,
+      "loss": 3.4137,
+      "step": 2744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5520222187042236,
+      "learning_rate": 0.0005970280490879076,
+      "loss": 3.5508,
+      "step": 2745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5935842990875244,
+      "learning_rate": 0.0005970258869231831,
+      "loss": 3.6917,
+      "step": 2746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6259281635284424,
+      "learning_rate": 0.0005970237239761496,
+      "loss": 3.5549,
+      "step": 2747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6221252679824829,
+      "learning_rate": 0.0005970215602468127,
+      "loss": 3.5163,
+      "step": 2748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.572911262512207,
+      "learning_rate": 0.0005970193957351782,
+      "loss": 3.6548,
+      "step": 2749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5745850801467896,
+      "learning_rate": 0.0005970172304412517,
+      "loss": 3.9044,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390734672546387,
+      "learning_rate": 0.0005970150643650391,
+      "loss": 3.635,
+      "step": 2751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6086849570274353,
+      "learning_rate": 0.0005970128975065459,
+      "loss": 3.527,
+      "step": 2752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5953769683837891,
+      "learning_rate": 0.000597010729865778,
+      "loss": 3.5134,
+      "step": 2753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5471598505973816,
+      "learning_rate": 0.0005970085614427409,
+      "loss": 3.4587,
+      "step": 2754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5561854243278503,
+      "learning_rate": 0.0005970063922374404,
+      "loss": 3.6646,
+      "step": 2755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5902842283248901,
+      "learning_rate": 0.0005970042222498823,
+      "loss": 3.4375,
+      "step": 2756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6033221483230591,
+      "learning_rate": 0.0005970020514800722,
+      "loss": 3.4188,
+      "step": 2757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5598551034927368,
+      "learning_rate": 0.0005969998799280158,
+      "loss": 3.5559,
+      "step": 2758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5618605017662048,
+      "learning_rate": 0.0005969977075937189,
+      "loss": 3.778,
+      "step": 2759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6076387763023376,
+      "learning_rate": 0.0005969955344771872,
+      "loss": 3.5922,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5863784551620483,
+      "learning_rate": 0.0005969933605784265,
+      "loss": 3.5362,
+      "step": 2761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5936189293861389,
+      "learning_rate": 0.0005969911858974424,
+      "loss": 3.6277,
+      "step": 2762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5283812880516052,
+      "learning_rate": 0.0005969890104342406,
+      "loss": 3.4249,
+      "step": 2763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6723504066467285,
+      "learning_rate": 0.000596986834188827,
+      "loss": 3.583,
+      "step": 2764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5366487503051758,
+      "learning_rate": 0.0005969846571612072,
+      "loss": 3.6369,
+      "step": 2765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6037803292274475,
+      "learning_rate": 0.0005969824793513871,
+      "loss": 3.4389,
+      "step": 2766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6620571613311768,
+      "learning_rate": 0.000596980300759372,
+      "loss": 3.6024,
+      "step": 2767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5733856558799744,
+      "learning_rate": 0.0005969781213851682,
+      "loss": 3.4694,
+      "step": 2768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5712093710899353,
+      "learning_rate": 0.000596975941228781,
+      "loss": 3.638,
+      "step": 2769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6224941611289978,
+      "learning_rate": 0.0005969737602902165,
+      "loss": 3.4212,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6067854166030884,
+      "learning_rate": 0.0005969715785694802,
+      "loss": 3.6626,
+      "step": 2771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5889960527420044,
+      "learning_rate": 0.0005969693960665779,
+      "loss": 3.641,
+      "step": 2772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6073998212814331,
+      "learning_rate": 0.0005969672127815153,
+      "loss": 3.3504,
+      "step": 2773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5916741490364075,
+      "learning_rate": 0.0005969650287142983,
+      "loss": 3.6483,
+      "step": 2774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5662715435028076,
+      "learning_rate": 0.0005969628438649326,
+      "loss": 3.3069,
+      "step": 2775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5828217267990112,
+      "learning_rate": 0.0005969606582334238,
+      "loss": 3.6519,
+      "step": 2776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5853772759437561,
+      "learning_rate": 0.0005969584718197779,
+      "loss": 3.5239,
+      "step": 2777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5956276655197144,
+      "learning_rate": 0.0005969562846240005,
+      "loss": 3.6071,
+      "step": 2778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5762497782707214,
+      "learning_rate": 0.0005969540966460974,
+      "loss": 3.738,
+      "step": 2779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5765044093132019,
+      "learning_rate": 0.0005969519078860742,
+      "loss": 3.575,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7366907596588135,
+      "learning_rate": 0.0005969497183439371,
+      "loss": 3.4483,
+      "step": 2781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5674998164176941,
+      "learning_rate": 0.0005969475280196915,
+      "loss": 3.6753,
+      "step": 2782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5822417140007019,
+      "learning_rate": 0.0005969453369133431,
+      "loss": 3.5733,
+      "step": 2783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5587081909179688,
+      "learning_rate": 0.000596943145024898,
+      "loss": 3.5532,
+      "step": 2784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5745068192481995,
+      "learning_rate": 0.0005969409523543618,
+      "loss": 3.5765,
+      "step": 2785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5741488933563232,
+      "learning_rate": 0.0005969387589017402,
+      "loss": 3.6993,
+      "step": 2786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5863240957260132,
+      "learning_rate": 0.0005969365646670391,
+      "loss": 3.6545,
+      "step": 2787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6137883067131042,
+      "learning_rate": 0.0005969343696502642,
+      "loss": 3.6242,
+      "step": 2788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.593112587928772,
+      "learning_rate": 0.0005969321738514213,
+      "loss": 3.5533,
+      "step": 2789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5809875130653381,
+      "learning_rate": 0.0005969299772705163,
+      "loss": 3.8685,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5593841671943665,
+      "learning_rate": 0.0005969277799075548,
+      "loss": 3.6496,
+      "step": 2791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5753694176673889,
+      "learning_rate": 0.0005969255817625428,
+      "loss": 3.4793,
+      "step": 2792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5818933248519897,
+      "learning_rate": 0.0005969233828354857,
+      "loss": 3.7245,
+      "step": 2793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5473413467407227,
+      "learning_rate": 0.0005969211831263899,
+      "loss": 3.3412,
+      "step": 2794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5579399466514587,
+      "learning_rate": 0.0005969189826352605,
+      "loss": 3.6375,
+      "step": 2795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6152604222297668,
+      "learning_rate": 0.0005969167813621038,
+      "loss": 3.5726,
+      "step": 2796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6010841727256775,
+      "learning_rate": 0.0005969145793069255,
+      "loss": 3.6222,
+      "step": 2797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5537547469139099,
+      "learning_rate": 0.0005969123764697313,
+      "loss": 3.7334,
+      "step": 2798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5809540748596191,
+      "learning_rate": 0.0005969101728505269,
+      "loss": 3.7945,
+      "step": 2799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5480958223342896,
+      "learning_rate": 0.0005969079684493184,
+      "loss": 3.6835,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5858062505722046,
+      "learning_rate": 0.0005969057632661115,
+      "loss": 3.5088,
+      "step": 2801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6234527230262756,
+      "learning_rate": 0.0005969035573009119,
+      "loss": 3.5747,
+      "step": 2802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5944379568099976,
+      "learning_rate": 0.0005969013505537254,
+      "loss": 3.6146,
+      "step": 2803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.544714093208313,
+      "learning_rate": 0.0005968991430245579,
+      "loss": 3.5981,
+      "step": 2804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5830975770950317,
+      "learning_rate": 0.0005968969347134154,
+      "loss": 3.5539,
+      "step": 2805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5846774578094482,
+      "learning_rate": 0.0005968947256203033,
+      "loss": 3.4609,
+      "step": 2806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5505405068397522,
+      "learning_rate": 0.0005968925157452276,
+      "loss": 3.5572,
+      "step": 2807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.584732174873352,
+      "learning_rate": 0.0005968903050881943,
+      "loss": 3.5472,
+      "step": 2808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6293777227401733,
+      "learning_rate": 0.0005968880936492089,
+      "loss": 3.5822,
+      "step": 2809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6841658353805542,
+      "learning_rate": 0.0005968858814282776,
+      "loss": 3.4138,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6176837086677551,
+      "learning_rate": 0.0005968836684254058,
+      "loss": 3.4622,
+      "step": 2811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5423951148986816,
+      "learning_rate": 0.0005968814546405998,
+      "loss": 3.7367,
+      "step": 2812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5896667242050171,
+      "learning_rate": 0.000596879240073865,
+      "loss": 3.8234,
+      "step": 2813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6115859150886536,
+      "learning_rate": 0.0005968770247252074,
+      "loss": 3.5844,
+      "step": 2814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5899752378463745,
+      "learning_rate": 0.000596874808594633,
+      "loss": 3.6039,
+      "step": 2815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5383777022361755,
+      "learning_rate": 0.0005968725916821474,
+      "loss": 3.5118,
+      "step": 2816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5764883756637573,
+      "learning_rate": 0.0005968703739877566,
+      "loss": 3.7468,
+      "step": 2817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.927655816078186,
+      "learning_rate": 0.0005968681555114663,
+      "loss": 3.5418,
+      "step": 2818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5765186548233032,
+      "learning_rate": 0.0005968659362532824,
+      "loss": 3.6529,
+      "step": 2819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5878188610076904,
+      "learning_rate": 0.0005968637162132108,
+      "loss": 3.6059,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.637100875377655,
+      "learning_rate": 0.0005968614953912574,
+      "loss": 3.6037,
+      "step": 2821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6326918601989746,
+      "learning_rate": 0.0005968592737874278,
+      "loss": 3.695,
+      "step": 2822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5842844843864441,
+      "learning_rate": 0.000596857051401728,
+      "loss": 3.7953,
+      "step": 2823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6204371452331543,
+      "learning_rate": 0.0005968548282341639,
+      "loss": 3.5894,
+      "step": 2824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.556685209274292,
+      "learning_rate": 0.0005968526042847413,
+      "loss": 3.606,
+      "step": 2825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5710826516151428,
+      "learning_rate": 0.0005968503795534661,
+      "loss": 3.5026,
+      "step": 2826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.569648802280426,
+      "learning_rate": 0.0005968481540403441,
+      "loss": 3.6895,
+      "step": 2827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.585290253162384,
+      "learning_rate": 0.0005968459277453813,
+      "loss": 3.5407,
+      "step": 2828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5872741937637329,
+      "learning_rate": 0.0005968437006685834,
+      "loss": 3.7963,
+      "step": 2829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5374106168746948,
+      "learning_rate": 0.0005968414728099563,
+      "loss": 3.6502,
+      "step": 2830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5915255546569824,
+      "learning_rate": 0.0005968392441695058,
+      "loss": 3.4514,
+      "step": 2831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6637536287307739,
+      "learning_rate": 0.000596837014747238,
+      "loss": 3.5065,
+      "step": 2832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5768295526504517,
+      "learning_rate": 0.0005968347845431586,
+      "loss": 3.5267,
+      "step": 2833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.595833957195282,
+      "learning_rate": 0.0005968325535572734,
+      "loss": 3.5768,
+      "step": 2834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5429056286811829,
+      "learning_rate": 0.0005968303217895885,
+      "loss": 3.569,
+      "step": 2835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6460859775543213,
+      "learning_rate": 0.0005968280892401095,
+      "loss": 3.503,
+      "step": 2836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6241177916526794,
+      "learning_rate": 0.0005968258559088427,
+      "loss": 3.6161,
+      "step": 2837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5874194502830505,
+      "learning_rate": 0.0005968236217957935,
+      "loss": 3.6232,
+      "step": 2838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5746603608131409,
+      "learning_rate": 0.0005968213869009681,
+      "loss": 3.5134,
+      "step": 2839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5978888273239136,
+      "learning_rate": 0.0005968191512243723,
+      "loss": 3.8172,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6809991002082825,
+      "learning_rate": 0.0005968169147660119,
+      "loss": 3.5159,
+      "step": 2841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6131600141525269,
+      "learning_rate": 0.0005968146775258929,
+      "loss": 3.5549,
+      "step": 2842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5964472889900208,
+      "learning_rate": 0.0005968124395040212,
+      "loss": 3.4967,
+      "step": 2843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.599494457244873,
+      "learning_rate": 0.0005968102007004026,
+      "loss": 3.5785,
+      "step": 2844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5741751790046692,
+      "learning_rate": 0.0005968079611150431,
+      "loss": 3.7153,
+      "step": 2845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5869643688201904,
+      "learning_rate": 0.0005968057207479486,
+      "loss": 3.5068,
+      "step": 2846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5856844186782837,
+      "learning_rate": 0.0005968034795991249,
+      "loss": 3.8081,
+      "step": 2847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6089456677436829,
+      "learning_rate": 0.000596801237668578,
+      "loss": 3.6488,
+      "step": 2848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5712987780570984,
+      "learning_rate": 0.0005967989949563136,
+      "loss": 3.4939,
+      "step": 2849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.560558021068573,
+      "learning_rate": 0.000596796751462338,
+      "loss": 3.4375,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6255397200584412,
+      "learning_rate": 0.0005967945071866567,
+      "loss": 3.4126,
+      "step": 2851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5810412764549255,
+      "learning_rate": 0.0005967922621292758,
+      "loss": 3.5257,
+      "step": 2852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6136355400085449,
+      "learning_rate": 0.0005967900162902013,
+      "loss": 3.5239,
+      "step": 2853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6601942777633667,
+      "learning_rate": 0.000596787769669439,
+      "loss": 3.6628,
+      "step": 2854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5993295311927795,
+      "learning_rate": 0.0005967855222669947,
+      "loss": 3.4613,
+      "step": 2855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6075938940048218,
+      "learning_rate": 0.0005967832740828746,
+      "loss": 3.4571,
+      "step": 2856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7354562878608704,
+      "learning_rate": 0.0005967810251170844,
+      "loss": 3.6329,
+      "step": 2857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5470739603042603,
+      "learning_rate": 0.0005967787753696301,
+      "loss": 3.309,
+      "step": 2858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7858545780181885,
+      "learning_rate": 0.0005967765248405177,
+      "loss": 3.6904,
+      "step": 2859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6135296821594238,
+      "learning_rate": 0.0005967742735297529,
+      "loss": 3.6427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5984460711479187,
+      "learning_rate": 0.0005967720214373419,
+      "loss": 3.8465,
+      "step": 2861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5901587009429932,
+      "learning_rate": 0.0005967697685632904,
+      "loss": 3.6866,
+      "step": 2862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5759127140045166,
+      "learning_rate": 0.0005967675149076046,
+      "loss": 3.6211,
+      "step": 2863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6124855279922485,
+      "learning_rate": 0.0005967652604702902,
+      "loss": 3.5057,
+      "step": 2864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5732986927032471,
+      "learning_rate": 0.0005967630052513532,
+      "loss": 3.6978,
+      "step": 2865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5618267059326172,
+      "learning_rate": 0.0005967607492507995,
+      "loss": 3.7548,
+      "step": 2866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6002089381217957,
+      "learning_rate": 0.0005967584924686353,
+      "loss": 3.7316,
+      "step": 2867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5847598314285278,
+      "learning_rate": 0.0005967562349048662,
+      "loss": 3.5734,
+      "step": 2868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5832062363624573,
+      "learning_rate": 0.0005967539765594983,
+      "loss": 3.7337,
+      "step": 2869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5734034776687622,
+      "learning_rate": 0.0005967517174325375,
+      "loss": 3.6052,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5718584656715393,
+      "learning_rate": 0.0005967494575239899,
+      "loss": 3.6527,
+      "step": 2871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6056542992591858,
+      "learning_rate": 0.0005967471968338612,
+      "loss": 3.7136,
+      "step": 2872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6068187952041626,
+      "learning_rate": 0.0005967449353621575,
+      "loss": 3.7391,
+      "step": 2873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5216867327690125,
+      "learning_rate": 0.0005967426731088849,
+      "loss": 3.6024,
+      "step": 2874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6825017929077148,
+      "learning_rate": 0.000596740410074049,
+      "loss": 3.6602,
+      "step": 2875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5960718393325806,
+      "learning_rate": 0.0005967381462576561,
+      "loss": 3.6202,
+      "step": 2876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.594954252243042,
+      "learning_rate": 0.0005967358816597121,
+      "loss": 3.5797,
+      "step": 2877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5977593660354614,
+      "learning_rate": 0.0005967336162802228,
+      "loss": 3.6304,
+      "step": 2878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5887445211410522,
+      "learning_rate": 0.0005967313501191942,
+      "loss": 3.5324,
+      "step": 2879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5791778564453125,
+      "learning_rate": 0.0005967290831766325,
+      "loss": 3.7551,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6645099520683289,
+      "learning_rate": 0.0005967268154525434,
+      "loss": 3.468,
+      "step": 2881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5834534168243408,
+      "learning_rate": 0.000596724546946933,
+      "loss": 3.7626,
+      "step": 2882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.600719153881073,
+      "learning_rate": 0.0005967222776598072,
+      "loss": 3.6733,
+      "step": 2883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6395867466926575,
+      "learning_rate": 0.0005967200075911722,
+      "loss": 3.5922,
+      "step": 2884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5920771360397339,
+      "learning_rate": 0.0005967177367410335,
+      "loss": 3.7339,
+      "step": 2885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5998722910881042,
+      "learning_rate": 0.0005967154651093977,
+      "loss": 3.6804,
+      "step": 2886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6235655546188354,
+      "learning_rate": 0.0005967131926962703,
+      "loss": 3.6184,
+      "step": 2887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5598371624946594,
+      "learning_rate": 0.0005967109195016575,
+      "loss": 3.4734,
+      "step": 2888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5574737787246704,
+      "learning_rate": 0.0005967086455255651,
+      "loss": 3.7455,
+      "step": 2889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.60200434923172,
+      "learning_rate": 0.0005967063707679994,
+      "loss": 3.5133,
+      "step": 2890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5801019668579102,
+      "learning_rate": 0.0005967040952289661,
+      "loss": 3.6737,
+      "step": 2891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.567669153213501,
+      "learning_rate": 0.0005967018189084715,
+      "loss": 3.7216,
+      "step": 2892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7055478096008301,
+      "learning_rate": 0.0005966995418065212,
+      "loss": 3.4534,
+      "step": 2893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.562245786190033,
+      "learning_rate": 0.0005966972639231215,
+      "loss": 3.7154,
+      "step": 2894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5916703343391418,
+      "learning_rate": 0.0005966949852582783,
+      "loss": 3.5089,
+      "step": 2895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5672451853752136,
+      "learning_rate": 0.0005966927058119976,
+      "loss": 3.7155,
+      "step": 2896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5992611646652222,
+      "learning_rate": 0.0005966904255842856,
+      "loss": 3.7891,
+      "step": 2897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5881920456886292,
+      "learning_rate": 0.0005966881445751478,
+      "loss": 3.6009,
+      "step": 2898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5571060180664062,
+      "learning_rate": 0.0005966858627845907,
+      "loss": 3.7047,
+      "step": 2899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6121320128440857,
+      "learning_rate": 0.0005966835802126201,
+      "loss": 3.518,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6062599420547485,
+      "learning_rate": 0.000596681296859242,
+      "loss": 3.3848,
+      "step": 2901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6015748381614685,
+      "learning_rate": 0.0005966790127244626,
+      "loss": 3.5153,
+      "step": 2902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5726820230484009,
+      "learning_rate": 0.0005966767278082877,
+      "loss": 3.7427,
+      "step": 2903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5963112115859985,
+      "learning_rate": 0.0005966744421107234,
+      "loss": 3.701,
+      "step": 2904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5752584934234619,
+      "learning_rate": 0.0005966721556317757,
+      "loss": 3.7769,
+      "step": 2905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56533282995224,
+      "learning_rate": 0.0005966698683714506,
+      "loss": 3.4874,
+      "step": 2906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5514774918556213,
+      "learning_rate": 0.0005966675803297542,
+      "loss": 3.4779,
+      "step": 2907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6327130794525146,
+      "learning_rate": 0.0005966652915066925,
+      "loss": 3.6413,
+      "step": 2908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.562054455280304,
+      "learning_rate": 0.0005966630019022715,
+      "loss": 3.7253,
+      "step": 2909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5362357497215271,
+      "learning_rate": 0.0005966607115164974,
+      "loss": 3.7303,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5918092131614685,
+      "learning_rate": 0.0005966584203493759,
+      "loss": 3.5367,
+      "step": 2911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6234297752380371,
+      "learning_rate": 0.0005966561284009132,
+      "loss": 3.6713,
+      "step": 2912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6105513572692871,
+      "learning_rate": 0.0005966538356711154,
+      "loss": 3.6219,
+      "step": 2913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.527632474899292,
+      "learning_rate": 0.0005966515421599885,
+      "loss": 3.4177,
+      "step": 2914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6240413188934326,
+      "learning_rate": 0.0005966492478675384,
+      "loss": 3.6584,
+      "step": 2915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.584805965423584,
+      "learning_rate": 0.0005966469527937716,
+      "loss": 3.5188,
+      "step": 2916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.581307053565979,
+      "learning_rate": 0.0005966446569386936,
+      "loss": 3.5399,
+      "step": 2917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6673707962036133,
+      "learning_rate": 0.0005966423603023105,
+      "loss": 3.5465,
+      "step": 2918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5942588448524475,
+      "learning_rate": 0.0005966400628846288,
+      "loss": 3.4065,
+      "step": 2919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5521858930587769,
+      "learning_rate": 0.0005966377646856541,
+      "loss": 3.5278,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5489894151687622,
+      "learning_rate": 0.0005966354657053926,
+      "loss": 3.805,
+      "step": 2921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0046864748001099,
+      "learning_rate": 0.0005966331659438505,
+      "loss": 3.5583,
+      "step": 2922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5688363909721375,
+      "learning_rate": 0.0005966308654010337,
+      "loss": 3.5806,
+      "step": 2923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5838121771812439,
+      "learning_rate": 0.0005966285640769482,
+      "loss": 3.404,
+      "step": 2924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6157076954841614,
+      "learning_rate": 0.0005966262619716003,
+      "loss": 3.6069,
+      "step": 2925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5905383229255676,
+      "learning_rate": 0.0005966239590849958,
+      "loss": 3.6116,
+      "step": 2926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5673978328704834,
+      "learning_rate": 0.0005966216554171409,
+      "loss": 3.3259,
+      "step": 2927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6059072613716125,
+      "learning_rate": 0.0005966193509680416,
+      "loss": 3.8699,
+      "step": 2928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5982714295387268,
+      "learning_rate": 0.000596617045737704,
+      "loss": 3.6271,
+      "step": 2929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5978359580039978,
+      "learning_rate": 0.0005966147397261344,
+      "loss": 3.55,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5846501588821411,
+      "learning_rate": 0.0005966124329333384,
+      "loss": 3.72,
+      "step": 2931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5485678911209106,
+      "learning_rate": 0.0005966101253593226,
+      "loss": 3.6435,
+      "step": 2932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6040960550308228,
+      "learning_rate": 0.0005966078170040925,
+      "loss": 3.4047,
+      "step": 2933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.629386842250824,
+      "learning_rate": 0.0005966055078676547,
+      "loss": 3.6221,
+      "step": 2934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5997804999351501,
+      "learning_rate": 0.000596603197950015,
+      "loss": 4.0703,
+      "step": 2935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6014472842216492,
+      "learning_rate": 0.0005966008872511796,
+      "loss": 3.6938,
+      "step": 2936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5625265836715698,
+      "learning_rate": 0.0005965985757711545,
+      "loss": 3.6621,
+      "step": 2937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5449039936065674,
+      "learning_rate": 0.0005965962635099459,
+      "loss": 3.648,
+      "step": 2938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5726627707481384,
+      "learning_rate": 0.0005965939504675598,
+      "loss": 3.6625,
+      "step": 2939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5476054549217224,
+      "learning_rate": 0.0005965916366440023,
+      "loss": 3.7134,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6098428964614868,
+      "learning_rate": 0.0005965893220392795,
+      "loss": 3.9104,
+      "step": 2941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6405155062675476,
+      "learning_rate": 0.0005965870066533975,
+      "loss": 3.6453,
+      "step": 2942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6072257161140442,
+      "learning_rate": 0.0005965846904863625,
+      "loss": 3.657,
+      "step": 2943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5281662940979004,
+      "learning_rate": 0.0005965823735381803,
+      "loss": 3.8624,
+      "step": 2944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5664449334144592,
+      "learning_rate": 0.0005965800558088573,
+      "loss": 3.7622,
+      "step": 2945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5815826654434204,
+      "learning_rate": 0.0005965777372983996,
+      "loss": 3.7028,
+      "step": 2946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5717793107032776,
+      "learning_rate": 0.0005965754180068132,
+      "loss": 3.5994,
+      "step": 2947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56984543800354,
+      "learning_rate": 0.0005965730979341042,
+      "loss": 3.4422,
+      "step": 2948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5990952849388123,
+      "learning_rate": 0.0005965707770802786,
+      "loss": 3.5145,
+      "step": 2949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5824059247970581,
+      "learning_rate": 0.0005965684554453428,
+      "loss": 3.5848,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5779900550842285,
+      "learning_rate": 0.0005965661330293027,
+      "loss": 3.6382,
+      "step": 2951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5480761528015137,
+      "learning_rate": 0.0005965638098321646,
+      "loss": 3.6473,
+      "step": 2952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5885190367698669,
+      "learning_rate": 0.0005965614858539344,
+      "loss": 3.5578,
+      "step": 2953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5397934913635254,
+      "learning_rate": 0.0005965591610946184,
+      "loss": 3.547,
+      "step": 2954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5664386749267578,
+      "learning_rate": 0.0005965568355542225,
+      "loss": 3.7429,
+      "step": 2955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5953521728515625,
+      "learning_rate": 0.0005965545092327531,
+      "loss": 3.6301,
+      "step": 2956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5471416711807251,
+      "learning_rate": 0.0005965521821302162,
+      "loss": 3.6399,
+      "step": 2957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5689468383789062,
+      "learning_rate": 0.0005965498542466178,
+      "loss": 3.6759,
+      "step": 2958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5762372612953186,
+      "learning_rate": 0.0005965475255819642,
+      "loss": 3.6877,
+      "step": 2959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5529139637947083,
+      "learning_rate": 0.0005965451961362615,
+      "loss": 3.4141,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5945937633514404,
+      "learning_rate": 0.0005965428659095158,
+      "loss": 3.5489,
+      "step": 2961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.548331081867218,
+      "learning_rate": 0.0005965405349017334,
+      "loss": 3.6366,
+      "step": 2962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6198789477348328,
+      "learning_rate": 0.0005965382031129201,
+      "loss": 3.5774,
+      "step": 2963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6308538913726807,
+      "learning_rate": 0.0005965358705430824,
+      "loss": 3.5777,
+      "step": 2964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5835420489311218,
+      "learning_rate": 0.0005965335371922261,
+      "loss": 3.4532,
+      "step": 2965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.56312495470047,
+      "learning_rate": 0.0005965312030603578,
+      "loss": 3.6189,
+      "step": 2966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6242998838424683,
+      "learning_rate": 0.0005965288681474832,
+      "loss": 3.4261,
+      "step": 2967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5804765224456787,
+      "learning_rate": 0.0005965265324536087,
+      "loss": 3.6305,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5844853520393372,
+      "learning_rate": 0.0005965241959787404,
+      "loss": 3.4267,
+      "step": 2969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5954545140266418,
+      "learning_rate": 0.0005965218587228844,
+      "loss": 3.615,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5783625245094299,
+      "learning_rate": 0.0005965195206860468,
+      "loss": 3.7189,
+      "step": 2971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5781705975532532,
+      "learning_rate": 0.0005965171818682339,
+      "loss": 3.6033,
+      "step": 2972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5989189147949219,
+      "learning_rate": 0.0005965148422694518,
+      "loss": 3.5753,
+      "step": 2973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6061220169067383,
+      "learning_rate": 0.0005965125018897068,
+      "loss": 3.6133,
+      "step": 2974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6106086373329163,
+      "learning_rate": 0.0005965101607290047,
+      "loss": 3.5082,
+      "step": 2975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.540738046169281,
+      "learning_rate": 0.0005965078187873522,
+      "loss": 3.3238,
+      "step": 2976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5824552774429321,
+      "learning_rate": 0.0005965054760647549,
+      "loss": 3.611,
+      "step": 2977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5599578619003296,
+      "learning_rate": 0.0005965031325612195,
+      "loss": 3.5603,
+      "step": 2978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5765003561973572,
+      "learning_rate": 0.0005965007882767516,
+      "loss": 3.4808,
+      "step": 2979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5761189460754395,
+      "learning_rate": 0.0005964984432113579,
+      "loss": 3.5618,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5370120406150818,
+      "learning_rate": 0.0005964960973650444,
+      "loss": 3.479,
+      "step": 2981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5539388060569763,
+      "learning_rate": 0.0005964937507378171,
+      "loss": 3.5543,
+      "step": 2982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5843010544776917,
+      "learning_rate": 0.0005964914033296824,
+      "loss": 3.4738,
+      "step": 2983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5479416251182556,
+      "learning_rate": 0.0005964890551406463,
+      "loss": 3.6649,
+      "step": 2984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.590008020401001,
+      "learning_rate": 0.0005964867061707152,
+      "loss": 3.6951,
+      "step": 2985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6158657073974609,
+      "learning_rate": 0.0005964843564198951,
+      "loss": 3.5394,
+      "step": 2986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5672131180763245,
+      "learning_rate": 0.0005964820058881925,
+      "loss": 3.5563,
+      "step": 2987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5937507748603821,
+      "learning_rate": 0.000596479654575613,
+      "loss": 3.7613,
+      "step": 2988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6074742078781128,
+      "learning_rate": 0.0005964773024821633,
+      "loss": 3.6806,
+      "step": 2989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5991113185882568,
+      "learning_rate": 0.0005964749496078495,
+      "loss": 3.528,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5741263031959534,
+      "learning_rate": 0.0005964725959526777,
+      "loss": 3.4517,
+      "step": 2991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6311984658241272,
+      "learning_rate": 0.0005964702415166541,
+      "loss": 3.5116,
+      "step": 2992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5775881409645081,
+      "learning_rate": 0.000596467886299785,
+      "loss": 3.4346,
+      "step": 2993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5953471660614014,
+      "learning_rate": 0.0005964655303020766,
+      "loss": 3.566,
+      "step": 2994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6620184779167175,
+      "learning_rate": 0.0005964631735235349,
+      "loss": 3.7845,
+      "step": 2995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5985296964645386,
+      "learning_rate": 0.0005964608159641664,
+      "loss": 3.6863,
+      "step": 2996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6449403166770935,
+      "learning_rate": 0.000596458457623977,
+      "loss": 3.5711,
+      "step": 2997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.583063006401062,
+      "learning_rate": 0.0005964560985029732,
+      "loss": 3.3602,
+      "step": 2998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5750587582588196,
+      "learning_rate": 0.0005964537386011611,
+      "loss": 3.6342,
+      "step": 2999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5808610320091248,
+      "learning_rate": 0.0005964513779185468,
+      "loss": 3.5019,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6237676739692688,
+      "learning_rate": 0.0005964490164551367,
+      "loss": 3.5213,
+      "step": 3001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5874052047729492,
+      "learning_rate": 0.000596446654210937,
+      "loss": 3.3836,
+      "step": 3002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5484369993209839,
+      "learning_rate": 0.0005964442911859538,
+      "loss": 3.2887,
+      "step": 3003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5622039437294006,
+      "learning_rate": 0.0005964419273801934,
+      "loss": 3.5959,
+      "step": 3004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.601517915725708,
+      "learning_rate": 0.000596439562793662,
+      "loss": 3.588,
+      "step": 3005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6053403615951538,
+      "learning_rate": 0.0005964371974263658,
+      "loss": 3.6096,
+      "step": 3006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5661147832870483,
+      "learning_rate": 0.0005964348312783111,
+      "loss": 3.52,
+      "step": 3007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6172541975975037,
+      "learning_rate": 0.0005964324643495042,
+      "loss": 3.6069,
+      "step": 3008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5625926852226257,
+      "learning_rate": 0.0005964300966399511,
+      "loss": 3.6741,
+      "step": 3009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5759730339050293,
+      "learning_rate": 0.0005964277281496583,
+      "loss": 3.5645,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5957322120666504,
+      "learning_rate": 0.0005964253588786318,
+      "loss": 3.5519,
+      "step": 3011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5785747170448303,
+      "learning_rate": 0.000596422988826878,
+      "loss": 3.4793,
+      "step": 3012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5878031849861145,
+      "learning_rate": 0.0005964206179944031,
+      "loss": 3.5527,
+      "step": 3013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5843365788459778,
+      "learning_rate": 0.0005964182463812133,
+      "loss": 3.5102,
+      "step": 3014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5953040719032288,
+      "learning_rate": 0.000596415873987315,
+      "loss": 3.6765,
+      "step": 3015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5682270526885986,
+      "learning_rate": 0.0005964135008127143,
+      "loss": 3.6247,
+      "step": 3016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5600559115409851,
+      "learning_rate": 0.0005964111268574174,
+      "loss": 3.5984,
+      "step": 3017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6236439943313599,
+      "learning_rate": 0.0005964087521214307,
+      "loss": 3.5485,
+      "step": 3018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5868257284164429,
+      "learning_rate": 0.0005964063766047602,
+      "loss": 3.6191,
+      "step": 3019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5846828818321228,
+      "learning_rate": 0.0005964040003074125,
+      "loss": 3.3374,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5583348274230957,
+      "learning_rate": 0.0005964016232293939,
+      "loss": 3.6922,
+      "step": 3021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6195686459541321,
+      "learning_rate": 0.0005963992453707102,
+      "loss": 3.708,
+      "step": 3022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6074361801147461,
+      "learning_rate": 0.0005963968667313681,
+      "loss": 3.6087,
+      "step": 3023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5756213665008545,
+      "learning_rate": 0.0005963944873113736,
+      "loss": 3.5665,
+      "step": 3024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6131848096847534,
+      "learning_rate": 0.0005963921071107331,
+      "loss": 3.1928,
+      "step": 3025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.550272524356842,
+      "learning_rate": 0.0005963897261294528,
+      "loss": 3.6607,
+      "step": 3026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6190971732139587,
+      "learning_rate": 0.000596387344367539,
+      "loss": 3.619,
+      "step": 3027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6038632392883301,
+      "learning_rate": 0.000596384961824998,
+      "loss": 3.775,
+      "step": 3028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5631963014602661,
+      "learning_rate": 0.0005963825785018361,
+      "loss": 3.5747,
+      "step": 3029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5642056465148926,
+      "learning_rate": 0.0005963801943980595,
+      "loss": 3.542,
+      "step": 3030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5921563506126404,
+      "learning_rate": 0.0005963778095136744,
+      "loss": 3.5717,
+      "step": 3031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5655155777931213,
+      "learning_rate": 0.0005963754238486873,
+      "loss": 3.5262,
+      "step": 3032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5852659940719604,
+      "learning_rate": 0.0005963730374031044,
+      "loss": 3.6345,
+      "step": 3033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5611869096755981,
+      "learning_rate": 0.0005963706501769319,
+      "loss": 3.6346,
+      "step": 3034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5403077602386475,
+      "learning_rate": 0.0005963682621701761,
+      "loss": 3.7496,
+      "step": 3035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5627409219741821,
+      "learning_rate": 0.0005963658733828433,
+      "loss": 3.6472,
+      "step": 3036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5713598132133484,
+      "learning_rate": 0.00059636348381494,
+      "loss": 3.3384,
+      "step": 3037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6267187595367432,
+      "learning_rate": 0.0005963610934664722,
+      "loss": 3.5897,
+      "step": 3038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6044197678565979,
+      "learning_rate": 0.0005963587023374463,
+      "loss": 3.7926,
+      "step": 3039
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6235830783843994,
+      "learning_rate": 0.0005963563104278687,
+      "loss": 3.7628,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6233013868331909,
+      "learning_rate": 0.0005963539177377456,
+      "loss": 3.4503,
+      "step": 3041
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5676363110542297,
+      "learning_rate": 0.0005963515242670833,
+      "loss": 3.5615,
+      "step": 3042
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6279856562614441,
+      "learning_rate": 0.0005963491300158881,
+      "loss": 3.6456,
+      "step": 3043
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5585012435913086,
+      "learning_rate": 0.0005963467349841663,
+      "loss": 3.5764,
+      "step": 3044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5672016143798828,
+      "learning_rate": 0.0005963443391719242,
+      "loss": 3.3576,
+      "step": 3045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6148715019226074,
+      "learning_rate": 0.0005963419425791683,
+      "loss": 3.7448,
+      "step": 3046
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5559766888618469,
+      "learning_rate": 0.0005963395452059046,
+      "loss": 3.7044,
+      "step": 3047
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5408651232719421,
+      "learning_rate": 0.0005963371470521396,
+      "loss": 3.5665,
+      "step": 3048
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5515286922454834,
+      "learning_rate": 0.0005963347481178796,
+      "loss": 3.5655,
+      "step": 3049
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5659925937652588,
+      "learning_rate": 0.0005963323484031309,
+      "loss": 3.7384,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5940973162651062,
+      "learning_rate": 0.0005963299479078999,
+      "loss": 3.6098,
+      "step": 3051
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5960044264793396,
+      "learning_rate": 0.0005963275466321928,
+      "loss": 3.5784,
+      "step": 3052
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5913186073303223,
+      "learning_rate": 0.000596325144576016,
+      "loss": 3.3696,
+      "step": 3053
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5472952723503113,
+      "learning_rate": 0.0005963227417393758,
+      "loss": 3.3607,
+      "step": 3054
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5894946455955505,
+      "learning_rate": 0.0005963203381222784,
+      "loss": 3.5893,
+      "step": 3055
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.586334228515625,
+      "learning_rate": 0.0005963179337247303,
+      "loss": 3.6586,
+      "step": 3056
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5724272131919861,
+      "learning_rate": 0.0005963155285467379,
+      "loss": 3.7376,
+      "step": 3057
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5854930281639099,
+      "learning_rate": 0.0005963131225883074,
+      "loss": 3.6825,
+      "step": 3058
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5693349242210388,
+      "learning_rate": 0.0005963107158494451,
+      "loss": 3.5495,
+      "step": 3059
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5654905438423157,
+      "learning_rate": 0.0005963083083301575,
+      "loss": 3.6469,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5747386813163757,
+      "learning_rate": 0.0005963059000304507,
+      "loss": 3.6497,
+      "step": 3061
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6350069046020508,
+      "learning_rate": 0.0005963034909503311,
+      "loss": 3.6941,
+      "step": 3062
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6059046387672424,
+      "learning_rate": 0.0005963010810898054,
+      "loss": 3.5363,
+      "step": 3063
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5842269062995911,
+      "learning_rate": 0.0005962986704488796,
+      "loss": 3.7521,
+      "step": 3064
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5611490607261658,
+      "learning_rate": 0.00059629625902756,
+      "loss": 3.6577,
+      "step": 3065
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5932784676551819,
+      "learning_rate": 0.0005962938468258532,
+      "loss": 3.5643,
+      "step": 3066
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.596534252166748,
+      "learning_rate": 0.0005962914338437654,
+      "loss": 3.7409,
+      "step": 3067
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5720816254615784,
+      "learning_rate": 0.0005962890200813029,
+      "loss": 3.7528,
+      "step": 3068
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5783271193504333,
+      "learning_rate": 0.0005962866055384722,
+      "loss": 3.8315,
+      "step": 3069
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5567073225975037,
+      "learning_rate": 0.0005962841902152796,
+      "loss": 3.5997,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5587060451507568,
+      "learning_rate": 0.0005962817741117316,
+      "loss": 3.7797,
+      "step": 3071
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5583739876747131,
+      "learning_rate": 0.0005962793572278342,
+      "loss": 3.768,
+      "step": 3072
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6153991222381592,
+      "learning_rate": 0.0005962769395635942,
+      "loss": 3.6803,
+      "step": 3073
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6166850328445435,
+      "learning_rate": 0.0005962745211190176,
+      "loss": 3.489,
+      "step": 3074
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6104187369346619,
+      "learning_rate": 0.000596272101894111,
+      "loss": 3.4945,
+      "step": 3075
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6122152805328369,
+      "learning_rate": 0.0005962696818888807,
+      "loss": 3.7251,
+      "step": 3076
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5618709921836853,
+      "learning_rate": 0.0005962672611033331,
+      "loss": 3.6053,
+      "step": 3077
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5944236516952515,
+      "learning_rate": 0.0005962648395374745,
+      "loss": 3.4591,
+      "step": 3078
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6525372266769409,
+      "learning_rate": 0.0005962624171913113,
+      "loss": 3.4751,
+      "step": 3079
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.64317387342453,
+      "learning_rate": 0.00059625999406485,
+      "loss": 3.6205,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6131284236907959,
+      "learning_rate": 0.0005962575701580968,
+      "loss": 3.6735,
+      "step": 3081
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5894636511802673,
+      "learning_rate": 0.0005962551454710583,
+      "loss": 3.8265,
+      "step": 3082
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6023474335670471,
+      "learning_rate": 0.0005962527200037407,
+      "loss": 3.573,
+      "step": 3083
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5374550223350525,
+      "learning_rate": 0.0005962502937561505,
+      "loss": 3.467,
+      "step": 3084
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5970792174339294,
+      "learning_rate": 0.0005962478667282939,
+      "loss": 3.6305,
+      "step": 3085
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6187454462051392,
+      "learning_rate": 0.0005962454389201776,
+      "loss": 3.5442,
+      "step": 3086
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5832275152206421,
+      "learning_rate": 0.0005962430103318077,
+      "loss": 3.5218,
+      "step": 3087
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6036730408668518,
+      "learning_rate": 0.0005962405809631908,
+      "loss": 3.3699,
+      "step": 3088
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.569485068321228,
+      "learning_rate": 0.0005962381508143332,
+      "loss": 3.6637,
+      "step": 3089
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.587161660194397,
+      "learning_rate": 0.0005962357198852412,
+      "loss": 3.8073,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5674685835838318,
+      "learning_rate": 0.0005962332881759215,
+      "loss": 3.5457,
+      "step": 3091
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5693333745002747,
+      "learning_rate": 0.0005962308556863803,
+      "loss": 3.6018,
+      "step": 3092
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6043726801872253,
+      "learning_rate": 0.000596228422416624,
+      "loss": 3.5762,
+      "step": 3093
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5639926791191101,
+      "learning_rate": 0.0005962259883666591,
+      "loss": 3.8223,
+      "step": 3094
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6272081136703491,
+      "learning_rate": 0.0005962235535364918,
+      "loss": 3.6311,
+      "step": 3095
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5490891933441162,
+      "learning_rate": 0.0005962211179261288,
+      "loss": 3.455,
+      "step": 3096
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6012369990348816,
+      "learning_rate": 0.0005962186815355765,
+      "loss": 3.4856,
+      "step": 3097
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5988452434539795,
+      "learning_rate": 0.000596216244364841,
+      "loss": 3.5496,
+      "step": 3098
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6350785493850708,
+      "learning_rate": 0.000596213806413929,
+      "loss": 3.5515,
+      "step": 3099
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6114810705184937,
+      "learning_rate": 0.0005962113676828468,
+      "loss": 3.4899,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5787695646286011,
+      "learning_rate": 0.000596208928171601,
+      "loss": 3.5458,
+      "step": 3101
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5776200890541077,
+      "learning_rate": 0.0005962064878801977,
+      "loss": 3.4334,
+      "step": 3102
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5516855716705322,
+      "learning_rate": 0.0005962040468086436,
+      "loss": 3.4205,
+      "step": 3103
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5812780857086182,
+      "learning_rate": 0.0005962016049569451,
+      "loss": 3.7321,
+      "step": 3104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5914744138717651,
+      "learning_rate": 0.0005961991623251085,
+      "loss": 3.6045,
+      "step": 3105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5617961287498474,
+      "learning_rate": 0.0005961967189131403,
+      "loss": 3.6301,
+      "step": 3106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6185020208358765,
+      "learning_rate": 0.0005961942747210469,
+      "loss": 3.5499,
+      "step": 3107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.609798014163971,
+      "learning_rate": 0.0005961918297488349,
+      "loss": 3.5542,
+      "step": 3108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5684676766395569,
+      "learning_rate": 0.0005961893839965106,
+      "loss": 3.5955,
+      "step": 3109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.584693193435669,
+      "learning_rate": 0.0005961869374640804,
+      "loss": 3.3336,
+      "step": 3110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6073812246322632,
+      "learning_rate": 0.0005961844901515509,
+      "loss": 3.669,
+      "step": 3111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5767044425010681,
+      "learning_rate": 0.0005961820420589285,
+      "loss": 3.63,
+      "step": 3112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5799092054367065,
+      "learning_rate": 0.0005961795931862194,
+      "loss": 3.4574,
+      "step": 3113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.545676052570343,
+      "learning_rate": 0.0005961771435334304,
+      "loss": 3.5919,
+      "step": 3114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.582565188407898,
+      "learning_rate": 0.0005961746931005678,
+      "loss": 3.664,
+      "step": 3115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5571921467781067,
+      "learning_rate": 0.000596172241887638,
+      "loss": 3.5959,
+      "step": 3116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5776621699333191,
+      "learning_rate": 0.0005961697898946475,
+      "loss": 3.5268,
+      "step": 3117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6262937188148499,
+      "learning_rate": 0.0005961673371216029,
+      "loss": 3.4795,
+      "step": 3118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6134164929389954,
+      "learning_rate": 0.0005961648835685105,
+      "loss": 3.782,
+      "step": 3119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5699538588523865,
+      "learning_rate": 0.0005961624292353766,
+      "loss": 3.8283,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5842909216880798,
+      "learning_rate": 0.0005961599741222081,
+      "loss": 3.4827,
+      "step": 3121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5857353806495667,
+      "learning_rate": 0.0005961575182290111,
+      "loss": 3.5638,
+      "step": 3122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5917581915855408,
+      "learning_rate": 0.0005961550615557923,
+      "loss": 3.5723,
+      "step": 3123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546079158782959,
+      "learning_rate": 0.000596152604102558,
+      "loss": 3.6738,
+      "step": 3124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5460621118545532,
+      "learning_rate": 0.0005961501458693147,
+      "loss": 3.6153,
+      "step": 3125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5750552415847778,
+      "learning_rate": 0.000596147686856069,
+      "loss": 3.5495,
+      "step": 3126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.55301833152771,
+      "learning_rate": 0.0005961452270628272,
+      "loss": 3.5873,
+      "step": 3127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5648224949836731,
+      "learning_rate": 0.0005961427664895959,
+      "loss": 3.6474,
+      "step": 3128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.581761360168457,
+      "learning_rate": 0.0005961403051363814,
+      "loss": 3.477,
+      "step": 3129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5733395218849182,
+      "learning_rate": 0.0005961378430031905,
+      "loss": 3.7321,
+      "step": 3130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5752854347229004,
+      "learning_rate": 0.0005961353800900296,
+      "loss": 3.7605,
+      "step": 3131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5976043343544006,
+      "learning_rate": 0.000596132916396905,
+      "loss": 3.7176,
+      "step": 3132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6109990477561951,
+      "learning_rate": 0.0005961304519238232,
+      "loss": 3.9394,
+      "step": 3133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6013369560241699,
+      "learning_rate": 0.0005961279866707909,
+      "loss": 3.4758,
+      "step": 3134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5654555559158325,
+      "learning_rate": 0.0005961255206378144,
+      "loss": 3.6548,
+      "step": 3135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5910729765892029,
+      "learning_rate": 0.0005961230538249003,
+      "loss": 3.5525,
+      "step": 3136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6182341575622559,
+      "learning_rate": 0.0005961205862320551,
+      "loss": 3.5527,
+      "step": 3137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5825470089912415,
+      "learning_rate": 0.0005961181178592852,
+      "loss": 3.6325,
+      "step": 3138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6601877808570862,
+      "learning_rate": 0.0005961156487065972,
+      "loss": 3.3948,
+      "step": 3139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5642703771591187,
+      "learning_rate": 0.0005961131787739976,
+      "loss": 3.5232,
+      "step": 3140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5936177372932434,
+      "learning_rate": 0.0005961107080614929,
+      "loss": 3.6665,
+      "step": 3141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5690498948097229,
+      "learning_rate": 0.0005961082365690894,
+      "loss": 3.4099,
+      "step": 3142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6080211400985718,
+      "learning_rate": 0.0005961057642967939,
+      "loss": 3.402,
+      "step": 3143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5955443382263184,
+      "learning_rate": 0.0005961032912446128,
+      "loss": 3.4212,
+      "step": 3144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5814092755317688,
+      "learning_rate": 0.0005961008174125527,
+      "loss": 3.3803,
+      "step": 3145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5517794489860535,
+      "learning_rate": 0.00059609834280062,
+      "loss": 3.5878,
+      "step": 3146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5653300881385803,
+      "learning_rate": 0.0005960958674088211,
+      "loss": 3.535,
+      "step": 3147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5799789428710938,
+      "learning_rate": 0.0005960933912371629,
+      "loss": 3.6731,
+      "step": 3148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5752755403518677,
+      "learning_rate": 0.0005960909142856516,
+      "loss": 3.6781,
+      "step": 3149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6208489537239075,
+      "learning_rate": 0.0005960884365542937,
+      "loss": 3.3908,
+      "step": 3150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6199827790260315,
+      "learning_rate": 0.000596085958043096,
+      "loss": 3.7275,
+      "step": 3151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6208821535110474,
+      "learning_rate": 0.0005960834787520648,
+      "loss": 3.4982,
+      "step": 3152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5410616993904114,
+      "learning_rate": 0.0005960809986812067,
+      "loss": 3.5705,
+      "step": 3153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5999816060066223,
+      "learning_rate": 0.0005960785178305283,
+      "loss": 3.4113,
+      "step": 3154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6006971001625061,
+      "learning_rate": 0.000596076036200036,
+      "loss": 3.7676,
+      "step": 3155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5484111309051514,
+      "learning_rate": 0.0005960735537897364,
+      "loss": 3.5441,
+      "step": 3156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.581990659236908,
+      "learning_rate": 0.000596071070599636,
+      "loss": 3.3492,
+      "step": 3157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6446429491043091,
+      "learning_rate": 0.0005960685866297415,
+      "loss": 3.6986,
+      "step": 3158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.564018726348877,
+      "learning_rate": 0.0005960661018800592,
+      "loss": 3.4903,
+      "step": 3159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5862210392951965,
+      "learning_rate": 0.0005960636163505958,
+      "loss": 3.5107,
+      "step": 3160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5642343163490295,
+      "learning_rate": 0.0005960611300413578,
+      "loss": 3.6318,
+      "step": 3161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5777455568313599,
+      "learning_rate": 0.0005960586429523518,
+      "loss": 3.417,
+      "step": 3162
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.586602509021759,
+      "learning_rate": 0.0005960561550835843,
+      "loss": 3.5432,
+      "step": 3163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5912312269210815,
+      "learning_rate": 0.0005960536664350619,
+      "loss": 3.6441,
+      "step": 3164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6185685396194458,
+      "learning_rate": 0.0005960511770067911,
+      "loss": 3.5537,
+      "step": 3165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6227981448173523,
+      "learning_rate": 0.0005960486867987784,
+      "loss": 3.4673,
+      "step": 3166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6355583071708679,
+      "learning_rate": 0.0005960461958110305,
+      "loss": 3.5603,
+      "step": 3167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6054497957229614,
+      "learning_rate": 0.0005960437040435539,
+      "loss": 3.5745,
+      "step": 3168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61248379945755,
+      "learning_rate": 0.0005960412114963552,
+      "loss": 3.5392,
+      "step": 3169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.600622832775116,
+      "learning_rate": 0.0005960387181694408,
+      "loss": 3.629,
+      "step": 3170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6262516379356384,
+      "learning_rate": 0.0005960362240628175,
+      "loss": 3.3767,
+      "step": 3171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5428594350814819,
+      "learning_rate": 0.0005960337291764918,
+      "loss": 3.6812,
+      "step": 3172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5798599123954773,
+      "learning_rate": 0.0005960312335104701,
+      "loss": 3.6121,
+      "step": 3173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.600669801235199,
+      "learning_rate": 0.0005960287370647591,
+      "loss": 3.5828,
+      "step": 3174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5855419635772705,
+      "learning_rate": 0.0005960262398393654,
+      "loss": 3.7861,
+      "step": 3175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5565866827964783,
+      "learning_rate": 0.0005960237418342957,
+      "loss": 3.7263,
+      "step": 3176
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5831587910652161,
+      "learning_rate": 0.0005960212430495564,
+      "loss": 3.7619,
+      "step": 3177
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6254531741142273,
+      "learning_rate": 0.000596018743485154,
+      "loss": 3.5224,
+      "step": 3178
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5760573744773865,
+      "learning_rate": 0.0005960162431410953,
+      "loss": 3.4772,
+      "step": 3179
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5323291420936584,
+      "learning_rate": 0.0005960137420173866,
+      "loss": 3.4991,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5438007116317749,
+      "learning_rate": 0.0005960112401140348,
+      "loss": 3.6223,
+      "step": 3181
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5264496207237244,
+      "learning_rate": 0.0005960087374310464,
+      "loss": 3.7063,
+      "step": 3182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5956478118896484,
+      "learning_rate": 0.0005960062339684279,
+      "loss": 3.564,
+      "step": 3183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6297171711921692,
+      "learning_rate": 0.000596003729726186,
+      "loss": 3.5353,
+      "step": 3184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5785334706306458,
+      "learning_rate": 0.0005960012247043272,
+      "loss": 3.7652,
+      "step": 3185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6340588331222534,
+      "learning_rate": 0.0005959987189028582,
+      "loss": 3.6217,
+      "step": 3186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5726503133773804,
+      "learning_rate": 0.0005959962123217855,
+      "loss": 3.585,
+      "step": 3187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5697149038314819,
+      "learning_rate": 0.0005959937049611156,
+      "loss": 3.4602,
+      "step": 3188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5911883115768433,
+      "learning_rate": 0.0005959911968208554,
+      "loss": 3.4454,
+      "step": 3189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.579783022403717,
+      "learning_rate": 0.0005959886879010113,
+      "loss": 3.6873,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5348889827728271,
+      "learning_rate": 0.00059598617820159,
+      "loss": 3.8555,
+      "step": 3191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5788812637329102,
+      "learning_rate": 0.000595983667722598,
+      "loss": 3.634,
+      "step": 3192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5206723809242249,
+      "learning_rate": 0.0005959811564640421,
+      "loss": 3.3991,
+      "step": 3193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5956487059593201,
+      "learning_rate": 0.0005959786444259287,
+      "loss": 3.4554,
+      "step": 3194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5612884759902954,
+      "learning_rate": 0.0005959761316082646,
+      "loss": 3.6328,
+      "step": 3195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5556085705757141,
+      "learning_rate": 0.0005959736180110563,
+      "loss": 3.6107,
+      "step": 3196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5669443607330322,
+      "learning_rate": 0.0005959711036343104,
+      "loss": 3.5104,
+      "step": 3197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5675809383392334,
+      "learning_rate": 0.0005959685884780337,
+      "loss": 3.5678,
+      "step": 3198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.605646014213562,
+      "learning_rate": 0.0005959660725422325,
+      "loss": 3.7298,
+      "step": 3199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5824044942855835,
+      "learning_rate": 0.0005959635558269137,
+      "loss": 3.416,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5763425230979919,
+      "learning_rate": 0.0005959610383320839,
+      "loss": 3.6143,
+      "step": 3201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5799658894538879,
+      "learning_rate": 0.0005959585200577497,
+      "loss": 3.5673,
+      "step": 3202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5830967426300049,
+      "learning_rate": 0.0005959560010039177,
+      "loss": 3.5561,
+      "step": 3203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5466955304145813,
+      "learning_rate": 0.0005959534811705946,
+      "loss": 3.6316,
+      "step": 3204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6080002784729004,
+      "learning_rate": 0.0005959509605577869,
+      "loss": 3.6567,
+      "step": 3205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5686078071594238,
+      "learning_rate": 0.0005959484391655013,
+      "loss": 3.5206,
+      "step": 3206
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6283111572265625,
+      "learning_rate": 0.0005959459169937445,
+      "loss": 3.669,
+      "step": 3207
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.551060140132904,
+      "learning_rate": 0.0005959433940425231,
+      "loss": 3.5432,
+      "step": 3208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5849263668060303,
+      "learning_rate": 0.0005959408703118437,
+      "loss": 3.463,
+      "step": 3209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5780369639396667,
+      "learning_rate": 0.0005959383458017133,
+      "loss": 3.4947,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6128699779510498,
+      "learning_rate": 0.000595935820512138,
+      "loss": 3.3344,
+      "step": 3211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5325568318367004,
+      "learning_rate": 0.0005959332944431246,
+      "loss": 3.5697,
+      "step": 3212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6106698513031006,
+      "learning_rate": 0.00059593076759468,
+      "loss": 3.4766,
+      "step": 3213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6008756160736084,
+      "learning_rate": 0.0005959282399668108,
+      "loss": 3.5924,
+      "step": 3214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5620996952056885,
+      "learning_rate": 0.0005959257115595235,
+      "loss": 3.548,
+      "step": 3215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6242149472236633,
+      "learning_rate": 0.0005959231823728247,
+      "loss": 3.6272,
+      "step": 3216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6121542453765869,
+      "learning_rate": 0.0005959206524067213,
+      "loss": 3.6008,
+      "step": 3217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6612833142280579,
+      "learning_rate": 0.0005959181216612199,
+      "loss": 3.3679,
+      "step": 3218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6306580305099487,
+      "learning_rate": 0.000595915590136327,
+      "loss": 3.6513,
+      "step": 3219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.615294337272644,
+      "learning_rate": 0.0005959130578320495,
+      "loss": 3.4633,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5863372683525085,
+      "learning_rate": 0.0005959105247483939,
+      "loss": 3.6524,
+      "step": 3221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.702987551689148,
+      "learning_rate": 0.0005959079908853669,
+      "loss": 3.7698,
+      "step": 3222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6178707480430603,
+      "learning_rate": 0.0005959054562429753,
+      "loss": 3.5101,
+      "step": 3223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5774521231651306,
+      "learning_rate": 0.0005959029208212255,
+      "loss": 3.5955,
+      "step": 3224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.587511420249939,
+      "learning_rate": 0.0005959003846201245,
+      "loss": 3.5713,
+      "step": 3225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5820519924163818,
+      "learning_rate": 0.0005958978476396788,
+      "loss": 3.1941,
+      "step": 3226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5678663849830627,
+      "learning_rate": 0.0005958953098798952,
+      "loss": 3.6527,
+      "step": 3227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.547321617603302,
+      "learning_rate": 0.0005958927713407801,
+      "loss": 3.5103,
+      "step": 3228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5541772246360779,
+      "learning_rate": 0.0005958902320223405,
+      "loss": 3.4871,
+      "step": 3229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5748250484466553,
+      "learning_rate": 0.000595887691924583,
+      "loss": 3.4653,
+      "step": 3230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6465830206871033,
+      "learning_rate": 0.0005958851510475142,
+      "loss": 3.5919,
+      "step": 3231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5622817277908325,
+      "learning_rate": 0.000595882609391141,
+      "loss": 3.6361,
+      "step": 3232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5414732098579407,
+      "learning_rate": 0.0005958800669554698,
+      "loss": 3.797,
+      "step": 3233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5994971990585327,
+      "learning_rate": 0.0005958775237405074,
+      "loss": 3.5292,
+      "step": 3234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6444474458694458,
+      "learning_rate": 0.0005958749797462607,
+      "loss": 3.5301,
+      "step": 3235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5784667730331421,
+      "learning_rate": 0.0005958724349727362,
+      "loss": 3.4053,
+      "step": 3236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5597053170204163,
+      "learning_rate": 0.0005958698894199406,
+      "loss": 3.5452,
+      "step": 3237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5689312815666199,
+      "learning_rate": 0.0005958673430878807,
+      "loss": 3.6081,
+      "step": 3238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5763968825340271,
+      "learning_rate": 0.0005958647959765632,
+      "loss": 3.3315,
+      "step": 3239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5735750794410706,
+      "learning_rate": 0.0005958622480859947,
+      "loss": 3.5241,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5359008312225342,
+      "learning_rate": 0.0005958596994161819,
+      "loss": 3.7146,
+      "step": 3241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6043967008590698,
+      "learning_rate": 0.0005958571499671318,
+      "loss": 3.6643,
+      "step": 3242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.588692843914032,
+      "learning_rate": 0.0005958545997388506,
+      "loss": 3.6953,
+      "step": 3243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5518408417701721,
+      "learning_rate": 0.0005958520487313456,
+      "loss": 3.49,
+      "step": 3244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6246479153633118,
+      "learning_rate": 0.0005958494969446231,
+      "loss": 3.5562,
+      "step": 3245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6223381757736206,
+      "learning_rate": 0.0005958469443786901,
+      "loss": 3.5703,
+      "step": 3246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5942324995994568,
+      "learning_rate": 0.0005958443910335531,
+      "loss": 3.4527,
+      "step": 3247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5596261620521545,
+      "learning_rate": 0.0005958418369092189,
+      "loss": 3.63,
+      "step": 3248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5686338543891907,
+      "learning_rate": 0.0005958392820056942,
+      "loss": 3.5783,
+      "step": 3249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5892527103424072,
+      "learning_rate": 0.0005958367263229859,
+      "loss": 3.5206,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5530284643173218,
+      "learning_rate": 0.0005958341698611004,
+      "loss": 3.7264,
+      "step": 3251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5221811532974243,
+      "learning_rate": 0.0005958316126200448,
+      "loss": 3.5932,
+      "step": 3252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5697766542434692,
+      "learning_rate": 0.0005958290545998255,
+      "loss": 3.6294,
+      "step": 3253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5952369570732117,
+      "learning_rate": 0.0005958264958004496,
+      "loss": 3.5551,
+      "step": 3254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5782142281532288,
+      "learning_rate": 0.0005958239362219235,
+      "loss": 3.4631,
+      "step": 3255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5828744173049927,
+      "learning_rate": 0.0005958213758642542,
+      "loss": 3.5134,
+      "step": 3256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5539057850837708,
+      "learning_rate": 0.0005958188147274483,
+      "loss": 3.3305,
+      "step": 3257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6173965930938721,
+      "learning_rate": 0.0005958162528115125,
+      "loss": 3.7218,
+      "step": 3258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5583602786064148,
+      "learning_rate": 0.0005958136901164537,
+      "loss": 3.6743,
+      "step": 3259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5693215131759644,
+      "learning_rate": 0.0005958111266422785,
+      "loss": 3.5812,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5246009826660156,
+      "learning_rate": 0.0005958085623889937,
+      "loss": 3.6991,
+      "step": 3261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5803536772727966,
+      "learning_rate": 0.0005958059973566062,
+      "loss": 3.7107,
+      "step": 3262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5437319874763489,
+      "learning_rate": 0.0005958034315451224,
+      "loss": 3.7399,
+      "step": 3263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5739423036575317,
+      "learning_rate": 0.0005958008649545496,
+      "loss": 3.522,
+      "step": 3264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6183177828788757,
+      "learning_rate": 0.000595798297584894,
+      "loss": 3.4463,
+      "step": 3265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.521632194519043,
+      "learning_rate": 0.0005957957294361628,
+      "loss": 3.57,
+      "step": 3266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5398718118667603,
+      "learning_rate": 0.0005957931605083624,
+      "loss": 3.5119,
+      "step": 3267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.556581437587738,
+      "learning_rate": 0.0005957905908014999,
+      "loss": 3.6029,
+      "step": 3268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5454385876655579,
+      "learning_rate": 0.0005957880203155818,
+      "loss": 3.4989,
+      "step": 3269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625463128089905,
+      "learning_rate": 0.000595785449050615,
+      "loss": 3.6914,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.538007915019989,
+      "learning_rate": 0.0005957828770066063,
+      "loss": 3.5866,
+      "step": 3271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5634467601776123,
+      "learning_rate": 0.0005957803041835623,
+      "loss": 3.8559,
+      "step": 3272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5901991128921509,
+      "learning_rate": 0.0005957777305814901,
+      "loss": 3.4086,
+      "step": 3273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.582669734954834,
+      "learning_rate": 0.0005957751562003961,
+      "loss": 3.506,
+      "step": 3274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5981149673461914,
+      "learning_rate": 0.0005957725810402874,
+      "loss": 3.717,
+      "step": 3275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6032350063323975,
+      "learning_rate": 0.0005957700051011706,
+      "loss": 3.6699,
+      "step": 3276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5754993557929993,
+      "learning_rate": 0.0005957674283830525,
+      "loss": 3.5732,
+      "step": 3277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6205776929855347,
+      "learning_rate": 0.00059576485088594,
+      "loss": 3.6692,
+      "step": 3278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5795102119445801,
+      "learning_rate": 0.0005957622726098397,
+      "loss": 3.574,
+      "step": 3279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.547251284122467,
+      "learning_rate": 0.0005957596935547586,
+      "loss": 3.7679,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5527950525283813,
+      "learning_rate": 0.0005957571137207032,
+      "loss": 3.4746,
+      "step": 3281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5651281476020813,
+      "learning_rate": 0.0005957545331076806,
+      "loss": 3.5383,
+      "step": 3282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.599833071231842,
+      "learning_rate": 0.0005957519517156975,
+      "loss": 3.5415,
+      "step": 3283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6276541352272034,
+      "learning_rate": 0.0005957493695447607,
+      "loss": 3.5067,
+      "step": 3284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5757126212120056,
+      "learning_rate": 0.0005957467865948769,
+      "loss": 3.6089,
+      "step": 3285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5513081550598145,
+      "learning_rate": 0.000595744202866053,
+      "loss": 3.6373,
+      "step": 3286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5885028839111328,
+      "learning_rate": 0.0005957416183582958,
+      "loss": 3.7294,
+      "step": 3287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5601800084114075,
+      "learning_rate": 0.0005957390330716121,
+      "loss": 3.5964,
+      "step": 3288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6039920449256897,
+      "learning_rate": 0.0005957364470060087,
+      "loss": 3.3979,
+      "step": 3289
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5872266888618469,
+      "learning_rate": 0.0005957338601614924,
+      "loss": 3.6418,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5281251072883606,
+      "learning_rate": 0.0005957312725380701,
+      "loss": 3.5281,
+      "step": 3291
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5793739557266235,
+      "learning_rate": 0.0005957286841357485,
+      "loss": 3.6252,
+      "step": 3292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6035564541816711,
+      "learning_rate": 0.0005957260949545345,
+      "loss": 3.4221,
+      "step": 3293
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5742641687393188,
+      "learning_rate": 0.0005957235049944347,
+      "loss": 3.7449,
+      "step": 3294
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5799552202224731,
+      "learning_rate": 0.0005957209142554563,
+      "loss": 3.7006,
+      "step": 3295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5929350256919861,
+      "learning_rate": 0.0005957183227376058,
+      "loss": 3.5188,
+      "step": 3296
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5783431529998779,
+      "learning_rate": 0.0005957157304408902,
+      "loss": 3.4988,
+      "step": 3297
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6081163883209229,
+      "learning_rate": 0.0005957131373653163,
+      "loss": 3.7293,
+      "step": 3298
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5771412253379822,
+      "learning_rate": 0.0005957105435108909,
+      "loss": 3.5091,
+      "step": 3299
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5864212512969971,
+      "learning_rate": 0.0005957079488776208,
+      "loss": 3.6186,
+      "step": 3300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5488468408584595,
+      "learning_rate": 0.0005957053534655129,
+      "loss": 3.617,
+      "step": 3301
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5773611068725586,
+      "learning_rate": 0.000595702757274574,
+      "loss": 3.6083,
+      "step": 3302
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5988139510154724,
+      "learning_rate": 0.0005957001603048109,
+      "loss": 3.4651,
+      "step": 3303
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5818193554878235,
+      "learning_rate": 0.0005956975625562305,
+      "loss": 3.3751,
+      "step": 3304
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5249636173248291,
+      "learning_rate": 0.0005956949640288397,
+      "loss": 3.6246,
+      "step": 3305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5715710520744324,
+      "learning_rate": 0.0005956923647226453,
+      "loss": 3.5151,
+      "step": 3306
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5765474438667297,
+      "learning_rate": 0.000595689764637654,
+      "loss": 3.4035,
+      "step": 3307
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6147971153259277,
+      "learning_rate": 0.0005956871637738728,
+      "loss": 3.6643,
+      "step": 3308
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6140459775924683,
+      "learning_rate": 0.0005956845621313086,
+      "loss": 3.3888,
+      "step": 3309
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6852372884750366,
+      "learning_rate": 0.0005956819597099681,
+      "loss": 3.9077,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5585691928863525,
+      "learning_rate": 0.0005956793565098582,
+      "loss": 3.4769,
+      "step": 3311
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5834951996803284,
+      "learning_rate": 0.0005956767525309858,
+      "loss": 3.588,
+      "step": 3312
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6259799599647522,
+      "learning_rate": 0.0005956741477733578,
+      "loss": 3.5964,
+      "step": 3313
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5601211190223694,
+      "learning_rate": 0.0005956715422369809,
+      "loss": 3.5845,
+      "step": 3314
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5854201316833496,
+      "learning_rate": 0.0005956689359218621,
+      "loss": 3.4226,
+      "step": 3315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5881748199462891,
+      "learning_rate": 0.0005956663288280082,
+      "loss": 3.4351,
+      "step": 3316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6546245813369751,
+      "learning_rate": 0.0005956637209554262,
+      "loss": 3.44,
+      "step": 3317
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5833501815795898,
+      "learning_rate": 0.0005956611123041229,
+      "loss": 3.6924,
+      "step": 3318
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5682247877120972,
+      "learning_rate": 0.0005956585028741049,
+      "loss": 3.7348,
+      "step": 3319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6428276896476746,
+      "learning_rate": 0.0005956558926653794,
+      "loss": 3.3296,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5545470714569092,
+      "learning_rate": 0.0005956532816779532,
+      "loss": 3.6842,
+      "step": 3321
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5805429220199585,
+      "learning_rate": 0.0005956506699118331,
+      "loss": 3.3973,
+      "step": 3322
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5832202434539795,
+      "learning_rate": 0.0005956480573670261,
+      "loss": 3.699,
+      "step": 3323
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5636213421821594,
+      "learning_rate": 0.000595645444043539,
+      "loss": 3.647,
+      "step": 3324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.599785327911377,
+      "learning_rate": 0.0005956428299413788,
+      "loss": 3.5346,
+      "step": 3325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5496143698692322,
+      "learning_rate": 0.0005956402150605522,
+      "loss": 3.6491,
+      "step": 3326
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5501871109008789,
+      "learning_rate": 0.0005956375994010661,
+      "loss": 3.5646,
+      "step": 3327
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6170458793640137,
+      "learning_rate": 0.0005956349829629276,
+      "loss": 3.5171,
+      "step": 3328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.554207980632782,
+      "learning_rate": 0.0005956323657461434,
+      "loss": 3.4183,
+      "step": 3329
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6159272193908691,
+      "learning_rate": 0.0005956297477507204,
+      "loss": 3.5173,
+      "step": 3330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5841159224510193,
+      "learning_rate": 0.0005956271289766654,
+      "loss": 3.5016,
+      "step": 3331
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5938801169395447,
+      "learning_rate": 0.0005956245094239857,
+      "loss": 3.557,
+      "step": 3332
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5655121207237244,
+      "learning_rate": 0.0005956218890926878,
+      "loss": 3.5488,
+      "step": 3333
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.581397294998169,
+      "learning_rate": 0.0005956192679827788,
+      "loss": 3.6558,
+      "step": 3334
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5534145832061768,
+      "learning_rate": 0.0005956166460942654,
+      "loss": 3.6431,
+      "step": 3335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6180419325828552,
+      "learning_rate": 0.0005956140234271549,
+      "loss": 3.5917,
+      "step": 3336
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5950748324394226,
+      "learning_rate": 0.0005956113999814537,
+      "loss": 3.5197,
+      "step": 3337
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5673226118087769,
+      "learning_rate": 0.0005956087757571691,
+      "loss": 3.465,
+      "step": 3338
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5869221091270447,
+      "learning_rate": 0.0005956061507543078,
+      "loss": 3.5244,
+      "step": 3339
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6073868870735168,
+      "learning_rate": 0.0005956035249728769,
+      "loss": 3.4927,
+      "step": 3340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.547737181186676,
+      "learning_rate": 0.0005956008984128831,
+      "loss": 3.4597,
+      "step": 3341
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5933374166488647,
+      "learning_rate": 0.0005955982710743336,
+      "loss": 3.5967,
+      "step": 3342
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.564926266670227,
+      "learning_rate": 0.000595595642957235,
+      "loss": 3.5223,
+      "step": 3343
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.625313937664032,
+      "learning_rate": 0.0005955930140615943,
+      "loss": 3.4885,
+      "step": 3344
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6144617199897766,
+      "learning_rate": 0.0005955903843874185,
+      "loss": 3.3671,
+      "step": 3345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6368928551673889,
+      "learning_rate": 0.0005955877539347146,
+      "loss": 3.5404,
+      "step": 3346
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5616483688354492,
+      "learning_rate": 0.0005955851227034894,
+      "loss": 3.5257,
+      "step": 3347
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5560274720191956,
+      "learning_rate": 0.00059558249069375,
+      "loss": 3.5856,
+      "step": 3348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6093305945396423,
+      "learning_rate": 0.0005955798579055031,
+      "loss": 3.6487,
+      "step": 3349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5649324655532837,
+      "learning_rate": 0.0005955772243387556,
+      "loss": 3.5062,
+      "step": 3350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5727810859680176,
+      "learning_rate": 0.0005955745899935147,
+      "loss": 3.2842,
+      "step": 3351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6266840100288391,
+      "learning_rate": 0.0005955719548697873,
+      "loss": 3.7879,
+      "step": 3352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6102510690689087,
+      "learning_rate": 0.0005955693189675802,
+      "loss": 3.71,
+      "step": 3353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5587088465690613,
+      "learning_rate": 0.0005955666822869004,
+      "loss": 3.4238,
+      "step": 3354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.617514431476593,
+      "learning_rate": 0.0005955640448277548,
+      "loss": 3.6391,
+      "step": 3355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5572774410247803,
+      "learning_rate": 0.0005955614065901504,
+      "loss": 3.6646,
+      "step": 3356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.598943829536438,
+      "learning_rate": 0.0005955587675740942,
+      "loss": 3.5632,
+      "step": 3357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6389573812484741,
+      "learning_rate": 0.000595556127779593,
+      "loss": 3.4648,
+      "step": 3358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5725786685943604,
+      "learning_rate": 0.0005955534872066538,
+      "loss": 3.387,
+      "step": 3359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625653862953186,
+      "learning_rate": 0.0005955508458552837,
+      "loss": 3.6279,
+      "step": 3360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5734360814094543,
+      "learning_rate": 0.0005955482037254895,
+      "loss": 3.5787,
+      "step": 3361
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5375578999519348,
+      "learning_rate": 0.0005955455608172783,
+      "loss": 3.7722,
+      "step": 3362
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6041198968887329,
+      "learning_rate": 0.0005955429171306569,
+      "loss": 3.5613,
+      "step": 3363
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5527943968772888,
+      "learning_rate": 0.0005955402726656323,
+      "loss": 3.4843,
+      "step": 3364
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.528593897819519,
+      "learning_rate": 0.0005955376274222116,
+      "loss": 3.5478,
+      "step": 3365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5430445671081543,
+      "learning_rate": 0.0005955349814004016,
+      "loss": 3.5002,
+      "step": 3366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5947743654251099,
+      "learning_rate": 0.0005955323346002094,
+      "loss": 3.454,
+      "step": 3367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5306910872459412,
+      "learning_rate": 0.0005955296870216419,
+      "loss": 3.5488,
+      "step": 3368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5531356930732727,
+      "learning_rate": 0.000595527038664706,
+      "loss": 3.7194,
+      "step": 3369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.601188600063324,
+      "learning_rate": 0.0005955243895294087,
+      "loss": 3.5034,
+      "step": 3370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5490877628326416,
+      "learning_rate": 0.0005955217396157571,
+      "loss": 3.8795,
+      "step": 3371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5771283507347107,
+      "learning_rate": 0.0005955190889237581,
+      "loss": 3.66,
+      "step": 3372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5651893019676208,
+      "learning_rate": 0.0005955164374534187,
+      "loss": 3.3696,
+      "step": 3373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5814728140830994,
+      "learning_rate": 0.0005955137852047459,
+      "loss": 3.6069,
+      "step": 3374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5790256261825562,
+      "learning_rate": 0.0005955111321777467,
+      "loss": 3.4807,
+      "step": 3375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5738055109977722,
+      "learning_rate": 0.0005955084783724281,
+      "loss": 3.4382,
+      "step": 3376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5678678154945374,
+      "learning_rate": 0.0005955058237887968,
+      "loss": 3.3348,
+      "step": 3377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5809141397476196,
+      "learning_rate": 0.0005955031684268603,
+      "loss": 3.5302,
+      "step": 3378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5668545961380005,
+      "learning_rate": 0.0005955005122866251,
+      "loss": 3.437,
+      "step": 3379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5487068891525269,
+      "learning_rate": 0.0005954978553680987,
+      "loss": 3.482,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5972830653190613,
+      "learning_rate": 0.0005954951976712876,
+      "loss": 3.5068,
+      "step": 3381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5785419940948486,
+      "learning_rate": 0.0005954925391961991,
+      "loss": 3.4964,
+      "step": 3382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5430408120155334,
+      "learning_rate": 0.00059548987994284,
+      "loss": 3.535,
+      "step": 3383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5556647181510925,
+      "learning_rate": 0.0005954872199112175,
+      "loss": 3.589,
+      "step": 3384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6139194369316101,
+      "learning_rate": 0.0005954845591013385,
+      "loss": 3.4409,
+      "step": 3385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5702038407325745,
+      "learning_rate": 0.0005954818975132102,
+      "loss": 3.5252,
+      "step": 3386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5292709469795227,
+      "learning_rate": 0.0005954792351468393,
+      "loss": 3.7446,
+      "step": 3387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5479516983032227,
+      "learning_rate": 0.000595476572002233,
+      "loss": 3.789,
+      "step": 3388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5670396089553833,
+      "learning_rate": 0.0005954739080793983,
+      "loss": 3.5434,
+      "step": 3389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5650784969329834,
+      "learning_rate": 0.0005954712433783421,
+      "loss": 3.4521,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.555449366569519,
+      "learning_rate": 0.0005954685778990715,
+      "loss": 3.6411,
+      "step": 3391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5617675185203552,
+      "learning_rate": 0.0005954659116415936,
+      "loss": 3.5113,
+      "step": 3392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.545922040939331,
+      "learning_rate": 0.0005954632446059153,
+      "loss": 3.6585,
+      "step": 3393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.562699019908905,
+      "learning_rate": 0.0005954605767920437,
+      "loss": 3.5584,
+      "step": 3394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5851036310195923,
+      "learning_rate": 0.0005954579081999859,
+      "loss": 3.3386,
+      "step": 3395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6002635359764099,
+      "learning_rate": 0.0005954552388297488,
+      "loss": 3.6444,
+      "step": 3396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5689345002174377,
+      "learning_rate": 0.0005954525686813394,
+      "loss": 3.5515,
+      "step": 3397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5846789479255676,
+      "learning_rate": 0.0005954498977547648,
+      "loss": 3.5281,
+      "step": 3398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5559812188148499,
+      "learning_rate": 0.0005954472260500321,
+      "loss": 3.4974,
+      "step": 3399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6159769296646118,
+      "learning_rate": 0.0005954445535671482,
+      "loss": 3.4743,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6169503927230835,
+      "learning_rate": 0.0005954418803061201,
+      "loss": 3.3966,
+      "step": 3401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5606275796890259,
+      "learning_rate": 0.0005954392062669551,
+      "loss": 3.6091,
+      "step": 3402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5616119503974915,
+      "learning_rate": 0.0005954365314496602,
+      "loss": 3.3921,
+      "step": 3403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5786251425743103,
+      "learning_rate": 0.0005954338558542421,
+      "loss": 3.5837,
+      "step": 3404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.645964503288269,
+      "learning_rate": 0.0005954311794807082,
+      "loss": 3.5519,
+      "step": 3405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6374315619468689,
+      "learning_rate": 0.0005954285023290653,
+      "loss": 3.6006,
+      "step": 3406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6593900322914124,
+      "learning_rate": 0.0005954258243993207,
+      "loss": 3.7227,
+      "step": 3407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5792526602745056,
+      "learning_rate": 0.0005954231456914813,
+      "loss": 3.665,
+      "step": 3408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5797976851463318,
+      "learning_rate": 0.0005954204662055542,
+      "loss": 3.3409,
+      "step": 3409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5604798197746277,
+      "learning_rate": 0.0005954177859415465,
+      "loss": 3.416,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5253710150718689,
+      "learning_rate": 0.0005954151048994652,
+      "loss": 3.5158,
+      "step": 3411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5831363797187805,
+      "learning_rate": 0.0005954124230793173,
+      "loss": 3.4324,
+      "step": 3412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.592633843421936,
+      "learning_rate": 0.00059540974048111,
+      "loss": 3.6425,
+      "step": 3413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5312915444374084,
+      "learning_rate": 0.0005954070571048503,
+      "loss": 3.5972,
+      "step": 3414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5762085914611816,
+      "learning_rate": 0.0005954043729505452,
+      "loss": 3.6361,
+      "step": 3415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5579561591148376,
+      "learning_rate": 0.0005954016880182018,
+      "loss": 3.5233,
+      "step": 3416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5626066327095032,
+      "learning_rate": 0.0005953990023078273,
+      "loss": 3.7441,
+      "step": 3417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5891571044921875,
+      "learning_rate": 0.0005953963158194285,
+      "loss": 3.7897,
+      "step": 3418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5446343421936035,
+      "learning_rate": 0.0005953936285530129,
+      "loss": 3.5544,
+      "step": 3419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6051816940307617,
+      "learning_rate": 0.0005953909405085872,
+      "loss": 3.5061,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5892947316169739,
+      "learning_rate": 0.0005953882516861586,
+      "loss": 3.5149,
+      "step": 3421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6093165874481201,
+      "learning_rate": 0.0005953855620857342,
+      "loss": 3.8573,
+      "step": 3422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5854846239089966,
+      "learning_rate": 0.0005953828717073212,
+      "loss": 3.3395,
+      "step": 3423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5802544951438904,
+      "learning_rate": 0.0005953801805509264,
+      "loss": 3.4692,
+      "step": 3424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5809839367866516,
+      "learning_rate": 0.0005953774886165572,
+      "loss": 3.5708,
+      "step": 3425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.602876603603363,
+      "learning_rate": 0.0005953747959042204,
+      "loss": 3.4044,
+      "step": 3426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6141166090965271,
+      "learning_rate": 0.0005953721024139233,
+      "loss": 3.3439,
+      "step": 3427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5813528895378113,
+      "learning_rate": 0.000595369408145673,
+      "loss": 3.4252,
+      "step": 3428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5566039085388184,
+      "learning_rate": 0.0005953667130994763,
+      "loss": 3.4517,
+      "step": 3429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5614877939224243,
+      "learning_rate": 0.0005953640172753407,
+      "loss": 3.4067,
+      "step": 3430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5762594938278198,
+      "learning_rate": 0.0005953613206732731,
+      "loss": 3.4889,
+      "step": 3431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5990262031555176,
+      "learning_rate": 0.0005953586232932806,
+      "loss": 3.4092,
+      "step": 3432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5427011251449585,
+      "learning_rate": 0.0005953559251353703,
+      "loss": 3.6497,
+      "step": 3433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5713863968849182,
+      "learning_rate": 0.0005953532261995494,
+      "loss": 3.4972,
+      "step": 3434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6282929182052612,
+      "learning_rate": 0.0005953505264858249,
+      "loss": 3.4623,
+      "step": 3435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5442312359809875,
+      "learning_rate": 0.0005953478259942041,
+      "loss": 3.4691,
+      "step": 3436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5707518458366394,
+      "learning_rate": 0.0005953451247246937,
+      "loss": 3.7231,
+      "step": 3437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5848592519760132,
+      "learning_rate": 0.0005953424226773013,
+      "loss": 3.6997,
+      "step": 3438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5771193504333496,
+      "learning_rate": 0.0005953397198520337,
+      "loss": 3.7628,
+      "step": 3439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6253427863121033,
+      "learning_rate": 0.0005953370162488981,
+      "loss": 3.3619,
+      "step": 3440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6061219573020935,
+      "learning_rate": 0.0005953343118679018,
+      "loss": 3.6352,
+      "step": 3441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5454443693161011,
+      "learning_rate": 0.0005953316067090516,
+      "loss": 3.4859,
+      "step": 3442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.617382287979126,
+      "learning_rate": 0.0005953289007723548,
+      "loss": 3.5843,
+      "step": 3443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5367785096168518,
+      "learning_rate": 0.0005953261940578186,
+      "loss": 3.4825,
+      "step": 3444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5802174210548401,
+      "learning_rate": 0.00059532348656545,
+      "loss": 3.4618,
+      "step": 3445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5566971302032471,
+      "learning_rate": 0.000595320778295256,
+      "loss": 3.5407,
+      "step": 3446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5934160351753235,
+      "learning_rate": 0.0005953180692472441,
+      "loss": 3.5885,
+      "step": 3447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5748066902160645,
+      "learning_rate": 0.0005953153594214212,
+      "loss": 3.8451,
+      "step": 3448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6022830009460449,
+      "learning_rate": 0.0005953126488177946,
+      "loss": 3.5088,
+      "step": 3449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6487731337547302,
+      "learning_rate": 0.0005953099374363711,
+      "loss": 3.4576,
+      "step": 3450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6260390877723694,
+      "learning_rate": 0.0005953072252771581,
+      "loss": 3.4172,
+      "step": 3451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5463585257530212,
+      "learning_rate": 0.0005953045123401628,
+      "loss": 3.6589,
+      "step": 3452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6311070322990417,
+      "learning_rate": 0.0005953017986253922,
+      "loss": 3.6357,
+      "step": 3453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5336967706680298,
+      "learning_rate": 0.0005952990841328536,
+      "loss": 3.4526,
+      "step": 3454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5945968627929688,
+      "learning_rate": 0.0005952963688625538,
+      "loss": 3.5846,
+      "step": 3455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5942106246948242,
+      "learning_rate": 0.0005952936528145004,
+      "loss": 3.5166,
+      "step": 3456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5890370607376099,
+      "learning_rate": 0.0005952909359887002,
+      "loss": 3.7746,
+      "step": 3457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5749627351760864,
+      "learning_rate": 0.0005952882183851606,
+      "loss": 3.4682,
+      "step": 3458
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59616619348526,
+      "learning_rate": 0.0005952855000038886,
+      "loss": 3.7142,
+      "step": 3459
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5930215716362,
+      "learning_rate": 0.0005952827808448916,
+      "loss": 3.7,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.564595103263855,
+      "learning_rate": 0.0005952800609081764,
+      "loss": 3.5487,
+      "step": 3461
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5829174518585205,
+      "learning_rate": 0.0005952773401937504,
+      "loss": 3.6912,
+      "step": 3462
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5580412149429321,
+      "learning_rate": 0.0005952746187016208,
+      "loss": 3.5539,
+      "step": 3463
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5945863127708435,
+      "learning_rate": 0.0005952718964317945,
+      "loss": 3.567,
+      "step": 3464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5504742860794067,
+      "learning_rate": 0.000595269173384279,
+      "loss": 3.6912,
+      "step": 3465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6526525616645813,
+      "learning_rate": 0.0005952664495590813,
+      "loss": 3.4437,
+      "step": 3466
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5954388976097107,
+      "learning_rate": 0.0005952637249562085,
+      "loss": 3.6307,
+      "step": 3467
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.712817907333374,
+      "learning_rate": 0.000595260999575668,
+      "loss": 3.6924,
+      "step": 3468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6235787868499756,
+      "learning_rate": 0.0005952582734174667,
+      "loss": 3.5941,
+      "step": 3469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5776662826538086,
+      "learning_rate": 0.0005952555464816122,
+      "loss": 3.8929,
+      "step": 3470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6015415787696838,
+      "learning_rate": 0.0005952528187681111,
+      "loss": 3.6784,
+      "step": 3471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5785424113273621,
+      "learning_rate": 0.0005952500902769711,
+      "loss": 3.7124,
+      "step": 3472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6219342947006226,
+      "learning_rate": 0.0005952473610081991,
+      "loss": 3.6605,
+      "step": 3473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5495757460594177,
+      "learning_rate": 0.0005952446309618023,
+      "loss": 3.5599,
+      "step": 3474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5983782410621643,
+      "learning_rate": 0.0005952419001377881,
+      "loss": 3.5188,
+      "step": 3475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6638477444648743,
+      "learning_rate": 0.0005952391685361634,
+      "loss": 3.6374,
+      "step": 3476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5729101896286011,
+      "learning_rate": 0.0005952364361569358,
+      "loss": 3.4728,
+      "step": 3477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5611720681190491,
+      "learning_rate": 0.000595233703000112,
+      "loss": 3.4283,
+      "step": 3478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.615308403968811,
+      "learning_rate": 0.0005952309690656997,
+      "loss": 3.6614,
+      "step": 3479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6057369709014893,
+      "learning_rate": 0.0005952282343537057,
+      "loss": 3.4623,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6732720136642456,
+      "learning_rate": 0.0005952254988641373,
+      "loss": 3.7109,
+      "step": 3481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5600842833518982,
+      "learning_rate": 0.0005952227625970018,
+      "loss": 3.6057,
+      "step": 3482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5803719162940979,
+      "learning_rate": 0.0005952200255523063,
+      "loss": 3.6392,
+      "step": 3483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6266722679138184,
+      "learning_rate": 0.0005952172877300582,
+      "loss": 3.5691,
+      "step": 3484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5460732579231262,
+      "learning_rate": 0.0005952145491302645,
+      "loss": 3.3658,
+      "step": 3485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5742864012718201,
+      "learning_rate": 0.0005952118097529325,
+      "loss": 3.7658,
+      "step": 3486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5541781187057495,
+      "learning_rate": 0.0005952090695980695,
+      "loss": 3.4799,
+      "step": 3487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6336858868598938,
+      "learning_rate": 0.0005952063286656824,
+      "loss": 3.2493,
+      "step": 3488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5907337069511414,
+      "learning_rate": 0.0005952035869557788,
+      "loss": 3.4372,
+      "step": 3489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5762500166893005,
+      "learning_rate": 0.0005952008444683657,
+      "loss": 3.8303,
+      "step": 3490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.580866813659668,
+      "learning_rate": 0.0005951981012034505,
+      "loss": 3.409,
+      "step": 3491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5785665512084961,
+      "learning_rate": 0.0005951953571610403,
+      "loss": 3.4152,
+      "step": 3492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5975906848907471,
+      "learning_rate": 0.0005951926123411422,
+      "loss": 3.554,
+      "step": 3493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5543960332870483,
+      "learning_rate": 0.0005951898667437637,
+      "loss": 3.4609,
+      "step": 3494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58534836769104,
+      "learning_rate": 0.0005951871203689118,
+      "loss": 3.5745,
+      "step": 3495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5618305802345276,
+      "learning_rate": 0.0005951843732165938,
+      "loss": 3.6903,
+      "step": 3496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6008855700492859,
+      "learning_rate": 0.0005951816252868172,
+      "loss": 3.5196,
+      "step": 3497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6193830370903015,
+      "learning_rate": 0.0005951788765795889,
+      "loss": 3.4215,
+      "step": 3498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5549221038818359,
+      "learning_rate": 0.0005951761270949161,
+      "loss": 3.4351,
+      "step": 3499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5623039603233337,
+      "learning_rate": 0.0005951733768328063,
+      "loss": 3.6452,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5464016795158386,
+      "learning_rate": 0.0005951706257932667,
+      "loss": 3.518,
+      "step": 3501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5889121294021606,
+      "learning_rate": 0.0005951678739763044,
+      "loss": 3.6592,
+      "step": 3502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6298267245292664,
+      "learning_rate": 0.0005951651213819268,
+      "loss": 3.5137,
+      "step": 3503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5713311433792114,
+      "learning_rate": 0.000595162368010141,
+      "loss": 3.5688,
+      "step": 3504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.610822856426239,
+      "learning_rate": 0.0005951596138609543,
+      "loss": 3.6497,
+      "step": 3505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5696738362312317,
+      "learning_rate": 0.000595156858934374,
+      "loss": 3.3546,
+      "step": 3506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5880289673805237,
+      "learning_rate": 0.0005951541032304075,
+      "loss": 3.7143,
+      "step": 3507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6033392548561096,
+      "learning_rate": 0.0005951513467490617,
+      "loss": 3.5457,
+      "step": 3508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5870084166526794,
+      "learning_rate": 0.0005951485894903441,
+      "loss": 3.4465,
+      "step": 3509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5881246328353882,
+      "learning_rate": 0.0005951458314542619,
+      "loss": 3.6633,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5996487140655518,
+      "learning_rate": 0.0005951430726408224,
+      "loss": 3.5734,
+      "step": 3511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5851162075996399,
+      "learning_rate": 0.0005951403130500329,
+      "loss": 3.5603,
+      "step": 3512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5705832242965698,
+      "learning_rate": 0.0005951375526819006,
+      "loss": 3.5581,
+      "step": 3513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5960780382156372,
+      "learning_rate": 0.0005951347915364327,
+      "loss": 3.4452,
+      "step": 3514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5697128176689148,
+      "learning_rate": 0.0005951320296136367,
+      "loss": 3.6832,
+      "step": 3515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5737714171409607,
+      "learning_rate": 0.0005951292669135197,
+      "loss": 3.6346,
+      "step": 3516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6310943365097046,
+      "learning_rate": 0.0005951265034360889,
+      "loss": 3.3565,
+      "step": 3517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5577863454818726,
+      "learning_rate": 0.0005951237391813518,
+      "loss": 3.4109,
+      "step": 3518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5705747008323669,
+      "learning_rate": 0.0005951209741493154,
+      "loss": 3.4384,
+      "step": 3519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5870479345321655,
+      "learning_rate": 0.0005951182083399874,
+      "loss": 3.5561,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5648481845855713,
+      "learning_rate": 0.0005951154417533748,
+      "loss": 3.3586,
+      "step": 3521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5373155474662781,
+      "learning_rate": 0.0005951126743894846,
+      "loss": 3.7483,
+      "step": 3522
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5593510270118713,
+      "learning_rate": 0.0005951099062483248,
+      "loss": 3.4208,
+      "step": 3523
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5978959202766418,
+      "learning_rate": 0.0005951071373299021,
+      "loss": 3.5202,
+      "step": 3524
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5732042789459229,
+      "learning_rate": 0.0005951043676342241,
+      "loss": 3.5725,
+      "step": 3525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5535354018211365,
+      "learning_rate": 0.000595101597161298,
+      "loss": 3.741,
+      "step": 3526
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5729936957359314,
+      "learning_rate": 0.0005950988259111309,
+      "loss": 3.7625,
+      "step": 3527
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5622599124908447,
+      "learning_rate": 0.0005950960538837305,
+      "loss": 3.3271,
+      "step": 3528
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5429548025131226,
+      "learning_rate": 0.0005950932810791038,
+      "loss": 3.5124,
+      "step": 3529
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5643887519836426,
+      "learning_rate": 0.0005950905074972581,
+      "loss": 3.4261,
+      "step": 3530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.548894464969635,
+      "learning_rate": 0.0005950877331382009,
+      "loss": 3.6266,
+      "step": 3531
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5614020228385925,
+      "learning_rate": 0.0005950849580019393,
+      "loss": 3.4156,
+      "step": 3532
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6196742653846741,
+      "learning_rate": 0.0005950821820884808,
+      "loss": 3.5224,
+      "step": 3533
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5436410903930664,
+      "learning_rate": 0.0005950794053978327,
+      "loss": 3.4044,
+      "step": 3534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5510107278823853,
+      "learning_rate": 0.0005950766279300021,
+      "loss": 3.438,
+      "step": 3535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6485077738761902,
+      "learning_rate": 0.0005950738496849965,
+      "loss": 3.4866,
+      "step": 3536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5946866869926453,
+      "learning_rate": 0.0005950710706628232,
+      "loss": 3.5581,
+      "step": 3537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6310849189758301,
+      "learning_rate": 0.0005950682908634895,
+      "loss": 3.4498,
+      "step": 3538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5689077973365784,
+      "learning_rate": 0.0005950655102870026,
+      "loss": 3.4855,
+      "step": 3539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5511724352836609,
+      "learning_rate": 0.00059506272893337,
+      "loss": 3.6545,
+      "step": 3540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5716314911842346,
+      "learning_rate": 0.000595059946802599,
+      "loss": 3.5961,
+      "step": 3541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5928378105163574,
+      "learning_rate": 0.0005950571638946968,
+      "loss": 3.3571,
+      "step": 3542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5479989051818848,
+      "learning_rate": 0.0005950543802096708,
+      "loss": 3.328,
+      "step": 3543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5652512311935425,
+      "learning_rate": 0.0005950515957475284,
+      "loss": 3.6244,
+      "step": 3544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6006174683570862,
+      "learning_rate": 0.000595048810508277,
+      "loss": 3.5703,
+      "step": 3545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5618359446525574,
+      "learning_rate": 0.0005950460244919236,
+      "loss": 3.5093,
+      "step": 3546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6154499053955078,
+      "learning_rate": 0.0005950432376984758,
+      "loss": 3.6431,
+      "step": 3547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5481139421463013,
+      "learning_rate": 0.000595040450127941,
+      "loss": 3.5728,
+      "step": 3548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5588555932044983,
+      "learning_rate": 0.0005950376617803264,
+      "loss": 3.5644,
+      "step": 3549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5340237021446228,
+      "learning_rate": 0.0005950348726556393,
+      "loss": 3.6245,
+      "step": 3550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5997275710105896,
+      "learning_rate": 0.0005950320827538873,
+      "loss": 3.4301,
+      "step": 3551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6511825919151306,
+      "learning_rate": 0.0005950292920750773,
+      "loss": 3.5778,
+      "step": 3552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6428468823432922,
+      "learning_rate": 0.0005950265006192171,
+      "loss": 3.6624,
+      "step": 3553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.576602041721344,
+      "learning_rate": 0.0005950237083863139,
+      "loss": 3.6925,
+      "step": 3554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5820669531822205,
+      "learning_rate": 0.0005950209153763751,
+      "loss": 3.4164,
+      "step": 3555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5869800448417664,
+      "learning_rate": 0.0005950181215894078,
+      "loss": 3.6757,
+      "step": 3556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5729996562004089,
+      "learning_rate": 0.0005950153270254196,
+      "loss": 3.4834,
+      "step": 3557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5958981513977051,
+      "learning_rate": 0.0005950125316844179,
+      "loss": 3.5674,
+      "step": 3558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6202664375305176,
+      "learning_rate": 0.0005950097355664098,
+      "loss": 3.4014,
+      "step": 3559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5707253813743591,
+      "learning_rate": 0.000595006938671403,
+      "loss": 3.5684,
+      "step": 3560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6256164908409119,
+      "learning_rate": 0.0005950041409994046,
+      "loss": 3.6933,
+      "step": 3561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5959612131118774,
+      "learning_rate": 0.0005950013425504221,
+      "loss": 3.553,
+      "step": 3562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8165073394775391,
+      "learning_rate": 0.0005949985433244628,
+      "loss": 3.6789,
+      "step": 3563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5742766261100769,
+      "learning_rate": 0.0005949957433215342,
+      "loss": 3.2581,
+      "step": 3564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5817901492118835,
+      "learning_rate": 0.0005949929425416435,
+      "loss": 3.4921,
+      "step": 3565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.561545729637146,
+      "learning_rate": 0.0005949901409847982,
+      "loss": 3.4838,
+      "step": 3566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5388302803039551,
+      "learning_rate": 0.0005949873386510056,
+      "loss": 3.5186,
+      "step": 3567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5865461826324463,
+      "learning_rate": 0.0005949845355402732,
+      "loss": 3.4544,
+      "step": 3568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5656617879867554,
+      "learning_rate": 0.0005949817316526083,
+      "loss": 3.6103,
+      "step": 3569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5684191584587097,
+      "learning_rate": 0.0005949789269880182,
+      "loss": 3.4794,
+      "step": 3570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5902529358863831,
+      "learning_rate": 0.0005949761215465104,
+      "loss": 3.5839,
+      "step": 3571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6156822443008423,
+      "learning_rate": 0.0005949733153280922,
+      "loss": 3.6531,
+      "step": 3572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5745266079902649,
+      "learning_rate": 0.0005949705083327713,
+      "loss": 3.7956,
+      "step": 3573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5861325860023499,
+      "learning_rate": 0.0005949677005605546,
+      "loss": 3.6641,
+      "step": 3574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5582994818687439,
+      "learning_rate": 0.0005949648920114499,
+      "loss": 3.4384,
+      "step": 3575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5631774663925171,
+      "learning_rate": 0.0005949620826854644,
+      "loss": 3.3741,
+      "step": 3576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5708392858505249,
+      "learning_rate": 0.0005949592725826055,
+      "loss": 3.4894,
+      "step": 3577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6094598770141602,
+      "learning_rate": 0.0005949564617028807,
+      "loss": 3.3773,
+      "step": 3578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6077072620391846,
+      "learning_rate": 0.0005949536500462972,
+      "loss": 3.4769,
+      "step": 3579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5706542134284973,
+      "learning_rate": 0.0005949508376128627,
+      "loss": 3.5485,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54489666223526,
+      "learning_rate": 0.0005949480244025844,
+      "loss": 3.4537,
+      "step": 3581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.560539960861206,
+      "learning_rate": 0.0005949452104154698,
+      "loss": 3.6606,
+      "step": 3582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6367251873016357,
+      "learning_rate": 0.0005949423956515263,
+      "loss": 3.4471,
+      "step": 3583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5641475915908813,
+      "learning_rate": 0.0005949395801107612,
+      "loss": 3.6803,
+      "step": 3584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5714133977890015,
+      "learning_rate": 0.0005949367637931821,
+      "loss": 3.6771,
+      "step": 3585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.539216935634613,
+      "learning_rate": 0.0005949339466987963,
+      "loss": 3.5072,
+      "step": 3586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5528108477592468,
+      "learning_rate": 0.0005949311288276113,
+      "loss": 3.6318,
+      "step": 3587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5910629630088806,
+      "learning_rate": 0.0005949283101796342,
+      "loss": 3.3507,
+      "step": 3588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.640604555606842,
+      "learning_rate": 0.000594925490754873,
+      "loss": 3.4874,
+      "step": 3589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6236400008201599,
+      "learning_rate": 0.0005949226705533348,
+      "loss": 3.4188,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5964713096618652,
+      "learning_rate": 0.0005949198495750268,
+      "loss": 3.4705,
+      "step": 3591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5554364323616028,
+      "learning_rate": 0.0005949170278199569,
+      "loss": 3.4395,
+      "step": 3592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5770800709724426,
+      "learning_rate": 0.0005949142052881322,
+      "loss": 3.5081,
+      "step": 3593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6268064975738525,
+      "learning_rate": 0.0005949113819795603,
+      "loss": 3.5902,
+      "step": 3594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5979000329971313,
+      "learning_rate": 0.0005949085578942486,
+      "loss": 3.3657,
+      "step": 3595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5864534974098206,
+      "learning_rate": 0.0005949057330322044,
+      "loss": 3.6838,
+      "step": 3596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5664005279541016,
+      "learning_rate": 0.0005949029073934354,
+      "loss": 3.4289,
+      "step": 3597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5910850167274475,
+      "learning_rate": 0.0005949000809779488,
+      "loss": 3.5872,
+      "step": 3598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.590392529964447,
+      "learning_rate": 0.0005948972537857521,
+      "loss": 3.4473,
+      "step": 3599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5588194727897644,
+      "learning_rate": 0.0005948944258168529,
+      "loss": 3.4485,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5453645586967468,
+      "learning_rate": 0.0005948915970712585,
+      "loss": 3.6381,
+      "step": 3601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5611460208892822,
+      "learning_rate": 0.0005948887675489763,
+      "loss": 3.7556,
+      "step": 3602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5924997329711914,
+      "learning_rate": 0.000594885937250014,
+      "loss": 3.5867,
+      "step": 3603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5849661231040955,
+      "learning_rate": 0.0005948831061743788,
+      "loss": 3.5556,
+      "step": 3604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5972135066986084,
+      "learning_rate": 0.0005948802743220782,
+      "loss": 3.4393,
+      "step": 3605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5754001140594482,
+      "learning_rate": 0.0005948774416931197,
+      "loss": 3.5528,
+      "step": 3606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.592771053314209,
+      "learning_rate": 0.0005948746082875109,
+      "loss": 3.4963,
+      "step": 3607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6338792443275452,
+      "learning_rate": 0.0005948717741052591,
+      "loss": 3.2424,
+      "step": 3608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5741003751754761,
+      "learning_rate": 0.0005948689391463718,
+      "loss": 3.6585,
+      "step": 3609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5677915811538696,
+      "learning_rate": 0.0005948661034108564,
+      "loss": 3.5649,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.59786057472229,
+      "learning_rate": 0.0005948632668987204,
+      "loss": 3.6715,
+      "step": 3611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5489624738693237,
+      "learning_rate": 0.0005948604296099714,
+      "loss": 3.3292,
+      "step": 3612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5429912209510803,
+      "learning_rate": 0.0005948575915446167,
+      "loss": 3.6417,
+      "step": 3613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.637424647808075,
+      "learning_rate": 0.0005948547527026639,
+      "loss": 3.4017,
+      "step": 3614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5805572867393494,
+      "learning_rate": 0.0005948519130841203,
+      "loss": 3.4764,
+      "step": 3615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6204468011856079,
+      "learning_rate": 0.0005948490726889937,
+      "loss": 3.5658,
+      "step": 3616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5520532131195068,
+      "learning_rate": 0.0005948462315172913,
+      "loss": 3.7156,
+      "step": 3617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5868279933929443,
+      "learning_rate": 0.0005948433895690207,
+      "loss": 3.5782,
+      "step": 3618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5735620856285095,
+      "learning_rate": 0.0005948405468441893,
+      "loss": 3.5567,
+      "step": 3619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5258257389068604,
+      "learning_rate": 0.0005948377033428046,
+      "loss": 3.4518,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5907759070396423,
+      "learning_rate": 0.0005948348590648743,
+      "loss": 3.4862,
+      "step": 3621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5513227581977844,
+      "learning_rate": 0.0005948320140104056,
+      "loss": 3.3782,
+      "step": 3622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.620806097984314,
+      "learning_rate": 0.0005948291681794062,
+      "loss": 3.4239,
+      "step": 3623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6315174698829651,
+      "learning_rate": 0.0005948263215718834,
+      "loss": 3.5383,
+      "step": 3624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6031731963157654,
+      "learning_rate": 0.0005948234741878449,
+      "loss": 3.4314,
+      "step": 3625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6577680110931396,
+      "learning_rate": 0.0005948206260272981,
+      "loss": 3.4231,
+      "step": 3626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5863077640533447,
+      "learning_rate": 0.0005948177770902504,
+      "loss": 3.5375,
+      "step": 3627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5604285597801208,
+      "learning_rate": 0.0005948149273767095,
+      "loss": 3.5875,
+      "step": 3628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6838080286979675,
+      "learning_rate": 0.0005948120768866828,
+      "loss": 3.4889,
+      "step": 3629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5543857216835022,
+      "learning_rate": 0.0005948092256201779,
+      "loss": 3.6624,
+      "step": 3630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5770386457443237,
+      "learning_rate": 0.0005948063735772022,
+      "loss": 3.6973,
+      "step": 3631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6096564531326294,
+      "learning_rate": 0.0005948035207577632,
+      "loss": 3.6492,
+      "step": 3632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6562588810920715,
+      "learning_rate": 0.0005948006671618685,
+      "loss": 3.6059,
+      "step": 3633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.576315701007843,
+      "learning_rate": 0.0005947978127895256,
+      "loss": 3.7477,
+      "step": 3634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5762778520584106,
+      "learning_rate": 0.000594794957640742,
+      "loss": 3.5309,
+      "step": 3635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.596248984336853,
+      "learning_rate": 0.0005947921017155251,
+      "loss": 3.5666,
+      "step": 3636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5929814577102661,
+      "learning_rate": 0.0005947892450138827,
+      "loss": 3.4502,
+      "step": 3637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.581706702709198,
+      "learning_rate": 0.000594786387535822,
+      "loss": 3.4126,
+      "step": 3638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.609158992767334,
+      "learning_rate": 0.0005947835292813507,
+      "loss": 3.6607,
+      "step": 3639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56010901927948,
+      "learning_rate": 0.0005947806702504765,
+      "loss": 3.5313,
+      "step": 3640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5166653990745544,
+      "learning_rate": 0.0005947778104432067,
+      "loss": 3.6868,
+      "step": 3641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6067258715629578,
+      "learning_rate": 0.0005947749498595487,
+      "loss": 3.4264,
+      "step": 3642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5833512544631958,
+      "learning_rate": 0.0005947720884995104,
+      "loss": 3.5246,
+      "step": 3643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5999512076377869,
+      "learning_rate": 0.000594769226363099,
+      "loss": 3.3084,
+      "step": 3644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5532363057136536,
+      "learning_rate": 0.0005947663634503223,
+      "loss": 3.6718,
+      "step": 3645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5565996170043945,
+      "learning_rate": 0.0005947634997611876,
+      "loss": 3.4042,
+      "step": 3646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6335086822509766,
+      "learning_rate": 0.0005947606352957027,
+      "loss": 3.5925,
+      "step": 3647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5904930830001831,
+      "learning_rate": 0.0005947577700538751,
+      "loss": 3.4858,
+      "step": 3648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5819961428642273,
+      "learning_rate": 0.000594754904035712,
+      "loss": 3.4756,
+      "step": 3649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5676370859146118,
+      "learning_rate": 0.0005947520372412214,
+      "loss": 3.3994,
+      "step": 3650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.528371274471283,
+      "learning_rate": 0.0005947491696704105,
+      "loss": 3.5782,
+      "step": 3651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5858977437019348,
+      "learning_rate": 0.0005947463013232872,
+      "loss": 3.5947,
+      "step": 3652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6060065627098083,
+      "learning_rate": 0.0005947434321998588,
+      "loss": 3.3542,
+      "step": 3653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6193976402282715,
+      "learning_rate": 0.000594740562300133,
+      "loss": 3.4906,
+      "step": 3654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6060580015182495,
+      "learning_rate": 0.000594737691624117,
+      "loss": 3.4806,
+      "step": 3655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6116194725036621,
+      "learning_rate": 0.000594734820171819,
+      "loss": 3.5376,
+      "step": 3656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5761888027191162,
+      "learning_rate": 0.000594731947943246,
+      "loss": 3.5894,
+      "step": 3657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5662036538124084,
+      "learning_rate": 0.0005947290749384059,
+      "loss": 3.5399,
+      "step": 3658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6075831055641174,
+      "learning_rate": 0.0005947262011573059,
+      "loss": 3.7756,
+      "step": 3659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5735023617744446,
+      "learning_rate": 0.000594723326599954,
+      "loss": 3.48,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5627445578575134,
+      "learning_rate": 0.0005947204512663575,
+      "loss": 3.6544,
+      "step": 3661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5845668911933899,
+      "learning_rate": 0.0005947175751565241,
+      "loss": 3.4011,
+      "step": 3662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5823236107826233,
+      "learning_rate": 0.0005947146982704613,
+      "loss": 3.2938,
+      "step": 3663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5699519515037537,
+      "learning_rate": 0.0005947118206081767,
+      "loss": 3.7,
+      "step": 3664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6400408744812012,
+      "learning_rate": 0.0005947089421696779,
+      "loss": 3.572,
+      "step": 3665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5679846405982971,
+      "learning_rate": 0.0005947060629549725,
+      "loss": 3.7148,
+      "step": 3666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5408234000205994,
+      "learning_rate": 0.0005947031829640681,
+      "loss": 3.35,
+      "step": 3667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5761770606040955,
+      "learning_rate": 0.0005947003021969721,
+      "loss": 3.5302,
+      "step": 3668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5421696305274963,
+      "learning_rate": 0.0005946974206536923,
+      "loss": 3.7173,
+      "step": 3669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5892155170440674,
+      "learning_rate": 0.0005946945383342361,
+      "loss": 3.5065,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6017743349075317,
+      "learning_rate": 0.0005946916552386113,
+      "loss": 3.4401,
+      "step": 3671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5975637435913086,
+      "learning_rate": 0.0005946887713668253,
+      "loss": 3.6127,
+      "step": 3672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5787088871002197,
+      "learning_rate": 0.000594685886718886,
+      "loss": 3.635,
+      "step": 3673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5916794538497925,
+      "learning_rate": 0.0005946830012948006,
+      "loss": 3.4734,
+      "step": 3674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5836448669433594,
+      "learning_rate": 0.000594680115094577,
+      "loss": 3.4839,
+      "step": 3675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5569721460342407,
+      "learning_rate": 0.0005946772281182226,
+      "loss": 3.3795,
+      "step": 3676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6041418313980103,
+      "learning_rate": 0.000594674340365745,
+      "loss": 3.8186,
+      "step": 3677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5707287788391113,
+      "learning_rate": 0.0005946714518371521,
+      "loss": 3.5052,
+      "step": 3678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5691279768943787,
+      "learning_rate": 0.0005946685625324512,
+      "loss": 3.693,
+      "step": 3679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.545986533164978,
+      "learning_rate": 0.00059466567245165,
+      "loss": 3.4695,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5766247510910034,
+      "learning_rate": 0.0005946627815947562,
+      "loss": 3.6347,
+      "step": 3681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.624789834022522,
+      "learning_rate": 0.0005946598899617773,
+      "loss": 3.4847,
+      "step": 3682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5886110067367554,
+      "learning_rate": 0.0005946569975527209,
+      "loss": 3.5243,
+      "step": 3683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6018827557563782,
+      "learning_rate": 0.0005946541043675948,
+      "loss": 3.4334,
+      "step": 3684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.595812201499939,
+      "learning_rate": 0.0005946512104064065,
+      "loss": 3.551,
+      "step": 3685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6566250920295715,
+      "learning_rate": 0.0005946483156691636,
+      "loss": 3.6164,
+      "step": 3686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5355682373046875,
+      "learning_rate": 0.0005946454201558736,
+      "loss": 3.5166,
+      "step": 3687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5876386761665344,
+      "learning_rate": 0.0005946425238665445,
+      "loss": 3.4562,
+      "step": 3688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5719892978668213,
+      "learning_rate": 0.0005946396268011834,
+      "loss": 3.3293,
+      "step": 3689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5640550255775452,
+      "learning_rate": 0.0005946367289597985,
+      "loss": 3.4951,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.580362856388092,
+      "learning_rate": 0.0005946338303423971,
+      "loss": 3.4677,
+      "step": 3691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.604098379611969,
+      "learning_rate": 0.0005946309309489868,
+      "loss": 3.556,
+      "step": 3692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6758443117141724,
+      "learning_rate": 0.0005946280307795753,
+      "loss": 3.2999,
+      "step": 3693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6679175496101379,
+      "learning_rate": 0.0005946251298341704,
+      "loss": 3.4755,
+      "step": 3694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5557239651679993,
+      "learning_rate": 0.0005946222281127796,
+      "loss": 3.712,
+      "step": 3695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5595870018005371,
+      "learning_rate": 0.0005946193256154105,
+      "loss": 3.5436,
+      "step": 3696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5639623403549194,
+      "learning_rate": 0.0005946164223420708,
+      "loss": 3.2984,
+      "step": 3697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6050103902816772,
+      "learning_rate": 0.0005946135182927681,
+      "loss": 3.4194,
+      "step": 3698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5879496335983276,
+      "learning_rate": 0.0005946106134675102,
+      "loss": 3.2472,
+      "step": 3699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5664472579956055,
+      "learning_rate": 0.0005946077078663046,
+      "loss": 3.4548,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.574516236782074,
+      "learning_rate": 0.000594604801489159,
+      "loss": 3.3961,
+      "step": 3701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.583021879196167,
+      "learning_rate": 0.000594601894336081,
+      "loss": 3.441,
+      "step": 3702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5696473121643066,
+      "learning_rate": 0.0005945989864070783,
+      "loss": 3.5425,
+      "step": 3703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5899128913879395,
+      "learning_rate": 0.0005945960777021587,
+      "loss": 3.449,
+      "step": 3704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5821749567985535,
+      "learning_rate": 0.0005945931682213296,
+      "loss": 3.6383,
+      "step": 3705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5889866352081299,
+      "learning_rate": 0.0005945902579645988,
+      "loss": 3.5592,
+      "step": 3706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6704166531562805,
+      "learning_rate": 0.000594587346931974,
+      "loss": 3.2622,
+      "step": 3707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5655115842819214,
+      "learning_rate": 0.0005945844351234628,
+      "loss": 3.5981,
+      "step": 3708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5271401405334473,
+      "learning_rate": 0.0005945815225390729,
+      "loss": 3.6283,
+      "step": 3709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5654386281967163,
+      "learning_rate": 0.0005945786091788119,
+      "loss": 3.3157,
+      "step": 3710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5483503341674805,
+      "learning_rate": 0.0005945756950426876,
+      "loss": 3.4272,
+      "step": 3711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.554814875125885,
+      "learning_rate": 0.0005945727801307077,
+      "loss": 3.5075,
+      "step": 3712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6500690579414368,
+      "learning_rate": 0.0005945698644428797,
+      "loss": 3.6308,
+      "step": 3713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5715118050575256,
+      "learning_rate": 0.0005945669479792113,
+      "loss": 3.6008,
+      "step": 3714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5506441593170166,
+      "learning_rate": 0.0005945640307397104,
+      "loss": 3.5656,
+      "step": 3715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5786412358283997,
+      "learning_rate": 0.0005945611127243844,
+      "loss": 3.575,
+      "step": 3716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6186723709106445,
+      "learning_rate": 0.0005945581939332412,
+      "loss": 3.6169,
+      "step": 3717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5763697624206543,
+      "learning_rate": 0.0005945552743662883,
+      "loss": 3.7148,
+      "step": 3718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.659016489982605,
+      "learning_rate": 0.0005945523540235336,
+      "loss": 3.5355,
+      "step": 3719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5776824951171875,
+      "learning_rate": 0.0005945494329049846,
+      "loss": 3.6071,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5591484904289246,
+      "learning_rate": 0.0005945465110106492,
+      "loss": 3.389,
+      "step": 3721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5636910200119019,
+      "learning_rate": 0.0005945435883405348,
+      "loss": 3.594,
+      "step": 3722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6116440296173096,
+      "learning_rate": 0.0005945406648946495,
+      "loss": 3.4926,
+      "step": 3723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.558946430683136,
+      "learning_rate": 0.0005945377406730007,
+      "loss": 3.6447,
+      "step": 3724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5685746669769287,
+      "learning_rate": 0.0005945348156755962,
+      "loss": 3.4761,
+      "step": 3725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9950370192527771,
+      "learning_rate": 0.0005945318899024435,
+      "loss": 3.7469,
+      "step": 3726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7801612615585327,
+      "learning_rate": 0.0005945289633535506,
+      "loss": 3.6114,
+      "step": 3727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6060027480125427,
+      "learning_rate": 0.0005945260360289251,
+      "loss": 3.4032,
+      "step": 3728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.580491840839386,
+      "learning_rate": 0.0005945231079285747,
+      "loss": 3.3195,
+      "step": 3729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7394986748695374,
+      "learning_rate": 0.000594520179052507,
+      "loss": 3.6785,
+      "step": 3730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5658180713653564,
+      "learning_rate": 0.00059451724940073,
+      "loss": 3.4843,
+      "step": 3731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5555374622344971,
+      "learning_rate": 0.0005945143189732512,
+      "loss": 3.4209,
+      "step": 3732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6545796990394592,
+      "learning_rate": 0.0005945113877700785,
+      "loss": 3.52,
+      "step": 3733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5792942047119141,
+      "learning_rate": 0.0005945084557912192,
+      "loss": 3.5676,
+      "step": 3734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.587958574295044,
+      "learning_rate": 0.0005945055230366815,
+      "loss": 3.5171,
+      "step": 3735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.570886492729187,
+      "learning_rate": 0.0005945025895064728,
+      "loss": 3.5358,
+      "step": 3736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5610484480857849,
+      "learning_rate": 0.0005944996552006011,
+      "loss": 3.6122,
+      "step": 3737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6265158653259277,
+      "learning_rate": 0.000594496720119074,
+      "loss": 3.3024,
+      "step": 3738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5752745866775513,
+      "learning_rate": 0.0005944937842618991,
+      "loss": 3.6183,
+      "step": 3739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6028185486793518,
+      "learning_rate": 0.0005944908476290843,
+      "loss": 3.4493,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5684720873832703,
+      "learning_rate": 0.0005944879102206372,
+      "loss": 3.5517,
+      "step": 3741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5804165005683899,
+      "learning_rate": 0.0005944849720365657,
+      "loss": 3.3608,
+      "step": 3742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5693696737289429,
+      "learning_rate": 0.0005944820330768775,
+      "loss": 3.5387,
+      "step": 3743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6432933807373047,
+      "learning_rate": 0.0005944790933415802,
+      "loss": 3.5484,
+      "step": 3744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6080823540687561,
+      "learning_rate": 0.0005944761528306817,
+      "loss": 3.5617,
+      "step": 3745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5980885028839111,
+      "learning_rate": 0.0005944732115441897,
+      "loss": 3.4468,
+      "step": 3746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6021443605422974,
+      "learning_rate": 0.0005944702694821119,
+      "loss": 3.4973,
+      "step": 3747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6065918803215027,
+      "learning_rate": 0.0005944673266444561,
+      "loss": 3.3864,
+      "step": 3748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.579862654209137,
+      "learning_rate": 0.00059446438303123,
+      "loss": 3.7255,
+      "step": 3749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5811945199966431,
+      "learning_rate": 0.0005944614386424414,
+      "loss": 3.5034,
+      "step": 3750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5766950249671936,
+      "learning_rate": 0.0005944584934780981,
+      "loss": 3.7607,
+      "step": 3751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6201891899108887,
+      "learning_rate": 0.0005944555475382077,
+      "loss": 3.342,
+      "step": 3752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5696931481361389,
+      "learning_rate": 0.0005944526008227781,
+      "loss": 3.419,
+      "step": 3753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6073271036148071,
+      "learning_rate": 0.000594449653331817,
+      "loss": 3.3548,
+      "step": 3754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5775006413459778,
+      "learning_rate": 0.0005944467050653323,
+      "loss": 3.3832,
+      "step": 3755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6126733422279358,
+      "learning_rate": 0.0005944437560233314,
+      "loss": 3.3462,
+      "step": 3756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7034265995025635,
+      "learning_rate": 0.0005944408062058225,
+      "loss": 3.4659,
+      "step": 3757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60416179895401,
+      "learning_rate": 0.0005944378556128132,
+      "loss": 3.2714,
+      "step": 3758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.640256404876709,
+      "learning_rate": 0.0005944349042443112,
+      "loss": 3.5203,
+      "step": 3759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.590840220451355,
+      "learning_rate": 0.0005944319521003243,
+      "loss": 3.5765,
+      "step": 3760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.561656653881073,
+      "learning_rate": 0.0005944289991808604,
+      "loss": 3.509,
+      "step": 3761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5961998105049133,
+      "learning_rate": 0.0005944260454859271,
+      "loss": 3.3526,
+      "step": 3762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5496677756309509,
+      "learning_rate": 0.0005944230910155322,
+      "loss": 3.7351,
+      "step": 3763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5810037851333618,
+      "learning_rate": 0.0005944201357696837,
+      "loss": 3.4154,
+      "step": 3764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6055195927619934,
+      "learning_rate": 0.0005944171797483891,
+      "loss": 3.3845,
+      "step": 3765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5705102682113647,
+      "learning_rate": 0.0005944142229516563,
+      "loss": 3.4328,
+      "step": 3766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5758143067359924,
+      "learning_rate": 0.0005944112653794931,
+      "loss": 3.31,
+      "step": 3767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5275263786315918,
+      "learning_rate": 0.0005944083070319074,
+      "loss": 3.6008,
+      "step": 3768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5653240084648132,
+      "learning_rate": 0.0005944053479089069,
+      "loss": 3.4932,
+      "step": 3769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5635776519775391,
+      "learning_rate": 0.0005944023880104992,
+      "loss": 3.5663,
+      "step": 3770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6032636165618896,
+      "learning_rate": 0.0005943994273366923,
+      "loss": 3.2103,
+      "step": 3771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6400607228279114,
+      "learning_rate": 0.000594396465887494,
+      "loss": 3.3427,
+      "step": 3772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5439187288284302,
+      "learning_rate": 0.0005943935036629122,
+      "loss": 3.5142,
+      "step": 3773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7817490100860596,
+      "learning_rate": 0.0005943905406629545,
+      "loss": 3.3834,
+      "step": 3774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5938287973403931,
+      "learning_rate": 0.0005943875768876286,
+      "loss": 3.4816,
+      "step": 3775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5887666940689087,
+      "learning_rate": 0.0005943846123369426,
+      "loss": 3.7435,
+      "step": 3776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5787019729614258,
+      "learning_rate": 0.0005943816470109042,
+      "loss": 3.5568,
+      "step": 3777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5176889896392822,
+      "learning_rate": 0.0005943786809095213,
+      "loss": 3.514,
+      "step": 3778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5439308285713196,
+      "learning_rate": 0.0005943757140328014,
+      "loss": 3.416,
+      "step": 3779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5765101909637451,
+      "learning_rate": 0.0005943727463807527,
+      "loss": 3.6599,
+      "step": 3780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5823919177055359,
+      "learning_rate": 0.0005943697779533828,
+      "loss": 3.4311,
+      "step": 3781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5826864242553711,
+      "learning_rate": 0.0005943668087506994,
+      "loss": 3.5703,
+      "step": 3782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5674658417701721,
+      "learning_rate": 0.0005943638387727107,
+      "loss": 3.3271,
+      "step": 3783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.564477801322937,
+      "learning_rate": 0.0005943608680194241,
+      "loss": 3.3989,
+      "step": 3784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5395248532295227,
+      "learning_rate": 0.0005943578964908478,
+      "loss": 3.5984,
+      "step": 3785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5394876599311829,
+      "learning_rate": 0.0005943549241869894,
+      "loss": 3.4793,
+      "step": 3786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5413987636566162,
+      "learning_rate": 0.0005943519511078568,
+      "loss": 3.4323,
+      "step": 3787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.57403963804245,
+      "learning_rate": 0.0005943489772534576,
+      "loss": 3.6065,
+      "step": 3788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5907705426216125,
+      "learning_rate": 0.0005943460026238001,
+      "loss": 3.5587,
+      "step": 3789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5364418625831604,
+      "learning_rate": 0.0005943430272188917,
+      "loss": 3.5492,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5795174241065979,
+      "learning_rate": 0.0005943400510387405,
+      "loss": 3.6131,
+      "step": 3791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5407680869102478,
+      "learning_rate": 0.0005943370740833543,
+      "loss": 3.5612,
+      "step": 3792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6448293924331665,
+      "learning_rate": 0.0005943340963527409,
+      "loss": 3.3383,
+      "step": 3793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.562232255935669,
+      "learning_rate": 0.0005943311178469079,
+      "loss": 3.4904,
+      "step": 3794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5268046259880066,
+      "learning_rate": 0.0005943281385658636,
+      "loss": 3.7969,
+      "step": 3795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5817152261734009,
+      "learning_rate": 0.0005943251585096156,
+      "loss": 3.5373,
+      "step": 3796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.523223876953125,
+      "learning_rate": 0.0005943221776781719,
+      "loss": 3.4779,
+      "step": 3797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6563833355903625,
+      "learning_rate": 0.0005943191960715399,
+      "loss": 3.5764,
+      "step": 3798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6011426448822021,
+      "learning_rate": 0.000594316213689728,
+      "loss": 3.5565,
+      "step": 3799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5760040283203125,
+      "learning_rate": 0.0005943132305327438,
+      "loss": 3.6364,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.551845371723175,
+      "learning_rate": 0.0005943102466005952,
+      "loss": 3.481,
+      "step": 3801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5699048638343811,
+      "learning_rate": 0.0005943072618932901,
+      "loss": 3.3893,
+      "step": 3802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5873308777809143,
+      "learning_rate": 0.0005943042764108363,
+      "loss": 3.6338,
+      "step": 3803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5894830226898193,
+      "learning_rate": 0.0005943012901532416,
+      "loss": 3.4884,
+      "step": 3804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5429089069366455,
+      "learning_rate": 0.0005942983031205139,
+      "loss": 3.4782,
+      "step": 3805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5496631264686584,
+      "learning_rate": 0.0005942953153126613,
+      "loss": 3.7168,
+      "step": 3806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5998590588569641,
+      "learning_rate": 0.0005942923267296913,
+      "loss": 3.7289,
+      "step": 3807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6006838083267212,
+      "learning_rate": 0.000594289337371612,
+      "loss": 3.4137,
+      "step": 3808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5551097393035889,
+      "learning_rate": 0.0005942863472384313,
+      "loss": 3.7467,
+      "step": 3809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5894075036048889,
+      "learning_rate": 0.0005942833563301569,
+      "loss": 3.3947,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5639740228652954,
+      "learning_rate": 0.0005942803646467968,
+      "loss": 3.6335,
+      "step": 3811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5757173299789429,
+      "learning_rate": 0.0005942773721883588,
+      "loss": 3.6161,
+      "step": 3812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.547919750213623,
+      "learning_rate": 0.000594274378954851,
+      "loss": 3.5195,
+      "step": 3813
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.543421745300293,
+      "learning_rate": 0.000594271384946281,
+      "loss": 3.2388,
+      "step": 3814
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5859833359718323,
+      "learning_rate": 0.0005942683901626569,
+      "loss": 3.5582,
+      "step": 3815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5319817066192627,
+      "learning_rate": 0.0005942653946039864,
+      "loss": 3.574,
+      "step": 3816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5493303537368774,
+      "learning_rate": 0.0005942623982702775,
+      "loss": 3.4146,
+      "step": 3817
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5733358860015869,
+      "learning_rate": 0.000594259401161538,
+      "loss": 3.543,
+      "step": 3818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5682793855667114,
+      "learning_rate": 0.000594256403277776,
+      "loss": 3.6779,
+      "step": 3819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5126398801803589,
+      "learning_rate": 0.0005942534046189992,
+      "loss": 3.5189,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5806546211242676,
+      "learning_rate": 0.0005942504051852155,
+      "loss": 3.4849,
+      "step": 3821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5452494025230408,
+      "learning_rate": 0.000594247404976433,
+      "loss": 3.5726,
+      "step": 3822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5558573007583618,
+      "learning_rate": 0.0005942444039926594,
+      "loss": 3.5881,
+      "step": 3823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5447998642921448,
+      "learning_rate": 0.0005942414022339027,
+      "loss": 3.4945,
+      "step": 3824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6169920563697815,
+      "learning_rate": 0.0005942383997001707,
+      "loss": 3.3484,
+      "step": 3825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5817279815673828,
+      "learning_rate": 0.0005942353963914714,
+      "loss": 3.4676,
+      "step": 3826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5804048776626587,
+      "learning_rate": 0.0005942323923078127,
+      "loss": 3.2754,
+      "step": 3827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5827670097351074,
+      "learning_rate": 0.0005942293874492025,
+      "loss": 3.4499,
+      "step": 3828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5710329413414001,
+      "learning_rate": 0.0005942263818156488,
+      "loss": 3.5819,
+      "step": 3829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6096348762512207,
+      "learning_rate": 0.0005942233754071594,
+      "loss": 3.3934,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5973950028419495,
+      "learning_rate": 0.0005942203682237422,
+      "loss": 3.4731,
+      "step": 3831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5598435401916504,
+      "learning_rate": 0.0005942173602654053,
+      "loss": 3.5913,
+      "step": 3832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6751794219017029,
+      "learning_rate": 0.0005942143515321563,
+      "loss": 3.576,
+      "step": 3833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5890194773674011,
+      "learning_rate": 0.0005942113420240034,
+      "loss": 3.6024,
+      "step": 3834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6194825768470764,
+      "learning_rate": 0.0005942083317409545,
+      "loss": 3.3712,
+      "step": 3835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5517776608467102,
+      "learning_rate": 0.0005942053206830175,
+      "loss": 3.4849,
+      "step": 3836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5492923855781555,
+      "learning_rate": 0.0005942023088502001,
+      "loss": 3.3793,
+      "step": 3837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5847030282020569,
+      "learning_rate": 0.0005941992962425107,
+      "loss": 3.6312,
+      "step": 3838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5900754928588867,
+      "learning_rate": 0.0005941962828599568,
+      "loss": 3.5081,
+      "step": 3839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5455581545829773,
+      "learning_rate": 0.0005941932687025467,
+      "loss": 3.4151,
+      "step": 3840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5705889463424683,
+      "learning_rate": 0.000594190253770288,
+      "loss": 3.6245,
+      "step": 3841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5608065724372864,
+      "learning_rate": 0.0005941872380631888,
+      "loss": 3.368,
+      "step": 3842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.579862117767334,
+      "learning_rate": 0.000594184221581257,
+      "loss": 3.5874,
+      "step": 3843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.570897102355957,
+      "learning_rate": 0.0005941812043245006,
+      "loss": 3.5095,
+      "step": 3844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5384157299995422,
+      "learning_rate": 0.0005941781862929275,
+      "loss": 3.6649,
+      "step": 3845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5707191824913025,
+      "learning_rate": 0.0005941751674865457,
+      "loss": 3.7034,
+      "step": 3846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5771855711936951,
+      "learning_rate": 0.0005941721479053632,
+      "loss": 3.5403,
+      "step": 3847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5326586961746216,
+      "learning_rate": 0.0005941691275493878,
+      "loss": 3.6345,
+      "step": 3848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6068933010101318,
+      "learning_rate": 0.0005941661064186276,
+      "loss": 3.5202,
+      "step": 3849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5824964642524719,
+      "learning_rate": 0.0005941630845130905,
+      "loss": 3.7016,
+      "step": 3850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5302342176437378,
+      "learning_rate": 0.0005941600618327843,
+      "loss": 3.5209,
+      "step": 3851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5523245334625244,
+      "learning_rate": 0.0005941570383777171,
+      "loss": 3.6873,
+      "step": 3852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5614722967147827,
+      "learning_rate": 0.0005941540141478971,
+      "loss": 3.5292,
+      "step": 3853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5354779958724976,
+      "learning_rate": 0.0005941509891433319,
+      "loss": 3.3006,
+      "step": 3854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6500388383865356,
+      "learning_rate": 0.0005941479633640296,
+      "loss": 3.5103,
+      "step": 3855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5664093494415283,
+      "learning_rate": 0.0005941449368099982,
+      "loss": 3.4756,
+      "step": 3856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5725468397140503,
+      "learning_rate": 0.0005941419094812456,
+      "loss": 3.4823,
+      "step": 3857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5943357348442078,
+      "learning_rate": 0.0005941388813777799,
+      "loss": 3.3334,
+      "step": 3858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6101706027984619,
+      "learning_rate": 0.000594135852499609,
+      "loss": 3.1905,
+      "step": 3859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5723959803581238,
+      "learning_rate": 0.0005941328228467408,
+      "loss": 3.4729,
+      "step": 3860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5894114971160889,
+      "learning_rate": 0.0005941297924191833,
+      "loss": 3.3462,
+      "step": 3861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5975579023361206,
+      "learning_rate": 0.0005941267612169447,
+      "loss": 3.4768,
+      "step": 3862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5968523621559143,
+      "learning_rate": 0.0005941237292400326,
+      "loss": 3.5171,
+      "step": 3863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5831944942474365,
+      "learning_rate": 0.0005941206964884554,
+      "loss": 3.2998,
+      "step": 3864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.581929624080658,
+      "learning_rate": 0.0005941176629622207,
+      "loss": 3.3597,
+      "step": 3865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.565826952457428,
+      "learning_rate": 0.0005941146286613368,
+      "loss": 3.4829,
+      "step": 3866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5366962552070618,
+      "learning_rate": 0.0005941115935858116,
+      "loss": 3.8651,
+      "step": 3867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5875154733657837,
+      "learning_rate": 0.000594108557735653,
+      "loss": 3.4664,
+      "step": 3868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5598399639129639,
+      "learning_rate": 0.0005941055211108691,
+      "loss": 3.6399,
+      "step": 3869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5694958567619324,
+      "learning_rate": 0.0005941024837114679,
+      "loss": 3.4511,
+      "step": 3870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5622062683105469,
+      "learning_rate": 0.0005940994455374574,
+      "loss": 3.4977,
+      "step": 3871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5713880062103271,
+      "learning_rate": 0.0005940964065888454,
+      "loss": 3.4234,
+      "step": 3872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5845184326171875,
+      "learning_rate": 0.0005940933668656402,
+      "loss": 3.4577,
+      "step": 3873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5796321034431458,
+      "learning_rate": 0.0005940903263678495,
+      "loss": 3.5307,
+      "step": 3874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5476338267326355,
+      "learning_rate": 0.0005940872850954816,
+      "loss": 3.4395,
+      "step": 3875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.579772412776947,
+      "learning_rate": 0.0005940842430485443,
+      "loss": 3.3787,
+      "step": 3876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5444009304046631,
+      "learning_rate": 0.0005940812002270458,
+      "loss": 3.6889,
+      "step": 3877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5588319301605225,
+      "learning_rate": 0.0005940781566309941,
+      "loss": 3.6946,
+      "step": 3878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5704074501991272,
+      "learning_rate": 0.0005940751122603971,
+      "loss": 3.5999,
+      "step": 3879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7025076746940613,
+      "learning_rate": 0.0005940720671152627,
+      "loss": 3.2713,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8407301306724548,
+      "learning_rate": 0.0005940690211955991,
+      "loss": 3.4089,
+      "step": 3881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5442054867744446,
+      "learning_rate": 0.0005940659745014143,
+      "loss": 3.5208,
+      "step": 3882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5735301375389099,
+      "learning_rate": 0.0005940629270327165,
+      "loss": 3.7339,
+      "step": 3883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5662720203399658,
+      "learning_rate": 0.0005940598787895132,
+      "loss": 3.5411,
+      "step": 3884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.635032057762146,
+      "learning_rate": 0.000594056829771813,
+      "loss": 3.6586,
+      "step": 3885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5698032379150391,
+      "learning_rate": 0.0005940537799796237,
+      "loss": 3.5767,
+      "step": 3886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5576255321502686,
+      "learning_rate": 0.0005940507294129534,
+      "loss": 3.6144,
+      "step": 3887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5743609070777893,
+      "learning_rate": 0.0005940476780718099,
+      "loss": 3.4329,
+      "step": 3888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5626585483551025,
+      "learning_rate": 0.0005940446259562015,
+      "loss": 3.62,
+      "step": 3889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5813789963722229,
+      "learning_rate": 0.0005940415730661361,
+      "loss": 3.4282,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6178430318832397,
+      "learning_rate": 0.0005940385194016219,
+      "loss": 3.612,
+      "step": 3891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.569376528263092,
+      "learning_rate": 0.0005940354649626667,
+      "loss": 3.4726,
+      "step": 3892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5537204742431641,
+      "learning_rate": 0.0005940324097492788,
+      "loss": 3.5413,
+      "step": 3893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5710381269454956,
+      "learning_rate": 0.000594029353761466,
+      "loss": 3.4135,
+      "step": 3894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6075487732887268,
+      "learning_rate": 0.0005940262969992365,
+      "loss": 3.3629,
+      "step": 3895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5630260705947876,
+      "learning_rate": 0.0005940232394625984,
+      "loss": 3.4259,
+      "step": 3896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5889552235603333,
+      "learning_rate": 0.0005940201811515595,
+      "loss": 3.4283,
+      "step": 3897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.593014657497406,
+      "learning_rate": 0.0005940171220661281,
+      "loss": 3.4631,
+      "step": 3898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5618011355400085,
+      "learning_rate": 0.0005940140622063123,
+      "loss": 3.4224,
+      "step": 3899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5981526970863342,
+      "learning_rate": 0.0005940110015721199,
+      "loss": 3.2224,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5683342814445496,
+      "learning_rate": 0.0005940079401635592,
+      "loss": 3.5997,
+      "step": 3901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5758498311042786,
+      "learning_rate": 0.0005940048779806381,
+      "loss": 3.498,
+      "step": 3902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.553376317024231,
+      "learning_rate": 0.0005940018150233647,
+      "loss": 3.613,
+      "step": 3903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5808395147323608,
+      "learning_rate": 0.0005939987512917472,
+      "loss": 3.5252,
+      "step": 3904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5820940732955933,
+      "learning_rate": 0.0005939956867857936,
+      "loss": 3.7314,
+      "step": 3905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5412302017211914,
+      "learning_rate": 0.0005939926215055119,
+      "loss": 3.5086,
+      "step": 3906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5737446546554565,
+      "learning_rate": 0.0005939895554509102,
+      "loss": 3.3745,
+      "step": 3907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5778170228004456,
+      "learning_rate": 0.0005939864886219966,
+      "loss": 3.6322,
+      "step": 3908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5643134713172913,
+      "learning_rate": 0.0005939834210187792,
+      "loss": 3.6112,
+      "step": 3909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5836457014083862,
+      "learning_rate": 0.0005939803526412658,
+      "loss": 3.6238,
+      "step": 3910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5769031047821045,
+      "learning_rate": 0.0005939772834894651,
+      "loss": 3.4665,
+      "step": 3911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5801408290863037,
+      "learning_rate": 0.0005939742135633846,
+      "loss": 3.1557,
+      "step": 3912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5858706831932068,
+      "learning_rate": 0.0005939711428630326,
+      "loss": 3.5962,
+      "step": 3913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6179161071777344,
+      "learning_rate": 0.0005939680713884173,
+      "loss": 3.5575,
+      "step": 3914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5316437482833862,
+      "learning_rate": 0.0005939649991395466,
+      "loss": 3.2445,
+      "step": 3915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5476315021514893,
+      "learning_rate": 0.0005939619261164286,
+      "loss": 3.5819,
+      "step": 3916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5380899906158447,
+      "learning_rate": 0.0005939588523190716,
+      "loss": 3.647,
+      "step": 3917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6630106568336487,
+      "learning_rate": 0.0005939557777474833,
+      "loss": 3.3215,
+      "step": 3918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5557861924171448,
+      "learning_rate": 0.0005939527024016723,
+      "loss": 3.5431,
+      "step": 3919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5520122647285461,
+      "learning_rate": 0.0005939496262816464,
+      "loss": 3.5015,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.593203604221344,
+      "learning_rate": 0.0005939465493874139,
+      "loss": 3.6008,
+      "step": 3921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5582143068313599,
+      "learning_rate": 0.0005939434717189825,
+      "loss": 3.3972,
+      "step": 3922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55299311876297,
+      "learning_rate": 0.0005939403932763607,
+      "loss": 3.5225,
+      "step": 3923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5243273377418518,
+      "learning_rate": 0.0005939373140595565,
+      "loss": 3.5162,
+      "step": 3924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5494229793548584,
+      "learning_rate": 0.0005939342340685778,
+      "loss": 3.4238,
+      "step": 3925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.56685471534729,
+      "learning_rate": 0.0005939311533034331,
+      "loss": 3.3776,
+      "step": 3926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5525105595588684,
+      "learning_rate": 0.0005939280717641302,
+      "loss": 3.5084,
+      "step": 3927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6117692589759827,
+      "learning_rate": 0.0005939249894506773,
+      "loss": 3.4447,
+      "step": 3928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6117774844169617,
+      "learning_rate": 0.0005939219063630826,
+      "loss": 3.5056,
+      "step": 3929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5624076724052429,
+      "learning_rate": 0.0005939188225013543,
+      "loss": 3.4238,
+      "step": 3930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.591431736946106,
+      "learning_rate": 0.0005939157378655002,
+      "loss": 3.4953,
+      "step": 3931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5721966624259949,
+      "learning_rate": 0.0005939126524555287,
+      "loss": 3.4561,
+      "step": 3932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5344234108924866,
+      "learning_rate": 0.0005939095662714479,
+      "loss": 3.4767,
+      "step": 3933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6061483025550842,
+      "learning_rate": 0.0005939064793132658,
+      "loss": 3.6087,
+      "step": 3934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5591382384300232,
+      "learning_rate": 0.0005939033915809905,
+      "loss": 3.5391,
+      "step": 3935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5948025584220886,
+      "learning_rate": 0.0005939003030746302,
+      "loss": 3.5729,
+      "step": 3936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6076129674911499,
+      "learning_rate": 0.0005938972137941933,
+      "loss": 3.4074,
+      "step": 3937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5574817061424255,
+      "learning_rate": 0.0005938941237396876,
+      "loss": 3.371,
+      "step": 3938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5657892823219299,
+      "learning_rate": 0.0005938910329111213,
+      "loss": 3.2757,
+      "step": 3939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5320510864257812,
+      "learning_rate": 0.0005938879413085026,
+      "loss": 3.5539,
+      "step": 3940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.594287097454071,
+      "learning_rate": 0.0005938848489318396,
+      "loss": 3.3186,
+      "step": 3941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5167534351348877,
+      "learning_rate": 0.0005938817557811405,
+      "loss": 3.4876,
+      "step": 3942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5355756282806396,
+      "learning_rate": 0.0005938786618564134,
+      "loss": 3.6319,
+      "step": 3943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5719373226165771,
+      "learning_rate": 0.0005938755671576664,
+      "loss": 3.6176,
+      "step": 3944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5378432869911194,
+      "learning_rate": 0.0005938724716849078,
+      "loss": 3.2131,
+      "step": 3945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5518651008605957,
+      "learning_rate": 0.0005938693754381456,
+      "loss": 3.3925,
+      "step": 3946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5597622990608215,
+      "learning_rate": 0.0005938662784173881,
+      "loss": 3.436,
+      "step": 3947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5614737868309021,
+      "learning_rate": 0.0005938631806226434,
+      "loss": 3.696,
+      "step": 3948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5528543591499329,
+      "learning_rate": 0.0005938600820539195,
+      "loss": 3.4416,
+      "step": 3949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.588862955570221,
+      "learning_rate": 0.0005938569827112247,
+      "loss": 3.5648,
+      "step": 3950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.590054452419281,
+      "learning_rate": 0.0005938538825945673,
+      "loss": 3.3912,
+      "step": 3951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5708290934562683,
+      "learning_rate": 0.0005938507817039553,
+      "loss": 3.5998,
+      "step": 3952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5861159563064575,
+      "learning_rate": 0.0005938476800393968,
+      "loss": 3.4476,
+      "step": 3953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5528053641319275,
+      "learning_rate": 0.0005938445776009002,
+      "loss": 3.4957,
+      "step": 3954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5492965579032898,
+      "learning_rate": 0.0005938414743884734,
+      "loss": 3.3417,
+      "step": 3955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.575512707233429,
+      "learning_rate": 0.0005938383704021247,
+      "loss": 3.398,
+      "step": 3956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5666360855102539,
+      "learning_rate": 0.0005938352656418624,
+      "loss": 3.5014,
+      "step": 3957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6097762584686279,
+      "learning_rate": 0.0005938321601076944,
+      "loss": 3.5334,
+      "step": 3958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5508065819740295,
+      "learning_rate": 0.0005938290537996292,
+      "loss": 3.4517,
+      "step": 3959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5451350808143616,
+      "learning_rate": 0.0005938259467176748,
+      "loss": 3.3272,
+      "step": 3960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5097694993019104,
+      "learning_rate": 0.0005938228388618393,
+      "loss": 3.3928,
+      "step": 3961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5869589447975159,
+      "learning_rate": 0.0005938197302321312,
+      "loss": 3.5485,
+      "step": 3962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5630464553833008,
+      "learning_rate": 0.0005938166208285582,
+      "loss": 3.41,
+      "step": 3963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6036396622657776,
+      "learning_rate": 0.0005938135106511289,
+      "loss": 3.6272,
+      "step": 3964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6252908706665039,
+      "learning_rate": 0.0005938103996998514,
+      "loss": 3.3919,
+      "step": 3965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5575089454650879,
+      "learning_rate": 0.0005938072879747338,
+      "loss": 3.47,
+      "step": 3966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5552845597267151,
+      "learning_rate": 0.0005938041754757844,
+      "loss": 3.384,
+      "step": 3967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5382720828056335,
+      "learning_rate": 0.0005938010622030112,
+      "loss": 3.4518,
+      "step": 3968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5792542695999146,
+      "learning_rate": 0.0005937979481564227,
+      "loss": 3.3024,
+      "step": 3969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.574998676776886,
+      "learning_rate": 0.000593794833336027,
+      "loss": 3.4806,
+      "step": 3970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6154451966285706,
+      "learning_rate": 0.000593791717741832,
+      "loss": 3.3694,
+      "step": 3971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5885059237480164,
+      "learning_rate": 0.0005937886013738464,
+      "loss": 3.3851,
+      "step": 3972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5663660168647766,
+      "learning_rate": 0.000593785484232078,
+      "loss": 3.5687,
+      "step": 3973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5735329389572144,
+      "learning_rate": 0.0005937823663165352,
+      "loss": 3.3934,
+      "step": 3974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6525985598564148,
+      "learning_rate": 0.0005937792476272262,
+      "loss": 3.328,
+      "step": 3975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.586041271686554,
+      "learning_rate": 0.0005937761281641594,
+      "loss": 3.4332,
+      "step": 3976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.563795804977417,
+      "learning_rate": 0.0005937730079273426,
+      "loss": 3.6729,
+      "step": 3977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5492075681686401,
+      "learning_rate": 0.0005937698869167842,
+      "loss": 3.4315,
+      "step": 3978
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6151273846626282,
+      "learning_rate": 0.0005937667651324925,
+      "loss": 3.7117,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5914956331253052,
+      "learning_rate": 0.0005937636425744757,
+      "loss": 3.4685,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5682784914970398,
+      "learning_rate": 0.0005937605192427422,
+      "loss": 3.2269,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5962632298469543,
+      "learning_rate": 0.0005937573951372997,
+      "loss": 3.4843,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6012601852416992,
+      "learning_rate": 0.0005937542702581569,
+      "loss": 3.4149,
+      "step": 3983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.556812047958374,
+      "learning_rate": 0.0005937511446053219,
+      "loss": 3.5567,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6080185174942017,
+      "learning_rate": 0.0005937480181788029,
+      "loss": 3.5601,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6003567576408386,
+      "learning_rate": 0.0005937448909786083,
+      "loss": 3.5517,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.575144350528717,
+      "learning_rate": 0.0005937417630047459,
+      "loss": 3.415,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5906019806861877,
+      "learning_rate": 0.0005937386342572244,
+      "loss": 3.6005,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.564968466758728,
+      "learning_rate": 0.0005937355047360519,
+      "loss": 3.4319,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5609432458877563,
+      "learning_rate": 0.0005937323744412365,
+      "loss": 3.3004,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5350099205970764,
+      "learning_rate": 0.0005937292433727866,
+      "loss": 3.5976,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5905284881591797,
+      "learning_rate": 0.0005937261115307104,
+      "loss": 3.5769,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5806431770324707,
+      "learning_rate": 0.0005937229789150162,
+      "loss": 3.6226,
+      "step": 3993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5419069528579712,
+      "learning_rate": 0.0005937198455257122,
+      "loss": 3.8644,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5670180916786194,
+      "learning_rate": 0.0005937167113628067,
+      "loss": 3.4316,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5351924896240234,
+      "learning_rate": 0.0005937135764263077,
+      "loss": 3.5708,
+      "step": 3996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5488182306289673,
+      "learning_rate": 0.0005937104407162238,
+      "loss": 3.5572,
+      "step": 3997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5685787200927734,
+      "learning_rate": 0.0005937073042325631,
+      "loss": 3.4464,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5353884696960449,
+      "learning_rate": 0.000593704166975334,
+      "loss": 3.4783,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6061869859695435,
+      "learning_rate": 0.0005937010289445444,
+      "loss": 3.2644,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.571092426776886,
+      "learning_rate": 0.000593697890140203,
+      "loss": 3.5456,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6306237578392029,
+      "learning_rate": 0.0005936947505623179,
+      "loss": 3.3862,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5207166075706482,
+      "learning_rate": 0.0005936916102108974,
+      "loss": 3.3985,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6106595396995544,
+      "learning_rate": 0.0005936884690859495,
+      "loss": 3.6107,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6063200235366821,
+      "learning_rate": 0.0005936853271874829,
+      "loss": 3.5122,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5411254167556763,
+      "learning_rate": 0.0005936821845155055,
+      "loss": 3.4418,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5724605321884155,
+      "learning_rate": 0.0005936790410700259,
+      "loss": 3.455,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5903576016426086,
+      "learning_rate": 0.0005936758968510521,
+      "loss": 3.4542,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5617760419845581,
+      "learning_rate": 0.0005936727518585925,
+      "loss": 3.5287,
+      "step": 4009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57075035572052,
+      "learning_rate": 0.0005936696060926555,
+      "loss": 3.54,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5305947065353394,
+      "learning_rate": 0.0005936664595532493,
+      "loss": 3.7036,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5543831586837769,
+      "learning_rate": 0.000593663312240382,
+      "loss": 3.4049,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5600700974464417,
+      "learning_rate": 0.0005936601641540622,
+      "loss": 3.8008,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5779480934143066,
+      "learning_rate": 0.0005936570152942979,
+      "loss": 3.7728,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5971601009368896,
+      "learning_rate": 0.0005936538656610977,
+      "loss": 3.4096,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5580905675888062,
+      "learning_rate": 0.0005936507152544695,
+      "loss": 3.6679,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6468621492385864,
+      "learning_rate": 0.0005936475640744221,
+      "loss": 3.6336,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6058899164199829,
+      "learning_rate": 0.0005936444121209634,
+      "loss": 3.4921,
+      "step": 4018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5818850994110107,
+      "learning_rate": 0.0005936412593941019,
+      "loss": 3.4677,
+      "step": 4019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5825283527374268,
+      "learning_rate": 0.0005936381058938457,
+      "loss": 3.4282,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5536349415779114,
+      "learning_rate": 0.0005936349516202034,
+      "loss": 3.3787,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.664323091506958,
+      "learning_rate": 0.000593631796573183,
+      "loss": 3.4861,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5751873254776001,
+      "learning_rate": 0.0005936286407527931,
+      "loss": 3.5405,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5620585083961487,
+      "learning_rate": 0.0005936254841590417,
+      "loss": 3.2407,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5601255893707275,
+      "learning_rate": 0.0005936223267919374,
+      "loss": 3.3897,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6205532550811768,
+      "learning_rate": 0.0005936191686514884,
+      "loss": 3.5952,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5451340675354004,
+      "learning_rate": 0.000593616009737703,
+      "loss": 3.3103,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5882269144058228,
+      "learning_rate": 0.0005936128500505896,
+      "loss": 3.274,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5673192739486694,
+      "learning_rate": 0.0005936096895901564,
+      "loss": 3.5805,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5817726850509644,
+      "learning_rate": 0.0005936065283564117,
+      "loss": 3.4446,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6373004913330078,
+      "learning_rate": 0.0005936033663493641,
+      "loss": 3.662,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5789350271224976,
+      "learning_rate": 0.0005936002035690216,
+      "loss": 3.4954,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5490628480911255,
+      "learning_rate": 0.0005935970400153927,
+      "loss": 3.4452,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.611883282661438,
+      "learning_rate": 0.0005935938756884857,
+      "loss": 3.5924,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5716962814331055,
+      "learning_rate": 0.000593590710588309,
+      "loss": 3.4642,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5935686230659485,
+      "learning_rate": 0.0005935875447148707,
+      "loss": 3.3395,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5877057909965515,
+      "learning_rate": 0.0005935843780681795,
+      "loss": 3.3124,
+      "step": 4037
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6173941493034363,
+      "learning_rate": 0.0005935812106482433,
+      "loss": 3.606,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5606823563575745,
+      "learning_rate": 0.0005935780424550709,
+      "loss": 3.6931,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6052460074424744,
+      "learning_rate": 0.0005935748734886704,
+      "loss": 3.4013,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.586379885673523,
+      "learning_rate": 0.0005935717037490501,
+      "loss": 3.4105,
+      "step": 4041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5679966807365417,
+      "learning_rate": 0.0005935685332362186,
+      "loss": 3.5497,
+      "step": 4042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5713095664978027,
+      "learning_rate": 0.000593565361950184,
+      "loss": 3.4705,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5471493005752563,
+      "learning_rate": 0.0005935621898909547,
+      "loss": 3.4439,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5692678093910217,
+      "learning_rate": 0.0005935590170585391,
+      "loss": 3.4251,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5359514951705933,
+      "learning_rate": 0.0005935558434529456,
+      "loss": 3.705,
+      "step": 4046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6026199460029602,
+      "learning_rate": 0.0005935526690741824,
+      "loss": 3.5045,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5974529385566711,
+      "learning_rate": 0.0005935494939222581,
+      "loss": 3.3607,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.583665132522583,
+      "learning_rate": 0.0005935463179971808,
+      "loss": 3.7261,
+      "step": 4049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5661903619766235,
+      "learning_rate": 0.0005935431412989591,
+      "loss": 3.591,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5681086778640747,
+      "learning_rate": 0.0005935399638276012,
+      "loss": 3.5113,
+      "step": 4051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5938650965690613,
+      "learning_rate": 0.0005935367855831155,
+      "loss": 3.4974,
+      "step": 4052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5689106583595276,
+      "learning_rate": 0.0005935336065655104,
+      "loss": 3.4858,
+      "step": 4053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.552765965461731,
+      "learning_rate": 0.0005935304267747943,
+      "loss": 3.6311,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5879870057106018,
+      "learning_rate": 0.0005935272462109756,
+      "loss": 3.4236,
+      "step": 4055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5558753609657288,
+      "learning_rate": 0.0005935240648740625,
+      "loss": 3.447,
+      "step": 4056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5875343084335327,
+      "learning_rate": 0.0005935208827640636,
+      "loss": 3.5598,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.580474853515625,
+      "learning_rate": 0.0005935176998809873,
+      "loss": 3.5608,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5907713770866394,
+      "learning_rate": 0.0005935145162248416,
+      "loss": 3.2178,
+      "step": 4059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5516918301582336,
+      "learning_rate": 0.0005935113317956353,
+      "loss": 3.4183,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5646939277648926,
+      "learning_rate": 0.0005935081465933767,
+      "loss": 3.4653,
+      "step": 4061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5440240502357483,
+      "learning_rate": 0.000593504960618074,
+      "loss": 3.7186,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5458465218544006,
+      "learning_rate": 0.0005935017738697357,
+      "loss": 3.4928,
+      "step": 4063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6689546704292297,
+      "learning_rate": 0.0005934985863483704,
+      "loss": 3.3547,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6123103499412537,
+      "learning_rate": 0.0005934953980539861,
+      "loss": 3.375,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5664320588111877,
+      "learning_rate": 0.0005934922089865915,
+      "loss": 3.449,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6381657719612122,
+      "learning_rate": 0.0005934890191461949,
+      "loss": 3.4441,
+      "step": 4067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5956173539161682,
+      "learning_rate": 0.0005934858285328048,
+      "loss": 3.6867,
+      "step": 4068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6595749855041504,
+      "learning_rate": 0.0005934826371464294,
+      "loss": 3.3421,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5597966313362122,
+      "learning_rate": 0.0005934794449870772,
+      "loss": 3.5673,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5948126912117004,
+      "learning_rate": 0.0005934762520547566,
+      "loss": 3.7263,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5999248027801514,
+      "learning_rate": 0.0005934730583494761,
+      "loss": 3.4787,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6411767601966858,
+      "learning_rate": 0.0005934698638712441,
+      "loss": 3.4928,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5557070374488831,
+      "learning_rate": 0.0005934666686200689,
+      "loss": 3.5091,
+      "step": 4074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5832461714744568,
+      "learning_rate": 0.0005934634725959589,
+      "loss": 3.5301,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5976592302322388,
+      "learning_rate": 0.0005934602757989227,
+      "loss": 3.4888,
+      "step": 4076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5408334732055664,
+      "learning_rate": 0.0005934570782289686,
+      "loss": 3.6005,
+      "step": 4077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5419313311576843,
+      "learning_rate": 0.000593453879886105,
+      "loss": 3.5205,
+      "step": 4078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6286296844482422,
+      "learning_rate": 0.0005934506807703403,
+      "loss": 3.4126,
+      "step": 4079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5880137085914612,
+      "learning_rate": 0.0005934474808816831,
+      "loss": 3.5638,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5868574976921082,
+      "learning_rate": 0.0005934442802201416,
+      "loss": 3.4796,
+      "step": 4081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5653226375579834,
+      "learning_rate": 0.0005934410787857245,
+      "loss": 3.0373,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5920324921607971,
+      "learning_rate": 0.0005934378765784399,
+      "loss": 3.5852,
+      "step": 4083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5790802836418152,
+      "learning_rate": 0.0005934346735982964,
+      "loss": 3.6043,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5777429938316345,
+      "learning_rate": 0.0005934314698453026,
+      "loss": 3.3492,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.569974422454834,
+      "learning_rate": 0.0005934282653194667,
+      "loss": 3.4926,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6177379488945007,
+      "learning_rate": 0.0005934250600207972,
+      "loss": 3.6035,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5291566848754883,
+      "learning_rate": 0.0005934218539493026,
+      "loss": 3.5606,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5401890873908997,
+      "learning_rate": 0.0005934186471049913,
+      "loss": 3.4682,
+      "step": 4089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5783869028091431,
+      "learning_rate": 0.0005934154394878717,
+      "loss": 3.4999,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5341452360153198,
+      "learning_rate": 0.0005934122310979524,
+      "loss": 3.4623,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5385696887969971,
+      "learning_rate": 0.0005934090219352416,
+      "loss": 3.6001,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6240125298500061,
+      "learning_rate": 0.000593405811999748,
+      "loss": 3.3475,
+      "step": 4093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5452415347099304,
+      "learning_rate": 0.00059340260129148,
+      "loss": 3.4982,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5494929552078247,
+      "learning_rate": 0.000593399389810446,
+      "loss": 3.4822,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55720055103302,
+      "learning_rate": 0.0005933961775566543,
+      "loss": 3.4194,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5275405049324036,
+      "learning_rate": 0.0005933929645301138,
+      "loss": 3.4621,
+      "step": 4097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.559727668762207,
+      "learning_rate": 0.0005933897507308325,
+      "loss": 3.5763,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5319221019744873,
+      "learning_rate": 0.0005933865361588192,
+      "loss": 3.3231,
+      "step": 4099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5701916217803955,
+      "learning_rate": 0.0005933833208140822,
+      "loss": 3.5887,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6000038981437683,
+      "learning_rate": 0.0005933801046966299,
+      "loss": 3.5659,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5542453527450562,
+      "learning_rate": 0.0005933768878064709,
+      "loss": 3.6189,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546266436576843,
+      "learning_rate": 0.0005933736701436137,
+      "loss": 3.4361,
+      "step": 4103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5334687232971191,
+      "learning_rate": 0.0005933704517080666,
+      "loss": 3.5126,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5554494857788086,
+      "learning_rate": 0.0005933672324998382,
+      "loss": 3.291,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5818085670471191,
+      "learning_rate": 0.0005933640125189371,
+      "loss": 3.2228,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5691803097724915,
+      "learning_rate": 0.0005933607917653715,
+      "loss": 3.4671,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5687941908836365,
+      "learning_rate": 0.0005933575702391501,
+      "loss": 3.3455,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5801530480384827,
+      "learning_rate": 0.0005933543479402814,
+      "loss": 3.5252,
+      "step": 4109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5499988198280334,
+      "learning_rate": 0.0005933511248687738,
+      "loss": 3.4909,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5364803671836853,
+      "learning_rate": 0.0005933479010246357,
+      "loss": 3.6644,
+      "step": 4111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5526125431060791,
+      "learning_rate": 0.0005933446764078758,
+      "loss": 3.3869,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5705166459083557,
+      "learning_rate": 0.0005933414510185023,
+      "loss": 3.4363,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6104483008384705,
+      "learning_rate": 0.0005933382248565239,
+      "loss": 3.4521,
+      "step": 4114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6703182458877563,
+      "learning_rate": 0.0005933349979219492,
+      "loss": 3.538,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5792509913444519,
+      "learning_rate": 0.0005933317702147865,
+      "loss": 3.4318,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5312807559967041,
+      "learning_rate": 0.0005933285417350444,
+      "loss": 3.477,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5642654895782471,
+      "learning_rate": 0.0005933253124827313,
+      "loss": 3.5236,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5508056879043579,
+      "learning_rate": 0.0005933220824578559,
+      "loss": 3.6048,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5829192399978638,
+      "learning_rate": 0.0005933188516604264,
+      "loss": 3.4721,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5596129298210144,
+      "learning_rate": 0.0005933156200904516,
+      "loss": 3.2922,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5868514776229858,
+      "learning_rate": 0.00059331238774794,
+      "loss": 3.4381,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5288069248199463,
+      "learning_rate": 0.0005933091546328999,
+      "loss": 3.3479,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5082468390464783,
+      "learning_rate": 0.00059330592074534,
+      "loss": 3.5534,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5732336640357971,
+      "learning_rate": 0.0005933026860852688,
+      "loss": 3.3907,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5370706915855408,
+      "learning_rate": 0.0005932994506526947,
+      "loss": 3.6477,
+      "step": 4126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5239009857177734,
+      "learning_rate": 0.0005932962144476262,
+      "loss": 3.6473,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.566159725189209,
+      "learning_rate": 0.000593292977470072,
+      "loss": 3.687,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5577471256256104,
+      "learning_rate": 0.0005932897397200405,
+      "loss": 3.5333,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6108332276344299,
+      "learning_rate": 0.0005932865011975403,
+      "loss": 3.4125,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5488579869270325,
+      "learning_rate": 0.00059328326190258,
+      "loss": 3.4491,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.522066056728363,
+      "learning_rate": 0.000593280021835168,
+      "loss": 3.4321,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5913230776786804,
+      "learning_rate": 0.0005932767809953128,
+      "loss": 3.6254,
+      "step": 4133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5286851525306702,
+      "learning_rate": 0.0005932735393830229,
+      "loss": 3.493,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5211114883422852,
+      "learning_rate": 0.0005932702969983071,
+      "loss": 3.7179,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5448617339134216,
+      "learning_rate": 0.0005932670538411737,
+      "loss": 3.5312,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5416504740715027,
+      "learning_rate": 0.0005932638099116314,
+      "loss": 3.523,
+      "step": 4137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5332238078117371,
+      "learning_rate": 0.0005932605652096887,
+      "loss": 3.5286,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5574766993522644,
+      "learning_rate": 0.0005932573197353539,
+      "loss": 3.2139,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5327970385551453,
+      "learning_rate": 0.000593254073488636,
+      "loss": 3.4693,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5722629427909851,
+      "learning_rate": 0.0005932508264695431,
+      "loss": 3.429,
+      "step": 4141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5846360921859741,
+      "learning_rate": 0.0005932475786780841,
+      "loss": 3.4747,
+      "step": 4142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5634985566139221,
+      "learning_rate": 0.0005932443301142673,
+      "loss": 3.4314,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5535609722137451,
+      "learning_rate": 0.0005932410807781015,
+      "loss": 3.4342,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5832380652427673,
+      "learning_rate": 0.0005932378306695951,
+      "loss": 3.2958,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5615767240524292,
+      "learning_rate": 0.0005932345797887566,
+      "loss": 3.4979,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5781761407852173,
+      "learning_rate": 0.0005932313281355947,
+      "loss": 3.5425,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5604632496833801,
+      "learning_rate": 0.000593228075710118,
+      "loss": 3.5805,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6127885580062866,
+      "learning_rate": 0.0005932248225123349,
+      "loss": 3.4723,
+      "step": 4149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5799558162689209,
+      "learning_rate": 0.000593221568542254,
+      "loss": 3.4249,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5683193802833557,
+      "learning_rate": 0.0005932183137998839,
+      "loss": 3.5735,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5393085479736328,
+      "learning_rate": 0.0005932150582852333,
+      "loss": 3.3668,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6012470126152039,
+      "learning_rate": 0.0005932118019983107,
+      "loss": 3.3904,
+      "step": 4153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5704240798950195,
+      "learning_rate": 0.0005932085449391245,
+      "loss": 3.4991,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6046105623245239,
+      "learning_rate": 0.0005932052871076835,
+      "loss": 3.531,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5393555760383606,
+      "learning_rate": 0.0005932020285039963,
+      "loss": 3.6832,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5603122115135193,
+      "learning_rate": 0.0005931987691280712,
+      "loss": 3.5048,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5598936676979065,
+      "learning_rate": 0.0005931955089799172,
+      "loss": 3.4841,
+      "step": 4158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6121480464935303,
+      "learning_rate": 0.0005931922480595424,
+      "loss": 3.4819,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5565993189811707,
+      "learning_rate": 0.0005931889863669558,
+      "loss": 3.5906,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5580215454101562,
+      "learning_rate": 0.0005931857239021657,
+      "loss": 3.4656,
+      "step": 4161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.604782223701477,
+      "learning_rate": 0.000593182460665181,
+      "loss": 3.3578,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5579813122749329,
+      "learning_rate": 0.00059317919665601,
+      "loss": 3.5514,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5879456400871277,
+      "learning_rate": 0.0005931759318746615,
+      "loss": 3.5933,
+      "step": 4164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5648056268692017,
+      "learning_rate": 0.000593172666321144,
+      "loss": 3.4556,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6041473150253296,
+      "learning_rate": 0.0005931693999954661,
+      "loss": 3.6027,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5803984999656677,
+      "learning_rate": 0.0005931661328976365,
+      "loss": 3.3593,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5941168665885925,
+      "learning_rate": 0.0005931628650276636,
+      "loss": 3.4784,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5952200293540955,
+      "learning_rate": 0.0005931595963855562,
+      "loss": 3.4204,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5529075860977173,
+      "learning_rate": 0.0005931563269713228,
+      "loss": 3.5687,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5690220594406128,
+      "learning_rate": 0.000593153056784972,
+      "loss": 3.5338,
+      "step": 4171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6204742789268494,
+      "learning_rate": 0.0005931497858265126,
+      "loss": 3.4925,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6174020171165466,
+      "learning_rate": 0.000593146514095953,
+      "loss": 3.4493,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5436307191848755,
+      "learning_rate": 0.000593143241593302,
+      "loss": 3.5383,
+      "step": 4174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5812988877296448,
+      "learning_rate": 0.0005931399683185679,
+      "loss": 3.265,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6355542540550232,
+      "learning_rate": 0.0005931366942717597,
+      "loss": 3.4879,
+      "step": 4176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5769966840744019,
+      "learning_rate": 0.0005931334194528859,
+      "loss": 3.4083,
+      "step": 4177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5602624416351318,
+      "learning_rate": 0.000593130143861955,
+      "loss": 3.4455,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5522497892379761,
+      "learning_rate": 0.0005931268674989758,
+      "loss": 3.589,
+      "step": 4179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6143437623977661,
+      "learning_rate": 0.0005931235903639568,
+      "loss": 3.39,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5582066774368286,
+      "learning_rate": 0.0005931203124569066,
+      "loss": 3.2465,
+      "step": 4181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5512022972106934,
+      "learning_rate": 0.000593117033777834,
+      "loss": 3.3506,
+      "step": 4182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5688873529434204,
+      "learning_rate": 0.0005931137543267476,
+      "loss": 3.6862,
+      "step": 4183
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5612702369689941,
+      "learning_rate": 0.0005931104741036558,
+      "loss": 3.4739,
+      "step": 4184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5855376720428467,
+      "learning_rate": 0.0005931071931085676,
+      "loss": 3.4008,
+      "step": 4185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5829062461853027,
+      "learning_rate": 0.0005931039113414914,
+      "loss": 3.3817,
+      "step": 4186
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5704607963562012,
+      "learning_rate": 0.0005931006288024358,
+      "loss": 3.5417,
+      "step": 4187
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5722986459732056,
+      "learning_rate": 0.0005930973454914097,
+      "loss": 3.6125,
+      "step": 4188
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5302977561950684,
+      "learning_rate": 0.0005930940614084216,
+      "loss": 3.2082,
+      "step": 4189
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5311842560768127,
+      "learning_rate": 0.00059309077655348,
+      "loss": 3.4686,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5523838996887207,
+      "learning_rate": 0.0005930874909265938,
+      "loss": 3.525,
+      "step": 4191
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6022278070449829,
+      "learning_rate": 0.0005930842045277716,
+      "loss": 3.3398,
+      "step": 4192
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.588112473487854,
+      "learning_rate": 0.000593080917357022,
+      "loss": 3.4084,
+      "step": 4193
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5542445182800293,
+      "learning_rate": 0.0005930776294143536,
+      "loss": 3.2733,
+      "step": 4194
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6074942350387573,
+      "learning_rate": 0.0005930743406997752,
+      "loss": 3.5434,
+      "step": 4195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5761960744857788,
+      "learning_rate": 0.0005930710512132954,
+      "loss": 3.6245,
+      "step": 4196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5809390544891357,
+      "learning_rate": 0.0005930677609549227,
+      "loss": 3.5375,
+      "step": 4197
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.572761595249176,
+      "learning_rate": 0.000593064469924666,
+      "loss": 3.5079,
+      "step": 4198
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5556721687316895,
+      "learning_rate": 0.000593061178122534,
+      "loss": 3.4619,
+      "step": 4199
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5726044178009033,
+      "learning_rate": 0.0005930578855485351,
+      "loss": 3.3863,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5410187840461731,
+      "learning_rate": 0.0005930545922026783,
+      "loss": 3.3838,
+      "step": 4201
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5719176530838013,
+      "learning_rate": 0.0005930512980849719,
+      "loss": 3.6133,
+      "step": 4202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5681977272033691,
+      "learning_rate": 0.0005930480031954249,
+      "loss": 3.5757,
+      "step": 4203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5701371431350708,
+      "learning_rate": 0.0005930447075340458,
+      "loss": 3.2612,
+      "step": 4204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8378312587738037,
+      "learning_rate": 0.0005930414111008435,
+      "loss": 3.5506,
+      "step": 4205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6150882840156555,
+      "learning_rate": 0.0005930381138958263,
+      "loss": 3.4457,
+      "step": 4206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5677423477172852,
+      "learning_rate": 0.0005930348159190031,
+      "loss": 3.3982,
+      "step": 4207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6335375308990479,
+      "learning_rate": 0.0005930315171703827,
+      "loss": 3.3809,
+      "step": 4208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5192340612411499,
+      "learning_rate": 0.0005930282176499738,
+      "loss": 3.4675,
+      "step": 4209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5766927599906921,
+      "learning_rate": 0.0005930249173577848,
+      "loss": 3.5231,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5748723149299622,
+      "learning_rate": 0.0005930216162938246,
+      "loss": 3.4761,
+      "step": 4211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.549691379070282,
+      "learning_rate": 0.0005930183144581019,
+      "loss": 3.5352,
+      "step": 4212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5617498755455017,
+      "learning_rate": 0.0005930150118506253,
+      "loss": 3.5251,
+      "step": 4213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5673946738243103,
+      "learning_rate": 0.0005930117084714036,
+      "loss": 3.4009,
+      "step": 4214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5976900458335876,
+      "learning_rate": 0.0005930084043204454,
+      "loss": 3.5733,
+      "step": 4215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5517022013664246,
+      "learning_rate": 0.0005930050993977594,
+      "loss": 3.4349,
+      "step": 4216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6215957999229431,
+      "learning_rate": 0.0005930017937033545,
+      "loss": 3.7329,
+      "step": 4217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5812935829162598,
+      "learning_rate": 0.0005929984872372391,
+      "loss": 3.5745,
+      "step": 4218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5533018112182617,
+      "learning_rate": 0.0005929951799994222,
+      "loss": 3.5439,
+      "step": 4219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.556189775466919,
+      "learning_rate": 0.0005929918719899123,
+      "loss": 3.6134,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5521323084831238,
+      "learning_rate": 0.0005929885632087183,
+      "loss": 3.3643,
+      "step": 4221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6353476643562317,
+      "learning_rate": 0.0005929852536558487,
+      "loss": 3.2529,
+      "step": 4222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6318731904029846,
+      "learning_rate": 0.0005929819433313124,
+      "loss": 3.5494,
+      "step": 4223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5776653289794922,
+      "learning_rate": 0.0005929786322351181,
+      "loss": 3.4537,
+      "step": 4224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5921760201454163,
+      "learning_rate": 0.0005929753203672743,
+      "loss": 3.5448,
+      "step": 4225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5609504580497742,
+      "learning_rate": 0.00059297200772779,
+      "loss": 3.4304,
+      "step": 4226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5797039866447449,
+      "learning_rate": 0.0005929686943166738,
+      "loss": 3.4262,
+      "step": 4227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.578444242477417,
+      "learning_rate": 0.0005929653801339344,
+      "loss": 3.2239,
+      "step": 4228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5497398376464844,
+      "learning_rate": 0.0005929620651795806,
+      "loss": 3.4644,
+      "step": 4229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5600916743278503,
+      "learning_rate": 0.0005929587494536212,
+      "loss": 3.4196,
+      "step": 4230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5911729335784912,
+      "learning_rate": 0.0005929554329560647,
+      "loss": 3.2735,
+      "step": 4231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5636935234069824,
+      "learning_rate": 0.00059295211568692,
+      "loss": 3.4634,
+      "step": 4232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5678104758262634,
+      "learning_rate": 0.0005929487976461959,
+      "loss": 3.5156,
+      "step": 4233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5603824853897095,
+      "learning_rate": 0.0005929454788339009,
+      "loss": 3.3868,
+      "step": 4234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5836139917373657,
+      "learning_rate": 0.000592942159250044,
+      "loss": 3.5746,
+      "step": 4235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7062385678291321,
+      "learning_rate": 0.0005929388388946338,
+      "loss": 3.3934,
+      "step": 4236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5844118595123291,
+      "learning_rate": 0.000592935517767679,
+      "loss": 3.494,
+      "step": 4237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5271828174591064,
+      "learning_rate": 0.0005929321958691885,
+      "loss": 3.4183,
+      "step": 4238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.528325617313385,
+      "learning_rate": 0.000592928873199171,
+      "loss": 3.3322,
+      "step": 4239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5398619174957275,
+      "learning_rate": 0.0005929255497576353,
+      "loss": 3.4529,
+      "step": 4240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5449317693710327,
+      "learning_rate": 0.0005929222255445899,
+      "loss": 3.6722,
+      "step": 4241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5887119770050049,
+      "learning_rate": 0.0005929189005600438,
+      "loss": 3.5835,
+      "step": 4242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5551977753639221,
+      "learning_rate": 0.0005929155748040057,
+      "loss": 3.3044,
+      "step": 4243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5693026781082153,
+      "learning_rate": 0.0005929122482764844,
+      "loss": 3.3308,
+      "step": 4244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6203973293304443,
+      "learning_rate": 0.0005929089209774883,
+      "loss": 3.5217,
+      "step": 4245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6212824583053589,
+      "learning_rate": 0.0005929055929070269,
+      "loss": 3.4606,
+      "step": 4246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5740115642547607,
+      "learning_rate": 0.0005929022640651083,
+      "loss": 3.4078,
+      "step": 4247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5441243052482605,
+      "learning_rate": 0.0005928989344517415,
+      "loss": 3.4294,
+      "step": 4248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6401486396789551,
+      "learning_rate": 0.0005928956040669353,
+      "loss": 3.4507,
+      "step": 4249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5190750360488892,
+      "learning_rate": 0.0005928922729106986,
+      "loss": 3.4062,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5887166857719421,
+      "learning_rate": 0.0005928889409830398,
+      "loss": 3.5247,
+      "step": 4251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5587601065635681,
+      "learning_rate": 0.0005928856082839681,
+      "loss": 3.462,
+      "step": 4252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5775784254074097,
+      "learning_rate": 0.0005928822748134919,
+      "loss": 3.5834,
+      "step": 4253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5799462795257568,
+      "learning_rate": 0.0005928789405716203,
+      "loss": 3.524,
+      "step": 4254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6155045628547668,
+      "learning_rate": 0.0005928756055583619,
+      "loss": 3.4856,
+      "step": 4255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5401473641395569,
+      "learning_rate": 0.0005928722697737254,
+      "loss": 3.4083,
+      "step": 4256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5645717978477478,
+      "learning_rate": 0.0005928689332177199,
+      "loss": 3.6458,
+      "step": 4257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5968051552772522,
+      "learning_rate": 0.0005928655958903538,
+      "loss": 3.6914,
+      "step": 4258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5574849843978882,
+      "learning_rate": 0.0005928622577916362,
+      "loss": 3.4648,
+      "step": 4259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5590183138847351,
+      "learning_rate": 0.0005928589189215757,
+      "loss": 3.5068,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5854844450950623,
+      "learning_rate": 0.0005928555792801812,
+      "loss": 3.4404,
+      "step": 4261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5542050004005432,
+      "learning_rate": 0.0005928522388674616,
+      "loss": 3.7172,
+      "step": 4262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5726750493049622,
+      "learning_rate": 0.0005928488976834254,
+      "loss": 3.5171,
+      "step": 4263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5503663420677185,
+      "learning_rate": 0.0005928455557280815,
+      "loss": 3.5306,
+      "step": 4264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5870798826217651,
+      "learning_rate": 0.000592842213001439,
+      "loss": 3.416,
+      "step": 4265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5509695410728455,
+      "learning_rate": 0.0005928388695035064,
+      "loss": 3.6081,
+      "step": 4266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.578829824924469,
+      "learning_rate": 0.0005928355252342925,
+      "loss": 3.4515,
+      "step": 4267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6161765456199646,
+      "learning_rate": 0.0005928321801938061,
+      "loss": 3.5032,
+      "step": 4268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5503228306770325,
+      "learning_rate": 0.0005928288343820563,
+      "loss": 3.3577,
+      "step": 4269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.550617516040802,
+      "learning_rate": 0.0005928254877990515,
+      "loss": 3.2324,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5668909549713135,
+      "learning_rate": 0.0005928221404448008,
+      "loss": 3.431,
+      "step": 4271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5765318274497986,
+      "learning_rate": 0.000592818792319313,
+      "loss": 3.5756,
+      "step": 4272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6077347993850708,
+      "learning_rate": 0.0005928154434225968,
+      "loss": 3.5527,
+      "step": 4273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5627986788749695,
+      "learning_rate": 0.0005928120937546611,
+      "loss": 3.8691,
+      "step": 4274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5574711561203003,
+      "learning_rate": 0.0005928087433155147,
+      "loss": 3.5428,
+      "step": 4275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5420995950698853,
+      "learning_rate": 0.0005928053921051664,
+      "loss": 3.4421,
+      "step": 4276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6338256001472473,
+      "learning_rate": 0.000592802040123625,
+      "loss": 3.3063,
+      "step": 4277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.572799026966095,
+      "learning_rate": 0.0005927986873708995,
+      "loss": 3.4216,
+      "step": 4278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5351685285568237,
+      "learning_rate": 0.0005927953338469986,
+      "loss": 3.391,
+      "step": 4279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5662745237350464,
+      "learning_rate": 0.0005927919795519311,
+      "loss": 3.5054,
+      "step": 4280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5973455905914307,
+      "learning_rate": 0.0005927886244857058,
+      "loss": 3.5656,
+      "step": 4281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5676031112670898,
+      "learning_rate": 0.0005927852686483317,
+      "loss": 3.2299,
+      "step": 4282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5569466352462769,
+      "learning_rate": 0.0005927819120398175,
+      "loss": 3.4992,
+      "step": 4283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5708233714103699,
+      "learning_rate": 0.0005927785546601721,
+      "loss": 3.5776,
+      "step": 4284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5976619124412537,
+      "learning_rate": 0.0005927751965094044,
+      "loss": 3.3507,
+      "step": 4285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5370445847511292,
+      "learning_rate": 0.0005927718375875231,
+      "loss": 3.3449,
+      "step": 4286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5878073573112488,
+      "learning_rate": 0.0005927684778945371,
+      "loss": 3.4856,
+      "step": 4287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5714808702468872,
+      "learning_rate": 0.0005927651174304553,
+      "loss": 3.6395,
+      "step": 4288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5233124494552612,
+      "learning_rate": 0.0005927617561952866,
+      "loss": 3.4073,
+      "step": 4289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5612361431121826,
+      "learning_rate": 0.0005927583941890398,
+      "loss": 3.3293,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5414638519287109,
+      "learning_rate": 0.0005927550314117235,
+      "loss": 3.4124,
+      "step": 4291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.586334764957428,
+      "learning_rate": 0.0005927516678633471,
+      "loss": 3.387,
+      "step": 4292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5697146654129028,
+      "learning_rate": 0.000592748303543919,
+      "loss": 3.6256,
+      "step": 4293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5857806205749512,
+      "learning_rate": 0.0005927449384534482,
+      "loss": 3.5125,
+      "step": 4294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5624878406524658,
+      "learning_rate": 0.0005927415725919435,
+      "loss": 3.5866,
+      "step": 4295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5444654226303101,
+      "learning_rate": 0.0005927382059594139,
+      "loss": 3.6301,
+      "step": 4296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5403390526771545,
+      "learning_rate": 0.0005927348385558682,
+      "loss": 3.4554,
+      "step": 4297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5376538634300232,
+      "learning_rate": 0.0005927314703813154,
+      "loss": 3.464,
+      "step": 4298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5038751363754272,
+      "learning_rate": 0.000592728101435764,
+      "loss": 3.5199,
+      "step": 4299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5603412389755249,
+      "learning_rate": 0.0005927247317192233,
+      "loss": 3.2781,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5391646027565002,
+      "learning_rate": 0.0005927213612317019,
+      "loss": 3.533,
+      "step": 4301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6210039258003235,
+      "learning_rate": 0.0005927179899732088,
+      "loss": 3.4641,
+      "step": 4302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5959204435348511,
+      "learning_rate": 0.0005927146179437529,
+      "loss": 3.5338,
+      "step": 4303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53571617603302,
+      "learning_rate": 0.000592711245143343,
+      "loss": 3.5324,
+      "step": 4304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5784791707992554,
+      "learning_rate": 0.000592707871571988,
+      "loss": 3.6262,
+      "step": 4305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5615180134773254,
+      "learning_rate": 0.0005927044972296969,
+      "loss": 3.4118,
+      "step": 4306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5299209952354431,
+      "learning_rate": 0.0005927011221164783,
+      "loss": 3.7879,
+      "step": 4307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5538467168807983,
+      "learning_rate": 0.0005926977462323414,
+      "loss": 3.3704,
+      "step": 4308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6038463711738586,
+      "learning_rate": 0.0005926943695772949,
+      "loss": 3.517,
+      "step": 4309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5968214869499207,
+      "learning_rate": 0.0005926909921513477,
+      "loss": 3.4321,
+      "step": 4310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5371940732002258,
+      "learning_rate": 0.0005926876139545089,
+      "loss": 3.4314,
+      "step": 4311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6479960680007935,
+      "learning_rate": 0.0005926842349867873,
+      "loss": 3.4106,
+      "step": 4312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5392265319824219,
+      "learning_rate": 0.0005926808552481917,
+      "loss": 3.6161,
+      "step": 4313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.572577714920044,
+      "learning_rate": 0.000592677474738731,
+      "loss": 3.5508,
+      "step": 4314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5596196055412292,
+      "learning_rate": 0.0005926740934584141,
+      "loss": 3.5483,
+      "step": 4315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5759965777397156,
+      "learning_rate": 0.0005926707114072501,
+      "loss": 3.411,
+      "step": 4316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53867506980896,
+      "learning_rate": 0.0005926673285852477,
+      "loss": 3.5226,
+      "step": 4317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5533181428909302,
+      "learning_rate": 0.0005926639449924158,
+      "loss": 3.5157,
+      "step": 4318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5320466756820679,
+      "learning_rate": 0.0005926605606287635,
+      "loss": 3.6545,
+      "step": 4319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5725945234298706,
+      "learning_rate": 0.0005926571754942996,
+      "loss": 3.4898,
+      "step": 4320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.633961021900177,
+      "learning_rate": 0.000592653789589033,
+      "loss": 3.4104,
+      "step": 4321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5438200831413269,
+      "learning_rate": 0.0005926504029129726,
+      "loss": 3.4986,
+      "step": 4322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5423779487609863,
+      "learning_rate": 0.0005926470154661275,
+      "loss": 3.2405,
+      "step": 4323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5243359804153442,
+      "learning_rate": 0.0005926436272485064,
+      "loss": 3.6668,
+      "step": 4324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5646772384643555,
+      "learning_rate": 0.0005926402382601183,
+      "loss": 3.2187,
+      "step": 4325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5363373756408691,
+      "learning_rate": 0.0005926368485009721,
+      "loss": 3.6333,
+      "step": 4326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5637429356575012,
+      "learning_rate": 0.0005926334579710768,
+      "loss": 3.7082,
+      "step": 4327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5417162775993347,
+      "learning_rate": 0.0005926300666704413,
+      "loss": 3.6215,
+      "step": 4328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5589267015457153,
+      "learning_rate": 0.0005926266745990745,
+      "loss": 3.4844,
+      "step": 4329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5703160762786865,
+      "learning_rate": 0.0005926232817569853,
+      "loss": 3.4402,
+      "step": 4330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5476822257041931,
+      "learning_rate": 0.0005926198881441828,
+      "loss": 3.4692,
+      "step": 4331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5585446357727051,
+      "learning_rate": 0.0005926164937606758,
+      "loss": 3.5845,
+      "step": 4332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.540649950504303,
+      "learning_rate": 0.0005926130986064733,
+      "loss": 3.4894,
+      "step": 4333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5446892976760864,
+      "learning_rate": 0.0005926097026815842,
+      "loss": 3.4582,
+      "step": 4334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5253145694732666,
+      "learning_rate": 0.0005926063059860173,
+      "loss": 3.4895,
+      "step": 4335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5359606742858887,
+      "learning_rate": 0.0005926029085197819,
+      "loss": 3.4371,
+      "step": 4336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5448620915412903,
+      "learning_rate": 0.0005925995102828867,
+      "loss": 3.4806,
+      "step": 4337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5616055130958557,
+      "learning_rate": 0.0005925961112753406,
+      "loss": 3.3458,
+      "step": 4338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.584311842918396,
+      "learning_rate": 0.0005925927114971527,
+      "loss": 3.4402,
+      "step": 4339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5829342603683472,
+      "learning_rate": 0.000592589310948332,
+      "loss": 3.4649,
+      "step": 4340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5744354128837585,
+      "learning_rate": 0.0005925859096288874,
+      "loss": 3.5193,
+      "step": 4341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5684018731117249,
+      "learning_rate": 0.0005925825075388277,
+      "loss": 3.4638,
+      "step": 4342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5490890741348267,
+      "learning_rate": 0.000592579104678162,
+      "loss": 3.593,
+      "step": 4343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6014746427536011,
+      "learning_rate": 0.0005925757010468993,
+      "loss": 3.2675,
+      "step": 4344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6437888145446777,
+      "learning_rate": 0.0005925722966450485,
+      "loss": 3.4341,
+      "step": 4345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5803835391998291,
+      "learning_rate": 0.0005925688914726185,
+      "loss": 3.368,
+      "step": 4346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5558992028236389,
+      "learning_rate": 0.0005925654855296183,
+      "loss": 3.5642,
+      "step": 4347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6441301703453064,
+      "learning_rate": 0.0005925620788160571,
+      "loss": 3.2692,
+      "step": 4348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5592367649078369,
+      "learning_rate": 0.0005925586713319436,
+      "loss": 3.5717,
+      "step": 4349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5774462223052979,
+      "learning_rate": 0.0005925552630772869,
+      "loss": 3.5024,
+      "step": 4350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6486428380012512,
+      "learning_rate": 0.0005925518540520958,
+      "loss": 3.4881,
+      "step": 4351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5728597044944763,
+      "learning_rate": 0.0005925484442563795,
+      "loss": 3.3829,
+      "step": 4352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5495771765708923,
+      "learning_rate": 0.000592545033690147,
+      "loss": 3.5991,
+      "step": 4353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5710741281509399,
+      "learning_rate": 0.0005925416223534071,
+      "loss": 3.3564,
+      "step": 4354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6084657311439514,
+      "learning_rate": 0.0005925382102461689,
+      "loss": 3.3813,
+      "step": 4355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5266708135604858,
+      "learning_rate": 0.0005925347973684414,
+      "loss": 3.3215,
+      "step": 4356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6160021424293518,
+      "learning_rate": 0.0005925313837202334,
+      "loss": 3.5516,
+      "step": 4357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5864583253860474,
+      "learning_rate": 0.0005925279693015541,
+      "loss": 3.5683,
+      "step": 4358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5891723036766052,
+      "learning_rate": 0.0005925245541124124,
+      "loss": 3.3166,
+      "step": 4359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5659295916557312,
+      "learning_rate": 0.0005925211381528175,
+      "loss": 3.4871,
+      "step": 4360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.609416127204895,
+      "learning_rate": 0.000592517721422778,
+      "loss": 3.4384,
+      "step": 4361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.518798828125,
+      "learning_rate": 0.0005925143039223032,
+      "loss": 3.3999,
+      "step": 4362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6183592081069946,
+      "learning_rate": 0.000592510885651402,
+      "loss": 3.5708,
+      "step": 4363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.522781491279602,
+      "learning_rate": 0.0005925074666100834,
+      "loss": 3.5036,
+      "step": 4364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.600847601890564,
+      "learning_rate": 0.0005925040467983564,
+      "loss": 3.4813,
+      "step": 4365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.556460440158844,
+      "learning_rate": 0.0005925006262162302,
+      "loss": 3.5809,
+      "step": 4366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5647059082984924,
+      "learning_rate": 0.0005924972048637135,
+      "loss": 3.5571,
+      "step": 4367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5725759267807007,
+      "learning_rate": 0.0005924937827408154,
+      "loss": 3.4548,
+      "step": 4368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5902033448219299,
+      "learning_rate": 0.0005924903598475451,
+      "loss": 3.4404,
+      "step": 4369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.723042905330658,
+      "learning_rate": 0.0005924869361839115,
+      "loss": 3.5888,
+      "step": 4370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7233627438545227,
+      "learning_rate": 0.0005924835117499235,
+      "loss": 3.3951,
+      "step": 4371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5561408996582031,
+      "learning_rate": 0.0005924800865455903,
+      "loss": 3.6309,
+      "step": 4372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5627115964889526,
+      "learning_rate": 0.0005924766605709209,
+      "loss": 3.5393,
+      "step": 4373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5121928453445435,
+      "learning_rate": 0.000592473233825924,
+      "loss": 3.5696,
+      "step": 4374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6076629757881165,
+      "learning_rate": 0.0005924698063106091,
+      "loss": 3.4886,
+      "step": 4375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5451004505157471,
+      "learning_rate": 0.0005924663780249851,
+      "loss": 3.3686,
+      "step": 4376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.582217276096344,
+      "learning_rate": 0.0005924629489690608,
+      "loss": 3.5121,
+      "step": 4377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5545997023582458,
+      "learning_rate": 0.0005924595191428454,
+      "loss": 3.5433,
+      "step": 4378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5341416001319885,
+      "learning_rate": 0.0005924560885463479,
+      "loss": 3.5336,
+      "step": 4379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6326627731323242,
+      "learning_rate": 0.0005924526571795774,
+      "loss": 3.4165,
+      "step": 4380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5663824677467346,
+      "learning_rate": 0.0005924492250425428,
+      "loss": 3.5157,
+      "step": 4381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5439797043800354,
+      "learning_rate": 0.0005924457921352533,
+      "loss": 3.4163,
+      "step": 4382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5896198749542236,
+      "learning_rate": 0.0005924423584577178,
+      "loss": 3.5275,
+      "step": 4383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5586664080619812,
+      "learning_rate": 0.0005924389240099454,
+      "loss": 3.576,
+      "step": 4384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5929638743400574,
+      "learning_rate": 0.0005924354887919452,
+      "loss": 3.5569,
+      "step": 4385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6475085020065308,
+      "learning_rate": 0.0005924320528037263,
+      "loss": 3.602,
+      "step": 4386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5904268622398376,
+      "learning_rate": 0.0005924286160452974,
+      "loss": 3.4897,
+      "step": 4387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5506060719490051,
+      "learning_rate": 0.000592425178516668,
+      "loss": 3.5867,
+      "step": 4388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6361829042434692,
+      "learning_rate": 0.000592421740217847,
+      "loss": 3.6221,
+      "step": 4389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5692065954208374,
+      "learning_rate": 0.0005924183011488433,
+      "loss": 3.6743,
+      "step": 4390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5525909066200256,
+      "learning_rate": 0.000592414861309666,
+      "loss": 3.4595,
+      "step": 4391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5622319579124451,
+      "learning_rate": 0.0005924114207003243,
+      "loss": 3.5213,
+      "step": 4392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5368682146072388,
+      "learning_rate": 0.0005924079793208273,
+      "loss": 3.5174,
+      "step": 4393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5590381622314453,
+      "learning_rate": 0.000592404537171184,
+      "loss": 3.4644,
+      "step": 4394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5523700714111328,
+      "learning_rate": 0.0005924010942514032,
+      "loss": 3.3879,
+      "step": 4395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5766987204551697,
+      "learning_rate": 0.0005923976505614944,
+      "loss": 3.7217,
+      "step": 4396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5824809670448303,
+      "learning_rate": 0.0005923942061014663,
+      "loss": 3.459,
+      "step": 4397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5317344665527344,
+      "learning_rate": 0.0005923907608713282,
+      "loss": 3.3727,
+      "step": 4398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5514039993286133,
+      "learning_rate": 0.0005923873148710892,
+      "loss": 3.3122,
+      "step": 4399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5581009387969971,
+      "learning_rate": 0.0005923838681007581,
+      "loss": 3.6301,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5631245374679565,
+      "learning_rate": 0.0005923804205603442,
+      "loss": 3.1971,
+      "step": 4401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5507313013076782,
+      "learning_rate": 0.0005923769722498566,
+      "loss": 3.4534,
+      "step": 4402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5625492334365845,
+      "learning_rate": 0.0005923735231693043,
+      "loss": 3.5629,
+      "step": 4403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.599851131439209,
+      "learning_rate": 0.0005923700733186965,
+      "loss": 3.6198,
+      "step": 4404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6086528897285461,
+      "learning_rate": 0.0005923666226980421,
+      "loss": 3.5182,
+      "step": 4405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5781559944152832,
+      "learning_rate": 0.0005923631713073503,
+      "loss": 3.3736,
+      "step": 4406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5884630680084229,
+      "learning_rate": 0.0005923597191466301,
+      "loss": 3.4333,
+      "step": 4407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5860154628753662,
+      "learning_rate": 0.0005923562662158907,
+      "loss": 3.5609,
+      "step": 4408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5450155138969421,
+      "learning_rate": 0.0005923528125151412,
+      "loss": 3.3551,
+      "step": 4409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6374961733818054,
+      "learning_rate": 0.0005923493580443907,
+      "loss": 3.4167,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.562510073184967,
+      "learning_rate": 0.0005923459028036482,
+      "loss": 3.3272,
+      "step": 4411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5314709544181824,
+      "learning_rate": 0.0005923424467929229,
+      "loss": 3.4734,
+      "step": 4412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5496847629547119,
+      "learning_rate": 0.0005923389900122239,
+      "loss": 3.0824,
+      "step": 4413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5986039638519287,
+      "learning_rate": 0.0005923355324615601,
+      "loss": 3.5452,
+      "step": 4414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5644450187683105,
+      "learning_rate": 0.0005923320741409409,
+      "loss": 3.3707,
+      "step": 4415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5469602346420288,
+      "learning_rate": 0.0005923286150503753,
+      "loss": 3.5989,
+      "step": 4416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5269728899002075,
+      "learning_rate": 0.0005923251551898722,
+      "loss": 3.6738,
+      "step": 4417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5269190073013306,
+      "learning_rate": 0.000592321694559441,
+      "loss": 3.6033,
+      "step": 4418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5562129020690918,
+      "learning_rate": 0.0005923182331590908,
+      "loss": 3.4572,
+      "step": 4419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5670672655105591,
+      "learning_rate": 0.0005923147709888305,
+      "loss": 3.4981,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5788316130638123,
+      "learning_rate": 0.0005923113080486695,
+      "loss": 3.3878,
+      "step": 4421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.585292637348175,
+      "learning_rate": 0.0005923078443386166,
+      "loss": 3.3861,
+      "step": 4422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5587526559829712,
+      "learning_rate": 0.0005923043798586812,
+      "loss": 3.3619,
+      "step": 4423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5967184901237488,
+      "learning_rate": 0.0005923009146088724,
+      "loss": 3.3487,
+      "step": 4424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5900785326957703,
+      "learning_rate": 0.0005922974485891991,
+      "loss": 3.4949,
+      "step": 4425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5535380840301514,
+      "learning_rate": 0.0005922939817996706,
+      "loss": 3.5115,
+      "step": 4426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5610527992248535,
+      "learning_rate": 0.000592290514240296,
+      "loss": 3.4378,
+      "step": 4427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5495424270629883,
+      "learning_rate": 0.0005922870459110844,
+      "loss": 3.2887,
+      "step": 4428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5786449313163757,
+      "learning_rate": 0.000592283576812045,
+      "loss": 3.4501,
+      "step": 4429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5528695583343506,
+      "learning_rate": 0.000592280106943187,
+      "loss": 3.58,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6839733719825745,
+      "learning_rate": 0.0005922766363045195,
+      "loss": 3.3403,
+      "step": 4431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.568471372127533,
+      "learning_rate": 0.0005922731648960514,
+      "loss": 3.4483,
+      "step": 4432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5260776877403259,
+      "learning_rate": 0.0005922696927177921,
+      "loss": 3.7186,
+      "step": 4433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5697051882743835,
+      "learning_rate": 0.0005922662197697507,
+      "loss": 3.3252,
+      "step": 4434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5838139057159424,
+      "learning_rate": 0.0005922627460519363,
+      "loss": 3.5432,
+      "step": 4435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5762350559234619,
+      "learning_rate": 0.0005922592715643582,
+      "loss": 3.4488,
+      "step": 4436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5904020071029663,
+      "learning_rate": 0.0005922557963070252,
+      "loss": 3.5981,
+      "step": 4437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5588435530662537,
+      "learning_rate": 0.0005922523202799468,
+      "loss": 3.6139,
+      "step": 4438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5399730205535889,
+      "learning_rate": 0.000592248843483132,
+      "loss": 3.3933,
+      "step": 4439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.57188481092453,
+      "learning_rate": 0.0005922453659165901,
+      "loss": 3.5267,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5735987424850464,
+      "learning_rate": 0.0005922418875803301,
+      "loss": 3.4599,
+      "step": 4441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5924454927444458,
+      "learning_rate": 0.0005922384084743611,
+      "loss": 3.754,
+      "step": 4442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5900004506111145,
+      "learning_rate": 0.0005922349285986925,
+      "loss": 3.3268,
+      "step": 4443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5811010003089905,
+      "learning_rate": 0.0005922314479533333,
+      "loss": 3.3628,
+      "step": 4444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5504472255706787,
+      "learning_rate": 0.0005922279665382927,
+      "loss": 3.5839,
+      "step": 4445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5668983459472656,
+      "learning_rate": 0.0005922244843535798,
+      "loss": 3.6179,
+      "step": 4446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5283928513526917,
+      "learning_rate": 0.000592221001399204,
+      "loss": 3.3998,
+      "step": 4447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5632615685462952,
+      "learning_rate": 0.0005922175176751742,
+      "loss": 3.3972,
+      "step": 4448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5324033498764038,
+      "learning_rate": 0.0005922140331814997,
+      "loss": 3.4857,
+      "step": 4449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5987648963928223,
+      "learning_rate": 0.0005922105479181898,
+      "loss": 3.5155,
+      "step": 4450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5266212821006775,
+      "learning_rate": 0.0005922070618852535,
+      "loss": 3.3693,
+      "step": 4451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5755586624145508,
+      "learning_rate": 0.0005922035750827,
+      "loss": 3.5271,
+      "step": 4452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6109917163848877,
+      "learning_rate": 0.0005922000875105385,
+      "loss": 3.3581,
+      "step": 4453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5387493968009949,
+      "learning_rate": 0.0005921965991687783,
+      "loss": 3.539,
+      "step": 4454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5238696336746216,
+      "learning_rate": 0.0005921931100574284,
+      "loss": 3.6753,
+      "step": 4455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.583580493927002,
+      "learning_rate": 0.0005921896201764981,
+      "loss": 3.4583,
+      "step": 4456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5637732148170471,
+      "learning_rate": 0.0005921861295259966,
+      "loss": 3.689,
+      "step": 4457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5128093957901001,
+      "learning_rate": 0.000592182638105933,
+      "loss": 3.5987,
+      "step": 4458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6136223673820496,
+      "learning_rate": 0.0005921791459163167,
+      "loss": 3.4421,
+      "step": 4459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.573216438293457,
+      "learning_rate": 0.0005921756529571566,
+      "loss": 3.3476,
+      "step": 4460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5862829685211182,
+      "learning_rate": 0.0005921721592284621,
+      "loss": 3.5071,
+      "step": 4461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5221441984176636,
+      "learning_rate": 0.0005921686647302424,
+      "loss": 3.4102,
+      "step": 4462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6147505640983582,
+      "learning_rate": 0.0005921651694625067,
+      "loss": 3.5414,
+      "step": 4463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5590131282806396,
+      "learning_rate": 0.0005921616734252641,
+      "loss": 3.4173,
+      "step": 4464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5548772811889648,
+      "learning_rate": 0.0005921581766185239,
+      "loss": 3.3825,
+      "step": 4465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6083531975746155,
+      "learning_rate": 0.0005921546790422953,
+      "loss": 3.3853,
+      "step": 4466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5427078604698181,
+      "learning_rate": 0.0005921511806965875,
+      "loss": 3.4407,
+      "step": 4467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5490184426307678,
+      "learning_rate": 0.0005921476815814098,
+      "loss": 3.521,
+      "step": 4468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5580472350120544,
+      "learning_rate": 0.0005921441816967711,
+      "loss": 3.4545,
+      "step": 4469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5435226559638977,
+      "learning_rate": 0.000592140681042681,
+      "loss": 3.5815,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5188127160072327,
+      "learning_rate": 0.0005921371796191486,
+      "loss": 3.5472,
+      "step": 4471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5608434081077576,
+      "learning_rate": 0.0005921336774261831,
+      "loss": 3.374,
+      "step": 4472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5581082105636597,
+      "learning_rate": 0.0005921301744637937,
+      "loss": 3.5685,
+      "step": 4473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5264663696289062,
+      "learning_rate": 0.0005921266707319896,
+      "loss": 3.5169,
+      "step": 4474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5787158608436584,
+      "learning_rate": 0.0005921231662307801,
+      "loss": 3.4266,
+      "step": 4475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5512024164199829,
+      "learning_rate": 0.0005921196609601744,
+      "loss": 3.6085,
+      "step": 4476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.559208333492279,
+      "learning_rate": 0.0005921161549201818,
+      "loss": 3.3556,
+      "step": 4477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5857759118080139,
+      "learning_rate": 0.0005921126481108112,
+      "loss": 3.5417,
+      "step": 4478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5337649583816528,
+      "learning_rate": 0.0005921091405320724,
+      "loss": 3.4392,
+      "step": 4479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5375748872756958,
+      "learning_rate": 0.0005921056321839742,
+      "loss": 3.4743,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5559386610984802,
+      "learning_rate": 0.0005921021230665261,
+      "loss": 3.4687,
+      "step": 4481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5421441793441772,
+      "learning_rate": 0.000592098613179737,
+      "loss": 3.3652,
+      "step": 4482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5195907354354858,
+      "learning_rate": 0.0005920951025236166,
+      "loss": 3.6266,
+      "step": 4483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.600489616394043,
+      "learning_rate": 0.0005920915910981739,
+      "loss": 3.3397,
+      "step": 4484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5409472584724426,
+      "learning_rate": 0.000592088078903418,
+      "loss": 3.2999,
+      "step": 4485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.522708535194397,
+      "learning_rate": 0.0005920845659393584,
+      "loss": 3.3678,
+      "step": 4486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53269362449646,
+      "learning_rate": 0.0005920810522060042,
+      "loss": 3.5437,
+      "step": 4487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5417482256889343,
+      "learning_rate": 0.0005920775377033648,
+      "loss": 3.5003,
+      "step": 4488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5512046813964844,
+      "learning_rate": 0.0005920740224314494,
+      "loss": 3.3525,
+      "step": 4489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5806989073753357,
+      "learning_rate": 0.0005920705063902672,
+      "loss": 3.4915,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5602362155914307,
+      "learning_rate": 0.0005920669895798275,
+      "loss": 3.5072,
+      "step": 4491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5372531414031982,
+      "learning_rate": 0.0005920634720001396,
+      "loss": 3.3132,
+      "step": 4492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6997660994529724,
+      "learning_rate": 0.0005920599536512126,
+      "loss": 3.2909,
+      "step": 4493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5167723894119263,
+      "learning_rate": 0.000592056434533056,
+      "loss": 3.498,
+      "step": 4494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5544456243515015,
+      "learning_rate": 0.0005920529146456789,
+      "loss": 3.393,
+      "step": 4495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5869852304458618,
+      "learning_rate": 0.0005920493939890907,
+      "loss": 3.5451,
+      "step": 4496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5493007898330688,
+      "learning_rate": 0.0005920458725633005,
+      "loss": 3.4858,
+      "step": 4497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5228421092033386,
+      "learning_rate": 0.0005920423503683178,
+      "loss": 3.5755,
+      "step": 4498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5734871029853821,
+      "learning_rate": 0.0005920388274041516,
+      "loss": 3.4937,
+      "step": 4499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6023586392402649,
+      "learning_rate": 0.0005920353036708115,
+      "loss": 3.3827,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5095198750495911,
+      "learning_rate": 0.0005920317791683065,
+      "loss": 3.4317,
+      "step": 4501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5262815356254578,
+      "learning_rate": 0.0005920282538966461,
+      "loss": 3.2033,
+      "step": 4502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5463591814041138,
+      "learning_rate": 0.0005920247278558394,
+      "loss": 3.4123,
+      "step": 4503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5262253880500793,
+      "learning_rate": 0.0005920212010458957,
+      "loss": 3.5664,
+      "step": 4504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5638479590415955,
+      "learning_rate": 0.0005920176734668244,
+      "loss": 3.3824,
+      "step": 4505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5075425505638123,
+      "learning_rate": 0.0005920141451186349,
+      "loss": 3.5525,
+      "step": 4506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5334774255752563,
+      "learning_rate": 0.0005920106160013362,
+      "loss": 3.4679,
+      "step": 4507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5502597689628601,
+      "learning_rate": 0.0005920070861149377,
+      "loss": 3.5242,
+      "step": 4508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5628333687782288,
+      "learning_rate": 0.0005920035554594489,
+      "loss": 3.4435,
+      "step": 4509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5412539839744568,
+      "learning_rate": 0.0005920000240348788,
+      "loss": 3.5723,
+      "step": 4510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5367388725280762,
+      "learning_rate": 0.0005919964918412368,
+      "loss": 3.3911,
+      "step": 4511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5794838070869446,
+      "learning_rate": 0.0005919929588785323,
+      "loss": 3.4213,
+      "step": 4512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5445544123649597,
+      "learning_rate": 0.0005919894251467746,
+      "loss": 3.4142,
+      "step": 4513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5294917225837708,
+      "learning_rate": 0.0005919858906459728,
+      "loss": 3.5728,
+      "step": 4514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5747779607772827,
+      "learning_rate": 0.0005919823553761365,
+      "loss": 3.7209,
+      "step": 4515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5731784105300903,
+      "learning_rate": 0.0005919788193372748,
+      "loss": 3.3981,
+      "step": 4516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5659194588661194,
+      "learning_rate": 0.000591975282529397,
+      "loss": 3.4661,
+      "step": 4517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6080614924430847,
+      "learning_rate": 0.0005919717449525127,
+      "loss": 3.448,
+      "step": 4518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5205955505371094,
+      "learning_rate": 0.0005919682066066309,
+      "loss": 3.7119,
+      "step": 4519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6132055521011353,
+      "learning_rate": 0.000591964667491761,
+      "loss": 3.5377,
+      "step": 4520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5384352207183838,
+      "learning_rate": 0.0005919611276079124,
+      "loss": 3.5972,
+      "step": 4521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5405988693237305,
+      "learning_rate": 0.0005919575869550944,
+      "loss": 3.4752,
+      "step": 4522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5446063280105591,
+      "learning_rate": 0.0005919540455333163,
+      "loss": 3.5515,
+      "step": 4523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6086515188217163,
+      "learning_rate": 0.0005919505033425875,
+      "loss": 3.5652,
+      "step": 4524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5298771262168884,
+      "learning_rate": 0.0005919469603829171,
+      "loss": 3.3091,
+      "step": 4525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5765856504440308,
+      "learning_rate": 0.0005919434166543146,
+      "loss": 3.3966,
+      "step": 4526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.569276750087738,
+      "learning_rate": 0.0005919398721567895,
+      "loss": 3.454,
+      "step": 4527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.568664014339447,
+      "learning_rate": 0.0005919363268903508,
+      "loss": 3.6629,
+      "step": 4528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.569466233253479,
+      "learning_rate": 0.000591932780855008,
+      "loss": 3.4888,
+      "step": 4529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5888025760650635,
+      "learning_rate": 0.0005919292340507706,
+      "loss": 3.3833,
+      "step": 4530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6375117897987366,
+      "learning_rate": 0.0005919256864776476,
+      "loss": 3.7058,
+      "step": 4531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5959801077842712,
+      "learning_rate": 0.0005919221381356486,
+      "loss": 3.3804,
+      "step": 4532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5718854069709778,
+      "learning_rate": 0.0005919185890247828,
+      "loss": 3.48,
+      "step": 4533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5459699630737305,
+      "learning_rate": 0.0005919150391450597,
+      "loss": 3.4685,
+      "step": 4534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5691114068031311,
+      "learning_rate": 0.0005919114884964886,
+      "loss": 3.4821,
+      "step": 4535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5469587445259094,
+      "learning_rate": 0.0005919079370790789,
+      "loss": 3.4524,
+      "step": 4536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5794917941093445,
+      "learning_rate": 0.0005919043848928397,
+      "loss": 3.5051,
+      "step": 4537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5750076770782471,
+      "learning_rate": 0.0005919008319377805,
+      "loss": 3.3637,
+      "step": 4538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5511584281921387,
+      "learning_rate": 0.0005918972782139108,
+      "loss": 3.3772,
+      "step": 4539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5298534631729126,
+      "learning_rate": 0.0005918937237212397,
+      "loss": 3.6845,
+      "step": 4540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5647959113121033,
+      "learning_rate": 0.0005918901684597769,
+      "loss": 3.2087,
+      "step": 4541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5687061548233032,
+      "learning_rate": 0.0005918866124295315,
+      "loss": 3.4084,
+      "step": 4542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5853356122970581,
+      "learning_rate": 0.000591883055630513,
+      "loss": 3.4805,
+      "step": 4543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5594838261604309,
+      "learning_rate": 0.0005918794980627305,
+      "loss": 3.728,
+      "step": 4544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5661375522613525,
+      "learning_rate": 0.0005918759397261936,
+      "loss": 3.4592,
+      "step": 4545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5649911165237427,
+      "learning_rate": 0.0005918723806209119,
+      "loss": 3.4811,
+      "step": 4546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.568805456161499,
+      "learning_rate": 0.0005918688207468942,
+      "loss": 3.3881,
+      "step": 4547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5687074661254883,
+      "learning_rate": 0.0005918652601041505,
+      "loss": 3.4324,
+      "step": 4548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5697916150093079,
+      "learning_rate": 0.0005918616986926898,
+      "loss": 3.5119,
+      "step": 4549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5741244554519653,
+      "learning_rate": 0.0005918581365125215,
+      "loss": 3.5548,
+      "step": 4550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5744023323059082,
+      "learning_rate": 0.000591854573563655,
+      "loss": 3.5327,
+      "step": 4551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5800146460533142,
+      "learning_rate": 0.0005918510098460997,
+      "loss": 3.4031,
+      "step": 4552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5622119307518005,
+      "learning_rate": 0.000591847445359865,
+      "loss": 3.4033,
+      "step": 4553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5262593626976013,
+      "learning_rate": 0.0005918438801049605,
+      "loss": 3.2841,
+      "step": 4554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5564959645271301,
+      "learning_rate": 0.0005918403140813952,
+      "loss": 3.6266,
+      "step": 4555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.574227511882782,
+      "learning_rate": 0.0005918367472891787,
+      "loss": 3.3961,
+      "step": 4556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.563045084476471,
+      "learning_rate": 0.0005918331797283204,
+      "loss": 3.6054,
+      "step": 4557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5304876565933228,
+      "learning_rate": 0.0005918296113988297,
+      "loss": 3.3966,
+      "step": 4558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5531924962997437,
+      "learning_rate": 0.0005918260423007159,
+      "loss": 3.2937,
+      "step": 4559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5169435739517212,
+      "learning_rate": 0.0005918224724339885,
+      "loss": 3.5655,
+      "step": 4560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6689804196357727,
+      "learning_rate": 0.0005918189017986569,
+      "loss": 3.6632,
+      "step": 4561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5568976998329163,
+      "learning_rate": 0.0005918153303947304,
+      "loss": 3.6612,
+      "step": 4562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5638584494590759,
+      "learning_rate": 0.0005918117582222185,
+      "loss": 3.4756,
+      "step": 4563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5415082573890686,
+      "learning_rate": 0.0005918081852811307,
+      "loss": 3.5922,
+      "step": 4564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5457674264907837,
+      "learning_rate": 0.0005918046115714762,
+      "loss": 3.6223,
+      "step": 4565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5880576968193054,
+      "learning_rate": 0.0005918010370932645,
+      "loss": 3.3749,
+      "step": 4566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.572864830493927,
+      "learning_rate": 0.0005917974618465051,
+      "loss": 3.428,
+      "step": 4567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5527423024177551,
+      "learning_rate": 0.0005917938858312073,
+      "loss": 3.3815,
+      "step": 4568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6350066661834717,
+      "learning_rate": 0.0005917903090473806,
+      "loss": 3.1924,
+      "step": 4569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5983766913414001,
+      "learning_rate": 0.0005917867314950343,
+      "loss": 3.4771,
+      "step": 4570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5613915324211121,
+      "learning_rate": 0.0005917831531741779,
+      "loss": 3.4837,
+      "step": 4571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5356848835945129,
+      "learning_rate": 0.000591779574084821,
+      "loss": 3.5368,
+      "step": 4572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5657932162284851,
+      "learning_rate": 0.0005917759942269727,
+      "loss": 3.3458,
+      "step": 4573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5904025435447693,
+      "learning_rate": 0.0005917724136006425,
+      "loss": 3.3554,
+      "step": 4574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5392588376998901,
+      "learning_rate": 0.00059176883220584,
+      "loss": 3.4159,
+      "step": 4575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5528022646903992,
+      "learning_rate": 0.0005917652500425747,
+      "loss": 3.5421,
+      "step": 4576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5817191004753113,
+      "learning_rate": 0.0005917616671108557,
+      "loss": 3.4776,
+      "step": 4577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5895118713378906,
+      "learning_rate": 0.0005917580834106927,
+      "loss": 3.4579,
+      "step": 4578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5503705143928528,
+      "learning_rate": 0.0005917544989420951,
+      "loss": 3.4782,
+      "step": 4579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5728640556335449,
+      "learning_rate": 0.0005917509137050723,
+      "loss": 3.2202,
+      "step": 4580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5362029671669006,
+      "learning_rate": 0.0005917473276996336,
+      "loss": 3.4599,
+      "step": 4581
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5472753643989563,
+      "learning_rate": 0.0005917437409257886,
+      "loss": 3.5085,
+      "step": 4582
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.575137197971344,
+      "learning_rate": 0.0005917401533835469,
+      "loss": 3.4713,
+      "step": 4583
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5501198768615723,
+      "learning_rate": 0.0005917365650729176,
+      "loss": 3.3051,
+      "step": 4584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5528760552406311,
+      "learning_rate": 0.0005917329759939105,
+      "loss": 3.4805,
+      "step": 4585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5414803624153137,
+      "learning_rate": 0.0005917293861465347,
+      "loss": 3.577,
+      "step": 4586
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5812093615531921,
+      "learning_rate": 0.0005917257955308,
+      "loss": 3.5206,
+      "step": 4587
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6059795022010803,
+      "learning_rate": 0.0005917222041467156,
+      "loss": 3.4236,
+      "step": 4588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5378461480140686,
+      "learning_rate": 0.0005917186119942912,
+      "loss": 3.4233,
+      "step": 4589
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5715885758399963,
+      "learning_rate": 0.000591715019073536,
+      "loss": 3.3245,
+      "step": 4590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5676418542861938,
+      "learning_rate": 0.0005917114253844596,
+      "loss": 3.1704,
+      "step": 4591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5657680034637451,
+      "learning_rate": 0.0005917078309270714,
+      "loss": 3.3586,
+      "step": 4592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5191043019294739,
+      "learning_rate": 0.000591704235701381,
+      "loss": 3.3325,
+      "step": 4593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.567353367805481,
+      "learning_rate": 0.0005917006397073977,
+      "loss": 3.6905,
+      "step": 4594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5378613471984863,
+      "learning_rate": 0.0005916970429451311,
+      "loss": 3.7788,
+      "step": 4595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.548173725605011,
+      "learning_rate": 0.0005916934454145906,
+      "loss": 3.3993,
+      "step": 4596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5689377188682556,
+      "learning_rate": 0.0005916898471157858,
+      "loss": 3.5089,
+      "step": 4597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5914126634597778,
+      "learning_rate": 0.000591686248048726,
+      "loss": 3.4533,
+      "step": 4598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.533582866191864,
+      "learning_rate": 0.0005916826482134207,
+      "loss": 3.5641,
+      "step": 4599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5263054370880127,
+      "learning_rate": 0.0005916790476098796,
+      "loss": 3.397,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5878801941871643,
+      "learning_rate": 0.000591675446238112,
+      "loss": 3.4843,
+      "step": 4601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7907947301864624,
+      "learning_rate": 0.0005916718440981272,
+      "loss": 3.5027,
+      "step": 4602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5525874495506287,
+      "learning_rate": 0.000591668241189935,
+      "loss": 3.3515,
+      "step": 4603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5373077988624573,
+      "learning_rate": 0.0005916646375135449,
+      "loss": 3.2774,
+      "step": 4604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5783985257148743,
+      "learning_rate": 0.0005916610330689661,
+      "loss": 3.3954,
+      "step": 4605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5470697283744812,
+      "learning_rate": 0.0005916574278562085,
+      "loss": 3.3739,
+      "step": 4606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5767296552658081,
+      "learning_rate": 0.0005916538218752812,
+      "loss": 3.3787,
+      "step": 4607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5586828589439392,
+      "learning_rate": 0.000591650215126194,
+      "loss": 3.4687,
+      "step": 4608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5796303153038025,
+      "learning_rate": 0.000591646607608956,
+      "loss": 3.7055,
+      "step": 4609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5355789065361023,
+      "learning_rate": 0.0005916429993235772,
+      "loss": 3.5942,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5896856188774109,
+      "learning_rate": 0.0005916393902700667,
+      "loss": 3.4948,
+      "step": 4611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5608619451522827,
+      "learning_rate": 0.0005916357804484344,
+      "loss": 3.5695,
+      "step": 4612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5785589218139648,
+      "learning_rate": 0.0005916321698586894,
+      "loss": 3.5872,
+      "step": 4613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5732633471488953,
+      "learning_rate": 0.0005916285585008415,
+      "loss": 3.3187,
+      "step": 4614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5910802483558655,
+      "learning_rate": 0.0005916249463748999,
+      "loss": 3.4223,
+      "step": 4615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5657915472984314,
+      "learning_rate": 0.0005916213334808745,
+      "loss": 3.5133,
+      "step": 4616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.522301435470581,
+      "learning_rate": 0.0005916177198187746,
+      "loss": 3.4913,
+      "step": 4617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5693953037261963,
+      "learning_rate": 0.0005916141053886097,
+      "loss": 3.3059,
+      "step": 4618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5592297911643982,
+      "learning_rate": 0.0005916104901903894,
+      "loss": 3.3731,
+      "step": 4619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5223372578620911,
+      "learning_rate": 0.0005916068742241232,
+      "loss": 3.1434,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5213208794593811,
+      "learning_rate": 0.0005916032574898206,
+      "loss": 3.3789,
+      "step": 4621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5590026378631592,
+      "learning_rate": 0.0005915996399874911,
+      "loss": 3.5919,
+      "step": 4622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5489098429679871,
+      "learning_rate": 0.0005915960217171444,
+      "loss": 3.3545,
+      "step": 4623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5755786299705505,
+      "learning_rate": 0.0005915924026787898,
+      "loss": 3.5858,
+      "step": 4624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5431007742881775,
+      "learning_rate": 0.0005915887828724369,
+      "loss": 3.5246,
+      "step": 4625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5259215235710144,
+      "learning_rate": 0.0005915851622980954,
+      "loss": 3.5859,
+      "step": 4626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5726738572120667,
+      "learning_rate": 0.0005915815409557745,
+      "loss": 3.7115,
+      "step": 4627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5531672239303589,
+      "learning_rate": 0.000591577918845484,
+      "loss": 3.4266,
+      "step": 4628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5267097353935242,
+      "learning_rate": 0.0005915742959672335,
+      "loss": 3.5795,
+      "step": 4629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5696015357971191,
+      "learning_rate": 0.0005915706723210323,
+      "loss": 3.2015,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5455690622329712,
+      "learning_rate": 0.00059156704790689,
+      "loss": 3.44,
+      "step": 4631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6064535975456238,
+      "learning_rate": 0.0005915634227248163,
+      "loss": 3.2651,
+      "step": 4632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5588998794555664,
+      "learning_rate": 0.0005915597967748207,
+      "loss": 3.383,
+      "step": 4633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5298521518707275,
+      "learning_rate": 0.0005915561700569127,
+      "loss": 3.5665,
+      "step": 4634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5695735812187195,
+      "learning_rate": 0.0005915525425711018,
+      "loss": 3.4372,
+      "step": 4635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5828069448471069,
+      "learning_rate": 0.0005915489143173976,
+      "loss": 3.4493,
+      "step": 4636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5570802092552185,
+      "learning_rate": 0.0005915452852958098,
+      "loss": 3.6557,
+      "step": 4637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5548609495162964,
+      "learning_rate": 0.0005915416555063478,
+      "loss": 3.5455,
+      "step": 4638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6037477850914001,
+      "learning_rate": 0.000591538024949021,
+      "loss": 3.5569,
+      "step": 4639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6033467054367065,
+      "learning_rate": 0.0005915343936238394,
+      "loss": 3.5211,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6203020811080933,
+      "learning_rate": 0.0005915307615308122,
+      "loss": 3.6181,
+      "step": 4641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5475682616233826,
+      "learning_rate": 0.000591527128669949,
+      "loss": 3.5266,
+      "step": 4642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5472739934921265,
+      "learning_rate": 0.0005915234950412596,
+      "loss": 3.6185,
+      "step": 4643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5932005047798157,
+      "learning_rate": 0.0005915198606447533,
+      "loss": 3.3851,
+      "step": 4644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5578650832176208,
+      "learning_rate": 0.0005915162254804398,
+      "loss": 3.4982,
+      "step": 4645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5666883587837219,
+      "learning_rate": 0.0005915125895483288,
+      "loss": 3.3541,
+      "step": 4646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6000686883926392,
+      "learning_rate": 0.0005915089528484295,
+      "loss": 3.376,
+      "step": 4647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.593292772769928,
+      "learning_rate": 0.0005915053153807519,
+      "loss": 3.5401,
+      "step": 4648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5368260145187378,
+      "learning_rate": 0.0005915016771453054,
+      "loss": 3.6079,
+      "step": 4649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6727637648582458,
+      "learning_rate": 0.0005914980381420996,
+      "loss": 3.4033,
+      "step": 4650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.541647732257843,
+      "learning_rate": 0.0005914943983711439,
+      "loss": 3.3189,
+      "step": 4651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5538272857666016,
+      "learning_rate": 0.0005914907578324481,
+      "loss": 3.4644,
+      "step": 4652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5518693923950195,
+      "learning_rate": 0.0005914871165260219,
+      "loss": 3.6281,
+      "step": 4653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5680675506591797,
+      "learning_rate": 0.0005914834744518747,
+      "loss": 3.4995,
+      "step": 4654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5549336671829224,
+      "learning_rate": 0.000591479831610016,
+      "loss": 3.264,
+      "step": 4655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5913617014884949,
+      "learning_rate": 0.0005914761880004555,
+      "loss": 3.2961,
+      "step": 4656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5764253735542297,
+      "learning_rate": 0.000591472543623203,
+      "loss": 3.5085,
+      "step": 4657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5552957057952881,
+      "learning_rate": 0.0005914688984782677,
+      "loss": 3.3739,
+      "step": 4658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5654916763305664,
+      "learning_rate": 0.0005914652525656596,
+      "loss": 3.3516,
+      "step": 4659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5623388886451721,
+      "learning_rate": 0.0005914616058853881,
+      "loss": 3.5072,
+      "step": 4660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5291627645492554,
+      "learning_rate": 0.0005914579584374627,
+      "loss": 3.4728,
+      "step": 4661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5866690874099731,
+      "learning_rate": 0.0005914543102218933,
+      "loss": 3.3841,
+      "step": 4662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5244299173355103,
+      "learning_rate": 0.0005914506612386891,
+      "loss": 3.4587,
+      "step": 4663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5474516153335571,
+      "learning_rate": 0.0005914470114878602,
+      "loss": 3.5114,
+      "step": 4664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56264728307724,
+      "learning_rate": 0.0005914433609694158,
+      "loss": 3.5291,
+      "step": 4665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5429409146308899,
+      "learning_rate": 0.0005914397096833658,
+      "loss": 3.7083,
+      "step": 4666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5317074656486511,
+      "learning_rate": 0.0005914360576297196,
+      "loss": 3.4968,
+      "step": 4667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5686256885528564,
+      "learning_rate": 0.0005914324048084869,
+      "loss": 3.4443,
+      "step": 4668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5332667827606201,
+      "learning_rate": 0.0005914287512196775,
+      "loss": 3.525,
+      "step": 4669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5805628299713135,
+      "learning_rate": 0.0005914250968633006,
+      "loss": 3.5423,
+      "step": 4670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5518739819526672,
+      "learning_rate": 0.0005914214417393663,
+      "loss": 3.5086,
+      "step": 4671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6050487160682678,
+      "learning_rate": 0.0005914177858478837,
+      "loss": 3.4918,
+      "step": 4672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5575416088104248,
+      "learning_rate": 0.000591414129188863,
+      "loss": 3.4099,
+      "step": 4673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5643271803855896,
+      "learning_rate": 0.0005914104717623136,
+      "loss": 3.2613,
+      "step": 4674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5856224298477173,
+      "learning_rate": 0.000591406813568245,
+      "loss": 3.3773,
+      "step": 4675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5721701383590698,
+      "learning_rate": 0.0005914031546066669,
+      "loss": 3.3944,
+      "step": 4676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5789268016815186,
+      "learning_rate": 0.000591399494877589,
+      "loss": 3.5411,
+      "step": 4677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5379511713981628,
+      "learning_rate": 0.000591395834381021,
+      "loss": 3.5688,
+      "step": 4678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5450015068054199,
+      "learning_rate": 0.0005913921731169724,
+      "loss": 3.5541,
+      "step": 4679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5848316550254822,
+      "learning_rate": 0.0005913885110854529,
+      "loss": 3.628,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5671524405479431,
+      "learning_rate": 0.0005913848482864721,
+      "loss": 3.629,
+      "step": 4681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5497124791145325,
+      "learning_rate": 0.0005913811847200397,
+      "loss": 3.6733,
+      "step": 4682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5373255610466003,
+      "learning_rate": 0.0005913775203861654,
+      "loss": 3.5201,
+      "step": 4683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5729827880859375,
+      "learning_rate": 0.0005913738552848587,
+      "loss": 3.3868,
+      "step": 4684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5580928921699524,
+      "learning_rate": 0.0005913701894161294,
+      "loss": 3.4114,
+      "step": 4685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5716227889060974,
+      "learning_rate": 0.000591366522779987,
+      "loss": 3.5178,
+      "step": 4686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.539944052696228,
+      "learning_rate": 0.0005913628553764413,
+      "loss": 3.4705,
+      "step": 4687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5653459429740906,
+      "learning_rate": 0.000591359187205502,
+      "loss": 3.3099,
+      "step": 4688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5607462525367737,
+      "learning_rate": 0.0005913555182671786,
+      "loss": 3.6437,
+      "step": 4689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5527281165122986,
+      "learning_rate": 0.0005913518485614809,
+      "loss": 3.4185,
+      "step": 4690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5707682967185974,
+      "learning_rate": 0.0005913481780884185,
+      "loss": 3.4184,
+      "step": 4691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5300978422164917,
+      "learning_rate": 0.000591344506848001,
+      "loss": 3.2963,
+      "step": 4692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.553466260433197,
+      "learning_rate": 0.0005913408348402382,
+      "loss": 3.3923,
+      "step": 4693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5744982361793518,
+      "learning_rate": 0.0005913371620651396,
+      "loss": 3.3072,
+      "step": 4694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5825265645980835,
+      "learning_rate": 0.0005913334885227151,
+      "loss": 3.3266,
+      "step": 4695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5940053462982178,
+      "learning_rate": 0.0005913298142129741,
+      "loss": 3.2624,
+      "step": 4696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5737009644508362,
+      "learning_rate": 0.0005913261391359266,
+      "loss": 3.2402,
+      "step": 4697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5772563219070435,
+      "learning_rate": 0.0005913224632915821,
+      "loss": 3.541,
+      "step": 4698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5606215000152588,
+      "learning_rate": 0.0005913187866799503,
+      "loss": 3.2515,
+      "step": 4699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6204817295074463,
+      "learning_rate": 0.0005913151093010408,
+      "loss": 3.6065,
+      "step": 4700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5818078517913818,
+      "learning_rate": 0.0005913114311548635,
+      "loss": 3.582,
+      "step": 4701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.556207001209259,
+      "learning_rate": 0.0005913077522414278,
+      "loss": 3.4011,
+      "step": 4702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5517778992652893,
+      "learning_rate": 0.0005913040725607437,
+      "loss": 3.362,
+      "step": 4703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5277376174926758,
+      "learning_rate": 0.0005913003921128206,
+      "loss": 3.5635,
+      "step": 4704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5681208968162537,
+      "learning_rate": 0.0005912967108976684,
+      "loss": 3.4729,
+      "step": 4705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.554221510887146,
+      "learning_rate": 0.0005912930289152967,
+      "loss": 3.5178,
+      "step": 4706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5068381428718567,
+      "learning_rate": 0.0005912893461657152,
+      "loss": 3.5696,
+      "step": 4707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5582959055900574,
+      "learning_rate": 0.0005912856626489337,
+      "loss": 3.3457,
+      "step": 4708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5499424338340759,
+      "learning_rate": 0.0005912819783649617,
+      "loss": 3.5449,
+      "step": 4709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5410943627357483,
+      "learning_rate": 0.0005912782933138091,
+      "loss": 3.5805,
+      "step": 4710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5599467754364014,
+      "learning_rate": 0.0005912746074954856,
+      "loss": 3.4815,
+      "step": 4711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6126967072486877,
+      "learning_rate": 0.0005912709209100007,
+      "loss": 3.2459,
+      "step": 4712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5697960257530212,
+      "learning_rate": 0.0005912672335573643,
+      "loss": 3.4214,
+      "step": 4713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5213540196418762,
+      "learning_rate": 0.0005912635454375861,
+      "loss": 3.6434,
+      "step": 4714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.667456328868866,
+      "learning_rate": 0.0005912598565506758,
+      "loss": 3.4364,
+      "step": 4715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5707606077194214,
+      "learning_rate": 0.0005912561668966431,
+      "loss": 3.4236,
+      "step": 4716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.551876962184906,
+      "learning_rate": 0.0005912524764754976,
+      "loss": 3.4234,
+      "step": 4717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5813961029052734,
+      "learning_rate": 0.0005912487852872492,
+      "loss": 3.4069,
+      "step": 4718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5820927023887634,
+      "learning_rate": 0.0005912450933319075,
+      "loss": 3.429,
+      "step": 4719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6033832430839539,
+      "learning_rate": 0.0005912414006094824,
+      "loss": 3.5748,
+      "step": 4720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5736304521560669,
+      "learning_rate": 0.0005912377071199834,
+      "loss": 3.2263,
+      "step": 4721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.555508553981781,
+      "learning_rate": 0.0005912340128634205,
+      "loss": 3.3534,
+      "step": 4722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5986608266830444,
+      "learning_rate": 0.000591230317839803,
+      "loss": 3.4823,
+      "step": 4723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6043596267700195,
+      "learning_rate": 0.000591226622049141,
+      "loss": 3.4663,
+      "step": 4724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.540687620639801,
+      "learning_rate": 0.0005912229254914442,
+      "loss": 3.3754,
+      "step": 4725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6195804476737976,
+      "learning_rate": 0.0005912192281667221,
+      "loss": 3.39,
+      "step": 4726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5378120541572571,
+      "learning_rate": 0.0005912155300749846,
+      "loss": 3.5522,
+      "step": 4727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5554460287094116,
+      "learning_rate": 0.0005912118312162416,
+      "loss": 3.3615,
+      "step": 4728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5363547205924988,
+      "learning_rate": 0.0005912081315905026,
+      "loss": 3.3284,
+      "step": 4729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6088721752166748,
+      "learning_rate": 0.0005912044311977775,
+      "loss": 3.4141,
+      "step": 4730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5486693382263184,
+      "learning_rate": 0.0005912007300380758,
+      "loss": 3.3685,
+      "step": 4731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.558586597442627,
+      "learning_rate": 0.0005911970281114075,
+      "loss": 3.4063,
+      "step": 4732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5668739676475525,
+      "learning_rate": 0.0005911933254177824,
+      "loss": 3.3792,
+      "step": 4733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5825808644294739,
+      "learning_rate": 0.00059118962195721,
+      "loss": 3.6577,
+      "step": 4734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5412962436676025,
+      "learning_rate": 0.0005911859177297002,
+      "loss": 3.2624,
+      "step": 4735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5681858658790588,
+      "learning_rate": 0.0005911822127352628,
+      "loss": 3.3996,
+      "step": 4736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6266666650772095,
+      "learning_rate": 0.0005911785069739073,
+      "loss": 3.5886,
+      "step": 4737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5447356700897217,
+      "learning_rate": 0.0005911748004456438,
+      "loss": 3.3494,
+      "step": 4738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5406894683837891,
+      "learning_rate": 0.0005911710931504818,
+      "loss": 3.4121,
+      "step": 4739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5392957329750061,
+      "learning_rate": 0.0005911673850884313,
+      "loss": 3.3898,
+      "step": 4740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5871279239654541,
+      "learning_rate": 0.000591163676259502,
+      "loss": 3.081,
+      "step": 4741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6096645593643188,
+      "learning_rate": 0.0005911599666637035,
+      "loss": 3.508,
+      "step": 4742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5398980379104614,
+      "learning_rate": 0.0005911562563010457,
+      "loss": 3.5063,
+      "step": 4743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5835353136062622,
+      "learning_rate": 0.0005911525451715383,
+      "loss": 3.477,
+      "step": 4744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5687283277511597,
+      "learning_rate": 0.0005911488332751911,
+      "loss": 3.4038,
+      "step": 4745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5824450254440308,
+      "learning_rate": 0.000591145120612014,
+      "loss": 3.5479,
+      "step": 4746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5424936413764954,
+      "learning_rate": 0.0005911414071820167,
+      "loss": 3.4889,
+      "step": 4747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5513694882392883,
+      "learning_rate": 0.000591137692985209,
+      "loss": 3.4297,
+      "step": 4748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5256475210189819,
+      "learning_rate": 0.0005911339780216005,
+      "loss": 3.6925,
+      "step": 4749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5420495867729187,
+      "learning_rate": 0.0005911302622912013,
+      "loss": 3.6353,
+      "step": 4750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5440043807029724,
+      "learning_rate": 0.000591126545794021,
+      "loss": 3.3512,
+      "step": 4751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5715298652648926,
+      "learning_rate": 0.0005911228285300692,
+      "loss": 3.3806,
+      "step": 4752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5849560499191284,
+      "learning_rate": 0.0005911191104993561,
+      "loss": 3.4816,
+      "step": 4753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5608013272285461,
+      "learning_rate": 0.0005911153917018912,
+      "loss": 3.527,
+      "step": 4754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.584721028804779,
+      "learning_rate": 0.0005911116721376844,
+      "loss": 3.284,
+      "step": 4755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5911849737167358,
+      "learning_rate": 0.0005911079518067455,
+      "loss": 3.4843,
+      "step": 4756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.569872260093689,
+      "learning_rate": 0.0005911042307090843,
+      "loss": 3.5176,
+      "step": 4757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6122729778289795,
+      "learning_rate": 0.0005911005088447105,
+      "loss": 3.6845,
+      "step": 4758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.552301287651062,
+      "learning_rate": 0.0005910967862136341,
+      "loss": 3.5472,
+      "step": 4759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5530393719673157,
+      "learning_rate": 0.0005910930628158647,
+      "loss": 3.6068,
+      "step": 4760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5497402548789978,
+      "learning_rate": 0.0005910893386514122,
+      "loss": 3.3727,
+      "step": 4761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6003914475440979,
+      "learning_rate": 0.0005910856137202865,
+      "loss": 3.5103,
+      "step": 4762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5288448333740234,
+      "learning_rate": 0.0005910818880224973,
+      "loss": 3.4836,
+      "step": 4763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5383844375610352,
+      "learning_rate": 0.0005910781615580542,
+      "loss": 3.3261,
+      "step": 4764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5371490120887756,
+      "learning_rate": 0.0005910744343269674,
+      "loss": 3.5052,
+      "step": 4765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5520905256271362,
+      "learning_rate": 0.0005910707063292466,
+      "loss": 3.5325,
+      "step": 4766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5796260237693787,
+      "learning_rate": 0.0005910669775649016,
+      "loss": 3.1702,
+      "step": 4767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6260196566581726,
+      "learning_rate": 0.0005910632480339421,
+      "loss": 3.4242,
+      "step": 4768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5613893270492554,
+      "learning_rate": 0.0005910595177363781,
+      "loss": 3.6865,
+      "step": 4769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5682705044746399,
+      "learning_rate": 0.0005910557866722193,
+      "loss": 3.1459,
+      "step": 4770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6285190582275391,
+      "learning_rate": 0.0005910520548414756,
+      "loss": 3.3947,
+      "step": 4771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5714618563652039,
+      "learning_rate": 0.0005910483222441568,
+      "loss": 3.5817,
+      "step": 4772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5577132105827332,
+      "learning_rate": 0.0005910445888802727,
+      "loss": 3.4281,
+      "step": 4773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6071662306785583,
+      "learning_rate": 0.0005910408547498332,
+      "loss": 3.3626,
+      "step": 4774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5966734886169434,
+      "learning_rate": 0.000591037119852848,
+      "loss": 3.4659,
+      "step": 4775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5586175322532654,
+      "learning_rate": 0.0005910333841893271,
+      "loss": 3.3256,
+      "step": 4776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5855554938316345,
+      "learning_rate": 0.0005910296477592803,
+      "loss": 3.7103,
+      "step": 4777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52793288230896,
+      "learning_rate": 0.0005910259105627174,
+      "loss": 3.6772,
+      "step": 4778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5505519509315491,
+      "learning_rate": 0.0005910221725996483,
+      "loss": 3.278,
+      "step": 4779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5763952136039734,
+      "learning_rate": 0.0005910184338700828,
+      "loss": 3.4688,
+      "step": 4780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6565443873405457,
+      "learning_rate": 0.0005910146943740308,
+      "loss": 3.5173,
+      "step": 4781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.582163393497467,
+      "learning_rate": 0.0005910109541115019,
+      "loss": 3.1918,
+      "step": 4782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5900031328201294,
+      "learning_rate": 0.0005910072130825063,
+      "loss": 3.4327,
+      "step": 4783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6079528331756592,
+      "learning_rate": 0.0005910034712870538,
+      "loss": 3.4477,
+      "step": 4784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5783188939094543,
+      "learning_rate": 0.0005909997287251539,
+      "loss": 3.5502,
+      "step": 4785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5941360592842102,
+      "learning_rate": 0.000590995985396817,
+      "loss": 3.351,
+      "step": 4786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6161750555038452,
+      "learning_rate": 0.0005909922413020525,
+      "loss": 3.5587,
+      "step": 4787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5669690370559692,
+      "learning_rate": 0.0005909884964408705,
+      "loss": 3.5122,
+      "step": 4788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5724545121192932,
+      "learning_rate": 0.0005909847508132808,
+      "loss": 3.3534,
+      "step": 4789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5895712375640869,
+      "learning_rate": 0.0005909810044192932,
+      "loss": 3.4805,
+      "step": 4790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.542855441570282,
+      "learning_rate": 0.0005909772572589177,
+      "loss": 3.5939,
+      "step": 4791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5481388568878174,
+      "learning_rate": 0.0005909735093321642,
+      "loss": 3.4121,
+      "step": 4792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5825479626655579,
+      "learning_rate": 0.0005909697606390423,
+      "loss": 3.5954,
+      "step": 4793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5931483507156372,
+      "learning_rate": 0.0005909660111795621,
+      "loss": 3.2799,
+      "step": 4794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.530947208404541,
+      "learning_rate": 0.0005909622609537334,
+      "loss": 3.5938,
+      "step": 4795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5414092540740967,
+      "learning_rate": 0.0005909585099615662,
+      "loss": 3.6877,
+      "step": 4796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5826252698898315,
+      "learning_rate": 0.0005909547582030702,
+      "loss": 3.4782,
+      "step": 4797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5572245717048645,
+      "learning_rate": 0.0005909510056782554,
+      "loss": 3.3905,
+      "step": 4798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.564272403717041,
+      "learning_rate": 0.0005909472523871317,
+      "loss": 3.5094,
+      "step": 4799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5508630871772766,
+      "learning_rate": 0.0005909434983297089,
+      "loss": 3.501,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5623564720153809,
+      "learning_rate": 0.0005909397435059967,
+      "loss": 3.3671,
+      "step": 4801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5778913497924805,
+      "learning_rate": 0.0005909359879160055,
+      "loss": 3.5242,
+      "step": 4802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5555720329284668,
+      "learning_rate": 0.0005909322315597448,
+      "loss": 3.4082,
+      "step": 4803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5474628806114197,
+      "learning_rate": 0.0005909284744372245,
+      "loss": 3.3297,
+      "step": 4804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.534807026386261,
+      "learning_rate": 0.0005909247165484547,
+      "loss": 3.3332,
+      "step": 4805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.524055540561676,
+      "learning_rate": 0.0005909209578934452,
+      "loss": 3.583,
+      "step": 4806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5375897288322449,
+      "learning_rate": 0.0005909171984722058,
+      "loss": 3.4536,
+      "step": 4807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5646499395370483,
+      "learning_rate": 0.0005909134382847466,
+      "loss": 3.3704,
+      "step": 4808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5710200071334839,
+      "learning_rate": 0.0005909096773310773,
+      "loss": 3.5032,
+      "step": 4809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5503785610198975,
+      "learning_rate": 0.000590905915611208,
+      "loss": 3.4326,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5581929087638855,
+      "learning_rate": 0.0005909021531251483,
+      "loss": 3.5753,
+      "step": 4811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6241434216499329,
+      "learning_rate": 0.0005908983898729084,
+      "loss": 3.4697,
+      "step": 4812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6147471070289612,
+      "learning_rate": 0.0005908946258544982,
+      "loss": 3.2685,
+      "step": 4813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6412365436553955,
+      "learning_rate": 0.0005908908610699275,
+      "loss": 3.4211,
+      "step": 4814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5611572861671448,
+      "learning_rate": 0.0005908870955192064,
+      "loss": 3.4937,
+      "step": 4815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5360367298126221,
+      "learning_rate": 0.0005908833292023445,
+      "loss": 3.6016,
+      "step": 4816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.546721339225769,
+      "learning_rate": 0.0005908795621193519,
+      "loss": 3.2601,
+      "step": 4817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5728215575218201,
+      "learning_rate": 0.0005908757942702385,
+      "loss": 3.4103,
+      "step": 4818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.556920051574707,
+      "learning_rate": 0.0005908720256550143,
+      "loss": 3.3727,
+      "step": 4819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5450972318649292,
+      "learning_rate": 0.0005908682562736892,
+      "loss": 3.4042,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5655497908592224,
+      "learning_rate": 0.000590864486126273,
+      "loss": 3.3996,
+      "step": 4821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5585508942604065,
+      "learning_rate": 0.0005908607152127757,
+      "loss": 3.4254,
+      "step": 4822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.542323648929596,
+      "learning_rate": 0.0005908569435332072,
+      "loss": 3.457,
+      "step": 4823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5362521409988403,
+      "learning_rate": 0.0005908531710875777,
+      "loss": 3.5526,
+      "step": 4824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5256375670433044,
+      "learning_rate": 0.0005908493978758969,
+      "loss": 3.4572,
+      "step": 4825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5696166157722473,
+      "learning_rate": 0.0005908456238981747,
+      "loss": 3.4514,
+      "step": 4826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5578446984291077,
+      "learning_rate": 0.000590841849154421,
+      "loss": 3.5851,
+      "step": 4827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5300297737121582,
+      "learning_rate": 0.0005908380736446459,
+      "loss": 3.5034,
+      "step": 4828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5648180246353149,
+      "learning_rate": 0.0005908342973688594,
+      "loss": 3.599,
+      "step": 4829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6014158129692078,
+      "learning_rate": 0.0005908305203270712,
+      "loss": 3.4826,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5411103367805481,
+      "learning_rate": 0.0005908267425192914,
+      "loss": 3.4111,
+      "step": 4831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625185370445251,
+      "learning_rate": 0.00059082296394553,
+      "loss": 3.403,
+      "step": 4832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6025658845901489,
+      "learning_rate": 0.0005908191846057968,
+      "loss": 3.329,
+      "step": 4833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5506299734115601,
+      "learning_rate": 0.0005908154045001019,
+      "loss": 3.6375,
+      "step": 4834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5933164358139038,
+      "learning_rate": 0.0005908116236284551,
+      "loss": 3.2049,
+      "step": 4835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5729756951332092,
+      "learning_rate": 0.0005908078419908666,
+      "loss": 3.3538,
+      "step": 4836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5527672171592712,
+      "learning_rate": 0.000590804059587346,
+      "loss": 3.3438,
+      "step": 4837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5585240721702576,
+      "learning_rate": 0.0005908002764179037,
+      "loss": 3.5065,
+      "step": 4838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5482339262962341,
+      "learning_rate": 0.0005907964924825493,
+      "loss": 3.425,
+      "step": 4839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5115832686424255,
+      "learning_rate": 0.000590792707781293,
+      "loss": 3.4516,
+      "step": 4840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5796265006065369,
+      "learning_rate": 0.0005907889223141446,
+      "loss": 3.4701,
+      "step": 4841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5543243288993835,
+      "learning_rate": 0.0005907851360811142,
+      "loss": 3.2507,
+      "step": 4842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5635089874267578,
+      "learning_rate": 0.0005907813490822116,
+      "loss": 3.3394,
+      "step": 4843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5561545491218567,
+      "learning_rate": 0.000590777561317447,
+      "loss": 3.3604,
+      "step": 4844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5824812650680542,
+      "learning_rate": 0.0005907737727868302,
+      "loss": 3.5275,
+      "step": 4845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5519053339958191,
+      "learning_rate": 0.0005907699834903713,
+      "loss": 3.5574,
+      "step": 4846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5963732004165649,
+      "learning_rate": 0.0005907661934280801,
+      "loss": 3.4281,
+      "step": 4847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5388736724853516,
+      "learning_rate": 0.0005907624025999669,
+      "loss": 3.3095,
+      "step": 4848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5201981067657471,
+      "learning_rate": 0.0005907586110060414,
+      "loss": 3.2326,
+      "step": 4849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5812944769859314,
+      "learning_rate": 0.0005907548186463137,
+      "loss": 3.4784,
+      "step": 4850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5888471603393555,
+      "learning_rate": 0.0005907510255207936,
+      "loss": 3.5884,
+      "step": 4851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5519949793815613,
+      "learning_rate": 0.0005907472316294914,
+      "loss": 3.3247,
+      "step": 4852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5584381818771362,
+      "learning_rate": 0.0005907434369724169,
+      "loss": 3.4411,
+      "step": 4853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5249735713005066,
+      "learning_rate": 0.0005907396415495802,
+      "loss": 3.4842,
+      "step": 4854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5697634816169739,
+      "learning_rate": 0.0005907358453609912,
+      "loss": 3.5065,
+      "step": 4855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5187658071517944,
+      "learning_rate": 0.00059073204840666,
+      "loss": 3.3634,
+      "step": 4856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5130595564842224,
+      "learning_rate": 0.0005907282506865965,
+      "loss": 3.2991,
+      "step": 4857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5928294658660889,
+      "learning_rate": 0.0005907244522008107,
+      "loss": 3.4294,
+      "step": 4858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5565031170845032,
+      "learning_rate": 0.0005907206529493127,
+      "loss": 3.7071,
+      "step": 4859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5738771557807922,
+      "learning_rate": 0.0005907168529321124,
+      "loss": 3.3815,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5372669696807861,
+      "learning_rate": 0.0005907130521492198,
+      "loss": 3.4133,
+      "step": 4861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7971889972686768,
+      "learning_rate": 0.0005907092506006452,
+      "loss": 3.5911,
+      "step": 4862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6166131496429443,
+      "learning_rate": 0.0005907054482863981,
+      "loss": 3.4924,
+      "step": 4863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6117991209030151,
+      "learning_rate": 0.0005907016452064889,
+      "loss": 3.4626,
+      "step": 4864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5576360821723938,
+      "learning_rate": 0.0005906978413609276,
+      "loss": 3.4898,
+      "step": 4865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5783229470252991,
+      "learning_rate": 0.000590694036749724,
+      "loss": 3.3548,
+      "step": 4866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5609897375106812,
+      "learning_rate": 0.0005906902313728885,
+      "loss": 3.3822,
+      "step": 4867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5815325379371643,
+      "learning_rate": 0.0005906864252304305,
+      "loss": 3.5328,
+      "step": 4868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5757078528404236,
+      "learning_rate": 0.0005906826183223607,
+      "loss": 3.4499,
+      "step": 4869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703591108322144,
+      "learning_rate": 0.0005906788106486888,
+      "loss": 3.4756,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.551869809627533,
+      "learning_rate": 0.0005906750022094246,
+      "loss": 3.6005,
+      "step": 4871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.643150269985199,
+      "learning_rate": 0.0005906711930045787,
+      "loss": 3.4636,
+      "step": 4872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5789183378219604,
+      "learning_rate": 0.0005906673830341606,
+      "loss": 3.5697,
+      "step": 4873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5472760796546936,
+      "learning_rate": 0.0005906635722981807,
+      "loss": 3.5206,
+      "step": 4874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6037294268608093,
+      "learning_rate": 0.0005906597607966488,
+      "loss": 3.4146,
+      "step": 4875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5475078225135803,
+      "learning_rate": 0.000590655948529575,
+      "loss": 3.5947,
+      "step": 4876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5711190700531006,
+      "learning_rate": 0.0005906521354969695,
+      "loss": 3.2974,
+      "step": 4877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5613605380058289,
+      "learning_rate": 0.0005906483216988421,
+      "loss": 3.2993,
+      "step": 4878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5774216055870056,
+      "learning_rate": 0.000590644507135203,
+      "loss": 3.3731,
+      "step": 4879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5997605323791504,
+      "learning_rate": 0.0005906406918060622,
+      "loss": 3.5623,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5883737206459045,
+      "learning_rate": 0.0005906368757114298,
+      "loss": 3.6096,
+      "step": 4881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5710089802742004,
+      "learning_rate": 0.0005906330588513157,
+      "loss": 3.4704,
+      "step": 4882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.516525149345398,
+      "learning_rate": 0.0005906292412257301,
+      "loss": 3.4233,
+      "step": 4883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.585911750793457,
+      "learning_rate": 0.000590625422834683,
+      "loss": 3.4795,
+      "step": 4884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5662615299224854,
+      "learning_rate": 0.0005906216036781845,
+      "loss": 3.5267,
+      "step": 4885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5360583066940308,
+      "learning_rate": 0.0005906177837562446,
+      "loss": 3.7512,
+      "step": 4886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5965997576713562,
+      "learning_rate": 0.0005906139630688735,
+      "loss": 3.4446,
+      "step": 4887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5111775994300842,
+      "learning_rate": 0.0005906101416160809,
+      "loss": 3.4171,
+      "step": 4888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5461533069610596,
+      "learning_rate": 0.0005906063193978773,
+      "loss": 3.5264,
+      "step": 4889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.543048620223999,
+      "learning_rate": 0.0005906024964142726,
+      "loss": 3.2279,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5315186977386475,
+      "learning_rate": 0.0005905986726652769,
+      "loss": 3.5112,
+      "step": 4891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5514733791351318,
+      "learning_rate": 0.0005905948481509,
+      "loss": 3.1409,
+      "step": 4892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5441703200340271,
+      "learning_rate": 0.0005905910228711524,
+      "loss": 3.3357,
+      "step": 4893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5863307118415833,
+      "learning_rate": 0.0005905871968260438,
+      "loss": 3.5144,
+      "step": 4894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5528578162193298,
+      "learning_rate": 0.0005905833700155847,
+      "loss": 3.3631,
+      "step": 4895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.565290093421936,
+      "learning_rate": 0.0005905795424397848,
+      "loss": 3.3386,
+      "step": 4896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5290336012840271,
+      "learning_rate": 0.0005905757140986542,
+      "loss": 3.6237,
+      "step": 4897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5270799398422241,
+      "learning_rate": 0.0005905718849922031,
+      "loss": 3.4527,
+      "step": 4898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5542150735855103,
+      "learning_rate": 0.0005905680551204416,
+      "loss": 3.3004,
+      "step": 4899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5722207427024841,
+      "learning_rate": 0.0005905642244833799,
+      "loss": 3.4647,
+      "step": 4900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5691162943840027,
+      "learning_rate": 0.000590560393081028,
+      "loss": 3.4091,
+      "step": 4901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5475846529006958,
+      "learning_rate": 0.0005905565609133957,
+      "loss": 3.6877,
+      "step": 4902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5672742128372192,
+      "learning_rate": 0.0005905527279804934,
+      "loss": 3.4401,
+      "step": 4903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49786025285720825,
+      "learning_rate": 0.0005905488942823312,
+      "loss": 3.4634,
+      "step": 4904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5762016773223877,
+      "learning_rate": 0.0005905450598189191,
+      "loss": 3.3126,
+      "step": 4905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5646076798439026,
+      "learning_rate": 0.0005905412245902673,
+      "loss": 3.4117,
+      "step": 4906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6327397227287292,
+      "learning_rate": 0.0005905373885963857,
+      "loss": 3.3765,
+      "step": 4907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5711735486984253,
+      "learning_rate": 0.0005905335518372846,
+      "loss": 3.3918,
+      "step": 4908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5844196677207947,
+      "learning_rate": 0.0005905297143129741,
+      "loss": 3.3744,
+      "step": 4909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.603782057762146,
+      "learning_rate": 0.0005905258760234641,
+      "loss": 3.3061,
+      "step": 4910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5405983328819275,
+      "learning_rate": 0.000590522036968765,
+      "loss": 3.6347,
+      "step": 4911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5765902400016785,
+      "learning_rate": 0.0005905181971488867,
+      "loss": 3.4907,
+      "step": 4912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6272974014282227,
+      "learning_rate": 0.0005905143565638394,
+      "loss": 3.2816,
+      "step": 4913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57854825258255,
+      "learning_rate": 0.0005905105152136332,
+      "loss": 3.4262,
+      "step": 4914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578022301197052,
+      "learning_rate": 0.0005905066730982782,
+      "loss": 3.321,
+      "step": 4915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5480340123176575,
+      "learning_rate": 0.0005905028302177845,
+      "loss": 3.5696,
+      "step": 4916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.529426634311676,
+      "learning_rate": 0.0005904989865721622,
+      "loss": 3.4581,
+      "step": 4917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.523251473903656,
+      "learning_rate": 0.0005904951421614216,
+      "loss": 3.5019,
+      "step": 4918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5666603446006775,
+      "learning_rate": 0.0005904912969855727,
+      "loss": 3.5147,
+      "step": 4919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6100955605506897,
+      "learning_rate": 0.0005904874510446256,
+      "loss": 3.4469,
+      "step": 4920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5637255907058716,
+      "learning_rate": 0.0005904836043385905,
+      "loss": 3.3005,
+      "step": 4921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5765202641487122,
+      "learning_rate": 0.0005904797568674774,
+      "loss": 3.4327,
+      "step": 4922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5553877353668213,
+      "learning_rate": 0.0005904759086312965,
+      "loss": 3.3015,
+      "step": 4923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5453650951385498,
+      "learning_rate": 0.0005904720596300581,
+      "loss": 3.4105,
+      "step": 4924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5456913113594055,
+      "learning_rate": 0.0005904682098637721,
+      "loss": 3.4166,
+      "step": 4925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5444839596748352,
+      "learning_rate": 0.0005904643593324487,
+      "loss": 3.5552,
+      "step": 4926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5601552724838257,
+      "learning_rate": 0.0005904605080360982,
+      "loss": 3.5365,
+      "step": 4927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5500659346580505,
+      "learning_rate": 0.0005904566559747305,
+      "loss": 3.4039,
+      "step": 4928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5786320567131042,
+      "learning_rate": 0.000590452803148356,
+      "loss": 3.4035,
+      "step": 4929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5525652766227722,
+      "learning_rate": 0.0005904489495569846,
+      "loss": 3.2293,
+      "step": 4930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5330731272697449,
+      "learning_rate": 0.0005904450952006267,
+      "loss": 3.5334,
+      "step": 4931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.551021158695221,
+      "learning_rate": 0.0005904412400792921,
+      "loss": 3.4606,
+      "step": 4932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5765814781188965,
+      "learning_rate": 0.0005904373841929913,
+      "loss": 3.5593,
+      "step": 4933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5668351054191589,
+      "learning_rate": 0.0005904335275417344,
+      "loss": 3.3573,
+      "step": 4934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5682421326637268,
+      "learning_rate": 0.0005904296701255314,
+      "loss": 3.3952,
+      "step": 4935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5964768528938293,
+      "learning_rate": 0.0005904258119443925,
+      "loss": 3.5121,
+      "step": 4936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5435137748718262,
+      "learning_rate": 0.000590421952998328,
+      "loss": 3.4751,
+      "step": 4937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703478455543518,
+      "learning_rate": 0.0005904180932873479,
+      "loss": 3.3983,
+      "step": 4938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5706227421760559,
+      "learning_rate": 0.0005904142328114624,
+      "loss": 3.2301,
+      "step": 4939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5686094164848328,
+      "learning_rate": 0.0005904103715706819,
+      "loss": 3.5421,
+      "step": 4940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6083675026893616,
+      "learning_rate": 0.0005904065095650162,
+      "loss": 3.4906,
+      "step": 4941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6154701113700867,
+      "learning_rate": 0.0005904026467944757,
+      "loss": 3.319,
+      "step": 4942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.589542031288147,
+      "learning_rate": 0.0005903987832590703,
+      "loss": 3.4153,
+      "step": 4943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5331887602806091,
+      "learning_rate": 0.0005903949189588107,
+      "loss": 3.37,
+      "step": 4944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5606480240821838,
+      "learning_rate": 0.0005903910538937067,
+      "loss": 3.4934,
+      "step": 4945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6116060614585876,
+      "learning_rate": 0.0005903871880637684,
+      "loss": 3.5101,
+      "step": 4946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5732858180999756,
+      "learning_rate": 0.0005903833214690064,
+      "loss": 3.4792,
+      "step": 4947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5375874638557434,
+      "learning_rate": 0.0005903794541094304,
+      "loss": 3.619,
+      "step": 4948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.552217423915863,
+      "learning_rate": 0.0005903755859850508,
+      "loss": 3.3901,
+      "step": 4949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.544503927230835,
+      "learning_rate": 0.0005903717170958779,
+      "loss": 3.3962,
+      "step": 4950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5237309336662292,
+      "learning_rate": 0.0005903678474419217,
+      "loss": 3.4463,
+      "step": 4951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5415375828742981,
+      "learning_rate": 0.0005903639770231925,
+      "loss": 3.2885,
+      "step": 4952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5329828858375549,
+      "learning_rate": 0.0005903601058397005,
+      "loss": 3.2622,
+      "step": 4953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5593845248222351,
+      "learning_rate": 0.0005903562338914558,
+      "loss": 3.6063,
+      "step": 4954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.689546525478363,
+      "learning_rate": 0.0005903523611784687,
+      "loss": 3.3436,
+      "step": 4955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5579712986946106,
+      "learning_rate": 0.0005903484877007493,
+      "loss": 3.392,
+      "step": 4956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5876078009605408,
+      "learning_rate": 0.0005903446134583079,
+      "loss": 3.5828,
+      "step": 4957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6026819348335266,
+      "learning_rate": 0.0005903407384511547,
+      "loss": 3.5553,
+      "step": 4958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5198916792869568,
+      "learning_rate": 0.0005903368626793,
+      "loss": 3.4045,
+      "step": 4959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5764784216880798,
+      "learning_rate": 0.0005903329861427537,
+      "loss": 3.3433,
+      "step": 4960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5851700305938721,
+      "learning_rate": 0.0005903291088415261,
+      "loss": 3.554,
+      "step": 4961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5475395321846008,
+      "learning_rate": 0.0005903252307756276,
+      "loss": 3.3761,
+      "step": 4962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6040002107620239,
+      "learning_rate": 0.0005903213519450683,
+      "loss": 3.7124,
+      "step": 4963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5730305910110474,
+      "learning_rate": 0.0005903174723498586,
+      "loss": 3.4037,
+      "step": 4964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5729190111160278,
+      "learning_rate": 0.0005903135919900084,
+      "loss": 3.1942,
+      "step": 4965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6518261432647705,
+      "learning_rate": 0.000590309710865528,
+      "loss": 3.5579,
+      "step": 4966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5518261790275574,
+      "learning_rate": 0.0005903058289764279,
+      "loss": 3.6733,
+      "step": 4967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5408662557601929,
+      "learning_rate": 0.000590301946322718,
+      "loss": 3.2613,
+      "step": 4968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5802552700042725,
+      "learning_rate": 0.0005902980629044085,
+      "loss": 3.5049,
+      "step": 4969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5875983238220215,
+      "learning_rate": 0.00059029417872151,
+      "loss": 3.3783,
+      "step": 4970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5996633768081665,
+      "learning_rate": 0.0005902902937740325,
+      "loss": 3.516,
+      "step": 4971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5604908466339111,
+      "learning_rate": 0.0005902864080619861,
+      "loss": 3.4725,
+      "step": 4972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5704894065856934,
+      "learning_rate": 0.0005902825215853812,
+      "loss": 3.3333,
+      "step": 4973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5649724006652832,
+      "learning_rate": 0.000590278634344228,
+      "loss": 3.3508,
+      "step": 4974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5423358678817749,
+      "learning_rate": 0.0005902747463385368,
+      "loss": 3.4517,
+      "step": 4975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5924620032310486,
+      "learning_rate": 0.0005902708575683177,
+      "loss": 3.4768,
+      "step": 4976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5734094381332397,
+      "learning_rate": 0.0005902669680335811,
+      "loss": 3.3845,
+      "step": 4977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5569986701011658,
+      "learning_rate": 0.0005902630777343372,
+      "loss": 3.5556,
+      "step": 4978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.555212676525116,
+      "learning_rate": 0.0005902591866705962,
+      "loss": 3.3604,
+      "step": 4979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5357699990272522,
+      "learning_rate": 0.0005902552948423682,
+      "loss": 3.4253,
+      "step": 4980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.539989173412323,
+      "learning_rate": 0.0005902514022496638,
+      "loss": 3.2653,
+      "step": 4981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.539340615272522,
+      "learning_rate": 0.0005902475088924929,
+      "loss": 3.6315,
+      "step": 4982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5658690929412842,
+      "learning_rate": 0.0005902436147708661,
+      "loss": 3.3263,
+      "step": 4983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5437087416648865,
+      "learning_rate": 0.0005902397198847934,
+      "loss": 3.3987,
+      "step": 4984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5789981484413147,
+      "learning_rate": 0.0005902358242342851,
+      "loss": 3.5408,
+      "step": 4985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5366504192352295,
+      "learning_rate": 0.0005902319278193516,
+      "loss": 3.4576,
+      "step": 4986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5638992190361023,
+      "learning_rate": 0.000590228030640003,
+      "loss": 3.4145,
+      "step": 4987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5469755530357361,
+      "learning_rate": 0.0005902241326962496,
+      "loss": 3.3786,
+      "step": 4988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5611241459846497,
+      "learning_rate": 0.0005902202339881017,
+      "loss": 3.4466,
+      "step": 4989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5863252282142639,
+      "learning_rate": 0.0005902163345155696,
+      "loss": 3.3491,
+      "step": 4990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.577661395072937,
+      "learning_rate": 0.0005902124342786635,
+      "loss": 3.1473,
+      "step": 4991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5453828573226929,
+      "learning_rate": 0.0005902085332773935,
+      "loss": 3.3708,
+      "step": 4992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5631303787231445,
+      "learning_rate": 0.0005902046315117704,
+      "loss": 3.4,
+      "step": 4993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5475294589996338,
+      "learning_rate": 0.000590200728981804,
+      "loss": 3.3217,
+      "step": 4994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5487865805625916,
+      "learning_rate": 0.0005901968256875047,
+      "loss": 3.3504,
+      "step": 4995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5545656085014343,
+      "learning_rate": 0.0005901929216288828,
+      "loss": 3.4953,
+      "step": 4996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5044790506362915,
+      "learning_rate": 0.0005901890168059487,
+      "loss": 3.142,
+      "step": 4997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5951768755912781,
+      "learning_rate": 0.0005901851112187125,
+      "loss": 3.4991,
+      "step": 4998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.540361225605011,
+      "learning_rate": 0.0005901812048671847,
+      "loss": 3.4398,
+      "step": 4999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5717666745185852,
+      "learning_rate": 0.0005901772977513754,
+      "loss": 3.3718,
+      "step": 5000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859256386756897,
+      "learning_rate": 0.0005901733898712948,
+      "loss": 3.2956,
+      "step": 5001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5552880764007568,
+      "learning_rate": 0.0005901694812269533,
+      "loss": 3.5566,
+      "step": 5002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5178077220916748,
+      "learning_rate": 0.0005901655718183615,
+      "loss": 3.503,
+      "step": 5003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5720773339271545,
+      "learning_rate": 0.0005901616616455292,
+      "loss": 3.2363,
+      "step": 5004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5559942126274109,
+      "learning_rate": 0.0005901577507084671,
+      "loss": 3.4998,
+      "step": 5005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5338559746742249,
+      "learning_rate": 0.0005901538390071853,
+      "loss": 3.6233,
+      "step": 5006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5544248223304749,
+      "learning_rate": 0.0005901499265416939,
+      "loss": 3.4523,
+      "step": 5007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5183117985725403,
+      "learning_rate": 0.0005901460133120037,
+      "loss": 3.4746,
+      "step": 5008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5503129959106445,
+      "learning_rate": 0.0005901420993181247,
+      "loss": 3.5572,
+      "step": 5009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6068310141563416,
+      "learning_rate": 0.000590138184560067,
+      "loss": 3.5988,
+      "step": 5010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5316990613937378,
+      "learning_rate": 0.0005901342690378413,
+      "loss": 3.6803,
+      "step": 5011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.516459047794342,
+      "learning_rate": 0.0005901303527514579,
+      "loss": 3.3902,
+      "step": 5012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5768346786499023,
+      "learning_rate": 0.0005901264357009269,
+      "loss": 3.572,
+      "step": 5013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5403907299041748,
+      "learning_rate": 0.0005901225178862588,
+      "loss": 3.5612,
+      "step": 5014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5107371807098389,
+      "learning_rate": 0.0005901185993074637,
+      "loss": 3.4745,
+      "step": 5015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.527996301651001,
+      "learning_rate": 0.0005901146799645521,
+      "loss": 3.5359,
+      "step": 5016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5773131847381592,
+      "learning_rate": 0.0005901107598575342,
+      "loss": 3.4109,
+      "step": 5017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5386675596237183,
+      "learning_rate": 0.0005901068389864204,
+      "loss": 3.4633,
+      "step": 5018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5442395806312561,
+      "learning_rate": 0.0005901029173512211,
+      "loss": 3.5547,
+      "step": 5019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5196928977966309,
+      "learning_rate": 0.0005900989949519465,
+      "loss": 3.5074,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5516992211341858,
+      "learning_rate": 0.000590095071788607,
+      "loss": 3.3833,
+      "step": 5021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5801973342895508,
+      "learning_rate": 0.0005900911478612129,
+      "loss": 3.2449,
+      "step": 5022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5835938453674316,
+      "learning_rate": 0.0005900872231697745,
+      "loss": 3.4805,
+      "step": 5023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5422643423080444,
+      "learning_rate": 0.0005900832977143022,
+      "loss": 3.347,
+      "step": 5024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5615981221199036,
+      "learning_rate": 0.0005900793714948065,
+      "loss": 3.234,
+      "step": 5025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.616595447063446,
+      "learning_rate": 0.0005900754445112974,
+      "loss": 3.3415,
+      "step": 5026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5145601034164429,
+      "learning_rate": 0.0005900715167637854,
+      "loss": 3.222,
+      "step": 5027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5435516834259033,
+      "learning_rate": 0.0005900675882522808,
+      "loss": 3.5811,
+      "step": 5028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5470258593559265,
+      "learning_rate": 0.0005900636589767941,
+      "loss": 3.4718,
+      "step": 5029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5767030119895935,
+      "learning_rate": 0.0005900597289373355,
+      "loss": 3.4194,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5893784761428833,
+      "learning_rate": 0.0005900557981339153,
+      "loss": 3.4108,
+      "step": 5031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5466517806053162,
+      "learning_rate": 0.0005900518665665441,
+      "loss": 3.439,
+      "step": 5032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.544185996055603,
+      "learning_rate": 0.000590047934235232,
+      "loss": 3.5536,
+      "step": 5033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5449085235595703,
+      "learning_rate": 0.0005900440011399895,
+      "loss": 3.527,
+      "step": 5034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5282718539237976,
+      "learning_rate": 0.000590040067280827,
+      "loss": 3.6722,
+      "step": 5035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5545849204063416,
+      "learning_rate": 0.0005900361326577547,
+      "loss": 3.5084,
+      "step": 5036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5555055141448975,
+      "learning_rate": 0.0005900321972707831,
+      "loss": 3.2376,
+      "step": 5037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5830196142196655,
+      "learning_rate": 0.0005900282611199225,
+      "loss": 3.3006,
+      "step": 5038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5802530646324158,
+      "learning_rate": 0.0005900243242051833,
+      "loss": 3.2766,
+      "step": 5039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5483977198600769,
+      "learning_rate": 0.0005900203865265757,
+      "loss": 3.5217,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5744450092315674,
+      "learning_rate": 0.0005900164480841104,
+      "loss": 3.4969,
+      "step": 5041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5572537779808044,
+      "learning_rate": 0.0005900125088777974,
+      "loss": 3.4738,
+      "step": 5042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5599087476730347,
+      "learning_rate": 0.0005900085689076474,
+      "loss": 3.4334,
+      "step": 5043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5435726046562195,
+      "learning_rate": 0.0005900046281736707,
+      "loss": 3.5784,
+      "step": 5044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5599098205566406,
+      "learning_rate": 0.0005900006866758774,
+      "loss": 3.4954,
+      "step": 5045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5361737608909607,
+      "learning_rate": 0.0005899967444142783,
+      "loss": 3.5627,
+      "step": 5046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5649425387382507,
+      "learning_rate": 0.0005899928013888835,
+      "loss": 3.447,
+      "step": 5047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.568342924118042,
+      "learning_rate": 0.0005899888575997035,
+      "loss": 3.4439,
+      "step": 5048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5820963382720947,
+      "learning_rate": 0.0005899849130467488,
+      "loss": 3.4019,
+      "step": 5049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5288267135620117,
+      "learning_rate": 0.0005899809677300294,
+      "loss": 3.6183,
+      "step": 5050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5664812326431274,
+      "learning_rate": 0.0005899770216495561,
+      "loss": 3.4155,
+      "step": 5051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5544307827949524,
+      "learning_rate": 0.000589973074805339,
+      "loss": 3.1679,
+      "step": 5052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5682546496391296,
+      "learning_rate": 0.0005899691271973888,
+      "loss": 3.3364,
+      "step": 5053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5925281643867493,
+      "learning_rate": 0.0005899651788257155,
+      "loss": 3.2977,
+      "step": 5054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5429676175117493,
+      "learning_rate": 0.00058996122969033,
+      "loss": 3.4784,
+      "step": 5055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5814798474311829,
+      "learning_rate": 0.0005899572797912422,
+      "loss": 3.2982,
+      "step": 5056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5307891964912415,
+      "learning_rate": 0.0005899533291284628,
+      "loss": 3.6792,
+      "step": 5057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.570961594581604,
+      "learning_rate": 0.0005899493777020021,
+      "loss": 3.4474,
+      "step": 5058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5177574157714844,
+      "learning_rate": 0.0005899454255118707,
+      "loss": 3.3287,
+      "step": 5059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5786210894584656,
+      "learning_rate": 0.0005899414725580787,
+      "loss": 3.4678,
+      "step": 5060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5470947623252869,
+      "learning_rate": 0.0005899375188406366,
+      "loss": 3.5344,
+      "step": 5061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5230193138122559,
+      "learning_rate": 0.000589933564359555,
+      "loss": 3.5553,
+      "step": 5062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5864706635475159,
+      "learning_rate": 0.0005899296091148442,
+      "loss": 3.3547,
+      "step": 5063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5561924576759338,
+      "learning_rate": 0.0005899256531065145,
+      "loss": 3.6271,
+      "step": 5064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5769221186637878,
+      "learning_rate": 0.0005899216963345766,
+      "loss": 3.4765,
+      "step": 5065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5741177201271057,
+      "learning_rate": 0.0005899177387990406,
+      "loss": 3.5332,
+      "step": 5066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.550760805606842,
+      "learning_rate": 0.000589913780499917,
+      "loss": 3.583,
+      "step": 5067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6087910532951355,
+      "learning_rate": 0.0005899098214372164,
+      "loss": 3.2774,
+      "step": 5068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6020672917366028,
+      "learning_rate": 0.0005899058616109492,
+      "loss": 3.5222,
+      "step": 5069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5288599133491516,
+      "learning_rate": 0.0005899019010211258,
+      "loss": 3.5523,
+      "step": 5070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6012958288192749,
+      "learning_rate": 0.0005898979396677564,
+      "loss": 3.5154,
+      "step": 5071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5777899622917175,
+      "learning_rate": 0.0005898939775508517,
+      "loss": 3.2862,
+      "step": 5072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5450665950775146,
+      "learning_rate": 0.000589890014670422,
+      "loss": 3.2721,
+      "step": 5073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5397341847419739,
+      "learning_rate": 0.0005898860510264778,
+      "loss": 3.3992,
+      "step": 5074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5461152195930481,
+      "learning_rate": 0.0005898820866190297,
+      "loss": 3.4282,
+      "step": 5075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5408706665039062,
+      "learning_rate": 0.0005898781214480878,
+      "loss": 3.5096,
+      "step": 5076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5572208166122437,
+      "learning_rate": 0.0005898741555136627,
+      "loss": 3.4326,
+      "step": 5077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5309155583381653,
+      "learning_rate": 0.0005898701888157649,
+      "loss": 3.3239,
+      "step": 5078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5635367035865784,
+      "learning_rate": 0.0005898662213544048,
+      "loss": 3.2986,
+      "step": 5079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5413909554481506,
+      "learning_rate": 0.0005898622531295929,
+      "loss": 3.5096,
+      "step": 5080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5198115110397339,
+      "learning_rate": 0.0005898582841413397,
+      "loss": 3.3075,
+      "step": 5081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5086342692375183,
+      "learning_rate": 0.0005898543143896555,
+      "loss": 3.3621,
+      "step": 5082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6958548426628113,
+      "learning_rate": 0.0005898503438745507,
+      "loss": 3.461,
+      "step": 5083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5100436210632324,
+      "learning_rate": 0.0005898463725960359,
+      "loss": 3.3348,
+      "step": 5084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6006938815116882,
+      "learning_rate": 0.0005898424005541216,
+      "loss": 3.4243,
+      "step": 5085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5753837823867798,
+      "learning_rate": 0.0005898384277488182,
+      "loss": 3.5554,
+      "step": 5086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5907991528511047,
+      "learning_rate": 0.0005898344541801362,
+      "loss": 3.4331,
+      "step": 5087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5146076679229736,
+      "learning_rate": 0.000589830479848086,
+      "loss": 3.3868,
+      "step": 5088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5464553833007812,
+      "learning_rate": 0.0005898265047526781,
+      "loss": 3.468,
+      "step": 5089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5454960465431213,
+      "learning_rate": 0.000589822528893923,
+      "loss": 3.4598,
+      "step": 5090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5484764575958252,
+      "learning_rate": 0.000589818552271831,
+      "loss": 3.3738,
+      "step": 5091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.554803729057312,
+      "learning_rate": 0.0005898145748864128,
+      "loss": 3.1957,
+      "step": 5092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5293271541595459,
+      "learning_rate": 0.0005898105967376788,
+      "loss": 3.5345,
+      "step": 5093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.56730717420578,
+      "learning_rate": 0.0005898066178256394,
+      "loss": 3.4173,
+      "step": 5094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5431801080703735,
+      "learning_rate": 0.0005898026381503051,
+      "loss": 3.3932,
+      "step": 5095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6341773867607117,
+      "learning_rate": 0.0005897986577116865,
+      "loss": 3.318,
+      "step": 5096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5350258946418762,
+      "learning_rate": 0.0005897946765097941,
+      "loss": 3.4459,
+      "step": 5097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5501015186309814,
+      "learning_rate": 0.0005897906945446382,
+      "loss": 3.2029,
+      "step": 5098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5346304178237915,
+      "learning_rate": 0.0005897867118162294,
+      "loss": 3.4928,
+      "step": 5099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6124736070632935,
+      "learning_rate": 0.0005897827283245781,
+      "loss": 3.65,
+      "step": 5100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5016577839851379,
+      "learning_rate": 0.000589778744069695,
+      "loss": 3.3579,
+      "step": 5101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.541294276714325,
+      "learning_rate": 0.0005897747590515904,
+      "loss": 3.1981,
+      "step": 5102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5059472322463989,
+      "learning_rate": 0.0005897707732702748,
+      "loss": 3.3077,
+      "step": 5103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5518838763237,
+      "learning_rate": 0.0005897667867257588,
+      "loss": 3.3557,
+      "step": 5104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5457679033279419,
+      "learning_rate": 0.0005897627994180529,
+      "loss": 3.3617,
+      "step": 5105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6038949489593506,
+      "learning_rate": 0.0005897588113471674,
+      "loss": 3.2915,
+      "step": 5106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.590950071811676,
+      "learning_rate": 0.000589754822513113,
+      "loss": 3.1439,
+      "step": 5107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5621350407600403,
+      "learning_rate": 0.0005897508329159003,
+      "loss": 3.2661,
+      "step": 5108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5230415463447571,
+      "learning_rate": 0.0005897468425555395,
+      "loss": 3.5213,
+      "step": 5109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5285176038742065,
+      "learning_rate": 0.0005897428514320414,
+      "loss": 3.3294,
+      "step": 5110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5499192476272583,
+      "learning_rate": 0.0005897388595454164,
+      "loss": 3.4348,
+      "step": 5111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5603438019752502,
+      "learning_rate": 0.000589734866895675,
+      "loss": 3.4721,
+      "step": 5112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5261187553405762,
+      "learning_rate": 0.0005897308734828276,
+      "loss": 3.4034,
+      "step": 5113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5853512287139893,
+      "learning_rate": 0.0005897268793068848,
+      "loss": 3.3507,
+      "step": 5114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.586706817150116,
+      "learning_rate": 0.0005897228843678574,
+      "loss": 3.2244,
+      "step": 5115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5517246127128601,
+      "learning_rate": 0.0005897188886657555,
+      "loss": 3.5185,
+      "step": 5116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5283602476119995,
+      "learning_rate": 0.00058971489220059,
+      "loss": 3.2524,
+      "step": 5117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5656755566596985,
+      "learning_rate": 0.000589710894972371,
+      "loss": 3.5088,
+      "step": 5118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5920532941818237,
+      "learning_rate": 0.0005897068969811094,
+      "loss": 3.5174,
+      "step": 5119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5770808458328247,
+      "learning_rate": 0.0005897028982268155,
+      "loss": 3.304,
+      "step": 5120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.542290210723877,
+      "learning_rate": 0.0005896988987094999,
+      "loss": 3.5272,
+      "step": 5121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5746126770973206,
+      "learning_rate": 0.0005896948984291733,
+      "loss": 3.6472,
+      "step": 5122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5588904023170471,
+      "learning_rate": 0.0005896908973858459,
+      "loss": 3.5337,
+      "step": 5123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5217200517654419,
+      "learning_rate": 0.0005896868955795286,
+      "loss": 3.3863,
+      "step": 5124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5870938897132874,
+      "learning_rate": 0.0005896828930102316,
+      "loss": 3.3279,
+      "step": 5125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6563788652420044,
+      "learning_rate": 0.0005896788896779658,
+      "loss": 3.2821,
+      "step": 5126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5480024218559265,
+      "learning_rate": 0.0005896748855827414,
+      "loss": 3.3585,
+      "step": 5127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5863211154937744,
+      "learning_rate": 0.0005896708807245693,
+      "loss": 3.4657,
+      "step": 5128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6124269366264343,
+      "learning_rate": 0.0005896668751034596,
+      "loss": 3.5115,
+      "step": 5129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6593347191810608,
+      "learning_rate": 0.0005896628687194232,
+      "loss": 3.3722,
+      "step": 5130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5311346054077148,
+      "learning_rate": 0.0005896588615724706,
+      "loss": 3.2849,
+      "step": 5131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5981443524360657,
+      "learning_rate": 0.0005896548536626123,
+      "loss": 3.2468,
+      "step": 5132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5324131846427917,
+      "learning_rate": 0.0005896508449898587,
+      "loss": 3.4398,
+      "step": 5133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5485259294509888,
+      "learning_rate": 0.0005896468355542206,
+      "loss": 3.3877,
+      "step": 5134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6059028506278992,
+      "learning_rate": 0.0005896428253557086,
+      "loss": 3.2956,
+      "step": 5135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5261827111244202,
+      "learning_rate": 0.0005896388143943331,
+      "loss": 3.5208,
+      "step": 5136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.596929669380188,
+      "learning_rate": 0.0005896348026701047,
+      "loss": 3.3708,
+      "step": 5137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5352926254272461,
+      "learning_rate": 0.0005896307901830339,
+      "loss": 3.3242,
+      "step": 5138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5876359939575195,
+      "learning_rate": 0.0005896267769331313,
+      "loss": 3.6109,
+      "step": 5139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5615381598472595,
+      "learning_rate": 0.0005896227629204077,
+      "loss": 3.2824,
+      "step": 5140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5441862940788269,
+      "learning_rate": 0.0005896187481448734,
+      "loss": 3.316,
+      "step": 5141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5573780536651611,
+      "learning_rate": 0.000589614732606539,
+      "loss": 3.3629,
+      "step": 5142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5792946815490723,
+      "learning_rate": 0.0005896107163054151,
+      "loss": 3.4812,
+      "step": 5143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5667078495025635,
+      "learning_rate": 0.0005896066992415125,
+      "loss": 3.4833,
+      "step": 5144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5536587834358215,
+      "learning_rate": 0.0005896026814148414,
+      "loss": 3.3131,
+      "step": 5145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5453284978866577,
+      "learning_rate": 0.0005895986628254126,
+      "loss": 3.518,
+      "step": 5146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6266303062438965,
+      "learning_rate": 0.0005895946434732366,
+      "loss": 3.3667,
+      "step": 5147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5294222831726074,
+      "learning_rate": 0.0005895906233583242,
+      "loss": 3.3286,
+      "step": 5148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6126430034637451,
+      "learning_rate": 0.0005895866024806858,
+      "loss": 3.3122,
+      "step": 5149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5357118248939514,
+      "learning_rate": 0.0005895825808403321,
+      "loss": 3.3943,
+      "step": 5150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5523558855056763,
+      "learning_rate": 0.0005895785584372734,
+      "loss": 3.5076,
+      "step": 5151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7012884020805359,
+      "learning_rate": 0.0005895745352715205,
+      "loss": 3.0987,
+      "step": 5152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5519027709960938,
+      "learning_rate": 0.0005895705113430842,
+      "loss": 3.6015,
+      "step": 5153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6840552091598511,
+      "learning_rate": 0.0005895664866519748,
+      "loss": 3.2324,
+      "step": 5154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5728353261947632,
+      "learning_rate": 0.000589562461198203,
+      "loss": 3.5145,
+      "step": 5155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5669745802879333,
+      "learning_rate": 0.0005895584349817794,
+      "loss": 3.2349,
+      "step": 5156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.551141083240509,
+      "learning_rate": 0.0005895544080027147,
+      "loss": 3.4344,
+      "step": 5157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5899537205696106,
+      "learning_rate": 0.0005895503802610193,
+      "loss": 3.5101,
+      "step": 5158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.541935384273529,
+      "learning_rate": 0.000589546351756704,
+      "loss": 3.6462,
+      "step": 5159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5857089757919312,
+      "learning_rate": 0.0005895423224897792,
+      "loss": 3.5987,
+      "step": 5160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5397253632545471,
+      "learning_rate": 0.0005895382924602558,
+      "loss": 3.4983,
+      "step": 5161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5339411497116089,
+      "learning_rate": 0.0005895342616681442,
+      "loss": 3.3218,
+      "step": 5162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5548738241195679,
+      "learning_rate": 0.0005895302301134551,
+      "loss": 3.5583,
+      "step": 5163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5891041159629822,
+      "learning_rate": 0.0005895261977961991,
+      "loss": 3.4377,
+      "step": 5164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.569484531879425,
+      "learning_rate": 0.0005895221647163867,
+      "loss": 3.3795,
+      "step": 5165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5073286890983582,
+      "learning_rate": 0.0005895181308740288,
+      "loss": 3.6983,
+      "step": 5166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.550396203994751,
+      "learning_rate": 0.0005895140962691358,
+      "loss": 3.3491,
+      "step": 5167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.502250611782074,
+      "learning_rate": 0.0005895100609017184,
+      "loss": 3.3437,
+      "step": 5168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5587743520736694,
+      "learning_rate": 0.0005895060247717871,
+      "loss": 3.6706,
+      "step": 5169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5262305736541748,
+      "learning_rate": 0.0005895019878793527,
+      "loss": 3.4108,
+      "step": 5170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5481409430503845,
+      "learning_rate": 0.0005894979502244259,
+      "loss": 3.4605,
+      "step": 5171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5923982858657837,
+      "learning_rate": 0.0005894939118070171,
+      "loss": 3.4484,
+      "step": 5172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5860485434532166,
+      "learning_rate": 0.0005894898726271371,
+      "loss": 3.0184,
+      "step": 5173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5705693364143372,
+      "learning_rate": 0.0005894858326847965,
+      "loss": 3.439,
+      "step": 5174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5961979031562805,
+      "learning_rate": 0.0005894817919800059,
+      "loss": 3.5159,
+      "step": 5175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5484743714332581,
+      "learning_rate": 0.000589477750512776,
+      "loss": 3.2929,
+      "step": 5176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.628601610660553,
+      "learning_rate": 0.0005894737082831173,
+      "loss": 3.3603,
+      "step": 5177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5901480317115784,
+      "learning_rate": 0.0005894696652910408,
+      "loss": 3.3222,
+      "step": 5178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.555828869342804,
+      "learning_rate": 0.0005894656215365567,
+      "loss": 3.4859,
+      "step": 5179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5342446565628052,
+      "learning_rate": 0.000589461577019676,
+      "loss": 3.7213,
+      "step": 5180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5517928004264832,
+      "learning_rate": 0.000589457531740409,
+      "loss": 3.5133,
+      "step": 5181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5539923906326294,
+      "learning_rate": 0.0005894534856987668,
+      "loss": 3.4874,
+      "step": 5182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5563471913337708,
+      "learning_rate": 0.0005894494388947598,
+      "loss": 3.4176,
+      "step": 5183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5729748606681824,
+      "learning_rate": 0.0005894453913283986,
+      "loss": 3.4297,
+      "step": 5184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.577461302280426,
+      "learning_rate": 0.0005894413429996939,
+      "loss": 3.0731,
+      "step": 5185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5624281764030457,
+      "learning_rate": 0.0005894372939086565,
+      "loss": 3.6104,
+      "step": 5186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.593137800693512,
+      "learning_rate": 0.0005894332440552969,
+      "loss": 3.4637,
+      "step": 5187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5267358422279358,
+      "learning_rate": 0.0005894291934396259,
+      "loss": 3.2419,
+      "step": 5188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6192349195480347,
+      "learning_rate": 0.0005894251420616541,
+      "loss": 3.5089,
+      "step": 5189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5877721905708313,
+      "learning_rate": 0.0005894210899213921,
+      "loss": 3.4892,
+      "step": 5190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7060837149620056,
+      "learning_rate": 0.0005894170370188508,
+      "loss": 3.3611,
+      "step": 5191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5571918487548828,
+      "learning_rate": 0.0005894129833540405,
+      "loss": 3.5587,
+      "step": 5192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5204548239707947,
+      "learning_rate": 0.0005894089289269724,
+      "loss": 3.4699,
+      "step": 5193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5898545384407043,
+      "learning_rate": 0.0005894048737376567,
+      "loss": 3.499,
+      "step": 5194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5386577248573303,
+      "learning_rate": 0.0005894008177861043,
+      "loss": 3.3448,
+      "step": 5195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5543986558914185,
+      "learning_rate": 0.0005893967610723257,
+      "loss": 3.4441,
+      "step": 5196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6119479537010193,
+      "learning_rate": 0.0005893927035963319,
+      "loss": 3.3068,
+      "step": 5197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5174164772033691,
+      "learning_rate": 0.0005893886453581334,
+      "loss": 3.2137,
+      "step": 5198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5318208932876587,
+      "learning_rate": 0.0005893845863577409,
+      "loss": 3.4068,
+      "step": 5199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.558030366897583,
+      "learning_rate": 0.000589380526595165,
+      "loss": 3.1989,
+      "step": 5200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5202679634094238,
+      "learning_rate": 0.0005893764660704166,
+      "loss": 3.3643,
+      "step": 5201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5347263813018799,
+      "learning_rate": 0.0005893724047835062,
+      "loss": 3.4693,
+      "step": 5202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5433123111724854,
+      "learning_rate": 0.0005893683427344445,
+      "loss": 3.2793,
+      "step": 5203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5789240598678589,
+      "learning_rate": 0.0005893642799232425,
+      "loss": 3.1991,
+      "step": 5204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5516834259033203,
+      "learning_rate": 0.0005893602163499105,
+      "loss": 3.4335,
+      "step": 5205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5537156462669373,
+      "learning_rate": 0.0005893561520144594,
+      "loss": 3.3853,
+      "step": 5206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5437060594558716,
+      "learning_rate": 0.0005893520869168999,
+      "loss": 3.301,
+      "step": 5207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6070511341094971,
+      "learning_rate": 0.0005893480210572427,
+      "loss": 3.4726,
+      "step": 5208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5556560754776001,
+      "learning_rate": 0.0005893439544354984,
+      "loss": 3.2371,
+      "step": 5209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5555329322814941,
+      "learning_rate": 0.000589339887051678,
+      "loss": 3.3675,
+      "step": 5210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.551421582698822,
+      "learning_rate": 0.0005893358189057919,
+      "loss": 3.3801,
+      "step": 5211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5763034224510193,
+      "learning_rate": 0.0005893317499978511,
+      "loss": 3.5617,
+      "step": 5212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5456756353378296,
+      "learning_rate": 0.0005893276803278659,
+      "loss": 3.5274,
+      "step": 5213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5538662672042847,
+      "learning_rate": 0.0005893236098958474,
+      "loss": 3.3474,
+      "step": 5214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5430862307548523,
+      "learning_rate": 0.0005893195387018062,
+      "loss": 3.3443,
+      "step": 5215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5130735039710999,
+      "learning_rate": 0.000589315466745753,
+      "loss": 3.637,
+      "step": 5216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5832960605621338,
+      "learning_rate": 0.0005893113940276985,
+      "loss": 3.319,
+      "step": 5217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5417028665542603,
+      "learning_rate": 0.0005893073205476535,
+      "loss": 3.3201,
+      "step": 5218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5512108206748962,
+      "learning_rate": 0.0005893032463056288,
+      "loss": 3.2845,
+      "step": 5219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5375921726226807,
+      "learning_rate": 0.0005892991713016348,
+      "loss": 3.642,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5193976759910583,
+      "learning_rate": 0.0005892950955356827,
+      "loss": 3.5162,
+      "step": 5221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5780069231987,
+      "learning_rate": 0.0005892910190077828,
+      "loss": 3.5228,
+      "step": 5222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5876449346542358,
+      "learning_rate": 0.0005892869417179461,
+      "loss": 3.2904,
+      "step": 5223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5421982407569885,
+      "learning_rate": 0.0005892828636661833,
+      "loss": 3.5897,
+      "step": 5224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6510359048843384,
+      "learning_rate": 0.0005892787848525052,
+      "loss": 3.2834,
+      "step": 5225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5625903010368347,
+      "learning_rate": 0.0005892747052769222,
+      "loss": 3.2265,
+      "step": 5226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5502780079841614,
+      "learning_rate": 0.0005892706249394455,
+      "loss": 3.3097,
+      "step": 5227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5297096967697144,
+      "learning_rate": 0.0005892665438400856,
+      "loss": 3.6185,
+      "step": 5228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5778887271881104,
+      "learning_rate": 0.0005892624619788533,
+      "loss": 3.4486,
+      "step": 5229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5754767656326294,
+      "learning_rate": 0.0005892583793557594,
+      "loss": 3.3823,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5356746315956116,
+      "learning_rate": 0.0005892542959708145,
+      "loss": 3.686,
+      "step": 5231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5547672510147095,
+      "learning_rate": 0.0005892502118240295,
+      "loss": 3.4749,
+      "step": 5232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5451996326446533,
+      "learning_rate": 0.0005892461269154151,
+      "loss": 3.2667,
+      "step": 5233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5528019070625305,
+      "learning_rate": 0.0005892420412449821,
+      "loss": 3.3068,
+      "step": 5234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5734788179397583,
+      "learning_rate": 0.0005892379548127411,
+      "loss": 3.4308,
+      "step": 5235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6340581774711609,
+      "learning_rate": 0.000589233867618703,
+      "loss": 3.4311,
+      "step": 5236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6577187180519104,
+      "learning_rate": 0.0005892297796628787,
+      "loss": 3.4902,
+      "step": 5237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5538560152053833,
+      "learning_rate": 0.0005892256909452788,
+      "loss": 3.5402,
+      "step": 5238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5807637572288513,
+      "learning_rate": 0.0005892216014659141,
+      "loss": 3.4806,
+      "step": 5239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6450725793838501,
+      "learning_rate": 0.0005892175112247952,
+      "loss": 3.3013,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5808170437812805,
+      "learning_rate": 0.0005892134202219332,
+      "loss": 3.4958,
+      "step": 5241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5749102234840393,
+      "learning_rate": 0.0005892093284573386,
+      "loss": 3.4446,
+      "step": 5242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5990297794342041,
+      "learning_rate": 0.0005892052359310224,
+      "loss": 3.2321,
+      "step": 5243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5670678019523621,
+      "learning_rate": 0.0005892011426429952,
+      "loss": 3.5992,
+      "step": 5244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5572386384010315,
+      "learning_rate": 0.0005891970485932679,
+      "loss": 3.3791,
+      "step": 5245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5236799120903015,
+      "learning_rate": 0.0005891929537818512,
+      "loss": 3.5792,
+      "step": 5246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.595807671546936,
+      "learning_rate": 0.0005891888582087559,
+      "loss": 3.1541,
+      "step": 5247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5384485125541687,
+      "learning_rate": 0.0005891847618739929,
+      "loss": 3.0939,
+      "step": 5248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5316618084907532,
+      "learning_rate": 0.0005891806647775727,
+      "loss": 3.6524,
+      "step": 5249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.527287483215332,
+      "learning_rate": 0.0005891765669195064,
+      "loss": 3.4803,
+      "step": 5250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5352149605751038,
+      "learning_rate": 0.0005891724682998046,
+      "loss": 3.4944,
+      "step": 5251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5439637303352356,
+      "learning_rate": 0.0005891683689184784,
+      "loss": 3.3703,
+      "step": 5252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5407761335372925,
+      "learning_rate": 0.0005891642687755382,
+      "loss": 3.635,
+      "step": 5253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5867874622344971,
+      "learning_rate": 0.0005891601678709949,
+      "loss": 3.3837,
+      "step": 5254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5279407501220703,
+      "learning_rate": 0.0005891560662048595,
+      "loss": 3.3589,
+      "step": 5255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.549512505531311,
+      "learning_rate": 0.0005891519637771426,
+      "loss": 3.3413,
+      "step": 5256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.566560685634613,
+      "learning_rate": 0.0005891478605878551,
+      "loss": 3.4066,
+      "step": 5257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5679735541343689,
+      "learning_rate": 0.0005891437566370078,
+      "loss": 3.3625,
+      "step": 5258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5230506062507629,
+      "learning_rate": 0.0005891396519246114,
+      "loss": 3.3698,
+      "step": 5259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5605278611183167,
+      "learning_rate": 0.0005891355464506769,
+      "loss": 3.3979,
+      "step": 5260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5670188069343567,
+      "learning_rate": 0.000589131440215215,
+      "loss": 3.5514,
+      "step": 5261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5283556580543518,
+      "learning_rate": 0.0005891273332182365,
+      "loss": 3.416,
+      "step": 5262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5201714038848877,
+      "learning_rate": 0.0005891232254597523,
+      "loss": 3.3487,
+      "step": 5263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6400845646858215,
+      "learning_rate": 0.0005891191169397731,
+      "loss": 3.2951,
+      "step": 5264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5510321259498596,
+      "learning_rate": 0.0005891150076583098,
+      "loss": 3.4685,
+      "step": 5265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5668927431106567,
+      "learning_rate": 0.0005891108976153732,
+      "loss": 3.4804,
+      "step": 5266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5134303569793701,
+      "learning_rate": 0.0005891067868109743,
+      "loss": 3.3349,
+      "step": 5267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6687813997268677,
+      "learning_rate": 0.0005891026752451235,
+      "loss": 3.4913,
+      "step": 5268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5363355278968811,
+      "learning_rate": 0.0005890985629178321,
+      "loss": 3.3884,
+      "step": 5269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5483888387680054,
+      "learning_rate": 0.0005890944498291106,
+      "loss": 3.2948,
+      "step": 5270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5730322599411011,
+      "learning_rate": 0.00058909033597897,
+      "loss": 3.4693,
+      "step": 5271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6208261847496033,
+      "learning_rate": 0.0005890862213674211,
+      "loss": 3.403,
+      "step": 5272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5502722263336182,
+      "learning_rate": 0.0005890821059944748,
+      "loss": 3.442,
+      "step": 5273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5494133234024048,
+      "learning_rate": 0.0005890779898601417,
+      "loss": 3.4079,
+      "step": 5274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5719727873802185,
+      "learning_rate": 0.0005890738729644329,
+      "loss": 3.4966,
+      "step": 5275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5802255868911743,
+      "learning_rate": 0.0005890697553073591,
+      "loss": 3.4647,
+      "step": 5276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5484011769294739,
+      "learning_rate": 0.0005890656368889313,
+      "loss": 3.4309,
+      "step": 5277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5723143219947815,
+      "learning_rate": 0.0005890615177091601,
+      "loss": 3.6246,
+      "step": 5278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5343371033668518,
+      "learning_rate": 0.0005890573977680567,
+      "loss": 3.6963,
+      "step": 5279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5718325972557068,
+      "learning_rate": 0.0005890532770656316,
+      "loss": 3.487,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5279609560966492,
+      "learning_rate": 0.0005890491556018958,
+      "loss": 3.4661,
+      "step": 5281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5968024134635925,
+      "learning_rate": 0.0005890450333768601,
+      "loss": 3.533,
+      "step": 5282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5400145053863525,
+      "learning_rate": 0.0005890409103905356,
+      "loss": 3.2575,
+      "step": 5283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5489025712013245,
+      "learning_rate": 0.0005890367866429329,
+      "loss": 3.36,
+      "step": 5284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.600946843624115,
+      "learning_rate": 0.0005890326621340628,
+      "loss": 3.5745,
+      "step": 5285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5801662802696228,
+      "learning_rate": 0.0005890285368639363,
+      "loss": 3.4846,
+      "step": 5286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5930183529853821,
+      "learning_rate": 0.0005890244108325645,
+      "loss": 3.3913,
+      "step": 5287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5204430818557739,
+      "learning_rate": 0.0005890202840399579,
+      "loss": 3.3981,
+      "step": 5288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5407831072807312,
+      "learning_rate": 0.0005890161564861274,
+      "loss": 3.6175,
+      "step": 5289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5350310206413269,
+      "learning_rate": 0.000589012028171084,
+      "loss": 3.6356,
+      "step": 5290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.601344645023346,
+      "learning_rate": 0.0005890078990948386,
+      "loss": 3.2246,
+      "step": 5291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5307313799858093,
+      "learning_rate": 0.000589003769257402,
+      "loss": 3.6117,
+      "step": 5292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5290243029594421,
+      "learning_rate": 0.000588999638658785,
+      "loss": 3.4505,
+      "step": 5293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5569483637809753,
+      "learning_rate": 0.0005889955072989986,
+      "loss": 3.3312,
+      "step": 5294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5312089920043945,
+      "learning_rate": 0.0005889913751780537,
+      "loss": 3.5256,
+      "step": 5295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6026173830032349,
+      "learning_rate": 0.0005889872422959612,
+      "loss": 3.4126,
+      "step": 5296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6224252581596375,
+      "learning_rate": 0.0005889831086527318,
+      "loss": 3.483,
+      "step": 5297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5267722606658936,
+      "learning_rate": 0.0005889789742483766,
+      "loss": 3.5036,
+      "step": 5298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5711367130279541,
+      "learning_rate": 0.0005889748390829064,
+      "loss": 3.3469,
+      "step": 5299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5197421312332153,
+      "learning_rate": 0.000588970703156332,
+      "loss": 3.3095,
+      "step": 5300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.552913248538971,
+      "learning_rate": 0.0005889665664686644,
+      "loss": 3.483,
+      "step": 5301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5599504113197327,
+      "learning_rate": 0.0005889624290199145,
+      "loss": 3.2881,
+      "step": 5302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5224385261535645,
+      "learning_rate": 0.0005889582908100932,
+      "loss": 3.6117,
+      "step": 5303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5339375734329224,
+      "learning_rate": 0.0005889541518392114,
+      "loss": 3.4773,
+      "step": 5304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5676855444908142,
+      "learning_rate": 0.0005889500121072799,
+      "loss": 3.365,
+      "step": 5305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6814284920692444,
+      "learning_rate": 0.0005889458716143097,
+      "loss": 3.5605,
+      "step": 5306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52662593126297,
+      "learning_rate": 0.0005889417303603117,
+      "loss": 3.5284,
+      "step": 5307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5496254563331604,
+      "learning_rate": 0.0005889375883452967,
+      "loss": 3.4624,
+      "step": 5308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5711499452590942,
+      "learning_rate": 0.0005889334455692758,
+      "loss": 3.4269,
+      "step": 5309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5623414516448975,
+      "learning_rate": 0.0005889293020322597,
+      "loss": 3.4268,
+      "step": 5310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5185620188713074,
+      "learning_rate": 0.0005889251577342597,
+      "loss": 3.3941,
+      "step": 5311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5413085222244263,
+      "learning_rate": 0.0005889210126752861,
+      "loss": 3.5135,
+      "step": 5312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5592106580734253,
+      "learning_rate": 0.0005889168668553504,
+      "loss": 3.3922,
+      "step": 5313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5406152606010437,
+      "learning_rate": 0.0005889127202744631,
+      "loss": 3.4864,
+      "step": 5314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5414677262306213,
+      "learning_rate": 0.0005889085729326354,
+      "loss": 3.3056,
+      "step": 5315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5268755555152893,
+      "learning_rate": 0.0005889044248298781,
+      "loss": 3.2041,
+      "step": 5316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5885822772979736,
+      "learning_rate": 0.0005889002759662021,
+      "loss": 3.4368,
+      "step": 5317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5720695853233337,
+      "learning_rate": 0.0005888961263416185,
+      "loss": 3.4827,
+      "step": 5318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5145799517631531,
+      "learning_rate": 0.000588891975956138,
+      "loss": 3.4205,
+      "step": 5319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5636258721351624,
+      "learning_rate": 0.0005888878248097717,
+      "loss": 3.4612,
+      "step": 5320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5309170484542847,
+      "learning_rate": 0.0005888836729025304,
+      "loss": 3.3744,
+      "step": 5321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5823953747749329,
+      "learning_rate": 0.0005888795202344251,
+      "loss": 3.1887,
+      "step": 5322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5687710046768188,
+      "learning_rate": 0.0005888753668054667,
+      "loss": 3.4677,
+      "step": 5323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5707094669342041,
+      "learning_rate": 0.0005888712126156663,
+      "loss": 3.3657,
+      "step": 5324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5369641184806824,
+      "learning_rate": 0.0005888670576650346,
+      "loss": 3.3792,
+      "step": 5325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.550995409488678,
+      "learning_rate": 0.0005888629019535828,
+      "loss": 3.3031,
+      "step": 5326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5561549067497253,
+      "learning_rate": 0.0005888587454813216,
+      "loss": 3.3467,
+      "step": 5327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.564005434513092,
+      "learning_rate": 0.0005888545882482622,
+      "loss": 3.262,
+      "step": 5328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5233489871025085,
+      "learning_rate": 0.0005888504302544152,
+      "loss": 3.1761,
+      "step": 5329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.586366593837738,
+      "learning_rate": 0.0005888462714997919,
+      "loss": 3.2845,
+      "step": 5330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6343865990638733,
+      "learning_rate": 0.000588842111984403,
+      "loss": 3.5545,
+      "step": 5331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.66603022813797,
+      "learning_rate": 0.0005888379517082597,
+      "loss": 3.4753,
+      "step": 5332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5965749621391296,
+      "learning_rate": 0.0005888337906713728,
+      "loss": 3.3929,
+      "step": 5333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5687329173088074,
+      "learning_rate": 0.0005888296288737531,
+      "loss": 3.4986,
+      "step": 5334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5994446277618408,
+      "learning_rate": 0.0005888254663154119,
+      "loss": 3.5974,
+      "step": 5335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5664075613021851,
+      "learning_rate": 0.00058882130299636,
+      "loss": 3.451,
+      "step": 5336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5647047758102417,
+      "learning_rate": 0.0005888171389166083,
+      "loss": 3.4053,
+      "step": 5337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5460997819900513,
+      "learning_rate": 0.0005888129740761679,
+      "loss": 3.2895,
+      "step": 5338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5719659328460693,
+      "learning_rate": 0.0005888088084750497,
+      "loss": 3.2872,
+      "step": 5339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5610400438308716,
+      "learning_rate": 0.0005888046421132647,
+      "loss": 3.3737,
+      "step": 5340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5502822995185852,
+      "learning_rate": 0.0005888004749908239,
+      "loss": 3.3927,
+      "step": 5341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6235296726226807,
+      "learning_rate": 0.0005887963071077382,
+      "loss": 3.4243,
+      "step": 5342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5775569081306458,
+      "learning_rate": 0.0005887921384640186,
+      "loss": 3.4377,
+      "step": 5343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5198445320129395,
+      "learning_rate": 0.000588787969059676,
+      "loss": 3.308,
+      "step": 5344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5459919571876526,
+      "learning_rate": 0.0005887837988947216,
+      "loss": 3.5653,
+      "step": 5345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5754595994949341,
+      "learning_rate": 0.0005887796279691661,
+      "loss": 3.226,
+      "step": 5346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5614638328552246,
+      "learning_rate": 0.0005887754562830207,
+      "loss": 3.4762,
+      "step": 5347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5696617960929871,
+      "learning_rate": 0.0005887712838362963,
+      "loss": 3.2895,
+      "step": 5348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.529895544052124,
+      "learning_rate": 0.000588767110629004,
+      "loss": 3.3031,
+      "step": 5349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5466099977493286,
+      "learning_rate": 0.0005887629366611547,
+      "loss": 3.6197,
+      "step": 5350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6054186820983887,
+      "learning_rate": 0.0005887587619327594,
+      "loss": 3.3346,
+      "step": 5351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.551434338092804,
+      "learning_rate": 0.0005887545864438291,
+      "loss": 3.6942,
+      "step": 5352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5595240592956543,
+      "learning_rate": 0.0005887504101943746,
+      "loss": 3.2481,
+      "step": 5353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5537325739860535,
+      "learning_rate": 0.0005887462331844072,
+      "loss": 3.3643,
+      "step": 5354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5237861275672913,
+      "learning_rate": 0.0005887420554139379,
+      "loss": 3.0849,
+      "step": 5355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5428269505500793,
+      "learning_rate": 0.0005887378768829776,
+      "loss": 3.6186,
+      "step": 5356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5731378793716431,
+      "learning_rate": 0.0005887336975915372,
+      "loss": 3.4173,
+      "step": 5357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5460600852966309,
+      "learning_rate": 0.0005887295175396279,
+      "loss": 3.294,
+      "step": 5358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5621720552444458,
+      "learning_rate": 0.0005887253367272605,
+      "loss": 3.19,
+      "step": 5359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5691437125205994,
+      "learning_rate": 0.0005887211551544462,
+      "loss": 3.346,
+      "step": 5360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6207178235054016,
+      "learning_rate": 0.0005887169728211959,
+      "loss": 3.4977,
+      "step": 5361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5653191804885864,
+      "learning_rate": 0.0005887127897275206,
+      "loss": 3.2567,
+      "step": 5362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6655686497688293,
+      "learning_rate": 0.0005887086058734316,
+      "loss": 3.4619,
+      "step": 5363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6186230182647705,
+      "learning_rate": 0.0005887044212589395,
+      "loss": 3.4998,
+      "step": 5364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5586534142494202,
+      "learning_rate": 0.0005887002358840556,
+      "loss": 3.4456,
+      "step": 5365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.561849057674408,
+      "learning_rate": 0.0005886960497487908,
+      "loss": 3.335,
+      "step": 5366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5541581511497498,
+      "learning_rate": 0.0005886918628531562,
+      "loss": 3.2474,
+      "step": 5367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5545457005500793,
+      "learning_rate": 0.0005886876751971628,
+      "loss": 3.3815,
+      "step": 5368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5654429197311401,
+      "learning_rate": 0.0005886834867808215,
+      "loss": 3.2011,
+      "step": 5369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.635900616645813,
+      "learning_rate": 0.0005886792976041436,
+      "loss": 3.4373,
+      "step": 5370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5833329558372498,
+      "learning_rate": 0.0005886751076671399,
+      "loss": 3.2561,
+      "step": 5371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.565444827079773,
+      "learning_rate": 0.0005886709169698216,
+      "loss": 3.3935,
+      "step": 5372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5286258459091187,
+      "learning_rate": 0.0005886667255121995,
+      "loss": 3.5707,
+      "step": 5373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5440570712089539,
+      "learning_rate": 0.000588662533294285,
+      "loss": 3.311,
+      "step": 5374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6369450092315674,
+      "learning_rate": 0.0005886583403160889,
+      "loss": 3.6501,
+      "step": 5375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5417674779891968,
+      "learning_rate": 0.0005886541465776223,
+      "loss": 3.288,
+      "step": 5376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6063926815986633,
+      "learning_rate": 0.0005886499520788961,
+      "loss": 3.2348,
+      "step": 5377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5412531495094299,
+      "learning_rate": 0.0005886457568199216,
+      "loss": 3.6838,
+      "step": 5378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5758473873138428,
+      "learning_rate": 0.0005886415608007096,
+      "loss": 3.5398,
+      "step": 5379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.577717661857605,
+      "learning_rate": 0.0005886373640212714,
+      "loss": 3.4554,
+      "step": 5380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5586448907852173,
+      "learning_rate": 0.000588633166481618,
+      "loss": 3.5677,
+      "step": 5381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.574151873588562,
+      "learning_rate": 0.0005886289681817601,
+      "loss": 3.3092,
+      "step": 5382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5311241149902344,
+      "learning_rate": 0.0005886247691217093,
+      "loss": 3.5122,
+      "step": 5383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5521003007888794,
+      "learning_rate": 0.0005886205693014764,
+      "loss": 3.4751,
+      "step": 5384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5857923030853271,
+      "learning_rate": 0.0005886163687210724,
+      "loss": 3.531,
+      "step": 5385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5671178102493286,
+      "learning_rate": 0.0005886121673805083,
+      "loss": 3.3799,
+      "step": 5386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.523662805557251,
+      "learning_rate": 0.0005886079652797954,
+      "loss": 3.6136,
+      "step": 5387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5966594219207764,
+      "learning_rate": 0.0005886037624189448,
+      "loss": 3.1987,
+      "step": 5388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5404354929924011,
+      "learning_rate": 0.0005885995587979672,
+      "loss": 3.469,
+      "step": 5389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.565900981426239,
+      "learning_rate": 0.0005885953544168741,
+      "loss": 3.4291,
+      "step": 5390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5991320610046387,
+      "learning_rate": 0.0005885911492756763,
+      "loss": 3.4353,
+      "step": 5391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5885935425758362,
+      "learning_rate": 0.0005885869433743849,
+      "loss": 3.3665,
+      "step": 5392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5500056743621826,
+      "learning_rate": 0.0005885827367130112,
+      "loss": 3.3849,
+      "step": 5393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6144444942474365,
+      "learning_rate": 0.000588578529291566,
+      "loss": 3.4082,
+      "step": 5394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5883260369300842,
+      "learning_rate": 0.0005885743211100603,
+      "loss": 3.5509,
+      "step": 5395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5883315205574036,
+      "learning_rate": 0.0005885701121685057,
+      "loss": 3.4782,
+      "step": 5396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5557137131690979,
+      "learning_rate": 0.0005885659024669128,
+      "loss": 3.3478,
+      "step": 5397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5621869564056396,
+      "learning_rate": 0.0005885616920052928,
+      "loss": 3.2406,
+      "step": 5398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6264198422431946,
+      "learning_rate": 0.000588557480783657,
+      "loss": 3.2804,
+      "step": 5399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.543046236038208,
+      "learning_rate": 0.0005885532688020163,
+      "loss": 3.3372,
+      "step": 5400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5603328943252563,
+      "learning_rate": 0.0005885490560603818,
+      "loss": 3.2055,
+      "step": 5401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5257130861282349,
+      "learning_rate": 0.0005885448425587645,
+      "loss": 3.3474,
+      "step": 5402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5107335448265076,
+      "learning_rate": 0.0005885406282971756,
+      "loss": 3.5721,
+      "step": 5403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5753149390220642,
+      "learning_rate": 0.0005885364132756264,
+      "loss": 3.4103,
+      "step": 5404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5694027543067932,
+      "learning_rate": 0.0005885321974941277,
+      "loss": 3.4066,
+      "step": 5405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5609138607978821,
+      "learning_rate": 0.0005885279809526908,
+      "loss": 3.3846,
+      "step": 5406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5206153988838196,
+      "learning_rate": 0.0005885237636513267,
+      "loss": 3.5455,
+      "step": 5407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6045171022415161,
+      "learning_rate": 0.0005885195455900465,
+      "loss": 3.2741,
+      "step": 5408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.565991997718811,
+      "learning_rate": 0.0005885153267688613,
+      "loss": 3.5252,
+      "step": 5409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5464850664138794,
+      "learning_rate": 0.0005885111071877823,
+      "loss": 3.2544,
+      "step": 5410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.567910373210907,
+      "learning_rate": 0.0005885068868468206,
+      "loss": 3.5726,
+      "step": 5411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5260792374610901,
+      "learning_rate": 0.0005885026657459873,
+      "loss": 3.3725,
+      "step": 5412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5899006128311157,
+      "learning_rate": 0.0005884984438852934,
+      "loss": 3.2411,
+      "step": 5413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6080503463745117,
+      "learning_rate": 0.0005884942212647502,
+      "loss": 3.32,
+      "step": 5414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5826466083526611,
+      "learning_rate": 0.0005884899978843688,
+      "loss": 3.2586,
+      "step": 5415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5410050749778748,
+      "learning_rate": 0.0005884857737441601,
+      "loss": 3.3174,
+      "step": 5416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5297449827194214,
+      "learning_rate": 0.0005884815488441356,
+      "loss": 3.3025,
+      "step": 5417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5570391416549683,
+      "learning_rate": 0.0005884773231843062,
+      "loss": 3.5041,
+      "step": 5418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5632867217063904,
+      "learning_rate": 0.0005884730967646828,
+      "loss": 3.4786,
+      "step": 5419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5535479784011841,
+      "learning_rate": 0.0005884688695852769,
+      "loss": 3.2442,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5231944918632507,
+      "learning_rate": 0.0005884646416460997,
+      "loss": 3.5603,
+      "step": 5421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5422855019569397,
+      "learning_rate": 0.000588460412947162,
+      "loss": 3.2527,
+      "step": 5422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5713430643081665,
+      "learning_rate": 0.000588456183488475,
+      "loss": 3.5105,
+      "step": 5423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.568607747554779,
+      "learning_rate": 0.0005884519532700501,
+      "loss": 3.363,
+      "step": 5424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5918583273887634,
+      "learning_rate": 0.0005884477222918981,
+      "loss": 3.3211,
+      "step": 5425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6431357264518738,
+      "learning_rate": 0.0005884434905540305,
+      "loss": 3.2004,
+      "step": 5426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5770518183708191,
+      "learning_rate": 0.0005884392580564581,
+      "loss": 3.4413,
+      "step": 5427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6105692982673645,
+      "learning_rate": 0.0005884350247991923,
+      "loss": 3.4229,
+      "step": 5428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5172119736671448,
+      "learning_rate": 0.0005884307907822441,
+      "loss": 3.3484,
+      "step": 5429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6026398539543152,
+      "learning_rate": 0.0005884265560056246,
+      "loss": 3.2793,
+      "step": 5430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5257851481437683,
+      "learning_rate": 0.0005884223204693453,
+      "loss": 3.324,
+      "step": 5431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5769349932670593,
+      "learning_rate": 0.000588418084173417,
+      "loss": 3.3327,
+      "step": 5432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5765005350112915,
+      "learning_rate": 0.000588413847117851,
+      "loss": 3.557,
+      "step": 5433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5662025213241577,
+      "learning_rate": 0.0005884096093026584,
+      "loss": 3.4275,
+      "step": 5434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5836684107780457,
+      "learning_rate": 0.0005884053707278504,
+      "loss": 3.4793,
+      "step": 5435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5512568354606628,
+      "learning_rate": 0.0005884011313934381,
+      "loss": 3.2952,
+      "step": 5436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5661933422088623,
+      "learning_rate": 0.0005883968912994328,
+      "loss": 3.4282,
+      "step": 5437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5790491700172424,
+      "learning_rate": 0.0005883926504458456,
+      "loss": 3.3611,
+      "step": 5438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5423797369003296,
+      "learning_rate": 0.0005883884088326877,
+      "loss": 3.3749,
+      "step": 5439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5996915698051453,
+      "learning_rate": 0.0005883841664599701,
+      "loss": 3.1857,
+      "step": 5440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5513800382614136,
+      "learning_rate": 0.0005883799233277042,
+      "loss": 3.2941,
+      "step": 5441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5614347457885742,
+      "learning_rate": 0.000588375679435901,
+      "loss": 3.4229,
+      "step": 5442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5885158777236938,
+      "learning_rate": 0.0005883714347845718,
+      "loss": 3.4624,
+      "step": 5443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5393982529640198,
+      "learning_rate": 0.0005883671893737279,
+      "loss": 3.2643,
+      "step": 5444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5851886868476868,
+      "learning_rate": 0.0005883629432033801,
+      "loss": 3.4327,
+      "step": 5445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5609333515167236,
+      "learning_rate": 0.0005883586962735399,
+      "loss": 3.2944,
+      "step": 5446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5521209239959717,
+      "learning_rate": 0.0005883544485842183,
+      "loss": 3.3134,
+      "step": 5447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5714461803436279,
+      "learning_rate": 0.0005883502001354267,
+      "loss": 3.5679,
+      "step": 5448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5831002593040466,
+      "learning_rate": 0.0005883459509271762,
+      "loss": 3.2632,
+      "step": 5449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5336947441101074,
+      "learning_rate": 0.0005883417009594778,
+      "loss": 3.3042,
+      "step": 5450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.575208306312561,
+      "learning_rate": 0.0005883374502323429,
+      "loss": 3.3451,
+      "step": 5451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5348706841468811,
+      "learning_rate": 0.0005883331987457827,
+      "loss": 3.4311,
+      "step": 5452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5652807950973511,
+      "learning_rate": 0.0005883289464998083,
+      "loss": 3.4136,
+      "step": 5453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6097990870475769,
+      "learning_rate": 0.000588324693494431,
+      "loss": 3.3811,
+      "step": 5454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5866896510124207,
+      "learning_rate": 0.0005883204397296619,
+      "loss": 3.2683,
+      "step": 5455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5691618323326111,
+      "learning_rate": 0.0005883161852055122,
+      "loss": 3.3863,
+      "step": 5456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6082558035850525,
+      "learning_rate": 0.0005883119299219932,
+      "loss": 3.3178,
+      "step": 5457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5235224962234497,
+      "learning_rate": 0.0005883076738791161,
+      "loss": 3.5059,
+      "step": 5458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.537198007106781,
+      "learning_rate": 0.0005883034170768921,
+      "loss": 3.3683,
+      "step": 5459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5196539163589478,
+      "learning_rate": 0.0005882991595153323,
+      "loss": 3.3288,
+      "step": 5460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5324885249137878,
+      "learning_rate": 0.0005882949011944481,
+      "loss": 3.3257,
+      "step": 5461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5442826747894287,
+      "learning_rate": 0.0005882906421142505,
+      "loss": 3.3675,
+      "step": 5462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5731666684150696,
+      "learning_rate": 0.0005882863822747508,
+      "loss": 3.4277,
+      "step": 5463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.598007321357727,
+      "learning_rate": 0.0005882821216759604,
+      "loss": 3.4906,
+      "step": 5464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5194315910339355,
+      "learning_rate": 0.0005882778603178903,
+      "loss": 3.2031,
+      "step": 5465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5207369327545166,
+      "learning_rate": 0.0005882735982005519,
+      "loss": 3.2482,
+      "step": 5466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5437124371528625,
+      "learning_rate": 0.0005882693353239562,
+      "loss": 3.5863,
+      "step": 5467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5464715957641602,
+      "learning_rate": 0.0005882650716881146,
+      "loss": 3.4643,
+      "step": 5468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5588339567184448,
+      "learning_rate": 0.0005882608072930381,
+      "loss": 3.3967,
+      "step": 5469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5152556896209717,
+      "learning_rate": 0.0005882565421387383,
+      "loss": 3.324,
+      "step": 5470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5201581120491028,
+      "learning_rate": 0.0005882522762252262,
+      "loss": 3.1271,
+      "step": 5471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5199649930000305,
+      "learning_rate": 0.0005882480095525132,
+      "loss": 3.491,
+      "step": 5472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5226982235908508,
+      "learning_rate": 0.0005882437421206102,
+      "loss": 3.4153,
+      "step": 5473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5411034226417542,
+      "learning_rate": 0.0005882394739295287,
+      "loss": 3.4042,
+      "step": 5474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.554895281791687,
+      "learning_rate": 0.0005882352049792801,
+      "loss": 3.384,
+      "step": 5475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.538006603717804,
+      "learning_rate": 0.0005882309352698752,
+      "loss": 3.4293,
+      "step": 5476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5578272342681885,
+      "learning_rate": 0.0005882266648013256,
+      "loss": 3.4783,
+      "step": 5477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5473056435585022,
+      "learning_rate": 0.0005882223935736424,
+      "loss": 3.5646,
+      "step": 5478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5610681176185608,
+      "learning_rate": 0.0005882181215868369,
+      "loss": 3.5254,
+      "step": 5479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5292629599571228,
+      "learning_rate": 0.0005882138488409204,
+      "loss": 3.376,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5061684846878052,
+      "learning_rate": 0.000588209575335904,
+      "loss": 3.5303,
+      "step": 5481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5533640384674072,
+      "learning_rate": 0.0005882053010717991,
+      "loss": 3.4842,
+      "step": 5482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5831032991409302,
+      "learning_rate": 0.0005882010260486169,
+      "loss": 3.5035,
+      "step": 5483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5656222701072693,
+      "learning_rate": 0.0005881967502663687,
+      "loss": 3.5032,
+      "step": 5484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6056988835334778,
+      "learning_rate": 0.0005881924737250655,
+      "loss": 3.5061,
+      "step": 5485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5542979836463928,
+      "learning_rate": 0.0005881881964247191,
+      "loss": 3.4491,
+      "step": 5486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5534904599189758,
+      "learning_rate": 0.0005881839183653402,
+      "loss": 3.4249,
+      "step": 5487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5802499651908875,
+      "learning_rate": 0.0005881796395469406,
+      "loss": 3.242,
+      "step": 5488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6037202477455139,
+      "learning_rate": 0.000588175359969531,
+      "loss": 3.4127,
+      "step": 5489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5480342507362366,
+      "learning_rate": 0.0005881710796331231,
+      "loss": 3.2675,
+      "step": 5490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5399625301361084,
+      "learning_rate": 0.000588166798537728,
+      "loss": 3.4004,
+      "step": 5491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5208690166473389,
+      "learning_rate": 0.000588162516683357,
+      "loss": 3.2553,
+      "step": 5492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5354910492897034,
+      "learning_rate": 0.0005881582340700215,
+      "loss": 3.2685,
+      "step": 5493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6398562788963318,
+      "learning_rate": 0.0005881539506977326,
+      "loss": 3.7163,
+      "step": 5494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6458652019500732,
+      "learning_rate": 0.0005881496665665016,
+      "loss": 3.4954,
+      "step": 5495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5615726113319397,
+      "learning_rate": 0.0005881453816763398,
+      "loss": 3.3978,
+      "step": 5496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5364488959312439,
+      "learning_rate": 0.0005881410960272587,
+      "loss": 3.3949,
+      "step": 5497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5817398428916931,
+      "learning_rate": 0.0005881368096192693,
+      "loss": 3.3621,
+      "step": 5498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8163599371910095,
+      "learning_rate": 0.000588132522452383,
+      "loss": 3.6545,
+      "step": 5499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5170142650604248,
+      "learning_rate": 0.0005881282345266111,
+      "loss": 3.4511,
+      "step": 5500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5624502897262573,
+      "learning_rate": 0.0005881239458419648,
+      "loss": 3.3277,
+      "step": 5501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5259974002838135,
+      "learning_rate": 0.0005881196563984555,
+      "loss": 3.3884,
+      "step": 5502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5127625465393066,
+      "learning_rate": 0.0005881153661960946,
+      "loss": 3.4118,
+      "step": 5503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5912742018699646,
+      "learning_rate": 0.0005881110752348931,
+      "loss": 3.5946,
+      "step": 5504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5364376902580261,
+      "learning_rate": 0.0005881067835148626,
+      "loss": 3.4842,
+      "step": 5505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5497884154319763,
+      "learning_rate": 0.0005881024910360143,
+      "loss": 3.3788,
+      "step": 5506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5624418258666992,
+      "learning_rate": 0.0005880981977983594,
+      "loss": 3.1812,
+      "step": 5507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.548305332660675,
+      "learning_rate": 0.0005880939038019093,
+      "loss": 3.2861,
+      "step": 5508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6290483474731445,
+      "learning_rate": 0.0005880896090466754,
+      "loss": 3.2389,
+      "step": 5509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6057081818580627,
+      "learning_rate": 0.0005880853135326688,
+      "loss": 3.102,
+      "step": 5510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.567807674407959,
+      "learning_rate": 0.0005880810172599009,
+      "loss": 3.3086,
+      "step": 5511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5722109079360962,
+      "learning_rate": 0.0005880767202283832,
+      "loss": 3.2491,
+      "step": 5512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5263040661811829,
+      "learning_rate": 0.0005880724224381267,
+      "loss": 3.4022,
+      "step": 5513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5402182340621948,
+      "learning_rate": 0.0005880681238891429,
+      "loss": 3.6702,
+      "step": 5514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5795789361000061,
+      "learning_rate": 0.0005880638245814433,
+      "loss": 3.3956,
+      "step": 5515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6076276302337646,
+      "learning_rate": 0.0005880595245150388,
+      "loss": 3.3292,
+      "step": 5516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5449411869049072,
+      "learning_rate": 0.0005880552236899411,
+      "loss": 3.4199,
+      "step": 5517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5408446192741394,
+      "learning_rate": 0.0005880509221061613,
+      "loss": 3.5047,
+      "step": 5518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.548406183719635,
+      "learning_rate": 0.0005880466197637108,
+      "loss": 3.639,
+      "step": 5519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6209466457366943,
+      "learning_rate": 0.000588042316662601,
+      "loss": 3.4597,
+      "step": 5520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5157639980316162,
+      "learning_rate": 0.0005880380128028431,
+      "loss": 3.3713,
+      "step": 5521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5012747645378113,
+      "learning_rate": 0.0005880337081844485,
+      "loss": 3.3632,
+      "step": 5522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5477715134620667,
+      "learning_rate": 0.0005880294028074286,
+      "loss": 3.4232,
+      "step": 5523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6210871338844299,
+      "learning_rate": 0.0005880250966717946,
+      "loss": 3.2222,
+      "step": 5524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0005880207897775581,
+      "loss": 3.5465,
+      "step": 5525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5438306331634521,
+      "learning_rate": 0.0005880164821247301,
+      "loss": 3.2486,
+      "step": 5526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5276567935943604,
+      "learning_rate": 0.0005880121737133221,
+      "loss": 3.3924,
+      "step": 5527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5347236394882202,
+      "learning_rate": 0.0005880078645433456,
+      "loss": 3.1702,
+      "step": 5528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.532122015953064,
+      "learning_rate": 0.0005880035546148118,
+      "loss": 3.6436,
+      "step": 5529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6177424788475037,
+      "learning_rate": 0.000587999243927732,
+      "loss": 3.4205,
+      "step": 5530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5172913670539856,
+      "learning_rate": 0.0005879949324821177,
+      "loss": 3.4277,
+      "step": 5531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5155248641967773,
+      "learning_rate": 0.0005879906202779801,
+      "loss": 3.3915,
+      "step": 5532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5313438177108765,
+      "learning_rate": 0.0005879863073153306,
+      "loss": 3.1238,
+      "step": 5533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5252282619476318,
+      "learning_rate": 0.0005879819935941806,
+      "loss": 3.5629,
+      "step": 5534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5261170268058777,
+      "learning_rate": 0.0005879776791145416,
+      "loss": 3.6691,
+      "step": 5535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5184322595596313,
+      "learning_rate": 0.0005879733638764246,
+      "loss": 3.5832,
+      "step": 5536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5469010472297668,
+      "learning_rate": 0.0005879690478798413,
+      "loss": 3.4247,
+      "step": 5537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5656020045280457,
+      "learning_rate": 0.0005879647311248029,
+      "loss": 3.3261,
+      "step": 5538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.526353657245636,
+      "learning_rate": 0.0005879604136113209,
+      "loss": 3.4339,
+      "step": 5539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5460639595985413,
+      "learning_rate": 0.0005879560953394066,
+      "loss": 3.2131,
+      "step": 5540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5069341659545898,
+      "learning_rate": 0.0005879517763090712,
+      "loss": 3.4322,
+      "step": 5541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5094472169876099,
+      "learning_rate": 0.0005879474565203263,
+      "loss": 3.36,
+      "step": 5542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.520026445388794,
+      "learning_rate": 0.0005879431359731833,
+      "loss": 3.4691,
+      "step": 5543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5083997249603271,
+      "learning_rate": 0.0005879388146676535,
+      "loss": 3.3143,
+      "step": 5544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5479220747947693,
+      "learning_rate": 0.0005879344926037482,
+      "loss": 3.5481,
+      "step": 5545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.561329185962677,
+      "learning_rate": 0.0005879301697814789,
+      "loss": 3.4544,
+      "step": 5546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5535502433776855,
+      "learning_rate": 0.000587925846200857,
+      "loss": 3.5186,
+      "step": 5547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5314511656761169,
+      "learning_rate": 0.0005879215218618937,
+      "loss": 3.5539,
+      "step": 5548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4937261641025543,
+      "learning_rate": 0.0005879171967646006,
+      "loss": 3.5703,
+      "step": 5549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5664466619491577,
+      "learning_rate": 0.0005879128709089889,
+      "loss": 3.454,
+      "step": 5550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.513667106628418,
+      "learning_rate": 0.0005879085442950703,
+      "loss": 3.2555,
+      "step": 5551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5058261156082153,
+      "learning_rate": 0.000587904216922856,
+      "loss": 3.5081,
+      "step": 5552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5401785373687744,
+      "learning_rate": 0.0005878998887923572,
+      "loss": 3.4128,
+      "step": 5553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5664740800857544,
+      "learning_rate": 0.0005878955599035857,
+      "loss": 3.5442,
+      "step": 5554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5653204321861267,
+      "learning_rate": 0.0005878912302565526,
+      "loss": 3.4864,
+      "step": 5555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5319250822067261,
+      "learning_rate": 0.0005878868998512694,
+      "loss": 3.5001,
+      "step": 5556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5475034713745117,
+      "learning_rate": 0.0005878825686877476,
+      "loss": 3.4573,
+      "step": 5557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5714825391769409,
+      "learning_rate": 0.0005878782367659984,
+      "loss": 3.3223,
+      "step": 5558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5734561681747437,
+      "learning_rate": 0.0005878739040860334,
+      "loss": 3.3125,
+      "step": 5559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5484195947647095,
+      "learning_rate": 0.0005878695706478639,
+      "loss": 3.4178,
+      "step": 5560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5906480550765991,
+      "learning_rate": 0.0005878652364515014,
+      "loss": 3.4166,
+      "step": 5561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5374751091003418,
+      "learning_rate": 0.0005878609014969572,
+      "loss": 3.4265,
+      "step": 5562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5320459008216858,
+      "learning_rate": 0.0005878565657842428,
+      "loss": 3.3379,
+      "step": 5563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5613856911659241,
+      "learning_rate": 0.0005878522293133697,
+      "loss": 3.3711,
+      "step": 5564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5332615971565247,
+      "learning_rate": 0.0005878478920843492,
+      "loss": 3.4358,
+      "step": 5565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5565606951713562,
+      "learning_rate": 0.0005878435540971926,
+      "loss": 3.3137,
+      "step": 5566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5280033946037292,
+      "learning_rate": 0.0005878392153519117,
+      "loss": 3.4746,
+      "step": 5567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5665175914764404,
+      "learning_rate": 0.0005878348758485176,
+      "loss": 3.4319,
+      "step": 5568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5359581708908081,
+      "learning_rate": 0.0005878305355870218,
+      "loss": 3.4903,
+      "step": 5569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5408956408500671,
+      "learning_rate": 0.0005878261945674358,
+      "loss": 3.4426,
+      "step": 5570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5556782484054565,
+      "learning_rate": 0.0005878218527897709,
+      "loss": 3.4188,
+      "step": 5571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.549187183380127,
+      "learning_rate": 0.0005878175102540387,
+      "loss": 3.4785,
+      "step": 5572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5007501244544983,
+      "learning_rate": 0.0005878131669602506,
+      "loss": 3.254,
+      "step": 5573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5330648422241211,
+      "learning_rate": 0.000587808822908418,
+      "loss": 3.5062,
+      "step": 5574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5661302208900452,
+      "learning_rate": 0.0005878044780985523,
+      "loss": 3.555,
+      "step": 5575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5426991581916809,
+      "learning_rate": 0.000587800132530665,
+      "loss": 3.7042,
+      "step": 5576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5796633958816528,
+      "learning_rate": 0.0005877957862047676,
+      "loss": 3.3673,
+      "step": 5577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5920323133468628,
+      "learning_rate": 0.0005877914391208716,
+      "loss": 3.3901,
+      "step": 5578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5278183817863464,
+      "learning_rate": 0.0005877870912789882,
+      "loss": 3.4738,
+      "step": 5579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.506721556186676,
+      "learning_rate": 0.0005877827426791289,
+      "loss": 3.2758,
+      "step": 5580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5367504954338074,
+      "learning_rate": 0.0005877783933213054,
+      "loss": 3.3765,
+      "step": 5581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5439772009849548,
+      "learning_rate": 0.0005877740432055288,
+      "loss": 3.6383,
+      "step": 5582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5529850721359253,
+      "learning_rate": 0.0005877696923318109,
+      "loss": 3.3865,
+      "step": 5583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5937544703483582,
+      "learning_rate": 0.000587765340700163,
+      "loss": 3.6044,
+      "step": 5584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5193443894386292,
+      "learning_rate": 0.0005877609883105966,
+      "loss": 3.2867,
+      "step": 5585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5450002551078796,
+      "learning_rate": 0.0005877566351631231,
+      "loss": 3.4058,
+      "step": 5586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5292603373527527,
+      "learning_rate": 0.0005877522812577539,
+      "loss": 3.5574,
+      "step": 5587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5539287328720093,
+      "learning_rate": 0.0005877479265945008,
+      "loss": 3.3719,
+      "step": 5588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5461103320121765,
+      "learning_rate": 0.0005877435711733748,
+      "loss": 3.3806,
+      "step": 5589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5659968256950378,
+      "learning_rate": 0.0005877392149943877,
+      "loss": 3.3854,
+      "step": 5590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6119036674499512,
+      "learning_rate": 0.0005877348580575509,
+      "loss": 3.2403,
+      "step": 5591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5469770431518555,
+      "learning_rate": 0.000587730500362876,
+      "loss": 3.428,
+      "step": 5592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5068824887275696,
+      "learning_rate": 0.0005877261419103741,
+      "loss": 3.5666,
+      "step": 5593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5368444919586182,
+      "learning_rate": 0.000587721782700057,
+      "loss": 3.2824,
+      "step": 5594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5606796145439148,
+      "learning_rate": 0.0005877174227319362,
+      "loss": 3.2626,
+      "step": 5595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5552815794944763,
+      "learning_rate": 0.0005877130620060229,
+      "loss": 3.2393,
+      "step": 5596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.591176450252533,
+      "learning_rate": 0.000587708700522329,
+      "loss": 3.339,
+      "step": 5597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5659617781639099,
+      "learning_rate": 0.0005877043382808655,
+      "loss": 3.582,
+      "step": 5598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5646779537200928,
+      "learning_rate": 0.0005876999752816443,
+      "loss": 3.5119,
+      "step": 5599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5375205278396606,
+      "learning_rate": 0.0005876956115246767,
+      "loss": 3.39,
+      "step": 5600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5745865106582642,
+      "learning_rate": 0.0005876912470099742,
+      "loss": 3.515,
+      "step": 5601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6667315363883972,
+      "learning_rate": 0.0005876868817375483,
+      "loss": 3.2686,
+      "step": 5602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5387046933174133,
+      "learning_rate": 0.0005876825157074107,
+      "loss": 3.3706,
+      "step": 5603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5582010746002197,
+      "learning_rate": 0.0005876781489195725,
+      "loss": 3.5323,
+      "step": 5604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5929326415061951,
+      "learning_rate": 0.0005876737813740456,
+      "loss": 3.3948,
+      "step": 5605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5750752091407776,
+      "learning_rate": 0.0005876694130708412,
+      "loss": 3.2972,
+      "step": 5606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5479739904403687,
+      "learning_rate": 0.0005876650440099709,
+      "loss": 3.5172,
+      "step": 5607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5554889440536499,
+      "learning_rate": 0.0005876606741914462,
+      "loss": 3.399,
+      "step": 5608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5355261564254761,
+      "learning_rate": 0.0005876563036152788,
+      "loss": 3.5163,
+      "step": 5609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5420413613319397,
+      "learning_rate": 0.00058765193228148,
+      "loss": 3.3165,
+      "step": 5610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5698093771934509,
+      "learning_rate": 0.0005876475601900614,
+      "loss": 3.0865,
+      "step": 5611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5109629034996033,
+      "learning_rate": 0.0005876431873410344,
+      "loss": 3.4757,
+      "step": 5612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5421175360679626,
+      "learning_rate": 0.0005876388137344107,
+      "loss": 3.422,
+      "step": 5613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5593693256378174,
+      "learning_rate": 0.0005876344393702016,
+      "loss": 3.1981,
+      "step": 5614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5609410405158997,
+      "learning_rate": 0.0005876300642484189,
+      "loss": 3.4627,
+      "step": 5615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5721429586410522,
+      "learning_rate": 0.0005876256883690739,
+      "loss": 3.5989,
+      "step": 5616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.531762957572937,
+      "learning_rate": 0.0005876213117321781,
+      "loss": 3.5232,
+      "step": 5617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6050277352333069,
+      "learning_rate": 0.0005876169343377432,
+      "loss": 3.2072,
+      "step": 5618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.536950945854187,
+      "learning_rate": 0.0005876125561857806,
+      "loss": 3.4877,
+      "step": 5619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5331342816352844,
+      "learning_rate": 0.0005876081772763019,
+      "loss": 3.3034,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5625576972961426,
+      "learning_rate": 0.0005876037976093188,
+      "loss": 3.2909,
+      "step": 5621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5438462495803833,
+      "learning_rate": 0.0005875994171848424,
+      "loss": 3.5022,
+      "step": 5622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5512886643409729,
+      "learning_rate": 0.0005875950360028846,
+      "loss": 3.4472,
+      "step": 5623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5511341094970703,
+      "learning_rate": 0.0005875906540634567,
+      "loss": 3.1134,
+      "step": 5624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5950229167938232,
+      "learning_rate": 0.0005875862713665705,
+      "loss": 3.5233,
+      "step": 5625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5315006375312805,
+      "learning_rate": 0.0005875818879122373,
+      "loss": 3.6534,
+      "step": 5626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5488346815109253,
+      "learning_rate": 0.0005875775037004689,
+      "loss": 3.3631,
+      "step": 5627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5246545672416687,
+      "learning_rate": 0.0005875731187312766,
+      "loss": 3.5114,
+      "step": 5628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5800779461860657,
+      "learning_rate": 0.000587568733004672,
+      "loss": 3.4159,
+      "step": 5629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5169866681098938,
+      "learning_rate": 0.0005875643465206668,
+      "loss": 3.2865,
+      "step": 5630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5645882487297058,
+      "learning_rate": 0.0005875599592792724,
+      "loss": 3.4028,
+      "step": 5631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5463460087776184,
+      "learning_rate": 0.0005875555712805005,
+      "loss": 3.4289,
+      "step": 5632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5568528771400452,
+      "learning_rate": 0.0005875511825243624,
+      "loss": 3.5229,
+      "step": 5633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.529748260974884,
+      "learning_rate": 0.00058754679301087,
+      "loss": 3.4635,
+      "step": 5634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5450758337974548,
+      "learning_rate": 0.0005875424027400346,
+      "loss": 3.1628,
+      "step": 5635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.536340594291687,
+      "learning_rate": 0.0005875380117118679,
+      "loss": 3.4757,
+      "step": 5636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5831730961799622,
+      "learning_rate": 0.0005875336199263814,
+      "loss": 3.4245,
+      "step": 5637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.532107949256897,
+      "learning_rate": 0.0005875292273835865,
+      "loss": 3.304,
+      "step": 5638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5280243754386902,
+      "learning_rate": 0.0005875248340834953,
+      "loss": 3.2755,
+      "step": 5639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5418915748596191,
+      "learning_rate": 0.0005875204400261187,
+      "loss": 3.337,
+      "step": 5640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5245290398597717,
+      "learning_rate": 0.0005875160452114689,
+      "loss": 3.528,
+      "step": 5641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5836740732192993,
+      "learning_rate": 0.0005875116496395569,
+      "loss": 3.3991,
+      "step": 5642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5641128420829773,
+      "learning_rate": 0.0005875072533103946,
+      "loss": 3.2812,
+      "step": 5643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5404115915298462,
+      "learning_rate": 0.0005875028562239936,
+      "loss": 3.3253,
+      "step": 5644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5600898861885071,
+      "learning_rate": 0.0005874984583803653,
+      "loss": 3.2992,
+      "step": 5645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5351735949516296,
+      "learning_rate": 0.0005874940597795215,
+      "loss": 3.5004,
+      "step": 5646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.561495840549469,
+      "learning_rate": 0.0005874896604214737,
+      "loss": 3.392,
+      "step": 5647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5557546615600586,
+      "learning_rate": 0.0005874852603062334,
+      "loss": 3.2979,
+      "step": 5648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5082600116729736,
+      "learning_rate": 0.0005874808594338122,
+      "loss": 3.5587,
+      "step": 5649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5817442536354065,
+      "learning_rate": 0.0005874764578042218,
+      "loss": 3.571,
+      "step": 5650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5154877305030823,
+      "learning_rate": 0.0005874720554174738,
+      "loss": 3.4611,
+      "step": 5651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5388178825378418,
+      "learning_rate": 0.0005874676522735796,
+      "loss": 3.2679,
+      "step": 5652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.568677544593811,
+      "learning_rate": 0.000587463248372551,
+      "loss": 3.3036,
+      "step": 5653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.520805835723877,
+      "learning_rate": 0.0005874588437143996,
+      "loss": 3.561,
+      "step": 5654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5892708897590637,
+      "learning_rate": 0.0005874544382991368,
+      "loss": 3.3356,
+      "step": 5655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5433066487312317,
+      "learning_rate": 0.0005874500321267743,
+      "loss": 3.4601,
+      "step": 5656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6382086873054504,
+      "learning_rate": 0.000587445625197324,
+      "loss": 3.5814,
+      "step": 5657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5490455031394958,
+      "learning_rate": 0.000587441217510797,
+      "loss": 3.4123,
+      "step": 5658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5612518191337585,
+      "learning_rate": 0.0005874368090672053,
+      "loss": 3.3378,
+      "step": 5659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578199565410614,
+      "learning_rate": 0.0005874323998665603,
+      "loss": 3.3815,
+      "step": 5660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5689168572425842,
+      "learning_rate": 0.0005874279899088735,
+      "loss": 3.3816,
+      "step": 5661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5449416637420654,
+      "learning_rate": 0.0005874235791941569,
+      "loss": 3.4455,
+      "step": 5662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5448278784751892,
+      "learning_rate": 0.0005874191677224218,
+      "loss": 3.6087,
+      "step": 5663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.521791934967041,
+      "learning_rate": 0.00058741475549368,
+      "loss": 3.3265,
+      "step": 5664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5685543417930603,
+      "learning_rate": 0.000587410342507943,
+      "loss": 3.4758,
+      "step": 5665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5723705291748047,
+      "learning_rate": 0.0005874059287652225,
+      "loss": 3.3845,
+      "step": 5666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5475627779960632,
+      "learning_rate": 0.0005874015142655302,
+      "loss": 3.4733,
+      "step": 5667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.565759539604187,
+      "learning_rate": 0.0005873970990088775,
+      "loss": 3.0086,
+      "step": 5668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5776841640472412,
+      "learning_rate": 0.0005873926829952762,
+      "loss": 3.4155,
+      "step": 5669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5339114665985107,
+      "learning_rate": 0.0005873882662247378,
+      "loss": 3.4948,
+      "step": 5670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5639076828956604,
+      "learning_rate": 0.000587383848697274,
+      "loss": 3.1924,
+      "step": 5671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5452633500099182,
+      "learning_rate": 0.0005873794304128966,
+      "loss": 3.151,
+      "step": 5672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5319305658340454,
+      "learning_rate": 0.000587375011371617,
+      "loss": 3.4559,
+      "step": 5673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.551712691783905,
+      "learning_rate": 0.000587370591573447,
+      "loss": 3.5398,
+      "step": 5674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5740674138069153,
+      "learning_rate": 0.000587366171018398,
+      "loss": 3.415,
+      "step": 5675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5360549688339233,
+      "learning_rate": 0.000587361749706482,
+      "loss": 3.3225,
+      "step": 5676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5476113557815552,
+      "learning_rate": 0.0005873573276377103,
+      "loss": 3.6012,
+      "step": 5677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5887551307678223,
+      "learning_rate": 0.0005873529048120948,
+      "loss": 3.3345,
+      "step": 5678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5409665107727051,
+      "learning_rate": 0.000587348481229647,
+      "loss": 3.3906,
+      "step": 5679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5526324510574341,
+      "learning_rate": 0.0005873440568903785,
+      "loss": 3.3438,
+      "step": 5680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5718168616294861,
+      "learning_rate": 0.0005873396317943013,
+      "loss": 3.4642,
+      "step": 5681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5631795525550842,
+      "learning_rate": 0.0005873352059414267,
+      "loss": 3.2674,
+      "step": 5682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6147184371948242,
+      "learning_rate": 0.0005873307793317664,
+      "loss": 3.2464,
+      "step": 5683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6067205667495728,
+      "learning_rate": 0.0005873263519653323,
+      "loss": 3.4046,
+      "step": 5684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.542724072933197,
+      "learning_rate": 0.0005873219238421356,
+      "loss": 3.5222,
+      "step": 5685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5695931315422058,
+      "learning_rate": 0.0005873174949621885,
+      "loss": 3.4056,
+      "step": 5686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5388402938842773,
+      "learning_rate": 0.0005873130653255023,
+      "loss": 3.4411,
+      "step": 5687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6163471341133118,
+      "learning_rate": 0.0005873086349320888,
+      "loss": 3.3085,
+      "step": 5688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5419004559516907,
+      "learning_rate": 0.0005873042037819597,
+      "loss": 3.5004,
+      "step": 5689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5461073517799377,
+      "learning_rate": 0.0005872997718751265,
+      "loss": 3.3721,
+      "step": 5690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5264971852302551,
+      "learning_rate": 0.0005872953392116011,
+      "loss": 3.3873,
+      "step": 5691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5653229355812073,
+      "learning_rate": 0.0005872909057913951,
+      "loss": 3.3069,
+      "step": 5692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5882543325424194,
+      "learning_rate": 0.00058728647161452,
+      "loss": 3.1904,
+      "step": 5693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5486871600151062,
+      "learning_rate": 0.0005872820366809877,
+      "loss": 3.2986,
+      "step": 5694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53853440284729,
+      "learning_rate": 0.0005872776009908099,
+      "loss": 3.4683,
+      "step": 5695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5672714710235596,
+      "learning_rate": 0.000587273164543998,
+      "loss": 3.4573,
+      "step": 5696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5596035122871399,
+      "learning_rate": 0.000587268727340564,
+      "loss": 3.556,
+      "step": 5697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5245596766471863,
+      "learning_rate": 0.0005872642893805194,
+      "loss": 3.3127,
+      "step": 5698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5364984273910522,
+      "learning_rate": 0.0005872598506638761,
+      "loss": 3.3348,
+      "step": 5699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5501704812049866,
+      "learning_rate": 0.0005872554111906454,
+      "loss": 3.352,
+      "step": 5700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5427786707878113,
+      "learning_rate": 0.0005872509709608394,
+      "loss": 3.4774,
+      "step": 5701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5833746194839478,
+      "learning_rate": 0.0005872465299744696,
+      "loss": 3.2135,
+      "step": 5702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5424846410751343,
+      "learning_rate": 0.0005872420882315476,
+      "loss": 3.4114,
+      "step": 5703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5223133563995361,
+      "learning_rate": 0.0005872376457320853,
+      "loss": 3.2703,
+      "step": 5704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5719478726387024,
+      "learning_rate": 0.0005872332024760944,
+      "loss": 3.5279,
+      "step": 5705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5417484641075134,
+      "learning_rate": 0.0005872287584635864,
+      "loss": 3.5221,
+      "step": 5706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5238466262817383,
+      "learning_rate": 0.0005872243136945732,
+      "loss": 3.3905,
+      "step": 5707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5247589349746704,
+      "learning_rate": 0.0005872198681690664,
+      "loss": 3.3364,
+      "step": 5708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5623005628585815,
+      "learning_rate": 0.0005872154218870778,
+      "loss": 3.3918,
+      "step": 5709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6546081900596619,
+      "learning_rate": 0.0005872109748486189,
+      "loss": 3.6369,
+      "step": 5710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5542550683021545,
+      "learning_rate": 0.0005872065270537017,
+      "loss": 3.4674,
+      "step": 5711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5579176545143127,
+      "learning_rate": 0.0005872020785023379,
+      "loss": 3.3265,
+      "step": 5712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6107970476150513,
+      "learning_rate": 0.0005871976291945388,
+      "loss": 3.5986,
+      "step": 5713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.534400463104248,
+      "learning_rate": 0.0005871931791303167,
+      "loss": 3.5339,
+      "step": 5714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5474011898040771,
+      "learning_rate": 0.000587188728309683,
+      "loss": 3.2851,
+      "step": 5715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5572850108146667,
+      "learning_rate": 0.0005871842767326492,
+      "loss": 3.3349,
+      "step": 5716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6158696413040161,
+      "learning_rate": 0.0005871798243992276,
+      "loss": 3.3417,
+      "step": 5717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5303359627723694,
+      "learning_rate": 0.0005871753713094294,
+      "loss": 3.5762,
+      "step": 5718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5638172030448914,
+      "learning_rate": 0.0005871709174632666,
+      "loss": 3.3489,
+      "step": 5719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5369972586631775,
+      "learning_rate": 0.0005871664628607509,
+      "loss": 3.2923,
+      "step": 5720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6055926084518433,
+      "learning_rate": 0.000587162007501894,
+      "loss": 3.145,
+      "step": 5721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5599731802940369,
+      "learning_rate": 0.0005871575513867076,
+      "loss": 3.5095,
+      "step": 5722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5579766035079956,
+      "learning_rate": 0.0005871530945152035,
+      "loss": 3.6883,
+      "step": 5723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5651584267616272,
+      "learning_rate": 0.0005871486368873934,
+      "loss": 3.4495,
+      "step": 5724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.000587144178503289,
+      "loss": 3.2469,
+      "step": 5725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.556201159954071,
+      "learning_rate": 0.0005871397193629022,
+      "loss": 3.5754,
+      "step": 5726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5447879433631897,
+      "learning_rate": 0.0005871352594662446,
+      "loss": 3.4798,
+      "step": 5727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5088903307914734,
+      "learning_rate": 0.000587130798813328,
+      "loss": 3.4272,
+      "step": 5728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5600134134292603,
+      "learning_rate": 0.0005871263374041642,
+      "loss": 3.3428,
+      "step": 5729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5825614333152771,
+      "learning_rate": 0.0005871218752387647,
+      "loss": 3.3385,
+      "step": 5730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5817743539810181,
+      "learning_rate": 0.0005871174123171415,
+      "loss": 3.3358,
+      "step": 5731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5553821325302124,
+      "learning_rate": 0.0005871129486393064,
+      "loss": 3.3266,
+      "step": 5732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.583930253982544,
+      "learning_rate": 0.0005871084842052711,
+      "loss": 3.3519,
+      "step": 5733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5544360280036926,
+      "learning_rate": 0.0005871040190150471,
+      "loss": 3.2843,
+      "step": 5734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5816003084182739,
+      "learning_rate": 0.0005870995530686465,
+      "loss": 3.5052,
+      "step": 5735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5861048698425293,
+      "learning_rate": 0.0005870950863660808,
+      "loss": 3.4918,
+      "step": 5736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5465466380119324,
+      "learning_rate": 0.000587090618907362,
+      "loss": 3.5772,
+      "step": 5737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5588349103927612,
+      "learning_rate": 0.0005870861506925018,
+      "loss": 3.5592,
+      "step": 5738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.558573842048645,
+      "learning_rate": 0.0005870816817215119,
+      "loss": 3.4306,
+      "step": 5739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5814093351364136,
+      "learning_rate": 0.0005870772119944041,
+      "loss": 3.3825,
+      "step": 5740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5688466429710388,
+      "learning_rate": 0.0005870727415111901,
+      "loss": 3.4397,
+      "step": 5741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5587722063064575,
+      "learning_rate": 0.0005870682702718817,
+      "loss": 3.4223,
+      "step": 5742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578432559967041,
+      "learning_rate": 0.000587063798276491,
+      "loss": 3.6142,
+      "step": 5743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5772233605384827,
+      "learning_rate": 0.0005870593255250293,
+      "loss": 3.4018,
+      "step": 5744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5719894170761108,
+      "learning_rate": 0.0005870548520175086,
+      "loss": 3.3818,
+      "step": 5745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5581337213516235,
+      "learning_rate": 0.0005870503777539406,
+      "loss": 3.6375,
+      "step": 5746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5520333051681519,
+      "learning_rate": 0.0005870459027343373,
+      "loss": 3.5316,
+      "step": 5747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.524156928062439,
+      "learning_rate": 0.0005870414269587102,
+      "loss": 3.2556,
+      "step": 5748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5664880871772766,
+      "learning_rate": 0.0005870369504270713,
+      "loss": 3.3596,
+      "step": 5749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5417056679725647,
+      "learning_rate": 0.0005870324731394323,
+      "loss": 3.4458,
+      "step": 5750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5310121774673462,
+      "learning_rate": 0.0005870279950958051,
+      "loss": 3.3986,
+      "step": 5751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5482550263404846,
+      "learning_rate": 0.0005870235162962012,
+      "loss": 3.2324,
+      "step": 5752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6072642207145691,
+      "learning_rate": 0.0005870190367406327,
+      "loss": 3.5553,
+      "step": 5753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5456782579421997,
+      "learning_rate": 0.0005870145564291113,
+      "loss": 3.18,
+      "step": 5754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5461390018463135,
+      "learning_rate": 0.0005870100753616488,
+      "loss": 3.4445,
+      "step": 5755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5472182035446167,
+      "learning_rate": 0.000587005593538257,
+      "loss": 3.4471,
+      "step": 5756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5583928823471069,
+      "learning_rate": 0.0005870011109589477,
+      "loss": 3.3076,
+      "step": 5757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5505167841911316,
+      "learning_rate": 0.0005869966276237327,
+      "loss": 3.3274,
+      "step": 5758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5618990659713745,
+      "learning_rate": 0.0005869921435326238,
+      "loss": 3.4508,
+      "step": 5759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5286376476287842,
+      "learning_rate": 0.0005869876586856328,
+      "loss": 3.207,
+      "step": 5760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5474581718444824,
+      "learning_rate": 0.0005869831730827715,
+      "loss": 3.6744,
+      "step": 5761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5827766060829163,
+      "learning_rate": 0.0005869786867240519,
+      "loss": 3.3691,
+      "step": 5762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5204379558563232,
+      "learning_rate": 0.0005869741996094856,
+      "loss": 3.7098,
+      "step": 5763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5318394899368286,
+      "learning_rate": 0.0005869697117390846,
+      "loss": 3.5282,
+      "step": 5764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.520143985748291,
+      "learning_rate": 0.0005869652231128604,
+      "loss": 3.621,
+      "step": 5765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5361225008964539,
+      "learning_rate": 0.0005869607337308251,
+      "loss": 3.4061,
+      "step": 5766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6079183220863342,
+      "learning_rate": 0.0005869562435929905,
+      "loss": 3.387,
+      "step": 5767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5326195359230042,
+      "learning_rate": 0.0005869517526993684,
+      "loss": 3.4427,
+      "step": 5768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.541530191898346,
+      "learning_rate": 0.0005869472610499706,
+      "loss": 3.4417,
+      "step": 5769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5230389833450317,
+      "learning_rate": 0.0005869427686448088,
+      "loss": 3.2022,
+      "step": 5770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.566440761089325,
+      "learning_rate": 0.0005869382754838951,
+      "loss": 3.3521,
+      "step": 5771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5397255420684814,
+      "learning_rate": 0.0005869337815672413,
+      "loss": 3.5468,
+      "step": 5772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5574547648429871,
+      "learning_rate": 0.0005869292868948589,
+      "loss": 3.13,
+      "step": 5773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5912591218948364,
+      "learning_rate": 0.0005869247914667601,
+      "loss": 3.2856,
+      "step": 5774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5619760751724243,
+      "learning_rate": 0.0005869202952829567,
+      "loss": 3.3057,
+      "step": 5775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5392544865608215,
+      "learning_rate": 0.0005869157983434604,
+      "loss": 3.4446,
+      "step": 5776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5367756485939026,
+      "learning_rate": 0.000586911300648283,
+      "loss": 3.4562,
+      "step": 5777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5628572702407837,
+      "learning_rate": 0.0005869068021974366,
+      "loss": 3.5371,
+      "step": 5778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5055538415908813,
+      "learning_rate": 0.0005869023029909328,
+      "loss": 3.5217,
+      "step": 5779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546389102935791,
+      "learning_rate": 0.0005868978030287836,
+      "loss": 3.4341,
+      "step": 5780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5764334201812744,
+      "learning_rate": 0.0005868933023110008,
+      "loss": 3.0859,
+      "step": 5781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6767985820770264,
+      "learning_rate": 0.0005868888008375963,
+      "loss": 3.332,
+      "step": 5782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5384335517883301,
+      "learning_rate": 0.0005868842986085818,
+      "loss": 3.4159,
+      "step": 5783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5437288284301758,
+      "learning_rate": 0.0005868797956239693,
+      "loss": 3.2161,
+      "step": 5784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5375307202339172,
+      "learning_rate": 0.0005868752918837707,
+      "loss": 3.3611,
+      "step": 5785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5247658491134644,
+      "learning_rate": 0.0005868707873879978,
+      "loss": 3.2946,
+      "step": 5786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5466168522834778,
+      "learning_rate": 0.0005868662821366624,
+      "loss": 3.4666,
+      "step": 5787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5654569864273071,
+      "learning_rate": 0.0005868617761297764,
+      "loss": 3.4604,
+      "step": 5788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5402109026908875,
+      "learning_rate": 0.0005868572693673516,
+      "loss": 3.359,
+      "step": 5789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.517817497253418,
+      "learning_rate": 0.0005868527618494001,
+      "loss": 3.3345,
+      "step": 5790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49960243701934814,
+      "learning_rate": 0.0005868482535759337,
+      "loss": 3.4715,
+      "step": 5791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.540615439414978,
+      "learning_rate": 0.000586843744546964,
+      "loss": 3.4621,
+      "step": 5792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5866577625274658,
+      "learning_rate": 0.0005868392347625032,
+      "loss": 3.2817,
+      "step": 5793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5393642783164978,
+      "learning_rate": 0.0005868347242225632,
+      "loss": 3.2075,
+      "step": 5794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5420916080474854,
+      "learning_rate": 0.0005868302129271555,
+      "loss": 3.4428,
+      "step": 5795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5432751774787903,
+      "learning_rate": 0.0005868257008762924,
+      "loss": 3.4497,
+      "step": 5796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5415337085723877,
+      "learning_rate": 0.0005868211880699854,
+      "loss": 3.3068,
+      "step": 5797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5475549697875977,
+      "learning_rate": 0.0005868166745082468,
+      "loss": 3.4784,
+      "step": 5798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5423837900161743,
+      "learning_rate": 0.0005868121601910882,
+      "loss": 3.2155,
+      "step": 5799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52655428647995,
+      "learning_rate": 0.0005868076451185215,
+      "loss": 3.4475,
+      "step": 5800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5762920379638672,
+      "learning_rate": 0.0005868031292905586,
+      "loss": 3.3285,
+      "step": 5801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5149837732315063,
+      "learning_rate": 0.0005867986127072116,
+      "loss": 3.5152,
+      "step": 5802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5225587487220764,
+      "learning_rate": 0.0005867940953684921,
+      "loss": 3.276,
+      "step": 5803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6301308870315552,
+      "learning_rate": 0.0005867895772744124,
+      "loss": 3.3187,
+      "step": 5804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5626131892204285,
+      "learning_rate": 0.0005867850584249841,
+      "loss": 3.4969,
+      "step": 5805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5513469576835632,
+      "learning_rate": 0.0005867805388202189,
+      "loss": 3.2877,
+      "step": 5806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5425329208374023,
+      "learning_rate": 0.0005867760184601292,
+      "loss": 3.5523,
+      "step": 5807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5266916155815125,
+      "learning_rate": 0.0005867714973447265,
+      "loss": 3.0832,
+      "step": 5808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5299080014228821,
+      "learning_rate": 0.000586766975474023,
+      "loss": 3.4578,
+      "step": 5809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.510150671005249,
+      "learning_rate": 0.0005867624528480303,
+      "loss": 3.3181,
+      "step": 5810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5414705872535706,
+      "learning_rate": 0.0005867579294667606,
+      "loss": 3.4233,
+      "step": 5811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5347448587417603,
+      "learning_rate": 0.0005867534053302258,
+      "loss": 3.6438,
+      "step": 5812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5466963648796082,
+      "learning_rate": 0.0005867488804384377,
+      "loss": 3.5877,
+      "step": 5813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5780187249183655,
+      "learning_rate": 0.0005867443547914081,
+      "loss": 3.4495,
+      "step": 5814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5186501145362854,
+      "learning_rate": 0.0005867398283891491,
+      "loss": 3.4562,
+      "step": 5815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5260766744613647,
+      "learning_rate": 0.0005867353012316725,
+      "loss": 3.2264,
+      "step": 5816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5569272041320801,
+      "learning_rate": 0.0005867307733189905,
+      "loss": 3.3634,
+      "step": 5817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5397284030914307,
+      "learning_rate": 0.0005867262446511147,
+      "loss": 3.2977,
+      "step": 5818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5917802453041077,
+      "learning_rate": 0.0005867217152280571,
+      "loss": 3.5556,
+      "step": 5819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5505132079124451,
+      "learning_rate": 0.0005867171850498298,
+      "loss": 3.4379,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5808700919151306,
+      "learning_rate": 0.0005867126541164445,
+      "loss": 3.3594,
+      "step": 5821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5809784531593323,
+      "learning_rate": 0.0005867081224279133,
+      "loss": 3.4292,
+      "step": 5822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5430020093917847,
+      "learning_rate": 0.0005867035899842481,
+      "loss": 3.3711,
+      "step": 5823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6083835959434509,
+      "learning_rate": 0.0005866990567854608,
+      "loss": 3.4043,
+      "step": 5824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5165572166442871,
+      "learning_rate": 0.0005866945228315634,
+      "loss": 3.6597,
+      "step": 5825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5558134913444519,
+      "learning_rate": 0.0005866899881225678,
+      "loss": 3.2103,
+      "step": 5826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5334741473197937,
+      "learning_rate": 0.0005866854526584859,
+      "loss": 3.087,
+      "step": 5827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5371214747428894,
+      "learning_rate": 0.0005866809164393297,
+      "loss": 3.4371,
+      "step": 5828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5531008839607239,
+      "learning_rate": 0.0005866763794651111,
+      "loss": 3.2865,
+      "step": 5829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5216488242149353,
+      "learning_rate": 0.0005866718417358421,
+      "loss": 3.4875,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.542599618434906,
+      "learning_rate": 0.0005866673032515347,
+      "loss": 3.3621,
+      "step": 5831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5730884075164795,
+      "learning_rate": 0.0005866627640122008,
+      "loss": 3.2273,
+      "step": 5832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5188013911247253,
+      "learning_rate": 0.0005866582240178523,
+      "loss": 3.2409,
+      "step": 5833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5457878708839417,
+      "learning_rate": 0.0005866536832685013,
+      "loss": 3.2904,
+      "step": 5834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.569889485836029,
+      "learning_rate": 0.0005866491417641595,
+      "loss": 3.4469,
+      "step": 5835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5618494153022766,
+      "learning_rate": 0.0005866445995048392,
+      "loss": 3.3618,
+      "step": 5836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5545337796211243,
+      "learning_rate": 0.0005866400564905521,
+      "loss": 3.3963,
+      "step": 5837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5397124290466309,
+      "learning_rate": 0.0005866355127213102,
+      "loss": 3.4804,
+      "step": 5838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5252190828323364,
+      "learning_rate": 0.0005866309681971256,
+      "loss": 3.5564,
+      "step": 5839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5801165699958801,
+      "learning_rate": 0.0005866264229180102,
+      "loss": 3.5835,
+      "step": 5840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.536383330821991,
+      "learning_rate": 0.000586621876883976,
+      "loss": 3.4126,
+      "step": 5841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5214142203330994,
+      "learning_rate": 0.0005866173300950349,
+      "loss": 3.2283,
+      "step": 5842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5730195641517639,
+      "learning_rate": 0.0005866127825511989,
+      "loss": 3.5788,
+      "step": 5843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.527881383895874,
+      "learning_rate": 0.00058660823425248,
+      "loss": 3.5447,
+      "step": 5844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5443788766860962,
+      "learning_rate": 0.0005866036851988901,
+      "loss": 3.2235,
+      "step": 5845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6077579855918884,
+      "learning_rate": 0.0005865991353904413,
+      "loss": 3.3991,
+      "step": 5846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578214704990387,
+      "learning_rate": 0.0005865945848271455,
+      "loss": 3.2017,
+      "step": 5847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6207023859024048,
+      "learning_rate": 0.0005865900335090149,
+      "loss": 3.4464,
+      "step": 5848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5436177253723145,
+      "learning_rate": 0.0005865854814360612,
+      "loss": 3.3529,
+      "step": 5849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6536390781402588,
+      "learning_rate": 0.0005865809286082963,
+      "loss": 3.4443,
+      "step": 5850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5636897087097168,
+      "learning_rate": 0.0005865763750257327,
+      "loss": 3.4031,
+      "step": 5851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5775346159934998,
+      "learning_rate": 0.0005865718206883819,
+      "loss": 3.2281,
+      "step": 5852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5930302143096924,
+      "learning_rate": 0.0005865672655962561,
+      "loss": 3.2219,
+      "step": 5853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5139269828796387,
+      "learning_rate": 0.0005865627097493672,
+      "loss": 3.5864,
+      "step": 5854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5036550760269165,
+      "learning_rate": 0.0005865581531477274,
+      "loss": 3.5422,
+      "step": 5855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6104763746261597,
+      "learning_rate": 0.0005865535957913486,
+      "loss": 3.4522,
+      "step": 5856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5676093697547913,
+      "learning_rate": 0.0005865490376802427,
+      "loss": 3.4295,
+      "step": 5857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.538794994354248,
+      "learning_rate": 0.0005865444788144217,
+      "loss": 3.2915,
+      "step": 5858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5480157732963562,
+      "learning_rate": 0.0005865399191938979,
+      "loss": 3.5967,
+      "step": 5859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5287188291549683,
+      "learning_rate": 0.000586535358818683,
+      "loss": 3.4753,
+      "step": 5860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5639832615852356,
+      "learning_rate": 0.0005865307976887891,
+      "loss": 3.1815,
+      "step": 5861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5048469305038452,
+      "learning_rate": 0.0005865262358042281,
+      "loss": 3.4191,
+      "step": 5862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5484185814857483,
+      "learning_rate": 0.0005865216731650123,
+      "loss": 3.4258,
+      "step": 5863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5514761209487915,
+      "learning_rate": 0.0005865171097711535,
+      "loss": 3.2908,
+      "step": 5864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6602954864501953,
+      "learning_rate": 0.0005865125456226638,
+      "loss": 3.5131,
+      "step": 5865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6218395829200745,
+      "learning_rate": 0.0005865079807195552,
+      "loss": 3.3947,
+      "step": 5866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5158183574676514,
+      "learning_rate": 0.0005865034150618397,
+      "loss": 3.2724,
+      "step": 5867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5524356961250305,
+      "learning_rate": 0.0005864988486495293,
+      "loss": 3.1983,
+      "step": 5868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5388345718383789,
+      "learning_rate": 0.000586494281482636,
+      "loss": 3.4331,
+      "step": 5869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.532199501991272,
+      "learning_rate": 0.0005864897135611721,
+      "loss": 3.3253,
+      "step": 5870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6542239189147949,
+      "learning_rate": 0.0005864851448851493,
+      "loss": 3.3342,
+      "step": 5871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5302449464797974,
+      "learning_rate": 0.0005864805754545798,
+      "loss": 3.2958,
+      "step": 5872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5232753753662109,
+      "learning_rate": 0.0005864760052694756,
+      "loss": 3.6324,
+      "step": 5873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6836511492729187,
+      "learning_rate": 0.0005864714343298488,
+      "loss": 3.4663,
+      "step": 5874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5641322135925293,
+      "learning_rate": 0.0005864668626357112,
+      "loss": 3.3574,
+      "step": 5875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5231789350509644,
+      "learning_rate": 0.0005864622901870753,
+      "loss": 3.4077,
+      "step": 5876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5678579807281494,
+      "learning_rate": 0.0005864577169839525,
+      "loss": 3.2796,
+      "step": 5877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5374036431312561,
+      "learning_rate": 0.0005864531430263555,
+      "loss": 3.342,
+      "step": 5878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5334300994873047,
+      "learning_rate": 0.000586448568314296,
+      "loss": 3.6034,
+      "step": 5879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5521325469017029,
+      "learning_rate": 0.0005864439928477859,
+      "loss": 3.1953,
+      "step": 5880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.563149094581604,
+      "learning_rate": 0.0005864394166268376,
+      "loss": 3.4593,
+      "step": 5881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5999089479446411,
+      "learning_rate": 0.000586434839651463,
+      "loss": 3.5447,
+      "step": 5882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5314249992370605,
+      "learning_rate": 0.0005864302619216742,
+      "loss": 3.0678,
+      "step": 5883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5752025842666626,
+      "learning_rate": 0.0005864256834374832,
+      "loss": 3.3008,
+      "step": 5884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6359145045280457,
+      "learning_rate": 0.0005864211041989021,
+      "loss": 3.1086,
+      "step": 5885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5294399857521057,
+      "learning_rate": 0.0005864165242059427,
+      "loss": 3.4525,
+      "step": 5886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078052878379822,
+      "learning_rate": 0.0005864119434586176,
+      "loss": 3.6026,
+      "step": 5887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5631575584411621,
+      "learning_rate": 0.0005864073619569383,
+      "loss": 3.3792,
+      "step": 5888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5679768323898315,
+      "learning_rate": 0.0005864027797009173,
+      "loss": 3.169,
+      "step": 5889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5411725044250488,
+      "learning_rate": 0.0005863981966905665,
+      "loss": 3.448,
+      "step": 5890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5465644598007202,
+      "learning_rate": 0.0005863936129258979,
+      "loss": 3.3566,
+      "step": 5891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5652481317520142,
+      "learning_rate": 0.0005863890284069236,
+      "loss": 3.1992,
+      "step": 5892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5486741065979004,
+      "learning_rate": 0.0005863844431336559,
+      "loss": 3.1683,
+      "step": 5893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5372270941734314,
+      "learning_rate": 0.0005863798571061065,
+      "loss": 3.3014,
+      "step": 5894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.609737753868103,
+      "learning_rate": 0.0005863752703242878,
+      "loss": 3.0947,
+      "step": 5895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5803235769271851,
+      "learning_rate": 0.0005863706827882116,
+      "loss": 3.4442,
+      "step": 5896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.560324490070343,
+      "learning_rate": 0.0005863660944978904,
+      "loss": 3.2564,
+      "step": 5897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.532012403011322,
+      "learning_rate": 0.0005863615054533357,
+      "loss": 3.5374,
+      "step": 5898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6232673525810242,
+      "learning_rate": 0.0005863569156545601,
+      "loss": 3.4147,
+      "step": 5899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5468286871910095,
+      "learning_rate": 0.0005863523251015755,
+      "loss": 3.4548,
+      "step": 5900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5706323981285095,
+      "learning_rate": 0.0005863477337943939,
+      "loss": 3.3646,
+      "step": 5901
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5574482083320618,
+      "learning_rate": 0.0005863431417330275,
+      "loss": 3.5875,
+      "step": 5902
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5878828763961792,
+      "learning_rate": 0.0005863385489174883,
+      "loss": 3.4706,
+      "step": 5903
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5340185761451721,
+      "learning_rate": 0.0005863339553477887,
+      "loss": 3.5311,
+      "step": 5904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5711298584938049,
+      "learning_rate": 0.0005863293610239404,
+      "loss": 3.2338,
+      "step": 5905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5839642286300659,
+      "learning_rate": 0.0005863247659459557,
+      "loss": 3.3848,
+      "step": 5906
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5418100953102112,
+      "learning_rate": 0.0005863201701138466,
+      "loss": 3.4485,
+      "step": 5907
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5416857600212097,
+      "learning_rate": 0.0005863155735276254,
+      "loss": 3.3346,
+      "step": 5908
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5420189499855042,
+      "learning_rate": 0.000586310976187304,
+      "loss": 3.391,
+      "step": 5909
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5418663024902344,
+      "learning_rate": 0.0005863063780928946,
+      "loss": 3.5192,
+      "step": 5910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5965156555175781,
+      "learning_rate": 0.0005863017792444092,
+      "loss": 3.3288,
+      "step": 5911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5291689038276672,
+      "learning_rate": 0.0005862971796418603,
+      "loss": 3.4977,
+      "step": 5912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5859507322311401,
+      "learning_rate": 0.0005862925792852595,
+      "loss": 3.3662,
+      "step": 5913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6109771728515625,
+      "learning_rate": 0.0005862879781746192,
+      "loss": 3.2146,
+      "step": 5914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5541035532951355,
+      "learning_rate": 0.0005862833763099515,
+      "loss": 3.3712,
+      "step": 5915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5736545324325562,
+      "learning_rate": 0.0005862787736912684,
+      "loss": 3.3146,
+      "step": 5916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6064590215682983,
+      "learning_rate": 0.0005862741703185822,
+      "loss": 3.2912,
+      "step": 5917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5344297885894775,
+      "learning_rate": 0.0005862695661919048,
+      "loss": 3.3818,
+      "step": 5918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5522978901863098,
+      "learning_rate": 0.0005862649613112486,
+      "loss": 3.325,
+      "step": 5919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5202211141586304,
+      "learning_rate": 0.0005862603556766254,
+      "loss": 3.3577,
+      "step": 5920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5692675113677979,
+      "learning_rate": 0.0005862557492880477,
+      "loss": 3.6632,
+      "step": 5921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5373372435569763,
+      "learning_rate": 0.0005862511421455274,
+      "loss": 3.3315,
+      "step": 5922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5706237554550171,
+      "learning_rate": 0.0005862465342490766,
+      "loss": 3.3529,
+      "step": 5923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5713226199150085,
+      "learning_rate": 0.0005862419255987076,
+      "loss": 3.1983,
+      "step": 5924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.567156195640564,
+      "learning_rate": 0.0005862373161944325,
+      "loss": 3.279,
+      "step": 5925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5587449669837952,
+      "learning_rate": 0.0005862327060362633,
+      "loss": 3.3058,
+      "step": 5926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5511670112609863,
+      "learning_rate": 0.0005862280951242123,
+      "loss": 3.5264,
+      "step": 5927
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5433431267738342,
+      "learning_rate": 0.0005862234834582916,
+      "loss": 3.3096,
+      "step": 5928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5212329030036926,
+      "learning_rate": 0.0005862188710385131,
+      "loss": 3.2682,
+      "step": 5929
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5597579479217529,
+      "learning_rate": 0.0005862142578648895,
+      "loss": 3.3191,
+      "step": 5930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5288726687431335,
+      "learning_rate": 0.0005862096439374325,
+      "loss": 3.4722,
+      "step": 5931
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5642876029014587,
+      "learning_rate": 0.0005862050292561544,
+      "loss": 3.4802,
+      "step": 5932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5844772458076477,
+      "learning_rate": 0.0005862004138210673,
+      "loss": 3.3419,
+      "step": 5933
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52292799949646,
+      "learning_rate": 0.0005861957976321834,
+      "loss": 3.49,
+      "step": 5934
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5201006531715393,
+      "learning_rate": 0.0005861911806895148,
+      "loss": 3.429,
+      "step": 5935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5457620024681091,
+      "learning_rate": 0.0005861865629930738,
+      "loss": 3.5125,
+      "step": 5936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5805454254150391,
+      "learning_rate": 0.0005861819445428724,
+      "loss": 3.5591,
+      "step": 5937
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5340372323989868,
+      "learning_rate": 0.0005861773253389228,
+      "loss": 3.2514,
+      "step": 5938
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5509132146835327,
+      "learning_rate": 0.0005861727053812373,
+      "loss": 3.3563,
+      "step": 5939
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5151408314704895,
+      "learning_rate": 0.0005861680846698279,
+      "loss": 3.291,
+      "step": 5940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.517645537853241,
+      "learning_rate": 0.0005861634632047069,
+      "loss": 3.3734,
+      "step": 5941
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5592371821403503,
+      "learning_rate": 0.0005861588409858864,
+      "loss": 3.3977,
+      "step": 5942
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5972470641136169,
+      "learning_rate": 0.0005861542180133785,
+      "loss": 3.274,
+      "step": 5943
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5444090366363525,
+      "learning_rate": 0.0005861495942871955,
+      "loss": 3.2464,
+      "step": 5944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5292772054672241,
+      "learning_rate": 0.0005861449698073497,
+      "loss": 3.4715,
+      "step": 5945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5135490298271179,
+      "learning_rate": 0.0005861403445738529,
+      "loss": 3.339,
+      "step": 5946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5963358283042908,
+      "learning_rate": 0.0005861357185867176,
+      "loss": 3.3574,
+      "step": 5947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5650749802589417,
+      "learning_rate": 0.0005861310918459559,
+      "loss": 3.2734,
+      "step": 5948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5786062479019165,
+      "learning_rate": 0.0005861264643515799,
+      "loss": 3.3048,
+      "step": 5949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578070878982544,
+      "learning_rate": 0.0005861218361036019,
+      "loss": 3.2631,
+      "step": 5950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5257975459098816,
+      "learning_rate": 0.000586117207102034,
+      "loss": 3.6505,
+      "step": 5951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5603009462356567,
+      "learning_rate": 0.0005861125773468886,
+      "loss": 3.297,
+      "step": 5952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6115564703941345,
+      "learning_rate": 0.0005861079468381776,
+      "loss": 3.466,
+      "step": 5953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5789349675178528,
+      "learning_rate": 0.0005861033155759133,
+      "loss": 3.4272,
+      "step": 5954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.565288245677948,
+      "learning_rate": 0.000586098683560108,
+      "loss": 3.5096,
+      "step": 5955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.563439667224884,
+      "learning_rate": 0.0005860940507907738,
+      "loss": 3.2744,
+      "step": 5956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5860108137130737,
+      "learning_rate": 0.0005860894172679229,
+      "loss": 3.4212,
+      "step": 5957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5446553826332092,
+      "learning_rate": 0.0005860847829915676,
+      "loss": 3.363,
+      "step": 5958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5710722208023071,
+      "learning_rate": 0.0005860801479617201,
+      "loss": 3.2973,
+      "step": 5959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5353807806968689,
+      "learning_rate": 0.0005860755121783924,
+      "loss": 3.5851,
+      "step": 5960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5594916939735413,
+      "learning_rate": 0.0005860708756415969,
+      "loss": 3.5233,
+      "step": 5961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5313096642494202,
+      "learning_rate": 0.0005860662383513458,
+      "loss": 3.3611,
+      "step": 5962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5173554420471191,
+      "learning_rate": 0.0005860616003076514,
+      "loss": 3.4742,
+      "step": 5963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5095874071121216,
+      "learning_rate": 0.0005860569615105256,
+      "loss": 3.3907,
+      "step": 5964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5513678193092346,
+      "learning_rate": 0.0005860523219599809,
+      "loss": 3.3048,
+      "step": 5965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5937461853027344,
+      "learning_rate": 0.0005860476816560294,
+      "loss": 3.3119,
+      "step": 5966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5583381056785583,
+      "learning_rate": 0.0005860430405986833,
+      "loss": 3.3584,
+      "step": 5967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5448343753814697,
+      "learning_rate": 0.000586038398787955,
+      "loss": 3.5657,
+      "step": 5968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5197991132736206,
+      "learning_rate": 0.0005860337562238566,
+      "loss": 3.3398,
+      "step": 5969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5560016632080078,
+      "learning_rate": 0.0005860291129064003,
+      "loss": 3.2854,
+      "step": 5970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5649117231369019,
+      "learning_rate": 0.0005860244688355984,
+      "loss": 3.3855,
+      "step": 5971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56504225730896,
+      "learning_rate": 0.000586019824011463,
+      "loss": 3.3499,
+      "step": 5972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5295236706733704,
+      "learning_rate": 0.0005860151784340065,
+      "loss": 3.2858,
+      "step": 5973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5536357164382935,
+      "learning_rate": 0.0005860105321032411,
+      "loss": 3.3838,
+      "step": 5974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5174973011016846,
+      "learning_rate": 0.0005860058850191789,
+      "loss": 3.3564,
+      "step": 5975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5602399110794067,
+      "learning_rate": 0.0005860012371818324,
+      "loss": 3.3537,
+      "step": 5976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5399395823478699,
+      "learning_rate": 0.0005859965885912135,
+      "loss": 3.35,
+      "step": 5977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5447744131088257,
+      "learning_rate": 0.0005859919392473348,
+      "loss": 3.2962,
+      "step": 5978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5593130588531494,
+      "learning_rate": 0.0005859872891502083,
+      "loss": 3.3472,
+      "step": 5979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5140863656997681,
+      "learning_rate": 0.0005859826382998462,
+      "loss": 3.303,
+      "step": 5980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5375498533248901,
+      "learning_rate": 0.0005859779866962609,
+      "loss": 3.3374,
+      "step": 5981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5679370164871216,
+      "learning_rate": 0.0005859733343394648,
+      "loss": 3.4518,
+      "step": 5982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5210931301116943,
+      "learning_rate": 0.0005859686812294698,
+      "loss": 3.4724,
+      "step": 5983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5706584453582764,
+      "learning_rate": 0.0005859640273662885,
+      "loss": 3.5289,
+      "step": 5984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5336747169494629,
+      "learning_rate": 0.0005859593727499329,
+      "loss": 3.4766,
+      "step": 5985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5442331433296204,
+      "learning_rate": 0.0005859547173804152,
+      "loss": 3.2875,
+      "step": 5986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5738904476165771,
+      "learning_rate": 0.000585950061257748,
+      "loss": 3.6571,
+      "step": 5987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7155367136001587,
+      "learning_rate": 0.0005859454043819433,
+      "loss": 3.3561,
+      "step": 5988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.526063084602356,
+      "learning_rate": 0.0005859407467530134,
+      "loss": 3.3691,
+      "step": 5989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5534820556640625,
+      "learning_rate": 0.0005859360883709707,
+      "loss": 3.3648,
+      "step": 5990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5640885829925537,
+      "learning_rate": 0.0005859314292358274,
+      "loss": 3.6089,
+      "step": 5991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5343618392944336,
+      "learning_rate": 0.0005859267693475956,
+      "loss": 3.5103,
+      "step": 5992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5363438725471497,
+      "learning_rate": 0.0005859221087062878,
+      "loss": 3.4727,
+      "step": 5993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5537747144699097,
+      "learning_rate": 0.0005859174473119162,
+      "loss": 3.1938,
+      "step": 5994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5269249081611633,
+      "learning_rate": 0.0005859127851644931,
+      "loss": 3.1289,
+      "step": 5995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5200663208961487,
+      "learning_rate": 0.0005859081222640306,
+      "loss": 3.3508,
+      "step": 5996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5903448462486267,
+      "learning_rate": 0.0005859034586105412,
+      "loss": 3.3433,
+      "step": 5997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5499140024185181,
+      "learning_rate": 0.0005858987942040371,
+      "loss": 3.4114,
+      "step": 5998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5757197737693787,
+      "learning_rate": 0.0005858941290445307,
+      "loss": 3.3508,
+      "step": 5999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5636634826660156,
+      "learning_rate": 0.000585889463132034,
+      "loss": 3.495,
+      "step": 6000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5337130427360535,
+      "learning_rate": 0.0005858847964665598,
+      "loss": 3.4583,
+      "step": 6001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5628725290298462,
+      "learning_rate": 0.0005858801290481197,
+      "loss": 3.6494,
+      "step": 6002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5964218974113464,
+      "learning_rate": 0.0005858754608767266,
+      "loss": 3.3848,
+      "step": 6003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5543503165245056,
+      "learning_rate": 0.0005858707919523924,
+      "loss": 3.3203,
+      "step": 6004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5465249419212341,
+      "learning_rate": 0.0005858661222751297,
+      "loss": 3.3155,
+      "step": 6005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5593113303184509,
+      "learning_rate": 0.0005858614518449506,
+      "loss": 3.268,
+      "step": 6006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5494705438613892,
+      "learning_rate": 0.0005858567806618673,
+      "loss": 3.4332,
+      "step": 6007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5505009293556213,
+      "learning_rate": 0.0005858521087258924,
+      "loss": 3.4089,
+      "step": 6008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5421280264854431,
+      "learning_rate": 0.000585847436037038,
+      "loss": 3.431,
+      "step": 6009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5362679362297058,
+      "learning_rate": 0.0005858427625953166,
+      "loss": 3.4296,
+      "step": 6010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5678845047950745,
+      "learning_rate": 0.0005858380884007403,
+      "loss": 3.2674,
+      "step": 6011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5496119856834412,
+      "learning_rate": 0.0005858334134533214,
+      "loss": 3.5641,
+      "step": 6012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5399487018585205,
+      "learning_rate": 0.0005858287377530723,
+      "loss": 3.2081,
+      "step": 6013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5608866214752197,
+      "learning_rate": 0.0005858240613000054,
+      "loss": 3.3,
+      "step": 6014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5753394365310669,
+      "learning_rate": 0.0005858193840941329,
+      "loss": 3.3895,
+      "step": 6015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.545498788356781,
+      "learning_rate": 0.0005858147061354672,
+      "loss": 3.5889,
+      "step": 6016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5470389723777771,
+      "learning_rate": 0.0005858100274240205,
+      "loss": 3.4151,
+      "step": 6017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5118675827980042,
+      "learning_rate": 0.0005858053479598053,
+      "loss": 3.2757,
+      "step": 6018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.520729660987854,
+      "learning_rate": 0.0005858006677428337,
+      "loss": 3.4047,
+      "step": 6019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5678765773773193,
+      "learning_rate": 0.0005857959867731181,
+      "loss": 3.2805,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5853541493415833,
+      "learning_rate": 0.000585791305050671,
+      "loss": 3.3365,
+      "step": 6021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5466600656509399,
+      "learning_rate": 0.0005857866225755045,
+      "loss": 3.2846,
+      "step": 6022
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5336818099021912,
+      "learning_rate": 0.0005857819393476312,
+      "loss": 3.4511,
+      "step": 6023
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5479405522346497,
+      "learning_rate": 0.0005857772553670631,
+      "loss": 3.2949,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5725945830345154,
+      "learning_rate": 0.0005857725706338129,
+      "loss": 3.2428,
+      "step": 6025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6460245847702026,
+      "learning_rate": 0.0005857678851478925,
+      "loss": 3.3048,
+      "step": 6026
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5461933016777039,
+      "learning_rate": 0.0005857631989093147,
+      "loss": 3.1917,
+      "step": 6027
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.615522563457489,
+      "learning_rate": 0.0005857585119180915,
+      "loss": 3.2753,
+      "step": 6028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5617178678512573,
+      "learning_rate": 0.0005857538241742354,
+      "loss": 3.2788,
+      "step": 6029
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5808992981910706,
+      "learning_rate": 0.0005857491356777587,
+      "loss": 3.4891,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49388301372528076,
+      "learning_rate": 0.0005857444464286739,
+      "loss": 3.6251,
+      "step": 6031
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5432161688804626,
+      "learning_rate": 0.0005857397564269931,
+      "loss": 3.35,
+      "step": 6032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62075275182724,
+      "learning_rate": 0.0005857350656727289,
+      "loss": 3.3793,
+      "step": 6033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5460484027862549,
+      "learning_rate": 0.0005857303741658933,
+      "loss": 3.3715,
+      "step": 6034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5340959429740906,
+      "learning_rate": 0.0005857256819064991,
+      "loss": 3.2426,
+      "step": 6035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.547926127910614,
+      "learning_rate": 0.0005857209888945583,
+      "loss": 3.6102,
+      "step": 6036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.540779709815979,
+      "learning_rate": 0.0005857162951300835,
+      "loss": 3.4833,
+      "step": 6037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5335472822189331,
+      "learning_rate": 0.0005857116006130869,
+      "loss": 3.438,
+      "step": 6038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5480554699897766,
+      "learning_rate": 0.0005857069053435809,
+      "loss": 3.4828,
+      "step": 6039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5643362998962402,
+      "learning_rate": 0.000585702209321578,
+      "loss": 3.3695,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5678135752677917,
+      "learning_rate": 0.0005856975125470904,
+      "loss": 3.3233,
+      "step": 6041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5982794165611267,
+      "learning_rate": 0.0005856928150201306,
+      "loss": 3.4825,
+      "step": 6042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5603371858596802,
+      "learning_rate": 0.0005856881167407109,
+      "loss": 3.4292,
+      "step": 6043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5519452691078186,
+      "learning_rate": 0.0005856834177088436,
+      "loss": 3.6237,
+      "step": 6044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5269837379455566,
+      "learning_rate": 0.0005856787179245412,
+      "loss": 3.5085,
+      "step": 6045
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5146785378456116,
+      "learning_rate": 0.000585674017387816,
+      "loss": 3.3698,
+      "step": 6046
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5637503862380981,
+      "learning_rate": 0.0005856693160986805,
+      "loss": 3.2,
+      "step": 6047
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5364857316017151,
+      "learning_rate": 0.000585664614057147,
+      "loss": 3.4152,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.562177300453186,
+      "learning_rate": 0.0005856599112632277,
+      "loss": 3.415,
+      "step": 6049
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.548433780670166,
+      "learning_rate": 0.0005856552077169354,
+      "loss": 3.2522,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5104949474334717,
+      "learning_rate": 0.0005856505034182822,
+      "loss": 3.5135,
+      "step": 6051
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5480595231056213,
+      "learning_rate": 0.0005856457983672805,
+      "loss": 3.6652,
+      "step": 6052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5037899017333984,
+      "learning_rate": 0.0005856410925639428,
+      "loss": 3.4469,
+      "step": 6053
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5276090502738953,
+      "learning_rate": 0.0005856363860082813,
+      "loss": 3.3164,
+      "step": 6054
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5726231336593628,
+      "learning_rate": 0.0005856316787003086,
+      "loss": 3.4393,
+      "step": 6055
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.530457079410553,
+      "learning_rate": 0.0005856269706400371,
+      "loss": 3.3204,
+      "step": 6056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5776810050010681,
+      "learning_rate": 0.000585622261827479,
+      "loss": 3.0907,
+      "step": 6057
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5454444289207458,
+      "learning_rate": 0.000585617552262647,
+      "loss": 3.1768,
+      "step": 6058
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5331177115440369,
+      "learning_rate": 0.0005856128419455532,
+      "loss": 3.5736,
+      "step": 6059
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5558076500892639,
+      "learning_rate": 0.0005856081308762102,
+      "loss": 3.5581,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6561071872711182,
+      "learning_rate": 0.0005856034190546304,
+      "loss": 3.4186,
+      "step": 6061
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5517882704734802,
+      "learning_rate": 0.000585598706480826,
+      "loss": 3.2384,
+      "step": 6062
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5598825812339783,
+      "learning_rate": 0.0005855939931548097,
+      "loss": 3.2194,
+      "step": 6063
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.508906900882721,
+      "learning_rate": 0.0005855892790765937,
+      "loss": 3.4494,
+      "step": 6064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5944477319717407,
+      "learning_rate": 0.0005855845642461907,
+      "loss": 3.4631,
+      "step": 6065
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5435997247695923,
+      "learning_rate": 0.0005855798486636127,
+      "loss": 3.2613,
+      "step": 6066
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5678590536117554,
+      "learning_rate": 0.0005855751323288724,
+      "loss": 3.4153,
+      "step": 6067
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5375375151634216,
+      "learning_rate": 0.0005855704152419822,
+      "loss": 3.378,
+      "step": 6068
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5496307611465454,
+      "learning_rate": 0.0005855656974029544,
+      "loss": 3.4264,
+      "step": 6069
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6323420405387878,
+      "learning_rate": 0.0005855609788118017,
+      "loss": 3.4229,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5859168171882629,
+      "learning_rate": 0.0005855562594685362,
+      "loss": 3.4765,
+      "step": 6071
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5254231095314026,
+      "learning_rate": 0.0005855515393731704,
+      "loss": 3.326,
+      "step": 6072
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49722519516944885,
+      "learning_rate": 0.0005855468185257169,
+      "loss": 3.4279,
+      "step": 6073
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5665181279182434,
+      "learning_rate": 0.0005855420969261881,
+      "loss": 3.4913,
+      "step": 6074
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5734388828277588,
+      "learning_rate": 0.0005855373745745962,
+      "loss": 3.1717,
+      "step": 6075
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5337647795677185,
+      "learning_rate": 0.0005855326514709539,
+      "loss": 3.2627,
+      "step": 6076
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.532321572303772,
+      "learning_rate": 0.0005855279276152736,
+      "loss": 3.3021,
+      "step": 6077
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5565383434295654,
+      "learning_rate": 0.0005855232030075677,
+      "loss": 3.3376,
+      "step": 6078
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5236389636993408,
+      "learning_rate": 0.0005855184776478485,
+      "loss": 3.5663,
+      "step": 6079
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5841774940490723,
+      "learning_rate": 0.0005855137515361286,
+      "loss": 3.2004,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5455382466316223,
+      "learning_rate": 0.0005855090246724205,
+      "loss": 3.3466,
+      "step": 6081
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6889823079109192,
+      "learning_rate": 0.0005855042970567366,
+      "loss": 3.4259,
+      "step": 6082
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5753301978111267,
+      "learning_rate": 0.0005854995686890893,
+      "loss": 3.2617,
+      "step": 6083
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5561864972114563,
+      "learning_rate": 0.0005854948395694911,
+      "loss": 3.2474,
+      "step": 6084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5513240694999695,
+      "learning_rate": 0.0005854901096979543,
+      "loss": 3.2588,
+      "step": 6085
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5519996285438538,
+      "learning_rate": 0.0005854853790744917,
+      "loss": 3.4927,
+      "step": 6086
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5103303790092468,
+      "learning_rate": 0.0005854806476991154,
+      "loss": 3.6893,
+      "step": 6087
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5609255433082581,
+      "learning_rate": 0.000585475915571838,
+      "loss": 3.2162,
+      "step": 6088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5276150703430176,
+      "learning_rate": 0.000585471182692672,
+      "loss": 3.3367,
+      "step": 6089
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5595441460609436,
+      "learning_rate": 0.00058546644906163,
+      "loss": 3.4075,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5784932971000671,
+      "learning_rate": 0.0005854617146787242,
+      "loss": 3.2866,
+      "step": 6091
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5267335176467896,
+      "learning_rate": 0.000585456979543967,
+      "loss": 3.5419,
+      "step": 6092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5266568064689636,
+      "learning_rate": 0.0005854522436573714,
+      "loss": 3.2447,
+      "step": 6093
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5461187958717346,
+      "learning_rate": 0.0005854475070189493,
+      "loss": 3.4585,
+      "step": 6094
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.594099760055542,
+      "learning_rate": 0.0005854427696287134,
+      "loss": 3.42,
+      "step": 6095
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5209415555000305,
+      "learning_rate": 0.0005854380314866763,
+      "loss": 3.6037,
+      "step": 6096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5774303078651428,
+      "learning_rate": 0.0005854332925928503,
+      "loss": 3.5351,
+      "step": 6097
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5695366263389587,
+      "learning_rate": 0.000585428552947248,
+      "loss": 3.3669,
+      "step": 6098
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5497203469276428,
+      "learning_rate": 0.0005854238125498817,
+      "loss": 3.3797,
+      "step": 6099
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5589160323143005,
+      "learning_rate": 0.0005854190714007641,
+      "loss": 3.3957,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8883815407752991,
+      "learning_rate": 0.0005854143294999075,
+      "loss": 3.3728,
+      "step": 6101
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5451207756996155,
+      "learning_rate": 0.0005854095868473246,
+      "loss": 3.6537,
+      "step": 6102
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5537768006324768,
+      "learning_rate": 0.0005854048434430277,
+      "loss": 3.3309,
+      "step": 6103
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5225555896759033,
+      "learning_rate": 0.0005854000992870295,
+      "loss": 3.4239,
+      "step": 6104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5561574101448059,
+      "learning_rate": 0.0005853953543793422,
+      "loss": 3.3518,
+      "step": 6105
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5772457122802734,
+      "learning_rate": 0.0005853906087199785,
+      "loss": 3.5618,
+      "step": 6106
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5690733194351196,
+      "learning_rate": 0.0005853858623089509,
+      "loss": 3.2823,
+      "step": 6107
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5398492813110352,
+      "learning_rate": 0.0005853811151462719,
+      "loss": 3.4709,
+      "step": 6108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5477753281593323,
+      "learning_rate": 0.0005853763672319538,
+      "loss": 3.4316,
+      "step": 6109
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5604354739189148,
+      "learning_rate": 0.0005853716185660095,
+      "loss": 3.4123,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5720632076263428,
+      "learning_rate": 0.0005853668691484512,
+      "loss": 3.3442,
+      "step": 6111
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5532767176628113,
+      "learning_rate": 0.0005853621189792913,
+      "loss": 3.2747,
+      "step": 6112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390136241912842,
+      "learning_rate": 0.0005853573680585427,
+      "loss": 3.2519,
+      "step": 6113
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5394711494445801,
+      "learning_rate": 0.0005853526163862177,
+      "loss": 3.4465,
+      "step": 6114
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5646769404411316,
+      "learning_rate": 0.0005853478639623287,
+      "loss": 3.2085,
+      "step": 6115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5330178737640381,
+      "learning_rate": 0.0005853431107868886,
+      "loss": 3.5491,
+      "step": 6116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.554105281829834,
+      "learning_rate": 0.0005853383568599094,
+      "loss": 3.7143,
+      "step": 6117
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5287169218063354,
+      "learning_rate": 0.000585333602181404,
+      "loss": 3.3217,
+      "step": 6118
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5341808199882507,
+      "learning_rate": 0.0005853288467513848,
+      "loss": 3.2725,
+      "step": 6119
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5601862072944641,
+      "learning_rate": 0.0005853240905698642,
+      "loss": 3.3502,
+      "step": 6120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58427894115448,
+      "learning_rate": 0.0005853193336368551,
+      "loss": 3.4909,
+      "step": 6121
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.523365318775177,
+      "learning_rate": 0.0005853145759523695,
+      "loss": 3.47,
+      "step": 6122
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5647141337394714,
+      "learning_rate": 0.0005853098175164204,
+      "loss": 3.2372,
+      "step": 6123
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5395342111587524,
+      "learning_rate": 0.0005853050583290202,
+      "loss": 3.4103,
+      "step": 6124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5265369415283203,
+      "learning_rate": 0.0005853002983901812,
+      "loss": 3.3473,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5595860481262207,
+      "learning_rate": 0.0005852955376999163,
+      "loss": 3.4987,
+      "step": 6126
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6001688241958618,
+      "learning_rate": 0.0005852907762582377,
+      "loss": 3.3633,
+      "step": 6127
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.517382800579071,
+      "learning_rate": 0.0005852860140651583,
+      "loss": 3.3587,
+      "step": 6128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5585851669311523,
+      "learning_rate": 0.0005852812511206902,
+      "loss": 3.4912,
+      "step": 6129
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5613790154457092,
+      "learning_rate": 0.0005852764874248464,
+      "loss": 3.2364,
+      "step": 6130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5714898109436035,
+      "learning_rate": 0.0005852717229776392,
+      "loss": 3.3002,
+      "step": 6131
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59192955493927,
+      "learning_rate": 0.000585266957779081,
+      "loss": 3.3714,
+      "step": 6132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5501763820648193,
+      "learning_rate": 0.0005852621918291846,
+      "loss": 3.2259,
+      "step": 6133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5687995553016663,
+      "learning_rate": 0.0005852574251279626,
+      "loss": 3.3774,
+      "step": 6134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5306016802787781,
+      "learning_rate": 0.0005852526576754274,
+      "loss": 3.319,
+      "step": 6135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5609244704246521,
+      "learning_rate": 0.0005852478894715917,
+      "loss": 3.363,
+      "step": 6136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5303353667259216,
+      "learning_rate": 0.0005852431205164678,
+      "loss": 3.3597,
+      "step": 6137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5196649432182312,
+      "learning_rate": 0.0005852383508100685,
+      "loss": 3.3383,
+      "step": 6138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5467498302459717,
+      "learning_rate": 0.0005852335803524062,
+      "loss": 3.4312,
+      "step": 6139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5493993163108826,
+      "learning_rate": 0.0005852288091434936,
+      "loss": 3.209,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.534140944480896,
+      "learning_rate": 0.0005852240371833432,
+      "loss": 3.3898,
+      "step": 6141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5165014863014221,
+      "learning_rate": 0.0005852192644719675,
+      "loss": 3.1214,
+      "step": 6142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6956821084022522,
+      "learning_rate": 0.0005852144910093792,
+      "loss": 3.2089,
+      "step": 6143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.544562816619873,
+      "learning_rate": 0.0005852097167955909,
+      "loss": 3.3128,
+      "step": 6144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6112951040267944,
+      "learning_rate": 0.0005852049418306151,
+      "loss": 3.3378,
+      "step": 6145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.545120358467102,
+      "learning_rate": 0.0005852001661144643,
+      "loss": 3.4138,
+      "step": 6146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5408572554588318,
+      "learning_rate": 0.0005851953896471512,
+      "loss": 3.3295,
+      "step": 6147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5871040225028992,
+      "learning_rate": 0.0005851906124286882,
+      "loss": 3.3294,
+      "step": 6148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5361684560775757,
+      "learning_rate": 0.0005851858344590881,
+      "loss": 3.4245,
+      "step": 6149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.541067361831665,
+      "learning_rate": 0.0005851810557383634,
+      "loss": 3.3993,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5347387194633484,
+      "learning_rate": 0.0005851762762665267,
+      "loss": 3.1216,
+      "step": 6151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5603194236755371,
+      "learning_rate": 0.0005851714960435906,
+      "loss": 3.416,
+      "step": 6152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7698899507522583,
+      "learning_rate": 0.0005851667150695676,
+      "loss": 3.2229,
+      "step": 6153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5307655930519104,
+      "learning_rate": 0.0005851619333444703,
+      "loss": 3.2918,
+      "step": 6154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5312530994415283,
+      "learning_rate": 0.0005851571508683115,
+      "loss": 3.5303,
+      "step": 6155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5486868619918823,
+      "learning_rate": 0.0005851523676411036,
+      "loss": 3.3959,
+      "step": 6156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5212375521659851,
+      "learning_rate": 0.0005851475836628591,
+      "loss": 3.2698,
+      "step": 6157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5626296997070312,
+      "learning_rate": 0.0005851427989335909,
+      "loss": 3.4387,
+      "step": 6158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5545399785041809,
+      "learning_rate": 0.0005851380134533114,
+      "loss": 3.3047,
+      "step": 6159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5794121026992798,
+      "learning_rate": 0.0005851332272220332,
+      "loss": 3.3047,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6324692964553833,
+      "learning_rate": 0.0005851284402397691,
+      "loss": 3.3857,
+      "step": 6161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.530517041683197,
+      "learning_rate": 0.0005851236525065314,
+      "loss": 3.4357,
+      "step": 6162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5403311252593994,
+      "learning_rate": 0.000585118864022333,
+      "loss": 3.3202,
+      "step": 6163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7685026526451111,
+      "learning_rate": 0.0005851140747871863,
+      "loss": 3.2808,
+      "step": 6164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6147613525390625,
+      "learning_rate": 0.0005851092848011039,
+      "loss": 3.5743,
+      "step": 6165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5476180911064148,
+      "learning_rate": 0.0005851044940640987,
+      "loss": 3.6459,
+      "step": 6166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5747238397598267,
+      "learning_rate": 0.000585099702576183,
+      "loss": 3.4143,
+      "step": 6167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.565980076789856,
+      "learning_rate": 0.0005850949103373697,
+      "loss": 3.3223,
+      "step": 6168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5255545973777771,
+      "learning_rate": 0.0005850901173476712,
+      "loss": 3.2993,
+      "step": 6169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5456163883209229,
+      "learning_rate": 0.0005850853236071003,
+      "loss": 3.4025,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5844244956970215,
+      "learning_rate": 0.0005850805291156692,
+      "loss": 3.6342,
+      "step": 6171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5095750689506531,
+      "learning_rate": 0.0005850757338733911,
+      "loss": 3.3355,
+      "step": 6172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6050037741661072,
+      "learning_rate": 0.0005850709378802785,
+      "loss": 3.4111,
+      "step": 6173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5732804536819458,
+      "learning_rate": 0.0005850661411363437,
+      "loss": 3.5586,
+      "step": 6174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5283846855163574,
+      "learning_rate": 0.0005850613436415996,
+      "loss": 3.237,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5460174083709717,
+      "learning_rate": 0.0005850565453960586,
+      "loss": 3.171,
+      "step": 6176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5901962518692017,
+      "learning_rate": 0.0005850517463997339,
+      "loss": 3.3324,
+      "step": 6177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5234120488166809,
+      "learning_rate": 0.0005850469466526376,
+      "loss": 3.4535,
+      "step": 6178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5157477855682373,
+      "learning_rate": 0.0005850421461547823,
+      "loss": 3.4048,
+      "step": 6179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.518635094165802,
+      "learning_rate": 0.000585037344906181,
+      "loss": 3.3394,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5400442481040955,
+      "learning_rate": 0.0005850325429068462,
+      "loss": 3.3305,
+      "step": 6181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5420893430709839,
+      "learning_rate": 0.0005850277401567906,
+      "loss": 3.4494,
+      "step": 6182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5632343292236328,
+      "learning_rate": 0.0005850229366560268,
+      "loss": 3.411,
+      "step": 6183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5508242249488831,
+      "learning_rate": 0.0005850181324045673,
+      "loss": 3.1892,
+      "step": 6184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5260775089263916,
+      "learning_rate": 0.0005850133274024249,
+      "loss": 3.0536,
+      "step": 6185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5268356800079346,
+      "learning_rate": 0.0005850085216496123,
+      "loss": 3.4394,
+      "step": 6186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4986163079738617,
+      "learning_rate": 0.0005850037151461421,
+      "loss": 3.4986,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5085906386375427,
+      "learning_rate": 0.0005849989078920271,
+      "loss": 3.6332,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5513754487037659,
+      "learning_rate": 0.0005849940998872796,
+      "loss": 3.1801,
+      "step": 6189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5337885022163391,
+      "learning_rate": 0.0005849892911319126,
+      "loss": 3.3828,
+      "step": 6190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5386898517608643,
+      "learning_rate": 0.0005849844816259387,
+      "loss": 3.4224,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5503227710723877,
+      "learning_rate": 0.0005849796713693704,
+      "loss": 3.2951,
+      "step": 6192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5732369422912598,
+      "learning_rate": 0.0005849748603622205,
+      "loss": 3.4267,
+      "step": 6193
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5297451615333557,
+      "learning_rate": 0.0005849700486045018,
+      "loss": 3.3835,
+      "step": 6194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5413727760314941,
+      "learning_rate": 0.0005849652360962268,
+      "loss": 3.3712,
+      "step": 6195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5539155006408691,
+      "learning_rate": 0.000584960422837408,
+      "loss": 3.3727,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5665326118469238,
+      "learning_rate": 0.0005849556088280585,
+      "loss": 3.3948,
+      "step": 6197
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5332841873168945,
+      "learning_rate": 0.0005849507940681907,
+      "loss": 3.5232,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5573357939720154,
+      "learning_rate": 0.0005849459785578174,
+      "loss": 3.3837,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5696620345115662,
+      "learning_rate": 0.0005849411622969511,
+      "loss": 3.4472,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5535147190093994,
+      "learning_rate": 0.0005849363452856048,
+      "loss": 3.5722,
+      "step": 6201
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5474180579185486,
+      "learning_rate": 0.0005849315275237908,
+      "loss": 3.2806,
+      "step": 6202
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.519481360912323,
+      "learning_rate": 0.0005849267090115222,
+      "loss": 3.3679,
+      "step": 6203
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5893501043319702,
+      "learning_rate": 0.0005849218897488115,
+      "loss": 3.3842,
+      "step": 6204
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5267314910888672,
+      "learning_rate": 0.0005849170697356711,
+      "loss": 3.5745,
+      "step": 6205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5750004649162292,
+      "learning_rate": 0.0005849122489721142,
+      "loss": 3.2855,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5196614265441895,
+      "learning_rate": 0.0005849074274581531,
+      "loss": 3.4182,
+      "step": 6207
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5150504112243652,
+      "learning_rate": 0.0005849026051938009,
+      "loss": 3.5894,
+      "step": 6208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.532528817653656,
+      "learning_rate": 0.0005848977821790699,
+      "loss": 3.3033,
+      "step": 6209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.63649982213974,
+      "learning_rate": 0.0005848929584139731,
+      "loss": 3.3009,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5199731588363647,
+      "learning_rate": 0.0005848881338985229,
+      "loss": 3.6216,
+      "step": 6211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5677109956741333,
+      "learning_rate": 0.0005848833086327323,
+      "loss": 3.5486,
+      "step": 6212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5644501447677612,
+      "learning_rate": 0.0005848784826166139,
+      "loss": 3.2385,
+      "step": 6213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6257779002189636,
+      "learning_rate": 0.0005848736558501804,
+      "loss": 3.3152,
+      "step": 6214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6935915350914001,
+      "learning_rate": 0.0005848688283334445,
+      "loss": 3.4666,
+      "step": 6215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5401797890663147,
+      "learning_rate": 0.0005848640000664188,
+      "loss": 3.4822,
+      "step": 6216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5422528982162476,
+      "learning_rate": 0.0005848591710491164,
+      "loss": 3.32,
+      "step": 6217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5878642201423645,
+      "learning_rate": 0.0005848543412815496,
+      "loss": 3.1503,
+      "step": 6218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5432202219963074,
+      "learning_rate": 0.0005848495107637312,
+      "loss": 3.3538,
+      "step": 6219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6053672432899475,
+      "learning_rate": 0.0005848446794956742,
+      "loss": 3.1726,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5622955560684204,
+      "learning_rate": 0.0005848398474773911,
+      "loss": 3.2979,
+      "step": 6221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5476831197738647,
+      "learning_rate": 0.0005848350147088946,
+      "loss": 3.5093,
+      "step": 6222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5248444676399231,
+      "learning_rate": 0.0005848301811901974,
+      "loss": 3.3824,
+      "step": 6223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5185007452964783,
+      "learning_rate": 0.0005848253469213125,
+      "loss": 3.4452,
+      "step": 6224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.618705153465271,
+      "learning_rate": 0.0005848205119022524,
+      "loss": 3.6843,
+      "step": 6225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5357537865638733,
+      "learning_rate": 0.0005848156761330298,
+      "loss": 3.4044,
+      "step": 6226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5574560761451721,
+      "learning_rate": 0.0005848108396136576,
+      "loss": 3.2504,
+      "step": 6227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5414639115333557,
+      "learning_rate": 0.0005848060023441484,
+      "loss": 3.439,
+      "step": 6228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5341112017631531,
+      "learning_rate": 0.0005848011643245151,
+      "loss": 3.283,
+      "step": 6229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5189714431762695,
+      "learning_rate": 0.0005847963255547704,
+      "loss": 3.3486,
+      "step": 6230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6320374608039856,
+      "learning_rate": 0.0005847914860349268,
+      "loss": 3.5425,
+      "step": 6231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5230095386505127,
+      "learning_rate": 0.0005847866457649973,
+      "loss": 3.3306,
+      "step": 6232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5646961331367493,
+      "learning_rate": 0.0005847818047449946,
+      "loss": 3.4545,
+      "step": 6233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5346679091453552,
+      "learning_rate": 0.0005847769629749314,
+      "loss": 3.4056,
+      "step": 6234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5118999481201172,
+      "learning_rate": 0.0005847721204548206,
+      "loss": 3.3682,
+      "step": 6235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5368824005126953,
+      "learning_rate": 0.0005847672771846748,
+      "loss": 3.37,
+      "step": 6236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.563027024269104,
+      "learning_rate": 0.0005847624331645067,
+      "loss": 3.4851,
+      "step": 6237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5190998911857605,
+      "learning_rate": 0.0005847575883943292,
+      "loss": 3.4153,
+      "step": 6238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5824965834617615,
+      "learning_rate": 0.000584752742874155,
+      "loss": 3.5383,
+      "step": 6239
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5584920048713684,
+      "learning_rate": 0.000584747896603997,
+      "loss": 3.2722,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5643733739852905,
+      "learning_rate": 0.0005847430495838677,
+      "loss": 3.4414,
+      "step": 6241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5031020045280457,
+      "learning_rate": 0.00058473820181378,
+      "loss": 3.3009,
+      "step": 6242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5333477854728699,
+      "learning_rate": 0.0005847333532937467,
+      "loss": 3.3855,
+      "step": 6243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5437228679656982,
+      "learning_rate": 0.0005847285040237807,
+      "loss": 3.2759,
+      "step": 6244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5025436878204346,
+      "learning_rate": 0.0005847236540038944,
+      "loss": 3.3027,
+      "step": 6245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5325117111206055,
+      "learning_rate": 0.0005847188032341009,
+      "loss": 3.5029,
+      "step": 6246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4821946918964386,
+      "learning_rate": 0.0005847139517144128,
+      "loss": 3.5097,
+      "step": 6247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5105711221694946,
+      "learning_rate": 0.0005847090994448431,
+      "loss": 3.3654,
+      "step": 6248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5041024088859558,
+      "learning_rate": 0.0005847042464254043,
+      "loss": 3.4881,
+      "step": 6249
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5586902499198914,
+      "learning_rate": 0.0005846993926561093,
+      "loss": 3.2771,
+      "step": 6250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5156290531158447,
+      "learning_rate": 0.0005846945381369709,
+      "loss": 3.3685,
+      "step": 6251
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5516939759254456,
+      "learning_rate": 0.0005846896828680019,
+      "loss": 3.4328,
+      "step": 6252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.628324031829834,
+      "learning_rate": 0.0005846848268492151,
+      "loss": 3.164,
+      "step": 6253
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5220317840576172,
+      "learning_rate": 0.0005846799700806232,
+      "loss": 3.5043,
+      "step": 6254
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.523783802986145,
+      "learning_rate": 0.000584675112562239,
+      "loss": 3.5469,
+      "step": 6255
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5619279146194458,
+      "learning_rate": 0.0005846702542940755,
+      "loss": 3.3155,
+      "step": 6256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6332241296768188,
+      "learning_rate": 0.0005846653952761452,
+      "loss": 3.1784,
+      "step": 6257
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5562300682067871,
+      "learning_rate": 0.0005846605355084611,
+      "loss": 3.2124,
+      "step": 6258
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5105780363082886,
+      "learning_rate": 0.0005846556749910358,
+      "loss": 3.3471,
+      "step": 6259
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5315980911254883,
+      "learning_rate": 0.0005846508137238825,
+      "loss": 3.2881,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5600804686546326,
+      "learning_rate": 0.0005846459517070135,
+      "loss": 3.4305,
+      "step": 6261
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5662427544593811,
+      "learning_rate": 0.000584641088940442,
+      "loss": 3.3034,
+      "step": 6262
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5419402122497559,
+      "learning_rate": 0.0005846362254241805,
+      "loss": 3.4267,
+      "step": 6263
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5330461263656616,
+      "learning_rate": 0.000584631361158242,
+      "loss": 3.4127,
+      "step": 6264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5632162690162659,
+      "learning_rate": 0.0005846264961426394,
+      "loss": 3.1738,
+      "step": 6265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5450085997581482,
+      "learning_rate": 0.0005846216303773853,
+      "loss": 3.4474,
+      "step": 6266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5258246660232544,
+      "learning_rate": 0.0005846167638624926,
+      "loss": 3.3843,
+      "step": 6267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5241590738296509,
+      "learning_rate": 0.0005846118965979741,
+      "loss": 3.4877,
+      "step": 6268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5271158218383789,
+      "learning_rate": 0.0005846070285838427,
+      "loss": 3.3302,
+      "step": 6269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5716249942779541,
+      "learning_rate": 0.0005846021598201111,
+      "loss": 3.6957,
+      "step": 6270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.546226441860199,
+      "learning_rate": 0.0005845972903067922,
+      "loss": 3.4328,
+      "step": 6271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.532817006111145,
+      "learning_rate": 0.0005845924200438987,
+      "loss": 3.4793,
+      "step": 6272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5522794723510742,
+      "learning_rate": 0.0005845875490314437,
+      "loss": 3.2225,
+      "step": 6273
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5274179577827454,
+      "learning_rate": 0.0005845826772694399,
+      "loss": 2.9809,
+      "step": 6274
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5768852829933167,
+      "learning_rate": 0.0005845778047578999,
+      "loss": 3.161,
+      "step": 6275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5117722749710083,
+      "learning_rate": 0.0005845729314968369,
+      "loss": 3.3208,
+      "step": 6276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6170226335525513,
+      "learning_rate": 0.0005845680574862635,
+      "loss": 3.3823,
+      "step": 6277
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.567700207233429,
+      "learning_rate": 0.0005845631827261927,
+      "loss": 3.5089,
+      "step": 6278
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5353153347969055,
+      "learning_rate": 0.0005845583072166371,
+      "loss": 3.3691,
+      "step": 6279
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5365519523620605,
+      "learning_rate": 0.0005845534309576097,
+      "loss": 3.432,
+      "step": 6280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5466122627258301,
+      "learning_rate": 0.0005845485539491234,
+      "loss": 3.3228,
+      "step": 6281
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6011998653411865,
+      "learning_rate": 0.0005845436761911909,
+      "loss": 3.0104,
+      "step": 6282
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234431624412537,
+      "learning_rate": 0.0005845387976838251,
+      "loss": 3.3694,
+      "step": 6283
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.540686309337616,
+      "learning_rate": 0.000584533918427039,
+      "loss": 3.6565,
+      "step": 6284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5174767971038818,
+      "learning_rate": 0.0005845290384208453,
+      "loss": 3.5169,
+      "step": 6285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5468271970748901,
+      "learning_rate": 0.0005845241576652567,
+      "loss": 3.3747,
+      "step": 6286
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5398919582366943,
+      "learning_rate": 0.0005845192761602864,
+      "loss": 3.6541,
+      "step": 6287
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5341367125511169,
+      "learning_rate": 0.000584514393905947,
+      "loss": 3.3902,
+      "step": 6288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5506887435913086,
+      "learning_rate": 0.0005845095109022514,
+      "loss": 3.4969,
+      "step": 6289
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5707558393478394,
+      "learning_rate": 0.0005845046271492127,
+      "loss": 3.2885,
+      "step": 6290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49530622363090515,
+      "learning_rate": 0.0005844997426468434,
+      "loss": 3.297,
+      "step": 6291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5162578821182251,
+      "learning_rate": 0.0005844948573951565,
+      "loss": 3.3581,
+      "step": 6292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5304474234580994,
+      "learning_rate": 0.000584489971394165,
+      "loss": 3.4776,
+      "step": 6293
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5445380806922913,
+      "learning_rate": 0.0005844850846438816,
+      "loss": 3.4076,
+      "step": 6294
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5633144974708557,
+      "learning_rate": 0.0005844801971443193,
+      "loss": 3.4512,
+      "step": 6295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5193982124328613,
+      "learning_rate": 0.0005844753088954908,
+      "loss": 3.404,
+      "step": 6296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5558420419692993,
+      "learning_rate": 0.0005844704198974093,
+      "loss": 3.5152,
+      "step": 6297
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.535561203956604,
+      "learning_rate": 0.0005844655301500873,
+      "loss": 3.3756,
+      "step": 6298
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5079712271690369,
+      "learning_rate": 0.0005844606396535378,
+      "loss": 3.5013,
+      "step": 6299
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5386590361595154,
+      "learning_rate": 0.0005844557484077738,
+      "loss": 3.1547,
+      "step": 6300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5513303875923157,
+      "learning_rate": 0.000584450856412808,
+      "loss": 3.5032,
+      "step": 6301
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5191980600357056,
+      "learning_rate": 0.0005844459636686535,
+      "loss": 3.388,
+      "step": 6302
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5659469366073608,
+      "learning_rate": 0.0005844410701753231,
+      "loss": 3.3341,
+      "step": 6303
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5289974808692932,
+      "learning_rate": 0.0005844361759328295,
+      "loss": 3.4693,
+      "step": 6304
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5283775925636292,
+      "learning_rate": 0.000584431280941186,
+      "loss": 3.4818,
+      "step": 6305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5795498490333557,
+      "learning_rate": 0.000584426385200405,
+      "loss": 3.4025,
+      "step": 6306
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5070279240608215,
+      "learning_rate": 0.0005844214887104998,
+      "loss": 3.2794,
+      "step": 6307
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5226815342903137,
+      "learning_rate": 0.000584416591471483,
+      "loss": 3.4417,
+      "step": 6308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5250857472419739,
+      "learning_rate": 0.0005844116934833678,
+      "loss": 3.5051,
+      "step": 6309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5460673570632935,
+      "learning_rate": 0.0005844067947461669,
+      "loss": 3.3849,
+      "step": 6310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5464668869972229,
+      "learning_rate": 0.0005844018952598931,
+      "loss": 3.2028,
+      "step": 6311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5466782450675964,
+      "learning_rate": 0.0005843969950245595,
+      "loss": 3.1737,
+      "step": 6312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5372828841209412,
+      "learning_rate": 0.0005843920940401792,
+      "loss": 3.5099,
+      "step": 6313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5676745176315308,
+      "learning_rate": 0.0005843871923067645,
+      "loss": 3.353,
+      "step": 6314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5780575275421143,
+      "learning_rate": 0.0005843822898243289,
+      "loss": 3.2481,
+      "step": 6315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7814470529556274,
+      "learning_rate": 0.000584377386592885,
+      "loss": 3.3843,
+      "step": 6316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5636649131774902,
+      "learning_rate": 0.0005843724826124457,
+      "loss": 3.5316,
+      "step": 6317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.548147976398468,
+      "learning_rate": 0.0005843675778830241,
+      "loss": 3.1795,
+      "step": 6318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5491986870765686,
+      "learning_rate": 0.000584362672404633,
+      "loss": 3.2468,
+      "step": 6319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5343765020370483,
+      "learning_rate": 0.0005843577661772854,
+      "loss": 3.3482,
+      "step": 6320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5225366950035095,
+      "learning_rate": 0.0005843528592009941,
+      "loss": 3.6224,
+      "step": 6321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5188947319984436,
+      "learning_rate": 0.0005843479514757721,
+      "loss": 3.2754,
+      "step": 6322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5342457294464111,
+      "learning_rate": 0.0005843430430016324,
+      "loss": 3.4175,
+      "step": 6323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5652335286140442,
+      "learning_rate": 0.0005843381337785877,
+      "loss": 3.2715,
+      "step": 6324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49999839067459106,
+      "learning_rate": 0.0005843332238066512,
+      "loss": 3.3006,
+      "step": 6325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5379924178123474,
+      "learning_rate": 0.0005843283130858357,
+      "loss": 3.2138,
+      "step": 6326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5372043251991272,
+      "learning_rate": 0.0005843234016161542,
+      "loss": 3.3537,
+      "step": 6327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5262093544006348,
+      "learning_rate": 0.0005843184893976194,
+      "loss": 3.3396,
+      "step": 6328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5059529542922974,
+      "learning_rate": 0.0005843135764302446,
+      "loss": 3.4341,
+      "step": 6329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5077370405197144,
+      "learning_rate": 0.0005843086627140425,
+      "loss": 3.1674,
+      "step": 6330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5159595012664795,
+      "learning_rate": 0.000584303748249026,
+      "loss": 3.429,
+      "step": 6331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5877441763877869,
+      "learning_rate": 0.0005842988330352082,
+      "loss": 3.3886,
+      "step": 6332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5474517941474915,
+      "learning_rate": 0.000584293917072602,
+      "loss": 3.3393,
+      "step": 6333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5149897336959839,
+      "learning_rate": 0.0005842890003612204,
+      "loss": 3.4596,
+      "step": 6334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5382705330848694,
+      "learning_rate": 0.0005842840829010762,
+      "loss": 3.428,
+      "step": 6335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5204023718833923,
+      "learning_rate": 0.0005842791646921825,
+      "loss": 3.5418,
+      "step": 6336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5262681245803833,
+      "learning_rate": 0.0005842742457345523,
+      "loss": 3.3259,
+      "step": 6337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5385437607765198,
+      "learning_rate": 0.0005842693260281981,
+      "loss": 3.2901,
+      "step": 6338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5480816960334778,
+      "learning_rate": 0.0005842644055731335,
+      "loss": 3.2381,
+      "step": 6339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5376442670822144,
+      "learning_rate": 0.0005842594843693711,
+      "loss": 3.0217,
+      "step": 6340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5409801602363586,
+      "learning_rate": 0.0005842545624169239,
+      "loss": 3.2004,
+      "step": 6341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5036441087722778,
+      "learning_rate": 0.0005842496397158049,
+      "loss": 3.4932,
+      "step": 6342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6348100304603577,
+      "learning_rate": 0.000584244716266027,
+      "loss": 3.4698,
+      "step": 6343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5811209082603455,
+      "learning_rate": 0.0005842397920676032,
+      "loss": 3.4319,
+      "step": 6344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.574324369430542,
+      "learning_rate": 0.0005842348671205466,
+      "loss": 3.3559,
+      "step": 6345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5267350077629089,
+      "learning_rate": 0.00058422994142487,
+      "loss": 3.3692,
+      "step": 6346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5143805742263794,
+      "learning_rate": 0.0005842250149805865,
+      "loss": 3.4085,
+      "step": 6347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390629172325134,
+      "learning_rate": 0.000584220087787709,
+      "loss": 3.2466,
+      "step": 6348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5482839941978455,
+      "learning_rate": 0.0005842151598462504,
+      "loss": 3.4562,
+      "step": 6349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5902968049049377,
+      "learning_rate": 0.0005842102311562238,
+      "loss": 3.4197,
+      "step": 6350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.561595618724823,
+      "learning_rate": 0.0005842053017176422,
+      "loss": 3.4425,
+      "step": 6351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5435536503791809,
+      "learning_rate": 0.0005842003715305185,
+      "loss": 3.595,
+      "step": 6352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5846390724182129,
+      "learning_rate": 0.0005841954405948656,
+      "loss": 3.5844,
+      "step": 6353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5653446912765503,
+      "learning_rate": 0.0005841905089106968,
+      "loss": 3.3855,
+      "step": 6354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5833756327629089,
+      "learning_rate": 0.0005841855764780248,
+      "loss": 3.1619,
+      "step": 6355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.565251350402832,
+      "learning_rate": 0.0005841806432968626,
+      "loss": 3.2028,
+      "step": 6356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5358942747116089,
+      "learning_rate": 0.0005841757093672234,
+      "loss": 3.2987,
+      "step": 6357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5292371511459351,
+      "learning_rate": 0.0005841707746891201,
+      "loss": 3.6083,
+      "step": 6358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.606774628162384,
+      "learning_rate": 0.0005841658392625656,
+      "loss": 3.3125,
+      "step": 6359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5326677560806274,
+      "learning_rate": 0.000584160903087573,
+      "loss": 3.3493,
+      "step": 6360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5448054671287537,
+      "learning_rate": 0.0005841559661641552,
+      "loss": 3.3649,
+      "step": 6361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5255556106567383,
+      "learning_rate": 0.0005841510284923253,
+      "loss": 3.3162,
+      "step": 6362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5699189305305481,
+      "learning_rate": 0.0005841460900720963,
+      "loss": 3.2386,
+      "step": 6363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5408067107200623,
+      "learning_rate": 0.0005841411509034812,
+      "loss": 3.5189,
+      "step": 6364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.561336100101471,
+      "learning_rate": 0.000584136210986493,
+      "loss": 3.4752,
+      "step": 6365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5111557245254517,
+      "learning_rate": 0.0005841312703211447,
+      "loss": 3.3528,
+      "step": 6366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.532857358455658,
+      "learning_rate": 0.0005841263289074493,
+      "loss": 3.4392,
+      "step": 6367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5691311359405518,
+      "learning_rate": 0.0005841213867454198,
+      "loss": 3.1776,
+      "step": 6368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5483730435371399,
+      "learning_rate": 0.0005841164438350693,
+      "loss": 3.3162,
+      "step": 6369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5207633376121521,
+      "learning_rate": 0.0005841115001764107,
+      "loss": 3.3971,
+      "step": 6370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5637426972389221,
+      "learning_rate": 0.0005841065557694572,
+      "loss": 3.3626,
+      "step": 6371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5457177758216858,
+      "learning_rate": 0.0005841016106142216,
+      "loss": 3.4505,
+      "step": 6372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5361736416816711,
+      "learning_rate": 0.0005840966647107171,
+      "loss": 3.5193,
+      "step": 6373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5250360369682312,
+      "learning_rate": 0.0005840917180589566,
+      "loss": 3.3164,
+      "step": 6374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5771521925926208,
+      "learning_rate": 0.0005840867706589531,
+      "loss": 3.1594,
+      "step": 6375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5609606504440308,
+      "learning_rate": 0.0005840818225107199,
+      "loss": 3.1954,
+      "step": 6376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5278244018554688,
+      "learning_rate": 0.0005840768736142698,
+      "loss": 3.6396,
+      "step": 6377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5072911381721497,
+      "learning_rate": 0.0005840719239696159,
+      "loss": 3.4169,
+      "step": 6378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5723280906677246,
+      "learning_rate": 0.0005840669735767712,
+      "loss": 3.2226,
+      "step": 6379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5588771104812622,
+      "learning_rate": 0.0005840620224357487,
+      "loss": 3.4331,
+      "step": 6380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5622850656509399,
+      "learning_rate": 0.0005840570705465616,
+      "loss": 3.5246,
+      "step": 6381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5287413597106934,
+      "learning_rate": 0.0005840521179092228,
+      "loss": 3.5396,
+      "step": 6382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6131024360656738,
+      "learning_rate": 0.0005840471645237454,
+      "loss": 3.4931,
+      "step": 6383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49073895812034607,
+      "learning_rate": 0.0005840422103901425,
+      "loss": 3.3222,
+      "step": 6384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5380456447601318,
+      "learning_rate": 0.000584037255508427,
+      "loss": 3.4926,
+      "step": 6385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.500093400478363,
+      "learning_rate": 0.000584032299878612,
+      "loss": 3.5383,
+      "step": 6386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5496594905853271,
+      "learning_rate": 0.0005840273435007106,
+      "loss": 3.4764,
+      "step": 6387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5659000277519226,
+      "learning_rate": 0.0005840223863747359,
+      "loss": 3.4011,
+      "step": 6388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5337103605270386,
+      "learning_rate": 0.0005840174285007009,
+      "loss": 3.3357,
+      "step": 6389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5471076965332031,
+      "learning_rate": 0.0005840124698786186,
+      "loss": 3.3584,
+      "step": 6390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5406061410903931,
+      "learning_rate": 0.0005840075105085021,
+      "loss": 3.5359,
+      "step": 6391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5463137626647949,
+      "learning_rate": 0.0005840025503903645,
+      "loss": 3.29,
+      "step": 6392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5283336043357849,
+      "learning_rate": 0.0005839975895242189,
+      "loss": 3.6057,
+      "step": 6393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5675161480903625,
+      "learning_rate": 0.0005839926279100783,
+      "loss": 3.2287,
+      "step": 6394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5471466183662415,
+      "learning_rate": 0.0005839876655479557,
+      "loss": 3.4739,
+      "step": 6395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5402804017066956,
+      "learning_rate": 0.0005839827024378643,
+      "loss": 3.3865,
+      "step": 6396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5770726799964905,
+      "learning_rate": 0.000583977738579817,
+      "loss": 3.3723,
+      "step": 6397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5275018811225891,
+      "learning_rate": 0.0005839727739738271,
+      "loss": 3.3587,
+      "step": 6398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5402635931968689,
+      "learning_rate": 0.0005839678086199076,
+      "loss": 3.6001,
+      "step": 6399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6085510849952698,
+      "learning_rate": 0.0005839628425180714,
+      "loss": 3.303,
+      "step": 6400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.626964271068573,
+      "learning_rate": 0.0005839578756683318,
+      "loss": 3.3263,
+      "step": 6401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5337768197059631,
+      "learning_rate": 0.0005839529080707019,
+      "loss": 3.4304,
+      "step": 6402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5228325128555298,
+      "learning_rate": 0.0005839479397251946,
+      "loss": 3.2449,
+      "step": 6403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.565669596195221,
+      "learning_rate": 0.000583942970631823,
+      "loss": 3.3662,
+      "step": 6404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5665708780288696,
+      "learning_rate": 0.0005839380007906003,
+      "loss": 3.339,
+      "step": 6405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.530850350856781,
+      "learning_rate": 0.0005839330302015396,
+      "loss": 3.4299,
+      "step": 6406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5689116716384888,
+      "learning_rate": 0.000583928058864654,
+      "loss": 3.4183,
+      "step": 6407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5822932720184326,
+      "learning_rate": 0.0005839230867799565,
+      "loss": 3.3617,
+      "step": 6408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5526020526885986,
+      "learning_rate": 0.0005839181139474601,
+      "loss": 3.3813,
+      "step": 6409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5911718606948853,
+      "learning_rate": 0.0005839131403671782,
+      "loss": 3.3916,
+      "step": 6410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.501130223274231,
+      "learning_rate": 0.0005839081660391236,
+      "loss": 3.2104,
+      "step": 6411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5649345517158508,
+      "learning_rate": 0.0005839031909633096,
+      "loss": 3.3394,
+      "step": 6412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5334262847900391,
+      "learning_rate": 0.0005838982151397492,
+      "loss": 3.2688,
+      "step": 6413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5680837035179138,
+      "learning_rate": 0.0005838932385684556,
+      "loss": 3.5027,
+      "step": 6414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5568577647209167,
+      "learning_rate": 0.0005838882612494417,
+      "loss": 3.3118,
+      "step": 6415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5233654379844666,
+      "learning_rate": 0.0005838832831827209,
+      "loss": 3.3249,
+      "step": 6416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5565677881240845,
+      "learning_rate": 0.0005838783043683062,
+      "loss": 3.257,
+      "step": 6417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.588458776473999,
+      "learning_rate": 0.0005838733248062105,
+      "loss": 3.3794,
+      "step": 6418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.543366551399231,
+      "learning_rate": 0.0005838683444964473,
+      "loss": 3.4565,
+      "step": 6419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.508150041103363,
+      "learning_rate": 0.0005838633634390293,
+      "loss": 3.2742,
+      "step": 6420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5089104175567627,
+      "learning_rate": 0.00058385838163397,
+      "loss": 3.1958,
+      "step": 6421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5183136463165283,
+      "learning_rate": 0.0005838533990812822,
+      "loss": 3.4799,
+      "step": 6422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5491008758544922,
+      "learning_rate": 0.0005838484157809794,
+      "loss": 3.281,
+      "step": 6423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5273602604866028,
+      "learning_rate": 0.0005838434317330743,
+      "loss": 3.5572,
+      "step": 6424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5344679951667786,
+      "learning_rate": 0.0005838384469375803,
+      "loss": 3.3987,
+      "step": 6425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5330580472946167,
+      "learning_rate": 0.0005838334613945105,
+      "loss": 3.1804,
+      "step": 6426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5547220706939697,
+      "learning_rate": 0.0005838284751038779,
+      "loss": 3.6428,
+      "step": 6427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5338019728660583,
+      "learning_rate": 0.0005838234880656957,
+      "loss": 3.3878,
+      "step": 6428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54743891954422,
+      "learning_rate": 0.0005838185002799771,
+      "loss": 3.4011,
+      "step": 6429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5494392514228821,
+      "learning_rate": 0.0005838135117467352,
+      "loss": 3.3812,
+      "step": 6430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6107029318809509,
+      "learning_rate": 0.0005838085224659832,
+      "loss": 3.3056,
+      "step": 6431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5496015548706055,
+      "learning_rate": 0.0005838035324377341,
+      "loss": 3.3015,
+      "step": 6432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5744660496711731,
+      "learning_rate": 0.0005837985416620011,
+      "loss": 3.6286,
+      "step": 6433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5377488732337952,
+      "learning_rate": 0.0005837935501387975,
+      "loss": 3.4111,
+      "step": 6434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5246378779411316,
+      "learning_rate": 0.0005837885578681361,
+      "loss": 3.4557,
+      "step": 6435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5327603220939636,
+      "learning_rate": 0.0005837835648500304,
+      "loss": 3.3453,
+      "step": 6436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5735540986061096,
+      "learning_rate": 0.0005837785710844934,
+      "loss": 3.4068,
+      "step": 6437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5326542854309082,
+      "learning_rate": 0.0005837735765715381,
+      "loss": 3.5506,
+      "step": 6438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6021717190742493,
+      "learning_rate": 0.000583768581311178,
+      "loss": 3.4768,
+      "step": 6439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5350591540336609,
+      "learning_rate": 0.0005837635853034259,
+      "loss": 3.3959,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5394638180732727,
+      "learning_rate": 0.0005837585885482953,
+      "loss": 3.4337,
+      "step": 6441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5362926125526428,
+      "learning_rate": 0.0005837535910457991,
+      "loss": 3.3943,
+      "step": 6442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5286582112312317,
+      "learning_rate": 0.0005837485927959505,
+      "loss": 3.4177,
+      "step": 6443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5231122970581055,
+      "learning_rate": 0.0005837435937987628,
+      "loss": 3.5713,
+      "step": 6444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.541363537311554,
+      "learning_rate": 0.0005837385940542491,
+      "loss": 3.2653,
+      "step": 6445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5405844449996948,
+      "learning_rate": 0.0005837335935624225,
+      "loss": 3.4653,
+      "step": 6446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5295756459236145,
+      "learning_rate": 0.0005837285923232963,
+      "loss": 3.3157,
+      "step": 6447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5377822518348694,
+      "learning_rate": 0.0005837235903368834,
+      "loss": 3.3217,
+      "step": 6448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5518754720687866,
+      "learning_rate": 0.0005837185876031973,
+      "loss": 3.2826,
+      "step": 6449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5562870502471924,
+      "learning_rate": 0.000583713584122251,
+      "loss": 3.1871,
+      "step": 6450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5364195704460144,
+      "learning_rate": 0.0005837085798940577,
+      "loss": 3.4522,
+      "step": 6451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5666096210479736,
+      "learning_rate": 0.0005837035749186307,
+      "loss": 3.5498,
+      "step": 6452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5489985942840576,
+      "learning_rate": 0.000583698569195983,
+      "loss": 3.2855,
+      "step": 6453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5230780839920044,
+      "learning_rate": 0.0005836935627261279,
+      "loss": 3.4296,
+      "step": 6454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5031773447990417,
+      "learning_rate": 0.0005836885555090786,
+      "loss": 3.455,
+      "step": 6455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5535476207733154,
+      "learning_rate": 0.0005836835475448482,
+      "loss": 3.2767,
+      "step": 6456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5760499835014343,
+      "learning_rate": 0.0005836785388334499,
+      "loss": 3.5469,
+      "step": 6457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5332911610603333,
+      "learning_rate": 0.0005836735293748969,
+      "loss": 3.4841,
+      "step": 6458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5543957352638245,
+      "learning_rate": 0.0005836685191692026,
+      "loss": 3.3279,
+      "step": 6459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.576471209526062,
+      "learning_rate": 0.0005836635082163798,
+      "loss": 3.4943,
+      "step": 6460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5583447813987732,
+      "learning_rate": 0.000583658496516442,
+      "loss": 3.3526,
+      "step": 6461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5250188112258911,
+      "learning_rate": 0.0005836534840694024,
+      "loss": 3.4248,
+      "step": 6462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5952808856964111,
+      "learning_rate": 0.0005836484708752739,
+      "loss": 3.368,
+      "step": 6463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5756708979606628,
+      "learning_rate": 0.0005836434569340701,
+      "loss": 3.3606,
+      "step": 6464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5491698384284973,
+      "learning_rate": 0.000583638442245804,
+      "loss": 3.4689,
+      "step": 6465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5349763631820679,
+      "learning_rate": 0.0005836334268104887,
+      "loss": 3.4542,
+      "step": 6466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5523869395256042,
+      "learning_rate": 0.0005836284106281377,
+      "loss": 3.3126,
+      "step": 6467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5572420358657837,
+      "learning_rate": 0.0005836233936987639,
+      "loss": 3.4631,
+      "step": 6468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5321294665336609,
+      "learning_rate": 0.0005836183760223808,
+      "loss": 3.3318,
+      "step": 6469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5739319324493408,
+      "learning_rate": 0.0005836133575990015,
+      "loss": 3.5455,
+      "step": 6470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5573357939720154,
+      "learning_rate": 0.0005836083384286391,
+      "loss": 3.3142,
+      "step": 6471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6728717684745789,
+      "learning_rate": 0.0005836033185113069,
+      "loss": 3.4165,
+      "step": 6472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5487765073776245,
+      "learning_rate": 0.0005835982978470182,
+      "loss": 3.3554,
+      "step": 6473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5465099215507507,
+      "learning_rate": 0.0005835932764357861,
+      "loss": 3.2763,
+      "step": 6474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5527744293212891,
+      "learning_rate": 0.0005835882542776239,
+      "loss": 3.2369,
+      "step": 6475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6429268717765808,
+      "learning_rate": 0.0005835832313725449,
+      "loss": 3.3702,
+      "step": 6476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5672469735145569,
+      "learning_rate": 0.0005835782077205623,
+      "loss": 3.4388,
+      "step": 6477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5245232582092285,
+      "learning_rate": 0.0005835731833216891,
+      "loss": 3.5464,
+      "step": 6478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5858415961265564,
+      "learning_rate": 0.0005835681581759387,
+      "loss": 3.3392,
+      "step": 6479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5938719511032104,
+      "learning_rate": 0.0005835631322833244,
+      "loss": 3.3036,
+      "step": 6480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5960376262664795,
+      "learning_rate": 0.0005835581056438595,
+      "loss": 3.3089,
+      "step": 6481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5122845768928528,
+      "learning_rate": 0.000583553078257557,
+      "loss": 3.4641,
+      "step": 6482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5184789299964905,
+      "learning_rate": 0.0005835480501244302,
+      "loss": 3.2639,
+      "step": 6483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5567031502723694,
+      "learning_rate": 0.0005835430212444927,
+      "loss": 3.6092,
+      "step": 6484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6219637393951416,
+      "learning_rate": 0.0005835379916177572,
+      "loss": 3.4196,
+      "step": 6485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5328686833381653,
+      "learning_rate": 0.0005835329612442372,
+      "loss": 3.523,
+      "step": 6486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5400885939598083,
+      "learning_rate": 0.000583527930123946,
+      "loss": 3.0661,
+      "step": 6487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5412009954452515,
+      "learning_rate": 0.0005835228982568968,
+      "loss": 3.2229,
+      "step": 6488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5066078901290894,
+      "learning_rate": 0.0005835178656431029,
+      "loss": 3.2135,
+      "step": 6489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6025939583778381,
+      "learning_rate": 0.0005835128322825774,
+      "loss": 3.5742,
+      "step": 6490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5483773350715637,
+      "learning_rate": 0.0005835077981753337,
+      "loss": 3.4047,
+      "step": 6491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5667142868041992,
+      "learning_rate": 0.0005835027633213851,
+      "loss": 3.3673,
+      "step": 6492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5416460633277893,
+      "learning_rate": 0.0005834977277207447,
+      "loss": 3.5253,
+      "step": 6493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5411655902862549,
+      "learning_rate": 0.0005834926913734259,
+      "loss": 3.2967,
+      "step": 6494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5534458756446838,
+      "learning_rate": 0.0005834876542794418,
+      "loss": 3.2466,
+      "step": 6495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5487003922462463,
+      "learning_rate": 0.0005834826164388059,
+      "loss": 3.4169,
+      "step": 6496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5103386044502258,
+      "learning_rate": 0.0005834775778515313,
+      "loss": 3.4592,
+      "step": 6497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5426819324493408,
+      "learning_rate": 0.0005834725385176312,
+      "loss": 3.4416,
+      "step": 6498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5770543217658997,
+      "learning_rate": 0.0005834674984371191,
+      "loss": 3.3999,
+      "step": 6499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5331725478172302,
+      "learning_rate": 0.0005834624576100082,
+      "loss": 3.4512,
+      "step": 6500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5609892010688782,
+      "learning_rate": 0.0005834574160363117,
+      "loss": 3.461,
+      "step": 6501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5909136533737183,
+      "learning_rate": 0.000583452373716043,
+      "loss": 3.3132,
+      "step": 6502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55914705991745,
+      "learning_rate": 0.0005834473306492152,
+      "loss": 3.5681,
+      "step": 6503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5517709255218506,
+      "learning_rate": 0.0005834422868358416,
+      "loss": 3.3071,
+      "step": 6504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5129667520523071,
+      "learning_rate": 0.0005834372422759358,
+      "loss": 3.2362,
+      "step": 6505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5354728102684021,
+      "learning_rate": 0.0005834321969695107,
+      "loss": 3.3641,
+      "step": 6506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5102858543395996,
+      "learning_rate": 0.0005834271509165798,
+      "loss": 3.1551,
+      "step": 6507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5487684011459351,
+      "learning_rate": 0.0005834221041171563,
+      "loss": 3.5077,
+      "step": 6508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5984534025192261,
+      "learning_rate": 0.0005834170565712535,
+      "loss": 3.3783,
+      "step": 6509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5358244180679321,
+      "learning_rate": 0.0005834120082788847,
+      "loss": 3.3893,
+      "step": 6510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5399655103683472,
+      "learning_rate": 0.0005834069592400632,
+      "loss": 3.3156,
+      "step": 6511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5290285348892212,
+      "learning_rate": 0.0005834019094548025,
+      "loss": 3.1687,
+      "step": 6512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5532265901565552,
+      "learning_rate": 0.0005833968589231155,
+      "loss": 3.4857,
+      "step": 6513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5471756458282471,
+      "learning_rate": 0.0005833918076450158,
+      "loss": 3.2192,
+      "step": 6514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5379656553268433,
+      "learning_rate": 0.0005833867556205165,
+      "loss": 3.4739,
+      "step": 6515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5359077453613281,
+      "learning_rate": 0.0005833817028496313,
+      "loss": 3.3655,
+      "step": 6516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5260913968086243,
+      "learning_rate": 0.000583376649332373,
+      "loss": 3.315,
+      "step": 6517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5894081592559814,
+      "learning_rate": 0.0005833715950687552,
+      "loss": 3.1694,
+      "step": 6518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5539742708206177,
+      "learning_rate": 0.0005833665400587911,
+      "loss": 3.3007,
+      "step": 6519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.506983757019043,
+      "learning_rate": 0.0005833614843024942,
+      "loss": 3.3166,
+      "step": 6520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5148934721946716,
+      "learning_rate": 0.0005833564277998776,
+      "loss": 3.3476,
+      "step": 6521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5183430910110474,
+      "learning_rate": 0.0005833513705509547,
+      "loss": 3.5182,
+      "step": 6522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5564296841621399,
+      "learning_rate": 0.0005833463125557389,
+      "loss": 3.2431,
+      "step": 6523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5783804655075073,
+      "learning_rate": 0.0005833412538142433,
+      "loss": 3.3268,
+      "step": 6524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5203677415847778,
+      "learning_rate": 0.0005833361943264815,
+      "loss": 3.3579,
+      "step": 6525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5871049165725708,
+      "learning_rate": 0.0005833311340924666,
+      "loss": 3.3097,
+      "step": 6526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.556999146938324,
+      "learning_rate": 0.0005833260731122121,
+      "loss": 3.254,
+      "step": 6527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5999124646186829,
+      "learning_rate": 0.0005833210113857313,
+      "loss": 3.439,
+      "step": 6528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5835153460502625,
+      "learning_rate": 0.0005833159489130373,
+      "loss": 3.4526,
+      "step": 6529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.521670401096344,
+      "learning_rate": 0.0005833108856941438,
+      "loss": 3.4603,
+      "step": 6530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5895057916641235,
+      "learning_rate": 0.0005833058217290638,
+      "loss": 3.4598,
+      "step": 6531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5201472043991089,
+      "learning_rate": 0.0005833007570178109,
+      "loss": 3.5078,
+      "step": 6532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5210258364677429,
+      "learning_rate": 0.0005832956915603982,
+      "loss": 3.3483,
+      "step": 6533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5231567621231079,
+      "learning_rate": 0.0005832906253568392,
+      "loss": 3.4721,
+      "step": 6534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5170723795890808,
+      "learning_rate": 0.0005832855584071474,
+      "loss": 3.3227,
+      "step": 6535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5135728716850281,
+      "learning_rate": 0.0005832804907113358,
+      "loss": 3.1659,
+      "step": 6536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5827286243438721,
+      "learning_rate": 0.0005832754222694179,
+      "loss": 3.3854,
+      "step": 6537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5649005174636841,
+      "learning_rate": 0.000583270353081407,
+      "loss": 3.2509,
+      "step": 6538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.585883378982544,
+      "learning_rate": 0.0005832652831473166,
+      "loss": 3.3499,
+      "step": 6539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5253201723098755,
+      "learning_rate": 0.00058326021246716,
+      "loss": 3.4087,
+      "step": 6540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4951026141643524,
+      "learning_rate": 0.0005832551410409505,
+      "loss": 3.4144,
+      "step": 6541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5574667453765869,
+      "learning_rate": 0.0005832500688687014,
+      "loss": 3.2957,
+      "step": 6542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5781322121620178,
+      "learning_rate": 0.0005832449959504262,
+      "loss": 3.3506,
+      "step": 6543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6145309209823608,
+      "learning_rate": 0.0005832399222861381,
+      "loss": 3.4637,
+      "step": 6544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5183820724487305,
+      "learning_rate": 0.0005832348478758507,
+      "loss": 3.5608,
+      "step": 6545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703016519546509,
+      "learning_rate": 0.0005832297727195771,
+      "loss": 3.3926,
+      "step": 6546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5798136591911316,
+      "learning_rate": 0.0005832246968173309,
+      "loss": 3.3352,
+      "step": 6547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5665884017944336,
+      "learning_rate": 0.0005832196201691252,
+      "loss": 3.2569,
+      "step": 6548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5170712471008301,
+      "learning_rate": 0.0005832145427749737,
+      "loss": 3.4392,
+      "step": 6549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5397904515266418,
+      "learning_rate": 0.0005832094646348894,
+      "loss": 3.5144,
+      "step": 6550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6113349795341492,
+      "learning_rate": 0.000583204385748886,
+      "loss": 3.3113,
+      "step": 6551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5789533853530884,
+      "learning_rate": 0.0005831993061169768,
+      "loss": 3.2817,
+      "step": 6552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5456516742706299,
+      "learning_rate": 0.000583194225739175,
+      "loss": 3.327,
+      "step": 6553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5421919822692871,
+      "learning_rate": 0.0005831891446154942,
+      "loss": 3.2764,
+      "step": 6554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5785422325134277,
+      "learning_rate": 0.0005831840627459476,
+      "loss": 3.0956,
+      "step": 6555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5162643194198608,
+      "learning_rate": 0.0005831789801305488,
+      "loss": 3.4063,
+      "step": 6556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6994883418083191,
+      "learning_rate": 0.0005831738967693109,
+      "loss": 3.3117,
+      "step": 6557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5696769952774048,
+      "learning_rate": 0.0005831688126622477,
+      "loss": 3.2208,
+      "step": 6558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5177335739135742,
+      "learning_rate": 0.0005831637278093722,
+      "loss": 3.2212,
+      "step": 6559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5111486315727234,
+      "learning_rate": 0.000583158642210698,
+      "loss": 3.4638,
+      "step": 6560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5310366153717041,
+      "learning_rate": 0.0005831535558662383,
+      "loss": 3.2656,
+      "step": 6561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5311713218688965,
+      "learning_rate": 0.0005831484687760067,
+      "loss": 3.6579,
+      "step": 6562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6105450987815857,
+      "learning_rate": 0.0005831433809400166,
+      "loss": 3.3503,
+      "step": 6563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5762481689453125,
+      "learning_rate": 0.0005831382923582812,
+      "loss": 3.4268,
+      "step": 6564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5664306879043579,
+      "learning_rate": 0.0005831332030308142,
+      "loss": 3.2995,
+      "step": 6565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5523332953453064,
+      "learning_rate": 0.0005831281129576286,
+      "loss": 3.0233,
+      "step": 6566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5818634033203125,
+      "learning_rate": 0.0005831230221387382,
+      "loss": 3.4713,
+      "step": 6567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5426694750785828,
+      "learning_rate": 0.0005831179305741562,
+      "loss": 3.3265,
+      "step": 6568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5816428661346436,
+      "learning_rate": 0.000583112838263896,
+      "loss": 3.1057,
+      "step": 6569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5406796932220459,
+      "learning_rate": 0.0005831077452079712,
+      "loss": 2.973,
+      "step": 6570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5280852317810059,
+      "learning_rate": 0.000583102651406395,
+      "loss": 3.3631,
+      "step": 6571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5578991174697876,
+      "learning_rate": 0.0005830975568591809,
+      "loss": 3.3991,
+      "step": 6572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.585803210735321,
+      "learning_rate": 0.0005830924615663423,
+      "loss": 3.4668,
+      "step": 6573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5431290864944458,
+      "learning_rate": 0.0005830873655278927,
+      "loss": 3.1047,
+      "step": 6574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5530149340629578,
+      "learning_rate": 0.0005830822687438455,
+      "loss": 3.4276,
+      "step": 6575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5447601675987244,
+      "learning_rate": 0.000583077171214214,
+      "loss": 3.2974,
+      "step": 6576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5636044144630432,
+      "learning_rate": 0.0005830720729390118,
+      "loss": 3.1732,
+      "step": 6577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5744757056236267,
+      "learning_rate": 0.0005830669739182522,
+      "loss": 3.3084,
+      "step": 6578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5662171840667725,
+      "learning_rate": 0.0005830618741519486,
+      "loss": 3.5397,
+      "step": 6579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5077491998672485,
+      "learning_rate": 0.0005830567736401145,
+      "loss": 2.9837,
+      "step": 6580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5053831338882446,
+      "learning_rate": 0.0005830516723827634,
+      "loss": 3.2128,
+      "step": 6581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5454685688018799,
+      "learning_rate": 0.0005830465703799086,
+      "loss": 3.5079,
+      "step": 6582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5580922961235046,
+      "learning_rate": 0.0005830414676315636,
+      "loss": 3.4775,
+      "step": 6583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5592604279518127,
+      "learning_rate": 0.000583036364137742,
+      "loss": 3.2971,
+      "step": 6584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5692855715751648,
+      "learning_rate": 0.0005830312598984569,
+      "loss": 3.1209,
+      "step": 6585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5731725692749023,
+      "learning_rate": 0.0005830261549137221,
+      "loss": 3.2845,
+      "step": 6586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5284132957458496,
+      "learning_rate": 0.0005830210491835508,
+      "loss": 3.438,
+      "step": 6587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5840345621109009,
+      "learning_rate": 0.0005830159427079565,
+      "loss": 3.4824,
+      "step": 6588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5384155511856079,
+      "learning_rate": 0.0005830108354869528,
+      "loss": 3.466,
+      "step": 6589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5243905782699585,
+      "learning_rate": 0.0005830057275205529,
+      "loss": 3.3626,
+      "step": 6590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4902263581752777,
+      "learning_rate": 0.0005830006188087704,
+      "loss": 3.22,
+      "step": 6591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5410398840904236,
+      "learning_rate": 0.0005829955093516187,
+      "loss": 3.317,
+      "step": 6592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5666494965553284,
+      "learning_rate": 0.0005829903991491113,
+      "loss": 3.5118,
+      "step": 6593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5500527024269104,
+      "learning_rate": 0.0005829852882012616,
+      "loss": 3.1609,
+      "step": 6594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5293498039245605,
+      "learning_rate": 0.0005829801765080833,
+      "loss": 3.5884,
+      "step": 6595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5184808969497681,
+      "learning_rate": 0.0005829750640695896,
+      "loss": 3.4994,
+      "step": 6596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5679014921188354,
+      "learning_rate": 0.000582969950885794,
+      "loss": 3.182,
+      "step": 6597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5611880421638489,
+      "learning_rate": 0.0005829648369567099,
+      "loss": 3.3356,
+      "step": 6598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5921120643615723,
+      "learning_rate": 0.0005829597222823511,
+      "loss": 3.2854,
+      "step": 6599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62581467628479,
+      "learning_rate": 0.0005829546068627306,
+      "loss": 3.3484,
+      "step": 6600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.554572343826294,
+      "learning_rate": 0.0005829494906978624,
+      "loss": 3.4792,
+      "step": 6601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.535103976726532,
+      "learning_rate": 0.0005829443737877595,
+      "loss": 3.5559,
+      "step": 6602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5401231050491333,
+      "learning_rate": 0.0005829392561324357,
+      "loss": 3.3311,
+      "step": 6603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5672370791435242,
+      "learning_rate": 0.0005829341377319042,
+      "loss": 3.452,
+      "step": 6604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5229573845863342,
+      "learning_rate": 0.0005829290185861788,
+      "loss": 3.1934,
+      "step": 6605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5550198554992676,
+      "learning_rate": 0.0005829238986952728,
+      "loss": 3.2653,
+      "step": 6606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4899026155471802,
+      "learning_rate": 0.0005829187780591995,
+      "loss": 3.5498,
+      "step": 6607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5529863238334656,
+      "learning_rate": 0.0005829136566779727,
+      "loss": 3.4041,
+      "step": 6608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5543391704559326,
+      "learning_rate": 0.0005829085345516058,
+      "loss": 3.5464,
+      "step": 6609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5163125991821289,
+      "learning_rate": 0.0005829034116801123,
+      "loss": 3.556,
+      "step": 6610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5830318331718445,
+      "learning_rate": 0.0005828982880635057,
+      "loss": 3.4059,
+      "step": 6611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5730868577957153,
+      "learning_rate": 0.0005828931637017992,
+      "loss": 3.2501,
+      "step": 6612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5114001631736755,
+      "learning_rate": 0.0005828880385950068,
+      "loss": 3.3564,
+      "step": 6613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5583293437957764,
+      "learning_rate": 0.0005828829127431418,
+      "loss": 3.1739,
+      "step": 6614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5196751952171326,
+      "learning_rate": 0.0005828777861462174,
+      "loss": 3.4315,
+      "step": 6615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5976796746253967,
+      "learning_rate": 0.0005828726588042475,
+      "loss": 3.3738,
+      "step": 6616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5232165455818176,
+      "learning_rate": 0.0005828675307172455,
+      "loss": 3.3511,
+      "step": 6617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5699440240859985,
+      "learning_rate": 0.0005828624018852247,
+      "loss": 3.4635,
+      "step": 6618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5138837099075317,
+      "learning_rate": 0.0005828572723081989,
+      "loss": 3.33,
+      "step": 6619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5367459654808044,
+      "learning_rate": 0.0005828521419861814,
+      "loss": 3.3618,
+      "step": 6620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5628336668014526,
+      "learning_rate": 0.0005828470109191859,
+      "loss": 3.4253,
+      "step": 6621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5682950615882874,
+      "learning_rate": 0.0005828418791072258,
+      "loss": 3.3819,
+      "step": 6622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5145696401596069,
+      "learning_rate": 0.0005828367465503145,
+      "loss": 3.409,
+      "step": 6623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5609208345413208,
+      "learning_rate": 0.0005828316132484657,
+      "loss": 3.471,
+      "step": 6624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.514314591884613,
+      "learning_rate": 0.0005828264792016929,
+      "loss": 3.4783,
+      "step": 6625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5764790773391724,
+      "learning_rate": 0.0005828213444100096,
+      "loss": 3.5846,
+      "step": 6626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5133751630783081,
+      "learning_rate": 0.0005828162088734293,
+      "loss": 3.4859,
+      "step": 6627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5235024094581604,
+      "learning_rate": 0.0005828110725919655,
+      "loss": 3.3398,
+      "step": 6628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5416542291641235,
+      "learning_rate": 0.0005828059355656317,
+      "loss": 3.279,
+      "step": 6629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5646631121635437,
+      "learning_rate": 0.0005828007977944416,
+      "loss": 3.3315,
+      "step": 6630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5523773431777954,
+      "learning_rate": 0.0005827956592784087,
+      "loss": 3.3439,
+      "step": 6631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5373170375823975,
+      "learning_rate": 0.0005827905200175463,
+      "loss": 3.3712,
+      "step": 6632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5419214367866516,
+      "learning_rate": 0.0005827853800118682,
+      "loss": 3.3848,
+      "step": 6633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5145456194877625,
+      "learning_rate": 0.0005827802392613877,
+      "loss": 3.4817,
+      "step": 6634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5141541957855225,
+      "learning_rate": 0.0005827750977661184,
+      "loss": 3.3013,
+      "step": 6635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.519413948059082,
+      "learning_rate": 0.0005827699555260742,
+      "loss": 3.2296,
+      "step": 6636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5240925550460815,
+      "learning_rate": 0.0005827648125412683,
+      "loss": 3.3045,
+      "step": 6637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5714395046234131,
+      "learning_rate": 0.0005827596688117141,
+      "loss": 3.1928,
+      "step": 6638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5431802868843079,
+      "learning_rate": 0.0005827545243374256,
+      "loss": 3.2881,
+      "step": 6639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.542210578918457,
+      "learning_rate": 0.0005827493791184159,
+      "loss": 3.4252,
+      "step": 6640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5310395956039429,
+      "learning_rate": 0.0005827442331546987,
+      "loss": 3.4693,
+      "step": 6641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5362105369567871,
+      "learning_rate": 0.0005827390864462878,
+      "loss": 3.5127,
+      "step": 6642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5233688354492188,
+      "learning_rate": 0.0005827339389931966,
+      "loss": 3.5151,
+      "step": 6643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5598340630531311,
+      "learning_rate": 0.0005827287907954385,
+      "loss": 3.3003,
+      "step": 6644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5226010680198669,
+      "learning_rate": 0.0005827236418530272,
+      "loss": 3.3553,
+      "step": 6645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6050033569335938,
+      "learning_rate": 0.0005827184921659761,
+      "loss": 3.1462,
+      "step": 6646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5034282207489014,
+      "learning_rate": 0.0005827133417342991,
+      "loss": 3.5462,
+      "step": 6647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5270561575889587,
+      "learning_rate": 0.0005827081905580095,
+      "loss": 3.5129,
+      "step": 6648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5257344245910645,
+      "learning_rate": 0.0005827030386371209,
+      "loss": 3.3749,
+      "step": 6649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6640146374702454,
+      "learning_rate": 0.0005826978859716469,
+      "loss": 3.0764,
+      "step": 6650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5421690940856934,
+      "learning_rate": 0.0005826927325616012,
+      "loss": 3.2901,
+      "step": 6651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5615790486335754,
+      "learning_rate": 0.0005826875784069971,
+      "loss": 3.3495,
+      "step": 6652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5812866687774658,
+      "learning_rate": 0.0005826824235078484,
+      "loss": 3.6013,
+      "step": 6653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5673434138298035,
+      "learning_rate": 0.0005826772678641685,
+      "loss": 3.313,
+      "step": 6654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5085522532463074,
+      "learning_rate": 0.0005826721114759711,
+      "loss": 3.3503,
+      "step": 6655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5811455249786377,
+      "learning_rate": 0.0005826669543432699,
+      "loss": 3.5832,
+      "step": 6656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.584208607673645,
+      "learning_rate": 0.0005826617964660783,
+      "loss": 3.2827,
+      "step": 6657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.547170877456665,
+      "learning_rate": 0.0005826566378444099,
+      "loss": 3.2459,
+      "step": 6658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5642180442810059,
+      "learning_rate": 0.0005826514784782783,
+      "loss": 3.3707,
+      "step": 6659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5796110033988953,
+      "learning_rate": 0.000582646318367697,
+      "loss": 3.4914,
+      "step": 6660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5225422978401184,
+      "learning_rate": 0.0005826411575126798,
+      "loss": 3.3454,
+      "step": 6661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5265496969223022,
+      "learning_rate": 0.0005826359959132402,
+      "loss": 3.2391,
+      "step": 6662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48403891921043396,
+      "learning_rate": 0.0005826308335693919,
+      "loss": 3.2319,
+      "step": 6663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5459921360015869,
+      "learning_rate": 0.0005826256704811481,
+      "loss": 3.488,
+      "step": 6664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.524363100528717,
+      "learning_rate": 0.0005826205066485229,
+      "loss": 3.5475,
+      "step": 6665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5568280816078186,
+      "learning_rate": 0.0005826153420715295,
+      "loss": 3.3387,
+      "step": 6666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5115675330162048,
+      "learning_rate": 0.0005826101767501818,
+      "loss": 3.5857,
+      "step": 6667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5281095504760742,
+      "learning_rate": 0.0005826050106844934,
+      "loss": 3.3025,
+      "step": 6668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5385881066322327,
+      "learning_rate": 0.0005825998438744775,
+      "loss": 3.244,
+      "step": 6669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5694405436515808,
+      "learning_rate": 0.0005825946763201482,
+      "loss": 3.2063,
+      "step": 6670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5485799908638,
+      "learning_rate": 0.0005825895080215188,
+      "loss": 3.3736,
+      "step": 6671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5174115300178528,
+      "learning_rate": 0.000582584338978603,
+      "loss": 3.3348,
+      "step": 6672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5306274890899658,
+      "learning_rate": 0.0005825791691914146,
+      "loss": 3.4455,
+      "step": 6673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5434350371360779,
+      "learning_rate": 0.000582573998659967,
+      "loss": 3.4376,
+      "step": 6674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.570429801940918,
+      "learning_rate": 0.0005825688273842738,
+      "loss": 3.3823,
+      "step": 6675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5130692720413208,
+      "learning_rate": 0.0005825636553643488,
+      "loss": 3.2729,
+      "step": 6676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5442194938659668,
+      "learning_rate": 0.0005825584826002054,
+      "loss": 3.547,
+      "step": 6677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6229523420333862,
+      "learning_rate": 0.0005825533090918574,
+      "loss": 3.3318,
+      "step": 6678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5505388975143433,
+      "learning_rate": 0.0005825481348393183,
+      "loss": 3.4383,
+      "step": 6679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5161044597625732,
+      "learning_rate": 0.0005825429598426018,
+      "loss": 3.3366,
+      "step": 6680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5270329117774963,
+      "learning_rate": 0.0005825377841017215,
+      "loss": 3.296,
+      "step": 6681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5343664884567261,
+      "learning_rate": 0.0005825326076166912,
+      "loss": 3.3817,
+      "step": 6682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5493436455726624,
+      "learning_rate": 0.0005825274303875242,
+      "loss": 3.3233,
+      "step": 6683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5526965260505676,
+      "learning_rate": 0.0005825222524142345,
+      "loss": 3.1901,
+      "step": 6684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5074105858802795,
+      "learning_rate": 0.0005825170736968354,
+      "loss": 3.3946,
+      "step": 6685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5705122947692871,
+      "learning_rate": 0.0005825118942353408,
+      "loss": 3.2233,
+      "step": 6686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5070473551750183,
+      "learning_rate": 0.0005825067140297641,
+      "loss": 3.2129,
+      "step": 6687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5425854325294495,
+      "learning_rate": 0.0005825015330801192,
+      "loss": 3.2749,
+      "step": 6688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5736327767372131,
+      "learning_rate": 0.0005824963513864197,
+      "loss": 3.2822,
+      "step": 6689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5530392527580261,
+      "learning_rate": 0.0005824911689486791,
+      "loss": 3.2252,
+      "step": 6690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5579785704612732,
+      "learning_rate": 0.0005824859857669111,
+      "loss": 3.6146,
+      "step": 6691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.671022891998291,
+      "learning_rate": 0.0005824808018411294,
+      "loss": 3.1428,
+      "step": 6692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5243225693702698,
+      "learning_rate": 0.0005824756171713477,
+      "loss": 3.3311,
+      "step": 6693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6820229291915894,
+      "learning_rate": 0.0005824704317575795,
+      "loss": 3.2211,
+      "step": 6694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5242649912834167,
+      "learning_rate": 0.0005824652455998385,
+      "loss": 3.3226,
+      "step": 6695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5282347202301025,
+      "learning_rate": 0.0005824600586981386,
+      "loss": 3.4453,
+      "step": 6696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5148776173591614,
+      "learning_rate": 0.0005824548710524931,
+      "loss": 3.3774,
+      "step": 6697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5305414795875549,
+      "learning_rate": 0.0005824496826629159,
+      "loss": 3.282,
+      "step": 6698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0000241994857788,
+      "learning_rate": 0.0005824444935294206,
+      "loss": 3.4245,
+      "step": 6699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5258133411407471,
+      "learning_rate": 0.0005824393036520208,
+      "loss": 3.4044,
+      "step": 6700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6157311201095581,
+      "learning_rate": 0.0005824341130307302,
+      "loss": 3.3356,
+      "step": 6701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5688565969467163,
+      "learning_rate": 0.0005824289216655626,
+      "loss": 3.3015,
+      "step": 6702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5846591591835022,
+      "learning_rate": 0.0005824237295565315,
+      "loss": 3.5027,
+      "step": 6703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5892431139945984,
+      "learning_rate": 0.0005824185367036507,
+      "loss": 3.3725,
+      "step": 6704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5611383318901062,
+      "learning_rate": 0.0005824133431069338,
+      "loss": 3.0607,
+      "step": 6705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5581179857254028,
+      "learning_rate": 0.0005824081487663945,
+      "loss": 3.1806,
+      "step": 6706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5461733341217041,
+      "learning_rate": 0.0005824029536820466,
+      "loss": 3.1481,
+      "step": 6707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5615188479423523,
+      "learning_rate": 0.0005823977578539035,
+      "loss": 3.3157,
+      "step": 6708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5543760061264038,
+      "learning_rate": 0.0005823925612819792,
+      "loss": 3.1724,
+      "step": 6709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.512069046497345,
+      "learning_rate": 0.0005823873639662871,
+      "loss": 3.476,
+      "step": 6710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7344093918800354,
+      "learning_rate": 0.0005823821659068411,
+      "loss": 3.3028,
+      "step": 6711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5303009152412415,
+      "learning_rate": 0.0005823769671036549,
+      "loss": 3.3448,
+      "step": 6712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5423128008842468,
+      "learning_rate": 0.0005823717675567419,
+      "loss": 3.2607,
+      "step": 6713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5232293009757996,
+      "learning_rate": 0.0005823665672661161,
+      "loss": 3.2688,
+      "step": 6714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5623641610145569,
+      "learning_rate": 0.0005823613662317912,
+      "loss": 3.2962,
+      "step": 6715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6031582355499268,
+      "learning_rate": 0.0005823561644537807,
+      "loss": 3.3641,
+      "step": 6716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5149546265602112,
+      "learning_rate": 0.0005823509619320986,
+      "loss": 3.4208,
+      "step": 6717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5367235541343689,
+      "learning_rate": 0.0005823457586667582,
+      "loss": 3.2141,
+      "step": 6718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5604205131530762,
+      "learning_rate": 0.0005823405546577735,
+      "loss": 3.5344,
+      "step": 6719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5116593241691589,
+      "learning_rate": 0.000582335349905158,
+      "loss": 3.5164,
+      "step": 6720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5447813272476196,
+      "learning_rate": 0.0005823301444089256,
+      "loss": 3.4071,
+      "step": 6721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5833255052566528,
+      "learning_rate": 0.0005823249381690899,
+      "loss": 3.3173,
+      "step": 6722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5849522352218628,
+      "learning_rate": 0.0005823197311856647,
+      "loss": 3.1083,
+      "step": 6723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5260932445526123,
+      "learning_rate": 0.0005823145234586638,
+      "loss": 3.3465,
+      "step": 6724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5599755644798279,
+      "learning_rate": 0.0005823093149881005,
+      "loss": 3.3558,
+      "step": 6725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5124487280845642,
+      "learning_rate": 0.0005823041057739889,
+      "loss": 3.589,
+      "step": 6726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5424213409423828,
+      "learning_rate": 0.0005822988958163427,
+      "loss": 3.3946,
+      "step": 6727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5511559844017029,
+      "learning_rate": 0.0005822936851151755,
+      "loss": 3.3763,
+      "step": 6728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5458391904830933,
+      "learning_rate": 0.0005822884736705011,
+      "loss": 3.2826,
+      "step": 6729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5722623467445374,
+      "learning_rate": 0.000582283261482333,
+      "loss": 3.2023,
+      "step": 6730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5854646563529968,
+      "learning_rate": 0.0005822780485506854,
+      "loss": 3.2832,
+      "step": 6731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5965773463249207,
+      "learning_rate": 0.0005822728348755716,
+      "loss": 3.1899,
+      "step": 6732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5229557156562805,
+      "learning_rate": 0.0005822676204570054,
+      "loss": 3.3824,
+      "step": 6733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.527096688747406,
+      "learning_rate": 0.0005822624052950006,
+      "loss": 3.4582,
+      "step": 6734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5202319025993347,
+      "learning_rate": 0.0005822571893895711,
+      "loss": 3.4314,
+      "step": 6735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5747625827789307,
+      "learning_rate": 0.0005822519727407304,
+      "loss": 3.345,
+      "step": 6736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.498573362827301,
+      "learning_rate": 0.0005822467553484923,
+      "loss": 3.4014,
+      "step": 6737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5384262800216675,
+      "learning_rate": 0.0005822415372128706,
+      "loss": 3.3868,
+      "step": 6738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5200839042663574,
+      "learning_rate": 0.000582236318333879,
+      "loss": 3.3441,
+      "step": 6739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5176569819450378,
+      "learning_rate": 0.0005822310987115314,
+      "loss": 3.2837,
+      "step": 6740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6018693447113037,
+      "learning_rate": 0.0005822258783458412,
+      "loss": 3.2213,
+      "step": 6741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5896228551864624,
+      "learning_rate": 0.0005822206572368224,
+      "loss": 3.339,
+      "step": 6742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5164976119995117,
+      "learning_rate": 0.0005822154353844887,
+      "loss": 3.4402,
+      "step": 6743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5089805722236633,
+      "learning_rate": 0.0005822102127888539,
+      "loss": 3.1868,
+      "step": 6744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5495955348014832,
+      "learning_rate": 0.0005822049894499316,
+      "loss": 3.5313,
+      "step": 6745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5287194848060608,
+      "learning_rate": 0.0005821997653677359,
+      "loss": 3.1508,
+      "step": 6746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6048151254653931,
+      "learning_rate": 0.0005821945405422802,
+      "loss": 3.1749,
+      "step": 6747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5822990536689758,
+      "learning_rate": 0.0005821893149735784,
+      "loss": 3.3297,
+      "step": 6748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.518795907497406,
+      "learning_rate": 0.0005821840886616441,
+      "loss": 3.3508,
+      "step": 6749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.608176052570343,
+      "learning_rate": 0.0005821788616064914,
+      "loss": 3.2698,
+      "step": 6750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5752763748168945,
+      "learning_rate": 0.0005821736338081339,
+      "loss": 3.3454,
+      "step": 6751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5169047117233276,
+      "learning_rate": 0.0005821684052665851,
+      "loss": 3.609,
+      "step": 6752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.529773473739624,
+      "learning_rate": 0.0005821631759818593,
+      "loss": 3.369,
+      "step": 6753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.543318510055542,
+      "learning_rate": 0.00058215794595397,
+      "loss": 3.4376,
+      "step": 6754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6167493462562561,
+      "learning_rate": 0.0005821527151829308,
+      "loss": 3.2169,
+      "step": 6755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5135380625724792,
+      "learning_rate": 0.0005821474836687559,
+      "loss": 3.2988,
+      "step": 6756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.597224235534668,
+      "learning_rate": 0.0005821422514114585,
+      "loss": 3.3903,
+      "step": 6757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5847658514976501,
+      "learning_rate": 0.0005821370184110529,
+      "loss": 3.6174,
+      "step": 6758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5715068578720093,
+      "learning_rate": 0.0005821317846675527,
+      "loss": 3.3343,
+      "step": 6759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5231801271438599,
+      "learning_rate": 0.0005821265501809716,
+      "loss": 3.3706,
+      "step": 6760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5947988033294678,
+      "learning_rate": 0.0005821213149513235,
+      "loss": 3.2535,
+      "step": 6761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5078067779541016,
+      "learning_rate": 0.0005821160789786222,
+      "loss": 3.412,
+      "step": 6762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5282332897186279,
+      "learning_rate": 0.0005821108422628813,
+      "loss": 3.2963,
+      "step": 6763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5280579924583435,
+      "learning_rate": 0.0005821056048041149,
+      "loss": 3.4438,
+      "step": 6764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5364978909492493,
+      "learning_rate": 0.0005821003666023366,
+      "loss": 3.3856,
+      "step": 6765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5377125144004822,
+      "learning_rate": 0.0005820951276575601,
+      "loss": 3.4609,
+      "step": 6766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5311643481254578,
+      "learning_rate": 0.0005820898879697994,
+      "loss": 3.3563,
+      "step": 6767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5513588190078735,
+      "learning_rate": 0.0005820846475390683,
+      "loss": 3.4952,
+      "step": 6768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5474352240562439,
+      "learning_rate": 0.0005820794063653805,
+      "loss": 3.6479,
+      "step": 6769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5823585987091064,
+      "learning_rate": 0.0005820741644487497,
+      "loss": 3.1314,
+      "step": 6770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5332926511764526,
+      "learning_rate": 0.0005820689217891899,
+      "loss": 3.4825,
+      "step": 6771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5201532244682312,
+      "learning_rate": 0.0005820636783867149,
+      "loss": 3.4215,
+      "step": 6772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.573883056640625,
+      "learning_rate": 0.0005820584342413385,
+      "loss": 3.3778,
+      "step": 6773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6504254937171936,
+      "learning_rate": 0.0005820531893530744,
+      "loss": 3.3315,
+      "step": 6774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5481976270675659,
+      "learning_rate": 0.0005820479437219365,
+      "loss": 3.4299,
+      "step": 6775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5290274024009705,
+      "learning_rate": 0.0005820426973479386,
+      "loss": 3.4469,
+      "step": 6776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5465989112854004,
+      "learning_rate": 0.0005820374502310945,
+      "loss": 3.4553,
+      "step": 6777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5408307313919067,
+      "learning_rate": 0.0005820322023714179,
+      "loss": 3.2675,
+      "step": 6778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5560708045959473,
+      "learning_rate": 0.0005820269537689229,
+      "loss": 3.235,
+      "step": 6779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5538557171821594,
+      "learning_rate": 0.0005820217044236232,
+      "loss": 3.3199,
+      "step": 6780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5382668972015381,
+      "learning_rate": 0.0005820164543355326,
+      "loss": 3.5514,
+      "step": 6781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5423834323883057,
+      "learning_rate": 0.0005820112035046648,
+      "loss": 3.2956,
+      "step": 6782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5152990221977234,
+      "learning_rate": 0.0005820059519310339,
+      "loss": 3.2701,
+      "step": 6783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5263355374336243,
+      "learning_rate": 0.0005820006996146536,
+      "loss": 3.4816,
+      "step": 6784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5338313579559326,
+      "learning_rate": 0.0005819954465555377,
+      "loss": 3.2517,
+      "step": 6785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5005643367767334,
+      "learning_rate": 0.0005819901927537,
+      "loss": 3.1299,
+      "step": 6786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5044447183609009,
+      "learning_rate": 0.0005819849382091545,
+      "loss": 3.501,
+      "step": 6787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5595723390579224,
+      "learning_rate": 0.0005819796829219149,
+      "loss": 3.4216,
+      "step": 6788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5340045690536499,
+      "learning_rate": 0.0005819744268919951,
+      "loss": 3.4126,
+      "step": 6789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5521731972694397,
+      "learning_rate": 0.000581969170119409,
+      "loss": 3.5361,
+      "step": 6790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546519935131073,
+      "learning_rate": 0.0005819639126041702,
+      "loss": 3.5096,
+      "step": 6791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5405811071395874,
+      "learning_rate": 0.0005819586543462928,
+      "loss": 3.2427,
+      "step": 6792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5262058973312378,
+      "learning_rate": 0.0005819533953457906,
+      "loss": 3.4118,
+      "step": 6793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5601271986961365,
+      "learning_rate": 0.0005819481356026774,
+      "loss": 3.2993,
+      "step": 6794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5209169387817383,
+      "learning_rate": 0.0005819428751169671,
+      "loss": 3.2435,
+      "step": 6795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5251837968826294,
+      "learning_rate": 0.0005819376138886734,
+      "loss": 3.2903,
+      "step": 6796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5272718667984009,
+      "learning_rate": 0.0005819323519178105,
+      "loss": 3.3486,
+      "step": 6797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6075021624565125,
+      "learning_rate": 0.0005819270892043919,
+      "loss": 3.413,
+      "step": 6798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5117523670196533,
+      "learning_rate": 0.0005819218257484315,
+      "loss": 3.4657,
+      "step": 6799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4947341978549957,
+      "learning_rate": 0.0005819165615499435,
+      "loss": 3.5324,
+      "step": 6800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5065931081771851,
+      "learning_rate": 0.0005819112966089415,
+      "loss": 3.4816,
+      "step": 6801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5459311604499817,
+      "learning_rate": 0.0005819060309254393,
+      "loss": 3.2451,
+      "step": 6802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5343260169029236,
+      "learning_rate": 0.0005819007644994509,
+      "loss": 3.4584,
+      "step": 6803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.514397382736206,
+      "learning_rate": 0.0005818954973309901,
+      "loss": 3.4568,
+      "step": 6804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5538575053215027,
+      "learning_rate": 0.0005818902294200708,
+      "loss": 3.208,
+      "step": 6805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5236798524856567,
+      "learning_rate": 0.0005818849607667069,
+      "loss": 3.3301,
+      "step": 6806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5536655187606812,
+      "learning_rate": 0.0005818796913709123,
+      "loss": 3.4512,
+      "step": 6807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5622990727424622,
+      "learning_rate": 0.0005818744212327009,
+      "loss": 3.5709,
+      "step": 6808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5513145327568054,
+      "learning_rate": 0.0005818691503520865,
+      "loss": 3.4054,
+      "step": 6809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5191030502319336,
+      "learning_rate": 0.0005818638787290829,
+      "loss": 3.3703,
+      "step": 6810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5005136132240295,
+      "learning_rate": 0.0005818586063637041,
+      "loss": 3.3643,
+      "step": 6811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5312431454658508,
+      "learning_rate": 0.0005818533332559642,
+      "loss": 3.4871,
+      "step": 6812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5487942099571228,
+      "learning_rate": 0.0005818480594058766,
+      "loss": 3.3361,
+      "step": 6813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5031387209892273,
+      "learning_rate": 0.0005818427848134556,
+      "loss": 3.5931,
+      "step": 6814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5074050426483154,
+      "learning_rate": 0.0005818375094787148,
+      "loss": 3.3569,
+      "step": 6815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5995979905128479,
+      "learning_rate": 0.0005818322334016684,
+      "loss": 3.412,
+      "step": 6816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5569592118263245,
+      "learning_rate": 0.0005818269565823301,
+      "loss": 3.4492,
+      "step": 6817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5373505353927612,
+      "learning_rate": 0.0005818216790207137,
+      "loss": 3.3863,
+      "step": 6818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.513282835483551,
+      "learning_rate": 0.0005818164007168335,
+      "loss": 3.1659,
+      "step": 6819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5331324338912964,
+      "learning_rate": 0.0005818111216707029,
+      "loss": 3.1737,
+      "step": 6820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5432617664337158,
+      "learning_rate": 0.0005818058418823361,
+      "loss": 3.5055,
+      "step": 6821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5340288877487183,
+      "learning_rate": 0.000581800561351747,
+      "loss": 3.3134,
+      "step": 6822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.523755669593811,
+      "learning_rate": 0.0005817952800789494,
+      "loss": 3.41,
+      "step": 6823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5807611346244812,
+      "learning_rate": 0.0005817899980639573,
+      "loss": 3.3649,
+      "step": 6824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5344754457473755,
+      "learning_rate": 0.0005817847153067846,
+      "loss": 3.3996,
+      "step": 6825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5475096106529236,
+      "learning_rate": 0.0005817794318074452,
+      "loss": 3.2395,
+      "step": 6826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5475925803184509,
+      "learning_rate": 0.0005817741475659529,
+      "loss": 3.0788,
+      "step": 6827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5353468060493469,
+      "learning_rate": 0.0005817688625823218,
+      "loss": 3.3761,
+      "step": 6828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5556272864341736,
+      "learning_rate": 0.0005817635768565657,
+      "loss": 3.4141,
+      "step": 6829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5247024297714233,
+      "learning_rate": 0.0005817582903886986,
+      "loss": 3.2435,
+      "step": 6830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5360187292098999,
+      "learning_rate": 0.0005817530031787344,
+      "loss": 3.3121,
+      "step": 6831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5367879867553711,
+      "learning_rate": 0.000581747715226687,
+      "loss": 3.3657,
+      "step": 6832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5176432728767395,
+      "learning_rate": 0.0005817424265325703,
+      "loss": 3.3106,
+      "step": 6833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5033512115478516,
+      "learning_rate": 0.0005817371370963984,
+      "loss": 3.3319,
+      "step": 6834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5678747892379761,
+      "learning_rate": 0.0005817318469181849,
+      "loss": 3.3628,
+      "step": 6835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5114911794662476,
+      "learning_rate": 0.000581726555997944,
+      "loss": 3.1957,
+      "step": 6836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5610429644584656,
+      "learning_rate": 0.0005817212643356897,
+      "loss": 3.1318,
+      "step": 6837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5097834467887878,
+      "learning_rate": 0.0005817159719314358,
+      "loss": 3.2942,
+      "step": 6838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5260331630706787,
+      "learning_rate": 0.0005817106787851962,
+      "loss": 3.3047,
+      "step": 6839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5727056860923767,
+      "learning_rate": 0.0005817053848969848,
+      "loss": 3.4154,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5135591626167297,
+      "learning_rate": 0.0005817000902668157,
+      "loss": 3.5154,
+      "step": 6841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5463142395019531,
+      "learning_rate": 0.0005816947948947028,
+      "loss": 3.2277,
+      "step": 6842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5426071882247925,
+      "learning_rate": 0.00058168949878066,
+      "loss": 3.4398,
+      "step": 6843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5238614082336426,
+      "learning_rate": 0.0005816842019247012,
+      "loss": 3.3057,
+      "step": 6844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4938346743583679,
+      "learning_rate": 0.0005816789043268406,
+      "loss": 3.3241,
+      "step": 6845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5384917259216309,
+      "learning_rate": 0.0005816736059870918,
+      "loss": 3.2526,
+      "step": 6846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5270363092422485,
+      "learning_rate": 0.000581668306905469,
+      "loss": 3.3041,
+      "step": 6847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.508563220500946,
+      "learning_rate": 0.0005816630070819861,
+      "loss": 3.3541,
+      "step": 6848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5612684488296509,
+      "learning_rate": 0.000581657706516657,
+      "loss": 3.463,
+      "step": 6849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5991032123565674,
+      "learning_rate": 0.0005816524052094957,
+      "loss": 3.1475,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5217996835708618,
+      "learning_rate": 0.0005816471031605163,
+      "loss": 3.565,
+      "step": 6851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5394172072410583,
+      "learning_rate": 0.0005816418003697324,
+      "loss": 3.2722,
+      "step": 6852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5643725395202637,
+      "learning_rate": 0.0005816364968371584,
+      "loss": 3.2096,
+      "step": 6853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5298509001731873,
+      "learning_rate": 0.000581631192562808,
+      "loss": 3.379,
+      "step": 6854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5548331141471863,
+      "learning_rate": 0.0005816258875466953,
+      "loss": 3.1849,
+      "step": 6855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4925016760826111,
+      "learning_rate": 0.0005816205817888342,
+      "loss": 3.3093,
+      "step": 6856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5486854910850525,
+      "learning_rate": 0.0005816152752892387,
+      "loss": 3.3507,
+      "step": 6857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5491653084754944,
+      "learning_rate": 0.0005816099680479226,
+      "loss": 3.3045,
+      "step": 6858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5340956449508667,
+      "learning_rate": 0.0005816046600649002,
+      "loss": 3.3908,
+      "step": 6859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5760660767555237,
+      "learning_rate": 0.0005815993513401853,
+      "loss": 3.2897,
+      "step": 6860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5405675768852234,
+      "learning_rate": 0.0005815940418737918,
+      "loss": 3.3984,
+      "step": 6861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5545721054077148,
+      "learning_rate": 0.0005815887316657339,
+      "loss": 3.6153,
+      "step": 6862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5303523540496826,
+      "learning_rate": 0.0005815834207160255,
+      "loss": 3.4198,
+      "step": 6863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5388019680976868,
+      "learning_rate": 0.0005815781090246805,
+      "loss": 3.3454,
+      "step": 6864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5352962613105774,
+      "learning_rate": 0.0005815727965917129,
+      "loss": 3.2491,
+      "step": 6865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5232275128364563,
+      "learning_rate": 0.0005815674834171368,
+      "loss": 3.3865,
+      "step": 6866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5305415987968445,
+      "learning_rate": 0.0005815621695009662,
+      "loss": 3.25,
+      "step": 6867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5452413558959961,
+      "learning_rate": 0.000581556854843215,
+      "loss": 3.3942,
+      "step": 6868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5450773239135742,
+      "learning_rate": 0.0005815515394438971,
+      "loss": 3.3143,
+      "step": 6869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.541786789894104,
+      "learning_rate": 0.0005815462233030267,
+      "loss": 3.5092,
+      "step": 6870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5301926136016846,
+      "learning_rate": 0.0005815409064206178,
+      "loss": 3.5044,
+      "step": 6871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5375513434410095,
+      "learning_rate": 0.0005815355887966843,
+      "loss": 3.2972,
+      "step": 6872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5402582883834839,
+      "learning_rate": 0.0005815302704312403,
+      "loss": 3.2579,
+      "step": 6873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5533377528190613,
+      "learning_rate": 0.0005815249513242996,
+      "loss": 3.5222,
+      "step": 6874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5436393022537231,
+      "learning_rate": 0.0005815196314758765,
+      "loss": 3.1836,
+      "step": 6875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5396994948387146,
+      "learning_rate": 0.0005815143108859848,
+      "loss": 3.374,
+      "step": 6876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5397756099700928,
+      "learning_rate": 0.0005815089895546386,
+      "loss": 3.2155,
+      "step": 6877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5924909710884094,
+      "learning_rate": 0.0005815036674818519,
+      "loss": 3.35,
+      "step": 6878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5628549456596375,
+      "learning_rate": 0.0005814983446676388,
+      "loss": 3.44,
+      "step": 6879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5356522798538208,
+      "learning_rate": 0.0005814930211120132,
+      "loss": 3.2106,
+      "step": 6880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5571137070655823,
+      "learning_rate": 0.0005814876968149891,
+      "loss": 3.4285,
+      "step": 6881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.579041600227356,
+      "learning_rate": 0.0005814823717765806,
+      "loss": 3.3484,
+      "step": 6882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4999987781047821,
+      "learning_rate": 0.0005814770459968016,
+      "loss": 3.3225,
+      "step": 6883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5245411396026611,
+      "learning_rate": 0.0005814717194756663,
+      "loss": 3.3605,
+      "step": 6884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5306369066238403,
+      "learning_rate": 0.0005814663922131888,
+      "loss": 3.4097,
+      "step": 6885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5047025084495544,
+      "learning_rate": 0.0005814610642093828,
+      "loss": 3.287,
+      "step": 6886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5388005971908569,
+      "learning_rate": 0.0005814557354642627,
+      "loss": 3.2558,
+      "step": 6887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5090034008026123,
+      "learning_rate": 0.0005814504059778422,
+      "loss": 3.2232,
+      "step": 6888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.523379921913147,
+      "learning_rate": 0.0005814450757501355,
+      "loss": 3.2021,
+      "step": 6889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5205253958702087,
+      "learning_rate": 0.0005814397447811568,
+      "loss": 3.5486,
+      "step": 6890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5232972502708435,
+      "learning_rate": 0.0005814344130709198,
+      "loss": 3.3859,
+      "step": 6891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.548040509223938,
+      "learning_rate": 0.0005814290806194388,
+      "loss": 3.2865,
+      "step": 6892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5398008227348328,
+      "learning_rate": 0.0005814237474267277,
+      "loss": 3.3062,
+      "step": 6893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5877861380577087,
+      "learning_rate": 0.0005814184134928008,
+      "loss": 3.3348,
+      "step": 6894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5505077242851257,
+      "learning_rate": 0.0005814130788176718,
+      "loss": 3.3898,
+      "step": 6895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5634656548500061,
+      "learning_rate": 0.0005814077434013549,
+      "loss": 3.2064,
+      "step": 6896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5058372616767883,
+      "learning_rate": 0.0005814024072438642,
+      "loss": 3.3888,
+      "step": 6897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5141379237174988,
+      "learning_rate": 0.0005813970703452138,
+      "loss": 3.174,
+      "step": 6898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5080946683883667,
+      "learning_rate": 0.0005813917327054175,
+      "loss": 3.1621,
+      "step": 6899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.533840000629425,
+      "learning_rate": 0.0005813863943244897,
+      "loss": 3.4227,
+      "step": 6900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5628681182861328,
+      "learning_rate": 0.0005813810552024441,
+      "loss": 3.1773,
+      "step": 6901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5388134717941284,
+      "learning_rate": 0.0005813757153392951,
+      "loss": 3.5305,
+      "step": 6902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5274122953414917,
+      "learning_rate": 0.0005813703747350566,
+      "loss": 3.3009,
+      "step": 6903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5223855972290039,
+      "learning_rate": 0.0005813650333897426,
+      "loss": 3.3918,
+      "step": 6904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5121346712112427,
+      "learning_rate": 0.0005813596913033673,
+      "loss": 3.3949,
+      "step": 6905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5447664260864258,
+      "learning_rate": 0.0005813543484759448,
+      "loss": 3.3467,
+      "step": 6906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5025432705879211,
+      "learning_rate": 0.000581349004907489,
+      "loss": 3.388,
+      "step": 6907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5353335738182068,
+      "learning_rate": 0.0005813436605980142,
+      "loss": 3.2766,
+      "step": 6908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.503914475440979,
+      "learning_rate": 0.0005813383155475342,
+      "loss": 3.2077,
+      "step": 6909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5467348098754883,
+      "learning_rate": 0.0005813329697560633,
+      "loss": 3.2751,
+      "step": 6910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5445327162742615,
+      "learning_rate": 0.0005813276232236155,
+      "loss": 3.3215,
+      "step": 6911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5562853217124939,
+      "learning_rate": 0.0005813222759502047,
+      "loss": 3.4057,
+      "step": 6912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5421975255012512,
+      "learning_rate": 0.0005813169279358454,
+      "loss": 3.2963,
+      "step": 6913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5326264500617981,
+      "learning_rate": 0.0005813115791805513,
+      "loss": 3.2552,
+      "step": 6914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5453374981880188,
+      "learning_rate": 0.0005813062296843368,
+      "loss": 3.5338,
+      "step": 6915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.531548023223877,
+      "learning_rate": 0.0005813008794472158,
+      "loss": 3.3438,
+      "step": 6916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.586500346660614,
+      "learning_rate": 0.0005812955284692022,
+      "loss": 3.2636,
+      "step": 6917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.523780345916748,
+      "learning_rate": 0.0005812901767503105,
+      "loss": 3.4048,
+      "step": 6918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5507670640945435,
+      "learning_rate": 0.0005812848242905547,
+      "loss": 3.3771,
+      "step": 6919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5263717174530029,
+      "learning_rate": 0.0005812794710899486,
+      "loss": 3.1895,
+      "step": 6920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5932714939117432,
+      "learning_rate": 0.0005812741171485066,
+      "loss": 3.3006,
+      "step": 6921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5689055919647217,
+      "learning_rate": 0.0005812687624662427,
+      "loss": 3.3535,
+      "step": 6922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5490853786468506,
+      "learning_rate": 0.000581263407043171,
+      "loss": 3.2321,
+      "step": 6923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5083225965499878,
+      "learning_rate": 0.0005812580508793056,
+      "loss": 3.2538,
+      "step": 6924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5350843071937561,
+      "learning_rate": 0.0005812526939746607,
+      "loss": 3.351,
+      "step": 6925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5658883452415466,
+      "learning_rate": 0.0005812473363292503,
+      "loss": 3.2112,
+      "step": 6926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5066301822662354,
+      "learning_rate": 0.0005812419779430885,
+      "loss": 3.5274,
+      "step": 6927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5520161986351013,
+      "learning_rate": 0.0005812366188161894,
+      "loss": 3.1782,
+      "step": 6928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5627199411392212,
+      "learning_rate": 0.0005812312589485673,
+      "loss": 3.2834,
+      "step": 6929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.544322669506073,
+      "learning_rate": 0.0005812258983402363,
+      "loss": 3.5101,
+      "step": 6930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5476049184799194,
+      "learning_rate": 0.0005812205369912102,
+      "loss": 3.2291,
+      "step": 6931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5459288358688354,
+      "learning_rate": 0.0005812151749015034,
+      "loss": 3.2593,
+      "step": 6932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.533470094203949,
+      "learning_rate": 0.0005812098120711299,
+      "loss": 3.3264,
+      "step": 6933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5248565077781677,
+      "learning_rate": 0.0005812044485001039,
+      "loss": 3.2764,
+      "step": 6934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5313982963562012,
+      "learning_rate": 0.0005811990841884395,
+      "loss": 3.4671,
+      "step": 6935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.695208728313446,
+      "learning_rate": 0.0005811937191361508,
+      "loss": 3.5434,
+      "step": 6936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5338233709335327,
+      "learning_rate": 0.0005811883533432521,
+      "loss": 3.249,
+      "step": 6937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5622416138648987,
+      "learning_rate": 0.0005811829868097572,
+      "loss": 3.4326,
+      "step": 6938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5586525797843933,
+      "learning_rate": 0.0005811776195356805,
+      "loss": 3.3476,
+      "step": 6939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5135695934295654,
+      "learning_rate": 0.0005811722515210361,
+      "loss": 3.3332,
+      "step": 6940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5429893732070923,
+      "learning_rate": 0.0005811668827658381,
+      "loss": 3.2199,
+      "step": 6941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6040959358215332,
+      "learning_rate": 0.0005811615132701005,
+      "loss": 3.493,
+      "step": 6942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5325751304626465,
+      "learning_rate": 0.0005811561430338378,
+      "loss": 3.2832,
+      "step": 6943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5440836548805237,
+      "learning_rate": 0.0005811507720570638,
+      "loss": 3.2621,
+      "step": 6944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5436229109764099,
+      "learning_rate": 0.0005811454003397928,
+      "loss": 3.3883,
+      "step": 6945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5637691020965576,
+      "learning_rate": 0.0005811400278820391,
+      "loss": 3.4486,
+      "step": 6946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5535193085670471,
+      "learning_rate": 0.0005811346546838165,
+      "loss": 3.4498,
+      "step": 6947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5341395139694214,
+      "learning_rate": 0.0005811292807451393,
+      "loss": 3.4532,
+      "step": 6948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6911645531654358,
+      "learning_rate": 0.0005811239060660217,
+      "loss": 3.2766,
+      "step": 6949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5011906027793884,
+      "learning_rate": 0.0005811185306464779,
+      "loss": 3.1654,
+      "step": 6950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5271753668785095,
+      "learning_rate": 0.0005811131544865218,
+      "loss": 3.3565,
+      "step": 6951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5180072784423828,
+      "learning_rate": 0.000581107777586168,
+      "loss": 3.4378,
+      "step": 6952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.541093111038208,
+      "learning_rate": 0.0005811023999454303,
+      "loss": 3.2264,
+      "step": 6953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5188407897949219,
+      "learning_rate": 0.000581097021564323,
+      "loss": 3.2942,
+      "step": 6954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5392166376113892,
+      "learning_rate": 0.0005810916424428602,
+      "loss": 3.5065,
+      "step": 6955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5347229242324829,
+      "learning_rate": 0.0005810862625810562,
+      "loss": 3.2041,
+      "step": 6956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6093183159828186,
+      "learning_rate": 0.000581080881978925,
+      "loss": 3.5634,
+      "step": 6957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5233674645423889,
+      "learning_rate": 0.0005810755006364809,
+      "loss": 3.2121,
+      "step": 6958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5367003679275513,
+      "learning_rate": 0.0005810701185537379,
+      "loss": 3.4925,
+      "step": 6959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5209534764289856,
+      "learning_rate": 0.0005810647357307105,
+      "loss": 3.292,
+      "step": 6960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5548787713050842,
+      "learning_rate": 0.0005810593521674125,
+      "loss": 3.1859,
+      "step": 6961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5430557131767273,
+      "learning_rate": 0.0005810539678638584,
+      "loss": 3.4861,
+      "step": 6962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5306611657142639,
+      "learning_rate": 0.0005810485828200622,
+      "loss": 3.4295,
+      "step": 6963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6463212966918945,
+      "learning_rate": 0.000581043197036038,
+      "loss": 3.4153,
+      "step": 6964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5019036531448364,
+      "learning_rate": 0.0005810378105118002,
+      "loss": 3.3514,
+      "step": 6965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5331908464431763,
+      "learning_rate": 0.0005810324232473629,
+      "loss": 3.4269,
+      "step": 6966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5343552827835083,
+      "learning_rate": 0.0005810270352427403,
+      "loss": 3.5078,
+      "step": 6967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.561147928237915,
+      "learning_rate": 0.0005810216464979466,
+      "loss": 3.233,
+      "step": 6968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5656723380088806,
+      "learning_rate": 0.0005810162570129958,
+      "loss": 3.1724,
+      "step": 6969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5117201805114746,
+      "learning_rate": 0.0005810108667879025,
+      "loss": 3.3606,
+      "step": 6970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5727672576904297,
+      "learning_rate": 0.0005810054758226805,
+      "loss": 3.4015,
+      "step": 6971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.594057559967041,
+      "learning_rate": 0.0005810000841173442,
+      "loss": 3.4109,
+      "step": 6972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5242696404457092,
+      "learning_rate": 0.0005809946916719077,
+      "loss": 3.1948,
+      "step": 6973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5410612225532532,
+      "learning_rate": 0.0005809892984863854,
+      "loss": 3.3574,
+      "step": 6974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5484248995780945,
+      "learning_rate": 0.0005809839045607913,
+      "loss": 3.1889,
+      "step": 6975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5360888242721558,
+      "learning_rate": 0.0005809785098951396,
+      "loss": 3.3528,
+      "step": 6976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6561615467071533,
+      "learning_rate": 0.0005809731144894445,
+      "loss": 3.3181,
+      "step": 6977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5585454106330872,
+      "learning_rate": 0.0005809677183437206,
+      "loss": 3.3861,
+      "step": 6978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.539271354675293,
+      "learning_rate": 0.0005809623214579816,
+      "loss": 3.3186,
+      "step": 6979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5142511129379272,
+      "learning_rate": 0.0005809569238322419,
+      "loss": 3.296,
+      "step": 6980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5410292148590088,
+      "learning_rate": 0.0005809515254665158,
+      "loss": 3.3056,
+      "step": 6981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5308331847190857,
+      "learning_rate": 0.0005809461263608175,
+      "loss": 3.3361,
+      "step": 6982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546271562576294,
+      "learning_rate": 0.000580940726515161,
+      "loss": 3.4194,
+      "step": 6983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.566936731338501,
+      "learning_rate": 0.0005809353259295608,
+      "loss": 3.3992,
+      "step": 6984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5126430988311768,
+      "learning_rate": 0.000580929924604031,
+      "loss": 3.2163,
+      "step": 6985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4886528551578522,
+      "learning_rate": 0.000580924522538586,
+      "loss": 3.3871,
+      "step": 6986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5845642685890198,
+      "learning_rate": 0.0005809191197332397,
+      "loss": 3.4001,
+      "step": 6987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5310481786727905,
+      "learning_rate": 0.0005809137161880065,
+      "loss": 3.5392,
+      "step": 6988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5669585466384888,
+      "learning_rate": 0.0005809083119029006,
+      "loss": 3.4265,
+      "step": 6989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5161418914794922,
+      "learning_rate": 0.0005809029068779365,
+      "loss": 3.3468,
+      "step": 6990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5583595037460327,
+      "learning_rate": 0.0005808975011131279,
+      "loss": 3.4198,
+      "step": 6991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5388359427452087,
+      "learning_rate": 0.0005808920946084896,
+      "loss": 3.4827,
+      "step": 6992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5691794157028198,
+      "learning_rate": 0.0005808866873640353,
+      "loss": 3.5184,
+      "step": 6993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6655071973800659,
+      "learning_rate": 0.0005808812793797797,
+      "loss": 3.3542,
+      "step": 6994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5387119054794312,
+      "learning_rate": 0.000580875870655737,
+      "loss": 3.2979,
+      "step": 6995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6775738596916199,
+      "learning_rate": 0.0005808704611919211,
+      "loss": 3.3308,
+      "step": 6996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.537527859210968,
+      "learning_rate": 0.0005808650509883465,
+      "loss": 3.4031,
+      "step": 6997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5527935028076172,
+      "learning_rate": 0.0005808596400450275,
+      "loss": 3.251,
+      "step": 6998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5826919674873352,
+      "learning_rate": 0.0005808542283619781,
+      "loss": 3.3252,
+      "step": 6999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5161682367324829,
+      "learning_rate": 0.0005808488159392129,
+      "loss": 3.0676,
+      "step": 7000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5332803130149841,
+      "learning_rate": 0.0005808434027767459,
+      "loss": 3.5714,
+      "step": 7001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5207781791687012,
+      "learning_rate": 0.0005808379888745914,
+      "loss": 3.3575,
+      "step": 7002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5248399376869202,
+      "learning_rate": 0.0005808325742327636,
+      "loss": 3.0861,
+      "step": 7003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.547214686870575,
+      "learning_rate": 0.0005808271588512771,
+      "loss": 3.4018,
+      "step": 7004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.49481138586997986,
+      "learning_rate": 0.0005808217427301456,
+      "loss": 3.2186,
+      "step": 7005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.567725658416748,
+      "learning_rate": 0.0005808163258693839,
+      "loss": 3.5048,
+      "step": 7006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6113198399543762,
+      "learning_rate": 0.000580810908269006,
+      "loss": 3.2416,
+      "step": 7007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5307554006576538,
+      "learning_rate": 0.0005808054899290262,
+      "loss": 3.2915,
+      "step": 7008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5398176312446594,
+      "learning_rate": 0.0005808000708494587,
+      "loss": 3.2426,
+      "step": 7009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5282324552536011,
+      "learning_rate": 0.0005807946510303179,
+      "loss": 3.5821,
+      "step": 7010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5628719925880432,
+      "learning_rate": 0.0005807892304716181,
+      "loss": 3.5204,
+      "step": 7011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5679787397384644,
+      "learning_rate": 0.0005807838091733734,
+      "loss": 3.3541,
+      "step": 7012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5467495322227478,
+      "learning_rate": 0.0005807783871355982,
+      "loss": 3.2906,
+      "step": 7013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5178360939025879,
+      "learning_rate": 0.0005807729643583069,
+      "loss": 3.3828,
+      "step": 7014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5609511137008667,
+      "learning_rate": 0.0005807675408415134,
+      "loss": 3.3343,
+      "step": 7015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5502193570137024,
+      "learning_rate": 0.0005807621165852325,
+      "loss": 3.3203,
+      "step": 7016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5327548980712891,
+      "learning_rate": 0.000580756691589478,
+      "loss": 3.3419,
+      "step": 7017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.506458044052124,
+      "learning_rate": 0.0005807512658542646,
+      "loss": 3.2494,
+      "step": 7018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5278050303459167,
+      "learning_rate": 0.0005807458393796062,
+      "loss": 3.3536,
+      "step": 7019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.530926525592804,
+      "learning_rate": 0.0005807404121655175,
+      "loss": 3.3645,
+      "step": 7020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6142529845237732,
+      "learning_rate": 0.0005807349842120124,
+      "loss": 3.3903,
+      "step": 7021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5566715002059937,
+      "learning_rate": 0.0005807295555191055,
+      "loss": 3.3908,
+      "step": 7022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5471014380455017,
+      "learning_rate": 0.0005807241260868109,
+      "loss": 3.4871,
+      "step": 7023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.584735095500946,
+      "learning_rate": 0.000580718695915143,
+      "loss": 3.3576,
+      "step": 7024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5210456848144531,
+      "learning_rate": 0.0005807132650041162,
+      "loss": 3.3165,
+      "step": 7025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5415894985198975,
+      "learning_rate": 0.0005807078333537445,
+      "loss": 3.4889,
+      "step": 7026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.519156813621521,
+      "learning_rate": 0.0005807024009640425,
+      "loss": 3.3567,
+      "step": 7027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5499240159988403,
+      "learning_rate": 0.0005806969678350243,
+      "loss": 3.3908,
+      "step": 7028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5356557369232178,
+      "learning_rate": 0.0005806915339667044,
+      "loss": 3.1221,
+      "step": 7029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.562337338924408,
+      "learning_rate": 0.000580686099359097,
+      "loss": 3.1642,
+      "step": 7030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5245587229728699,
+      "learning_rate": 0.0005806806640122164,
+      "loss": 3.277,
+      "step": 7031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.548549473285675,
+      "learning_rate": 0.000580675227926077,
+      "loss": 3.497,
+      "step": 7032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5365231037139893,
+      "learning_rate": 0.0005806697911006931,
+      "loss": 3.257,
+      "step": 7033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5649599432945251,
+      "learning_rate": 0.0005806643535360789,
+      "loss": 3.2649,
+      "step": 7034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5196275115013123,
+      "learning_rate": 0.0005806589152322489,
+      "loss": 3.2268,
+      "step": 7035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.509590744972229,
+      "learning_rate": 0.0005806534761892172,
+      "loss": 3.2333,
+      "step": 7036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5706533789634705,
+      "learning_rate": 0.0005806480364069983,
+      "loss": 3.011,
+      "step": 7037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5456781983375549,
+      "learning_rate": 0.0005806425958856065,
+      "loss": 3.5429,
+      "step": 7038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5550026893615723,
+      "learning_rate": 0.0005806371546250562,
+      "loss": 3.4187,
+      "step": 7039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5485456585884094,
+      "learning_rate": 0.0005806317126253616,
+      "loss": 3.2612,
+      "step": 7040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5675588250160217,
+      "learning_rate": 0.0005806262698865369,
+      "loss": 3.4519,
+      "step": 7041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5755375623703003,
+      "learning_rate": 0.0005806208264085968,
+      "loss": 3.2874,
+      "step": 7042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5225034952163696,
+      "learning_rate": 0.0005806153821915554,
+      "loss": 3.317,
+      "step": 7043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5519276261329651,
+      "learning_rate": 0.0005806099372354271,
+      "loss": 3.4839,
+      "step": 7044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5228083729743958,
+      "learning_rate": 0.0005806044915402262,
+      "loss": 3.334,
+      "step": 7045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5322780013084412,
+      "learning_rate": 0.0005805990451059671,
+      "loss": 3.397,
+      "step": 7046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5328549742698669,
+      "learning_rate": 0.0005805935979326641,
+      "loss": 3.4732,
+      "step": 7047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48566094040870667,
+      "learning_rate": 0.0005805881500203316,
+      "loss": 3.3144,
+      "step": 7048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5601107478141785,
+      "learning_rate": 0.0005805827013689839,
+      "loss": 3.3341,
+      "step": 7049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5418280363082886,
+      "learning_rate": 0.0005805772519786353,
+      "loss": 3.4561,
+      "step": 7050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5339364409446716,
+      "learning_rate": 0.0005805718018493003,
+      "loss": 3.4457,
+      "step": 7051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6148756146430969,
+      "learning_rate": 0.0005805663509809932,
+      "loss": 3.3762,
+      "step": 7052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5323485136032104,
+      "learning_rate": 0.0005805608993737282,
+      "loss": 3.3379,
+      "step": 7053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5185231566429138,
+      "learning_rate": 0.0005805554470275199,
+      "loss": 3.1767,
+      "step": 7054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.54229736328125,
+      "learning_rate": 0.0005805499939423826,
+      "loss": 3.3344,
+      "step": 7055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5611851215362549,
+      "learning_rate": 0.0005805445401183305,
+      "loss": 3.4762,
+      "step": 7056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5249059200286865,
+      "learning_rate": 0.000580539085555378,
+      "loss": 3.4588,
+      "step": 7057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5460208654403687,
+      "learning_rate": 0.0005805336302535397,
+      "loss": 3.3557,
+      "step": 7058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6470723152160645,
+      "learning_rate": 0.0005805281742128297,
+      "loss": 3.0215,
+      "step": 7059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5494178533554077,
+      "learning_rate": 0.0005805227174332625,
+      "loss": 3.3288,
+      "step": 7060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5545464754104614,
+      "learning_rate": 0.0005805172599148525,
+      "loss": 3.222,
+      "step": 7061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5664929747581482,
+      "learning_rate": 0.0005805118016576139,
+      "loss": 3.2427,
+      "step": 7062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.589393138885498,
+      "learning_rate": 0.0005805063426615613,
+      "loss": 3.3364,
+      "step": 7063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5010703802108765,
+      "learning_rate": 0.000580500882926709,
+      "loss": 3.4066,
+      "step": 7064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5556477904319763,
+      "learning_rate": 0.0005804954224530712,
+      "loss": 3.4683,
+      "step": 7065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5511776208877563,
+      "learning_rate": 0.0005804899612406627,
+      "loss": 3.3322,
+      "step": 7066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5432040095329285,
+      "learning_rate": 0.0005804844992894975,
+      "loss": 3.2741,
+      "step": 7067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5588909983634949,
+      "learning_rate": 0.00058047903659959,
+      "loss": 3.6012,
+      "step": 7068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5545442700386047,
+      "learning_rate": 0.0005804735731709548,
+      "loss": 3.373,
+      "step": 7069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5283875465393066,
+      "learning_rate": 0.0005804681090036062,
+      "loss": 3.2354,
+      "step": 7070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4935014843940735,
+      "learning_rate": 0.0005804626440975585,
+      "loss": 3.3775,
+      "step": 7071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5301677584648132,
+      "learning_rate": 0.0005804571784528262,
+      "loss": 3.3929,
+      "step": 7072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5176547169685364,
+      "learning_rate": 0.0005804517120694238,
+      "loss": 3.4168,
+      "step": 7073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5569431781768799,
+      "learning_rate": 0.0005804462449473654,
+      "loss": 3.4039,
+      "step": 7074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5683629512786865,
+      "learning_rate": 0.0005804407770866656,
+      "loss": 3.4013,
+      "step": 7075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5122573375701904,
+      "learning_rate": 0.0005804353084873388,
+      "loss": 3.2245,
+      "step": 7076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5609269142150879,
+      "learning_rate": 0.0005804298391493993,
+      "loss": 3.2241,
+      "step": 7077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5435025691986084,
+      "learning_rate": 0.0005804243690728617,
+      "loss": 3.2807,
+      "step": 7078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5618283152580261,
+      "learning_rate": 0.0005804188982577402,
+      "loss": 3.1999,
+      "step": 7079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5037075877189636,
+      "learning_rate": 0.0005804134267040494,
+      "loss": 3.3382,
+      "step": 7080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5446479916572571,
+      "learning_rate": 0.0005804079544118034,
+      "loss": 3.1628,
+      "step": 7081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5279077291488647,
+      "learning_rate": 0.0005804024813810169,
+      "loss": 3.2108,
+      "step": 7082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.556573748588562,
+      "learning_rate": 0.0005803970076117043,
+      "loss": 3.47,
+      "step": 7083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5409570932388306,
+      "learning_rate": 0.0005803915331038799,
+      "loss": 3.2393,
+      "step": 7084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5191570520401001,
+      "learning_rate": 0.0005803860578575581,
+      "loss": 3.3679,
+      "step": 7085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5882177352905273,
+      "learning_rate": 0.0005803805818727535,
+      "loss": 3.1786,
+      "step": 7086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5755952000617981,
+      "learning_rate": 0.0005803751051494803,
+      "loss": 3.4671,
+      "step": 7087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5264267325401306,
+      "learning_rate": 0.000580369627687753,
+      "loss": 3.1961,
+      "step": 7088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5818674564361572,
+      "learning_rate": 0.0005803641494875861,
+      "loss": 3.334,
+      "step": 7089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.529707670211792,
+      "learning_rate": 0.000580358670548994,
+      "loss": 3.3026,
+      "step": 7090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5402799844741821,
+      "learning_rate": 0.0005803531908719912,
+      "loss": 3.4426,
+      "step": 7091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5018097162246704,
+      "learning_rate": 0.000580347710456592,
+      "loss": 3.1922,
+      "step": 7092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5327255129814148,
+      "learning_rate": 0.0005803422293028109,
+      "loss": 3.2424,
+      "step": 7093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5543902516365051,
+      "learning_rate": 0.0005803367474106623,
+      "loss": 3.463,
+      "step": 7094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5371736288070679,
+      "learning_rate": 0.0005803312647801607,
+      "loss": 3.2514,
+      "step": 7095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5503165125846863,
+      "learning_rate": 0.0005803257814113204,
+      "loss": 3.5273,
+      "step": 7096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5683706998825073,
+      "learning_rate": 0.0005803202973041561,
+      "loss": 3.189,
+      "step": 7097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5162700414657593,
+      "learning_rate": 0.0005803148124586819,
+      "loss": 3.3167,
+      "step": 7098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49133065342903137,
+      "learning_rate": 0.0005803093268749125,
+      "loss": 3.1603,
+      "step": 7099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5067523717880249,
+      "learning_rate": 0.0005803038405528622,
+      "loss": 3.4655,
+      "step": 7100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5425522923469543,
+      "learning_rate": 0.0005802983534925457,
+      "loss": 3.4017,
+      "step": 7101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5662040114402771,
+      "learning_rate": 0.0005802928656939773,
+      "loss": 3.4728,
+      "step": 7102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5476053953170776,
+      "learning_rate": 0.0005802873771571712,
+      "loss": 3.5488,
+      "step": 7103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5239982604980469,
+      "learning_rate": 0.0005802818878821424,
+      "loss": 3.3556,
+      "step": 7104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5249889492988586,
+      "learning_rate": 0.0005802763978689048,
+      "loss": 3.5085,
+      "step": 7105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5688517689704895,
+      "learning_rate": 0.0005802709071174732,
+      "loss": 3.3199,
+      "step": 7106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5531274080276489,
+      "learning_rate": 0.0005802654156278619,
+      "loss": 3.3684,
+      "step": 7107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5954955220222473,
+      "learning_rate": 0.0005802599234000855,
+      "loss": 3.3379,
+      "step": 7108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5915130376815796,
+      "learning_rate": 0.0005802544304341584,
+      "loss": 3.4219,
+      "step": 7109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5060703754425049,
+      "learning_rate": 0.000580248936730095,
+      "loss": 3.2657,
+      "step": 7110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5294564366340637,
+      "learning_rate": 0.0005802434422879099,
+      "loss": 3.4222,
+      "step": 7111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5669923424720764,
+      "learning_rate": 0.0005802379471076175,
+      "loss": 3.4991,
+      "step": 7112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.550014317035675,
+      "learning_rate": 0.0005802324511892323,
+      "loss": 3.3429,
+      "step": 7113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5213134288787842,
+      "learning_rate": 0.0005802269545327688,
+      "loss": 3.4838,
+      "step": 7114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5911935567855835,
+      "learning_rate": 0.0005802214571382413,
+      "loss": 3.267,
+      "step": 7115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5569819808006287,
+      "learning_rate": 0.0005802159590056644,
+      "loss": 3.4411,
+      "step": 7116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5387932658195496,
+      "learning_rate": 0.0005802104601350528,
+      "loss": 3.5086,
+      "step": 7117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5420069694519043,
+      "learning_rate": 0.0005802049605264205,
+      "loss": 3.3445,
+      "step": 7118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5610042214393616,
+      "learning_rate": 0.0005801994601797825,
+      "loss": 3.3451,
+      "step": 7119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5179895162582397,
+      "learning_rate": 0.000580193959095153,
+      "loss": 3.3824,
+      "step": 7120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.545372486114502,
+      "learning_rate": 0.0005801884572725464,
+      "loss": 3.5493,
+      "step": 7121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5903968214988708,
+      "learning_rate": 0.0005801829547119775,
+      "loss": 3.4681,
+      "step": 7122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6884878873825073,
+      "learning_rate": 0.0005801774514134605,
+      "loss": 3.4401,
+      "step": 7123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5330374836921692,
+      "learning_rate": 0.00058017194737701,
+      "loss": 3.5123,
+      "step": 7124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5485833883285522,
+      "learning_rate": 0.0005801664426026405,
+      "loss": 3.4519,
+      "step": 7125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6042609214782715,
+      "learning_rate": 0.0005801609370903666,
+      "loss": 3.3484,
+      "step": 7126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5192358493804932,
+      "learning_rate": 0.0005801554308402028,
+      "loss": 3.3385,
+      "step": 7127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5295864343643188,
+      "learning_rate": 0.0005801499238521634,
+      "loss": 3.4082,
+      "step": 7128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5500878691673279,
+      "learning_rate": 0.000580144416126263,
+      "loss": 3.2593,
+      "step": 7129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5443781614303589,
+      "learning_rate": 0.0005801389076625161,
+      "loss": 3.3759,
+      "step": 7130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5370346307754517,
+      "learning_rate": 0.0005801333984609372,
+      "loss": 3.1128,
+      "step": 7131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5285878777503967,
+      "learning_rate": 0.0005801278885215409,
+      "loss": 3.3864,
+      "step": 7132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6192882657051086,
+      "learning_rate": 0.0005801223778443417,
+      "loss": 3.2203,
+      "step": 7133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5949166417121887,
+      "learning_rate": 0.000580116866429354,
+      "loss": 3.527,
+      "step": 7134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5258962512016296,
+      "learning_rate": 0.0005801113542765925,
+      "loss": 3.4491,
+      "step": 7135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.524845540523529,
+      "learning_rate": 0.0005801058413860714,
+      "loss": 3.2764,
+      "step": 7136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5003640651702881,
+      "learning_rate": 0.0005801003277578055,
+      "loss": 3.3281,
+      "step": 7137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5418901443481445,
+      "learning_rate": 0.0005800948133918094,
+      "loss": 3.2208,
+      "step": 7138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5504566431045532,
+      "learning_rate": 0.0005800892982880973,
+      "loss": 3.1584,
+      "step": 7139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5194109082221985,
+      "learning_rate": 0.000580083782446684,
+      "loss": 3.641,
+      "step": 7140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5327237248420715,
+      "learning_rate": 0.0005800782658675838,
+      "loss": 3.1637,
+      "step": 7141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6268619894981384,
+      "learning_rate": 0.0005800727485508114,
+      "loss": 3.264,
+      "step": 7142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5085632801055908,
+      "learning_rate": 0.0005800672304963813,
+      "loss": 3.5367,
+      "step": 7143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6155527830123901,
+      "learning_rate": 0.000580061711704308,
+      "loss": 3.3654,
+      "step": 7144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5261454582214355,
+      "learning_rate": 0.0005800561921746061,
+      "loss": 3.4477,
+      "step": 7145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5312758684158325,
+      "learning_rate": 0.0005800506719072899,
+      "loss": 3.4262,
+      "step": 7146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5167414546012878,
+      "learning_rate": 0.0005800451509023744,
+      "loss": 3.4457,
+      "step": 7147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5222433805465698,
+      "learning_rate": 0.0005800396291598737,
+      "loss": 3.4803,
+      "step": 7148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49773696064949036,
+      "learning_rate": 0.0005800341066798025,
+      "loss": 3.1294,
+      "step": 7149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5385370254516602,
+      "learning_rate": 0.0005800285834621754,
+      "loss": 3.4034,
+      "step": 7150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5338417887687683,
+      "learning_rate": 0.000580023059507007,
+      "loss": 3.1577,
+      "step": 7151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5068842768669128,
+      "learning_rate": 0.0005800175348143116,
+      "loss": 3.5412,
+      "step": 7152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49130746722221375,
+      "learning_rate": 0.0005800120093841039,
+      "loss": 3.4468,
+      "step": 7153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5758814215660095,
+      "learning_rate": 0.0005800064832163985,
+      "loss": 3.3889,
+      "step": 7154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5376800298690796,
+      "learning_rate": 0.00058000095631121,
+      "loss": 3.4071,
+      "step": 7155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5141198635101318,
+      "learning_rate": 0.0005799954286685527,
+      "loss": 3.5247,
+      "step": 7156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5330366492271423,
+      "learning_rate": 0.0005799899002884415,
+      "loss": 3.0816,
+      "step": 7157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4993706941604614,
+      "learning_rate": 0.0005799843711708908,
+      "loss": 3.3283,
+      "step": 7158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.520387589931488,
+      "learning_rate": 0.0005799788413159149,
+      "loss": 3.4417,
+      "step": 7159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5464620590209961,
+      "learning_rate": 0.0005799733107235288,
+      "loss": 3.5151,
+      "step": 7160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5394158363342285,
+      "learning_rate": 0.0005799677793937469,
+      "loss": 3.4495,
+      "step": 7161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.505730390548706,
+      "learning_rate": 0.0005799622473265837,
+      "loss": 3.3232,
+      "step": 7162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5446467995643616,
+      "learning_rate": 0.0005799567145220539,
+      "loss": 3.3166,
+      "step": 7163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5053635835647583,
+      "learning_rate": 0.0005799511809801719,
+      "loss": 3.2405,
+      "step": 7164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5116280317306519,
+      "learning_rate": 0.0005799456467009523,
+      "loss": 3.0451,
+      "step": 7165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.575804591178894,
+      "learning_rate": 0.0005799401116844099,
+      "loss": 3.4171,
+      "step": 7166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5386016368865967,
+      "learning_rate": 0.000579934575930559,
+      "loss": 3.5537,
+      "step": 7167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58470219373703,
+      "learning_rate": 0.0005799290394394144,
+      "loss": 3.4352,
+      "step": 7168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5334919691085815,
+      "learning_rate": 0.0005799235022109906,
+      "loss": 3.4018,
+      "step": 7169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.513592004776001,
+      "learning_rate": 0.000579917964245302,
+      "loss": 3.5992,
+      "step": 7170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5190378427505493,
+      "learning_rate": 0.0005799124255423634,
+      "loss": 3.3433,
+      "step": 7171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5345826148986816,
+      "learning_rate": 0.0005799068861021895,
+      "loss": 3.469,
+      "step": 7172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5253705382347107,
+      "learning_rate": 0.0005799013459247946,
+      "loss": 3.2886,
+      "step": 7173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5313470363616943,
+      "learning_rate": 0.0005798958050101935,
+      "loss": 3.2108,
+      "step": 7174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.545667290687561,
+      "learning_rate": 0.0005798902633584006,
+      "loss": 3.5013,
+      "step": 7175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5079378485679626,
+      "learning_rate": 0.0005798847209694308,
+      "loss": 3.3649,
+      "step": 7176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.555418074131012,
+      "learning_rate": 0.0005798791778432984,
+      "loss": 3.406,
+      "step": 7177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5127460956573486,
+      "learning_rate": 0.0005798736339800181,
+      "loss": 3.333,
+      "step": 7178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5296385288238525,
+      "learning_rate": 0.0005798680893796045,
+      "loss": 3.2033,
+      "step": 7179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5440773367881775,
+      "learning_rate": 0.0005798625440420721,
+      "loss": 3.533,
+      "step": 7180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5760077834129333,
+      "learning_rate": 0.0005798569979674358,
+      "loss": 3.3711,
+      "step": 7181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.550422728061676,
+      "learning_rate": 0.00057985145115571,
+      "loss": 3.3078,
+      "step": 7182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5506702661514282,
+      "learning_rate": 0.0005798459036069094,
+      "loss": 3.2914,
+      "step": 7183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5331199169158936,
+      "learning_rate": 0.0005798403553210484,
+      "loss": 3.2955,
+      "step": 7184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5018191337585449,
+      "learning_rate": 0.0005798348062981419,
+      "loss": 3.2215,
+      "step": 7185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5439038872718811,
+      "learning_rate": 0.0005798292565382042,
+      "loss": 3.3194,
+      "step": 7186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5371338725090027,
+      "learning_rate": 0.0005798237060412502,
+      "loss": 3.4208,
+      "step": 7187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5246415138244629,
+      "learning_rate": 0.0005798181548072943,
+      "loss": 3.2842,
+      "step": 7188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.519935131072998,
+      "learning_rate": 0.0005798126028363514,
+      "loss": 3.1714,
+      "step": 7189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5213209390640259,
+      "learning_rate": 0.0005798070501284359,
+      "loss": 3.1316,
+      "step": 7190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5461356043815613,
+      "learning_rate": 0.0005798014966835625,
+      "loss": 3.3031,
+      "step": 7191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5040096044540405,
+      "learning_rate": 0.0005797959425017457,
+      "loss": 3.3874,
+      "step": 7192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5159637331962585,
+      "learning_rate": 0.0005797903875830004,
+      "loss": 3.2251,
+      "step": 7193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5395895838737488,
+      "learning_rate": 0.000579784831927341,
+      "loss": 3.4672,
+      "step": 7194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5256237387657166,
+      "learning_rate": 0.000579779275534782,
+      "loss": 3.4937,
+      "step": 7195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5334175825119019,
+      "learning_rate": 0.0005797737184053385,
+      "loss": 3.2622,
+      "step": 7196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5539845824241638,
+      "learning_rate": 0.0005797681605390248,
+      "loss": 3.3592,
+      "step": 7197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5329002737998962,
+      "learning_rate": 0.0005797626019358556,
+      "loss": 3.2035,
+      "step": 7198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5471001267433167,
+      "learning_rate": 0.0005797570425958454,
+      "loss": 3.4235,
+      "step": 7199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5086562633514404,
+      "learning_rate": 0.0005797514825190092,
+      "loss": 3.2384,
+      "step": 7200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5643491744995117,
+      "learning_rate": 0.0005797459217053613,
+      "loss": 3.4946,
+      "step": 7201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5237635374069214,
+      "learning_rate": 0.0005797403601549166,
+      "loss": 3.3411,
+      "step": 7202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5038819909095764,
+      "learning_rate": 0.0005797347978676895,
+      "loss": 3.5791,
+      "step": 7203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49819880723953247,
+      "learning_rate": 0.0005797292348436949,
+      "loss": 3.4183,
+      "step": 7204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5045859813690186,
+      "learning_rate": 0.0005797236710829473,
+      "loss": 3.3744,
+      "step": 7205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5460410714149475,
+      "learning_rate": 0.0005797181065854613,
+      "loss": 3.3562,
+      "step": 7206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5919173955917358,
+      "learning_rate": 0.0005797125413512517,
+      "loss": 3.3013,
+      "step": 7207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5470105409622192,
+      "learning_rate": 0.0005797069753803332,
+      "loss": 3.2812,
+      "step": 7208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5546271204948425,
+      "learning_rate": 0.0005797014086727201,
+      "loss": 3.5578,
+      "step": 7209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5407612323760986,
+      "learning_rate": 0.0005796958412284275,
+      "loss": 3.5534,
+      "step": 7210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5482301115989685,
+      "learning_rate": 0.0005796902730474698,
+      "loss": 3.2728,
+      "step": 7211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5478767156600952,
+      "learning_rate": 0.0005796847041298619,
+      "loss": 3.1991,
+      "step": 7212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5303657054901123,
+      "learning_rate": 0.0005796791344756182,
+      "loss": 3.3981,
+      "step": 7213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5501422882080078,
+      "learning_rate": 0.0005796735640847535,
+      "loss": 3.2946,
+      "step": 7214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5230575203895569,
+      "learning_rate": 0.0005796679929572826,
+      "loss": 3.2646,
+      "step": 7215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5465124249458313,
+      "learning_rate": 0.0005796624210932197,
+      "loss": 3.0865,
+      "step": 7216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5589669346809387,
+      "learning_rate": 0.00057965684849258,
+      "loss": 3.1481,
+      "step": 7217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5898823142051697,
+      "learning_rate": 0.000579651275155378,
+      "loss": 3.25,
+      "step": 7218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5371426939964294,
+      "learning_rate": 0.0005796457010816284,
+      "loss": 3.1897,
+      "step": 7219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5486535429954529,
+      "learning_rate": 0.0005796401262713457,
+      "loss": 3.1218,
+      "step": 7220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5335965156555176,
+      "learning_rate": 0.0005796345507245448,
+      "loss": 3.2851,
+      "step": 7221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5595109462738037,
+      "learning_rate": 0.0005796289744412404,
+      "loss": 3.4765,
+      "step": 7222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5471145510673523,
+      "learning_rate": 0.000579623397421447,
+      "loss": 3.4348,
+      "step": 7223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5449061393737793,
+      "learning_rate": 0.0005796178196651794,
+      "loss": 3.2696,
+      "step": 7224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5270853042602539,
+      "learning_rate": 0.0005796122411724523,
+      "loss": 3.5372,
+      "step": 7225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5229194164276123,
+      "learning_rate": 0.0005796066619432803,
+      "loss": 3.3925,
+      "step": 7226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.564228892326355,
+      "learning_rate": 0.0005796010819776782,
+      "loss": 3.325,
+      "step": 7227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55915766954422,
+      "learning_rate": 0.0005795955012756607,
+      "loss": 3.4594,
+      "step": 7228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.513458251953125,
+      "learning_rate": 0.0005795899198372423,
+      "loss": 3.3471,
+      "step": 7229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5892640948295593,
+      "learning_rate": 0.0005795843376624381,
+      "loss": 3.2332,
+      "step": 7230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5395854115486145,
+      "learning_rate": 0.0005795787547512624,
+      "loss": 3.1295,
+      "step": 7231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5873109698295593,
+      "learning_rate": 0.0005795731711037301,
+      "loss": 3.2344,
+      "step": 7232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5463156700134277,
+      "learning_rate": 0.0005795675867198559,
+      "loss": 3.2807,
+      "step": 7233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5152543783187866,
+      "learning_rate": 0.0005795620015996545,
+      "loss": 3.3083,
+      "step": 7234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.630656361579895,
+      "learning_rate": 0.0005795564157431405,
+      "loss": 3.5325,
+      "step": 7235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5535502433776855,
+      "learning_rate": 0.0005795508291503288,
+      "loss": 3.4074,
+      "step": 7236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5285792946815491,
+      "learning_rate": 0.0005795452418212339,
+      "loss": 3.3968,
+      "step": 7237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.548895537853241,
+      "learning_rate": 0.0005795396537558707,
+      "loss": 3.3346,
+      "step": 7238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5185685157775879,
+      "learning_rate": 0.0005795340649542539,
+      "loss": 3.5812,
+      "step": 7239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5098907351493835,
+      "learning_rate": 0.0005795284754163981,
+      "loss": 3.3364,
+      "step": 7240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5207344889640808,
+      "learning_rate": 0.0005795228851423182,
+      "loss": 3.1888,
+      "step": 7241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5404914021492004,
+      "learning_rate": 0.0005795172941320287,
+      "loss": 3.4547,
+      "step": 7242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6177796125411987,
+      "learning_rate": 0.0005795117023855446,
+      "loss": 3.2094,
+      "step": 7243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546385049819946,
+      "learning_rate": 0.0005795061099028802,
+      "loss": 3.4063,
+      "step": 7244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5809313058853149,
+      "learning_rate": 0.0005795005166840507,
+      "loss": 3.3204,
+      "step": 7245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6135513186454773,
+      "learning_rate": 0.0005794949227290705,
+      "loss": 3.2965,
+      "step": 7246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5403055548667908,
+      "learning_rate": 0.0005794893280379546,
+      "loss": 3.3973,
+      "step": 7247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5268582105636597,
+      "learning_rate": 0.0005794837326107175,
+      "loss": 3.4849,
+      "step": 7248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5323728919029236,
+      "learning_rate": 0.0005794781364473741,
+      "loss": 3.5033,
+      "step": 7249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5301493406295776,
+      "learning_rate": 0.000579472539547939,
+      "loss": 3.3231,
+      "step": 7250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5442898273468018,
+      "learning_rate": 0.0005794669419124271,
+      "loss": 3.1438,
+      "step": 7251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6318520903587341,
+      "learning_rate": 0.0005794613435408531,
+      "loss": 3.1192,
+      "step": 7252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5194680094718933,
+      "learning_rate": 0.0005794557444332316,
+      "loss": 3.3424,
+      "step": 7253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5083394646644592,
+      "learning_rate": 0.0005794501445895774,
+      "loss": 3.1855,
+      "step": 7254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.512860119342804,
+      "learning_rate": 0.0005794445440099054,
+      "loss": 3.291,
+      "step": 7255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5411530137062073,
+      "learning_rate": 0.0005794389426942302,
+      "loss": 3.3483,
+      "step": 7256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5959362387657166,
+      "learning_rate": 0.0005794333406425667,
+      "loss": 3.4056,
+      "step": 7257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.596243143081665,
+      "learning_rate": 0.0005794277378549296,
+      "loss": 3.4625,
+      "step": 7258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.523287832736969,
+      "learning_rate": 0.0005794221343313334,
+      "loss": 3.4901,
+      "step": 7259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5272862315177917,
+      "learning_rate": 0.0005794165300717932,
+      "loss": 3.3664,
+      "step": 7260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5093209147453308,
+      "learning_rate": 0.0005794109250763236,
+      "loss": 3.3563,
+      "step": 7261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5650598406791687,
+      "learning_rate": 0.0005794053193449394,
+      "loss": 3.0961,
+      "step": 7262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5295237898826599,
+      "learning_rate": 0.0005793997128776554,
+      "loss": 3.4869,
+      "step": 7263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5343896150588989,
+      "learning_rate": 0.0005793941056744863,
+      "loss": 3.316,
+      "step": 7264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5440590977668762,
+      "learning_rate": 0.000579388497735447,
+      "loss": 3.2155,
+      "step": 7265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4939824938774109,
+      "learning_rate": 0.000579382889060552,
+      "loss": 3.3011,
+      "step": 7266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49455365538597107,
+      "learning_rate": 0.0005793772796498163,
+      "loss": 3.4065,
+      "step": 7267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5292657613754272,
+      "learning_rate": 0.0005793716695032546,
+      "loss": 3.4599,
+      "step": 7268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5179243683815002,
+      "learning_rate": 0.0005793660586208818,
+      "loss": 3.5574,
+      "step": 7269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5624231100082397,
+      "learning_rate": 0.0005793604470027124,
+      "loss": 3.2575,
+      "step": 7270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5110817551612854,
+      "learning_rate": 0.0005793548346487614,
+      "loss": 3.4008,
+      "step": 7271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5481024384498596,
+      "learning_rate": 0.0005793492215590435,
+      "loss": 3.1985,
+      "step": 7272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5263863205909729,
+      "learning_rate": 0.0005793436077335736,
+      "loss": 3.2327,
+      "step": 7273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5095751881599426,
+      "learning_rate": 0.0005793379931723664,
+      "loss": 3.5152,
+      "step": 7274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5422204732894897,
+      "learning_rate": 0.0005793323778754367,
+      "loss": 3.341,
+      "step": 7275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5029591917991638,
+      "learning_rate": 0.0005793267618427991,
+      "loss": 3.4303,
+      "step": 7276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5158606767654419,
+      "learning_rate": 0.0005793211450744688,
+      "loss": 3.4689,
+      "step": 7277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.565763533115387,
+      "learning_rate": 0.0005793155275704601,
+      "loss": 3.2644,
+      "step": 7278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5302228331565857,
+      "learning_rate": 0.0005793099093307883,
+      "loss": 3.4085,
+      "step": 7279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5519428849220276,
+      "learning_rate": 0.0005793042903554679,
+      "loss": 3.349,
+      "step": 7280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5244438648223877,
+      "learning_rate": 0.0005792986706445137,
+      "loss": 3.2967,
+      "step": 7281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5649428367614746,
+      "learning_rate": 0.0005792930501979406,
+      "loss": 3.2556,
+      "step": 7282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.512695848941803,
+      "learning_rate": 0.0005792874290157633,
+      "loss": 3.3095,
+      "step": 7283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5236402750015259,
+      "learning_rate": 0.0005792818070979967,
+      "loss": 3.4541,
+      "step": 7284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5257241725921631,
+      "learning_rate": 0.0005792761844446555,
+      "loss": 3.3157,
+      "step": 7285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5090532898902893,
+      "learning_rate": 0.0005792705610557546,
+      "loss": 3.4076,
+      "step": 7286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5210464000701904,
+      "learning_rate": 0.0005792649369313088,
+      "loss": 3.1769,
+      "step": 7287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5331592559814453,
+      "learning_rate": 0.0005792593120713329,
+      "loss": 3.4404,
+      "step": 7288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5243934988975525,
+      "learning_rate": 0.0005792536864758418,
+      "loss": 3.2593,
+      "step": 7289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5213882327079773,
+      "learning_rate": 0.0005792480601448502,
+      "loss": 3.3763,
+      "step": 7290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5196213722229004,
+      "learning_rate": 0.0005792424330783729,
+      "loss": 3.3627,
+      "step": 7291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5319061875343323,
+      "learning_rate": 0.0005792368052764248,
+      "loss": 3.3825,
+      "step": 7292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5190185308456421,
+      "learning_rate": 0.0005792311767390207,
+      "loss": 3.3911,
+      "step": 7293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5538293123245239,
+      "learning_rate": 0.0005792255474661753,
+      "loss": 3.5731,
+      "step": 7294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5294688940048218,
+      "learning_rate": 0.0005792199174579038,
+      "loss": 3.4866,
+      "step": 7295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5441796183586121,
+      "learning_rate": 0.0005792142867142206,
+      "loss": 3.4322,
+      "step": 7296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49350860714912415,
+      "learning_rate": 0.0005792086552351407,
+      "loss": 3.1275,
+      "step": 7297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.569159746170044,
+      "learning_rate": 0.0005792030230206789,
+      "loss": 3.3663,
+      "step": 7298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5132843255996704,
+      "learning_rate": 0.0005791973900708502,
+      "loss": 3.1675,
+      "step": 7299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6056221723556519,
+      "learning_rate": 0.0005791917563856692,
+      "loss": 3.4195,
+      "step": 7300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.532596230506897,
+      "learning_rate": 0.000579186121965151,
+      "loss": 3.183,
+      "step": 7301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5601078867912292,
+      "learning_rate": 0.0005791804868093101,
+      "loss": 3.228,
+      "step": 7302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5013242959976196,
+      "learning_rate": 0.0005791748509181616,
+      "loss": 3.269,
+      "step": 7303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5339405536651611,
+      "learning_rate": 0.0005791692142917203,
+      "loss": 3.416,
+      "step": 7304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5372835993766785,
+      "learning_rate": 0.000579163576930001,
+      "loss": 3.4051,
+      "step": 7305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5756890773773193,
+      "learning_rate": 0.0005791579388330186,
+      "loss": 3.5151,
+      "step": 7306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5298717617988586,
+      "learning_rate": 0.0005791523000007878,
+      "loss": 3.3441,
+      "step": 7307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5848795175552368,
+      "learning_rate": 0.0005791466604333237,
+      "loss": 3.4685,
+      "step": 7308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5432910323143005,
+      "learning_rate": 0.0005791410201306409,
+      "loss": 3.4025,
+      "step": 7309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5559453964233398,
+      "learning_rate": 0.0005791353790927545,
+      "loss": 3.3999,
+      "step": 7310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5492426753044128,
+      "learning_rate": 0.0005791297373196791,
+      "loss": 3.4373,
+      "step": 7311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5173240900039673,
+      "learning_rate": 0.0005791240948114297,
+      "loss": 3.2916,
+      "step": 7312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5640772581100464,
+      "learning_rate": 0.0005791184515680213,
+      "loss": 3.447,
+      "step": 7313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5233004093170166,
+      "learning_rate": 0.0005791128075894685,
+      "loss": 3.5497,
+      "step": 7314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5180065631866455,
+      "learning_rate": 0.0005791071628757862,
+      "loss": 3.3526,
+      "step": 7315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5473960041999817,
+      "learning_rate": 0.0005791015174269895,
+      "loss": 3.1132,
+      "step": 7316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5282230377197266,
+      "learning_rate": 0.0005790958712430931,
+      "loss": 3.4481,
+      "step": 7317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5316844582557678,
+      "learning_rate": 0.0005790902243241119,
+      "loss": 3.4213,
+      "step": 7318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5286957621574402,
+      "learning_rate": 0.0005790845766700608,
+      "loss": 3.5678,
+      "step": 7319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49466997385025024,
+      "learning_rate": 0.0005790789282809545,
+      "loss": 3.1566,
+      "step": 7320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5194963812828064,
+      "learning_rate": 0.0005790732791568081,
+      "loss": 3.2,
+      "step": 7321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5403347611427307,
+      "learning_rate": 0.0005790676292976363,
+      "loss": 3.5393,
+      "step": 7322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5521409511566162,
+      "learning_rate": 0.0005790619787034542,
+      "loss": 3.1492,
+      "step": 7323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5029959678649902,
+      "learning_rate": 0.0005790563273742764,
+      "loss": 3.2783,
+      "step": 7324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5743463039398193,
+      "learning_rate": 0.0005790506753101181,
+      "loss": 3.3343,
+      "step": 7325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5631399154663086,
+      "learning_rate": 0.0005790450225109939,
+      "loss": 3.4623,
+      "step": 7326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5515002608299255,
+      "learning_rate": 0.0005790393689769188,
+      "loss": 3.5293,
+      "step": 7327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5547970533370972,
+      "learning_rate": 0.0005790337147079078,
+      "loss": 3.3389,
+      "step": 7328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5851017832756042,
+      "learning_rate": 0.0005790280597039757,
+      "loss": 3.4403,
+      "step": 7329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5544925332069397,
+      "learning_rate": 0.0005790224039651374,
+      "loss": 3.4946,
+      "step": 7330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5337487459182739,
+      "learning_rate": 0.0005790167474914077,
+      "loss": 3.379,
+      "step": 7331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5269622206687927,
+      "learning_rate": 0.0005790110902828017,
+      "loss": 3.2976,
+      "step": 7332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5398240089416504,
+      "learning_rate": 0.000579005432339334,
+      "loss": 3.4284,
+      "step": 7333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5220947861671448,
+      "learning_rate": 0.0005789997736610199,
+      "loss": 3.2636,
+      "step": 7334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5519540309906006,
+      "learning_rate": 0.000578994114247874,
+      "loss": 3.1985,
+      "step": 7335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.604718029499054,
+      "learning_rate": 0.0005789884540999112,
+      "loss": 3.5494,
+      "step": 7336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5395439863204956,
+      "learning_rate": 0.0005789827932171466,
+      "loss": 3.4658,
+      "step": 7337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5418725609779358,
+      "learning_rate": 0.0005789771315995951,
+      "loss": 3.4172,
+      "step": 7338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5764614343643188,
+      "learning_rate": 0.0005789714692472714,
+      "loss": 3.2426,
+      "step": 7339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5681049227714539,
+      "learning_rate": 0.0005789658061601905,
+      "loss": 3.3262,
+      "step": 7340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5404443144798279,
+      "learning_rate": 0.0005789601423383675,
+      "loss": 3.6112,
+      "step": 7341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5424375534057617,
+      "learning_rate": 0.000578954477781817,
+      "loss": 3.5564,
+      "step": 7342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5254672765731812,
+      "learning_rate": 0.0005789488124905542,
+      "loss": 3.271,
+      "step": 7343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5020593404769897,
+      "learning_rate": 0.000578943146464594,
+      "loss": 3.3267,
+      "step": 7344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5463967323303223,
+      "learning_rate": 0.0005789374797039511,
+      "loss": 3.4191,
+      "step": 7345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5019806623458862,
+      "learning_rate": 0.0005789318122086406,
+      "loss": 3.2696,
+      "step": 7346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5514928698539734,
+      "learning_rate": 0.0005789261439786774,
+      "loss": 3.1895,
+      "step": 7347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.512698769569397,
+      "learning_rate": 0.0005789204750140764,
+      "loss": 3.4189,
+      "step": 7348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5055906772613525,
+      "learning_rate": 0.0005789148053148525,
+      "loss": 3.2841,
+      "step": 7349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.527269721031189,
+      "learning_rate": 0.0005789091348810208,
+      "loss": 3.5163,
+      "step": 7350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5826881527900696,
+      "learning_rate": 0.0005789034637125961,
+      "loss": 3.4445,
+      "step": 7351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5765799283981323,
+      "learning_rate": 0.0005788977918095932,
+      "loss": 3.2363,
+      "step": 7352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5874758362770081,
+      "learning_rate": 0.0005788921191720273,
+      "loss": 3.5972,
+      "step": 7353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5443326234817505,
+      "learning_rate": 0.0005788864457999133,
+      "loss": 3.3818,
+      "step": 7354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5976613759994507,
+      "learning_rate": 0.000578880771693266,
+      "loss": 3.1729,
+      "step": 7355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.571810781955719,
+      "learning_rate": 0.0005788750968521003,
+      "loss": 3.3752,
+      "step": 7356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5518274307250977,
+      "learning_rate": 0.0005788694212764314,
+      "loss": 3.3239,
+      "step": 7357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5681546330451965,
+      "learning_rate": 0.0005788637449662742,
+      "loss": 3.3544,
+      "step": 7358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5265949368476868,
+      "learning_rate": 0.0005788580679216434,
+      "loss": 3.3936,
+      "step": 7359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5548690557479858,
+      "learning_rate": 0.0005788523901425541,
+      "loss": 3.2901,
+      "step": 7360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5343828201293945,
+      "learning_rate": 0.0005788467116290214,
+      "loss": 3.2131,
+      "step": 7361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5076448321342468,
+      "learning_rate": 0.0005788410323810602,
+      "loss": 3.3964,
+      "step": 7362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49037474393844604,
+      "learning_rate": 0.0005788353523986852,
+      "loss": 3.2415,
+      "step": 7363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5263360142707825,
+      "learning_rate": 0.0005788296716819116,
+      "loss": 3.4041,
+      "step": 7364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5553779006004333,
+      "learning_rate": 0.0005788239902307543,
+      "loss": 3.3971,
+      "step": 7365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5806408524513245,
+      "learning_rate": 0.0005788183080452283,
+      "loss": 3.1735,
+      "step": 7366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5436879396438599,
+      "learning_rate": 0.0005788126251253486,
+      "loss": 3.5634,
+      "step": 7367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5230838656425476,
+      "learning_rate": 0.00057880694147113,
+      "loss": 3.2974,
+      "step": 7368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.630568265914917,
+      "learning_rate": 0.0005788012570825875,
+      "loss": 3.2782,
+      "step": 7369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5829651951789856,
+      "learning_rate": 0.0005787955719597362,
+      "loss": 3.0793,
+      "step": 7370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5027291774749756,
+      "learning_rate": 0.0005787898861025912,
+      "loss": 3.4173,
+      "step": 7371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5039538145065308,
+      "learning_rate": 0.0005787841995111671,
+      "loss": 3.2034,
+      "step": 7372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5734065175056458,
+      "learning_rate": 0.0005787785121854791,
+      "loss": 3.459,
+      "step": 7373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5559008717536926,
+      "learning_rate": 0.0005787728241255422,
+      "loss": 3.1911,
+      "step": 7374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5591650605201721,
+      "learning_rate": 0.0005787671353313712,
+      "loss": 3.1949,
+      "step": 7375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52269047498703,
+      "learning_rate": 0.0005787614458029813,
+      "loss": 3.3678,
+      "step": 7376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5631672143936157,
+      "learning_rate": 0.0005787557555403875,
+      "loss": 3.3729,
+      "step": 7377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5107264518737793,
+      "learning_rate": 0.0005787500645436047,
+      "loss": 3.3954,
+      "step": 7378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5732921361923218,
+      "learning_rate": 0.0005787443728126478,
+      "loss": 3.4472,
+      "step": 7379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5226930975914001,
+      "learning_rate": 0.0005787386803475318,
+      "loss": 3.4802,
+      "step": 7380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5663575530052185,
+      "learning_rate": 0.0005787329871482717,
+      "loss": 3.3462,
+      "step": 7381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5493120551109314,
+      "learning_rate": 0.0005787272932148827,
+      "loss": 3.2889,
+      "step": 7382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5059551000595093,
+      "learning_rate": 0.0005787215985473797,
+      "loss": 3.3501,
+      "step": 7383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5528244972229004,
+      "learning_rate": 0.0005787159031457776,
+      "loss": 3.3196,
+      "step": 7384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.519409716129303,
+      "learning_rate": 0.0005787102070100914,
+      "loss": 3.2856,
+      "step": 7385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5072803497314453,
+      "learning_rate": 0.0005787045101403362,
+      "loss": 3.4766,
+      "step": 7386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5262261033058167,
+      "learning_rate": 0.0005786988125365269,
+      "loss": 3.3746,
+      "step": 7387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5350049734115601,
+      "learning_rate": 0.0005786931141986786,
+      "loss": 3.2329,
+      "step": 7388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5463743805885315,
+      "learning_rate": 0.0005786874151268062,
+      "loss": 3.2722,
+      "step": 7389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5149025321006775,
+      "learning_rate": 0.0005786817153209249,
+      "loss": 3.243,
+      "step": 7390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5367695093154907,
+      "learning_rate": 0.0005786760147810496,
+      "loss": 3.4026,
+      "step": 7391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625688433647156,
+      "learning_rate": 0.0005786703135071952,
+      "loss": 3.3234,
+      "step": 7392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5635698437690735,
+      "learning_rate": 0.000578664611499377,
+      "loss": 3.3182,
+      "step": 7393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5578081607818604,
+      "learning_rate": 0.0005786589087576097,
+      "loss": 3.1293,
+      "step": 7394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5266342163085938,
+      "learning_rate": 0.0005786532052819085,
+      "loss": 3.4332,
+      "step": 7395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.524929940700531,
+      "learning_rate": 0.0005786475010722883,
+      "loss": 3.1655,
+      "step": 7396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.521431028842926,
+      "learning_rate": 0.0005786417961287643,
+      "loss": 3.2723,
+      "step": 7397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5528025031089783,
+      "learning_rate": 0.0005786360904513515,
+      "loss": 3.495,
+      "step": 7398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47876498103141785,
+      "learning_rate": 0.0005786303840400647,
+      "loss": 3.2928,
+      "step": 7399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5592080354690552,
+      "learning_rate": 0.0005786246768949193,
+      "loss": 3.1918,
+      "step": 7400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5398855209350586,
+      "learning_rate": 0.0005786189690159299,
+      "loss": 3.3853,
+      "step": 7401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5718713402748108,
+      "learning_rate": 0.0005786132604031119,
+      "loss": 3.3739,
+      "step": 7402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5837875604629517,
+      "learning_rate": 0.0005786075510564801,
+      "loss": 3.3774,
+      "step": 7403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5221856236457825,
+      "learning_rate": 0.0005786018409760497,
+      "loss": 3.4436,
+      "step": 7404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5874915719032288,
+      "learning_rate": 0.0005785961301618356,
+      "loss": 3.3879,
+      "step": 7405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5608260035514832,
+      "learning_rate": 0.000578590418613853,
+      "loss": 3.3331,
+      "step": 7406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5786241292953491,
+      "learning_rate": 0.0005785847063321168,
+      "loss": 3.2905,
+      "step": 7407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5414553284645081,
+      "learning_rate": 0.0005785789933166422,
+      "loss": 3.2087,
+      "step": 7408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5226951837539673,
+      "learning_rate": 0.000578573279567444,
+      "loss": 3.4015,
+      "step": 7409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5430212616920471,
+      "learning_rate": 0.0005785675650845374,
+      "loss": 3.208,
+      "step": 7410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5425985455513,
+      "learning_rate": 0.0005785618498679375,
+      "loss": 3.3109,
+      "step": 7411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5451919436454773,
+      "learning_rate": 0.0005785561339176593,
+      "loss": 3.1937,
+      "step": 7412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5355823636054993,
+      "learning_rate": 0.0005785504172337178,
+      "loss": 3.2868,
+      "step": 7413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4789738059043884,
+      "learning_rate": 0.0005785446998161282,
+      "loss": 3.4312,
+      "step": 7414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5494072437286377,
+      "learning_rate": 0.0005785389816649054,
+      "loss": 3.481,
+      "step": 7415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.529240608215332,
+      "learning_rate": 0.0005785332627800645,
+      "loss": 3.231,
+      "step": 7416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5651838779449463,
+      "learning_rate": 0.0005785275431616207,
+      "loss": 3.3466,
+      "step": 7417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5189810395240784,
+      "learning_rate": 0.0005785218228095889,
+      "loss": 3.2454,
+      "step": 7418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5465260744094849,
+      "learning_rate": 0.0005785161017239842,
+      "loss": 3.3845,
+      "step": 7419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5511559844017029,
+      "learning_rate": 0.0005785103799048218,
+      "loss": 3.4355,
+      "step": 7420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5183231830596924,
+      "learning_rate": 0.0005785046573521165,
+      "loss": 3.3434,
+      "step": 7421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5259350538253784,
+      "learning_rate": 0.0005784989340658837,
+      "loss": 3.3595,
+      "step": 7422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5619202852249146,
+      "learning_rate": 0.000578493210046138,
+      "loss": 3.4631,
+      "step": 7423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5505042672157288,
+      "learning_rate": 0.000578487485292895,
+      "loss": 3.3808,
+      "step": 7424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5139535665512085,
+      "learning_rate": 0.0005784817598061696,
+      "loss": 3.2009,
+      "step": 7425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.580228328704834,
+      "learning_rate": 0.0005784760335859769,
+      "loss": 3.1475,
+      "step": 7426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.535790205001831,
+      "learning_rate": 0.0005784703066323317,
+      "loss": 3.2759,
+      "step": 7427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5042516589164734,
+      "learning_rate": 0.0005784645789452494,
+      "loss": 3.1972,
+      "step": 7428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5909183025360107,
+      "learning_rate": 0.000578458850524745,
+      "loss": 3.3301,
+      "step": 7429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5318119525909424,
+      "learning_rate": 0.0005784531213708336,
+      "loss": 3.4631,
+      "step": 7430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4836869537830353,
+      "learning_rate": 0.0005784473914835302,
+      "loss": 3.5986,
+      "step": 7431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234432816505432,
+      "learning_rate": 0.0005784416608628501,
+      "loss": 3.1308,
+      "step": 7432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6105952858924866,
+      "learning_rate": 0.0005784359295088081,
+      "loss": 3.3292,
+      "step": 7433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.600591242313385,
+      "learning_rate": 0.0005784301974214195,
+      "loss": 3.3737,
+      "step": 7434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48450472950935364,
+      "learning_rate": 0.0005784244646006993,
+      "loss": 3.2875,
+      "step": 7435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6669772863388062,
+      "learning_rate": 0.0005784187310466628,
+      "loss": 3.239,
+      "step": 7436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5766210556030273,
+      "learning_rate": 0.0005784129967593249,
+      "loss": 3.2816,
+      "step": 7437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5472173690795898,
+      "learning_rate": 0.0005784072617387006,
+      "loss": 3.3769,
+      "step": 7438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5345730185508728,
+      "learning_rate": 0.0005784015259848053,
+      "loss": 3.4177,
+      "step": 7439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5472462177276611,
+      "learning_rate": 0.0005783957894976538,
+      "loss": 3.3872,
+      "step": 7440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5371647477149963,
+      "learning_rate": 0.0005783900522772615,
+      "loss": 3.4841,
+      "step": 7441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5290404558181763,
+      "learning_rate": 0.0005783843143236433,
+      "loss": 3.2963,
+      "step": 7442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5215094685554504,
+      "learning_rate": 0.0005783785756368145,
+      "loss": 3.3373,
+      "step": 7443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61955726146698,
+      "learning_rate": 0.0005783728362167901,
+      "loss": 3.3818,
+      "step": 7444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5387982130050659,
+      "learning_rate": 0.0005783670960635851,
+      "loss": 3.3528,
+      "step": 7445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.553604245185852,
+      "learning_rate": 0.0005783613551772149,
+      "loss": 3.3234,
+      "step": 7446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5161128640174866,
+      "learning_rate": 0.0005783556135576942,
+      "loss": 3.2385,
+      "step": 7447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5468826293945312,
+      "learning_rate": 0.0005783498712050386,
+      "loss": 3.3768,
+      "step": 7448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5205458402633667,
+      "learning_rate": 0.000578344128119263,
+      "loss": 3.4234,
+      "step": 7449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5659615397453308,
+      "learning_rate": 0.0005783383843003825,
+      "loss": 3.2642,
+      "step": 7450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6167935132980347,
+      "learning_rate": 0.0005783326397484123,
+      "loss": 3.4605,
+      "step": 7451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5576913356781006,
+      "learning_rate": 0.0005783268944633675,
+      "loss": 3.4278,
+      "step": 7452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5447071194648743,
+      "learning_rate": 0.0005783211484452632,
+      "loss": 3.2907,
+      "step": 7453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49073418974876404,
+      "learning_rate": 0.0005783154016941145,
+      "loss": 3.3534,
+      "step": 7454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5985206365585327,
+      "learning_rate": 0.0005783096542099366,
+      "loss": 3.1996,
+      "step": 7455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5504509806632996,
+      "learning_rate": 0.0005783039059927448,
+      "loss": 3.3993,
+      "step": 7456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5634692907333374,
+      "learning_rate": 0.0005782981570425539,
+      "loss": 3.3579,
+      "step": 7457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49616754055023193,
+      "learning_rate": 0.0005782924073593792,
+      "loss": 3.3781,
+      "step": 7458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.533212423324585,
+      "learning_rate": 0.0005782866569432359,
+      "loss": 3.1977,
+      "step": 7459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5390735864639282,
+      "learning_rate": 0.0005782809057941392,
+      "loss": 3.358,
+      "step": 7460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5744284391403198,
+      "learning_rate": 0.000578275153912104,
+      "loss": 3.3979,
+      "step": 7461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5401898622512817,
+      "learning_rate": 0.0005782694012971458,
+      "loss": 3.557,
+      "step": 7462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5606440305709839,
+      "learning_rate": 0.0005782636479492793,
+      "loss": 3.2477,
+      "step": 7463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48706990480422974,
+      "learning_rate": 0.0005782578938685201,
+      "loss": 3.5734,
+      "step": 7464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5521231293678284,
+      "learning_rate": 0.0005782521390548831,
+      "loss": 3.3121,
+      "step": 7465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5380448698997498,
+      "learning_rate": 0.0005782463835083834,
+      "loss": 3.4478,
+      "step": 7466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5551396608352661,
+      "learning_rate": 0.0005782406272290364,
+      "loss": 3.2938,
+      "step": 7467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5721957683563232,
+      "learning_rate": 0.0005782348702168572,
+      "loss": 3.3066,
+      "step": 7468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5749885439872742,
+      "learning_rate": 0.0005782291124718608,
+      "loss": 3.3865,
+      "step": 7469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5412680506706238,
+      "learning_rate": 0.0005782233539940625,
+      "loss": 3.1974,
+      "step": 7470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.530744731426239,
+      "learning_rate": 0.0005782175947834774,
+      "loss": 3.42,
+      "step": 7471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5663665533065796,
+      "learning_rate": 0.0005782118348401207,
+      "loss": 3.4735,
+      "step": 7472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5773357152938843,
+      "learning_rate": 0.0005782060741640075,
+      "loss": 3.287,
+      "step": 7473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5321745872497559,
+      "learning_rate": 0.000578200312755153,
+      "loss": 3.2528,
+      "step": 7474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.543752908706665,
+      "learning_rate": 0.0005781945506135726,
+      "loss": 3.1304,
+      "step": 7475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5263063311576843,
+      "learning_rate": 0.0005781887877392812,
+      "loss": 3.5356,
+      "step": 7476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5132307410240173,
+      "learning_rate": 0.0005781830241322942,
+      "loss": 3.0732,
+      "step": 7477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5528070330619812,
+      "learning_rate": 0.0005781772597926266,
+      "loss": 3.5516,
+      "step": 7478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5620551109313965,
+      "learning_rate": 0.0005781714947202935,
+      "loss": 3.5392,
+      "step": 7479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5856842398643494,
+      "learning_rate": 0.0005781657289153103,
+      "loss": 3.4985,
+      "step": 7480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49245211482048035,
+      "learning_rate": 0.0005781599623776922,
+      "loss": 3.2934,
+      "step": 7481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5112742781639099,
+      "learning_rate": 0.0005781541951074541,
+      "loss": 3.39,
+      "step": 7482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.567556619644165,
+      "learning_rate": 0.0005781484271046115,
+      "loss": 3.3748,
+      "step": 7483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5500913262367249,
+      "learning_rate": 0.0005781426583691794,
+      "loss": 3.1794,
+      "step": 7484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.588733434677124,
+      "learning_rate": 0.0005781368889011731,
+      "loss": 3.3915,
+      "step": 7485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5132182240486145,
+      "learning_rate": 0.0005781311187006079,
+      "loss": 3.2548,
+      "step": 7486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5105894804000854,
+      "learning_rate": 0.0005781253477674987,
+      "loss": 3.4658,
+      "step": 7487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5359848737716675,
+      "learning_rate": 0.0005781195761018609,
+      "loss": 3.2688,
+      "step": 7488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5214976072311401,
+      "learning_rate": 0.0005781138037037096,
+      "loss": 3.3225,
+      "step": 7489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5228166580200195,
+      "learning_rate": 0.0005781080305730603,
+      "loss": 3.2782,
+      "step": 7490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5464044809341431,
+      "learning_rate": 0.0005781022567099277,
+      "loss": 3.2469,
+      "step": 7491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5573117733001709,
+      "learning_rate": 0.0005780964821143274,
+      "loss": 3.0187,
+      "step": 7492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5702837109565735,
+      "learning_rate": 0.0005780907067862744,
+      "loss": 3.2428,
+      "step": 7493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6031702160835266,
+      "learning_rate": 0.0005780849307257842,
+      "loss": 3.4579,
+      "step": 7494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5179644823074341,
+      "learning_rate": 0.0005780791539328716,
+      "loss": 3.1622,
+      "step": 7495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5007571578025818,
+      "learning_rate": 0.0005780733764075521,
+      "loss": 3.4954,
+      "step": 7496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5171201825141907,
+      "learning_rate": 0.0005780675981498409,
+      "loss": 3.379,
+      "step": 7497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5398955941200256,
+      "learning_rate": 0.0005780618191597531,
+      "loss": 3.3197,
+      "step": 7498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5368615984916687,
+      "learning_rate": 0.000578056039437304,
+      "loss": 3.5917,
+      "step": 7499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5254148244857788,
+      "learning_rate": 0.0005780502589825087,
+      "loss": 3.3423,
+      "step": 7500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49917274713516235,
+      "learning_rate": 0.0005780444777953827,
+      "loss": 3.3729,
+      "step": 7501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.579575777053833,
+      "learning_rate": 0.000578038695875941,
+      "loss": 3.1839,
+      "step": 7502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5500216484069824,
+      "learning_rate": 0.0005780329132241989,
+      "loss": 3.4427,
+      "step": 7503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5117226243019104,
+      "learning_rate": 0.0005780271298401715,
+      "loss": 3.043,
+      "step": 7504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.57242751121521,
+      "learning_rate": 0.0005780213457238743,
+      "loss": 3.2203,
+      "step": 7505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.538989245891571,
+      "learning_rate": 0.0005780155608753223,
+      "loss": 3.4584,
+      "step": 7506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5599838495254517,
+      "learning_rate": 0.0005780097752945307,
+      "loss": 3.4769,
+      "step": 7507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5817453265190125,
+      "learning_rate": 0.0005780039889815151,
+      "loss": 3.4676,
+      "step": 7508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5717339515686035,
+      "learning_rate": 0.0005779982019362903,
+      "loss": 3.1948,
+      "step": 7509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5515158772468567,
+      "learning_rate": 0.0005779924141588718,
+      "loss": 3.3845,
+      "step": 7510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5656583905220032,
+      "learning_rate": 0.0005779866256492749,
+      "loss": 3.3043,
+      "step": 7511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5461367964744568,
+      "learning_rate": 0.0005779808364075146,
+      "loss": 3.2265,
+      "step": 7512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5367287397384644,
+      "learning_rate": 0.0005779750464336062,
+      "loss": 3.3777,
+      "step": 7513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5329413414001465,
+      "learning_rate": 0.0005779692557275651,
+      "loss": 3.3883,
+      "step": 7514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5147910714149475,
+      "learning_rate": 0.0005779634642894066,
+      "loss": 3.1705,
+      "step": 7515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5768789052963257,
+      "learning_rate": 0.0005779576721191457,
+      "loss": 3.2288,
+      "step": 7516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5086293816566467,
+      "learning_rate": 0.0005779518792167978,
+      "loss": 3.3029,
+      "step": 7517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5314181447029114,
+      "learning_rate": 0.0005779460855823782,
+      "loss": 3.5192,
+      "step": 7518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5063535571098328,
+      "learning_rate": 0.0005779402912159021,
+      "loss": 3.3481,
+      "step": 7519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5236873030662537,
+      "learning_rate": 0.0005779344961173847,
+      "loss": 3.3957,
+      "step": 7520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5629249215126038,
+      "learning_rate": 0.0005779287002868413,
+      "loss": 3.2262,
+      "step": 7521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6165677905082703,
+      "learning_rate": 0.0005779229037242873,
+      "loss": 3.184,
+      "step": 7522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5311621427536011,
+      "learning_rate": 0.0005779171064297378,
+      "loss": 3.3883,
+      "step": 7523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5682266354560852,
+      "learning_rate": 0.0005779113084032082,
+      "loss": 3.4386,
+      "step": 7524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5127180814743042,
+      "learning_rate": 0.0005779055096447136,
+      "loss": 3.3266,
+      "step": 7525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49226856231689453,
+      "learning_rate": 0.0005778997101542694,
+      "loss": 3.3149,
+      "step": 7526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5309831500053406,
+      "learning_rate": 0.0005778939099318908,
+      "loss": 3.2658,
+      "step": 7527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5468692779541016,
+      "learning_rate": 0.0005778881089775933,
+      "loss": 3.3354,
+      "step": 7528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5492680072784424,
+      "learning_rate": 0.0005778823072913918,
+      "loss": 3.3651,
+      "step": 7529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5209020972251892,
+      "learning_rate": 0.0005778765048733019,
+      "loss": 3.3828,
+      "step": 7530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5251026153564453,
+      "learning_rate": 0.0005778707017233387,
+      "loss": 3.4814,
+      "step": 7531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5343338847160339,
+      "learning_rate": 0.0005778648978415176,
+      "loss": 3.3883,
+      "step": 7532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5628687143325806,
+      "learning_rate": 0.0005778590932278537,
+      "loss": 3.4574,
+      "step": 7533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5103763341903687,
+      "learning_rate": 0.0005778532878823625,
+      "loss": 3.1286,
+      "step": 7534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5265288949012756,
+      "learning_rate": 0.0005778474818050593,
+      "loss": 3.2203,
+      "step": 7535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5499712824821472,
+      "learning_rate": 0.0005778416749959592,
+      "loss": 3.1883,
+      "step": 7536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5209925174713135,
+      "learning_rate": 0.0005778358674550778,
+      "loss": 3.2346,
+      "step": 7537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5306970477104187,
+      "learning_rate": 0.00057783005918243,
+      "loss": 3.3883,
+      "step": 7538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.506366491317749,
+      "learning_rate": 0.0005778242501780313,
+      "loss": 3.2689,
+      "step": 7539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5261942744255066,
+      "learning_rate": 0.0005778184404418971,
+      "loss": 3.5414,
+      "step": 7540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.539635181427002,
+      "learning_rate": 0.0005778126299740425,
+      "loss": 3.2619,
+      "step": 7541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5381528735160828,
+      "learning_rate": 0.0005778068187744829,
+      "loss": 3.2378,
+      "step": 7542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5434399843215942,
+      "learning_rate": 0.0005778010068432336,
+      "loss": 3.1842,
+      "step": 7543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5563649535179138,
+      "learning_rate": 0.0005777951941803099,
+      "loss": 3.3198,
+      "step": 7544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5185420513153076,
+      "learning_rate": 0.0005777893807857273,
+      "loss": 3.3414,
+      "step": 7545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5348244905471802,
+      "learning_rate": 0.0005777835666595007,
+      "loss": 3.2746,
+      "step": 7546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5372639894485474,
+      "learning_rate": 0.0005777777518016458,
+      "loss": 3.2459,
+      "step": 7547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5764162540435791,
+      "learning_rate": 0.0005777719362121777,
+      "loss": 3.3808,
+      "step": 7548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.533909022808075,
+      "learning_rate": 0.0005777661198911118,
+      "loss": 3.4382,
+      "step": 7549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5675473809242249,
+      "learning_rate": 0.0005777603028384634,
+      "loss": 3.1793,
+      "step": 7550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5419423580169678,
+      "learning_rate": 0.0005777544850542477,
+      "loss": 3.309,
+      "step": 7551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5336681008338928,
+      "learning_rate": 0.0005777486665384802,
+      "loss": 3.4056,
+      "step": 7552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5690992474555969,
+      "learning_rate": 0.0005777428472911763,
+      "loss": 3.4831,
+      "step": 7553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5622754693031311,
+      "learning_rate": 0.000577737027312351,
+      "loss": 3.3529,
+      "step": 7554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5477057695388794,
+      "learning_rate": 0.0005777312066020199,
+      "loss": 3.4004,
+      "step": 7555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5189183950424194,
+      "learning_rate": 0.0005777253851601984,
+      "loss": 3.319,
+      "step": 7556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5486992001533508,
+      "learning_rate": 0.0005777195629869015,
+      "loss": 3.3856,
+      "step": 7557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5244187712669373,
+      "learning_rate": 0.0005777137400821448,
+      "loss": 3.2891,
+      "step": 7558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5623378157615662,
+      "learning_rate": 0.0005777079164459436,
+      "loss": 3.3083,
+      "step": 7559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8148874044418335,
+      "learning_rate": 0.000577702092078313,
+      "loss": 3.3492,
+      "step": 7560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5145521759986877,
+      "learning_rate": 0.0005776962669792687,
+      "loss": 3.2571,
+      "step": 7561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5764961242675781,
+      "learning_rate": 0.0005776904411488259,
+      "loss": 3.4779,
+      "step": 7562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.518203616142273,
+      "learning_rate": 0.0005776846145869997,
+      "loss": 3.4448,
+      "step": 7563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6169568300247192,
+      "learning_rate": 0.0005776787872938059,
+      "loss": 3.3586,
+      "step": 7564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47566235065460205,
+      "learning_rate": 0.0005776729592692596,
+      "loss": 3.5285,
+      "step": 7565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5244961380958557,
+      "learning_rate": 0.0005776671305133761,
+      "loss": 3.3522,
+      "step": 7566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5160118937492371,
+      "learning_rate": 0.0005776613010261708,
+      "loss": 3.2136,
+      "step": 7567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6301264762878418,
+      "learning_rate": 0.0005776554708076591,
+      "loss": 3.1727,
+      "step": 7568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48011279106140137,
+      "learning_rate": 0.0005776496398578562,
+      "loss": 3.3688,
+      "step": 7569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5248762369155884,
+      "learning_rate": 0.0005776438081767778,
+      "loss": 3.159,
+      "step": 7570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5746804475784302,
+      "learning_rate": 0.000577637975764439,
+      "loss": 3.5644,
+      "step": 7571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5382371544837952,
+      "learning_rate": 0.0005776321426208551,
+      "loss": 3.3489,
+      "step": 7572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5431787371635437,
+      "learning_rate": 0.0005776263087460416,
+      "loss": 3.0477,
+      "step": 7573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5550752878189087,
+      "learning_rate": 0.0005776204741400138,
+      "loss": 3.3814,
+      "step": 7574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5394303798675537,
+      "learning_rate": 0.0005776146388027872,
+      "loss": 3.32,
+      "step": 7575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5841784477233887,
+      "learning_rate": 0.000577608802734377,
+      "loss": 3.4877,
+      "step": 7576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5106384754180908,
+      "learning_rate": 0.0005776029659347986,
+      "loss": 3.2927,
+      "step": 7577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5529481172561646,
+      "learning_rate": 0.0005775971284040675,
+      "loss": 3.414,
+      "step": 7578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5255731344223022,
+      "learning_rate": 0.0005775912901421989,
+      "loss": 3.2171,
+      "step": 7579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4985384941101074,
+      "learning_rate": 0.0005775854511492084,
+      "loss": 3.4437,
+      "step": 7580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5763261318206787,
+      "learning_rate": 0.0005775796114251111,
+      "loss": 3.3172,
+      "step": 7581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5372400879859924,
+      "learning_rate": 0.0005775737709699227,
+      "loss": 3.5312,
+      "step": 7582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5711855888366699,
+      "learning_rate": 0.0005775679297836582,
+      "loss": 3.3675,
+      "step": 7583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5218439698219299,
+      "learning_rate": 0.0005775620878663333,
+      "loss": 3.2173,
+      "step": 7584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5360540151596069,
+      "learning_rate": 0.0005775562452179632,
+      "loss": 3.2145,
+      "step": 7585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5195031762123108,
+      "learning_rate": 0.0005775504018385635,
+      "loss": 3.5188,
+      "step": 7586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5842971801757812,
+      "learning_rate": 0.0005775445577281494,
+      "loss": 3.3013,
+      "step": 7587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5489726662635803,
+      "learning_rate": 0.0005775387128867363,
+      "loss": 3.1806,
+      "step": 7588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.529917299747467,
+      "learning_rate": 0.0005775328673143398,
+      "loss": 3.1544,
+      "step": 7589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5455032587051392,
+      "learning_rate": 0.0005775270210109751,
+      "loss": 3.2344,
+      "step": 7590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5468717217445374,
+      "learning_rate": 0.0005775211739766574,
+      "loss": 3.3418,
+      "step": 7591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5684468746185303,
+      "learning_rate": 0.0005775153262114025,
+      "loss": 3.2833,
+      "step": 7592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.502007246017456,
+      "learning_rate": 0.0005775094777152257,
+      "loss": 3.2564,
+      "step": 7593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5651172995567322,
+      "learning_rate": 0.0005775036284881422,
+      "loss": 3.4975,
+      "step": 7594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.525607705116272,
+      "learning_rate": 0.0005774977785301677,
+      "loss": 3.3124,
+      "step": 7595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5294498205184937,
+      "learning_rate": 0.0005774919278413174,
+      "loss": 3.3201,
+      "step": 7596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5207985639572144,
+      "learning_rate": 0.0005774860764216068,
+      "loss": 3.4086,
+      "step": 7597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5538904666900635,
+      "learning_rate": 0.0005774802242710513,
+      "loss": 3.2833,
+      "step": 7598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5455402135848999,
+      "learning_rate": 0.0005774743713896661,
+      "loss": 3.3564,
+      "step": 7599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48715144395828247,
+      "learning_rate": 0.000577468517777467,
+      "loss": 3.3534,
+      "step": 7600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.520661473274231,
+      "learning_rate": 0.0005774626634344692,
+      "loss": 3.5091,
+      "step": 7601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5360146760940552,
+      "learning_rate": 0.000577456808360688,
+      "loss": 3.4542,
+      "step": 7602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.511243462562561,
+      "learning_rate": 0.0005774509525561392,
+      "loss": 3.2027,
+      "step": 7603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5435131788253784,
+      "learning_rate": 0.0005774450960208378,
+      "loss": 3.2947,
+      "step": 7604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5104213953018188,
+      "learning_rate": 0.0005774392387547995,
+      "loss": 3.3127,
+      "step": 7605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5034766793251038,
+      "learning_rate": 0.0005774333807580395,
+      "loss": 3.3856,
+      "step": 7606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5869839787483215,
+      "learning_rate": 0.0005774275220305735,
+      "loss": 3.308,
+      "step": 7607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4974309206008911,
+      "learning_rate": 0.0005774216625724168,
+      "loss": 3.4021,
+      "step": 7608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5040432810783386,
+      "learning_rate": 0.0005774158023835848,
+      "loss": 2.9991,
+      "step": 7609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5178181529045105,
+      "learning_rate": 0.000577409941464093,
+      "loss": 3.2486,
+      "step": 7610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5407399535179138,
+      "learning_rate": 0.0005774040798139567,
+      "loss": 3.3858,
+      "step": 7611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5059396028518677,
+      "learning_rate": 0.0005773982174331915,
+      "loss": 3.431,
+      "step": 7612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5154821276664734,
+      "learning_rate": 0.0005773923543218128,
+      "loss": 3.2791,
+      "step": 7613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5808987021446228,
+      "learning_rate": 0.000577386490479836,
+      "loss": 3.3297,
+      "step": 7614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5294091701507568,
+      "learning_rate": 0.0005773806259072766,
+      "loss": 3.3628,
+      "step": 7615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49198436737060547,
+      "learning_rate": 0.00057737476060415,
+      "loss": 3.441,
+      "step": 7616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5616235733032227,
+      "learning_rate": 0.0005773688945704717,
+      "loss": 3.4603,
+      "step": 7617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5396645665168762,
+      "learning_rate": 0.0005773630278062571,
+      "loss": 3.3925,
+      "step": 7618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5407871007919312,
+      "learning_rate": 0.0005773571603115216,
+      "loss": 3.4043,
+      "step": 7619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5107448697090149,
+      "learning_rate": 0.0005773512920862808,
+      "loss": 3.4727,
+      "step": 7620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5386175513267517,
+      "learning_rate": 0.00057734542313055,
+      "loss": 3.2539,
+      "step": 7621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5494099855422974,
+      "learning_rate": 0.0005773395534443448,
+      "loss": 3.3771,
+      "step": 7622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5173628926277161,
+      "learning_rate": 0.0005773336830276805,
+      "loss": 3.379,
+      "step": 7623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5712496638298035,
+      "learning_rate": 0.0005773278118805727,
+      "loss": 3.1727,
+      "step": 7624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.524523138999939,
+      "learning_rate": 0.0005773219400030369,
+      "loss": 3.1427,
+      "step": 7625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5556202530860901,
+      "learning_rate": 0.0005773160673950883,
+      "loss": 3.4616,
+      "step": 7626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5599361658096313,
+      "learning_rate": 0.0005773101940567427,
+      "loss": 3.3032,
+      "step": 7627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5155829191207886,
+      "learning_rate": 0.0005773043199880154,
+      "loss": 3.3193,
+      "step": 7628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5319119095802307,
+      "learning_rate": 0.0005772984451889219,
+      "loss": 3.3039,
+      "step": 7629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5668798089027405,
+      "learning_rate": 0.0005772925696594776,
+      "loss": 3.3281,
+      "step": 7630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5063618421554565,
+      "learning_rate": 0.000577286693399698,
+      "loss": 3.2928,
+      "step": 7631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5319681763648987,
+      "learning_rate": 0.0005772808164095988,
+      "loss": 3.2887,
+      "step": 7632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5006822347640991,
+      "learning_rate": 0.0005772749386891951,
+      "loss": 3.5356,
+      "step": 7633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5074585676193237,
+      "learning_rate": 0.0005772690602385026,
+      "loss": 3.4013,
+      "step": 7634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5565813779830933,
+      "learning_rate": 0.0005772631810575369,
+      "loss": 3.4353,
+      "step": 7635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.549435019493103,
+      "learning_rate": 0.0005772573011463131,
+      "loss": 3.0845,
+      "step": 7636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5290880799293518,
+      "learning_rate": 0.0005772514205048472,
+      "loss": 3.459,
+      "step": 7637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5667818784713745,
+      "learning_rate": 0.0005772455391331542,
+      "loss": 3.2927,
+      "step": 7638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5045775175094604,
+      "learning_rate": 0.0005772396570312499,
+      "loss": 3.2672,
+      "step": 7639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5503361821174622,
+      "learning_rate": 0.0005772337741991497,
+      "loss": 3.2494,
+      "step": 7640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5238841772079468,
+      "learning_rate": 0.0005772278906368692,
+      "loss": 3.2805,
+      "step": 7641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5158736109733582,
+      "learning_rate": 0.0005772220063444236,
+      "loss": 3.3447,
+      "step": 7642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5552991032600403,
+      "learning_rate": 0.0005772161213218286,
+      "loss": 3.1852,
+      "step": 7643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5120370388031006,
+      "learning_rate": 0.0005772102355690998,
+      "loss": 3.2797,
+      "step": 7644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47802504897117615,
+      "learning_rate": 0.0005772043490862525,
+      "loss": 3.2269,
+      "step": 7645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5293357372283936,
+      "learning_rate": 0.0005771984618733024,
+      "loss": 3.5585,
+      "step": 7646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5576381683349609,
+      "learning_rate": 0.0005771925739302649,
+      "loss": 3.5653,
+      "step": 7647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6383746266365051,
+      "learning_rate": 0.0005771866852571554,
+      "loss": 3.4087,
+      "step": 7648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5233350992202759,
+      "learning_rate": 0.0005771807958539895,
+      "loss": 3.5551,
+      "step": 7649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6100677847862244,
+      "learning_rate": 0.0005771749057207828,
+      "loss": 3.2505,
+      "step": 7650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49944281578063965,
+      "learning_rate": 0.0005771690148575508,
+      "loss": 3.3978,
+      "step": 7651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5267223119735718,
+      "learning_rate": 0.0005771631232643088,
+      "loss": 3.1858,
+      "step": 7652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5045062899589539,
+      "learning_rate": 0.0005771572309410726,
+      "loss": 3.4002,
+      "step": 7653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5398784279823303,
+      "learning_rate": 0.0005771513378878576,
+      "loss": 3.6054,
+      "step": 7654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5356148481369019,
+      "learning_rate": 0.0005771454441046793,
+      "loss": 3.5048,
+      "step": 7655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5511022806167603,
+      "learning_rate": 0.0005771395495915531,
+      "loss": 3.1728,
+      "step": 7656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4920971691608429,
+      "learning_rate": 0.0005771336543484948,
+      "loss": 3.2441,
+      "step": 7657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5204853415489197,
+      "learning_rate": 0.0005771277583755198,
+      "loss": 3.4215,
+      "step": 7658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5585867762565613,
+      "learning_rate": 0.0005771218616726436,
+      "loss": 3.2756,
+      "step": 7659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5565081238746643,
+      "learning_rate": 0.0005771159642398817,
+      "loss": 3.4225,
+      "step": 7660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5324299335479736,
+      "learning_rate": 0.0005771100660772497,
+      "loss": 3.127,
+      "step": 7661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.516318678855896,
+      "learning_rate": 0.0005771041671847632,
+      "loss": 3.4088,
+      "step": 7662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5677664279937744,
+      "learning_rate": 0.0005770982675624374,
+      "loss": 3.3326,
+      "step": 7663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5322689414024353,
+      "learning_rate": 0.0005770923672102883,
+      "loss": 3.3283,
+      "step": 7664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6116513013839722,
+      "learning_rate": 0.0005770864661283312,
+      "loss": 3.3218,
+      "step": 7665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5484536290168762,
+      "learning_rate": 0.0005770805643165819,
+      "loss": 3.2955,
+      "step": 7666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5231142044067383,
+      "learning_rate": 0.0005770746617750553,
+      "loss": 3.4475,
+      "step": 7667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5095486640930176,
+      "learning_rate": 0.0005770687585037676,
+      "loss": 3.4287,
+      "step": 7668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.544468343257904,
+      "learning_rate": 0.0005770628545027341,
+      "loss": 3.3266,
+      "step": 7669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5009661912918091,
+      "learning_rate": 0.0005770569497719703,
+      "loss": 3.4781,
+      "step": 7670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.510226845741272,
+      "learning_rate": 0.0005770510443114918,
+      "loss": 3.3332,
+      "step": 7671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5290097594261169,
+      "learning_rate": 0.0005770451381213144,
+      "loss": 3.4185,
+      "step": 7672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5284835696220398,
+      "learning_rate": 0.0005770392312014532,
+      "loss": 3.2657,
+      "step": 7673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4978867471218109,
+      "learning_rate": 0.000577033323551924,
+      "loss": 3.3484,
+      "step": 7674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5280537009239197,
+      "learning_rate": 0.0005770274151727424,
+      "loss": 3.3948,
+      "step": 7675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5177137851715088,
+      "learning_rate": 0.0005770215060639239,
+      "loss": 3.2166,
+      "step": 7676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5431517362594604,
+      "learning_rate": 0.000577015596225484,
+      "loss": 3.1223,
+      "step": 7677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5485773682594299,
+      "learning_rate": 0.0005770096856574383,
+      "loss": 3.2481,
+      "step": 7678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5501081347465515,
+      "learning_rate": 0.0005770037743598025,
+      "loss": 3.5521,
+      "step": 7679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5404449701309204,
+      "learning_rate": 0.000576997862332592,
+      "loss": 3.4757,
+      "step": 7680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5793304443359375,
+      "learning_rate": 0.0005769919495758225,
+      "loss": 3.1084,
+      "step": 7681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.559897243976593,
+      "learning_rate": 0.0005769860360895096,
+      "loss": 3.0807,
+      "step": 7682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5876379609107971,
+      "learning_rate": 0.0005769801218736686,
+      "loss": 3.2023,
+      "step": 7683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.548350989818573,
+      "learning_rate": 0.0005769742069283154,
+      "loss": 3.2818,
+      "step": 7684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5384724140167236,
+      "learning_rate": 0.0005769682912534653,
+      "loss": 3.2043,
+      "step": 7685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5407583713531494,
+      "learning_rate": 0.0005769623748491342,
+      "loss": 3.3275,
+      "step": 7686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5203924179077148,
+      "learning_rate": 0.0005769564577153374,
+      "loss": 3.3591,
+      "step": 7687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5495003461837769,
+      "learning_rate": 0.0005769505398520907,
+      "loss": 3.2688,
+      "step": 7688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5348663926124573,
+      "learning_rate": 0.0005769446212594094,
+      "loss": 3.3437,
+      "step": 7689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5982791185379028,
+      "learning_rate": 0.0005769387019373094,
+      "loss": 3.398,
+      "step": 7690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5527102947235107,
+      "learning_rate": 0.0005769327818858062,
+      "loss": 3.4327,
+      "step": 7691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5997951626777649,
+      "learning_rate": 0.0005769268611049152,
+      "loss": 3.4901,
+      "step": 7692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5478419065475464,
+      "learning_rate": 0.0005769209395946522,
+      "loss": 3.3142,
+      "step": 7693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49550652503967285,
+      "learning_rate": 0.0005769150173550328,
+      "loss": 3.5203,
+      "step": 7694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5617064833641052,
+      "learning_rate": 0.0005769090943860724,
+      "loss": 3.5058,
+      "step": 7695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4888365566730499,
+      "learning_rate": 0.0005769031706877869,
+      "loss": 3.4375,
+      "step": 7696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5538713335990906,
+      "learning_rate": 0.0005768972462601916,
+      "loss": 3.3487,
+      "step": 7697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5713244080543518,
+      "learning_rate": 0.0005768913211033024,
+      "loss": 3.2288,
+      "step": 7698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5322604775428772,
+      "learning_rate": 0.0005768853952171346,
+      "loss": 3.5587,
+      "step": 7699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5296763777732849,
+      "learning_rate": 0.000576879468601704,
+      "loss": 3.3886,
+      "step": 7700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.581745982170105,
+      "learning_rate": 0.0005768735412570262,
+      "loss": 3.3111,
+      "step": 7701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5159407258033752,
+      "learning_rate": 0.0005768676131831168,
+      "loss": 3.2317,
+      "step": 7702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.527847945690155,
+      "learning_rate": 0.0005768616843799913,
+      "loss": 3.2123,
+      "step": 7703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5626876354217529,
+      "learning_rate": 0.0005768557548476654,
+      "loss": 3.1141,
+      "step": 7704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5151354670524597,
+      "learning_rate": 0.0005768498245861548,
+      "loss": 3.2901,
+      "step": 7705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.538934051990509,
+      "learning_rate": 0.000576843893595475,
+      "loss": 3.3142,
+      "step": 7706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5391485095024109,
+      "learning_rate": 0.0005768379618756417,
+      "loss": 3.1658,
+      "step": 7707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5370870232582092,
+      "learning_rate": 0.0005768320294266705,
+      "loss": 3.2983,
+      "step": 7708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5493991374969482,
+      "learning_rate": 0.0005768260962485769,
+      "loss": 3.3358,
+      "step": 7709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5690867900848389,
+      "learning_rate": 0.0005768201623413768,
+      "loss": 3.4018,
+      "step": 7710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5145018100738525,
+      "learning_rate": 0.0005768142277050856,
+      "loss": 3.3555,
+      "step": 7711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.526140034198761,
+      "learning_rate": 0.000576808292339719,
+      "loss": 3.4113,
+      "step": 7712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5112320780754089,
+      "learning_rate": 0.0005768023562452926,
+      "loss": 3.3697,
+      "step": 7713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5290682315826416,
+      "learning_rate": 0.0005767964194218221,
+      "loss": 3.5244,
+      "step": 7714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5239760279655457,
+      "learning_rate": 0.0005767904818693231,
+      "loss": 3.4218,
+      "step": 7715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5848803520202637,
+      "learning_rate": 0.0005767845435878113,
+      "loss": 3.298,
+      "step": 7716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5328184366226196,
+      "learning_rate": 0.0005767786045773021,
+      "loss": 3.2251,
+      "step": 7717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5391929745674133,
+      "learning_rate": 0.0005767726648378115,
+      "loss": 3.2584,
+      "step": 7718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5395445823669434,
+      "learning_rate": 0.0005767667243693548,
+      "loss": 3.3085,
+      "step": 7719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5276076197624207,
+      "learning_rate": 0.0005767607831719479,
+      "loss": 3.3174,
+      "step": 7720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4999656677246094,
+      "learning_rate": 0.0005767548412456064,
+      "loss": 3.3946,
+      "step": 7721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5640438795089722,
+      "learning_rate": 0.0005767488985903459,
+      "loss": 3.4772,
+      "step": 7722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5889086723327637,
+      "learning_rate": 0.000576742955206182,
+      "loss": 3.4004,
+      "step": 7723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5023418068885803,
+      "learning_rate": 0.0005767370110931306,
+      "loss": 3.4566,
+      "step": 7724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5705462694168091,
+      "learning_rate": 0.000576731066251207,
+      "loss": 3.2547,
+      "step": 7725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6652705073356628,
+      "learning_rate": 0.000576725120680427,
+      "loss": 3.1061,
+      "step": 7726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5656840205192566,
+      "learning_rate": 0.0005767191743808064,
+      "loss": 3.1697,
+      "step": 7727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5446447134017944,
+      "learning_rate": 0.0005767132273523606,
+      "loss": 3.2713,
+      "step": 7728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5088135004043579,
+      "learning_rate": 0.0005767072795951056,
+      "loss": 3.439,
+      "step": 7729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.573800802230835,
+      "learning_rate": 0.0005767013311090567,
+      "loss": 3.5233,
+      "step": 7730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5233531594276428,
+      "learning_rate": 0.0005766953818942299,
+      "loss": 3.368,
+      "step": 7731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5602032542228699,
+      "learning_rate": 0.0005766894319506406,
+      "loss": 3.5343,
+      "step": 7732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5444862246513367,
+      "learning_rate": 0.0005766834812783047,
+      "loss": 3.3899,
+      "step": 7733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5610346794128418,
+      "learning_rate": 0.0005766775298772377,
+      "loss": 3.263,
+      "step": 7734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5168694257736206,
+      "learning_rate": 0.0005766715777474553,
+      "loss": 3.3323,
+      "step": 7735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5316835641860962,
+      "learning_rate": 0.0005766656248889732,
+      "loss": 3.4178,
+      "step": 7736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5672079920768738,
+      "learning_rate": 0.0005766596713018072,
+      "loss": 3.3652,
+      "step": 7737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5037599205970764,
+      "learning_rate": 0.0005766537169859728,
+      "loss": 3.3349,
+      "step": 7738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.538493812084198,
+      "learning_rate": 0.0005766477619414858,
+      "loss": 3.4587,
+      "step": 7739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5200104713439941,
+      "learning_rate": 0.0005766418061683618,
+      "loss": 3.3614,
+      "step": 7740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5429470539093018,
+      "learning_rate": 0.0005766358496666165,
+      "loss": 3.2853,
+      "step": 7741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5296519994735718,
+      "learning_rate": 0.0005766298924362656,
+      "loss": 3.3077,
+      "step": 7742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5855775475502014,
+      "learning_rate": 0.0005766239344773249,
+      "loss": 3.3771,
+      "step": 7743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6540366411209106,
+      "learning_rate": 0.0005766179757898098,
+      "loss": 3.432,
+      "step": 7744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5268910527229309,
+      "learning_rate": 0.0005766120163737364,
+      "loss": 3.3832,
+      "step": 7745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5402625799179077,
+      "learning_rate": 0.00057660605622912,
+      "loss": 3.2677,
+      "step": 7746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.545888364315033,
+      "learning_rate": 0.0005766000953559767,
+      "loss": 3.1326,
+      "step": 7747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5113720893859863,
+      "learning_rate": 0.0005765941337543218,
+      "loss": 3.3382,
+      "step": 7748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5178648829460144,
+      "learning_rate": 0.0005765881714241714,
+      "loss": 3.3942,
+      "step": 7749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5671051740646362,
+      "learning_rate": 0.0005765822083655407,
+      "loss": 3.3048,
+      "step": 7750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.501137375831604,
+      "learning_rate": 0.0005765762445784459,
+      "loss": 3.2334,
+      "step": 7751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5226545333862305,
+      "learning_rate": 0.0005765702800629023,
+      "loss": 3.3512,
+      "step": 7752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5014787912368774,
+      "learning_rate": 0.000576564314818926,
+      "loss": 3.318,
+      "step": 7753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5254672169685364,
+      "learning_rate": 0.0005765583488465324,
+      "loss": 3.3825,
+      "step": 7754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5159661173820496,
+      "learning_rate": 0.0005765523821457374,
+      "loss": 3.3095,
+      "step": 7755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5051149725914001,
+      "learning_rate": 0.0005765464147165566,
+      "loss": 3.2341,
+      "step": 7756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5197758078575134,
+      "learning_rate": 0.0005765404465590059,
+      "loss": 3.2489,
+      "step": 7757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5094234943389893,
+      "learning_rate": 0.0005765344776731008,
+      "loss": 3.2445,
+      "step": 7758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5342963337898254,
+      "learning_rate": 0.000576528508058857,
+      "loss": 3.5146,
+      "step": 7759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5151779055595398,
+      "learning_rate": 0.0005765225377162904,
+      "loss": 3.1479,
+      "step": 7760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5261240601539612,
+      "learning_rate": 0.0005765165666454167,
+      "loss": 3.0576,
+      "step": 7761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5607361793518066,
+      "learning_rate": 0.0005765105948462516,
+      "loss": 3.211,
+      "step": 7762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.534424364566803,
+      "learning_rate": 0.0005765046223188108,
+      "loss": 3.3502,
+      "step": 7763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5433931946754456,
+      "learning_rate": 0.00057649864906311,
+      "loss": 3.3541,
+      "step": 7764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5713850855827332,
+      "learning_rate": 0.0005764926750791649,
+      "loss": 3.1849,
+      "step": 7765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5363174676895142,
+      "learning_rate": 0.0005764867003669914,
+      "loss": 3.297,
+      "step": 7766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5349758267402649,
+      "learning_rate": 0.0005764807249266052,
+      "loss": 3.2353,
+      "step": 7767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5053718686103821,
+      "learning_rate": 0.0005764747487580218,
+      "loss": 3.0331,
+      "step": 7768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5225927233695984,
+      "learning_rate": 0.0005764687718612572,
+      "loss": 3.2927,
+      "step": 7769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5191579461097717,
+      "learning_rate": 0.0005764627942363271,
+      "loss": 3.2522,
+      "step": 7770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.530640721321106,
+      "learning_rate": 0.0005764568158832472,
+      "loss": 3.1888,
+      "step": 7771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5546700358390808,
+      "learning_rate": 0.0005764508368020333,
+      "loss": 2.9471,
+      "step": 7772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.541641354560852,
+      "learning_rate": 0.0005764448569927009,
+      "loss": 3.1619,
+      "step": 7773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5144984722137451,
+      "learning_rate": 0.0005764388764552662,
+      "loss": 3.1562,
+      "step": 7774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5483419299125671,
+      "learning_rate": 0.0005764328951897446,
+      "loss": 3.5068,
+      "step": 7775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5148709416389465,
+      "learning_rate": 0.0005764269131961519,
+      "loss": 3.3056,
+      "step": 7776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.535964846611023,
+      "learning_rate": 0.000576420930474504,
+      "loss": 3.4626,
+      "step": 7777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5585061311721802,
+      "learning_rate": 0.0005764149470248166,
+      "loss": 3.3747,
+      "step": 7778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5796385407447815,
+      "learning_rate": 0.0005764089628471054,
+      "loss": 3.3358,
+      "step": 7779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5621188282966614,
+      "learning_rate": 0.000576402977941386,
+      "loss": 3.5115,
+      "step": 7780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5454630255699158,
+      "learning_rate": 0.0005763969923076746,
+      "loss": 3.4531,
+      "step": 7781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5305492877960205,
+      "learning_rate": 0.0005763910059459865,
+      "loss": 3.3743,
+      "step": 7782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5111590027809143,
+      "learning_rate": 0.0005763850188563378,
+      "loss": 3.1609,
+      "step": 7783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.520412266254425,
+      "learning_rate": 0.0005763790310387441,
+      "loss": 3.5552,
+      "step": 7784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5311888456344604,
+      "learning_rate": 0.0005763730424932213,
+      "loss": 3.3538,
+      "step": 7785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5347929000854492,
+      "learning_rate": 0.0005763670532197851,
+      "loss": 3.4564,
+      "step": 7786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5463716983795166,
+      "learning_rate": 0.0005763610632184512,
+      "loss": 3.3186,
+      "step": 7787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5599271655082703,
+      "learning_rate": 0.0005763550724892355,
+      "loss": 3.3339,
+      "step": 7788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.545688271522522,
+      "learning_rate": 0.0005763490810321537,
+      "loss": 3.3977,
+      "step": 7789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5224574208259583,
+      "learning_rate": 0.0005763430888472217,
+      "loss": 3.2594,
+      "step": 7790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49941286444664,
+      "learning_rate": 0.0005763370959344549,
+      "loss": 3.185,
+      "step": 7791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5086252689361572,
+      "learning_rate": 0.0005763311022938696,
+      "loss": 3.3799,
+      "step": 7792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5339295268058777,
+      "learning_rate": 0.0005763251079254814,
+      "loss": 3.5027,
+      "step": 7793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5528396368026733,
+      "learning_rate": 0.000576319112829306,
+      "loss": 3.3822,
+      "step": 7794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5613994002342224,
+      "learning_rate": 0.0005763131170053591,
+      "loss": 3.439,
+      "step": 7795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5107088685035706,
+      "learning_rate": 0.0005763071204536568,
+      "loss": 3.243,
+      "step": 7796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5415818691253662,
+      "learning_rate": 0.0005763011231742146,
+      "loss": 3.1863,
+      "step": 7797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.550011396408081,
+      "learning_rate": 0.0005762951251670485,
+      "loss": 3.3939,
+      "step": 7798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5543647408485413,
+      "learning_rate": 0.0005762891264321742,
+      "loss": 3.4366,
+      "step": 7799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5226086974143982,
+      "learning_rate": 0.0005762831269696075,
+      "loss": 3.3135,
+      "step": 7800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5125158429145813,
+      "learning_rate": 0.0005762771267793642,
+      "loss": 3.5191,
+      "step": 7801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.485068678855896,
+      "learning_rate": 0.0005762711258614602,
+      "loss": 3.5,
+      "step": 7802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49382463097572327,
+      "learning_rate": 0.0005762651242159111,
+      "loss": 3.3202,
+      "step": 7803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5490413904190063,
+      "learning_rate": 0.0005762591218427328,
+      "loss": 3.1602,
+      "step": 7804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.496092826128006,
+      "learning_rate": 0.0005762531187419413,
+      "loss": 3.511,
+      "step": 7805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5007348656654358,
+      "learning_rate": 0.0005762471149135522,
+      "loss": 3.2596,
+      "step": 7806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5107442736625671,
+      "learning_rate": 0.0005762411103575813,
+      "loss": 3.3431,
+      "step": 7807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5263485908508301,
+      "learning_rate": 0.0005762351050740445,
+      "loss": 3.3426,
+      "step": 7808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5867173671722412,
+      "learning_rate": 0.0005762290990629576,
+      "loss": 3.3641,
+      "step": 7809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5284026861190796,
+      "learning_rate": 0.0005762230923243365,
+      "loss": 3.2272,
+      "step": 7810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5352510809898376,
+      "learning_rate": 0.0005762170848581968,
+      "loss": 3.447,
+      "step": 7811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4992859959602356,
+      "learning_rate": 0.0005762110766645546,
+      "loss": 3.4186,
+      "step": 7812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.514480471611023,
+      "learning_rate": 0.0005762050677434254,
+      "loss": 3.2831,
+      "step": 7813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5168455243110657,
+      "learning_rate": 0.0005761990580948254,
+      "loss": 3.3549,
+      "step": 7814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5284212827682495,
+      "learning_rate": 0.0005761930477187701,
+      "loss": 3.3554,
+      "step": 7815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5353116393089294,
+      "learning_rate": 0.0005761870366152755,
+      "loss": 3.2887,
+      "step": 7816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5134118795394897,
+      "learning_rate": 0.0005761810247843574,
+      "loss": 3.4749,
+      "step": 7817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49037235975265503,
+      "learning_rate": 0.0005761750122260317,
+      "loss": 3.1841,
+      "step": 7818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5198505520820618,
+      "learning_rate": 0.0005761689989403141,
+      "loss": 3.4457,
+      "step": 7819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5264098048210144,
+      "learning_rate": 0.0005761629849272205,
+      "loss": 3.3493,
+      "step": 7820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5358461141586304,
+      "learning_rate": 0.0005761569701867668,
+      "loss": 3.309,
+      "step": 7821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5349216461181641,
+      "learning_rate": 0.0005761509547189688,
+      "loss": 3.2732,
+      "step": 7822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.498106986284256,
+      "learning_rate": 0.0005761449385238422,
+      "loss": 3.1092,
+      "step": 7823
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5381754040718079,
+      "learning_rate": 0.0005761389216014031,
+      "loss": 3.3537,
+      "step": 7824
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.535024106502533,
+      "learning_rate": 0.0005761329039516671,
+      "loss": 3.269,
+      "step": 7825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5659539103507996,
+      "learning_rate": 0.0005761268855746503,
+      "loss": 3.3276,
+      "step": 7826
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5511694550514221,
+      "learning_rate": 0.0005761208664703684,
+      "loss": 3.3624,
+      "step": 7827
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5096381902694702,
+      "learning_rate": 0.0005761148466388373,
+      "loss": 3.2467,
+      "step": 7828
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5487003326416016,
+      "learning_rate": 0.0005761088260800728,
+      "loss": 3.361,
+      "step": 7829
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5437702536582947,
+      "learning_rate": 0.0005761028047940907,
+      "loss": 3.4645,
+      "step": 7830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5526220798492432,
+      "learning_rate": 0.0005760967827809072,
+      "loss": 3.1801,
+      "step": 7831
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5293183326721191,
+      "learning_rate": 0.0005760907600405377,
+      "loss": 3.2272,
+      "step": 7832
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5198914408683777,
+      "learning_rate": 0.0005760847365729984,
+      "loss": 3.3003,
+      "step": 7833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5536202788352966,
+      "learning_rate": 0.0005760787123783049,
+      "loss": 2.9891,
+      "step": 7834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5597622990608215,
+      "learning_rate": 0.0005760726874564732,
+      "loss": 3.0627,
+      "step": 7835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5665072202682495,
+      "learning_rate": 0.0005760666618075192,
+      "loss": 3.2544,
+      "step": 7836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6050432324409485,
+      "learning_rate": 0.0005760606354314588,
+      "loss": 3.3163,
+      "step": 7837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5545769929885864,
+      "learning_rate": 0.0005760546083283079,
+      "loss": 3.1321,
+      "step": 7838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5636788010597229,
+      "learning_rate": 0.0005760485804980821,
+      "loss": 3.1955,
+      "step": 7839
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5258702635765076,
+      "learning_rate": 0.0005760425519407976,
+      "loss": 3.274,
+      "step": 7840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5311688184738159,
+      "learning_rate": 0.0005760365226564701,
+      "loss": 3.2431,
+      "step": 7841
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5252965092658997,
+      "learning_rate": 0.0005760304926451155,
+      "loss": 3.3003,
+      "step": 7842
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5277718901634216,
+      "learning_rate": 0.0005760244619067498,
+      "loss": 3.3049,
+      "step": 7843
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5588723421096802,
+      "learning_rate": 0.0005760184304413887,
+      "loss": 3.0993,
+      "step": 7844
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5348266363143921,
+      "learning_rate": 0.0005760123982490481,
+      "loss": 3.3248,
+      "step": 7845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5221610069274902,
+      "learning_rate": 0.0005760063653297441,
+      "loss": 3.1574,
+      "step": 7846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5521312355995178,
+      "learning_rate": 0.0005760003316834924,
+      "loss": 3.2215,
+      "step": 7847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.570512056350708,
+      "learning_rate": 0.0005759942973103089,
+      "loss": 3.478,
+      "step": 7848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5131711363792419,
+      "learning_rate": 0.0005759882622102096,
+      "loss": 3.3638,
+      "step": 7849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5566141605377197,
+      "learning_rate": 0.0005759822263832103,
+      "loss": 3.0577,
+      "step": 7850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5762760043144226,
+      "learning_rate": 0.0005759761898293269,
+      "loss": 3.275,
+      "step": 7851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5507886409759521,
+      "learning_rate": 0.0005759701525485754,
+      "loss": 3.1812,
+      "step": 7852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5555328726768494,
+      "learning_rate": 0.0005759641145409716,
+      "loss": 3.3363,
+      "step": 7853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5610153079032898,
+      "learning_rate": 0.0005759580758065315,
+      "loss": 3.3007,
+      "step": 7854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5262787938117981,
+      "learning_rate": 0.0005759520363452709,
+      "loss": 3.5069,
+      "step": 7855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5241034030914307,
+      "learning_rate": 0.0005759459961572057,
+      "loss": 3.3264,
+      "step": 7856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5599466562271118,
+      "learning_rate": 0.0005759399552423518,
+      "loss": 3.4544,
+      "step": 7857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5439044833183289,
+      "learning_rate": 0.0005759339136007253,
+      "loss": 3.3155,
+      "step": 7858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5432597994804382,
+      "learning_rate": 0.0005759278712323419,
+      "loss": 3.2769,
+      "step": 7859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.550878643989563,
+      "learning_rate": 0.0005759218281372175,
+      "loss": 3.5132,
+      "step": 7860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5362837910652161,
+      "learning_rate": 0.0005759157843153683,
+      "loss": 3.2445,
+      "step": 7861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5157976150512695,
+      "learning_rate": 0.00057590973976681,
+      "loss": 3.3654,
+      "step": 7862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5195072293281555,
+      "learning_rate": 0.0005759036944915585,
+      "loss": 3.2124,
+      "step": 7863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5618158578872681,
+      "learning_rate": 0.0005758976484896296,
+      "loss": 3.1182,
+      "step": 7864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5115683078765869,
+      "learning_rate": 0.0005758916017610396,
+      "loss": 3.4023,
+      "step": 7865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5243606567382812,
+      "learning_rate": 0.0005758855543058042,
+      "loss": 3.2927,
+      "step": 7866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5262304544448853,
+      "learning_rate": 0.0005758795061239393,
+      "loss": 3.3996,
+      "step": 7867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5342805981636047,
+      "learning_rate": 0.0005758734572154609,
+      "loss": 3.443,
+      "step": 7868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49876853823661804,
+      "learning_rate": 0.0005758674075803848,
+      "loss": 3.5409,
+      "step": 7869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.534772515296936,
+      "learning_rate": 0.000575861357218727,
+      "loss": 3.3427,
+      "step": 7870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5307782292366028,
+      "learning_rate": 0.0005758553061305038,
+      "loss": 3.3184,
+      "step": 7871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5947427153587341,
+      "learning_rate": 0.0005758492543157305,
+      "loss": 3.1305,
+      "step": 7872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5112699270248413,
+      "learning_rate": 0.0005758432017744235,
+      "loss": 3.2861,
+      "step": 7873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5101796388626099,
+      "learning_rate": 0.0005758371485065986,
+      "loss": 3.3975,
+      "step": 7874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5383381247520447,
+      "learning_rate": 0.0005758310945122717,
+      "loss": 3.2371,
+      "step": 7875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5568742156028748,
+      "learning_rate": 0.0005758250397914587,
+      "loss": 3.3188,
+      "step": 7876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5260884761810303,
+      "learning_rate": 0.0005758189843441757,
+      "loss": 3.3591,
+      "step": 7877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5004744529724121,
+      "learning_rate": 0.0005758129281704386,
+      "loss": 3.2018,
+      "step": 7878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5695815682411194,
+      "learning_rate": 0.0005758068712702633,
+      "loss": 3.0783,
+      "step": 7879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5590498447418213,
+      "learning_rate": 0.0005758008136436658,
+      "loss": 3.3908,
+      "step": 7880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5661531090736389,
+      "learning_rate": 0.0005757947552906621,
+      "loss": 3.2036,
+      "step": 7881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5372530817985535,
+      "learning_rate": 0.0005757886962112679,
+      "loss": 3.3096,
+      "step": 7882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5631973147392273,
+      "learning_rate": 0.0005757826364054995,
+      "loss": 3.3011,
+      "step": 7883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5350229740142822,
+      "learning_rate": 0.0005757765758733727,
+      "loss": 3.4572,
+      "step": 7884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5403389930725098,
+      "learning_rate": 0.0005757705146149034,
+      "loss": 3.3266,
+      "step": 7885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5325131416320801,
+      "learning_rate": 0.0005757644526301078,
+      "loss": 3.3954,
+      "step": 7886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5278269052505493,
+      "learning_rate": 0.0005757583899190016,
+      "loss": 3.3926,
+      "step": 7887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5290004014968872,
+      "learning_rate": 0.0005757523264816009,
+      "loss": 3.5282,
+      "step": 7888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5585773587226868,
+      "learning_rate": 0.0005757462623179215,
+      "loss": 3.5418,
+      "step": 7889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5469244122505188,
+      "learning_rate": 0.0005757401974279796,
+      "loss": 3.2753,
+      "step": 7890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6067792177200317,
+      "learning_rate": 0.0005757341318117911,
+      "loss": 3.3368,
+      "step": 7891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5292524099349976,
+      "learning_rate": 0.0005757280654693721,
+      "loss": 3.2878,
+      "step": 7892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6546759605407715,
+      "learning_rate": 0.0005757219984007382,
+      "loss": 3.1904,
+      "step": 7893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5419657826423645,
+      "learning_rate": 0.0005757159306059057,
+      "loss": 3.4508,
+      "step": 7894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5196870565414429,
+      "learning_rate": 0.0005757098620848905,
+      "loss": 3.4328,
+      "step": 7895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5152660608291626,
+      "learning_rate": 0.0005757037928377087,
+      "loss": 3.3568,
+      "step": 7896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5874919891357422,
+      "learning_rate": 0.0005756977228643761,
+      "loss": 3.4204,
+      "step": 7897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5220939517021179,
+      "learning_rate": 0.0005756916521649088,
+      "loss": 3.5503,
+      "step": 7898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5331912636756897,
+      "learning_rate": 0.0005756855807393227,
+      "loss": 3.2373,
+      "step": 7899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5303639769554138,
+      "learning_rate": 0.0005756795085876338,
+      "loss": 3.2296,
+      "step": 7900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.522990882396698,
+      "learning_rate": 0.0005756734357098581,
+      "loss": 3.2013,
+      "step": 7901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5241119861602783,
+      "learning_rate": 0.0005756673621060117,
+      "loss": 3.0806,
+      "step": 7902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5303024649620056,
+      "learning_rate": 0.0005756612877761105,
+      "loss": 3.3074,
+      "step": 7903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5631526112556458,
+      "learning_rate": 0.0005756552127201706,
+      "loss": 3.3211,
+      "step": 7904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5668905377388,
+      "learning_rate": 0.0005756491369382078,
+      "loss": 3.2826,
+      "step": 7905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5107238292694092,
+      "learning_rate": 0.0005756430604302383,
+      "loss": 3.2575,
+      "step": 7906
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5243762731552124,
+      "learning_rate": 0.0005756369831962779,
+      "loss": 3.294,
+      "step": 7907
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5488375425338745,
+      "learning_rate": 0.0005756309052363429,
+      "loss": 3.6128,
+      "step": 7908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5057998299598694,
+      "learning_rate": 0.0005756248265504491,
+      "loss": 3.294,
+      "step": 7909
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.544693648815155,
+      "learning_rate": 0.0005756187471386126,
+      "loss": 3.2293,
+      "step": 7910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5097287893295288,
+      "learning_rate": 0.0005756126670008492,
+      "loss": 3.1767,
+      "step": 7911
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5124967098236084,
+      "learning_rate": 0.0005756065861371751,
+      "loss": 3.0868,
+      "step": 7912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5405759215354919,
+      "learning_rate": 0.0005756005045476064,
+      "loss": 3.4881,
+      "step": 7913
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5262176394462585,
+      "learning_rate": 0.000575594422232159,
+      "loss": 3.4049,
+      "step": 7914
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5306533575057983,
+      "learning_rate": 0.0005755883391908489,
+      "loss": 3.3627,
+      "step": 7915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5325959920883179,
+      "learning_rate": 0.0005755822554236921,
+      "loss": 3.4184,
+      "step": 7916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5592959523200989,
+      "learning_rate": 0.0005755761709307048,
+      "loss": 3.1438,
+      "step": 7917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5043314695358276,
+      "learning_rate": 0.0005755700857119028,
+      "loss": 3.2669,
+      "step": 7918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5433175563812256,
+      "learning_rate": 0.0005755639997673022,
+      "loss": 3.2691,
+      "step": 7919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5013863444328308,
+      "learning_rate": 0.0005755579130969191,
+      "loss": 3.1457,
+      "step": 7920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5583698749542236,
+      "learning_rate": 0.0005755518257007696,
+      "loss": 3.6578,
+      "step": 7921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5930216312408447,
+      "learning_rate": 0.0005755457375788694,
+      "loss": 3.2728,
+      "step": 7922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5604255795478821,
+      "learning_rate": 0.000575539648731235,
+      "loss": 3.2173,
+      "step": 7923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5618053674697876,
+      "learning_rate": 0.000575533559157882,
+      "loss": 3.3222,
+      "step": 7924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5377988219261169,
+      "learning_rate": 0.0005755274688588268,
+      "loss": 3.3963,
+      "step": 7925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5500739216804504,
+      "learning_rate": 0.0005755213778340852,
+      "loss": 3.6868,
+      "step": 7926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5406518578529358,
+      "learning_rate": 0.0005755152860836733,
+      "loss": 3.374,
+      "step": 7927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5654966235160828,
+      "learning_rate": 0.0005755091936076071,
+      "loss": 3.3569,
+      "step": 7928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5690451860427856,
+      "learning_rate": 0.0005755031004059028,
+      "loss": 3.2013,
+      "step": 7929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6072752475738525,
+      "learning_rate": 0.0005754970064785763,
+      "loss": 3.3478,
+      "step": 7930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5231303572654724,
+      "learning_rate": 0.0005754909118256438,
+      "loss": 3.1507,
+      "step": 7931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5188618898391724,
+      "learning_rate": 0.0005754848164471211,
+      "loss": 3.3347,
+      "step": 7932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5124830007553101,
+      "learning_rate": 0.0005754787203430245,
+      "loss": 3.5574,
+      "step": 7933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5399743914604187,
+      "learning_rate": 0.0005754726235133701,
+      "loss": 3.4946,
+      "step": 7934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.526694655418396,
+      "learning_rate": 0.0005754665259581736,
+      "loss": 3.2302,
+      "step": 7935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5251884460449219,
+      "learning_rate": 0.0005754604276774515,
+      "loss": 3.5666,
+      "step": 7936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5409374237060547,
+      "learning_rate": 0.0005754543286712193,
+      "loss": 3.2799,
+      "step": 7937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49520036578178406,
+      "learning_rate": 0.0005754482289394938,
+      "loss": 3.3421,
+      "step": 7938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5165755152702332,
+      "learning_rate": 0.0005754421284822905,
+      "loss": 3.0666,
+      "step": 7939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5416868329048157,
+      "learning_rate": 0.0005754360272996256,
+      "loss": 3.2758,
+      "step": 7940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0005754299253915153,
+      "loss": 3.2919,
+      "step": 7941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5226501822471619,
+      "learning_rate": 0.0005754238227579755,
+      "loss": 3.2847,
+      "step": 7942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5210117101669312,
+      "learning_rate": 0.0005754177193990225,
+      "loss": 3.4445,
+      "step": 7943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5392207503318787,
+      "learning_rate": 0.0005754116153146721,
+      "loss": 3.2395,
+      "step": 7944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5218811631202698,
+      "learning_rate": 0.0005754055105049406,
+      "loss": 3.3672,
+      "step": 7945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5551583170890808,
+      "learning_rate": 0.0005753994049698439,
+      "loss": 3.483,
+      "step": 7946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5441166758537292,
+      "learning_rate": 0.0005753932987093983,
+      "loss": 3.2976,
+      "step": 7947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5118980407714844,
+      "learning_rate": 0.0005753871917236195,
+      "loss": 3.258,
+      "step": 7948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51288902759552,
+      "learning_rate": 0.000575381084012524,
+      "loss": 3.1806,
+      "step": 7949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5351516008377075,
+      "learning_rate": 0.0005753749755761277,
+      "loss": 3.2706,
+      "step": 7950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5109082460403442,
+      "learning_rate": 0.0005753688664144467,
+      "loss": 3.0689,
+      "step": 7951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49430152773857117,
+      "learning_rate": 0.0005753627565274972,
+      "loss": 3.5631,
+      "step": 7952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5566909909248352,
+      "learning_rate": 0.0005753566459152952,
+      "loss": 3.3404,
+      "step": 7953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.576958179473877,
+      "learning_rate": 0.0005753505345778567,
+      "loss": 3.2608,
+      "step": 7954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5318806767463684,
+      "learning_rate": 0.0005753444225151978,
+      "loss": 3.2429,
+      "step": 7955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5288330316543579,
+      "learning_rate": 0.0005753383097273348,
+      "loss": 3.4162,
+      "step": 7956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5240846276283264,
+      "learning_rate": 0.0005753321962142837,
+      "loss": 3.142,
+      "step": 7957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5280718207359314,
+      "learning_rate": 0.0005753260819760605,
+      "loss": 3.2296,
+      "step": 7958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5125327110290527,
+      "learning_rate": 0.0005753199670126815,
+      "loss": 3.3329,
+      "step": 7959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5324898362159729,
+      "learning_rate": 0.0005753138513241626,
+      "loss": 3.3372,
+      "step": 7960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.519921064376831,
+      "learning_rate": 0.00057530773491052,
+      "loss": 3.653,
+      "step": 7961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5081679821014404,
+      "learning_rate": 0.0005753016177717699,
+      "loss": 3.2574,
+      "step": 7962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5309244990348816,
+      "learning_rate": 0.0005752954999079282,
+      "loss": 3.3269,
+      "step": 7963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5274438261985779,
+      "learning_rate": 0.0005752893813190112,
+      "loss": 3.3909,
+      "step": 7964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5856743454933167,
+      "learning_rate": 0.000575283262005035,
+      "loss": 3.2714,
+      "step": 7965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.528982400894165,
+      "learning_rate": 0.0005752771419660156,
+      "loss": 3.2405,
+      "step": 7966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5737332701683044,
+      "learning_rate": 0.0005752710212019692,
+      "loss": 3.5454,
+      "step": 7967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5274087190628052,
+      "learning_rate": 0.000575264899712912,
+      "loss": 3.3418,
+      "step": 7968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5080131888389587,
+      "learning_rate": 0.0005752587774988598,
+      "loss": 3.3738,
+      "step": 7969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4994853138923645,
+      "learning_rate": 0.0005752526545598291,
+      "loss": 3.4255,
+      "step": 7970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5325427651405334,
+      "learning_rate": 0.0005752465308958358,
+      "loss": 3.3933,
+      "step": 7971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5056576728820801,
+      "learning_rate": 0.0005752404065068963,
+      "loss": 3.4345,
+      "step": 7972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5503882765769958,
+      "learning_rate": 0.0005752342813930264,
+      "loss": 3.1064,
+      "step": 7973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.537910521030426,
+      "learning_rate": 0.0005752281555542423,
+      "loss": 3.264,
+      "step": 7974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5261189937591553,
+      "learning_rate": 0.0005752220289905604,
+      "loss": 3.375,
+      "step": 7975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5032466650009155,
+      "learning_rate": 0.0005752159017019965,
+      "loss": 3.2479,
+      "step": 7976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5199475884437561,
+      "learning_rate": 0.0005752097736885669,
+      "loss": 3.4072,
+      "step": 7977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5200490951538086,
+      "learning_rate": 0.0005752036449502878,
+      "loss": 3.2218,
+      "step": 7978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5323976874351501,
+      "learning_rate": 0.0005751975154871752,
+      "loss": 3.3384,
+      "step": 7979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5439580082893372,
+      "learning_rate": 0.0005751913852992454,
+      "loss": 3.3075,
+      "step": 7980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5520688891410828,
+      "learning_rate": 0.0005751852543865143,
+      "loss": 3.2518,
+      "step": 7981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5149745345115662,
+      "learning_rate": 0.0005751791227489983,
+      "loss": 3.2855,
+      "step": 7982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6031532287597656,
+      "learning_rate": 0.0005751729903867134,
+      "loss": 2.9604,
+      "step": 7983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5211678147315979,
+      "learning_rate": 0.0005751668572996758,
+      "loss": 3.5004,
+      "step": 7984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5391950011253357,
+      "learning_rate": 0.0005751607234879017,
+      "loss": 3.3339,
+      "step": 7985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5196065306663513,
+      "learning_rate": 0.0005751545889514072,
+      "loss": 3.2076,
+      "step": 7986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5143944025039673,
+      "learning_rate": 0.0005751484536902084,
+      "loss": 3.1234,
+      "step": 7987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5726907253265381,
+      "learning_rate": 0.0005751423177043216,
+      "loss": 3.3739,
+      "step": 7988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5271036028862,
+      "learning_rate": 0.0005751361809937629,
+      "loss": 3.4486,
+      "step": 7989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5144461393356323,
+      "learning_rate": 0.0005751300435585483,
+      "loss": 3.5016,
+      "step": 7990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5567069053649902,
+      "learning_rate": 0.0005751239053986944,
+      "loss": 3.3089,
+      "step": 7991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5412687659263611,
+      "learning_rate": 0.0005751177665142169,
+      "loss": 3.2718,
+      "step": 7992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5389328598976135,
+      "learning_rate": 0.0005751116269051321,
+      "loss": 3.1432,
+      "step": 7993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5082566142082214,
+      "learning_rate": 0.0005751054865714562,
+      "loss": 3.3403,
+      "step": 7994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51983243227005,
+      "learning_rate": 0.0005750993455132055,
+      "loss": 3.2146,
+      "step": 7995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5300633311271667,
+      "learning_rate": 0.0005750932037303961,
+      "loss": 3.234,
+      "step": 7996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5436065196990967,
+      "learning_rate": 0.0005750870612230442,
+      "loss": 3.2246,
+      "step": 7997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5409985184669495,
+      "learning_rate": 0.0005750809179911657,
+      "loss": 3.1161,
+      "step": 7998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5061076879501343,
+      "learning_rate": 0.0005750747740347771,
+      "loss": 3.2811,
+      "step": 7999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5530107617378235,
+      "learning_rate": 0.0005750686293538945,
+      "loss": 3.2673,
+      "step": 8000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5269939303398132,
+      "learning_rate": 0.0005750624839485341,
+      "loss": 3.3215,
+      "step": 8001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5110950469970703,
+      "learning_rate": 0.000575056337818712,
+      "loss": 3.5774,
+      "step": 8002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5932079553604126,
+      "learning_rate": 0.0005750501909644445,
+      "loss": 3.5213,
+      "step": 8003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5554304122924805,
+      "learning_rate": 0.0005750440433857477,
+      "loss": 3.2986,
+      "step": 8004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5720034241676331,
+      "learning_rate": 0.0005750378950826378,
+      "loss": 3.3181,
+      "step": 8005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5665091276168823,
+      "learning_rate": 0.000575031746055131,
+      "loss": 2.9911,
+      "step": 8006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5483243465423584,
+      "learning_rate": 0.0005750255963032436,
+      "loss": 3.3377,
+      "step": 8007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5081855058670044,
+      "learning_rate": 0.0005750194458269916,
+      "loss": 3.2402,
+      "step": 8008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5034002065658569,
+      "learning_rate": 0.0005750132946263914,
+      "loss": 3.2496,
+      "step": 8009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6338850259780884,
+      "learning_rate": 0.0005750071427014591,
+      "loss": 3.3337,
+      "step": 8010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5657280683517456,
+      "learning_rate": 0.0005750009900522109,
+      "loss": 3.4191,
+      "step": 8011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5098117589950562,
+      "learning_rate": 0.000574994836678663,
+      "loss": 3.3851,
+      "step": 8012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5298253297805786,
+      "learning_rate": 0.0005749886825808316,
+      "loss": 3.3719,
+      "step": 8013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5312026143074036,
+      "learning_rate": 0.000574982527758733,
+      "loss": 3.4547,
+      "step": 8014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5396552085876465,
+      "learning_rate": 0.0005749763722123832,
+      "loss": 3.2734,
+      "step": 8015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5509809255599976,
+      "learning_rate": 0.0005749702159417988,
+      "loss": 3.4702,
+      "step": 8016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5270836353302002,
+      "learning_rate": 0.0005749640589469956,
+      "loss": 3.3203,
+      "step": 8017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.540736973285675,
+      "learning_rate": 0.00057495790122799,
+      "loss": 3.1879,
+      "step": 8018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.531613290309906,
+      "learning_rate": 0.0005749517427847982,
+      "loss": 3.5024,
+      "step": 8019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5209075808525085,
+      "learning_rate": 0.0005749455836174365,
+      "loss": 3.5551,
+      "step": 8020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.536415159702301,
+      "learning_rate": 0.0005749394237259209,
+      "loss": 3.2383,
+      "step": 8021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5413359999656677,
+      "learning_rate": 0.0005749332631102679,
+      "loss": 3.1601,
+      "step": 8022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5244982242584229,
+      "learning_rate": 0.0005749271017704935,
+      "loss": 3.2752,
+      "step": 8023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5341219305992126,
+      "learning_rate": 0.0005749209397066142,
+      "loss": 3.1301,
+      "step": 8024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53799968957901,
+      "learning_rate": 0.0005749147769186459,
+      "loss": 3.2056,
+      "step": 8025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5490437746047974,
+      "learning_rate": 0.0005749086134066051,
+      "loss": 3.5046,
+      "step": 8026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5136128664016724,
+      "learning_rate": 0.0005749024491705078,
+      "loss": 3.2401,
+      "step": 8027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.566075325012207,
+      "learning_rate": 0.0005748962842103703,
+      "loss": 3.2781,
+      "step": 8028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5798686742782593,
+      "learning_rate": 0.0005748901185262091,
+      "loss": 3.3664,
+      "step": 8029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.541175365447998,
+      "learning_rate": 0.0005748839521180401,
+      "loss": 3.4175,
+      "step": 8030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5209834575653076,
+      "learning_rate": 0.0005748777849858797,
+      "loss": 3.3311,
+      "step": 8031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5575250387191772,
+      "learning_rate": 0.0005748716171297441,
+      "loss": 3.3556,
+      "step": 8032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5445090532302856,
+      "learning_rate": 0.0005748654485496496,
+      "loss": 3.2137,
+      "step": 8033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5050318241119385,
+      "learning_rate": 0.0005748592792456124,
+      "loss": 3.5883,
+      "step": 8034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5381130576133728,
+      "learning_rate": 0.0005748531092176487,
+      "loss": 3.3816,
+      "step": 8035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5211299657821655,
+      "learning_rate": 0.0005748469384657749,
+      "loss": 3.405,
+      "step": 8036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.543992817401886,
+      "learning_rate": 0.0005748407669900071,
+      "loss": 3.1935,
+      "step": 8037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5342070460319519,
+      "learning_rate": 0.0005748345947903615,
+      "loss": 3.2672,
+      "step": 8038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5374672412872314,
+      "learning_rate": 0.0005748284218668546,
+      "loss": 3.6805,
+      "step": 8039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5109804272651672,
+      "learning_rate": 0.0005748222482195026,
+      "loss": 3.3104,
+      "step": 8040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4903632402420044,
+      "learning_rate": 0.0005748160738483216,
+      "loss": 3.3117,
+      "step": 8041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.525119423866272,
+      "learning_rate": 0.0005748098987533279,
+      "loss": 3.3496,
+      "step": 8042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5801721215248108,
+      "learning_rate": 0.0005748037229345379,
+      "loss": 3.2369,
+      "step": 8043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.514396071434021,
+      "learning_rate": 0.0005747975463919677,
+      "loss": 3.2263,
+      "step": 8044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5463517904281616,
+      "learning_rate": 0.0005747913691256338,
+      "loss": 3.296,
+      "step": 8045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5206927061080933,
+      "learning_rate": 0.0005747851911355521,
+      "loss": 3.3216,
+      "step": 8046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5000367760658264,
+      "learning_rate": 0.0005747790124217393,
+      "loss": 3.2014,
+      "step": 8047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5487034320831299,
+      "learning_rate": 0.0005747728329842113,
+      "loss": 3.0561,
+      "step": 8048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5461079478263855,
+      "learning_rate": 0.0005747666528229847,
+      "loss": 3.3097,
+      "step": 8049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4785720407962799,
+      "learning_rate": 0.0005747604719380754,
+      "loss": 3.4331,
+      "step": 8050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5695480704307556,
+      "learning_rate": 0.0005747542903295,
+      "loss": 3.2418,
+      "step": 8051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5708943009376526,
+      "learning_rate": 0.0005747481079972749,
+      "loss": 3.3734,
+      "step": 8052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5381008386611938,
+      "learning_rate": 0.0005747419249414159,
+      "loss": 3.4608,
+      "step": 8053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5541656017303467,
+      "learning_rate": 0.0005747357411619396,
+      "loss": 3.3497,
+      "step": 8054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5573221445083618,
+      "learning_rate": 0.0005747295566588622,
+      "loss": 3.3627,
+      "step": 8055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5886656641960144,
+      "learning_rate": 0.0005747233714322001,
+      "loss": 3.399,
+      "step": 8056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5136348605155945,
+      "learning_rate": 0.0005747171854819694,
+      "loss": 3.3561,
+      "step": 8057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5364802479743958,
+      "learning_rate": 0.0005747109988081865,
+      "loss": 3.4321,
+      "step": 8058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5847378969192505,
+      "learning_rate": 0.000574704811410868,
+      "loss": 3.3716,
+      "step": 8059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5139754414558411,
+      "learning_rate": 0.0005746986232900295,
+      "loss": 3.3843,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5183747410774231,
+      "learning_rate": 0.0005746924344456879,
+      "loss": 3.3141,
+      "step": 8061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5344926714897156,
+      "learning_rate": 0.0005746862448778593,
+      "loss": 3.4683,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5372681021690369,
+      "learning_rate": 0.0005746800545865599,
+      "loss": 3.4784,
+      "step": 8063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5549917817115784,
+      "learning_rate": 0.0005746738635718061,
+      "loss": 3.2525,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5214190483093262,
+      "learning_rate": 0.0005746676718336143,
+      "loss": 3.4208,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5115858912467957,
+      "learning_rate": 0.0005746614793720007,
+      "loss": 3.1698,
+      "step": 8066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.571573793888092,
+      "learning_rate": 0.0005746552861869815,
+      "loss": 3.3276,
+      "step": 8067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5278741121292114,
+      "learning_rate": 0.0005746490922785733,
+      "loss": 3.3125,
+      "step": 8068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5263270139694214,
+      "learning_rate": 0.0005746428976467922,
+      "loss": 3.39,
+      "step": 8069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.553209125995636,
+      "learning_rate": 0.0005746367022916545,
+      "loss": 3.1092,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5743881464004517,
+      "learning_rate": 0.0005746305062131765,
+      "loss": 3.3103,
+      "step": 8071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5375149846076965,
+      "learning_rate": 0.0005746243094113748,
+      "loss": 3.4892,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5758310556411743,
+      "learning_rate": 0.0005746181118862655,
+      "loss": 3.199,
+      "step": 8073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.522111177444458,
+      "learning_rate": 0.0005746119136378648,
+      "loss": 3.2247,
+      "step": 8074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5190290808677673,
+      "learning_rate": 0.0005746057146661892,
+      "loss": 3.1963,
+      "step": 8075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5556468367576599,
+      "learning_rate": 0.0005745995149712552,
+      "loss": 3.0931,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5247904658317566,
+      "learning_rate": 0.0005745933145530787,
+      "loss": 3.28,
+      "step": 8077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5039224624633789,
+      "learning_rate": 0.0005745871134116763,
+      "loss": 3.3245,
+      "step": 8078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5405135750770569,
+      "learning_rate": 0.0005745809115470643,
+      "loss": 3.1882,
+      "step": 8079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5452626347541809,
+      "learning_rate": 0.000574574708959259,
+      "loss": 3.3268,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5153411626815796,
+      "learning_rate": 0.000574568505648277,
+      "loss": 3.3195,
+      "step": 8081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5305858254432678,
+      "learning_rate": 0.0005745623016141341,
+      "loss": 3.4117,
+      "step": 8082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5426940321922302,
+      "learning_rate": 0.000574556096856847,
+      "loss": 3.1285,
+      "step": 8083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4963018596172333,
+      "learning_rate": 0.000574549891376432,
+      "loss": 3.2248,
+      "step": 8084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5105463266372681,
+      "learning_rate": 0.0005745436851729055,
+      "loss": 3.2956,
+      "step": 8085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5522632598876953,
+      "learning_rate": 0.0005745374782462837,
+      "loss": 3.0467,
+      "step": 8086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5475056767463684,
+      "learning_rate": 0.000574531270596583,
+      "loss": 3.382,
+      "step": 8087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.501108705997467,
+      "learning_rate": 0.0005745250622238198,
+      "loss": 3.5365,
+      "step": 8088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5344222187995911,
+      "learning_rate": 0.0005745188531280105,
+      "loss": 3.2848,
+      "step": 8089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5425450801849365,
+      "learning_rate": 0.0005745126433091712,
+      "loss": 3.5062,
+      "step": 8090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5173064470291138,
+      "learning_rate": 0.0005745064327673185,
+      "loss": 3.1953,
+      "step": 8091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5049318671226501,
+      "learning_rate": 0.0005745002215024687,
+      "loss": 3.4188,
+      "step": 8092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5095267295837402,
+      "learning_rate": 0.0005744940095146381,
+      "loss": 3.5542,
+      "step": 8093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5197057127952576,
+      "learning_rate": 0.0005744877968038432,
+      "loss": 3.341,
+      "step": 8094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4955214262008667,
+      "learning_rate": 0.0005744815833701003,
+      "loss": 3.3152,
+      "step": 8095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5291468501091003,
+      "learning_rate": 0.0005744753692134256,
+      "loss": 3.3581,
+      "step": 8096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5096137523651123,
+      "learning_rate": 0.0005744691543338357,
+      "loss": 3.4981,
+      "step": 8097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47097456455230713,
+      "learning_rate": 0.0005744629387313469,
+      "loss": 3.2238,
+      "step": 8098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4965920150279999,
+      "learning_rate": 0.0005744567224059754,
+      "loss": 3.5343,
+      "step": 8099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5325508713722229,
+      "learning_rate": 0.0005744505053577379,
+      "loss": 3.2636,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5449921488761902,
+      "learning_rate": 0.0005744442875866504,
+      "loss": 3.4296,
+      "step": 8101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5376118421554565,
+      "learning_rate": 0.0005744380690927295,
+      "loss": 3.2763,
+      "step": 8102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5433290600776672,
+      "learning_rate": 0.0005744318498759917,
+      "loss": 3.2533,
+      "step": 8103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5353100299835205,
+      "learning_rate": 0.000574425629936453,
+      "loss": 3.3458,
+      "step": 8104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5136173963546753,
+      "learning_rate": 0.0005744194092741302,
+      "loss": 3.3949,
+      "step": 8105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5367599725723267,
+      "learning_rate": 0.0005744131878890394,
+      "loss": 3.2921,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5330642461776733,
+      "learning_rate": 0.0005744069657811971,
+      "loss": 3.3031,
+      "step": 8107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5547467470169067,
+      "learning_rate": 0.0005744007429506195,
+      "loss": 3.5355,
+      "step": 8108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47983911633491516,
+      "learning_rate": 0.0005743945193973233,
+      "loss": 3.4107,
+      "step": 8109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5580479502677917,
+      "learning_rate": 0.0005743882951213247,
+      "loss": 3.2559,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5408802628517151,
+      "learning_rate": 0.0005743820701226402,
+      "loss": 3.2648,
+      "step": 8111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5535223484039307,
+      "learning_rate": 0.0005743758444012861,
+      "loss": 3.3305,
+      "step": 8112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5365628600120544,
+      "learning_rate": 0.0005743696179572788,
+      "loss": 3.4065,
+      "step": 8113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6094711422920227,
+      "learning_rate": 0.0005743633907906348,
+      "loss": 3.5076,
+      "step": 8114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5201326608657837,
+      "learning_rate": 0.0005743571629013704,
+      "loss": 3.4203,
+      "step": 8115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5079147219657898,
+      "learning_rate": 0.000574350934289502,
+      "loss": 3.2888,
+      "step": 8116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5209977030754089,
+      "learning_rate": 0.0005743447049550461,
+      "loss": 3.292,
+      "step": 8117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.524752140045166,
+      "learning_rate": 0.000574338474898019,
+      "loss": 3.129,
+      "step": 8118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5393475890159607,
+      "learning_rate": 0.0005743322441184372,
+      "loss": 3.1805,
+      "step": 8119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5741223096847534,
+      "learning_rate": 0.000574326012616317,
+      "loss": 3.3835,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.531087338924408,
+      "learning_rate": 0.0005743197803916749,
+      "loss": 3.4494,
+      "step": 8121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5499753952026367,
+      "learning_rate": 0.0005743135474445273,
+      "loss": 3.1866,
+      "step": 8122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5074490904808044,
+      "learning_rate": 0.0005743073137748906,
+      "loss": 3.3362,
+      "step": 8123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4964973032474518,
+      "learning_rate": 0.0005743010793827813,
+      "loss": 3.1807,
+      "step": 8124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4979309141635895,
+      "learning_rate": 0.0005742948442682157,
+      "loss": 3.3788,
+      "step": 8125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5132046341896057,
+      "learning_rate": 0.0005742886084312102,
+      "loss": 3.2557,
+      "step": 8126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5504202246665955,
+      "learning_rate": 0.0005742823718717814,
+      "loss": 3.3969,
+      "step": 8127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5701044797897339,
+      "learning_rate": 0.0005742761345899455,
+      "loss": 3.2887,
+      "step": 8128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5031047463417053,
+      "learning_rate": 0.0005742698965857192,
+      "loss": 3.2875,
+      "step": 8129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.512702465057373,
+      "learning_rate": 0.0005742636578591186,
+      "loss": 3.2581,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5461293458938599,
+      "learning_rate": 0.0005742574184101605,
+      "loss": 3.3255,
+      "step": 8131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5259397625923157,
+      "learning_rate": 0.000574251178238861,
+      "loss": 2.9756,
+      "step": 8132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5428778529167175,
+      "learning_rate": 0.0005742449373452367,
+      "loss": 3.2215,
+      "step": 8133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5212182402610779,
+      "learning_rate": 0.0005742386957293041,
+      "loss": 3.4854,
+      "step": 8134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.537875235080719,
+      "learning_rate": 0.0005742324533910794,
+      "loss": 3.467,
+      "step": 8135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5199194550514221,
+      "learning_rate": 0.0005742262103305793,
+      "loss": 3.432,
+      "step": 8136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.498753160238266,
+      "learning_rate": 0.0005742199665478201,
+      "loss": 3.2888,
+      "step": 8137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5349212288856506,
+      "learning_rate": 0.0005742137220428184,
+      "loss": 3.3626,
+      "step": 8138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5592776536941528,
+      "learning_rate": 0.0005742074768155903,
+      "loss": 3.5535,
+      "step": 8139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5438594818115234,
+      "learning_rate": 0.0005742012308661527,
+      "loss": 3.3263,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5591511726379395,
+      "learning_rate": 0.0005741949841945217,
+      "loss": 3.191,
+      "step": 8141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5086604356765747,
+      "learning_rate": 0.000574188736800714,
+      "loss": 3.4601,
+      "step": 8142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5508185029029846,
+      "learning_rate": 0.0005741824886847458,
+      "loss": 3.2594,
+      "step": 8143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5638971328735352,
+      "learning_rate": 0.0005741762398466337,
+      "loss": 3.366,
+      "step": 8144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5249466300010681,
+      "learning_rate": 0.0005741699902863943,
+      "loss": 3.3142,
+      "step": 8145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5074349045753479,
+      "learning_rate": 0.0005741637400040438,
+      "loss": 3.4865,
+      "step": 8146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5369219779968262,
+      "learning_rate": 0.0005741574889995987,
+      "loss": 3.2189,
+      "step": 8147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.546882688999176,
+      "learning_rate": 0.0005741512372730757,
+      "loss": 3.2913,
+      "step": 8148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5176022052764893,
+      "learning_rate": 0.000574144984824491,
+      "loss": 3.2776,
+      "step": 8149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5151086449623108,
+      "learning_rate": 0.0005741387316538612,
+      "loss": 3.3433,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5456839799880981,
+      "learning_rate": 0.0005741324777612027,
+      "loss": 3.3781,
+      "step": 8151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.591968834400177,
+      "learning_rate": 0.0005741262231465321,
+      "loss": 3.3347,
+      "step": 8152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5454823970794678,
+      "learning_rate": 0.0005741199678098657,
+      "loss": 3.6877,
+      "step": 8153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5745152831077576,
+      "learning_rate": 0.00057411371175122,
+      "loss": 3.3773,
+      "step": 8154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5442945957183838,
+      "learning_rate": 0.0005741074549706117,
+      "loss": 3.4401,
+      "step": 8155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5125526785850525,
+      "learning_rate": 0.000574101197468057,
+      "loss": 3.3032,
+      "step": 8156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5886013507843018,
+      "learning_rate": 0.0005740949392435726,
+      "loss": 3.3589,
+      "step": 8157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5841769576072693,
+      "learning_rate": 0.0005740886802971748,
+      "loss": 3.1734,
+      "step": 8158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5075914263725281,
+      "learning_rate": 0.0005740824206288801,
+      "loss": 3.2588,
+      "step": 8159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5713216662406921,
+      "learning_rate": 0.0005740761602387052,
+      "loss": 3.1835,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5094101428985596,
+      "learning_rate": 0.0005740698991266664,
+      "loss": 3.5436,
+      "step": 8161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49177950620651245,
+      "learning_rate": 0.0005740636372927802,
+      "loss": 3.3805,
+      "step": 8162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5621153116226196,
+      "learning_rate": 0.0005740573747370632,
+      "loss": 3.1933,
+      "step": 8163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5192760229110718,
+      "learning_rate": 0.0005740511114595317,
+      "loss": 3.2768,
+      "step": 8164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5471800565719604,
+      "learning_rate": 0.0005740448474602024,
+      "loss": 3.3567,
+      "step": 8165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.541714608669281,
+      "learning_rate": 0.0005740385827390916,
+      "loss": 3.1748,
+      "step": 8166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4933028519153595,
+      "learning_rate": 0.000574032317296216,
+      "loss": 3.3155,
+      "step": 8167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5285734534263611,
+      "learning_rate": 0.0005740260511315921,
+      "loss": 3.0752,
+      "step": 8168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5299739837646484,
+      "learning_rate": 0.0005740197842452362,
+      "loss": 3.3437,
+      "step": 8169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5052409768104553,
+      "learning_rate": 0.000574013516637165,
+      "loss": 3.4159,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.522113561630249,
+      "learning_rate": 0.0005740072483073948,
+      "loss": 3.3624,
+      "step": 8171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5488013625144958,
+      "learning_rate": 0.0005740009792559424,
+      "loss": 3.338,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5112448334693909,
+      "learning_rate": 0.000573994709482824,
+      "loss": 3.3995,
+      "step": 8173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5400604605674744,
+      "learning_rate": 0.0005739884389880564,
+      "loss": 3.0325,
+      "step": 8174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5453541874885559,
+      "learning_rate": 0.000573982167771656,
+      "loss": 3.333,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5471153855323792,
+      "learning_rate": 0.0005739758958336392,
+      "loss": 3.2572,
+      "step": 8176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5193156003952026,
+      "learning_rate": 0.0005739696231740226,
+      "loss": 3.1392,
+      "step": 8177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5482691526412964,
+      "learning_rate": 0.0005739633497928228,
+      "loss": 3.2852,
+      "step": 8178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5617150664329529,
+      "learning_rate": 0.0005739570756900563,
+      "loss": 3.3936,
+      "step": 8179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5217198133468628,
+      "learning_rate": 0.0005739508008657394,
+      "loss": 3.2155,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5516659021377563,
+      "learning_rate": 0.000573944525319889,
+      "loss": 3.2566,
+      "step": 8181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5190207362174988,
+      "learning_rate": 0.0005739382490525214,
+      "loss": 3.6173,
+      "step": 8182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5515297055244446,
+      "learning_rate": 0.0005739319720636532,
+      "loss": 3.1461,
+      "step": 8183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4930954873561859,
+      "learning_rate": 0.0005739256943533007,
+      "loss": 3.3663,
+      "step": 8184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5274080038070679,
+      "learning_rate": 0.0005739194159214808,
+      "loss": 3.5295,
+      "step": 8185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5591486096382141,
+      "learning_rate": 0.00057391313676821,
+      "loss": 3.4101,
+      "step": 8186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5552020072937012,
+      "learning_rate": 0.0005739068568935046,
+      "loss": 3.3131,
+      "step": 8187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5581715106964111,
+      "learning_rate": 0.0005739005762973813,
+      "loss": 3.1061,
+      "step": 8188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5154845714569092,
+      "learning_rate": 0.0005738942949798564,
+      "loss": 3.3115,
+      "step": 8189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49115267395973206,
+      "learning_rate": 0.0005738880129409469,
+      "loss": 3.395,
+      "step": 8190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5890899896621704,
+      "learning_rate": 0.0005738817301806689,
+      "loss": 3.1336,
+      "step": 8191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5730346441268921,
+      "learning_rate": 0.0005738754466990393,
+      "loss": 3.2823,
+      "step": 8192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5348773002624512,
+      "learning_rate": 0.0005738691624960743,
+      "loss": 3.3428,
+      "step": 8193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5359728336334229,
+      "learning_rate": 0.0005738628775717907,
+      "loss": 3.3814,
+      "step": 8194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5193948149681091,
+      "learning_rate": 0.000573856591926205,
+      "loss": 3.3435,
+      "step": 8195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5473117232322693,
+      "learning_rate": 0.0005738503055593338,
+      "loss": 3.5604,
+      "step": 8196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5555245280265808,
+      "learning_rate": 0.0005738440184711934,
+      "loss": 3.3051,
+      "step": 8197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6440548896789551,
+      "learning_rate": 0.0005738377306618008,
+      "loss": 3.357,
+      "step": 8198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4919479489326477,
+      "learning_rate": 0.0005738314421311722,
+      "loss": 3.1436,
+      "step": 8199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5028539299964905,
+      "learning_rate": 0.0005738251528793244,
+      "loss": 3.3776,
+      "step": 8200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5393859148025513,
+      "learning_rate": 0.0005738188629062737,
+      "loss": 3.3237,
+      "step": 8201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5415965914726257,
+      "learning_rate": 0.0005738125722120368,
+      "loss": 3.3788,
+      "step": 8202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5266854763031006,
+      "learning_rate": 0.0005738062807966303,
+      "loss": 3.3418,
+      "step": 8203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5485467314720154,
+      "learning_rate": 0.0005737999886600709,
+      "loss": 3.3907,
+      "step": 8204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5207034945487976,
+      "learning_rate": 0.0005737936958023749,
+      "loss": 3.3886,
+      "step": 8205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5115718841552734,
+      "learning_rate": 0.0005737874022235589,
+      "loss": 3.4131,
+      "step": 8206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52555251121521,
+      "learning_rate": 0.0005737811079236397,
+      "loss": 3.308,
+      "step": 8207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5632022023200989,
+      "learning_rate": 0.0005737748129026338,
+      "loss": 3.2812,
+      "step": 8208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5216935873031616,
+      "learning_rate": 0.0005737685171605577,
+      "loss": 3.3546,
+      "step": 8209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5155556201934814,
+      "learning_rate": 0.0005737622206974279,
+      "loss": 3.3297,
+      "step": 8210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5012245178222656,
+      "learning_rate": 0.0005737559235132612,
+      "loss": 3.4148,
+      "step": 8211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8418143391609192,
+      "learning_rate": 0.000573749625608074,
+      "loss": 3.4831,
+      "step": 8212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5532170534133911,
+      "learning_rate": 0.0005737433269818831,
+      "loss": 3.3624,
+      "step": 8213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5478039979934692,
+      "learning_rate": 0.0005737370276347049,
+      "loss": 3.1837,
+      "step": 8214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5057279467582703,
+      "learning_rate": 0.0005737307275665561,
+      "loss": 3.2443,
+      "step": 8215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5062552690505981,
+      "learning_rate": 0.0005737244267774531,
+      "loss": 3.423,
+      "step": 8216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5578494071960449,
+      "learning_rate": 0.0005737181252674127,
+      "loss": 3.2908,
+      "step": 8217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5365298390388489,
+      "learning_rate": 0.0005737118230364515,
+      "loss": 3.4519,
+      "step": 8218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4922030568122864,
+      "learning_rate": 0.0005737055200845861,
+      "loss": 3.3189,
+      "step": 8219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5830522775650024,
+      "learning_rate": 0.0005736992164118328,
+      "loss": 3.1601,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5145552158355713,
+      "learning_rate": 0.0005736929120182086,
+      "loss": 3.4044,
+      "step": 8221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5255774259567261,
+      "learning_rate": 0.0005736866069037299,
+      "loss": 3.2494,
+      "step": 8222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.518876314163208,
+      "learning_rate": 0.0005736803010684134,
+      "loss": 3.1429,
+      "step": 8223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49960964918136597,
+      "learning_rate": 0.0005736739945122756,
+      "loss": 3.1982,
+      "step": 8224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5569823980331421,
+      "learning_rate": 0.0005736676872353331,
+      "loss": 3.2566,
+      "step": 8225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5008841753005981,
+      "learning_rate": 0.0005736613792376027,
+      "loss": 3.3866,
+      "step": 8226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5104358196258545,
+      "learning_rate": 0.0005736550705191009,
+      "loss": 3.291,
+      "step": 8227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5353780388832092,
+      "learning_rate": 0.0005736487610798443,
+      "loss": 3.2402,
+      "step": 8228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5313364863395691,
+      "learning_rate": 0.0005736424509198495,
+      "loss": 3.5111,
+      "step": 8229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.520453155040741,
+      "learning_rate": 0.0005736361400391332,
+      "loss": 3.4149,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5445863008499146,
+      "learning_rate": 0.0005736298284377119,
+      "loss": 3.2695,
+      "step": 8231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5336853265762329,
+      "learning_rate": 0.0005736235161156023,
+      "loss": 3.4905,
+      "step": 8232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5520970225334167,
+      "learning_rate": 0.000573617203072821,
+      "loss": 3.3523,
+      "step": 8233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5282114744186401,
+      "learning_rate": 0.0005736108893093847,
+      "loss": 3.2539,
+      "step": 8234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.560897946357727,
+      "learning_rate": 0.00057360457482531,
+      "loss": 3.3655,
+      "step": 8235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5395894646644592,
+      "learning_rate": 0.0005735982596206133,
+      "loss": 3.3413,
+      "step": 8236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5417824983596802,
+      "learning_rate": 0.0005735919436953117,
+      "loss": 3.2808,
+      "step": 8237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5005999207496643,
+      "learning_rate": 0.0005735856270494214,
+      "loss": 3.3514,
+      "step": 8238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5334196090698242,
+      "learning_rate": 0.0005735793096829593,
+      "loss": 3.4646,
+      "step": 8239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.571017861366272,
+      "learning_rate": 0.0005735729915959419,
+      "loss": 3.3096,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.560081422328949,
+      "learning_rate": 0.0005735666727883861,
+      "loss": 3.3187,
+      "step": 8241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5151913166046143,
+      "learning_rate": 0.000573560353260308,
+      "loss": 3.3688,
+      "step": 8242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5449792742729187,
+      "learning_rate": 0.0005735540330117247,
+      "loss": 3.1677,
+      "step": 8243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.530566394329071,
+      "learning_rate": 0.0005735477120426528,
+      "loss": 3.0801,
+      "step": 8244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5325320363044739,
+      "learning_rate": 0.0005735413903531089,
+      "loss": 3.2273,
+      "step": 8245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5118095874786377,
+      "learning_rate": 0.0005735350679431095,
+      "loss": 3.2628,
+      "step": 8246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5123176574707031,
+      "learning_rate": 0.0005735287448126714,
+      "loss": 3.2,
+      "step": 8247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5000588297843933,
+      "learning_rate": 0.0005735224209618113,
+      "loss": 3.2256,
+      "step": 8248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6550585627555847,
+      "learning_rate": 0.0005735160963905458,
+      "loss": 3.3236,
+      "step": 8249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5188483595848083,
+      "learning_rate": 0.0005735097710988914,
+      "loss": 3.2484,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5245037078857422,
+      "learning_rate": 0.000573503445086865,
+      "loss": 3.1947,
+      "step": 8251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5324363708496094,
+      "learning_rate": 0.0005734971183544832,
+      "loss": 3.2969,
+      "step": 8252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5693351626396179,
+      "learning_rate": 0.0005734907909017624,
+      "loss": 3.4193,
+      "step": 8253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5394246578216553,
+      "learning_rate": 0.0005734844627287197,
+      "loss": 3.3065,
+      "step": 8254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5450414419174194,
+      "learning_rate": 0.0005734781338353716,
+      "loss": 3.4544,
+      "step": 8255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5223867297172546,
+      "learning_rate": 0.0005734718042217345,
+      "loss": 3.3402,
+      "step": 8256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5268417000770569,
+      "learning_rate": 0.0005734654738878256,
+      "loss": 3.254,
+      "step": 8257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5623879432678223,
+      "learning_rate": 0.0005734591428336611,
+      "loss": 3.2588,
+      "step": 8258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5085955858230591,
+      "learning_rate": 0.0005734528110592577,
+      "loss": 3.3621,
+      "step": 8259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5223268270492554,
+      "learning_rate": 0.0005734464785646325,
+      "loss": 3.201,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4991510808467865,
+      "learning_rate": 0.0005734401453498018,
+      "loss": 3.2424,
+      "step": 8261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5398638248443604,
+      "learning_rate": 0.0005734338114147823,
+      "loss": 3.4962,
+      "step": 8262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5050841569900513,
+      "learning_rate": 0.0005734274767595908,
+      "loss": 3.4429,
+      "step": 8263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5685188174247742,
+      "learning_rate": 0.000573421141384244,
+      "loss": 3.1578,
+      "step": 8264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5070170164108276,
+      "learning_rate": 0.0005734148052887585,
+      "loss": 3.3676,
+      "step": 8265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5665714144706726,
+      "learning_rate": 0.0005734084684731511,
+      "loss": 3.2815,
+      "step": 8266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.498685359954834,
+      "learning_rate": 0.0005734021309374383,
+      "loss": 3.3352,
+      "step": 8267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.509119987487793,
+      "learning_rate": 0.000573395792681637,
+      "loss": 3.2356,
+      "step": 8268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5538164973258972,
+      "learning_rate": 0.0005733894537057638,
+      "loss": 3.2961,
+      "step": 8269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5612331628799438,
+      "learning_rate": 0.0005733831140098353,
+      "loss": 3.3455,
+      "step": 8270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5053423047065735,
+      "learning_rate": 0.0005733767735938683,
+      "loss": 3.3879,
+      "step": 8271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56037837266922,
+      "learning_rate": 0.0005733704324578795,
+      "loss": 3.0991,
+      "step": 8272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5380472540855408,
+      "learning_rate": 0.0005733640906018856,
+      "loss": 3.2985,
+      "step": 8273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5246527791023254,
+      "learning_rate": 0.0005733577480259034,
+      "loss": 3.3638,
+      "step": 8274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5476314425468445,
+      "learning_rate": 0.0005733514047299494,
+      "loss": 3.4797,
+      "step": 8275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5380063652992249,
+      "learning_rate": 0.0005733450607140404,
+      "loss": 3.3618,
+      "step": 8276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5321431159973145,
+      "learning_rate": 0.0005733387159781931,
+      "loss": 3.3288,
+      "step": 8277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6225263476371765,
+      "learning_rate": 0.0005733323705224243,
+      "loss": 3.3563,
+      "step": 8278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5668192505836487,
+      "learning_rate": 0.0005733260243467506,
+      "loss": 3.2277,
+      "step": 8279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5005346536636353,
+      "learning_rate": 0.0005733196774511888,
+      "loss": 3.4365,
+      "step": 8280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.597763180732727,
+      "learning_rate": 0.0005733133298357555,
+      "loss": 3.102,
+      "step": 8281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5184804797172546,
+      "learning_rate": 0.0005733069815004675,
+      "loss": 3.5517,
+      "step": 8282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.540427565574646,
+      "learning_rate": 0.0005733006324453415,
+      "loss": 3.3953,
+      "step": 8283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.595478355884552,
+      "learning_rate": 0.0005732942826703943,
+      "loss": 3.3344,
+      "step": 8284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.551795482635498,
+      "learning_rate": 0.0005732879321756426,
+      "loss": 3.3396,
+      "step": 8285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5238455533981323,
+      "learning_rate": 0.0005732815809611029,
+      "loss": 3.5336,
+      "step": 8286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5102473497390747,
+      "learning_rate": 0.0005732752290267921,
+      "loss": 3.5597,
+      "step": 8287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5258164405822754,
+      "learning_rate": 0.0005732688763727273,
+      "loss": 3.4711,
+      "step": 8288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5059826970100403,
+      "learning_rate": 0.0005732625229989245,
+      "loss": 3.2522,
+      "step": 8289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5301188826560974,
+      "learning_rate": 0.0005732561689054009,
+      "loss": 3.5621,
+      "step": 8290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5123043060302734,
+      "learning_rate": 0.0005732498140921732,
+      "loss": 3.4555,
+      "step": 8291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5111595988273621,
+      "learning_rate": 0.000573243458559258,
+      "loss": 3.3539,
+      "step": 8292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6053318977355957,
+      "learning_rate": 0.0005732371023066721,
+      "loss": 3.4818,
+      "step": 8293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5144967436790466,
+      "learning_rate": 0.0005732307453344323,
+      "loss": 3.4808,
+      "step": 8294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5194787383079529,
+      "learning_rate": 0.0005732243876425554,
+      "loss": 3.1705,
+      "step": 8295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5327216386795044,
+      "learning_rate": 0.0005732180292310579,
+      "loss": 3.1993,
+      "step": 8296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49364081025123596,
+      "learning_rate": 0.0005732116700999568,
+      "loss": 3.4751,
+      "step": 8297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5248584747314453,
+      "learning_rate": 0.0005732053102492687,
+      "loss": 3.7122,
+      "step": 8298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5066429376602173,
+      "learning_rate": 0.0005731989496790105,
+      "loss": 3.073,
+      "step": 8299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5377814173698425,
+      "learning_rate": 0.0005731925883891986,
+      "loss": 3.3839,
+      "step": 8300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5316858291625977,
+      "learning_rate": 0.0005731862263798502,
+      "loss": 3.0971,
+      "step": 8301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5127855539321899,
+      "learning_rate": 0.0005731798636509817,
+      "loss": 3.1861,
+      "step": 8302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5574333071708679,
+      "learning_rate": 0.0005731735002026102,
+      "loss": 3.3133,
+      "step": 8303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5213868021965027,
+      "learning_rate": 0.0005731671360347521,
+      "loss": 3.1549,
+      "step": 8304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5839649438858032,
+      "learning_rate": 0.0005731607711474244,
+      "loss": 3.3224,
+      "step": 8305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.524940013885498,
+      "learning_rate": 0.0005731544055406439,
+      "loss": 3.5019,
+      "step": 8306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5284970998764038,
+      "learning_rate": 0.0005731480392144272,
+      "loss": 3.2947,
+      "step": 8307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5189017057418823,
+      "learning_rate": 0.000573141672168791,
+      "loss": 3.0949,
+      "step": 8308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5156539678573608,
+      "learning_rate": 0.0005731353044037524,
+      "loss": 3.3861,
+      "step": 8309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49928221106529236,
+      "learning_rate": 0.0005731289359193278,
+      "loss": 3.4869,
+      "step": 8310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546218991279602,
+      "learning_rate": 0.0005731225667155343,
+      "loss": 3.2345,
+      "step": 8311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5048373937606812,
+      "learning_rate": 0.0005731161967923885,
+      "loss": 3.1884,
+      "step": 8312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5357444286346436,
+      "learning_rate": 0.0005731098261499071,
+      "loss": 3.3281,
+      "step": 8313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5371036529541016,
+      "learning_rate": 0.0005731034547881071,
+      "loss": 3.3279,
+      "step": 8314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.539368748664856,
+      "learning_rate": 0.000573097082707005,
+      "loss": 3.4323,
+      "step": 8315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48763561248779297,
+      "learning_rate": 0.0005730907099066179,
+      "loss": 3.5516,
+      "step": 8316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5571647882461548,
+      "learning_rate": 0.0005730843363869624,
+      "loss": 3.4282,
+      "step": 8317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51971834897995,
+      "learning_rate": 0.0005730779621480552,
+      "loss": 3.2306,
+      "step": 8318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5173705816268921,
+      "learning_rate": 0.0005730715871899133,
+      "loss": 3.3458,
+      "step": 8319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5432358980178833,
+      "learning_rate": 0.0005730652115125533,
+      "loss": 3.3075,
+      "step": 8320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5046690106391907,
+      "learning_rate": 0.0005730588351159922,
+      "loss": 3.2926,
+      "step": 8321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5098605155944824,
+      "learning_rate": 0.0005730524580002468,
+      "loss": 3.21,
+      "step": 8322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5081766247749329,
+      "learning_rate": 0.0005730460801653335,
+      "loss": 3.401,
+      "step": 8323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5278263092041016,
+      "learning_rate": 0.0005730397016112694,
+      "loss": 3.2034,
+      "step": 8324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5464687347412109,
+      "learning_rate": 0.0005730333223380714,
+      "loss": 3.4792,
+      "step": 8325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5445109009742737,
+      "learning_rate": 0.0005730269423457561,
+      "loss": 3.5362,
+      "step": 8326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5181647539138794,
+      "learning_rate": 0.0005730205616343405,
+      "loss": 3.1952,
+      "step": 8327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5573515892028809,
+      "learning_rate": 0.0005730141802038413,
+      "loss": 3.1039,
+      "step": 8328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5178354978561401,
+      "learning_rate": 0.0005730077980542751,
+      "loss": 3.536,
+      "step": 8329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5695675015449524,
+      "learning_rate": 0.000573001415185659,
+      "loss": 3.3917,
+      "step": 8330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5261296629905701,
+      "learning_rate": 0.0005729950315980098,
+      "loss": 3.4801,
+      "step": 8331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5165058374404907,
+      "learning_rate": 0.0005729886472913441,
+      "loss": 3.3878,
+      "step": 8332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.513191819190979,
+      "learning_rate": 0.0005729822622656788,
+      "loss": 3.4338,
+      "step": 8333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5528197884559631,
+      "learning_rate": 0.0005729758765210309,
+      "loss": 3.5439,
+      "step": 8334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5654256343841553,
+      "learning_rate": 0.000572969490057417,
+      "loss": 3.2453,
+      "step": 8335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5060741901397705,
+      "learning_rate": 0.0005729631028748539,
+      "loss": 3.3054,
+      "step": 8336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5616364479064941,
+      "learning_rate": 0.0005729567149733587,
+      "loss": 3.2475,
+      "step": 8337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4996776878833771,
+      "learning_rate": 0.0005729503263529479,
+      "loss": 3.3329,
+      "step": 8338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.519364595413208,
+      "learning_rate": 0.0005729439370136387,
+      "loss": 3.1648,
+      "step": 8339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5395146012306213,
+      "learning_rate": 0.0005729375469554474,
+      "loss": 3.2874,
+      "step": 8340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5185738205909729,
+      "learning_rate": 0.0005729311561783913,
+      "loss": 3.092,
+      "step": 8341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5481864809989929,
+      "learning_rate": 0.0005729247646824871,
+      "loss": 3.2024,
+      "step": 8342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5273303389549255,
+      "learning_rate": 0.0005729183724677515,
+      "loss": 3.1049,
+      "step": 8343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5579976439476013,
+      "learning_rate": 0.0005729119795342014,
+      "loss": 3.2024,
+      "step": 8344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5214023590087891,
+      "learning_rate": 0.0005729055858818537,
+      "loss": 3.4417,
+      "step": 8345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5415693521499634,
+      "learning_rate": 0.0005728991915107252,
+      "loss": 3.4384,
+      "step": 8346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5267033576965332,
+      "learning_rate": 0.0005728927964208329,
+      "loss": 3.1189,
+      "step": 8347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5356072783470154,
+      "learning_rate": 0.0005728864006121934,
+      "loss": 3.3506,
+      "step": 8348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5404013395309448,
+      "learning_rate": 0.0005728800040848235,
+      "loss": 3.3664,
+      "step": 8349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5271034836769104,
+      "learning_rate": 0.0005728736068387404,
+      "loss": 3.3377,
+      "step": 8350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5246227383613586,
+      "learning_rate": 0.0005728672088739606,
+      "loss": 3.4265,
+      "step": 8351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4726375937461853,
+      "learning_rate": 0.0005728608101905012,
+      "loss": 3.2613,
+      "step": 8352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5876821875572205,
+      "learning_rate": 0.0005728544107883788,
+      "loss": 3.246,
+      "step": 8353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5646312832832336,
+      "learning_rate": 0.0005728480106676105,
+      "loss": 3.5403,
+      "step": 8354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5237587690353394,
+      "learning_rate": 0.000572841609828213,
+      "loss": 3.3111,
+      "step": 8355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5066969394683838,
+      "learning_rate": 0.0005728352082702032,
+      "loss": 3.2042,
+      "step": 8356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5408262610435486,
+      "learning_rate": 0.000572828805993598,
+      "loss": 3.1741,
+      "step": 8357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4968174695968628,
+      "learning_rate": 0.0005728224029984142,
+      "loss": 3.2532,
+      "step": 8358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5312142372131348,
+      "learning_rate": 0.0005728159992846687,
+      "loss": 3.1769,
+      "step": 8359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5489519834518433,
+      "learning_rate": 0.0005728095948523784,
+      "loss": 3.2044,
+      "step": 8360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5215625166893005,
+      "learning_rate": 0.0005728031897015601,
+      "loss": 3.2395,
+      "step": 8361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4845183491706848,
+      "learning_rate": 0.0005727967838322307,
+      "loss": 3.3525,
+      "step": 8362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5335796475410461,
+      "learning_rate": 0.0005727903772444071,
+      "loss": 3.3049,
+      "step": 8363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5164490342140198,
+      "learning_rate": 0.0005727839699381062,
+      "loss": 3.1817,
+      "step": 8364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5204142332077026,
+      "learning_rate": 0.0005727775619133446,
+      "loss": 3.0639,
+      "step": 8365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5255699753761292,
+      "learning_rate": 0.0005727711531701396,
+      "loss": 3.285,
+      "step": 8366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.527921736240387,
+      "learning_rate": 0.0005727647437085078,
+      "loss": 3.4429,
+      "step": 8367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5462525486946106,
+      "learning_rate": 0.0005727583335284662,
+      "loss": 3.5266,
+      "step": 8368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5473728179931641,
+      "learning_rate": 0.0005727519226300317,
+      "loss": 3.3949,
+      "step": 8369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5657687187194824,
+      "learning_rate": 0.000572745511013221,
+      "loss": 3.1723,
+      "step": 8370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5343332886695862,
+      "learning_rate": 0.0005727390986780513,
+      "loss": 3.2042,
+      "step": 8371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4889446794986725,
+      "learning_rate": 0.0005727326856245391,
+      "loss": 3.2929,
+      "step": 8372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5153390169143677,
+      "learning_rate": 0.0005727262718527014,
+      "loss": 3.5404,
+      "step": 8373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5464727282524109,
+      "learning_rate": 0.0005727198573625555,
+      "loss": 3.449,
+      "step": 8374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5315839648246765,
+      "learning_rate": 0.0005727134421541179,
+      "loss": 3.2474,
+      "step": 8375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5293451547622681,
+      "learning_rate": 0.0005727070262274054,
+      "loss": 3.3761,
+      "step": 8376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5146985054016113,
+      "learning_rate": 0.0005727006095824352,
+      "loss": 3.2667,
+      "step": 8377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5047792196273804,
+      "learning_rate": 0.0005726941922192242,
+      "loss": 3.4058,
+      "step": 8378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5004801154136658,
+      "learning_rate": 0.0005726877741377889,
+      "loss": 3.2643,
+      "step": 8379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5195822715759277,
+      "learning_rate": 0.0005726813553381467,
+      "loss": 3.354,
+      "step": 8380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5108522772789001,
+      "learning_rate": 0.0005726749358203142,
+      "loss": 3.2618,
+      "step": 8381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5223830342292786,
+      "learning_rate": 0.0005726685155843083,
+      "loss": 3.3946,
+      "step": 8382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49974021315574646,
+      "learning_rate": 0.0005726620946301462,
+      "loss": 3.1381,
+      "step": 8383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5291284918785095,
+      "learning_rate": 0.0005726556729578445,
+      "loss": 3.3743,
+      "step": 8384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5420185327529907,
+      "learning_rate": 0.0005726492505674204,
+      "loss": 3.245,
+      "step": 8385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5729690790176392,
+      "learning_rate": 0.0005726428274588904,
+      "loss": 3.1877,
+      "step": 8386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5272477269172668,
+      "learning_rate": 0.0005726364036322718,
+      "loss": 3.1366,
+      "step": 8387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5126119256019592,
+      "learning_rate": 0.0005726299790875814,
+      "loss": 3.2684,
+      "step": 8388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5218641757965088,
+      "learning_rate": 0.000572623553824836,
+      "loss": 3.1348,
+      "step": 8389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.505780041217804,
+      "learning_rate": 0.0005726171278440527,
+      "loss": 3.2774,
+      "step": 8390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5334300994873047,
+      "learning_rate": 0.0005726107011452484,
+      "loss": 3.3003,
+      "step": 8391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5299844741821289,
+      "learning_rate": 0.0005726042737284399,
+      "loss": 3.1663,
+      "step": 8392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5310911536216736,
+      "learning_rate": 0.0005725978455936443,
+      "loss": 3.3384,
+      "step": 8393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5203625559806824,
+      "learning_rate": 0.0005725914167408784,
+      "loss": 3.3432,
+      "step": 8394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5236132144927979,
+      "learning_rate": 0.000572584987170159,
+      "loss": 3.2808,
+      "step": 8395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5143657326698303,
+      "learning_rate": 0.0005725785568815034,
+      "loss": 3.4455,
+      "step": 8396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5301834344863892,
+      "learning_rate": 0.0005725721258749282,
+      "loss": 3.3219,
+      "step": 8397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5046020150184631,
+      "learning_rate": 0.0005725656941504507,
+      "loss": 3.4859,
+      "step": 8398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49806228280067444,
+      "learning_rate": 0.0005725592617080873,
+      "loss": 3.3874,
+      "step": 8399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5187866687774658,
+      "learning_rate": 0.0005725528285478555,
+      "loss": 3.5443,
+      "step": 8400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5415318608283997,
+      "learning_rate": 0.000572546394669772,
+      "loss": 3.2546,
+      "step": 8401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4997556209564209,
+      "learning_rate": 0.0005725399600738537,
+      "loss": 3.4427,
+      "step": 8402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5891625285148621,
+      "learning_rate": 0.0005725335247601176,
+      "loss": 3.5401,
+      "step": 8403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6055923104286194,
+      "learning_rate": 0.0005725270887285806,
+      "loss": 3.1657,
+      "step": 8404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5357984304428101,
+      "learning_rate": 0.0005725206519792597,
+      "loss": 3.3858,
+      "step": 8405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5380067825317383,
+      "learning_rate": 0.0005725142145121719,
+      "loss": 3.2251,
+      "step": 8406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5393832325935364,
+      "learning_rate": 0.0005725077763273341,
+      "loss": 3.3881,
+      "step": 8407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4998052716255188,
+      "learning_rate": 0.0005725013374247633,
+      "loss": 3.5416,
+      "step": 8408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5088112950325012,
+      "learning_rate": 0.0005724948978044763,
+      "loss": 3.5218,
+      "step": 8409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5030485987663269,
+      "learning_rate": 0.0005724884574664903,
+      "loss": 3.4079,
+      "step": 8410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5525428652763367,
+      "learning_rate": 0.000572482016410822,
+      "loss": 3.2823,
+      "step": 8411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5205178260803223,
+      "learning_rate": 0.0005724755746374887,
+      "loss": 3.3983,
+      "step": 8412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5077677369117737,
+      "learning_rate": 0.0005724691321465071,
+      "loss": 3.2847,
+      "step": 8413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.568877100944519,
+      "learning_rate": 0.0005724626889378942,
+      "loss": 3.323,
+      "step": 8414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.563330352306366,
+      "learning_rate": 0.0005724562450116669,
+      "loss": 3.1247,
+      "step": 8415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5267123579978943,
+      "learning_rate": 0.0005724498003678425,
+      "loss": 3.296,
+      "step": 8416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.483675479888916,
+      "learning_rate": 0.0005724433550064376,
+      "loss": 3.4229,
+      "step": 8417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5056666731834412,
+      "learning_rate": 0.0005724369089274695,
+      "loss": 3.3706,
+      "step": 8418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5181419253349304,
+      "learning_rate": 0.000572430462130955,
+      "loss": 3.2534,
+      "step": 8419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.502122700214386,
+      "learning_rate": 0.000572424014616911,
+      "loss": 3.5067,
+      "step": 8420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5571039319038391,
+      "learning_rate": 0.0005724175663853545,
+      "loss": 2.9924,
+      "step": 8421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5149720907211304,
+      "learning_rate": 0.0005724111174363026,
+      "loss": 3.1874,
+      "step": 8422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5191240310668945,
+      "learning_rate": 0.0005724046677697724,
+      "loss": 3.4348,
+      "step": 8423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5127260684967041,
+      "learning_rate": 0.0005723982173857806,
+      "loss": 3.1189,
+      "step": 8424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5314476490020752,
+      "learning_rate": 0.0005723917662843444,
+      "loss": 3.1581,
+      "step": 8425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5084034204483032,
+      "learning_rate": 0.0005723853144654806,
+      "loss": 3.2397,
+      "step": 8426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4968039095401764,
+      "learning_rate": 0.0005723788619292064,
+      "loss": 3.2453,
+      "step": 8427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5596404671669006,
+      "learning_rate": 0.0005723724086755386,
+      "loss": 3.329,
+      "step": 8428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5200359225273132,
+      "learning_rate": 0.0005723659547044944,
+      "loss": 3.4476,
+      "step": 8429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48586541414260864,
+      "learning_rate": 0.0005723595000160906,
+      "loss": 3.3514,
+      "step": 8430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5171741843223572,
+      "learning_rate": 0.0005723530446103443,
+      "loss": 3.2959,
+      "step": 8431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4921620190143585,
+      "learning_rate": 0.0005723465884872726,
+      "loss": 3.3176,
+      "step": 8432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5125603675842285,
+      "learning_rate": 0.0005723401316468923,
+      "loss": 3.377,
+      "step": 8433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5400583744049072,
+      "learning_rate": 0.0005723336740892206,
+      "loss": 3.1951,
+      "step": 8434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5424138903617859,
+      "learning_rate": 0.0005723272158142744,
+      "loss": 3.1451,
+      "step": 8435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5870009064674377,
+      "learning_rate": 0.0005723207568220707,
+      "loss": 3.5986,
+      "step": 8436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5539098381996155,
+      "learning_rate": 0.0005723142971126265,
+      "loss": 3.2467,
+      "step": 8437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.520626425743103,
+      "learning_rate": 0.0005723078366859588,
+      "loss": 3.4634,
+      "step": 8438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.501064658164978,
+      "learning_rate": 0.0005723013755420847,
+      "loss": 3.1619,
+      "step": 8439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5047374963760376,
+      "learning_rate": 0.0005722949136810212,
+      "loss": 3.2784,
+      "step": 8440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5190777778625488,
+      "learning_rate": 0.0005722884511027853,
+      "loss": 3.4443,
+      "step": 8441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5325513482093811,
+      "learning_rate": 0.0005722819878073942,
+      "loss": 3.2189,
+      "step": 8442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5083960294723511,
+      "learning_rate": 0.0005722755237948645,
+      "loss": 3.4314,
+      "step": 8443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5294110178947449,
+      "learning_rate": 0.0005722690590652136,
+      "loss": 3.4491,
+      "step": 8444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5367835164070129,
+      "learning_rate": 0.0005722625936184583,
+      "loss": 3.3437,
+      "step": 8445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5289018750190735,
+      "learning_rate": 0.0005722561274546158,
+      "loss": 3.4518,
+      "step": 8446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5518658757209778,
+      "learning_rate": 0.0005722496605737031,
+      "loss": 3.4819,
+      "step": 8447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5337439179420471,
+      "learning_rate": 0.0005722431929757371,
+      "loss": 3.3336,
+      "step": 8448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5443679094314575,
+      "learning_rate": 0.0005722367246607349,
+      "loss": 3.2325,
+      "step": 8449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5037354230880737,
+      "learning_rate": 0.0005722302556287137,
+      "loss": 3.1895,
+      "step": 8450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51902174949646,
+      "learning_rate": 0.0005722237858796903,
+      "loss": 3.3569,
+      "step": 8451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5206155776977539,
+      "learning_rate": 0.0005722173154136818,
+      "loss": 3.2948,
+      "step": 8452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5317664742469788,
+      "learning_rate": 0.0005722108442307054,
+      "loss": 3.4704,
+      "step": 8453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.527557909488678,
+      "learning_rate": 0.000572204372330778,
+      "loss": 3.2911,
+      "step": 8454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5252202153205872,
+      "learning_rate": 0.0005721978997139165,
+      "loss": 3.3554,
+      "step": 8455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5341652035713196,
+      "learning_rate": 0.0005721914263801382,
+      "loss": 3.0762,
+      "step": 8456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5101057887077332,
+      "learning_rate": 0.0005721849523294602,
+      "loss": 3.3628,
+      "step": 8457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5079684853553772,
+      "learning_rate": 0.0005721784775618993,
+      "loss": 3.4156,
+      "step": 8458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5565444231033325,
+      "learning_rate": 0.0005721720020774727,
+      "loss": 3.3037,
+      "step": 8459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5274984240531921,
+      "learning_rate": 0.0005721655258761973,
+      "loss": 3.2268,
+      "step": 8460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5040982365608215,
+      "learning_rate": 0.0005721590489580904,
+      "loss": 3.4673,
+      "step": 8461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5225953459739685,
+      "learning_rate": 0.000572152571323169,
+      "loss": 3.4349,
+      "step": 8462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5098921656608582,
+      "learning_rate": 0.0005721460929714501,
+      "loss": 3.4011,
+      "step": 8463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5205847024917603,
+      "learning_rate": 0.0005721396139029507,
+      "loss": 3.2664,
+      "step": 8464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5714619755744934,
+      "learning_rate": 0.000572133134117688,
+      "loss": 3.2306,
+      "step": 8465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5325881838798523,
+      "learning_rate": 0.0005721266536156789,
+      "loss": 3.219,
+      "step": 8466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5481316447257996,
+      "learning_rate": 0.0005721201723969407,
+      "loss": 3.3076,
+      "step": 8467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4882158637046814,
+      "learning_rate": 0.0005721136904614901,
+      "loss": 3.3831,
+      "step": 8468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5075830817222595,
+      "learning_rate": 0.0005721072078093447,
+      "loss": 3.3366,
+      "step": 8469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5224390029907227,
+      "learning_rate": 0.0005721007244405211,
+      "loss": 3.4743,
+      "step": 8470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5543192028999329,
+      "learning_rate": 0.0005720942403550366,
+      "loss": 3.2743,
+      "step": 8471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.572250485420227,
+      "learning_rate": 0.0005720877555529082,
+      "loss": 3.2516,
+      "step": 8472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5291645526885986,
+      "learning_rate": 0.0005720812700341531,
+      "loss": 3.2301,
+      "step": 8473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5106573104858398,
+      "learning_rate": 0.0005720747837987882,
+      "loss": 3.3192,
+      "step": 8474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5043050646781921,
+      "learning_rate": 0.0005720682968468308,
+      "loss": 3.3815,
+      "step": 8475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5045797824859619,
+      "learning_rate": 0.0005720618091782978,
+      "loss": 3.2098,
+      "step": 8476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5489840507507324,
+      "learning_rate": 0.0005720553207932064,
+      "loss": 3.3597,
+      "step": 8477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5759950876235962,
+      "learning_rate": 0.0005720488316915736,
+      "loss": 3.2866,
+      "step": 8478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4919251799583435,
+      "learning_rate": 0.0005720423418734164,
+      "loss": 3.3322,
+      "step": 8479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5042932033538818,
+      "learning_rate": 0.0005720358513387522,
+      "loss": 3.3563,
+      "step": 8480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5451899766921997,
+      "learning_rate": 0.0005720293600875979,
+      "loss": 3.3275,
+      "step": 8481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5713815689086914,
+      "learning_rate": 0.0005720228681199707,
+      "loss": 3.2709,
+      "step": 8482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5565445423126221,
+      "learning_rate": 0.0005720163754358874,
+      "loss": 3.3623,
+      "step": 8483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5319048166275024,
+      "learning_rate": 0.0005720098820353655,
+      "loss": 3.2325,
+      "step": 8484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5170350670814514,
+      "learning_rate": 0.0005720033879184219,
+      "loss": 3.2892,
+      "step": 8485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5210342407226562,
+      "learning_rate": 0.0005719968930850736,
+      "loss": 3.2944,
+      "step": 8486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5216629505157471,
+      "learning_rate": 0.000571990397535338,
+      "loss": 3.4078,
+      "step": 8487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5704594254493713,
+      "learning_rate": 0.0005719839012692319,
+      "loss": 3.3264,
+      "step": 8488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5190528035163879,
+      "learning_rate": 0.0005719774042867726,
+      "loss": 3.11,
+      "step": 8489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5578228235244751,
+      "learning_rate": 0.0005719709065879771,
+      "loss": 3.4529,
+      "step": 8490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5134179592132568,
+      "learning_rate": 0.0005719644081728627,
+      "loss": 3.1287,
+      "step": 8491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5582857131958008,
+      "learning_rate": 0.0005719579090414464,
+      "loss": 3.1355,
+      "step": 8492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5763298273086548,
+      "learning_rate": 0.0005719514091937451,
+      "loss": 3.2548,
+      "step": 8493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5173172950744629,
+      "learning_rate": 0.0005719449086297762,
+      "loss": 3.3526,
+      "step": 8494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0055021047592163,
+      "learning_rate": 0.0005719384073495569,
+      "loss": 3.1682,
+      "step": 8495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49112802743911743,
+      "learning_rate": 0.000571931905353104,
+      "loss": 3.5877,
+      "step": 8496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5321609377861023,
+      "learning_rate": 0.0005719254026404349,
+      "loss": 3.2739,
+      "step": 8497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5343760848045349,
+      "learning_rate": 0.0005719188992115667,
+      "loss": 3.4689,
+      "step": 8498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5921098589897156,
+      "learning_rate": 0.0005719123950665162,
+      "loss": 3.229,
+      "step": 8499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5534712672233582,
+      "learning_rate": 0.0005719058902053009,
+      "loss": 3.2989,
+      "step": 8500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.513491690158844,
+      "learning_rate": 0.000571899384627938,
+      "loss": 3.183,
+      "step": 8501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5323536396026611,
+      "learning_rate": 0.0005718928783344442,
+      "loss": 3.2725,
+      "step": 8502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6444754004478455,
+      "learning_rate": 0.0005718863713248371,
+      "loss": 3.3666,
+      "step": 8503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48075416684150696,
+      "learning_rate": 0.0005718798635991334,
+      "loss": 3.1333,
+      "step": 8504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.507790207862854,
+      "learning_rate": 0.0005718733551573506,
+      "loss": 3.3297,
+      "step": 8505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5401341915130615,
+      "learning_rate": 0.0005718668459995056,
+      "loss": 3.3617,
+      "step": 8506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5216659307479858,
+      "learning_rate": 0.0005718603361256157,
+      "loss": 3.3293,
+      "step": 8507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5784571170806885,
+      "learning_rate": 0.0005718538255356981,
+      "loss": 3.2978,
+      "step": 8508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5394245982170105,
+      "learning_rate": 0.0005718473142297697,
+      "loss": 3.1971,
+      "step": 8509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.547707200050354,
+      "learning_rate": 0.0005718408022078479,
+      "loss": 3.4488,
+      "step": 8510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5773078203201294,
+      "learning_rate": 0.0005718342894699497,
+      "loss": 3.4022,
+      "step": 8511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5116841197013855,
+      "learning_rate": 0.0005718277760160922,
+      "loss": 3.5378,
+      "step": 8512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5021461844444275,
+      "learning_rate": 0.0005718212618462928,
+      "loss": 3.2106,
+      "step": 8513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5065045952796936,
+      "learning_rate": 0.0005718147469605684,
+      "loss": 3.2924,
+      "step": 8514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5146911144256592,
+      "learning_rate": 0.0005718082313589363,
+      "loss": 3.4854,
+      "step": 8515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5704720616340637,
+      "learning_rate": 0.0005718017150414137,
+      "loss": 3.1635,
+      "step": 8516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5921428799629211,
+      "learning_rate": 0.0005717951980080176,
+      "loss": 3.3375,
+      "step": 8517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5015804767608643,
+      "learning_rate": 0.0005717886802587653,
+      "loss": 3.2446,
+      "step": 8518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5033897161483765,
+      "learning_rate": 0.0005717821617936739,
+      "loss": 3.4058,
+      "step": 8519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5112386345863342,
+      "learning_rate": 0.0005717756426127606,
+      "loss": 3.4671,
+      "step": 8520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5142545104026794,
+      "learning_rate": 0.0005717691227160426,
+      "loss": 3.3429,
+      "step": 8521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5595449209213257,
+      "learning_rate": 0.000571762602103537,
+      "loss": 3.4415,
+      "step": 8522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5375891327857971,
+      "learning_rate": 0.000571756080775261,
+      "loss": 3.2085,
+      "step": 8523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5820460319519043,
+      "learning_rate": 0.0005717495587312318,
+      "loss": 3.2163,
+      "step": 8524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5265014171600342,
+      "learning_rate": 0.0005717430359714666,
+      "loss": 3.0536,
+      "step": 8525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5012522339820862,
+      "learning_rate": 0.0005717365124959824,
+      "loss": 3.2223,
+      "step": 8526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5129319429397583,
+      "learning_rate": 0.0005717299883047967,
+      "loss": 3.1573,
+      "step": 8527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5185990333557129,
+      "learning_rate": 0.0005717234633979265,
+      "loss": 3.1524,
+      "step": 8528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5384962558746338,
+      "learning_rate": 0.0005717169377753888,
+      "loss": 3.2809,
+      "step": 8529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5573104023933411,
+      "learning_rate": 0.0005717104114372012,
+      "loss": 3.3692,
+      "step": 8530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5535567402839661,
+      "learning_rate": 0.0005717038843833805,
+      "loss": 3.2744,
+      "step": 8531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.543593168258667,
+      "learning_rate": 0.0005716973566139441,
+      "loss": 3.3109,
+      "step": 8532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5009624361991882,
+      "learning_rate": 0.0005716908281289092,
+      "loss": 3.4917,
+      "step": 8533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5481317043304443,
+      "learning_rate": 0.000571684298928293,
+      "loss": 3.3068,
+      "step": 8534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.538657546043396,
+      "learning_rate": 0.0005716777690121125,
+      "loss": 3.409,
+      "step": 8535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5731958150863647,
+      "learning_rate": 0.0005716712383803851,
+      "loss": 3.1841,
+      "step": 8536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5232669115066528,
+      "learning_rate": 0.000571664707033128,
+      "loss": 3.231,
+      "step": 8537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5260790586471558,
+      "learning_rate": 0.0005716581749703583,
+      "loss": 3.3928,
+      "step": 8538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5380903482437134,
+      "learning_rate": 0.0005716516421920932,
+      "loss": 3.2905,
+      "step": 8539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5121652483940125,
+      "learning_rate": 0.00057164510869835,
+      "loss": 3.3719,
+      "step": 8540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5237699747085571,
+      "learning_rate": 0.0005716385744891459,
+      "loss": 3.273,
+      "step": 8541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6156141757965088,
+      "learning_rate": 0.000571632039564498,
+      "loss": 3.1406,
+      "step": 8542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48015645146369934,
+      "learning_rate": 0.0005716255039244235,
+      "loss": 3.4731,
+      "step": 8543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5515682697296143,
+      "learning_rate": 0.0005716189675689399,
+      "loss": 3.1267,
+      "step": 8544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5330041646957397,
+      "learning_rate": 0.0005716124304980642,
+      "loss": 3.2132,
+      "step": 8545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.537356972694397,
+      "learning_rate": 0.0005716058927118135,
+      "loss": 3.3974,
+      "step": 8546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5180966854095459,
+      "learning_rate": 0.0005715993542102052,
+      "loss": 3.3525,
+      "step": 8547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5257582664489746,
+      "learning_rate": 0.0005715928149932565,
+      "loss": 3.2291,
+      "step": 8548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5318938493728638,
+      "learning_rate": 0.0005715862750609845,
+      "loss": 2.958,
+      "step": 8549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5303199887275696,
+      "learning_rate": 0.0005715797344134067,
+      "loss": 3.3393,
+      "step": 8550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5268381834030151,
+      "learning_rate": 0.0005715731930505401,
+      "loss": 3.3426,
+      "step": 8551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6098791360855103,
+      "learning_rate": 0.0005715666509724019,
+      "loss": 3.0801,
+      "step": 8552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5364266037940979,
+      "learning_rate": 0.0005715601081790094,
+      "loss": 3.0393,
+      "step": 8553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5054658651351929,
+      "learning_rate": 0.0005715535646703798,
+      "loss": 3.3621,
+      "step": 8554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5238931179046631,
+      "learning_rate": 0.0005715470204465305,
+      "loss": 3.2215,
+      "step": 8555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390506386756897,
+      "learning_rate": 0.0005715404755074785,
+      "loss": 3.4286,
+      "step": 8556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5367797613143921,
+      "learning_rate": 0.0005715339298532412,
+      "loss": 3.428,
+      "step": 8557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5821202397346497,
+      "learning_rate": 0.0005715273834838358,
+      "loss": 3.3585,
+      "step": 8558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.523921549320221,
+      "learning_rate": 0.0005715208363992794,
+      "loss": 3.2719,
+      "step": 8559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5809316635131836,
+      "learning_rate": 0.0005715142885995895,
+      "loss": 3.1228,
+      "step": 8560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5245625376701355,
+      "learning_rate": 0.0005715077400847832,
+      "loss": 3.771,
+      "step": 8561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5554155707359314,
+      "learning_rate": 0.0005715011908548778,
+      "loss": 3.1731,
+      "step": 8562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.530573308467865,
+      "learning_rate": 0.0005714946409098905,
+      "loss": 3.2805,
+      "step": 8563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5505214333534241,
+      "learning_rate": 0.0005714880902498385,
+      "loss": 3.2432,
+      "step": 8564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5658882260322571,
+      "learning_rate": 0.0005714815388747391,
+      "loss": 3.2784,
+      "step": 8565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5260425209999084,
+      "learning_rate": 0.0005714749867846097,
+      "loss": 3.2037,
+      "step": 8566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4990243911743164,
+      "learning_rate": 0.0005714684339794674,
+      "loss": 3.3172,
+      "step": 8567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5792827606201172,
+      "learning_rate": 0.0005714618804593295,
+      "loss": 3.5224,
+      "step": 8568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5181724429130554,
+      "learning_rate": 0.0005714553262242131,
+      "loss": 3.3583,
+      "step": 8569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.564866304397583,
+      "learning_rate": 0.0005714487712741357,
+      "loss": 3.4391,
+      "step": 8570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4975354075431824,
+      "learning_rate": 0.0005714422156091146,
+      "loss": 3.2768,
+      "step": 8571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5749588012695312,
+      "learning_rate": 0.0005714356592291668,
+      "loss": 3.2569,
+      "step": 8572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49790215492248535,
+      "learning_rate": 0.0005714291021343097,
+      "loss": 3.2321,
+      "step": 8573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5509989857673645,
+      "learning_rate": 0.0005714225443245607,
+      "loss": 3.2786,
+      "step": 8574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5187534093856812,
+      "learning_rate": 0.0005714159857999368,
+      "loss": 3.241,
+      "step": 8575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5033349394798279,
+      "learning_rate": 0.0005714094265604556,
+      "loss": 3.3021,
+      "step": 8576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5202603936195374,
+      "learning_rate": 0.0005714028666061341,
+      "loss": 3.1149,
+      "step": 8577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.544248104095459,
+      "learning_rate": 0.0005713963059369898,
+      "loss": 3.5272,
+      "step": 8578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5000052452087402,
+      "learning_rate": 0.0005713897445530396,
+      "loss": 3.278,
+      "step": 8579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5294511318206787,
+      "learning_rate": 0.0005713831824543013,
+      "loss": 3.3534,
+      "step": 8580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5449314713478088,
+      "learning_rate": 0.0005713766196407919,
+      "loss": 3.1439,
+      "step": 8581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5595104694366455,
+      "learning_rate": 0.0005713700561125286,
+      "loss": 3.2132,
+      "step": 8582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5275298953056335,
+      "learning_rate": 0.0005713634918695288,
+      "loss": 3.5211,
+      "step": 8583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5486336946487427,
+      "learning_rate": 0.0005713569269118099,
+      "loss": 3.3394,
+      "step": 8584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5621734261512756,
+      "learning_rate": 0.0005713503612393889,
+      "loss": 3.2691,
+      "step": 8585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5324231386184692,
+      "learning_rate": 0.0005713437948522834,
+      "loss": 3.1905,
+      "step": 8586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5473950505256653,
+      "learning_rate": 0.0005713372277505106,
+      "loss": 3.2829,
+      "step": 8587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5286551713943481,
+      "learning_rate": 0.0005713306599340877,
+      "loss": 3.2689,
+      "step": 8588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5888059139251709,
+      "learning_rate": 0.000571324091403032,
+      "loss": 3.4355,
+      "step": 8589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5260636210441589,
+      "learning_rate": 0.000571317522157361,
+      "loss": 3.4292,
+      "step": 8590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5505531430244446,
+      "learning_rate": 0.0005713109521970918,
+      "loss": 3.3409,
+      "step": 8591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5252068638801575,
+      "learning_rate": 0.0005713043815222418,
+      "loss": 3.3137,
+      "step": 8592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5074268579483032,
+      "learning_rate": 0.0005712978101328281,
+      "loss": 3.2255,
+      "step": 8593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4985024631023407,
+      "learning_rate": 0.0005712912380288683,
+      "loss": 3.3656,
+      "step": 8594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5729339122772217,
+      "learning_rate": 0.0005712846652103796,
+      "loss": 3.1283,
+      "step": 8595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5986528396606445,
+      "learning_rate": 0.0005712780916773794,
+      "loss": 3.2123,
+      "step": 8596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5517857670783997,
+      "learning_rate": 0.0005712715174298848,
+      "loss": 3.3586,
+      "step": 8597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5476078391075134,
+      "learning_rate": 0.0005712649424679132,
+      "loss": 3.2001,
+      "step": 8598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5720165371894836,
+      "learning_rate": 0.000571258366791482,
+      "loss": 2.9817,
+      "step": 8599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5489450693130493,
+      "learning_rate": 0.0005712517904006085,
+      "loss": 3.4279,
+      "step": 8600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5727310180664062,
+      "learning_rate": 0.0005712452132953099,
+      "loss": 3.1009,
+      "step": 8601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5166281461715698,
+      "learning_rate": 0.0005712386354756037,
+      "loss": 3.3062,
+      "step": 8602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5071576237678528,
+      "learning_rate": 0.0005712320569415071,
+      "loss": 3.4458,
+      "step": 8603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5760843753814697,
+      "learning_rate": 0.0005712254776930374,
+      "loss": 3.3133,
+      "step": 8604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5532389283180237,
+      "learning_rate": 0.0005712188977302121,
+      "loss": 3.1873,
+      "step": 8605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5423989295959473,
+      "learning_rate": 0.0005712123170530484,
+      "loss": 3.0871,
+      "step": 8606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5491202473640442,
+      "learning_rate": 0.0005712057356615637,
+      "loss": 3.176,
+      "step": 8607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5447011590003967,
+      "learning_rate": 0.0005711991535557751,
+      "loss": 3.1241,
+      "step": 8608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5630657076835632,
+      "learning_rate": 0.0005711925707357002,
+      "loss": 3.3227,
+      "step": 8609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4960578978061676,
+      "learning_rate": 0.0005711859872013563,
+      "loss": 3.3174,
+      "step": 8610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5437517166137695,
+      "learning_rate": 0.0005711794029527607,
+      "loss": 3.3928,
+      "step": 8611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5181750059127808,
+      "learning_rate": 0.0005711728179899308,
+      "loss": 3.3892,
+      "step": 8612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5108785033226013,
+      "learning_rate": 0.0005711662323128838,
+      "loss": 3.3299,
+      "step": 8613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4989277720451355,
+      "learning_rate": 0.0005711596459216372,
+      "loss": 3.3234,
+      "step": 8614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5421345233917236,
+      "learning_rate": 0.0005711530588162082,
+      "loss": 3.3554,
+      "step": 8615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5189564228057861,
+      "learning_rate": 0.0005711464709966142,
+      "loss": 3.2916,
+      "step": 8616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5410189628601074,
+      "learning_rate": 0.0005711398824628727,
+      "loss": 3.3517,
+      "step": 8617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5351953506469727,
+      "learning_rate": 0.0005711332932150008,
+      "loss": 3.1984,
+      "step": 8618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5187792181968689,
+      "learning_rate": 0.0005711267032530161,
+      "loss": 3.3928,
+      "step": 8619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5344939231872559,
+      "learning_rate": 0.0005711201125769358,
+      "loss": 3.3316,
+      "step": 8620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5304107666015625,
+      "learning_rate": 0.0005711135211867773,
+      "loss": 3.4969,
+      "step": 8621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5163986682891846,
+      "learning_rate": 0.0005711069290825579,
+      "loss": 3.117,
+      "step": 8622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5484994649887085,
+      "learning_rate": 0.0005711003362642951,
+      "loss": 3.3076,
+      "step": 8623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5722675323486328,
+      "learning_rate": 0.0005710937427320062,
+      "loss": 3.437,
+      "step": 8624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5870904326438904,
+      "learning_rate": 0.0005710871484857085,
+      "loss": 3.3008,
+      "step": 8625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5617459416389465,
+      "learning_rate": 0.0005710805535254195,
+      "loss": 3.244,
+      "step": 8626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5233004689216614,
+      "learning_rate": 0.0005710739578511564,
+      "loss": 3.3681,
+      "step": 8627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5104897618293762,
+      "learning_rate": 0.0005710673614629367,
+      "loss": 3.3192,
+      "step": 8628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5213035345077515,
+      "learning_rate": 0.0005710607643607778,
+      "loss": 3.4488,
+      "step": 8629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5398159623146057,
+      "learning_rate": 0.0005710541665446969,
+      "loss": 3.4134,
+      "step": 8630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5116217136383057,
+      "learning_rate": 0.0005710475680147115,
+      "loss": 3.1915,
+      "step": 8631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5108386278152466,
+      "learning_rate": 0.0005710409687708391,
+      "loss": 3.2113,
+      "step": 8632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5582148432731628,
+      "learning_rate": 0.0005710343688130968,
+      "loss": 3.2279,
+      "step": 8633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5281783938407898,
+      "learning_rate": 0.0005710277681415022,
+      "loss": 3.4379,
+      "step": 8634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4925077259540558,
+      "learning_rate": 0.0005710211667560726,
+      "loss": 3.2618,
+      "step": 8635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5662412047386169,
+      "learning_rate": 0.0005710145646568254,
+      "loss": 3.2086,
+      "step": 8636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.528595507144928,
+      "learning_rate": 0.0005710079618437781,
+      "loss": 3.452,
+      "step": 8637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5423139333724976,
+      "learning_rate": 0.000571001358316948,
+      "loss": 3.2282,
+      "step": 8638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5846478939056396,
+      "learning_rate": 0.0005709947540763524,
+      "loss": 3.3114,
+      "step": 8639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.543500542640686,
+      "learning_rate": 0.0005709881491220087,
+      "loss": 3.2656,
+      "step": 8640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5225181579589844,
+      "learning_rate": 0.0005709815434539344,
+      "loss": 3.2979,
+      "step": 8641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5145890712738037,
+      "learning_rate": 0.0005709749370721469,
+      "loss": 3.0179,
+      "step": 8642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.527047336101532,
+      "learning_rate": 0.0005709683299766635,
+      "loss": 3.1937,
+      "step": 8643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.558195948600769,
+      "learning_rate": 0.0005709617221675017,
+      "loss": 3.3409,
+      "step": 8644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5157774686813354,
+      "learning_rate": 0.000570955113644679,
+      "loss": 3.4396,
+      "step": 8645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5344122052192688,
+      "learning_rate": 0.0005709485044082125,
+      "loss": 3.3056,
+      "step": 8646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5069317817687988,
+      "learning_rate": 0.0005709418944581199,
+      "loss": 3.2913,
+      "step": 8647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5138603448867798,
+      "learning_rate": 0.0005709352837944184,
+      "loss": 3.3517,
+      "step": 8648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5492541193962097,
+      "learning_rate": 0.0005709286724171256,
+      "loss": 3.3809,
+      "step": 8649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6118084192276001,
+      "learning_rate": 0.0005709220603262587,
+      "loss": 3.1827,
+      "step": 8650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.522902250289917,
+      "learning_rate": 0.0005709154475218354,
+      "loss": 3.2828,
+      "step": 8651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5477918386459351,
+      "learning_rate": 0.0005709088340038729,
+      "loss": 3.431,
+      "step": 8652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0005709022197723886,
+      "loss": 3.0517,
+      "step": 8653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.563598096370697,
+      "learning_rate": 0.0005708956048273999,
+      "loss": 3.1275,
+      "step": 8654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5188009738922119,
+      "learning_rate": 0.0005708889891689245,
+      "loss": 3.1314,
+      "step": 8655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5175809264183044,
+      "learning_rate": 0.0005708823727969796,
+      "loss": 3.1442,
+      "step": 8656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5397357940673828,
+      "learning_rate": 0.0005708757557115826,
+      "loss": 3.1639,
+      "step": 8657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0455191135406494,
+      "learning_rate": 0.000570869137912751,
+      "loss": 3.2676,
+      "step": 8658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.571491003036499,
+      "learning_rate": 0.0005708625194005023,
+      "loss": 3.4802,
+      "step": 8659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5649874210357666,
+      "learning_rate": 0.0005708559001748538,
+      "loss": 3.1757,
+      "step": 8660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234768986701965,
+      "learning_rate": 0.000570849280235823,
+      "loss": 3.3455,
+      "step": 8661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5115209817886353,
+      "learning_rate": 0.0005708426595834273,
+      "loss": 3.167,
+      "step": 8662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5697651505470276,
+      "learning_rate": 0.0005708360382176841,
+      "loss": 3.4317,
+      "step": 8663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5905675292015076,
+      "learning_rate": 0.0005708294161386109,
+      "loss": 3.2357,
+      "step": 8664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5598018765449524,
+      "learning_rate": 0.0005708227933462252,
+      "loss": 3.2358,
+      "step": 8665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5009368658065796,
+      "learning_rate": 0.0005708161698405444,
+      "loss": 3.2861,
+      "step": 8666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5382165312767029,
+      "learning_rate": 0.0005708095456215859,
+      "loss": 3.3066,
+      "step": 8667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5542304515838623,
+      "learning_rate": 0.0005708029206893672,
+      "loss": 3.4175,
+      "step": 8668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5310438871383667,
+      "learning_rate": 0.0005707962950439057,
+      "loss": 3.6001,
+      "step": 8669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5058013200759888,
+      "learning_rate": 0.0005707896686852189,
+      "loss": 3.3482,
+      "step": 8670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5328607559204102,
+      "learning_rate": 0.0005707830416133243,
+      "loss": 3.2293,
+      "step": 8671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4994487464427948,
+      "learning_rate": 0.0005707764138282391,
+      "loss": 3.3414,
+      "step": 8672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.547566294670105,
+      "learning_rate": 0.0005707697853299811,
+      "loss": 3.4422,
+      "step": 8673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5558493137359619,
+      "learning_rate": 0.0005707631561185675,
+      "loss": 3.3349,
+      "step": 8674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5128504633903503,
+      "learning_rate": 0.0005707565261940158,
+      "loss": 3.3234,
+      "step": 8675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.543178141117096,
+      "learning_rate": 0.0005707498955563437,
+      "loss": 3.3392,
+      "step": 8676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5692942142486572,
+      "learning_rate": 0.0005707432642055683,
+      "loss": 3.4206,
+      "step": 8677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49869203567504883,
+      "learning_rate": 0.0005707366321417073,
+      "loss": 3.104,
+      "step": 8678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5554705262184143,
+      "learning_rate": 0.0005707299993647782,
+      "loss": 3.2966,
+      "step": 8679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5459063053131104,
+      "learning_rate": 0.0005707233658747983,
+      "loss": 3.4036,
+      "step": 8680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5476892590522766,
+      "learning_rate": 0.0005707167316717852,
+      "loss": 3.3779,
+      "step": 8681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5351525545120239,
+      "learning_rate": 0.0005707100967557563,
+      "loss": 3.26,
+      "step": 8682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5344018936157227,
+      "learning_rate": 0.0005707034611267291,
+      "loss": 3.2521,
+      "step": 8683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5126794576644897,
+      "learning_rate": 0.0005706968247847212,
+      "loss": 3.086,
+      "step": 8684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5328701734542847,
+      "learning_rate": 0.0005706901877297498,
+      "loss": 3.2423,
+      "step": 8685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6240466237068176,
+      "learning_rate": 0.0005706835499618326,
+      "loss": 3.3378,
+      "step": 8686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49643224477767944,
+      "learning_rate": 0.000570676911480987,
+      "loss": 3.2525,
+      "step": 8687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5192905068397522,
+      "learning_rate": 0.0005706702722872305,
+      "loss": 3.369,
+      "step": 8688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5646792650222778,
+      "learning_rate": 0.0005706636323805807,
+      "loss": 3.3023,
+      "step": 8689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5123174786567688,
+      "learning_rate": 0.0005706569917610548,
+      "loss": 3.2993,
+      "step": 8690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5409421324729919,
+      "learning_rate": 0.0005706503504286707,
+      "loss": 3.163,
+      "step": 8691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5075324773788452,
+      "learning_rate": 0.0005706437083834456,
+      "loss": 3.4438,
+      "step": 8692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4756852686405182,
+      "learning_rate": 0.000570637065625397,
+      "loss": 3.3624,
+      "step": 8693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5012529492378235,
+      "learning_rate": 0.0005706304221545424,
+      "loss": 3.2455,
+      "step": 8694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5726290345191956,
+      "learning_rate": 0.0005706237779708994,
+      "loss": 3.11,
+      "step": 8695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5796265006065369,
+      "learning_rate": 0.0005706171330744854,
+      "loss": 3.2597,
+      "step": 8696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5194861888885498,
+      "learning_rate": 0.000570610487465318,
+      "loss": 3.0619,
+      "step": 8697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5499964952468872,
+      "learning_rate": 0.0005706038411434147,
+      "loss": 3.4847,
+      "step": 8698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5542723536491394,
+      "learning_rate": 0.000570597194108793,
+      "loss": 3.3495,
+      "step": 8699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5347164273262024,
+      "learning_rate": 0.0005705905463614702,
+      "loss": 3.288,
+      "step": 8700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5562871098518372,
+      "learning_rate": 0.0005705838979014642,
+      "loss": 3.3801,
+      "step": 8701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5251455307006836,
+      "learning_rate": 0.0005705772487287921,
+      "loss": 3.3115,
+      "step": 8702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.526947021484375,
+      "learning_rate": 0.0005705705988434716,
+      "loss": 3.2882,
+      "step": 8703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5250048637390137,
+      "learning_rate": 0.0005705639482455204,
+      "loss": 3.1308,
+      "step": 8704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5624373555183411,
+      "learning_rate": 0.0005705572969349556,
+      "loss": 3.2379,
+      "step": 8705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.600304365158081,
+      "learning_rate": 0.000570550644911795,
+      "loss": 3.2652,
+      "step": 8706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5497757196426392,
+      "learning_rate": 0.0005705439921760562,
+      "loss": 3.2458,
+      "step": 8707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5430121421813965,
+      "learning_rate": 0.0005705373387277566,
+      "loss": 3.4981,
+      "step": 8708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5202553868293762,
+      "learning_rate": 0.0005705306845669137,
+      "loss": 3.3059,
+      "step": 8709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53725266456604,
+      "learning_rate": 0.0005705240296935448,
+      "loss": 3.321,
+      "step": 8710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5087431073188782,
+      "learning_rate": 0.000570517374107668,
+      "loss": 3.255,
+      "step": 8711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5445797443389893,
+      "learning_rate": 0.0005705107178093004,
+      "loss": 3.2109,
+      "step": 8712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.563112199306488,
+      "learning_rate": 0.0005705040607984595,
+      "loss": 3.322,
+      "step": 8713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.544439435005188,
+      "learning_rate": 0.000570497403075163,
+      "loss": 3.0042,
+      "step": 8714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5439892411231995,
+      "learning_rate": 0.0005704907446394285,
+      "loss": 3.3405,
+      "step": 8715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.554537296295166,
+      "learning_rate": 0.0005704840854912732,
+      "loss": 3.3383,
+      "step": 8716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.506321907043457,
+      "learning_rate": 0.0005704774256307151,
+      "loss": 3.3718,
+      "step": 8717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5402343273162842,
+      "learning_rate": 0.0005704707650577716,
+      "loss": 3.1319,
+      "step": 8718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5468218326568604,
+      "learning_rate": 0.0005704641037724599,
+      "loss": 3.2299,
+      "step": 8719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5793018341064453,
+      "learning_rate": 0.000570457441774798,
+      "loss": 3.1532,
+      "step": 8720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5519994497299194,
+      "learning_rate": 0.0005704507790648031,
+      "loss": 3.2421,
+      "step": 8721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5348733067512512,
+      "learning_rate": 0.0005704441156424931,
+      "loss": 3.1543,
+      "step": 8722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5524995923042297,
+      "learning_rate": 0.0005704374515078853,
+      "loss": 3.2966,
+      "step": 8723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5177032351493835,
+      "learning_rate": 0.0005704307866609971,
+      "loss": 3.4132,
+      "step": 8724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5006693005561829,
+      "learning_rate": 0.0005704241211018464,
+      "loss": 3.2054,
+      "step": 8725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5278719663619995,
+      "learning_rate": 0.0005704174548304506,
+      "loss": 3.2603,
+      "step": 8726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4858294129371643,
+      "learning_rate": 0.0005704107878468272,
+      "loss": 3.3033,
+      "step": 8727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5394824147224426,
+      "learning_rate": 0.0005704041201509939,
+      "loss": 3.3758,
+      "step": 8728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5565510988235474,
+      "learning_rate": 0.0005703974517429681,
+      "loss": 3.0167,
+      "step": 8729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.533441960811615,
+      "learning_rate": 0.0005703907826227676,
+      "loss": 3.5257,
+      "step": 8730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5071123242378235,
+      "learning_rate": 0.0005703841127904097,
+      "loss": 3.347,
+      "step": 8731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5355621576309204,
+      "learning_rate": 0.000570377442245912,
+      "loss": 3.074,
+      "step": 8732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5217567086219788,
+      "learning_rate": 0.0005703707709892923,
+      "loss": 3.4603,
+      "step": 8733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5563116669654846,
+      "learning_rate": 0.0005703640990205681,
+      "loss": 3.2609,
+      "step": 8734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5034108757972717,
+      "learning_rate": 0.0005703574263397566,
+      "loss": 2.9131,
+      "step": 8735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5095783472061157,
+      "learning_rate": 0.000570350752946876,
+      "loss": 3.3354,
+      "step": 8736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.542126476764679,
+      "learning_rate": 0.0005703440788419435,
+      "loss": 3.272,
+      "step": 8737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5572932362556458,
+      "learning_rate": 0.0005703374040249765,
+      "loss": 3.3902,
+      "step": 8738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5626958012580872,
+      "learning_rate": 0.000570330728495993,
+      "loss": 3.2746,
+      "step": 8739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5667341947555542,
+      "learning_rate": 0.0005703240522550102,
+      "loss": 3.3586,
+      "step": 8740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.556536078453064,
+      "learning_rate": 0.0005703173753020461,
+      "loss": 3.227,
+      "step": 8741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.530508816242218,
+      "learning_rate": 0.0005703106976371179,
+      "loss": 3.5442,
+      "step": 8742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5346199870109558,
+      "learning_rate": 0.0005703040192602435,
+      "loss": 3.5053,
+      "step": 8743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5241973996162415,
+      "learning_rate": 0.0005702973401714402,
+      "loss": 3.4193,
+      "step": 8744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.590114414691925,
+      "learning_rate": 0.0005702906603707256,
+      "loss": 3.1247,
+      "step": 8745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5349878668785095,
+      "learning_rate": 0.0005702839798581176,
+      "loss": 3.3134,
+      "step": 8746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.598888099193573,
+      "learning_rate": 0.0005702772986336337,
+      "loss": 3.546,
+      "step": 8747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4993423521518707,
+      "learning_rate": 0.0005702706166972912,
+      "loss": 3.3099,
+      "step": 8748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5180084705352783,
+      "learning_rate": 0.000570263934049108,
+      "loss": 3.175,
+      "step": 8749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5331127643585205,
+      "learning_rate": 0.0005702572506891017,
+      "loss": 3.0252,
+      "step": 8750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5388214588165283,
+      "learning_rate": 0.0005702505666172897,
+      "loss": 3.2869,
+      "step": 8751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5273463129997253,
+      "learning_rate": 0.0005702438818336897,
+      "loss": 3.2632,
+      "step": 8752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5084626078605652,
+      "learning_rate": 0.0005702371963383194,
+      "loss": 3.3897,
+      "step": 8753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5808398127555847,
+      "learning_rate": 0.0005702305101311963,
+      "loss": 3.3616,
+      "step": 8754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5917160511016846,
+      "learning_rate": 0.000570223823212338,
+      "loss": 3.1535,
+      "step": 8755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5147940516471863,
+      "learning_rate": 0.0005702171355817623,
+      "loss": 3.1924,
+      "step": 8756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5295524597167969,
+      "learning_rate": 0.0005702104472394866,
+      "loss": 3.2706,
+      "step": 8757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5461099743843079,
+      "learning_rate": 0.0005702037581855285,
+      "loss": 3.4574,
+      "step": 8758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5407048463821411,
+      "learning_rate": 0.0005701970684199057,
+      "loss": 3.1396,
+      "step": 8759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5329182744026184,
+      "learning_rate": 0.000570190377942636,
+      "loss": 3.1328,
+      "step": 8760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5270434617996216,
+      "learning_rate": 0.0005701836867537367,
+      "loss": 3.4657,
+      "step": 8761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5661664009094238,
+      "learning_rate": 0.0005701769948532257,
+      "loss": 3.1477,
+      "step": 8762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.526901364326477,
+      "learning_rate": 0.0005701703022411203,
+      "loss": 3.3944,
+      "step": 8763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5473800897598267,
+      "learning_rate": 0.0005701636089174384,
+      "loss": 3.4588,
+      "step": 8764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5030317902565002,
+      "learning_rate": 0.0005701569148821976,
+      "loss": 3.2712,
+      "step": 8765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5436576008796692,
+      "learning_rate": 0.0005701502201354154,
+      "loss": 3.2773,
+      "step": 8766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5438995361328125,
+      "learning_rate": 0.0005701435246771095,
+      "loss": 3.1112,
+      "step": 8767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51812344789505,
+      "learning_rate": 0.0005701368285072977,
+      "loss": 3.4097,
+      "step": 8768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5479652285575867,
+      "learning_rate": 0.0005701301316259973,
+      "loss": 3.1205,
+      "step": 8769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5035589337348938,
+      "learning_rate": 0.0005701234340332262,
+      "loss": 3.3524,
+      "step": 8770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5446529388427734,
+      "learning_rate": 0.000570116735729002,
+      "loss": 3.2749,
+      "step": 8771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5573705434799194,
+      "learning_rate": 0.0005701100367133422,
+      "loss": 3.4812,
+      "step": 8772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.517190158367157,
+      "learning_rate": 0.0005701033369862647,
+      "loss": 3.1171,
+      "step": 8773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5003644824028015,
+      "learning_rate": 0.0005700966365477869,
+      "loss": 3.3361,
+      "step": 8774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6054016947746277,
+      "learning_rate": 0.0005700899353979265,
+      "loss": 3.0853,
+      "step": 8775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46385738253593445,
+      "learning_rate": 0.0005700832335367012,
+      "loss": 3.4381,
+      "step": 8776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.495850145816803,
+      "learning_rate": 0.0005700765309641287,
+      "loss": 3.2427,
+      "step": 8777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5276760458946228,
+      "learning_rate": 0.0005700698276802266,
+      "loss": 3.2379,
+      "step": 8778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5193692445755005,
+      "learning_rate": 0.0005700631236850124,
+      "loss": 3.1412,
+      "step": 8779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.537988543510437,
+      "learning_rate": 0.0005700564189785041,
+      "loss": 3.1968,
+      "step": 8780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5210827589035034,
+      "learning_rate": 0.0005700497135607191,
+      "loss": 3.3336,
+      "step": 8781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5374914407730103,
+      "learning_rate": 0.0005700430074316751,
+      "loss": 3.2019,
+      "step": 8782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5302662253379822,
+      "learning_rate": 0.0005700363005913898,
+      "loss": 3.2672,
+      "step": 8783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5213940143585205,
+      "learning_rate": 0.0005700295930398809,
+      "loss": 3.3535,
+      "step": 8784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.509855329990387,
+      "learning_rate": 0.000570022884777166,
+      "loss": 3.3465,
+      "step": 8785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5186008214950562,
+      "learning_rate": 0.0005700161758032628,
+      "loss": 3.3766,
+      "step": 8786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.558765709400177,
+      "learning_rate": 0.0005700094661181889,
+      "loss": 3.1944,
+      "step": 8787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5422574877738953,
+      "learning_rate": 0.000570002755721962,
+      "loss": 3.2641,
+      "step": 8788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5169627070426941,
+      "learning_rate": 0.0005699960446145999,
+      "loss": 3.2564,
+      "step": 8789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5492352843284607,
+      "learning_rate": 0.0005699893327961201,
+      "loss": 3.0885,
+      "step": 8790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.553865909576416,
+      "learning_rate": 0.0005699826202665405,
+      "loss": 3.4303,
+      "step": 8791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5586493611335754,
+      "learning_rate": 0.0005699759070258785,
+      "loss": 3.1598,
+      "step": 8792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5349708199501038,
+      "learning_rate": 0.000569969193074152,
+      "loss": 3.3447,
+      "step": 8793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5063890814781189,
+      "learning_rate": 0.0005699624784113785,
+      "loss": 3.3572,
+      "step": 8794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5007393956184387,
+      "learning_rate": 0.0005699557630375759,
+      "loss": 3.3862,
+      "step": 8795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4956451654434204,
+      "learning_rate": 0.0005699490469527617,
+      "loss": 3.3052,
+      "step": 8796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5122515559196472,
+      "learning_rate": 0.0005699423301569536,
+      "loss": 3.3706,
+      "step": 8797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5375229716300964,
+      "learning_rate": 0.0005699356126501695,
+      "loss": 3.3361,
+      "step": 8798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48476994037628174,
+      "learning_rate": 0.0005699288944324268,
+      "loss": 3.3599,
+      "step": 8799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5187641978263855,
+      "learning_rate": 0.0005699221755037435,
+      "loss": 3.2034,
+      "step": 8800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5315556526184082,
+      "learning_rate": 0.000569915455864137,
+      "loss": 3.5175,
+      "step": 8801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5142885446548462,
+      "learning_rate": 0.0005699087355136252,
+      "loss": 3.1606,
+      "step": 8802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5275764465332031,
+      "learning_rate": 0.0005699020144522257,
+      "loss": 3.4446,
+      "step": 8803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5340292453765869,
+      "learning_rate": 0.0005698952926799563,
+      "loss": 3.3616,
+      "step": 8804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50909823179245,
+      "learning_rate": 0.0005698885701968347,
+      "loss": 3.388,
+      "step": 8805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5391430258750916,
+      "learning_rate": 0.0005698818470028784,
+      "loss": 3.2484,
+      "step": 8806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5389807224273682,
+      "learning_rate": 0.0005698751230981053,
+      "loss": 3.1581,
+      "step": 8807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4988233745098114,
+      "learning_rate": 0.0005698683984825331,
+      "loss": 3.0371,
+      "step": 8808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5417897701263428,
+      "learning_rate": 0.0005698616731561794,
+      "loss": 3.3176,
+      "step": 8809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5258243680000305,
+      "learning_rate": 0.0005698549471190621,
+      "loss": 3.5196,
+      "step": 8810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48488008975982666,
+      "learning_rate": 0.0005698482203711988,
+      "loss": 3.2598,
+      "step": 8811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48625877499580383,
+      "learning_rate": 0.0005698414929126071,
+      "loss": 3.3822,
+      "step": 8812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5102300047874451,
+      "learning_rate": 0.0005698347647433049,
+      "loss": 3.3212,
+      "step": 8813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5392574667930603,
+      "learning_rate": 0.0005698280358633099,
+      "loss": 3.1794,
+      "step": 8814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5915513038635254,
+      "learning_rate": 0.0005698213062726397,
+      "loss": 3.3302,
+      "step": 8815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5354930758476257,
+      "learning_rate": 0.0005698145759713122,
+      "loss": 3.2355,
+      "step": 8816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5436181426048279,
+      "learning_rate": 0.000569807844959345,
+      "loss": 3.3668,
+      "step": 8817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5166468024253845,
+      "learning_rate": 0.0005698011132367558,
+      "loss": 3.1715,
+      "step": 8818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5697653889656067,
+      "learning_rate": 0.0005697943808035625,
+      "loss": 3.271,
+      "step": 8819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5263750553131104,
+      "learning_rate": 0.0005697876476597826,
+      "loss": 3.5375,
+      "step": 8820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5435929894447327,
+      "learning_rate": 0.0005697809138054341,
+      "loss": 3.3333,
+      "step": 8821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5712611079216003,
+      "learning_rate": 0.0005697741792405344,
+      "loss": 3.453,
+      "step": 8822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5347647666931152,
+      "learning_rate": 0.0005697674439651017,
+      "loss": 3.164,
+      "step": 8823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.549929141998291,
+      "learning_rate": 0.0005697607079791533,
+      "loss": 3.4472,
+      "step": 8824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5421603918075562,
+      "learning_rate": 0.0005697539712827071,
+      "loss": 3.0966,
+      "step": 8825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5252081751823425,
+      "learning_rate": 0.0005697472338757808,
+      "loss": 3.1905,
+      "step": 8826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5157675743103027,
+      "learning_rate": 0.0005697404957583923,
+      "loss": 3.2351,
+      "step": 8827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5952437520027161,
+      "learning_rate": 0.0005697337569305594,
+      "loss": 3.0903,
+      "step": 8828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5244645476341248,
+      "learning_rate": 0.0005697270173922994,
+      "loss": 3.2001,
+      "step": 8829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5517503619194031,
+      "learning_rate": 0.0005697202771436305,
+      "loss": 3.1902,
+      "step": 8830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5599244236946106,
+      "learning_rate": 0.0005697135361845703,
+      "loss": 3.2742,
+      "step": 8831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5345979332923889,
+      "learning_rate": 0.0005697067945151365,
+      "loss": 3.5683,
+      "step": 8832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5085006356239319,
+      "learning_rate": 0.000569700052135347,
+      "loss": 3.2911,
+      "step": 8833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5625810027122498,
+      "learning_rate": 0.0005696933090452193,
+      "loss": 3.1879,
+      "step": 8834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5771826505661011,
+      "learning_rate": 0.0005696865652447715,
+      "loss": 3.2465,
+      "step": 8835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5013545155525208,
+      "learning_rate": 0.0005696798207340211,
+      "loss": 3.1855,
+      "step": 8836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5615394115447998,
+      "learning_rate": 0.000569673075512986,
+      "loss": 3.4138,
+      "step": 8837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5521506071090698,
+      "learning_rate": 0.0005696663295816839,
+      "loss": 3.1211,
+      "step": 8838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.713085412979126,
+      "learning_rate": 0.0005696595829401325,
+      "loss": 3.1726,
+      "step": 8839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5880343914031982,
+      "learning_rate": 0.0005696528355883497,
+      "loss": 3.3201,
+      "step": 8840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6011033654212952,
+      "learning_rate": 0.0005696460875263534,
+      "loss": 3.243,
+      "step": 8841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5444574356079102,
+      "learning_rate": 0.000569639338754161,
+      "loss": 3.5191,
+      "step": 8842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5401180386543274,
+      "learning_rate": 0.0005696325892717906,
+      "loss": 3.2096,
+      "step": 8843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5681575536727905,
+      "learning_rate": 0.0005696258390792598,
+      "loss": 3.2179,
+      "step": 8844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.505577027797699,
+      "learning_rate": 0.0005696190881765864,
+      "loss": 3.5161,
+      "step": 8845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5092687606811523,
+      "learning_rate": 0.0005696123365637882,
+      "loss": 3.3775,
+      "step": 8846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4876072108745575,
+      "learning_rate": 0.000569605584240883,
+      "loss": 3.3778,
+      "step": 8847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5532077550888062,
+      "learning_rate": 0.0005695988312078886,
+      "loss": 3.2693,
+      "step": 8848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5383113622665405,
+      "learning_rate": 0.0005695920774648227,
+      "loss": 3.2532,
+      "step": 8849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5278736352920532,
+      "learning_rate": 0.0005695853230117033,
+      "loss": 3.286,
+      "step": 8850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5102056264877319,
+      "learning_rate": 0.000569578567848548,
+      "loss": 3.493,
+      "step": 8851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5252404808998108,
+      "learning_rate": 0.0005695718119753746,
+      "loss": 3.2592,
+      "step": 8852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5343819856643677,
+      "learning_rate": 0.0005695650553922009,
+      "loss": 3.2568,
+      "step": 8853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5407352447509766,
+      "learning_rate": 0.0005695582980990448,
+      "loss": 3.2887,
+      "step": 8854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5392612218856812,
+      "learning_rate": 0.0005695515400959239,
+      "loss": 3.4683,
+      "step": 8855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5008077025413513,
+      "learning_rate": 0.0005695447813828562,
+      "loss": 3.4033,
+      "step": 8856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5120012164115906,
+      "learning_rate": 0.0005695380219598594,
+      "loss": 3.3343,
+      "step": 8857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.545326292514801,
+      "learning_rate": 0.0005695312618269513,
+      "loss": 3.2657,
+      "step": 8858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.550786018371582,
+      "learning_rate": 0.0005695245009841497,
+      "loss": 3.4312,
+      "step": 8859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5332869291305542,
+      "learning_rate": 0.0005695177394314725,
+      "loss": 3.4009,
+      "step": 8860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5279282927513123,
+      "learning_rate": 0.0005695109771689375,
+      "loss": 3.1493,
+      "step": 8861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5188350081443787,
+      "learning_rate": 0.0005695042141965624,
+      "loss": 3.4996,
+      "step": 8862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5265781283378601,
+      "learning_rate": 0.000569497450514365,
+      "loss": 3.4578,
+      "step": 8863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5063434839248657,
+      "learning_rate": 0.0005694906861223632,
+      "loss": 3.4383,
+      "step": 8864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5559127926826477,
+      "learning_rate": 0.0005694839210205749,
+      "loss": 3.2548,
+      "step": 8865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5212278962135315,
+      "learning_rate": 0.0005694771552090177,
+      "loss": 3.1999,
+      "step": 8866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5368652939796448,
+      "learning_rate": 0.0005694703886877097,
+      "loss": 3.1888,
+      "step": 8867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5175959467887878,
+      "learning_rate": 0.0005694636214566684,
+      "loss": 3.3606,
+      "step": 8868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5428808927536011,
+      "learning_rate": 0.0005694568535159118,
+      "loss": 3.2056,
+      "step": 8869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5059577822685242,
+      "learning_rate": 0.0005694500848654578,
+      "loss": 3.1956,
+      "step": 8870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5455737113952637,
+      "learning_rate": 0.000569443315505324,
+      "loss": 3.5257,
+      "step": 8871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5467769503593445,
+      "learning_rate": 0.0005694365454355284,
+      "loss": 3.4259,
+      "step": 8872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5153681635856628,
+      "learning_rate": 0.0005694297746560888,
+      "loss": 3.5655,
+      "step": 8873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4906521141529083,
+      "learning_rate": 0.0005694230031670231,
+      "loss": 3.5068,
+      "step": 8874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5056790113449097,
+      "learning_rate": 0.0005694162309683489,
+      "loss": 3.2567,
+      "step": 8875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5721079111099243,
+      "learning_rate": 0.0005694094580600843,
+      "loss": 3.158,
+      "step": 8876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5457658767700195,
+      "learning_rate": 0.0005694026844422471,
+      "loss": 3.2658,
+      "step": 8877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5211271047592163,
+      "learning_rate": 0.000569395910114855,
+      "loss": 3.3445,
+      "step": 8878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5187593102455139,
+      "learning_rate": 0.000569389135077926,
+      "loss": 3.2324,
+      "step": 8879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49315252900123596,
+      "learning_rate": 0.0005693823593314778,
+      "loss": 3.2866,
+      "step": 8880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5174716711044312,
+      "learning_rate": 0.0005693755828755283,
+      "loss": 3.0987,
+      "step": 8881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5270132422447205,
+      "learning_rate": 0.0005693688057100953,
+      "loss": 3.3338,
+      "step": 8882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5709814429283142,
+      "learning_rate": 0.0005693620278351968,
+      "loss": 3.2527,
+      "step": 8883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.50661700963974,
+      "learning_rate": 0.0005693552492508505,
+      "loss": 3.1687,
+      "step": 8884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5201055407524109,
+      "learning_rate": 0.0005693484699570744,
+      "loss": 3.2431,
+      "step": 8885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5174275040626526,
+      "learning_rate": 0.0005693416899538861,
+      "loss": 3.1476,
+      "step": 8886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5571829676628113,
+      "learning_rate": 0.0005693349092413038,
+      "loss": 3.349,
+      "step": 8887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5371866226196289,
+      "learning_rate": 0.0005693281278193452,
+      "loss": 3.1509,
+      "step": 8888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5208328366279602,
+      "learning_rate": 0.000569321345688028,
+      "loss": 3.4085,
+      "step": 8889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5333773493766785,
+      "learning_rate": 0.0005693145628473703,
+      "loss": 3.3804,
+      "step": 8890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5705960392951965,
+      "learning_rate": 0.0005693077792973899,
+      "loss": 3.4182,
+      "step": 8891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5110260844230652,
+      "learning_rate": 0.0005693009950381046,
+      "loss": 3.2802,
+      "step": 8892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4911125600337982,
+      "learning_rate": 0.0005692942100695322,
+      "loss": 3.244,
+      "step": 8893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6147732138633728,
+      "learning_rate": 0.0005692874243916908,
+      "loss": 3.2172,
+      "step": 8894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5848454236984253,
+      "learning_rate": 0.0005692806380045981,
+      "loss": 3.4469,
+      "step": 8895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5394555330276489,
+      "learning_rate": 0.000569273850908272,
+      "loss": 3.4887,
+      "step": 8896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5056954622268677,
+      "learning_rate": 0.0005692670631027304,
+      "loss": 3.3544,
+      "step": 8897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4951151907444,
+      "learning_rate": 0.0005692602745879913,
+      "loss": 3.2587,
+      "step": 8898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5151358246803284,
+      "learning_rate": 0.0005692534853640723,
+      "loss": 3.3862,
+      "step": 8899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5135519504547119,
+      "learning_rate": 0.0005692466954309915,
+      "loss": 3.2097,
+      "step": 8900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5229550004005432,
+      "learning_rate": 0.0005692399047887667,
+      "loss": 3.291,
+      "step": 8901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5192024111747742,
+      "learning_rate": 0.000569233113437416,
+      "loss": 3.2481,
+      "step": 8902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5763697624206543,
+      "learning_rate": 0.0005692263213769569,
+      "loss": 3.1955,
+      "step": 8903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5331725478172302,
+      "learning_rate": 0.0005692195286074075,
+      "loss": 3.2931,
+      "step": 8904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5367273092269897,
+      "learning_rate": 0.0005692127351287857,
+      "loss": 3.2269,
+      "step": 8905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5095797777175903,
+      "learning_rate": 0.0005692059409411094,
+      "loss": 3.3072,
+      "step": 8906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5251290202140808,
+      "learning_rate": 0.0005691991460443964,
+      "loss": 3.3444,
+      "step": 8907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5581363439559937,
+      "learning_rate": 0.0005691923504386646,
+      "loss": 3.3821,
+      "step": 8908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.536482572555542,
+      "learning_rate": 0.000569185554123932,
+      "loss": 3.3329,
+      "step": 8909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5598181486129761,
+      "learning_rate": 0.0005691787571002165,
+      "loss": 3.3854,
+      "step": 8910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5411391854286194,
+      "learning_rate": 0.000569171959367536,
+      "loss": 3.2254,
+      "step": 8911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5239635109901428,
+      "learning_rate": 0.0005691651609259083,
+      "loss": 3.4488,
+      "step": 8912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5380442142486572,
+      "learning_rate": 0.0005691583617753514,
+      "loss": 3.601,
+      "step": 8913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5128017663955688,
+      "learning_rate": 0.0005691515619158831,
+      "loss": 3.39,
+      "step": 8914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5668172240257263,
+      "learning_rate": 0.0005691447613475214,
+      "loss": 3.4607,
+      "step": 8915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5155226588249207,
+      "learning_rate": 0.0005691379600702843,
+      "loss": 3.1785,
+      "step": 8916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53493332862854,
+      "learning_rate": 0.0005691311580841896,
+      "loss": 3.3619,
+      "step": 8917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5263345241546631,
+      "learning_rate": 0.0005691243553892551,
+      "loss": 3.3008,
+      "step": 8918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5928928852081299,
+      "learning_rate": 0.0005691175519854989,
+      "loss": 3.1524,
+      "step": 8919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5617950558662415,
+      "learning_rate": 0.0005691107478729389,
+      "loss": 3.0835,
+      "step": 8920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5107870697975159,
+      "learning_rate": 0.0005691039430515929,
+      "loss": 3.248,
+      "step": 8921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5091202259063721,
+      "learning_rate": 0.000569097137521479,
+      "loss": 3.3436,
+      "step": 8922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5600295066833496,
+      "learning_rate": 0.000569090331282615,
+      "loss": 3.1758,
+      "step": 8923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5212362408638,
+      "learning_rate": 0.0005690835243350188,
+      "loss": 3.4895,
+      "step": 8924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5386121273040771,
+      "learning_rate": 0.0005690767166787084,
+      "loss": 3.277,
+      "step": 8925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5397125482559204,
+      "learning_rate": 0.0005690699083137018,
+      "loss": 3.354,
+      "step": 8926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5409417152404785,
+      "learning_rate": 0.0005690630992400168,
+      "loss": 3.2076,
+      "step": 8927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5249380469322205,
+      "learning_rate": 0.0005690562894576713,
+      "loss": 3.3417,
+      "step": 8928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4952598810195923,
+      "learning_rate": 0.0005690494789666834,
+      "loss": 3.2278,
+      "step": 8929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5684417486190796,
+      "learning_rate": 0.0005690426677670709,
+      "loss": 3.0598,
+      "step": 8930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5358139872550964,
+      "learning_rate": 0.0005690358558588519,
+      "loss": 3.4284,
+      "step": 8931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5555673241615295,
+      "learning_rate": 0.0005690290432420441,
+      "loss": 3.2636,
+      "step": 8932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5253645777702332,
+      "learning_rate": 0.0005690222299166656,
+      "loss": 3.4941,
+      "step": 8933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5221039056777954,
+      "learning_rate": 0.0005690154158827344,
+      "loss": 3.3307,
+      "step": 8934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5121043920516968,
+      "learning_rate": 0.0005690086011402683,
+      "loss": 3.2696,
+      "step": 8935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5969629883766174,
+      "learning_rate": 0.0005690017856892853,
+      "loss": 3.1725,
+      "step": 8936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5254464745521545,
+      "learning_rate": 0.0005689949695298034,
+      "loss": 3.3222,
+      "step": 8937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5058313608169556,
+      "learning_rate": 0.0005689881526618405,
+      "loss": 3.3466,
+      "step": 8938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49000418186187744,
+      "learning_rate": 0.0005689813350854147,
+      "loss": 3.4226,
+      "step": 8939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5364489555358887,
+      "learning_rate": 0.0005689745168005437,
+      "loss": 3.3661,
+      "step": 8940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5190027952194214,
+      "learning_rate": 0.0005689676978072456,
+      "loss": 3.2002,
+      "step": 8941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5191941261291504,
+      "learning_rate": 0.0005689608781055383,
+      "loss": 3.0918,
+      "step": 8942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5378019213676453,
+      "learning_rate": 0.0005689540576954399,
+      "loss": 3.3341,
+      "step": 8943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6198469400405884,
+      "learning_rate": 0.0005689472365769683,
+      "loss": 3.3351,
+      "step": 8944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6274257302284241,
+      "learning_rate": 0.0005689404147501413,
+      "loss": 3.1836,
+      "step": 8945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5766898393630981,
+      "learning_rate": 0.000568933592214977,
+      "loss": 3.3648,
+      "step": 8946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5012130737304688,
+      "learning_rate": 0.0005689267689714934,
+      "loss": 3.3627,
+      "step": 8947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5235632061958313,
+      "learning_rate": 0.0005689199450197086,
+      "loss": 3.3629,
+      "step": 8948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4902224838733673,
+      "learning_rate": 0.0005689131203596404,
+      "loss": 3.4249,
+      "step": 8949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.555653989315033,
+      "learning_rate": 0.0005689062949913067,
+      "loss": 3.2133,
+      "step": 8950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49776995182037354,
+      "learning_rate": 0.0005688994689147256,
+      "loss": 3.3663,
+      "step": 8951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5030352473258972,
+      "learning_rate": 0.000568892642129915,
+      "loss": 3.1925,
+      "step": 8952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4929159879684448,
+      "learning_rate": 0.000568885814636893,
+      "loss": 3.2832,
+      "step": 8953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5239042639732361,
+      "learning_rate": 0.0005688789864356775,
+      "loss": 3.2955,
+      "step": 8954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5698967576026917,
+      "learning_rate": 0.0005688721575262865,
+      "loss": 3.205,
+      "step": 8955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5118188858032227,
+      "learning_rate": 0.0005688653279087379,
+      "loss": 3.1984,
+      "step": 8956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6158947944641113,
+      "learning_rate": 0.0005688584975830498,
+      "loss": 3.1382,
+      "step": 8957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.571174144744873,
+      "learning_rate": 0.0005688516665492403,
+      "loss": 3.1952,
+      "step": 8958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5016655921936035,
+      "learning_rate": 0.0005688448348073271,
+      "loss": 3.2333,
+      "step": 8959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5244669914245605,
+      "learning_rate": 0.0005688380023573284,
+      "loss": 3.3077,
+      "step": 8960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.542945146560669,
+      "learning_rate": 0.0005688311691992621,
+      "loss": 3.1278,
+      "step": 8961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5659850835800171,
+      "learning_rate": 0.0005688243353331462,
+      "loss": 3.2479,
+      "step": 8962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5096282362937927,
+      "learning_rate": 0.0005688175007589989,
+      "loss": 3.3169,
+      "step": 8963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5422225594520569,
+      "learning_rate": 0.0005688106654768379,
+      "loss": 3.412,
+      "step": 8964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5580344200134277,
+      "learning_rate": 0.0005688038294866814,
+      "loss": 3.2899,
+      "step": 8965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5104421973228455,
+      "learning_rate": 0.0005687969927885474,
+      "loss": 3.1282,
+      "step": 8966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5075674057006836,
+      "learning_rate": 0.0005687901553824537,
+      "loss": 3.4159,
+      "step": 8967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5359368324279785,
+      "learning_rate": 0.0005687833172684186,
+      "loss": 3.4478,
+      "step": 8968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5409534573554993,
+      "learning_rate": 0.0005687764784464599,
+      "loss": 3.3085,
+      "step": 8969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5197051763534546,
+      "learning_rate": 0.0005687696389165956,
+      "loss": 3.0945,
+      "step": 8970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5289903283119202,
+      "learning_rate": 0.000568762798678844,
+      "loss": 3.4001,
+      "step": 8971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5642632842063904,
+      "learning_rate": 0.0005687559577332228,
+      "loss": 3.5348,
+      "step": 8972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6010560989379883,
+      "learning_rate": 0.0005687491160797501,
+      "loss": 3.2148,
+      "step": 8973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5142990350723267,
+      "learning_rate": 0.000568742273718444,
+      "loss": 3.2754,
+      "step": 8974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5444654822349548,
+      "learning_rate": 0.0005687354306493225,
+      "loss": 3.4776,
+      "step": 8975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5075336694717407,
+      "learning_rate": 0.0005687285868724035,
+      "loss": 3.1729,
+      "step": 8976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5050414204597473,
+      "learning_rate": 0.0005687217423877051,
+      "loss": 3.244,
+      "step": 8977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49637553095817566,
+      "learning_rate": 0.0005687148971952455,
+      "loss": 3.3921,
+      "step": 8978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5068024396896362,
+      "learning_rate": 0.0005687080512950426,
+      "loss": 3.228,
+      "step": 8979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.512059211730957,
+      "learning_rate": 0.0005687012046871143,
+      "loss": 3.2777,
+      "step": 8980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5337809920310974,
+      "learning_rate": 0.0005686943573714787,
+      "loss": 3.4469,
+      "step": 8981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5400496125221252,
+      "learning_rate": 0.0005686875093481539,
+      "loss": 3.2248,
+      "step": 8982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5225552916526794,
+      "learning_rate": 0.0005686806606171579,
+      "loss": 3.4751,
+      "step": 8983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5124357342720032,
+      "learning_rate": 0.0005686738111785088,
+      "loss": 3.1647,
+      "step": 8984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5976647734642029,
+      "learning_rate": 0.0005686669610322246,
+      "loss": 3.3468,
+      "step": 8985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5431848764419556,
+      "learning_rate": 0.0005686601101783232,
+      "loss": 3.3075,
+      "step": 8986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.522193193435669,
+      "learning_rate": 0.000568653258616823,
+      "loss": 3.2333,
+      "step": 8987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5003888607025146,
+      "learning_rate": 0.0005686464063477416,
+      "loss": 3.4794,
+      "step": 8988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5476301312446594,
+      "learning_rate": 0.0005686395533710973,
+      "loss": 3.1188,
+      "step": 8989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5447575449943542,
+      "learning_rate": 0.0005686326996869083,
+      "loss": 3.3467,
+      "step": 8990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5251429677009583,
+      "learning_rate": 0.0005686258452951923,
+      "loss": 3.2945,
+      "step": 8991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5147581100463867,
+      "learning_rate": 0.0005686189901959675,
+      "loss": 3.4831,
+      "step": 8992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5492880344390869,
+      "learning_rate": 0.000568612134389252,
+      "loss": 3.4656,
+      "step": 8993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.511838436126709,
+      "learning_rate": 0.000568605277875064,
+      "loss": 3.2636,
+      "step": 8994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5325186848640442,
+      "learning_rate": 0.0005685984206534212,
+      "loss": 3.3853,
+      "step": 8995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.528266966342926,
+      "learning_rate": 0.000568591562724342,
+      "loss": 3.3297,
+      "step": 8996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5761194229125977,
+      "learning_rate": 0.0005685847040878441,
+      "loss": 3.1765,
+      "step": 8997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4952467978000641,
+      "learning_rate": 0.000568577844743946,
+      "loss": 3.4474,
+      "step": 8998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5307065844535828,
+      "learning_rate": 0.0005685709846926654,
+      "loss": 3.3741,
+      "step": 8999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5305284857749939,
+      "learning_rate": 0.0005685641239340206,
+      "loss": 3.2478,
+      "step": 9000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5165554881095886,
+      "learning_rate": 0.0005685572624680295,
+      "loss": 3.1295,
+      "step": 9001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5124072432518005,
+      "learning_rate": 0.0005685504002947104,
+      "loss": 3.3352,
+      "step": 9002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5218518376350403,
+      "learning_rate": 0.0005685435374140811,
+      "loss": 3.2078,
+      "step": 9003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.574737548828125,
+      "learning_rate": 0.0005685366738261598,
+      "loss": 3.2878,
+      "step": 9004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5492660403251648,
+      "learning_rate": 0.0005685298095309646,
+      "loss": 3.2467,
+      "step": 9005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5594704151153564,
+      "learning_rate": 0.0005685229445285137,
+      "loss": 3.4172,
+      "step": 9006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5028058290481567,
+      "learning_rate": 0.000568516078818825,
+      "loss": 3.3881,
+      "step": 9007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5461245179176331,
+      "learning_rate": 0.0005685092124019165,
+      "loss": 3.5241,
+      "step": 9008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5365086793899536,
+      "learning_rate": 0.0005685023452778065,
+      "loss": 3.3182,
+      "step": 9009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.530041515827179,
+      "learning_rate": 0.0005684954774465129,
+      "loss": 3.4351,
+      "step": 9010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5450629591941833,
+      "learning_rate": 0.000568488608908054,
+      "loss": 3.386,
+      "step": 9011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5048748254776001,
+      "learning_rate": 0.0005684817396624476,
+      "loss": 3.6001,
+      "step": 9012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5259529948234558,
+      "learning_rate": 0.0005684748697097121,
+      "loss": 3.0533,
+      "step": 9013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5435417294502258,
+      "learning_rate": 0.0005684679990498655,
+      "loss": 3.4664,
+      "step": 9014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5544543266296387,
+      "learning_rate": 0.0005684611276829259,
+      "loss": 3.151,
+      "step": 9015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.536688506603241,
+      "learning_rate": 0.0005684542556089112,
+      "loss": 3.3264,
+      "step": 9016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5237370133399963,
+      "learning_rate": 0.0005684473828278398,
+      "loss": 3.3703,
+      "step": 9017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5127825140953064,
+      "learning_rate": 0.0005684405093397296,
+      "loss": 3.2351,
+      "step": 9018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5225836634635925,
+      "learning_rate": 0.0005684336351445987,
+      "loss": 3.2658,
+      "step": 9019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5963146686553955,
+      "learning_rate": 0.0005684267602424652,
+      "loss": 3.2942,
+      "step": 9020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5185006260871887,
+      "learning_rate": 0.0005684198846333475,
+      "loss": 3.1149,
+      "step": 9021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5315390229225159,
+      "learning_rate": 0.0005684130083172634,
+      "loss": 3.1857,
+      "step": 9022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5717655420303345,
+      "learning_rate": 0.0005684061312942309,
+      "loss": 3.3572,
+      "step": 9023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.559047520160675,
+      "learning_rate": 0.0005683992535642686,
+      "loss": 3.3638,
+      "step": 9024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5396157503128052,
+      "learning_rate": 0.0005683923751273941,
+      "loss": 3.2502,
+      "step": 9025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5662112832069397,
+      "learning_rate": 0.0005683854959836259,
+      "loss": 3.244,
+      "step": 9026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49871206283569336,
+      "learning_rate": 0.0005683786161329819,
+      "loss": 3.1904,
+      "step": 9027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5973621606826782,
+      "learning_rate": 0.0005683717355754802,
+      "loss": 3.2459,
+      "step": 9028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5342198014259338,
+      "learning_rate": 0.000568364854311139,
+      "loss": 3.1354,
+      "step": 9029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.54073566198349,
+      "learning_rate": 0.0005683579723399765,
+      "loss": 3.2937,
+      "step": 9030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5744414329528809,
+      "learning_rate": 0.0005683510896620108,
+      "loss": 3.2309,
+      "step": 9031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5294274687767029,
+      "learning_rate": 0.0005683442062772598,
+      "loss": 3.1342,
+      "step": 9032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5062122941017151,
+      "learning_rate": 0.0005683373221857419,
+      "loss": 3.3708,
+      "step": 9033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5301145315170288,
+      "learning_rate": 0.0005683304373874752,
+      "loss": 3.3389,
+      "step": 9034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5067840814590454,
+      "learning_rate": 0.0005683235518824777,
+      "loss": 3.3437,
+      "step": 9035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5595483183860779,
+      "learning_rate": 0.0005683166656707676,
+      "loss": 3.315,
+      "step": 9036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5125157833099365,
+      "learning_rate": 0.0005683097787523631,
+      "loss": 3.3297,
+      "step": 9037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5199974179267883,
+      "learning_rate": 0.0005683028911272822,
+      "loss": 3.23,
+      "step": 9038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390259027481079,
+      "learning_rate": 0.0005682960027955431,
+      "loss": 3.2136,
+      "step": 9039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.500478982925415,
+      "learning_rate": 0.0005682891137571641,
+      "loss": 3.3919,
+      "step": 9040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5096060037612915,
+      "learning_rate": 0.0005682822240121631,
+      "loss": 3.5207,
+      "step": 9041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4984170198440552,
+      "learning_rate": 0.0005682753335605583,
+      "loss": 3.3042,
+      "step": 9042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5067201852798462,
+      "learning_rate": 0.000568268442402368,
+      "loss": 3.231,
+      "step": 9043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5161800384521484,
+      "learning_rate": 0.0005682615505376103,
+      "loss": 3.296,
+      "step": 9044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5189632773399353,
+      "learning_rate": 0.0005682546579663032,
+      "loss": 3.2647,
+      "step": 9045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5524479150772095,
+      "learning_rate": 0.000568247764688465,
+      "loss": 3.1366,
+      "step": 9046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5149971842765808,
+      "learning_rate": 0.0005682408707041138,
+      "loss": 3.3692,
+      "step": 9047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5103289484977722,
+      "learning_rate": 0.0005682339760132677,
+      "loss": 3.6341,
+      "step": 9048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5169664621353149,
+      "learning_rate": 0.000568227080615945,
+      "loss": 3.3884,
+      "step": 9049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5336402058601379,
+      "learning_rate": 0.0005682201845121638,
+      "loss": 3.3195,
+      "step": 9050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4936128556728363,
+      "learning_rate": 0.0005682132877019424,
+      "loss": 3.2705,
+      "step": 9051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49011480808258057,
+      "learning_rate": 0.0005682063901852986,
+      "loss": 3.3899,
+      "step": 9052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5118272304534912,
+      "learning_rate": 0.0005681994919622508,
+      "loss": 3.3134,
+      "step": 9053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5226544737815857,
+      "learning_rate": 0.0005681925930328172,
+      "loss": 3.2044,
+      "step": 9054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5248229503631592,
+      "learning_rate": 0.0005681856933970159,
+      "loss": 3.0951,
+      "step": 9055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.533872127532959,
+      "learning_rate": 0.0005681787930548652,
+      "loss": 3.3359,
+      "step": 9056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5588023066520691,
+      "learning_rate": 0.0005681718920063831,
+      "loss": 3.3904,
+      "step": 9057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5113193988800049,
+      "learning_rate": 0.0005681649902515878,
+      "loss": 3.3799,
+      "step": 9058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5519219040870667,
+      "learning_rate": 0.0005681580877904977,
+      "loss": 3.2737,
+      "step": 9059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5289116501808167,
+      "learning_rate": 0.0005681511846231306,
+      "loss": 3.4338,
+      "step": 9060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6075866222381592,
+      "learning_rate": 0.000568144280749505,
+      "loss": 3.1228,
+      "step": 9061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5013885498046875,
+      "learning_rate": 0.0005681373761696389,
+      "loss": 3.1611,
+      "step": 9062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5552061200141907,
+      "learning_rate": 0.0005681304708835506,
+      "loss": 3.2345,
+      "step": 9063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5864258408546448,
+      "learning_rate": 0.0005681235648912581,
+      "loss": 3.2384,
+      "step": 9064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5120783448219299,
+      "learning_rate": 0.00056811665819278,
+      "loss": 3.3637,
+      "step": 9065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49182936549186707,
+      "learning_rate": 0.0005681097507881342,
+      "loss": 3.3427,
+      "step": 9066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49345386028289795,
+      "learning_rate": 0.0005681028426773387,
+      "loss": 3.2404,
+      "step": 9067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5188148617744446,
+      "learning_rate": 0.0005680959338604121,
+      "loss": 3.4539,
+      "step": 9068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5250046849250793,
+      "learning_rate": 0.0005680890243373722,
+      "loss": 3.4335,
+      "step": 9069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5325458645820618,
+      "learning_rate": 0.0005680821141082375,
+      "loss": 3.18,
+      "step": 9070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5744656324386597,
+      "learning_rate": 0.0005680752031730262,
+      "loss": 3.3223,
+      "step": 9071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5349627733230591,
+      "learning_rate": 0.0005680682915317562,
+      "loss": 3.1908,
+      "step": 9072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5127280354499817,
+      "learning_rate": 0.0005680613791844461,
+      "loss": 3.29,
+      "step": 9073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5042176246643066,
+      "learning_rate": 0.000568054466131114,
+      "loss": 3.3658,
+      "step": 9074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5064307451248169,
+      "learning_rate": 0.0005680475523717778,
+      "loss": 3.3205,
+      "step": 9075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48436474800109863,
+      "learning_rate": 0.0005680406379064561,
+      "loss": 3.2207,
+      "step": 9076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5284563302993774,
+      "learning_rate": 0.0005680337227351668,
+      "loss": 3.3994,
+      "step": 9077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5303294658660889,
+      "learning_rate": 0.0005680268068579282,
+      "loss": 3.4003,
+      "step": 9078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5509895086288452,
+      "learning_rate": 0.0005680198902747589,
+      "loss": 3.2329,
+      "step": 9079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5539824366569519,
+      "learning_rate": 0.0005680129729856765,
+      "loss": 3.1104,
+      "step": 9080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6219602823257446,
+      "learning_rate": 0.0005680060549906996,
+      "loss": 3.4123,
+      "step": 9081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5088189244270325,
+      "learning_rate": 0.0005679991362898463,
+      "loss": 3.2834,
+      "step": 9082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.557954728603363,
+      "learning_rate": 0.0005679922168831348,
+      "loss": 3.2455,
+      "step": 9083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6321285367012024,
+      "learning_rate": 0.0005679852967705836,
+      "loss": 3.1699,
+      "step": 9084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.585474967956543,
+      "learning_rate": 0.0005679783759522105,
+      "loss": 3.1827,
+      "step": 9085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5444757342338562,
+      "learning_rate": 0.000567971454428034,
+      "loss": 3.3533,
+      "step": 9086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4752892553806305,
+      "learning_rate": 0.0005679645321980723,
+      "loss": 3.256,
+      "step": 9087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5205745100975037,
+      "learning_rate": 0.0005679576092623435,
+      "loss": 3.4033,
+      "step": 9088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5516600608825684,
+      "learning_rate": 0.000567950685620866,
+      "loss": 3.3387,
+      "step": 9089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5286070704460144,
+      "learning_rate": 0.0005679437612736578,
+      "loss": 3.2433,
+      "step": 9090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5709437131881714,
+      "learning_rate": 0.0005679368362207375,
+      "loss": 3.2315,
+      "step": 9091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5004037022590637,
+      "learning_rate": 0.0005679299104621231,
+      "loss": 3.3357,
+      "step": 9092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5113794207572937,
+      "learning_rate": 0.0005679229839978329,
+      "loss": 3.2806,
+      "step": 9093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5075574517250061,
+      "learning_rate": 0.0005679160568278851,
+      "loss": 3.6071,
+      "step": 9094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5023056268692017,
+      "learning_rate": 0.0005679091289522979,
+      "loss": 3.3308,
+      "step": 9095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5051378607749939,
+      "learning_rate": 0.0005679022003710896,
+      "loss": 3.4196,
+      "step": 9096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5265752077102661,
+      "learning_rate": 0.0005678952710842785,
+      "loss": 3.5432,
+      "step": 9097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49984413385391235,
+      "learning_rate": 0.0005678883410918828,
+      "loss": 3.4198,
+      "step": 9098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5776906609535217,
+      "learning_rate": 0.0005678814103939209,
+      "loss": 3.4116,
+      "step": 9099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5122084617614746,
+      "learning_rate": 0.0005678744789904108,
+      "loss": 3.3746,
+      "step": 9100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5504657626152039,
+      "learning_rate": 0.0005678675468813709,
+      "loss": 3.4364,
+      "step": 9101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5166839361190796,
+      "learning_rate": 0.0005678606140668194,
+      "loss": 3.4628,
+      "step": 9102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4547307789325714,
+      "learning_rate": 0.0005678536805467746,
+      "loss": 3.5267,
+      "step": 9103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5224431157112122,
+      "learning_rate": 0.0005678467463212549,
+      "loss": 3.2876,
+      "step": 9104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.501089334487915,
+      "learning_rate": 0.0005678398113902782,
+      "loss": 3.4004,
+      "step": 9105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5165853500366211,
+      "learning_rate": 0.0005678328757538631,
+      "loss": 3.3895,
+      "step": 9106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5630010366439819,
+      "learning_rate": 0.0005678259394120277,
+      "loss": 3.4256,
+      "step": 9107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5082736015319824,
+      "learning_rate": 0.0005678190023647903,
+      "loss": 3.3324,
+      "step": 9108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5279152989387512,
+      "learning_rate": 0.0005678120646121694,
+      "loss": 3.4226,
+      "step": 9109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5158643126487732,
+      "learning_rate": 0.0005678051261541828,
+      "loss": 3.1352,
+      "step": 9110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5294057130813599,
+      "learning_rate": 0.0005677981869908491,
+      "loss": 3.0135,
+      "step": 9111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5090487599372864,
+      "learning_rate": 0.0005677912471221866,
+      "loss": 3.396,
+      "step": 9112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5615133047103882,
+      "learning_rate": 0.0005677843065482135,
+      "loss": 3.1692,
+      "step": 9113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5106341242790222,
+      "learning_rate": 0.000567777365268948,
+      "loss": 3.3025,
+      "step": 9114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5407706499099731,
+      "learning_rate": 0.0005677704232844084,
+      "loss": 3.2903,
+      "step": 9115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5446071028709412,
+      "learning_rate": 0.0005677634805946131,
+      "loss": 3.311,
+      "step": 9116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5780032277107239,
+      "learning_rate": 0.0005677565371995804,
+      "loss": 3.4569,
+      "step": 9117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49284398555755615,
+      "learning_rate": 0.0005677495930993284,
+      "loss": 3.1733,
+      "step": 9118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48607781529426575,
+      "learning_rate": 0.0005677426482938756,
+      "loss": 3.3819,
+      "step": 9119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.520286500453949,
+      "learning_rate": 0.0005677357027832401,
+      "loss": 3.0716,
+      "step": 9120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5000066757202148,
+      "learning_rate": 0.0005677287565674404,
+      "loss": 3.2683,
+      "step": 9121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5120892524719238,
+      "learning_rate": 0.0005677218096464946,
+      "loss": 3.349,
+      "step": 9122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5167591571807861,
+      "learning_rate": 0.0005677148620204211,
+      "loss": 3.4684,
+      "step": 9123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50194251537323,
+      "learning_rate": 0.0005677079136892381,
+      "loss": 3.2571,
+      "step": 9124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5270057320594788,
+      "learning_rate": 0.000567700964652964,
+      "loss": 3.3654,
+      "step": 9125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5042098760604858,
+      "learning_rate": 0.0005676940149116172,
+      "loss": 3.3709,
+      "step": 9126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4862194061279297,
+      "learning_rate": 0.0005676870644652158,
+      "loss": 3.3156,
+      "step": 9127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5414255261421204,
+      "learning_rate": 0.0005676801133137782,
+      "loss": 3.191,
+      "step": 9128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5115101933479309,
+      "learning_rate": 0.0005676731614573228,
+      "loss": 3.276,
+      "step": 9129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4912344515323639,
+      "learning_rate": 0.0005676662088958676,
+      "loss": 3.393,
+      "step": 9130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5179233551025391,
+      "learning_rate": 0.0005676592556294312,
+      "loss": 3.2392,
+      "step": 9131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5112375617027283,
+      "learning_rate": 0.0005676523016580319,
+      "loss": 3.3681,
+      "step": 9132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5025057792663574,
+      "learning_rate": 0.0005676453469816878,
+      "loss": 3.3225,
+      "step": 9133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4907090961933136,
+      "learning_rate": 0.0005676383916004174,
+      "loss": 3.3381,
+      "step": 9134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5072884559631348,
+      "learning_rate": 0.0005676314355142392,
+      "loss": 3.2997,
+      "step": 9135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5344531536102295,
+      "learning_rate": 0.0005676244787231711,
+      "loss": 3.1497,
+      "step": 9136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5543556809425354,
+      "learning_rate": 0.0005676175212272317,
+      "loss": 3.5104,
+      "step": 9137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5010618567466736,
+      "learning_rate": 0.0005676105630264392,
+      "loss": 3.6381,
+      "step": 9138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5100370645523071,
+      "learning_rate": 0.000567603604120812,
+      "loss": 3.1735,
+      "step": 9139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5232434868812561,
+      "learning_rate": 0.0005675966445103684,
+      "loss": 3.5975,
+      "step": 9140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5133650898933411,
+      "learning_rate": 0.0005675896841951268,
+      "loss": 3.4042,
+      "step": 9141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.526075541973114,
+      "learning_rate": 0.0005675827231751055,
+      "loss": 3.4472,
+      "step": 9142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5056864619255066,
+      "learning_rate": 0.0005675757614503227,
+      "loss": 3.1239,
+      "step": 9143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5270462036132812,
+      "learning_rate": 0.0005675687990207969,
+      "loss": 3.4077,
+      "step": 9144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5691565275192261,
+      "learning_rate": 0.0005675618358865463,
+      "loss": 3.3887,
+      "step": 9145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47389543056488037,
+      "learning_rate": 0.0005675548720475894,
+      "loss": 3.3965,
+      "step": 9146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5528371930122375,
+      "learning_rate": 0.0005675479075039444,
+      "loss": 3.1438,
+      "step": 9147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48816508054733276,
+      "learning_rate": 0.0005675409422556297,
+      "loss": 3.2555,
+      "step": 9148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5429815649986267,
+      "learning_rate": 0.0005675339763026638,
+      "loss": 3.1632,
+      "step": 9149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5055248141288757,
+      "learning_rate": 0.0005675270096450648,
+      "loss": 3.3858,
+      "step": 9150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5444815158843994,
+      "learning_rate": 0.000567520042282851,
+      "loss": 3.1408,
+      "step": 9151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5034661889076233,
+      "learning_rate": 0.000567513074216041,
+      "loss": 3.4106,
+      "step": 9152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5205017924308777,
+      "learning_rate": 0.0005675061054446531,
+      "loss": 3.3771,
+      "step": 9153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5044941306114197,
+      "learning_rate": 0.0005674991359687055,
+      "loss": 3.1945,
+      "step": 9154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5016582608222961,
+      "learning_rate": 0.0005674921657882168,
+      "loss": 3.1752,
+      "step": 9155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.524653434753418,
+      "learning_rate": 0.0005674851949032052,
+      "loss": 3.1112,
+      "step": 9156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.551464319229126,
+      "learning_rate": 0.0005674782233136889,
+      "loss": 3.4564,
+      "step": 9157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6055057644844055,
+      "learning_rate": 0.0005674712510196865,
+      "loss": 3.235,
+      "step": 9158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390385389328003,
+      "learning_rate": 0.0005674642780212165,
+      "loss": 3.2609,
+      "step": 9159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5527321100234985,
+      "learning_rate": 0.0005674573043182969,
+      "loss": 3.1562,
+      "step": 9160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5559296607971191,
+      "learning_rate": 0.0005674503299109462,
+      "loss": 3.1802,
+      "step": 9161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5836023092269897,
+      "learning_rate": 0.0005674433547991828,
+      "loss": 3.2304,
+      "step": 9162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5389158725738525,
+      "learning_rate": 0.0005674363789830251,
+      "loss": 3.1442,
+      "step": 9163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48464417457580566,
+      "learning_rate": 0.0005674294024624914,
+      "loss": 3.4488,
+      "step": 9164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5225753784179688,
+      "learning_rate": 0.0005674224252376003,
+      "loss": 3.2688,
+      "step": 9165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.516726553440094,
+      "learning_rate": 0.0005674154473083699,
+      "loss": 3.3803,
+      "step": 9166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5179518461227417,
+      "learning_rate": 0.0005674084686748186,
+      "loss": 3.438,
+      "step": 9167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5502138733863831,
+      "learning_rate": 0.000567401489336965,
+      "loss": 3.1833,
+      "step": 9168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.523209273815155,
+      "learning_rate": 0.0005673945092948273,
+      "loss": 3.2305,
+      "step": 9169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48779064416885376,
+      "learning_rate": 0.0005673875285484238,
+      "loss": 3.3943,
+      "step": 9170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5196689963340759,
+      "learning_rate": 0.0005673805470977732,
+      "loss": 3.2378,
+      "step": 9171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5514273643493652,
+      "learning_rate": 0.0005673735649428934,
+      "loss": 3.2909,
+      "step": 9172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5025139451026917,
+      "learning_rate": 0.0005673665820838034,
+      "loss": 3.3022,
+      "step": 9173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5027960538864136,
+      "learning_rate": 0.000567359598520521,
+      "loss": 3.2242,
+      "step": 9174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5085102319717407,
+      "learning_rate": 0.0005673526142530651,
+      "loss": 3.2399,
+      "step": 9175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5565027594566345,
+      "learning_rate": 0.0005673456292814539,
+      "loss": 3.2157,
+      "step": 9176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5032663941383362,
+      "learning_rate": 0.0005673386436057056,
+      "loss": 3.2285,
+      "step": 9177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5043344497680664,
+      "learning_rate": 0.0005673316572258389,
+      "loss": 3.3707,
+      "step": 9178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5071136951446533,
+      "learning_rate": 0.000567324670141872,
+      "loss": 3.1895,
+      "step": 9179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4878983795642853,
+      "learning_rate": 0.0005673176823538232,
+      "loss": 3.1579,
+      "step": 9180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5063034892082214,
+      "learning_rate": 0.0005673106938617113,
+      "loss": 3.3481,
+      "step": 9181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5148180723190308,
+      "learning_rate": 0.0005673037046655544,
+      "loss": 3.4275,
+      "step": 9182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.529994547367096,
+      "learning_rate": 0.0005672967147653709,
+      "loss": 3.321,
+      "step": 9183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5083233118057251,
+      "learning_rate": 0.0005672897241611795,
+      "loss": 3.2562,
+      "step": 9184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5259013772010803,
+      "learning_rate": 0.0005672827328529982,
+      "loss": 3.4605,
+      "step": 9185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5973625779151917,
+      "learning_rate": 0.0005672757408408458,
+      "loss": 3.2049,
+      "step": 9186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5016379952430725,
+      "learning_rate": 0.0005672687481247404,
+      "loss": 2.9897,
+      "step": 9187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4864937961101532,
+      "learning_rate": 0.0005672617547047006,
+      "loss": 3.4482,
+      "step": 9188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5325981974601746,
+      "learning_rate": 0.0005672547605807449,
+      "loss": 3.1332,
+      "step": 9189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5618669390678406,
+      "learning_rate": 0.0005672477657528913,
+      "loss": 3.6335,
+      "step": 9190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5770689249038696,
+      "learning_rate": 0.0005672407702211588,
+      "loss": 3.4217,
+      "step": 9191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5329024791717529,
+      "learning_rate": 0.0005672337739855654,
+      "loss": 3.2829,
+      "step": 9192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5350424647331238,
+      "learning_rate": 0.0005672267770461297,
+      "loss": 3.2882,
+      "step": 9193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5432996153831482,
+      "learning_rate": 0.00056721977940287,
+      "loss": 3.2945,
+      "step": 9194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5327097773551941,
+      "learning_rate": 0.000567212781055805,
+      "loss": 3.4755,
+      "step": 9195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5117045044898987,
+      "learning_rate": 0.000567205782004953,
+      "loss": 3.1168,
+      "step": 9196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5237148404121399,
+      "learning_rate": 0.0005671987822503322,
+      "loss": 3.2491,
+      "step": 9197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5158869624137878,
+      "learning_rate": 0.0005671917817919613,
+      "loss": 3.2337,
+      "step": 9198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5397765040397644,
+      "learning_rate": 0.0005671847806298587,
+      "loss": 3.3685,
+      "step": 9199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5293693542480469,
+      "learning_rate": 0.0005671777787640428,
+      "loss": 3.0961,
+      "step": 9200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.522372305393219,
+      "learning_rate": 0.000567170776194532,
+      "loss": 3.3425,
+      "step": 9201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5064888596534729,
+      "learning_rate": 0.000567163772921345,
+      "loss": 3.3638,
+      "step": 9202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5614068508148193,
+      "learning_rate": 0.0005671567689444997,
+      "loss": 3.305,
+      "step": 9203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5083727240562439,
+      "learning_rate": 0.0005671497642640151,
+      "loss": 3.1742,
+      "step": 9204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5845643281936646,
+      "learning_rate": 0.0005671427588799094,
+      "loss": 3.2797,
+      "step": 9205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5405354499816895,
+      "learning_rate": 0.000567135752792201,
+      "loss": 3.4243,
+      "step": 9206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5075390934944153,
+      "learning_rate": 0.0005671287460009086,
+      "loss": 3.3293,
+      "step": 9207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5443362593650818,
+      "learning_rate": 0.0005671217385060504,
+      "loss": 3.3097,
+      "step": 9208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5043942332267761,
+      "learning_rate": 0.0005671147303076449,
+      "loss": 3.2578,
+      "step": 9209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5327485799789429,
+      "learning_rate": 0.0005671077214057107,
+      "loss": 3.3638,
+      "step": 9210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5666617751121521,
+      "learning_rate": 0.0005671007118002662,
+      "loss": 3.3362,
+      "step": 9211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5105660557746887,
+      "learning_rate": 0.0005670937014913297,
+      "loss": 3.3562,
+      "step": 9212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5232953429222107,
+      "learning_rate": 0.0005670866904789199,
+      "loss": 3.3715,
+      "step": 9213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5395106077194214,
+      "learning_rate": 0.0005670796787630552,
+      "loss": 3.213,
+      "step": 9214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48984065651893616,
+      "learning_rate": 0.0005670726663437539,
+      "loss": 3.3097,
+      "step": 9215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5296399593353271,
+      "learning_rate": 0.0005670656532210346,
+      "loss": 3.3367,
+      "step": 9216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4997063875198364,
+      "learning_rate": 0.0005670586393949159,
+      "loss": 3.4825,
+      "step": 9217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5573795437812805,
+      "learning_rate": 0.000567051624865416,
+      "loss": 3.2277,
+      "step": 9218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5867475271224976,
+      "learning_rate": 0.0005670446096325536,
+      "loss": 3.4567,
+      "step": 9219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5648623704910278,
+      "learning_rate": 0.000567037593696347,
+      "loss": 3.3481,
+      "step": 9220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5602989792823792,
+      "learning_rate": 0.0005670305770568148,
+      "loss": 3.4067,
+      "step": 9221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5342902541160583,
+      "learning_rate": 0.0005670235597139755,
+      "loss": 3.1211,
+      "step": 9222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5073335766792297,
+      "learning_rate": 0.0005670165416678476,
+      "loss": 3.3586,
+      "step": 9223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5396925806999207,
+      "learning_rate": 0.0005670095229184494,
+      "loss": 3.4748,
+      "step": 9224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5194887518882751,
+      "learning_rate": 0.0005670025034657995,
+      "loss": 3.4222,
+      "step": 9225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.56588214635849,
+      "learning_rate": 0.0005669954833099164,
+      "loss": 3.3498,
+      "step": 9226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5211643576622009,
+      "learning_rate": 0.0005669884624508186,
+      "loss": 3.431,
+      "step": 9227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5741928219795227,
+      "learning_rate": 0.0005669814408885246,
+      "loss": 3.4504,
+      "step": 9228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5376482009887695,
+      "learning_rate": 0.0005669744186230528,
+      "loss": 3.3266,
+      "step": 9229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5111083388328552,
+      "learning_rate": 0.0005669673956544219,
+      "loss": 3.049,
+      "step": 9230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390723943710327,
+      "learning_rate": 0.0005669603719826501,
+      "loss": 3.1554,
+      "step": 9231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5268939733505249,
+      "learning_rate": 0.0005669533476077561,
+      "loss": 3.4162,
+      "step": 9232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.528060793876648,
+      "learning_rate": 0.0005669463225297585,
+      "loss": 3.2039,
+      "step": 9233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5250667333602905,
+      "learning_rate": 0.0005669392967486755,
+      "loss": 3.2844,
+      "step": 9234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.557022213935852,
+      "learning_rate": 0.0005669322702645257,
+      "loss": 3.3297,
+      "step": 9235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5029047131538391,
+      "learning_rate": 0.0005669252430773279,
+      "loss": 3.2711,
+      "step": 9236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5264648199081421,
+      "learning_rate": 0.0005669182151871003,
+      "loss": 3.3115,
+      "step": 9237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5282554626464844,
+      "learning_rate": 0.0005669111865938614,
+      "loss": 3.3636,
+      "step": 9238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4783882200717926,
+      "learning_rate": 0.0005669041572976299,
+      "loss": 3.4196,
+      "step": 9239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5388532280921936,
+      "learning_rate": 0.0005668971272984242,
+      "loss": 3.3577,
+      "step": 9240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5190770030021667,
+      "learning_rate": 0.0005668900965962627,
+      "loss": 3.252,
+      "step": 9241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4901539087295532,
+      "learning_rate": 0.0005668830651911642,
+      "loss": 3.4138,
+      "step": 9242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5431410074234009,
+      "learning_rate": 0.000566876033083147,
+      "loss": 3.2324,
+      "step": 9243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5126324892044067,
+      "learning_rate": 0.0005668690002722297,
+      "loss": 3.2226,
+      "step": 9244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5156897306442261,
+      "learning_rate": 0.0005668619667584309,
+      "loss": 3.2097,
+      "step": 9245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49389657378196716,
+      "learning_rate": 0.000566854932541769,
+      "loss": 3.284,
+      "step": 9246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5490319728851318,
+      "learning_rate": 0.0005668478976222624,
+      "loss": 3.4974,
+      "step": 9247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5475131273269653,
+      "learning_rate": 0.00056684086199993,
+      "loss": 3.4208,
+      "step": 9248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5155820846557617,
+      "learning_rate": 0.00056683382567479,
+      "loss": 3.4763,
+      "step": 9249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5059611201286316,
+      "learning_rate": 0.0005668267886468611,
+      "loss": 3.223,
+      "step": 9250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5430670380592346,
+      "learning_rate": 0.0005668197509161618,
+      "loss": 3.1563,
+      "step": 9251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5147794485092163,
+      "learning_rate": 0.0005668127124827106,
+      "loss": 3.218,
+      "step": 9252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5046300888061523,
+      "learning_rate": 0.0005668056733465262,
+      "loss": 3.118,
+      "step": 9253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5530743598937988,
+      "learning_rate": 0.0005667986335076269,
+      "loss": 3.2957,
+      "step": 9254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5372881889343262,
+      "learning_rate": 0.0005667915929660313,
+      "loss": 3.3998,
+      "step": 9255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5331935286521912,
+      "learning_rate": 0.000566784551721758,
+      "loss": 3.4255,
+      "step": 9256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5113346576690674,
+      "learning_rate": 0.0005667775097748255,
+      "loss": 3.1996,
+      "step": 9257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4899602234363556,
+      "learning_rate": 0.0005667704671252525,
+      "loss": 3.484,
+      "step": 9258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.535860538482666,
+      "learning_rate": 0.0005667634237730573,
+      "loss": 3.1899,
+      "step": 9259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5384031534194946,
+      "learning_rate": 0.0005667563797182586,
+      "loss": 3.3316,
+      "step": 9260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5427521467208862,
+      "learning_rate": 0.0005667493349608751,
+      "loss": 3.3506,
+      "step": 9261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.543279230594635,
+      "learning_rate": 0.000566742289500925,
+      "loss": 3.2267,
+      "step": 9262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5359872579574585,
+      "learning_rate": 0.0005667352433384272,
+      "loss": 3.2173,
+      "step": 9263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5161938667297363,
+      "learning_rate": 0.0005667281964734,
+      "loss": 3.2177,
+      "step": 9264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5487397313117981,
+      "learning_rate": 0.0005667211489058621,
+      "loss": 3.3548,
+      "step": 9265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5496209263801575,
+      "learning_rate": 0.0005667141006358321,
+      "loss": 3.1292,
+      "step": 9266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5191566348075867,
+      "learning_rate": 0.0005667070516633283,
+      "loss": 3.31,
+      "step": 9267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5273240804672241,
+      "learning_rate": 0.0005667000019883696,
+      "loss": 2.9271,
+      "step": 9268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5322637557983398,
+      "learning_rate": 0.0005666929516109744,
+      "loss": 3.2176,
+      "step": 9269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5463858246803284,
+      "learning_rate": 0.0005666859005311612,
+      "loss": 3.1465,
+      "step": 9270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5420017838478088,
+      "learning_rate": 0.0005666788487489488,
+      "loss": 3.1499,
+      "step": 9271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5984858274459839,
+      "learning_rate": 0.0005666717962643555,
+      "loss": 3.293,
+      "step": 9272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5361172556877136,
+      "learning_rate": 0.0005666647430774001,
+      "loss": 3.3239,
+      "step": 9273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5419718623161316,
+      "learning_rate": 0.000566657689188101,
+      "loss": 3.3363,
+      "step": 9274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6194941997528076,
+      "learning_rate": 0.000566650634596477,
+      "loss": 3.2487,
+      "step": 9275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5028418302536011,
+      "learning_rate": 0.0005666435793025465,
+      "loss": 3.5507,
+      "step": 9276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5127885937690735,
+      "learning_rate": 0.0005666365233063281,
+      "loss": 3.4255,
+      "step": 9277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5219062566757202,
+      "learning_rate": 0.0005666294666078404,
+      "loss": 3.1941,
+      "step": 9278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.515256404876709,
+      "learning_rate": 0.0005666224092071019,
+      "loss": 3.3369,
+      "step": 9279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5556721687316895,
+      "learning_rate": 0.0005666153511041315,
+      "loss": 3.3815,
+      "step": 9280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5763710737228394,
+      "learning_rate": 0.0005666082922989475,
+      "loss": 3.3639,
+      "step": 9281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5201305747032166,
+      "learning_rate": 0.0005666012327915686,
+      "loss": 3.2809,
+      "step": 9282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5183270573616028,
+      "learning_rate": 0.0005665941725820133,
+      "loss": 3.1675,
+      "step": 9283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.533555805683136,
+      "learning_rate": 0.0005665871116703003,
+      "loss": 3.195,
+      "step": 9284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5322220325469971,
+      "learning_rate": 0.000566580050056448,
+      "loss": 3.3518,
+      "step": 9285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5173676013946533,
+      "learning_rate": 0.0005665729877404753,
+      "loss": 3.3363,
+      "step": 9286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5588550567626953,
+      "learning_rate": 0.0005665659247224006,
+      "loss": 3.3185,
+      "step": 9287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5288047790527344,
+      "learning_rate": 0.0005665588610022426,
+      "loss": 3.2531,
+      "step": 9288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4969140589237213,
+      "learning_rate": 0.0005665517965800199,
+      "loss": 3.3127,
+      "step": 9289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5432141423225403,
+      "learning_rate": 0.000566544731455751,
+      "loss": 3.082,
+      "step": 9290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5413832664489746,
+      "learning_rate": 0.0005665376656294545,
+      "loss": 3.3992,
+      "step": 9291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5215909481048584,
+      "learning_rate": 0.0005665305991011492,
+      "loss": 3.5054,
+      "step": 9292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5591144561767578,
+      "learning_rate": 0.0005665235318708537,
+      "loss": 3.3934,
+      "step": 9293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6027560234069824,
+      "learning_rate": 0.0005665164639385863,
+      "loss": 3.252,
+      "step": 9294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5260102152824402,
+      "learning_rate": 0.0005665093953043658,
+      "loss": 3.2683,
+      "step": 9295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.525118887424469,
+      "learning_rate": 0.0005665023259682111,
+      "loss": 3.277,
+      "step": 9296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4957008957862854,
+      "learning_rate": 0.0005664952559301403,
+      "loss": 3.1469,
+      "step": 9297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5985009074211121,
+      "learning_rate": 0.0005664881851901725,
+      "loss": 3.195,
+      "step": 9298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5201311111450195,
+      "learning_rate": 0.0005664811137483259,
+      "loss": 3.4488,
+      "step": 9299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5100109577178955,
+      "learning_rate": 0.0005664740416046195,
+      "loss": 3.1562,
+      "step": 9300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49282416701316833,
+      "learning_rate": 0.0005664669687590717,
+      "loss": 3.3542,
+      "step": 9301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5214360356330872,
+      "learning_rate": 0.0005664598952117012,
+      "loss": 3.2439,
+      "step": 9302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5333632826805115,
+      "learning_rate": 0.0005664528209625265,
+      "loss": 3.2928,
+      "step": 9303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5121089816093445,
+      "learning_rate": 0.0005664457460115665,
+      "loss": 3.3875,
+      "step": 9304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48593372106552124,
+      "learning_rate": 0.0005664386703588396,
+      "loss": 3.3829,
+      "step": 9305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5539550185203552,
+      "learning_rate": 0.0005664315940043645,
+      "loss": 3.427,
+      "step": 9306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5157627463340759,
+      "learning_rate": 0.00056642451694816,
+      "loss": 3.3568,
+      "step": 9307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5772610902786255,
+      "learning_rate": 0.0005664174391902444,
+      "loss": 3.0817,
+      "step": 9308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5100268125534058,
+      "learning_rate": 0.0005664103607306367,
+      "loss": 3.1593,
+      "step": 9309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5215173363685608,
+      "learning_rate": 0.0005664032815693553,
+      "loss": 3.1953,
+      "step": 9310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5628623962402344,
+      "learning_rate": 0.000566396201706419,
+      "loss": 3.1241,
+      "step": 9311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5145275592803955,
+      "learning_rate": 0.0005663891211418463,
+      "loss": 3.2781,
+      "step": 9312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.545166015625,
+      "learning_rate": 0.0005663820398756559,
+      "loss": 3.3879,
+      "step": 9313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5119915008544922,
+      "learning_rate": 0.0005663749579078665,
+      "loss": 3.0349,
+      "step": 9314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5298326015472412,
+      "learning_rate": 0.0005663678752384968,
+      "loss": 3.3834,
+      "step": 9315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5123884081840515,
+      "learning_rate": 0.0005663607918675654,
+      "loss": 3.3531,
+      "step": 9316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5344944000244141,
+      "learning_rate": 0.0005663537077950908,
+      "loss": 3.4021,
+      "step": 9317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5218607187271118,
+      "learning_rate": 0.0005663466230210919,
+      "loss": 3.1947,
+      "step": 9318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5089520215988159,
+      "learning_rate": 0.0005663395375455872,
+      "loss": 3.3224,
+      "step": 9319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5154445171356201,
+      "learning_rate": 0.0005663324513685954,
+      "loss": 3.4781,
+      "step": 9320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4960334002971649,
+      "learning_rate": 0.0005663253644901351,
+      "loss": 3.316,
+      "step": 9321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5327979326248169,
+      "learning_rate": 0.0005663182769102252,
+      "loss": 3.3396,
+      "step": 9322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5505029559135437,
+      "learning_rate": 0.0005663111886288842,
+      "loss": 3.2301,
+      "step": 9323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5501486659049988,
+      "learning_rate": 0.0005663040996461308,
+      "loss": 3.4768,
+      "step": 9324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5302287936210632,
+      "learning_rate": 0.0005662970099619835,
+      "loss": 3.1638,
+      "step": 9325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5448341965675354,
+      "learning_rate": 0.0005662899195764612,
+      "loss": 3.17,
+      "step": 9326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5440930724143982,
+      "learning_rate": 0.0005662828284895826,
+      "loss": 3.4745,
+      "step": 9327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5151364803314209,
+      "learning_rate": 0.0005662757367013662,
+      "loss": 3.1387,
+      "step": 9328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4982331395149231,
+      "learning_rate": 0.0005662686442118308,
+      "loss": 3.285,
+      "step": 9329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5508311986923218,
+      "learning_rate": 0.0005662615510209949,
+      "loss": 3.2912,
+      "step": 9330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4659712016582489,
+      "learning_rate": 0.0005662544571288775,
+      "loss": 3.4545,
+      "step": 9331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5672176480293274,
+      "learning_rate": 0.0005662473625354969,
+      "loss": 3.2206,
+      "step": 9332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5322652459144592,
+      "learning_rate": 0.0005662402672408722,
+      "loss": 3.2292,
+      "step": 9333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5741870403289795,
+      "learning_rate": 0.0005662331712450216,
+      "loss": 3.2488,
+      "step": 9334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5742843151092529,
+      "learning_rate": 0.0005662260745479643,
+      "loss": 3.36,
+      "step": 9335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5701401233673096,
+      "learning_rate": 0.0005662189771497187,
+      "loss": 3.166,
+      "step": 9336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5188326239585876,
+      "learning_rate": 0.0005662118790503035,
+      "loss": 3.4782,
+      "step": 9337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5277048945426941,
+      "learning_rate": 0.0005662047802497376,
+      "loss": 3.4702,
+      "step": 9338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5392853021621704,
+      "learning_rate": 0.0005661976807480394,
+      "loss": 3.5907,
+      "step": 9339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5241400599479675,
+      "learning_rate": 0.0005661905805452277,
+      "loss": 3.5178,
+      "step": 9340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5726556181907654,
+      "learning_rate": 0.0005661834796413214,
+      "loss": 3.3445,
+      "step": 9341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5296681523323059,
+      "learning_rate": 0.0005661763780363389,
+      "loss": 3.3437,
+      "step": 9342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5572839379310608,
+      "learning_rate": 0.0005661692757302991,
+      "loss": 3.4079,
+      "step": 9343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5110607147216797,
+      "learning_rate": 0.0005661621727232206,
+      "loss": 3.2675,
+      "step": 9344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4931510388851166,
+      "learning_rate": 0.0005661550690151222,
+      "loss": 3.3693,
+      "step": 9345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5317081212997437,
+      "learning_rate": 0.0005661479646060227,
+      "loss": 3.2439,
+      "step": 9346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5172111392021179,
+      "learning_rate": 0.0005661408594959405,
+      "loss": 3.3955,
+      "step": 9347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.567136824131012,
+      "learning_rate": 0.0005661337536848946,
+      "loss": 3.4301,
+      "step": 9348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5417160987854004,
+      "learning_rate": 0.0005661266471729035,
+      "loss": 3.268,
+      "step": 9349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5237059593200684,
+      "learning_rate": 0.0005661195399599861,
+      "loss": 3.0861,
+      "step": 9350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5109018087387085,
+      "learning_rate": 0.0005661124320461611,
+      "loss": 3.3038,
+      "step": 9351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4800640940666199,
+      "learning_rate": 0.0005661053234314471,
+      "loss": 3.2938,
+      "step": 9352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5349900126457214,
+      "learning_rate": 0.000566098214115863,
+      "loss": 3.1207,
+      "step": 9353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49598759412765503,
+      "learning_rate": 0.0005660911040994272,
+      "loss": 3.2266,
+      "step": 9354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5334964394569397,
+      "learning_rate": 0.0005660839933821588,
+      "loss": 3.2708,
+      "step": 9355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6042339205741882,
+      "learning_rate": 0.0005660768819640764,
+      "loss": 3.1651,
+      "step": 9356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5126491785049438,
+      "learning_rate": 0.0005660697698451985,
+      "loss": 3.1133,
+      "step": 9357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5190067291259766,
+      "learning_rate": 0.0005660626570255442,
+      "loss": 3.5437,
+      "step": 9358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5068770051002502,
+      "learning_rate": 0.0005660555435051321,
+      "loss": 3.4367,
+      "step": 9359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5406444668769836,
+      "learning_rate": 0.0005660484292839807,
+      "loss": 3.3291,
+      "step": 9360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46386462450027466,
+      "learning_rate": 0.0005660413143621091,
+      "loss": 3.2917,
+      "step": 9361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5372816920280457,
+      "learning_rate": 0.000566034198739536,
+      "loss": 3.2199,
+      "step": 9362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5300079584121704,
+      "learning_rate": 0.0005660270824162798,
+      "loss": 3.2802,
+      "step": 9363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5637562274932861,
+      "learning_rate": 0.0005660199653923594,
+      "loss": 3.2477,
+      "step": 9364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5250990390777588,
+      "learning_rate": 0.0005660128476677939,
+      "loss": 3.078,
+      "step": 9365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5225361585617065,
+      "learning_rate": 0.0005660057292426016,
+      "loss": 3.4664,
+      "step": 9366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4957283139228821,
+      "learning_rate": 0.0005659986101168013,
+      "loss": 3.3938,
+      "step": 9367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5098510384559631,
+      "learning_rate": 0.0005659914902904121,
+      "loss": 3.2841,
+      "step": 9368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5317561030387878,
+      "learning_rate": 0.0005659843697634522,
+      "loss": 3.2659,
+      "step": 9369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4984907805919647,
+      "learning_rate": 0.0005659772485359409,
+      "loss": 3.1747,
+      "step": 9370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5244981646537781,
+      "learning_rate": 0.0005659701266078966,
+      "loss": 3.2435,
+      "step": 9371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5228347182273865,
+      "learning_rate": 0.0005659630039793383,
+      "loss": 3.2482,
+      "step": 9372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5450955629348755,
+      "learning_rate": 0.0005659558806502845,
+      "loss": 3.3684,
+      "step": 9373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5306031107902527,
+      "learning_rate": 0.0005659487566207542,
+      "loss": 2.9565,
+      "step": 9374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5437656044960022,
+      "learning_rate": 0.000565941631890766,
+      "loss": 3.2759,
+      "step": 9375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5219473838806152,
+      "learning_rate": 0.0005659345064603387,
+      "loss": 3.3767,
+      "step": 9376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5270979404449463,
+      "learning_rate": 0.000565927380329491,
+      "loss": 3.4866,
+      "step": 9377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5120222568511963,
+      "learning_rate": 0.0005659202534982419,
+      "loss": 3.1391,
+      "step": 9378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.520464301109314,
+      "learning_rate": 0.00056591312596661,
+      "loss": 3.2149,
+      "step": 9379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5516645908355713,
+      "learning_rate": 0.0005659059977346141,
+      "loss": 3.2735,
+      "step": 9380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.519219696521759,
+      "learning_rate": 0.000565898868802273,
+      "loss": 3.242,
+      "step": 9381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5575689077377319,
+      "learning_rate": 0.0005658917391696054,
+      "loss": 3.366,
+      "step": 9382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5153242349624634,
+      "learning_rate": 0.0005658846088366302,
+      "loss": 3.2188,
+      "step": 9383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5371017456054688,
+      "learning_rate": 0.000565877477803366,
+      "loss": 3.3019,
+      "step": 9384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.57710200548172,
+      "learning_rate": 0.0005658703460698318,
+      "loss": 3.4646,
+      "step": 9385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5319957137107849,
+      "learning_rate": 0.0005658632136360461,
+      "loss": 3.1126,
+      "step": 9386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5224001407623291,
+      "learning_rate": 0.0005658560805020281,
+      "loss": 3.384,
+      "step": 9387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5222063660621643,
+      "learning_rate": 0.0005658489466677963,
+      "loss": 3.3125,
+      "step": 9388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5301948189735413,
+      "learning_rate": 0.0005658418121333694,
+      "loss": 3.3855,
+      "step": 9389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.506697952747345,
+      "learning_rate": 0.0005658346768987664,
+      "loss": 3.4046,
+      "step": 9390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5180991291999817,
+      "learning_rate": 0.0005658275409640062,
+      "loss": 3.2995,
+      "step": 9391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5402610898017883,
+      "learning_rate": 0.0005658204043291072,
+      "loss": 3.412,
+      "step": 9392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5179954171180725,
+      "learning_rate": 0.0005658132669940885,
+      "loss": 3.4606,
+      "step": 9393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5407754778862,
+      "learning_rate": 0.0005658061289589687,
+      "loss": 3.0892,
+      "step": 9394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5019816756248474,
+      "learning_rate": 0.0005657989902237669,
+      "loss": 3.2621,
+      "step": 9395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5457536578178406,
+      "learning_rate": 0.0005657918507885016,
+      "loss": 3.1853,
+      "step": 9396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5361157059669495,
+      "learning_rate": 0.0005657847106531916,
+      "loss": 3.1843,
+      "step": 9397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5209142565727234,
+      "learning_rate": 0.000565777569817856,
+      "loss": 3.4549,
+      "step": 9398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49910151958465576,
+      "learning_rate": 0.0005657704282825133,
+      "loss": 3.4455,
+      "step": 9399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5092660188674927,
+      "learning_rate": 0.0005657632860471826,
+      "loss": 3.1643,
+      "step": 9400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5319451093673706,
+      "learning_rate": 0.0005657561431118824,
+      "loss": 3.1362,
+      "step": 9401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5203292369842529,
+      "learning_rate": 0.0005657489994766318,
+      "loss": 3.3171,
+      "step": 9402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5382462739944458,
+      "learning_rate": 0.0005657418551414494,
+      "loss": 3.3471,
+      "step": 9403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5401602387428284,
+      "learning_rate": 0.0005657347101063541,
+      "loss": 3.415,
+      "step": 9404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5223503112792969,
+      "learning_rate": 0.0005657275643713648,
+      "loss": 3.2212,
+      "step": 9405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4994029998779297,
+      "learning_rate": 0.0005657204179365001,
+      "loss": 3.1509,
+      "step": 9406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5694056749343872,
+      "learning_rate": 0.000565713270801779,
+      "loss": 3.3337,
+      "step": 9407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5182304978370667,
+      "learning_rate": 0.0005657061229672203,
+      "loss": 3.4504,
+      "step": 9408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5066438317298889,
+      "learning_rate": 0.0005656989744328428,
+      "loss": 2.98,
+      "step": 9409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5442813634872437,
+      "learning_rate": 0.0005656918251986654,
+      "loss": 3.2891,
+      "step": 9410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5112155675888062,
+      "learning_rate": 0.0005656846752647068,
+      "loss": 3.4086,
+      "step": 9411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4860506057739258,
+      "learning_rate": 0.0005656775246309859,
+      "loss": 3.1496,
+      "step": 9412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5639499425888062,
+      "learning_rate": 0.0005656703732975215,
+      "loss": 3.4534,
+      "step": 9413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49762552976608276,
+      "learning_rate": 0.0005656632212643326,
+      "loss": 3.0931,
+      "step": 9414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5720370411872864,
+      "learning_rate": 0.0005656560685314378,
+      "loss": 3.3915,
+      "step": 9415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5175341963768005,
+      "learning_rate": 0.000565648915098856,
+      "loss": 3.3709,
+      "step": 9416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5201534032821655,
+      "learning_rate": 0.0005656417609666061,
+      "loss": 3.2131,
+      "step": 9417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5602389574050903,
+      "learning_rate": 0.000565634606134707,
+      "loss": 3.4146,
+      "step": 9418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5673112869262695,
+      "learning_rate": 0.0005656274506031775,
+      "loss": 3.29,
+      "step": 9419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5292515754699707,
+      "learning_rate": 0.0005656202943720363,
+      "loss": 3.2578,
+      "step": 9420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4988608658313751,
+      "learning_rate": 0.0005656131374413024,
+      "loss": 3.1482,
+      "step": 9421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5025928020477295,
+      "learning_rate": 0.0005656059798109947,
+      "loss": 3.3671,
+      "step": 9422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5336331725120544,
+      "learning_rate": 0.0005655988214811318,
+      "loss": 3.1554,
+      "step": 9423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4792225956916809,
+      "learning_rate": 0.0005655916624517328,
+      "loss": 3.3616,
+      "step": 9424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5044949650764465,
+      "learning_rate": 0.0005655845027228164,
+      "loss": 3.2776,
+      "step": 9425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.519375205039978,
+      "learning_rate": 0.0005655773422944017,
+      "loss": 3.4197,
+      "step": 9426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5419018268585205,
+      "learning_rate": 0.0005655701811665073,
+      "loss": 3.1409,
+      "step": 9427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5148293972015381,
+      "learning_rate": 0.0005655630193391522,
+      "loss": 3.2837,
+      "step": 9428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.522705078125,
+      "learning_rate": 0.0005655558568123551,
+      "loss": 3.2122,
+      "step": 9429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4861201345920563,
+      "learning_rate": 0.0005655486935861352,
+      "loss": 3.1855,
+      "step": 9430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4865737557411194,
+      "learning_rate": 0.0005655415296605109,
+      "loss": 3.1497,
+      "step": 9431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5147886276245117,
+      "learning_rate": 0.0005655343650355014,
+      "loss": 3.5226,
+      "step": 9432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5273894667625427,
+      "learning_rate": 0.0005655271997111256,
+      "loss": 3.3109,
+      "step": 9433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5883224010467529,
+      "learning_rate": 0.0005655200336874021,
+      "loss": 3.2307,
+      "step": 9434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.535631537437439,
+      "learning_rate": 0.00056551286696435,
+      "loss": 3.1471,
+      "step": 9435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5643936991691589,
+      "learning_rate": 0.0005655056995419881,
+      "loss": 3.1523,
+      "step": 9436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5332732200622559,
+      "learning_rate": 0.0005654985314203354,
+      "loss": 3.3777,
+      "step": 9437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4967159628868103,
+      "learning_rate": 0.0005654913625994105,
+      "loss": 3.5025,
+      "step": 9438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5050669312477112,
+      "learning_rate": 0.0005654841930792325,
+      "loss": 3.2085,
+      "step": 9439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5131103992462158,
+      "learning_rate": 0.0005654770228598202,
+      "loss": 3.3931,
+      "step": 9440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5723204612731934,
+      "learning_rate": 0.0005654698519411925,
+      "loss": 3.2395,
+      "step": 9441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.521449089050293,
+      "learning_rate": 0.0005654626803233684,
+      "loss": 3.2252,
+      "step": 9442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.559038519859314,
+      "learning_rate": 0.0005654555080063665,
+      "loss": 3.2708,
+      "step": 9443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5154927968978882,
+      "learning_rate": 0.000565448334990206,
+      "loss": 3.03,
+      "step": 9444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6243354678153992,
+      "learning_rate": 0.0005654411612749058,
+      "loss": 3.3827,
+      "step": 9445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6051532030105591,
+      "learning_rate": 0.0005654339868604846,
+      "loss": 3.2296,
+      "step": 9446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5143436789512634,
+      "learning_rate": 0.0005654268117469613,
+      "loss": 3.2727,
+      "step": 9447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.485236257314682,
+      "learning_rate": 0.0005654196359343548,
+      "loss": 3.2742,
+      "step": 9448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.536080539226532,
+      "learning_rate": 0.0005654124594226841,
+      "loss": 3.2754,
+      "step": 9449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.648050844669342,
+      "learning_rate": 0.0005654052822119681,
+      "loss": 3.0665,
+      "step": 9450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5583487749099731,
+      "learning_rate": 0.0005653981043022257,
+      "loss": 3.2576,
+      "step": 9451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5086358785629272,
+      "learning_rate": 0.0005653909256934757,
+      "loss": 3.3566,
+      "step": 9452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5181884765625,
+      "learning_rate": 0.0005653837463857371,
+      "loss": 3.1578,
+      "step": 9453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4989009499549866,
+      "learning_rate": 0.0005653765663790288,
+      "loss": 2.9139,
+      "step": 9454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5821312069892883,
+      "learning_rate": 0.0005653693856733698,
+      "loss": 3.2885,
+      "step": 9455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5479455590248108,
+      "learning_rate": 0.0005653622042687788,
+      "loss": 3.0113,
+      "step": 9456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5062962770462036,
+      "learning_rate": 0.0005653550221652747,
+      "loss": 3.2562,
+      "step": 9457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5756667256355286,
+      "learning_rate": 0.0005653478393628767,
+      "loss": 3.1515,
+      "step": 9458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5819968581199646,
+      "learning_rate": 0.0005653406558616034,
+      "loss": 3.3712,
+      "step": 9459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53989577293396,
+      "learning_rate": 0.0005653334716614741,
+      "loss": 3.1972,
+      "step": 9460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5572423934936523,
+      "learning_rate": 0.0005653262867625074,
+      "loss": 3.3386,
+      "step": 9461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5422435998916626,
+      "learning_rate": 0.0005653191011647223,
+      "loss": 3.1865,
+      "step": 9462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5119712352752686,
+      "learning_rate": 0.0005653119148681378,
+      "loss": 3.346,
+      "step": 9463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5083188414573669,
+      "learning_rate": 0.0005653047278727728,
+      "loss": 3.3056,
+      "step": 9464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5463467836380005,
+      "learning_rate": 0.0005652975401786461,
+      "loss": 3.3122,
+      "step": 9465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5052841305732727,
+      "learning_rate": 0.0005652903517857768,
+      "loss": 3.3023,
+      "step": 9466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5321139097213745,
+      "learning_rate": 0.0005652831626941838,
+      "loss": 3.2202,
+      "step": 9467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5242906808853149,
+      "learning_rate": 0.0005652759729038859,
+      "loss": 3.1611,
+      "step": 9468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5150046944618225,
+      "learning_rate": 0.0005652687824149022,
+      "loss": 3.3922,
+      "step": 9469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5529156923294067,
+      "learning_rate": 0.0005652615912272516,
+      "loss": 3.2372,
+      "step": 9470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4883171021938324,
+      "learning_rate": 0.000565254399340953,
+      "loss": 3.1305,
+      "step": 9471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5113022923469543,
+      "learning_rate": 0.0005652472067560254,
+      "loss": 3.4073,
+      "step": 9472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5019404292106628,
+      "learning_rate": 0.0005652400134724877,
+      "loss": 3.2087,
+      "step": 9473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5218349695205688,
+      "learning_rate": 0.0005652328194903587,
+      "loss": 3.2981,
+      "step": 9474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6199938058853149,
+      "learning_rate": 0.0005652256248096577,
+      "loss": 3.3046,
+      "step": 9475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5042975544929504,
+      "learning_rate": 0.0005652184294304034,
+      "loss": 3.3814,
+      "step": 9476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5005037784576416,
+      "learning_rate": 0.0005652112333526147,
+      "loss": 3.316,
+      "step": 9477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.507548451423645,
+      "learning_rate": 0.0005652040365763107,
+      "loss": 3.2449,
+      "step": 9478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5326326489448547,
+      "learning_rate": 0.0005651968391015104,
+      "loss": 3.333,
+      "step": 9479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5457335114479065,
+      "learning_rate": 0.0005651896409282326,
+      "loss": 3.2439,
+      "step": 9480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5438616275787354,
+      "learning_rate": 0.0005651824420564962,
+      "loss": 3.0847,
+      "step": 9481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5413506627082825,
+      "learning_rate": 0.0005651752424863205,
+      "loss": 3.3258,
+      "step": 9482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5376286506652832,
+      "learning_rate": 0.0005651680422177241,
+      "loss": 3.0409,
+      "step": 9483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5599796772003174,
+      "learning_rate": 0.0005651608412507262,
+      "loss": 3.2176,
+      "step": 9484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5151841044425964,
+      "learning_rate": 0.0005651536395853456,
+      "loss": 3.1449,
+      "step": 9485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5488869547843933,
+      "learning_rate": 0.0005651464372216015,
+      "loss": 3.2857,
+      "step": 9486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5191092491149902,
+      "learning_rate": 0.0005651392341595125,
+      "loss": 3.3297,
+      "step": 9487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5657668709754944,
+      "learning_rate": 0.0005651320303990978,
+      "loss": 3.1369,
+      "step": 9488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5233826637268066,
+      "learning_rate": 0.0005651248259403765,
+      "loss": 3.2528,
+      "step": 9489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49316510558128357,
+      "learning_rate": 0.0005651176207833673,
+      "loss": 3.2797,
+      "step": 9490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5391998291015625,
+      "learning_rate": 0.0005651104149280894,
+      "loss": 3.2407,
+      "step": 9491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7097063660621643,
+      "learning_rate": 0.0005651032083745616,
+      "loss": 3.3675,
+      "step": 9492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5312466621398926,
+      "learning_rate": 0.0005650960011228031,
+      "loss": 3.2063,
+      "step": 9493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5219338536262512,
+      "learning_rate": 0.0005650887931728326,
+      "loss": 3.3838,
+      "step": 9494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5633711218833923,
+      "learning_rate": 0.0005650815845246694,
+      "loss": 3.084,
+      "step": 9495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5363718867301941,
+      "learning_rate": 0.0005650743751783321,
+      "loss": 3.2394,
+      "step": 9496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5066879987716675,
+      "learning_rate": 0.0005650671651338401,
+      "loss": 3.3002,
+      "step": 9497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5164818167686462,
+      "learning_rate": 0.0005650599543912121,
+      "loss": 2.9735,
+      "step": 9498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5876210331916809,
+      "learning_rate": 0.0005650527429504673,
+      "loss": 3.4418,
+      "step": 9499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4980337917804718,
+      "learning_rate": 0.0005650455308116245,
+      "loss": 3.3927,
+      "step": 9500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5478733777999878,
+      "learning_rate": 0.0005650383179747028,
+      "loss": 3.2894,
+      "step": 9501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5227766036987305,
+      "learning_rate": 0.0005650311044397212,
+      "loss": 3.3741,
+      "step": 9502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4926789700984955,
+      "learning_rate": 0.0005650238902066987,
+      "loss": 3.3118,
+      "step": 9503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5030087828636169,
+      "learning_rate": 0.0005650166752756542,
+      "loss": 3.4107,
+      "step": 9504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5236660838127136,
+      "learning_rate": 0.0005650094596466068,
+      "loss": 3.2949,
+      "step": 9505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5276412963867188,
+      "learning_rate": 0.0005650022433195755,
+      "loss": 3.2227,
+      "step": 9506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5344411730766296,
+      "learning_rate": 0.0005649950262945794,
+      "loss": 3.1652,
+      "step": 9507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5086687207221985,
+      "learning_rate": 0.0005649878085716372,
+      "loss": 3.1716,
+      "step": 9508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5827347636222839,
+      "learning_rate": 0.0005649805901507682,
+      "loss": 3.265,
+      "step": 9509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5286895036697388,
+      "learning_rate": 0.0005649733710319913,
+      "loss": 3.3054,
+      "step": 9510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5271221399307251,
+      "learning_rate": 0.0005649661512153256,
+      "loss": 3.2053,
+      "step": 9511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4985242784023285,
+      "learning_rate": 0.00056495893070079,
+      "loss": 3.1904,
+      "step": 9512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5197968482971191,
+      "learning_rate": 0.0005649517094884036,
+      "loss": 3.2541,
+      "step": 9513
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5603257417678833,
+      "learning_rate": 0.0005649444875781853,
+      "loss": 3.3489,
+      "step": 9514
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5142351388931274,
+      "learning_rate": 0.0005649372649701544,
+      "loss": 3.2883,
+      "step": 9515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6001418232917786,
+      "learning_rate": 0.0005649300416643296,
+      "loss": 3.1696,
+      "step": 9516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5239596962928772,
+      "learning_rate": 0.0005649228176607301,
+      "loss": 3.3235,
+      "step": 9517
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48557403683662415,
+      "learning_rate": 0.0005649155929593748,
+      "loss": 3.1674,
+      "step": 9518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5428034663200378,
+      "learning_rate": 0.0005649083675602829,
+      "loss": 3.1312,
+      "step": 9519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5654087662696838,
+      "learning_rate": 0.0005649011414634733,
+      "loss": 3.5056,
+      "step": 9520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4970563054084778,
+      "learning_rate": 0.000564893914668965,
+      "loss": 3.3086,
+      "step": 9521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5339492559432983,
+      "learning_rate": 0.0005648866871767772,
+      "loss": 3.2905,
+      "step": 9522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5164199471473694,
+      "learning_rate": 0.0005648794589869289,
+      "loss": 3.1285,
+      "step": 9523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.476354718208313,
+      "learning_rate": 0.000564872230099439,
+      "loss": 3.2268,
+      "step": 9524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5380869507789612,
+      "learning_rate": 0.0005648650005143267,
+      "loss": 3.0886,
+      "step": 9525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5686689019203186,
+      "learning_rate": 0.0005648577702316108,
+      "loss": 3.353,
+      "step": 9526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5547523498535156,
+      "learning_rate": 0.0005648505392513107,
+      "loss": 3.2541,
+      "step": 9527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5288465619087219,
+      "learning_rate": 0.0005648433075734451,
+      "loss": 3.3771,
+      "step": 9528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5291219353675842,
+      "learning_rate": 0.0005648360751980332,
+      "loss": 3.2426,
+      "step": 9529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5199901461601257,
+      "learning_rate": 0.0005648288421250942,
+      "loss": 3.245,
+      "step": 9530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5697982311248779,
+      "learning_rate": 0.0005648216083546469,
+      "loss": 3.3015,
+      "step": 9531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5655345916748047,
+      "learning_rate": 0.0005648143738867104,
+      "loss": 3.3584,
+      "step": 9532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5142018795013428,
+      "learning_rate": 0.0005648071387213039,
+      "loss": 3.2386,
+      "step": 9533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5273463129997253,
+      "learning_rate": 0.0005647999028584463,
+      "loss": 3.211,
+      "step": 9534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58909672498703,
+      "learning_rate": 0.0005647926662981568,
+      "loss": 3.4849,
+      "step": 9535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7461726665496826,
+      "learning_rate": 0.0005647854290404543,
+      "loss": 3.036,
+      "step": 9536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5585423707962036,
+      "learning_rate": 0.0005647781910853579,
+      "loss": 3.2072,
+      "step": 9537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5416821241378784,
+      "learning_rate": 0.0005647709524328867,
+      "loss": 3.3082,
+      "step": 9538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5571542382240295,
+      "learning_rate": 0.0005647637130830599,
+      "loss": 3.4454,
+      "step": 9539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5479061007499695,
+      "learning_rate": 0.0005647564730358963,
+      "loss": 3.3225,
+      "step": 9540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5047709941864014,
+      "learning_rate": 0.0005647492322914152,
+      "loss": 3.4261,
+      "step": 9541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5049424171447754,
+      "learning_rate": 0.0005647419908496355,
+      "loss": 3.3516,
+      "step": 9542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5358158349990845,
+      "learning_rate": 0.0005647347487105764,
+      "loss": 3.309,
+      "step": 9543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5347208380699158,
+      "learning_rate": 0.0005647275058742569,
+      "loss": 3.3036,
+      "step": 9544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5153633952140808,
+      "learning_rate": 0.000564720262340696,
+      "loss": 3.3403,
+      "step": 9545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5642601847648621,
+      "learning_rate": 0.0005647130181099131,
+      "loss": 3.3031,
+      "step": 9546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5254157185554504,
+      "learning_rate": 0.0005647057731819269,
+      "loss": 3.3091,
+      "step": 9547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5136579275131226,
+      "learning_rate": 0.0005646985275567566,
+      "loss": 3.1616,
+      "step": 9548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.504160463809967,
+      "learning_rate": 0.0005646912812344214,
+      "loss": 3.2302,
+      "step": 9549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5003314018249512,
+      "learning_rate": 0.0005646840342149403,
+      "loss": 3.0769,
+      "step": 9550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5534130334854126,
+      "learning_rate": 0.0005646767864983325,
+      "loss": 3.3755,
+      "step": 9551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5688060522079468,
+      "learning_rate": 0.0005646695380846168,
+      "loss": 3.3327,
+      "step": 9552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5155788660049438,
+      "learning_rate": 0.0005646622889738125,
+      "loss": 3.2615,
+      "step": 9553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5471298098564148,
+      "learning_rate": 0.0005646550391659387,
+      "loss": 3.3885,
+      "step": 9554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5412099957466125,
+      "learning_rate": 0.0005646477886610145,
+      "loss": 3.215,
+      "step": 9555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4741232693195343,
+      "learning_rate": 0.0005646405374590589,
+      "loss": 3.4638,
+      "step": 9556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5171995162963867,
+      "learning_rate": 0.0005646332855600911,
+      "loss": 3.1037,
+      "step": 9557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48581480979919434,
+      "learning_rate": 0.0005646260329641302,
+      "loss": 3.3405,
+      "step": 9558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5740549564361572,
+      "learning_rate": 0.0005646187796711951,
+      "loss": 2.8654,
+      "step": 9559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5170774459838867,
+      "learning_rate": 0.0005646115256813053,
+      "loss": 3.2295,
+      "step": 9560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5482763648033142,
+      "learning_rate": 0.0005646042709944794,
+      "loss": 3.3797,
+      "step": 9561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4993712604045868,
+      "learning_rate": 0.0005645970156107369,
+      "loss": 3.3476,
+      "step": 9562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390969514846802,
+      "learning_rate": 0.0005645897595300967,
+      "loss": 3.2817,
+      "step": 9563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53498375415802,
+      "learning_rate": 0.0005645825027525781,
+      "loss": 3.3194,
+      "step": 9564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.528544008731842,
+      "learning_rate": 0.0005645752452782001,
+      "loss": 3.3329,
+      "step": 9565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5482951402664185,
+      "learning_rate": 0.0005645679871069817,
+      "loss": 3.2478,
+      "step": 9566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5523584485054016,
+      "learning_rate": 0.0005645607282389423,
+      "loss": 3.0492,
+      "step": 9567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5092670321464539,
+      "learning_rate": 0.0005645534686741009,
+      "loss": 3.1282,
+      "step": 9568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5134148597717285,
+      "learning_rate": 0.0005645462084124765,
+      "loss": 3.2474,
+      "step": 9569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5111011862754822,
+      "learning_rate": 0.0005645389474540882,
+      "loss": 3.1094,
+      "step": 9570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5071569085121155,
+      "learning_rate": 0.0005645316857989553,
+      "loss": 3.3619,
+      "step": 9571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49778759479522705,
+      "learning_rate": 0.0005645244234470969,
+      "loss": 3.1761,
+      "step": 9572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5030775666236877,
+      "learning_rate": 0.0005645171603985321,
+      "loss": 3.3071,
+      "step": 9573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5095931887626648,
+      "learning_rate": 0.0005645098966532798,
+      "loss": 3.4006,
+      "step": 9574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5154526233673096,
+      "learning_rate": 0.0005645026322113596,
+      "loss": 3.3653,
+      "step": 9575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4855976700782776,
+      "learning_rate": 0.0005644953670727902,
+      "loss": 3.4481,
+      "step": 9576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5199698805809021,
+      "learning_rate": 0.0005644881012375909,
+      "loss": 3.3039,
+      "step": 9577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5196504592895508,
+      "learning_rate": 0.0005644808347057809,
+      "loss": 3.4343,
+      "step": 9578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5275589227676392,
+      "learning_rate": 0.0005644735674773793,
+      "loss": 3.3105,
+      "step": 9579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4889765977859497,
+      "learning_rate": 0.0005644662995524051,
+      "loss": 3.2342,
+      "step": 9580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5421981811523438,
+      "learning_rate": 0.0005644590309308778,
+      "loss": 3.4189,
+      "step": 9581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5198723077774048,
+      "learning_rate": 0.0005644517616128161,
+      "loss": 3.3944,
+      "step": 9582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.518722653388977,
+      "learning_rate": 0.0005644444915982394,
+      "loss": 3.4588,
+      "step": 9583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5175428986549377,
+      "learning_rate": 0.0005644372208871668,
+      "loss": 3.2401,
+      "step": 9584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5273626446723938,
+      "learning_rate": 0.0005644299494796175,
+      "loss": 3.3321,
+      "step": 9585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5376038551330566,
+      "learning_rate": 0.0005644226773756107,
+      "loss": 3.1802,
+      "step": 9586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5051103234291077,
+      "learning_rate": 0.0005644154045751652,
+      "loss": 3.2906,
+      "step": 9587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5152342319488525,
+      "learning_rate": 0.0005644081310783006,
+      "loss": 3.231,
+      "step": 9588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.490278959274292,
+      "learning_rate": 0.0005644008568850359,
+      "loss": 3.333,
+      "step": 9589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5564194321632385,
+      "learning_rate": 0.0005643935819953901,
+      "loss": 3.1443,
+      "step": 9590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5331935286521912,
+      "learning_rate": 0.0005643863064093825,
+      "loss": 3.2971,
+      "step": 9591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5207473635673523,
+      "learning_rate": 0.0005643790301270323,
+      "loss": 3.3384,
+      "step": 9592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5474972724914551,
+      "learning_rate": 0.0005643717531483586,
+      "loss": 3.2845,
+      "step": 9593
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.534555196762085,
+      "learning_rate": 0.0005643644754733805,
+      "loss": 3.2346,
+      "step": 9594
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5441542267799377,
+      "learning_rate": 0.0005643571971021174,
+      "loss": 3.3468,
+      "step": 9595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5864543914794922,
+      "learning_rate": 0.0005643499180345882,
+      "loss": 3.3281,
+      "step": 9596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5105965733528137,
+      "learning_rate": 0.0005643426382708124,
+      "loss": 3.2396,
+      "step": 9597
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5301392674446106,
+      "learning_rate": 0.0005643353578108088,
+      "loss": 3.3751,
+      "step": 9598
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5022128820419312,
+      "learning_rate": 0.0005643280766545967,
+      "loss": 3.393,
+      "step": 9599
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5763443112373352,
+      "learning_rate": 0.0005643207948021954,
+      "loss": 3.0997,
+      "step": 9600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5480446815490723,
+      "learning_rate": 0.000564313512253624,
+      "loss": 3.2441,
+      "step": 9601
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5295218229293823,
+      "learning_rate": 0.0005643062290089017,
+      "loss": 3.4164,
+      "step": 9602
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.524186909198761,
+      "learning_rate": 0.0005642989450680474,
+      "loss": 3.4485,
+      "step": 9603
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5457528829574585,
+      "learning_rate": 0.0005642916604310809,
+      "loss": 3.5014,
+      "step": 9604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5163187384605408,
+      "learning_rate": 0.0005642843750980209,
+      "loss": 3.359,
+      "step": 9605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5899530053138733,
+      "learning_rate": 0.0005642770890688866,
+      "loss": 3.4003,
+      "step": 9606
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5038844347000122,
+      "learning_rate": 0.0005642698023436974,
+      "loss": 3.2742,
+      "step": 9607
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5313959717750549,
+      "learning_rate": 0.0005642625149224724,
+      "loss": 3.1566,
+      "step": 9608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5297811627388,
+      "learning_rate": 0.0005642552268052309,
+      "loss": 3.3764,
+      "step": 9609
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5239699482917786,
+      "learning_rate": 0.0005642479379919918,
+      "loss": 3.2483,
+      "step": 9610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5261852145195007,
+      "learning_rate": 0.0005642406484827746,
+      "loss": 3.2288,
+      "step": 9611
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5086889266967773,
+      "learning_rate": 0.0005642333582775984,
+      "loss": 3.2811,
+      "step": 9612
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5237385034561157,
+      "learning_rate": 0.0005642260673764822,
+      "loss": 3.2338,
+      "step": 9613
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5552691221237183,
+      "learning_rate": 0.0005642187757794456,
+      "loss": 3.2998,
+      "step": 9614
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5657722353935242,
+      "learning_rate": 0.0005642114834865076,
+      "loss": 3.2386,
+      "step": 9615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49646058678627014,
+      "learning_rate": 0.0005642041904976873,
+      "loss": 3.404,
+      "step": 9616
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5342143177986145,
+      "learning_rate": 0.000564196896813004,
+      "loss": 3.1347,
+      "step": 9617
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49228495359420776,
+      "learning_rate": 0.0005641896024324769,
+      "loss": 3.4549,
+      "step": 9618
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5277411937713623,
+      "learning_rate": 0.0005641823073561253,
+      "loss": 3.385,
+      "step": 9619
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5513384938240051,
+      "learning_rate": 0.0005641750115839685,
+      "loss": 3.2151,
+      "step": 9620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5188488960266113,
+      "learning_rate": 0.0005641677151160253,
+      "loss": 3.2605,
+      "step": 9621
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5301359295845032,
+      "learning_rate": 0.0005641604179523153,
+      "loss": 3.0863,
+      "step": 9622
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5934592485427856,
+      "learning_rate": 0.0005641531200928575,
+      "loss": 3.33,
+      "step": 9623
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5196259617805481,
+      "learning_rate": 0.0005641458215376713,
+      "loss": 3.1144,
+      "step": 9624
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5472568273544312,
+      "learning_rate": 0.0005641385222867758,
+      "loss": 3.197,
+      "step": 9625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5343920588493347,
+      "learning_rate": 0.0005641312223401904,
+      "loss": 3.3338,
+      "step": 9626
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5346314907073975,
+      "learning_rate": 0.000564123921697934,
+      "loss": 3.0962,
+      "step": 9627
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5246593356132507,
+      "learning_rate": 0.0005641166203600262,
+      "loss": 3.342,
+      "step": 9628
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5096229910850525,
+      "learning_rate": 0.000564109318326486,
+      "loss": 3.1208,
+      "step": 9629
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5147344470024109,
+      "learning_rate": 0.0005641020155973326,
+      "loss": 3.4918,
+      "step": 9630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5902569890022278,
+      "learning_rate": 0.0005640947121725853,
+      "loss": 3.4023,
+      "step": 9631
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5098241567611694,
+      "learning_rate": 0.0005640874080522635,
+      "loss": 3.2977,
+      "step": 9632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5633296966552734,
+      "learning_rate": 0.0005640801032363862,
+      "loss": 3.3381,
+      "step": 9633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5749189257621765,
+      "learning_rate": 0.0005640727977249728,
+      "loss": 3.3493,
+      "step": 9634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5577384829521179,
+      "learning_rate": 0.0005640654915180424,
+      "loss": 3.2943,
+      "step": 9635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5394460558891296,
+      "learning_rate": 0.0005640581846156143,
+      "loss": 3.1971,
+      "step": 9636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6581524610519409,
+      "learning_rate": 0.0005640508770177079,
+      "loss": 3.4347,
+      "step": 9637
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5236822962760925,
+      "learning_rate": 0.0005640435687243421,
+      "loss": 3.273,
+      "step": 9638
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5210537910461426,
+      "learning_rate": 0.0005640362597355365,
+      "loss": 3.2335,
+      "step": 9639
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5516870021820068,
+      "learning_rate": 0.0005640289500513101,
+      "loss": 3.3469,
+      "step": 9640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5240686535835266,
+      "learning_rate": 0.0005640216396716824,
+      "loss": 3.5318,
+      "step": 9641
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48787635564804077,
+      "learning_rate": 0.0005640143285966724,
+      "loss": 3.423,
+      "step": 9642
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234872102737427,
+      "learning_rate": 0.0005640070168262996,
+      "loss": 3.292,
+      "step": 9643
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5378441214561462,
+      "learning_rate": 0.000563999704360583,
+      "loss": 3.2868,
+      "step": 9644
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5410236716270447,
+      "learning_rate": 0.000563992391199542,
+      "loss": 3.2939,
+      "step": 9645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5149595737457275,
+      "learning_rate": 0.0005639850773431959,
+      "loss": 3.3569,
+      "step": 9646
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5150244235992432,
+      "learning_rate": 0.0005639777627915639,
+      "loss": 3.2811,
+      "step": 9647
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5107737183570862,
+      "learning_rate": 0.0005639704475446653,
+      "loss": 3.2627,
+      "step": 9648
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5012353658676147,
+      "learning_rate": 0.0005639631316025193,
+      "loss": 3.2458,
+      "step": 9649
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5054120421409607,
+      "learning_rate": 0.0005639558149651452,
+      "loss": 3.141,
+      "step": 9650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5207725763320923,
+      "learning_rate": 0.0005639484976325623,
+      "loss": 3.3233,
+      "step": 9651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5636855363845825,
+      "learning_rate": 0.0005639411796047898,
+      "loss": 3.1278,
+      "step": 9652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5159328579902649,
+      "learning_rate": 0.0005639338608818471,
+      "loss": 3.422,
+      "step": 9653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5192365050315857,
+      "learning_rate": 0.0005639265414637534,
+      "loss": 3.2066,
+      "step": 9654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5930947065353394,
+      "learning_rate": 0.0005639192213505279,
+      "loss": 3.2735,
+      "step": 9655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5183570981025696,
+      "learning_rate": 0.0005639119005421901,
+      "loss": 3.1819,
+      "step": 9656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5415933132171631,
+      "learning_rate": 0.000563904579038759,
+      "loss": 3.3708,
+      "step": 9657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5286021828651428,
+      "learning_rate": 0.0005638972568402542,
+      "loss": 3.3341,
+      "step": 9658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.523844838142395,
+      "learning_rate": 0.0005638899339466948,
+      "loss": 3.3128,
+      "step": 9659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5575231909751892,
+      "learning_rate": 0.0005638826103580999,
+      "loss": 3.2247,
+      "step": 9660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5189796686172485,
+      "learning_rate": 0.0005638752860744891,
+      "loss": 3.3183,
+      "step": 9661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5277138352394104,
+      "learning_rate": 0.0005638679610958817,
+      "loss": 3.2104,
+      "step": 9662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.500163733959198,
+      "learning_rate": 0.0005638606354222967,
+      "loss": 3.3856,
+      "step": 9663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.555261492729187,
+      "learning_rate": 0.0005638533090537536,
+      "loss": 3.3742,
+      "step": 9664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5257711410522461,
+      "learning_rate": 0.0005638459819902718,
+      "loss": 3.2025,
+      "step": 9665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5524579286575317,
+      "learning_rate": 0.0005638386542318703,
+      "loss": 3.2449,
+      "step": 9666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.561862051486969,
+      "learning_rate": 0.0005638313257785685,
+      "loss": 3.1967,
+      "step": 9667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5200432538986206,
+      "learning_rate": 0.0005638239966303859,
+      "loss": 3.2598,
+      "step": 9668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.524965226650238,
+      "learning_rate": 0.0005638166667873417,
+      "loss": 3.4027,
+      "step": 9669
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5578727722167969,
+      "learning_rate": 0.0005638093362494551,
+      "loss": 3.0414,
+      "step": 9670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5654264092445374,
+      "learning_rate": 0.0005638020050167456,
+      "loss": 3.3246,
+      "step": 9671
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5263561606407166,
+      "learning_rate": 0.0005637946730892323,
+      "loss": 3.2548,
+      "step": 9672
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5265440940856934,
+      "learning_rate": 0.0005637873404669345,
+      "loss": 3.1688,
+      "step": 9673
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5504220128059387,
+      "learning_rate": 0.0005637800071498717,
+      "loss": 3.1825,
+      "step": 9674
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5497219562530518,
+      "learning_rate": 0.0005637726731380631,
+      "loss": 3.1871,
+      "step": 9675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5197966694831848,
+      "learning_rate": 0.0005637653384315281,
+      "loss": 3.5599,
+      "step": 9676
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5388951897621155,
+      "learning_rate": 0.0005637580030302859,
+      "loss": 3.3868,
+      "step": 9677
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5066190361976624,
+      "learning_rate": 0.000563750666934356,
+      "loss": 3.1101,
+      "step": 9678
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5020130276679993,
+      "learning_rate": 0.0005637433301437575,
+      "loss": 3.2261,
+      "step": 9679
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5464184880256653,
+      "learning_rate": 0.0005637359926585099,
+      "loss": 3.2281,
+      "step": 9680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5577077865600586,
+      "learning_rate": 0.0005637286544786323,
+      "loss": 3.3123,
+      "step": 9681
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.564420759677887,
+      "learning_rate": 0.0005637213156041443,
+      "loss": 3.0541,
+      "step": 9682
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4688968360424042,
+      "learning_rate": 0.000563713976035065,
+      "loss": 3.2077,
+      "step": 9683
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4943731129169464,
+      "learning_rate": 0.000563706635771414,
+      "loss": 3.2392,
+      "step": 9684
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5055580139160156,
+      "learning_rate": 0.0005636992948132103,
+      "loss": 3.3449,
+      "step": 9685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4915640950202942,
+      "learning_rate": 0.0005636919531604736,
+      "loss": 3.3726,
+      "step": 9686
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4966067969799042,
+      "learning_rate": 0.0005636846108132229,
+      "loss": 3.322,
+      "step": 9687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5646604299545288,
+      "learning_rate": 0.0005636772677714777,
+      "loss": 3.1713,
+      "step": 9688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5298094153404236,
+      "learning_rate": 0.0005636699240352574,
+      "loss": 3.2453,
+      "step": 9689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48152562975883484,
+      "learning_rate": 0.0005636625796045813,
+      "loss": 3.3521,
+      "step": 9690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5157354474067688,
+      "learning_rate": 0.0005636552344794685,
+      "loss": 3.3635,
+      "step": 9691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5314420461654663,
+      "learning_rate": 0.0005636478886599387,
+      "loss": 3.1824,
+      "step": 9692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49262553453445435,
+      "learning_rate": 0.0005636405421460112,
+      "loss": 3.0643,
+      "step": 9693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5675431489944458,
+      "learning_rate": 0.000563633194937705,
+      "loss": 3.4652,
+      "step": 9694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5083454847335815,
+      "learning_rate": 0.0005636258470350399,
+      "loss": 3.1061,
+      "step": 9695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5256954431533813,
+      "learning_rate": 0.0005636184984380349,
+      "loss": 3.2333,
+      "step": 9696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5341222286224365,
+      "learning_rate": 0.0005636111491467097,
+      "loss": 3.3005,
+      "step": 9697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5235113501548767,
+      "learning_rate": 0.0005636037991610833,
+      "loss": 3.3963,
+      "step": 9698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5327295064926147,
+      "learning_rate": 0.0005635964484811753,
+      "loss": 3.0668,
+      "step": 9699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5342981815338135,
+      "learning_rate": 0.000563589097107005,
+      "loss": 3.271,
+      "step": 9700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5108175277709961,
+      "learning_rate": 0.0005635817450385918,
+      "loss": 3.4472,
+      "step": 9701
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5050833225250244,
+      "learning_rate": 0.0005635743922759548,
+      "loss": 3.4441,
+      "step": 9702
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5103448033332825,
+      "learning_rate": 0.0005635670388191137,
+      "loss": 3.4166,
+      "step": 9703
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5414007902145386,
+      "learning_rate": 0.0005635596846680878,
+      "loss": 3.2219,
+      "step": 9704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5395253300666809,
+      "learning_rate": 0.0005635523298228964,
+      "loss": 3.3887,
+      "step": 9705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5145513415336609,
+      "learning_rate": 0.0005635449742835588,
+      "loss": 3.3582,
+      "step": 9706
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5107079148292542,
+      "learning_rate": 0.0005635376180500945,
+      "loss": 3.2202,
+      "step": 9707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.496855229139328,
+      "learning_rate": 0.0005635302611225228,
+      "loss": 3.3866,
+      "step": 9708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5205982327461243,
+      "learning_rate": 0.0005635229035008632,
+      "loss": 3.0983,
+      "step": 9709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5210973024368286,
+      "learning_rate": 0.0005635155451851349,
+      "loss": 3.2638,
+      "step": 9710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5027773380279541,
+      "learning_rate": 0.0005635081861753575,
+      "loss": 3.3372,
+      "step": 9711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5695503950119019,
+      "learning_rate": 0.0005635008264715501,
+      "loss": 3.3357,
+      "step": 9712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5400940179824829,
+      "learning_rate": 0.0005634934660737323,
+      "loss": 3.0366,
+      "step": 9713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6115588545799255,
+      "learning_rate": 0.0005634861049819234,
+      "loss": 3.3567,
+      "step": 9714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5351577401161194,
+      "learning_rate": 0.0005634787431961428,
+      "loss": 3.1951,
+      "step": 9715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5367692112922668,
+      "learning_rate": 0.00056347138071641,
+      "loss": 3.216,
+      "step": 9716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5334668159484863,
+      "learning_rate": 0.0005634640175427441,
+      "loss": 3.4465,
+      "step": 9717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5270911455154419,
+      "learning_rate": 0.0005634566536751648,
+      "loss": 3.2577,
+      "step": 9718
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.547909677028656,
+      "learning_rate": 0.0005634492891136914,
+      "loss": 3.154,
+      "step": 9719
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5077096819877625,
+      "learning_rate": 0.0005634419238583433,
+      "loss": 3.2528,
+      "step": 9720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5063366889953613,
+      "learning_rate": 0.0005634345579091398,
+      "loss": 3.2683,
+      "step": 9721
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5194924473762512,
+      "learning_rate": 0.0005634271912661003,
+      "loss": 3.1052,
+      "step": 9722
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4994848966598511,
+      "learning_rate": 0.0005634198239292444,
+      "loss": 3.3142,
+      "step": 9723
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5550405383110046,
+      "learning_rate": 0.0005634124558985913,
+      "loss": 3.217,
+      "step": 9724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5168113708496094,
+      "learning_rate": 0.0005634050871741606,
+      "loss": 3.0289,
+      "step": 9725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5467432737350464,
+      "learning_rate": 0.0005633977177559715,
+      "loss": 3.2294,
+      "step": 9726
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5153245329856873,
+      "learning_rate": 0.0005633903476440434,
+      "loss": 3.3251,
+      "step": 9727
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5289653539657593,
+      "learning_rate": 0.0005633829768383961,
+      "loss": 3.0288,
+      "step": 9728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5196574330329895,
+      "learning_rate": 0.0005633756053390485,
+      "loss": 3.0948,
+      "step": 9729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5112874507904053,
+      "learning_rate": 0.0005633682331460204,
+      "loss": 3.3054,
+      "step": 9730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5316709876060486,
+      "learning_rate": 0.000563360860259331,
+      "loss": 3.2422,
+      "step": 9731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5173203349113464,
+      "learning_rate": 0.0005633534866789997,
+      "loss": 3.3232,
+      "step": 9732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5282646417617798,
+      "learning_rate": 0.000563346112405046,
+      "loss": 3.2662,
+      "step": 9733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.598261296749115,
+      "learning_rate": 0.0005633387374374894,
+      "loss": 2.9985,
+      "step": 9734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4993882179260254,
+      "learning_rate": 0.0005633313617763493,
+      "loss": 3.2471,
+      "step": 9735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5630433559417725,
+      "learning_rate": 0.0005633239854216449,
+      "loss": 3.095,
+      "step": 9736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47922441363334656,
+      "learning_rate": 0.0005633166083733959,
+      "loss": 3.1201,
+      "step": 9737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5252969264984131,
+      "learning_rate": 0.0005633092306316216,
+      "loss": 3.2614,
+      "step": 9738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.513140857219696,
+      "learning_rate": 0.0005633018521963415,
+      "loss": 3.1924,
+      "step": 9739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6242732405662537,
+      "learning_rate": 0.0005632944730675749,
+      "loss": 3.2182,
+      "step": 9740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5124731659889221,
+      "learning_rate": 0.0005632870932453415,
+      "loss": 3.3061,
+      "step": 9741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49745190143585205,
+      "learning_rate": 0.0005632797127296605,
+      "loss": 3.3783,
+      "step": 9742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5272019505500793,
+      "learning_rate": 0.0005632723315205513,
+      "loss": 3.2325,
+      "step": 9743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5917782187461853,
+      "learning_rate": 0.0005632649496180336,
+      "loss": 3.1377,
+      "step": 9744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5247296094894409,
+      "learning_rate": 0.0005632575670221266,
+      "loss": 3.1735,
+      "step": 9745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546837329864502,
+      "learning_rate": 0.0005632501837328498,
+      "loss": 3.6283,
+      "step": 9746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5424903035163879,
+      "learning_rate": 0.0005632427997502227,
+      "loss": 3.3488,
+      "step": 9747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5044673681259155,
+      "learning_rate": 0.0005632354150742648,
+      "loss": 3.2579,
+      "step": 9748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5371702909469604,
+      "learning_rate": 0.0005632280297049954,
+      "loss": 3.1453,
+      "step": 9749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5304502844810486,
+      "learning_rate": 0.0005632206436424342,
+      "loss": 3.2807,
+      "step": 9750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6061875820159912,
+      "learning_rate": 0.0005632132568866002,
+      "loss": 2.9629,
+      "step": 9751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5296114683151245,
+      "learning_rate": 0.0005632058694375134,
+      "loss": 3.2677,
+      "step": 9752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5466916561126709,
+      "learning_rate": 0.0005631984812951928,
+      "loss": 3.4509,
+      "step": 9753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5157197117805481,
+      "learning_rate": 0.0005631910924596582,
+      "loss": 3.1786,
+      "step": 9754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5921932458877563,
+      "learning_rate": 0.0005631837029309288,
+      "loss": 3.1227,
+      "step": 9755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4948883652687073,
+      "learning_rate": 0.0005631763127090242,
+      "loss": 3.3028,
+      "step": 9756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.543792724609375,
+      "learning_rate": 0.0005631689217939639,
+      "loss": 3.1953,
+      "step": 9757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4762703478336334,
+      "learning_rate": 0.0005631615301857673,
+      "loss": 3.0992,
+      "step": 9758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5200537443161011,
+      "learning_rate": 0.0005631541378844538,
+      "loss": 3.2991,
+      "step": 9759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5312259793281555,
+      "learning_rate": 0.000563146744890043,
+      "loss": 3.2549,
+      "step": 9760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.547489583492279,
+      "learning_rate": 0.0005631393512025544,
+      "loss": 3.392,
+      "step": 9761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5420605540275574,
+      "learning_rate": 0.0005631319568220072,
+      "loss": 3.5638,
+      "step": 9762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5814469456672668,
+      "learning_rate": 0.0005631245617484211,
+      "loss": 3.1061,
+      "step": 9763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48665496706962585,
+      "learning_rate": 0.0005631171659818158,
+      "loss": 3.3398,
+      "step": 9764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.506272554397583,
+      "learning_rate": 0.0005631097695222103,
+      "loss": 3.3197,
+      "step": 9765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49559804797172546,
+      "learning_rate": 0.0005631023723696243,
+      "loss": 3.3218,
+      "step": 9766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5244449377059937,
+      "learning_rate": 0.0005630949745240773,
+      "loss": 3.3369,
+      "step": 9767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5261397957801819,
+      "learning_rate": 0.0005630875759855889,
+      "loss": 3.3897,
+      "step": 9768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5492829084396362,
+      "learning_rate": 0.0005630801767541782,
+      "loss": 3.1185,
+      "step": 9769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5077092051506042,
+      "learning_rate": 0.0005630727768298652,
+      "loss": 3.4211,
+      "step": 9770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5366680026054382,
+      "learning_rate": 0.000563065376212669,
+      "loss": 3.2115,
+      "step": 9771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5005910992622375,
+      "learning_rate": 0.0005630579749026093,
+      "loss": 3.2202,
+      "step": 9772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5190508365631104,
+      "learning_rate": 0.0005630505728997055,
+      "loss": 3.3077,
+      "step": 9773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5151124000549316,
+      "learning_rate": 0.000563043170203977,
+      "loss": 3.2981,
+      "step": 9774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5408289432525635,
+      "learning_rate": 0.0005630357668154435,
+      "loss": 3.288,
+      "step": 9775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5253680944442749,
+      "learning_rate": 0.0005630283627341245,
+      "loss": 3.4114,
+      "step": 9776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5189976096153259,
+      "learning_rate": 0.0005630209579600393,
+      "loss": 3.1448,
+      "step": 9777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5061872601509094,
+      "learning_rate": 0.0005630135524932076,
+      "loss": 3.1726,
+      "step": 9778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5295987725257874,
+      "learning_rate": 0.0005630061463336488,
+      "loss": 3.1084,
+      "step": 9779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5112553238868713,
+      "learning_rate": 0.0005629987394813824,
+      "loss": 3.4138,
+      "step": 9780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5010356903076172,
+      "learning_rate": 0.0005629913319364279,
+      "loss": 3.0226,
+      "step": 9781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.497075617313385,
+      "learning_rate": 0.0005629839236988048,
+      "loss": 3.3942,
+      "step": 9782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5749728679656982,
+      "learning_rate": 0.0005629765147685328,
+      "loss": 3.2406,
+      "step": 9783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5109654664993286,
+      "learning_rate": 0.0005629691051456312,
+      "loss": 3.4464,
+      "step": 9784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4870873689651489,
+      "learning_rate": 0.0005629616948301196,
+      "loss": 3.3477,
+      "step": 9785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5036649107933044,
+      "learning_rate": 0.0005629542838220175,
+      "loss": 3.3398,
+      "step": 9786
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4647713303565979,
+      "learning_rate": 0.0005629468721213444,
+      "loss": 3.4662,
+      "step": 9787
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5171986818313599,
+      "learning_rate": 0.0005629394597281199,
+      "loss": 3.3284,
+      "step": 9788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5243804454803467,
+      "learning_rate": 0.0005629320466423634,
+      "loss": 3.3182,
+      "step": 9789
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5532283186912537,
+      "learning_rate": 0.0005629246328640945,
+      "loss": 3.3757,
+      "step": 9790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5032833814620972,
+      "learning_rate": 0.0005629172183933327,
+      "loss": 3.3534,
+      "step": 9791
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5246580839157104,
+      "learning_rate": 0.0005629098032300978,
+      "loss": 3.408,
+      "step": 9792
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5527105927467346,
+      "learning_rate": 0.0005629023873744087,
+      "loss": 3.2838,
+      "step": 9793
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5229583978652954,
+      "learning_rate": 0.0005628949708262856,
+      "loss": 3.3617,
+      "step": 9794
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5168598294258118,
+      "learning_rate": 0.0005628875535857476,
+      "loss": 3.4933,
+      "step": 9795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4852995276451111,
+      "learning_rate": 0.0005628801356528144,
+      "loss": 3.2868,
+      "step": 9796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5106096267700195,
+      "learning_rate": 0.0005628727170275055,
+      "loss": 3.188,
+      "step": 9797
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5226311683654785,
+      "learning_rate": 0.0005628652977098405,
+      "loss": 3.4544,
+      "step": 9798
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5285156965255737,
+      "learning_rate": 0.0005628578776998389,
+      "loss": 2.9966,
+      "step": 9799
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5192917585372925,
+      "learning_rate": 0.0005628504569975201,
+      "loss": 3.5259,
+      "step": 9800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5123999714851379,
+      "learning_rate": 0.0005628430356029039,
+      "loss": 3.1751,
+      "step": 9801
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.484468013048172,
+      "learning_rate": 0.0005628356135160097,
+      "loss": 3.3334,
+      "step": 9802
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5368399620056152,
+      "learning_rate": 0.0005628281907368571,
+      "loss": 3.4312,
+      "step": 9803
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5847996473312378,
+      "learning_rate": 0.0005628207672654656,
+      "loss": 3.4378,
+      "step": 9804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.487942099571228,
+      "learning_rate": 0.0005628133431018548,
+      "loss": 3.1906,
+      "step": 9805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4921421706676483,
+      "learning_rate": 0.0005628059182460442,
+      "loss": 3.1767,
+      "step": 9806
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5178766250610352,
+      "learning_rate": 0.0005627984926980535,
+      "loss": 3.2761,
+      "step": 9807
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5124492049217224,
+      "learning_rate": 0.000562791066457902,
+      "loss": 3.4485,
+      "step": 9808
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5111625790596008,
+      "learning_rate": 0.0005627836395256095,
+      "loss": 3.3804,
+      "step": 9809
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5112327337265015,
+      "learning_rate": 0.0005627762119011955,
+      "loss": 3.3067,
+      "step": 9810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5200767517089844,
+      "learning_rate": 0.0005627687835846794,
+      "loss": 3.2413,
+      "step": 9811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5200941562652588,
+      "learning_rate": 0.000562761354576081,
+      "loss": 3.3482,
+      "step": 9812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5365529656410217,
+      "learning_rate": 0.0005627539248754196,
+      "loss": 3.281,
+      "step": 9813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5436131954193115,
+      "learning_rate": 0.0005627464944827151,
+      "loss": 3.3693,
+      "step": 9814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.545846164226532,
+      "learning_rate": 0.0005627390633979869,
+      "loss": 3.3813,
+      "step": 9815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5612288117408752,
+      "learning_rate": 0.0005627316316212544,
+      "loss": 3.1317,
+      "step": 9816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5418693423271179,
+      "learning_rate": 0.0005627241991525376,
+      "loss": 3.3768,
+      "step": 9817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5217850804328918,
+      "learning_rate": 0.0005627167659918557,
+      "loss": 3.2675,
+      "step": 9818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5639944076538086,
+      "learning_rate": 0.0005627093321392283,
+      "loss": 3.313,
+      "step": 9819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5243082642555237,
+      "learning_rate": 0.0005627018975946752,
+      "loss": 3.4244,
+      "step": 9820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4913342595100403,
+      "learning_rate": 0.0005626944623582158,
+      "loss": 3.3463,
+      "step": 9821
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5058574080467224,
+      "learning_rate": 0.0005626870264298698,
+      "loss": 3.402,
+      "step": 9822
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5007573962211609,
+      "learning_rate": 0.0005626795898096568,
+      "loss": 3.3769,
+      "step": 9823
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5387548804283142,
+      "learning_rate": 0.0005626721524975962,
+      "loss": 3.2397,
+      "step": 9824
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5633268356323242,
+      "learning_rate": 0.0005626647144937076,
+      "loss": 3.4537,
+      "step": 9825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48922309279441833,
+      "learning_rate": 0.0005626572757980109,
+      "loss": 3.2293,
+      "step": 9826
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5135512948036194,
+      "learning_rate": 0.0005626498364105254,
+      "loss": 3.2582,
+      "step": 9827
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5401769280433655,
+      "learning_rate": 0.0005626423963312707,
+      "loss": 3.4899,
+      "step": 9828
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5234643816947937,
+      "learning_rate": 0.0005626349555602666,
+      "loss": 3.3206,
+      "step": 9829
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5322685837745667,
+      "learning_rate": 0.0005626275140975326,
+      "loss": 3.2296,
+      "step": 9830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5047280192375183,
+      "learning_rate": 0.0005626200719430881,
+      "loss": 3.2615,
+      "step": 9831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5207504630088806,
+      "learning_rate": 0.0005626126290969529,
+      "loss": 3.2849,
+      "step": 9832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4990766942501068,
+      "learning_rate": 0.0005626051855591467,
+      "loss": 3.21,
+      "step": 9833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5085784792900085,
+      "learning_rate": 0.0005625977413296889,
+      "loss": 3.4235,
+      "step": 9834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5034537315368652,
+      "learning_rate": 0.0005625902964085992,
+      "loss": 3.2626,
+      "step": 9835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6476481556892395,
+      "learning_rate": 0.0005625828507958973,
+      "loss": 3.1521,
+      "step": 9836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5102798938751221,
+      "learning_rate": 0.0005625754044916025,
+      "loss": 3.2906,
+      "step": 9837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.510310173034668,
+      "learning_rate": 0.0005625679574957349,
+      "loss": 2.9805,
+      "step": 9838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5330062508583069,
+      "learning_rate": 0.0005625605098083135,
+      "loss": 3.287,
+      "step": 9839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5602695941925049,
+      "learning_rate": 0.0005625530614293584,
+      "loss": 3.1953,
+      "step": 9840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5196672677993774,
+      "learning_rate": 0.0005625456123588892,
+      "loss": 3.2402,
+      "step": 9841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6540876626968384,
+      "learning_rate": 0.0005625381625969252,
+      "loss": 3.3599,
+      "step": 9842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5737609267234802,
+      "learning_rate": 0.0005625307121434862,
+      "loss": 3.2897,
+      "step": 9843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5284690260887146,
+      "learning_rate": 0.0005625232609985919,
+      "loss": 3.2414,
+      "step": 9844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5034304261207581,
+      "learning_rate": 0.0005625158091622619,
+      "loss": 3.2033,
+      "step": 9845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49428412318229675,
+      "learning_rate": 0.0005625083566345158,
+      "loss": 3.1545,
+      "step": 9846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5276312828063965,
+      "learning_rate": 0.0005625009034153732,
+      "loss": 3.1025,
+      "step": 9847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5318641066551208,
+      "learning_rate": 0.0005624934495048535,
+      "loss": 3.2542,
+      "step": 9848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5393041372299194,
+      "learning_rate": 0.0005624859949029768,
+      "loss": 3.2845,
+      "step": 9849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5345759391784668,
+      "learning_rate": 0.0005624785396097625,
+      "loss": 3.2756,
+      "step": 9850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5293691754341125,
+      "learning_rate": 0.0005624710836252302,
+      "loss": 3.2409,
+      "step": 9851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5240424275398254,
+      "learning_rate": 0.0005624636269493995,
+      "loss": 3.1066,
+      "step": 9852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5183370113372803,
+      "learning_rate": 0.0005624561695822903,
+      "loss": 3.2444,
+      "step": 9853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5233088731765747,
+      "learning_rate": 0.0005624487115239219,
+      "loss": 3.0997,
+      "step": 9854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5566701889038086,
+      "learning_rate": 0.0005624412527743142,
+      "loss": 3.6284,
+      "step": 9855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5460590720176697,
+      "learning_rate": 0.0005624337933334867,
+      "loss": 3.2746,
+      "step": 9856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5318386554718018,
+      "learning_rate": 0.0005624263332014591,
+      "loss": 3.38,
+      "step": 9857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5089846253395081,
+      "learning_rate": 0.0005624188723782511,
+      "loss": 3.2773,
+      "step": 9858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5393161177635193,
+      "learning_rate": 0.0005624114108638822,
+      "loss": 2.969,
+      "step": 9859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5289279818534851,
+      "learning_rate": 0.0005624039486583721,
+      "loss": 3.1732,
+      "step": 9860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5223667621612549,
+      "learning_rate": 0.0005623964857617407,
+      "loss": 3.0991,
+      "step": 9861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5194090604782104,
+      "learning_rate": 0.0005623890221740074,
+      "loss": 3.1106,
+      "step": 9862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5405828952789307,
+      "learning_rate": 0.0005623815578951918,
+      "loss": 2.9179,
+      "step": 9863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49917301535606384,
+      "learning_rate": 0.0005623740929253136,
+      "loss": 3.1937,
+      "step": 9864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5121216773986816,
+      "learning_rate": 0.0005623666272643927,
+      "loss": 3.483,
+      "step": 9865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5729875564575195,
+      "learning_rate": 0.0005623591609124486,
+      "loss": 3.2528,
+      "step": 9866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47562557458877563,
+      "learning_rate": 0.0005623516938695009,
+      "loss": 3.1612,
+      "step": 9867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5255392789840698,
+      "learning_rate": 0.0005623442261355694,
+      "loss": 3.2327,
+      "step": 9868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5365827083587646,
+      "learning_rate": 0.0005623367577106736,
+      "loss": 3.2066,
+      "step": 9869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5231798887252808,
+      "learning_rate": 0.0005623292885948333,
+      "loss": 3.2322,
+      "step": 9870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5149793028831482,
+      "learning_rate": 0.0005623218187880682,
+      "loss": 3.4431,
+      "step": 9871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5159502625465393,
+      "learning_rate": 0.0005623143482903979,
+      "loss": 3.103,
+      "step": 9872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.523520290851593,
+      "learning_rate": 0.0005623068771018419,
+      "loss": 3.2667,
+      "step": 9873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5203685760498047,
+      "learning_rate": 0.0005622994052224203,
+      "loss": 3.3393,
+      "step": 9874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5236057639122009,
+      "learning_rate": 0.0005622919326521525,
+      "loss": 3.3034,
+      "step": 9875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5350485444068909,
+      "learning_rate": 0.0005622844593910583,
+      "loss": 3.4209,
+      "step": 9876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48950937390327454,
+      "learning_rate": 0.0005622769854391571,
+      "loss": 3.3428,
+      "step": 9877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5686964988708496,
+      "learning_rate": 0.0005622695107964689,
+      "loss": 3.0339,
+      "step": 9878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5592756271362305,
+      "learning_rate": 0.0005622620354630133,
+      "loss": 3.3535,
+      "step": 9879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4988493025302887,
+      "learning_rate": 0.00056225455943881,
+      "loss": 3.3169,
+      "step": 9880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5318952202796936,
+      "learning_rate": 0.0005622470827238786,
+      "loss": 3.171,
+      "step": 9881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5253759622573853,
+      "learning_rate": 0.000562239605318239,
+      "loss": 3.4359,
+      "step": 9882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4880097806453705,
+      "learning_rate": 0.0005622321272219105,
+      "loss": 3.3283,
+      "step": 9883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49652519822120667,
+      "learning_rate": 0.0005622246484349132,
+      "loss": 3.1138,
+      "step": 9884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.554382860660553,
+      "learning_rate": 0.0005622171689572666,
+      "loss": 3.1591,
+      "step": 9885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5065361857414246,
+      "learning_rate": 0.0005622096887889905,
+      "loss": 3.1288,
+      "step": 9886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5004228353500366,
+      "learning_rate": 0.0005622022079301045,
+      "loss": 3.1548,
+      "step": 9887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5251879096031189,
+      "learning_rate": 0.0005621947263806284,
+      "loss": 3.2474,
+      "step": 9888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5283453464508057,
+      "learning_rate": 0.0005621872441405818,
+      "loss": 3.3388,
+      "step": 9889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5356844067573547,
+      "learning_rate": 0.0005621797612099845,
+      "loss": 3.3577,
+      "step": 9890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5235812067985535,
+      "learning_rate": 0.0005621722775888561,
+      "loss": 3.1009,
+      "step": 9891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5591863989830017,
+      "learning_rate": 0.0005621647932772164,
+      "loss": 3.3538,
+      "step": 9892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48452454805374146,
+      "learning_rate": 0.000562157308275085,
+      "loss": 3.429,
+      "step": 9893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546316504478455,
+      "learning_rate": 0.0005621498225824818,
+      "loss": 3.214,
+      "step": 9894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48921647667884827,
+      "learning_rate": 0.0005621423361994264,
+      "loss": 3.2568,
+      "step": 9895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5067411661148071,
+      "learning_rate": 0.0005621348491259386,
+      "loss": 3.5054,
+      "step": 9896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5032419562339783,
+      "learning_rate": 0.000562127361362038,
+      "loss": 3.3825,
+      "step": 9897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48625296354293823,
+      "learning_rate": 0.0005621198729077444,
+      "loss": 3.0214,
+      "step": 9898
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5369133949279785,
+      "learning_rate": 0.0005621123837630776,
+      "loss": 3.1147,
+      "step": 9899
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5231306552886963,
+      "learning_rate": 0.000562104893928057,
+      "loss": 3.3092,
+      "step": 9900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5136045217514038,
+      "learning_rate": 0.0005620974034027026,
+      "loss": 3.3411,
+      "step": 9901
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5047800540924072,
+      "learning_rate": 0.0005620899121870342,
+      "loss": 3.3716,
+      "step": 9902
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5415762662887573,
+      "learning_rate": 0.0005620824202810713,
+      "loss": 3.2862,
+      "step": 9903
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5388754606246948,
+      "learning_rate": 0.0005620749276848339,
+      "loss": 3.3603,
+      "step": 9904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4989524781703949,
+      "learning_rate": 0.0005620674343983415,
+      "loss": 3.2319,
+      "step": 9905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5338003039360046,
+      "learning_rate": 0.0005620599404216138,
+      "loss": 3.2099,
+      "step": 9906
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.483467698097229,
+      "learning_rate": 0.0005620524457546708,
+      "loss": 3.3824,
+      "step": 9907
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5273773670196533,
+      "learning_rate": 0.000562044950397532,
+      "loss": 3.2784,
+      "step": 9908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5121115446090698,
+      "learning_rate": 0.0005620374543502173,
+      "loss": 3.2377,
+      "step": 9909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5008372068405151,
+      "learning_rate": 0.0005620299576127463,
+      "loss": 3.3862,
+      "step": 9910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49955323338508606,
+      "learning_rate": 0.0005620224601851389,
+      "loss": 3.4437,
+      "step": 9911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5031404495239258,
+      "learning_rate": 0.0005620149620674147,
+      "loss": 3.3564,
+      "step": 9912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5203202962875366,
+      "learning_rate": 0.0005620074632595936,
+      "loss": 3.1263,
+      "step": 9913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48752161860466003,
+      "learning_rate": 0.0005619999637616951,
+      "loss": 3.476,
+      "step": 9914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5217284560203552,
+      "learning_rate": 0.0005619924635737393,
+      "loss": 3.3237,
+      "step": 9915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5151270031929016,
+      "learning_rate": 0.0005619849626957457,
+      "loss": 3.2331,
+      "step": 9916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.549656093120575,
+      "learning_rate": 0.0005619774611277342,
+      "loss": 3.2498,
+      "step": 9917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.568009078502655,
+      "learning_rate": 0.0005619699588697243,
+      "loss": 3.3618,
+      "step": 9918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4984912872314453,
+      "learning_rate": 0.0005619624559217361,
+      "loss": 3.3981,
+      "step": 9919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5053259134292603,
+      "learning_rate": 0.0005619549522837891,
+      "loss": 3.5771,
+      "step": 9920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5003093481063843,
+      "learning_rate": 0.0005619474479559033,
+      "loss": 3.3099,
+      "step": 9921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5107256770133972,
+      "learning_rate": 0.0005619399429380983,
+      "loss": 3.2729,
+      "step": 9922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5822942852973938,
+      "learning_rate": 0.0005619324372303938,
+      "loss": 3.1418,
+      "step": 9923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6355964541435242,
+      "learning_rate": 0.0005619249308328098,
+      "loss": 3.2895,
+      "step": 9924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5443882942199707,
+      "learning_rate": 0.0005619174237453658,
+      "loss": 3.3586,
+      "step": 9925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5277007818222046,
+      "learning_rate": 0.0005619099159680818,
+      "loss": 3.2578,
+      "step": 9926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49704602360725403,
+      "learning_rate": 0.0005619024075009775,
+      "loss": 3.2784,
+      "step": 9927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5114836692810059,
+      "learning_rate": 0.0005618948983440727,
+      "loss": 3.0513,
+      "step": 9928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5052823424339294,
+      "learning_rate": 0.0005618873884973871,
+      "loss": 3.4613,
+      "step": 9929
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5057597756385803,
+      "learning_rate": 0.0005618798779609405,
+      "loss": 3.2938,
+      "step": 9930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5410943031311035,
+      "learning_rate": 0.0005618723667347526,
+      "loss": 3.2756,
+      "step": 9931
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5320307612419128,
+      "learning_rate": 0.0005618648548188434,
+      "loss": 3.2589,
+      "step": 9932
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5032922029495239,
+      "learning_rate": 0.0005618573422132327,
+      "loss": 3.308,
+      "step": 9933
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49512460827827454,
+      "learning_rate": 0.0005618498289179399,
+      "loss": 3.2621,
+      "step": 9934
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5111963152885437,
+      "learning_rate": 0.0005618423149329853,
+      "loss": 3.2512,
+      "step": 9935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5099007487297058,
+      "learning_rate": 0.0005618348002583883,
+      "loss": 3.4597,
+      "step": 9936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5154657363891602,
+      "learning_rate": 0.0005618272848941687,
+      "loss": 3.4915,
+      "step": 9937
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5354853272438049,
+      "learning_rate": 0.0005618197688403466,
+      "loss": 3.3357,
+      "step": 9938
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5375204682350159,
+      "learning_rate": 0.0005618122520969416,
+      "loss": 3.2745,
+      "step": 9939
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5439879894256592,
+      "learning_rate": 0.0005618047346639735,
+      "loss": 3.3741,
+      "step": 9940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5084054470062256,
+      "learning_rate": 0.0005617972165414621,
+      "loss": 3.4072,
+      "step": 9941
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5047259330749512,
+      "learning_rate": 0.0005617896977294271,
+      "loss": 3.1825,
+      "step": 9942
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5263615846633911,
+      "learning_rate": 0.0005617821782278886,
+      "loss": 3.3154,
+      "step": 9943
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5237258076667786,
+      "learning_rate": 0.0005617746580368661,
+      "loss": 3.3176,
+      "step": 9944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.535914957523346,
+      "learning_rate": 0.0005617671371563797,
+      "loss": 3.2534,
+      "step": 9945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5265949368476868,
+      "learning_rate": 0.0005617596155864489,
+      "loss": 3.3145,
+      "step": 9946
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49210605025291443,
+      "learning_rate": 0.0005617520933270937,
+      "loss": 3.4768,
+      "step": 9947
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5354414582252502,
+      "learning_rate": 0.0005617445703783337,
+      "loss": 3.0916,
+      "step": 9948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.525729775428772,
+      "learning_rate": 0.0005617370467401891,
+      "loss": 3.5307,
+      "step": 9949
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5057360529899597,
+      "learning_rate": 0.0005617295224126794,
+      "loss": 3.1895,
+      "step": 9950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5312747955322266,
+      "learning_rate": 0.0005617219973958244,
+      "loss": 3.3153,
+      "step": 9951
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5353907942771912,
+      "learning_rate": 0.0005617144716896441,
+      "loss": 3.3866,
+      "step": 9952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4996834397315979,
+      "learning_rate": 0.0005617069452941584,
+      "loss": 3.0625,
+      "step": 9953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5687936544418335,
+      "learning_rate": 0.0005616994182093869,
+      "loss": 3.1249,
+      "step": 9954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5160571336746216,
+      "learning_rate": 0.0005616918904353494,
+      "loss": 3.2989,
+      "step": 9955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5176610350608826,
+      "learning_rate": 0.0005616843619720658,
+      "loss": 3.2189,
+      "step": 9956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5297330021858215,
+      "learning_rate": 0.000561676832819556,
+      "loss": 3.2166,
+      "step": 9957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5623044371604919,
+      "learning_rate": 0.00056166930297784,
+      "loss": 3.4027,
+      "step": 9958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5169060230255127,
+      "learning_rate": 0.0005616617724469371,
+      "loss": 3.3397,
+      "step": 9959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5446101427078247,
+      "learning_rate": 0.0005616542412268677,
+      "loss": 3.3629,
+      "step": 9960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5260056853294373,
+      "learning_rate": 0.0005616467093176511,
+      "loss": 3.0683,
+      "step": 9961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5235419869422913,
+      "learning_rate": 0.0005616391767193077,
+      "loss": 3.1902,
+      "step": 9962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.514287531375885,
+      "learning_rate": 0.0005616316434318569,
+      "loss": 3.2299,
+      "step": 9963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5356868505477905,
+      "learning_rate": 0.0005616241094553188,
+      "loss": 3.4199,
+      "step": 9964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5147073864936829,
+      "learning_rate": 0.000561616574789713,
+      "loss": 3.2718,
+      "step": 9965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5343377590179443,
+      "learning_rate": 0.0005616090394350596,
+      "loss": 3.2803,
+      "step": 9966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5024563670158386,
+      "learning_rate": 0.0005616015033913784,
+      "loss": 3.204,
+      "step": 9967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5427229404449463,
+      "learning_rate": 0.0005615939666586891,
+      "loss": 3.395,
+      "step": 9968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5278394222259521,
+      "learning_rate": 0.0005615864292370116,
+      "loss": 3.07,
+      "step": 9969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5011979937553406,
+      "learning_rate": 0.0005615788911263659,
+      "loss": 3.3395,
+      "step": 9970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5226938724517822,
+      "learning_rate": 0.0005615713523267716,
+      "loss": 3.1389,
+      "step": 9971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5529884696006775,
+      "learning_rate": 0.0005615638128382488,
+      "loss": 3.2872,
+      "step": 9972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5180261731147766,
+      "learning_rate": 0.0005615562726608173,
+      "loss": 2.9213,
+      "step": 9973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49697425961494446,
+      "learning_rate": 0.0005615487317944969,
+      "loss": 3.3555,
+      "step": 9974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5160676836967468,
+      "learning_rate": 0.0005615411902393073,
+      "loss": 3.175,
+      "step": 9975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49578121304512024,
+      "learning_rate": 0.0005615336479952687,
+      "loss": 3.1123,
+      "step": 9976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5369651317596436,
+      "learning_rate": 0.0005615261050624007,
+      "loss": 3.2317,
+      "step": 9977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.557094156742096,
+      "learning_rate": 0.0005615185614407234,
+      "loss": 3.2406,
+      "step": 9978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5667169094085693,
+      "learning_rate": 0.0005615110171302565,
+      "loss": 3.2702,
+      "step": 9979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5118554830551147,
+      "learning_rate": 0.0005615034721310199,
+      "loss": 3.2947,
+      "step": 9980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4968426823616028,
+      "learning_rate": 0.0005614959264430335,
+      "loss": 3.251,
+      "step": 9981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5141085386276245,
+      "learning_rate": 0.000561488380066317,
+      "loss": 3.1599,
+      "step": 9982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49446555972099304,
+      "learning_rate": 0.0005614808330008906,
+      "loss": 3.3511,
+      "step": 9983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5145710110664368,
+      "learning_rate": 0.0005614732852467741,
+      "loss": 3.2855,
+      "step": 9984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5225349068641663,
+      "learning_rate": 0.0005614657368039871,
+      "loss": 3.3159,
+      "step": 9985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5085193514823914,
+      "learning_rate": 0.0005614581876725497,
+      "loss": 3.1745,
+      "step": 9986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5428699851036072,
+      "learning_rate": 0.0005614506378524818,
+      "loss": 3.2747,
+      "step": 9987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.495169073343277,
+      "learning_rate": 0.0005614430873438032,
+      "loss": 3.3072,
+      "step": 9988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.558083176612854,
+      "learning_rate": 0.0005614355361465338,
+      "loss": 3.3659,
+      "step": 9989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5748363733291626,
+      "learning_rate": 0.0005614279842606936,
+      "loss": 3.3423,
+      "step": 9990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5683549046516418,
+      "learning_rate": 0.0005614204316863023,
+      "loss": 3.1963,
+      "step": 9991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.576992392539978,
+      "learning_rate": 0.00056141287842338,
+      "loss": 3.1108,
+      "step": 9992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.545221209526062,
+      "learning_rate": 0.0005614053244719464,
+      "loss": 3.2083,
+      "step": 9993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5298340320587158,
+      "learning_rate": 0.0005613977698320215,
+      "loss": 3.231,
+      "step": 9994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5359015464782715,
+      "learning_rate": 0.0005613902145036253,
+      "loss": 3.2543,
+      "step": 9995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.518440842628479,
+      "learning_rate": 0.0005613826584867775,
+      "loss": 3.2654,
+      "step": 9996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5107799768447876,
+      "learning_rate": 0.000561375101781498,
+      "loss": 3.2014,
+      "step": 9997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5043929219245911,
+      "learning_rate": 0.0005613675443878069,
+      "loss": 3.167,
+      "step": 9998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5088926553726196,
+      "learning_rate": 0.0005613599863057238,
+      "loss": 3.1162,
+      "step": 9999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5218754410743713,
+      "learning_rate": 0.000561352427535269,
+      "loss": 3.3379,
+      "step": 10000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.521481990814209,
+      "learning_rate": 0.0005613448680764621,
+      "loss": 3.3661,
+      "step": 10001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5064173936843872,
+      "learning_rate": 0.0005613373079293232,
+      "loss": 3.3448,
+      "step": 10002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5323746204376221,
+      "learning_rate": 0.0005613297470938721,
+      "loss": 3.0224,
+      "step": 10003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5698615312576294,
+      "learning_rate": 0.0005613221855701287,
+      "loss": 3.3162,
+      "step": 10004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5549126267433167,
+      "learning_rate": 0.000561314623358113,
+      "loss": 2.9439,
+      "step": 10005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5148659944534302,
+      "learning_rate": 0.0005613070604578448,
+      "loss": 3.1785,
+      "step": 10006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5279960632324219,
+      "learning_rate": 0.0005612994968693442,
+      "loss": 3.1706,
+      "step": 10007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5391876697540283,
+      "learning_rate": 0.0005612919325926308,
+      "loss": 3.3005,
+      "step": 10008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5231080651283264,
+      "learning_rate": 0.0005612843676277249,
+      "loss": 3.3432,
+      "step": 10009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5668114423751831,
+      "learning_rate": 0.0005612768019746464,
+      "loss": 3.3572,
+      "step": 10010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5466375946998596,
+      "learning_rate": 0.0005612692356334149,
+      "loss": 3.2859,
+      "step": 10011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5328760147094727,
+      "learning_rate": 0.0005612616686040505,
+      "loss": 3.1786,
+      "step": 10012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5143555998802185,
+      "learning_rate": 0.0005612541008865733,
+      "loss": 3.2706,
+      "step": 10013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5146706104278564,
+      "learning_rate": 0.0005612465324810029,
+      "loss": 3.4123,
+      "step": 10014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4874396324157715,
+      "learning_rate": 0.0005612389633873594,
+      "loss": 3.3607,
+      "step": 10015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5126667022705078,
+      "learning_rate": 0.0005612313936056629,
+      "loss": 3.2169,
+      "step": 10016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5450606346130371,
+      "learning_rate": 0.0005612238231359331,
+      "loss": 3.3098,
+      "step": 10017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5393866896629333,
+      "learning_rate": 0.0005612162519781901,
+      "loss": 3.3113,
+      "step": 10018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5736554265022278,
+      "learning_rate": 0.0005612086801324536,
+      "loss": 3.3441,
+      "step": 10019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5407893061637878,
+      "learning_rate": 0.0005612011075987439,
+      "loss": 3.3276,
+      "step": 10020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5435847043991089,
+      "learning_rate": 0.0005611935343770806,
+      "loss": 3.1054,
+      "step": 10021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5143806338310242,
+      "learning_rate": 0.0005611859604674839,
+      "loss": 3.2321,
+      "step": 10022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5631576776504517,
+      "learning_rate": 0.0005611783858699736,
+      "loss": 3.2555,
+      "step": 10023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5366818904876709,
+      "learning_rate": 0.0005611708105845697,
+      "loss": 3.1924,
+      "step": 10024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5211443901062012,
+      "learning_rate": 0.0005611632346112921,
+      "loss": 3.3326,
+      "step": 10025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5157060623168945,
+      "learning_rate": 0.000561155657950161,
+      "loss": 3.3635,
+      "step": 10026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5200645327568054,
+      "learning_rate": 0.0005611480806011959,
+      "loss": 3.3153,
+      "step": 10027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5269324779510498,
+      "learning_rate": 0.0005611405025644171,
+      "loss": 3.2965,
+      "step": 10028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4967326521873474,
+      "learning_rate": 0.0005611329238398446,
+      "loss": 3.1998,
+      "step": 10029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5370751619338989,
+      "learning_rate": 0.0005611253444274981,
+      "loss": 3.2197,
+      "step": 10030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.581263542175293,
+      "learning_rate": 0.0005611177643273977,
+      "loss": 3.2003,
+      "step": 10031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5175623297691345,
+      "learning_rate": 0.0005611101835395634,
+      "loss": 3.2288,
+      "step": 10032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5144805312156677,
+      "learning_rate": 0.0005611026020640151,
+      "loss": 3.2167,
+      "step": 10033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5305891633033752,
+      "learning_rate": 0.0005610950199007728,
+      "loss": 3.3823,
+      "step": 10034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5469315052032471,
+      "learning_rate": 0.0005610874370498566,
+      "loss": 3.0891,
+      "step": 10035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.54054856300354,
+      "learning_rate": 0.0005610798535112862,
+      "loss": 3.3481,
+      "step": 10036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5348857641220093,
+      "learning_rate": 0.0005610722692850817,
+      "loss": 3.3687,
+      "step": 10037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5133238434791565,
+      "learning_rate": 0.0005610646843712631,
+      "loss": 3.3272,
+      "step": 10038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5264869928359985,
+      "learning_rate": 0.0005610570987698504,
+      "loss": 3.1792,
+      "step": 10039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5270009636878967,
+      "learning_rate": 0.0005610495124808635,
+      "loss": 3.1658,
+      "step": 10040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5224774479866028,
+      "learning_rate": 0.0005610419255043225,
+      "loss": 3.1901,
+      "step": 10041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5436100363731384,
+      "learning_rate": 0.0005610343378402473,
+      "loss": 3.2561,
+      "step": 10042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5261750817298889,
+      "learning_rate": 0.0005610267494886578,
+      "loss": 3.3767,
+      "step": 10043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5026256442070007,
+      "learning_rate": 0.0005610191604495741,
+      "loss": 3.0808,
+      "step": 10044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5132651329040527,
+      "learning_rate": 0.0005610115707230162,
+      "loss": 3.2861,
+      "step": 10045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5228791236877441,
+      "learning_rate": 0.0005610039803090041,
+      "loss": 3.1605,
+      "step": 10046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5274104475975037,
+      "learning_rate": 0.0005609963892075577,
+      "loss": 3.2428,
+      "step": 10047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53509920835495,
+      "learning_rate": 0.0005609887974186969,
+      "loss": 3.4265,
+      "step": 10048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5528433918952942,
+      "learning_rate": 0.000560981204942442,
+      "loss": 3.2642,
+      "step": 10049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5362628102302551,
+      "learning_rate": 0.0005609736117788129,
+      "loss": 3.0798,
+      "step": 10050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5504554510116577,
+      "learning_rate": 0.0005609660179278295,
+      "loss": 3.1879,
+      "step": 10051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5313317179679871,
+      "learning_rate": 0.0005609584233895117,
+      "loss": 3.3346,
+      "step": 10052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5195586085319519,
+      "learning_rate": 0.0005609508281638798,
+      "loss": 3.3072,
+      "step": 10053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5247373580932617,
+      "learning_rate": 0.0005609432322509534,
+      "loss": 3.222,
+      "step": 10054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5279762744903564,
+      "learning_rate": 0.0005609356356507529,
+      "loss": 3.2947,
+      "step": 10055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.509883463382721,
+      "learning_rate": 0.0005609280383632981,
+      "loss": 3.4618,
+      "step": 10056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5072588324546814,
+      "learning_rate": 0.0005609204403886092,
+      "loss": 3.0462,
+      "step": 10057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5225916504859924,
+      "learning_rate": 0.000560912841726706,
+      "loss": 3.3883,
+      "step": 10058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5001477599143982,
+      "learning_rate": 0.0005609052423776085,
+      "loss": 3.3117,
+      "step": 10059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5187198519706726,
+      "learning_rate": 0.0005608976423413369,
+      "loss": 3.2966,
+      "step": 10060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5253342390060425,
+      "learning_rate": 0.0005608900416179111,
+      "loss": 3.3214,
+      "step": 10061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5546568632125854,
+      "learning_rate": 0.0005608824402073512,
+      "loss": 3.4283,
+      "step": 10062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5639400482177734,
+      "learning_rate": 0.0005608748381096771,
+      "loss": 3.3312,
+      "step": 10063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5478060245513916,
+      "learning_rate": 0.0005608672353249089,
+      "loss": 3.4401,
+      "step": 10064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5374152660369873,
+      "learning_rate": 0.0005608596318530665,
+      "loss": 3.4056,
+      "step": 10065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5391674041748047,
+      "learning_rate": 0.0005608520276941701,
+      "loss": 3.2503,
+      "step": 10066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.504303514957428,
+      "learning_rate": 0.0005608444228482398,
+      "loss": 3.4028,
+      "step": 10067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5163530111312866,
+      "learning_rate": 0.0005608368173152953,
+      "loss": 3.2145,
+      "step": 10068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5681028366088867,
+      "learning_rate": 0.0005608292110953569,
+      "loss": 3.2051,
+      "step": 10069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5483840703964233,
+      "learning_rate": 0.0005608216041884446,
+      "loss": 3.1229,
+      "step": 10070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5046508312225342,
+      "learning_rate": 0.0005608139965945783,
+      "loss": 3.1301,
+      "step": 10071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5168729424476624,
+      "learning_rate": 0.0005608063883137782,
+      "loss": 3.4912,
+      "step": 10072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5135037302970886,
+      "learning_rate": 0.0005607987793460642,
+      "loss": 3.4096,
+      "step": 10073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.524517834186554,
+      "learning_rate": 0.0005607911696914565,
+      "loss": 3.1052,
+      "step": 10074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4915313422679901,
+      "learning_rate": 0.0005607835593499749,
+      "loss": 3.2567,
+      "step": 10075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.503527045249939,
+      "learning_rate": 0.0005607759483216398,
+      "loss": 3.0298,
+      "step": 10076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5238627791404724,
+      "learning_rate": 0.0005607683366064709,
+      "loss": 3.1529,
+      "step": 10077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5229913592338562,
+      "learning_rate": 0.0005607607242044884,
+      "loss": 3.3084,
+      "step": 10078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5081816911697388,
+      "learning_rate": 0.0005607531111157123,
+      "loss": 3.2387,
+      "step": 10079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5463985800743103,
+      "learning_rate": 0.0005607454973401627,
+      "loss": 3.1542,
+      "step": 10080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5712737441062927,
+      "learning_rate": 0.0005607378828778598,
+      "loss": 3.3672,
+      "step": 10081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48129820823669434,
+      "learning_rate": 0.0005607302677288233,
+      "loss": 3.5522,
+      "step": 10082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5331260561943054,
+      "learning_rate": 0.0005607226518930735,
+      "loss": 3.2012,
+      "step": 10083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5193986296653748,
+      "learning_rate": 0.0005607150353706305,
+      "loss": 3.3374,
+      "step": 10084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5116517543792725,
+      "learning_rate": 0.0005607074181615141,
+      "loss": 3.2151,
+      "step": 10085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5041857957839966,
+      "learning_rate": 0.0005606998002657447,
+      "loss": 3.2215,
+      "step": 10086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5375064015388489,
+      "learning_rate": 0.0005606921816833422,
+      "loss": 3.3488,
+      "step": 10087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5213835835456848,
+      "learning_rate": 0.0005606845624143266,
+      "loss": 3.346,
+      "step": 10088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5121163725852966,
+      "learning_rate": 0.000560676942458718,
+      "loss": 3.0301,
+      "step": 10089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5614414811134338,
+      "learning_rate": 0.0005606693218165366,
+      "loss": 3.2334,
+      "step": 10090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5325416326522827,
+      "learning_rate": 0.0005606617004878024,
+      "loss": 3.3862,
+      "step": 10091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5521125793457031,
+      "learning_rate": 0.0005606540784725353,
+      "loss": 3.3762,
+      "step": 10092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5565793514251709,
+      "learning_rate": 0.0005606464557707555,
+      "loss": 3.1573,
+      "step": 10093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5189270973205566,
+      "learning_rate": 0.0005606388323824832,
+      "loss": 3.4819,
+      "step": 10094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5117472410202026,
+      "learning_rate": 0.0005606312083077384,
+      "loss": 3.2461,
+      "step": 10095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5243008732795715,
+      "learning_rate": 0.0005606235835465412,
+      "loss": 3.0837,
+      "step": 10096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5307744145393372,
+      "learning_rate": 0.0005606159580989115,
+      "loss": 3.2632,
+      "step": 10097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5260905027389526,
+      "learning_rate": 0.0005606083319648695,
+      "loss": 3.2347,
+      "step": 10098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5161594152450562,
+      "learning_rate": 0.0005606007051444354,
+      "loss": 3.214,
+      "step": 10099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5047897100448608,
+      "learning_rate": 0.0005605930776376293,
+      "loss": 3.1914,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4992629885673523,
+      "learning_rate": 0.000560585449444471,
+      "loss": 3.4533,
+      "step": 10101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4943458139896393,
+      "learning_rate": 0.0005605778205649808,
+      "loss": 3.217,
+      "step": 10102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5585146546363831,
+      "learning_rate": 0.0005605701909991789,
+      "loss": 3.0695,
+      "step": 10103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5517671704292297,
+      "learning_rate": 0.0005605625607470851,
+      "loss": 3.0032,
+      "step": 10104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5213403105735779,
+      "learning_rate": 0.0005605549298087197,
+      "loss": 3.3822,
+      "step": 10105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5177978277206421,
+      "learning_rate": 0.0005605472981841028,
+      "loss": 3.4062,
+      "step": 10106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5317561030387878,
+      "learning_rate": 0.0005605396658732545,
+      "loss": 3.2718,
+      "step": 10107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5072512626647949,
+      "learning_rate": 0.0005605320328761948,
+      "loss": 3.3612,
+      "step": 10108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.519949734210968,
+      "learning_rate": 0.000560524399192944,
+      "loss": 3.4358,
+      "step": 10109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4792541563510895,
+      "learning_rate": 0.0005605167648235218,
+      "loss": 3.2888,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5226483345031738,
+      "learning_rate": 0.0005605091297679488,
+      "loss": 3.1335,
+      "step": 10111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5207281112670898,
+      "learning_rate": 0.0005605014940262449,
+      "loss": 3.3549,
+      "step": 10112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5150095224380493,
+      "learning_rate": 0.0005604938575984301,
+      "loss": 3.0484,
+      "step": 10113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5432024002075195,
+      "learning_rate": 0.0005604862204845245,
+      "loss": 3.1677,
+      "step": 10114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5010020732879639,
+      "learning_rate": 0.0005604785826845486,
+      "loss": 3.2577,
+      "step": 10115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49979767203330994,
+      "learning_rate": 0.0005604709441985222,
+      "loss": 3.3097,
+      "step": 10116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5717746615409851,
+      "learning_rate": 0.0005604633050264652,
+      "loss": 3.1078,
+      "step": 10117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.530775785446167,
+      "learning_rate": 0.0005604556651683982,
+      "loss": 3.0593,
+      "step": 10118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5330826640129089,
+      "learning_rate": 0.000560448024624341,
+      "loss": 3.3647,
+      "step": 10119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4870454668998718,
+      "learning_rate": 0.000560440383394314,
+      "loss": 3.3083,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5176739692687988,
+      "learning_rate": 0.000560432741478337,
+      "loss": 3.3033,
+      "step": 10121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5079305768013,
+      "learning_rate": 0.0005604250988764303,
+      "loss": 3.3905,
+      "step": 10122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5102031826972961,
+      "learning_rate": 0.000560417455588614,
+      "loss": 3.2952,
+      "step": 10123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234565138816833,
+      "learning_rate": 0.0005604098116149083,
+      "loss": 3.4796,
+      "step": 10124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48953551054000854,
+      "learning_rate": 0.0005604021669553332,
+      "loss": 3.135,
+      "step": 10125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5201693773269653,
+      "learning_rate": 0.000560394521609909,
+      "loss": 3.2396,
+      "step": 10126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5162103176116943,
+      "learning_rate": 0.0005603868755786557,
+      "loss": 3.3356,
+      "step": 10127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.535090982913971,
+      "learning_rate": 0.0005603792288615935,
+      "loss": 3.2915,
+      "step": 10128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5197641253471375,
+      "learning_rate": 0.0005603715814587425,
+      "loss": 3.1518,
+      "step": 10129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5615101456642151,
+      "learning_rate": 0.0005603639333701228,
+      "loss": 3.3298,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5671878457069397,
+      "learning_rate": 0.0005603562845957548,
+      "loss": 3.2195,
+      "step": 10131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.654164731502533,
+      "learning_rate": 0.0005603486351356582,
+      "loss": 3.2873,
+      "step": 10132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5772500038146973,
+      "learning_rate": 0.0005603409849898535,
+      "loss": 3.2613,
+      "step": 10133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5550999641418457,
+      "learning_rate": 0.0005603333341583607,
+      "loss": 3.2879,
+      "step": 10134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49901139736175537,
+      "learning_rate": 0.0005603256826412002,
+      "loss": 3.1976,
+      "step": 10135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5248491764068604,
+      "learning_rate": 0.0005603180304383917,
+      "loss": 3.4905,
+      "step": 10136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5227196216583252,
+      "learning_rate": 0.0005603103775499559,
+      "loss": 3.3214,
+      "step": 10137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5142403841018677,
+      "learning_rate": 0.0005603027239759124,
+      "loss": 3.1934,
+      "step": 10138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5276763439178467,
+      "learning_rate": 0.0005602950697162817,
+      "loss": 3.1553,
+      "step": 10139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4911149740219116,
+      "learning_rate": 0.0005602874147710838,
+      "loss": 3.3715,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5246410965919495,
+      "learning_rate": 0.0005602797591403391,
+      "loss": 3.1999,
+      "step": 10141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5695706605911255,
+      "learning_rate": 0.0005602721028240675,
+      "loss": 3.4837,
+      "step": 10142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5213264226913452,
+      "learning_rate": 0.0005602644458222893,
+      "loss": 3.258,
+      "step": 10143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5553370714187622,
+      "learning_rate": 0.0005602567881350248,
+      "loss": 3.3955,
+      "step": 10144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5400976538658142,
+      "learning_rate": 0.0005602491297622939,
+      "loss": 3.2611,
+      "step": 10145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5228300094604492,
+      "learning_rate": 0.0005602414707041168,
+      "loss": 3.2218,
+      "step": 10146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5486548542976379,
+      "learning_rate": 0.0005602338109605137,
+      "loss": 3.2305,
+      "step": 10147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5071641802787781,
+      "learning_rate": 0.0005602261505315049,
+      "loss": 3.2951,
+      "step": 10148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5746123790740967,
+      "learning_rate": 0.0005602184894171106,
+      "loss": 3.3812,
+      "step": 10149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5090572237968445,
+      "learning_rate": 0.0005602108276173507,
+      "loss": 3.2406,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5613428354263306,
+      "learning_rate": 0.0005602031651322457,
+      "loss": 3.248,
+      "step": 10151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49390625953674316,
+      "learning_rate": 0.0005601955019618156,
+      "loss": 3.1618,
+      "step": 10152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5180720686912537,
+      "learning_rate": 0.0005601878381060807,
+      "loss": 3.3309,
+      "step": 10153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4854278266429901,
+      "learning_rate": 0.0005601801735650609,
+      "loss": 3.4998,
+      "step": 10154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.539308488368988,
+      "learning_rate": 0.0005601725083387767,
+      "loss": 3.3767,
+      "step": 10155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5451698899269104,
+      "learning_rate": 0.0005601648424272482,
+      "loss": 3.0673,
+      "step": 10156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5229968428611755,
+      "learning_rate": 0.0005601571758304957,
+      "loss": 3.2044,
+      "step": 10157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5304858088493347,
+      "learning_rate": 0.0005601495085485391,
+      "loss": 3.4131,
+      "step": 10158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5545448064804077,
+      "learning_rate": 0.0005601418405813989,
+      "loss": 3.3089,
+      "step": 10159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47743499279022217,
+      "learning_rate": 0.000560134171929095,
+      "loss": 3.2294,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5139665007591248,
+      "learning_rate": 0.0005601265025916478,
+      "loss": 3.2909,
+      "step": 10161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54677414894104,
+      "learning_rate": 0.0005601188325690774,
+      "loss": 3.0912,
+      "step": 10162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5394816994667053,
+      "learning_rate": 0.0005601111618614041,
+      "loss": 3.1764,
+      "step": 10163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5438968539237976,
+      "learning_rate": 0.0005601034904686482,
+      "loss": 3.1558,
+      "step": 10164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5133037567138672,
+      "learning_rate": 0.0005600958183908296,
+      "loss": 3.2835,
+      "step": 10165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5024428963661194,
+      "learning_rate": 0.0005600881456279687,
+      "loss": 3.1066,
+      "step": 10166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5286892652511597,
+      "learning_rate": 0.0005600804721800857,
+      "loss": 3.1596,
+      "step": 10167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5346733927726746,
+      "learning_rate": 0.0005600727980472009,
+      "loss": 3.4603,
+      "step": 10168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51430344581604,
+      "learning_rate": 0.0005600651232293343,
+      "loss": 3.3469,
+      "step": 10169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4911240041255951,
+      "learning_rate": 0.0005600574477265062,
+      "loss": 3.4604,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5240756273269653,
+      "learning_rate": 0.0005600497715387368,
+      "loss": 3.4024,
+      "step": 10171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5592271685600281,
+      "learning_rate": 0.0005600420946660464,
+      "loss": 2.9897,
+      "step": 10172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49334201216697693,
+      "learning_rate": 0.0005600344171084551,
+      "loss": 3.098,
+      "step": 10173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5433202385902405,
+      "learning_rate": 0.0005600267388659833,
+      "loss": 3.0928,
+      "step": 10174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5227448344230652,
+      "learning_rate": 0.0005600190599386511,
+      "loss": 3.0703,
+      "step": 10175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5543997287750244,
+      "learning_rate": 0.0005600113803264787,
+      "loss": 3.0732,
+      "step": 10176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4942059814929962,
+      "learning_rate": 0.0005600037000294863,
+      "loss": 3.2383,
+      "step": 10177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5727730393409729,
+      "learning_rate": 0.0005599960190476943,
+      "loss": 3.1809,
+      "step": 10178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5260288119316101,
+      "learning_rate": 0.0005599883373811228,
+      "loss": 3.3584,
+      "step": 10179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5285677313804626,
+      "learning_rate": 0.0005599806550297921,
+      "loss": 3.2441,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5523196458816528,
+      "learning_rate": 0.0005599729719937222,
+      "loss": 3.2402,
+      "step": 10181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5142799019813538,
+      "learning_rate": 0.0005599652882729337,
+      "loss": 3.1881,
+      "step": 10182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5352577567100525,
+      "learning_rate": 0.0005599576038674466,
+      "loss": 3.3204,
+      "step": 10183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5273776650428772,
+      "learning_rate": 0.0005599499187772813,
+      "loss": 3.105,
+      "step": 10184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5525439381599426,
+      "learning_rate": 0.0005599422330024578,
+      "loss": 3.2628,
+      "step": 10185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5069778561592102,
+      "learning_rate": 0.0005599345465429965,
+      "loss": 3.0637,
+      "step": 10186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5232580900192261,
+      "learning_rate": 0.0005599268593989177,
+      "loss": 3.1459,
+      "step": 10187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5141770839691162,
+      "learning_rate": 0.0005599191715702416,
+      "loss": 3.2485,
+      "step": 10188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5252525210380554,
+      "learning_rate": 0.0005599114830569884,
+      "loss": 3.5268,
+      "step": 10189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.552005410194397,
+      "learning_rate": 0.0005599037938591782,
+      "loss": 3.3599,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4965636134147644,
+      "learning_rate": 0.0005598961039768316,
+      "loss": 3.1345,
+      "step": 10191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5195544362068176,
+      "learning_rate": 0.0005598884134099687,
+      "loss": 3.2845,
+      "step": 10192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5534344911575317,
+      "learning_rate": 0.0005598807221586097,
+      "loss": 3.2648,
+      "step": 10193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5047644376754761,
+      "learning_rate": 0.0005598730302227749,
+      "loss": 3.363,
+      "step": 10194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5294030904769897,
+      "learning_rate": 0.0005598653376024845,
+      "loss": 3.3243,
+      "step": 10195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4891115725040436,
+      "learning_rate": 0.0005598576442977588,
+      "loss": 3.1471,
+      "step": 10196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4774000942707062,
+      "learning_rate": 0.0005598499503086182,
+      "loss": 3.5423,
+      "step": 10197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5198843479156494,
+      "learning_rate": 0.0005598422556350827,
+      "loss": 3.0888,
+      "step": 10198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5236388444900513,
+      "learning_rate": 0.0005598345602771728,
+      "loss": 3.2707,
+      "step": 10199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5683999061584473,
+      "learning_rate": 0.0005598268642349086,
+      "loss": 3.4979,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5080311298370361,
+      "learning_rate": 0.0005598191675083105,
+      "loss": 3.2974,
+      "step": 10201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5282857418060303,
+      "learning_rate": 0.0005598114700973988,
+      "loss": 3.2676,
+      "step": 10202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5361818671226501,
+      "learning_rate": 0.0005598037720021935,
+      "loss": 3.1697,
+      "step": 10203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5248768925666809,
+      "learning_rate": 0.0005597960732227152,
+      "loss": 3.2677,
+      "step": 10204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5688296556472778,
+      "learning_rate": 0.000559788373758984,
+      "loss": 3.2665,
+      "step": 10205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5076455473899841,
+      "learning_rate": 0.0005597806736110202,
+      "loss": 3.2625,
+      "step": 10206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5241527557373047,
+      "learning_rate": 0.0005597729727788442,
+      "loss": 3.3101,
+      "step": 10207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.561117947101593,
+      "learning_rate": 0.0005597652712624761,
+      "loss": 3.3028,
+      "step": 10208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49607518315315247,
+      "learning_rate": 0.0005597575690619362,
+      "loss": 3.3582,
+      "step": 10209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5208153128623962,
+      "learning_rate": 0.000559749866177245,
+      "loss": 3.2614,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47716373205184937,
+      "learning_rate": 0.0005597421626084225,
+      "loss": 3.0946,
+      "step": 10211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5305400490760803,
+      "learning_rate": 0.0005597344583554893,
+      "loss": 3.3251,
+      "step": 10212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49316057562828064,
+      "learning_rate": 0.0005597267534184654,
+      "loss": 3.4436,
+      "step": 10213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5044198036193848,
+      "learning_rate": 0.0005597190477973712,
+      "loss": 3.4866,
+      "step": 10214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5129861831665039,
+      "learning_rate": 0.0005597113414922271,
+      "loss": 3.1051,
+      "step": 10215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5174407362937927,
+      "learning_rate": 0.0005597036345030532,
+      "loss": 3.2929,
+      "step": 10216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5471522808074951,
+      "learning_rate": 0.00055969592682987,
+      "loss": 3.3473,
+      "step": 10217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5389727354049683,
+      "learning_rate": 0.0005596882184726976,
+      "loss": 3.4021,
+      "step": 10218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5008918642997742,
+      "learning_rate": 0.0005596805094315565,
+      "loss": 3.2045,
+      "step": 10219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5358201265335083,
+      "learning_rate": 0.0005596727997064669,
+      "loss": 3.5891,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5327726602554321,
+      "learning_rate": 0.0005596650892974492,
+      "loss": 3.3117,
+      "step": 10221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.549594521522522,
+      "learning_rate": 0.0005596573782045236,
+      "loss": 3.2777,
+      "step": 10222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4677071273326874,
+      "learning_rate": 0.0005596496664277104,
+      "loss": 3.1865,
+      "step": 10223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5113584995269775,
+      "learning_rate": 0.0005596419539670299,
+      "loss": 3.4198,
+      "step": 10224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5231432914733887,
+      "learning_rate": 0.0005596342408225024,
+      "loss": 3.2867,
+      "step": 10225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5117649435997009,
+      "learning_rate": 0.0005596265269941485,
+      "loss": 3.266,
+      "step": 10226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5720617771148682,
+      "learning_rate": 0.0005596188124819881,
+      "loss": 3.4351,
+      "step": 10227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5223726034164429,
+      "learning_rate": 0.0005596110972860419,
+      "loss": 3.2824,
+      "step": 10228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5558088421821594,
+      "learning_rate": 0.0005596033814063299,
+      "loss": 3.3895,
+      "step": 10229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5451314449310303,
+      "learning_rate": 0.0005595956648428726,
+      "loss": 3.3244,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5551238059997559,
+      "learning_rate": 0.0005595879475956902,
+      "loss": 3.2968,
+      "step": 10231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5394368171691895,
+      "learning_rate": 0.0005595802296648031,
+      "loss": 3.0322,
+      "step": 10232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5566186904907227,
+      "learning_rate": 0.0005595725110502317,
+      "loss": 3.4952,
+      "step": 10233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.508893609046936,
+      "learning_rate": 0.0005595647917519962,
+      "loss": 3.338,
+      "step": 10234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5071843266487122,
+      "learning_rate": 0.0005595570717701172,
+      "loss": 3.3411,
+      "step": 10235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5460813045501709,
+      "learning_rate": 0.0005595493511046146,
+      "loss": 3.3642,
+      "step": 10236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5000070929527283,
+      "learning_rate": 0.000559541629755509,
+      "loss": 3.4009,
+      "step": 10237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5307058691978455,
+      "learning_rate": 0.0005595339077228207,
+      "loss": 3.2637,
+      "step": 10238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5203930735588074,
+      "learning_rate": 0.0005595261850065701,
+      "loss": 3.3635,
+      "step": 10239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5376899242401123,
+      "learning_rate": 0.0005595184616067775,
+      "loss": 3.3816,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.661049485206604,
+      "learning_rate": 0.0005595107375234631,
+      "loss": 3.3454,
+      "step": 10241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.505993664264679,
+      "learning_rate": 0.0005595030127566474,
+      "loss": 3.274,
+      "step": 10242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4986911118030548,
+      "learning_rate": 0.0005594952873063507,
+      "loss": 3.2775,
+      "step": 10243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5380048155784607,
+      "learning_rate": 0.0005594875611725934,
+      "loss": 3.2546,
+      "step": 10244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5448837280273438,
+      "learning_rate": 0.0005594798343553959,
+      "loss": 3.0941,
+      "step": 10245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5074895024299622,
+      "learning_rate": 0.0005594721068547783,
+      "loss": 3.5444,
+      "step": 10246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5164600014686584,
+      "learning_rate": 0.0005594643786707611,
+      "loss": 3.3968,
+      "step": 10247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49650079011917114,
+      "learning_rate": 0.0005594566498033647,
+      "loss": 3.4343,
+      "step": 10248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4930596351623535,
+      "learning_rate": 0.0005594489202526094,
+      "loss": 3.3165,
+      "step": 10249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5038466453552246,
+      "learning_rate": 0.0005594411900185156,
+      "loss": 3.3183,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5192158818244934,
+      "learning_rate": 0.0005594334591011036,
+      "loss": 3.2953,
+      "step": 10251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5114601254463196,
+      "learning_rate": 0.0005594257275003938,
+      "loss": 3.4671,
+      "step": 10252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5503695011138916,
+      "learning_rate": 0.0005594179952164067,
+      "loss": 3.4004,
+      "step": 10253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49898719787597656,
+      "learning_rate": 0.0005594102622491624,
+      "loss": 3.2809,
+      "step": 10254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5273557305335999,
+      "learning_rate": 0.0005594025285986814,
+      "loss": 3.3543,
+      "step": 10255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5539382100105286,
+      "learning_rate": 0.0005593947942649841,
+      "loss": 3.2133,
+      "step": 10256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5285389423370361,
+      "learning_rate": 0.0005593870592480908,
+      "loss": 3.2378,
+      "step": 10257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.504951000213623,
+      "learning_rate": 0.000559379323548022,
+      "loss": 3.3102,
+      "step": 10258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5255095362663269,
+      "learning_rate": 0.0005593715871647978,
+      "loss": 3.2966,
+      "step": 10259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48558107018470764,
+      "learning_rate": 0.0005593638500984389,
+      "loss": 3.4101,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5173567533493042,
+      "learning_rate": 0.0005593561123489654,
+      "loss": 3.4406,
+      "step": 10261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5314114093780518,
+      "learning_rate": 0.0005593483739163979,
+      "loss": 3.244,
+      "step": 10262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.502029538154602,
+      "learning_rate": 0.0005593406348007567,
+      "loss": 3.39,
+      "step": 10263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5283157825469971,
+      "learning_rate": 0.0005593328950020623,
+      "loss": 3.2418,
+      "step": 10264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5121799111366272,
+      "learning_rate": 0.0005593251545203348,
+      "loss": 3.2263,
+      "step": 10265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.517483651638031,
+      "learning_rate": 0.0005593174133555948,
+      "loss": 3.3841,
+      "step": 10266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5155675411224365,
+      "learning_rate": 0.0005593096715078626,
+      "loss": 3.0287,
+      "step": 10267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5659499168395996,
+      "learning_rate": 0.0005593019289771587,
+      "loss": 3.11,
+      "step": 10268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5271885395050049,
+      "learning_rate": 0.0005592941857635034,
+      "loss": 3.3701,
+      "step": 10269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5610504746437073,
+      "learning_rate": 0.0005592864418669171,
+      "loss": 3.2141,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5438143014907837,
+      "learning_rate": 0.0005592786972874203,
+      "loss": 3.1933,
+      "step": 10271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5469740033149719,
+      "learning_rate": 0.0005592709520250331,
+      "loss": 3.3807,
+      "step": 10272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.530945360660553,
+      "learning_rate": 0.0005592632060797762,
+      "loss": 3.3132,
+      "step": 10273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5154049396514893,
+      "learning_rate": 0.00055925545945167,
+      "loss": 3.2693,
+      "step": 10274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5086045861244202,
+      "learning_rate": 0.0005592477121407347,
+      "loss": 3.267,
+      "step": 10275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5159934163093567,
+      "learning_rate": 0.000559239964146991,
+      "loss": 3.1684,
+      "step": 10276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4987911283969879,
+      "learning_rate": 0.000559232215470459,
+      "loss": 3.238,
+      "step": 10277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5027976632118225,
+      "learning_rate": 0.0005592244661111592,
+      "loss": 3.2168,
+      "step": 10278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5337111353874207,
+      "learning_rate": 0.0005592167160691121,
+      "loss": 3.2408,
+      "step": 10279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4724106192588806,
+      "learning_rate": 0.000559208965344338,
+      "loss": 3.201,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5122838616371155,
+      "learning_rate": 0.0005592012139368574,
+      "loss": 3.1215,
+      "step": 10281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5199552774429321,
+      "learning_rate": 0.0005591934618466908,
+      "loss": 3.2236,
+      "step": 10282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.506280243396759,
+      "learning_rate": 0.0005591857090738584,
+      "loss": 3.0693,
+      "step": 10283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5122705101966858,
+      "learning_rate": 0.0005591779556183807,
+      "loss": 3.2787,
+      "step": 10284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47597208619117737,
+      "learning_rate": 0.0005591702014802782,
+      "loss": 3.3259,
+      "step": 10285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5174579620361328,
+      "learning_rate": 0.0005591624466595712,
+      "loss": 3.2819,
+      "step": 10286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5105884075164795,
+      "learning_rate": 0.0005591546911562802,
+      "loss": 3.4321,
+      "step": 10287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5002108812332153,
+      "learning_rate": 0.0005591469349704257,
+      "loss": 3.2793,
+      "step": 10288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5126345157623291,
+      "learning_rate": 0.000559139178102028,
+      "loss": 3.4978,
+      "step": 10289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5179390907287598,
+      "learning_rate": 0.0005591314205511075,
+      "loss": 3.3352,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49281585216522217,
+      "learning_rate": 0.0005591236623176849,
+      "loss": 3.2815,
+      "step": 10291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5143146514892578,
+      "learning_rate": 0.0005591159034017803,
+      "loss": 3.4356,
+      "step": 10292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5127323865890503,
+      "learning_rate": 0.0005591081438034143,
+      "loss": 3.2153,
+      "step": 10293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5019605755805969,
+      "learning_rate": 0.0005591003835226073,
+      "loss": 3.3087,
+      "step": 10294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5273397564888,
+      "learning_rate": 0.0005590926225593798,
+      "loss": 3.2319,
+      "step": 10295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.528471052646637,
+      "learning_rate": 0.0005590848609137521,
+      "loss": 3.3845,
+      "step": 10296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5653498768806458,
+      "learning_rate": 0.0005590770985857448,
+      "loss": 3.2638,
+      "step": 10297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5153144001960754,
+      "learning_rate": 0.0005590693355753784,
+      "loss": 3.3797,
+      "step": 10298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5238440632820129,
+      "learning_rate": 0.0005590615718826731,
+      "loss": 3.3346,
+      "step": 10299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5217507481575012,
+      "learning_rate": 0.0005590538075076494,
+      "loss": 3.0556,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5109837055206299,
+      "learning_rate": 0.000559046042450328,
+      "loss": 3.4186,
+      "step": 10301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5007416009902954,
+      "learning_rate": 0.0005590382767107291,
+      "loss": 3.403,
+      "step": 10302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5148555040359497,
+      "learning_rate": 0.0005590305102888732,
+      "loss": 3.2634,
+      "step": 10303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5685636401176453,
+      "learning_rate": 0.0005590227431847809,
+      "loss": 3.3305,
+      "step": 10304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5556630492210388,
+      "learning_rate": 0.0005590149753984724,
+      "loss": 3.3192,
+      "step": 10305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.545971691608429,
+      "learning_rate": 0.0005590072069299684,
+      "loss": 3.1871,
+      "step": 10306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5093954801559448,
+      "learning_rate": 0.0005589994377792893,
+      "loss": 3.3295,
+      "step": 10307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5161018967628479,
+      "learning_rate": 0.0005589916679464554,
+      "loss": 3.2058,
+      "step": 10308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.527915358543396,
+      "learning_rate": 0.0005589838974314874,
+      "loss": 3.2002,
+      "step": 10309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5070471167564392,
+      "learning_rate": 0.0005589761262344056,
+      "loss": 3.0882,
+      "step": 10310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4909028708934784,
+      "learning_rate": 0.0005589683543552305,
+      "loss": 3.2828,
+      "step": 10311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5422300696372986,
+      "learning_rate": 0.0005589605817939826,
+      "loss": 3.3773,
+      "step": 10312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5772907137870789,
+      "learning_rate": 0.0005589528085506824,
+      "loss": 3.2922,
+      "step": 10313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5355245471000671,
+      "learning_rate": 0.0005589450346253504,
+      "loss": 3.1342,
+      "step": 10314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5125998258590698,
+      "learning_rate": 0.0005589372600180069,
+      "loss": 3.4883,
+      "step": 10315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5136677622795105,
+      "learning_rate": 0.0005589294847286726,
+      "loss": 3.3296,
+      "step": 10316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.514228105545044,
+      "learning_rate": 0.0005589217087573678,
+      "loss": 3.2156,
+      "step": 10317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5587817430496216,
+      "learning_rate": 0.000558913932104113,
+      "loss": 3.2759,
+      "step": 10318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6269665360450745,
+      "learning_rate": 0.0005589061547689288,
+      "loss": 3.2579,
+      "step": 10319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5368645787239075,
+      "learning_rate": 0.0005588983767518356,
+      "loss": 3.3245,
+      "step": 10320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5467808842658997,
+      "learning_rate": 0.0005588905980528539,
+      "loss": 3.272,
+      "step": 10321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5453903079032898,
+      "learning_rate": 0.0005588828186720041,
+      "loss": 3.2743,
+      "step": 10322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5781856179237366,
+      "learning_rate": 0.000558875038609307,
+      "loss": 3.2912,
+      "step": 10323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6446077227592468,
+      "learning_rate": 0.0005588672578647827,
+      "loss": 3.2426,
+      "step": 10324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5255220532417297,
+      "learning_rate": 0.0005588594764384519,
+      "loss": 3.4687,
+      "step": 10325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5455769896507263,
+      "learning_rate": 0.0005588516943303352,
+      "loss": 3.2468,
+      "step": 10326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5492979288101196,
+      "learning_rate": 0.0005588439115404527,
+      "loss": 3.4958,
+      "step": 10327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5501328110694885,
+      "learning_rate": 0.0005588361280688252,
+      "loss": 3.2044,
+      "step": 10328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5918759107589722,
+      "learning_rate": 0.0005588283439154733,
+      "loss": 3.4665,
+      "step": 10329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5502012968063354,
+      "learning_rate": 0.0005588205590804173,
+      "loss": 3.2085,
+      "step": 10330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49773234128952026,
+      "learning_rate": 0.0005588127735636776,
+      "loss": 3.3999,
+      "step": 10331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.542114794254303,
+      "learning_rate": 0.0005588049873652751,
+      "loss": 3.321,
+      "step": 10332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5828530788421631,
+      "learning_rate": 0.00055879720048523,
+      "loss": 3.2922,
+      "step": 10333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5640475153923035,
+      "learning_rate": 0.0005587894129235628,
+      "loss": 3.1788,
+      "step": 10334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5116139650344849,
+      "learning_rate": 0.0005587816246802941,
+      "loss": 3.1212,
+      "step": 10335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4903753399848938,
+      "learning_rate": 0.0005587738357554445,
+      "loss": 3.1048,
+      "step": 10336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5303298234939575,
+      "learning_rate": 0.0005587660461490343,
+      "loss": 3.1958,
+      "step": 10337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5109275579452515,
+      "learning_rate": 0.0005587582558610843,
+      "loss": 3.3446,
+      "step": 10338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5509363412857056,
+      "learning_rate": 0.0005587504648916147,
+      "loss": 3.2626,
+      "step": 10339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.510272204875946,
+      "learning_rate": 0.0005587426732406463,
+      "loss": 3.3572,
+      "step": 10340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5146096348762512,
+      "learning_rate": 0.0005587348809081993,
+      "loss": 3.1036,
+      "step": 10341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48640674352645874,
+      "learning_rate": 0.0005587270878942947,
+      "loss": 3.4293,
+      "step": 10342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5620465278625488,
+      "learning_rate": 0.0005587192941989525,
+      "loss": 3.2765,
+      "step": 10343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5385037064552307,
+      "learning_rate": 0.0005587114998221935,
+      "loss": 3.0954,
+      "step": 10344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.547918975353241,
+      "learning_rate": 0.0005587037047640383,
+      "loss": 3.2205,
+      "step": 10345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5199856758117676,
+      "learning_rate": 0.0005586959090245074,
+      "loss": 3.3202,
+      "step": 10346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49615347385406494,
+      "learning_rate": 0.000558688112603621,
+      "loss": 3.1684,
+      "step": 10347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5756198763847351,
+      "learning_rate": 0.0005586803155014001,
+      "loss": 3.2479,
+      "step": 10348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5044625997543335,
+      "learning_rate": 0.000558672517717865,
+      "loss": 3.2301,
+      "step": 10349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5165237784385681,
+      "learning_rate": 0.0005586647192530362,
+      "loss": 3.1831,
+      "step": 10350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5043236613273621,
+      "learning_rate": 0.0005586569201069344,
+      "loss": 3.2906,
+      "step": 10351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5174095034599304,
+      "learning_rate": 0.00055864912027958,
+      "loss": 3.3525,
+      "step": 10352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.539576530456543,
+      "learning_rate": 0.0005586413197709936,
+      "loss": 3.2209,
+      "step": 10353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49933311343193054,
+      "learning_rate": 0.0005586335185811957,
+      "loss": 3.3751,
+      "step": 10354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48638248443603516,
+      "learning_rate": 0.0005586257167102069,
+      "loss": 3.1653,
+      "step": 10355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5505256652832031,
+      "learning_rate": 0.0005586179141580478,
+      "loss": 2.9672,
+      "step": 10356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5323838591575623,
+      "learning_rate": 0.0005586101109247389,
+      "loss": 3.0727,
+      "step": 10357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5171400904655457,
+      "learning_rate": 0.0005586023070103006,
+      "loss": 3.1937,
+      "step": 10358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5195598006248474,
+      "learning_rate": 0.0005585945024147537,
+      "loss": 3.6333,
+      "step": 10359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49022892117500305,
+      "learning_rate": 0.0005585866971381186,
+      "loss": 3.1178,
+      "step": 10360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5279939770698547,
+      "learning_rate": 0.000558578891180416,
+      "loss": 3.2818,
+      "step": 10361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5240474343299866,
+      "learning_rate": 0.0005585710845416663,
+      "loss": 3.2201,
+      "step": 10362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5336095094680786,
+      "learning_rate": 0.0005585632772218901,
+      "loss": 3.1314,
+      "step": 10363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4804680347442627,
+      "learning_rate": 0.000558555469221108,
+      "loss": 3.27,
+      "step": 10364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5072349309921265,
+      "learning_rate": 0.0005585476605393406,
+      "loss": 3.3179,
+      "step": 10365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6415109634399414,
+      "learning_rate": 0.0005585398511766084,
+      "loss": 3.493,
+      "step": 10366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49252402782440186,
+      "learning_rate": 0.0005585320411329321,
+      "loss": 3.1779,
+      "step": 10367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5037877559661865,
+      "learning_rate": 0.000558524230408332,
+      "loss": 3.2647,
+      "step": 10368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5191027522087097,
+      "learning_rate": 0.0005585164190028287,
+      "loss": 3.3537,
+      "step": 10369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49495670199394226,
+      "learning_rate": 0.0005585086069164432,
+      "loss": 3.3338,
+      "step": 10370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5146889090538025,
+      "learning_rate": 0.0005585007941491957,
+      "loss": 3.2084,
+      "step": 10371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5505595207214355,
+      "learning_rate": 0.0005584929807011068,
+      "loss": 3.2504,
+      "step": 10372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4919992983341217,
+      "learning_rate": 0.0005584851665721972,
+      "loss": 3.2404,
+      "step": 10373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49141108989715576,
+      "learning_rate": 0.0005584773517624873,
+      "loss": 3.3036,
+      "step": 10374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5041980743408203,
+      "learning_rate": 0.0005584695362719979,
+      "loss": 3.4077,
+      "step": 10375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5285553336143494,
+      "learning_rate": 0.0005584617201007494,
+      "loss": 3.3163,
+      "step": 10376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5297923684120178,
+      "learning_rate": 0.0005584539032487626,
+      "loss": 3.0486,
+      "step": 10377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4896916151046753,
+      "learning_rate": 0.0005584460857160578,
+      "loss": 3.5066,
+      "step": 10378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.528682291507721,
+      "learning_rate": 0.0005584382675026558,
+      "loss": 3.2385,
+      "step": 10379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4898264706134796,
+      "learning_rate": 0.0005584304486085772,
+      "loss": 3.1054,
+      "step": 10380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5090500116348267,
+      "learning_rate": 0.0005584226290338425,
+      "loss": 3.3509,
+      "step": 10381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5073500871658325,
+      "learning_rate": 0.0005584148087784724,
+      "loss": 3.526,
+      "step": 10382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5196703672409058,
+      "learning_rate": 0.0005584069878424873,
+      "loss": 3.2417,
+      "step": 10383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4907332956790924,
+      "learning_rate": 0.0005583991662259082,
+      "loss": 3.2636,
+      "step": 10384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47070619463920593,
+      "learning_rate": 0.000558391343928755,
+      "loss": 3.2755,
+      "step": 10385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5050802230834961,
+      "learning_rate": 0.000558383520951049,
+      "loss": 3.3161,
+      "step": 10386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5075258612632751,
+      "learning_rate": 0.0005583756972928106,
+      "loss": 3.1821,
+      "step": 10387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390806198120117,
+      "learning_rate": 0.0005583678729540602,
+      "loss": 3.2685,
+      "step": 10388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5114220380783081,
+      "learning_rate": 0.0005583600479348185,
+      "loss": 3.1876,
+      "step": 10389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5140728950500488,
+      "learning_rate": 0.0005583522222351062,
+      "loss": 3.2796,
+      "step": 10390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5219741463661194,
+      "learning_rate": 0.0005583443958549439,
+      "loss": 3.3553,
+      "step": 10391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5452612042427063,
+      "learning_rate": 0.0005583365687943521,
+      "loss": 3.4334,
+      "step": 10392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6287804841995239,
+      "learning_rate": 0.0005583287410533516,
+      "loss": 3.1848,
+      "step": 10393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5615096092224121,
+      "learning_rate": 0.0005583209126319629,
+      "loss": 3.4768,
+      "step": 10394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5138780474662781,
+      "learning_rate": 0.0005583130835302066,
+      "loss": 3.3886,
+      "step": 10395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5071299076080322,
+      "learning_rate": 0.0005583052537481034,
+      "loss": 3.3302,
+      "step": 10396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5131478905677795,
+      "learning_rate": 0.0005582974232856738,
+      "loss": 3.4358,
+      "step": 10397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5130304098129272,
+      "learning_rate": 0.0005582895921429385,
+      "loss": 3.1993,
+      "step": 10398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5018404722213745,
+      "learning_rate": 0.0005582817603199182,
+      "loss": 3.4178,
+      "step": 10399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5256871581077576,
+      "learning_rate": 0.0005582739278166334,
+      "loss": 3.0488,
+      "step": 10400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5261520743370056,
+      "learning_rate": 0.0005582660946331047,
+      "loss": 3.1773,
+      "step": 10401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5217165946960449,
+      "learning_rate": 0.0005582582607693529,
+      "loss": 3.1666,
+      "step": 10402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5378689169883728,
+      "learning_rate": 0.0005582504262253986,
+      "loss": 3.199,
+      "step": 10403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5194031596183777,
+      "learning_rate": 0.0005582425910012624,
+      "loss": 3.3709,
+      "step": 10404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5214276909828186,
+      "learning_rate": 0.0005582347550969648,
+      "loss": 3.2128,
+      "step": 10405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5072099566459656,
+      "learning_rate": 0.0005582269185125264,
+      "loss": 3.1792,
+      "step": 10406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6110155582427979,
+      "learning_rate": 0.0005582190812479683,
+      "loss": 3.3741,
+      "step": 10407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.517043948173523,
+      "learning_rate": 0.0005582112433033107,
+      "loss": 3.3555,
+      "step": 10408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5081747174263,
+      "learning_rate": 0.0005582034046785745,
+      "loss": 3.4135,
+      "step": 10409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5313897132873535,
+      "learning_rate": 0.0005581955653737801,
+      "loss": 3.2883,
+      "step": 10410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5237902998924255,
+      "learning_rate": 0.0005581877253889482,
+      "loss": 3.3033,
+      "step": 10411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49405860900878906,
+      "learning_rate": 0.0005581798847240997,
+      "loss": 3.0508,
+      "step": 10412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49449262022972107,
+      "learning_rate": 0.000558172043379255,
+      "loss": 3.268,
+      "step": 10413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5103479623794556,
+      "learning_rate": 0.0005581642013544347,
+      "loss": 3.18,
+      "step": 10414
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4870624244213104,
+      "learning_rate": 0.0005581563586496597,
+      "loss": 3.2905,
+      "step": 10415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5494616031646729,
+      "learning_rate": 0.0005581485152649506,
+      "loss": 3.3204,
+      "step": 10416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5629361271858215,
+      "learning_rate": 0.0005581406712003279,
+      "loss": 3.3171,
+      "step": 10417
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48867589235305786,
+      "learning_rate": 0.0005581328264558123,
+      "loss": 3.3654,
+      "step": 10418
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5334605574607849,
+      "learning_rate": 0.0005581249810314245,
+      "loss": 3.0235,
+      "step": 10419
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5167070031166077,
+      "learning_rate": 0.0005581171349271853,
+      "loss": 3.2379,
+      "step": 10420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5437859892845154,
+      "learning_rate": 0.0005581092881431152,
+      "loss": 3.2161,
+      "step": 10421
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4782818853855133,
+      "learning_rate": 0.0005581014406792348,
+      "loss": 3.2091,
+      "step": 10422
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5339621901512146,
+      "learning_rate": 0.0005580935925355652,
+      "loss": 3.1981,
+      "step": 10423
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5722216963768005,
+      "learning_rate": 0.0005580857437121264,
+      "loss": 3.2892,
+      "step": 10424
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5274963974952698,
+      "learning_rate": 0.0005580778942089396,
+      "loss": 3.4175,
+      "step": 10425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5084352493286133,
+      "learning_rate": 0.0005580700440260253,
+      "loss": 3.3634,
+      "step": 10426
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49948909878730774,
+      "learning_rate": 0.000558062193163404,
+      "loss": 3.2063,
+      "step": 10427
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5016568899154663,
+      "learning_rate": 0.0005580543416210967,
+      "loss": 3.0554,
+      "step": 10428
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49527838826179504,
+      "learning_rate": 0.000558046489399124,
+      "loss": 3.0898,
+      "step": 10429
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5031401515007019,
+      "learning_rate": 0.0005580386364975064,
+      "loss": 3.2988,
+      "step": 10430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5411937832832336,
+      "learning_rate": 0.0005580307829162646,
+      "loss": 3.3959,
+      "step": 10431
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48796623945236206,
+      "learning_rate": 0.0005580229286554195,
+      "loss": 3.2403,
+      "step": 10432
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.502121090888977,
+      "learning_rate": 0.0005580150737149916,
+      "loss": 3.2067,
+      "step": 10433
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5142828822135925,
+      "learning_rate": 0.0005580072180950017,
+      "loss": 3.2503,
+      "step": 10434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4943016767501831,
+      "learning_rate": 0.0005579993617954704,
+      "loss": 3.4234,
+      "step": 10435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4971558153629303,
+      "learning_rate": 0.0005579915048164186,
+      "loss": 3.2003,
+      "step": 10436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.508538544178009,
+      "learning_rate": 0.0005579836471578665,
+      "loss": 3.299,
+      "step": 10437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5330371856689453,
+      "learning_rate": 0.0005579757888198353,
+      "loss": 3.2828,
+      "step": 10438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48706308007240295,
+      "learning_rate": 0.0005579679298023456,
+      "loss": 3.369,
+      "step": 10439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.519385039806366,
+      "learning_rate": 0.0005579600701054179,
+      "loss": 3.3164,
+      "step": 10440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.504608154296875,
+      "learning_rate": 0.0005579522097290732,
+      "loss": 3.338,
+      "step": 10441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5118958353996277,
+      "learning_rate": 0.0005579443486733318,
+      "loss": 3.3114,
+      "step": 10442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49348047375679016,
+      "learning_rate": 0.0005579364869382148,
+      "loss": 3.2707,
+      "step": 10443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5011881589889526,
+      "learning_rate": 0.0005579286245237426,
+      "loss": 3.2466,
+      "step": 10444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49209505319595337,
+      "learning_rate": 0.0005579207614299361,
+      "loss": 3.1962,
+      "step": 10445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5165194869041443,
+      "learning_rate": 0.000557912897656816,
+      "loss": 3.2352,
+      "step": 10446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48565101623535156,
+      "learning_rate": 0.0005579050332044029,
+      "loss": 3.4099,
+      "step": 10447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5325493812561035,
+      "learning_rate": 0.0005578971680727177,
+      "loss": 3.5164,
+      "step": 10448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5225394368171692,
+      "learning_rate": 0.0005578893022617808,
+      "loss": 3.5059,
+      "step": 10449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5450068712234497,
+      "learning_rate": 0.0005578814357716132,
+      "loss": 3.3856,
+      "step": 10450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4797419011592865,
+      "learning_rate": 0.0005578735686022355,
+      "loss": 3.2904,
+      "step": 10451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5698765516281128,
+      "learning_rate": 0.0005578657007536685,
+      "loss": 3.274,
+      "step": 10452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5215989947319031,
+      "learning_rate": 0.0005578578322259328,
+      "loss": 3.422,
+      "step": 10453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5165630578994751,
+      "learning_rate": 0.0005578499630190493,
+      "loss": 3.3273,
+      "step": 10454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5286875367164612,
+      "learning_rate": 0.0005578420931330385,
+      "loss": 3.302,
+      "step": 10455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5140888690948486,
+      "learning_rate": 0.0005578342225679213,
+      "loss": 3.1989,
+      "step": 10456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4854901432991028,
+      "learning_rate": 0.0005578263513237185,
+      "loss": 3.417,
+      "step": 10457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5346885323524475,
+      "learning_rate": 0.0005578184794004506,
+      "loss": 3.4183,
+      "step": 10458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5289451479911804,
+      "learning_rate": 0.0005578106067981384,
+      "loss": 3.2724,
+      "step": 10459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.536060631275177,
+      "learning_rate": 0.0005578027335168027,
+      "loss": 3.3203,
+      "step": 10460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5051462650299072,
+      "learning_rate": 0.0005577948595564643,
+      "loss": 3.2259,
+      "step": 10461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.483619749546051,
+      "learning_rate": 0.0005577869849171437,
+      "loss": 3.3036,
+      "step": 10462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5087090730667114,
+      "learning_rate": 0.0005577791095988619,
+      "loss": 3.4634,
+      "step": 10463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5173831582069397,
+      "learning_rate": 0.0005577712336016395,
+      "loss": 3.4216,
+      "step": 10464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.504951536655426,
+      "learning_rate": 0.0005577633569254974,
+      "loss": 3.3021,
+      "step": 10465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5104922652244568,
+      "learning_rate": 0.0005577554795704561,
+      "loss": 3.2196,
+      "step": 10466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4943409264087677,
+      "learning_rate": 0.0005577476015365365,
+      "loss": 3.1075,
+      "step": 10467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5140161514282227,
+      "learning_rate": 0.0005577397228237593,
+      "loss": 3.2358,
+      "step": 10468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5718799829483032,
+      "learning_rate": 0.0005577318434321453,
+      "loss": 3.1215,
+      "step": 10469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5070701241493225,
+      "learning_rate": 0.0005577239633617153,
+      "loss": 3.1489,
+      "step": 10470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5295852422714233,
+      "learning_rate": 0.0005577160826124899,
+      "loss": 3.2577,
+      "step": 10471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5276656150817871,
+      "learning_rate": 0.0005577082011844898,
+      "loss": 3.2257,
+      "step": 10472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6001366376876831,
+      "learning_rate": 0.0005577003190777361,
+      "loss": 3.268,
+      "step": 10473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5033782720565796,
+      "learning_rate": 0.0005576924362922493,
+      "loss": 3.3577,
+      "step": 10474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5481687784194946,
+      "learning_rate": 0.0005576845528280503,
+      "loss": 3.1024,
+      "step": 10475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5187335014343262,
+      "learning_rate": 0.0005576766686851595,
+      "loss": 3.2345,
+      "step": 10476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5212278962135315,
+      "learning_rate": 0.0005576687838635983,
+      "loss": 3.2455,
+      "step": 10477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5137622356414795,
+      "learning_rate": 0.0005576608983633868,
+      "loss": 3.3298,
+      "step": 10478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.544165313243866,
+      "learning_rate": 0.0005576530121845463,
+      "loss": 3.0631,
+      "step": 10479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5154332518577576,
+      "learning_rate": 0.0005576451253270973,
+      "loss": 3.3821,
+      "step": 10480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5267524719238281,
+      "learning_rate": 0.0005576372377910605,
+      "loss": 3.2283,
+      "step": 10481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5340297818183899,
+      "learning_rate": 0.000557629349576457,
+      "loss": 3.1936,
+      "step": 10482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5131736397743225,
+      "learning_rate": 0.0005576214606833073,
+      "loss": 3.2363,
+      "step": 10483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5175976753234863,
+      "learning_rate": 0.0005576135711116322,
+      "loss": 3.3296,
+      "step": 10484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5071966052055359,
+      "learning_rate": 0.0005576056808614526,
+      "loss": 3.3664,
+      "step": 10485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5538497567176819,
+      "learning_rate": 0.0005575977899327892,
+      "loss": 3.1944,
+      "step": 10486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5503365993499756,
+      "learning_rate": 0.0005575898983256627,
+      "loss": 3.2746,
+      "step": 10487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5471153855323792,
+      "learning_rate": 0.000557582006040094,
+      "loss": 3.4136,
+      "step": 10488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5127497911453247,
+      "learning_rate": 0.000557574113076104,
+      "loss": 3.4066,
+      "step": 10489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5495885610580444,
+      "learning_rate": 0.0005575662194337133,
+      "loss": 3.1525,
+      "step": 10490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5071145296096802,
+      "learning_rate": 0.0005575583251129426,
+      "loss": 3.3041,
+      "step": 10491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5305414199829102,
+      "learning_rate": 0.000557550430113813,
+      "loss": 3.3928,
+      "step": 10492
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5200974345207214,
+      "learning_rate": 0.0005575425344363452,
+      "loss": 3.4053,
+      "step": 10493
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5152339339256287,
+      "learning_rate": 0.0005575346380805598,
+      "loss": 3.306,
+      "step": 10494
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5076627135276794,
+      "learning_rate": 0.0005575267410464778,
+      "loss": 3.414,
+      "step": 10495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5367875695228577,
+      "learning_rate": 0.0005575188433341198,
+      "loss": 3.3654,
+      "step": 10496
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5150578618049622,
+      "learning_rate": 0.0005575109449435068,
+      "loss": 3.2771,
+      "step": 10497
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5308785438537598,
+      "learning_rate": 0.0005575030458746595,
+      "loss": 3.4856,
+      "step": 10498
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5408422946929932,
+      "learning_rate": 0.0005574951461275989,
+      "loss": 3.018,
+      "step": 10499
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5197494626045227,
+      "learning_rate": 0.0005574872457023455,
+      "loss": 3.2124,
+      "step": 10500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5191333293914795,
+      "learning_rate": 0.0005574793445989202,
+      "loss": 3.4525,
+      "step": 10501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5601697564125061,
+      "learning_rate": 0.000557471442817344,
+      "loss": 3.3425,
+      "step": 10502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5880317687988281,
+      "learning_rate": 0.0005574635403576374,
+      "loss": 3.363,
+      "step": 10503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.528934121131897,
+      "learning_rate": 0.0005574556372198215,
+      "loss": 3.2255,
+      "step": 10504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5256865620613098,
+      "learning_rate": 0.000557447733403917,
+      "loss": 3.3382,
+      "step": 10505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.534885823726654,
+      "learning_rate": 0.0005574398289099448,
+      "loss": 3.1408,
+      "step": 10506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5375804901123047,
+      "learning_rate": 0.0005574319237379255,
+      "loss": 3.2966,
+      "step": 10507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5356113314628601,
+      "learning_rate": 0.00055742401788788,
+      "loss": 2.8656,
+      "step": 10508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5177241563796997,
+      "learning_rate": 0.0005574161113598293,
+      "loss": 3.3429,
+      "step": 10509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.619787871837616,
+      "learning_rate": 0.0005574082041537941,
+      "loss": 3.3243,
+      "step": 10510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078191161155701,
+      "learning_rate": 0.0005574002962697953,
+      "loss": 3.2323,
+      "step": 10511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49133309721946716,
+      "learning_rate": 0.0005573923877078534,
+      "loss": 3.2236,
+      "step": 10512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5388320088386536,
+      "learning_rate": 0.0005573844784679897,
+      "loss": 3.1184,
+      "step": 10513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5303979516029358,
+      "learning_rate": 0.0005573765685502247,
+      "loss": 3.5108,
+      "step": 10514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5063710808753967,
+      "learning_rate": 0.0005573686579545795,
+      "loss": 3.2074,
+      "step": 10515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.533225417137146,
+      "learning_rate": 0.0005573607466810747,
+      "loss": 2.9851,
+      "step": 10516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48221254348754883,
+      "learning_rate": 0.0005573528347297312,
+      "loss": 3.3277,
+      "step": 10517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4855428636074066,
+      "learning_rate": 0.0005573449221005699,
+      "loss": 3.3781,
+      "step": 10518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5387424826622009,
+      "learning_rate": 0.0005573370087936117,
+      "loss": 3.4274,
+      "step": 10519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5350310802459717,
+      "learning_rate": 0.0005573290948088772,
+      "loss": 3.2629,
+      "step": 10520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5230748653411865,
+      "learning_rate": 0.0005573211801463874,
+      "loss": 3.1636,
+      "step": 10521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5564438104629517,
+      "learning_rate": 0.000557313264806163,
+      "loss": 3.2858,
+      "step": 10522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4955500066280365,
+      "learning_rate": 0.0005573053487882252,
+      "loss": 3.2284,
+      "step": 10523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5798243284225464,
+      "learning_rate": 0.0005572974320925946,
+      "loss": 3.3389,
+      "step": 10524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5217620134353638,
+      "learning_rate": 0.000557289514719292,
+      "loss": 3.2003,
+      "step": 10525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49824416637420654,
+      "learning_rate": 0.0005572815966683385,
+      "loss": 3.1126,
+      "step": 10526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5172311663627625,
+      "learning_rate": 0.0005572736779397546,
+      "loss": 3.2446,
+      "step": 10527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5369932651519775,
+      "learning_rate": 0.0005572657585335614,
+      "loss": 3.3792,
+      "step": 10528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5192260146141052,
+      "learning_rate": 0.0005572578384497797,
+      "loss": 3.4216,
+      "step": 10529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5092270374298096,
+      "learning_rate": 0.0005572499176884305,
+      "loss": 3.4087,
+      "step": 10530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5477033853530884,
+      "learning_rate": 0.0005572419962495344,
+      "loss": 3.2542,
+      "step": 10531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49869391322135925,
+      "learning_rate": 0.0005572340741331124,
+      "loss": 3.5372,
+      "step": 10532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5513463616371155,
+      "learning_rate": 0.0005572261513391854,
+      "loss": 3.2842,
+      "step": 10533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5208010673522949,
+      "learning_rate": 0.0005572182278677741,
+      "loss": 3.2494,
+      "step": 10534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5104653239250183,
+      "learning_rate": 0.0005572103037188996,
+      "loss": 3.4666,
+      "step": 10535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49619725346565247,
+      "learning_rate": 0.0005572023788925827,
+      "loss": 3.3879,
+      "step": 10536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5280769467353821,
+      "learning_rate": 0.000557194453388844,
+      "loss": 3.2656,
+      "step": 10537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5171984434127808,
+      "learning_rate": 0.0005571865272077049,
+      "loss": 3.3996,
+      "step": 10538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4976312816143036,
+      "learning_rate": 0.0005571786003491858,
+      "loss": 3.3502,
+      "step": 10539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5273638963699341,
+      "learning_rate": 0.0005571706728133078,
+      "loss": 3.3182,
+      "step": 10540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5085924863815308,
+      "learning_rate": 0.0005571627446000917,
+      "loss": 3.1397,
+      "step": 10541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5273411870002747,
+      "learning_rate": 0.0005571548157095585,
+      "loss": 3.1939,
+      "step": 10542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5445072054862976,
+      "learning_rate": 0.000557146886141729,
+      "loss": 3.1008,
+      "step": 10543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5404632091522217,
+      "learning_rate": 0.0005571389558966241,
+      "loss": 3.4246,
+      "step": 10544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5002350211143494,
+      "learning_rate": 0.0005571310249742647,
+      "loss": 3.4459,
+      "step": 10545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5209913849830627,
+      "learning_rate": 0.0005571230933746716,
+      "loss": 3.0648,
+      "step": 10546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5363181233406067,
+      "learning_rate": 0.0005571151610978658,
+      "loss": 3.2458,
+      "step": 10547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5626364350318909,
+      "learning_rate": 0.0005571072281438681,
+      "loss": 2.9742,
+      "step": 10548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5303053259849548,
+      "learning_rate": 0.0005570992945126994,
+      "loss": 3.2247,
+      "step": 10549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5701689720153809,
+      "learning_rate": 0.0005570913602043808,
+      "loss": 3.2445,
+      "step": 10550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49203193187713623,
+      "learning_rate": 0.0005570834252189329,
+      "loss": 3.265,
+      "step": 10551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5791632533073425,
+      "learning_rate": 0.0005570754895563767,
+      "loss": 3.2336,
+      "step": 10552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5477674007415771,
+      "learning_rate": 0.0005570675532167333,
+      "loss": 3.1324,
+      "step": 10553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49030929803848267,
+      "learning_rate": 0.0005570596162000233,
+      "loss": 3.076,
+      "step": 10554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5654690861701965,
+      "learning_rate": 0.0005570516785062678,
+      "loss": 3.1474,
+      "step": 10555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5409621000289917,
+      "learning_rate": 0.0005570437401354877,
+      "loss": 3.0283,
+      "step": 10556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4912637174129486,
+      "learning_rate": 0.0005570358010877038,
+      "loss": 3.0622,
+      "step": 10557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5478519797325134,
+      "learning_rate": 0.000557027861362937,
+      "loss": 3.1291,
+      "step": 10558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5594099760055542,
+      "learning_rate": 0.0005570199209612084,
+      "loss": 3.2931,
+      "step": 10559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5263882279396057,
+      "learning_rate": 0.0005570119798825388,
+      "loss": 3.3384,
+      "step": 10560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5052241683006287,
+      "learning_rate": 0.000557004038126949,
+      "loss": 3.4222,
+      "step": 10561
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5351318120956421,
+      "learning_rate": 0.0005569960956944601,
+      "loss": 3.2503,
+      "step": 10562
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5089155435562134,
+      "learning_rate": 0.0005569881525850929,
+      "loss": 3.1974,
+      "step": 10563
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5822809338569641,
+      "learning_rate": 0.0005569802087988684,
+      "loss": 3.2786,
+      "step": 10564
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5087013840675354,
+      "learning_rate": 0.0005569722643358075,
+      "loss": 3.3187,
+      "step": 10565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5044753551483154,
+      "learning_rate": 0.000556964319195931,
+      "loss": 3.4414,
+      "step": 10566
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5216333866119385,
+      "learning_rate": 0.0005569563733792601,
+      "loss": 3.3294,
+      "step": 10567
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5563536882400513,
+      "learning_rate": 0.0005569484268858155,
+      "loss": 2.9519,
+      "step": 10568
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5203812122344971,
+      "learning_rate": 0.000556940479715618,
+      "loss": 3.3726,
+      "step": 10569
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.545968770980835,
+      "learning_rate": 0.000556932531868689,
+      "loss": 3.1397,
+      "step": 10570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5787644386291504,
+      "learning_rate": 0.000556924583345049,
+      "loss": 3.2113,
+      "step": 10571
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5733756422996521,
+      "learning_rate": 0.0005569166341447192,
+      "loss": 3.5598,
+      "step": 10572
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.541483998298645,
+      "learning_rate": 0.0005569086842677203,
+      "loss": 3.2303,
+      "step": 10573
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.549351692199707,
+      "learning_rate": 0.0005569007337140736,
+      "loss": 3.354,
+      "step": 10574
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5365926623344421,
+      "learning_rate": 0.0005568927824837996,
+      "loss": 3.2604,
+      "step": 10575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5326177477836609,
+      "learning_rate": 0.0005568848305769195,
+      "loss": 3.2394,
+      "step": 10576
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.523600697517395,
+      "learning_rate": 0.0005568768779934542,
+      "loss": 3.1083,
+      "step": 10577
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5898261666297913,
+      "learning_rate": 0.0005568689247334247,
+      "loss": 3.1423,
+      "step": 10578
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5717616677284241,
+      "learning_rate": 0.0005568609707968518,
+      "loss": 3.1968,
+      "step": 10579
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49297675490379333,
+      "learning_rate": 0.0005568530161837566,
+      "loss": 3.2594,
+      "step": 10580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.505689263343811,
+      "learning_rate": 0.0005568450608941599,
+      "loss": 3.187,
+      "step": 10581
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5439349412918091,
+      "learning_rate": 0.0005568371049280827,
+      "loss": 3.2561,
+      "step": 10582
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5405676960945129,
+      "learning_rate": 0.0005568291482855462,
+      "loss": 3.353,
+      "step": 10583
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4949883818626404,
+      "learning_rate": 0.000556821190966571,
+      "loss": 3.2305,
+      "step": 10584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5206522941589355,
+      "learning_rate": 0.0005568132329711783,
+      "loss": 3.2696,
+      "step": 10585
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5137814283370972,
+      "learning_rate": 0.0005568052742993889,
+      "loss": 3.1519,
+      "step": 10586
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.577721118927002,
+      "learning_rate": 0.0005567973149512239,
+      "loss": 3.2501,
+      "step": 10587
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4995252192020416,
+      "learning_rate": 0.0005567893549267042,
+      "loss": 3.2882,
+      "step": 10588
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5364012122154236,
+      "learning_rate": 0.0005567813942258506,
+      "loss": 3.3206,
+      "step": 10589
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5193110108375549,
+      "learning_rate": 0.0005567734328486844,
+      "loss": 3.1889,
+      "step": 10590
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.510722815990448,
+      "learning_rate": 0.0005567654707952262,
+      "loss": 3.3038,
+      "step": 10591
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5020973682403564,
+      "learning_rate": 0.0005567575080654974,
+      "loss": 3.25,
+      "step": 10592
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5226955413818359,
+      "learning_rate": 0.0005567495446595187,
+      "loss": 3.3096,
+      "step": 10593
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5112230777740479,
+      "learning_rate": 0.000556741580577311,
+      "loss": 3.2976,
+      "step": 10594
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5107260942459106,
+      "learning_rate": 0.0005567336158188955,
+      "loss": 3.3089,
+      "step": 10595
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5062245726585388,
+      "learning_rate": 0.0005567256503842929,
+      "loss": 3.1591,
+      "step": 10596
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.625127375125885,
+      "learning_rate": 0.0005567176842735244,
+      "loss": 3.3238,
+      "step": 10597
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48830580711364746,
+      "learning_rate": 0.000556709717486611,
+      "loss": 3.2201,
+      "step": 10598
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5055554509162903,
+      "learning_rate": 0.0005567017500235736,
+      "loss": 3.2346,
+      "step": 10599
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.558253824710846,
+      "learning_rate": 0.0005566937818844332,
+      "loss": 3.3789,
+      "step": 10600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5219190120697021,
+      "learning_rate": 0.0005566858130692107,
+      "loss": 3.1847,
+      "step": 10601
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5320719480514526,
+      "learning_rate": 0.0005566778435779272,
+      "loss": 3.3797,
+      "step": 10602
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5657633543014526,
+      "learning_rate": 0.0005566698734106037,
+      "loss": 3.2031,
+      "step": 10603
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6036418080329895,
+      "learning_rate": 0.0005566619025672611,
+      "loss": 3.3975,
+      "step": 10604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5115952491760254,
+      "learning_rate": 0.0005566539310479206,
+      "loss": 3.058,
+      "step": 10605
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5792054533958435,
+      "learning_rate": 0.0005566459588526028,
+      "loss": 3.3311,
+      "step": 10606
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5313816070556641,
+      "learning_rate": 0.000556637985981329,
+      "loss": 3.428,
+      "step": 10607
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5156925320625305,
+      "learning_rate": 0.0005566300124341203,
+      "loss": 3.2343,
+      "step": 10608
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5339927077293396,
+      "learning_rate": 0.0005566220382109974,
+      "loss": 3.0971,
+      "step": 10609
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5031298995018005,
+      "learning_rate": 0.0005566140633119814,
+      "loss": 3.2465,
+      "step": 10610
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5161144137382507,
+      "learning_rate": 0.0005566060877370934,
+      "loss": 3.1803,
+      "step": 10611
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49942952394485474,
+      "learning_rate": 0.0005565981114863544,
+      "loss": 3.3168,
+      "step": 10612
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5302490592002869,
+      "learning_rate": 0.0005565901345597853,
+      "loss": 3.2728,
+      "step": 10613
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49967092275619507,
+      "learning_rate": 0.0005565821569574072,
+      "loss": 3.3966,
+      "step": 10614
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.516776978969574,
+      "learning_rate": 0.000556574178679241,
+      "loss": 3.1585,
+      "step": 10615
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5194513201713562,
+      "learning_rate": 0.0005565661997253079,
+      "loss": 3.3124,
+      "step": 10616
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5056691765785217,
+      "learning_rate": 0.0005565582200956288,
+      "loss": 3.3491,
+      "step": 10617
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5017831921577454,
+      "learning_rate": 0.0005565502397902246,
+      "loss": 3.2153,
+      "step": 10618
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5000698566436768,
+      "learning_rate": 0.0005565422588091165,
+      "loss": 3.4327,
+      "step": 10619
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5435037016868591,
+      "learning_rate": 0.0005565342771523255,
+      "loss": 3.0531,
+      "step": 10620
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5276373624801636,
+      "learning_rate": 0.0005565262948198726,
+      "loss": 3.4252,
+      "step": 10621
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5036376118659973,
+      "learning_rate": 0.0005565183118117787,
+      "loss": 3.1393,
+      "step": 10622
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5199164748191833,
+      "learning_rate": 0.0005565103281280652,
+      "loss": 3.2729,
+      "step": 10623
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5530470013618469,
+      "learning_rate": 0.0005565023437687526,
+      "loss": 3.4069,
+      "step": 10624
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5330382585525513,
+      "learning_rate": 0.0005564943587338622,
+      "loss": 3.2335,
+      "step": 10625
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.530340313911438,
+      "learning_rate": 0.0005564863730234151,
+      "loss": 3.1747,
+      "step": 10626
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5398244261741638,
+      "learning_rate": 0.0005564783866374323,
+      "loss": 3.407,
+      "step": 10627
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4917842447757721,
+      "learning_rate": 0.0005564703995759347,
+      "loss": 3.144,
+      "step": 10628
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5320751667022705,
+      "learning_rate": 0.0005564624118389435,
+      "loss": 3.1363,
+      "step": 10629
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5649660229682922,
+      "learning_rate": 0.0005564544234264797,
+      "loss": 3.283,
+      "step": 10630
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5005009174346924,
+      "learning_rate": 0.0005564464343385642,
+      "loss": 3.2569,
+      "step": 10631
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5064327120780945,
+      "learning_rate": 0.0005564384445752181,
+      "loss": 3.066,
+      "step": 10632
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5261727571487427,
+      "learning_rate": 0.0005564304541364626,
+      "loss": 3.186,
+      "step": 10633
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49180030822753906,
+      "learning_rate": 0.0005564224630223186,
+      "loss": 3.3107,
+      "step": 10634
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5282382369041443,
+      "learning_rate": 0.0005564144712328072,
+      "loss": 3.2852,
+      "step": 10635
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5133998394012451,
+      "learning_rate": 0.0005564064787679494,
+      "loss": 3.2587,
+      "step": 10636
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5127143859863281,
+      "learning_rate": 0.0005563984856277662,
+      "loss": 3.475,
+      "step": 10637
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5018812417984009,
+      "learning_rate": 0.000556390491812279,
+      "loss": 2.9832,
+      "step": 10638
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5890460014343262,
+      "learning_rate": 0.0005563824973215083,
+      "loss": 3.3091,
+      "step": 10639
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5279850363731384,
+      "learning_rate": 0.0005563745021554756,
+      "loss": 3.4503,
+      "step": 10640
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5249155759811401,
+      "learning_rate": 0.0005563665063142018,
+      "loss": 3.0548,
+      "step": 10641
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5137816667556763,
+      "learning_rate": 0.0005563585097977079,
+      "loss": 3.4102,
+      "step": 10642
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5169107913970947,
+      "learning_rate": 0.000556350512606015,
+      "loss": 3.3964,
+      "step": 10643
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5247167944908142,
+      "learning_rate": 0.0005563425147391442,
+      "loss": 3.1934,
+      "step": 10644
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5554961562156677,
+      "learning_rate": 0.0005563345161971165,
+      "loss": 3.6094,
+      "step": 10645
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5095405578613281,
+      "learning_rate": 0.0005563265169799532,
+      "loss": 3.5217,
+      "step": 10646
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5305556058883667,
+      "learning_rate": 0.0005563185170876751,
+      "loss": 3.1786,
+      "step": 10647
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5047399997711182,
+      "learning_rate": 0.0005563105165203034,
+      "loss": 3.1298,
+      "step": 10648
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5853273272514343,
+      "learning_rate": 0.0005563025152778591,
+      "loss": 3.0866,
+      "step": 10649
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5093114972114563,
+      "learning_rate": 0.0005562945133603633,
+      "loss": 3.273,
+      "step": 10650
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4893055558204651,
+      "learning_rate": 0.000556286510767837,
+      "loss": 3.3782,
+      "step": 10651
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6072062849998474,
+      "learning_rate": 0.0005562785075003013,
+      "loss": 3.24,
+      "step": 10652
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49338966608047485,
+      "learning_rate": 0.0005562705035577775,
+      "loss": 3.4271,
+      "step": 10653
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5136591792106628,
+      "learning_rate": 0.0005562624989402864,
+      "loss": 3.2597,
+      "step": 10654
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5117157101631165,
+      "learning_rate": 0.0005562544936478492,
+      "loss": 3.2494,
+      "step": 10655
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4980575442314148,
+      "learning_rate": 0.0005562464876804871,
+      "loss": 3.3992,
+      "step": 10656
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5555065870285034,
+      "learning_rate": 0.000556238481038221,
+      "loss": 3.2717,
+      "step": 10657
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5585774779319763,
+      "learning_rate": 0.000556230473721072,
+      "loss": 3.2388,
+      "step": 10658
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6022112965583801,
+      "learning_rate": 0.0005562224657290613,
+      "loss": 3.0722,
+      "step": 10659
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5127832889556885,
+      "learning_rate": 0.0005562144570622099,
+      "loss": 3.3021,
+      "step": 10660
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5011223554611206,
+      "learning_rate": 0.000556206447720539,
+      "loss": 3.3003,
+      "step": 10661
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5553777813911438,
+      "learning_rate": 0.0005561984377040695,
+      "loss": 3.1519,
+      "step": 10662
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5582361817359924,
+      "learning_rate": 0.0005561904270128227,
+      "loss": 3.1381,
+      "step": 10663
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5032232403755188,
+      "learning_rate": 0.0005561824156468196,
+      "loss": 3.1779,
+      "step": 10664
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.516594409942627,
+      "learning_rate": 0.0005561744036060814,
+      "loss": 3.3612,
+      "step": 10665
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5512834191322327,
+      "learning_rate": 0.0005561663908906291,
+      "loss": 3.2972,
+      "step": 10666
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5587075352668762,
+      "learning_rate": 0.0005561583775004837,
+      "loss": 3.1885,
+      "step": 10667
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6395102739334106,
+      "learning_rate": 0.0005561503634356666,
+      "loss": 3.368,
+      "step": 10668
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.572404146194458,
+      "learning_rate": 0.0005561423486961987,
+      "loss": 3.0949,
+      "step": 10669
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5496918559074402,
+      "learning_rate": 0.000556134333282101,
+      "loss": 3.1419,
+      "step": 10670
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5133854150772095,
+      "learning_rate": 0.0005561263171933949,
+      "loss": 3.3315,
+      "step": 10671
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5696301460266113,
+      "learning_rate": 0.0005561183004301015,
+      "loss": 3.1278,
+      "step": 10672
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5422386527061462,
+      "learning_rate": 0.0005561102829922415,
+      "loss": 3.2289,
+      "step": 10673
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5507988333702087,
+      "learning_rate": 0.0005561022648798364,
+      "loss": 3.2526,
+      "step": 10674
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5414395332336426,
+      "learning_rate": 0.0005560942460929074,
+      "loss": 3.0932,
+      "step": 10675
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.611114501953125,
+      "learning_rate": 0.0005560862266314752,
+      "loss": 3.0416,
+      "step": 10676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5457938313484192,
+      "learning_rate": 0.0005560782064955613,
+      "loss": 3.1892,
+      "step": 10677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5073613524436951,
+      "learning_rate": 0.0005560701856851866,
+      "loss": 3.3985,
+      "step": 10678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5243719220161438,
+      "learning_rate": 0.0005560621642003723,
+      "loss": 3.1556,
+      "step": 10679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5766971707344055,
+      "learning_rate": 0.0005560541420411398,
+      "loss": 3.3581,
+      "step": 10680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5084385871887207,
+      "learning_rate": 0.0005560461192075097,
+      "loss": 3.1972,
+      "step": 10681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5689043402671814,
+      "learning_rate": 0.0005560380956995035,
+      "loss": 3.2245,
+      "step": 10682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.506496787071228,
+      "learning_rate": 0.0005560300715171421,
+      "loss": 3.2089,
+      "step": 10683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5448225736618042,
+      "learning_rate": 0.0005560220466604469,
+      "loss": 3.1088,
+      "step": 10684
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5289160013198853,
+      "learning_rate": 0.0005560140211294389,
+      "loss": 3.3693,
+      "step": 10685
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5228204131126404,
+      "learning_rate": 0.0005560059949241392,
+      "loss": 3.4617,
+      "step": 10686
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49474701285362244,
+      "learning_rate": 0.000555997968044569,
+      "loss": 3.2549,
+      "step": 10687
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5213100910186768,
+      "learning_rate": 0.0005559899404907493,
+      "loss": 3.2791,
+      "step": 10688
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5579019784927368,
+      "learning_rate": 0.0005559819122627016,
+      "loss": 3.0853,
+      "step": 10689
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5246168375015259,
+      "learning_rate": 0.0005559738833604466,
+      "loss": 3.1685,
+      "step": 10690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5287438631057739,
+      "learning_rate": 0.0005559658537840058,
+      "loss": 3.2581,
+      "step": 10691
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.576284646987915,
+      "learning_rate": 0.0005559578235334002,
+      "loss": 3.3451,
+      "step": 10692
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5133945345878601,
+      "learning_rate": 0.0005559497926086508,
+      "loss": 3.2493,
+      "step": 10693
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5034992098808289,
+      "learning_rate": 0.000555941761009779,
+      "loss": 3.3255,
+      "step": 10694
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5439068078994751,
+      "learning_rate": 0.0005559337287368058,
+      "loss": 3.2783,
+      "step": 10695
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5081217288970947,
+      "learning_rate": 0.0005559256957897525,
+      "loss": 3.2928,
+      "step": 10696
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5664759278297424,
+      "learning_rate": 0.0005559176621686402,
+      "loss": 3.2541,
+      "step": 10697
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5075600743293762,
+      "learning_rate": 0.0005559096278734899,
+      "loss": 3.1877,
+      "step": 10698
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5304969549179077,
+      "learning_rate": 0.000555901592904323,
+      "loss": 3.347,
+      "step": 10699
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5134689211845398,
+      "learning_rate": 0.0005558935572611605,
+      "loss": 3.2153,
+      "step": 10700
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5504531264305115,
+      "learning_rate": 0.0005558855209440236,
+      "loss": 3.1492,
+      "step": 10701
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5000514984130859,
+      "learning_rate": 0.0005558774839529335,
+      "loss": 3.472,
+      "step": 10702
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.532200276851654,
+      "learning_rate": 0.0005558694462879113,
+      "loss": 3.1464,
+      "step": 10703
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5475481152534485,
+      "learning_rate": 0.0005558614079489784,
+      "loss": 2.9702,
+      "step": 10704
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5486966967582703,
+      "learning_rate": 0.0005558533689361557,
+      "loss": 3.5168,
+      "step": 10705
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5467349886894226,
+      "learning_rate": 0.0005558453292494644,
+      "loss": 3.4869,
+      "step": 10706
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.720486581325531,
+      "learning_rate": 0.0005558372888889258,
+      "loss": 3.0787,
+      "step": 10707
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5721949338912964,
+      "learning_rate": 0.0005558292478545611,
+      "loss": 3.5098,
+      "step": 10708
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4928334355354309,
+      "learning_rate": 0.0005558212061463912,
+      "loss": 3.5137,
+      "step": 10709
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5080165863037109,
+      "learning_rate": 0.0005558131637644376,
+      "loss": 3.3198,
+      "step": 10710
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5456277132034302,
+      "learning_rate": 0.0005558051207087214,
+      "loss": 3.2869,
+      "step": 10711
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48766282200813293,
+      "learning_rate": 0.0005557970769792636,
+      "loss": 3.1597,
+      "step": 10712
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5692100524902344,
+      "learning_rate": 0.0005557890325760856,
+      "loss": 3.2611,
+      "step": 10713
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5719293355941772,
+      "learning_rate": 0.0005557809874992086,
+      "loss": 3.0606,
+      "step": 10714
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5092300772666931,
+      "learning_rate": 0.0005557729417486536,
+      "loss": 3.1642,
+      "step": 10715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5151323676109314,
+      "learning_rate": 0.0005557648953244419,
+      "loss": 3.2684,
+      "step": 10716
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.48842179775238037,
+      "learning_rate": 0.0005557568482265947,
+      "loss": 3.3302,
+      "step": 10717
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5149058699607849,
+      "learning_rate": 0.0005557488004551332,
+      "loss": 3.3089,
+      "step": 10718
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5239120125770569,
+      "learning_rate": 0.0005557407520100785,
+      "loss": 3.2215,
+      "step": 10719
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5721972584724426,
+      "learning_rate": 0.000555732702891452,
+      "loss": 3.2324,
+      "step": 10720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5425095558166504,
+      "learning_rate": 0.0005557246530992748,
+      "loss": 3.3185,
+      "step": 10721
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5574762225151062,
+      "learning_rate": 0.000555716602633568,
+      "loss": 3.4277,
+      "step": 10722
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5534297823905945,
+      "learning_rate": 0.0005557085514943529,
+      "loss": 3.3156,
+      "step": 10723
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5003551244735718,
+      "learning_rate": 0.0005557004996816507,
+      "loss": 3.3441,
+      "step": 10724
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5311870574951172,
+      "learning_rate": 0.0005556924471954826,
+      "loss": 3.235,
+      "step": 10725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5108062624931335,
+      "learning_rate": 0.0005556843940358698,
+      "loss": 3.1934,
+      "step": 10726
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5077083110809326,
+      "learning_rate": 0.0005556763402028334,
+      "loss": 3.2491,
+      "step": 10727
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5326809883117676,
+      "learning_rate": 0.0005556682856963949,
+      "loss": 3.3736,
+      "step": 10728
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49083372950553894,
+      "learning_rate": 0.0005556602305165752,
+      "loss": 3.0318,
+      "step": 10729
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49490422010421753,
+      "learning_rate": 0.0005556521746633957,
+      "loss": 3.1,
+      "step": 10730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.48100271821022034,
+      "learning_rate": 0.0005556441181368776,
+      "loss": 3.2817,
+      "step": 10731
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5259497761726379,
+      "learning_rate": 0.0005556360609370421,
+      "loss": 3.3625,
+      "step": 10732
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5028544068336487,
+      "learning_rate": 0.0005556280030639103,
+      "loss": 3.2937,
+      "step": 10733
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.628638744354248,
+      "learning_rate": 0.0005556199445175037,
+      "loss": 3.2895,
+      "step": 10734
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.517248809337616,
+      "learning_rate": 0.0005556118852978433,
+      "loss": 3.4441,
+      "step": 10735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5241116285324097,
+      "learning_rate": 0.0005556038254049503,
+      "loss": 3.071,
+      "step": 10736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5196915864944458,
+      "learning_rate": 0.0005555957648388461,
+      "loss": 3.4301,
+      "step": 10737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5392457246780396,
+      "learning_rate": 0.000555587703599552,
+      "loss": 3.0426,
+      "step": 10738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49311745166778564,
+      "learning_rate": 0.0005555796416870888,
+      "loss": 3.3097,
+      "step": 10739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.643158495426178,
+      "learning_rate": 0.0005555715791014781,
+      "loss": 3.2897,
+      "step": 10740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5130305290222168,
+      "learning_rate": 0.0005555635158427412,
+      "loss": 3.2176,
+      "step": 10741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5505951046943665,
+      "learning_rate": 0.000555555451910899,
+      "loss": 3.2898,
+      "step": 10742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5373878479003906,
+      "learning_rate": 0.000555547387305973,
+      "loss": 3.2094,
+      "step": 10743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4722157418727875,
+      "learning_rate": 0.0005555393220279843,
+      "loss": 3.1176,
+      "step": 10744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5164155960083008,
+      "learning_rate": 0.0005555312560769542,
+      "loss": 3.1844,
+      "step": 10745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49586886167526245,
+      "learning_rate": 0.000555523189452904,
+      "loss": 3.0007,
+      "step": 10746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5387834906578064,
+      "learning_rate": 0.0005555151221558549,
+      "loss": 3.2779,
+      "step": 10747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5266426205635071,
+      "learning_rate": 0.0005555070541858281,
+      "loss": 3.1793,
+      "step": 10748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5374503135681152,
+      "learning_rate": 0.000555498985542845,
+      "loss": 3.2667,
+      "step": 10749
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5462484359741211,
+      "learning_rate": 0.0005554909162269266,
+      "loss": 3.189,
+      "step": 10750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5269687175750732,
+      "learning_rate": 0.0005554828462380945,
+      "loss": 3.3992,
+      "step": 10751
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5425649881362915,
+      "learning_rate": 0.0005554747755763695,
+      "loss": 3.2771,
+      "step": 10752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5352199077606201,
+      "learning_rate": 0.0005554667042417733,
+      "loss": 3.5624,
+      "step": 10753
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5189509987831116,
+      "learning_rate": 0.000555458632234327,
+      "loss": 3.2905,
+      "step": 10754
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5382845401763916,
+      "learning_rate": 0.0005554505595540516,
+      "loss": 3.1669,
+      "step": 10755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5350643396377563,
+      "learning_rate": 0.0005554424862009688,
+      "loss": 3.202,
+      "step": 10756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5131836533546448,
+      "learning_rate": 0.0005554344121750994,
+      "loss": 3.1861,
+      "step": 10757
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5483985543251038,
+      "learning_rate": 0.0005554263374764651,
+      "loss": 3.2958,
+      "step": 10758
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5102946758270264,
+      "learning_rate": 0.000555418262105087,
+      "loss": 3.2614,
+      "step": 10759
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5368705987930298,
+      "learning_rate": 0.0005554101860609864,
+      "loss": 3.3872,
+      "step": 10760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49759748578071594,
+      "learning_rate": 0.0005554021093441844,
+      "loss": 3.3422,
+      "step": 10761
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5840714573860168,
+      "learning_rate": 0.0005553940319547024,
+      "loss": 3.3937,
+      "step": 10762
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5512921214103699,
+      "learning_rate": 0.0005553859538925617,
+      "loss": 3.3002,
+      "step": 10763
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49647557735443115,
+      "learning_rate": 0.0005553778751577836,
+      "loss": 3.3018,
+      "step": 10764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.550879955291748,
+      "learning_rate": 0.0005553697957503893,
+      "loss": 3.3197,
+      "step": 10765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5185007452964783,
+      "learning_rate": 0.0005553617156704001,
+      "loss": 3.2382,
+      "step": 10766
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.510525643825531,
+      "learning_rate": 0.0005553536349178372,
+      "loss": 3.3413,
+      "step": 10767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5063631534576416,
+      "learning_rate": 0.0005553455534927221,
+      "loss": 3.1335,
+      "step": 10768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5001246929168701,
+      "learning_rate": 0.000555337471395076,
+      "loss": 3.1726,
+      "step": 10769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5542845129966736,
+      "learning_rate": 0.00055532938862492,
+      "loss": 3.1714,
+      "step": 10770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5402297377586365,
+      "learning_rate": 0.0005553213051822755,
+      "loss": 3.2763,
+      "step": 10771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5092653632164001,
+      "learning_rate": 0.0005553132210671639,
+      "loss": 3.2897,
+      "step": 10772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5712873935699463,
+      "learning_rate": 0.0005553051362796064,
+      "loss": 3.3383,
+      "step": 10773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5029581189155579,
+      "learning_rate": 0.0005552970508196243,
+      "loss": 3.1195,
+      "step": 10774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.558640718460083,
+      "learning_rate": 0.0005552889646872389,
+      "loss": 3.1996,
+      "step": 10775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5104977488517761,
+      "learning_rate": 0.0005552808778824715,
+      "loss": 3.1922,
+      "step": 10776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4940243363380432,
+      "learning_rate": 0.0005552727904053435,
+      "loss": 2.9714,
+      "step": 10777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5271364450454712,
+      "learning_rate": 0.0005552647022558759,
+      "loss": 3.2721,
+      "step": 10778
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5337794423103333,
+      "learning_rate": 0.0005552566134340904,
+      "loss": 3.3958,
+      "step": 10779
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4681585133075714,
+      "learning_rate": 0.000555248523940008,
+      "loss": 3.3627,
+      "step": 10780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5498515367507935,
+      "learning_rate": 0.0005552404337736501,
+      "loss": 3.3218,
+      "step": 10781
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4920918941497803,
+      "learning_rate": 0.000555232342935038,
+      "loss": 3.2476,
+      "step": 10782
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5251589417457581,
+      "learning_rate": 0.0005552242514241931,
+      "loss": 3.4136,
+      "step": 10783
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49055927991867065,
+      "learning_rate": 0.0005552161592411366,
+      "loss": 3.4924,
+      "step": 10784
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5254443287849426,
+      "learning_rate": 0.0005552080663858899,
+      "loss": 3.1118,
+      "step": 10785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5273293852806091,
+      "learning_rate": 0.0005551999728584742,
+      "loss": 3.2376,
+      "step": 10786
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.46797534823417664,
+      "learning_rate": 0.000555191878658911,
+      "loss": 3.1969,
+      "step": 10787
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5711906552314758,
+      "learning_rate": 0.0005551837837872213,
+      "loss": 3.3446,
+      "step": 10788
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5303103923797607,
+      "learning_rate": 0.0005551756882434268,
+      "loss": 3.1252,
+      "step": 10789
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515155017375946,
+      "learning_rate": 0.0005551675920275486,
+      "loss": 3.2225,
+      "step": 10790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5145278573036194,
+      "learning_rate": 0.0005551594951396081,
+      "loss": 3.5122,
+      "step": 10791
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4963783025741577,
+      "learning_rate": 0.0005551513975796265,
+      "loss": 3.4213,
+      "step": 10792
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5196395516395569,
+      "learning_rate": 0.0005551432993476254,
+      "loss": 3.2661,
+      "step": 10793
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5125848054885864,
+      "learning_rate": 0.0005551352004436258,
+      "loss": 3.2938,
+      "step": 10794
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5345282554626465,
+      "learning_rate": 0.0005551271008676492,
+      "loss": 3.1414,
+      "step": 10795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4970857799053192,
+      "learning_rate": 0.0005551190006197169,
+      "loss": 3.3573,
+      "step": 10796
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49079298973083496,
+      "learning_rate": 0.0005551108996998503,
+      "loss": 3.3713,
+      "step": 10797
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.502100944519043,
+      "learning_rate": 0.0005551027981080707,
+      "loss": 3.2045,
+      "step": 10798
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5889394879341125,
+      "learning_rate": 0.0005550946958443994,
+      "loss": 3.3565,
+      "step": 10799
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5061367154121399,
+      "learning_rate": 0.0005550865929088577,
+      "loss": 3.349,
+      "step": 10800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5013979077339172,
+      "learning_rate": 0.000555078489301467,
+      "loss": 3.3525,
+      "step": 10801
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49582046270370483,
+      "learning_rate": 0.0005550703850222487,
+      "loss": 3.2322,
+      "step": 10802
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5003193020820618,
+      "learning_rate": 0.0005550622800712242,
+      "loss": 3.3309,
+      "step": 10803
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5031837821006775,
+      "learning_rate": 0.0005550541744484147,
+      "loss": 3.1939,
+      "step": 10804
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5380825400352478,
+      "learning_rate": 0.0005550460681538415,
+      "loss": 3.2945,
+      "step": 10805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5140364766120911,
+      "learning_rate": 0.0005550379611875261,
+      "loss": 3.3461,
+      "step": 10806
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5664733052253723,
+      "learning_rate": 0.0005550298535494898,
+      "loss": 3.0924,
+      "step": 10807
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5468794703483582,
+      "learning_rate": 0.000555021745239754,
+      "loss": 3.0776,
+      "step": 10808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5097938776016235,
+      "learning_rate": 0.00055501363625834,
+      "loss": 3.4317,
+      "step": 10809
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5171865820884705,
+      "learning_rate": 0.000555005526605269,
+      "loss": 3.1843,
+      "step": 10810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.544694721698761,
+      "learning_rate": 0.0005549974162805626,
+      "loss": 3.1868,
+      "step": 10811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4788232147693634,
+      "learning_rate": 0.0005549893052842421,
+      "loss": 3.4821,
+      "step": 10812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5630718469619751,
+      "learning_rate": 0.0005549811936163288,
+      "loss": 3.2703,
+      "step": 10813
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5108457207679749,
+      "learning_rate": 0.0005549730812768441,
+      "loss": 3.2386,
+      "step": 10814
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5172486305236816,
+      "learning_rate": 0.0005549649682658096,
+      "loss": 3.4601,
+      "step": 10815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5107886791229248,
+      "learning_rate": 0.0005549568545832462,
+      "loss": 3.1461,
+      "step": 10816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5882880091667175,
+      "learning_rate": 0.0005549487402291756,
+      "loss": 3.2976,
+      "step": 10817
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5461258292198181,
+      "learning_rate": 0.000554940625203619,
+      "loss": 3.3833,
+      "step": 10818
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5405899882316589,
+      "learning_rate": 0.0005549325095065979,
+      "loss": 3.3489,
+      "step": 10819
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5153059959411621,
+      "learning_rate": 0.0005549243931381336,
+      "loss": 2.9776,
+      "step": 10820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49118930101394653,
+      "learning_rate": 0.0005549162760982477,
+      "loss": 3.0449,
+      "step": 10821
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5056452751159668,
+      "learning_rate": 0.0005549081583869612,
+      "loss": 3.4947,
+      "step": 10822
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5282566547393799,
+      "learning_rate": 0.0005549000400042958,
+      "loss": 3.3881,
+      "step": 10823
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5301047563552856,
+      "learning_rate": 0.0005548919209502726,
+      "loss": 3.1563,
+      "step": 10824
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5251234769821167,
+      "learning_rate": 0.0005548838012249132,
+      "loss": 3.2803,
+      "step": 10825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.523571252822876,
+      "learning_rate": 0.000554875680828239,
+      "loss": 3.318,
+      "step": 10826
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5182716846466064,
+      "learning_rate": 0.0005548675597602711,
+      "loss": 3.3921,
+      "step": 10827
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5157158970832825,
+      "learning_rate": 0.0005548594380210313,
+      "loss": 3.0689,
+      "step": 10828
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5221959948539734,
+      "learning_rate": 0.0005548513156105407,
+      "loss": 3.1948,
+      "step": 10829
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49166935682296753,
+      "learning_rate": 0.0005548431925288208,
+      "loss": 3.1451,
+      "step": 10830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5118061900138855,
+      "learning_rate": 0.000554835068775893,
+      "loss": 3.2432,
+      "step": 10831
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49349334836006165,
+      "learning_rate": 0.0005548269443517787,
+      "loss": 3.2898,
+      "step": 10832
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5066019892692566,
+      "learning_rate": 0.0005548188192564991,
+      "loss": 3.2891,
+      "step": 10833
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5938020944595337,
+      "learning_rate": 0.000554810693490076,
+      "loss": 3.2116,
+      "step": 10834
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5334991216659546,
+      "learning_rate": 0.0005548025670525304,
+      "loss": 3.1081,
+      "step": 10835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.502778947353363,
+      "learning_rate": 0.000554794439943884,
+      "loss": 3.233,
+      "step": 10836
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5045210123062134,
+      "learning_rate": 0.0005547863121641581,
+      "loss": 3.314,
+      "step": 10837
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5293318033218384,
+      "learning_rate": 0.000554778183713374,
+      "loss": 3.3784,
+      "step": 10838
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.518131673336029,
+      "learning_rate": 0.0005547700545915531,
+      "loss": 3.3426,
+      "step": 10839
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4948231279850006,
+      "learning_rate": 0.0005547619247987171,
+      "loss": 3.0953,
+      "step": 10840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5470211505889893,
+      "learning_rate": 0.0005547537943348871,
+      "loss": 3.1605,
+      "step": 10841
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5280324816703796,
+      "learning_rate": 0.0005547456632000846,
+      "loss": 3.2811,
+      "step": 10842
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5659511089324951,
+      "learning_rate": 0.0005547375313943312,
+      "loss": 3.2445,
+      "step": 10843
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5293766856193542,
+      "learning_rate": 0.000554729398917648,
+      "loss": 3.2949,
+      "step": 10844
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5941742062568665,
+      "learning_rate": 0.0005547212657700568,
+      "loss": 3.202,
+      "step": 10845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5199886560440063,
+      "learning_rate": 0.0005547131319515787,
+      "loss": 3.1408,
+      "step": 10846
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5088639259338379,
+      "learning_rate": 0.0005547049974622351,
+      "loss": 3.2688,
+      "step": 10847
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5248528718948364,
+      "learning_rate": 0.0005546968623020477,
+      "loss": 3.2334,
+      "step": 10848
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5154558420181274,
+      "learning_rate": 0.0005546887264710377,
+      "loss": 3.2585,
+      "step": 10849
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5221732258796692,
+      "learning_rate": 0.0005546805899692267,
+      "loss": 3.2253,
+      "step": 10850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5415873527526855,
+      "learning_rate": 0.000554672452796636,
+      "loss": 3.1749,
+      "step": 10851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4975447952747345,
+      "learning_rate": 0.000554664314953287,
+      "loss": 3.0391,
+      "step": 10852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4918574094772339,
+      "learning_rate": 0.0005546561764392014,
+      "loss": 3.1746,
+      "step": 10853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.516654372215271,
+      "learning_rate": 0.0005546480372544003,
+      "loss": 3.2713,
+      "step": 10854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5436745285987854,
+      "learning_rate": 0.0005546398973989053,
+      "loss": 3.0698,
+      "step": 10855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5498283505439758,
+      "learning_rate": 0.0005546317568727379,
+      "loss": 3.3623,
+      "step": 10856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5308879613876343,
+      "learning_rate": 0.0005546236156759194,
+      "loss": 3.2788,
+      "step": 10857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.535848081111908,
+      "learning_rate": 0.0005546154738084711,
+      "loss": 3.3993,
+      "step": 10858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6059902310371399,
+      "learning_rate": 0.0005546073312704149,
+      "loss": 3.2072,
+      "step": 10859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5028658509254456,
+      "learning_rate": 0.0005545991880617719,
+      "loss": 3.2419,
+      "step": 10860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5084434747695923,
+      "learning_rate": 0.0005545910441825636,
+      "loss": 3.2826,
+      "step": 10861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5181224942207336,
+      "learning_rate": 0.0005545828996328116,
+      "loss": 3.1966,
+      "step": 10862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5151437520980835,
+      "learning_rate": 0.0005545747544125371,
+      "loss": 3.169,
+      "step": 10863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4924545884132385,
+      "learning_rate": 0.0005545666085217619,
+      "loss": 3.254,
+      "step": 10864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5671659708023071,
+      "learning_rate": 0.000554558461960507,
+      "loss": 3.4059,
+      "step": 10865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5134097337722778,
+      "learning_rate": 0.0005545503147287941,
+      "loss": 3.1668,
+      "step": 10866
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5386476516723633,
+      "learning_rate": 0.0005545421668266448,
+      "loss": 3.2502,
+      "step": 10867
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5015349984169006,
+      "learning_rate": 0.0005545340182540805,
+      "loss": 3.3447,
+      "step": 10868
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4915396273136139,
+      "learning_rate": 0.0005545258690111224,
+      "loss": 3.302,
+      "step": 10869
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5859756469726562,
+      "learning_rate": 0.0005545177190977923,
+      "loss": 3.1863,
+      "step": 10870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.500820517539978,
+      "learning_rate": 0.0005545095685141113,
+      "loss": 3.3541,
+      "step": 10871
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5108267068862915,
+      "learning_rate": 0.0005545014172601014,
+      "loss": 3.0616,
+      "step": 10872
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5519191026687622,
+      "learning_rate": 0.0005544932653357835,
+      "loss": 3.3225,
+      "step": 10873
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4729250967502594,
+      "learning_rate": 0.0005544851127411793,
+      "loss": 3.2311,
+      "step": 10874
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5374777317047119,
+      "learning_rate": 0.0005544769594763104,
+      "loss": 3.1297,
+      "step": 10875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4953976273536682,
+      "learning_rate": 0.0005544688055411981,
+      "loss": 3.0189,
+      "step": 10876
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6086330413818359,
+      "learning_rate": 0.0005544606509358639,
+      "loss": 3.0814,
+      "step": 10877
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5771964192390442,
+      "learning_rate": 0.0005544524956603294,
+      "loss": 3.194,
+      "step": 10878
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5086356997489929,
+      "learning_rate": 0.000554444339714616,
+      "loss": 3.1714,
+      "step": 10879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4892556071281433,
+      "learning_rate": 0.0005544361830987451,
+      "loss": 3.2823,
+      "step": 10880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5008576512336731,
+      "learning_rate": 0.0005544280258127383,
+      "loss": 3.2336,
+      "step": 10881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5014019012451172,
+      "learning_rate": 0.0005544198678566171,
+      "loss": 3.321,
+      "step": 10882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5316650867462158,
+      "learning_rate": 0.0005544117092304029,
+      "loss": 3.4712,
+      "step": 10883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4876498878002167,
+      "learning_rate": 0.0005544035499341172,
+      "loss": 3.4199,
+      "step": 10884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5168114304542542,
+      "learning_rate": 0.0005543953899677815,
+      "loss": 3.1632,
+      "step": 10885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5220661163330078,
+      "learning_rate": 0.0005543872293314174,
+      "loss": 3.0831,
+      "step": 10886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4988704025745392,
+      "learning_rate": 0.0005543790680250461,
+      "loss": 3.2857,
+      "step": 10887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4903241991996765,
+      "learning_rate": 0.0005543709060486895,
+      "loss": 3.2408,
+      "step": 10888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5027443170547485,
+      "learning_rate": 0.0005543627434023688,
+      "loss": 3.2762,
+      "step": 10889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4971694350242615,
+      "learning_rate": 0.0005543545800861055,
+      "loss": 3.1649,
+      "step": 10890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5153796672821045,
+      "learning_rate": 0.0005543464160999214,
+      "loss": 3.3079,
+      "step": 10891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5482482314109802,
+      "learning_rate": 0.0005543382514438376,
+      "loss": 3.0886,
+      "step": 10892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5379815101623535,
+      "learning_rate": 0.0005543300861178759,
+      "loss": 3.1402,
+      "step": 10893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5130733251571655,
+      "learning_rate": 0.0005543219201220576,
+      "loss": 3.5,
+      "step": 10894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47875428199768066,
+      "learning_rate": 0.0005543137534564044,
+      "loss": 3.1322,
+      "step": 10895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5776912569999695,
+      "learning_rate": 0.0005543055861209376,
+      "loss": 3.1212,
+      "step": 10896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5165498852729797,
+      "learning_rate": 0.0005542974181156789,
+      "loss": 3.1643,
+      "step": 10897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5151985883712769,
+      "learning_rate": 0.0005542892494406499,
+      "loss": 3.0556,
+      "step": 10898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5367438793182373,
+      "learning_rate": 0.0005542810800958718,
+      "loss": 3.2911,
+      "step": 10899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49575287103652954,
+      "learning_rate": 0.0005542729100813662,
+      "loss": 3.3014,
+      "step": 10900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5376871824264526,
+      "learning_rate": 0.0005542647393971548,
+      "loss": 3.2976,
+      "step": 10901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5368502140045166,
+      "learning_rate": 0.000554256568043259,
+      "loss": 3.4037,
+      "step": 10902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5264731049537659,
+      "learning_rate": 0.0005542483960197002,
+      "loss": 3.3177,
+      "step": 10903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5258291363716125,
+      "learning_rate": 0.0005542402233265003,
+      "loss": 3.0652,
+      "step": 10904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5016001462936401,
+      "learning_rate": 0.0005542320499636804,
+      "loss": 3.3955,
+      "step": 10905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5093668103218079,
+      "learning_rate": 0.0005542238759312623,
+      "loss": 3.1226,
+      "step": 10906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.504332959651947,
+      "learning_rate": 0.0005542157012292673,
+      "loss": 3.2307,
+      "step": 10907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5304985642433167,
+      "learning_rate": 0.0005542075258577172,
+      "loss": 3.1603,
+      "step": 10908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5355924367904663,
+      "learning_rate": 0.0005541993498166334,
+      "loss": 3.4154,
+      "step": 10909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.598781943321228,
+      "learning_rate": 0.0005541911731060374,
+      "loss": 3.4022,
+      "step": 10910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5250735282897949,
+      "learning_rate": 0.0005541829957259508,
+      "loss": 3.0233,
+      "step": 10911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5747032761573792,
+      "learning_rate": 0.0005541748176763951,
+      "loss": 3.1773,
+      "step": 10912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5299064517021179,
+      "learning_rate": 0.0005541666389573918,
+      "loss": 2.9314,
+      "step": 10913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4818550646305084,
+      "learning_rate": 0.0005541584595689624,
+      "loss": 3.2298,
+      "step": 10914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5289633274078369,
+      "learning_rate": 0.0005541502795111288,
+      "loss": 3.26,
+      "step": 10915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5091275572776794,
+      "learning_rate": 0.000554142098783912,
+      "loss": 3.2175,
+      "step": 10916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5333389639854431,
+      "learning_rate": 0.000554133917387334,
+      "loss": 3.6551,
+      "step": 10917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4829789400100708,
+      "learning_rate": 0.0005541257353214161,
+      "loss": 3.1711,
+      "step": 10918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5595918297767639,
+      "learning_rate": 0.00055411755258618,
+      "loss": 3.2935,
+      "step": 10919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5571411848068237,
+      "learning_rate": 0.000554109369181647,
+      "loss": 3.2141,
+      "step": 10920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.534190833568573,
+      "learning_rate": 0.000554101185107839,
+      "loss": 3.3717,
+      "step": 10921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.502724289894104,
+      "learning_rate": 0.0005540930003647773,
+      "loss": 3.2407,
+      "step": 10922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5559344291687012,
+      "learning_rate": 0.0005540848149524835,
+      "loss": 3.1769,
+      "step": 10923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5261440277099609,
+      "learning_rate": 0.0005540766288709792,
+      "loss": 3.3852,
+      "step": 10924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5682399272918701,
+      "learning_rate": 0.000554068442120286,
+      "loss": 3.2593,
+      "step": 10925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4791615605354309,
+      "learning_rate": 0.0005540602547004255,
+      "loss": 3.428,
+      "step": 10926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5266600847244263,
+      "learning_rate": 0.000554052066611419,
+      "loss": 3.3388,
+      "step": 10927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5460447072982788,
+      "learning_rate": 0.0005540438778532885,
+      "loss": 3.0518,
+      "step": 10928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.534444272518158,
+      "learning_rate": 0.0005540356884260551,
+      "loss": 3.2383,
+      "step": 10929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5190091729164124,
+      "learning_rate": 0.0005540274983297407,
+      "loss": 3.2742,
+      "step": 10930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5553072094917297,
+      "learning_rate": 0.0005540193075643668,
+      "loss": 3.304,
+      "step": 10931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5311570167541504,
+      "learning_rate": 0.0005540111161299547,
+      "loss": 3.3071,
+      "step": 10932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.523281455039978,
+      "learning_rate": 0.0005540029240265265,
+      "loss": 3.2171,
+      "step": 10933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5270448327064514,
+      "learning_rate": 0.0005539947312541033,
+      "loss": 3.3145,
+      "step": 10934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5168893933296204,
+      "learning_rate": 0.0005539865378127069,
+      "loss": 3.2543,
+      "step": 10935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5505146384239197,
+      "learning_rate": 0.0005539783437023588,
+      "loss": 3.0689,
+      "step": 10936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48953521251678467,
+      "learning_rate": 0.0005539701489230807,
+      "loss": 3.2071,
+      "step": 10937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5777466297149658,
+      "learning_rate": 0.000553961953474894,
+      "loss": 3.228,
+      "step": 10938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.538305938243866,
+      "learning_rate": 0.0005539537573578205,
+      "loss": 3.2121,
+      "step": 10939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5043644905090332,
+      "learning_rate": 0.0005539455605718817,
+      "loss": 3.3311,
+      "step": 10940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5230538845062256,
+      "learning_rate": 0.000553937363117099,
+      "loss": 3.0433,
+      "step": 10941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49755704402923584,
+      "learning_rate": 0.0005539291649934943,
+      "loss": 3.2049,
+      "step": 10942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49073851108551025,
+      "learning_rate": 0.0005539209662010889,
+      "loss": 3.0794,
+      "step": 10943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5318127274513245,
+      "learning_rate": 0.0005539127667399048,
+      "loss": 3.3468,
+      "step": 10944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5030101537704468,
+      "learning_rate": 0.0005539045666099632,
+      "loss": 3.3754,
+      "step": 10945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5165680646896362,
+      "learning_rate": 0.0005538963658112858,
+      "loss": 3.102,
+      "step": 10946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48056358098983765,
+      "learning_rate": 0.0005538881643438943,
+      "loss": 3.3966,
+      "step": 10947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.517524242401123,
+      "learning_rate": 0.0005538799622078101,
+      "loss": 3.3516,
+      "step": 10948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5031284689903259,
+      "learning_rate": 0.000553871759403055,
+      "loss": 3.0759,
+      "step": 10949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5510088801383972,
+      "learning_rate": 0.0005538635559296507,
+      "loss": 3.1852,
+      "step": 10950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48190462589263916,
+      "learning_rate": 0.0005538553517876185,
+      "loss": 3.5855,
+      "step": 10951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.518423318862915,
+      "learning_rate": 0.0005538471469769802,
+      "loss": 3.4513,
+      "step": 10952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5671064257621765,
+      "learning_rate": 0.0005538389414977573,
+      "loss": 3.1931,
+      "step": 10953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5123745799064636,
+      "learning_rate": 0.0005538307353499715,
+      "loss": 3.2022,
+      "step": 10954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4641689658164978,
+      "learning_rate": 0.0005538225285336445,
+      "loss": 3.3417,
+      "step": 10955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5528790950775146,
+      "learning_rate": 0.0005538143210487977,
+      "loss": 3.3302,
+      "step": 10956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5319774150848389,
+      "learning_rate": 0.000553806112895453,
+      "loss": 3.1984,
+      "step": 10957
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5352564454078674,
+      "learning_rate": 0.0005537979040736317,
+      "loss": 3.2959,
+      "step": 10958
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5923522710800171,
+      "learning_rate": 0.0005537896945833555,
+      "loss": 3.2073,
+      "step": 10959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5047934651374817,
+      "learning_rate": 0.0005537814844246462,
+      "loss": 3.5024,
+      "step": 10960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5356642007827759,
+      "learning_rate": 0.0005537732735975252,
+      "loss": 3.2996,
+      "step": 10961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5496769547462463,
+      "learning_rate": 0.0005537650621020143,
+      "loss": 3.2524,
+      "step": 10962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5538159608840942,
+      "learning_rate": 0.0005537568499381351,
+      "loss": 3.1377,
+      "step": 10963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5073235630989075,
+      "learning_rate": 0.0005537486371059092,
+      "loss": 3.4908,
+      "step": 10964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5352749228477478,
+      "learning_rate": 0.0005537404236053582,
+      "loss": 3.3659,
+      "step": 10965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.501549482345581,
+      "learning_rate": 0.0005537322094365038,
+      "loss": 3.4548,
+      "step": 10966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5620731115341187,
+      "learning_rate": 0.0005537239945993675,
+      "loss": 3.1772,
+      "step": 10967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5060946345329285,
+      "learning_rate": 0.0005537157790939711,
+      "loss": 3.1628,
+      "step": 10968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5136206150054932,
+      "learning_rate": 0.0005537075629203361,
+      "loss": 3.1592,
+      "step": 10969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5491225123405457,
+      "learning_rate": 0.0005536993460784843,
+      "loss": 3.1226,
+      "step": 10970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5510766506195068,
+      "learning_rate": 0.0005536911285684372,
+      "loss": 3.2825,
+      "step": 10971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59326171875,
+      "learning_rate": 0.0005536829103902164,
+      "loss": 3.2038,
+      "step": 10972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5343856811523438,
+      "learning_rate": 0.0005536746915438438,
+      "loss": 3.2069,
+      "step": 10973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.524446964263916,
+      "learning_rate": 0.0005536664720293408,
+      "loss": 3.1553,
+      "step": 10974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5818180441856384,
+      "learning_rate": 0.0005536582518467292,
+      "loss": 3.2898,
+      "step": 10975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49687114357948303,
+      "learning_rate": 0.0005536500309960304,
+      "loss": 3.4254,
+      "step": 10976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5242406725883484,
+      "learning_rate": 0.0005536418094772664,
+      "loss": 3.1369,
+      "step": 10977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5363040566444397,
+      "learning_rate": 0.0005536335872904587,
+      "loss": 3.2857,
+      "step": 10978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5277572870254517,
+      "learning_rate": 0.0005536253644356289,
+      "loss": 3.2198,
+      "step": 10979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5251345634460449,
+      "learning_rate": 0.0005536171409127986,
+      "loss": 3.3776,
+      "step": 10980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4895996153354645,
+      "learning_rate": 0.0005536089167219897,
+      "loss": 3.1118,
+      "step": 10981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.550209105014801,
+      "learning_rate": 0.0005536006918632236,
+      "loss": 3.3635,
+      "step": 10982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5238742232322693,
+      "learning_rate": 0.0005535924663365221,
+      "loss": 3.37,
+      "step": 10983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5147226452827454,
+      "learning_rate": 0.0005535842401419071,
+      "loss": 3.3436,
+      "step": 10984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5709726214408875,
+      "learning_rate": 0.0005535760132793997,
+      "loss": 3.2941,
+      "step": 10985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5367663502693176,
+      "learning_rate": 0.000553567785749022,
+      "loss": 3.2,
+      "step": 10986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5051316022872925,
+      "learning_rate": 0.0005535595575507955,
+      "loss": 3.2523,
+      "step": 10987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.502051830291748,
+      "learning_rate": 0.0005535513286847421,
+      "loss": 3.2888,
+      "step": 10988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5447306632995605,
+      "learning_rate": 0.0005535430991508831,
+      "loss": 3.2808,
+      "step": 10989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5220550894737244,
+      "learning_rate": 0.0005535348689492404,
+      "loss": 3.1534,
+      "step": 10990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6093209385871887,
+      "learning_rate": 0.0005535266380798358,
+      "loss": 3.1002,
+      "step": 10991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51113361120224,
+      "learning_rate": 0.0005535184065426907,
+      "loss": 3.2006,
+      "step": 10992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5078108310699463,
+      "learning_rate": 0.000553510174337827,
+      "loss": 3.1692,
+      "step": 10993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5096676349639893,
+      "learning_rate": 0.0005535019414652662,
+      "loss": 3.2649,
+      "step": 10994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6298714280128479,
+      "learning_rate": 0.0005534937079250301,
+      "loss": 3.0877,
+      "step": 10995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5195075869560242,
+      "learning_rate": 0.0005534854737171402,
+      "loss": 3.3914,
+      "step": 10996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5262744426727295,
+      "learning_rate": 0.0005534772388416186,
+      "loss": 3.3941,
+      "step": 10997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5323526263237,
+      "learning_rate": 0.0005534690032984866,
+      "loss": 3.2896,
+      "step": 10998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.567234456539154,
+      "learning_rate": 0.0005534607670877661,
+      "loss": 3.1214,
+      "step": 10999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49547919631004333,
+      "learning_rate": 0.0005534525302094787,
+      "loss": 3.4712,
+      "step": 11000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5253562331199646,
+      "learning_rate": 0.0005534442926636461,
+      "loss": 3.4296,
+      "step": 11001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.536155104637146,
+      "learning_rate": 0.00055343605445029,
+      "loss": 3.2442,
+      "step": 11002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5386752486228943,
+      "learning_rate": 0.0005534278155694321,
+      "loss": 3.1203,
+      "step": 11003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5196104049682617,
+      "learning_rate": 0.0005534195760210941,
+      "loss": 3.2681,
+      "step": 11004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5381636023521423,
+      "learning_rate": 0.0005534113358052977,
+      "loss": 3.2436,
+      "step": 11005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5484285950660706,
+      "learning_rate": 0.0005534030949220646,
+      "loss": 3.2857,
+      "step": 11006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5643191933631897,
+      "learning_rate": 0.0005533948533714166,
+      "loss": 3.2674,
+      "step": 11007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5526049733161926,
+      "learning_rate": 0.0005533866111533753,
+      "loss": 3.2277,
+      "step": 11008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5223939418792725,
+      "learning_rate": 0.0005533783682679624,
+      "loss": 3.3033,
+      "step": 11009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5286209583282471,
+      "learning_rate": 0.0005533701247151996,
+      "loss": 3.2965,
+      "step": 11010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5266342759132385,
+      "learning_rate": 0.0005533618804951087,
+      "loss": 3.2701,
+      "step": 11011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5430400967597961,
+      "learning_rate": 0.0005533536356077113,
+      "loss": 3.0469,
+      "step": 11012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49393025040626526,
+      "learning_rate": 0.0005533453900530294,
+      "loss": 3.1746,
+      "step": 11013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5422281622886658,
+      "learning_rate": 0.0005533371438310843,
+      "loss": 3.3271,
+      "step": 11014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5507173538208008,
+      "learning_rate": 0.000553328896941898,
+      "loss": 3.2458,
+      "step": 11015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4943542778491974,
+      "learning_rate": 0.000553320649385492,
+      "loss": 3.2087,
+      "step": 11016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5512304902076721,
+      "learning_rate": 0.0005533124011618884,
+      "loss": 3.1788,
+      "step": 11017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45192670822143555,
+      "learning_rate": 0.0005533041522711085,
+      "loss": 3.1242,
+      "step": 11018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5453776717185974,
+      "learning_rate": 0.0005532959027131743,
+      "loss": 3.3347,
+      "step": 11019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5177527070045471,
+      "learning_rate": 0.0005532876524881075,
+      "loss": 3.2072,
+      "step": 11020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5608527660369873,
+      "learning_rate": 0.0005532794015959296,
+      "loss": 3.2669,
+      "step": 11021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5439431667327881,
+      "learning_rate": 0.0005532711500366625,
+      "loss": 3.1555,
+      "step": 11022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5050414204597473,
+      "learning_rate": 0.000553262897810328,
+      "loss": 3.3673,
+      "step": 11023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6056579947471619,
+      "learning_rate": 0.0005532546449169478,
+      "loss": 3.0847,
+      "step": 11024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50806725025177,
+      "learning_rate": 0.0005532463913565436,
+      "loss": 3.1318,
+      "step": 11025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47920647263526917,
+      "learning_rate": 0.0005532381371291372,
+      "loss": 3.307,
+      "step": 11026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49545514583587646,
+      "learning_rate": 0.0005532298822347501,
+      "loss": 3.369,
+      "step": 11027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4789004921913147,
+      "learning_rate": 0.0005532216266734044,
+      "loss": 3.2807,
+      "step": 11028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5694847106933594,
+      "learning_rate": 0.0005532133704451216,
+      "loss": 3.2739,
+      "step": 11029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5591051578521729,
+      "learning_rate": 0.0005532051135499236,
+      "loss": 3.2335,
+      "step": 11030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5124754309654236,
+      "learning_rate": 0.000553196855987832,
+      "loss": 3.0694,
+      "step": 11031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5239121317863464,
+      "learning_rate": 0.0005531885977588686,
+      "loss": 3.3521,
+      "step": 11032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5368714332580566,
+      "learning_rate": 0.0005531803388630551,
+      "loss": 3.2614,
+      "step": 11033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5138024687767029,
+      "learning_rate": 0.0005531720793004135,
+      "loss": 3.2224,
+      "step": 11034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5147234201431274,
+      "learning_rate": 0.0005531638190709651,
+      "loss": 3.2612,
+      "step": 11035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5324978828430176,
+      "learning_rate": 0.0005531555581747321,
+      "loss": 3.4883,
+      "step": 11036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.501106321811676,
+      "learning_rate": 0.0005531472966117361,
+      "loss": 3.2922,
+      "step": 11037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5278347134590149,
+      "learning_rate": 0.0005531390343819987,
+      "loss": 3.2189,
+      "step": 11038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5315659046173096,
+      "learning_rate": 0.0005531307714855419,
+      "loss": 3.2379,
+      "step": 11039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.523786187171936,
+      "learning_rate": 0.0005531225079223874,
+      "loss": 3.2896,
+      "step": 11040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5235302448272705,
+      "learning_rate": 0.0005531142436925569,
+      "loss": 3.5168,
+      "step": 11041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.492496132850647,
+      "learning_rate": 0.0005531059787960721,
+      "loss": 3.2794,
+      "step": 11042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5163391828536987,
+      "learning_rate": 0.0005530977132329549,
+      "loss": 3.0546,
+      "step": 11043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5126833915710449,
+      "learning_rate": 0.0005530894470032271,
+      "loss": 3.1981,
+      "step": 11044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5161365270614624,
+      "learning_rate": 0.0005530811801069104,
+      "loss": 3.2133,
+      "step": 11045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5676143169403076,
+      "learning_rate": 0.0005530729125440265,
+      "loss": 3.117,
+      "step": 11046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5352276563644409,
+      "learning_rate": 0.0005530646443145973,
+      "loss": 3.2627,
+      "step": 11047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5018179416656494,
+      "learning_rate": 0.0005530563754186444,
+      "loss": 3.2552,
+      "step": 11048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5012510418891907,
+      "learning_rate": 0.0005530481058561899,
+      "loss": 3.2785,
+      "step": 11049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5244871973991394,
+      "learning_rate": 0.0005530398356272554,
+      "loss": 3.3138,
+      "step": 11050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5109242796897888,
+      "learning_rate": 0.0005530315647318626,
+      "loss": 3.3587,
+      "step": 11051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5416744351387024,
+      "learning_rate": 0.0005530232931700333,
+      "loss": 3.182,
+      "step": 11052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5090746283531189,
+      "learning_rate": 0.0005530150209417894,
+      "loss": 3.1963,
+      "step": 11053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5169641971588135,
+      "learning_rate": 0.0005530067480471526,
+      "loss": 3.3494,
+      "step": 11054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4892723560333252,
+      "learning_rate": 0.0005529984744861448,
+      "loss": 3.3338,
+      "step": 11055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5152621865272522,
+      "learning_rate": 0.0005529902002587877,
+      "loss": 3.3415,
+      "step": 11056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4981408417224884,
+      "learning_rate": 0.000552981925365103,
+      "loss": 3.2972,
+      "step": 11057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5364402532577515,
+      "learning_rate": 0.0005529736498051127,
+      "loss": 3.2753,
+      "step": 11058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4937974214553833,
+      "learning_rate": 0.0005529653735788383,
+      "loss": 3.2908,
+      "step": 11059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5432580709457397,
+      "learning_rate": 0.0005529570966863021,
+      "loss": 3.2897,
+      "step": 11060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5275623798370361,
+      "learning_rate": 0.0005529488191275253,
+      "loss": 3.3915,
+      "step": 11061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5411865711212158,
+      "learning_rate": 0.0005529405409025302,
+      "loss": 3.2582,
+      "step": 11062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5429436564445496,
+      "learning_rate": 0.0005529322620113382,
+      "loss": 3.267,
+      "step": 11063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5083034038543701,
+      "learning_rate": 0.0005529239824539715,
+      "loss": 3.3395,
+      "step": 11064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49724969267845154,
+      "learning_rate": 0.0005529157022304516,
+      "loss": 3.0817,
+      "step": 11065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48776689171791077,
+      "learning_rate": 0.0005529074213408004,
+      "loss": 3.2054,
+      "step": 11066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4902147054672241,
+      "learning_rate": 0.0005528991397850397,
+      "loss": 3.3422,
+      "step": 11067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5294910073280334,
+      "learning_rate": 0.0005528908575631914,
+      "loss": 3.3306,
+      "step": 11068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5070090293884277,
+      "learning_rate": 0.0005528825746752773,
+      "loss": 3.4023,
+      "step": 11069
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5285083055496216,
+      "learning_rate": 0.0005528742911213191,
+      "loss": 3.2324,
+      "step": 11070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.513079822063446,
+      "learning_rate": 0.0005528660069013387,
+      "loss": 3.2539,
+      "step": 11071
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5381377339363098,
+      "learning_rate": 0.0005528577220153579,
+      "loss": 3.4106,
+      "step": 11072
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4675285518169403,
+      "learning_rate": 0.0005528494364633985,
+      "loss": 3.4012,
+      "step": 11073
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5249191522598267,
+      "learning_rate": 0.0005528411502454824,
+      "loss": 3.1627,
+      "step": 11074
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49168458580970764,
+      "learning_rate": 0.0005528328633616313,
+      "loss": 3.2428,
+      "step": 11075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4863361716270447,
+      "learning_rate": 0.0005528245758118671,
+      "loss": 3.2086,
+      "step": 11076
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5234729647636414,
+      "learning_rate": 0.0005528162875962117,
+      "loss": 3.4566,
+      "step": 11077
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5098854303359985,
+      "learning_rate": 0.0005528079987146868,
+      "loss": 3.3018,
+      "step": 11078
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5080091953277588,
+      "learning_rate": 0.0005527997091673143,
+      "loss": 3.3924,
+      "step": 11079
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5155888795852661,
+      "learning_rate": 0.000552791418954116,
+      "loss": 3.1819,
+      "step": 11080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5143377780914307,
+      "learning_rate": 0.0005527831280751138,
+      "loss": 3.5005,
+      "step": 11081
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5565736293792725,
+      "learning_rate": 0.0005527748365303295,
+      "loss": 3.198,
+      "step": 11082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5252052545547485,
+      "learning_rate": 0.0005527665443197849,
+      "loss": 3.1299,
+      "step": 11083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5647584199905396,
+      "learning_rate": 0.0005527582514435017,
+      "loss": 3.5219,
+      "step": 11084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.581468939781189,
+      "learning_rate": 0.0005527499579015021,
+      "loss": 3.3298,
+      "step": 11085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5266270041465759,
+      "learning_rate": 0.0005527416636938077,
+      "loss": 3.1852,
+      "step": 11086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5025612115859985,
+      "learning_rate": 0.0005527333688204405,
+      "loss": 3.3013,
+      "step": 11087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5126590132713318,
+      "learning_rate": 0.0005527250732814222,
+      "loss": 3.4074,
+      "step": 11088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5125699043273926,
+      "learning_rate": 0.0005527167770767746,
+      "loss": 3.0508,
+      "step": 11089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5985158085823059,
+      "learning_rate": 0.0005527084802065197,
+      "loss": 3.5684,
+      "step": 11090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4821082353591919,
+      "learning_rate": 0.0005527001826706793,
+      "loss": 3.2999,
+      "step": 11091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5150079131126404,
+      "learning_rate": 0.0005526918844692752,
+      "loss": 3.3448,
+      "step": 11092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5428050756454468,
+      "learning_rate": 0.0005526835856023294,
+      "loss": 3.0969,
+      "step": 11093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5639786720275879,
+      "learning_rate": 0.0005526752860698636,
+      "loss": 3.2585,
+      "step": 11094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5006402730941772,
+      "learning_rate": 0.0005526669858718998,
+      "loss": 3.2274,
+      "step": 11095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5905807614326477,
+      "learning_rate": 0.0005526586850084596,
+      "loss": 3.3441,
+      "step": 11096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5085511207580566,
+      "learning_rate": 0.0005526503834795654,
+      "loss": 3.416,
+      "step": 11097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54239422082901,
+      "learning_rate": 0.0005526420812852384,
+      "loss": 2.8477,
+      "step": 11098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5083803534507751,
+      "learning_rate": 0.0005526337784255009,
+      "loss": 3.3743,
+      "step": 11099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4984433352947235,
+      "learning_rate": 0.0005526254749003746,
+      "loss": 3.4834,
+      "step": 11100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5210543870925903,
+      "learning_rate": 0.0005526171707098814,
+      "loss": 3.336,
+      "step": 11101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5217922925949097,
+      "learning_rate": 0.0005526088658540433,
+      "loss": 3.1881,
+      "step": 11102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4963856339454651,
+      "learning_rate": 0.000552600560332882,
+      "loss": 3.1313,
+      "step": 11103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5491219758987427,
+      "learning_rate": 0.0005525922541464195,
+      "loss": 3.2162,
+      "step": 11104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5014486908912659,
+      "learning_rate": 0.0005525839472946774,
+      "loss": 3.2393,
+      "step": 11105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5988680720329285,
+      "learning_rate": 0.000552575639777678,
+      "loss": 3.3308,
+      "step": 11106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5203366279602051,
+      "learning_rate": 0.0005525673315954428,
+      "loss": 3.3049,
+      "step": 11107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5146699547767639,
+      "learning_rate": 0.0005525590227479941,
+      "loss": 3.0487,
+      "step": 11108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5197408199310303,
+      "learning_rate": 0.0005525507132353533,
+      "loss": 3.1794,
+      "step": 11109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5432579517364502,
+      "learning_rate": 0.0005525424030575427,
+      "loss": 3.0737,
+      "step": 11110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5454299449920654,
+      "learning_rate": 0.000552534092214584,
+      "loss": 3.1482,
+      "step": 11111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5839552283287048,
+      "learning_rate": 0.000552525780706499,
+      "loss": 3.2161,
+      "step": 11112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5181378126144409,
+      "learning_rate": 0.0005525174685333098,
+      "loss": 3.2954,
+      "step": 11113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5586148500442505,
+      "learning_rate": 0.0005525091556950381,
+      "loss": 3.1837,
+      "step": 11114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5175696015357971,
+      "learning_rate": 0.000552500842191706,
+      "loss": 3.1286,
+      "step": 11115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5837456583976746,
+      "learning_rate": 0.0005524925280233351,
+      "loss": 3.1588,
+      "step": 11116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5126967430114746,
+      "learning_rate": 0.0005524842131899476,
+      "loss": 3.4051,
+      "step": 11117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49251213669776917,
+      "learning_rate": 0.0005524758976915652,
+      "loss": 3.2309,
+      "step": 11118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5178259015083313,
+      "learning_rate": 0.0005524675815282099,
+      "loss": 3.2385,
+      "step": 11119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5161910653114319,
+      "learning_rate": 0.0005524592646999035,
+      "loss": 3.1784,
+      "step": 11120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5630325078964233,
+      "learning_rate": 0.0005524509472066683,
+      "loss": 3.0025,
+      "step": 11121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5052022337913513,
+      "learning_rate": 0.0005524426290485255,
+      "loss": 3.1637,
+      "step": 11122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5926064252853394,
+      "learning_rate": 0.0005524343102254976,
+      "loss": 3.3637,
+      "step": 11123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5273416638374329,
+      "learning_rate": 0.0005524259907376063,
+      "loss": 3.2147,
+      "step": 11124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.552270233631134,
+      "learning_rate": 0.0005524176705848735,
+      "loss": 3.1198,
+      "step": 11125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5399708151817322,
+      "learning_rate": 0.0005524093497673211,
+      "loss": 3.0797,
+      "step": 11126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5599607229232788,
+      "learning_rate": 0.000552401028284971,
+      "loss": 3.0212,
+      "step": 11127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.536972165107727,
+      "learning_rate": 0.0005523927061378453,
+      "loss": 2.9896,
+      "step": 11128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5732285976409912,
+      "learning_rate": 0.0005523843833259658,
+      "loss": 3.1458,
+      "step": 11129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5276511311531067,
+      "learning_rate": 0.0005523760598493544,
+      "loss": 3.5314,
+      "step": 11130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.506087064743042,
+      "learning_rate": 0.0005523677357080329,
+      "loss": 3.3586,
+      "step": 11131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5099312663078308,
+      "learning_rate": 0.0005523594109020233,
+      "loss": 3.328,
+      "step": 11132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51421719789505,
+      "learning_rate": 0.0005523510854313478,
+      "loss": 3.3486,
+      "step": 11133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5033894181251526,
+      "learning_rate": 0.000552342759296028,
+      "loss": 3.2523,
+      "step": 11134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5557559132575989,
+      "learning_rate": 0.000552334432496086,
+      "loss": 3.4129,
+      "step": 11135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5153203010559082,
+      "learning_rate": 0.0005523261050315435,
+      "loss": 3.4043,
+      "step": 11136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5677420496940613,
+      "learning_rate": 0.0005523177769024228,
+      "loss": 3.0835,
+      "step": 11137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.555735170841217,
+      "learning_rate": 0.0005523094481087455,
+      "loss": 3.2663,
+      "step": 11138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5250891447067261,
+      "learning_rate": 0.0005523011186505338,
+      "loss": 3.1598,
+      "step": 11139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.501929759979248,
+      "learning_rate": 0.0005522927885278094,
+      "loss": 3.3389,
+      "step": 11140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5323585271835327,
+      "learning_rate": 0.0005522844577405943,
+      "loss": 3.2071,
+      "step": 11141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5253964066505432,
+      "learning_rate": 0.0005522761262889106,
+      "loss": 3.0697,
+      "step": 11142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5121307969093323,
+      "learning_rate": 0.0005522677941727801,
+      "loss": 3.1454,
+      "step": 11143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5765665173530579,
+      "learning_rate": 0.0005522594613922248,
+      "loss": 3.3376,
+      "step": 11144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5566869974136353,
+      "learning_rate": 0.0005522511279472666,
+      "loss": 3.2357,
+      "step": 11145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.540515124797821,
+      "learning_rate": 0.0005522427938379275,
+      "loss": 3.2131,
+      "step": 11146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5782142877578735,
+      "learning_rate": 0.0005522344590642294,
+      "loss": 3.3415,
+      "step": 11147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5115688443183899,
+      "learning_rate": 0.0005522261236261943,
+      "loss": 3.1543,
+      "step": 11148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5313143730163574,
+      "learning_rate": 0.0005522177875238441,
+      "loss": 3.3948,
+      "step": 11149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.563582718372345,
+      "learning_rate": 0.0005522094507572009,
+      "loss": 3.2437,
+      "step": 11150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.514954686164856,
+      "learning_rate": 0.0005522011133262864,
+      "loss": 3.3182,
+      "step": 11151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5384813547134399,
+      "learning_rate": 0.0005521927752311227,
+      "loss": 3.309,
+      "step": 11152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5301558971405029,
+      "learning_rate": 0.0005521844364717319,
+      "loss": 3.1738,
+      "step": 11153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5261275768280029,
+      "learning_rate": 0.0005521760970481357,
+      "loss": 3.2791,
+      "step": 11154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.539004385471344,
+      "learning_rate": 0.0005521677569603563,
+      "loss": 3.172,
+      "step": 11155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5053558945655823,
+      "learning_rate": 0.0005521594162084155,
+      "loss": 3.207,
+      "step": 11156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4946291744709015,
+      "learning_rate": 0.0005521510747923353,
+      "loss": 3.1878,
+      "step": 11157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5328413844108582,
+      "learning_rate": 0.0005521427327121376,
+      "loss": 3.2071,
+      "step": 11158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49959659576416016,
+      "learning_rate": 0.0005521343899678447,
+      "loss": 3.3371,
+      "step": 11159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.535815417766571,
+      "learning_rate": 0.0005521260465594782,
+      "loss": 3.2255,
+      "step": 11160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5187474489212036,
+      "learning_rate": 0.0005521177024870602,
+      "loss": 3.2528,
+      "step": 11161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5385481715202332,
+      "learning_rate": 0.0005521093577506128,
+      "loss": 3.2155,
+      "step": 11162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5287864804267883,
+      "learning_rate": 0.0005521010123501578,
+      "loss": 3.2346,
+      "step": 11163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.501727819442749,
+      "learning_rate": 0.0005520926662857174,
+      "loss": 3.2213,
+      "step": 11164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5046778321266174,
+      "learning_rate": 0.0005520843195573132,
+      "loss": 3.3777,
+      "step": 11165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5383339524269104,
+      "learning_rate": 0.0005520759721649676,
+      "loss": 3.3155,
+      "step": 11166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49202319979667664,
+      "learning_rate": 0.0005520676241087023,
+      "loss": 3.0838,
+      "step": 11167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4806743562221527,
+      "learning_rate": 0.0005520592753885394,
+      "loss": 3.3125,
+      "step": 11168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5282479524612427,
+      "learning_rate": 0.0005520509260045009,
+      "loss": 3.1906,
+      "step": 11169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5135144591331482,
+      "learning_rate": 0.0005520425759566087,
+      "loss": 3.1592,
+      "step": 11170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5296515822410583,
+      "learning_rate": 0.000552034225244885,
+      "loss": 3.3004,
+      "step": 11171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5206128358840942,
+      "learning_rate": 0.0005520258738693516,
+      "loss": 3.308,
+      "step": 11172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5016621351242065,
+      "learning_rate": 0.0005520175218300305,
+      "loss": 3.2374,
+      "step": 11173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5277829170227051,
+      "learning_rate": 0.0005520091691269438,
+      "loss": 3.2679,
+      "step": 11174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48501524329185486,
+      "learning_rate": 0.0005520008157601134,
+      "loss": 3.3124,
+      "step": 11175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5397695899009705,
+      "learning_rate": 0.0005519924617295613,
+      "loss": 3.0827,
+      "step": 11176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4927702844142914,
+      "learning_rate": 0.0005519841070353097,
+      "loss": 3.2994,
+      "step": 11177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48005515336990356,
+      "learning_rate": 0.0005519757516773804,
+      "loss": 3.4319,
+      "step": 11178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5230315327644348,
+      "learning_rate": 0.0005519673956557954,
+      "loss": 3.2013,
+      "step": 11179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5246223211288452,
+      "learning_rate": 0.0005519590389705769,
+      "loss": 3.2824,
+      "step": 11180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5252644419670105,
+      "learning_rate": 0.0005519506816217466,
+      "loss": 3.323,
+      "step": 11181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5019978284835815,
+      "learning_rate": 0.0005519423236093268,
+      "loss": 3.3634,
+      "step": 11182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5068698525428772,
+      "learning_rate": 0.0005519339649333394,
+      "loss": 3.2813,
+      "step": 11183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5238425731658936,
+      "learning_rate": 0.0005519256055938064,
+      "loss": 3.2488,
+      "step": 11184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51482093334198,
+      "learning_rate": 0.0005519172455907499,
+      "loss": 3.3138,
+      "step": 11185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5190244317054749,
+      "learning_rate": 0.0005519088849241918,
+      "loss": 3.246,
+      "step": 11186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49023398756980896,
+      "learning_rate": 0.0005519005235941542,
+      "loss": 3.2134,
+      "step": 11187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5314106345176697,
+      "learning_rate": 0.0005518921616006591,
+      "loss": 3.134,
+      "step": 11188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4904386103153229,
+      "learning_rate": 0.0005518837989437285,
+      "loss": 3.3674,
+      "step": 11189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49414244294166565,
+      "learning_rate": 0.0005518754356233845,
+      "loss": 3.0914,
+      "step": 11190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5230140089988708,
+      "learning_rate": 0.000551867071639649,
+      "loss": 3.1609,
+      "step": 11191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5059942007064819,
+      "learning_rate": 0.0005518587069925442,
+      "loss": 3.2041,
+      "step": 11192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5030388832092285,
+      "learning_rate": 0.000551850341682092,
+      "loss": 3.1284,
+      "step": 11193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5302931666374207,
+      "learning_rate": 0.0005518419757083145,
+      "loss": 3.2589,
+      "step": 11194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5348262786865234,
+      "learning_rate": 0.0005518336090712337,
+      "loss": 3.3606,
+      "step": 11195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5146247148513794,
+      "learning_rate": 0.0005518252417708716,
+      "loss": 3.214,
+      "step": 11196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5075222253799438,
+      "learning_rate": 0.0005518168738072502,
+      "loss": 3.259,
+      "step": 11197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5404617190361023,
+      "learning_rate": 0.0005518085051803918,
+      "loss": 3.3196,
+      "step": 11198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5376437306404114,
+      "learning_rate": 0.0005518001358903182,
+      "loss": 3.0802,
+      "step": 11199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5305649638175964,
+      "learning_rate": 0.0005517917659370513,
+      "loss": 3.2274,
+      "step": 11200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.533257246017456,
+      "learning_rate": 0.0005517833953206135,
+      "loss": 3.1717,
+      "step": 11201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5252943634986877,
+      "learning_rate": 0.0005517750240410268,
+      "loss": 3.0743,
+      "step": 11202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4910027086734772,
+      "learning_rate": 0.000551766652098313,
+      "loss": 3.5248,
+      "step": 11203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5113924741744995,
+      "learning_rate": 0.0005517582794924943,
+      "loss": 3.1122,
+      "step": 11204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5466858744621277,
+      "learning_rate": 0.0005517499062235928,
+      "loss": 3.2754,
+      "step": 11205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5288564562797546,
+      "learning_rate": 0.0005517415322916305,
+      "loss": 3.3129,
+      "step": 11206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5409068465232849,
+      "learning_rate": 0.0005517331576966294,
+      "loss": 3.2236,
+      "step": 11207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.581696093082428,
+      "learning_rate": 0.0005517247824386117,
+      "loss": 3.5911,
+      "step": 11208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4891887903213501,
+      "learning_rate": 0.0005517164065175993,
+      "loss": 3.3352,
+      "step": 11209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5119558572769165,
+      "learning_rate": 0.0005517080299336143,
+      "loss": 3.1011,
+      "step": 11210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5541062355041504,
+      "learning_rate": 0.0005516996526866788,
+      "loss": 3.2571,
+      "step": 11211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5030975341796875,
+      "learning_rate": 0.0005516912747768149,
+      "loss": 3.1197,
+      "step": 11212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5093387365341187,
+      "learning_rate": 0.0005516828962040446,
+      "loss": 3.2713,
+      "step": 11213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.525682806968689,
+      "learning_rate": 0.00055167451696839,
+      "loss": 3.2748,
+      "step": 11214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.565834105014801,
+      "learning_rate": 0.0005516661370698732,
+      "loss": 3.0386,
+      "step": 11215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5513453483581543,
+      "learning_rate": 0.0005516577565085162,
+      "loss": 3.2004,
+      "step": 11216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5197407007217407,
+      "learning_rate": 0.0005516493752843411,
+      "loss": 3.1911,
+      "step": 11217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5003480315208435,
+      "learning_rate": 0.00055164099339737,
+      "loss": 3.3038,
+      "step": 11218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5131300687789917,
+      "learning_rate": 0.0005516326108476248,
+      "loss": 3.2264,
+      "step": 11219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.528200626373291,
+      "learning_rate": 0.0005516242276351279,
+      "loss": 3.1164,
+      "step": 11220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5266100764274597,
+      "learning_rate": 0.0005516158437599011,
+      "loss": 3.2815,
+      "step": 11221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5277635455131531,
+      "learning_rate": 0.0005516074592219666,
+      "loss": 3.2759,
+      "step": 11222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5396936535835266,
+      "learning_rate": 0.0005515990740213466,
+      "loss": 3.2679,
+      "step": 11223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5293653607368469,
+      "learning_rate": 0.000551590688158063,
+      "loss": 3.111,
+      "step": 11224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.539237380027771,
+      "learning_rate": 0.0005515823016321379,
+      "loss": 3.2653,
+      "step": 11225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5173228979110718,
+      "learning_rate": 0.0005515739144435934,
+      "loss": 3.2804,
+      "step": 11226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.569195032119751,
+      "learning_rate": 0.0005515655265924518,
+      "loss": 3.2265,
+      "step": 11227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5007268786430359,
+      "learning_rate": 0.0005515571380787348,
+      "loss": 3.4115,
+      "step": 11228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5260988473892212,
+      "learning_rate": 0.0005515487489024649,
+      "loss": 3.2786,
+      "step": 11229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5149096250534058,
+      "learning_rate": 0.0005515403590636639,
+      "loss": 3.382,
+      "step": 11230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5453810691833496,
+      "learning_rate": 0.000551531968562354,
+      "loss": 3.4435,
+      "step": 11231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5204135775566101,
+      "learning_rate": 0.0005515235773985573,
+      "loss": 3.2987,
+      "step": 11232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5300283432006836,
+      "learning_rate": 0.000551515185572296,
+      "loss": 3.264,
+      "step": 11233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5270791053771973,
+      "learning_rate": 0.000551506793083592,
+      "loss": 3.3112,
+      "step": 11234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5587947368621826,
+      "learning_rate": 0.0005514983999324676,
+      "loss": 3.2582,
+      "step": 11235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5228792428970337,
+      "learning_rate": 0.0005514900061189447,
+      "loss": 3.2563,
+      "step": 11236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.507921040058136,
+      "learning_rate": 0.0005514816116430456,
+      "loss": 3.4942,
+      "step": 11237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5501565337181091,
+      "learning_rate": 0.0005514732165047923,
+      "loss": 3.2608,
+      "step": 11238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5317707657814026,
+      "learning_rate": 0.000551464820704207,
+      "loss": 3.058,
+      "step": 11239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5452500581741333,
+      "learning_rate": 0.0005514564242413118,
+      "loss": 2.9297,
+      "step": 11240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5187104344367981,
+      "learning_rate": 0.0005514480271161287,
+      "loss": 3.1697,
+      "step": 11241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5452839732170105,
+      "learning_rate": 0.0005514396293286798,
+      "loss": 3.2642,
+      "step": 11242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5251529812812805,
+      "learning_rate": 0.0005514312308789875,
+      "loss": 3.0369,
+      "step": 11243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5177854299545288,
+      "learning_rate": 0.0005514228317670736,
+      "loss": 3.2987,
+      "step": 11244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5245817303657532,
+      "learning_rate": 0.0005514144319929604,
+      "loss": 3.1808,
+      "step": 11245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5251028537750244,
+      "learning_rate": 0.0005514060315566699,
+      "loss": 3.0786,
+      "step": 11246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.513904333114624,
+      "learning_rate": 0.0005513976304582243,
+      "loss": 3.1704,
+      "step": 11247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5236141681671143,
+      "learning_rate": 0.0005513892286976458,
+      "loss": 3.3471,
+      "step": 11248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5166938304901123,
+      "learning_rate": 0.0005513808262749564,
+      "loss": 3.2649,
+      "step": 11249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5092709064483643,
+      "learning_rate": 0.0005513724231901783,
+      "loss": 3.2016,
+      "step": 11250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5671427845954895,
+      "learning_rate": 0.0005513640194433336,
+      "loss": 3.0024,
+      "step": 11251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4978698194026947,
+      "learning_rate": 0.0005513556150344445,
+      "loss": 3.2126,
+      "step": 11252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5027517676353455,
+      "learning_rate": 0.000551347209963533,
+      "loss": 3.2967,
+      "step": 11253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5220968723297119,
+      "learning_rate": 0.0005513388042306214,
+      "loss": 3.4618,
+      "step": 11254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49995917081832886,
+      "learning_rate": 0.0005513303978357317,
+      "loss": 3.2012,
+      "step": 11255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5419469475746155,
+      "learning_rate": 0.0005513219907788861,
+      "loss": 3.134,
+      "step": 11256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5099649429321289,
+      "learning_rate": 0.0005513135830601068,
+      "loss": 3.1283,
+      "step": 11257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5278074741363525,
+      "learning_rate": 0.0005513051746794159,
+      "loss": 3.3438,
+      "step": 11258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5176796317100525,
+      "learning_rate": 0.0005512967656368355,
+      "loss": 3.0982,
+      "step": 11259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5078525543212891,
+      "learning_rate": 0.0005512883559323878,
+      "loss": 3.4361,
+      "step": 11260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4989159405231476,
+      "learning_rate": 0.0005512799455660948,
+      "loss": 3.2102,
+      "step": 11261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5165753960609436,
+      "learning_rate": 0.000551271534537979,
+      "loss": 3.2573,
+      "step": 11262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5196847319602966,
+      "learning_rate": 0.0005512631228480622,
+      "loss": 3.3376,
+      "step": 11263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49779194593429565,
+      "learning_rate": 0.0005512547104963668,
+      "loss": 3.5134,
+      "step": 11264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5193706750869751,
+      "learning_rate": 0.0005512462974829147,
+      "loss": 3.2742,
+      "step": 11265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5080481171607971,
+      "learning_rate": 0.0005512378838077283,
+      "loss": 3.1623,
+      "step": 11266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5434474945068359,
+      "learning_rate": 0.0005512294694708296,
+      "loss": 3.3106,
+      "step": 11267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4940679967403412,
+      "learning_rate": 0.0005512210544722408,
+      "loss": 3.2186,
+      "step": 11268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5241498947143555,
+      "learning_rate": 0.0005512126388119842,
+      "loss": 3.1095,
+      "step": 11269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5188072323799133,
+      "learning_rate": 0.0005512042224900818,
+      "loss": 3.3153,
+      "step": 11270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.492482990026474,
+      "learning_rate": 0.000551195805506556,
+      "loss": 3.2382,
+      "step": 11271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49130314588546753,
+      "learning_rate": 0.0005511873878614286,
+      "loss": 3.236,
+      "step": 11272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.517808735370636,
+      "learning_rate": 0.0005511789695547219,
+      "loss": 3.3371,
+      "step": 11273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5449178218841553,
+      "learning_rate": 0.0005511705505864582,
+      "loss": 3.0906,
+      "step": 11274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5152904987335205,
+      "learning_rate": 0.0005511621309566597,
+      "loss": 3.2692,
+      "step": 11275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.554939866065979,
+      "learning_rate": 0.0005511537106653484,
+      "loss": 3.3965,
+      "step": 11276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5003484487533569,
+      "learning_rate": 0.0005511452897125467,
+      "loss": 3.1112,
+      "step": 11277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5301329493522644,
+      "learning_rate": 0.0005511368680982765,
+      "loss": 3.3394,
+      "step": 11278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4994280934333801,
+      "learning_rate": 0.0005511284458225601,
+      "loss": 3.1309,
+      "step": 11279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5375825762748718,
+      "learning_rate": 0.0005511200228854197,
+      "loss": 3.2027,
+      "step": 11280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.532985270023346,
+      "learning_rate": 0.0005511115992868776,
+      "loss": 3.2824,
+      "step": 11281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48315703868865967,
+      "learning_rate": 0.0005511031750269558,
+      "loss": 3.2923,
+      "step": 11282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5368021726608276,
+      "learning_rate": 0.0005510947501056766,
+      "loss": 3.2745,
+      "step": 11283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5001078844070435,
+      "learning_rate": 0.0005510863245230621,
+      "loss": 3.2778,
+      "step": 11284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49867165088653564,
+      "learning_rate": 0.0005510778982791345,
+      "loss": 3.0507,
+      "step": 11285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5319203734397888,
+      "learning_rate": 0.0005510694713739162,
+      "loss": 3.1747,
+      "step": 11286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5031033158302307,
+      "learning_rate": 0.0005510610438074291,
+      "loss": 3.2447,
+      "step": 11287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4784884452819824,
+      "learning_rate": 0.0005510526155796956,
+      "loss": 3.4212,
+      "step": 11288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5477344989776611,
+      "learning_rate": 0.0005510441866907377,
+      "loss": 3.1972,
+      "step": 11289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49922096729278564,
+      "learning_rate": 0.000551035757140578,
+      "loss": 3.3556,
+      "step": 11290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4894360899925232,
+      "learning_rate": 0.0005510273269292382,
+      "loss": 3.2473,
+      "step": 11291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5014991760253906,
+      "learning_rate": 0.0005510188960567408,
+      "loss": 3.2122,
+      "step": 11292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5156797170639038,
+      "learning_rate": 0.0005510104645231079,
+      "loss": 3.3523,
+      "step": 11293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.510955274105072,
+      "learning_rate": 0.0005510020323283618,
+      "loss": 3.1125,
+      "step": 11294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5010038614273071,
+      "learning_rate": 0.0005509935994725245,
+      "loss": 3.2618,
+      "step": 11295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5552043318748474,
+      "learning_rate": 0.0005509851659556185,
+      "loss": 3.0575,
+      "step": 11296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48965537548065186,
+      "learning_rate": 0.0005509767317776659,
+      "loss": 3.1812,
+      "step": 11297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.521115243434906,
+      "learning_rate": 0.0005509682969386888,
+      "loss": 3.2456,
+      "step": 11298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.544768214225769,
+      "learning_rate": 0.0005509598614387096,
+      "loss": 3.3468,
+      "step": 11299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5138852596282959,
+      "learning_rate": 0.0005509514252777503,
+      "loss": 3.4115,
+      "step": 11300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5543062090873718,
+      "learning_rate": 0.0005509429884558334,
+      "loss": 3.3333,
+      "step": 11301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5786649584770203,
+      "learning_rate": 0.0005509345509729808,
+      "loss": 3.0643,
+      "step": 11302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5284539461135864,
+      "learning_rate": 0.000550926112829215,
+      "loss": 3.4618,
+      "step": 11303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5231096744537354,
+      "learning_rate": 0.000550917674024558,
+      "loss": 3.213,
+      "step": 11304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5191011428833008,
+      "learning_rate": 0.0005509092345590323,
+      "loss": 3.1695,
+      "step": 11305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5704165101051331,
+      "learning_rate": 0.0005509007944326598,
+      "loss": 3.2244,
+      "step": 11306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5259694457054138,
+      "learning_rate": 0.000550892353645463,
+      "loss": 3.0671,
+      "step": 11307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5437265634536743,
+      "learning_rate": 0.0005508839121974641,
+      "loss": 3.2964,
+      "step": 11308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5166676044464111,
+      "learning_rate": 0.0005508754700886851,
+      "loss": 3.16,
+      "step": 11309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5371460318565369,
+      "learning_rate": 0.0005508670273191485,
+      "loss": 3.0633,
+      "step": 11310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5051896572113037,
+      "learning_rate": 0.0005508585838888764,
+      "loss": 3.2427,
+      "step": 11311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6367805600166321,
+      "learning_rate": 0.000550850139797891,
+      "loss": 3.3534,
+      "step": 11312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5592988729476929,
+      "learning_rate": 0.0005508416950462146,
+      "loss": 3.0948,
+      "step": 11313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5036086440086365,
+      "learning_rate": 0.0005508332496338696,
+      "loss": 3.1853,
+      "step": 11314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49412891268730164,
+      "learning_rate": 0.000550824803560878,
+      "loss": 3.1354,
+      "step": 11315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5339794158935547,
+      "learning_rate": 0.0005508163568272622,
+      "loss": 3.2264,
+      "step": 11316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5197786092758179,
+      "learning_rate": 0.0005508079094330443,
+      "loss": 3.388,
+      "step": 11317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5386886596679688,
+      "learning_rate": 0.0005507994613782466,
+      "loss": 3.0609,
+      "step": 11318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5509684085845947,
+      "learning_rate": 0.0005507910126628915,
+      "loss": 3.131,
+      "step": 11319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5496838688850403,
+      "learning_rate": 0.0005507825632870011,
+      "loss": 3.4422,
+      "step": 11320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5109300017356873,
+      "learning_rate": 0.0005507741132505977,
+      "loss": 3.2252,
+      "step": 11321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48941949009895325,
+      "learning_rate": 0.0005507656625537035,
+      "loss": 3.2382,
+      "step": 11322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.545982301235199,
+      "learning_rate": 0.0005507572111963408,
+      "loss": 3.2181,
+      "step": 11323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5077481269836426,
+      "learning_rate": 0.000550748759178532,
+      "loss": 3.2736,
+      "step": 11324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5178395509719849,
+      "learning_rate": 0.0005507403065002991,
+      "loss": 3.249,
+      "step": 11325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5448976159095764,
+      "learning_rate": 0.0005507318531616645,
+      "loss": 3.2074,
+      "step": 11326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4983932375907898,
+      "learning_rate": 0.0005507233991626504,
+      "loss": 3.309,
+      "step": 11327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5093890428543091,
+      "learning_rate": 0.0005507149445032793,
+      "loss": 3.2261,
+      "step": 11328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4824458658695221,
+      "learning_rate": 0.0005507064891835732,
+      "loss": 3.1697,
+      "step": 11329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5757153034210205,
+      "learning_rate": 0.0005506980332035543,
+      "loss": 3.1732,
+      "step": 11330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5348081588745117,
+      "learning_rate": 0.0005506895765632452,
+      "loss": 2.9841,
+      "step": 11331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.526430070400238,
+      "learning_rate": 0.0005506811192626679,
+      "loss": 3.1298,
+      "step": 11332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5291725993156433,
+      "learning_rate": 0.0005506726613018449,
+      "loss": 3.2061,
+      "step": 11333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5114479064941406,
+      "learning_rate": 0.0005506642026807983,
+      "loss": 3.3891,
+      "step": 11334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544650673866272,
+      "learning_rate": 0.0005506557433995505,
+      "loss": 3.2341,
+      "step": 11335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5388392806053162,
+      "learning_rate": 0.0005506472834581236,
+      "loss": 3.1413,
+      "step": 11336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5075357556343079,
+      "learning_rate": 0.0005506388228565401,
+      "loss": 3.1682,
+      "step": 11337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5130732655525208,
+      "learning_rate": 0.000550630361594822,
+      "loss": 3.2566,
+      "step": 11338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5059003233909607,
+      "learning_rate": 0.0005506218996729919,
+      "loss": 3.1209,
+      "step": 11339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5154292583465576,
+      "learning_rate": 0.0005506134370910719,
+      "loss": 3.1587,
+      "step": 11340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5048404932022095,
+      "learning_rate": 0.0005506049738490844,
+      "loss": 3.2718,
+      "step": 11341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6062512397766113,
+      "learning_rate": 0.0005505965099470516,
+      "loss": 3.2463,
+      "step": 11342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5517935156822205,
+      "learning_rate": 0.0005505880453849958,
+      "loss": 3.2156,
+      "step": 11343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49947217106819153,
+      "learning_rate": 0.0005505795801629393,
+      "loss": 3.1784,
+      "step": 11344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4772196114063263,
+      "learning_rate": 0.0005505711142809043,
+      "loss": 3.2223,
+      "step": 11345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49474474787712097,
+      "learning_rate": 0.0005505626477389134,
+      "loss": 3.4315,
+      "step": 11346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.527829110622406,
+      "learning_rate": 0.0005505541805369888,
+      "loss": 3.2979,
+      "step": 11347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5038018822669983,
+      "learning_rate": 0.0005505457126751524,
+      "loss": 3.0596,
+      "step": 11348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5035970211029053,
+      "learning_rate": 0.000550537244153427,
+      "loss": 3.267,
+      "step": 11349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5311546325683594,
+      "learning_rate": 0.0005505287749718348,
+      "loss": 3.2461,
+      "step": 11350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5109077095985413,
+      "learning_rate": 0.0005505203051303978,
+      "loss": 3.3966,
+      "step": 11351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5348523855209351,
+      "learning_rate": 0.0005505118346291386,
+      "loss": 3.3345,
+      "step": 11352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5685103535652161,
+      "learning_rate": 0.0005505033634680795,
+      "loss": 3.1709,
+      "step": 11353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.533107578754425,
+      "learning_rate": 0.0005504948916472427,
+      "loss": 3.1036,
+      "step": 11354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5173946619033813,
+      "learning_rate": 0.0005504864191666506,
+      "loss": 3.3519,
+      "step": 11355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5562966465950012,
+      "learning_rate": 0.0005504779460263255,
+      "loss": 3.2072,
+      "step": 11356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5066145062446594,
+      "learning_rate": 0.0005504694722262897,
+      "loss": 3.1163,
+      "step": 11357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5840992331504822,
+      "learning_rate": 0.0005504609977665655,
+      "loss": 3.5294,
+      "step": 11358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5080714225769043,
+      "learning_rate": 0.0005504525226471752,
+      "loss": 3.2118,
+      "step": 11359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5689548254013062,
+      "learning_rate": 0.0005504440468681412,
+      "loss": 3.426,
+      "step": 11360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5232775211334229,
+      "learning_rate": 0.0005504355704294857,
+      "loss": 3.3447,
+      "step": 11361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.530229926109314,
+      "learning_rate": 0.0005504270933312311,
+      "loss": 3.1767,
+      "step": 11362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4973139762878418,
+      "learning_rate": 0.0005504186155733998,
+      "loss": 3.5491,
+      "step": 11363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5430886149406433,
+      "learning_rate": 0.0005504101371560141,
+      "loss": 3.0946,
+      "step": 11364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5411336421966553,
+      "learning_rate": 0.0005504016580790963,
+      "loss": 3.306,
+      "step": 11365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5088608860969543,
+      "learning_rate": 0.0005503931783426686,
+      "loss": 3.3453,
+      "step": 11366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5100044012069702,
+      "learning_rate": 0.0005503846979467535,
+      "loss": 3.177,
+      "step": 11367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49586084485054016,
+      "learning_rate": 0.0005503762168913733,
+      "loss": 3.3249,
+      "step": 11368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5439696311950684,
+      "learning_rate": 0.0005503677351765503,
+      "loss": 3.2196,
+      "step": 11369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5079057216644287,
+      "learning_rate": 0.0005503592528023069,
+      "loss": 3.3114,
+      "step": 11370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5126676559448242,
+      "learning_rate": 0.0005503507697686654,
+      "loss": 3.4438,
+      "step": 11371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5350159406661987,
+      "learning_rate": 0.0005503422860756482,
+      "loss": 3.2688,
+      "step": 11372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5482049584388733,
+      "learning_rate": 0.0005503338017232776,
+      "loss": 3.2605,
+      "step": 11373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5451943278312683,
+      "learning_rate": 0.0005503253167115759,
+      "loss": 3.1051,
+      "step": 11374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.610254168510437,
+      "learning_rate": 0.0005503168310405655,
+      "loss": 3.2985,
+      "step": 11375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5316271185874939,
+      "learning_rate": 0.0005503083447102688,
+      "loss": 3.3086,
+      "step": 11376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5323793292045593,
+      "learning_rate": 0.000550299857720708,
+      "loss": 3.2379,
+      "step": 11377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5014329552650452,
+      "learning_rate": 0.0005502913700719056,
+      "loss": 3.2504,
+      "step": 11378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.548941433429718,
+      "learning_rate": 0.0005502828817638838,
+      "loss": 3.0864,
+      "step": 11379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5736619234085083,
+      "learning_rate": 0.0005502743927966652,
+      "loss": 3.2717,
+      "step": 11380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5364370942115784,
+      "learning_rate": 0.0005502659031702721,
+      "loss": 3.0777,
+      "step": 11381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6484993100166321,
+      "learning_rate": 0.0005502574128847266,
+      "loss": 3.3747,
+      "step": 11382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5159845948219299,
+      "learning_rate": 0.0005502489219400512,
+      "loss": 3.377,
+      "step": 11383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48935675621032715,
+      "learning_rate": 0.0005502404303362684,
+      "loss": 3.2342,
+      "step": 11384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5167993307113647,
+      "learning_rate": 0.0005502319380734005,
+      "loss": 3.5881,
+      "step": 11385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5177289843559265,
+      "learning_rate": 0.0005502234451514697,
+      "loss": 3.392,
+      "step": 11386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5180376172065735,
+      "learning_rate": 0.0005502149515704985,
+      "loss": 2.9029,
+      "step": 11387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5249106884002686,
+      "learning_rate": 0.0005502064573305094,
+      "loss": 3.2103,
+      "step": 11388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5171917676925659,
+      "learning_rate": 0.0005501979624315247,
+      "loss": 3.109,
+      "step": 11389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5251325368881226,
+      "learning_rate": 0.0005501894668735666,
+      "loss": 3.3081,
+      "step": 11390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.518473207950592,
+      "learning_rate": 0.0005501809706566575,
+      "loss": 3.0674,
+      "step": 11391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5217577815055847,
+      "learning_rate": 0.00055017247378082,
+      "loss": 3.2443,
+      "step": 11392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.498160719871521,
+      "learning_rate": 0.0005501639762460764,
+      "loss": 3.1407,
+      "step": 11393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5372119545936584,
+      "learning_rate": 0.000550155478052449,
+      "loss": 3.3184,
+      "step": 11394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.550507664680481,
+      "learning_rate": 0.0005501469791999601,
+      "loss": 3.2081,
+      "step": 11395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5311989784240723,
+      "learning_rate": 0.0005501384796886323,
+      "loss": 3.3632,
+      "step": 11396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5252476334571838,
+      "learning_rate": 0.0005501299795184878,
+      "loss": 3.1677,
+      "step": 11397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5009608268737793,
+      "learning_rate": 0.0005501214786895491,
+      "loss": 3.2492,
+      "step": 11398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5100443959236145,
+      "learning_rate": 0.0005501129772018387,
+      "loss": 3.4272,
+      "step": 11399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5339254140853882,
+      "learning_rate": 0.0005501044750553788,
+      "loss": 3.2937,
+      "step": 11400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5647916793823242,
+      "learning_rate": 0.0005500959722501917,
+      "loss": 3.2429,
+      "step": 11401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48013341426849365,
+      "learning_rate": 0.0005500874687863,
+      "loss": 3.2782,
+      "step": 11402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5085362195968628,
+      "learning_rate": 0.000550078964663726,
+      "loss": 3.148,
+      "step": 11403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5075708031654358,
+      "learning_rate": 0.0005500704598824923,
+      "loss": 3.2333,
+      "step": 11404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5158693194389343,
+      "learning_rate": 0.0005500619544426211,
+      "loss": 3.276,
+      "step": 11405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5510281920433044,
+      "learning_rate": 0.0005500534483441347,
+      "loss": 3.0602,
+      "step": 11406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.492349237203598,
+      "learning_rate": 0.0005500449415870558,
+      "loss": 3.3378,
+      "step": 11407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.502224862575531,
+      "learning_rate": 0.0005500364341714066,
+      "loss": 2.86,
+      "step": 11408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5376916527748108,
+      "learning_rate": 0.0005500279260972095,
+      "loss": 3.1087,
+      "step": 11409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5204702615737915,
+      "learning_rate": 0.0005500194173644869,
+      "loss": 3.235,
+      "step": 11410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5284196734428406,
+      "learning_rate": 0.0005500109079732613,
+      "loss": 3.2917,
+      "step": 11411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5423819422721863,
+      "learning_rate": 0.0005500023979235552,
+      "loss": 3.2564,
+      "step": 11412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7053835988044739,
+      "learning_rate": 0.0005499938872153908,
+      "loss": 3.2405,
+      "step": 11413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5075749754905701,
+      "learning_rate": 0.0005499853758487907,
+      "loss": 3.2424,
+      "step": 11414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5351574420928955,
+      "learning_rate": 0.0005499768638237771,
+      "loss": 3.2685,
+      "step": 11415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49611949920654297,
+      "learning_rate": 0.0005499683511403727,
+      "loss": 3.2747,
+      "step": 11416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5512820482254028,
+      "learning_rate": 0.0005499598377985997,
+      "loss": 3.2599,
+      "step": 11417
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5308202505111694,
+      "learning_rate": 0.0005499513237984807,
+      "loss": 3.134,
+      "step": 11418
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5064956545829773,
+      "learning_rate": 0.0005499428091400378,
+      "loss": 3.2349,
+      "step": 11419
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5472379326820374,
+      "learning_rate": 0.0005499342938232938,
+      "loss": 3.1336,
+      "step": 11420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5334680676460266,
+      "learning_rate": 0.0005499257778482709,
+      "loss": 3.2006,
+      "step": 11421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5283361077308655,
+      "learning_rate": 0.0005499172612149916,
+      "loss": 3.5046,
+      "step": 11422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5037579536437988,
+      "learning_rate": 0.0005499087439234784,
+      "loss": 3.2281,
+      "step": 11423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5362147688865662,
+      "learning_rate": 0.0005499002259737536,
+      "loss": 3.2373,
+      "step": 11424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5454772710800171,
+      "learning_rate": 0.0005498917073658397,
+      "loss": 3.1557,
+      "step": 11425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5577088594436646,
+      "learning_rate": 0.0005498831880997591,
+      "loss": 3.3548,
+      "step": 11426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.688560962677002,
+      "learning_rate": 0.0005498746681755343,
+      "loss": 3.0268,
+      "step": 11427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5097750425338745,
+      "learning_rate": 0.0005498661475931879,
+      "loss": 3.438,
+      "step": 11428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5364587903022766,
+      "learning_rate": 0.000549857626352742,
+      "loss": 3.2104,
+      "step": 11429
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5260832905769348,
+      "learning_rate": 0.0005498491044542191,
+      "loss": 3.4208,
+      "step": 11430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49136224389076233,
+      "learning_rate": 0.0005498405818976418,
+      "loss": 3.4387,
+      "step": 11431
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5791602730751038,
+      "learning_rate": 0.0005498320586830326,
+      "loss": 3.4698,
+      "step": 11432
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5048266053199768,
+      "learning_rate": 0.0005498235348104137,
+      "loss": 3.1668,
+      "step": 11433
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5733696818351746,
+      "learning_rate": 0.0005498150102798078,
+      "loss": 3.261,
+      "step": 11434
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5145251750946045,
+      "learning_rate": 0.0005498064850912373,
+      "loss": 3.3864,
+      "step": 11435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5857335329055786,
+      "learning_rate": 0.0005497979592447244,
+      "loss": 3.1253,
+      "step": 11436
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5221096277236938,
+      "learning_rate": 0.0005497894327402918,
+      "loss": 3.1466,
+      "step": 11437
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4997885525226593,
+      "learning_rate": 0.0005497809055779619,
+      "loss": 3.1797,
+      "step": 11438
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5484127998352051,
+      "learning_rate": 0.0005497723777577573,
+      "loss": 3.1413,
+      "step": 11439
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5325717926025391,
+      "learning_rate": 0.0005497638492797002,
+      "loss": 3.3583,
+      "step": 11440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5324811339378357,
+      "learning_rate": 0.0005497553201438132,
+      "loss": 3.111,
+      "step": 11441
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5136557221412659,
+      "learning_rate": 0.0005497467903501188,
+      "loss": 3.2754,
+      "step": 11442
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5221689343452454,
+      "learning_rate": 0.0005497382598986394,
+      "loss": 3.3697,
+      "step": 11443
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5226287245750427,
+      "learning_rate": 0.0005497297287893975,
+      "loss": 3.5221,
+      "step": 11444
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5215665102005005,
+      "learning_rate": 0.0005497211970224156,
+      "loss": 3.2144,
+      "step": 11445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5560000538825989,
+      "learning_rate": 0.000549712664597716,
+      "loss": 3.0434,
+      "step": 11446
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5306897163391113,
+      "learning_rate": 0.0005497041315153215,
+      "loss": 3.2474,
+      "step": 11447
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5063650012016296,
+      "learning_rate": 0.0005496955977752541,
+      "loss": 3.2657,
+      "step": 11448
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5555657148361206,
+      "learning_rate": 0.0005496870633775367,
+      "loss": 3.2211,
+      "step": 11449
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5415468811988831,
+      "learning_rate": 0.0005496785283221917,
+      "loss": 3.3836,
+      "step": 11450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6452375054359436,
+      "learning_rate": 0.0005496699926092415,
+      "loss": 3.2545,
+      "step": 11451
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5206488966941833,
+      "learning_rate": 0.0005496614562387085,
+      "loss": 3.2159,
+      "step": 11452
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4810924232006073,
+      "learning_rate": 0.0005496529192106153,
+      "loss": 3.2473,
+      "step": 11453
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5150770545005798,
+      "learning_rate": 0.0005496443815249843,
+      "loss": 3.1742,
+      "step": 11454
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5128140449523926,
+      "learning_rate": 0.0005496358431818381,
+      "loss": 3.0848,
+      "step": 11455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5324752926826477,
+      "learning_rate": 0.0005496273041811991,
+      "loss": 3.066,
+      "step": 11456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5085709095001221,
+      "learning_rate": 0.0005496187645230898,
+      "loss": 3.216,
+      "step": 11457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.500564455986023,
+      "learning_rate": 0.0005496102242075328,
+      "loss": 3.2649,
+      "step": 11458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5043260455131531,
+      "learning_rate": 0.0005496016832345505,
+      "loss": 3.0378,
+      "step": 11459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.496572345495224,
+      "learning_rate": 0.0005495931416041654,
+      "loss": 3.1889,
+      "step": 11460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5041834712028503,
+      "learning_rate": 0.0005495845993163999,
+      "loss": 3.3908,
+      "step": 11461
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50934898853302,
+      "learning_rate": 0.0005495760563712768,
+      "loss": 3.3334,
+      "step": 11462
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5718405246734619,
+      "learning_rate": 0.0005495675127688181,
+      "loss": 3.1535,
+      "step": 11463
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48958179354667664,
+      "learning_rate": 0.0005495589685090468,
+      "loss": 3.0113,
+      "step": 11464
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5216549038887024,
+      "learning_rate": 0.0005495504235919852,
+      "loss": 3.3366,
+      "step": 11465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4936378300189972,
+      "learning_rate": 0.0005495418780176558,
+      "loss": 3.3092,
+      "step": 11466
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5067612528800964,
+      "learning_rate": 0.0005495333317860812,
+      "loss": 3.3701,
+      "step": 11467
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5563675165176392,
+      "learning_rate": 0.0005495247848972837,
+      "loss": 3.2486,
+      "step": 11468
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5286310911178589,
+      "learning_rate": 0.0005495162373512859,
+      "loss": 3.2765,
+      "step": 11469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5466699004173279,
+      "learning_rate": 0.0005495076891481104,
+      "loss": 3.1424,
+      "step": 11470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5529357194900513,
+      "learning_rate": 0.0005494991402877797,
+      "loss": 3.1241,
+      "step": 11471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5391454696655273,
+      "learning_rate": 0.0005494905907703164,
+      "loss": 3.1807,
+      "step": 11472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5178135633468628,
+      "learning_rate": 0.0005494820405957427,
+      "loss": 3.2315,
+      "step": 11473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5371395945549011,
+      "learning_rate": 0.0005494734897640815,
+      "loss": 3.3146,
+      "step": 11474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4930516183376312,
+      "learning_rate": 0.0005494649382753549,
+      "loss": 3.1017,
+      "step": 11475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.518638551235199,
+      "learning_rate": 0.0005494563861295858,
+      "loss": 3.098,
+      "step": 11476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515612781047821,
+      "learning_rate": 0.0005494478333267966,
+      "loss": 3.2221,
+      "step": 11477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5380525588989258,
+      "learning_rate": 0.0005494392798670099,
+      "loss": 3.3513,
+      "step": 11478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5430471301078796,
+      "learning_rate": 0.0005494307257502479,
+      "loss": 3.0634,
+      "step": 11479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5702585577964783,
+      "learning_rate": 0.0005494221709765335,
+      "loss": 3.1669,
+      "step": 11480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5334585905075073,
+      "learning_rate": 0.0005494136155458892,
+      "loss": 3.167,
+      "step": 11481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5363363027572632,
+      "learning_rate": 0.0005494050594583373,
+      "loss": 3.2432,
+      "step": 11482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5021289587020874,
+      "learning_rate": 0.0005493965027139005,
+      "loss": 3.4602,
+      "step": 11483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5429815649986267,
+      "learning_rate": 0.0005493879453126013,
+      "loss": 3.1518,
+      "step": 11484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48367413878440857,
+      "learning_rate": 0.0005493793872544625,
+      "loss": 3.1878,
+      "step": 11485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5013885498046875,
+      "learning_rate": 0.0005493708285395061,
+      "loss": 3.0526,
+      "step": 11486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5268964767456055,
+      "learning_rate": 0.000549362269167755,
+      "loss": 3.109,
+      "step": 11487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5033860802650452,
+      "learning_rate": 0.0005493537091392316,
+      "loss": 3.3082,
+      "step": 11488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.535088300704956,
+      "learning_rate": 0.0005493451484539586,
+      "loss": 3.2002,
+      "step": 11489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49352458119392395,
+      "learning_rate": 0.0005493365871119584,
+      "loss": 3.4408,
+      "step": 11490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5212691426277161,
+      "learning_rate": 0.0005493280251132538,
+      "loss": 3.0987,
+      "step": 11491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.500873327255249,
+      "learning_rate": 0.000549319462457867,
+      "loss": 3.3256,
+      "step": 11492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5535351037979126,
+      "learning_rate": 0.0005493108991458207,
+      "loss": 3.1447,
+      "step": 11493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.495970219373703,
+      "learning_rate": 0.0005493023351771376,
+      "loss": 3.2712,
+      "step": 11494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5443682670593262,
+      "learning_rate": 0.0005492937705518401,
+      "loss": 3.0544,
+      "step": 11495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5153055787086487,
+      "learning_rate": 0.0005492852052699507,
+      "loss": 3.3102,
+      "step": 11496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48461559414863586,
+      "learning_rate": 0.0005492766393314921,
+      "loss": 3.2207,
+      "step": 11497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.541000247001648,
+      "learning_rate": 0.0005492680727364868,
+      "loss": 3.2273,
+      "step": 11498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53602534532547,
+      "learning_rate": 0.0005492595054849572,
+      "loss": 3.2255,
+      "step": 11499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46997231245040894,
+      "learning_rate": 0.0005492509375769264,
+      "loss": 3.197,
+      "step": 11500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4768153429031372,
+      "learning_rate": 0.0005492423690124164,
+      "loss": 3.2706,
+      "step": 11501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5005391836166382,
+      "learning_rate": 0.0005492337997914499,
+      "loss": 3.0595,
+      "step": 11502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5174161195755005,
+      "learning_rate": 0.0005492252299140497,
+      "loss": 3.493,
+      "step": 11503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5118815898895264,
+      "learning_rate": 0.000549216659380238,
+      "loss": 3.2988,
+      "step": 11504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5185562372207642,
+      "learning_rate": 0.0005492080881900377,
+      "loss": 3.3501,
+      "step": 11505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5072935819625854,
+      "learning_rate": 0.0005491995163434712,
+      "loss": 3.4849,
+      "step": 11506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49706950783729553,
+      "learning_rate": 0.0005491909438405612,
+      "loss": 3.4126,
+      "step": 11507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5042622685432434,
+      "learning_rate": 0.0005491823706813303,
+      "loss": 3.1778,
+      "step": 11508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5073032379150391,
+      "learning_rate": 0.0005491737968658009,
+      "loss": 3.041,
+      "step": 11509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5095604658126831,
+      "learning_rate": 0.0005491652223939956,
+      "loss": 3.3084,
+      "step": 11510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5435943603515625,
+      "learning_rate": 0.0005491566472659372,
+      "loss": 3.1515,
+      "step": 11511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51333087682724,
+      "learning_rate": 0.0005491480714816479,
+      "loss": 3.2505,
+      "step": 11512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.528915286064148,
+      "learning_rate": 0.0005491394950411508,
+      "loss": 3.1406,
+      "step": 11513
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5109853744506836,
+      "learning_rate": 0.000549130917944468,
+      "loss": 3.2142,
+      "step": 11514
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5281078219413757,
+      "learning_rate": 0.0005491223401916225,
+      "loss": 3.0415,
+      "step": 11515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.506549596786499,
+      "learning_rate": 0.0005491137617826366,
+      "loss": 3.6062,
+      "step": 11516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5103456974029541,
+      "learning_rate": 0.000549105182717533,
+      "loss": 3.3239,
+      "step": 11517
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5172756910324097,
+      "learning_rate": 0.0005490966029963343,
+      "loss": 3.1329,
+      "step": 11518
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5389090776443481,
+      "learning_rate": 0.0005490880226190632,
+      "loss": 3.1909,
+      "step": 11519
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5944215059280396,
+      "learning_rate": 0.000549079441585742,
+      "loss": 3.2218,
+      "step": 11520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.525435745716095,
+      "learning_rate": 0.0005490708598963936,
+      "loss": 3.2495,
+      "step": 11521
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.511767566204071,
+      "learning_rate": 0.0005490622775510404,
+      "loss": 3.2823,
+      "step": 11522
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5037621855735779,
+      "learning_rate": 0.0005490536945497052,
+      "loss": 3.3753,
+      "step": 11523
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5477989315986633,
+      "learning_rate": 0.0005490451108924104,
+      "loss": 3.299,
+      "step": 11524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5245879292488098,
+      "learning_rate": 0.0005490365265791787,
+      "loss": 3.1515,
+      "step": 11525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48814061284065247,
+      "learning_rate": 0.0005490279416100328,
+      "loss": 3.154,
+      "step": 11526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5687476992607117,
+      "learning_rate": 0.0005490193559849953,
+      "loss": 3.1192,
+      "step": 11527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49493011832237244,
+      "learning_rate": 0.0005490107697040886,
+      "loss": 3.0917,
+      "step": 11528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5287251472473145,
+      "learning_rate": 0.0005490021827673355,
+      "loss": 3.2482,
+      "step": 11529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5165507197380066,
+      "learning_rate": 0.0005489935951747586,
+      "loss": 3.2542,
+      "step": 11530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48569759726524353,
+      "learning_rate": 0.0005489850069263804,
+      "loss": 3.2343,
+      "step": 11531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5181214809417725,
+      "learning_rate": 0.0005489764180222237,
+      "loss": 3.5893,
+      "step": 11532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5267981886863708,
+      "learning_rate": 0.0005489678284623109,
+      "loss": 3.2945,
+      "step": 11533
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5052371025085449,
+      "learning_rate": 0.0005489592382466649,
+      "loss": 3.2504,
+      "step": 11534
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5270872116088867,
+      "learning_rate": 0.0005489506473753082,
+      "loss": 3.1425,
+      "step": 11535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5012282133102417,
+      "learning_rate": 0.0005489420558482634,
+      "loss": 3.3208,
+      "step": 11536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5245894193649292,
+      "learning_rate": 0.0005489334636655529,
+      "loss": 3.1897,
+      "step": 11537
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5544889569282532,
+      "learning_rate": 0.0005489248708271999,
+      "loss": 3.4318,
+      "step": 11538
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5470820069313049,
+      "learning_rate": 0.0005489162773332265,
+      "loss": 3.0173,
+      "step": 11539
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5013363361358643,
+      "learning_rate": 0.0005489076831836555,
+      "loss": 3.3224,
+      "step": 11540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.514118492603302,
+      "learning_rate": 0.0005488990883785097,
+      "loss": 3.0869,
+      "step": 11541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5037225484848022,
+      "learning_rate": 0.0005488904929178115,
+      "loss": 3.1977,
+      "step": 11542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5562716722488403,
+      "learning_rate": 0.0005488818968015836,
+      "loss": 3.3027,
+      "step": 11543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4860294759273529,
+      "learning_rate": 0.0005488733000298488,
+      "loss": 3.2274,
+      "step": 11544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5283397436141968,
+      "learning_rate": 0.0005488647026026296,
+      "loss": 3.3057,
+      "step": 11545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5059245824813843,
+      "learning_rate": 0.0005488561045199487,
+      "loss": 3.3249,
+      "step": 11546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5175250172615051,
+      "learning_rate": 0.0005488475057818286,
+      "loss": 3.2407,
+      "step": 11547
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.505158007144928,
+      "learning_rate": 0.0005488389063882922,
+      "loss": 3.3951,
+      "step": 11548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48427003622055054,
+      "learning_rate": 0.0005488303063393619,
+      "loss": 3.2815,
+      "step": 11549
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49035337567329407,
+      "learning_rate": 0.0005488217056350605,
+      "loss": 3.3064,
+      "step": 11550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5698608756065369,
+      "learning_rate": 0.0005488131042754107,
+      "loss": 3.2583,
+      "step": 11551
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5469956994056702,
+      "learning_rate": 0.0005488045022604349,
+      "loss": 3.324,
+      "step": 11552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5337932109832764,
+      "learning_rate": 0.000548795899590156,
+      "loss": 3.0973,
+      "step": 11553
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5492282509803772,
+      "learning_rate": 0.0005487872962645966,
+      "loss": 3.218,
+      "step": 11554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5477828979492188,
+      "learning_rate": 0.0005487786922837793,
+      "loss": 3.224,
+      "step": 11555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4647465944290161,
+      "learning_rate": 0.0005487700876477268,
+      "loss": 3.1174,
+      "step": 11556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5192469954490662,
+      "learning_rate": 0.0005487614823564619,
+      "loss": 3.1721,
+      "step": 11557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5096564888954163,
+      "learning_rate": 0.000548752876410007,
+      "loss": 3.0154,
+      "step": 11558
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5024121999740601,
+      "learning_rate": 0.0005487442698083848,
+      "loss": 3.3643,
+      "step": 11559
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5501764416694641,
+      "learning_rate": 0.0005487356625516183,
+      "loss": 3.3475,
+      "step": 11560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5160467624664307,
+      "learning_rate": 0.0005487270546397299,
+      "loss": 3.1848,
+      "step": 11561
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5589426159858704,
+      "learning_rate": 0.0005487184460727422,
+      "loss": 2.9357,
+      "step": 11562
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5194653868675232,
+      "learning_rate": 0.000548709836850678,
+      "loss": 3.4209,
+      "step": 11563
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5185741782188416,
+      "learning_rate": 0.0005487012269735599,
+      "loss": 3.1228,
+      "step": 11564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4845937192440033,
+      "learning_rate": 0.0005486926164414108,
+      "loss": 3.3671,
+      "step": 11565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.561070442199707,
+      "learning_rate": 0.0005486840052542531,
+      "loss": 3.0674,
+      "step": 11566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5006464719772339,
+      "learning_rate": 0.0005486753934121095,
+      "loss": 3.2313,
+      "step": 11567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49923497438430786,
+      "learning_rate": 0.0005486667809150029,
+      "loss": 3.453,
+      "step": 11568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5308060050010681,
+      "learning_rate": 0.0005486581677629558,
+      "loss": 3.2375,
+      "step": 11569
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5206416249275208,
+      "learning_rate": 0.000548649553955991,
+      "loss": 3.1938,
+      "step": 11570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5626609325408936,
+      "learning_rate": 0.0005486409394941311,
+      "loss": 3.2765,
+      "step": 11571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5191032290458679,
+      "learning_rate": 0.0005486323243773988,
+      "loss": 3.3281,
+      "step": 11572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.503855288028717,
+      "learning_rate": 0.0005486237086058169,
+      "loss": 3.0798,
+      "step": 11573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5209179520606995,
+      "learning_rate": 0.0005486150921794079,
+      "loss": 3.1159,
+      "step": 11574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.514619767665863,
+      "learning_rate": 0.0005486064750981946,
+      "loss": 3.2129,
+      "step": 11575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5100206136703491,
+      "learning_rate": 0.0005485978573621997,
+      "loss": 3.2254,
+      "step": 11576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48534801602363586,
+      "learning_rate": 0.0005485892389714459,
+      "loss": 3.2878,
+      "step": 11577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5108082294464111,
+      "learning_rate": 0.000548580619925956,
+      "loss": 3.3976,
+      "step": 11578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5104441046714783,
+      "learning_rate": 0.0005485720002257524,
+      "loss": 3.0612,
+      "step": 11579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5448250770568848,
+      "learning_rate": 0.000548563379870858,
+      "loss": 3.1129,
+      "step": 11580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4960390329360962,
+      "learning_rate": 0.0005485547588612956,
+      "loss": 3.0571,
+      "step": 11581
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515153706073761,
+      "learning_rate": 0.0005485461371970878,
+      "loss": 3.1635,
+      "step": 11582
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5046236515045166,
+      "learning_rate": 0.0005485375148782572,
+      "loss": 3.1507,
+      "step": 11583
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5140905976295471,
+      "learning_rate": 0.0005485288919048266,
+      "loss": 3.2966,
+      "step": 11584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5744099617004395,
+      "learning_rate": 0.0005485202682768189,
+      "loss": 3.2489,
+      "step": 11585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49185919761657715,
+      "learning_rate": 0.0005485116439942566,
+      "loss": 3.2854,
+      "step": 11586
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5120055675506592,
+      "learning_rate": 0.0005485030190571623,
+      "loss": 3.2128,
+      "step": 11587
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5046406388282776,
+      "learning_rate": 0.000548494393465559,
+      "loss": 2.9364,
+      "step": 11588
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5551308393478394,
+      "learning_rate": 0.0005484857672194693,
+      "loss": 3.1919,
+      "step": 11589
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5424243807792664,
+      "learning_rate": 0.0005484771403189158,
+      "loss": 3.4059,
+      "step": 11590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5322892069816589,
+      "learning_rate": 0.0005484685127639215,
+      "loss": 3.1558,
+      "step": 11591
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5228599905967712,
+      "learning_rate": 0.0005484598845545089,
+      "loss": 3.4993,
+      "step": 11592
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5182601809501648,
+      "learning_rate": 0.0005484512556907007,
+      "loss": 3.2109,
+      "step": 11593
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5158981084823608,
+      "learning_rate": 0.0005484426261725198,
+      "loss": 3.4712,
+      "step": 11594
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.565634548664093,
+      "learning_rate": 0.0005484339959999887,
+      "loss": 3.5278,
+      "step": 11595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5130377411842346,
+      "learning_rate": 0.0005484253651731305,
+      "loss": 3.1502,
+      "step": 11596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5539126992225647,
+      "learning_rate": 0.0005484167336919675,
+      "loss": 3.3652,
+      "step": 11597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5684749484062195,
+      "learning_rate": 0.0005484081015565226,
+      "loss": 3.0461,
+      "step": 11598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5369781851768494,
+      "learning_rate": 0.0005483994687668187,
+      "loss": 3.3655,
+      "step": 11599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5240543484687805,
+      "learning_rate": 0.0005483908353228784,
+      "loss": 3.387,
+      "step": 11600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.520671010017395,
+      "learning_rate": 0.0005483822012247244,
+      "loss": 3.1823,
+      "step": 11601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4907285273075104,
+      "learning_rate": 0.0005483735664723795,
+      "loss": 3.2514,
+      "step": 11602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.519844651222229,
+      "learning_rate": 0.0005483649310658665,
+      "loss": 3.4297,
+      "step": 11603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5508689880371094,
+      "learning_rate": 0.0005483562950052079,
+      "loss": 3.2559,
+      "step": 11604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.595680832862854,
+      "learning_rate": 0.0005483476582904268,
+      "loss": 3.2321,
+      "step": 11605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4791205823421478,
+      "learning_rate": 0.0005483390209215456,
+      "loss": 3.0654,
+      "step": 11606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4992697238922119,
+      "learning_rate": 0.0005483303828985873,
+      "loss": 3.3038,
+      "step": 11607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5372535586357117,
+      "learning_rate": 0.0005483217442215745,
+      "loss": 3.349,
+      "step": 11608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5181151032447815,
+      "learning_rate": 0.0005483131048905301,
+      "loss": 3.1583,
+      "step": 11609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5052180886268616,
+      "learning_rate": 0.0005483044649054767,
+      "loss": 3.4163,
+      "step": 11610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5012765526771545,
+      "learning_rate": 0.0005482958242664373,
+      "loss": 3.4786,
+      "step": 11611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5442771911621094,
+      "learning_rate": 0.0005482871829734342,
+      "loss": 3.1562,
+      "step": 11612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5303670167922974,
+      "learning_rate": 0.0005482785410264907,
+      "loss": 3.2916,
+      "step": 11613
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5204977989196777,
+      "learning_rate": 0.0005482698984256291,
+      "loss": 3.2238,
+      "step": 11614
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4961937367916107,
+      "learning_rate": 0.0005482612551708725,
+      "loss": 3.3505,
+      "step": 11615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5263996124267578,
+      "learning_rate": 0.0005482526112622435,
+      "loss": 3.1029,
+      "step": 11616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5191607475280762,
+      "learning_rate": 0.0005482439666997648,
+      "loss": 3.1744,
+      "step": 11617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5280757546424866,
+      "learning_rate": 0.0005482353214834594,
+      "loss": 3.1277,
+      "step": 11618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.583758533000946,
+      "learning_rate": 0.0005482266756133498,
+      "loss": 3.2388,
+      "step": 11619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49690064787864685,
+      "learning_rate": 0.0005482180290894592,
+      "loss": 3.2565,
+      "step": 11620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4999449849128723,
+      "learning_rate": 0.0005482093819118098,
+      "loss": 3.1643,
+      "step": 11621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4750107228755951,
+      "learning_rate": 0.0005482007340804248,
+      "loss": 2.9841,
+      "step": 11622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5504389405250549,
+      "learning_rate": 0.0005481920855953268,
+      "loss": 3.0009,
+      "step": 11623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5175572037696838,
+      "learning_rate": 0.0005481834364565386,
+      "loss": 3.3435,
+      "step": 11624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5125237107276917,
+      "learning_rate": 0.000548174786664083,
+      "loss": 3.5311,
+      "step": 11625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.556337296962738,
+      "learning_rate": 0.0005481661362179827,
+      "loss": 3.2554,
+      "step": 11626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5328450202941895,
+      "learning_rate": 0.0005481574851182606,
+      "loss": 3.2114,
+      "step": 11627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5400496125221252,
+      "learning_rate": 0.0005481488333649395,
+      "loss": 3.2479,
+      "step": 11628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.529948890209198,
+      "learning_rate": 0.0005481401809580421,
+      "loss": 3.2482,
+      "step": 11629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5066934823989868,
+      "learning_rate": 0.0005481315278975911,
+      "loss": 3.4325,
+      "step": 11630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4977806806564331,
+      "learning_rate": 0.0005481228741836096,
+      "loss": 3.3261,
+      "step": 11631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.526506245136261,
+      "learning_rate": 0.0005481142198161201,
+      "loss": 3.1154,
+      "step": 11632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6684457659721375,
+      "learning_rate": 0.0005481055647951456,
+      "loss": 3.2972,
+      "step": 11633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49738237261772156,
+      "learning_rate": 0.0005480969091207086,
+      "loss": 3.2738,
+      "step": 11634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5682615637779236,
+      "learning_rate": 0.0005480882527928322,
+      "loss": 3.3088,
+      "step": 11635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5242875814437866,
+      "learning_rate": 0.0005480795958115391,
+      "loss": 3.2557,
+      "step": 11636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5072360038757324,
+      "learning_rate": 0.0005480709381768521,
+      "loss": 3.0437,
+      "step": 11637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5002800822257996,
+      "learning_rate": 0.000548062279888794,
+      "loss": 3.2534,
+      "step": 11638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5526995062828064,
+      "learning_rate": 0.0005480536209473874,
+      "loss": 3.0043,
+      "step": 11639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5290579199790955,
+      "learning_rate": 0.0005480449613526555,
+      "loss": 3.2353,
+      "step": 11640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5533804893493652,
+      "learning_rate": 0.0005480363011046208,
+      "loss": 3.3581,
+      "step": 11641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4938673973083496,
+      "learning_rate": 0.0005480276402033064,
+      "loss": 3.2501,
+      "step": 11642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5148259401321411,
+      "learning_rate": 0.0005480189786487348,
+      "loss": 3.3157,
+      "step": 11643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5525953769683838,
+      "learning_rate": 0.0005480103164409289,
+      "loss": 3.0652,
+      "step": 11644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5157591104507446,
+      "learning_rate": 0.0005480016535799117,
+      "loss": 3.2691,
+      "step": 11645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4905431866645813,
+      "learning_rate": 0.0005479929900657057,
+      "loss": 3.1825,
+      "step": 11646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49346473813056946,
+      "learning_rate": 0.000547984325898334,
+      "loss": 3.2634,
+      "step": 11647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5140097141265869,
+      "learning_rate": 0.0005479756610778194,
+      "loss": 3.1073,
+      "step": 11648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4972335398197174,
+      "learning_rate": 0.0005479669956041844,
+      "loss": 3.2337,
+      "step": 11649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5696159601211548,
+      "learning_rate": 0.0005479583294774522,
+      "loss": 3.2024,
+      "step": 11650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5201365947723389,
+      "learning_rate": 0.0005479496626976455,
+      "loss": 3.4275,
+      "step": 11651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5507444143295288,
+      "learning_rate": 0.000547940995264787,
+      "loss": 3.0116,
+      "step": 11652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5264685153961182,
+      "learning_rate": 0.0005479323271788997,
+      "loss": 3.2018,
+      "step": 11653
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5618388652801514,
+      "learning_rate": 0.0005479236584400065,
+      "loss": 3.2548,
+      "step": 11654
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5223483443260193,
+      "learning_rate": 0.0005479149890481299,
+      "loss": 3.2424,
+      "step": 11655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5244974493980408,
+      "learning_rate": 0.000547906319003293,
+      "loss": 3.221,
+      "step": 11656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5352416634559631,
+      "learning_rate": 0.0005478976483055185,
+      "loss": 3.1539,
+      "step": 11657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5786653757095337,
+      "learning_rate": 0.0005478889769548295,
+      "loss": 3.2923,
+      "step": 11658
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5236338376998901,
+      "learning_rate": 0.0005478803049512484,
+      "loss": 3.1554,
+      "step": 11659
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.529576301574707,
+      "learning_rate": 0.0005478716322947985,
+      "loss": 3.2727,
+      "step": 11660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5126217007637024,
+      "learning_rate": 0.0005478629589855022,
+      "loss": 3.3197,
+      "step": 11661
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5071096420288086,
+      "learning_rate": 0.0005478542850233827,
+      "loss": 3.4426,
+      "step": 11662
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5255385637283325,
+      "learning_rate": 0.0005478456104084627,
+      "loss": 3.2004,
+      "step": 11663
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5327421426773071,
+      "learning_rate": 0.0005478369351407651,
+      "loss": 3.1227,
+      "step": 11664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5336896181106567,
+      "learning_rate": 0.0005478282592203126,
+      "loss": 3.2945,
+      "step": 11665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4557580351829529,
+      "learning_rate": 0.0005478195826471282,
+      "loss": 3.2835,
+      "step": 11666
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5098157525062561,
+      "learning_rate": 0.0005478109054212349,
+      "loss": 3.3144,
+      "step": 11667
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5083901286125183,
+      "learning_rate": 0.0005478022275426551,
+      "loss": 3.1161,
+      "step": 11668
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5286469459533691,
+      "learning_rate": 0.000547793549011412,
+      "loss": 3.2938,
+      "step": 11669
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5404655933380127,
+      "learning_rate": 0.0005477848698275285,
+      "loss": 3.1669,
+      "step": 11670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5109778046607971,
+      "learning_rate": 0.0005477761899910272,
+      "loss": 3.2569,
+      "step": 11671
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234958529472351,
+      "learning_rate": 0.0005477675095019312,
+      "loss": 3.2914,
+      "step": 11672
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5528802871704102,
+      "learning_rate": 0.0005477588283602632,
+      "loss": 3.2845,
+      "step": 11673
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5386664867401123,
+      "learning_rate": 0.0005477501465660461,
+      "loss": 3.344,
+      "step": 11674
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5150508284568787,
+      "learning_rate": 0.0005477414641193028,
+      "loss": 3.3954,
+      "step": 11675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4998857378959656,
+      "learning_rate": 0.0005477327810200562,
+      "loss": 3.3447,
+      "step": 11676
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5428280234336853,
+      "learning_rate": 0.0005477240972683292,
+      "loss": 3.1212,
+      "step": 11677
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5053800940513611,
+      "learning_rate": 0.0005477154128641445,
+      "loss": 3.2313,
+      "step": 11678
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.586479663848877,
+      "learning_rate": 0.0005477067278075251,
+      "loss": 3.3651,
+      "step": 11679
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5042681694030762,
+      "learning_rate": 0.000547698042098494,
+      "loss": 3.2991,
+      "step": 11680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5273053646087646,
+      "learning_rate": 0.0005476893557370737,
+      "loss": 3.1918,
+      "step": 11681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5238834023475647,
+      "learning_rate": 0.0005476806687232874,
+      "loss": 3.3427,
+      "step": 11682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5385345220565796,
+      "learning_rate": 0.0005476719810571579,
+      "loss": 3.2813,
+      "step": 11683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4971548318862915,
+      "learning_rate": 0.000547663292738708,
+      "loss": 3.2153,
+      "step": 11684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5180023908615112,
+      "learning_rate": 0.0005476546037679608,
+      "loss": 3.3399,
+      "step": 11685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.510409414768219,
+      "learning_rate": 0.0005476459141449388,
+      "loss": 3.1909,
+      "step": 11686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.555425763130188,
+      "learning_rate": 0.0005476372238696653,
+      "loss": 3.1216,
+      "step": 11687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5270985960960388,
+      "learning_rate": 0.0005476285329421629,
+      "loss": 3.1847,
+      "step": 11688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5697065591812134,
+      "learning_rate": 0.0005476198413624548,
+      "loss": 3.1772,
+      "step": 11689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.514995813369751,
+      "learning_rate": 0.0005476111491305635,
+      "loss": 3.3174,
+      "step": 11690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5724436044692993,
+      "learning_rate": 0.0005476024562465121,
+      "loss": 3.3164,
+      "step": 11691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5155641436576843,
+      "learning_rate": 0.0005475937627103237,
+      "loss": 3.0523,
+      "step": 11692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4877888560295105,
+      "learning_rate": 0.0005475850685220208,
+      "loss": 3.2307,
+      "step": 11693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49294334650039673,
+      "learning_rate": 0.0005475763736816264,
+      "loss": 3.0084,
+      "step": 11694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4982193410396576,
+      "learning_rate": 0.0005475676781891636,
+      "loss": 3.3148,
+      "step": 11695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6280152797698975,
+      "learning_rate": 0.0005475589820446552,
+      "loss": 3.3086,
+      "step": 11696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5116281509399414,
+      "learning_rate": 0.000547550285248124,
+      "loss": 3.2566,
+      "step": 11697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5089260339736938,
+      "learning_rate": 0.000547541587799593,
+      "loss": 3.1644,
+      "step": 11698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5232623815536499,
+      "learning_rate": 0.0005475328896990851,
+      "loss": 3.2846,
+      "step": 11699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5016197562217712,
+      "learning_rate": 0.0005475241909466234,
+      "loss": 3.0528,
+      "step": 11700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5107477307319641,
+      "learning_rate": 0.0005475154915422304,
+      "loss": 3.0808,
+      "step": 11701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5307403206825256,
+      "learning_rate": 0.0005475067914859292,
+      "loss": 3.2766,
+      "step": 11702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.50713050365448,
+      "learning_rate": 0.0005474980907777428,
+      "loss": 3.3566,
+      "step": 11703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4886563718318939,
+      "learning_rate": 0.0005474893894176941,
+      "loss": 3.3369,
+      "step": 11704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5123024582862854,
+      "learning_rate": 0.000547480687405806,
+      "loss": 3.1893,
+      "step": 11705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.510797917842865,
+      "learning_rate": 0.0005474719847421015,
+      "loss": 3.0718,
+      "step": 11706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.558925449848175,
+      "learning_rate": 0.0005474632814266031,
+      "loss": 2.9562,
+      "step": 11707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5396509170532227,
+      "learning_rate": 0.0005474545774593343,
+      "loss": 3.4154,
+      "step": 11708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5025230050086975,
+      "learning_rate": 0.0005474458728403176,
+      "loss": 3.0728,
+      "step": 11709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5266216993331909,
+      "learning_rate": 0.0005474371675695762,
+      "loss": 3.4107,
+      "step": 11710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5457305312156677,
+      "learning_rate": 0.000547428461647133,
+      "loss": 3.1522,
+      "step": 11711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.522469162940979,
+      "learning_rate": 0.0005474197550730107,
+      "loss": 3.5021,
+      "step": 11712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4818090796470642,
+      "learning_rate": 0.0005474110478472325,
+      "loss": 3.0651,
+      "step": 11713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.520215630531311,
+      "learning_rate": 0.0005474023399698212,
+      "loss": 3.3468,
+      "step": 11714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5050302743911743,
+      "learning_rate": 0.0005473936314407996,
+      "loss": 3.144,
+      "step": 11715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5783846378326416,
+      "learning_rate": 0.0005473849222601909,
+      "loss": 3.2924,
+      "step": 11716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4962663948535919,
+      "learning_rate": 0.0005473762124280179,
+      "loss": 3.0763,
+      "step": 11717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48002344369888306,
+      "learning_rate": 0.0005473675019443036,
+      "loss": 3.1622,
+      "step": 11718
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5629715919494629,
+      "learning_rate": 0.0005473587908090709,
+      "loss": 3.351,
+      "step": 11719
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5525310039520264,
+      "learning_rate": 0.0005473500790223428,
+      "loss": 3.2598,
+      "step": 11720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47623834013938904,
+      "learning_rate": 0.000547341366584142,
+      "loss": 3.2051,
+      "step": 11721
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.546389639377594,
+      "learning_rate": 0.0005473326534944918,
+      "loss": 3.1717,
+      "step": 11722
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4834481477737427,
+      "learning_rate": 0.000547323939753415,
+      "loss": 3.2224,
+      "step": 11723
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5237901210784912,
+      "learning_rate": 0.0005473152253609345,
+      "loss": 3.1499,
+      "step": 11724
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5636874437332153,
+      "learning_rate": 0.0005473065103170733,
+      "loss": 3.1321,
+      "step": 11725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5102803707122803,
+      "learning_rate": 0.0005472977946218543,
+      "loss": 3.3097,
+      "step": 11726
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5490740537643433,
+      "learning_rate": 0.0005472890782753006,
+      "loss": 3.3188,
+      "step": 11727
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5862623453140259,
+      "learning_rate": 0.000547280361277435,
+      "loss": 3.1409,
+      "step": 11728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48724859952926636,
+      "learning_rate": 0.0005472716436282806,
+      "loss": 3.4519,
+      "step": 11729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5731695294380188,
+      "learning_rate": 0.0005472629253278601,
+      "loss": 3.2008,
+      "step": 11730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5652557611465454,
+      "learning_rate": 0.0005472542063761968,
+      "loss": 3.2627,
+      "step": 11731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49549639225006104,
+      "learning_rate": 0.0005472454867733134,
+      "loss": 3.2696,
+      "step": 11732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5207010507583618,
+      "learning_rate": 0.0005472367665192331,
+      "loss": 3.2412,
+      "step": 11733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5095499157905579,
+      "learning_rate": 0.0005472280456139786,
+      "loss": 3.066,
+      "step": 11734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5058441162109375,
+      "learning_rate": 0.000547219324057573,
+      "loss": 3.3581,
+      "step": 11735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.536139190196991,
+      "learning_rate": 0.0005472106018500394,
+      "loss": 3.2069,
+      "step": 11736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5336572527885437,
+      "learning_rate": 0.0005472018789914007,
+      "loss": 3.1781,
+      "step": 11737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5066656470298767,
+      "learning_rate": 0.0005471931554816797,
+      "loss": 3.0132,
+      "step": 11738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53529292345047,
+      "learning_rate": 0.0005471844313208995,
+      "loss": 3.2795,
+      "step": 11739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4876529276371002,
+      "learning_rate": 0.0005471757065090831,
+      "loss": 3.2349,
+      "step": 11740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5300586819648743,
+      "learning_rate": 0.0005471669810462534,
+      "loss": 3.1156,
+      "step": 11741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5383363962173462,
+      "learning_rate": 0.0005471582549324336,
+      "loss": 3.3422,
+      "step": 11742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5305922031402588,
+      "learning_rate": 0.0005471495281676464,
+      "loss": 3.1445,
+      "step": 11743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5225808620452881,
+      "learning_rate": 0.000547140800751915,
+      "loss": 3.3356,
+      "step": 11744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5275900959968567,
+      "learning_rate": 0.0005471320726852621,
+      "loss": 3.3428,
+      "step": 11745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.553205132484436,
+      "learning_rate": 0.000547123343967711,
+      "loss": 3.2428,
+      "step": 11746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5640978217124939,
+      "learning_rate": 0.0005471146145992847,
+      "loss": 3.0525,
+      "step": 11747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5505227446556091,
+      "learning_rate": 0.0005471058845800059,
+      "loss": 3.059,
+      "step": 11748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5284337997436523,
+      "learning_rate": 0.0005470971539098978,
+      "loss": 3.3115,
+      "step": 11749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5046818256378174,
+      "learning_rate": 0.0005470884225889834,
+      "loss": 3.2677,
+      "step": 11750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5719025135040283,
+      "learning_rate": 0.0005470796906172855,
+      "loss": 3.1408,
+      "step": 11751
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5198563933372498,
+      "learning_rate": 0.0005470709579948274,
+      "loss": 3.3498,
+      "step": 11752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5236194133758545,
+      "learning_rate": 0.0005470622247216319,
+      "loss": 3.255,
+      "step": 11753
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5159012675285339,
+      "learning_rate": 0.000547053490797722,
+      "loss": 3.4876,
+      "step": 11754
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4822523891925812,
+      "learning_rate": 0.0005470447562231209,
+      "loss": 3.2904,
+      "step": 11755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5354002118110657,
+      "learning_rate": 0.0005470360209978513,
+      "loss": 3.123,
+      "step": 11756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5133538246154785,
+      "learning_rate": 0.0005470272851219364,
+      "loss": 3.412,
+      "step": 11757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5027098655700684,
+      "learning_rate": 0.0005470185485953992,
+      "loss": 3.2522,
+      "step": 11758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5310078859329224,
+      "learning_rate": 0.0005470098114182627,
+      "loss": 3.2146,
+      "step": 11759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5589718818664551,
+      "learning_rate": 0.0005470010735905499,
+      "loss": 3.2976,
+      "step": 11760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5355909466743469,
+      "learning_rate": 0.0005469923351122837,
+      "loss": 3.2431,
+      "step": 11761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5391919016838074,
+      "learning_rate": 0.0005469835959834873,
+      "loss": 3.0428,
+      "step": 11762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5123609304428101,
+      "learning_rate": 0.0005469748562041837,
+      "loss": 3.2564,
+      "step": 11763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4937504827976227,
+      "learning_rate": 0.0005469661157743958,
+      "loss": 3.2483,
+      "step": 11764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5278680920600891,
+      "learning_rate": 0.0005469573746941467,
+      "loss": 3.2399,
+      "step": 11765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5318676233291626,
+      "learning_rate": 0.0005469486329634593,
+      "loss": 3.1255,
+      "step": 11766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5315147638320923,
+      "learning_rate": 0.0005469398905823569,
+      "loss": 3.3234,
+      "step": 11767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5593608617782593,
+      "learning_rate": 0.0005469311475508622,
+      "loss": 3.253,
+      "step": 11768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5003385543823242,
+      "learning_rate": 0.0005469224038689985,
+      "loss": 2.979,
+      "step": 11769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.530859649181366,
+      "learning_rate": 0.0005469136595367887,
+      "loss": 3.1236,
+      "step": 11770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5268099308013916,
+      "learning_rate": 0.0005469049145542558,
+      "loss": 3.4322,
+      "step": 11771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5292220115661621,
+      "learning_rate": 0.0005468961689214228,
+      "loss": 3.2846,
+      "step": 11772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5274138450622559,
+      "learning_rate": 0.0005468874226383128,
+      "loss": 3.1513,
+      "step": 11773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5348718762397766,
+      "learning_rate": 0.000546878675704949,
+      "loss": 3.2605,
+      "step": 11774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5165255665779114,
+      "learning_rate": 0.0005468699281213541,
+      "loss": 3.1522,
+      "step": 11775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5076016783714294,
+      "learning_rate": 0.0005468611798875515,
+      "loss": 3.2869,
+      "step": 11776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5093604922294617,
+      "learning_rate": 0.0005468524310035639,
+      "loss": 3.0047,
+      "step": 11777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5292362570762634,
+      "learning_rate": 0.0005468436814694146,
+      "loss": 3.1524,
+      "step": 11778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5151748657226562,
+      "learning_rate": 0.0005468349312851265,
+      "loss": 3.1051,
+      "step": 11779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5270185470581055,
+      "learning_rate": 0.0005468261804507228,
+      "loss": 3.2477,
+      "step": 11780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5115005970001221,
+      "learning_rate": 0.0005468174289662264,
+      "loss": 3.265,
+      "step": 11781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5105668902397156,
+      "learning_rate": 0.0005468086768316604,
+      "loss": 3.189,
+      "step": 11782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4958377778530121,
+      "learning_rate": 0.0005467999240470478,
+      "loss": 3.2936,
+      "step": 11783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5410537719726562,
+      "learning_rate": 0.0005467911706124117,
+      "loss": 3.2262,
+      "step": 11784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5103792548179626,
+      "learning_rate": 0.0005467824165277752,
+      "loss": 3.2767,
+      "step": 11785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.623610258102417,
+      "learning_rate": 0.0005467736617931613,
+      "loss": 3.2318,
+      "step": 11786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5323020219802856,
+      "learning_rate": 0.000546764906408593,
+      "loss": 3.1768,
+      "step": 11787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5189688205718994,
+      "learning_rate": 0.0005467561503740934,
+      "loss": 3.1885,
+      "step": 11788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5025799870491028,
+      "learning_rate": 0.0005467473936896857,
+      "loss": 3.3244,
+      "step": 11789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5281747579574585,
+      "learning_rate": 0.0005467386363553927,
+      "loss": 3.3322,
+      "step": 11790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5055098533630371,
+      "learning_rate": 0.0005467298783712378,
+      "loss": 3.1912,
+      "step": 11791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5109471678733826,
+      "learning_rate": 0.0005467211197372438,
+      "loss": 3.3548,
+      "step": 11792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49919557571411133,
+      "learning_rate": 0.0005467123604534338,
+      "loss": 3.2776,
+      "step": 11793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5183613300323486,
+      "learning_rate": 0.0005467036005198311,
+      "loss": 3.3152,
+      "step": 11794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49261006712913513,
+      "learning_rate": 0.0005466948399364584,
+      "loss": 3.0799,
+      "step": 11795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078513622283936,
+      "learning_rate": 0.0005466860787033391,
+      "loss": 3.2225,
+      "step": 11796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5038175582885742,
+      "learning_rate": 0.000546677316820496,
+      "loss": 3.3051,
+      "step": 11797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5619712471961975,
+      "learning_rate": 0.0005466685542879525,
+      "loss": 3.2478,
+      "step": 11798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4969761371612549,
+      "learning_rate": 0.0005466597911057314,
+      "loss": 3.2508,
+      "step": 11799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5520473122596741,
+      "learning_rate": 0.0005466510272738559,
+      "loss": 3.3678,
+      "step": 11800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5152010917663574,
+      "learning_rate": 0.000546642262792349,
+      "loss": 3.2858,
+      "step": 11801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5011783242225647,
+      "learning_rate": 0.000546633497661234,
+      "loss": 3.2719,
+      "step": 11802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5367514491081238,
+      "learning_rate": 0.0005466247318805336,
+      "loss": 3.3372,
+      "step": 11803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5518689155578613,
+      "learning_rate": 0.0005466159654502713,
+      "loss": 3.2038,
+      "step": 11804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5553038716316223,
+      "learning_rate": 0.00054660719837047,
+      "loss": 3.082,
+      "step": 11805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5008841753005981,
+      "learning_rate": 0.0005465984306411528,
+      "loss": 3.0853,
+      "step": 11806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48236528038978577,
+      "learning_rate": 0.0005465896622623427,
+      "loss": 3.2213,
+      "step": 11807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5906662940979004,
+      "learning_rate": 0.000546580893234063,
+      "loss": 3.1436,
+      "step": 11808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5112290382385254,
+      "learning_rate": 0.0005465721235563365,
+      "loss": 3.2082,
+      "step": 11809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5187637805938721,
+      "learning_rate": 0.0005465633532291867,
+      "loss": 3.2795,
+      "step": 11810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5141165852546692,
+      "learning_rate": 0.0005465545822526364,
+      "loss": 3.191,
+      "step": 11811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6304354071617126,
+      "learning_rate": 0.0005465458106267087,
+      "loss": 3.3437,
+      "step": 11812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5287776589393616,
+      "learning_rate": 0.0005465370383514269,
+      "loss": 3.1787,
+      "step": 11813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5059935450553894,
+      "learning_rate": 0.000546528265426814,
+      "loss": 3.3609,
+      "step": 11814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5362207293510437,
+      "learning_rate": 0.000546519491852893,
+      "loss": 3.1434,
+      "step": 11815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5879662036895752,
+      "learning_rate": 0.0005465107176296872,
+      "loss": 3.4093,
+      "step": 11816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5185505747795105,
+      "learning_rate": 0.0005465019427572195,
+      "loss": 3.2697,
+      "step": 11817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5067775845527649,
+      "learning_rate": 0.0005464931672355131,
+      "loss": 3.4235,
+      "step": 11818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5255836248397827,
+      "learning_rate": 0.0005464843910645913,
+      "loss": 3.3745,
+      "step": 11819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6073008179664612,
+      "learning_rate": 0.0005464756142444769,
+      "loss": 3.182,
+      "step": 11820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48989614844322205,
+      "learning_rate": 0.0005464668367751933,
+      "loss": 3.4873,
+      "step": 11821
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6135048270225525,
+      "learning_rate": 0.0005464580586567634,
+      "loss": 3.3022,
+      "step": 11822
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5023560523986816,
+      "learning_rate": 0.0005464492798892104,
+      "loss": 3.2013,
+      "step": 11823
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5405312180519104,
+      "learning_rate": 0.0005464405004725574,
+      "loss": 3.3605,
+      "step": 11824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5376933217048645,
+      "learning_rate": 0.0005464317204068276,
+      "loss": 3.1126,
+      "step": 11825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5398737788200378,
+      "learning_rate": 0.0005464229396920441,
+      "loss": 3.1471,
+      "step": 11826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5017811059951782,
+      "learning_rate": 0.0005464141583282299,
+      "loss": 3.3562,
+      "step": 11827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49591362476348877,
+      "learning_rate": 0.0005464053763154083,
+      "loss": 3.1727,
+      "step": 11828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5274151563644409,
+      "learning_rate": 0.0005463965936536024,
+      "loss": 3.0793,
+      "step": 11829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4952261745929718,
+      "learning_rate": 0.0005463878103428353,
+      "loss": 3.4623,
+      "step": 11830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5281954407691956,
+      "learning_rate": 0.0005463790263831301,
+      "loss": 3.3092,
+      "step": 11831
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5254419445991516,
+      "learning_rate": 0.0005463702417745099,
+      "loss": 3.333,
+      "step": 11832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5291889309883118,
+      "learning_rate": 0.0005463614565169979,
+      "loss": 3.2975,
+      "step": 11833
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5549447536468506,
+      "learning_rate": 0.0005463526706106173,
+      "loss": 3.3199,
+      "step": 11834
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5520808696746826,
+      "learning_rate": 0.0005463438840553912,
+      "loss": 3.1179,
+      "step": 11835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5752543807029724,
+      "learning_rate": 0.0005463350968513426,
+      "loss": 3.4448,
+      "step": 11836
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5210770964622498,
+      "learning_rate": 0.0005463263089984948,
+      "loss": 3.3781,
+      "step": 11837
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5394595265388489,
+      "learning_rate": 0.000546317520496871,
+      "loss": 3.0691,
+      "step": 11838
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5369167327880859,
+      "learning_rate": 0.0005463087313464942,
+      "loss": 3.299,
+      "step": 11839
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5101680755615234,
+      "learning_rate": 0.0005462999415473877,
+      "loss": 3.1371,
+      "step": 11840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4878643751144409,
+      "learning_rate": 0.0005462911510995744,
+      "loss": 3.277,
+      "step": 11841
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.513389527797699,
+      "learning_rate": 0.0005462823600030776,
+      "loss": 3.2535,
+      "step": 11842
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5329154133796692,
+      "learning_rate": 0.0005462735682579205,
+      "loss": 3.4121,
+      "step": 11843
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4917752146720886,
+      "learning_rate": 0.0005462647758641263,
+      "loss": 3.3547,
+      "step": 11844
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49476316571235657,
+      "learning_rate": 0.000546255982821718,
+      "loss": 3.0001,
+      "step": 11845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4978811740875244,
+      "learning_rate": 0.0005462471891307189,
+      "loss": 3.2179,
+      "step": 11846
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5276978611946106,
+      "learning_rate": 0.000546238394791152,
+      "loss": 3.4435,
+      "step": 11847
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5227974653244019,
+      "learning_rate": 0.0005462295998030406,
+      "loss": 3.4436,
+      "step": 11848
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5526250004768372,
+      "learning_rate": 0.0005462208041664079,
+      "loss": 3.4602,
+      "step": 11849
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5186265110969543,
+      "learning_rate": 0.0005462120078812769,
+      "loss": 3.1297,
+      "step": 11850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5635437965393066,
+      "learning_rate": 0.0005462032109476709,
+      "loss": 2.9613,
+      "step": 11851
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5280086398124695,
+      "learning_rate": 0.000546194413365613,
+      "loss": 3.1061,
+      "step": 11852
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5283723473548889,
+      "learning_rate": 0.0005461856151351264,
+      "loss": 3.1436,
+      "step": 11853
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5015434622764587,
+      "learning_rate": 0.0005461768162562342,
+      "loss": 3.307,
+      "step": 11854
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5084378719329834,
+      "learning_rate": 0.0005461680167289598,
+      "loss": 3.2405,
+      "step": 11855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5142874717712402,
+      "learning_rate": 0.0005461592165533261,
+      "loss": 3.2759,
+      "step": 11856
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5148090720176697,
+      "learning_rate": 0.0005461504157293563,
+      "loss": 2.9513,
+      "step": 11857
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5030063390731812,
+      "learning_rate": 0.0005461416142570739,
+      "loss": 3.3158,
+      "step": 11858
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4925903379917145,
+      "learning_rate": 0.0005461328121365018,
+      "loss": 3.3714,
+      "step": 11859
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5264309644699097,
+      "learning_rate": 0.0005461240093676632,
+      "loss": 3.271,
+      "step": 11860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5105083584785461,
+      "learning_rate": 0.0005461152059505814,
+      "loss": 3.2343,
+      "step": 11861
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48054060339927673,
+      "learning_rate": 0.0005461064018852794,
+      "loss": 3.1617,
+      "step": 11862
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5289759635925293,
+      "learning_rate": 0.0005460975971717805,
+      "loss": 3.2449,
+      "step": 11863
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5321834087371826,
+      "learning_rate": 0.000546088791810108,
+      "loss": 3.2072,
+      "step": 11864
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6084837913513184,
+      "learning_rate": 0.0005460799858002849,
+      "loss": 3.2506,
+      "step": 11865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5481894016265869,
+      "learning_rate": 0.0005460711791423344,
+      "loss": 3.1303,
+      "step": 11866
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5393480658531189,
+      "learning_rate": 0.0005460623718362799,
+      "loss": 3.2954,
+      "step": 11867
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5366723537445068,
+      "learning_rate": 0.0005460535638821444,
+      "loss": 3.1237,
+      "step": 11868
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.505320131778717,
+      "learning_rate": 0.0005460447552799512,
+      "loss": 3.2153,
+      "step": 11869
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5377161502838135,
+      "learning_rate": 0.0005460359460297235,
+      "loss": 3.1654,
+      "step": 11870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49337872862815857,
+      "learning_rate": 0.0005460271361314845,
+      "loss": 3.2592,
+      "step": 11871
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5618535280227661,
+      "learning_rate": 0.0005460183255852573,
+      "loss": 3.2642,
+      "step": 11872
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4871905446052551,
+      "learning_rate": 0.0005460095143910652,
+      "loss": 3.3339,
+      "step": 11873
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5062762498855591,
+      "learning_rate": 0.0005460007025489313,
+      "loss": 3.3657,
+      "step": 11874
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.530390739440918,
+      "learning_rate": 0.0005459918900588789,
+      "loss": 3.303,
+      "step": 11875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5175230503082275,
+      "learning_rate": 0.0005459830769209314,
+      "loss": 3.0723,
+      "step": 11876
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5422409176826477,
+      "learning_rate": 0.0005459742631351118,
+      "loss": 3.2174,
+      "step": 11877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5116949081420898,
+      "learning_rate": 0.0005459654487014431,
+      "loss": 3.0453,
+      "step": 11878
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49938473105430603,
+      "learning_rate": 0.0005459566336199488,
+      "loss": 3.021,
+      "step": 11879
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5277251601219177,
+      "learning_rate": 0.0005459478178906522,
+      "loss": 3.2052,
+      "step": 11880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5035902261734009,
+      "learning_rate": 0.0005459390015135762,
+      "loss": 3.2446,
+      "step": 11881
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5442661643028259,
+      "learning_rate": 0.0005459301844887444,
+      "loss": 3.1998,
+      "step": 11882
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5180203914642334,
+      "learning_rate": 0.0005459213668161797,
+      "loss": 3.4836,
+      "step": 11883
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5293683409690857,
+      "learning_rate": 0.0005459125484959054,
+      "loss": 3.3354,
+      "step": 11884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4947066307067871,
+      "learning_rate": 0.0005459037295279449,
+      "loss": 2.9769,
+      "step": 11885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5257008671760559,
+      "learning_rate": 0.0005458949099123214,
+      "loss": 3.3676,
+      "step": 11886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4895974397659302,
+      "learning_rate": 0.0005458860896490577,
+      "loss": 3.1573,
+      "step": 11887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56827712059021,
+      "learning_rate": 0.0005458772687381776,
+      "loss": 3.0959,
+      "step": 11888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5436473488807678,
+      "learning_rate": 0.000545868447179704,
+      "loss": 3.3533,
+      "step": 11889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5245484113693237,
+      "learning_rate": 0.0005458596249736604,
+      "loss": 3.3021,
+      "step": 11890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.555914044380188,
+      "learning_rate": 0.0005458508021200697,
+      "loss": 3.2511,
+      "step": 11891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5292685031890869,
+      "learning_rate": 0.0005458419786189552,
+      "loss": 3.3015,
+      "step": 11892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5257068276405334,
+      "learning_rate": 0.0005458331544703405,
+      "loss": 3.1262,
+      "step": 11893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5110155940055847,
+      "learning_rate": 0.0005458243296742485,
+      "loss": 3.2781,
+      "step": 11894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5501812100410461,
+      "learning_rate": 0.0005458155042307024,
+      "loss": 3.0971,
+      "step": 11895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5484451651573181,
+      "learning_rate": 0.0005458066781397257,
+      "loss": 3.0403,
+      "step": 11896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546119213104248,
+      "learning_rate": 0.0005457978514013415,
+      "loss": 3.3563,
+      "step": 11897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5238593220710754,
+      "learning_rate": 0.000545789024015573,
+      "loss": 3.3929,
+      "step": 11898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5623632669448853,
+      "learning_rate": 0.0005457801959824435,
+      "loss": 3.2798,
+      "step": 11899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5106022357940674,
+      "learning_rate": 0.0005457713673019764,
+      "loss": 3.1605,
+      "step": 11900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5641065835952759,
+      "learning_rate": 0.0005457625379741947,
+      "loss": 3.364,
+      "step": 11901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5487107634544373,
+      "learning_rate": 0.0005457537079991218,
+      "loss": 3.1397,
+      "step": 11902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5087454915046692,
+      "learning_rate": 0.000545744877376781,
+      "loss": 3.1786,
+      "step": 11903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5044289827346802,
+      "learning_rate": 0.0005457360461071953,
+      "loss": 3.2765,
+      "step": 11904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5639960169792175,
+      "learning_rate": 0.0005457272141903884,
+      "loss": 3.3457,
+      "step": 11905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5312978029251099,
+      "learning_rate": 0.0005457183816263831,
+      "loss": 3.1708,
+      "step": 11906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.562142550945282,
+      "learning_rate": 0.000545709548415203,
+      "loss": 3.4045,
+      "step": 11907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5207469463348389,
+      "learning_rate": 0.0005457007145568712,
+      "loss": 3.2396,
+      "step": 11908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.513844907283783,
+      "learning_rate": 0.0005456918800514109,
+      "loss": 3.2508,
+      "step": 11909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5175647735595703,
+      "learning_rate": 0.0005456830448988456,
+      "loss": 3.0191,
+      "step": 11910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5360447764396667,
+      "learning_rate": 0.0005456742090991984,
+      "loss": 3.2997,
+      "step": 11911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5037137866020203,
+      "learning_rate": 0.0005456653726524926,
+      "loss": 3.2219,
+      "step": 11912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5330907106399536,
+      "learning_rate": 0.0005456565355587515,
+      "loss": 3.4176,
+      "step": 11913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.511675238609314,
+      "learning_rate": 0.0005456476978179984,
+      "loss": 3.3674,
+      "step": 11914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5179082751274109,
+      "learning_rate": 0.0005456388594302565,
+      "loss": 3.3472,
+      "step": 11915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5486281514167786,
+      "learning_rate": 0.0005456300203955491,
+      "loss": 3.4349,
+      "step": 11916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.516790509223938,
+      "learning_rate": 0.0005456211807138996,
+      "loss": 2.9798,
+      "step": 11917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5260696411132812,
+      "learning_rate": 0.0005456123403853311,
+      "loss": 3.0199,
+      "step": 11918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4837647080421448,
+      "learning_rate": 0.0005456034994098671,
+      "loss": 3.19,
+      "step": 11919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5303565263748169,
+      "learning_rate": 0.0005455946577875304,
+      "loss": 3.3105,
+      "step": 11920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4975842535495758,
+      "learning_rate": 0.0005455858155183449,
+      "loss": 3.4884,
+      "step": 11921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5344429612159729,
+      "learning_rate": 0.0005455769726023336,
+      "loss": 3.1137,
+      "step": 11922
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5470472574234009,
+      "learning_rate": 0.0005455681290395199,
+      "loss": 3.0215,
+      "step": 11923
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49586766958236694,
+      "learning_rate": 0.0005455592848299269,
+      "loss": 3.1258,
+      "step": 11924
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49502456188201904,
+      "learning_rate": 0.000545550439973578,
+      "loss": 3.188,
+      "step": 11925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5571303963661194,
+      "learning_rate": 0.0005455415944704966,
+      "loss": 3.2725,
+      "step": 11926
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4836958944797516,
+      "learning_rate": 0.0005455327483207057,
+      "loss": 3.2533,
+      "step": 11927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49969571828842163,
+      "learning_rate": 0.0005455239015242289,
+      "loss": 3.3009,
+      "step": 11928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5281708240509033,
+      "learning_rate": 0.0005455150540810894,
+      "loss": 3.2464,
+      "step": 11929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5075505375862122,
+      "learning_rate": 0.0005455062059913106,
+      "loss": 3.2988,
+      "step": 11930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5302159190177917,
+      "learning_rate": 0.0005454973572549156,
+      "loss": 3.2058,
+      "step": 11931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5521146059036255,
+      "learning_rate": 0.0005454885078719277,
+      "loss": 3.3473,
+      "step": 11932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5224472880363464,
+      "learning_rate": 0.0005454796578423705,
+      "loss": 3.1948,
+      "step": 11933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5724395513534546,
+      "learning_rate": 0.000545470807166267,
+      "loss": 3.4027,
+      "step": 11934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5843696594238281,
+      "learning_rate": 0.0005454619558436407,
+      "loss": 3.277,
+      "step": 11935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5131762027740479,
+      "learning_rate": 0.0005454531038745148,
+      "loss": 3.3245,
+      "step": 11936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5358636975288391,
+      "learning_rate": 0.0005454442512589127,
+      "loss": 3.1341,
+      "step": 11937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5540893077850342,
+      "learning_rate": 0.0005454353979968576,
+      "loss": 3.2948,
+      "step": 11938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48805323243141174,
+      "learning_rate": 0.000545426544088373,
+      "loss": 3.2973,
+      "step": 11939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4978611171245575,
+      "learning_rate": 0.0005454176895334822,
+      "loss": 3.3075,
+      "step": 11940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49105244874954224,
+      "learning_rate": 0.0005454088343322083,
+      "loss": 3.202,
+      "step": 11941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4822368621826172,
+      "learning_rate": 0.0005453999784845747,
+      "loss": 3.2363,
+      "step": 11942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5346906185150146,
+      "learning_rate": 0.0005453911219906049,
+      "loss": 3.2,
+      "step": 11943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5279293060302734,
+      "learning_rate": 0.0005453822648503221,
+      "loss": 3.1051,
+      "step": 11944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5374941825866699,
+      "learning_rate": 0.0005453734070637496,
+      "loss": 3.2868,
+      "step": 11945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5096423625946045,
+      "learning_rate": 0.0005453645486309109,
+      "loss": 3.2491,
+      "step": 11946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5027688145637512,
+      "learning_rate": 0.000545355689551829,
+      "loss": 3.252,
+      "step": 11947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5206319689750671,
+      "learning_rate": 0.0005453468298265275,
+      "loss": 3.3398,
+      "step": 11948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5208166241645813,
+      "learning_rate": 0.0005453379694550297,
+      "loss": 3.3007,
+      "step": 11949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5166978240013123,
+      "learning_rate": 0.0005453291084373589,
+      "loss": 3.2176,
+      "step": 11950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5170909762382507,
+      "learning_rate": 0.0005453202467735384,
+      "loss": 3.1967,
+      "step": 11951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5434339642524719,
+      "learning_rate": 0.0005453113844635916,
+      "loss": 3.2238,
+      "step": 11952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4791112244129181,
+      "learning_rate": 0.0005453025215075419,
+      "loss": 3.1606,
+      "step": 11953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49351251125335693,
+      "learning_rate": 0.0005452936579054125,
+      "loss": 3.0225,
+      "step": 11954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5247668623924255,
+      "learning_rate": 0.0005452847936572268,
+      "loss": 3.0996,
+      "step": 11955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5701345801353455,
+      "learning_rate": 0.0005452759287630081,
+      "loss": 3.3449,
+      "step": 11956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5012340545654297,
+      "learning_rate": 0.00054526706322278,
+      "loss": 3.1187,
+      "step": 11957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5431010723114014,
+      "learning_rate": 0.0005452581970365655,
+      "loss": 3.3408,
+      "step": 11958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5214460492134094,
+      "learning_rate": 0.0005452493302043882,
+      "loss": 3.1243,
+      "step": 11959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49427273869514465,
+      "learning_rate": 0.0005452404627262713,
+      "loss": 3.08,
+      "step": 11960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5280971527099609,
+      "learning_rate": 0.0005452315946022383,
+      "loss": 3.253,
+      "step": 11961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5085782408714294,
+      "learning_rate": 0.0005452227258323124,
+      "loss": 3.1777,
+      "step": 11962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5629658102989197,
+      "learning_rate": 0.000545213856416517,
+      "loss": 3.2163,
+      "step": 11963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49250122904777527,
+      "learning_rate": 0.0005452049863548756,
+      "loss": 3.1596,
+      "step": 11964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5141192078590393,
+      "learning_rate": 0.0005451961156474113,
+      "loss": 3.182,
+      "step": 11965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4984816908836365,
+      "learning_rate": 0.0005451872442941478,
+      "loss": 3.0893,
+      "step": 11966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4933840334415436,
+      "learning_rate": 0.0005451783722951082,
+      "loss": 3.0771,
+      "step": 11967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.564106822013855,
+      "learning_rate": 0.000545169499650316,
+      "loss": 3.4349,
+      "step": 11968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4997556805610657,
+      "learning_rate": 0.0005451606263597945,
+      "loss": 3.2064,
+      "step": 11969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4757207930088043,
+      "learning_rate": 0.000545151752423567,
+      "loss": 3.1864,
+      "step": 11970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4952181577682495,
+      "learning_rate": 0.0005451428778416571,
+      "loss": 3.0965,
+      "step": 11971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5092923045158386,
+      "learning_rate": 0.0005451340026140879,
+      "loss": 3.2239,
+      "step": 11972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5279220342636108,
+      "learning_rate": 0.000545125126740883,
+      "loss": 3.4003,
+      "step": 11973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4941803216934204,
+      "learning_rate": 0.0005451162502220657,
+      "loss": 3.2384,
+      "step": 11974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5303682684898376,
+      "learning_rate": 0.0005451073730576594,
+      "loss": 3.373,
+      "step": 11975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.548751175403595,
+      "learning_rate": 0.0005450984952476874,
+      "loss": 3.1797,
+      "step": 11976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.509413480758667,
+      "learning_rate": 0.0005450896167921731,
+      "loss": 3.1192,
+      "step": 11977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5299882888793945,
+      "learning_rate": 0.00054508073769114,
+      "loss": 2.747,
+      "step": 11978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5232790112495422,
+      "learning_rate": 0.0005450718579446113,
+      "loss": 3.3651,
+      "step": 11979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5589876174926758,
+      "learning_rate": 0.0005450629775526106,
+      "loss": 3.2521,
+      "step": 11980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4836844205856323,
+      "learning_rate": 0.0005450540965151612,
+      "loss": 3.0705,
+      "step": 11981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5362475514411926,
+      "learning_rate": 0.0005450452148322864,
+      "loss": 3.1405,
+      "step": 11982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5328556895256042,
+      "learning_rate": 0.0005450363325040096,
+      "loss": 3.2575,
+      "step": 11983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5442496538162231,
+      "learning_rate": 0.0005450274495303544,
+      "loss": 3.1584,
+      "step": 11984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5376067757606506,
+      "learning_rate": 0.0005450185659113439,
+      "loss": 3.3907,
+      "step": 11985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5082165598869324,
+      "learning_rate": 0.0005450096816470017,
+      "loss": 3.1008,
+      "step": 11986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4951854944229126,
+      "learning_rate": 0.0005450007967373512,
+      "loss": 3.1832,
+      "step": 11987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5797527432441711,
+      "learning_rate": 0.0005449919111824157,
+      "loss": 3.1662,
+      "step": 11988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5416431427001953,
+      "learning_rate": 0.0005449830249822187,
+      "loss": 3.1162,
+      "step": 11989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5033136010169983,
+      "learning_rate": 0.0005449741381367836,
+      "loss": 3.1388,
+      "step": 11990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5002380609512329,
+      "learning_rate": 0.0005449652506461337,
+      "loss": 3.31,
+      "step": 11991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49132394790649414,
+      "learning_rate": 0.0005449563625102926,
+      "loss": 3.1882,
+      "step": 11992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4861402213573456,
+      "learning_rate": 0.0005449474737292834,
+      "loss": 3.4393,
+      "step": 11993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5278277397155762,
+      "learning_rate": 0.0005449385843031298,
+      "loss": 3.2842,
+      "step": 11994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5361764430999756,
+      "learning_rate": 0.000544929694231855,
+      "loss": 3.2113,
+      "step": 11995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5028777122497559,
+      "learning_rate": 0.0005449208035154827,
+      "loss": 3.294,
+      "step": 11996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5272602438926697,
+      "learning_rate": 0.000544911912154036,
+      "loss": 3.3859,
+      "step": 11997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5307137966156006,
+      "learning_rate": 0.0005449030201475385,
+      "loss": 3.1559,
+      "step": 11998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.510999858379364,
+      "learning_rate": 0.0005448941274960136,
+      "loss": 3.1829,
+      "step": 11999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5812216997146606,
+      "learning_rate": 0.0005448852341994846,
+      "loss": 3.1065,
+      "step": 12000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5590925812721252,
+      "learning_rate": 0.0005448763402579752,
+      "loss": 3.1302,
+      "step": 12001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5319423079490662,
+      "learning_rate": 0.0005448674456715085,
+      "loss": 2.961,
+      "step": 12002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5174445509910583,
+      "learning_rate": 0.0005448585504401082,
+      "loss": 2.9525,
+      "step": 12003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5351834893226624,
+      "learning_rate": 0.0005448496545637975,
+      "loss": 3.1183,
+      "step": 12004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5270566940307617,
+      "learning_rate": 0.0005448407580425999,
+      "loss": 3.2682,
+      "step": 12005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5064947009086609,
+      "learning_rate": 0.000544831860876539,
+      "loss": 3.3034,
+      "step": 12006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5744919180870056,
+      "learning_rate": 0.0005448229630656379,
+      "loss": 3.0479,
+      "step": 12007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5197696089744568,
+      "learning_rate": 0.0005448140646099204,
+      "loss": 3.0092,
+      "step": 12008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52765291929245,
+      "learning_rate": 0.0005448051655094098,
+      "loss": 3.1997,
+      "step": 12009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5932416915893555,
+      "learning_rate": 0.0005447962657641294,
+      "loss": 3.3712,
+      "step": 12010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5595377087593079,
+      "learning_rate": 0.0005447873653741027,
+      "loss": 3.0408,
+      "step": 12011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5186452269554138,
+      "learning_rate": 0.0005447784643393533,
+      "loss": 3.5143,
+      "step": 12012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5053780674934387,
+      "learning_rate": 0.0005447695626599045,
+      "loss": 3.2042,
+      "step": 12013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5170993208885193,
+      "learning_rate": 0.0005447606603357798,
+      "loss": 3.0966,
+      "step": 12014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5181581377983093,
+      "learning_rate": 0.0005447517573670026,
+      "loss": 3.1413,
+      "step": 12015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48989567160606384,
+      "learning_rate": 0.0005447428537535963,
+      "loss": 3.1706,
+      "step": 12016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5295213460922241,
+      "learning_rate": 0.0005447339494955845,
+      "loss": 3.2038,
+      "step": 12017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5177062749862671,
+      "learning_rate": 0.0005447250445929906,
+      "loss": 3.3209,
+      "step": 12018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5109267234802246,
+      "learning_rate": 0.0005447161390458381,
+      "loss": 3.2872,
+      "step": 12019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5133575201034546,
+      "learning_rate": 0.0005447072328541502,
+      "loss": 3.2706,
+      "step": 12020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5542067289352417,
+      "learning_rate": 0.0005446983260179506,
+      "loss": 3.3886,
+      "step": 12021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.537536084651947,
+      "learning_rate": 0.0005446894185372628,
+      "loss": 3.4365,
+      "step": 12022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5550865530967712,
+      "learning_rate": 0.0005446805104121101,
+      "loss": 3.3174,
+      "step": 12023
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5242879390716553,
+      "learning_rate": 0.000544671601642516,
+      "loss": 3.1944,
+      "step": 12024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5210665464401245,
+      "learning_rate": 0.000544662692228504,
+      "loss": 3.1458,
+      "step": 12025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48247066140174866,
+      "learning_rate": 0.0005446537821700977,
+      "loss": 3.376,
+      "step": 12026
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5202993154525757,
+      "learning_rate": 0.0005446448714673202,
+      "loss": 3.286,
+      "step": 12027
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5529611706733704,
+      "learning_rate": 0.0005446359601201953,
+      "loss": 3.0672,
+      "step": 12028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.536654531955719,
+      "learning_rate": 0.0005446270481287463,
+      "loss": 3.1053,
+      "step": 12029
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5165926814079285,
+      "learning_rate": 0.0005446181354929969,
+      "loss": 3.3346,
+      "step": 12030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5090608596801758,
+      "learning_rate": 0.0005446092222129703,
+      "loss": 3.1495,
+      "step": 12031
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5462985038757324,
+      "learning_rate": 0.00054460030828869,
+      "loss": 3.0608,
+      "step": 12032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48308515548706055,
+      "learning_rate": 0.0005445913937201797,
+      "loss": 2.9185,
+      "step": 12033
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5395585894584656,
+      "learning_rate": 0.0005445824785074627,
+      "loss": 3.3225,
+      "step": 12034
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4958081543445587,
+      "learning_rate": 0.0005445735626505624,
+      "loss": 3.2987,
+      "step": 12035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5074117183685303,
+      "learning_rate": 0.0005445646461495025,
+      "loss": 3.2805,
+      "step": 12036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5241624712944031,
+      "learning_rate": 0.0005445557290043064,
+      "loss": 3.3391,
+      "step": 12037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.520176112651825,
+      "learning_rate": 0.0005445468112149977,
+      "loss": 3.4373,
+      "step": 12038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5940264463424683,
+      "learning_rate": 0.0005445378927815995,
+      "loss": 3.0772,
+      "step": 12039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5212997794151306,
+      "learning_rate": 0.0005445289737041357,
+      "loss": 3.279,
+      "step": 12040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5169647336006165,
+      "learning_rate": 0.0005445200539826297,
+      "loss": 3.2505,
+      "step": 12041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5246007442474365,
+      "learning_rate": 0.0005445111336171048,
+      "loss": 3.4347,
+      "step": 12042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.524836003780365,
+      "learning_rate": 0.0005445022126075847,
+      "loss": 3.0207,
+      "step": 12043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.490573525428772,
+      "learning_rate": 0.0005444932909540928,
+      "loss": 3.2743,
+      "step": 12044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5303407907485962,
+      "learning_rate": 0.0005444843686566528,
+      "loss": 3.2337,
+      "step": 12045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5297519564628601,
+      "learning_rate": 0.0005444754457152878,
+      "loss": 3.3453,
+      "step": 12046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6097574830055237,
+      "learning_rate": 0.0005444665221300216,
+      "loss": 3.0906,
+      "step": 12047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5111755132675171,
+      "learning_rate": 0.0005444575979008777,
+      "loss": 3.4244,
+      "step": 12048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5661916136741638,
+      "learning_rate": 0.0005444486730278794,
+      "loss": 3.2577,
+      "step": 12049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5134075880050659,
+      "learning_rate": 0.0005444397475110504,
+      "loss": 3.1203,
+      "step": 12050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5311046242713928,
+      "learning_rate": 0.0005444308213504143,
+      "loss": 3.376,
+      "step": 12051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5159074068069458,
+      "learning_rate": 0.0005444218945459943,
+      "loss": 3.2704,
+      "step": 12052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5199549794197083,
+      "learning_rate": 0.000544412967097814,
+      "loss": 3.3863,
+      "step": 12053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5141023397445679,
+      "learning_rate": 0.000544404039005897,
+      "loss": 3.1942,
+      "step": 12054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5393960475921631,
+      "learning_rate": 0.000544395110270267,
+      "loss": 3.2088,
+      "step": 12055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5226369500160217,
+      "learning_rate": 0.0005443861808909472,
+      "loss": 3.2853,
+      "step": 12056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5066924691200256,
+      "learning_rate": 0.0005443772508679611,
+      "loss": 3.0143,
+      "step": 12057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5037077069282532,
+      "learning_rate": 0.0005443683202013325,
+      "loss": 3.4013,
+      "step": 12058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.534941554069519,
+      "learning_rate": 0.0005443593888910847,
+      "loss": 3.2482,
+      "step": 12059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5512545108795166,
+      "learning_rate": 0.0005443504569372414,
+      "loss": 3.177,
+      "step": 12060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5608600974082947,
+      "learning_rate": 0.000544341524339826,
+      "loss": 3.3516,
+      "step": 12061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5026220679283142,
+      "learning_rate": 0.0005443325910988619,
+      "loss": 3.2668,
+      "step": 12062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5782676935195923,
+      "learning_rate": 0.0005443236572143729,
+      "loss": 3.1726,
+      "step": 12063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5567470788955688,
+      "learning_rate": 0.0005443147226863824,
+      "loss": 3.1268,
+      "step": 12064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4905277192592621,
+      "learning_rate": 0.000544305787514914,
+      "loss": 3.0573,
+      "step": 12065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5269580483436584,
+      "learning_rate": 0.000544296851699991,
+      "loss": 3.2479,
+      "step": 12066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5122688412666321,
+      "learning_rate": 0.0005442879152416373,
+      "loss": 3.2095,
+      "step": 12067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4919382631778717,
+      "learning_rate": 0.0005442789781398761,
+      "loss": 3.3365,
+      "step": 12068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5227946639060974,
+      "learning_rate": 0.0005442700403947312,
+      "loss": 2.9657,
+      "step": 12069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5085049867630005,
+      "learning_rate": 0.0005442611020062259,
+      "loss": 3.358,
+      "step": 12070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5325589776039124,
+      "learning_rate": 0.0005442521629743839,
+      "loss": 3.1012,
+      "step": 12071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5010466575622559,
+      "learning_rate": 0.0005442432232992289,
+      "loss": 3.1746,
+      "step": 12072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5115364193916321,
+      "learning_rate": 0.0005442342829807841,
+      "loss": 3.1592,
+      "step": 12073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4746970236301422,
+      "learning_rate": 0.0005442253420190732,
+      "loss": 3.3611,
+      "step": 12074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5140778422355652,
+      "learning_rate": 0.0005442164004141196,
+      "loss": 3.2885,
+      "step": 12075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.480934202671051,
+      "learning_rate": 0.0005442074581659472,
+      "loss": 3.3753,
+      "step": 12076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4761304557323456,
+      "learning_rate": 0.0005441985152745794,
+      "loss": 3.2269,
+      "step": 12077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5257875323295593,
+      "learning_rate": 0.0005441895717400396,
+      "loss": 3.1432,
+      "step": 12078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5360661745071411,
+      "learning_rate": 0.0005441806275623515,
+      "loss": 3.1401,
+      "step": 12079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48718738555908203,
+      "learning_rate": 0.0005441716827415385,
+      "loss": 3.3443,
+      "step": 12080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4796687066555023,
+      "learning_rate": 0.0005441627372776244,
+      "loss": 3.2142,
+      "step": 12081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47105666995048523,
+      "learning_rate": 0.0005441537911706327,
+      "loss": 3.0479,
+      "step": 12082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5267301201820374,
+      "learning_rate": 0.0005441448444205868,
+      "loss": 3.3318,
+      "step": 12083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5082873106002808,
+      "learning_rate": 0.0005441358970275104,
+      "loss": 3.2415,
+      "step": 12084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5294603705406189,
+      "learning_rate": 0.000544126948991427,
+      "loss": 3.1729,
+      "step": 12085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.536239504814148,
+      "learning_rate": 0.0005441180003123603,
+      "loss": 3.1646,
+      "step": 12086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5099353194236755,
+      "learning_rate": 0.0005441090509903336,
+      "loss": 3.2599,
+      "step": 12087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5732430219650269,
+      "learning_rate": 0.0005441001010253707,
+      "loss": 3.2231,
+      "step": 12088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5284761786460876,
+      "learning_rate": 0.0005440911504174952,
+      "loss": 3.5178,
+      "step": 12089
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5436829924583435,
+      "learning_rate": 0.0005440821991667306,
+      "loss": 3.4769,
+      "step": 12090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5140102505683899,
+      "learning_rate": 0.0005440732472731004,
+      "loss": 3.2641,
+      "step": 12091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5060403347015381,
+      "learning_rate": 0.0005440642947366281,
+      "loss": 3.1533,
+      "step": 12092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5052133798599243,
+      "learning_rate": 0.0005440553415573375,
+      "loss": 3.3999,
+      "step": 12093
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6156617999076843,
+      "learning_rate": 0.0005440463877352522,
+      "loss": 3.3402,
+      "step": 12094
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5013222098350525,
+      "learning_rate": 0.0005440374332703956,
+      "loss": 3.1037,
+      "step": 12095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.515588641166687,
+      "learning_rate": 0.0005440284781627915,
+      "loss": 3.2543,
+      "step": 12096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49556267261505127,
+      "learning_rate": 0.0005440195224124633,
+      "loss": 3.2262,
+      "step": 12097
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.528473436832428,
+      "learning_rate": 0.0005440105660194345,
+      "loss": 3.1852,
+      "step": 12098
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5471101999282837,
+      "learning_rate": 0.0005440016089837289,
+      "loss": 3.2124,
+      "step": 12099
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5259342193603516,
+      "learning_rate": 0.0005439926513053701,
+      "loss": 3.318,
+      "step": 12100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4963459372520447,
+      "learning_rate": 0.0005439836929843815,
+      "loss": 3.2905,
+      "step": 12101
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5480161309242249,
+      "learning_rate": 0.0005439747340207869,
+      "loss": 3.1982,
+      "step": 12102
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4799657464027405,
+      "learning_rate": 0.0005439657744146098,
+      "loss": 3.4103,
+      "step": 12103
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4825492203235626,
+      "learning_rate": 0.0005439568141658738,
+      "loss": 3.2936,
+      "step": 12104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5499624609947205,
+      "learning_rate": 0.0005439478532746024,
+      "loss": 3.1542,
+      "step": 12105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5210323333740234,
+      "learning_rate": 0.0005439388917408194,
+      "loss": 3.364,
+      "step": 12106
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5199629664421082,
+      "learning_rate": 0.0005439299295645482,
+      "loss": 2.9365,
+      "step": 12107
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5032206773757935,
+      "learning_rate": 0.0005439209667458127,
+      "loss": 3.0098,
+      "step": 12108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5243465304374695,
+      "learning_rate": 0.0005439120032846361,
+      "loss": 3.1939,
+      "step": 12109
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4972172975540161,
+      "learning_rate": 0.0005439030391810424,
+      "loss": 3.5505,
+      "step": 12110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5937232971191406,
+      "learning_rate": 0.000543894074435055,
+      "loss": 3.0929,
+      "step": 12111
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5489984154701233,
+      "learning_rate": 0.0005438851090466975,
+      "loss": 3.2707,
+      "step": 12112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5671371221542358,
+      "learning_rate": 0.0005438761430159936,
+      "loss": 3.1847,
+      "step": 12113
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5630182027816772,
+      "learning_rate": 0.0005438671763429668,
+      "loss": 3.1924,
+      "step": 12114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5923741459846497,
+      "learning_rate": 0.0005438582090276408,
+      "loss": 3.3754,
+      "step": 12115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5446171164512634,
+      "learning_rate": 0.0005438492410700394,
+      "loss": 3.2884,
+      "step": 12116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5040573477745056,
+      "learning_rate": 0.0005438402724701857,
+      "loss": 3.4074,
+      "step": 12117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5180044174194336,
+      "learning_rate": 0.0005438313032281039,
+      "loss": 3.0427,
+      "step": 12118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5119923949241638,
+      "learning_rate": 0.0005438223333438173,
+      "loss": 3.0863,
+      "step": 12119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5041249394416809,
+      "learning_rate": 0.0005438133628173496,
+      "loss": 3.0335,
+      "step": 12120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5423731207847595,
+      "learning_rate": 0.0005438043916487245,
+      "loss": 3.2203,
+      "step": 12121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5120278000831604,
+      "learning_rate": 0.0005437954198379654,
+      "loss": 3.2297,
+      "step": 12122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5203986167907715,
+      "learning_rate": 0.0005437864473850962,
+      "loss": 3.1436,
+      "step": 12123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5062956213951111,
+      "learning_rate": 0.0005437774742901403,
+      "loss": 3.225,
+      "step": 12124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5282447934150696,
+      "learning_rate": 0.0005437685005531216,
+      "loss": 3.3557,
+      "step": 12125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5864943861961365,
+      "learning_rate": 0.0005437595261740635,
+      "loss": 3.3101,
+      "step": 12126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5011259913444519,
+      "learning_rate": 0.0005437505511529897,
+      "loss": 3.2331,
+      "step": 12127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5472565293312073,
+      "learning_rate": 0.0005437415754899239,
+      "loss": 3.4375,
+      "step": 12128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49552929401397705,
+      "learning_rate": 0.0005437325991848896,
+      "loss": 3.3217,
+      "step": 12129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5005829334259033,
+      "learning_rate": 0.0005437236222379107,
+      "loss": 3.3369,
+      "step": 12130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5018550753593445,
+      "learning_rate": 0.0005437146446490107,
+      "loss": 3.3297,
+      "step": 12131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5203503370285034,
+      "learning_rate": 0.000543705666418213,
+      "loss": 3.2202,
+      "step": 12132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.568457841873169,
+      "learning_rate": 0.0005436966875455417,
+      "loss": 3.3885,
+      "step": 12133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5276675224304199,
+      "learning_rate": 0.0005436877080310202,
+      "loss": 3.2712,
+      "step": 12134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5073103308677673,
+      "learning_rate": 0.000543678727874672,
+      "loss": 3.3254,
+      "step": 12135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48787346482276917,
+      "learning_rate": 0.0005436697470765212,
+      "loss": 3.3454,
+      "step": 12136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5986006259918213,
+      "learning_rate": 0.000543660765636591,
+      "loss": 3.2068,
+      "step": 12137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5651094317436218,
+      "learning_rate": 0.0005436517835549052,
+      "loss": 3.1003,
+      "step": 12138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49831265211105347,
+      "learning_rate": 0.0005436428008314876,
+      "loss": 3.2614,
+      "step": 12139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5459215044975281,
+      "learning_rate": 0.0005436338174663617,
+      "loss": 3.2675,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546776652336121,
+      "learning_rate": 0.0005436248334595512,
+      "loss": 3.3903,
+      "step": 12141
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.518644392490387,
+      "learning_rate": 0.0005436158488110798,
+      "loss": 3.2527,
+      "step": 12142
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5478095412254333,
+      "learning_rate": 0.0005436068635209712,
+      "loss": 3.0529,
+      "step": 12143
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.497527152299881,
+      "learning_rate": 0.0005435978775892489,
+      "loss": 3.2598,
+      "step": 12144
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5034574270248413,
+      "learning_rate": 0.0005435888910159367,
+      "loss": 3.1994,
+      "step": 12145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5118417739868164,
+      "learning_rate": 0.0005435799038010583,
+      "loss": 3.4717,
+      "step": 12146
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5360592007637024,
+      "learning_rate": 0.0005435709159446374,
+      "loss": 3.2304,
+      "step": 12147
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4851125478744507,
+      "learning_rate": 0.0005435619274466974,
+      "loss": 3.1554,
+      "step": 12148
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4884193539619446,
+      "learning_rate": 0.0005435529383072622,
+      "loss": 3.0128,
+      "step": 12149
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5373410582542419,
+      "learning_rate": 0.0005435439485263554,
+      "loss": 3.152,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5832005143165588,
+      "learning_rate": 0.0005435349581040007,
+      "loss": 3.1407,
+      "step": 12151
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5295130014419556,
+      "learning_rate": 0.0005435259670402219,
+      "loss": 3.1837,
+      "step": 12152
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5200740694999695,
+      "learning_rate": 0.0005435169753350425,
+      "loss": 3.1467,
+      "step": 12153
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5249675512313843,
+      "learning_rate": 0.0005435079829884862,
+      "loss": 3.1414,
+      "step": 12154
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5205420851707458,
+      "learning_rate": 0.0005434989900005769,
+      "loss": 3.291,
+      "step": 12155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4993976950645447,
+      "learning_rate": 0.000543489996371338,
+      "loss": 3.1936,
+      "step": 12156
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49298644065856934,
+      "learning_rate": 0.0005434810021007933,
+      "loss": 3.2909,
+      "step": 12157
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5064563751220703,
+      "learning_rate": 0.0005434720071889666,
+      "loss": 3.2605,
+      "step": 12158
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4936967194080353,
+      "learning_rate": 0.0005434630116358814,
+      "loss": 3.1549,
+      "step": 12159
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4993574917316437,
+      "learning_rate": 0.0005434540154415615,
+      "loss": 3.1781,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5252702236175537,
+      "learning_rate": 0.0005434450186060305,
+      "loss": 3.3435,
+      "step": 12161
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5163652896881104,
+      "learning_rate": 0.0005434360211293123,
+      "loss": 3.0923,
+      "step": 12162
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5049394965171814,
+      "learning_rate": 0.0005434270230114305,
+      "loss": 3.1294,
+      "step": 12163
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578566312789917,
+      "learning_rate": 0.0005434180242524086,
+      "loss": 3.3308,
+      "step": 12164
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5135256052017212,
+      "learning_rate": 0.0005434090248522706,
+      "loss": 3.1645,
+      "step": 12165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4964609146118164,
+      "learning_rate": 0.00054340002481104,
+      "loss": 3.3597,
+      "step": 12166
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5104126334190369,
+      "learning_rate": 0.0005433910241287407,
+      "loss": 2.9874,
+      "step": 12167
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6239469051361084,
+      "learning_rate": 0.0005433820228053962,
+      "loss": 3.2153,
+      "step": 12168
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4951239824295044,
+      "learning_rate": 0.0005433730208410303,
+      "loss": 3.0318,
+      "step": 12169
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.506628155708313,
+      "learning_rate": 0.0005433640182356666,
+      "loss": 3.3427,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6572966575622559,
+      "learning_rate": 0.0005433550149893291,
+      "loss": 3.2291,
+      "step": 12171
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5308460593223572,
+      "learning_rate": 0.0005433460111020413,
+      "loss": 3.1009,
+      "step": 12172
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5671391487121582,
+      "learning_rate": 0.0005433370065738268,
+      "loss": 3.1697,
+      "step": 12173
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5267753005027771,
+      "learning_rate": 0.0005433280014047095,
+      "loss": 3.2779,
+      "step": 12174
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5052096247673035,
+      "learning_rate": 0.0005433189955947131,
+      "loss": 3.2854,
+      "step": 12175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5118441581726074,
+      "learning_rate": 0.0005433099891438614,
+      "loss": 3.276,
+      "step": 12176
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.504606306552887,
+      "learning_rate": 0.0005433009820521779,
+      "loss": 3.3353,
+      "step": 12177
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5112560391426086,
+      "learning_rate": 0.0005432919743196865,
+      "loss": 3.2776,
+      "step": 12178
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5266879200935364,
+      "learning_rate": 0.0005432829659464107,
+      "loss": 3.3279,
+      "step": 12179
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5236815214157104,
+      "learning_rate": 0.0005432739569323745,
+      "loss": 3.1972,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5193027257919312,
+      "learning_rate": 0.0005432649472776015,
+      "loss": 3.1793,
+      "step": 12181
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5326626300811768,
+      "learning_rate": 0.0005432559369821156,
+      "loss": 3.2743,
+      "step": 12182
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5254206657409668,
+      "learning_rate": 0.0005432469260459402,
+      "loss": 3.4511,
+      "step": 12183
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5383450388908386,
+      "learning_rate": 0.0005432379144690992,
+      "loss": 3.1277,
+      "step": 12184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48813480138778687,
+      "learning_rate": 0.0005432289022516164,
+      "loss": 2.9519,
+      "step": 12185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5338407158851624,
+      "learning_rate": 0.0005432198893935155,
+      "loss": 3.1829,
+      "step": 12186
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5344692468643188,
+      "learning_rate": 0.0005432108758948201,
+      "loss": 3.2572,
+      "step": 12187
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5389733910560608,
+      "learning_rate": 0.0005432018617555542,
+      "loss": 3.1464,
+      "step": 12188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49581897258758545,
+      "learning_rate": 0.0005431928469757415,
+      "loss": 3.1573,
+      "step": 12189
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5306205749511719,
+      "learning_rate": 0.0005431838315554054,
+      "loss": 3.3115,
+      "step": 12190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5148359537124634,
+      "learning_rate": 0.0005431748154945701,
+      "loss": 3.2685,
+      "step": 12191
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625180602073669,
+      "learning_rate": 0.0005431657987932589,
+      "loss": 3.1569,
+      "step": 12192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48063063621520996,
+      "learning_rate": 0.000543156781451496,
+      "loss": 3.3465,
+      "step": 12193
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49699029326438904,
+      "learning_rate": 0.0005431477634693048,
+      "loss": 3.2766,
+      "step": 12194
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4957977533340454,
+      "learning_rate": 0.0005431387448467091,
+      "loss": 3.4118,
+      "step": 12195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5320689678192139,
+      "learning_rate": 0.000543129725583733,
+      "loss": 3.2341,
+      "step": 12196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4783723056316376,
+      "learning_rate": 0.0005431207056803999,
+      "loss": 3.1873,
+      "step": 12197
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5043808817863464,
+      "learning_rate": 0.0005431116851367336,
+      "loss": 3.0182,
+      "step": 12198
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5587702393531799,
+      "learning_rate": 0.000543102663952758,
+      "loss": 3.2452,
+      "step": 12199
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5152283310890198,
+      "learning_rate": 0.0005430936421284967,
+      "loss": 3.2618,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5216172337532043,
+      "learning_rate": 0.0005430846196639734,
+      "loss": 3.1672,
+      "step": 12201
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5034514665603638,
+      "learning_rate": 0.0005430755965592121,
+      "loss": 3.1348,
+      "step": 12202
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5430858731269836,
+      "learning_rate": 0.0005430665728142366,
+      "loss": 3.2768,
+      "step": 12203
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5175026059150696,
+      "learning_rate": 0.0005430575484290704,
+      "loss": 3.2739,
+      "step": 12204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5324364304542542,
+      "learning_rate": 0.0005430485234037373,
+      "loss": 3.3467,
+      "step": 12205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5634534955024719,
+      "learning_rate": 0.0005430394977382613,
+      "loss": 3.0756,
+      "step": 12206
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5331823229789734,
+      "learning_rate": 0.000543030471432666,
+      "loss": 3.3163,
+      "step": 12207
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5143170356750488,
+      "learning_rate": 0.0005430214444869752,
+      "loss": 3.16,
+      "step": 12208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5276603102684021,
+      "learning_rate": 0.0005430124169012127,
+      "loss": 3.182,
+      "step": 12209
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5391786694526672,
+      "learning_rate": 0.0005430033886754022,
+      "loss": 3.0836,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5072814226150513,
+      "learning_rate": 0.0005429943598095677,
+      "loss": 3.2794,
+      "step": 12211
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5004345774650574,
+      "learning_rate": 0.0005429853303037328,
+      "loss": 3.3006,
+      "step": 12212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5201953053474426,
+      "learning_rate": 0.0005429763001579211,
+      "loss": 3.2392,
+      "step": 12213
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6965551972389221,
+      "learning_rate": 0.0005429672693721567,
+      "loss": 3.0654,
+      "step": 12214
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5072376132011414,
+      "learning_rate": 0.0005429582379464634,
+      "loss": 3.4563,
+      "step": 12215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47854286432266235,
+      "learning_rate": 0.0005429492058808647,
+      "loss": 3.2517,
+      "step": 12216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5493162870407104,
+      "learning_rate": 0.0005429401731753846,
+      "loss": 3.3376,
+      "step": 12217
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5450829863548279,
+      "learning_rate": 0.0005429311398300468,
+      "loss": 3.2878,
+      "step": 12218
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4950675368309021,
+      "learning_rate": 0.0005429221058448752,
+      "loss": 3.0321,
+      "step": 12219
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5706702470779419,
+      "learning_rate": 0.0005429130712198935,
+      "loss": 3.4247,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4959094226360321,
+      "learning_rate": 0.0005429040359551256,
+      "loss": 3.1983,
+      "step": 12221
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5183280110359192,
+      "learning_rate": 0.0005428950000505951,
+      "loss": 3.0507,
+      "step": 12222
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4983315169811249,
+      "learning_rate": 0.0005428859635063258,
+      "loss": 3.3056,
+      "step": 12223
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5146493315696716,
+      "learning_rate": 0.0005428769263223418,
+      "loss": 3.3479,
+      "step": 12224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4943070709705353,
+      "learning_rate": 0.0005428678884986667,
+      "loss": 3.1996,
+      "step": 12225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49658817052841187,
+      "learning_rate": 0.0005428588500353243,
+      "loss": 3.1671,
+      "step": 12226
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5471540689468384,
+      "learning_rate": 0.0005428498109323384,
+      "loss": 3.1609,
+      "step": 12227
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5084481835365295,
+      "learning_rate": 0.0005428407711897329,
+      "loss": 3.2232,
+      "step": 12228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5170078277587891,
+      "learning_rate": 0.0005428317308075315,
+      "loss": 3.3208,
+      "step": 12229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4958844780921936,
+      "learning_rate": 0.0005428226897857581,
+      "loss": 3.4691,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5121991634368896,
+      "learning_rate": 0.0005428136481244363,
+      "loss": 3.3307,
+      "step": 12231
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5569097399711609,
+      "learning_rate": 0.0005428046058235903,
+      "loss": 3.2827,
+      "step": 12232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5006715655326843,
+      "learning_rate": 0.0005427955628832436,
+      "loss": 3.4895,
+      "step": 12233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5267584919929504,
+      "learning_rate": 0.0005427865193034202,
+      "loss": 3.0273,
+      "step": 12234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5398766994476318,
+      "learning_rate": 0.0005427774750841437,
+      "loss": 3.2227,
+      "step": 12235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49153032898902893,
+      "learning_rate": 0.0005427684302254381,
+      "loss": 3.2155,
+      "step": 12236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55174720287323,
+      "learning_rate": 0.0005427593847273272,
+      "loss": 3.3676,
+      "step": 12237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5181941986083984,
+      "learning_rate": 0.0005427503385898348,
+      "loss": 3.2572,
+      "step": 12238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5273623466491699,
+      "learning_rate": 0.0005427412918129847,
+      "loss": 3.0727,
+      "step": 12239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5499759912490845,
+      "learning_rate": 0.0005427322443968007,
+      "loss": 3.2162,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4892823100090027,
+      "learning_rate": 0.0005427231963413069,
+      "loss": 3.2164,
+      "step": 12241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49758782982826233,
+      "learning_rate": 0.0005427141476465266,
+      "loss": 3.1819,
+      "step": 12242
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5292083024978638,
+      "learning_rate": 0.0005427050983124842,
+      "loss": 3.241,
+      "step": 12243
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5434751510620117,
+      "learning_rate": 0.0005426960483392031,
+      "loss": 3.2821,
+      "step": 12244
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5042057037353516,
+      "learning_rate": 0.0005426869977267075,
+      "loss": 3.2146,
+      "step": 12245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5056325197219849,
+      "learning_rate": 0.000542677946475021,
+      "loss": 3.3742,
+      "step": 12246
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5204561352729797,
+      "learning_rate": 0.0005426688945841674,
+      "loss": 3.2397,
+      "step": 12247
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5179333686828613,
+      "learning_rate": 0.0005426598420541707,
+      "loss": 3.2824,
+      "step": 12248
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5086974501609802,
+      "learning_rate": 0.0005426507888850547,
+      "loss": 3.2344,
+      "step": 12249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49645212292671204,
+      "learning_rate": 0.0005426417350768431,
+      "loss": 3.0371,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4945175051689148,
+      "learning_rate": 0.00054263268062956,
+      "loss": 3.2515,
+      "step": 12251
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5622349381446838,
+      "learning_rate": 0.0005426236255432291,
+      "loss": 3.3922,
+      "step": 12252
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5245631337165833,
+      "learning_rate": 0.0005426145698178742,
+      "loss": 3.1202,
+      "step": 12253
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5517042279243469,
+      "learning_rate": 0.0005426055134535192,
+      "loss": 3.0991,
+      "step": 12254
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5217844843864441,
+      "learning_rate": 0.000542596456450188,
+      "loss": 3.2649,
+      "step": 12255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47473424673080444,
+      "learning_rate": 0.0005425873988079045,
+      "loss": 3.2014,
+      "step": 12256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5011487007141113,
+      "learning_rate": 0.0005425783405266924,
+      "loss": 3.3579,
+      "step": 12257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5660378336906433,
+      "learning_rate": 0.0005425692816065757,
+      "loss": 3.3866,
+      "step": 12258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5503248572349548,
+      "learning_rate": 0.0005425602220475781,
+      "loss": 3.3629,
+      "step": 12259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5156776309013367,
+      "learning_rate": 0.0005425511618497237,
+      "loss": 3.1339,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5958825349807739,
+      "learning_rate": 0.0005425421010130361,
+      "loss": 3.1596,
+      "step": 12261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5076464414596558,
+      "learning_rate": 0.0005425330395375392,
+      "loss": 3.287,
+      "step": 12262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4932123124599457,
+      "learning_rate": 0.0005425239774232572,
+      "loss": 3.3516,
+      "step": 12263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5037664175033569,
+      "learning_rate": 0.0005425149146702135,
+      "loss": 3.1989,
+      "step": 12264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5109959244728088,
+      "learning_rate": 0.0005425058512784323,
+      "loss": 3.3902,
+      "step": 12265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49323901534080505,
+      "learning_rate": 0.0005424967872479373,
+      "loss": 3.4398,
+      "step": 12266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49260270595550537,
+      "learning_rate": 0.0005424877225787525,
+      "loss": 3.293,
+      "step": 12267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5285913348197937,
+      "learning_rate": 0.0005424786572709015,
+      "loss": 3.452,
+      "step": 12268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5172354578971863,
+      "learning_rate": 0.0005424695913244087,
+      "loss": 3.3493,
+      "step": 12269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5332759022712708,
+      "learning_rate": 0.0005424605247392975,
+      "loss": 3.1187,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5140502452850342,
+      "learning_rate": 0.0005424514575155919,
+      "loss": 3.3081,
+      "step": 12271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49481186270713806,
+      "learning_rate": 0.0005424423896533159,
+      "loss": 3.1223,
+      "step": 12272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5081273913383484,
+      "learning_rate": 0.0005424333211524932,
+      "loss": 3.1317,
+      "step": 12273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4956912100315094,
+      "learning_rate": 0.0005424242520131479,
+      "loss": 3.4277,
+      "step": 12274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5238935947418213,
+      "learning_rate": 0.0005424151822353038,
+      "loss": 3.2959,
+      "step": 12275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5190920233726501,
+      "learning_rate": 0.0005424061118189847,
+      "loss": 3.226,
+      "step": 12276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5076124668121338,
+      "learning_rate": 0.0005423970407642145,
+      "loss": 3.2697,
+      "step": 12277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4859389066696167,
+      "learning_rate": 0.0005423879690710172,
+      "loss": 3.1284,
+      "step": 12278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5037886500358582,
+      "learning_rate": 0.0005423788967394166,
+      "loss": 3.3524,
+      "step": 12279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5088861584663391,
+      "learning_rate": 0.0005423698237694368,
+      "loss": 3.3835,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4993892312049866,
+      "learning_rate": 0.0005423607501611013,
+      "loss": 3.1905,
+      "step": 12281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5249192714691162,
+      "learning_rate": 0.0005423516759144343,
+      "loss": 3.4946,
+      "step": 12282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5226430296897888,
+      "learning_rate": 0.0005423426010294597,
+      "loss": 3.1759,
+      "step": 12283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5157943964004517,
+      "learning_rate": 0.0005423335255062013,
+      "loss": 3.0393,
+      "step": 12284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4765145182609558,
+      "learning_rate": 0.000542324449344683,
+      "loss": 3.2868,
+      "step": 12285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.539223849773407,
+      "learning_rate": 0.0005423153725449287,
+      "loss": 3.2323,
+      "step": 12286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5330362319946289,
+      "learning_rate": 0.0005423062951069624,
+      "loss": 3.2413,
+      "step": 12287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5498042106628418,
+      "learning_rate": 0.000542297217030808,
+      "loss": 2.9149,
+      "step": 12288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5433230400085449,
+      "learning_rate": 0.0005422881383164893,
+      "loss": 3.1215,
+      "step": 12289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5228878855705261,
+      "learning_rate": 0.0005422790589640303,
+      "loss": 3.2642,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5186182260513306,
+      "learning_rate": 0.0005422699789734548,
+      "loss": 3.4114,
+      "step": 12291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5254524946212769,
+      "learning_rate": 0.0005422608983447868,
+      "loss": 3.4245,
+      "step": 12292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5335171222686768,
+      "learning_rate": 0.0005422518170780504,
+      "loss": 3.0748,
+      "step": 12293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5440496206283569,
+      "learning_rate": 0.0005422427351732692,
+      "loss": 3.0979,
+      "step": 12294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5325355529785156,
+      "learning_rate": 0.0005422336526304673,
+      "loss": 3.2867,
+      "step": 12295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.536151647567749,
+      "learning_rate": 0.0005422245694496685,
+      "loss": 3.3285,
+      "step": 12296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.516470730304718,
+      "learning_rate": 0.0005422154856308969,
+      "loss": 3.3174,
+      "step": 12297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4913516938686371,
+      "learning_rate": 0.0005422064011741763,
+      "loss": 3.1424,
+      "step": 12298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4844023883342743,
+      "learning_rate": 0.0005421973160795306,
+      "loss": 3.3294,
+      "step": 12299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5266075134277344,
+      "learning_rate": 0.0005421882303469839,
+      "loss": 3.1666,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5134689807891846,
+      "learning_rate": 0.0005421791439765599,
+      "loss": 3.1568,
+      "step": 12301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5349195599555969,
+      "learning_rate": 0.0005421700569682827,
+      "loss": 3.3157,
+      "step": 12302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5083217024803162,
+      "learning_rate": 0.0005421609693221761,
+      "loss": 3.0988,
+      "step": 12303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49216586351394653,
+      "learning_rate": 0.0005421518810382641,
+      "loss": 3.3646,
+      "step": 12304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5193181037902832,
+      "learning_rate": 0.0005421427921165708,
+      "loss": 3.3176,
+      "step": 12305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5635499358177185,
+      "learning_rate": 0.0005421337025571199,
+      "loss": 3.1367,
+      "step": 12306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5421886444091797,
+      "learning_rate": 0.0005421246123599354,
+      "loss": 3.2597,
+      "step": 12307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49263355135917664,
+      "learning_rate": 0.0005421155215250413,
+      "loss": 3.3185,
+      "step": 12308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4998137652873993,
+      "learning_rate": 0.0005421064300524615,
+      "loss": 3.2735,
+      "step": 12309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5159027576446533,
+      "learning_rate": 0.00054209733794222,
+      "loss": 3.1562,
+      "step": 12310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48767736554145813,
+      "learning_rate": 0.0005420882451943406,
+      "loss": 3.0237,
+      "step": 12311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5056929588317871,
+      "learning_rate": 0.0005420791518088474,
+      "loss": 3.2519,
+      "step": 12312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5624472498893738,
+      "learning_rate": 0.0005420700577857643,
+      "loss": 3.0888,
+      "step": 12313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5267937183380127,
+      "learning_rate": 0.0005420609631251154,
+      "loss": 3.1991,
+      "step": 12314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5424567461013794,
+      "learning_rate": 0.0005420518678269244,
+      "loss": 3.2991,
+      "step": 12315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4930018484592438,
+      "learning_rate": 0.0005420427718912153,
+      "loss": 3.123,
+      "step": 12316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5012913346290588,
+      "learning_rate": 0.0005420336753180122,
+      "loss": 3.3452,
+      "step": 12317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5235104560852051,
+      "learning_rate": 0.0005420245781073389,
+      "loss": 3.1449,
+      "step": 12318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5059866905212402,
+      "learning_rate": 0.0005420154802592195,
+      "loss": 3.213,
+      "step": 12319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5268304944038391,
+      "learning_rate": 0.0005420063817736778,
+      "loss": 3.2597,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5179598331451416,
+      "learning_rate": 0.000541997282650738,
+      "loss": 3.0205,
+      "step": 12321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.501977264881134,
+      "learning_rate": 0.0005419881828904239,
+      "loss": 3.1745,
+      "step": 12322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5240980386734009,
+      "learning_rate": 0.0005419790824927595,
+      "loss": 3.2458,
+      "step": 12323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49430304765701294,
+      "learning_rate": 0.0005419699814577687,
+      "loss": 3.4038,
+      "step": 12324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5261852145195007,
+      "learning_rate": 0.0005419608797854757,
+      "loss": 3.3711,
+      "step": 12325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5073800683021545,
+      "learning_rate": 0.0005419517774759041,
+      "loss": 3.1167,
+      "step": 12326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5025413036346436,
+      "learning_rate": 0.0005419426745290782,
+      "loss": 3.3158,
+      "step": 12327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5052799582481384,
+      "learning_rate": 0.0005419335709450218,
+      "loss": 3.131,
+      "step": 12328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5107625722885132,
+      "learning_rate": 0.000541924466723759,
+      "loss": 3.4689,
+      "step": 12329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5340820550918579,
+      "learning_rate": 0.0005419153618653137,
+      "loss": 3.3352,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5223778486251831,
+      "learning_rate": 0.0005419062563697098,
+      "loss": 3.2638,
+      "step": 12331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5212966203689575,
+      "learning_rate": 0.0005418971502369716,
+      "loss": 3.2023,
+      "step": 12332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5198991298675537,
+      "learning_rate": 0.0005418880434671226,
+      "loss": 3.2943,
+      "step": 12333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5650694966316223,
+      "learning_rate": 0.0005418789360601872,
+      "loss": 3.0997,
+      "step": 12334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5771616101264954,
+      "learning_rate": 0.0005418698280161892,
+      "loss": 3.07,
+      "step": 12335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5297648906707764,
+      "learning_rate": 0.0005418607193351526,
+      "loss": 3.3583,
+      "step": 12336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5116552710533142,
+      "learning_rate": 0.0005418516100171015,
+      "loss": 3.1286,
+      "step": 12337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5247451663017273,
+      "learning_rate": 0.0005418425000620597,
+      "loss": 2.9956,
+      "step": 12338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5303729772567749,
+      "learning_rate": 0.0005418333894700513,
+      "loss": 3.1847,
+      "step": 12339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5321176648139954,
+      "learning_rate": 0.0005418242782411004,
+      "loss": 3.2272,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5363842844963074,
+      "learning_rate": 0.0005418151663752307,
+      "loss": 3.1136,
+      "step": 12341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.543472409248352,
+      "learning_rate": 0.0005418060538724666,
+      "loss": 3.3665,
+      "step": 12342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5340068936347961,
+      "learning_rate": 0.0005417969407328318,
+      "loss": 3.1082,
+      "step": 12343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49749499559402466,
+      "learning_rate": 0.0005417878269563504,
+      "loss": 3.0852,
+      "step": 12344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5240546464920044,
+      "learning_rate": 0.0005417787125430464,
+      "loss": 3.1989,
+      "step": 12345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5031521320343018,
+      "learning_rate": 0.0005417695974929438,
+      "loss": 3.1801,
+      "step": 12346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5448225736618042,
+      "learning_rate": 0.0005417604818060666,
+      "loss": 3.4148,
+      "step": 12347
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5050467848777771,
+      "learning_rate": 0.0005417513654824387,
+      "loss": 3.1923,
+      "step": 12348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5299510359764099,
+      "learning_rate": 0.0005417422485220844,
+      "loss": 3.1741,
+      "step": 12349
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5421825647354126,
+      "learning_rate": 0.0005417331309250275,
+      "loss": 3.2135,
+      "step": 12350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5428870916366577,
+      "learning_rate": 0.0005417240126912921,
+      "loss": 3.1527,
+      "step": 12351
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6039544939994812,
+      "learning_rate": 0.0005417148938209021,
+      "loss": 3.3517,
+      "step": 12352
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5433862805366516,
+      "learning_rate": 0.0005417057743138816,
+      "loss": 3.2371,
+      "step": 12353
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48752960562705994,
+      "learning_rate": 0.0005416966541702546,
+      "loss": 3.1304,
+      "step": 12354
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.529007613658905,
+      "learning_rate": 0.0005416875333900451,
+      "loss": 3.1325,
+      "step": 12355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5171936750411987,
+      "learning_rate": 0.0005416784119732773,
+      "loss": 3.1296,
+      "step": 12356
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5062511563301086,
+      "learning_rate": 0.0005416692899199749,
+      "loss": 3.3456,
+      "step": 12357
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5134553909301758,
+      "learning_rate": 0.0005416601672301622,
+      "loss": 3.0025,
+      "step": 12358
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5461899042129517,
+      "learning_rate": 0.0005416510439038631,
+      "loss": 3.4576,
+      "step": 12359
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6372771859169006,
+      "learning_rate": 0.0005416419199411016,
+      "loss": 3.4149,
+      "step": 12360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5487025380134583,
+      "learning_rate": 0.0005416327953419018,
+      "loss": 3.2196,
+      "step": 12361
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5107635855674744,
+      "learning_rate": 0.0005416236701062878,
+      "loss": 3.1029,
+      "step": 12362
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.556648850440979,
+      "learning_rate": 0.0005416145442342836,
+      "loss": 3.2373,
+      "step": 12363
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5657925605773926,
+      "learning_rate": 0.0005416054177259131,
+      "loss": 3.1103,
+      "step": 12364
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49510687589645386,
+      "learning_rate": 0.0005415962905812004,
+      "loss": 3.2521,
+      "step": 12365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5163629651069641,
+      "learning_rate": 0.0005415871628001696,
+      "loss": 3.2231,
+      "step": 12366
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5414348244667053,
+      "learning_rate": 0.0005415780343828447,
+      "loss": 3.0145,
+      "step": 12367
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5388548374176025,
+      "learning_rate": 0.0005415689053292497,
+      "loss": 3.0301,
+      "step": 12368
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49389225244522095,
+      "learning_rate": 0.0005415597756394089,
+      "loss": 3.2222,
+      "step": 12369
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5336668491363525,
+      "learning_rate": 0.0005415506453133459,
+      "loss": 3.2878,
+      "step": 12370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.500521719455719,
+      "learning_rate": 0.000541541514351085,
+      "loss": 3.113,
+      "step": 12371
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4869195520877838,
+      "learning_rate": 0.0005415323827526505,
+      "loss": 3.1353,
+      "step": 12372
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5147513747215271,
+      "learning_rate": 0.000541523250518066,
+      "loss": 3.3804,
+      "step": 12373
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5055763721466064,
+      "learning_rate": 0.0005415141176473557,
+      "loss": 3.3596,
+      "step": 12374
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5058522820472717,
+      "learning_rate": 0.0005415049841405437,
+      "loss": 3.0117,
+      "step": 12375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5772499442100525,
+      "learning_rate": 0.0005414958499976541,
+      "loss": 3.1889,
+      "step": 12376
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5016292333602905,
+      "learning_rate": 0.0005414867152187108,
+      "loss": 3.1536,
+      "step": 12377
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5072944760322571,
+      "learning_rate": 0.0005414775798037382,
+      "loss": 3.2119,
+      "step": 12378
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5283709168434143,
+      "learning_rate": 0.0005414684437527599,
+      "loss": 2.9134,
+      "step": 12379
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5712698101997375,
+      "learning_rate": 0.0005414593070658004,
+      "loss": 3.2677,
+      "step": 12380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5036458373069763,
+      "learning_rate": 0.0005414501697428833,
+      "loss": 3.2085,
+      "step": 12381
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5280167460441589,
+      "learning_rate": 0.0005414410317840331,
+      "loss": 3.1025,
+      "step": 12382
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.548110842704773,
+      "learning_rate": 0.0005414318931892737,
+      "loss": 3.3154,
+      "step": 12383
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5638526082038879,
+      "learning_rate": 0.0005414227539586291,
+      "loss": 3.189,
+      "step": 12384
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.519481897354126,
+      "learning_rate": 0.0005414136140921234,
+      "loss": 3.1508,
+      "step": 12385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49351656436920166,
+      "learning_rate": 0.0005414044735897808,
+      "loss": 3.2086,
+      "step": 12386
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47887444496154785,
+      "learning_rate": 0.0005413953324516251,
+      "loss": 3.0731,
+      "step": 12387
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5507675409317017,
+      "learning_rate": 0.0005413861906776807,
+      "loss": 3.3115,
+      "step": 12388
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.518807590007782,
+      "learning_rate": 0.0005413770482679714,
+      "loss": 3.5088,
+      "step": 12389
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5343676209449768,
+      "learning_rate": 0.0005413679052225216,
+      "loss": 3.2752,
+      "step": 12390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5148628354072571,
+      "learning_rate": 0.0005413587615413551,
+      "loss": 3.1195,
+      "step": 12391
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5338741540908813,
+      "learning_rate": 0.000541349617224496,
+      "loss": 3.0901,
+      "step": 12392
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.508698046207428,
+      "learning_rate": 0.0005413404722719686,
+      "loss": 3.3288,
+      "step": 12393
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5255823731422424,
+      "learning_rate": 0.0005413313266837968,
+      "loss": 3.2126,
+      "step": 12394
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5159070491790771,
+      "learning_rate": 0.0005413221804600047,
+      "loss": 3.0548,
+      "step": 12395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5951764583587646,
+      "learning_rate": 0.0005413130336006164,
+      "loss": 3.3417,
+      "step": 12396
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5076818466186523,
+      "learning_rate": 0.000541303886105656,
+      "loss": 3.3209,
+      "step": 12397
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5986706018447876,
+      "learning_rate": 0.0005412947379751477,
+      "loss": 3.2027,
+      "step": 12398
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.506370484828949,
+      "learning_rate": 0.0005412855892091154,
+      "loss": 3.1447,
+      "step": 12399
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5274400115013123,
+      "learning_rate": 0.0005412764398075833,
+      "loss": 3.3458,
+      "step": 12400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5394174456596375,
+      "learning_rate": 0.0005412672897705755,
+      "loss": 3.3062,
+      "step": 12401
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49921727180480957,
+      "learning_rate": 0.0005412581390981161,
+      "loss": 3.3473,
+      "step": 12402
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5664904713630676,
+      "learning_rate": 0.0005412489877902292,
+      "loss": 3.2196,
+      "step": 12403
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5725209712982178,
+      "learning_rate": 0.0005412398358469389,
+      "loss": 3.3237,
+      "step": 12404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5521411299705505,
+      "learning_rate": 0.0005412306832682693,
+      "loss": 3.1473,
+      "step": 12405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5140300393104553,
+      "learning_rate": 0.0005412215300542446,
+      "loss": 3.2564,
+      "step": 12406
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5515618920326233,
+      "learning_rate": 0.0005412123762048885,
+      "loss": 3.029,
+      "step": 12407
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5705258846282959,
+      "learning_rate": 0.0005412032217202257,
+      "loss": 3.2977,
+      "step": 12408
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5424354672431946,
+      "learning_rate": 0.00054119406660028,
+      "loss": 3.1524,
+      "step": 12409
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5054551959037781,
+      "learning_rate": 0.0005411849108450756,
+      "loss": 3.2744,
+      "step": 12410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5675023794174194,
+      "learning_rate": 0.0005411757544546364,
+      "loss": 3.1539,
+      "step": 12411
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5317476391792297,
+      "learning_rate": 0.0005411665974289867,
+      "loss": 3.2369,
+      "step": 12412
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5437188744544983,
+      "learning_rate": 0.0005411574397681507,
+      "loss": 3.3225,
+      "step": 12413
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5426942706108093,
+      "learning_rate": 0.0005411482814721523,
+      "loss": 3.2466,
+      "step": 12414
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5315844416618347,
+      "learning_rate": 0.0005411391225410159,
+      "loss": 3.1621,
+      "step": 12415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5135928392410278,
+      "learning_rate": 0.0005411299629747654,
+      "loss": 3.1404,
+      "step": 12416
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5293831825256348,
+      "learning_rate": 0.000541120802773425,
+      "loss": 3.1356,
+      "step": 12417
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5011104941368103,
+      "learning_rate": 0.0005411116419370188,
+      "loss": 3.281,
+      "step": 12418
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.524491012096405,
+      "learning_rate": 0.0005411024804655708,
+      "loss": 3.1678,
+      "step": 12419
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5337167978286743,
+      "learning_rate": 0.0005410933183591053,
+      "loss": 3.1052,
+      "step": 12420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49512872099876404,
+      "learning_rate": 0.0005410841556176465,
+      "loss": 3.3314,
+      "step": 12421
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5637658834457397,
+      "learning_rate": 0.0005410749922412184,
+      "loss": 3.0071,
+      "step": 12422
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5093816518783569,
+      "learning_rate": 0.0005410658282298451,
+      "loss": 3.38,
+      "step": 12423
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5493909120559692,
+      "learning_rate": 0.0005410566635835509,
+      "loss": 3.4419,
+      "step": 12424
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6119068264961243,
+      "learning_rate": 0.0005410474983023599,
+      "loss": 3.3356,
+      "step": 12425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5353181958198547,
+      "learning_rate": 0.000541038332386296,
+      "loss": 3.2801,
+      "step": 12426
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5060677528381348,
+      "learning_rate": 0.0005410291658353837,
+      "loss": 3.1649,
+      "step": 12427
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5227444171905518,
+      "learning_rate": 0.0005410199986496467,
+      "loss": 3.4991,
+      "step": 12428
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5346323251724243,
+      "learning_rate": 0.0005410108308291097,
+      "loss": 3.2646,
+      "step": 12429
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5016912221908569,
+      "learning_rate": 0.0005410016623737964,
+      "loss": 3.328,
+      "step": 12430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5151358246803284,
+      "learning_rate": 0.0005409924932837312,
+      "loss": 3.2546,
+      "step": 12431
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5453892350196838,
+      "learning_rate": 0.0005409833235589381,
+      "loss": 3.1258,
+      "step": 12432
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5162804126739502,
+      "learning_rate": 0.0005409741531994413,
+      "loss": 3.212,
+      "step": 12433
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5229930877685547,
+      "learning_rate": 0.0005409649822052648,
+      "loss": 2.9828,
+      "step": 12434
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5364524722099304,
+      "learning_rate": 0.0005409558105764332,
+      "loss": 3.2532,
+      "step": 12435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5143020153045654,
+      "learning_rate": 0.0005409466383129701,
+      "loss": 3.1768,
+      "step": 12436
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.523679256439209,
+      "learning_rate": 0.0005409374654149001,
+      "loss": 3.3917,
+      "step": 12437
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5435823798179626,
+      "learning_rate": 0.0005409282918822472,
+      "loss": 3.1854,
+      "step": 12438
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5507246255874634,
+      "learning_rate": 0.0005409191177150355,
+      "loss": 3.1831,
+      "step": 12439
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5211560726165771,
+      "learning_rate": 0.0005409099429132891,
+      "loss": 3.2665,
+      "step": 12440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49892207980155945,
+      "learning_rate": 0.0005409007674770324,
+      "loss": 2.9391,
+      "step": 12441
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5270420908927917,
+      "learning_rate": 0.0005408915914062894,
+      "loss": 3.2581,
+      "step": 12442
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.518926203250885,
+      "learning_rate": 0.0005408824147010844,
+      "loss": 3.2074,
+      "step": 12443
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4994022846221924,
+      "learning_rate": 0.0005408732373614414,
+      "loss": 3.0999,
+      "step": 12444
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5447221994400024,
+      "learning_rate": 0.0005408640593873846,
+      "loss": 3.1956,
+      "step": 12445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5384201407432556,
+      "learning_rate": 0.0005408548807789383,
+      "loss": 3.3842,
+      "step": 12446
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5086652040481567,
+      "learning_rate": 0.0005408457015361266,
+      "loss": 3.2858,
+      "step": 12447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49078619480133057,
+      "learning_rate": 0.0005408365216589736,
+      "loss": 3.143,
+      "step": 12448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5547512769699097,
+      "learning_rate": 0.0005408273411475036,
+      "loss": 3.374,
+      "step": 12449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49764230847358704,
+      "learning_rate": 0.0005408181600017407,
+      "loss": 3.1671,
+      "step": 12450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5081911087036133,
+      "learning_rate": 0.0005408089782217092,
+      "loss": 3.1078,
+      "step": 12451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5074054002761841,
+      "learning_rate": 0.0005407997958074331,
+      "loss": 3.1706,
+      "step": 12452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47185125946998596,
+      "learning_rate": 0.0005407906127589368,
+      "loss": 3.1016,
+      "step": 12453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5095272064208984,
+      "learning_rate": 0.0005407814290762442,
+      "loss": 3.3671,
+      "step": 12454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5170059204101562,
+      "learning_rate": 0.0005407722447593798,
+      "loss": 3.1382,
+      "step": 12455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5567418336868286,
+      "learning_rate": 0.0005407630598083676,
+      "loss": 3.2116,
+      "step": 12456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48573851585388184,
+      "learning_rate": 0.0005407538742232317,
+      "loss": 3.3303,
+      "step": 12457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5352392792701721,
+      "learning_rate": 0.0005407446880039966,
+      "loss": 2.8769,
+      "step": 12458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6195808053016663,
+      "learning_rate": 0.0005407355011506862,
+      "loss": 3.3433,
+      "step": 12459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5522502660751343,
+      "learning_rate": 0.000540726313663325,
+      "loss": 3.2809,
+      "step": 12460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5253701210021973,
+      "learning_rate": 0.0005407171255419369,
+      "loss": 3.3492,
+      "step": 12461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5422989130020142,
+      "learning_rate": 0.0005407079367865462,
+      "loss": 3.2108,
+      "step": 12462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.478427916765213,
+      "learning_rate": 0.0005406987473971772,
+      "loss": 3.0953,
+      "step": 12463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.616083562374115,
+      "learning_rate": 0.0005406895573738539,
+      "loss": 3.2778,
+      "step": 12464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5093544125556946,
+      "learning_rate": 0.0005406803667166008,
+      "loss": 3.3096,
+      "step": 12465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5077793598175049,
+      "learning_rate": 0.0005406711754254418,
+      "loss": 3.2769,
+      "step": 12466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52507483959198,
+      "learning_rate": 0.0005406619835004012,
+      "loss": 3.2368,
+      "step": 12467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.500822126865387,
+      "learning_rate": 0.0005406527909415033,
+      "loss": 3.068,
+      "step": 12468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5564792156219482,
+      "learning_rate": 0.0005406435977487723,
+      "loss": 3.1324,
+      "step": 12469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5204517245292664,
+      "learning_rate": 0.0005406344039222325,
+      "loss": 3.0969,
+      "step": 12470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5361521244049072,
+      "learning_rate": 0.0005406252094619079,
+      "loss": 3.3305,
+      "step": 12471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5398157238960266,
+      "learning_rate": 0.0005406160143678226,
+      "loss": 3.0892,
+      "step": 12472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5248488187789917,
+      "learning_rate": 0.0005406068186400013,
+      "loss": 3.1276,
+      "step": 12473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5540500283241272,
+      "learning_rate": 0.0005405976222784678,
+      "loss": 3.1021,
+      "step": 12474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5601980686187744,
+      "learning_rate": 0.0005405884252832465,
+      "loss": 2.9468,
+      "step": 12475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5255300998687744,
+      "learning_rate": 0.0005405792276543616,
+      "loss": 3.3761,
+      "step": 12476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5468989014625549,
+      "learning_rate": 0.0005405700293918373,
+      "loss": 3.3636,
+      "step": 12477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5150614976882935,
+      "learning_rate": 0.000540560830495698,
+      "loss": 3.3567,
+      "step": 12478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.555590033531189,
+      "learning_rate": 0.0005405516309659674,
+      "loss": 3.0297,
+      "step": 12479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5096856951713562,
+      "learning_rate": 0.0005405424308026704,
+      "loss": 3.2316,
+      "step": 12480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5092878341674805,
+      "learning_rate": 0.0005405332300058308,
+      "loss": 3.5083,
+      "step": 12481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5776546597480774,
+      "learning_rate": 0.0005405240285754731,
+      "loss": 3.1405,
+      "step": 12482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5385265946388245,
+      "learning_rate": 0.0005405148265116213,
+      "loss": 3.3177,
+      "step": 12483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5115643739700317,
+      "learning_rate": 0.0005405056238142997,
+      "loss": 3.3791,
+      "step": 12484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4903174340724945,
+      "learning_rate": 0.0005404964204835327,
+      "loss": 3.3797,
+      "step": 12485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5272238254547119,
+      "learning_rate": 0.0005404872165193443,
+      "loss": 3.2622,
+      "step": 12486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5495355725288391,
+      "learning_rate": 0.0005404780119217589,
+      "loss": 3.0021,
+      "step": 12487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5022799968719482,
+      "learning_rate": 0.0005404688066908006,
+      "loss": 3.26,
+      "step": 12488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5563308596611023,
+      "learning_rate": 0.0005404596008264939,
+      "loss": 3.3463,
+      "step": 12489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5790652632713318,
+      "learning_rate": 0.0005404503943288628,
+      "loss": 3.3775,
+      "step": 12490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5320955514907837,
+      "learning_rate": 0.0005404411871979316,
+      "loss": 3.2666,
+      "step": 12491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5261737704277039,
+      "learning_rate": 0.0005404319794337246,
+      "loss": 3.1438,
+      "step": 12492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4732470214366913,
+      "learning_rate": 0.0005404227710362661,
+      "loss": 3.1054,
+      "step": 12493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5220340490341187,
+      "learning_rate": 0.0005404135620055803,
+      "loss": 3.2321,
+      "step": 12494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4781638979911804,
+      "learning_rate": 0.0005404043523416914,
+      "loss": 3.1854,
+      "step": 12495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5066351294517517,
+      "learning_rate": 0.0005403951420446237,
+      "loss": 3.2492,
+      "step": 12496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49951955676078796,
+      "learning_rate": 0.0005403859311144015,
+      "loss": 3.1887,
+      "step": 12497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5178000330924988,
+      "learning_rate": 0.000540376719551049,
+      "loss": 3.2451,
+      "step": 12498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5040697455406189,
+      "learning_rate": 0.0005403675073545906,
+      "loss": 3.2617,
+      "step": 12499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48904526233673096,
+      "learning_rate": 0.0005403582945250503,
+      "loss": 3.0489,
+      "step": 12500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5385371446609497,
+      "learning_rate": 0.0005403490810624525,
+      "loss": 3.1477,
+      "step": 12501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5342295169830322,
+      "learning_rate": 0.0005403398669668215,
+      "loss": 3.1246,
+      "step": 12502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5917371511459351,
+      "learning_rate": 0.0005403306522381815,
+      "loss": 3.241,
+      "step": 12503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5648188591003418,
+      "learning_rate": 0.0005403214368765569,
+      "loss": 3.3202,
+      "step": 12504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.534196674823761,
+      "learning_rate": 0.0005403122208819718,
+      "loss": 3.3616,
+      "step": 12505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5107386112213135,
+      "learning_rate": 0.0005403030042544506,
+      "loss": 3.179,
+      "step": 12506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5261774659156799,
+      "learning_rate": 0.0005402937869940177,
+      "loss": 3.5787,
+      "step": 12507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4821683168411255,
+      "learning_rate": 0.000540284569100697,
+      "loss": 3.2806,
+      "step": 12508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5442585349082947,
+      "learning_rate": 0.0005402753505745131,
+      "loss": 3.3793,
+      "step": 12509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5089293122291565,
+      "learning_rate": 0.0005402661314154901,
+      "loss": 3.3455,
+      "step": 12510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5478642582893372,
+      "learning_rate": 0.0005402569116236525,
+      "loss": 3.1172,
+      "step": 12511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4844317138195038,
+      "learning_rate": 0.0005402476911990242,
+      "loss": 3.3636,
+      "step": 12512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5121679902076721,
+      "learning_rate": 0.0005402384701416298,
+      "loss": 3.1472,
+      "step": 12513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4980253279209137,
+      "learning_rate": 0.0005402292484514935,
+      "loss": 3.3732,
+      "step": 12514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49577051401138306,
+      "learning_rate": 0.0005402200261286396,
+      "loss": 3.2947,
+      "step": 12515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5190374255180359,
+      "learning_rate": 0.0005402108031730923,
+      "loss": 3.0081,
+      "step": 12516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5585235953330994,
+      "learning_rate": 0.0005402015795848762,
+      "loss": 3.2278,
+      "step": 12517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5044002532958984,
+      "learning_rate": 0.0005401923553640152,
+      "loss": 3.397,
+      "step": 12518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5685398578643799,
+      "learning_rate": 0.0005401831305105339,
+      "loss": 3.2287,
+      "step": 12519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5105231404304504,
+      "learning_rate": 0.0005401739050244563,
+      "loss": 3.2115,
+      "step": 12520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49602973461151123,
+      "learning_rate": 0.000540164678905807,
+      "loss": 3.2298,
+      "step": 12521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5373128056526184,
+      "learning_rate": 0.00054015545215461,
+      "loss": 2.917,
+      "step": 12522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5047098994255066,
+      "learning_rate": 0.0005401462247708899,
+      "loss": 3.4197,
+      "step": 12523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5524264574050903,
+      "learning_rate": 0.0005401369967546708,
+      "loss": 3.0935,
+      "step": 12524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5501293540000916,
+      "learning_rate": 0.000540127768105977,
+      "loss": 3.1648,
+      "step": 12525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5153632760047913,
+      "learning_rate": 0.0005401185388248329,
+      "loss": 3.1659,
+      "step": 12526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48313668370246887,
+      "learning_rate": 0.0005401093089112628,
+      "loss": 3.0783,
+      "step": 12527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5056579113006592,
+      "learning_rate": 0.0005401000783652911,
+      "loss": 3.2297,
+      "step": 12528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5305315256118774,
+      "learning_rate": 0.0005400908471869419,
+      "loss": 3.1893,
+      "step": 12529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5503982901573181,
+      "learning_rate": 0.0005400816153762396,
+      "loss": 3.1965,
+      "step": 12530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4872680902481079,
+      "learning_rate": 0.0005400723829332085,
+      "loss": 3.1287,
+      "step": 12531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5678309798240662,
+      "learning_rate": 0.000540063149857873,
+      "loss": 3.1886,
+      "step": 12532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4857582151889801,
+      "learning_rate": 0.0005400539161502574,
+      "loss": 3.208,
+      "step": 12533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5303822755813599,
+      "learning_rate": 0.000540044681810386,
+      "loss": 3.339,
+      "step": 12534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5294134616851807,
+      "learning_rate": 0.000540035446838283,
+      "loss": 3.2367,
+      "step": 12535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49902549386024475,
+      "learning_rate": 0.0005400262112339728,
+      "loss": 3.2169,
+      "step": 12536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4956739544868469,
+      "learning_rate": 0.0005400169749974798,
+      "loss": 3.1918,
+      "step": 12537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5438934564590454,
+      "learning_rate": 0.0005400077381288284,
+      "loss": 3.1434,
+      "step": 12538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4825858771800995,
+      "learning_rate": 0.0005399985006280427,
+      "loss": 3.0899,
+      "step": 12539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7768726944923401,
+      "learning_rate": 0.0005399892624951472,
+      "loss": 3.2963,
+      "step": 12540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5147074460983276,
+      "learning_rate": 0.0005399800237301661,
+      "loss": 3.1657,
+      "step": 12541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5385922789573669,
+      "learning_rate": 0.0005399707843331238,
+      "loss": 3.1703,
+      "step": 12542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5428267121315002,
+      "learning_rate": 0.0005399615443040447,
+      "loss": 3.0775,
+      "step": 12543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5166685581207275,
+      "learning_rate": 0.000539952303642953,
+      "loss": 3.2899,
+      "step": 12544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5144116282463074,
+      "learning_rate": 0.0005399430623498732,
+      "loss": 3.3296,
+      "step": 12545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49154841899871826,
+      "learning_rate": 0.0005399338204248295,
+      "loss": 3.0634,
+      "step": 12546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5445835590362549,
+      "learning_rate": 0.0005399245778678464,
+      "loss": 3.3533,
+      "step": 12547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4968005418777466,
+      "learning_rate": 0.0005399153346789479,
+      "loss": 3.3812,
+      "step": 12548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5119567513465881,
+      "learning_rate": 0.0005399060908581587,
+      "loss": 3.2191,
+      "step": 12549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5225988626480103,
+      "learning_rate": 0.0005398968464055031,
+      "loss": 3.2854,
+      "step": 12550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5341635942459106,
+      "learning_rate": 0.0005398876013210053,
+      "loss": 3.2697,
+      "step": 12551
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5326067805290222,
+      "learning_rate": 0.0005398783556046897,
+      "loss": 3.535,
+      "step": 12552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5414754152297974,
+      "learning_rate": 0.0005398691092565808,
+      "loss": 3.3331,
+      "step": 12553
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5281270146369934,
+      "learning_rate": 0.0005398598622767027,
+      "loss": 3.454,
+      "step": 12554
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5076268911361694,
+      "learning_rate": 0.0005398506146650799,
+      "loss": 3.4165,
+      "step": 12555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5204314589500427,
+      "learning_rate": 0.0005398413664217368,
+      "loss": 3.1195,
+      "step": 12556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5105859041213989,
+      "learning_rate": 0.0005398321175466977,
+      "loss": 3.2604,
+      "step": 12557
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5473527908325195,
+      "learning_rate": 0.0005398228680399869,
+      "loss": 3.1417,
+      "step": 12558
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4897322356700897,
+      "learning_rate": 0.0005398136179016288,
+      "loss": 3.2681,
+      "step": 12559
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.532136857509613,
+      "learning_rate": 0.000539804367131648,
+      "loss": 3.2423,
+      "step": 12560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.532255232334137,
+      "learning_rate": 0.0005397951157300684,
+      "loss": 3.3202,
+      "step": 12561
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.531471312046051,
+      "learning_rate": 0.0005397858636969148,
+      "loss": 3.0657,
+      "step": 12562
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5550201535224915,
+      "learning_rate": 0.0005397766110322112,
+      "loss": 3.203,
+      "step": 12563
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.519936740398407,
+      "learning_rate": 0.0005397673577359822,
+      "loss": 3.2974,
+      "step": 12564
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4963712990283966,
+      "learning_rate": 0.0005397581038082521,
+      "loss": 3.2021,
+      "step": 12565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5781660676002502,
+      "learning_rate": 0.0005397488492490455,
+      "loss": 3.1871,
+      "step": 12566
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49295535683631897,
+      "learning_rate": 0.0005397395940583864,
+      "loss": 3.127,
+      "step": 12567
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5267174243927002,
+      "learning_rate": 0.0005397303382362994,
+      "loss": 2.8856,
+      "step": 12568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5607599020004272,
+      "learning_rate": 0.0005397210817828088,
+      "loss": 3.1008,
+      "step": 12569
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5201016068458557,
+      "learning_rate": 0.000539711824697939,
+      "loss": 3.1056,
+      "step": 12570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5651845932006836,
+      "learning_rate": 0.0005397025669817144,
+      "loss": 3.2542,
+      "step": 12571
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5444151163101196,
+      "learning_rate": 0.0005396933086341593,
+      "loss": 3.0831,
+      "step": 12572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5270144939422607,
+      "learning_rate": 0.0005396840496552982,
+      "loss": 2.9548,
+      "step": 12573
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5137320756912231,
+      "learning_rate": 0.0005396747900451555,
+      "loss": 3.1978,
+      "step": 12574
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5519542694091797,
+      "learning_rate": 0.0005396655298037555,
+      "loss": 3.2202,
+      "step": 12575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5144039392471313,
+      "learning_rate": 0.0005396562689311226,
+      "loss": 3.1693,
+      "step": 12576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5353249907493591,
+      "learning_rate": 0.0005396470074272812,
+      "loss": 3.404,
+      "step": 12577
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5302972793579102,
+      "learning_rate": 0.0005396377452922558,
+      "loss": 3.4043,
+      "step": 12578
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4982793927192688,
+      "learning_rate": 0.0005396284825260705,
+      "loss": 3.2489,
+      "step": 12579
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5199756026268005,
+      "learning_rate": 0.0005396192191287502,
+      "loss": 3.1789,
+      "step": 12580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5148780345916748,
+      "learning_rate": 0.0005396099551003187,
+      "loss": 3.2521,
+      "step": 12581
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5244265198707581,
+      "learning_rate": 0.0005396006904408009,
+      "loss": 3.3569,
+      "step": 12582
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5343596935272217,
+      "learning_rate": 0.0005395914251502208,
+      "loss": 3.3348,
+      "step": 12583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.514910101890564,
+      "learning_rate": 0.0005395821592286031,
+      "loss": 3.4142,
+      "step": 12584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5111855864524841,
+      "learning_rate": 0.0005395728926759721,
+      "loss": 3.0777,
+      "step": 12585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5080044269561768,
+      "learning_rate": 0.0005395636254923522,
+      "loss": 3.3565,
+      "step": 12586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47200343012809753,
+      "learning_rate": 0.0005395543576777679,
+      "loss": 3.1417,
+      "step": 12587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5408815741539001,
+      "learning_rate": 0.0005395450892322433,
+      "loss": 3.3413,
+      "step": 12588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5425565242767334,
+      "learning_rate": 0.0005395358201558032,
+      "loss": 3.2498,
+      "step": 12589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5002942681312561,
+      "learning_rate": 0.0005395265504484719,
+      "loss": 3.2378,
+      "step": 12590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5082552433013916,
+      "learning_rate": 0.0005395172801102736,
+      "loss": 3.2024,
+      "step": 12591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5224912166595459,
+      "learning_rate": 0.000539508009141233,
+      "loss": 3.215,
+      "step": 12592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7015027403831482,
+      "learning_rate": 0.0005394987375413745,
+      "loss": 3.2016,
+      "step": 12593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4700780510902405,
+      "learning_rate": 0.0005394894653107222,
+      "loss": 3.3461,
+      "step": 12594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49604517221450806,
+      "learning_rate": 0.0005394801924493008,
+      "loss": 3.2246,
+      "step": 12595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5033439993858337,
+      "learning_rate": 0.0005394709189571347,
+      "loss": 3.2227,
+      "step": 12596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5192400217056274,
+      "learning_rate": 0.0005394616448342483,
+      "loss": 3.3658,
+      "step": 12597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5239402055740356,
+      "learning_rate": 0.0005394523700806659,
+      "loss": 2.9431,
+      "step": 12598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5193118453025818,
+      "learning_rate": 0.0005394430946964122,
+      "loss": 3.3063,
+      "step": 12599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5492883324623108,
+      "learning_rate": 0.0005394338186815114,
+      "loss": 3.1588,
+      "step": 12600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48858433961868286,
+      "learning_rate": 0.000539424542035988,
+      "loss": 3.3016,
+      "step": 12601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5181791186332703,
+      "learning_rate": 0.0005394152647598664,
+      "loss": 3.2007,
+      "step": 12602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5329108238220215,
+      "learning_rate": 0.0005394059868531711,
+      "loss": 3.1491,
+      "step": 12603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5719881653785706,
+      "learning_rate": 0.0005393967083159266,
+      "loss": 3.2346,
+      "step": 12604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.509117841720581,
+      "learning_rate": 0.0005393874291481571,
+      "loss": 3.1653,
+      "step": 12605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5386404395103455,
+      "learning_rate": 0.0005393781493498872,
+      "loss": 3.3633,
+      "step": 12606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5088648200035095,
+      "learning_rate": 0.0005393688689211413,
+      "loss": 3.1949,
+      "step": 12607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5584010481834412,
+      "learning_rate": 0.000539359587861944,
+      "loss": 3.4235,
+      "step": 12608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6642793416976929,
+      "learning_rate": 0.0005393503061723196,
+      "loss": 3.3751,
+      "step": 12609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5091199278831482,
+      "learning_rate": 0.0005393410238522924,
+      "loss": 3.1341,
+      "step": 12610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5205773711204529,
+      "learning_rate": 0.0005393317409018871,
+      "loss": 3.2806,
+      "step": 12611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5145056247711182,
+      "learning_rate": 0.0005393224573211281,
+      "loss": 3.3279,
+      "step": 12612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5004215240478516,
+      "learning_rate": 0.0005393131731100398,
+      "loss": 3.3095,
+      "step": 12613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5290515422821045,
+      "learning_rate": 0.0005393038882686466,
+      "loss": 3.2428,
+      "step": 12614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5009429454803467,
+      "learning_rate": 0.000539294602796973,
+      "loss": 3.0593,
+      "step": 12615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6031820178031921,
+      "learning_rate": 0.0005392853166950436,
+      "loss": 3.1133,
+      "step": 12616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5044199824333191,
+      "learning_rate": 0.0005392760299628825,
+      "loss": 3.3295,
+      "step": 12617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.582173228263855,
+      "learning_rate": 0.0005392667426005146,
+      "loss": 3.172,
+      "step": 12618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5244518518447876,
+      "learning_rate": 0.000539257454607964,
+      "loss": 3.4616,
+      "step": 12619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5106006264686584,
+      "learning_rate": 0.0005392481659852554,
+      "loss": 3.453,
+      "step": 12620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5275260806083679,
+      "learning_rate": 0.0005392388767324132,
+      "loss": 3.2022,
+      "step": 12621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6484015583992004,
+      "learning_rate": 0.0005392295868494617,
+      "loss": 3.1695,
+      "step": 12622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5001370310783386,
+      "learning_rate": 0.0005392202963364256,
+      "loss": 3.1173,
+      "step": 12623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5128294229507446,
+      "learning_rate": 0.0005392110051933293,
+      "loss": 3.3133,
+      "step": 12624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48935312032699585,
+      "learning_rate": 0.0005392017134201973,
+      "loss": 3.2482,
+      "step": 12625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546270489692688,
+      "learning_rate": 0.0005391924210170539,
+      "loss": 3.2341,
+      "step": 12626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5206765532493591,
+      "learning_rate": 0.0005391831279839237,
+      "loss": 3.1757,
+      "step": 12627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5980457067489624,
+      "learning_rate": 0.0005391738343208313,
+      "loss": 3.1083,
+      "step": 12628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5202022194862366,
+      "learning_rate": 0.0005391645400278009,
+      "loss": 3.256,
+      "step": 12629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5010168552398682,
+      "learning_rate": 0.0005391552451048571,
+      "loss": 3.1127,
+      "step": 12630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5590848326683044,
+      "learning_rate": 0.0005391459495520247,
+      "loss": 2.9533,
+      "step": 12631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49834656715393066,
+      "learning_rate": 0.0005391366533693276,
+      "loss": 3.1844,
+      "step": 12632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5144838690757751,
+      "learning_rate": 0.0005391273565567906,
+      "loss": 3.1728,
+      "step": 12633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5475999712944031,
+      "learning_rate": 0.0005391180591144383,
+      "loss": 3.3422,
+      "step": 12634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5091504454612732,
+      "learning_rate": 0.000539108761042295,
+      "loss": 3.1376,
+      "step": 12635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5275076031684875,
+      "learning_rate": 0.0005390994623403853,
+      "loss": 3.3421,
+      "step": 12636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.530569851398468,
+      "learning_rate": 0.0005390901630087336,
+      "loss": 3.1941,
+      "step": 12637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5221152901649475,
+      "learning_rate": 0.0005390808630473643,
+      "loss": 3.2616,
+      "step": 12638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5298863649368286,
+      "learning_rate": 0.0005390715624563021,
+      "loss": 3.2261,
+      "step": 12639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.513576865196228,
+      "learning_rate": 0.0005390622612355715,
+      "loss": 3.2288,
+      "step": 12640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4933810234069824,
+      "learning_rate": 0.0005390529593851969,
+      "loss": 3.261,
+      "step": 12641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5063499808311462,
+      "learning_rate": 0.0005390436569052027,
+      "loss": 3.1588,
+      "step": 12642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4862244725227356,
+      "learning_rate": 0.0005390343537956135,
+      "loss": 3.3587,
+      "step": 12643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49645307660102844,
+      "learning_rate": 0.0005390250500564539,
+      "loss": 3.0493,
+      "step": 12644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.506386399269104,
+      "learning_rate": 0.0005390157456877483,
+      "loss": 3.3449,
+      "step": 12645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5540480613708496,
+      "learning_rate": 0.0005390064406895212,
+      "loss": 3.3795,
+      "step": 12646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5100905895233154,
+      "learning_rate": 0.0005389971350617972,
+      "loss": 3.2195,
+      "step": 12647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5141370296478271,
+      "learning_rate": 0.0005389878288046007,
+      "loss": 3.3391,
+      "step": 12648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49485722184181213,
+      "learning_rate": 0.0005389785219179562,
+      "loss": 3.0948,
+      "step": 12649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5505667328834534,
+      "learning_rate": 0.0005389692144018883,
+      "loss": 3.1628,
+      "step": 12650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5638545751571655,
+      "learning_rate": 0.0005389599062564216,
+      "loss": 3.0784,
+      "step": 12651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5764116048812866,
+      "learning_rate": 0.0005389505974815803,
+      "loss": 3.0416,
+      "step": 12652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.516578197479248,
+      "learning_rate": 0.0005389412880773891,
+      "loss": 3.4401,
+      "step": 12653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5281229019165039,
+      "learning_rate": 0.0005389319780438728,
+      "loss": 3.1235,
+      "step": 12654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5611839294433594,
+      "learning_rate": 0.0005389226673810554,
+      "loss": 3.2922,
+      "step": 12655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5751029849052429,
+      "learning_rate": 0.0005389133560889617,
+      "loss": 3.0524,
+      "step": 12656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5245303511619568,
+      "learning_rate": 0.0005389040441676164,
+      "loss": 3.2482,
+      "step": 12657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5082685351371765,
+      "learning_rate": 0.0005388947316170437,
+      "loss": 3.2612,
+      "step": 12658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5107821822166443,
+      "learning_rate": 0.0005388854184372682,
+      "loss": 3.3678,
+      "step": 12659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5098000168800354,
+      "learning_rate": 0.0005388761046283146,
+      "loss": 3.1828,
+      "step": 12660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.506803035736084,
+      "learning_rate": 0.0005388667901902071,
+      "loss": 3.3027,
+      "step": 12661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5258462429046631,
+      "learning_rate": 0.0005388574751229707,
+      "loss": 3.2059,
+      "step": 12662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5080270767211914,
+      "learning_rate": 0.0005388481594266295,
+      "loss": 3.2258,
+      "step": 12663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5246100425720215,
+      "learning_rate": 0.0005388388431012083,
+      "loss": 3.2327,
+      "step": 12664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5078802108764648,
+      "learning_rate": 0.0005388295261467315,
+      "loss": 3.1975,
+      "step": 12665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5111520290374756,
+      "learning_rate": 0.0005388202085632237,
+      "loss": 3.2364,
+      "step": 12666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5226684212684631,
+      "learning_rate": 0.0005388108903507095,
+      "loss": 3.118,
+      "step": 12667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5708491802215576,
+      "learning_rate": 0.0005388015715092133,
+      "loss": 3.3362,
+      "step": 12668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5695480108261108,
+      "learning_rate": 0.0005387922520387597,
+      "loss": 3.2954,
+      "step": 12669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4956531822681427,
+      "learning_rate": 0.0005387829319393735,
+      "loss": 3.1396,
+      "step": 12670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5610358715057373,
+      "learning_rate": 0.0005387736112110787,
+      "loss": 3.1429,
+      "step": 12671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5279606580734253,
+      "learning_rate": 0.0005387642898539004,
+      "loss": 3.1699,
+      "step": 12672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5159085988998413,
+      "learning_rate": 0.0005387549678678627,
+      "loss": 3.1701,
+      "step": 12673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5232102870941162,
+      "learning_rate": 0.0005387456452529904,
+      "loss": 3.2938,
+      "step": 12674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5625350475311279,
+      "learning_rate": 0.0005387363220093082,
+      "loss": 3.1154,
+      "step": 12675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.524451732635498,
+      "learning_rate": 0.0005387269981368403,
+      "loss": 3.2826,
+      "step": 12676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5030291080474854,
+      "learning_rate": 0.0005387176736356116,
+      "loss": 3.0269,
+      "step": 12677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5304322242736816,
+      "learning_rate": 0.0005387083485056463,
+      "loss": 3.3728,
+      "step": 12678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5527458786964417,
+      "learning_rate": 0.0005386990227469693,
+      "loss": 3.2158,
+      "step": 12679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5265601277351379,
+      "learning_rate": 0.0005386896963596049,
+      "loss": 3.3758,
+      "step": 12680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5212921500205994,
+      "learning_rate": 0.0005386803693435778,
+      "loss": 3.3079,
+      "step": 12681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5151424407958984,
+      "learning_rate": 0.0005386710416989126,
+      "loss": 3.2643,
+      "step": 12682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5002523064613342,
+      "learning_rate": 0.0005386617134256337,
+      "loss": 3.2302,
+      "step": 12683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5380643010139465,
+      "learning_rate": 0.0005386523845237659,
+      "loss": 3.1046,
+      "step": 12684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5118758082389832,
+      "learning_rate": 0.0005386430549933336,
+      "loss": 3.192,
+      "step": 12685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5095431804656982,
+      "learning_rate": 0.0005386337248343613,
+      "loss": 3.3801,
+      "step": 12686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5566089749336243,
+      "learning_rate": 0.0005386243940468738,
+      "loss": 3.0442,
+      "step": 12687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4650208055973053,
+      "learning_rate": 0.0005386150626308957,
+      "loss": 3.0336,
+      "step": 12688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5460222959518433,
+      "learning_rate": 0.0005386057305864513,
+      "loss": 3.206,
+      "step": 12689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5706539154052734,
+      "learning_rate": 0.0005385963979135653,
+      "loss": 3.3201,
+      "step": 12690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5206764936447144,
+      "learning_rate": 0.0005385870646122624,
+      "loss": 3.2854,
+      "step": 12691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4984837770462036,
+      "learning_rate": 0.000538577730682567,
+      "loss": 3.2708,
+      "step": 12692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5224167108535767,
+      "learning_rate": 0.0005385683961245039,
+      "loss": 3.2557,
+      "step": 12693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5294681787490845,
+      "learning_rate": 0.0005385590609380974,
+      "loss": 3.1294,
+      "step": 12694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.544699490070343,
+      "learning_rate": 0.0005385497251233724,
+      "loss": 3.3876,
+      "step": 12695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5481709241867065,
+      "learning_rate": 0.0005385403886803532,
+      "loss": 3.2215,
+      "step": 12696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4763781428337097,
+      "learning_rate": 0.0005385310516090646,
+      "loss": 3.4355,
+      "step": 12697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48930293321609497,
+      "learning_rate": 0.0005385217139095311,
+      "loss": 3.1579,
+      "step": 12698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48442885279655457,
+      "learning_rate": 0.0005385123755817773,
+      "loss": 3.3198,
+      "step": 12699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.551783561706543,
+      "learning_rate": 0.0005385030366258278,
+      "loss": 3.3245,
+      "step": 12700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4957975149154663,
+      "learning_rate": 0.0005384936970417073,
+      "loss": 3.1174,
+      "step": 12701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5431187152862549,
+      "learning_rate": 0.0005384843568294401,
+      "loss": 3.0538,
+      "step": 12702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5256251096725464,
+      "learning_rate": 0.0005384750159890512,
+      "loss": 3.1746,
+      "step": 12703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5222598910331726,
+      "learning_rate": 0.0005384656745205649,
+      "loss": 3.4051,
+      "step": 12704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5458401441574097,
+      "learning_rate": 0.000538456332424006,
+      "loss": 3.1197,
+      "step": 12705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4959386885166168,
+      "learning_rate": 0.0005384469896993989,
+      "loss": 3.2583,
+      "step": 12706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5014752745628357,
+      "learning_rate": 0.0005384376463467683,
+      "loss": 3.1556,
+      "step": 12707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.491812527179718,
+      "learning_rate": 0.0005384283023661389,
+      "loss": 3.2109,
+      "step": 12708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5517925024032593,
+      "learning_rate": 0.0005384189577575352,
+      "loss": 3.2206,
+      "step": 12709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.504330039024353,
+      "learning_rate": 0.000538409612520982,
+      "loss": 3.1117,
+      "step": 12710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5130345225334167,
+      "learning_rate": 0.0005384002666565036,
+      "loss": 3.1259,
+      "step": 12711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.511404275894165,
+      "learning_rate": 0.0005383909201641247,
+      "loss": 3.2433,
+      "step": 12712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5336930155754089,
+      "learning_rate": 0.0005383815730438702,
+      "loss": 3.3069,
+      "step": 12713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5323059558868408,
+      "learning_rate": 0.0005383722252957644,
+      "loss": 3.0323,
+      "step": 12714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.540348470211029,
+      "learning_rate": 0.000538362876919832,
+      "loss": 2.9597,
+      "step": 12715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5608701109886169,
+      "learning_rate": 0.0005383535279160978,
+      "loss": 3.2359,
+      "step": 12716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5046235918998718,
+      "learning_rate": 0.0005383441782845863,
+      "loss": 3.0872,
+      "step": 12717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5153566002845764,
+      "learning_rate": 0.0005383348280253219,
+      "loss": 3.1468,
+      "step": 12718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.521328866481781,
+      "learning_rate": 0.0005383254771383296,
+      "loss": 3.3589,
+      "step": 12719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.505582332611084,
+      "learning_rate": 0.0005383161256236337,
+      "loss": 3.3211,
+      "step": 12720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5521395206451416,
+      "learning_rate": 0.0005383067734812592,
+      "loss": 3.3091,
+      "step": 12721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5566513538360596,
+      "learning_rate": 0.0005382974207112304,
+      "loss": 3.1199,
+      "step": 12722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5230540633201599,
+      "learning_rate": 0.000538288067313572,
+      "loss": 3.3551,
+      "step": 12723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5009475350379944,
+      "learning_rate": 0.0005382787132883087,
+      "loss": 3.4286,
+      "step": 12724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5285469889640808,
+      "learning_rate": 0.0005382693586354653,
+      "loss": 3.1925,
+      "step": 12725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5307457447052002,
+      "learning_rate": 0.000538260003355066,
+      "loss": 3.2882,
+      "step": 12726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5530621409416199,
+      "learning_rate": 0.0005382506474471359,
+      "loss": 3.1743,
+      "step": 12727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5210942625999451,
+      "learning_rate": 0.0005382412909116993,
+      "loss": 3.0778,
+      "step": 12728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4954610764980316,
+      "learning_rate": 0.0005382319337487812,
+      "loss": 3.3454,
+      "step": 12729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5154182314872742,
+      "learning_rate": 0.0005382225759584058,
+      "loss": 3.3812,
+      "step": 12730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5248116850852966,
+      "learning_rate": 0.0005382132175405982,
+      "loss": 3.4016,
+      "step": 12731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.524318277835846,
+      "learning_rate": 0.0005382038584953828,
+      "loss": 3.1419,
+      "step": 12732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5205380916595459,
+      "learning_rate": 0.0005381944988227842,
+      "loss": 3.2683,
+      "step": 12733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5063703060150146,
+      "learning_rate": 0.000538185138522827,
+      "loss": 3.3365,
+      "step": 12734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5397539734840393,
+      "learning_rate": 0.0005381757775955362,
+      "loss": 3.23,
+      "step": 12735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5201455950737,
+      "learning_rate": 0.0005381664160409362,
+      "loss": 3.2598,
+      "step": 12736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5498931407928467,
+      "learning_rate": 0.0005381570538590517,
+      "loss": 3.2865,
+      "step": 12737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4814511239528656,
+      "learning_rate": 0.0005381476910499073,
+      "loss": 3.4717,
+      "step": 12738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49921301007270813,
+      "learning_rate": 0.0005381383276135277,
+      "loss": 3.3063,
+      "step": 12739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5252735614776611,
+      "learning_rate": 0.0005381289635499376,
+      "loss": 3.1865,
+      "step": 12740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5268913507461548,
+      "learning_rate": 0.0005381195988591617,
+      "loss": 3.1435,
+      "step": 12741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4931754171848297,
+      "learning_rate": 0.0005381102335412245,
+      "loss": 3.4524,
+      "step": 12742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5258716940879822,
+      "learning_rate": 0.0005381008675961509,
+      "loss": 3.3626,
+      "step": 12743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5764732360839844,
+      "learning_rate": 0.0005380915010239654,
+      "loss": 3.3538,
+      "step": 12744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49837490916252136,
+      "learning_rate": 0.0005380821338246926,
+      "loss": 3.1224,
+      "step": 12745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5089791417121887,
+      "learning_rate": 0.0005380727659983573,
+      "loss": 3.1517,
+      "step": 12746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5005397796630859,
+      "learning_rate": 0.0005380633975449842,
+      "loss": 3.1022,
+      "step": 12747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5430983901023865,
+      "learning_rate": 0.0005380540284645979,
+      "loss": 3.3593,
+      "step": 12748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5346560478210449,
+      "learning_rate": 0.0005380446587572231,
+      "loss": 3.2986,
+      "step": 12749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5033416152000427,
+      "learning_rate": 0.0005380352884228846,
+      "loss": 3.4088,
+      "step": 12750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49493205547332764,
+      "learning_rate": 0.0005380259174616068,
+      "loss": 3.3954,
+      "step": 12751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49495062232017517,
+      "learning_rate": 0.0005380165458734147,
+      "loss": 3.3133,
+      "step": 12752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5404918193817139,
+      "learning_rate": 0.0005380071736583327,
+      "loss": 3.1616,
+      "step": 12753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5051382184028625,
+      "learning_rate": 0.0005379978008163857,
+      "loss": 3.103,
+      "step": 12754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5010060667991638,
+      "learning_rate": 0.0005379884273475982,
+      "loss": 3.4863,
+      "step": 12755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4935106933116913,
+      "learning_rate": 0.0005379790532519951,
+      "loss": 3.2361,
+      "step": 12756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5263615846633911,
+      "learning_rate": 0.0005379696785296008,
+      "loss": 3.1684,
+      "step": 12757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5205957293510437,
+      "learning_rate": 0.0005379603031804404,
+      "loss": 3.1844,
+      "step": 12758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5138325095176697,
+      "learning_rate": 0.0005379509272045381,
+      "loss": 3.2793,
+      "step": 12759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5030960440635681,
+      "learning_rate": 0.000537941550601919,
+      "loss": 3.3753,
+      "step": 12760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49174514412879944,
+      "learning_rate": 0.0005379321733726077,
+      "loss": 3.2894,
+      "step": 12761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5075502395629883,
+      "learning_rate": 0.0005379227955166287,
+      "loss": 3.1492,
+      "step": 12762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47649258375167847,
+      "learning_rate": 0.0005379134170340069,
+      "loss": 3.2046,
+      "step": 12763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5589198470115662,
+      "learning_rate": 0.000537904037924767,
+      "loss": 3.3309,
+      "step": 12764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47373253107070923,
+      "learning_rate": 0.0005378946581889336,
+      "loss": 3.1731,
+      "step": 12765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5284726023674011,
+      "learning_rate": 0.0005378852778265315,
+      "loss": 3.2825,
+      "step": 12766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5329920053482056,
+      "learning_rate": 0.0005378758968375854,
+      "loss": 3.1192,
+      "step": 12767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5024005770683289,
+      "learning_rate": 0.0005378665152221198,
+      "loss": 3.2978,
+      "step": 12768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49619409441947937,
+      "learning_rate": 0.0005378571329801596,
+      "loss": 3.3436,
+      "step": 12769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5017001628875732,
+      "learning_rate": 0.0005378477501117296,
+      "loss": 3.3643,
+      "step": 12770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5197016596794128,
+      "learning_rate": 0.0005378383666168545,
+      "loss": 3.1311,
+      "step": 12771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5079033374786377,
+      "learning_rate": 0.0005378289824955587,
+      "loss": 3.3026,
+      "step": 12772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4886915981769562,
+      "learning_rate": 0.0005378195977478672,
+      "loss": 3.2003,
+      "step": 12773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51219242811203,
+      "learning_rate": 0.0005378102123738046,
+      "loss": 3.3336,
+      "step": 12774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5035862326622009,
+      "learning_rate": 0.0005378008263733959,
+      "loss": 3.1417,
+      "step": 12775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5228604078292847,
+      "learning_rate": 0.0005377914397466653,
+      "loss": 3.2611,
+      "step": 12776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5679159760475159,
+      "learning_rate": 0.000537782052493638,
+      "loss": 3.4281,
+      "step": 12777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5072061419487,
+      "learning_rate": 0.0005377726646143384,
+      "loss": 3.1225,
+      "step": 12778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4947572350502014,
+      "learning_rate": 0.0005377632761087915,
+      "loss": 3.2278,
+      "step": 12779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5147088170051575,
+      "learning_rate": 0.0005377538869770218,
+      "loss": 3.0943,
+      "step": 12780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4995141923427582,
+      "learning_rate": 0.0005377444972190541,
+      "loss": 3.4808,
+      "step": 12781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5687293410301208,
+      "learning_rate": 0.0005377351068349132,
+      "loss": 3.3684,
+      "step": 12782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4961227476596832,
+      "learning_rate": 0.0005377257158246237,
+      "loss": 3.145,
+      "step": 12783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5262210965156555,
+      "learning_rate": 0.0005377163241882105,
+      "loss": 3.2739,
+      "step": 12784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47961992025375366,
+      "learning_rate": 0.0005377069319256983,
+      "loss": 3.0336,
+      "step": 12785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5090808272361755,
+      "learning_rate": 0.0005376975390371116,
+      "loss": 3.2362,
+      "step": 12786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5485349297523499,
+      "learning_rate": 0.0005376881455224755,
+      "loss": 2.8677,
+      "step": 12787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5236116647720337,
+      "learning_rate": 0.0005376787513818145,
+      "loss": 3.3714,
+      "step": 12788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5215224623680115,
+      "learning_rate": 0.0005376693566151535,
+      "loss": 3.1176,
+      "step": 12789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1414709091186523,
+      "learning_rate": 0.0005376599612225171,
+      "loss": 3.4685,
+      "step": 12790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5067881345748901,
+      "learning_rate": 0.00053765056520393,
+      "loss": 3.2304,
+      "step": 12791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5385640263557434,
+      "learning_rate": 0.0005376411685594171,
+      "loss": 3.2995,
+      "step": 12792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6865431666374207,
+      "learning_rate": 0.0005376317712890032,
+      "loss": 3.2136,
+      "step": 12793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5289443135261536,
+      "learning_rate": 0.0005376223733927129,
+      "loss": 3.317,
+      "step": 12794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5437031984329224,
+      "learning_rate": 0.0005376129748705709,
+      "loss": 3.1993,
+      "step": 12795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53926020860672,
+      "learning_rate": 0.0005376035757226022,
+      "loss": 3.2593,
+      "step": 12796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6048049330711365,
+      "learning_rate": 0.0005375941759488313,
+      "loss": 3.2746,
+      "step": 12797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5292719602584839,
+      "learning_rate": 0.0005375847755492831,
+      "loss": 3.1617,
+      "step": 12798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5354344844818115,
+      "learning_rate": 0.0005375753745239823,
+      "loss": 3.1443,
+      "step": 12799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5445627570152283,
+      "learning_rate": 0.0005375659728729537,
+      "loss": 3.2476,
+      "step": 12800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5120712518692017,
+      "learning_rate": 0.0005375565705962222,
+      "loss": 3.1865,
+      "step": 12801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5278573036193848,
+      "learning_rate": 0.0005375471676938123,
+      "loss": 3.3421,
+      "step": 12802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4721059501171112,
+      "learning_rate": 0.0005375377641657487,
+      "loss": 3.3415,
+      "step": 12803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5143515467643738,
+      "learning_rate": 0.0005375283600120565,
+      "loss": 3.4199,
+      "step": 12804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5790268778800964,
+      "learning_rate": 0.0005375189552327604,
+      "loss": 3.0928,
+      "step": 12805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5175043940544128,
+      "learning_rate": 0.000537509549827885,
+      "loss": 3.5718,
+      "step": 12806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5538207292556763,
+      "learning_rate": 0.0005375001437974552,
+      "loss": 3.1262,
+      "step": 12807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49472522735595703,
+      "learning_rate": 0.0005374907371414956,
+      "loss": 3.3913,
+      "step": 12808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5046636462211609,
+      "learning_rate": 0.0005374813298600312,
+      "loss": 3.2723,
+      "step": 12809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.538169264793396,
+      "learning_rate": 0.0005374719219530867,
+      "loss": 3.3067,
+      "step": 12810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5092310309410095,
+      "learning_rate": 0.0005374625134206868,
+      "loss": 3.0894,
+      "step": 12811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5437710881233215,
+      "learning_rate": 0.0005374531042628564,
+      "loss": 3.0988,
+      "step": 12812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5265883207321167,
+      "learning_rate": 0.0005374436944796202,
+      "loss": 3.1499,
+      "step": 12813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5095208287239075,
+      "learning_rate": 0.0005374342840710029,
+      "loss": 3.3266,
+      "step": 12814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5680007338523865,
+      "learning_rate": 0.0005374248730370295,
+      "loss": 3.3421,
+      "step": 12815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5302740931510925,
+      "learning_rate": 0.0005374154613777246,
+      "loss": 3.314,
+      "step": 12816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5018924474716187,
+      "learning_rate": 0.0005374060490931132,
+      "loss": 3.4018,
+      "step": 12817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5072476267814636,
+      "learning_rate": 0.0005373966361832199,
+      "loss": 3.1629,
+      "step": 12818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5976406931877136,
+      "learning_rate": 0.0005373872226480695,
+      "loss": 3.2877,
+      "step": 12819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5225663781166077,
+      "learning_rate": 0.0005373778084876869,
+      "loss": 3.3186,
+      "step": 12820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5068392753601074,
+      "learning_rate": 0.0005373683937020967,
+      "loss": 3.1388,
+      "step": 12821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5025680065155029,
+      "learning_rate": 0.000537358978291324,
+      "loss": 3.3463,
+      "step": 12822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5425643920898438,
+      "learning_rate": 0.0005373495622553934,
+      "loss": 3.1294,
+      "step": 12823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5345323085784912,
+      "learning_rate": 0.0005373401455943298,
+      "loss": 3.2662,
+      "step": 12824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5626019239425659,
+      "learning_rate": 0.0005373307283081577,
+      "loss": 3.4571,
+      "step": 12825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5290101170539856,
+      "learning_rate": 0.0005373213103969024,
+      "loss": 3.3035,
+      "step": 12826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5548025965690613,
+      "learning_rate": 0.0005373118918605883,
+      "loss": 3.1378,
+      "step": 12827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5364861488342285,
+      "learning_rate": 0.0005373024726992403,
+      "loss": 3.3674,
+      "step": 12828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5761748552322388,
+      "learning_rate": 0.0005372930529128833,
+      "loss": 3.2908,
+      "step": 12829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5271551012992859,
+      "learning_rate": 0.0005372836325015422,
+      "loss": 3.2864,
+      "step": 12830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5387271046638489,
+      "learning_rate": 0.0005372742114652415,
+      "loss": 3.0969,
+      "step": 12831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48338204622268677,
+      "learning_rate": 0.0005372647898040062,
+      "loss": 3.1388,
+      "step": 12832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5032204389572144,
+      "learning_rate": 0.0005372553675178612,
+      "loss": 3.0124,
+      "step": 12833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5275563597679138,
+      "learning_rate": 0.0005372459446068312,
+      "loss": 3.3673,
+      "step": 12834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5216962695121765,
+      "learning_rate": 0.000537236521070941,
+      "loss": 3.2244,
+      "step": 12835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4898160398006439,
+      "learning_rate": 0.0005372270969102156,
+      "loss": 3.2835,
+      "step": 12836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5112550854682922,
+      "learning_rate": 0.0005372176721246795,
+      "loss": 3.2138,
+      "step": 12837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.541236937046051,
+      "learning_rate": 0.0005372082467143578,
+      "loss": 3.1371,
+      "step": 12838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5265771150588989,
+      "learning_rate": 0.0005371988206792752,
+      "loss": 3.2119,
+      "step": 12839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5151164531707764,
+      "learning_rate": 0.0005371893940194566,
+      "loss": 3.1975,
+      "step": 12840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4924558997154236,
+      "learning_rate": 0.0005371799667349267,
+      "loss": 3.1876,
+      "step": 12841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5264354348182678,
+      "learning_rate": 0.0005371705388257105,
+      "loss": 3.152,
+      "step": 12842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5298112034797668,
+      "learning_rate": 0.0005371611102918327,
+      "loss": 3.111,
+      "step": 12843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5277727842330933,
+      "learning_rate": 0.0005371516811333182,
+      "loss": 3.1264,
+      "step": 12844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5544726252555847,
+      "learning_rate": 0.0005371422513501919,
+      "loss": 3.0475,
+      "step": 12845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4931381940841675,
+      "learning_rate": 0.0005371328209424783,
+      "loss": 3.3639,
+      "step": 12846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4993634819984436,
+      "learning_rate": 0.0005371233899102027,
+      "loss": 3.1516,
+      "step": 12847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5138599276542664,
+      "learning_rate": 0.0005371139582533896,
+      "loss": 3.198,
+      "step": 12848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5093522071838379,
+      "learning_rate": 0.000537104525972064,
+      "loss": 3.03,
+      "step": 12849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4839041531085968,
+      "learning_rate": 0.0005370950930662508,
+      "loss": 3.2278,
+      "step": 12850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48701995611190796,
+      "learning_rate": 0.0005370856595359746,
+      "loss": 3.2649,
+      "step": 12851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5446614027023315,
+      "learning_rate": 0.0005370762253812605,
+      "loss": 3.0454,
+      "step": 12852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.500033438205719,
+      "learning_rate": 0.0005370667906021332,
+      "loss": 3.0629,
+      "step": 12853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5227843523025513,
+      "learning_rate": 0.0005370573551986178,
+      "loss": 3.3169,
+      "step": 12854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4906866252422333,
+      "learning_rate": 0.0005370479191707387,
+      "loss": 3.2001,
+      "step": 12855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5372707843780518,
+      "learning_rate": 0.0005370384825185211,
+      "loss": 3.0848,
+      "step": 12856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5140169262886047,
+      "learning_rate": 0.0005370290452419898,
+      "loss": 3.3082,
+      "step": 12857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5246527791023254,
+      "learning_rate": 0.0005370196073411696,
+      "loss": 3.0989,
+      "step": 12858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5074948668479919,
+      "learning_rate": 0.0005370101688160852,
+      "loss": 3.1239,
+      "step": 12859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5050709843635559,
+      "learning_rate": 0.0005370007296667617,
+      "loss": 3.1019,
+      "step": 12860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5377639532089233,
+      "learning_rate": 0.000536991289893224,
+      "loss": 3.1435,
+      "step": 12861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.526069164276123,
+      "learning_rate": 0.0005369818494954968,
+      "loss": 3.48,
+      "step": 12862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47031062841415405,
+      "learning_rate": 0.0005369724084736051,
+      "loss": 3.3851,
+      "step": 12863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.510369598865509,
+      "learning_rate": 0.0005369629668275736,
+      "loss": 3.2161,
+      "step": 12864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5503350496292114,
+      "learning_rate": 0.0005369535245574272,
+      "loss": 3.0545,
+      "step": 12865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48866593837738037,
+      "learning_rate": 0.0005369440816631909,
+      "loss": 3.3215,
+      "step": 12866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.553887665271759,
+      "learning_rate": 0.0005369346381448894,
+      "loss": 3.1592,
+      "step": 12867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5026930570602417,
+      "learning_rate": 0.0005369251940025478,
+      "loss": 3.1913,
+      "step": 12868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49965840578079224,
+      "learning_rate": 0.0005369157492361907,
+      "loss": 3.1695,
+      "step": 12869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5119639039039612,
+      "learning_rate": 0.0005369063038458432,
+      "loss": 3.1687,
+      "step": 12870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5644335746765137,
+      "learning_rate": 0.00053689685783153,
+      "loss": 3.3533,
+      "step": 12871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5100526213645935,
+      "learning_rate": 0.0005368874111932761,
+      "loss": 3.2576,
+      "step": 12872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5507298111915588,
+      "learning_rate": 0.0005368779639311064,
+      "loss": 3.2868,
+      "step": 12873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.519503116607666,
+      "learning_rate": 0.0005368685160450457,
+      "loss": 3.2348,
+      "step": 12874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5227647423744202,
+      "learning_rate": 0.000536859067535119,
+      "loss": 3.1734,
+      "step": 12875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5177075862884521,
+      "learning_rate": 0.000536849618401351,
+      "loss": 2.9512,
+      "step": 12876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49735453724861145,
+      "learning_rate": 0.0005368401686437667,
+      "loss": 3.137,
+      "step": 12877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5045885443687439,
+      "learning_rate": 0.0005368307182623909,
+      "loss": 3.2829,
+      "step": 12878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4974973499774933,
+      "learning_rate": 0.0005368212672572487,
+      "loss": 3.4202,
+      "step": 12879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49687159061431885,
+      "learning_rate": 0.0005368118156283648,
+      "loss": 3.3512,
+      "step": 12880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5176152586936951,
+      "learning_rate": 0.0005368023633757642,
+      "loss": 3.3083,
+      "step": 12881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4925430417060852,
+      "learning_rate": 0.0005367929104994717,
+      "loss": 3.1931,
+      "step": 12882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5005999207496643,
+      "learning_rate": 0.0005367834569995122,
+      "loss": 3.2917,
+      "step": 12883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5244042277336121,
+      "learning_rate": 0.0005367740028759108,
+      "loss": 3.0702,
+      "step": 12884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.567849338054657,
+      "learning_rate": 0.0005367645481286921,
+      "loss": 3.2763,
+      "step": 12885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5182596445083618,
+      "learning_rate": 0.0005367550927578812,
+      "loss": 3.2015,
+      "step": 12886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5369324684143066,
+      "learning_rate": 0.0005367456367635029,
+      "loss": 3.1494,
+      "step": 12887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5310854911804199,
+      "learning_rate": 0.0005367361801455823,
+      "loss": 3.4733,
+      "step": 12888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4995528757572174,
+      "learning_rate": 0.0005367267229041441,
+      "loss": 3.316,
+      "step": 12889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234096646308899,
+      "learning_rate": 0.000536717265039213,
+      "loss": 3.3181,
+      "step": 12890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5279734134674072,
+      "learning_rate": 0.0005367078065508146,
+      "loss": 3.3797,
+      "step": 12891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.563451886177063,
+      "learning_rate": 0.0005366983474389732,
+      "loss": 3.2402,
+      "step": 12892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5074998140335083,
+      "learning_rate": 0.0005366888877037138,
+      "loss": 3.2033,
+      "step": 12893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5854350924491882,
+      "learning_rate": 0.0005366794273450615,
+      "loss": 3.3255,
+      "step": 12894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5305065512657166,
+      "learning_rate": 0.0005366699663630413,
+      "loss": 3.1053,
+      "step": 12895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5451061725616455,
+      "learning_rate": 0.0005366605047576778,
+      "loss": 3.2225,
+      "step": 12896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48388218879699707,
+      "learning_rate": 0.000536651042528996,
+      "loss": 3.3632,
+      "step": 12897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5109940767288208,
+      "learning_rate": 0.000536641579677021,
+      "loss": 3.1846,
+      "step": 12898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5055460333824158,
+      "learning_rate": 0.0005366321162017778,
+      "loss": 3.1477,
+      "step": 12899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5371001958847046,
+      "learning_rate": 0.0005366226521032908,
+      "loss": 3.1253,
+      "step": 12900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5191012024879456,
+      "learning_rate": 0.0005366131873815855,
+      "loss": 3.2262,
+      "step": 12901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5225028991699219,
+      "learning_rate": 0.0005366037220366866,
+      "loss": 3.5542,
+      "step": 12902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49432340264320374,
+      "learning_rate": 0.0005365942560686189,
+      "loss": 3.3985,
+      "step": 12903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5132206082344055,
+      "learning_rate": 0.0005365847894774076,
+      "loss": 3.161,
+      "step": 12904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.550641655921936,
+      "learning_rate": 0.0005365753222630774,
+      "loss": 3.4166,
+      "step": 12905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5244418382644653,
+      "learning_rate": 0.0005365658544256533,
+      "loss": 3.1923,
+      "step": 12906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5510836839675903,
+      "learning_rate": 0.0005365563859651604,
+      "loss": 3.1081,
+      "step": 12907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5539757013320923,
+      "learning_rate": 0.0005365469168816235,
+      "loss": 3.3491,
+      "step": 12908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5009015798568726,
+      "learning_rate": 0.0005365374471750674,
+      "loss": 3.2821,
+      "step": 12909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4771251976490021,
+      "learning_rate": 0.0005365279768455173,
+      "loss": 3.1718,
+      "step": 12910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5260068774223328,
+      "learning_rate": 0.000536518505892998,
+      "loss": 3.1779,
+      "step": 12911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.526455283164978,
+      "learning_rate": 0.0005365090343175345,
+      "loss": 3.057,
+      "step": 12912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5597702860832214,
+      "learning_rate": 0.0005364995621191516,
+      "loss": 3.1074,
+      "step": 12913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5866166353225708,
+      "learning_rate": 0.0005364900892978746,
+      "loss": 3.0222,
+      "step": 12914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49002647399902344,
+      "learning_rate": 0.000536480615853728,
+      "loss": 3.3856,
+      "step": 12915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5181958675384521,
+      "learning_rate": 0.0005364711417867371,
+      "loss": 3.288,
+      "step": 12916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5336577892303467,
+      "learning_rate": 0.0005364616670969266,
+      "loss": 3.3546,
+      "step": 12917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5243266224861145,
+      "learning_rate": 0.0005364521917843217,
+      "loss": 3.3334,
+      "step": 12918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5071207880973816,
+      "learning_rate": 0.0005364427158489472,
+      "loss": 3.1865,
+      "step": 12919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5273388028144836,
+      "learning_rate": 0.000536433239290828,
+      "loss": 3.3596,
+      "step": 12920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5106172561645508,
+      "learning_rate": 0.0005364237621099893,
+      "loss": 3.2821,
+      "step": 12921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5156787037849426,
+      "learning_rate": 0.0005364142843064558,
+      "loss": 3.2522,
+      "step": 12922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.593102216720581,
+      "learning_rate": 0.0005364048058802527,
+      "loss": 3.1004,
+      "step": 12923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5650702118873596,
+      "learning_rate": 0.0005363953268314048,
+      "loss": 3.3883,
+      "step": 12924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5169339179992676,
+      "learning_rate": 0.0005363858471599369,
+      "loss": 3.2,
+      "step": 12925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5006090998649597,
+      "learning_rate": 0.0005363763668658744,
+      "loss": 3.3361,
+      "step": 12926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.533066987991333,
+      "learning_rate": 0.000536366885949242,
+      "loss": 3.225,
+      "step": 12927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5449780225753784,
+      "learning_rate": 0.0005363574044100647,
+      "loss": 3.1673,
+      "step": 12928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49450409412384033,
+      "learning_rate": 0.0005363479222483674,
+      "loss": 3.3605,
+      "step": 12929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49364954233169556,
+      "learning_rate": 0.0005363384394641753,
+      "loss": 3.1942,
+      "step": 12930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4991125762462616,
+      "learning_rate": 0.0005363289560575131,
+      "loss": 3.0315,
+      "step": 12931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6006268858909607,
+      "learning_rate": 0.000536319472028406,
+      "loss": 2.9982,
+      "step": 12932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49599677324295044,
+      "learning_rate": 0.0005363099873768787,
+      "loss": 3.0228,
+      "step": 12933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5204678177833557,
+      "learning_rate": 0.0005363005021029566,
+      "loss": 3.0848,
+      "step": 12934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5052371621131897,
+      "learning_rate": 0.0005362910162066644,
+      "loss": 3.0953,
+      "step": 12935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5127009153366089,
+      "learning_rate": 0.0005362815296880272,
+      "loss": 3.5021,
+      "step": 12936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5664146542549133,
+      "learning_rate": 0.0005362720425470698,
+      "loss": 3.1467,
+      "step": 12937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5321229100227356,
+      "learning_rate": 0.0005362625547838173,
+      "loss": 3.3785,
+      "step": 12938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5562167167663574,
+      "learning_rate": 0.0005362530663982948,
+      "loss": 3.4932,
+      "step": 12939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.508590817451477,
+      "learning_rate": 0.0005362435773905271,
+      "loss": 3.2824,
+      "step": 12940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.511133074760437,
+      "learning_rate": 0.0005362340877605394,
+      "loss": 3.0221,
+      "step": 12941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5249322652816772,
+      "learning_rate": 0.0005362245975083566,
+      "loss": 3.1149,
+      "step": 12942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5249238014221191,
+      "learning_rate": 0.0005362151066340035,
+      "loss": 3.0869,
+      "step": 12943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5448753237724304,
+      "learning_rate": 0.0005362056151375054,
+      "loss": 3.0745,
+      "step": 12944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49821531772613525,
+      "learning_rate": 0.0005361961230188871,
+      "loss": 3.2544,
+      "step": 12945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5204089879989624,
+      "learning_rate": 0.0005361866302781736,
+      "loss": 3.0721,
+      "step": 12946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5314644575119019,
+      "learning_rate": 0.0005361771369153901,
+      "loss": 3.1426,
+      "step": 12947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6273903250694275,
+      "learning_rate": 0.0005361676429305615,
+      "loss": 3.1625,
+      "step": 12948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.510073184967041,
+      "learning_rate": 0.0005361581483237127,
+      "loss": 3.3129,
+      "step": 12949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5302249193191528,
+      "learning_rate": 0.0005361486530948688,
+      "loss": 3.4637,
+      "step": 12950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49914729595184326,
+      "learning_rate": 0.0005361391572440547,
+      "loss": 3.4225,
+      "step": 12951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5495375394821167,
+      "learning_rate": 0.0005361296607712956,
+      "loss": 3.0738,
+      "step": 12952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5367246270179749,
+      "learning_rate": 0.0005361201636766165,
+      "loss": 3.1163,
+      "step": 12953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5100537538528442,
+      "learning_rate": 0.0005361106659600423,
+      "loss": 3.1204,
+      "step": 12954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49203306436538696,
+      "learning_rate": 0.000536101167621598,
+      "loss": 3.3834,
+      "step": 12955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5398481488227844,
+      "learning_rate": 0.0005360916686613087,
+      "loss": 3.2211,
+      "step": 12956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5405081510543823,
+      "learning_rate": 0.0005360821690791992,
+      "loss": 3.1053,
+      "step": 12957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5752179622650146,
+      "learning_rate": 0.000536072668875295,
+      "loss": 2.9739,
+      "step": 12958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4982735514640808,
+      "learning_rate": 0.0005360631680496206,
+      "loss": 3.289,
+      "step": 12959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.500749409198761,
+      "learning_rate": 0.0005360536666022014,
+      "loss": 3.1595,
+      "step": 12960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5099477171897888,
+      "learning_rate": 0.000536044164533062,
+      "loss": 3.2093,
+      "step": 12961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5565930604934692,
+      "learning_rate": 0.000536034661842228,
+      "loss": 3.0367,
+      "step": 12962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5251843929290771,
+      "learning_rate": 0.0005360251585297239,
+      "loss": 3.1134,
+      "step": 12963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5190328359603882,
+      "learning_rate": 0.0005360156545955752,
+      "loss": 3.377,
+      "step": 12964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4984540343284607,
+      "learning_rate": 0.0005360061500398065,
+      "loss": 3.4222,
+      "step": 12965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5332692861557007,
+      "learning_rate": 0.0005359966448624431,
+      "loss": 3.0324,
+      "step": 12966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5812960863113403,
+      "learning_rate": 0.00053598713906351,
+      "loss": 3.3179,
+      "step": 12967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.498300164937973,
+      "learning_rate": 0.0005359776326430321,
+      "loss": 3.3108,
+      "step": 12968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4898775517940521,
+      "learning_rate": 0.0005359681256010345,
+      "loss": 3.215,
+      "step": 12969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5427769422531128,
+      "learning_rate": 0.0005359586179375424,
+      "loss": 3.2795,
+      "step": 12970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.517987072467804,
+      "learning_rate": 0.0005359491096525806,
+      "loss": 2.9976,
+      "step": 12971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5325692296028137,
+      "learning_rate": 0.0005359396007461743,
+      "loss": 3.327,
+      "step": 12972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5042824149131775,
+      "learning_rate": 0.0005359300912183485,
+      "loss": 3.1492,
+      "step": 12973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5307673811912537,
+      "learning_rate": 0.0005359205810691282,
+      "loss": 3.3302,
+      "step": 12974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5268490314483643,
+      "learning_rate": 0.0005359110702985385,
+      "loss": 3.3567,
+      "step": 12975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5066551566123962,
+      "learning_rate": 0.0005359015589066046,
+      "loss": 3.2215,
+      "step": 12976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.503842294216156,
+      "learning_rate": 0.0005358920468933511,
+      "loss": 3.2734,
+      "step": 12977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5022884011268616,
+      "learning_rate": 0.0005358825342588035,
+      "loss": 3.2803,
+      "step": 12978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5310874581336975,
+      "learning_rate": 0.0005358730210029869,
+      "loss": 3.2037,
+      "step": 12979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.509129524230957,
+      "learning_rate": 0.0005358635071259259,
+      "loss": 3.3409,
+      "step": 12980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.548210620880127,
+      "learning_rate": 0.0005358539926276459,
+      "loss": 3.2471,
+      "step": 12981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5100131630897522,
+      "learning_rate": 0.0005358444775081718,
+      "loss": 3.1354,
+      "step": 12982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5340960621833801,
+      "learning_rate": 0.0005358349617675289,
+      "loss": 3.2023,
+      "step": 12983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5364866256713867,
+      "learning_rate": 0.000535825445405742,
+      "loss": 3.2475,
+      "step": 12984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5168478488922119,
+      "learning_rate": 0.0005358159284228363,
+      "loss": 3.0523,
+      "step": 12985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49745577573776245,
+      "learning_rate": 0.0005358064108188366,
+      "loss": 3.1681,
+      "step": 12986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5580443739891052,
+      "learning_rate": 0.0005357968925937685,
+      "loss": 3.4102,
+      "step": 12987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5218451023101807,
+      "learning_rate": 0.0005357873737476565,
+      "loss": 3.279,
+      "step": 12988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5404812693595886,
+      "learning_rate": 0.0005357778542805262,
+      "loss": 3.4815,
+      "step": 12989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5017077326774597,
+      "learning_rate": 0.0005357683341924023,
+      "loss": 3.3535,
+      "step": 12990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49366968870162964,
+      "learning_rate": 0.00053575881348331,
+      "loss": 3.1195,
+      "step": 12991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5663856863975525,
+      "learning_rate": 0.0005357492921532743,
+      "loss": 3.3108,
+      "step": 12992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5067052841186523,
+      "learning_rate": 0.0005357397702023204,
+      "loss": 3.3169,
+      "step": 12993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5267317295074463,
+      "learning_rate": 0.0005357302476304732,
+      "loss": 3.0451,
+      "step": 12994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5492493510246277,
+      "learning_rate": 0.000535720724437758,
+      "loss": 3.091,
+      "step": 12995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4951799511909485,
+      "learning_rate": 0.0005357112006241998,
+      "loss": 3.1907,
+      "step": 12996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5452933311462402,
+      "learning_rate": 0.0005357016761898236,
+      "loss": 3.1006,
+      "step": 12997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5037853121757507,
+      "learning_rate": 0.0005356921511346545,
+      "loss": 3.2203,
+      "step": 12998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48112764954566956,
+      "learning_rate": 0.0005356826254587177,
+      "loss": 3.2942,
+      "step": 12999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49758175015449524,
+      "learning_rate": 0.0005356730991620382,
+      "loss": 3.2869,
+      "step": 13000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5027998089790344,
+      "learning_rate": 0.0005356635722446412,
+      "loss": 3.1453,
+      "step": 13001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5159273147583008,
+      "learning_rate": 0.0005356540447065516,
+      "loss": 3.162,
+      "step": 13002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5522950887680054,
+      "learning_rate": 0.0005356445165477947,
+      "loss": 3.164,
+      "step": 13003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5712302923202515,
+      "learning_rate": 0.0005356349877683954,
+      "loss": 3.1857,
+      "step": 13004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4907442331314087,
+      "learning_rate": 0.0005356254583683789,
+      "loss": 3.4676,
+      "step": 13005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49316293001174927,
+      "learning_rate": 0.0005356159283477703,
+      "loss": 3.274,
+      "step": 13006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5303720831871033,
+      "learning_rate": 0.0005356063977065948,
+      "loss": 3.1756,
+      "step": 13007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4843452572822571,
+      "learning_rate": 0.0005355968664448772,
+      "loss": 3.0976,
+      "step": 13008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5406621098518372,
+      "learning_rate": 0.0005355873345626429,
+      "loss": 3.216,
+      "step": 13009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5066521167755127,
+      "learning_rate": 0.0005355778020599168,
+      "loss": 3.3491,
+      "step": 13010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5344280004501343,
+      "learning_rate": 0.0005355682689367243,
+      "loss": 3.3312,
+      "step": 13011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5455307364463806,
+      "learning_rate": 0.0005355587351930902,
+      "loss": 3.2931,
+      "step": 13012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46568596363067627,
+      "learning_rate": 0.0005355492008290397,
+      "loss": 3.2364,
+      "step": 13013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4989795684814453,
+      "learning_rate": 0.000535539665844598,
+      "loss": 3.3095,
+      "step": 13014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5595158338546753,
+      "learning_rate": 0.0005355301302397901,
+      "loss": 3.3839,
+      "step": 13015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4899807870388031,
+      "learning_rate": 0.0005355205940146412,
+      "loss": 3.3119,
+      "step": 13016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48893171548843384,
+      "learning_rate": 0.0005355110571691764,
+      "loss": 3.2628,
+      "step": 13017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5232262015342712,
+      "learning_rate": 0.0005355015197034207,
+      "loss": 3.281,
+      "step": 13018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5392224788665771,
+      "learning_rate": 0.0005354919816173995,
+      "loss": 3.2029,
+      "step": 13019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5342455506324768,
+      "learning_rate": 0.0005354824429111376,
+      "loss": 3.2832,
+      "step": 13020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48179611563682556,
+      "learning_rate": 0.0005354729035846603,
+      "loss": 3.0406,
+      "step": 13021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4980888068675995,
+      "learning_rate": 0.0005354633636379927,
+      "loss": 3.2632,
+      "step": 13022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5115153193473816,
+      "learning_rate": 0.0005354538230711598,
+      "loss": 3.1863,
+      "step": 13023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5516042709350586,
+      "learning_rate": 0.0005354442818841869,
+      "loss": 3.2411,
+      "step": 13024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4984360933303833,
+      "learning_rate": 0.0005354347400770992,
+      "loss": 3.1063,
+      "step": 13025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5270858407020569,
+      "learning_rate": 0.0005354251976499217,
+      "loss": 3.2525,
+      "step": 13026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5250088572502136,
+      "learning_rate": 0.0005354156546026794,
+      "loss": 3.3425,
+      "step": 13027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5461434125900269,
+      "learning_rate": 0.0005354061109353976,
+      "loss": 3.3327,
+      "step": 13028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5004782676696777,
+      "learning_rate": 0.0005353965666481015,
+      "loss": 3.1885,
+      "step": 13029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5391425490379333,
+      "learning_rate": 0.0005353870217408161,
+      "loss": 3.0705,
+      "step": 13030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5242264866828918,
+      "learning_rate": 0.0005353774762135666,
+      "loss": 3.1418,
+      "step": 13031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5164439678192139,
+      "learning_rate": 0.000535367930066378,
+      "loss": 3.2949,
+      "step": 13032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5032666921615601,
+      "learning_rate": 0.0005353583832992758,
+      "loss": 3.257,
+      "step": 13033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5135819315910339,
+      "learning_rate": 0.0005353488359122848,
+      "loss": 3.1992,
+      "step": 13034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5079220533370972,
+      "learning_rate": 0.0005353392879054302,
+      "loss": 3.2749,
+      "step": 13035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5061142444610596,
+      "learning_rate": 0.0005353297392787373,
+      "loss": 3.1821,
+      "step": 13036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5289643406867981,
+      "learning_rate": 0.000535320190032231,
+      "loss": 3.29,
+      "step": 13037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5101862549781799,
+      "learning_rate": 0.0005353106401659367,
+      "loss": 3.2464,
+      "step": 13038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5174379348754883,
+      "learning_rate": 0.0005353010896798796,
+      "loss": 3.1531,
+      "step": 13039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5504981875419617,
+      "learning_rate": 0.0005352915385740845,
+      "loss": 2.8933,
+      "step": 13040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.534186840057373,
+      "learning_rate": 0.0005352819868485769,
+      "loss": 3.2163,
+      "step": 13041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5198753476142883,
+      "learning_rate": 0.0005352724345033818,
+      "loss": 3.1988,
+      "step": 13042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5213879942893982,
+      "learning_rate": 0.0005352628815385244,
+      "loss": 3.1679,
+      "step": 13043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5233698487281799,
+      "learning_rate": 0.0005352533279540298,
+      "loss": 3.0413,
+      "step": 13044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5285971164703369,
+      "learning_rate": 0.0005352437737499232,
+      "loss": 3.0021,
+      "step": 13045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5171172618865967,
+      "learning_rate": 0.0005352342189262298,
+      "loss": 3.1727,
+      "step": 13046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5061637163162231,
+      "learning_rate": 0.0005352246634829748,
+      "loss": 3.5304,
+      "step": 13047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.535328209400177,
+      "learning_rate": 0.0005352151074201832,
+      "loss": 3.1083,
+      "step": 13048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5753775835037231,
+      "learning_rate": 0.0005352055507378804,
+      "loss": 3.189,
+      "step": 13049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5696597695350647,
+      "learning_rate": 0.0005351959934360913,
+      "loss": 3.121,
+      "step": 13050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5169010758399963,
+      "learning_rate": 0.0005351864355148413,
+      "loss": 3.13,
+      "step": 13051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6170527935028076,
+      "learning_rate": 0.0005351768769741556,
+      "loss": 3.1525,
+      "step": 13052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4995497465133667,
+      "learning_rate": 0.000535167317814059,
+      "loss": 3.2603,
+      "step": 13053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5279785990715027,
+      "learning_rate": 0.0005351577580345771,
+      "loss": 3.3865,
+      "step": 13054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5221437811851501,
+      "learning_rate": 0.000535148197635735,
+      "loss": 3.3428,
+      "step": 13055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5271676182746887,
+      "learning_rate": 0.0005351386366175577,
+      "loss": 3.1217,
+      "step": 13056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5257646441459656,
+      "learning_rate": 0.0005351290749800705,
+      "loss": 3.2126,
+      "step": 13057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5410515666007996,
+      "learning_rate": 0.0005351195127232986,
+      "loss": 3.0734,
+      "step": 13058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5229355692863464,
+      "learning_rate": 0.0005351099498472671,
+      "loss": 3.1239,
+      "step": 13059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5615711212158203,
+      "learning_rate": 0.0005351003863520013,
+      "loss": 3.0884,
+      "step": 13060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5424404740333557,
+      "learning_rate": 0.0005350908222375263,
+      "loss": 3.2056,
+      "step": 13061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.537496030330658,
+      "learning_rate": 0.0005350812575038673,
+      "loss": 3.3297,
+      "step": 13062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49330636858940125,
+      "learning_rate": 0.0005350716921510495,
+      "loss": 3.385,
+      "step": 13063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5436633229255676,
+      "learning_rate": 0.0005350621261790982,
+      "loss": 3.198,
+      "step": 13064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4999907910823822,
+      "learning_rate": 0.0005350525595880384,
+      "loss": 3.4466,
+      "step": 13065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5513013005256653,
+      "learning_rate": 0.0005350429923778954,
+      "loss": 3.4655,
+      "step": 13066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5286356806755066,
+      "learning_rate": 0.0005350334245486943,
+      "loss": 3.3866,
+      "step": 13067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5245810747146606,
+      "learning_rate": 0.0005350238561004606,
+      "loss": 2.9812,
+      "step": 13068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.534548282623291,
+      "learning_rate": 0.0005350142870332192,
+      "loss": 3.1293,
+      "step": 13069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5076059699058533,
+      "learning_rate": 0.0005350047173469953,
+      "loss": 3.3001,
+      "step": 13070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5420432686805725,
+      "learning_rate": 0.0005349951470418144,
+      "loss": 3.0998,
+      "step": 13071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5068901777267456,
+      "learning_rate": 0.0005349855761177014,
+      "loss": 3.2598,
+      "step": 13072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5439526438713074,
+      "learning_rate": 0.0005349760045746816,
+      "loss": 3.4108,
+      "step": 13073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4926943778991699,
+      "learning_rate": 0.0005349664324127803,
+      "loss": 3.2626,
+      "step": 13074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5149872899055481,
+      "learning_rate": 0.0005349568596320225,
+      "loss": 3.2164,
+      "step": 13075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5110384225845337,
+      "learning_rate": 0.0005349472862324337,
+      "loss": 3.1819,
+      "step": 13076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5247209072113037,
+      "learning_rate": 0.0005349377122140388,
+      "loss": 3.2829,
+      "step": 13077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.553516149520874,
+      "learning_rate": 0.0005349281375768634,
+      "loss": 3.2516,
+      "step": 13078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5458715558052063,
+      "learning_rate": 0.0005349185623209324,
+      "loss": 3.2085,
+      "step": 13079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4971589744091034,
+      "learning_rate": 0.0005349089864462711,
+      "loss": 3.255,
+      "step": 13080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5303791761398315,
+      "learning_rate": 0.0005348994099529047,
+      "loss": 3.2376,
+      "step": 13081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5025544762611389,
+      "learning_rate": 0.0005348898328408584,
+      "loss": 3.2325,
+      "step": 13082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49780189990997314,
+      "learning_rate": 0.0005348802551101578,
+      "loss": 2.9287,
+      "step": 13083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5410318970680237,
+      "learning_rate": 0.0005348706767608275,
+      "loss": 3.2634,
+      "step": 13084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5140213966369629,
+      "learning_rate": 0.0005348610977928931,
+      "loss": 3.1781,
+      "step": 13085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5628545880317688,
+      "learning_rate": 0.0005348515182063799,
+      "loss": 3.239,
+      "step": 13086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5181366205215454,
+      "learning_rate": 0.0005348419380013128,
+      "loss": 3.2054,
+      "step": 13087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5096615552902222,
+      "learning_rate": 0.0005348323571777174,
+      "loss": 3.1634,
+      "step": 13088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5191645622253418,
+      "learning_rate": 0.0005348227757356187,
+      "loss": 3.2104,
+      "step": 13089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5057798624038696,
+      "learning_rate": 0.0005348131936750419,
+      "loss": 3.177,
+      "step": 13090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4993162453174591,
+      "learning_rate": 0.0005348036109960125,
+      "loss": 3.3898,
+      "step": 13091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5280262231826782,
+      "learning_rate": 0.0005347940276985555,
+      "loss": 3.3019,
+      "step": 13092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5497944951057434,
+      "learning_rate": 0.0005347844437826962,
+      "loss": 3.2199,
+      "step": 13093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.47965291142463684,
+      "learning_rate": 0.0005347748592484599,
+      "loss": 3.2582,
+      "step": 13094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5250326991081238,
+      "learning_rate": 0.0005347652740958718,
+      "loss": 3.3034,
+      "step": 13095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5235711336135864,
+      "learning_rate": 0.0005347556883249572,
+      "loss": 3.5181,
+      "step": 13096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4954131245613098,
+      "learning_rate": 0.0005347461019357412,
+      "loss": 3.2653,
+      "step": 13097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5206896066665649,
+      "learning_rate": 0.0005347365149282492,
+      "loss": 3.1443,
+      "step": 13098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5158017873764038,
+      "learning_rate": 0.0005347269273025064,
+      "loss": 3.2635,
+      "step": 13099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5218068957328796,
+      "learning_rate": 0.0005347173390585381,
+      "loss": 3.2463,
+      "step": 13100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5204295516014099,
+      "learning_rate": 0.0005347077501963694,
+      "loss": 3.2756,
+      "step": 13101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5045682191848755,
+      "learning_rate": 0.0005346981607160257,
+      "loss": 3.2225,
+      "step": 13102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5521380305290222,
+      "learning_rate": 0.0005346885706175321,
+      "loss": 3.0098,
+      "step": 13103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5172122716903687,
+      "learning_rate": 0.0005346789799009141,
+      "loss": 3.2474,
+      "step": 13104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5108321905136108,
+      "learning_rate": 0.0005346693885661968,
+      "loss": 3.132,
+      "step": 13105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4934992790222168,
+      "learning_rate": 0.0005346597966134056,
+      "loss": 3.0254,
+      "step": 13106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5101711750030518,
+      "learning_rate": 0.0005346502040425655,
+      "loss": 3.4279,
+      "step": 13107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5210503935813904,
+      "learning_rate": 0.000534640610853702,
+      "loss": 3.2178,
+      "step": 13108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.496926873922348,
+      "learning_rate": 0.0005346310170468402,
+      "loss": 3.1698,
+      "step": 13109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49992474913597107,
+      "learning_rate": 0.0005346214226220055,
+      "loss": 3.2576,
+      "step": 13110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5001792311668396,
+      "learning_rate": 0.0005346118275792232,
+      "loss": 3.4211,
+      "step": 13111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5183314681053162,
+      "learning_rate": 0.0005346022319185185,
+      "loss": 3.19,
+      "step": 13112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4935111403465271,
+      "learning_rate": 0.0005345926356399166,
+      "loss": 3.3512,
+      "step": 13113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5626166462898254,
+      "learning_rate": 0.0005345830387434428,
+      "loss": 3.3638,
+      "step": 13114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5216866731643677,
+      "learning_rate": 0.0005345734412291226,
+      "loss": 3.244,
+      "step": 13115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5332128405570984,
+      "learning_rate": 0.000534563843096981,
+      "loss": 3.2552,
+      "step": 13116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.47880974411964417,
+      "learning_rate": 0.0005345542443470434,
+      "loss": 3.1028,
+      "step": 13117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4967505931854248,
+      "learning_rate": 0.000534544644979335,
+      "loss": 3.004,
+      "step": 13118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5178172588348389,
+      "learning_rate": 0.0005345350449938811,
+      "loss": 3.2698,
+      "step": 13119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5368995070457458,
+      "learning_rate": 0.0005345254443907072,
+      "loss": 3.3812,
+      "step": 13120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4906540513038635,
+      "learning_rate": 0.0005345158431698383,
+      "loss": 3.1979,
+      "step": 13121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5134233236312866,
+      "learning_rate": 0.0005345062413312998,
+      "loss": 3.3588,
+      "step": 13122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5044988989830017,
+      "learning_rate": 0.0005344966388751171,
+      "loss": 3.1731,
+      "step": 13123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5189638137817383,
+      "learning_rate": 0.0005344870358013152,
+      "loss": 3.1494,
+      "step": 13124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6163929104804993,
+      "learning_rate": 0.0005344774321099197,
+      "loss": 3.2465,
+      "step": 13125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5382804870605469,
+      "learning_rate": 0.0005344678278009557,
+      "loss": 3.1525,
+      "step": 13126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5891967415809631,
+      "learning_rate": 0.0005344582228744486,
+      "loss": 3.1272,
+      "step": 13127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5315770506858826,
+      "learning_rate": 0.0005344486173304236,
+      "loss": 2.9923,
+      "step": 13128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5364977121353149,
+      "learning_rate": 0.0005344390111689061,
+      "loss": 3.2317,
+      "step": 13129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5116824507713318,
+      "learning_rate": 0.0005344294043899215,
+      "loss": 3.1648,
+      "step": 13130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5573520660400391,
+      "learning_rate": 0.0005344197969934949,
+      "loss": 3.353,
+      "step": 13131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5192342400550842,
+      "learning_rate": 0.0005344101889796516,
+      "loss": 3.1479,
+      "step": 13132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49575215578079224,
+      "learning_rate": 0.0005344005803484171,
+      "loss": 3.4167,
+      "step": 13133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5160652995109558,
+      "learning_rate": 0.0005343909710998164,
+      "loss": 3.2124,
+      "step": 13134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5108956098556519,
+      "learning_rate": 0.0005343813612338751,
+      "loss": 3.3316,
+      "step": 13135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.528973400592804,
+      "learning_rate": 0.0005343717507506184,
+      "loss": 3.0333,
+      "step": 13136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5308817028999329,
+      "learning_rate": 0.0005343621396500716,
+      "loss": 3.1145,
+      "step": 13137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5422894358634949,
+      "learning_rate": 0.00053435252793226,
+      "loss": 3.2858,
+      "step": 13138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4965519607067108,
+      "learning_rate": 0.000534342915597209,
+      "loss": 3.2758,
+      "step": 13139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5743324160575867,
+      "learning_rate": 0.0005343333026449438,
+      "loss": 3.2823,
+      "step": 13140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4840207099914551,
+      "learning_rate": 0.0005343236890754898,
+      "loss": 3.0271,
+      "step": 13141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5335875749588013,
+      "learning_rate": 0.0005343140748888724,
+      "loss": 3.2321,
+      "step": 13142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5198099613189697,
+      "learning_rate": 0.0005343044600851166,
+      "loss": 3.1172,
+      "step": 13143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5045928359031677,
+      "learning_rate": 0.0005342948446642481,
+      "loss": 3.2568,
+      "step": 13144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49999576807022095,
+      "learning_rate": 0.0005342852286262921,
+      "loss": 3.2756,
+      "step": 13145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5080724954605103,
+      "learning_rate": 0.0005342756119712737,
+      "loss": 3.3996,
+      "step": 13146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48446428775787354,
+      "learning_rate": 0.0005342659946992187,
+      "loss": 3.1447,
+      "step": 13147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5702810883522034,
+      "learning_rate": 0.000534256376810152,
+      "loss": 3.2459,
+      "step": 13148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5255053639411926,
+      "learning_rate": 0.0005342467583040991,
+      "loss": 3.0702,
+      "step": 13149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5071660280227661,
+      "learning_rate": 0.0005342371391810853,
+      "loss": 3.3868,
+      "step": 13150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5754165053367615,
+      "learning_rate": 0.000534227519441136,
+      "loss": 3.2203,
+      "step": 13151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5076290965080261,
+      "learning_rate": 0.0005342178990842765,
+      "loss": 3.4923,
+      "step": 13152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5148435831069946,
+      "learning_rate": 0.0005342082781105321,
+      "loss": 3.3192,
+      "step": 13153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5352534651756287,
+      "learning_rate": 0.0005341986565199281,
+      "loss": 3.2992,
+      "step": 13154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5055733323097229,
+      "learning_rate": 0.0005341890343124902,
+      "loss": 3.1765,
+      "step": 13155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5165096521377563,
+      "learning_rate": 0.0005341794114882431,
+      "loss": 3.4136,
+      "step": 13156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53326416015625,
+      "learning_rate": 0.0005341697880472126,
+      "loss": 3.2336,
+      "step": 13157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.512313187122345,
+      "learning_rate": 0.000534160163989424,
+      "loss": 3.2201,
+      "step": 13158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5026753544807434,
+      "learning_rate": 0.0005341505393149026,
+      "loss": 3.3028,
+      "step": 13159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4956597685813904,
+      "learning_rate": 0.0005341409140236738,
+      "loss": 3.1856,
+      "step": 13160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5383108854293823,
+      "learning_rate": 0.0005341312881157628,
+      "loss": 3.2079,
+      "step": 13161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5031716823577881,
+      "learning_rate": 0.000534121661591195,
+      "loss": 3.363,
+      "step": 13162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.564416766166687,
+      "learning_rate": 0.0005341120344499959,
+      "loss": 3.1272,
+      "step": 13163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5095632672309875,
+      "learning_rate": 0.0005341024066921907,
+      "loss": 3.2337,
+      "step": 13164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.520420253276825,
+      "learning_rate": 0.0005340927783178049,
+      "loss": 3.2769,
+      "step": 13165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5509207248687744,
+      "learning_rate": 0.0005340831493268637,
+      "loss": 3.1734,
+      "step": 13166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5348342657089233,
+      "learning_rate": 0.0005340735197193924,
+      "loss": 3.1599,
+      "step": 13167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5090433955192566,
+      "learning_rate": 0.0005340638894954167,
+      "loss": 3.352,
+      "step": 13168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5051810145378113,
+      "learning_rate": 0.0005340542586549617,
+      "loss": 3.0708,
+      "step": 13169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5810332298278809,
+      "learning_rate": 0.0005340446271980528,
+      "loss": 3.34,
+      "step": 13170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5068036913871765,
+      "learning_rate": 0.0005340349951247156,
+      "loss": 3.2238,
+      "step": 13171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5000148415565491,
+      "learning_rate": 0.000534025362434975,
+      "loss": 3.3234,
+      "step": 13172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5451956987380981,
+      "learning_rate": 0.0005340157291288568,
+      "loss": 2.9447,
+      "step": 13173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.524332582950592,
+      "learning_rate": 0.000534006095206386,
+      "loss": 3.1001,
+      "step": 13174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48256099224090576,
+      "learning_rate": 0.0005339964606675883,
+      "loss": 3.1865,
+      "step": 13175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5315820574760437,
+      "learning_rate": 0.000533986825512489,
+      "loss": 3.0933,
+      "step": 13176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5178393721580505,
+      "learning_rate": 0.0005339771897411134,
+      "loss": 3.1253,
+      "step": 13177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48656055331230164,
+      "learning_rate": 0.0005339675533534869,
+      "loss": 3.2523,
+      "step": 13178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5785982608795166,
+      "learning_rate": 0.000533957916349635,
+      "loss": 3.1421,
+      "step": 13179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5141577124595642,
+      "learning_rate": 0.0005339482787295828,
+      "loss": 3.4436,
+      "step": 13180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5167346000671387,
+      "learning_rate": 0.000533938640493356,
+      "loss": 3.1199,
+      "step": 13181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5306900143623352,
+      "learning_rate": 0.0005339290016409797,
+      "loss": 3.3397,
+      "step": 13182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5225565433502197,
+      "learning_rate": 0.0005339193621724795,
+      "loss": 3.1148,
+      "step": 13183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5168359875679016,
+      "learning_rate": 0.0005339097220878808,
+      "loss": 3.0261,
+      "step": 13184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5109102725982666,
+      "learning_rate": 0.0005339000813872088,
+      "loss": 3.1654,
+      "step": 13185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5324611663818359,
+      "learning_rate": 0.0005338904400704891,
+      "loss": 3.291,
+      "step": 13186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5554729104042053,
+      "learning_rate": 0.0005338807981377469,
+      "loss": 3.4504,
+      "step": 13187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5127044320106506,
+      "learning_rate": 0.0005338711555890077,
+      "loss": 3.3947,
+      "step": 13188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5299892425537109,
+      "learning_rate": 0.000533861512424297,
+      "loss": 3.148,
+      "step": 13189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5451264381408691,
+      "learning_rate": 0.0005338518686436399,
+      "loss": 3.4564,
+      "step": 13190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5090279579162598,
+      "learning_rate": 0.0005338422242470621,
+      "loss": 3.114,
+      "step": 13191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.519496500492096,
+      "learning_rate": 0.0005338325792345888,
+      "loss": 2.9807,
+      "step": 13192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5295326113700867,
+      "learning_rate": 0.0005338229336062456,
+      "loss": 3.3167,
+      "step": 13193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5315036177635193,
+      "learning_rate": 0.0005338132873620577,
+      "loss": 3.0778,
+      "step": 13194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5226171016693115,
+      "learning_rate": 0.0005338036405020507,
+      "loss": 3.1866,
+      "step": 13195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49867236614227295,
+      "learning_rate": 0.0005337939930262497,
+      "loss": 3.1654,
+      "step": 13196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5161440372467041,
+      "learning_rate": 0.0005337843449346807,
+      "loss": 3.1951,
+      "step": 13197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49722862243652344,
+      "learning_rate": 0.0005337746962273684,
+      "loss": 3.2434,
+      "step": 13198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5370376706123352,
+      "learning_rate": 0.0005337650469043387,
+      "loss": 3.2291,
+      "step": 13199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5049730539321899,
+      "learning_rate": 0.0005337553969656166,
+      "loss": 3.291,
+      "step": 13200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5508888363838196,
+      "learning_rate": 0.000533745746411228,
+      "loss": 3.1669,
+      "step": 13201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49294355511665344,
+      "learning_rate": 0.0005337360952411981,
+      "loss": 3.0508,
+      "step": 13202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5044154524803162,
+      "learning_rate": 0.0005337264434555522,
+      "loss": 3.2915,
+      "step": 13203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5183781385421753,
+      "learning_rate": 0.000533716791054316,
+      "loss": 3.0519,
+      "step": 13204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.559165358543396,
+      "learning_rate": 0.0005337071380375146,
+      "loss": 3.2492,
+      "step": 13205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5487591624259949,
+      "learning_rate": 0.0005336974844051736,
+      "loss": 3.2115,
+      "step": 13206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5180480480194092,
+      "learning_rate": 0.0005336878301573185,
+      "loss": 3.2974,
+      "step": 13207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4945024251937866,
+      "learning_rate": 0.0005336781752939744,
+      "loss": 3.1486,
+      "step": 13208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4666735529899597,
+      "learning_rate": 0.0005336685198151672,
+      "loss": 3.1456,
+      "step": 13209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5270425081253052,
+      "learning_rate": 0.0005336588637209219,
+      "loss": 3.1666,
+      "step": 13210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5256056785583496,
+      "learning_rate": 0.0005336492070112642,
+      "loss": 3.1653,
+      "step": 13211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4974687695503235,
+      "learning_rate": 0.0005336395496862195,
+      "loss": 3.209,
+      "step": 13212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5208483934402466,
+      "learning_rate": 0.0005336298917458131,
+      "loss": 3.1261,
+      "step": 13213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5191110372543335,
+      "learning_rate": 0.0005336202331900707,
+      "loss": 3.0473,
+      "step": 13214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5726637244224548,
+      "learning_rate": 0.0005336105740190175,
+      "loss": 2.9428,
+      "step": 13215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.520617663860321,
+      "learning_rate": 0.000533600914232679,
+      "loss": 3.0333,
+      "step": 13216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51925128698349,
+      "learning_rate": 0.0005335912538310805,
+      "loss": 3.0626,
+      "step": 13217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5106421113014221,
+      "learning_rate": 0.0005335815928142478,
+      "loss": 3.2478,
+      "step": 13218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4910167455673218,
+      "learning_rate": 0.000533571931182206,
+      "loss": 3.4039,
+      "step": 13219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5045503973960876,
+      "learning_rate": 0.0005335622689349807,
+      "loss": 3.1105,
+      "step": 13220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53183513879776,
+      "learning_rate": 0.0005335526060725975,
+      "loss": 3.3134,
+      "step": 13221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390097498893738,
+      "learning_rate": 0.0005335429425950814,
+      "loss": 3.1459,
+      "step": 13222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53877192735672,
+      "learning_rate": 0.0005335332785024583,
+      "loss": 3.1017,
+      "step": 13223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5238897800445557,
+      "learning_rate": 0.0005335236137947536,
+      "loss": 3.1823,
+      "step": 13224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5160094499588013,
+      "learning_rate": 0.0005335139484719925,
+      "loss": 3.0258,
+      "step": 13225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.512133002281189,
+      "learning_rate": 0.0005335042825342005,
+      "loss": 3.1001,
+      "step": 13226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4885791838169098,
+      "learning_rate": 0.0005334946159814033,
+      "loss": 3.3233,
+      "step": 13227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5112274289131165,
+      "learning_rate": 0.0005334849488136263,
+      "loss": 3.269,
+      "step": 13228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4955732524394989,
+      "learning_rate": 0.0005334752810308948,
+      "loss": 3.3469,
+      "step": 13229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5273498296737671,
+      "learning_rate": 0.0005334656126332343,
+      "loss": 3.2339,
+      "step": 13230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5362980961799622,
+      "learning_rate": 0.0005334559436206702,
+      "loss": 3.2441,
+      "step": 13231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5081391334533691,
+      "learning_rate": 0.0005334462739932282,
+      "loss": 3.2659,
+      "step": 13232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5640556812286377,
+      "learning_rate": 0.0005334366037509337,
+      "loss": 3.192,
+      "step": 13233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5298064351081848,
+      "learning_rate": 0.000533426932893812,
+      "loss": 3.1988,
+      "step": 13234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5060878396034241,
+      "learning_rate": 0.0005334172614218887,
+      "loss": 3.3089,
+      "step": 13235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5255646705627441,
+      "learning_rate": 0.0005334075893351893,
+      "loss": 3.212,
+      "step": 13236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5083935856819153,
+      "learning_rate": 0.0005333979166337393,
+      "loss": 3.2419,
+      "step": 13237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5213572382926941,
+      "learning_rate": 0.000533388243317564,
+      "loss": 3.3522,
+      "step": 13238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.495538592338562,
+      "learning_rate": 0.000533378569386689,
+      "loss": 3.3924,
+      "step": 13239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5392500758171082,
+      "learning_rate": 0.0005333688948411398,
+      "loss": 2.9577,
+      "step": 13240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5462858080863953,
+      "learning_rate": 0.0005333592196809418,
+      "loss": 3.1583,
+      "step": 13241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.584762454032898,
+      "learning_rate": 0.0005333495439061206,
+      "loss": 3.1343,
+      "step": 13242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.520191490650177,
+      "learning_rate": 0.0005333398675167015,
+      "loss": 3.3106,
+      "step": 13243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5066478252410889,
+      "learning_rate": 0.0005333301905127101,
+      "loss": 3.3081,
+      "step": 13244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5514070391654968,
+      "learning_rate": 0.000533320512894172,
+      "loss": 3.1183,
+      "step": 13245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5034546256065369,
+      "learning_rate": 0.0005333108346611124,
+      "loss": 3.0596,
+      "step": 13246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5097323656082153,
+      "learning_rate": 0.0005333011558135572,
+      "loss": 3.1469,
+      "step": 13247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5640053153038025,
+      "learning_rate": 0.0005332914763515314,
+      "loss": 3.1606,
+      "step": 13248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5292096138000488,
+      "learning_rate": 0.0005332817962750609,
+      "loss": 3.1932,
+      "step": 13249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4695495367050171,
+      "learning_rate": 0.0005332721155841711,
+      "loss": 3.147,
+      "step": 13250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.512503445148468,
+      "learning_rate": 0.0005332624342788873,
+      "loss": 3.1208,
+      "step": 13251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5124956965446472,
+      "learning_rate": 0.0005332527523592353,
+      "loss": 3.1443,
+      "step": 13252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5241851210594177,
+      "learning_rate": 0.0005332430698252403,
+      "loss": 3.1076,
+      "step": 13253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5366250872612,
+      "learning_rate": 0.0005332333866769279,
+      "loss": 3.1135,
+      "step": 13254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.506049394607544,
+      "learning_rate": 0.0005332237029143238,
+      "loss": 3.411,
+      "step": 13255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5145257711410522,
+      "learning_rate": 0.0005332140185374532,
+      "loss": 3.2594,
+      "step": 13256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5252572894096375,
+      "learning_rate": 0.0005332043335463419,
+      "loss": 3.2745,
+      "step": 13257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4891444146633148,
+      "learning_rate": 0.0005331946479410152,
+      "loss": 3.124,
+      "step": 13258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.569198727607727,
+      "learning_rate": 0.0005331849617214987,
+      "loss": 3.3658,
+      "step": 13259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.627887487411499,
+      "learning_rate": 0.0005331752748878179,
+      "loss": 3.0741,
+      "step": 13260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5502414703369141,
+      "learning_rate": 0.0005331655874399982,
+      "loss": 3.1345,
+      "step": 13261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5486642122268677,
+      "learning_rate": 0.0005331558993780653,
+      "loss": 3.1981,
+      "step": 13262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5511811375617981,
+      "learning_rate": 0.0005331462107020446,
+      "loss": 3.1109,
+      "step": 13263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5167766809463501,
+      "learning_rate": 0.0005331365214119617,
+      "loss": 3.2288,
+      "step": 13264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5129167437553406,
+      "learning_rate": 0.000533126831507842,
+      "loss": 3.3583,
+      "step": 13265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5186481475830078,
+      "learning_rate": 0.0005331171409897112,
+      "loss": 3.18,
+      "step": 13266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5730112195014954,
+      "learning_rate": 0.0005331074498575946,
+      "loss": 3.0874,
+      "step": 13267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5106200575828552,
+      "learning_rate": 0.000533097758111518,
+      "loss": 3.2888,
+      "step": 13268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5146166682243347,
+      "learning_rate": 0.0005330880657515066,
+      "loss": 3.0715,
+      "step": 13269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49323368072509766,
+      "learning_rate": 0.0005330783727775861,
+      "loss": 3.1591,
+      "step": 13270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5173508524894714,
+      "learning_rate": 0.000533068679189782,
+      "loss": 3.092,
+      "step": 13271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51826411485672,
+      "learning_rate": 0.00053305898498812,
+      "loss": 3.228,
+      "step": 13272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5417203307151794,
+      "learning_rate": 0.0005330492901726255,
+      "loss": 3.1268,
+      "step": 13273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5598623752593994,
+      "learning_rate": 0.0005330395947433238,
+      "loss": 3.1918,
+      "step": 13274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5296295881271362,
+      "learning_rate": 0.0005330298987002408,
+      "loss": 3.1156,
+      "step": 13275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4983462989330292,
+      "learning_rate": 0.0005330202020434019,
+      "loss": 3.0815,
+      "step": 13276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.604824423789978,
+      "learning_rate": 0.0005330105047728326,
+      "loss": 3.241,
+      "step": 13277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5088874101638794,
+      "learning_rate": 0.0005330008068885585,
+      "loss": 3.255,
+      "step": 13278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5178427696228027,
+      "learning_rate": 0.0005329911083906051,
+      "loss": 3.2774,
+      "step": 13279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.510136604309082,
+      "learning_rate": 0.0005329814092789979,
+      "loss": 3.4761,
+      "step": 13280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48275598883628845,
+      "learning_rate": 0.0005329717095537627,
+      "loss": 3.2948,
+      "step": 13281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5280422568321228,
+      "learning_rate": 0.0005329620092149247,
+      "loss": 3.0466,
+      "step": 13282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5315813422203064,
+      "learning_rate": 0.0005329523082625097,
+      "loss": 3.0598,
+      "step": 13283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49926483631134033,
+      "learning_rate": 0.0005329426066965431,
+      "loss": 3.2156,
+      "step": 13284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5536932349205017,
+      "learning_rate": 0.0005329329045170503,
+      "loss": 3.3147,
+      "step": 13285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5208083391189575,
+      "learning_rate": 0.0005329232017240573,
+      "loss": 3.1502,
+      "step": 13286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5383280515670776,
+      "learning_rate": 0.0005329134983175894,
+      "loss": 3.2438,
+      "step": 13287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5244266986846924,
+      "learning_rate": 0.0005329037942976721,
+      "loss": 3.4281,
+      "step": 13288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47546955943107605,
+      "learning_rate": 0.0005328940896643311,
+      "loss": 3.2289,
+      "step": 13289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5083588361740112,
+      "learning_rate": 0.0005328843844175918,
+      "loss": 3.138,
+      "step": 13290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.522491455078125,
+      "learning_rate": 0.0005328746785574799,
+      "loss": 3.1468,
+      "step": 13291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49923422932624817,
+      "learning_rate": 0.0005328649720840209,
+      "loss": 3.1271,
+      "step": 13292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5488773584365845,
+      "learning_rate": 0.0005328552649972405,
+      "loss": 3.1722,
+      "step": 13293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5081570744514465,
+      "learning_rate": 0.0005328455572971639,
+      "loss": 3.3069,
+      "step": 13294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48642170429229736,
+      "learning_rate": 0.0005328358489838171,
+      "loss": 3.3906,
+      "step": 13295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5340069532394409,
+      "learning_rate": 0.0005328261400572254,
+      "loss": 3.1015,
+      "step": 13296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5175917744636536,
+      "learning_rate": 0.0005328164305174146,
+      "loss": 3.2171,
+      "step": 13297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5808233022689819,
+      "learning_rate": 0.00053280672036441,
+      "loss": 3.2978,
+      "step": 13298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4955079257488251,
+      "learning_rate": 0.0005327970095982372,
+      "loss": 3.1519,
+      "step": 13299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6096027493476868,
+      "learning_rate": 0.0005327872982189221,
+      "loss": 3.2015,
+      "step": 13300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5957887172698975,
+      "learning_rate": 0.0005327775862264899,
+      "loss": 3.286,
+      "step": 13301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5077028274536133,
+      "learning_rate": 0.0005327678736209664,
+      "loss": 3.1815,
+      "step": 13302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5380714535713196,
+      "learning_rate": 0.0005327581604023772,
+      "loss": 3.4044,
+      "step": 13303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6547355651855469,
+      "learning_rate": 0.0005327484465707477,
+      "loss": 3.2944,
+      "step": 13304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5356435179710388,
+      "learning_rate": 0.0005327387321261035,
+      "loss": 2.9592,
+      "step": 13305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48445892333984375,
+      "learning_rate": 0.0005327290170684705,
+      "loss": 3.4262,
+      "step": 13306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5008205771446228,
+      "learning_rate": 0.0005327193013978739,
+      "loss": 3.0121,
+      "step": 13307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5113664269447327,
+      "learning_rate": 0.0005327095851143394,
+      "loss": 3.3788,
+      "step": 13308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5016259551048279,
+      "learning_rate": 0.0005326998682178927,
+      "loss": 3.0934,
+      "step": 13309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48941028118133545,
+      "learning_rate": 0.0005326901507085594,
+      "loss": 3.364,
+      "step": 13310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49986913800239563,
+      "learning_rate": 0.000532680432586365,
+      "loss": 3.3927,
+      "step": 13311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5154104828834534,
+      "learning_rate": 0.0005326707138513351,
+      "loss": 3.3246,
+      "step": 13312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5213533639907837,
+      "learning_rate": 0.0005326609945034953,
+      "loss": 3.1087,
+      "step": 13313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5237452387809753,
+      "learning_rate": 0.0005326512745428713,
+      "loss": 3.347,
+      "step": 13314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49681347608566284,
+      "learning_rate": 0.0005326415539694885,
+      "loss": 3.1432,
+      "step": 13315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5009694695472717,
+      "learning_rate": 0.0005326318327833726,
+      "loss": 3.307,
+      "step": 13316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5398823618888855,
+      "learning_rate": 0.0005326221109845493,
+      "loss": 3.1889,
+      "step": 13317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4797592759132385,
+      "learning_rate": 0.0005326123885730441,
+      "loss": 3.2475,
+      "step": 13318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5444583296775818,
+      "learning_rate": 0.0005326026655488827,
+      "loss": 3.0342,
+      "step": 13319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5243876576423645,
+      "learning_rate": 0.0005325929419120906,
+      "loss": 3.2708,
+      "step": 13320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46254366636276245,
+      "learning_rate": 0.0005325832176626934,
+      "loss": 2.979,
+      "step": 13321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5210396647453308,
+      "learning_rate": 0.0005325734928007168,
+      "loss": 3.1684,
+      "step": 13322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5120729207992554,
+      "learning_rate": 0.0005325637673261864,
+      "loss": 3.1615,
+      "step": 13323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5329714417457581,
+      "learning_rate": 0.0005325540412391279,
+      "loss": 3.3009,
+      "step": 13324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5156061053276062,
+      "learning_rate": 0.0005325443145395666,
+      "loss": 3.1694,
+      "step": 13325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5308995842933655,
+      "learning_rate": 0.0005325345872275285,
+      "loss": 3.163,
+      "step": 13326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5136690735816956,
+      "learning_rate": 0.0005325248593030389,
+      "loss": 3.5507,
+      "step": 13327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5340861082077026,
+      "learning_rate": 0.0005325151307661237,
+      "loss": 3.4215,
+      "step": 13328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6588315367698669,
+      "learning_rate": 0.0005325054016168083,
+      "loss": 3.4133,
+      "step": 13329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49800702929496765,
+      "learning_rate": 0.0005324956718551185,
+      "loss": 3.2623,
+      "step": 13330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5641326308250427,
+      "learning_rate": 0.0005324859414810798,
+      "loss": 3.0368,
+      "step": 13331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49139222502708435,
+      "learning_rate": 0.0005324762104947179,
+      "loss": 3.432,
+      "step": 13332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5017745494842529,
+      "learning_rate": 0.0005324664788960583,
+      "loss": 3.1099,
+      "step": 13333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5510955452919006,
+      "learning_rate": 0.0005324567466851269,
+      "loss": 3.08,
+      "step": 13334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5165101885795593,
+      "learning_rate": 0.0005324470138619492,
+      "loss": 3.2788,
+      "step": 13335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5383206605911255,
+      "learning_rate": 0.0005324372804265505,
+      "loss": 3.1441,
+      "step": 13336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5414730906486511,
+      "learning_rate": 0.000532427546378957,
+      "loss": 3.2794,
+      "step": 13337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5007972717285156,
+      "learning_rate": 0.000532417811719194,
+      "loss": 3.2736,
+      "step": 13338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.543908953666687,
+      "learning_rate": 0.0005324080764472871,
+      "loss": 2.9989,
+      "step": 13339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5582568049430847,
+      "learning_rate": 0.0005323983405632623,
+      "loss": 3.1517,
+      "step": 13340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4982832968235016,
+      "learning_rate": 0.0005323886040671448,
+      "loss": 3.2608,
+      "step": 13341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49134111404418945,
+      "learning_rate": 0.0005323788669589606,
+      "loss": 3.1525,
+      "step": 13342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5180835723876953,
+      "learning_rate": 0.000532369129238735,
+      "loss": 3.0439,
+      "step": 13343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49943089485168457,
+      "learning_rate": 0.000532359390906494,
+      "loss": 3.311,
+      "step": 13344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5632797479629517,
+      "learning_rate": 0.0005323496519622629,
+      "loss": 3.1605,
+      "step": 13345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4958808720111847,
+      "learning_rate": 0.0005323399124060677,
+      "loss": 3.3853,
+      "step": 13346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5369716286659241,
+      "learning_rate": 0.0005323301722379338,
+      "loss": 3.6639,
+      "step": 13347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5046685934066772,
+      "learning_rate": 0.000532320431457887,
+      "loss": 3.2543,
+      "step": 13348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.528698205947876,
+      "learning_rate": 0.0005323106900659529,
+      "loss": 2.9939,
+      "step": 13349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5998996496200562,
+      "learning_rate": 0.0005323009480621571,
+      "loss": 3.2342,
+      "step": 13350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5694158673286438,
+      "learning_rate": 0.0005322912054465253,
+      "loss": 3.0568,
+      "step": 13351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4973612129688263,
+      "learning_rate": 0.0005322814622190831,
+      "loss": 3.1143,
+      "step": 13352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5280294418334961,
+      "learning_rate": 0.0005322717183798564,
+      "loss": 3.1217,
+      "step": 13353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.507739245891571,
+      "learning_rate": 0.0005322619739288706,
+      "loss": 3.0354,
+      "step": 13354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5100513696670532,
+      "learning_rate": 0.0005322522288661515,
+      "loss": 3.0251,
+      "step": 13355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49843087792396545,
+      "learning_rate": 0.0005322424831917247,
+      "loss": 3.152,
+      "step": 13356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5560824275016785,
+      "learning_rate": 0.000532232736905616,
+      "loss": 3.3028,
+      "step": 13357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5524668097496033,
+      "learning_rate": 0.0005322229900078507,
+      "loss": 3.2832,
+      "step": 13358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5191108584403992,
+      "learning_rate": 0.0005322132424984549,
+      "loss": 3.136,
+      "step": 13359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49663254618644714,
+      "learning_rate": 0.0005322034943774542,
+      "loss": 3.0543,
+      "step": 13360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5162781476974487,
+      "learning_rate": 0.0005321937456448741,
+      "loss": 3.2929,
+      "step": 13361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5527589321136475,
+      "learning_rate": 0.0005321839963007402,
+      "loss": 3.4472,
+      "step": 13362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.539014458656311,
+      "learning_rate": 0.0005321742463450786,
+      "loss": 3.2797,
+      "step": 13363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5228904485702515,
+      "learning_rate": 0.0005321644957779146,
+      "loss": 3.2703,
+      "step": 13364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5191890597343445,
+      "learning_rate": 0.000532154744599274,
+      "loss": 3.118,
+      "step": 13365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5481828451156616,
+      "learning_rate": 0.0005321449928091825,
+      "loss": 3.2102,
+      "step": 13366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5154120326042175,
+      "learning_rate": 0.0005321352404076659,
+      "loss": 3.273,
+      "step": 13367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5242748260498047,
+      "learning_rate": 0.0005321254873947495,
+      "loss": 3.2006,
+      "step": 13368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5365203619003296,
+      "learning_rate": 0.0005321157337704594,
+      "loss": 2.8698,
+      "step": 13369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5272378325462341,
+      "learning_rate": 0.000532105979534821,
+      "loss": 3.1841,
+      "step": 13370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5126152634620667,
+      "learning_rate": 0.0005320962246878602,
+      "loss": 3.1408,
+      "step": 13371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5221778154373169,
+      "learning_rate": 0.0005320864692296026,
+      "loss": 3.2598,
+      "step": 13372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5323341488838196,
+      "learning_rate": 0.000532076713160074,
+      "loss": 3.3795,
+      "step": 13373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5273595452308655,
+      "learning_rate": 0.0005320669564792999,
+      "loss": 3.5628,
+      "step": 13374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5162096619606018,
+      "learning_rate": 0.0005320571991873061,
+      "loss": 3.1631,
+      "step": 13375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49325329065322876,
+      "learning_rate": 0.0005320474412841183,
+      "loss": 3.4696,
+      "step": 13376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5000008344650269,
+      "learning_rate": 0.0005320376827697622,
+      "loss": 3.1889,
+      "step": 13377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5023176074028015,
+      "learning_rate": 0.0005320279236442635,
+      "loss": 3.2494,
+      "step": 13378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4893326759338379,
+      "learning_rate": 0.000532018163907648,
+      "loss": 3.1743,
+      "step": 13379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5529983639717102,
+      "learning_rate": 0.0005320084035599413,
+      "loss": 3.0792,
+      "step": 13380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5161319375038147,
+      "learning_rate": 0.000531998642601169,
+      "loss": 3.2193,
+      "step": 13381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4804127812385559,
+      "learning_rate": 0.000531988881031357,
+      "loss": 3.3955,
+      "step": 13382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5218368172645569,
+      "learning_rate": 0.0005319791188505309,
+      "loss": 3.2941,
+      "step": 13383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5038884878158569,
+      "learning_rate": 0.0005319693560587164,
+      "loss": 3.384,
+      "step": 13384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.503178060054779,
+      "learning_rate": 0.0005319595926559392,
+      "loss": 3.2356,
+      "step": 13385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48960548639297485,
+      "learning_rate": 0.0005319498286422252,
+      "loss": 3.1029,
+      "step": 13386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5522580146789551,
+      "learning_rate": 0.0005319400640176,
+      "loss": 3.2378,
+      "step": 13387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5147445201873779,
+      "learning_rate": 0.0005319302987820894,
+      "loss": 3.3786,
+      "step": 13388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5028669238090515,
+      "learning_rate": 0.0005319205329357188,
+      "loss": 3.2985,
+      "step": 13389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5062679648399353,
+      "learning_rate": 0.0005319107664785144,
+      "loss": 3.2392,
+      "step": 13390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5048500895500183,
+      "learning_rate": 0.0005319009994105014,
+      "loss": 3.3131,
+      "step": 13391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5031437277793884,
+      "learning_rate": 0.000531891231731706,
+      "loss": 3.485,
+      "step": 13392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5089054703712463,
+      "learning_rate": 0.0005318814634421537,
+      "loss": 3.082,
+      "step": 13393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5406572222709656,
+      "learning_rate": 0.0005318716945418701,
+      "loss": 3.1211,
+      "step": 13394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5037758350372314,
+      "learning_rate": 0.0005318619250308812,
+      "loss": 3.055,
+      "step": 13395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48916196823120117,
+      "learning_rate": 0.0005318521549092126,
+      "loss": 3.3232,
+      "step": 13396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5082511305809021,
+      "learning_rate": 0.0005318423841768901,
+      "loss": 3.4669,
+      "step": 13397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5099027156829834,
+      "learning_rate": 0.0005318326128339393,
+      "loss": 3.3092,
+      "step": 13398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5375298261642456,
+      "learning_rate": 0.0005318228408803861,
+      "loss": 3.4794,
+      "step": 13399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5250174403190613,
+      "learning_rate": 0.0005318130683162561,
+      "loss": 3.2933,
+      "step": 13400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5176973938941956,
+      "learning_rate": 0.0005318032951415751,
+      "loss": 3.2749,
+      "step": 13401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5219639539718628,
+      "learning_rate": 0.0005317935213563687,
+      "loss": 3.4369,
+      "step": 13402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49903932213783264,
+      "learning_rate": 0.000531783746960663,
+      "loss": 3.1772,
+      "step": 13403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5053322911262512,
+      "learning_rate": 0.0005317739719544834,
+      "loss": 3.415,
+      "step": 13404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5507119297981262,
+      "learning_rate": 0.0005317641963378557,
+      "loss": 3.0963,
+      "step": 13405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5222508311271667,
+      "learning_rate": 0.0005317544201108058,
+      "loss": 3.2718,
+      "step": 13406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.509367048740387,
+      "learning_rate": 0.0005317446432733594,
+      "loss": 3.0838,
+      "step": 13407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5182288885116577,
+      "learning_rate": 0.0005317348658255421,
+      "loss": 3.1657,
+      "step": 13408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5248708128929138,
+      "learning_rate": 0.0005317250877673798,
+      "loss": 3.3512,
+      "step": 13409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5264621376991272,
+      "learning_rate": 0.0005317153090988983,
+      "loss": 3.1901,
+      "step": 13410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5019949078559875,
+      "learning_rate": 0.0005317055298201232,
+      "loss": 3.4335,
+      "step": 13411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5084396004676819,
+      "learning_rate": 0.0005316957499310802,
+      "loss": 3.3224,
+      "step": 13412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5858707427978516,
+      "learning_rate": 0.0005316859694317954,
+      "loss": 3.0917,
+      "step": 13413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5144307613372803,
+      "learning_rate": 0.0005316761883222943,
+      "loss": 3.4099,
+      "step": 13414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5141854286193848,
+      "learning_rate": 0.0005316664066026026,
+      "loss": 3.2635,
+      "step": 13415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5495926737785339,
+      "learning_rate": 0.0005316566242727463,
+      "loss": 3.3313,
+      "step": 13416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5578013062477112,
+      "learning_rate": 0.000531646841332751,
+      "loss": 3.4108,
+      "step": 13417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5008851885795593,
+      "learning_rate": 0.0005316370577826424,
+      "loss": 3.2526,
+      "step": 13418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5154222846031189,
+      "learning_rate": 0.0005316272736224464,
+      "loss": 3.2379,
+      "step": 13419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.527387261390686,
+      "learning_rate": 0.0005316174888521888,
+      "loss": 3.3232,
+      "step": 13420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5717757344245911,
+      "learning_rate": 0.0005316077034718952,
+      "loss": 3.1741,
+      "step": 13421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5230823159217834,
+      "learning_rate": 0.0005315979174815916,
+      "loss": 3.273,
+      "step": 13422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.521920382976532,
+      "learning_rate": 0.0005315881308813038,
+      "loss": 3.2103,
+      "step": 13423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5073776245117188,
+      "learning_rate": 0.0005315783436710572,
+      "loss": 3.25,
+      "step": 13424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5398444533348083,
+      "learning_rate": 0.0005315685558508779,
+      "loss": 3.1903,
+      "step": 13425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4858630299568176,
+      "learning_rate": 0.0005315587674207914,
+      "loss": 3.2599,
+      "step": 13426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5114261507987976,
+      "learning_rate": 0.0005315489783808239,
+      "loss": 3.1523,
+      "step": 13427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5169461965560913,
+      "learning_rate": 0.0005315391887310009,
+      "loss": 3.1635,
+      "step": 13428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49738162755966187,
+      "learning_rate": 0.0005315293984713482,
+      "loss": 3.0964,
+      "step": 13429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5174451470375061,
+      "learning_rate": 0.0005315196076018917,
+      "loss": 3.2218,
+      "step": 13430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5436557531356812,
+      "learning_rate": 0.000531509816122657,
+      "loss": 3.0565,
+      "step": 13431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5209143161773682,
+      "learning_rate": 0.0005315000240336702,
+      "loss": 3.1762,
+      "step": 13432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5072414875030518,
+      "learning_rate": 0.0005314902313349566,
+      "loss": 3.4239,
+      "step": 13433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4978218674659729,
+      "learning_rate": 0.0005314804380265425,
+      "loss": 3.3271,
+      "step": 13434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5045718550682068,
+      "learning_rate": 0.0005314706441084535,
+      "loss": 3.0421,
+      "step": 13435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5093064904212952,
+      "learning_rate": 0.0005314608495807151,
+      "loss": 3.2584,
+      "step": 13436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5230816602706909,
+      "learning_rate": 0.0005314510544433536,
+      "loss": 3.1959,
+      "step": 13437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.547500491142273,
+      "learning_rate": 0.0005314412586963945,
+      "loss": 3.1111,
+      "step": 13438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5420114398002625,
+      "learning_rate": 0.0005314314623398637,
+      "loss": 3.2622,
+      "step": 13439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5222113728523254,
+      "learning_rate": 0.0005314216653737869,
+      "loss": 3.1762,
+      "step": 13440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5479421615600586,
+      "learning_rate": 0.00053141186779819,
+      "loss": 3.0739,
+      "step": 13441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5246508121490479,
+      "learning_rate": 0.0005314020696130989,
+      "loss": 3.2307,
+      "step": 13442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5006526112556458,
+      "learning_rate": 0.0005313922708185391,
+      "loss": 3.3417,
+      "step": 13443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5077061653137207,
+      "learning_rate": 0.0005313824714145367,
+      "loss": 3.0799,
+      "step": 13444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4841029942035675,
+      "learning_rate": 0.0005313726714011173,
+      "loss": 3.1433,
+      "step": 13445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5092795491218567,
+      "learning_rate": 0.000531362870778307,
+      "loss": 3.2824,
+      "step": 13446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5002308487892151,
+      "learning_rate": 0.0005313530695461313,
+      "loss": 3.2313,
+      "step": 13447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5429970026016235,
+      "learning_rate": 0.0005313432677046162,
+      "loss": 3.2798,
+      "step": 13448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5142826437950134,
+      "learning_rate": 0.0005313334652537873,
+      "loss": 3.0374,
+      "step": 13449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5216500759124756,
+      "learning_rate": 0.0005313236621936707,
+      "loss": 3.401,
+      "step": 13450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5026384592056274,
+      "learning_rate": 0.0005313138585242921,
+      "loss": 3.2493,
+      "step": 13451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5115340352058411,
+      "learning_rate": 0.0005313040542456772,
+      "loss": 3.0881,
+      "step": 13452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5016939043998718,
+      "learning_rate": 0.0005312942493578519,
+      "loss": 3.0879,
+      "step": 13453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.459194540977478,
+      "learning_rate": 0.0005312844438608423,
+      "loss": 2.9753,
+      "step": 13454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5059602856636047,
+      "learning_rate": 0.0005312746377546739,
+      "loss": 3.1455,
+      "step": 13455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49334749579429626,
+      "learning_rate": 0.0005312648310393726,
+      "loss": 3.1937,
+      "step": 13456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5341232419013977,
+      "learning_rate": 0.0005312550237149641,
+      "loss": 3.2977,
+      "step": 13457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5050813555717468,
+      "learning_rate": 0.0005312452157814746,
+      "loss": 3.3302,
+      "step": 13458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5001168251037598,
+      "learning_rate": 0.0005312354072389296,
+      "loss": 3.0763,
+      "step": 13459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5147970914840698,
+      "learning_rate": 0.000531225598087355,
+      "loss": 3.235,
+      "step": 13460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.529277503490448,
+      "learning_rate": 0.0005312157883267767,
+      "loss": 3.2548,
+      "step": 13461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49942076206207275,
+      "learning_rate": 0.0005312059779572205,
+      "loss": 3.106,
+      "step": 13462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5275735259056091,
+      "learning_rate": 0.0005311961669787124,
+      "loss": 3.0986,
+      "step": 13463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5004652738571167,
+      "learning_rate": 0.0005311863553912778,
+      "loss": 3.2899,
+      "step": 13464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5354148149490356,
+      "learning_rate": 0.0005311765431949431,
+      "loss": 3.2583,
+      "step": 13465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5163960456848145,
+      "learning_rate": 0.0005311667303897337,
+      "loss": 3.1102,
+      "step": 13466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5226514935493469,
+      "learning_rate": 0.0005311569169756757,
+      "loss": 3.1032,
+      "step": 13467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5380831956863403,
+      "learning_rate": 0.000531147102952795,
+      "loss": 3.3771,
+      "step": 13468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5179706811904907,
+      "learning_rate": 0.0005311372883211171,
+      "loss": 3.2395,
+      "step": 13469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5135579705238342,
+      "learning_rate": 0.0005311274730806681,
+      "loss": 3.3656,
+      "step": 13470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.500043511390686,
+      "learning_rate": 0.0005311176572314739,
+      "loss": 3.354,
+      "step": 13471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5214130282402039,
+      "learning_rate": 0.0005311078407735602,
+      "loss": 3.2447,
+      "step": 13472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5328757762908936,
+      "learning_rate": 0.000531098023706953,
+      "loss": 3.0472,
+      "step": 13473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49766138195991516,
+      "learning_rate": 0.000531088206031678,
+      "loss": 3.2254,
+      "step": 13474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5517077445983887,
+      "learning_rate": 0.0005310783877477612,
+      "loss": 3.3586,
+      "step": 13475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5423430800437927,
+      "learning_rate": 0.0005310685688552284,
+      "loss": 3.0909,
+      "step": 13476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.519158124923706,
+      "learning_rate": 0.0005310587493541054,
+      "loss": 3.1953,
+      "step": 13477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5957249999046326,
+      "learning_rate": 0.0005310489292444182,
+      "loss": 3.1486,
+      "step": 13478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5332064032554626,
+      "learning_rate": 0.0005310391085261926,
+      "loss": 3.3535,
+      "step": 13479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5151717066764832,
+      "learning_rate": 0.0005310292871994544,
+      "loss": 3.2729,
+      "step": 13480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5478836297988892,
+      "learning_rate": 0.0005310194652642295,
+      "loss": 3.1139,
+      "step": 13481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5319055318832397,
+      "learning_rate": 0.0005310096427205437,
+      "loss": 3.1867,
+      "step": 13482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5295076966285706,
+      "learning_rate": 0.0005309998195684231,
+      "loss": 3.4644,
+      "step": 13483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5261130332946777,
+      "learning_rate": 0.0005309899958078934,
+      "loss": 3.1089,
+      "step": 13484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.504030168056488,
+      "learning_rate": 0.0005309801714389805,
+      "loss": 3.1605,
+      "step": 13485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5086562037467957,
+      "learning_rate": 0.0005309703464617103,
+      "loss": 3.2354,
+      "step": 13486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49703478813171387,
+      "learning_rate": 0.0005309605208761087,
+      "loss": 3.2152,
+      "step": 13487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5037059783935547,
+      "learning_rate": 0.0005309506946822015,
+      "loss": 3.1068,
+      "step": 13488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49587151408195496,
+      "learning_rate": 0.0005309408678800145,
+      "loss": 3.2943,
+      "step": 13489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5075664520263672,
+      "learning_rate": 0.0005309310404695739,
+      "loss": 3.2823,
+      "step": 13490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5008799433708191,
+      "learning_rate": 0.0005309212124509052,
+      "loss": 3.1233,
+      "step": 13491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49640950560569763,
+      "learning_rate": 0.0005309113838240346,
+      "loss": 3.316,
+      "step": 13492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5615620613098145,
+      "learning_rate": 0.0005309015545889878,
+      "loss": 3.3059,
+      "step": 13493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5070755481719971,
+      "learning_rate": 0.0005308917247457907,
+      "loss": 3.2173,
+      "step": 13494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48888513445854187,
+      "learning_rate": 0.0005308818942944693,
+      "loss": 3.1687,
+      "step": 13495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5695043802261353,
+      "learning_rate": 0.0005308720632350494,
+      "loss": 3.2426,
+      "step": 13496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.533906102180481,
+      "learning_rate": 0.000530862231567557,
+      "loss": 3.1716,
+      "step": 13497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5260034203529358,
+      "learning_rate": 0.0005308523992920178,
+      "loss": 3.2352,
+      "step": 13498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5359194278717041,
+      "learning_rate": 0.0005308425664084579,
+      "loss": 3.2188,
+      "step": 13499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49645182490348816,
+      "learning_rate": 0.0005308327329169029,
+      "loss": 3.2868,
+      "step": 13500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5202005505561829,
+      "learning_rate": 0.0005308228988173792,
+      "loss": 3.3061,
+      "step": 13501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5067334771156311,
+      "learning_rate": 0.0005308130641099122,
+      "loss": 3.0534,
+      "step": 13502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5037301778793335,
+      "learning_rate": 0.0005308032287945281,
+      "loss": 3.3825,
+      "step": 13503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6401285529136658,
+      "learning_rate": 0.0005307933928712527,
+      "loss": 3.1316,
+      "step": 13504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.516437292098999,
+      "learning_rate": 0.000530783556340112,
+      "loss": 3.095,
+      "step": 13505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5353245139122009,
+      "learning_rate": 0.0005307737192011316,
+      "loss": 3.0556,
+      "step": 13506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5425240993499756,
+      "learning_rate": 0.0005307638814543378,
+      "loss": 3.2073,
+      "step": 13507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5350566506385803,
+      "learning_rate": 0.0005307540430997563,
+      "loss": 3.1884,
+      "step": 13508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5366530418395996,
+      "learning_rate": 0.0005307442041374131,
+      "loss": 3.2401,
+      "step": 13509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5320677161216736,
+      "learning_rate": 0.0005307343645673342,
+      "loss": 3.1042,
+      "step": 13510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5887209177017212,
+      "learning_rate": 0.0005307245243895451,
+      "loss": 3.0003,
+      "step": 13511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5200484991073608,
+      "learning_rate": 0.0005307146836040722,
+      "loss": 3.3779,
+      "step": 13512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5377318263053894,
+      "learning_rate": 0.0005307048422109412,
+      "loss": 3.0319,
+      "step": 13513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5469724535942078,
+      "learning_rate": 0.000530695000210178,
+      "loss": 3.1924,
+      "step": 13514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5550596117973328,
+      "learning_rate": 0.0005306851576018086,
+      "loss": 3.2543,
+      "step": 13515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49396204948425293,
+      "learning_rate": 0.0005306753143858588,
+      "loss": 3.2256,
+      "step": 13516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49809572100639343,
+      "learning_rate": 0.0005306654705623547,
+      "loss": 3.362,
+      "step": 13517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4874090254306793,
+      "learning_rate": 0.0005306556261313222,
+      "loss": 3.0973,
+      "step": 13518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5111352801322937,
+      "learning_rate": 0.0005306457810927872,
+      "loss": 3.4853,
+      "step": 13519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5032990574836731,
+      "learning_rate": 0.0005306359354467754,
+      "loss": 3.2901,
+      "step": 13520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5103004574775696,
+      "learning_rate": 0.0005306260891933131,
+      "loss": 3.3242,
+      "step": 13521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5028958916664124,
+      "learning_rate": 0.000530616242332426,
+      "loss": 3.232,
+      "step": 13522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5463312864303589,
+      "learning_rate": 0.0005306063948641401,
+      "loss": 3.2748,
+      "step": 13523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5075322985649109,
+      "learning_rate": 0.0005305965467884813,
+      "loss": 3.1649,
+      "step": 13524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5177689790725708,
+      "learning_rate": 0.0005305866981054757,
+      "loss": 3.1198,
+      "step": 13525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5131239295005798,
+      "learning_rate": 0.000530576848815149,
+      "loss": 3.29,
+      "step": 13526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4989679455757141,
+      "learning_rate": 0.0005305669989175273,
+      "loss": 3.0434,
+      "step": 13527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5141971111297607,
+      "learning_rate": 0.0005305571484126365,
+      "loss": 3.1655,
+      "step": 13528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5180733799934387,
+      "learning_rate": 0.0005305472973005025,
+      "loss": 3.3093,
+      "step": 13529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5635950565338135,
+      "learning_rate": 0.0005305374455811514,
+      "loss": 3.1966,
+      "step": 13530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5000056624412537,
+      "learning_rate": 0.0005305275932546089,
+      "loss": 3.1471,
+      "step": 13531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5147111415863037,
+      "learning_rate": 0.0005305177403209011,
+      "loss": 3.0226,
+      "step": 13532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48060378432273865,
+      "learning_rate": 0.0005305078867800541,
+      "loss": 3.1014,
+      "step": 13533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.569724440574646,
+      "learning_rate": 0.0005304980326320935,
+      "loss": 3.2077,
+      "step": 13534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5264111161231995,
+      "learning_rate": 0.0005304881778770455,
+      "loss": 3.0713,
+      "step": 13535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.512325644493103,
+      "learning_rate": 0.000530478322514936,
+      "loss": 3.3984,
+      "step": 13536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5474256873130798,
+      "learning_rate": 0.000530468466545791,
+      "loss": 3.0077,
+      "step": 13537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5356321334838867,
+      "learning_rate": 0.0005304586099696364,
+      "loss": 3.3486,
+      "step": 13538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5049594044685364,
+      "learning_rate": 0.0005304487527864982,
+      "loss": 3.3997,
+      "step": 13539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5442296862602234,
+      "learning_rate": 0.0005304388949964022,
+      "loss": 3.3842,
+      "step": 13540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5284900665283203,
+      "learning_rate": 0.0005304290365993747,
+      "loss": 3.0788,
+      "step": 13541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0005304191775954414,
+      "loss": 3.157,
+      "step": 13542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49122077226638794,
+      "learning_rate": 0.0005304093179846281,
+      "loss": 3.0097,
+      "step": 13543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5494361519813538,
+      "learning_rate": 0.0005303994577669612,
+      "loss": 3.0168,
+      "step": 13544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5453931093215942,
+      "learning_rate": 0.0005303895969424665,
+      "loss": 3.2503,
+      "step": 13545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5414052605628967,
+      "learning_rate": 0.0005303797355111699,
+      "loss": 3.0531,
+      "step": 13546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5246322751045227,
+      "learning_rate": 0.0005303698734730974,
+      "loss": 3.3351,
+      "step": 13547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5333435535430908,
+      "learning_rate": 0.0005303600108282749,
+      "loss": 3.4273,
+      "step": 13548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.511567234992981,
+      "learning_rate": 0.0005303501475767287,
+      "loss": 3.1325,
+      "step": 13549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4871523082256317,
+      "learning_rate": 0.0005303402837184844,
+      "loss": 3.2711,
+      "step": 13550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5091209411621094,
+      "learning_rate": 0.0005303304192535681,
+      "loss": 3.3666,
+      "step": 13551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5344579815864563,
+      "learning_rate": 0.0005303205541820058,
+      "loss": 3.1501,
+      "step": 13552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5157179832458496,
+      "learning_rate": 0.0005303106885038235,
+      "loss": 3.2942,
+      "step": 13553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4973103404045105,
+      "learning_rate": 0.0005303008222190472,
+      "loss": 3.1057,
+      "step": 13554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5082616209983826,
+      "learning_rate": 0.0005302909553277029,
+      "loss": 3.079,
+      "step": 13555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.491454154253006,
+      "learning_rate": 0.0005302810878298165,
+      "loss": 3.2827,
+      "step": 13556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5277991890907288,
+      "learning_rate": 0.000530271219725414,
+      "loss": 3.3536,
+      "step": 13557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49939367175102234,
+      "learning_rate": 0.0005302613510145215,
+      "loss": 3.1075,
+      "step": 13558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5459829568862915,
+      "learning_rate": 0.0005302514816971648,
+      "loss": 3.2634,
+      "step": 13559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5232791304588318,
+      "learning_rate": 0.0005302416117733701,
+      "loss": 3.1575,
+      "step": 13560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5249155163764954,
+      "learning_rate": 0.0005302317412431632,
+      "loss": 3.1753,
+      "step": 13561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5318909883499146,
+      "learning_rate": 0.0005302218701065703,
+      "loss": 3.177,
+      "step": 13562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49357926845550537,
+      "learning_rate": 0.0005302119983636174,
+      "loss": 3.2127,
+      "step": 13563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5603306293487549,
+      "learning_rate": 0.0005302021260143303,
+      "loss": 3.2759,
+      "step": 13564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49928614497184753,
+      "learning_rate": 0.0005301922530587351,
+      "loss": 3.3354,
+      "step": 13565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5590007901191711,
+      "learning_rate": 0.0005301823794968577,
+      "loss": 3.1456,
+      "step": 13566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5050297975540161,
+      "learning_rate": 0.0005301725053287243,
+      "loss": 3.1094,
+      "step": 13567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.532940149307251,
+      "learning_rate": 0.0005301626305543608,
+      "loss": 3.2555,
+      "step": 13568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.511152982711792,
+      "learning_rate": 0.0005301527551737933,
+      "loss": 3.3773,
+      "step": 13569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5149011611938477,
+      "learning_rate": 0.0005301428791870476,
+      "loss": 3.2087,
+      "step": 13570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5097518563270569,
+      "learning_rate": 0.00053013300259415,
+      "loss": 3.2675,
+      "step": 13571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5337220430374146,
+      "learning_rate": 0.0005301231253951263,
+      "loss": 3.3959,
+      "step": 13572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5316744446754456,
+      "learning_rate": 0.0005301132475900026,
+      "loss": 3.4082,
+      "step": 13573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48296254873275757,
+      "learning_rate": 0.0005301033691788048,
+      "loss": 3.2966,
+      "step": 13574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5001648664474487,
+      "learning_rate": 0.0005300934901615591,
+      "loss": 3.2072,
+      "step": 13575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5572800040245056,
+      "learning_rate": 0.0005300836105382914,
+      "loss": 3.2663,
+      "step": 13576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5312752723693848,
+      "learning_rate": 0.0005300737303090277,
+      "loss": 3.2437,
+      "step": 13577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.528778076171875,
+      "learning_rate": 0.0005300638494737941,
+      "loss": 3.2888,
+      "step": 13578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5192068219184875,
+      "learning_rate": 0.0005300539680326168,
+      "loss": 3.2725,
+      "step": 13579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5534399747848511,
+      "learning_rate": 0.0005300440859855214,
+      "loss": 3.1588,
+      "step": 13580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5094475746154785,
+      "learning_rate": 0.0005300342033325342,
+      "loss": 3.4536,
+      "step": 13581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48778852820396423,
+      "learning_rate": 0.0005300243200736811,
+      "loss": 3.2469,
+      "step": 13582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47840654850006104,
+      "learning_rate": 0.0005300144362089883,
+      "loss": 3.2045,
+      "step": 13583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5369465947151184,
+      "learning_rate": 0.0005300045517384818,
+      "loss": 3.1993,
+      "step": 13584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5591325759887695,
+      "learning_rate": 0.0005299946666621875,
+      "loss": 3.1054,
+      "step": 13585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5063976645469666,
+      "learning_rate": 0.0005299847809801314,
+      "loss": 2.9926,
+      "step": 13586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5150634050369263,
+      "learning_rate": 0.0005299748946923399,
+      "loss": 3.2566,
+      "step": 13587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5185797214508057,
+      "learning_rate": 0.0005299650077988386,
+      "loss": 3.0262,
+      "step": 13588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5387712717056274,
+      "learning_rate": 0.0005299551202996537,
+      "loss": 3.0643,
+      "step": 13589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5186799764633179,
+      "learning_rate": 0.0005299452321948114,
+      "loss": 3.2493,
+      "step": 13590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5232559442520142,
+      "learning_rate": 0.0005299353434843376,
+      "loss": 3.5282,
+      "step": 13591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5263158679008484,
+      "learning_rate": 0.0005299254541682583,
+      "loss": 3.0349,
+      "step": 13592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5348055362701416,
+      "learning_rate": 0.0005299155642465996,
+      "loss": 3.1144,
+      "step": 13593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.515724241733551,
+      "learning_rate": 0.0005299056737193876,
+      "loss": 3.3351,
+      "step": 13594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5149891972541809,
+      "learning_rate": 0.0005298957825866482,
+      "loss": 3.1768,
+      "step": 13595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.509054958820343,
+      "learning_rate": 0.0005298858908484076,
+      "loss": 3.3885,
+      "step": 13596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5293143391609192,
+      "learning_rate": 0.0005298759985046919,
+      "loss": 3.3963,
+      "step": 13597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5564351677894592,
+      "learning_rate": 0.0005298661055555269,
+      "loss": 3.2376,
+      "step": 13598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5585691928863525,
+      "learning_rate": 0.000529856212000939,
+      "loss": 3.3014,
+      "step": 13599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5429052114486694,
+      "learning_rate": 0.000529846317840954,
+      "loss": 3.2396,
+      "step": 13600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5204060673713684,
+      "learning_rate": 0.000529836423075598,
+      "loss": 3.3343,
+      "step": 13601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4938051998615265,
+      "learning_rate": 0.0005298265277048971,
+      "loss": 3.2435,
+      "step": 13602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4976741373538971,
+      "learning_rate": 0.0005298166317288774,
+      "loss": 3.2764,
+      "step": 13603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5170921683311462,
+      "learning_rate": 0.0005298067351475649,
+      "loss": 3.1395,
+      "step": 13604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4929213225841522,
+      "learning_rate": 0.0005297968379609858,
+      "loss": 3.2229,
+      "step": 13605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5679933428764343,
+      "learning_rate": 0.0005297869401691658,
+      "loss": 3.3445,
+      "step": 13606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5538800954818726,
+      "learning_rate": 0.0005297770417721314,
+      "loss": 3.2866,
+      "step": 13607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5196000933647156,
+      "learning_rate": 0.0005297671427699084,
+      "loss": 3.204,
+      "step": 13608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5372135043144226,
+      "learning_rate": 0.0005297572431625229,
+      "loss": 3.2732,
+      "step": 13609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5452011227607727,
+      "learning_rate": 0.0005297473429500013,
+      "loss": 3.1766,
+      "step": 13610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4952123463153839,
+      "learning_rate": 0.0005297374421323692,
+      "loss": 3.1995,
+      "step": 13611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.506626307964325,
+      "learning_rate": 0.000529727540709653,
+      "loss": 3.1581,
+      "step": 13612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5172677636146545,
+      "learning_rate": 0.0005297176386818786,
+      "loss": 2.8921,
+      "step": 13613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5049740076065063,
+      "learning_rate": 0.0005297077360490722,
+      "loss": 3.208,
+      "step": 13614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5029763579368591,
+      "learning_rate": 0.0005296978328112598,
+      "loss": 3.1353,
+      "step": 13615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5007016658782959,
+      "learning_rate": 0.0005296879289684675,
+      "loss": 3.1918,
+      "step": 13616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5236932039260864,
+      "learning_rate": 0.0005296780245207215,
+      "loss": 3.0716,
+      "step": 13617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5022386908531189,
+      "learning_rate": 0.0005296681194680477,
+      "loss": 3.3136,
+      "step": 13618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.50746089220047,
+      "learning_rate": 0.0005296582138104723,
+      "loss": 3.3146,
+      "step": 13619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.482505738735199,
+      "learning_rate": 0.0005296483075480213,
+      "loss": 3.3466,
+      "step": 13620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49940067529678345,
+      "learning_rate": 0.0005296384006807209,
+      "loss": 3.2996,
+      "step": 13621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49896591901779175,
+      "learning_rate": 0.0005296284932085972,
+      "loss": 3.1868,
+      "step": 13622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5380591750144958,
+      "learning_rate": 0.0005296185851316761,
+      "loss": 3.0733,
+      "step": 13623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5520894527435303,
+      "learning_rate": 0.0005296086764499839,
+      "loss": 3.1411,
+      "step": 13624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5142804980278015,
+      "learning_rate": 0.0005295987671635468,
+      "loss": 3.4073,
+      "step": 13625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5468404293060303,
+      "learning_rate": 0.0005295888572723906,
+      "loss": 3.0009,
+      "step": 13626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5099403858184814,
+      "learning_rate": 0.0005295789467765414,
+      "loss": 3.3411,
+      "step": 13627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4918704330921173,
+      "learning_rate": 0.0005295690356760256,
+      "loss": 3.0391,
+      "step": 13628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5046103596687317,
+      "learning_rate": 0.0005295591239708691,
+      "loss": 3.4971,
+      "step": 13629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5209892392158508,
+      "learning_rate": 0.0005295492116610982,
+      "loss": 2.7844,
+      "step": 13630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5438646078109741,
+      "learning_rate": 0.0005295392987467387,
+      "loss": 3.1362,
+      "step": 13631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5239036083221436,
+      "learning_rate": 0.0005295293852278168,
+      "loss": 3.3859,
+      "step": 13632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.529381275177002,
+      "learning_rate": 0.0005295194711043588,
+      "loss": 3.2098,
+      "step": 13633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5509489178657532,
+      "learning_rate": 0.0005295095563763907,
+      "loss": 3.2301,
+      "step": 13634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5186215043067932,
+      "learning_rate": 0.0005294996410439384,
+      "loss": 3.159,
+      "step": 13635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5146812796592712,
+      "learning_rate": 0.0005294897251070283,
+      "loss": 3.2299,
+      "step": 13636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5205144882202148,
+      "learning_rate": 0.0005294798085656865,
+      "loss": 3.1246,
+      "step": 13637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5245778560638428,
+      "learning_rate": 0.0005294698914199391,
+      "loss": 3.3632,
+      "step": 13638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5007055997848511,
+      "learning_rate": 0.0005294599736698121,
+      "loss": 3.1628,
+      "step": 13639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5703233480453491,
+      "learning_rate": 0.0005294500553153316,
+      "loss": 3.3017,
+      "step": 13640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5297021269798279,
+      "learning_rate": 0.0005294401363565239,
+      "loss": 3.3155,
+      "step": 13641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5470933318138123,
+      "learning_rate": 0.000529430216793415,
+      "loss": 3.1739,
+      "step": 13642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5373654961585999,
+      "learning_rate": 0.0005294202966260312,
+      "loss": 3.4286,
+      "step": 13643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5070244669914246,
+      "learning_rate": 0.0005294103758543983,
+      "loss": 3.1327,
+      "step": 13644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5400650501251221,
+      "learning_rate": 0.0005294004544785428,
+      "loss": 3.1624,
+      "step": 13645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5214609503746033,
+      "learning_rate": 0.0005293905324984905,
+      "loss": 3.1664,
+      "step": 13646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5696268081665039,
+      "learning_rate": 0.0005293806099142677,
+      "loss": 3.1741,
+      "step": 13647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5091151595115662,
+      "learning_rate": 0.0005293706867259006,
+      "loss": 3.3379,
+      "step": 13648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5270219445228577,
+      "learning_rate": 0.0005293607629334152,
+      "loss": 3.1869,
+      "step": 13649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5341525077819824,
+      "learning_rate": 0.0005293508385368378,
+      "loss": 3.1583,
+      "step": 13650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5189785361289978,
+      "learning_rate": 0.0005293409135361943,
+      "loss": 3.286,
+      "step": 13651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5022042393684387,
+      "learning_rate": 0.0005293309879315111,
+      "loss": 3.2813,
+      "step": 13652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4763084053993225,
+      "learning_rate": 0.0005293210617228141,
+      "loss": 3.3161,
+      "step": 13653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.510564386844635,
+      "learning_rate": 0.0005293111349101296,
+      "loss": 3.1931,
+      "step": 13654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.531646728515625,
+      "learning_rate": 0.0005293012074934836,
+      "loss": 3.2027,
+      "step": 13655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5853457450866699,
+      "learning_rate": 0.0005292912794729025,
+      "loss": 3.1992,
+      "step": 13656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5224567651748657,
+      "learning_rate": 0.0005292813508484122,
+      "loss": 3.3783,
+      "step": 13657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5682322978973389,
+      "learning_rate": 0.000529271421620039,
+      "loss": 3.0409,
+      "step": 13658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48867014050483704,
+      "learning_rate": 0.000529261491787809,
+      "loss": 3.2267,
+      "step": 13659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5091564059257507,
+      "learning_rate": 0.0005292515613517483,
+      "loss": 3.2667,
+      "step": 13660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5306949615478516,
+      "learning_rate": 0.0005292416303118832,
+      "loss": 3.027,
+      "step": 13661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48959165811538696,
+      "learning_rate": 0.0005292316986682396,
+      "loss": 3.1372,
+      "step": 13662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5132994055747986,
+      "learning_rate": 0.000529221766420844,
+      "loss": 3.0894,
+      "step": 13663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48111510276794434,
+      "learning_rate": 0.0005292118335697223,
+      "loss": 3.4498,
+      "step": 13664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.522191047668457,
+      "learning_rate": 0.0005292019001149008,
+      "loss": 3.1612,
+      "step": 13665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4841899275779724,
+      "learning_rate": 0.0005291919660564055,
+      "loss": 3.299,
+      "step": 13666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5242030620574951,
+      "learning_rate": 0.0005291820313942627,
+      "loss": 3.4106,
+      "step": 13667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5671795010566711,
+      "learning_rate": 0.0005291720961284986,
+      "loss": 3.073,
+      "step": 13668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.537661612033844,
+      "learning_rate": 0.0005291621602591393,
+      "loss": 3.2446,
+      "step": 13669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.530839204788208,
+      "learning_rate": 0.0005291522237862109,
+      "loss": 3.369,
+      "step": 13670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5125889182090759,
+      "learning_rate": 0.0005291422867097397,
+      "loss": 3.1252,
+      "step": 13671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5530248880386353,
+      "learning_rate": 0.0005291323490297518,
+      "loss": 3.025,
+      "step": 13672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5228330492973328,
+      "learning_rate": 0.0005291224107462734,
+      "loss": 3.2608,
+      "step": 13673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5220538973808289,
+      "learning_rate": 0.0005291124718593307,
+      "loss": 3.1958,
+      "step": 13674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5133023858070374,
+      "learning_rate": 0.0005291025323689497,
+      "loss": 3.341,
+      "step": 13675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5261237621307373,
+      "learning_rate": 0.0005290925922751569,
+      "loss": 3.1028,
+      "step": 13676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5116633176803589,
+      "learning_rate": 0.0005290826515779782,
+      "loss": 3.0981,
+      "step": 13677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4918628931045532,
+      "learning_rate": 0.00052907271027744,
+      "loss": 3.0695,
+      "step": 13678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5067062973976135,
+      "learning_rate": 0.0005290627683735682,
+      "loss": 3.1897,
+      "step": 13679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5566603541374207,
+      "learning_rate": 0.0005290528258663892,
+      "loss": 3.2876,
+      "step": 13680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5340036153793335,
+      "learning_rate": 0.0005290428827559292,
+      "loss": 3.0107,
+      "step": 13681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5094714164733887,
+      "learning_rate": 0.0005290329390422142,
+      "loss": 3.1796,
+      "step": 13682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5311017632484436,
+      "learning_rate": 0.0005290229947252707,
+      "loss": 3.1173,
+      "step": 13683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5469943881034851,
+      "learning_rate": 0.0005290130498051246,
+      "loss": 3.2645,
+      "step": 13684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49231579899787903,
+      "learning_rate": 0.0005290031042818022,
+      "loss": 3.0617,
+      "step": 13685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5386189222335815,
+      "learning_rate": 0.0005289931581553297,
+      "loss": 3.2606,
+      "step": 13686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5233540534973145,
+      "learning_rate": 0.0005289832114257333,
+      "loss": 3.2946,
+      "step": 13687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5471063256263733,
+      "learning_rate": 0.0005289732640930393,
+      "loss": 3.0275,
+      "step": 13688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5296842455863953,
+      "learning_rate": 0.0005289633161572737,
+      "loss": 3.3652,
+      "step": 13689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.541083037853241,
+      "learning_rate": 0.0005289533676184627,
+      "loss": 3.2787,
+      "step": 13690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.527202308177948,
+      "learning_rate": 0.0005289434184766326,
+      "loss": 3.1486,
+      "step": 13691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5069360136985779,
+      "learning_rate": 0.0005289334687318098,
+      "loss": 3.2396,
+      "step": 13692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5278681516647339,
+      "learning_rate": 0.00052892351838402,
+      "loss": 3.3923,
+      "step": 13693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5264055728912354,
+      "learning_rate": 0.0005289135674332899,
+      "loss": 3.342,
+      "step": 13694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5079107284545898,
+      "learning_rate": 0.0005289036158796455,
+      "loss": 3.0714,
+      "step": 13695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5393988490104675,
+      "learning_rate": 0.000528893663723113,
+      "loss": 3.156,
+      "step": 13696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5057147145271301,
+      "learning_rate": 0.0005288837109637187,
+      "loss": 3.2109,
+      "step": 13697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5622931718826294,
+      "learning_rate": 0.0005288737576014887,
+      "loss": 3.3168,
+      "step": 13698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5340046286582947,
+      "learning_rate": 0.0005288638036364493,
+      "loss": 3.1038,
+      "step": 13699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5570096969604492,
+      "learning_rate": 0.0005288538490686267,
+      "loss": 3.2489,
+      "step": 13700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5238069891929626,
+      "learning_rate": 0.0005288438938980471,
+      "loss": 3.2243,
+      "step": 13701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48438480496406555,
+      "learning_rate": 0.0005288339381247367,
+      "loss": 3.119,
+      "step": 13702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5663520693778992,
+      "learning_rate": 0.0005288239817487217,
+      "loss": 3.2317,
+      "step": 13703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5262908935546875,
+      "learning_rate": 0.0005288140247700285,
+      "loss": 3.1207,
+      "step": 13704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5518969893455505,
+      "learning_rate": 0.0005288040671886831,
+      "loss": 3.1587,
+      "step": 13705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5191221833229065,
+      "learning_rate": 0.0005287941090047118,
+      "loss": 3.2533,
+      "step": 13706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5094031095504761,
+      "learning_rate": 0.0005287841502181409,
+      "loss": 3.2238,
+      "step": 13707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.512729823589325,
+      "learning_rate": 0.0005287741908289967,
+      "loss": 3.2058,
+      "step": 13708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5242785215377808,
+      "learning_rate": 0.0005287642308373051,
+      "loss": 3.2444,
+      "step": 13709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5172874927520752,
+      "learning_rate": 0.0005287542702430926,
+      "loss": 3.3319,
+      "step": 13710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48693451285362244,
+      "learning_rate": 0.0005287443090463854,
+      "loss": 3.3414,
+      "step": 13711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5203180313110352,
+      "learning_rate": 0.0005287343472472097,
+      "loss": 3.1379,
+      "step": 13712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5015645623207092,
+      "learning_rate": 0.0005287243848455918,
+      "loss": 3.2083,
+      "step": 13713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5029230117797852,
+      "learning_rate": 0.0005287144218415579,
+      "loss": 2.9887,
+      "step": 13714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.523784875869751,
+      "learning_rate": 0.0005287044582351341,
+      "loss": 2.9404,
+      "step": 13715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49073922634124756,
+      "learning_rate": 0.000528694494026347,
+      "loss": 3.119,
+      "step": 13716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5473636984825134,
+      "learning_rate": 0.0005286845292152224,
+      "loss": 3.2291,
+      "step": 13717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4929855763912201,
+      "learning_rate": 0.0005286745638017868,
+      "loss": 3.1537,
+      "step": 13718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5156190991401672,
+      "learning_rate": 0.0005286645977860664,
+      "loss": 3.1542,
+      "step": 13719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4985933303833008,
+      "learning_rate": 0.0005286546311680876,
+      "loss": 3.0746,
+      "step": 13720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5488167405128479,
+      "learning_rate": 0.0005286446639478764,
+      "loss": 3.2484,
+      "step": 13721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5215466022491455,
+      "learning_rate": 0.0005286346961254591,
+      "loss": 3.024,
+      "step": 13722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5182087421417236,
+      "learning_rate": 0.0005286247277008621,
+      "loss": 3.2172,
+      "step": 13723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5041150450706482,
+      "learning_rate": 0.0005286147586741115,
+      "loss": 3.0219,
+      "step": 13724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5143710970878601,
+      "learning_rate": 0.0005286047890452337,
+      "loss": 3.1054,
+      "step": 13725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5302610993385315,
+      "learning_rate": 0.0005285948188142549,
+      "loss": 3.19,
+      "step": 13726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5241024494171143,
+      "learning_rate": 0.0005285848479812012,
+      "loss": 3.2024,
+      "step": 13727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5108481645584106,
+      "learning_rate": 0.0005285748765460991,
+      "loss": 3.3537,
+      "step": 13728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4921933710575104,
+      "learning_rate": 0.0005285649045089748,
+      "loss": 3.1969,
+      "step": 13729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5298369526863098,
+      "learning_rate": 0.0005285549318698544,
+      "loss": 3.1773,
+      "step": 13730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5141018033027649,
+      "learning_rate": 0.0005285449586287644,
+      "loss": 3.2114,
+      "step": 13731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5361664295196533,
+      "learning_rate": 0.0005285349847857309,
+      "loss": 3.0577,
+      "step": 13732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5357168912887573,
+      "learning_rate": 0.0005285250103407804,
+      "loss": 3.2206,
+      "step": 13733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5472400784492493,
+      "learning_rate": 0.0005285150352939388,
+      "loss": 3.0433,
+      "step": 13734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49933162331581116,
+      "learning_rate": 0.0005285050596452326,
+      "loss": 3.4927,
+      "step": 13735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5269633531570435,
+      "learning_rate": 0.0005284950833946882,
+      "loss": 3.1318,
+      "step": 13736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5392086505889893,
+      "learning_rate": 0.0005284851065423316,
+      "loss": 3.3404,
+      "step": 13737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4980263411998749,
+      "learning_rate": 0.0005284751290881893,
+      "loss": 3.2672,
+      "step": 13738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49644413590431213,
+      "learning_rate": 0.0005284651510322874,
+      "loss": 3.2669,
+      "step": 13739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5201795697212219,
+      "learning_rate": 0.0005284551723746523,
+      "loss": 3.4163,
+      "step": 13740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5354999899864197,
+      "learning_rate": 0.0005284451931153102,
+      "loss": 3.2036,
+      "step": 13741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5267964601516724,
+      "learning_rate": 0.0005284352132542873,
+      "loss": 3.118,
+      "step": 13742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5029120445251465,
+      "learning_rate": 0.0005284252327916102,
+      "loss": 3.2955,
+      "step": 13743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5081202983856201,
+      "learning_rate": 0.000528415251727305,
+      "loss": 3.2133,
+      "step": 13744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.517963707447052,
+      "learning_rate": 0.000528405270061398,
+      "loss": 2.9726,
+      "step": 13745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55925452709198,
+      "learning_rate": 0.0005283952877939153,
+      "loss": 3.3627,
+      "step": 13746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5444297790527344,
+      "learning_rate": 0.0005283853049248834,
+      "loss": 3.1062,
+      "step": 13747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48813849687576294,
+      "learning_rate": 0.0005283753214543287,
+      "loss": 3.2675,
+      "step": 13748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234923362731934,
+      "learning_rate": 0.0005283653373822773,
+      "loss": 3.2118,
+      "step": 13749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.487904816865921,
+      "learning_rate": 0.0005283553527087556,
+      "loss": 3.2319,
+      "step": 13750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5277583599090576,
+      "learning_rate": 0.0005283453674337898,
+      "loss": 3.3372,
+      "step": 13751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5574771165847778,
+      "learning_rate": 0.0005283353815574063,
+      "loss": 3.1972,
+      "step": 13752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5329998731613159,
+      "learning_rate": 0.0005283253950796312,
+      "loss": 3.1427,
+      "step": 13753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4901135265827179,
+      "learning_rate": 0.000528315408000491,
+      "loss": 3.5074,
+      "step": 13754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.524285078048706,
+      "learning_rate": 0.000528305420320012,
+      "loss": 3.2045,
+      "step": 13755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5120623707771301,
+      "learning_rate": 0.0005282954320382205,
+      "loss": 3.2821,
+      "step": 13756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4949445426464081,
+      "learning_rate": 0.0005282854431551428,
+      "loss": 3.238,
+      "step": 13757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5326241850852966,
+      "learning_rate": 0.0005282754536708051,
+      "loss": 3.193,
+      "step": 13758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5440043807029724,
+      "learning_rate": 0.0005282654635852339,
+      "loss": 3.0779,
+      "step": 13759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5041990876197815,
+      "learning_rate": 0.0005282554728984551,
+      "loss": 3.1848,
+      "step": 13760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5535484552383423,
+      "learning_rate": 0.0005282454816104956,
+      "loss": 3.2104,
+      "step": 13761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.503193736076355,
+      "learning_rate": 0.0005282354897213812,
+      "loss": 3.0778,
+      "step": 13762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49220454692840576,
+      "learning_rate": 0.0005282254972311386,
+      "loss": 3.2279,
+      "step": 13763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.516409695148468,
+      "learning_rate": 0.000528215504139794,
+      "loss": 3.0826,
+      "step": 13764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5269769430160522,
+      "learning_rate": 0.0005282055104473736,
+      "loss": 3.3278,
+      "step": 13765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5283230543136597,
+      "learning_rate": 0.0005281955161539039,
+      "loss": 3.1647,
+      "step": 13766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5059794187545776,
+      "learning_rate": 0.0005281855212594111,
+      "loss": 3.2336,
+      "step": 13767
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53988116979599,
+      "learning_rate": 0.0005281755257639214,
+      "loss": 3.0983,
+      "step": 13768
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.520753800868988,
+      "learning_rate": 0.0005281655296674615,
+      "loss": 3.1827,
+      "step": 13769
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5291000008583069,
+      "learning_rate": 0.0005281555329700574,
+      "loss": 3.2596,
+      "step": 13770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5507010817527771,
+      "learning_rate": 0.0005281455356717355,
+      "loss": 3.1211,
+      "step": 13771
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49522608518600464,
+      "learning_rate": 0.0005281355377725223,
+      "loss": 3.1505,
+      "step": 13772
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5150662660598755,
+      "learning_rate": 0.0005281255392724438,
+      "loss": 3.3192,
+      "step": 13773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5103153586387634,
+      "learning_rate": 0.0005281155401715267,
+      "loss": 3.0925,
+      "step": 13774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5034166574478149,
+      "learning_rate": 0.000528105540469797,
+      "loss": 3.3163,
+      "step": 13775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5109154582023621,
+      "learning_rate": 0.0005280955401672814,
+      "loss": 3.3177,
+      "step": 13776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4964618682861328,
+      "learning_rate": 0.0005280855392640059,
+      "loss": 3.0376,
+      "step": 13777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47001469135284424,
+      "learning_rate": 0.000528075537759997,
+      "loss": 3.3065,
+      "step": 13778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49114564061164856,
+      "learning_rate": 0.0005280655356552811,
+      "loss": 3.3961,
+      "step": 13779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5325284600257874,
+      "learning_rate": 0.0005280555329498845,
+      "loss": 3.2993,
+      "step": 13780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5186975598335266,
+      "learning_rate": 0.0005280455296438333,
+      "loss": 3.1416,
+      "step": 13781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5972892045974731,
+      "learning_rate": 0.0005280355257371544,
+      "loss": 3.4019,
+      "step": 13782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5253905653953552,
+      "learning_rate": 0.0005280255212298735,
+      "loss": 3.1386,
+      "step": 13783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5274996161460876,
+      "learning_rate": 0.0005280155161220174,
+      "loss": 3.2277,
+      "step": 13784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5044057369232178,
+      "learning_rate": 0.0005280055104136123,
+      "loss": 3.2136,
+      "step": 13785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49711552262306213,
+      "learning_rate": 0.0005279955041046845,
+      "loss": 3.3523,
+      "step": 13786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5024986267089844,
+      "learning_rate": 0.0005279854971952606,
+      "loss": 3.0944,
+      "step": 13787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5386544466018677,
+      "learning_rate": 0.0005279754896853667,
+      "loss": 3.4201,
+      "step": 13788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5329682230949402,
+      "learning_rate": 0.0005279654815750291,
+      "loss": 3.145,
+      "step": 13789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5430646538734436,
+      "learning_rate": 0.0005279554728642744,
+      "loss": 3.4528,
+      "step": 13790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5068038105964661,
+      "learning_rate": 0.0005279454635531289,
+      "loss": 3.2641,
+      "step": 13791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4926636219024658,
+      "learning_rate": 0.0005279354536416188,
+      "loss": 3.1893,
+      "step": 13792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5283586382865906,
+      "learning_rate": 0.0005279254431297707,
+      "loss": 3.173,
+      "step": 13793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5072885155677795,
+      "learning_rate": 0.0005279154320176107,
+      "loss": 3.0823,
+      "step": 13794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4954681694507599,
+      "learning_rate": 0.0005279054203051655,
+      "loss": 3.2519,
+      "step": 13795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.50550377368927,
+      "learning_rate": 0.0005278954079924611,
+      "loss": 3.2629,
+      "step": 13796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5151498913764954,
+      "learning_rate": 0.0005278853950795242,
+      "loss": 3.0185,
+      "step": 13797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5135114192962646,
+      "learning_rate": 0.0005278753815663811,
+      "loss": 3.1437,
+      "step": 13798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49339157342910767,
+      "learning_rate": 0.0005278653674530579,
+      "loss": 3.2054,
+      "step": 13799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5128827691078186,
+      "learning_rate": 0.0005278553527395813,
+      "loss": 3.1548,
+      "step": 13800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5171065330505371,
+      "learning_rate": 0.0005278453374259776,
+      "loss": 3.2015,
+      "step": 13801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5121150612831116,
+      "learning_rate": 0.0005278353215122729,
+      "loss": 3.3808,
+      "step": 13802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5693663954734802,
+      "learning_rate": 0.0005278253049984939,
+      "loss": 3.0008,
+      "step": 13803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.506459653377533,
+      "learning_rate": 0.000527815287884667,
+      "loss": 3.16,
+      "step": 13804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5166449546813965,
+      "learning_rate": 0.0005278052701708184,
+      "loss": 3.1424,
+      "step": 13805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5361778736114502,
+      "learning_rate": 0.0005277952518569746,
+      "loss": 3.4089,
+      "step": 13806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5384201407432556,
+      "learning_rate": 0.000527785232943162,
+      "loss": 3.3915,
+      "step": 13807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.581087052822113,
+      "learning_rate": 0.0005277752134294067,
+      "loss": 3.2422,
+      "step": 13808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4882732331752777,
+      "learning_rate": 0.0005277651933157355,
+      "loss": 3.3192,
+      "step": 13809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5098223686218262,
+      "learning_rate": 0.0005277551726021746,
+      "loss": 3.0037,
+      "step": 13810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.522487223148346,
+      "learning_rate": 0.0005277451512887504,
+      "loss": 3.3347,
+      "step": 13811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5181809663772583,
+      "learning_rate": 0.0005277351293754894,
+      "loss": 3.2624,
+      "step": 13812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5491748452186584,
+      "learning_rate": 0.0005277251068624178,
+      "loss": 3.2306,
+      "step": 13813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5368053913116455,
+      "learning_rate": 0.0005277150837495621,
+      "loss": 3.2325,
+      "step": 13814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5411176085472107,
+      "learning_rate": 0.0005277050600369487,
+      "loss": 3.1715,
+      "step": 13815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49559369683265686,
+      "learning_rate": 0.000527695035724604,
+      "loss": 3.0924,
+      "step": 13816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5154446363449097,
+      "learning_rate": 0.0005276850108125544,
+      "loss": 3.6344,
+      "step": 13817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5071130990982056,
+      "learning_rate": 0.0005276749853008263,
+      "loss": 3.0258,
+      "step": 13818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5003767013549805,
+      "learning_rate": 0.0005276649591894459,
+      "loss": 3.0506,
+      "step": 13819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5963944792747498,
+      "learning_rate": 0.0005276549324784401,
+      "loss": 3.0792,
+      "step": 13820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5032363533973694,
+      "learning_rate": 0.0005276449051678349,
+      "loss": 3.1895,
+      "step": 13821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48575350642204285,
+      "learning_rate": 0.000527634877257657,
+      "loss": 3.0814,
+      "step": 13822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5336856245994568,
+      "learning_rate": 0.0005276248487479325,
+      "loss": 3.0062,
+      "step": 13823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.519405722618103,
+      "learning_rate": 0.0005276148196386879,
+      "loss": 3.1346,
+      "step": 13824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5641560554504395,
+      "learning_rate": 0.0005276047899299497,
+      "loss": 3.1063,
+      "step": 13825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49971362948417664,
+      "learning_rate": 0.0005275947596217442,
+      "loss": 3.1934,
+      "step": 13826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5345155596733093,
+      "learning_rate": 0.0005275847287140981,
+      "loss": 3.1373,
+      "step": 13827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5394874811172485,
+      "learning_rate": 0.0005275746972070376,
+      "loss": 3.301,
+      "step": 13828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5135223865509033,
+      "learning_rate": 0.0005275646651005891,
+      "loss": 3.301,
+      "step": 13829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5040108561515808,
+      "learning_rate": 0.000527554632394779,
+      "loss": 3.2598,
+      "step": 13830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5344460010528564,
+      "learning_rate": 0.0005275445990896338,
+      "loss": 3.1245,
+      "step": 13831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4787933826446533,
+      "learning_rate": 0.00052753456518518,
+      "loss": 3.3899,
+      "step": 13832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5142536759376526,
+      "learning_rate": 0.0005275245306814439,
+      "loss": 3.1408,
+      "step": 13833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5277648568153381,
+      "learning_rate": 0.000527514495578452,
+      "loss": 3.1087,
+      "step": 13834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5267074704170227,
+      "learning_rate": 0.0005275044598762307,
+      "loss": 3.192,
+      "step": 13835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5166609883308411,
+      "learning_rate": 0.0005274944235748064,
+      "loss": 3.3189,
+      "step": 13836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5297030210494995,
+      "learning_rate": 0.0005274843866742056,
+      "loss": 3.0285,
+      "step": 13837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5255849361419678,
+      "learning_rate": 0.0005274743491744548,
+      "loss": 3.2142,
+      "step": 13838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49405017495155334,
+      "learning_rate": 0.0005274643110755801,
+      "loss": 3.3999,
+      "step": 13839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4815720319747925,
+      "learning_rate": 0.0005274542723776083,
+      "loss": 3.4118,
+      "step": 13840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5125225186347961,
+      "learning_rate": 0.0005274442330805658,
+      "loss": 3.1375,
+      "step": 13841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5525969862937927,
+      "learning_rate": 0.0005274341931844788,
+      "loss": 3.3452,
+      "step": 13842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5333914756774902,
+      "learning_rate": 0.0005274241526893741,
+      "loss": 3.3247,
+      "step": 13843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5027710199356079,
+      "learning_rate": 0.0005274141115952779,
+      "loss": 3.228,
+      "step": 13844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5438841581344604,
+      "learning_rate": 0.0005274040699022167,
+      "loss": 3.2234,
+      "step": 13845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5225579142570496,
+      "learning_rate": 0.0005273940276102168,
+      "loss": 3.165,
+      "step": 13846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5789764523506165,
+      "learning_rate": 0.000527383984719305,
+      "loss": 3.1446,
+      "step": 13847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5264090895652771,
+      "learning_rate": 0.0005273739412295073,
+      "loss": 3.2548,
+      "step": 13848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5331341028213501,
+      "learning_rate": 0.0005273638971408506,
+      "loss": 3.1594,
+      "step": 13849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5145845413208008,
+      "learning_rate": 0.0005273538524533612,
+      "loss": 3.1882,
+      "step": 13850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5996436476707458,
+      "learning_rate": 0.0005273438071670653,
+      "loss": 3.359,
+      "step": 13851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5678809881210327,
+      "learning_rate": 0.0005273337612819897,
+      "loss": 3.3849,
+      "step": 13852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49239474534988403,
+      "learning_rate": 0.0005273237147981607,
+      "loss": 3.1734,
+      "step": 13853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5626592040061951,
+      "learning_rate": 0.0005273136677156048,
+      "loss": 3.0927,
+      "step": 13854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49832597374916077,
+      "learning_rate": 0.0005273036200343484,
+      "loss": 3.3677,
+      "step": 13855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6042376160621643,
+      "learning_rate": 0.000527293571754418,
+      "loss": 3.2694,
+      "step": 13856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5671773552894592,
+      "learning_rate": 0.0005272835228758401,
+      "loss": 2.8949,
+      "step": 13857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5112889409065247,
+      "learning_rate": 0.0005272734733986411,
+      "loss": 3.0336,
+      "step": 13858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5109499096870422,
+      "learning_rate": 0.0005272634233228476,
+      "loss": 3.3543,
+      "step": 13859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5124879479408264,
+      "learning_rate": 0.000527253372648486,
+      "loss": 3.2364,
+      "step": 13860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5452317595481873,
+      "learning_rate": 0.0005272433213755827,
+      "loss": 3.1102,
+      "step": 13861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5369084477424622,
+      "learning_rate": 0.0005272332695041642,
+      "loss": 3.2783,
+      "step": 13862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5925998091697693,
+      "learning_rate": 0.0005272232170342569,
+      "loss": 3.0603,
+      "step": 13863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5368659496307373,
+      "learning_rate": 0.0005272131639658876,
+      "loss": 3.1947,
+      "step": 13864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.516013503074646,
+      "learning_rate": 0.0005272031102990824,
+      "loss": 3.3695,
+      "step": 13865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5412189960479736,
+      "learning_rate": 0.000527193056033868,
+      "loss": 2.8892,
+      "step": 13866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5142117142677307,
+      "learning_rate": 0.0005271830011702708,
+      "loss": 3.292,
+      "step": 13867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4763852059841156,
+      "learning_rate": 0.0005271729457083173,
+      "loss": 3.2973,
+      "step": 13868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5418897271156311,
+      "learning_rate": 0.000527162889648034,
+      "loss": 3.2097,
+      "step": 13869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5203564763069153,
+      "learning_rate": 0.0005271528329894473,
+      "loss": 3.0707,
+      "step": 13870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5115311741828918,
+      "learning_rate": 0.0005271427757325839,
+      "loss": 3.2188,
+      "step": 13871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5365582704544067,
+      "learning_rate": 0.00052713271787747,
+      "loss": 3.2721,
+      "step": 13872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.529913067817688,
+      "learning_rate": 0.0005271226594241323,
+      "loss": 3.2094,
+      "step": 13873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5000056028366089,
+      "learning_rate": 0.0005271126003725972,
+      "loss": 2.963,
+      "step": 13874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5235258936882019,
+      "learning_rate": 0.0005271025407228912,
+      "loss": 3.294,
+      "step": 13875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5482226610183716,
+      "learning_rate": 0.0005270924804750408,
+      "loss": 3.2941,
+      "step": 13876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5100428462028503,
+      "learning_rate": 0.0005270824196290726,
+      "loss": 3.2611,
+      "step": 13877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5015908479690552,
+      "learning_rate": 0.000527072358185013,
+      "loss": 3.1427,
+      "step": 13878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5557044744491577,
+      "learning_rate": 0.0005270622961428883,
+      "loss": 3.1989,
+      "step": 13879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.529284656047821,
+      "learning_rate": 0.0005270522335027255,
+      "loss": 3.1541,
+      "step": 13880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5141466856002808,
+      "learning_rate": 0.0005270421702645506,
+      "loss": 3.3865,
+      "step": 13881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5499520897865295,
+      "learning_rate": 0.0005270321064283905,
+      "loss": 3.1756,
+      "step": 13882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5323184728622437,
+      "learning_rate": 0.0005270220419942714,
+      "loss": 3.2529,
+      "step": 13883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5083839893341064,
+      "learning_rate": 0.00052701197696222,
+      "loss": 3.2344,
+      "step": 13884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5136963129043579,
+      "learning_rate": 0.0005270019113322626,
+      "loss": 3.2175,
+      "step": 13885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5176054239273071,
+      "learning_rate": 0.000526991845104426,
+      "loss": 3.0358,
+      "step": 13886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49073687195777893,
+      "learning_rate": 0.0005269817782787365,
+      "loss": 3.1841,
+      "step": 13887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.495728999376297,
+      "learning_rate": 0.0005269717108552208,
+      "loss": 3.3771,
+      "step": 13888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5247365832328796,
+      "learning_rate": 0.0005269616428339052,
+      "loss": 3.1361,
+      "step": 13889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5086815357208252,
+      "learning_rate": 0.0005269515742148163,
+      "loss": 3.1449,
+      "step": 13890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5526536703109741,
+      "learning_rate": 0.0005269415049979807,
+      "loss": 3.1725,
+      "step": 13891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5365892052650452,
+      "learning_rate": 0.0005269314351834247,
+      "loss": 3.2791,
+      "step": 13892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5318396687507629,
+      "learning_rate": 0.000526921364771175,
+      "loss": 3.0982,
+      "step": 13893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5801922082901001,
+      "learning_rate": 0.0005269112937612582,
+      "loss": 3.0354,
+      "step": 13894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4854373335838318,
+      "learning_rate": 0.0005269012221537008,
+      "loss": 3.0311,
+      "step": 13895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5157737731933594,
+      "learning_rate": 0.0005268911499485291,
+      "loss": 3.2959,
+      "step": 13896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5271645784378052,
+      "learning_rate": 0.0005268810771457698,
+      "loss": 3.4125,
+      "step": 13897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5193085074424744,
+      "learning_rate": 0.0005268710037454494,
+      "loss": 3.1424,
+      "step": 13898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5354040861129761,
+      "learning_rate": 0.0005268609297475944,
+      "loss": 3.0197,
+      "step": 13899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5036852955818176,
+      "learning_rate": 0.0005268508551522315,
+      "loss": 3.3795,
+      "step": 13900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.527779757976532,
+      "learning_rate": 0.000526840779959387,
+      "loss": 3.1321,
+      "step": 13901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5030642747879028,
+      "learning_rate": 0.0005268307041690877,
+      "loss": 3.3091,
+      "step": 13902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4996885061264038,
+      "learning_rate": 0.0005268206277813598,
+      "loss": 3.2489,
+      "step": 13903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5028959512710571,
+      "learning_rate": 0.0005268105507962301,
+      "loss": 3.2326,
+      "step": 13904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49086108803749084,
+      "learning_rate": 0.000526800473213725,
+      "loss": 3.3486,
+      "step": 13905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.508350133895874,
+      "learning_rate": 0.0005267903950338711,
+      "loss": 3.2829,
+      "step": 13906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54703289270401,
+      "learning_rate": 0.000526780316256695,
+      "loss": 3.3257,
+      "step": 13907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5806416273117065,
+      "learning_rate": 0.0005267702368822232,
+      "loss": 3.2072,
+      "step": 13908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4797351062297821,
+      "learning_rate": 0.0005267601569104823,
+      "loss": 3.5585,
+      "step": 13909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5201489329338074,
+      "learning_rate": 0.0005267500763414986,
+      "loss": 3.0275,
+      "step": 13910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5400049686431885,
+      "learning_rate": 0.000526739995175299,
+      "loss": 3.2039,
+      "step": 13911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5291928052902222,
+      "learning_rate": 0.0005267299134119098,
+      "loss": 3.1484,
+      "step": 13912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49172693490982056,
+      "learning_rate": 0.0005267198310513577,
+      "loss": 3.0451,
+      "step": 13913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5199759006500244,
+      "learning_rate": 0.0005267097480936691,
+      "loss": 3.1473,
+      "step": 13914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5097964406013489,
+      "learning_rate": 0.0005266996645388708,
+      "loss": 3.0738,
+      "step": 13915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5464935898780823,
+      "learning_rate": 0.0005266895803869891,
+      "loss": 3.2276,
+      "step": 13916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4840008616447449,
+      "learning_rate": 0.0005266794956380507,
+      "loss": 3.1427,
+      "step": 13917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5694568753242493,
+      "learning_rate": 0.0005266694102920822,
+      "loss": 3.1599,
+      "step": 13918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5026246905326843,
+      "learning_rate": 0.00052665932434911,
+      "loss": 3.2966,
+      "step": 13919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5052033066749573,
+      "learning_rate": 0.0005266492378091608,
+      "loss": 3.3566,
+      "step": 13920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5257059931755066,
+      "learning_rate": 0.0005266391506722611,
+      "loss": 3.1304,
+      "step": 13921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5384652018547058,
+      "learning_rate": 0.0005266290629384376,
+      "loss": 3.2555,
+      "step": 13922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5102568864822388,
+      "learning_rate": 0.0005266189746077167,
+      "loss": 3.2415,
+      "step": 13923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5261817574501038,
+      "learning_rate": 0.000526608885680125,
+      "loss": 3.3374,
+      "step": 13924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4975242018699646,
+      "learning_rate": 0.0005265987961556891,
+      "loss": 3.2718,
+      "step": 13925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5032918453216553,
+      "learning_rate": 0.0005265887060344356,
+      "loss": 3.2327,
+      "step": 13926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6491032242774963,
+      "learning_rate": 0.000526578615316391,
+      "loss": 3.183,
+      "step": 13927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5199459791183472,
+      "learning_rate": 0.0005265685240015821,
+      "loss": 3.2972,
+      "step": 13928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5117924809455872,
+      "learning_rate": 0.0005265584320900353,
+      "loss": 3.1965,
+      "step": 13929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5264638662338257,
+      "learning_rate": 0.000526548339581777,
+      "loss": 3.0494,
+      "step": 13930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5094579458236694,
+      "learning_rate": 0.0005265382464768341,
+      "loss": 3.18,
+      "step": 13931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4896301329135895,
+      "learning_rate": 0.0005265281527752331,
+      "loss": 3.3233,
+      "step": 13932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4931809604167938,
+      "learning_rate": 0.0005265180584770004,
+      "loss": 3.3084,
+      "step": 13933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5263796448707581,
+      "learning_rate": 0.0005265079635821628,
+      "loss": 3.3991,
+      "step": 13934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5130126476287842,
+      "learning_rate": 0.0005264978680907469,
+      "loss": 3.2796,
+      "step": 13935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5254062414169312,
+      "learning_rate": 0.000526487772002779,
+      "loss": 3.2386,
+      "step": 13936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5328313708305359,
+      "learning_rate": 0.0005264776753182861,
+      "loss": 3.0323,
+      "step": 13937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4820843040943146,
+      "learning_rate": 0.0005264675780372945,
+      "loss": 3.3463,
+      "step": 13938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5069538950920105,
+      "learning_rate": 0.0005264574801598309,
+      "loss": 3.1807,
+      "step": 13939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5414431691169739,
+      "learning_rate": 0.0005264473816859219,
+      "loss": 3.4286,
+      "step": 13940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5100494623184204,
+      "learning_rate": 0.0005264372826155941,
+      "loss": 3.1482,
+      "step": 13941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5338406562805176,
+      "learning_rate": 0.000526427182948874,
+      "loss": 3.3412,
+      "step": 13942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48375827074050903,
+      "learning_rate": 0.0005264170826857883,
+      "loss": 3.1934,
+      "step": 13943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5023521184921265,
+      "learning_rate": 0.0005264069818263636,
+      "loss": 3.1973,
+      "step": 13944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5352455377578735,
+      "learning_rate": 0.0005263968803706265,
+      "loss": 3.0357,
+      "step": 13945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4851732552051544,
+      "learning_rate": 0.0005263867783186036,
+      "loss": 3.5018,
+      "step": 13946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5262126326560974,
+      "learning_rate": 0.0005263766756703213,
+      "loss": 3.0649,
+      "step": 13947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51038658618927,
+      "learning_rate": 0.0005263665724258066,
+      "loss": 3.3052,
+      "step": 13948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49130168557167053,
+      "learning_rate": 0.000526356468585086,
+      "loss": 3.0464,
+      "step": 13949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.525030791759491,
+      "learning_rate": 0.0005263463641481858,
+      "loss": 3.4954,
+      "step": 13950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5531750917434692,
+      "learning_rate": 0.000526336259115133,
+      "loss": 3.3464,
+      "step": 13951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5086334347724915,
+      "learning_rate": 0.0005263261534859539,
+      "loss": 3.421,
+      "step": 13952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5100414156913757,
+      "learning_rate": 0.0005263160472606754,
+      "loss": 3.261,
+      "step": 13953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5146250128746033,
+      "learning_rate": 0.0005263059404393239,
+      "loss": 3.0504,
+      "step": 13954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5014898180961609,
+      "learning_rate": 0.0005262958330219262,
+      "loss": 3.3744,
+      "step": 13955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5280792117118835,
+      "learning_rate": 0.0005262857250085088,
+      "loss": 3.1879,
+      "step": 13956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48315441608428955,
+      "learning_rate": 0.0005262756163990982,
+      "loss": 3.1047,
+      "step": 13957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.508493959903717,
+      "learning_rate": 0.0005262655071937214,
+      "loss": 3.3101,
+      "step": 13958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5163034796714783,
+      "learning_rate": 0.0005262553973924047,
+      "loss": 3.4717,
+      "step": 13959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5181001424789429,
+      "learning_rate": 0.0005262452869951749,
+      "loss": 3.2164,
+      "step": 13960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.526516854763031,
+      "learning_rate": 0.0005262351760020584,
+      "loss": 3.2027,
+      "step": 13961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5265016555786133,
+      "learning_rate": 0.000526225064413082,
+      "loss": 3.2067,
+      "step": 13962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.508037269115448,
+      "learning_rate": 0.0005262149522282724,
+      "loss": 3.2605,
+      "step": 13963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5375155806541443,
+      "learning_rate": 0.0005262048394476562,
+      "loss": 3.3783,
+      "step": 13964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5235868096351624,
+      "learning_rate": 0.0005261947260712599,
+      "loss": 3.3954,
+      "step": 13965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5014271140098572,
+      "learning_rate": 0.0005261846120991103,
+      "loss": 3.2368,
+      "step": 13966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5021702647209167,
+      "learning_rate": 0.0005261744975312339,
+      "loss": 3.1191,
+      "step": 13967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5597658753395081,
+      "learning_rate": 0.0005261643823676574,
+      "loss": 3.0137,
+      "step": 13968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4978190064430237,
+      "learning_rate": 0.0005261542666084074,
+      "loss": 3.3998,
+      "step": 13969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5252299308776855,
+      "learning_rate": 0.0005261441502535107,
+      "loss": 3.0559,
+      "step": 13970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5657221674919128,
+      "learning_rate": 0.0005261340333029938,
+      "loss": 3.1933,
+      "step": 13971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5250703692436218,
+      "learning_rate": 0.0005261239157568833,
+      "loss": 3.3627,
+      "step": 13972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5377254486083984,
+      "learning_rate": 0.0005261137976152059,
+      "loss": 3.1583,
+      "step": 13973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49435439705848694,
+      "learning_rate": 0.0005261036788779884,
+      "loss": 3.2259,
+      "step": 13974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5244966149330139,
+      "learning_rate": 0.0005260935595452573,
+      "loss": 3.2558,
+      "step": 13975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4984242916107178,
+      "learning_rate": 0.0005260834396170393,
+      "loss": 3.1271,
+      "step": 13976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5036216974258423,
+      "learning_rate": 0.000526073319093361,
+      "loss": 3.2616,
+      "step": 13977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5375218987464905,
+      "learning_rate": 0.000526063197974249,
+      "loss": 3.0446,
+      "step": 13978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5373899936676025,
+      "learning_rate": 0.00052605307625973,
+      "loss": 3.123,
+      "step": 13979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5408071279525757,
+      "learning_rate": 0.0005260429539498308,
+      "loss": 3.0805,
+      "step": 13980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5188453197479248,
+      "learning_rate": 0.000526032831044578,
+      "loss": 3.1694,
+      "step": 13981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5271825194358826,
+      "learning_rate": 0.0005260227075439983,
+      "loss": 3.1811,
+      "step": 13982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47875532507896423,
+      "learning_rate": 0.0005260125834481181,
+      "loss": 3.494,
+      "step": 13983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5747538208961487,
+      "learning_rate": 0.0005260024587569644,
+      "loss": 3.1184,
+      "step": 13984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5151486992835999,
+      "learning_rate": 0.0005259923334705636,
+      "loss": 2.9681,
+      "step": 13985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.534481406211853,
+      "learning_rate": 0.0005259822075889426,
+      "loss": 3.2856,
+      "step": 13986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51065993309021,
+      "learning_rate": 0.0005259720811121278,
+      "loss": 3.1268,
+      "step": 13987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5007278323173523,
+      "learning_rate": 0.0005259619540401462,
+      "loss": 3.0136,
+      "step": 13988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5178660154342651,
+      "learning_rate": 0.0005259518263730242,
+      "loss": 3.1685,
+      "step": 13989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5047518610954285,
+      "learning_rate": 0.0005259416981107886,
+      "loss": 3.2531,
+      "step": 13990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5168188810348511,
+      "learning_rate": 0.000525931569253466,
+      "loss": 3.3957,
+      "step": 13991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5354238152503967,
+      "learning_rate": 0.0005259214398010833,
+      "loss": 3.3066,
+      "step": 13992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5306437611579895,
+      "learning_rate": 0.0005259113097536667,
+      "loss": 3.2913,
+      "step": 13993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5510126948356628,
+      "learning_rate": 0.0005259011791112435,
+      "loss": 3.0978,
+      "step": 13994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5774776339530945,
+      "learning_rate": 0.00052589104787384,
+      "loss": 3.133,
+      "step": 13995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5207593441009521,
+      "learning_rate": 0.0005258809160414827,
+      "loss": 3.1222,
+      "step": 13996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5279645919799805,
+      "learning_rate": 0.0005258707836141989,
+      "loss": 3.2507,
+      "step": 13997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5136027336120605,
+      "learning_rate": 0.0005258606505920147,
+      "loss": 3.4416,
+      "step": 13998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49309471249580383,
+      "learning_rate": 0.0005258505169749571,
+      "loss": 3.1352,
+      "step": 13999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5222823023796082,
+      "learning_rate": 0.0005258403827630527,
+      "loss": 3.3458,
+      "step": 14000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4934551417827606,
+      "learning_rate": 0.0005258302479563282,
+      "loss": 3.129,
+      "step": 14001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5131190419197083,
+      "learning_rate": 0.0005258201125548103,
+      "loss": 3.2386,
+      "step": 14002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6161550879478455,
+      "learning_rate": 0.0005258099765585256,
+      "loss": 3.0293,
+      "step": 14003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5266488194465637,
+      "learning_rate": 0.0005257998399675009,
+      "loss": 3.2425,
+      "step": 14004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5947927832603455,
+      "learning_rate": 0.0005257897027817629,
+      "loss": 3.2695,
+      "step": 14005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5191298723220825,
+      "learning_rate": 0.0005257795650013382,
+      "loss": 3.0016,
+      "step": 14006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5333731770515442,
+      "learning_rate": 0.0005257694266262537,
+      "loss": 3.1303,
+      "step": 14007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5200604200363159,
+      "learning_rate": 0.000525759287656536,
+      "loss": 3.3124,
+      "step": 14008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5588985681533813,
+      "learning_rate": 0.0005257491480922117,
+      "loss": 3.262,
+      "step": 14009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49545398354530334,
+      "learning_rate": 0.0005257390079333077,
+      "loss": 3.2407,
+      "step": 14010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.550879180431366,
+      "learning_rate": 0.0005257288671798504,
+      "loss": 3.0096,
+      "step": 14011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5142801403999329,
+      "learning_rate": 0.0005257187258318668,
+      "loss": 3.261,
+      "step": 14012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5323050022125244,
+      "learning_rate": 0.0005257085838893836,
+      "loss": 3.0495,
+      "step": 14013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5185598731040955,
+      "learning_rate": 0.0005256984413524274,
+      "loss": 3.1516,
+      "step": 14014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5257108807563782,
+      "learning_rate": 0.0005256882982210249,
+      "loss": 3.1726,
+      "step": 14015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5254387259483337,
+      "learning_rate": 0.0005256781544952028,
+      "loss": 3.2881,
+      "step": 14016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.514013409614563,
+      "learning_rate": 0.000525668010174988,
+      "loss": 3.06,
+      "step": 14017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5110748410224915,
+      "learning_rate": 0.000525657865260407,
+      "loss": 3.283,
+      "step": 14018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5237401127815247,
+      "learning_rate": 0.0005256477197514866,
+      "loss": 3.1188,
+      "step": 14019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4896359145641327,
+      "learning_rate": 0.0005256375736482536,
+      "loss": 3.2941,
+      "step": 14020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5119338035583496,
+      "learning_rate": 0.0005256274269507346,
+      "loss": 3.1067,
+      "step": 14021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5138943791389465,
+      "learning_rate": 0.0005256172796589564,
+      "loss": 3.2943,
+      "step": 14022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5650306940078735,
+      "learning_rate": 0.0005256071317729457,
+      "loss": 3.0597,
+      "step": 14023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5293331146240234,
+      "learning_rate": 0.0005255969832927293,
+      "loss": 3.2667,
+      "step": 14024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5199422836303711,
+      "learning_rate": 0.0005255868342183338,
+      "loss": 3.1008,
+      "step": 14025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53669273853302,
+      "learning_rate": 0.000525576684549786,
+      "loss": 3.1676,
+      "step": 14026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5880951285362244,
+      "learning_rate": 0.0005255665342871126,
+      "loss": 3.3137,
+      "step": 14027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5066351890563965,
+      "learning_rate": 0.0005255563834303404,
+      "loss": 3.1104,
+      "step": 14028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.495430052280426,
+      "learning_rate": 0.000525546231979496,
+      "loss": 3.0237,
+      "step": 14029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47963449358940125,
+      "learning_rate": 0.0005255360799346064,
+      "loss": 3.1957,
+      "step": 14030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48525768518447876,
+      "learning_rate": 0.0005255259272956981,
+      "loss": 3.1917,
+      "step": 14031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5137525796890259,
+      "learning_rate": 0.000525515774062798,
+      "loss": 2.9276,
+      "step": 14032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4915049076080322,
+      "learning_rate": 0.0005255056202359325,
+      "loss": 2.9671,
+      "step": 14033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5103715062141418,
+      "learning_rate": 0.0005254954658151288,
+      "loss": 3.0266,
+      "step": 14034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5029316544532776,
+      "learning_rate": 0.0005254853108004134,
+      "loss": 3.164,
+      "step": 14035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5544018149375916,
+      "learning_rate": 0.0005254751551918131,
+      "loss": 3.2264,
+      "step": 14036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5042638182640076,
+      "learning_rate": 0.0005254649989893547,
+      "loss": 2.8606,
+      "step": 14037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49873411655426025,
+      "learning_rate": 0.0005254548421930647,
+      "loss": 3.2909,
+      "step": 14038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5256275534629822,
+      "learning_rate": 0.0005254446848029701,
+      "loss": 3.1487,
+      "step": 14039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5733166337013245,
+      "learning_rate": 0.0005254345268190976,
+      "loss": 2.9514,
+      "step": 14040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4876699447631836,
+      "learning_rate": 0.000525424368241474,
+      "loss": 3.1855,
+      "step": 14041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5447889566421509,
+      "learning_rate": 0.000525414209070126,
+      "loss": 3.2617,
+      "step": 14042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48302027583122253,
+      "learning_rate": 0.0005254040493050802,
+      "loss": 3.1729,
+      "step": 14043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4758540093898773,
+      "learning_rate": 0.0005253938889463638,
+      "loss": 3.3485,
+      "step": 14044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.489118754863739,
+      "learning_rate": 0.000525383727994003,
+      "loss": 3.2733,
+      "step": 14045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5639674067497253,
+      "learning_rate": 0.000525373566448025,
+      "loss": 3.0852,
+      "step": 14046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5075635313987732,
+      "learning_rate": 0.0005253634043084562,
+      "loss": 3.3726,
+      "step": 14047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5024601817131042,
+      "learning_rate": 0.0005253532415753238,
+      "loss": 3.2864,
+      "step": 14048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5034321546554565,
+      "learning_rate": 0.0005253430782486542,
+      "loss": 3.3487,
+      "step": 14049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5592278242111206,
+      "learning_rate": 0.0005253329143284744,
+      "loss": 3.247,
+      "step": 14050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5167191028594971,
+      "learning_rate": 0.000525322749814811,
+      "loss": 3.0507,
+      "step": 14051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49541527032852173,
+      "learning_rate": 0.0005253125847076908,
+      "loss": 3.1837,
+      "step": 14052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5572511553764343,
+      "learning_rate": 0.0005253024190071407,
+      "loss": 3.1698,
+      "step": 14053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47061023116111755,
+      "learning_rate": 0.0005252922527131873,
+      "loss": 3.1254,
+      "step": 14054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5373715162277222,
+      "learning_rate": 0.0005252820858258576,
+      "loss": 3.251,
+      "step": 14055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5261591672897339,
+      "learning_rate": 0.000525271918345178,
+      "loss": 3.2149,
+      "step": 14056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.525307834148407,
+      "learning_rate": 0.0005252617502711758,
+      "loss": 3.1451,
+      "step": 14057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5310854315757751,
+      "learning_rate": 0.0005252515816038773,
+      "loss": 3.1049,
+      "step": 14058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5353348255157471,
+      "learning_rate": 0.0005252414123433095,
+      "loss": 3.3399,
+      "step": 14059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5103008151054382,
+      "learning_rate": 0.0005252312424894993,
+      "loss": 3.5181,
+      "step": 14060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5458347201347351,
+      "learning_rate": 0.0005252210720424733,
+      "loss": 3.3634,
+      "step": 14061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4898107647895813,
+      "learning_rate": 0.0005252109010022583,
+      "loss": 3.2373,
+      "step": 14062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5220617055892944,
+      "learning_rate": 0.0005252007293688812,
+      "loss": 3.1624,
+      "step": 14063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5146247148513794,
+      "learning_rate": 0.0005251905571423686,
+      "loss": 3.045,
+      "step": 14064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5100995302200317,
+      "learning_rate": 0.0005251803843227475,
+      "loss": 3.2578,
+      "step": 14065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.490809828042984,
+      "learning_rate": 0.0005251702109100447,
+      "loss": 3.16,
+      "step": 14066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5795995593070984,
+      "learning_rate": 0.0005251600369042868,
+      "loss": 3.0737,
+      "step": 14067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5179703831672668,
+      "learning_rate": 0.0005251498623055006,
+      "loss": 3.2451,
+      "step": 14068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.518638014793396,
+      "learning_rate": 0.0005251396871137131,
+      "loss": 3.3138,
+      "step": 14069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49439534544944763,
+      "learning_rate": 0.000525129511328951,
+      "loss": 3.1177,
+      "step": 14070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5000263452529907,
+      "learning_rate": 0.0005251193349512411,
+      "loss": 3.3391,
+      "step": 14071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.513812780380249,
+      "learning_rate": 0.0005251091579806101,
+      "loss": 3.2079,
+      "step": 14072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5052744746208191,
+      "learning_rate": 0.000525098980417085,
+      "loss": 2.9954,
+      "step": 14073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48377129435539246,
+      "learning_rate": 0.0005250888022606925,
+      "loss": 3.0341,
+      "step": 14074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5143635869026184,
+      "learning_rate": 0.0005250786235114594,
+      "loss": 3.0281,
+      "step": 14075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5064332485198975,
+      "learning_rate": 0.0005250684441694125,
+      "loss": 3.2199,
+      "step": 14076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5439924001693726,
+      "learning_rate": 0.0005250582642345786,
+      "loss": 3.0897,
+      "step": 14077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50617915391922,
+      "learning_rate": 0.0005250480837069846,
+      "loss": 3.278,
+      "step": 14078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48605844378471375,
+      "learning_rate": 0.0005250379025866572,
+      "loss": 3.1856,
+      "step": 14079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49857380986213684,
+      "learning_rate": 0.0005250277208736233,
+      "loss": 3.1719,
+      "step": 14080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6834864020347595,
+      "learning_rate": 0.0005250175385679098,
+      "loss": 3.2893,
+      "step": 14081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5134391188621521,
+      "learning_rate": 0.0005250073556695433,
+      "loss": 3.1539,
+      "step": 14082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5330235958099365,
+      "learning_rate": 0.0005249971721785506,
+      "loss": 3.1599,
+      "step": 14083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5358902812004089,
+      "learning_rate": 0.0005249869880949589,
+      "loss": 3.1398,
+      "step": 14084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5489894151687622,
+      "learning_rate": 0.0005249768034187947,
+      "loss": 3.195,
+      "step": 14085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.505240261554718,
+      "learning_rate": 0.0005249666181500848,
+      "loss": 3.2595,
+      "step": 14086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5279876589775085,
+      "learning_rate": 0.0005249564322888562,
+      "loss": 3.1531,
+      "step": 14087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5321412682533264,
+      "learning_rate": 0.0005249462458351358,
+      "loss": 3.1561,
+      "step": 14088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.506250262260437,
+      "learning_rate": 0.00052493605878895,
+      "loss": 3.286,
+      "step": 14089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.528944730758667,
+      "learning_rate": 0.0005249258711503261,
+      "loss": 2.914,
+      "step": 14090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5110275149345398,
+      "learning_rate": 0.0005249156829192906,
+      "loss": 3.1901,
+      "step": 14091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4941984713077545,
+      "learning_rate": 0.0005249054940958707,
+      "loss": 3.1032,
+      "step": 14092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5150231719017029,
+      "learning_rate": 0.0005248953046800928,
+      "loss": 3.1427,
+      "step": 14093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4935740530490875,
+      "learning_rate": 0.000524885114671984,
+      "loss": 3.1807,
+      "step": 14094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5428426265716553,
+      "learning_rate": 0.0005248749240715712,
+      "loss": 3.2272,
+      "step": 14095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4966670274734497,
+      "learning_rate": 0.000524864732878881,
+      "loss": 3.1482,
+      "step": 14096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5251358151435852,
+      "learning_rate": 0.0005248545410939404,
+      "loss": 3.2162,
+      "step": 14097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5258474946022034,
+      "learning_rate": 0.0005248443487167762,
+      "loss": 3.1545,
+      "step": 14098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5385447144508362,
+      "learning_rate": 0.0005248341557474152,
+      "loss": 3.1439,
+      "step": 14099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5289866328239441,
+      "learning_rate": 0.0005248239621858845,
+      "loss": 3.294,
+      "step": 14100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5030156373977661,
+      "learning_rate": 0.0005248137680322106,
+      "loss": 3.2187,
+      "step": 14101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48551294207572937,
+      "learning_rate": 0.0005248035732864205,
+      "loss": 3.4024,
+      "step": 14102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5139172673225403,
+      "learning_rate": 0.000524793377948541,
+      "loss": 3.0436,
+      "step": 14103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.529897928237915,
+      "learning_rate": 0.0005247831820185991,
+      "loss": 3.2084,
+      "step": 14104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5231662392616272,
+      "learning_rate": 0.0005247729854966215,
+      "loss": 3.0257,
+      "step": 14105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.569053053855896,
+      "learning_rate": 0.0005247627883826353,
+      "loss": 3.2904,
+      "step": 14106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49440523982048035,
+      "learning_rate": 0.0005247525906766669,
+      "loss": 3.3116,
+      "step": 14107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5238437056541443,
+      "learning_rate": 0.0005247423923787437,
+      "loss": 3.3411,
+      "step": 14108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5570396780967712,
+      "learning_rate": 0.0005247321934888921,
+      "loss": 3.2735,
+      "step": 14109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5721766352653503,
+      "learning_rate": 0.0005247219940071392,
+      "loss": 3.2418,
+      "step": 14110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5046465396881104,
+      "learning_rate": 0.0005247117939335119,
+      "loss": 3.1536,
+      "step": 14111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.522420346736908,
+      "learning_rate": 0.0005247015932680368,
+      "loss": 3.1374,
+      "step": 14112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5400798320770264,
+      "learning_rate": 0.0005246913920107411,
+      "loss": 3.3477,
+      "step": 14113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5389352440834045,
+      "learning_rate": 0.0005246811901616514,
+      "loss": 3.0954,
+      "step": 14114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6275335550308228,
+      "learning_rate": 0.0005246709877207948,
+      "loss": 3.3585,
+      "step": 14115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4925820827484131,
+      "learning_rate": 0.000524660784688198,
+      "loss": 3.2751,
+      "step": 14116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5348528027534485,
+      "learning_rate": 0.000524650581063888,
+      "loss": 3.2773,
+      "step": 14117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5362456440925598,
+      "learning_rate": 0.0005246403768478916,
+      "loss": 3.1904,
+      "step": 14118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5220507979393005,
+      "learning_rate": 0.0005246301720402356,
+      "loss": 3.2062,
+      "step": 14119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5358459949493408,
+      "learning_rate": 0.0005246199666409469,
+      "loss": 3.21,
+      "step": 14120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5036259293556213,
+      "learning_rate": 0.0005246097606500526,
+      "loss": 3.1651,
+      "step": 14121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5509597063064575,
+      "learning_rate": 0.0005245995540675793,
+      "loss": 3.1898,
+      "step": 14122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5109611749649048,
+      "learning_rate": 0.000524589346893554,
+      "loss": 3.3053,
+      "step": 14123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.563474714756012,
+      "learning_rate": 0.0005245791391280037,
+      "loss": 3.2851,
+      "step": 14124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49557414650917053,
+      "learning_rate": 0.000524568930770955,
+      "loss": 3.2507,
+      "step": 14125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4812037944793701,
+      "learning_rate": 0.0005245587218224351,
+      "loss": 3.3388,
+      "step": 14126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5291796326637268,
+      "learning_rate": 0.0005245485122824707,
+      "loss": 3.0031,
+      "step": 14127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5801912546157837,
+      "learning_rate": 0.0005245383021510887,
+      "loss": 3.3179,
+      "step": 14128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4930591881275177,
+      "learning_rate": 0.0005245280914283161,
+      "loss": 3.2406,
+      "step": 14129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5279679894447327,
+      "learning_rate": 0.0005245178801141796,
+      "loss": 3.3534,
+      "step": 14130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4926796555519104,
+      "learning_rate": 0.0005245076682087064,
+      "loss": 3.3052,
+      "step": 14131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6082884073257446,
+      "learning_rate": 0.000524497455711923,
+      "loss": 3.1572,
+      "step": 14132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4980970025062561,
+      "learning_rate": 0.0005244872426238567,
+      "loss": 3.0342,
+      "step": 14133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5105233192443848,
+      "learning_rate": 0.000524477028944534,
+      "loss": 3.1551,
+      "step": 14134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5064019560813904,
+      "learning_rate": 0.0005244668146739822,
+      "loss": 3.1884,
+      "step": 14135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5264427661895752,
+      "learning_rate": 0.0005244565998122278,
+      "loss": 3.3935,
+      "step": 14136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5206676721572876,
+      "learning_rate": 0.0005244463843592981,
+      "loss": 3.1003,
+      "step": 14137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5028414726257324,
+      "learning_rate": 0.0005244361683152197,
+      "loss": 3.3872,
+      "step": 14138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5728408694267273,
+      "learning_rate": 0.0005244259516800198,
+      "loss": 3.1309,
+      "step": 14139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234643816947937,
+      "learning_rate": 0.0005244157344537249,
+      "loss": 3.1336,
+      "step": 14140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5184975862503052,
+      "learning_rate": 0.0005244055166363623,
+      "loss": 3.2717,
+      "step": 14141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5001710057258606,
+      "learning_rate": 0.0005243952982279587,
+      "loss": 3.249,
+      "step": 14142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5176953673362732,
+      "learning_rate": 0.0005243850792285411,
+      "loss": 3.1428,
+      "step": 14143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5145552754402161,
+      "learning_rate": 0.0005243748596381364,
+      "loss": 3.0941,
+      "step": 14144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5094621777534485,
+      "learning_rate": 0.0005243646394567714,
+      "loss": 3.3651,
+      "step": 14145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4982603192329407,
+      "learning_rate": 0.0005243544186844733,
+      "loss": 3.4314,
+      "step": 14146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48291024565696716,
+      "learning_rate": 0.0005243441973212687,
+      "loss": 3.4426,
+      "step": 14147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5711972713470459,
+      "learning_rate": 0.0005243339753671847,
+      "loss": 3.273,
+      "step": 14148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5552250146865845,
+      "learning_rate": 0.0005243237528222482,
+      "loss": 3.2662,
+      "step": 14149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5337475538253784,
+      "learning_rate": 0.0005243135296864861,
+      "loss": 3.1347,
+      "step": 14150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4788720905780792,
+      "learning_rate": 0.0005243033059599253,
+      "loss": 3.1474,
+      "step": 14151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5156588554382324,
+      "learning_rate": 0.0005242930816425928,
+      "loss": 3.2473,
+      "step": 14152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49395278096199036,
+      "learning_rate": 0.0005242828567345154,
+      "loss": 3.2504,
+      "step": 14153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5037361979484558,
+      "learning_rate": 0.0005242726312357202,
+      "loss": 3.1234,
+      "step": 14154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5027718544006348,
+      "learning_rate": 0.0005242624051462341,
+      "loss": 3.1226,
+      "step": 14155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5126724243164062,
+      "learning_rate": 0.0005242521784660839,
+      "loss": 3.1279,
+      "step": 14156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5130273103713989,
+      "learning_rate": 0.0005242419511952966,
+      "loss": 3.2485,
+      "step": 14157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5258603692054749,
+      "learning_rate": 0.0005242317233338992,
+      "loss": 3.1893,
+      "step": 14158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.503662645816803,
+      "learning_rate": 0.0005242214948819186,
+      "loss": 3.2815,
+      "step": 14159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4841233789920807,
+      "learning_rate": 0.0005242112658393817,
+      "loss": 3.1095,
+      "step": 14160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5432723760604858,
+      "learning_rate": 0.0005242010362063156,
+      "loss": 3.4093,
+      "step": 14161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5304633378982544,
+      "learning_rate": 0.0005241908059827469,
+      "loss": 3.2416,
+      "step": 14162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5919955372810364,
+      "learning_rate": 0.0005241805751687029,
+      "loss": 3.2266,
+      "step": 14163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5040702819824219,
+      "learning_rate": 0.0005241703437642104,
+      "loss": 3.0357,
+      "step": 14164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5021560788154602,
+      "learning_rate": 0.0005241601117692964,
+      "loss": 3.1437,
+      "step": 14165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.513043224811554,
+      "learning_rate": 0.0005241498791839877,
+      "loss": 3.0822,
+      "step": 14166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4927295446395874,
+      "learning_rate": 0.0005241396460083114,
+      "loss": 3.3873,
+      "step": 14167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49543529748916626,
+      "learning_rate": 0.0005241294122422945,
+      "loss": 3.126,
+      "step": 14168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.528400719165802,
+      "learning_rate": 0.0005241191778859637,
+      "loss": 3.4217,
+      "step": 14169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5334087014198303,
+      "learning_rate": 0.0005241089429393462,
+      "loss": 3.2066,
+      "step": 14170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4597061574459076,
+      "learning_rate": 0.0005240987074024689,
+      "loss": 3.4921,
+      "step": 14171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5277274250984192,
+      "learning_rate": 0.0005240884712753588,
+      "loss": 3.4269,
+      "step": 14172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5084794759750366,
+      "learning_rate": 0.0005240782345580427,
+      "loss": 3.0579,
+      "step": 14173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5290378332138062,
+      "learning_rate": 0.0005240679972505476,
+      "loss": 3.2095,
+      "step": 14174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5595150589942932,
+      "learning_rate": 0.0005240577593529006,
+      "loss": 3.1068,
+      "step": 14175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4827249050140381,
+      "learning_rate": 0.0005240475208651286,
+      "loss": 3.3068,
+      "step": 14176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5036142468452454,
+      "learning_rate": 0.0005240372817872585,
+      "loss": 3.2103,
+      "step": 14177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5806594491004944,
+      "learning_rate": 0.0005240270421193173,
+      "loss": 3.1885,
+      "step": 14178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5014117360115051,
+      "learning_rate": 0.0005240168018613321,
+      "loss": 3.2471,
+      "step": 14179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5045899152755737,
+      "learning_rate": 0.0005240065610133296,
+      "loss": 3.3355,
+      "step": 14180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5164101719856262,
+      "learning_rate": 0.000523996319575337,
+      "loss": 3.111,
+      "step": 14181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.504895031452179,
+      "learning_rate": 0.0005239860775473811,
+      "loss": 3.2782,
+      "step": 14182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5253392457962036,
+      "learning_rate": 0.0005239758349294891,
+      "loss": 3.3066,
+      "step": 14183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5302526950836182,
+      "learning_rate": 0.0005239655917216879,
+      "loss": 3.176,
+      "step": 14184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5363248586654663,
+      "learning_rate": 0.0005239553479240044,
+      "loss": 3.1524,
+      "step": 14185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5327321290969849,
+      "learning_rate": 0.0005239451035364656,
+      "loss": 3.3454,
+      "step": 14186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5333127975463867,
+      "learning_rate": 0.0005239348585590985,
+      "loss": 3.0064,
+      "step": 14187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5116428732872009,
+      "learning_rate": 0.0005239246129919299,
+      "loss": 3.0169,
+      "step": 14188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5070028305053711,
+      "learning_rate": 0.0005239143668349872,
+      "loss": 3.157,
+      "step": 14189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5059760808944702,
+      "learning_rate": 0.0005239041200882971,
+      "loss": 3.4573,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5042681097984314,
+      "learning_rate": 0.0005238938727518867,
+      "loss": 3.2877,
+      "step": 14191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8668105602264404,
+      "learning_rate": 0.0005238836248257829,
+      "loss": 3.3942,
+      "step": 14192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5118007659912109,
+      "learning_rate": 0.0005238733763100126,
+      "loss": 3.0267,
+      "step": 14193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.527990460395813,
+      "learning_rate": 0.0005238631272046029,
+      "loss": 3.4337,
+      "step": 14194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5032305717468262,
+      "learning_rate": 0.0005238528775095808,
+      "loss": 3.2735,
+      "step": 14195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48045584559440613,
+      "learning_rate": 0.0005238426272249735,
+      "loss": 3.2052,
+      "step": 14196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5078832507133484,
+      "learning_rate": 0.0005238323763508077,
+      "loss": 2.9881,
+      "step": 14197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49898645281791687,
+      "learning_rate": 0.0005238221248871105,
+      "loss": 3.0997,
+      "step": 14198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5208724737167358,
+      "learning_rate": 0.0005238118728339089,
+      "loss": 3.1622,
+      "step": 14199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49348101019859314,
+      "learning_rate": 0.0005238016201912298,
+      "loss": 3.2991,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5580134391784668,
+      "learning_rate": 0.0005237913669591004,
+      "loss": 3.2781,
+      "step": 14201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5383898615837097,
+      "learning_rate": 0.0005237811131375476,
+      "loss": 3.2633,
+      "step": 14202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5041082501411438,
+      "learning_rate": 0.0005237708587265984,
+      "loss": 3.2228,
+      "step": 14203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5141589641571045,
+      "learning_rate": 0.0005237606037262799,
+      "loss": 3.0724,
+      "step": 14204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.509434700012207,
+      "learning_rate": 0.0005237503481366189,
+      "loss": 3.1054,
+      "step": 14205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5451233386993408,
+      "learning_rate": 0.0005237400919576426,
+      "loss": 3.0023,
+      "step": 14206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5041611790657043,
+      "learning_rate": 0.000523729835189378,
+      "loss": 3.2864,
+      "step": 14207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48671042919158936,
+      "learning_rate": 0.0005237195778318521,
+      "loss": 3.2833,
+      "step": 14208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5079613924026489,
+      "learning_rate": 0.0005237093198850917,
+      "loss": 3.0954,
+      "step": 14209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5360010862350464,
+      "learning_rate": 0.0005236990613491242,
+      "loss": 3.2204,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5029088258743286,
+      "learning_rate": 0.0005236888022239763,
+      "loss": 3.1848,
+      "step": 14211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5785709023475647,
+      "learning_rate": 0.0005236785425096753,
+      "loss": 3.1963,
+      "step": 14212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5002698302268982,
+      "learning_rate": 0.0005236682822062479,
+      "loss": 3.2937,
+      "step": 14213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46872082352638245,
+      "learning_rate": 0.0005236580213137214,
+      "loss": 3.2813,
+      "step": 14214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5846667289733887,
+      "learning_rate": 0.0005236477598321227,
+      "loss": 3.2123,
+      "step": 14215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5419370532035828,
+      "learning_rate": 0.0005236374977614788,
+      "loss": 3.1255,
+      "step": 14216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5296844840049744,
+      "learning_rate": 0.0005236272351018167,
+      "loss": 3.2611,
+      "step": 14217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5379822254180908,
+      "learning_rate": 0.0005236169718531637,
+      "loss": 3.15,
+      "step": 14218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5059786438941956,
+      "learning_rate": 0.0005236067080155465,
+      "loss": 3.3715,
+      "step": 14219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48545315861701965,
+      "learning_rate": 0.0005235964435889923,
+      "loss": 3.2246,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5349769592285156,
+      "learning_rate": 0.0005235861785735281,
+      "loss": 3.1605,
+      "step": 14221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5449730157852173,
+      "learning_rate": 0.0005235759129691809,
+      "loss": 3.2828,
+      "step": 14222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5503379106521606,
+      "learning_rate": 0.0005235656467759778,
+      "loss": 3.2192,
+      "step": 14223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5100115537643433,
+      "learning_rate": 0.0005235553799939458,
+      "loss": 3.2576,
+      "step": 14224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5590293407440186,
+      "learning_rate": 0.0005235451126231119,
+      "loss": 3.2798,
+      "step": 14225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5304601788520813,
+      "learning_rate": 0.0005235348446635034,
+      "loss": 3.0178,
+      "step": 14226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5042637586593628,
+      "learning_rate": 0.0005235245761151468,
+      "loss": 3.1506,
+      "step": 14227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5219356417655945,
+      "learning_rate": 0.0005235143069780698,
+      "loss": 3.2972,
+      "step": 14228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5481365919113159,
+      "learning_rate": 0.000523504037252299,
+      "loss": 3.2286,
+      "step": 14229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48297563195228577,
+      "learning_rate": 0.0005234937669378615,
+      "loss": 3.3976,
+      "step": 14230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5044030547142029,
+      "learning_rate": 0.0005234834960347846,
+      "loss": 3.2173,
+      "step": 14231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5015493035316467,
+      "learning_rate": 0.0005234732245430952,
+      "loss": 3.3385,
+      "step": 14232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5298082232475281,
+      "learning_rate": 0.0005234629524628201,
+      "loss": 3.3677,
+      "step": 14233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5369641780853271,
+      "learning_rate": 0.0005234526797939868,
+      "loss": 3.42,
+      "step": 14234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6291206479072571,
+      "learning_rate": 0.000523442406536622,
+      "loss": 3.1479,
+      "step": 14235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4945876896381378,
+      "learning_rate": 0.000523432132690753,
+      "loss": 3.1374,
+      "step": 14236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.519692063331604,
+      "learning_rate": 0.0005234218582564067,
+      "loss": 3.0743,
+      "step": 14237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5264421701431274,
+      "learning_rate": 0.0005234115832336103,
+      "loss": 3.3572,
+      "step": 14238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49894416332244873,
+      "learning_rate": 0.0005234013076223907,
+      "loss": 3.1509,
+      "step": 14239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.519243597984314,
+      "learning_rate": 0.0005233910314227751,
+      "loss": 3.2018,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5167336463928223,
+      "learning_rate": 0.0005233807546347906,
+      "loss": 3.3301,
+      "step": 14241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5419256687164307,
+      "learning_rate": 0.000523370477258464,
+      "loss": 3.3605,
+      "step": 14242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5188645124435425,
+      "learning_rate": 0.0005233601992938228,
+      "loss": 3.3297,
+      "step": 14243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5342188477516174,
+      "learning_rate": 0.0005233499207408936,
+      "loss": 3.2517,
+      "step": 14244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5165455341339111,
+      "learning_rate": 0.0005233396415997038,
+      "loss": 3.2128,
+      "step": 14245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5181511640548706,
+      "learning_rate": 0.0005233293618702804,
+      "loss": 3.0307,
+      "step": 14246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5308560729026794,
+      "learning_rate": 0.0005233190815526502,
+      "loss": 3.439,
+      "step": 14247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4866609573364258,
+      "learning_rate": 0.0005233088006468407,
+      "loss": 3.0744,
+      "step": 14248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5222994089126587,
+      "learning_rate": 0.0005232985191528788,
+      "loss": 3.1265,
+      "step": 14249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.495818555355072,
+      "learning_rate": 0.0005232882370707916,
+      "loss": 3.4966,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5098391175270081,
+      "learning_rate": 0.0005232779544006061,
+      "loss": 3.2485,
+      "step": 14251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5358620882034302,
+      "learning_rate": 0.0005232676711423495,
+      "loss": 3.2363,
+      "step": 14252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5330485105514526,
+      "learning_rate": 0.0005232573872960488,
+      "loss": 3.2469,
+      "step": 14253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5173335075378418,
+      "learning_rate": 0.0005232471028617311,
+      "loss": 3.0878,
+      "step": 14254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.510068953037262,
+      "learning_rate": 0.0005232368178394234,
+      "loss": 3.1082,
+      "step": 14255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.567221462726593,
+      "learning_rate": 0.000523226532229153,
+      "loss": 3.1399,
+      "step": 14256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5452208518981934,
+      "learning_rate": 0.0005232162460309468,
+      "loss": 3.119,
+      "step": 14257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5163204669952393,
+      "learning_rate": 0.000523205959244832,
+      "loss": 3.2187,
+      "step": 14258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4994533061981201,
+      "learning_rate": 0.0005231956718708357,
+      "loss": 3.1825,
+      "step": 14259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5001878142356873,
+      "learning_rate": 0.0005231853839089849,
+      "loss": 3.2424,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5246791243553162,
+      "learning_rate": 0.0005231750953593068,
+      "loss": 3.2498,
+      "step": 14261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4972473084926605,
+      "learning_rate": 0.0005231648062218285,
+      "loss": 3.193,
+      "step": 14262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5774008631706238,
+      "learning_rate": 0.0005231545164965769,
+      "loss": 3.0555,
+      "step": 14263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5512023568153381,
+      "learning_rate": 0.0005231442261835794,
+      "loss": 3.2012,
+      "step": 14264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5204340815544128,
+      "learning_rate": 0.0005231339352828628,
+      "loss": 3.0874,
+      "step": 14265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5237892866134644,
+      "learning_rate": 0.0005231236437944545,
+      "loss": 3.1603,
+      "step": 14266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5312855839729309,
+      "learning_rate": 0.0005231133517183814,
+      "loss": 3.1177,
+      "step": 14267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48745495080947876,
+      "learning_rate": 0.0005231030590546707,
+      "loss": 3.268,
+      "step": 14268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5135250091552734,
+      "learning_rate": 0.0005230927658033494,
+      "loss": 3.395,
+      "step": 14269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5581345558166504,
+      "learning_rate": 0.0005230824719644448,
+      "loss": 3.1017,
+      "step": 14270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.47780802845954895,
+      "learning_rate": 0.0005230721775379837,
+      "loss": 3.1137,
+      "step": 14271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49343234300613403,
+      "learning_rate": 0.0005230618825239937,
+      "loss": 3.2004,
+      "step": 14272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5608261823654175,
+      "learning_rate": 0.0005230515869225013,
+      "loss": 3.1291,
+      "step": 14273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.510046124458313,
+      "learning_rate": 0.0005230412907335343,
+      "loss": 3.3241,
+      "step": 14274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.506389856338501,
+      "learning_rate": 0.0005230309939571193,
+      "loss": 3.1,
+      "step": 14275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4831939935684204,
+      "learning_rate": 0.0005230206965932836,
+      "loss": 3.2499,
+      "step": 14276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.506610095500946,
+      "learning_rate": 0.0005230103986420542,
+      "loss": 3.371,
+      "step": 14277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49668189883232117,
+      "learning_rate": 0.0005230001001034585,
+      "loss": 3.105,
+      "step": 14278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5084254145622253,
+      "learning_rate": 0.0005229898009775233,
+      "loss": 3.1564,
+      "step": 14279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5354201793670654,
+      "learning_rate": 0.000522979501264276,
+      "loss": 3.2659,
+      "step": 14280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5142485499382019,
+      "learning_rate": 0.0005229692009637437,
+      "loss": 3.0802,
+      "step": 14281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5173102617263794,
+      "learning_rate": 0.0005229589000759531,
+      "loss": 3.0241,
+      "step": 14282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5071712732315063,
+      "learning_rate": 0.0005229485986009319,
+      "loss": 3.2652,
+      "step": 14283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4926835596561432,
+      "learning_rate": 0.0005229382965387069,
+      "loss": 3.1913,
+      "step": 14284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.503774106502533,
+      "learning_rate": 0.0005229279938893055,
+      "loss": 3.2452,
+      "step": 14285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4997362196445465,
+      "learning_rate": 0.0005229176906527545,
+      "loss": 3.3595,
+      "step": 14286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5188296437263489,
+      "learning_rate": 0.0005229073868290813,
+      "loss": 3.1387,
+      "step": 14287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5256169438362122,
+      "learning_rate": 0.000522897082418313,
+      "loss": 3.1277,
+      "step": 14288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5132431387901306,
+      "learning_rate": 0.0005228867774204765,
+      "loss": 3.1984,
+      "step": 14289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5352613925933838,
+      "learning_rate": 0.0005228764718355993,
+      "loss": 3.301,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5036006569862366,
+      "learning_rate": 0.0005228661656637082,
+      "loss": 3.2298,
+      "step": 14291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4817676544189453,
+      "learning_rate": 0.0005228558589048306,
+      "loss": 3.3918,
+      "step": 14292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.528864324092865,
+      "learning_rate": 0.0005228455515589935,
+      "loss": 3.0981,
+      "step": 14293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5127184391021729,
+      "learning_rate": 0.0005228352436262243,
+      "loss": 3.3858,
+      "step": 14294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5853077173233032,
+      "learning_rate": 0.0005228249351065498,
+      "loss": 3.2205,
+      "step": 14295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5480331182479858,
+      "learning_rate": 0.0005228146259999972,
+      "loss": 3.0766,
+      "step": 14296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5229378342628479,
+      "learning_rate": 0.000522804316306594,
+      "loss": 3.0965,
+      "step": 14297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5020848512649536,
+      "learning_rate": 0.0005227940060263669,
+      "loss": 3.3201,
+      "step": 14298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5315490365028381,
+      "learning_rate": 0.0005227836951593434,
+      "loss": 3.2358,
+      "step": 14299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5193626880645752,
+      "learning_rate": 0.0005227733837055504,
+      "loss": 3.0192,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5127216577529907,
+      "learning_rate": 0.0005227630716650152,
+      "loss": 3.179,
+      "step": 14301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5541197061538696,
+      "learning_rate": 0.0005227527590377651,
+      "loss": 3.1317,
+      "step": 14302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5353065133094788,
+      "learning_rate": 0.0005227424458238269,
+      "loss": 3.1438,
+      "step": 14303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4878837764263153,
+      "learning_rate": 0.0005227321320232281,
+      "loss": 3.3127,
+      "step": 14304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5021807551383972,
+      "learning_rate": 0.0005227218176359957,
+      "loss": 3.2301,
+      "step": 14305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4997186064720154,
+      "learning_rate": 0.0005227115026621568,
+      "loss": 3.1486,
+      "step": 14306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5215864181518555,
+      "learning_rate": 0.0005227011871017388,
+      "loss": 3.1568,
+      "step": 14307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5061655044555664,
+      "learning_rate": 0.0005226908709547687,
+      "loss": 3.1592,
+      "step": 14308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5245349407196045,
+      "learning_rate": 0.0005226805542212737,
+      "loss": 3.3514,
+      "step": 14309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4831959307193756,
+      "learning_rate": 0.000522670236901281,
+      "loss": 3.2534,
+      "step": 14310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.47621917724609375,
+      "learning_rate": 0.0005226599189948176,
+      "loss": 3.0129,
+      "step": 14311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5170857310295105,
+      "learning_rate": 0.000522649600501911,
+      "loss": 3.073,
+      "step": 14312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5005858540534973,
+      "learning_rate": 0.0005226392814225881,
+      "loss": 3.114,
+      "step": 14313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5574570894241333,
+      "learning_rate": 0.0005226289617568763,
+      "loss": 3.3234,
+      "step": 14314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5117168426513672,
+      "learning_rate": 0.0005226186415048026,
+      "loss": 3.1659,
+      "step": 14315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49425193667411804,
+      "learning_rate": 0.0005226083206663941,
+      "loss": 2.9738,
+      "step": 14316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5152395963668823,
+      "learning_rate": 0.0005225979992416783,
+      "loss": 3.2,
+      "step": 14317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5114201903343201,
+      "learning_rate": 0.0005225876772306822,
+      "loss": 3.2103,
+      "step": 14318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48314422369003296,
+      "learning_rate": 0.0005225773546334328,
+      "loss": 3.2442,
+      "step": 14319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234989523887634,
+      "learning_rate": 0.0005225670314499577,
+      "loss": 3.1864,
+      "step": 14320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5077502727508545,
+      "learning_rate": 0.0005225567076802838,
+      "loss": 3.2501,
+      "step": 14321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4953576624393463,
+      "learning_rate": 0.0005225463833244384,
+      "loss": 3.2259,
+      "step": 14322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5094617605209351,
+      "learning_rate": 0.0005225360583824487,
+      "loss": 3.1856,
+      "step": 14323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5310658812522888,
+      "learning_rate": 0.0005225257328543417,
+      "loss": 3.2715,
+      "step": 14324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5143598318099976,
+      "learning_rate": 0.0005225154067401448,
+      "loss": 3.2914,
+      "step": 14325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5416876673698425,
+      "learning_rate": 0.0005225050800398851,
+      "loss": 3.3223,
+      "step": 14326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5177731513977051,
+      "learning_rate": 0.00052249475275359,
+      "loss": 3.1994,
+      "step": 14327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4947225749492645,
+      "learning_rate": 0.0005224844248812864,
+      "loss": 3.2301,
+      "step": 14328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5090954303741455,
+      "learning_rate": 0.0005224740964230017,
+      "loss": 3.1208,
+      "step": 14329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5198858976364136,
+      "learning_rate": 0.0005224637673787631,
+      "loss": 3.2342,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5442454814910889,
+      "learning_rate": 0.0005224534377485977,
+      "loss": 3.0242,
+      "step": 14331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6403926610946655,
+      "learning_rate": 0.0005224431075325327,
+      "loss": 3.0215,
+      "step": 14332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49326714873313904,
+      "learning_rate": 0.0005224327767305954,
+      "loss": 3.4625,
+      "step": 14333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.527711808681488,
+      "learning_rate": 0.000522422445342813,
+      "loss": 3.3671,
+      "step": 14334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5429787039756775,
+      "learning_rate": 0.0005224121133692127,
+      "loss": 3.3924,
+      "step": 14335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5077711343765259,
+      "learning_rate": 0.0005224017808098217,
+      "loss": 3.2152,
+      "step": 14336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48497462272644043,
+      "learning_rate": 0.0005223914476646671,
+      "loss": 3.1022,
+      "step": 14337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5042962431907654,
+      "learning_rate": 0.0005223811139337763,
+      "loss": 3.3529,
+      "step": 14338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5740790963172913,
+      "learning_rate": 0.0005223707796171765,
+      "loss": 3.0065,
+      "step": 14339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5047106742858887,
+      "learning_rate": 0.0005223604447148947,
+      "loss": 3.2321,
+      "step": 14340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5074807405471802,
+      "learning_rate": 0.0005223501092269584,
+      "loss": 3.2166,
+      "step": 14341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49849480390548706,
+      "learning_rate": 0.0005223397731533947,
+      "loss": 2.8486,
+      "step": 14342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4887546896934509,
+      "learning_rate": 0.0005223294364942309,
+      "loss": 3.2793,
+      "step": 14343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5052506923675537,
+      "learning_rate": 0.0005223190992494941,
+      "loss": 3.2666,
+      "step": 14344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.530553936958313,
+      "learning_rate": 0.0005223087614192116,
+      "loss": 3.3803,
+      "step": 14345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5107871294021606,
+      "learning_rate": 0.0005222984230034105,
+      "loss": 3.2682,
+      "step": 14346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5242406725883484,
+      "learning_rate": 0.0005222880840021183,
+      "loss": 3.1691,
+      "step": 14347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5011206269264221,
+      "learning_rate": 0.0005222777444153618,
+      "loss": 3.5911,
+      "step": 14348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5178399085998535,
+      "learning_rate": 0.0005222674042431688,
+      "loss": 3.2566,
+      "step": 14349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5099117159843445,
+      "learning_rate": 0.000522257063485566,
+      "loss": 3.159,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5087069869041443,
+      "learning_rate": 0.000522246722142581,
+      "loss": 3.0981,
+      "step": 14351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49150705337524414,
+      "learning_rate": 0.0005222363802142409,
+      "loss": 3.2154,
+      "step": 14352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5360817313194275,
+      "learning_rate": 0.0005222260377005729,
+      "loss": 3.0334,
+      "step": 14353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7099424600601196,
+      "learning_rate": 0.0005222156946016043,
+      "loss": 3.1963,
+      "step": 14354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5025527477264404,
+      "learning_rate": 0.0005222053509173623,
+      "loss": 3.0433,
+      "step": 14355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.543745219707489,
+      "learning_rate": 0.0005221950066478742,
+      "loss": 3.0136,
+      "step": 14356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5021082758903503,
+      "learning_rate": 0.0005221846617931672,
+      "loss": 2.9967,
+      "step": 14357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5673491358757019,
+      "learning_rate": 0.0005221743163532686,
+      "loss": 3.2684,
+      "step": 14358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5165929198265076,
+      "learning_rate": 0.0005221639703282057,
+      "loss": 3.236,
+      "step": 14359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5241513848304749,
+      "learning_rate": 0.0005221536237180054,
+      "loss": 3.178,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48851609230041504,
+      "learning_rate": 0.0005221432765226955,
+      "loss": 3.2317,
+      "step": 14361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5319517254829407,
+      "learning_rate": 0.0005221329287423027,
+      "loss": 3.3219,
+      "step": 14362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5038327574729919,
+      "learning_rate": 0.0005221225803768546,
+      "loss": 3.2557,
+      "step": 14363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5223231911659241,
+      "learning_rate": 0.0005221122314263785,
+      "loss": 3.1857,
+      "step": 14364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4955352544784546,
+      "learning_rate": 0.0005221018818909014,
+      "loss": 3.1251,
+      "step": 14365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5417662262916565,
+      "learning_rate": 0.0005220915317704507,
+      "loss": 3.0374,
+      "step": 14366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4988352060317993,
+      "learning_rate": 0.0005220811810650537,
+      "loss": 3.1441,
+      "step": 14367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4997967779636383,
+      "learning_rate": 0.0005220708297747375,
+      "loss": 3.267,
+      "step": 14368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5113928914070129,
+      "learning_rate": 0.0005220604778995296,
+      "loss": 3.1211,
+      "step": 14369
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5090401768684387,
+      "learning_rate": 0.0005220501254394571,
+      "loss": 3.2306,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390316843986511,
+      "learning_rate": 0.0005220397723945473,
+      "loss": 3.3519,
+      "step": 14371
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5376928448677063,
+      "learning_rate": 0.0005220294187648275,
+      "loss": 3.0629,
+      "step": 14372
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5060685276985168,
+      "learning_rate": 0.0005220190645503248,
+      "loss": 3.2316,
+      "step": 14373
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5190165638923645,
+      "learning_rate": 0.0005220087097510668,
+      "loss": 3.1309,
+      "step": 14374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5150943398475647,
+      "learning_rate": 0.0005219983543670805,
+      "loss": 3.2837,
+      "step": 14375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4947223663330078,
+      "learning_rate": 0.0005219879983983933,
+      "loss": 3.376,
+      "step": 14376
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48892343044281006,
+      "learning_rate": 0.0005219776418450323,
+      "loss": 3.24,
+      "step": 14377
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5334641337394714,
+      "learning_rate": 0.0005219672847070251,
+      "loss": 3.2448,
+      "step": 14378
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5088352560997009,
+      "learning_rate": 0.0005219569269843987,
+      "loss": 3.1581,
+      "step": 14379
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49996280670166016,
+      "learning_rate": 0.0005219465686771805,
+      "loss": 3.2747,
+      "step": 14380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5581712126731873,
+      "learning_rate": 0.0005219362097853977,
+      "loss": 3.135,
+      "step": 14381
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5300365686416626,
+      "learning_rate": 0.0005219258503090776,
+      "loss": 3.2328,
+      "step": 14382
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4860009551048279,
+      "learning_rate": 0.0005219154902482476,
+      "loss": 3.2522,
+      "step": 14383
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5062441825866699,
+      "learning_rate": 0.0005219051296029349,
+      "loss": 3.3731,
+      "step": 14384
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5869806408882141,
+      "learning_rate": 0.0005218947683731667,
+      "loss": 3.0537,
+      "step": 14385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5176935791969299,
+      "learning_rate": 0.0005218844065589704,
+      "loss": 2.9489,
+      "step": 14386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5050328373908997,
+      "learning_rate": 0.0005218740441603735,
+      "loss": 3.3333,
+      "step": 14387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5312533378601074,
+      "learning_rate": 0.0005218636811774029,
+      "loss": 3.2162,
+      "step": 14388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5098238587379456,
+      "learning_rate": 0.000521853317610086,
+      "loss": 3.1103,
+      "step": 14389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48738306760787964,
+      "learning_rate": 0.0005218429534584502,
+      "loss": 3.2526,
+      "step": 14390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5178094506263733,
+      "learning_rate": 0.0005218325887225228,
+      "loss": 3.2081,
+      "step": 14391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5126038789749146,
+      "learning_rate": 0.000521822223402331,
+      "loss": 3.1075,
+      "step": 14392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49259766936302185,
+      "learning_rate": 0.0005218118574979023,
+      "loss": 3.2309,
+      "step": 14393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5296828746795654,
+      "learning_rate": 0.0005218014910092636,
+      "loss": 3.294,
+      "step": 14394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5134131908416748,
+      "learning_rate": 0.0005217911239364427,
+      "loss": 3.2609,
+      "step": 14395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5110315680503845,
+      "learning_rate": 0.0005217807562794666,
+      "loss": 3.2463,
+      "step": 14396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5391361713409424,
+      "learning_rate": 0.0005217703880383626,
+      "loss": 2.8263,
+      "step": 14397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5444517135620117,
+      "learning_rate": 0.0005217600192131582,
+      "loss": 3.1325,
+      "step": 14398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5090447664260864,
+      "learning_rate": 0.0005217496498038805,
+      "loss": 2.9955,
+      "step": 14399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5326402187347412,
+      "learning_rate": 0.0005217392798105569,
+      "loss": 2.8706,
+      "step": 14400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.586733877658844,
+      "learning_rate": 0.0005217289092332147,
+      "loss": 3.1483,
+      "step": 14401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.534562349319458,
+      "learning_rate": 0.0005217185380718812,
+      "loss": 3.0548,
+      "step": 14402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49788811802864075,
+      "learning_rate": 0.0005217081663265838,
+      "loss": 3.1239,
+      "step": 14403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5514092445373535,
+      "learning_rate": 0.0005216977939973498,
+      "loss": 3.2406,
+      "step": 14404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5975626707077026,
+      "learning_rate": 0.0005216874210842064,
+      "loss": 3.0378,
+      "step": 14405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5058104395866394,
+      "learning_rate": 0.0005216770475871811,
+      "loss": 3.1995,
+      "step": 14406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5320433974266052,
+      "learning_rate": 0.0005216666735063009,
+      "loss": 2.9181,
+      "step": 14407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5036849975585938,
+      "learning_rate": 0.0005216562988415935,
+      "loss": 3.3894,
+      "step": 14408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5307196378707886,
+      "learning_rate": 0.0005216459235930861,
+      "loss": 3.0559,
+      "step": 14409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5142438411712646,
+      "learning_rate": 0.0005216355477608059,
+      "loss": 3.0828,
+      "step": 14410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5658327341079712,
+      "learning_rate": 0.0005216251713447804,
+      "loss": 3.2546,
+      "step": 14411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5084229111671448,
+      "learning_rate": 0.0005216147943450368,
+      "loss": 3.3371,
+      "step": 14412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5430961847305298,
+      "learning_rate": 0.0005216044167616025,
+      "loss": 3.2217,
+      "step": 14413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5450170636177063,
+      "learning_rate": 0.0005215940385945048,
+      "loss": 3.1991,
+      "step": 14414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49537381529808044,
+      "learning_rate": 0.000521583659843771,
+      "loss": 3.4065,
+      "step": 14415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5271323919296265,
+      "learning_rate": 0.0005215732805094286,
+      "loss": 3.2777,
+      "step": 14416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5401394367218018,
+      "learning_rate": 0.0005215629005915047,
+      "loss": 3.1833,
+      "step": 14417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234045386314392,
+      "learning_rate": 0.0005215525200900269,
+      "loss": 3.0229,
+      "step": 14418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5828486680984497,
+      "learning_rate": 0.0005215421390050224,
+      "loss": 3.1473,
+      "step": 14419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5277825593948364,
+      "learning_rate": 0.0005215317573365185,
+      "loss": 3.0845,
+      "step": 14420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5510663390159607,
+      "learning_rate": 0.0005215213750845425,
+      "loss": 3.2722,
+      "step": 14421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5259878039360046,
+      "learning_rate": 0.0005215109922491218,
+      "loss": 3.0543,
+      "step": 14422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4986727237701416,
+      "learning_rate": 0.000521500608830284,
+      "loss": 3.2939,
+      "step": 14423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.558155357837677,
+      "learning_rate": 0.000521490224828056,
+      "loss": 3.0415,
+      "step": 14424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5337893962860107,
+      "learning_rate": 0.0005214798402424655,
+      "loss": 2.949,
+      "step": 14425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5568438172340393,
+      "learning_rate": 0.0005214694550735396,
+      "loss": 3.4036,
+      "step": 14426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52046138048172,
+      "learning_rate": 0.000521459069321306,
+      "loss": 3.054,
+      "step": 14427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4952624440193176,
+      "learning_rate": 0.0005214486829857916,
+      "loss": 3.2659,
+      "step": 14428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.584016740322113,
+      "learning_rate": 0.0005214382960670241,
+      "loss": 3.0175,
+      "step": 14429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5094707608222961,
+      "learning_rate": 0.0005214279085650308,
+      "loss": 3.0569,
+      "step": 14430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5051601529121399,
+      "learning_rate": 0.0005214175204798388,
+      "loss": 3.0209,
+      "step": 14431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4987492263317108,
+      "learning_rate": 0.0005214071318114759,
+      "loss": 3.2559,
+      "step": 14432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4821774661540985,
+      "learning_rate": 0.0005213967425599692,
+      "loss": 3.23,
+      "step": 14433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5031430125236511,
+      "learning_rate": 0.0005213863527253459,
+      "loss": 3.2185,
+      "step": 14434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52392578125,
+      "learning_rate": 0.0005213759623076337,
+      "loss": 3.1812,
+      "step": 14435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4971693754196167,
+      "learning_rate": 0.0005213655713068598,
+      "loss": 3.0771,
+      "step": 14436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5507715344429016,
+      "learning_rate": 0.0005213551797230516,
+      "loss": 3.0654,
+      "step": 14437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5817131996154785,
+      "learning_rate": 0.0005213447875562365,
+      "loss": 3.1453,
+      "step": 14438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5093923211097717,
+      "learning_rate": 0.0005213343948064417,
+      "loss": 2.9989,
+      "step": 14439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5584835410118103,
+      "learning_rate": 0.0005213240014736947,
+      "loss": 3.1431,
+      "step": 14440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.534837007522583,
+      "learning_rate": 0.000521313607558023,
+      "loss": 3.1652,
+      "step": 14441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.532855749130249,
+      "learning_rate": 0.0005213032130594537,
+      "loss": 3.1471,
+      "step": 14442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5114381909370422,
+      "learning_rate": 0.0005212928179780144,
+      "loss": 3.1387,
+      "step": 14443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5551025867462158,
+      "learning_rate": 0.0005212824223137325,
+      "loss": 3.2062,
+      "step": 14444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5081254839897156,
+      "learning_rate": 0.0005212720260666352,
+      "loss": 3.4413,
+      "step": 14445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5733745098114014,
+      "learning_rate": 0.00052126162923675,
+      "loss": 3.2314,
+      "step": 14446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5209436416625977,
+      "learning_rate": 0.0005212512318241042,
+      "loss": 3.4682,
+      "step": 14447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5001239776611328,
+      "learning_rate": 0.0005212408338287252,
+      "loss": 3.1519,
+      "step": 14448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5506751537322998,
+      "learning_rate": 0.0005212304352506405,
+      "loss": 2.9797,
+      "step": 14449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5498661398887634,
+      "learning_rate": 0.0005212200360898775,
+      "loss": 3.0644,
+      "step": 14450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5015347599983215,
+      "learning_rate": 0.0005212096363464633,
+      "loss": 3.1071,
+      "step": 14451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5187351107597351,
+      "learning_rate": 0.0005211992360204256,
+      "loss": 3.1041,
+      "step": 14452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5389336943626404,
+      "learning_rate": 0.0005211888351117917,
+      "loss": 3.1903,
+      "step": 14453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5327998995780945,
+      "learning_rate": 0.0005211784336205889,
+      "loss": 3.3395,
+      "step": 14454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5216488838195801,
+      "learning_rate": 0.0005211680315468448,
+      "loss": 3.2868,
+      "step": 14455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5144593715667725,
+      "learning_rate": 0.0005211576288905865,
+      "loss": 3.35,
+      "step": 14456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5322148203849792,
+      "learning_rate": 0.0005211472256518416,
+      "loss": 3.0846,
+      "step": 14457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4917519688606262,
+      "learning_rate": 0.0005211368218306377,
+      "loss": 3.1299,
+      "step": 14458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5577586889266968,
+      "learning_rate": 0.0005211264174270016,
+      "loss": 3.156,
+      "step": 14459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5201316475868225,
+      "learning_rate": 0.0005211160124409613,
+      "loss": 3.2895,
+      "step": 14460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49748873710632324,
+      "learning_rate": 0.000521105606872544,
+      "loss": 3.2544,
+      "step": 14461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5247087478637695,
+      "learning_rate": 0.000521095200721777,
+      "loss": 3.1305,
+      "step": 14462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5503666996955872,
+      "learning_rate": 0.0005210847939886878,
+      "loss": 3.2872,
+      "step": 14463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5127733945846558,
+      "learning_rate": 0.0005210743866733039,
+      "loss": 3.3832,
+      "step": 14464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5337607860565186,
+      "learning_rate": 0.0005210639787756525,
+      "loss": 3.2879,
+      "step": 14465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4939156770706177,
+      "learning_rate": 0.0005210535702957612,
+      "loss": 3.2596,
+      "step": 14466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5159709453582764,
+      "learning_rate": 0.0005210431612336573,
+      "loss": 3.3191,
+      "step": 14467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5352545976638794,
+      "learning_rate": 0.0005210327515893683,
+      "loss": 3.1018,
+      "step": 14468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5087549686431885,
+      "learning_rate": 0.0005210223413629214,
+      "loss": 3.0488,
+      "step": 14469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5398578643798828,
+      "learning_rate": 0.0005210119305543443,
+      "loss": 3.1926,
+      "step": 14470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9343258738517761,
+      "learning_rate": 0.0005210015191636643,
+      "loss": 3.0479,
+      "step": 14471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5741486549377441,
+      "learning_rate": 0.0005209911071909089,
+      "loss": 3.1191,
+      "step": 14472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5154914855957031,
+      "learning_rate": 0.0005209806946361055,
+      "loss": 3.2,
+      "step": 14473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5743831992149353,
+      "learning_rate": 0.0005209702814992813,
+      "loss": 3.374,
+      "step": 14474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5344333648681641,
+      "learning_rate": 0.000520959867780464,
+      "loss": 3.467,
+      "step": 14475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5519991517066956,
+      "learning_rate": 0.0005209494534796809,
+      "loss": 2.9471,
+      "step": 14476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5173640251159668,
+      "learning_rate": 0.0005209390385969595,
+      "loss": 3.24,
+      "step": 14477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5470741987228394,
+      "learning_rate": 0.0005209286231323271,
+      "loss": 3.1831,
+      "step": 14478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6194097399711609,
+      "learning_rate": 0.0005209182070858114,
+      "loss": 3.1974,
+      "step": 14479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4889790415763855,
+      "learning_rate": 0.0005209077904574396,
+      "loss": 2.9937,
+      "step": 14480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5154456496238708,
+      "learning_rate": 0.0005208973732472391,
+      "loss": 3.2068,
+      "step": 14481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4870891273021698,
+      "learning_rate": 0.0005208869554552375,
+      "loss": 3.0451,
+      "step": 14482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5939205288887024,
+      "learning_rate": 0.0005208765370814622,
+      "loss": 3.2206,
+      "step": 14483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4910772442817688,
+      "learning_rate": 0.0005208661181259406,
+      "loss": 3.1084,
+      "step": 14484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5324827432632446,
+      "learning_rate": 0.0005208556985887001,
+      "loss": 3.1209,
+      "step": 14485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.497478723526001,
+      "learning_rate": 0.0005208452784697681,
+      "loss": 3.2166,
+      "step": 14486
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5420011878013611,
+      "learning_rate": 0.0005208348577691723,
+      "loss": 3.2863,
+      "step": 14487
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5050036311149597,
+      "learning_rate": 0.0005208244364869399,
+      "loss": 3.0082,
+      "step": 14488
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4819994568824768,
+      "learning_rate": 0.0005208140146230985,
+      "loss": 3.2324,
+      "step": 14489
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5056858658790588,
+      "learning_rate": 0.0005208035921776755,
+      "loss": 3.2479,
+      "step": 14490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5183698534965515,
+      "learning_rate": 0.0005207931691506982,
+      "loss": 3.206,
+      "step": 14491
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5309738516807556,
+      "learning_rate": 0.0005207827455421943,
+      "loss": 3.2727,
+      "step": 14492
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5266025066375732,
+      "learning_rate": 0.0005207723213521911,
+      "loss": 3.1433,
+      "step": 14493
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5019662976264954,
+      "learning_rate": 0.0005207618965807161,
+      "loss": 3.224,
+      "step": 14494
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5374035239219666,
+      "learning_rate": 0.0005207514712277968,
+      "loss": 3.1059,
+      "step": 14495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5742165446281433,
+      "learning_rate": 0.0005207410452934605,
+      "loss": 3.2328,
+      "step": 14496
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4858763813972473,
+      "learning_rate": 0.0005207306187777348,
+      "loss": 3.0767,
+      "step": 14497
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5236046314239502,
+      "learning_rate": 0.0005207201916806473,
+      "loss": 3.3342,
+      "step": 14498
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5222908854484558,
+      "learning_rate": 0.000520709764002225,
+      "loss": 3.2661,
+      "step": 14499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5480522513389587,
+      "learning_rate": 0.0005206993357424959,
+      "loss": 3.132,
+      "step": 14500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5132131576538086,
+      "learning_rate": 0.0005206889069014871,
+      "loss": 3.3595,
+      "step": 14501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.494584858417511,
+      "learning_rate": 0.0005206784774792263,
+      "loss": 3.3245,
+      "step": 14502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4954333007335663,
+      "learning_rate": 0.0005206680474757407,
+      "loss": 3.2417,
+      "step": 14503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5081412196159363,
+      "learning_rate": 0.0005206576168910581,
+      "loss": 3.0736,
+      "step": 14504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5705395340919495,
+      "learning_rate": 0.0005206471857252057,
+      "loss": 3.0512,
+      "step": 14505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5100386142730713,
+      "learning_rate": 0.000520636753978211,
+      "loss": 3.082,
+      "step": 14506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5036544799804688,
+      "learning_rate": 0.0005206263216501018,
+      "loss": 3.2025,
+      "step": 14507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5994665026664734,
+      "learning_rate": 0.0005206158887409052,
+      "loss": 3.2004,
+      "step": 14508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.50201815366745,
+      "learning_rate": 0.0005206054552506487,
+      "loss": 3.1203,
+      "step": 14509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5317625999450684,
+      "learning_rate": 0.00052059502117936,
+      "loss": 2.9033,
+      "step": 14510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5502082705497742,
+      "learning_rate": 0.0005205845865270664,
+      "loss": 3.1199,
+      "step": 14511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5232048034667969,
+      "learning_rate": 0.0005205741512937955,
+      "loss": 3.1681,
+      "step": 14512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5692151784896851,
+      "learning_rate": 0.0005205637154795748,
+      "loss": 3.1412,
+      "step": 14513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47583895921707153,
+      "learning_rate": 0.0005205532790844317,
+      "loss": 3.3911,
+      "step": 14514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234794616699219,
+      "learning_rate": 0.0005205428421083936,
+      "loss": 3.0717,
+      "step": 14515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5292587876319885,
+      "learning_rate": 0.0005205324045514882,
+      "loss": 3.2203,
+      "step": 14516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4843212068080902,
+      "learning_rate": 0.0005205219664137428,
+      "loss": 3.317,
+      "step": 14517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5709378719329834,
+      "learning_rate": 0.000520511527695185,
+      "loss": 3.3233,
+      "step": 14518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4811621308326721,
+      "learning_rate": 0.0005205010883958423,
+      "loss": 3.2231,
+      "step": 14519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5761101841926575,
+      "learning_rate": 0.0005204906485157423,
+      "loss": 3.431,
+      "step": 14520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5028489828109741,
+      "learning_rate": 0.0005204802080549122,
+      "loss": 3.1521,
+      "step": 14521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5608251690864563,
+      "learning_rate": 0.0005204697670133798,
+      "loss": 2.9437,
+      "step": 14522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5372280478477478,
+      "learning_rate": 0.0005204593253911724,
+      "loss": 3.249,
+      "step": 14523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5417999029159546,
+      "learning_rate": 0.0005204488831883174,
+      "loss": 3.2695,
+      "step": 14524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5193753242492676,
+      "learning_rate": 0.0005204384404048426,
+      "loss": 3.1652,
+      "step": 14525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.528472900390625,
+      "learning_rate": 0.0005204279970407754,
+      "loss": 3.1903,
+      "step": 14526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5966113805770874,
+      "learning_rate": 0.0005204175530961432,
+      "loss": 3.2464,
+      "step": 14527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5065594911575317,
+      "learning_rate": 0.0005204071085709737,
+      "loss": 3.4082,
+      "step": 14528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5471876859664917,
+      "learning_rate": 0.0005203966634652942,
+      "loss": 3.0131,
+      "step": 14529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5287893414497375,
+      "learning_rate": 0.0005203862177791324,
+      "loss": 3.1887,
+      "step": 14530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4954758882522583,
+      "learning_rate": 0.0005203757715125157,
+      "loss": 3.2469,
+      "step": 14531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4946754276752472,
+      "learning_rate": 0.0005203653246654715,
+      "loss": 3.3233,
+      "step": 14532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5238100290298462,
+      "learning_rate": 0.0005203548772380276,
+      "loss": 3.1324,
+      "step": 14533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5408486127853394,
+      "learning_rate": 0.0005203444292302112,
+      "loss": 3.1249,
+      "step": 14534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4843558073043823,
+      "learning_rate": 0.0005203339806420501,
+      "loss": 3.1595,
+      "step": 14535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5062605142593384,
+      "learning_rate": 0.0005203235314735717,
+      "loss": 3.1967,
+      "step": 14536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5224248766899109,
+      "learning_rate": 0.0005203130817248035,
+      "loss": 3.4137,
+      "step": 14537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5459402799606323,
+      "learning_rate": 0.000520302631395773,
+      "loss": 2.9693,
+      "step": 14538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5118008852005005,
+      "learning_rate": 0.0005202921804865078,
+      "loss": 3.4552,
+      "step": 14539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.516243040561676,
+      "learning_rate": 0.0005202817289970353,
+      "loss": 3.3095,
+      "step": 14540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4901047945022583,
+      "learning_rate": 0.0005202712769273833,
+      "loss": 3.1625,
+      "step": 14541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5035889148712158,
+      "learning_rate": 0.0005202608242775791,
+      "loss": 3.2717,
+      "step": 14542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5227107405662537,
+      "learning_rate": 0.0005202503710476502,
+      "loss": 3.3589,
+      "step": 14543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5514225959777832,
+      "learning_rate": 0.0005202399172376242,
+      "loss": 3.1658,
+      "step": 14544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5554363131523132,
+      "learning_rate": 0.0005202294628475287,
+      "loss": 3.0967,
+      "step": 14545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47724464535713196,
+      "learning_rate": 0.0005202190078773912,
+      "loss": 3.2781,
+      "step": 14546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5094786882400513,
+      "learning_rate": 0.0005202085523272393,
+      "loss": 3.3368,
+      "step": 14547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4955451190471649,
+      "learning_rate": 0.0005201980961971002,
+      "loss": 3.1882,
+      "step": 14548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5031975507736206,
+      "learning_rate": 0.0005201876394870018,
+      "loss": 3.0168,
+      "step": 14549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5426763892173767,
+      "learning_rate": 0.0005201771821969716,
+      "loss": 3.14,
+      "step": 14550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4851762056350708,
+      "learning_rate": 0.0005201667243270371,
+      "loss": 3.2685,
+      "step": 14551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5311375856399536,
+      "learning_rate": 0.0005201562658772258,
+      "loss": 3.4027,
+      "step": 14552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5157365798950195,
+      "learning_rate": 0.0005201458068475652,
+      "loss": 3.1562,
+      "step": 14553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5381741523742676,
+      "learning_rate": 0.0005201353472380829,
+      "loss": 3.1626,
+      "step": 14554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.532410204410553,
+      "learning_rate": 0.0005201248870488065,
+      "loss": 3.3552,
+      "step": 14555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5098461508750916,
+      "learning_rate": 0.0005201144262797636,
+      "loss": 3.1537,
+      "step": 14556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5130876898765564,
+      "learning_rate": 0.0005201039649309815,
+      "loss": 3.2291,
+      "step": 14557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.512928307056427,
+      "learning_rate": 0.0005200935030024881,
+      "loss": 3.0487,
+      "step": 14558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5555957555770874,
+      "learning_rate": 0.0005200830404943106,
+      "loss": 3.1302,
+      "step": 14559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49892672896385193,
+      "learning_rate": 0.0005200725774064768,
+      "loss": 3.2873,
+      "step": 14560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5405455827713013,
+      "learning_rate": 0.0005200621137390141,
+      "loss": 3.1677,
+      "step": 14561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48926350474357605,
+      "learning_rate": 0.0005200516494919502,
+      "loss": 3.2558,
+      "step": 14562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5370481014251709,
+      "learning_rate": 0.0005200411846653127,
+      "loss": 3.2475,
+      "step": 14563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5003460049629211,
+      "learning_rate": 0.000520030719259129,
+      "loss": 3.1576,
+      "step": 14564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48471590876579285,
+      "learning_rate": 0.0005200202532734267,
+      "loss": 3.1899,
+      "step": 14565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5002859234809875,
+      "learning_rate": 0.0005200097867082335,
+      "loss": 3.2503,
+      "step": 14566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5285899639129639,
+      "learning_rate": 0.0005199993195635768,
+      "loss": 3.1374,
+      "step": 14567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5553933382034302,
+      "learning_rate": 0.0005199888518394841,
+      "loss": 3.1874,
+      "step": 14568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7586709856987,
+      "learning_rate": 0.0005199783835359833,
+      "loss": 3.2705,
+      "step": 14569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5392246842384338,
+      "learning_rate": 0.0005199679146531017,
+      "loss": 3.0995,
+      "step": 14570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5427459478378296,
+      "learning_rate": 0.000519957445190867,
+      "loss": 3.0778,
+      "step": 14571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5674456357955933,
+      "learning_rate": 0.0005199469751493065,
+      "loss": 3.1027,
+      "step": 14572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5392853021621704,
+      "learning_rate": 0.0005199365045284482,
+      "loss": 3.2362,
+      "step": 14573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5736402869224548,
+      "learning_rate": 0.0005199260333283195,
+      "loss": 3.0357,
+      "step": 14574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5310801267623901,
+      "learning_rate": 0.0005199155615489478,
+      "loss": 3.3081,
+      "step": 14575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5260907411575317,
+      "learning_rate": 0.000519905089190361,
+      "loss": 3.1554,
+      "step": 14576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4972694218158722,
+      "learning_rate": 0.0005198946162525864,
+      "loss": 3.1766,
+      "step": 14577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.472507506608963,
+      "learning_rate": 0.0005198841427356517,
+      "loss": 3.1027,
+      "step": 14578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6582967042922974,
+      "learning_rate": 0.0005198736686395846,
+      "loss": 3.3213,
+      "step": 14579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5455946326255798,
+      "learning_rate": 0.0005198631939644124,
+      "loss": 3.1553,
+      "step": 14580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5628792643547058,
+      "learning_rate": 0.000519852718710163,
+      "loss": 3.0593,
+      "step": 14581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5313958525657654,
+      "learning_rate": 0.0005198422428768639,
+      "loss": 3.2098,
+      "step": 14582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5243238806724548,
+      "learning_rate": 0.0005198317664645424,
+      "loss": 3.2207,
+      "step": 14583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5564098358154297,
+      "learning_rate": 0.0005198212894732266,
+      "loss": 3.0649,
+      "step": 14584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5318445563316345,
+      "learning_rate": 0.0005198108119029437,
+      "loss": 3.2918,
+      "step": 14585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5206500291824341,
+      "learning_rate": 0.0005198003337537214,
+      "loss": 3.3708,
+      "step": 14586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5864635109901428,
+      "learning_rate": 0.0005197898550255874,
+      "loss": 3.2293,
+      "step": 14587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.497435986995697,
+      "learning_rate": 0.0005197793757185692,
+      "loss": 3.2024,
+      "step": 14588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5075864195823669,
+      "learning_rate": 0.0005197688958326944,
+      "loss": 3.2853,
+      "step": 14589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5199374556541443,
+      "learning_rate": 0.0005197584153679906,
+      "loss": 3.2211,
+      "step": 14590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5085850358009338,
+      "learning_rate": 0.0005197479343244854,
+      "loss": 3.43,
+      "step": 14591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5507273077964783,
+      "learning_rate": 0.0005197374527022065,
+      "loss": 3.3325,
+      "step": 14592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5263931155204773,
+      "learning_rate": 0.0005197269705011815,
+      "loss": 3.1009,
+      "step": 14593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5148525834083557,
+      "learning_rate": 0.0005197164877214379,
+      "loss": 3.0378,
+      "step": 14594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5128089189529419,
+      "learning_rate": 0.0005197060043630032,
+      "loss": 3.2473,
+      "step": 14595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4934292733669281,
+      "learning_rate": 0.0005196955204259053,
+      "loss": 3.1324,
+      "step": 14596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5595125555992126,
+      "learning_rate": 0.0005196850359101716,
+      "loss": 3.201,
+      "step": 14597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5288695693016052,
+      "learning_rate": 0.0005196745508158299,
+      "loss": 3.0758,
+      "step": 14598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5381522178649902,
+      "learning_rate": 0.0005196640651429078,
+      "loss": 3.0904,
+      "step": 14599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.502053439617157,
+      "learning_rate": 0.0005196535788914326,
+      "loss": 3.2181,
+      "step": 14600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49978283047676086,
+      "learning_rate": 0.0005196430920614323,
+      "loss": 3.274,
+      "step": 14601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5054643750190735,
+      "learning_rate": 0.0005196326046529344,
+      "loss": 3.2331,
+      "step": 14602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5227307677268982,
+      "learning_rate": 0.0005196221166659663,
+      "loss": 3.3232,
+      "step": 14603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5008553862571716,
+      "learning_rate": 0.000519611628100556,
+      "loss": 3.2415,
+      "step": 14604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5128350853919983,
+      "learning_rate": 0.0005196011389567308,
+      "loss": 3.219,
+      "step": 14605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5377113223075867,
+      "learning_rate": 0.0005195906492345186,
+      "loss": 3.0948,
+      "step": 14606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5005555152893066,
+      "learning_rate": 0.0005195801589339468,
+      "loss": 3.3787,
+      "step": 14607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5084679126739502,
+      "learning_rate": 0.0005195696680550431,
+      "loss": 3.2606,
+      "step": 14608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4670720100402832,
+      "learning_rate": 0.0005195591765978352,
+      "loss": 3.2102,
+      "step": 14609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5008261799812317,
+      "learning_rate": 0.0005195486845623507,
+      "loss": 3.3554,
+      "step": 14610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5201194286346436,
+      "learning_rate": 0.0005195381919486171,
+      "loss": 3.1835,
+      "step": 14611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5437430143356323,
+      "learning_rate": 0.0005195276987566623,
+      "loss": 3.1677,
+      "step": 14612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5281556844711304,
+      "learning_rate": 0.0005195172049865138,
+      "loss": 3.3595,
+      "step": 14613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5061221122741699,
+      "learning_rate": 0.0005195067106381992,
+      "loss": 3.2528,
+      "step": 14614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5027444362640381,
+      "learning_rate": 0.000519496215711746,
+      "loss": 3.115,
+      "step": 14615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5213714241981506,
+      "learning_rate": 0.0005194857202071822,
+      "loss": 3.3346,
+      "step": 14616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5137941837310791,
+      "learning_rate": 0.0005194752241245352,
+      "loss": 3.2455,
+      "step": 14617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.520012617111206,
+      "learning_rate": 0.0005194647274638327,
+      "loss": 3.3542,
+      "step": 14618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5151776075363159,
+      "learning_rate": 0.0005194542302251024,
+      "loss": 3.2251,
+      "step": 14619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5593071579933167,
+      "learning_rate": 0.0005194437324083718,
+      "loss": 2.9776,
+      "step": 14620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4902730882167816,
+      "learning_rate": 0.0005194332340136686,
+      "loss": 3.2225,
+      "step": 14621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5338214039802551,
+      "learning_rate": 0.0005194227350410205,
+      "loss": 3.3136,
+      "step": 14622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5151000022888184,
+      "learning_rate": 0.0005194122354904553,
+      "loss": 3.224,
+      "step": 14623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5185823440551758,
+      "learning_rate": 0.0005194017353620004,
+      "loss": 3.149,
+      "step": 14624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5511698126792908,
+      "learning_rate": 0.0005193912346556836,
+      "loss": 3.1918,
+      "step": 14625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5243788361549377,
+      "learning_rate": 0.0005193807333715324,
+      "loss": 3.1177,
+      "step": 14626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5167205333709717,
+      "learning_rate": 0.0005193702315095746,
+      "loss": 2.923,
+      "step": 14627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6894581317901611,
+      "learning_rate": 0.0005193597290698379,
+      "loss": 3.3533,
+      "step": 14628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5190028548240662,
+      "learning_rate": 0.0005193492260523499,
+      "loss": 3.2836,
+      "step": 14629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.508630633354187,
+      "learning_rate": 0.0005193387224571381,
+      "loss": 3.4301,
+      "step": 14630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5017585158348083,
+      "learning_rate": 0.0005193282182842305,
+      "loss": 3.2973,
+      "step": 14631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5430968403816223,
+      "learning_rate": 0.0005193177135336544,
+      "loss": 3.1365,
+      "step": 14632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5301803350448608,
+      "learning_rate": 0.0005193072082054379,
+      "loss": 3.2685,
+      "step": 14633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5263997912406921,
+      "learning_rate": 0.0005192967022996083,
+      "loss": 3.4263,
+      "step": 14634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5006689429283142,
+      "learning_rate": 0.0005192861958161933,
+      "loss": 3.3041,
+      "step": 14635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5174351334571838,
+      "learning_rate": 0.0005192756887552207,
+      "loss": 3.1699,
+      "step": 14636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6178990602493286,
+      "learning_rate": 0.0005192651811167183,
+      "loss": 3.1885,
+      "step": 14637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5171335339546204,
+      "learning_rate": 0.0005192546729007135,
+      "loss": 3.1908,
+      "step": 14638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5851891040802002,
+      "learning_rate": 0.000519244164107234,
+      "loss": 3.2078,
+      "step": 14639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49289432168006897,
+      "learning_rate": 0.0005192336547363076,
+      "loss": 3.2048,
+      "step": 14640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4947859048843384,
+      "learning_rate": 0.0005192231447879621,
+      "loss": 3.1892,
+      "step": 14641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5246251821517944,
+      "learning_rate": 0.0005192126342622249,
+      "loss": 2.9959,
+      "step": 14642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5482833385467529,
+      "learning_rate": 0.0005192021231591238,
+      "loss": 3.3374,
+      "step": 14643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5282585620880127,
+      "learning_rate": 0.0005191916114786865,
+      "loss": 3.2919,
+      "step": 14644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5180543661117554,
+      "learning_rate": 0.0005191810992209407,
+      "loss": 3.2924,
+      "step": 14645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5106507539749146,
+      "learning_rate": 0.0005191705863859141,
+      "loss": 3.0478,
+      "step": 14646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.501225471496582,
+      "learning_rate": 0.0005191600729736343,
+      "loss": 3.3096,
+      "step": 14647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5608596205711365,
+      "learning_rate": 0.000519149558984129,
+      "loss": 3.1157,
+      "step": 14648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.521439790725708,
+      "learning_rate": 0.0005191390444174261,
+      "loss": 3.2124,
+      "step": 14649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4991474747657776,
+      "learning_rate": 0.0005191285292735529,
+      "loss": 3.2962,
+      "step": 14650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5133585333824158,
+      "learning_rate": 0.0005191180135525375,
+      "loss": 3.3118,
+      "step": 14651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4970352053642273,
+      "learning_rate": 0.0005191074972544073,
+      "loss": 3.3128,
+      "step": 14652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5411221385002136,
+      "learning_rate": 0.0005190969803791903,
+      "loss": 3.2144,
+      "step": 14653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5432132482528687,
+      "learning_rate": 0.0005190864629269139,
+      "loss": 3.1416,
+      "step": 14654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.599610447883606,
+      "learning_rate": 0.0005190759448976059,
+      "loss": 2.9844,
+      "step": 14655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49764925241470337,
+      "learning_rate": 0.0005190654262912941,
+      "loss": 3.1174,
+      "step": 14656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49829450249671936,
+      "learning_rate": 0.000519054907108006,
+      "loss": 3.2279,
+      "step": 14657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5487377643585205,
+      "learning_rate": 0.0005190443873477696,
+      "loss": 3.0803,
+      "step": 14658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5312570333480835,
+      "learning_rate": 0.0005190338670106124,
+      "loss": 3.2438,
+      "step": 14659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5211238265037537,
+      "learning_rate": 0.0005190233460965621,
+      "loss": 3.2646,
+      "step": 14660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5218424797058105,
+      "learning_rate": 0.0005190128246056465,
+      "loss": 3.0871,
+      "step": 14661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5989399552345276,
+      "learning_rate": 0.0005190023025378932,
+      "loss": 3.2677,
+      "step": 14662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5372099876403809,
+      "learning_rate": 0.00051899177989333,
+      "loss": 3.164,
+      "step": 14663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4821583926677704,
+      "learning_rate": 0.0005189812566719847,
+      "loss": 3.1532,
+      "step": 14664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5003272891044617,
+      "learning_rate": 0.0005189707328738848,
+      "loss": 3.3264,
+      "step": 14665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.521091103553772,
+      "learning_rate": 0.0005189602084990581,
+      "loss": 3.1722,
+      "step": 14666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5140360593795776,
+      "learning_rate": 0.0005189496835475325,
+      "loss": 3.2693,
+      "step": 14667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5883575081825256,
+      "learning_rate": 0.0005189391580193354,
+      "loss": 3.0789,
+      "step": 14668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4931708574295044,
+      "learning_rate": 0.0005189286319144949,
+      "loss": 3.1453,
+      "step": 14669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5307677984237671,
+      "learning_rate": 0.0005189181052330384,
+      "loss": 3.3826,
+      "step": 14670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4849635362625122,
+      "learning_rate": 0.0005189075779749937,
+      "loss": 3.0497,
+      "step": 14671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49971646070480347,
+      "learning_rate": 0.0005188970501403886,
+      "loss": 3.0944,
+      "step": 14672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.521876871585846,
+      "learning_rate": 0.0005188865217292508,
+      "loss": 3.1896,
+      "step": 14673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4860338568687439,
+      "learning_rate": 0.000518875992741608,
+      "loss": 3.2229,
+      "step": 14674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5542223453521729,
+      "learning_rate": 0.0005188654631774881,
+      "loss": 3.0625,
+      "step": 14675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5281686186790466,
+      "learning_rate": 0.0005188549330369186,
+      "loss": 3.1883,
+      "step": 14676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49884727597236633,
+      "learning_rate": 0.0005188444023199272,
+      "loss": 3.141,
+      "step": 14677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5445032715797424,
+      "learning_rate": 0.0005188338710265419,
+      "loss": 3.2494,
+      "step": 14678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5325313806533813,
+      "learning_rate": 0.0005188233391567903,
+      "loss": 3.1469,
+      "step": 14679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5177390575408936,
+      "learning_rate": 0.0005188128067107002,
+      "loss": 2.9586,
+      "step": 14680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5098357200622559,
+      "learning_rate": 0.0005188022736882991,
+      "loss": 3.1629,
+      "step": 14681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4955054521560669,
+      "learning_rate": 0.0005187917400896149,
+      "loss": 3.3163,
+      "step": 14682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5325889587402344,
+      "learning_rate": 0.0005187812059146756,
+      "loss": 3.1758,
+      "step": 14683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5700262188911438,
+      "learning_rate": 0.0005187706711635086,
+      "loss": 3.1304,
+      "step": 14684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.517220139503479,
+      "learning_rate": 0.0005187601358361417,
+      "loss": 3.0101,
+      "step": 14685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5129212737083435,
+      "learning_rate": 0.0005187495999326027,
+      "loss": 2.9816,
+      "step": 14686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5115025043487549,
+      "learning_rate": 0.0005187390634529194,
+      "loss": 3.1781,
+      "step": 14687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5292808413505554,
+      "learning_rate": 0.0005187285263971196,
+      "loss": 3.129,
+      "step": 14688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5144400596618652,
+      "learning_rate": 0.0005187179887652307,
+      "loss": 3.1161,
+      "step": 14689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.525768518447876,
+      "learning_rate": 0.000518707450557281,
+      "loss": 3.2922,
+      "step": 14690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5286670327186584,
+      "learning_rate": 0.0005186969117732977,
+      "loss": 3.4953,
+      "step": 14691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5114057660102844,
+      "learning_rate": 0.000518686372413309,
+      "loss": 3.3082,
+      "step": 14692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.513731062412262,
+      "learning_rate": 0.0005186758324773425,
+      "loss": 3.048,
+      "step": 14693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5025843977928162,
+      "learning_rate": 0.0005186652919654259,
+      "loss": 3.2755,
+      "step": 14694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.615983247756958,
+      "learning_rate": 0.0005186547508775869,
+      "loss": 3.111,
+      "step": 14695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5085950493812561,
+      "learning_rate": 0.0005186442092138535,
+      "loss": 3.2742,
+      "step": 14696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5130026936531067,
+      "learning_rate": 0.0005186336669742533,
+      "loss": 3.1467,
+      "step": 14697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5810296535491943,
+      "learning_rate": 0.000518623124158814,
+      "loss": 3.2415,
+      "step": 14698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5049729943275452,
+      "learning_rate": 0.0005186125807675636,
+      "loss": 3.09,
+      "step": 14699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5015254616737366,
+      "learning_rate": 0.0005186020368005297,
+      "loss": 3.0512,
+      "step": 14700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5339839458465576,
+      "learning_rate": 0.00051859149225774,
+      "loss": 3.2614,
+      "step": 14701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5117117762565613,
+      "learning_rate": 0.0005185809471392226,
+      "loss": 3.4624,
+      "step": 14702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5019668936729431,
+      "learning_rate": 0.0005185704014450048,
+      "loss": 3.3371,
+      "step": 14703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5273573994636536,
+      "learning_rate": 0.0005185598551751148,
+      "loss": 3.3187,
+      "step": 14704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5079138278961182,
+      "learning_rate": 0.0005185493083295802,
+      "loss": 3.2523,
+      "step": 14705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5488572120666504,
+      "learning_rate": 0.0005185387609084286,
+      "loss": 3.1907,
+      "step": 14706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.47864294052124023,
+      "learning_rate": 0.0005185282129116882,
+      "loss": 2.9153,
+      "step": 14707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.516621470451355,
+      "learning_rate": 0.0005185176643393864,
+      "loss": 3.2739,
+      "step": 14708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5150627493858337,
+      "learning_rate": 0.0005185071151915512,
+      "loss": 3.2769,
+      "step": 14709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5592693090438843,
+      "learning_rate": 0.0005184965654682103,
+      "loss": 3.2601,
+      "step": 14710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5248083472251892,
+      "learning_rate": 0.0005184860151693914,
+      "loss": 3.0784,
+      "step": 14711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5441768765449524,
+      "learning_rate": 0.0005184754642951224,
+      "loss": 3.0752,
+      "step": 14712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4891248047351837,
+      "learning_rate": 0.0005184649128454312,
+      "loss": 3.2164,
+      "step": 14713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5104601383209229,
+      "learning_rate": 0.0005184543608203454,
+      "loss": 2.9591,
+      "step": 14714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5093756318092346,
+      "learning_rate": 0.0005184438082198929,
+      "loss": 3.1289,
+      "step": 14715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5416193604469299,
+      "learning_rate": 0.0005184332550441013,
+      "loss": 3.1514,
+      "step": 14716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5121513605117798,
+      "learning_rate": 0.0005184227012929987,
+      "loss": 3.3509,
+      "step": 14717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5216866135597229,
+      "learning_rate": 0.0005184121469666127,
+      "loss": 3.1625,
+      "step": 14718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5179713368415833,
+      "learning_rate": 0.0005184015920649711,
+      "loss": 3.0819,
+      "step": 14719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4934249222278595,
+      "learning_rate": 0.0005183910365881018,
+      "loss": 3.0533,
+      "step": 14720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5014771819114685,
+      "learning_rate": 0.0005183804805360324,
+      "loss": 3.4464,
+      "step": 14721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4989439845085144,
+      "learning_rate": 0.000518369923908791,
+      "loss": 3.0768,
+      "step": 14722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5312348008155823,
+      "learning_rate": 0.0005183593667064052,
+      "loss": 3.1842,
+      "step": 14723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5148173570632935,
+      "learning_rate": 0.0005183488089289029,
+      "loss": 3.1817,
+      "step": 14724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5735609531402588,
+      "learning_rate": 0.0005183382505763117,
+      "loss": 3.1671,
+      "step": 14725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4953717887401581,
+      "learning_rate": 0.0005183276916486597,
+      "loss": 3.0436,
+      "step": 14726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5508190393447876,
+      "learning_rate": 0.0005183171321459745,
+      "loss": 3.1293,
+      "step": 14727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4912906289100647,
+      "learning_rate": 0.0005183065720682842,
+      "loss": 3.2873,
+      "step": 14728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5511575937271118,
+      "learning_rate": 0.0005182960114156162,
+      "loss": 3.1012,
+      "step": 14729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4874035716056824,
+      "learning_rate": 0.0005182854501879986,
+      "loss": 3.3522,
+      "step": 14730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5335097312927246,
+      "learning_rate": 0.000518274888385459,
+      "loss": 3.0398,
+      "step": 14731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5298565626144409,
+      "learning_rate": 0.0005182643260080254,
+      "loss": 2.8668,
+      "step": 14732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5011436939239502,
+      "learning_rate": 0.0005182537630557257,
+      "loss": 3.2198,
+      "step": 14733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5561090111732483,
+      "learning_rate": 0.0005182431995285875,
+      "loss": 2.8162,
+      "step": 14734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5102478265762329,
+      "learning_rate": 0.0005182326354266387,
+      "loss": 3.1414,
+      "step": 14735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.472723126411438,
+      "learning_rate": 0.0005182220707499072,
+      "loss": 2.9028,
+      "step": 14736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5202345252037048,
+      "learning_rate": 0.0005182115054984207,
+      "loss": 3.2935,
+      "step": 14737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5131343007087708,
+      "learning_rate": 0.0005182009396722071,
+      "loss": 3.2739,
+      "step": 14738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5101212859153748,
+      "learning_rate": 0.0005181903732712943,
+      "loss": 3.2873,
+      "step": 14739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5080827474594116,
+      "learning_rate": 0.0005181798062957099,
+      "loss": 3.0049,
+      "step": 14740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.526172935962677,
+      "learning_rate": 0.000518169238745482,
+      "loss": 3.2314,
+      "step": 14741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5397858619689941,
+      "learning_rate": 0.0005181586706206382,
+      "loss": 3.241,
+      "step": 14742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5016712546348572,
+      "learning_rate": 0.0005181481019212066,
+      "loss": 3.3635,
+      "step": 14743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5157707929611206,
+      "learning_rate": 0.0005181375326472147,
+      "loss": 3.2207,
+      "step": 14744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5005051493644714,
+      "learning_rate": 0.0005181269627986908,
+      "loss": 3.1462,
+      "step": 14745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.47533586621284485,
+      "learning_rate": 0.0005181163923756622,
+      "loss": 3.1661,
+      "step": 14746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.537551760673523,
+      "learning_rate": 0.000518105821378157,
+      "loss": 3.185,
+      "step": 14747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5036092400550842,
+      "learning_rate": 0.0005180952498062032,
+      "loss": 3.1514,
+      "step": 14748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5005072355270386,
+      "learning_rate": 0.0005180846776598285,
+      "loss": 3.1253,
+      "step": 14749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5564939379692078,
+      "learning_rate": 0.0005180741049390605,
+      "loss": 3.2356,
+      "step": 14750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.584366500377655,
+      "learning_rate": 0.0005180635316439275,
+      "loss": 3.0674,
+      "step": 14751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5277496576309204,
+      "learning_rate": 0.000518052957774457,
+      "loss": 3.2329,
+      "step": 14752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5059505105018616,
+      "learning_rate": 0.000518042383330677,
+      "loss": 3.2356,
+      "step": 14753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5543511509895325,
+      "learning_rate": 0.0005180318083126154,
+      "loss": 3.39,
+      "step": 14754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.537144124507904,
+      "learning_rate": 0.0005180212327202999,
+      "loss": 3.078,
+      "step": 14755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5184658765792847,
+      "learning_rate": 0.0005180106565537585,
+      "loss": 3.2036,
+      "step": 14756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5322108268737793,
+      "learning_rate": 0.000518000079813019,
+      "loss": 3.2073,
+      "step": 14757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.511556088924408,
+      "learning_rate": 0.0005179895024981092,
+      "loss": 3.1481,
+      "step": 14758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5586161017417908,
+      "learning_rate": 0.0005179789246090568,
+      "loss": 3.3132,
+      "step": 14759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5228317379951477,
+      "learning_rate": 0.0005179683461458901,
+      "loss": 3.1418,
+      "step": 14760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4944567084312439,
+      "learning_rate": 0.0005179577671086367,
+      "loss": 3.1762,
+      "step": 14761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6001697182655334,
+      "learning_rate": 0.0005179471874973245,
+      "loss": 3.1919,
+      "step": 14762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5296616554260254,
+      "learning_rate": 0.0005179366073119813,
+      "loss": 3.1888,
+      "step": 14763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390952825546265,
+      "learning_rate": 0.000517926026552635,
+      "loss": 3.0015,
+      "step": 14764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5541010499000549,
+      "learning_rate": 0.0005179154452193135,
+      "loss": 3.1304,
+      "step": 14765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49909520149230957,
+      "learning_rate": 0.0005179048633120447,
+      "loss": 3.1821,
+      "step": 14766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5527284145355225,
+      "learning_rate": 0.0005178942808308564,
+      "loss": 3.2751,
+      "step": 14767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5055661201477051,
+      "learning_rate": 0.0005178836977757765,
+      "loss": 3.2823,
+      "step": 14768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5182890892028809,
+      "learning_rate": 0.0005178731141468329,
+      "loss": 3.1544,
+      "step": 14769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4902885854244232,
+      "learning_rate": 0.0005178625299440534,
+      "loss": 3.0724,
+      "step": 14770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5040931701660156,
+      "learning_rate": 0.0005178519451674659,
+      "loss": 3.378,
+      "step": 14771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5123295187950134,
+      "learning_rate": 0.0005178413598170984,
+      "loss": 3.3552,
+      "step": 14772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.533523440361023,
+      "learning_rate": 0.0005178307738929786,
+      "loss": 3.1523,
+      "step": 14773
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5177558064460754,
+      "learning_rate": 0.0005178201873951346,
+      "loss": 3.139,
+      "step": 14774
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5482215285301208,
+      "learning_rate": 0.000517809600323594,
+      "loss": 2.9881,
+      "step": 14775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5305909514427185,
+      "learning_rate": 0.0005177990126783849,
+      "loss": 3.1347,
+      "step": 14776
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5373917818069458,
+      "learning_rate": 0.0005177884244595352,
+      "loss": 3.3311,
+      "step": 14777
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5221949815750122,
+      "learning_rate": 0.0005177778356670725,
+      "loss": 3.154,
+      "step": 14778
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5249523520469666,
+      "learning_rate": 0.000517767246301025,
+      "loss": 3.026,
+      "step": 14779
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5519317388534546,
+      "learning_rate": 0.0005177566563614205,
+      "loss": 3.0845,
+      "step": 14780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48520752787590027,
+      "learning_rate": 0.0005177460658482868,
+      "loss": 3.1916,
+      "step": 14781
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5459604263305664,
+      "learning_rate": 0.0005177354747616519,
+      "loss": 3.2222,
+      "step": 14782
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4783353805541992,
+      "learning_rate": 0.0005177248831015437,
+      "loss": 3.1407,
+      "step": 14783
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5260900259017944,
+      "learning_rate": 0.00051771429086799,
+      "loss": 3.2664,
+      "step": 14784
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5184537768363953,
+      "learning_rate": 0.0005177036980610188,
+      "loss": 3.3091,
+      "step": 14785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.523328959941864,
+      "learning_rate": 0.000517693104680658,
+      "loss": 3.2127,
+      "step": 14786
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4887447953224182,
+      "learning_rate": 0.0005176825107269353,
+      "loss": 3.177,
+      "step": 14787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5406696200370789,
+      "learning_rate": 0.000517671916199879,
+      "loss": 3.3842,
+      "step": 14788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5218521952629089,
+      "learning_rate": 0.0005176613210995166,
+      "loss": 3.0918,
+      "step": 14789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5580902695655823,
+      "learning_rate": 0.0005176507254258763,
+      "loss": 3.2155,
+      "step": 14790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5313786268234253,
+      "learning_rate": 0.0005176401291789857,
+      "loss": 3.1422,
+      "step": 14791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5215513110160828,
+      "learning_rate": 0.000517629532358873,
+      "loss": 3.1058,
+      "step": 14792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.490644633769989,
+      "learning_rate": 0.000517618934965566,
+      "loss": 3.084,
+      "step": 14793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5126657485961914,
+      "learning_rate": 0.0005176083369990925,
+      "loss": 2.951,
+      "step": 14794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5562359690666199,
+      "learning_rate": 0.0005175977384594807,
+      "loss": 3.1351,
+      "step": 14795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5439273715019226,
+      "learning_rate": 0.0005175871393467581,
+      "loss": 3.0672,
+      "step": 14796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5346477031707764,
+      "learning_rate": 0.000517576539660953,
+      "loss": 3.1833,
+      "step": 14797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5201503038406372,
+      "learning_rate": 0.0005175659394020932,
+      "loss": 3.2397,
+      "step": 14798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5136426091194153,
+      "learning_rate": 0.0005175553385702065,
+      "loss": 3.2395,
+      "step": 14799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5076440572738647,
+      "learning_rate": 0.000517544737165321,
+      "loss": 3.1516,
+      "step": 14800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.531037449836731,
+      "learning_rate": 0.0005175341351874645,
+      "loss": 3.0899,
+      "step": 14801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5360001921653748,
+      "learning_rate": 0.0005175235326366649,
+      "loss": 3.3677,
+      "step": 14802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5268040299415588,
+      "learning_rate": 0.0005175129295129503,
+      "loss": 3.1516,
+      "step": 14803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5175145864486694,
+      "learning_rate": 0.0005175023258163483,
+      "loss": 3.1313,
+      "step": 14804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5107941627502441,
+      "learning_rate": 0.000517491721546887,
+      "loss": 3.1514,
+      "step": 14805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0005174811167045946,
+      "loss": 3.1506,
+      "step": 14806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49850302934646606,
+      "learning_rate": 0.0005174705112894987,
+      "loss": 3.4543,
+      "step": 14807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5240212678909302,
+      "learning_rate": 0.0005174599053016273,
+      "loss": 3.2027,
+      "step": 14808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5271751880645752,
+      "learning_rate": 0.0005174492987410084,
+      "loss": 3.2392,
+      "step": 14809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4971039295196533,
+      "learning_rate": 0.00051743869160767,
+      "loss": 3.1203,
+      "step": 14810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5178433060646057,
+      "learning_rate": 0.0005174280839016398,
+      "loss": 3.1699,
+      "step": 14811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5342043042182922,
+      "learning_rate": 0.0005174174756229458,
+      "loss": 3.0835,
+      "step": 14812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5262744426727295,
+      "learning_rate": 0.0005174068667716162,
+      "loss": 3.2602,
+      "step": 14813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5131629109382629,
+      "learning_rate": 0.0005173962573476786,
+      "loss": 3.2553,
+      "step": 14814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5120997428894043,
+      "learning_rate": 0.0005173856473511612,
+      "loss": 3.2564,
+      "step": 14815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5193552374839783,
+      "learning_rate": 0.0005173750367820917,
+      "loss": 3.2926,
+      "step": 14816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4874288737773895,
+      "learning_rate": 0.0005173644256404983,
+      "loss": 3.1219,
+      "step": 14817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5450944304466248,
+      "learning_rate": 0.0005173538139264087,
+      "loss": 3.2068,
+      "step": 14818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5270337462425232,
+      "learning_rate": 0.0005173432016398513,
+      "loss": 3.3498,
+      "step": 14819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5095805525779724,
+      "learning_rate": 0.0005173325887808535,
+      "loss": 3.1008,
+      "step": 14820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5138509273529053,
+      "learning_rate": 0.0005173219753494435,
+      "loss": 3.0766,
+      "step": 14821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5434461236000061,
+      "learning_rate": 0.0005173113613456493,
+      "loss": 3.1396,
+      "step": 14822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5392305850982666,
+      "learning_rate": 0.0005173007467694986,
+      "loss": 3.1582,
+      "step": 14823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5202492475509644,
+      "learning_rate": 0.0005172901316210197,
+      "loss": 3.1764,
+      "step": 14824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5307274460792542,
+      "learning_rate": 0.0005172795159002405,
+      "loss": 3.3271,
+      "step": 14825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5426248908042908,
+      "learning_rate": 0.0005172688996071889,
+      "loss": 3.0903,
+      "step": 14826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5029236078262329,
+      "learning_rate": 0.0005172582827418927,
+      "loss": 3.386,
+      "step": 14827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.522150993347168,
+      "learning_rate": 0.0005172476653043799,
+      "loss": 3.2013,
+      "step": 14828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5246309638023376,
+      "learning_rate": 0.0005172370472946787,
+      "loss": 3.2427,
+      "step": 14829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5152744650840759,
+      "learning_rate": 0.000517226428712817,
+      "loss": 3.1461,
+      "step": 14830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5182197690010071,
+      "learning_rate": 0.0005172158095588227,
+      "loss": 3.0218,
+      "step": 14831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5244132280349731,
+      "learning_rate": 0.0005172051898327235,
+      "loss": 3.2129,
+      "step": 14832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5407620072364807,
+      "learning_rate": 0.0005171945695345478,
+      "loss": 3.0372,
+      "step": 14833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5058204531669617,
+      "learning_rate": 0.0005171839486643234,
+      "loss": 2.9253,
+      "step": 14834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5572190284729004,
+      "learning_rate": 0.0005171733272220783,
+      "loss": 3.2299,
+      "step": 14835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5125101208686829,
+      "learning_rate": 0.0005171627052078404,
+      "loss": 3.3526,
+      "step": 14836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5376558899879456,
+      "learning_rate": 0.0005171520826216377,
+      "loss": 3.2156,
+      "step": 14837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5481821298599243,
+      "learning_rate": 0.0005171414594634983,
+      "loss": 3.2858,
+      "step": 14838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4958679974079132,
+      "learning_rate": 0.0005171308357334501,
+      "loss": 3.258,
+      "step": 14839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5072594285011292,
+      "learning_rate": 0.000517120211431521,
+      "loss": 3.2469,
+      "step": 14840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5196676850318909,
+      "learning_rate": 0.0005171095865577391,
+      "loss": 3.1494,
+      "step": 14841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5403569936752319,
+      "learning_rate": 0.0005170989611121323,
+      "loss": 3.1397,
+      "step": 14842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4967591464519501,
+      "learning_rate": 0.0005170883350947286,
+      "loss": 3.2614,
+      "step": 14843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5076732635498047,
+      "learning_rate": 0.000517077708505556,
+      "loss": 2.9743,
+      "step": 14844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5110801458358765,
+      "learning_rate": 0.0005170670813446425,
+      "loss": 3.1071,
+      "step": 14845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48647382855415344,
+      "learning_rate": 0.0005170564536120161,
+      "loss": 3.2916,
+      "step": 14846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5668714046478271,
+      "learning_rate": 0.0005170458253077048,
+      "loss": 3.0199,
+      "step": 14847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48818862438201904,
+      "learning_rate": 0.0005170351964317364,
+      "loss": 3.1778,
+      "step": 14848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5142984390258789,
+      "learning_rate": 0.0005170245669841393,
+      "loss": 3.3728,
+      "step": 14849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5312206149101257,
+      "learning_rate": 0.0005170139369649412,
+      "loss": 3.0934,
+      "step": 14850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5142450928688049,
+      "learning_rate": 0.0005170033063741701,
+      "loss": 3.1013,
+      "step": 14851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5436277985572815,
+      "learning_rate": 0.0005169926752118541,
+      "loss": 3.2664,
+      "step": 14852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.545096218585968,
+      "learning_rate": 0.0005169820434780211,
+      "loss": 3.1664,
+      "step": 14853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5273535847663879,
+      "learning_rate": 0.0005169714111726992,
+      "loss": 2.9818,
+      "step": 14854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5400977730751038,
+      "learning_rate": 0.0005169607782959163,
+      "loss": 3.1989,
+      "step": 14855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5369027853012085,
+      "learning_rate": 0.0005169501448477005,
+      "loss": 3.1274,
+      "step": 14856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48228007555007935,
+      "learning_rate": 0.0005169395108280797,
+      "loss": 3.3068,
+      "step": 14857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.562002420425415,
+      "learning_rate": 0.0005169288762370821,
+      "loss": 3.0936,
+      "step": 14858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5516139268875122,
+      "learning_rate": 0.0005169182410747356,
+      "loss": 3.0671,
+      "step": 14859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5317583680152893,
+      "learning_rate": 0.0005169076053410681,
+      "loss": 3.0629,
+      "step": 14860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5046291947364807,
+      "learning_rate": 0.0005168969690361077,
+      "loss": 3.1502,
+      "step": 14861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5158634781837463,
+      "learning_rate": 0.0005168863321598825,
+      "loss": 3.157,
+      "step": 14862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5672610402107239,
+      "learning_rate": 0.0005168756947124204,
+      "loss": 3.0771,
+      "step": 14863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5045456290245056,
+      "learning_rate": 0.0005168650566937496,
+      "loss": 3.2721,
+      "step": 14864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5558025240898132,
+      "learning_rate": 0.0005168544181038978,
+      "loss": 3.1883,
+      "step": 14865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5100647807121277,
+      "learning_rate": 0.0005168437789428932,
+      "loss": 3.1412,
+      "step": 14866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4676547050476074,
+      "learning_rate": 0.0005168331392107639,
+      "loss": 3.1914,
+      "step": 14867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5415322780609131,
+      "learning_rate": 0.0005168224989075378,
+      "loss": 3.0172,
+      "step": 14868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5330214500427246,
+      "learning_rate": 0.000516811858033243,
+      "loss": 3.2398,
+      "step": 14869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6251876354217529,
+      "learning_rate": 0.0005168012165879074,
+      "loss": 3.0238,
+      "step": 14870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5353378653526306,
+      "learning_rate": 0.0005167905745715592,
+      "loss": 3.2086,
+      "step": 14871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.537301242351532,
+      "learning_rate": 0.0005167799319842264,
+      "loss": 2.9889,
+      "step": 14872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5349966883659363,
+      "learning_rate": 0.0005167692888259368,
+      "loss": 3.3271,
+      "step": 14873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5379777550697327,
+      "learning_rate": 0.0005167586450967189,
+      "loss": 2.9083,
+      "step": 14874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5365452766418457,
+      "learning_rate": 0.0005167480007966002,
+      "loss": 3.2886,
+      "step": 14875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5008100867271423,
+      "learning_rate": 0.000516737355925609,
+      "loss": 3.1322,
+      "step": 14876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5189768671989441,
+      "learning_rate": 0.0005167267104837735,
+      "loss": 3.1345,
+      "step": 14877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5536531209945679,
+      "learning_rate": 0.0005167160644711214,
+      "loss": 3.0225,
+      "step": 14878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4946756660938263,
+      "learning_rate": 0.000516705417887681,
+      "loss": 3.4478,
+      "step": 14879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5488638877868652,
+      "learning_rate": 0.00051669477073348,
+      "loss": 2.9697,
+      "step": 14880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.520703911781311,
+      "learning_rate": 0.0005166841230085469,
+      "loss": 3.1385,
+      "step": 14881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5039485096931458,
+      "learning_rate": 0.0005166734747129094,
+      "loss": 3.1356,
+      "step": 14882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4801254868507385,
+      "learning_rate": 0.0005166628258465958,
+      "loss": 3.2317,
+      "step": 14883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5044119358062744,
+      "learning_rate": 0.0005166521764096339,
+      "loss": 2.9742,
+      "step": 14884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5282803773880005,
+      "learning_rate": 0.0005166415264020519,
+      "loss": 3.0444,
+      "step": 14885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5237149596214294,
+      "learning_rate": 0.0005166308758238779,
+      "loss": 3.3579,
+      "step": 14886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5436862111091614,
+      "learning_rate": 0.0005166202246751397,
+      "loss": 3.1705,
+      "step": 14887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5225011110305786,
+      "learning_rate": 0.0005166095729558656,
+      "loss": 3.145,
+      "step": 14888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5690309405326843,
+      "learning_rate": 0.0005165989206660836,
+      "loss": 2.8944,
+      "step": 14889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5105469226837158,
+      "learning_rate": 0.0005165882678058217,
+      "loss": 3.1488,
+      "step": 14890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5054097175598145,
+      "learning_rate": 0.000516577614375108,
+      "loss": 3.4983,
+      "step": 14891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.570755660533905,
+      "learning_rate": 0.0005165669603739704,
+      "loss": 3.2432,
+      "step": 14892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5138176679611206,
+      "learning_rate": 0.0005165563058024373,
+      "loss": 3.1639,
+      "step": 14893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4894483685493469,
+      "learning_rate": 0.0005165456506605365,
+      "loss": 2.9985,
+      "step": 14894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4852317273616791,
+      "learning_rate": 0.0005165349949482962,
+      "loss": 3.2036,
+      "step": 14895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5104283094406128,
+      "learning_rate": 0.0005165243386657442,
+      "loss": 3.2489,
+      "step": 14896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5486336350440979,
+      "learning_rate": 0.000516513681812909,
+      "loss": 3.2546,
+      "step": 14897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5464646816253662,
+      "learning_rate": 0.0005165030243898183,
+      "loss": 3.1661,
+      "step": 14898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.494334876537323,
+      "learning_rate": 0.0005164923663965003,
+      "loss": 3.3491,
+      "step": 14899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51130211353302,
+      "learning_rate": 0.0005164817078329831,
+      "loss": 3.4166,
+      "step": 14900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5445241928100586,
+      "learning_rate": 0.0005164710486992947,
+      "loss": 3.2104,
+      "step": 14901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5161874294281006,
+      "learning_rate": 0.0005164603889954633,
+      "loss": 3.291,
+      "step": 14902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5364497303962708,
+      "learning_rate": 0.0005164497287215169,
+      "loss": 3.3265,
+      "step": 14903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4792921841144562,
+      "learning_rate": 0.0005164390678774834,
+      "loss": 3.2952,
+      "step": 14904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5327908396720886,
+      "learning_rate": 0.0005164284064633912,
+      "loss": 3.1207,
+      "step": 14905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5640168190002441,
+      "learning_rate": 0.0005164177444792683,
+      "loss": 3.2393,
+      "step": 14906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5080165863037109,
+      "learning_rate": 0.0005164070819251425,
+      "loss": 3.1347,
+      "step": 14907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4810307025909424,
+      "learning_rate": 0.0005163964188010424,
+      "loss": 3.2237,
+      "step": 14908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5038141012191772,
+      "learning_rate": 0.0005163857551069954,
+      "loss": 3.2509,
+      "step": 14909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5256493091583252,
+      "learning_rate": 0.0005163750908430303,
+      "loss": 3.1796,
+      "step": 14910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5214911699295044,
+      "learning_rate": 0.0005163644260091746,
+      "loss": 3.211,
+      "step": 14911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5438799858093262,
+      "learning_rate": 0.0005163537606054568,
+      "loss": 3.1117,
+      "step": 14912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5240724682807922,
+      "learning_rate": 0.0005163430946319047,
+      "loss": 3.3282,
+      "step": 14913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5152091979980469,
+      "learning_rate": 0.0005163324280885467,
+      "loss": 3.1758,
+      "step": 14914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4732866585254669,
+      "learning_rate": 0.0005163217609754105,
+      "loss": 3.0374,
+      "step": 14915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48373696208000183,
+      "learning_rate": 0.0005163110932925245,
+      "loss": 3.2967,
+      "step": 14916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5080105662345886,
+      "learning_rate": 0.0005163004250399168,
+      "loss": 3.1916,
+      "step": 14917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4790027141571045,
+      "learning_rate": 0.0005162897562176154,
+      "loss": 3.2397,
+      "step": 14918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.491432249546051,
+      "learning_rate": 0.0005162790868256484,
+      "loss": 3.2743,
+      "step": 14919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5296748876571655,
+      "learning_rate": 0.0005162684168640439,
+      "loss": 2.8877,
+      "step": 14920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5081189274787903,
+      "learning_rate": 0.00051625774633283,
+      "loss": 3.2914,
+      "step": 14921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5037716031074524,
+      "learning_rate": 0.0005162470752320347,
+      "loss": 3.2631,
+      "step": 14922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5493853688240051,
+      "learning_rate": 0.0005162364035616863,
+      "loss": 3.2527,
+      "step": 14923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48433980345726013,
+      "learning_rate": 0.0005162257313218128,
+      "loss": 3.1472,
+      "step": 14924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5239858627319336,
+      "learning_rate": 0.0005162150585124425,
+      "loss": 3.2365,
+      "step": 14925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5162671208381653,
+      "learning_rate": 0.0005162043851336032,
+      "loss": 3.1116,
+      "step": 14926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.532788097858429,
+      "learning_rate": 0.0005161937111853232,
+      "loss": 3.185,
+      "step": 14927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5161779522895813,
+      "learning_rate": 0.0005161830366676305,
+      "loss": 3.2383,
+      "step": 14928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5137925744056702,
+      "learning_rate": 0.0005161723615805534,
+      "loss": 3.2344,
+      "step": 14929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5238530039787292,
+      "learning_rate": 0.0005161616859241199,
+      "loss": 3.2037,
+      "step": 14930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5177023410797119,
+      "learning_rate": 0.0005161510096983581,
+      "loss": 3.035,
+      "step": 14931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49137774109840393,
+      "learning_rate": 0.0005161403329032961,
+      "loss": 3.2177,
+      "step": 14932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5207857489585876,
+      "learning_rate": 0.000516129655538962,
+      "loss": 3.2774,
+      "step": 14933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5206941962242126,
+      "learning_rate": 0.0005161189776053841,
+      "loss": 3.054,
+      "step": 14934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5176581144332886,
+      "learning_rate": 0.0005161082991025904,
+      "loss": 3.2821,
+      "step": 14935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5066308975219727,
+      "learning_rate": 0.000516097620030609,
+      "loss": 3.3941,
+      "step": 14936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5001108646392822,
+      "learning_rate": 0.000516086940389468,
+      "loss": 3.1497,
+      "step": 14937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5629680156707764,
+      "learning_rate": 0.0005160762601791956,
+      "loss": 3.4361,
+      "step": 14938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5518137216567993,
+      "learning_rate": 0.00051606557939982,
+      "loss": 3.2953,
+      "step": 14939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5315781831741333,
+      "learning_rate": 0.0005160548980513692,
+      "loss": 3.1599,
+      "step": 14940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4899667501449585,
+      "learning_rate": 0.0005160442161338713,
+      "loss": 3.2092,
+      "step": 14941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5403282046318054,
+      "learning_rate": 0.0005160335336473547,
+      "loss": 3.0352,
+      "step": 14942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5120472311973572,
+      "learning_rate": 0.0005160228505918472,
+      "loss": 3.3255,
+      "step": 14943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5195842385292053,
+      "learning_rate": 0.0005160121669673771,
+      "loss": 3.2938,
+      "step": 14944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5495728254318237,
+      "learning_rate": 0.0005160014827739726,
+      "loss": 3.2412,
+      "step": 14945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5019406080245972,
+      "learning_rate": 0.0005159907980116617,
+      "loss": 3.3542,
+      "step": 14946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5161232352256775,
+      "learning_rate": 0.0005159801126804726,
+      "loss": 3.064,
+      "step": 14947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4725062847137451,
+      "learning_rate": 0.0005159694267804335,
+      "loss": 2.9515,
+      "step": 14948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5322039127349854,
+      "learning_rate": 0.0005159587403115724,
+      "loss": 3.1031,
+      "step": 14949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48917528986930847,
+      "learning_rate": 0.0005159480532739176,
+      "loss": 3.3152,
+      "step": 14950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5044992566108704,
+      "learning_rate": 0.0005159373656674972,
+      "loss": 3.304,
+      "step": 14951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5023238658905029,
+      "learning_rate": 0.0005159266774923393,
+      "loss": 2.9814,
+      "step": 14952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49678921699523926,
+      "learning_rate": 0.0005159159887484721,
+      "loss": 3.2713,
+      "step": 14953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5398493409156799,
+      "learning_rate": 0.0005159052994359239,
+      "loss": 2.8908,
+      "step": 14954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5190320014953613,
+      "learning_rate": 0.0005158946095547225,
+      "loss": 3.1101,
+      "step": 14955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4837747812271118,
+      "learning_rate": 0.0005158839191048963,
+      "loss": 3.3732,
+      "step": 14956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5262941718101501,
+      "learning_rate": 0.0005158732280864735,
+      "loss": 3.095,
+      "step": 14957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5336409211158752,
+      "learning_rate": 0.000515862536499482,
+      "loss": 3.3913,
+      "step": 14958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5454764366149902,
+      "learning_rate": 0.0005158518443439502,
+      "loss": 3.3525,
+      "step": 14959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.536488950252533,
+      "learning_rate": 0.0005158411516199061,
+      "loss": 3.2649,
+      "step": 14960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5544682741165161,
+      "learning_rate": 0.000515830458327378,
+      "loss": 3.0752,
+      "step": 14961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5361686944961548,
+      "learning_rate": 0.0005158197644663941,
+      "loss": 3.163,
+      "step": 14962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5054564476013184,
+      "learning_rate": 0.0005158090700369824,
+      "loss": 3.0706,
+      "step": 14963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5609951019287109,
+      "learning_rate": 0.0005157983750391711,
+      "loss": 3.2753,
+      "step": 14964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5623591542243958,
+      "learning_rate": 0.0005157876794729885,
+      "loss": 3.1528,
+      "step": 14965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5499200224876404,
+      "learning_rate": 0.0005157769833384626,
+      "loss": 3.0385,
+      "step": 14966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47295448184013367,
+      "learning_rate": 0.0005157662866356217,
+      "loss": 3.0952,
+      "step": 14967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5192577838897705,
+      "learning_rate": 0.0005157555893644939,
+      "loss": 3.3865,
+      "step": 14968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51607346534729,
+      "learning_rate": 0.0005157448915251074,
+      "loss": 3.0292,
+      "step": 14969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5210020542144775,
+      "learning_rate": 0.0005157341931174904,
+      "loss": 3.3449,
+      "step": 14970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5266844034194946,
+      "learning_rate": 0.0005157234941416711,
+      "loss": 3.2106,
+      "step": 14971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6120948791503906,
+      "learning_rate": 0.0005157127945976776,
+      "loss": 2.9999,
+      "step": 14972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5157740116119385,
+      "learning_rate": 0.000515702094485538,
+      "loss": 3.1832,
+      "step": 14973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49761828780174255,
+      "learning_rate": 0.0005156913938052808,
+      "loss": 3.1624,
+      "step": 14974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5079160928726196,
+      "learning_rate": 0.0005156806925569338,
+      "loss": 3.2627,
+      "step": 14975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5453952550888062,
+      "learning_rate": 0.0005156699907405255,
+      "loss": 2.9036,
+      "step": 14976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5448424220085144,
+      "learning_rate": 0.0005156592883560838,
+      "loss": 3.1669,
+      "step": 14977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49395671486854553,
+      "learning_rate": 0.0005156485854036371,
+      "loss": 3.2462,
+      "step": 14978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5351203680038452,
+      "learning_rate": 0.0005156378818832136,
+      "loss": 3.3148,
+      "step": 14979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5157082080841064,
+      "learning_rate": 0.0005156271777948414,
+      "loss": 3.3561,
+      "step": 14980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5183131098747253,
+      "learning_rate": 0.0005156164731385487,
+      "loss": 3.1139,
+      "step": 14981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5393839478492737,
+      "learning_rate": 0.0005156057679143636,
+      "loss": 3.2187,
+      "step": 14982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49937084317207336,
+      "learning_rate": 0.0005155950621223145,
+      "loss": 3.1017,
+      "step": 14983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5061917901039124,
+      "learning_rate": 0.0005155843557624296,
+      "loss": 3.0941,
+      "step": 14984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5656406283378601,
+      "learning_rate": 0.0005155736488347369,
+      "loss": 3.0482,
+      "step": 14985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5063902735710144,
+      "learning_rate": 0.0005155629413392646,
+      "loss": 3.0954,
+      "step": 14986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5024520754814148,
+      "learning_rate": 0.0005155522332760411,
+      "loss": 3.0767,
+      "step": 14987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.83547043800354,
+      "learning_rate": 0.0005155415246450945,
+      "loss": 3.0165,
+      "step": 14988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5294227600097656,
+      "learning_rate": 0.000515530815446453,
+      "loss": 3.2607,
+      "step": 14989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5271424651145935,
+      "learning_rate": 0.0005155201056801449,
+      "loss": 3.1893,
+      "step": 14990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5043739080429077,
+      "learning_rate": 0.0005155093953461981,
+      "loss": 3.1544,
+      "step": 14991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5016458034515381,
+      "learning_rate": 0.0005154986844446411,
+      "loss": 3.099,
+      "step": 14992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234816670417786,
+      "learning_rate": 0.0005154879729755022,
+      "loss": 3.2483,
+      "step": 14993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5132349133491516,
+      "learning_rate": 0.0005154772609388093,
+      "loss": 3.215,
+      "step": 14994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5096055865287781,
+      "learning_rate": 0.0005154665483345909,
+      "loss": 3.289,
+      "step": 14995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5225401520729065,
+      "learning_rate": 0.0005154558351628749,
+      "loss": 3.285,
+      "step": 14996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5399203300476074,
+      "learning_rate": 0.0005154451214236898,
+      "loss": 3.0405,
+      "step": 14997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5019733905792236,
+      "learning_rate": 0.0005154344071170637,
+      "loss": 3.2109,
+      "step": 14998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5068507790565491,
+      "learning_rate": 0.0005154236922430248,
+      "loss": 3.0886,
+      "step": 14999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5227283835411072,
+      "learning_rate": 0.0005154129768016013,
+      "loss": 3.0491,
+      "step": 15000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5454354882240295,
+      "learning_rate": 0.0005154022607928217,
+      "loss": 3.2926,
+      "step": 15001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5321445465087891,
+      "learning_rate": 0.0005153915442167138,
+      "loss": 3.1433,
+      "step": 15002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5160802006721497,
+      "learning_rate": 0.0005153808270733061,
+      "loss": 3.0392,
+      "step": 15003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5624576210975647,
+      "learning_rate": 0.0005153701093626267,
+      "loss": 3.2465,
+      "step": 15004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5116503238677979,
+      "learning_rate": 0.0005153593910847039,
+      "loss": 3.4115,
+      "step": 15005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4815455377101898,
+      "learning_rate": 0.000515348672239566,
+      "loss": 3.2247,
+      "step": 15006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5008863806724548,
+      "learning_rate": 0.0005153379528272411,
+      "loss": 2.9155,
+      "step": 15007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5305734276771545,
+      "learning_rate": 0.0005153272328477573,
+      "loss": 3.3725,
+      "step": 15008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49309682846069336,
+      "learning_rate": 0.0005153165123011432,
+      "loss": 3.2633,
+      "step": 15009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5189629793167114,
+      "learning_rate": 0.0005153057911874267,
+      "loss": 3.1755,
+      "step": 15010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5381687879562378,
+      "learning_rate": 0.0005152950695066364,
+      "loss": 3.222,
+      "step": 15011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4915025234222412,
+      "learning_rate": 0.0005152843472588001,
+      "loss": 3.1748,
+      "step": 15012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5047193169593811,
+      "learning_rate": 0.0005152736244439464,
+      "loss": 3.1085,
+      "step": 15013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5417593717575073,
+      "learning_rate": 0.0005152629010621033,
+      "loss": 3.1649,
+      "step": 15014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5217851400375366,
+      "learning_rate": 0.0005152521771132993,
+      "loss": 3.2755,
+      "step": 15015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.517413854598999,
+      "learning_rate": 0.0005152414525975625,
+      "loss": 3.0616,
+      "step": 15016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5072711110115051,
+      "learning_rate": 0.0005152307275149209,
+      "loss": 3.0893,
+      "step": 15017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5205687284469604,
+      "learning_rate": 0.0005152200018654032,
+      "loss": 3.1728,
+      "step": 15018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5183899402618408,
+      "learning_rate": 0.0005152092756490373,
+      "loss": 3.3177,
+      "step": 15019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5392817854881287,
+      "learning_rate": 0.0005151985488658517,
+      "loss": 3.2724,
+      "step": 15020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49955472350120544,
+      "learning_rate": 0.0005151878215158745,
+      "loss": 3.1084,
+      "step": 15021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5460142493247986,
+      "learning_rate": 0.0005151770935991339,
+      "loss": 3.2581,
+      "step": 15022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5164291262626648,
+      "learning_rate": 0.0005151663651156584,
+      "loss": 3.2137,
+      "step": 15023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4972623586654663,
+      "learning_rate": 0.000515155636065476,
+      "loss": 3.1696,
+      "step": 15024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5117320418357849,
+      "learning_rate": 0.0005151449064486151,
+      "loss": 2.9165,
+      "step": 15025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5329243540763855,
+      "learning_rate": 0.000515134176265104,
+      "loss": 3.2211,
+      "step": 15026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.548600971698761,
+      "learning_rate": 0.0005151234455149707,
+      "loss": 3.4036,
+      "step": 15027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5310418009757996,
+      "learning_rate": 0.0005151127141982437,
+      "loss": 3.1994,
+      "step": 15028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5383056998252869,
+      "learning_rate": 0.0005151019823149513,
+      "loss": 3.2994,
+      "step": 15029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4978289008140564,
+      "learning_rate": 0.0005150912498651216,
+      "loss": 3.2983,
+      "step": 15030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5403746366500854,
+      "learning_rate": 0.000515080516848783,
+      "loss": 3.1833,
+      "step": 15031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5366250276565552,
+      "learning_rate": 0.0005150697832659636,
+      "loss": 3.0881,
+      "step": 15032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49605393409729004,
+      "learning_rate": 0.0005150590491166919,
+      "loss": 3.3317,
+      "step": 15033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5022369027137756,
+      "learning_rate": 0.0005150483144009961,
+      "loss": 3.1123,
+      "step": 15034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5486195683479309,
+      "learning_rate": 0.0005150375791189043,
+      "loss": 3.368,
+      "step": 15035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.511721670627594,
+      "learning_rate": 0.0005150268432704449,
+      "loss": 3.1523,
+      "step": 15036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5076295137405396,
+      "learning_rate": 0.0005150161068556463,
+      "loss": 3.0981,
+      "step": 15037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5491887331008911,
+      "learning_rate": 0.0005150053698745365,
+      "loss": 3.1993,
+      "step": 15038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5024232268333435,
+      "learning_rate": 0.000514994632327144,
+      "loss": 3.4245,
+      "step": 15039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5337858200073242,
+      "learning_rate": 0.0005149838942134971,
+      "loss": 3.2517,
+      "step": 15040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5012475252151489,
+      "learning_rate": 0.0005149731555336239,
+      "loss": 3.2314,
+      "step": 15041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5080512166023254,
+      "learning_rate": 0.0005149624162875528,
+      "loss": 3.432,
+      "step": 15042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5350621938705444,
+      "learning_rate": 0.0005149516764753121,
+      "loss": 3.232,
+      "step": 15043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5167108774185181,
+      "learning_rate": 0.00051494093609693,
+      "loss": 3.1149,
+      "step": 15044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5356556177139282,
+      "learning_rate": 0.0005149301951524348,
+      "loss": 3.3853,
+      "step": 15045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48991358280181885,
+      "learning_rate": 0.0005149194536418551,
+      "loss": 3.0879,
+      "step": 15046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5196684002876282,
+      "learning_rate": 0.0005149087115652185,
+      "loss": 2.9087,
+      "step": 15047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4956427216529846,
+      "learning_rate": 0.000514897968922554,
+      "loss": 3.2385,
+      "step": 15048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48819366097450256,
+      "learning_rate": 0.0005148872257138895,
+      "loss": 3.2224,
+      "step": 15049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5472261905670166,
+      "learning_rate": 0.0005148764819392535,
+      "loss": 3.1346,
+      "step": 15050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5290123820304871,
+      "learning_rate": 0.0005148657375986741,
+      "loss": 3.1584,
+      "step": 15051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5149974226951599,
+      "learning_rate": 0.0005148549926921798,
+      "loss": 3.3616,
+      "step": 15052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5451234579086304,
+      "learning_rate": 0.0005148442472197986,
+      "loss": 3.145,
+      "step": 15053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4977348744869232,
+      "learning_rate": 0.0005148335011815592,
+      "loss": 3.1043,
+      "step": 15054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4974806010723114,
+      "learning_rate": 0.0005148227545774898,
+      "loss": 3.2151,
+      "step": 15055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5258920192718506,
+      "learning_rate": 0.0005148120074076184,
+      "loss": 3.2716,
+      "step": 15056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5211778879165649,
+      "learning_rate": 0.0005148012596719735,
+      "loss": 3.1984,
+      "step": 15057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5030432939529419,
+      "learning_rate": 0.0005147905113705835,
+      "loss": 3.1843,
+      "step": 15058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4971194267272949,
+      "learning_rate": 0.0005147797625034766,
+      "loss": 3.1692,
+      "step": 15059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5721365213394165,
+      "learning_rate": 0.0005147690130706811,
+      "loss": 3.0595,
+      "step": 15060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5956337451934814,
+      "learning_rate": 0.0005147582630722255,
+      "loss": 3.2497,
+      "step": 15061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5455291271209717,
+      "learning_rate": 0.0005147475125081379,
+      "loss": 3.286,
+      "step": 15062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49531859159469604,
+      "learning_rate": 0.0005147367613784465,
+      "loss": 3.2845,
+      "step": 15063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5478885769844055,
+      "learning_rate": 0.00051472600968318,
+      "loss": 3.4411,
+      "step": 15064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5322989821434021,
+      "learning_rate": 0.0005147152574223665,
+      "loss": 3.4172,
+      "step": 15065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5252341628074646,
+      "learning_rate": 0.0005147045045960344,
+      "loss": 3.2743,
+      "step": 15066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5184382796287537,
+      "learning_rate": 0.0005146937512042118,
+      "loss": 3.1589,
+      "step": 15067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5355132818222046,
+      "learning_rate": 0.0005146829972469272,
+      "loss": 3.0857,
+      "step": 15068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5309262275695801,
+      "learning_rate": 0.000514672242724209,
+      "loss": 3.3667,
+      "step": 15069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.529289186000824,
+      "learning_rate": 0.0005146614876360853,
+      "loss": 3.4135,
+      "step": 15070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48815011978149414,
+      "learning_rate": 0.0005146507319825846,
+      "loss": 3.4187,
+      "step": 15071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.526040256023407,
+      "learning_rate": 0.0005146399757637352,
+      "loss": 3.1328,
+      "step": 15072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5177284479141235,
+      "learning_rate": 0.0005146292189795654,
+      "loss": 3.2615,
+      "step": 15073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4890044033527374,
+      "learning_rate": 0.0005146184616301036,
+      "loss": 3.0297,
+      "step": 15074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.511398434638977,
+      "learning_rate": 0.000514607703715378,
+      "loss": 3.5007,
+      "step": 15075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5235434174537659,
+      "learning_rate": 0.000514596945235417,
+      "loss": 3.3912,
+      "step": 15076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5152917504310608,
+      "learning_rate": 0.0005145861861902491,
+      "loss": 3.0959,
+      "step": 15077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.555122971534729,
+      "learning_rate": 0.0005145754265799023,
+      "loss": 3.1187,
+      "step": 15078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5248861908912659,
+      "learning_rate": 0.0005145646664044053,
+      "loss": 3.3088,
+      "step": 15079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48319074511528015,
+      "learning_rate": 0.0005145539056637861,
+      "loss": 3.0406,
+      "step": 15080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5107579827308655,
+      "learning_rate": 0.0005145431443580732,
+      "loss": 2.9253,
+      "step": 15081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5582626461982727,
+      "learning_rate": 0.000514532382487295,
+      "loss": 3.3489,
+      "step": 15082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5222228765487671,
+      "learning_rate": 0.00051452162005148,
+      "loss": 3.2836,
+      "step": 15083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49412620067596436,
+      "learning_rate": 0.0005145108570506561,
+      "loss": 3.163,
+      "step": 15084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.530225932598114,
+      "learning_rate": 0.0005145000934848519,
+      "loss": 3.144,
+      "step": 15085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5154251456260681,
+      "learning_rate": 0.0005144893293540957,
+      "loss": 3.0688,
+      "step": 15086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5582703948020935,
+      "learning_rate": 0.0005144785646584159,
+      "loss": 3.3622,
+      "step": 15087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5435081720352173,
+      "learning_rate": 0.000514467799397841,
+      "loss": 3.1106,
+      "step": 15088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5430107712745667,
+      "learning_rate": 0.000514457033572399,
+      "loss": 3.0035,
+      "step": 15089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.543735682964325,
+      "learning_rate": 0.0005144462671821186,
+      "loss": 3.1732,
+      "step": 15090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5661927461624146,
+      "learning_rate": 0.0005144355002270278,
+      "loss": 2.9093,
+      "step": 15091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5020666122436523,
+      "learning_rate": 0.0005144247327071553,
+      "loss": 3.2608,
+      "step": 15092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5958285331726074,
+      "learning_rate": 0.0005144139646225293,
+      "loss": 3.0192,
+      "step": 15093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5555652976036072,
+      "learning_rate": 0.0005144031959731783,
+      "loss": 3.2023,
+      "step": 15094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5096257925033569,
+      "learning_rate": 0.0005143924267591304,
+      "loss": 3.2049,
+      "step": 15095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5613263249397278,
+      "learning_rate": 0.0005143816569804141,
+      "loss": 3.074,
+      "step": 15096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4865977168083191,
+      "learning_rate": 0.0005143708866370579,
+      "loss": 3.0918,
+      "step": 15097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.509009838104248,
+      "learning_rate": 0.00051436011572909,
+      "loss": 3.2596,
+      "step": 15098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5192385315895081,
+      "learning_rate": 0.0005143493442565387,
+      "loss": 3.0407,
+      "step": 15099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5306475758552551,
+      "learning_rate": 0.0005143385722194326,
+      "loss": 3.2282,
+      "step": 15100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5092197060585022,
+      "learning_rate": 0.0005143277996177998,
+      "loss": 3.2562,
+      "step": 15101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49553215503692627,
+      "learning_rate": 0.000514317026451669,
+      "loss": 3.114,
+      "step": 15102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5059915781021118,
+      "learning_rate": 0.0005143062527210683,
+      "loss": 3.3094,
+      "step": 15103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4825346767902374,
+      "learning_rate": 0.0005142954784260261,
+      "loss": 3.2117,
+      "step": 15104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5263687968254089,
+      "learning_rate": 0.000514284703566571,
+      "loss": 3.0489,
+      "step": 15105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.50020432472229,
+      "learning_rate": 0.0005142739281427313,
+      "loss": 3.2894,
+      "step": 15106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5188915133476257,
+      "learning_rate": 0.0005142631521545351,
+      "loss": 3.2912,
+      "step": 15107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5241820812225342,
+      "learning_rate": 0.0005142523756020111,
+      "loss": 3.1568,
+      "step": 15108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47457194328308105,
+      "learning_rate": 0.0005142415984851875,
+      "loss": 3.2036,
+      "step": 15109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46626847982406616,
+      "learning_rate": 0.0005142308208040928,
+      "loss": 3.0613,
+      "step": 15110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5026150941848755,
+      "learning_rate": 0.0005142200425587555,
+      "loss": 3.0074,
+      "step": 15111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.531330406665802,
+      "learning_rate": 0.0005142092637492036,
+      "loss": 3.3118,
+      "step": 15112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5424619913101196,
+      "learning_rate": 0.0005141984843754658,
+      "loss": 3.2401,
+      "step": 15113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5372337102890015,
+      "learning_rate": 0.0005141877044375705,
+      "loss": 3.2051,
+      "step": 15114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5131272673606873,
+      "learning_rate": 0.000514176923935546,
+      "loss": 3.0717,
+      "step": 15115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4956575334072113,
+      "learning_rate": 0.0005141661428694206,
+      "loss": 3.1742,
+      "step": 15116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5434887409210205,
+      "learning_rate": 0.0005141553612392229,
+      "loss": 3.1218,
+      "step": 15117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5086133480072021,
+      "learning_rate": 0.0005141445790449811,
+      "loss": 3.0653,
+      "step": 15118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5095211863517761,
+      "learning_rate": 0.0005141337962867238,
+      "loss": 3.2132,
+      "step": 15119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5019669532775879,
+      "learning_rate": 0.0005141230129644792,
+      "loss": 3.1334,
+      "step": 15120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5907220244407654,
+      "learning_rate": 0.0005141122290782758,
+      "loss": 3.0343,
+      "step": 15121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.531029462814331,
+      "learning_rate": 0.000514101444628142,
+      "loss": 3.2226,
+      "step": 15122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5512033104896545,
+      "learning_rate": 0.0005140906596141063,
+      "loss": 3.3203,
+      "step": 15123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5198848843574524,
+      "learning_rate": 0.0005140798740361968,
+      "loss": 3.1245,
+      "step": 15124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5568894147872925,
+      "learning_rate": 0.0005140690878944423,
+      "loss": 3.2085,
+      "step": 15125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5185664296150208,
+      "learning_rate": 0.0005140583011888709,
+      "loss": 3.3906,
+      "step": 15126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5689828991889954,
+      "learning_rate": 0.0005140475139195112,
+      "loss": 3.0878,
+      "step": 15127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5163256525993347,
+      "learning_rate": 0.0005140367260863916,
+      "loss": 3.3418,
+      "step": 15128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.487690269947052,
+      "learning_rate": 0.0005140259376895404,
+      "loss": 3.4939,
+      "step": 15129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5142568349838257,
+      "learning_rate": 0.000514015148728986,
+      "loss": 3.1635,
+      "step": 15130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546401739120483,
+      "learning_rate": 0.000514004359204757,
+      "loss": 3.1504,
+      "step": 15131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5494998693466187,
+      "learning_rate": 0.0005139935691168816,
+      "loss": 3.1133,
+      "step": 15132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5090645551681519,
+      "learning_rate": 0.0005139827784653884,
+      "loss": 3.1497,
+      "step": 15133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5398714542388916,
+      "learning_rate": 0.0005139719872503057,
+      "loss": 3.2604,
+      "step": 15134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5296069383621216,
+      "learning_rate": 0.0005139611954716619,
+      "loss": 3.4238,
+      "step": 15135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.507793664932251,
+      "learning_rate": 0.0005139504031294855,
+      "loss": 3.1279,
+      "step": 15136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5212898850440979,
+      "learning_rate": 0.000513939610223805,
+      "loss": 3.2929,
+      "step": 15137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5241724848747253,
+      "learning_rate": 0.0005139288167546487,
+      "loss": 3.2657,
+      "step": 15138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5247151255607605,
+      "learning_rate": 0.0005139180227220451,
+      "loss": 3.2378,
+      "step": 15139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4941554665565491,
+      "learning_rate": 0.0005139072281260226,
+      "loss": 3.0256,
+      "step": 15140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5236513018608093,
+      "learning_rate": 0.0005138964329666096,
+      "loss": 3.1188,
+      "step": 15141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5333306789398193,
+      "learning_rate": 0.0005138856372438347,
+      "loss": 3.3072,
+      "step": 15142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5609138607978821,
+      "learning_rate": 0.000513874840957726,
+      "loss": 3.1602,
+      "step": 15143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49884551763534546,
+      "learning_rate": 0.0005138640441083122,
+      "loss": 3.1501,
+      "step": 15144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5260482430458069,
+      "learning_rate": 0.0005138532466956216,
+      "loss": 3.3145,
+      "step": 15145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.543743908405304,
+      "learning_rate": 0.0005138424487196829,
+      "loss": 3.0801,
+      "step": 15146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5195195078849792,
+      "learning_rate": 0.0005138316501805242,
+      "loss": 3.2929,
+      "step": 15147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5003941655158997,
+      "learning_rate": 0.0005138208510781741,
+      "loss": 3.1971,
+      "step": 15148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5132508873939514,
+      "learning_rate": 0.000513810051412661,
+      "loss": 3.271,
+      "step": 15149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5287436246871948,
+      "learning_rate": 0.0005137992511840134,
+      "loss": 3.1374,
+      "step": 15150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5131794810295105,
+      "learning_rate": 0.0005137884503922597,
+      "loss": 3.1865,
+      "step": 15151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4891071021556854,
+      "learning_rate": 0.0005137776490374284,
+      "loss": 3.1813,
+      "step": 15152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.518698513507843,
+      "learning_rate": 0.0005137668471195478,
+      "loss": 2.9667,
+      "step": 15153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5069079995155334,
+      "learning_rate": 0.0005137560446386466,
+      "loss": 3.1667,
+      "step": 15154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5019563436508179,
+      "learning_rate": 0.0005137452415947531,
+      "loss": 3.048,
+      "step": 15155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5977448225021362,
+      "learning_rate": 0.0005137344379878958,
+      "loss": 3.0489,
+      "step": 15156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.526675283908844,
+      "learning_rate": 0.000513723633818103,
+      "loss": 3.2784,
+      "step": 15157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5488605499267578,
+      "learning_rate": 0.0005137128290854035,
+      "loss": 3.3835,
+      "step": 15158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5367397665977478,
+      "learning_rate": 0.0005137020237898254,
+      "loss": 3.3152,
+      "step": 15159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5273241400718689,
+      "learning_rate": 0.0005136912179313973,
+      "loss": 2.9809,
+      "step": 15160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.555223822593689,
+      "learning_rate": 0.0005136804115101475,
+      "loss": 3.1469,
+      "step": 15161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.533078670501709,
+      "learning_rate": 0.0005136696045261049,
+      "loss": 3.1378,
+      "step": 15162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5096789598464966,
+      "learning_rate": 0.0005136587969792975,
+      "loss": 3.2697,
+      "step": 15163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4871804416179657,
+      "learning_rate": 0.000513647988869754,
+      "loss": 3.1497,
+      "step": 15164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47410327196121216,
+      "learning_rate": 0.0005136371801975028,
+      "loss": 3.0985,
+      "step": 15165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5203956961631775,
+      "learning_rate": 0.0005136263709625724,
+      "loss": 3.1768,
+      "step": 15166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5182998180389404,
+      "learning_rate": 0.0005136155611649912,
+      "loss": 3.0054,
+      "step": 15167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5807198882102966,
+      "learning_rate": 0.0005136047508047879,
+      "loss": 3.1998,
+      "step": 15168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49342507123947144,
+      "learning_rate": 0.0005135939398819906,
+      "loss": 3.1338,
+      "step": 15169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5400848984718323,
+      "learning_rate": 0.000513583128396628,
+      "loss": 3.1402,
+      "step": 15170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7351841926574707,
+      "learning_rate": 0.0005135723163487286,
+      "loss": 3.0862,
+      "step": 15171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.601331353187561,
+      "learning_rate": 0.0005135615037383209,
+      "loss": 3.1345,
+      "step": 15172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5336458086967468,
+      "learning_rate": 0.0005135506905654331,
+      "loss": 3.0461,
+      "step": 15173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5114117860794067,
+      "learning_rate": 0.000513539876830094,
+      "loss": 3.3398,
+      "step": 15174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.521041750907898,
+      "learning_rate": 0.0005135290625323319,
+      "loss": 3.2127,
+      "step": 15175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5434523224830627,
+      "learning_rate": 0.0005135182476721754,
+      "loss": 3.2585,
+      "step": 15176
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.556186854839325,
+      "learning_rate": 0.0005135074322496529,
+      "loss": 3.1948,
+      "step": 15177
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.540410041809082,
+      "learning_rate": 0.0005134966162647929,
+      "loss": 2.9405,
+      "step": 15178
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.540989100933075,
+      "learning_rate": 0.0005134857997176241,
+      "loss": 3.3179,
+      "step": 15179
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5502685308456421,
+      "learning_rate": 0.0005134749826081745,
+      "loss": 2.9779,
+      "step": 15180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5441862344741821,
+      "learning_rate": 0.000513464164936473,
+      "loss": 3.2497,
+      "step": 15181
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5529954433441162,
+      "learning_rate": 0.0005134533467025479,
+      "loss": 3.1787,
+      "step": 15182
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5215994715690613,
+      "learning_rate": 0.0005134425279064279,
+      "loss": 3.2406,
+      "step": 15183
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4941001832485199,
+      "learning_rate": 0.0005134317085481413,
+      "loss": 2.8891,
+      "step": 15184
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5015299320220947,
+      "learning_rate": 0.0005134208886277167,
+      "loss": 3.1398,
+      "step": 15185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4704887866973877,
+      "learning_rate": 0.0005134100681451825,
+      "loss": 3.2913,
+      "step": 15186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4975797235965729,
+      "learning_rate": 0.0005133992471005672,
+      "loss": 3.055,
+      "step": 15187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5469768047332764,
+      "learning_rate": 0.0005133884254938994,
+      "loss": 3.1481,
+      "step": 15188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5250169634819031,
+      "learning_rate": 0.0005133776033252076,
+      "loss": 3.2758,
+      "step": 15189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5461866855621338,
+      "learning_rate": 0.0005133667805945202,
+      "loss": 3.2773,
+      "step": 15190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.501222550868988,
+      "learning_rate": 0.0005133559573018658,
+      "loss": 3.1817,
+      "step": 15191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5163980722427368,
+      "learning_rate": 0.0005133451334472729,
+      "loss": 3.3022,
+      "step": 15192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49753910303115845,
+      "learning_rate": 0.0005133343090307699,
+      "loss": 3.3661,
+      "step": 15193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5362524390220642,
+      "learning_rate": 0.0005133234840523854,
+      "loss": 3.2704,
+      "step": 15194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5319569110870361,
+      "learning_rate": 0.000513312658512148,
+      "loss": 3.2043,
+      "step": 15195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5191274881362915,
+      "learning_rate": 0.0005133018324100859,
+      "loss": 3.3014,
+      "step": 15196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6559003591537476,
+      "learning_rate": 0.000513291005746228,
+      "loss": 3.253,
+      "step": 15197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5224505066871643,
+      "learning_rate": 0.0005132801785206026,
+      "loss": 3.2519,
+      "step": 15198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5321853756904602,
+      "learning_rate": 0.0005132693507332383,
+      "loss": 2.9671,
+      "step": 15199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5000855326652527,
+      "learning_rate": 0.0005132585223841635,
+      "loss": 3.1668,
+      "step": 15200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5224664211273193,
+      "learning_rate": 0.0005132476934734068,
+      "loss": 3.2821,
+      "step": 15201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4886454641819,
+      "learning_rate": 0.0005132368640009968,
+      "loss": 3.0804,
+      "step": 15202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5113134384155273,
+      "learning_rate": 0.0005132260339669618,
+      "loss": 3.0912,
+      "step": 15203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5165402293205261,
+      "learning_rate": 0.0005132152033713305,
+      "loss": 3.1487,
+      "step": 15204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5087444186210632,
+      "learning_rate": 0.0005132043722141314,
+      "loss": 3.0623,
+      "step": 15205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49701055884361267,
+      "learning_rate": 0.0005131935404953931,
+      "loss": 3.1004,
+      "step": 15206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5303621888160706,
+      "learning_rate": 0.000513182708215144,
+      "loss": 3.1908,
+      "step": 15207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48327744007110596,
+      "learning_rate": 0.0005131718753734127,
+      "loss": 3.1967,
+      "step": 15208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5533356666564941,
+      "learning_rate": 0.0005131610419702276,
+      "loss": 3.2278,
+      "step": 15209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5282249450683594,
+      "learning_rate": 0.0005131502080056174,
+      "loss": 3.2681,
+      "step": 15210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5078888535499573,
+      "learning_rate": 0.0005131393734796106,
+      "loss": 3.3327,
+      "step": 15211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49346816539764404,
+      "learning_rate": 0.0005131285383922357,
+      "loss": 3.1853,
+      "step": 15212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5162724256515503,
+      "learning_rate": 0.0005131177027435213,
+      "loss": 3.0953,
+      "step": 15213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5229780673980713,
+      "learning_rate": 0.0005131068665334957,
+      "loss": 3.0968,
+      "step": 15214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5405413508415222,
+      "learning_rate": 0.0005130960297621877,
+      "loss": 2.9953,
+      "step": 15215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5310748219490051,
+      "learning_rate": 0.000513085192429626,
+      "loss": 3.1261,
+      "step": 15216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5561037659645081,
+      "learning_rate": 0.0005130743545358386,
+      "loss": 3.3495,
+      "step": 15217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5247962474822998,
+      "learning_rate": 0.0005130635160808545,
+      "loss": 3.3857,
+      "step": 15218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5569313168525696,
+      "learning_rate": 0.0005130526770647021,
+      "loss": 3.169,
+      "step": 15219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.516389012336731,
+      "learning_rate": 0.0005130418374874099,
+      "loss": 3.2518,
+      "step": 15220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5360453128814697,
+      "learning_rate": 0.0005130309973490066,
+      "loss": 3.1335,
+      "step": 15221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4962994158267975,
+      "learning_rate": 0.0005130201566495206,
+      "loss": 3.3244,
+      "step": 15222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5042166113853455,
+      "learning_rate": 0.0005130093153889805,
+      "loss": 3.1213,
+      "step": 15223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.504375696182251,
+      "learning_rate": 0.0005129984735674149,
+      "loss": 3.1488,
+      "step": 15224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5094754695892334,
+      "learning_rate": 0.0005129876311848522,
+      "loss": 3.4237,
+      "step": 15225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49719002842903137,
+      "learning_rate": 0.0005129767882413211,
+      "loss": 3.2614,
+      "step": 15226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5632519721984863,
+      "learning_rate": 0.0005129659447368502,
+      "loss": 3.1523,
+      "step": 15227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5078555345535278,
+      "learning_rate": 0.0005129551006714678,
+      "loss": 3.2035,
+      "step": 15228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5147547721862793,
+      "learning_rate": 0.0005129442560452029,
+      "loss": 3.3097,
+      "step": 15229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5100898742675781,
+      "learning_rate": 0.0005129334108580837,
+      "loss": 3.2415,
+      "step": 15230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49058064818382263,
+      "learning_rate": 0.0005129225651101389,
+      "loss": 3.1857,
+      "step": 15231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4862991273403168,
+      "learning_rate": 0.0005129117188013971,
+      "loss": 3.4175,
+      "step": 15232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5109384059906006,
+      "learning_rate": 0.0005129008719318867,
+      "loss": 3.3401,
+      "step": 15233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5978837013244629,
+      "learning_rate": 0.0005128900245016365,
+      "loss": 2.9597,
+      "step": 15234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5241730809211731,
+      "learning_rate": 0.0005128791765106747,
+      "loss": 3.3489,
+      "step": 15235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.544468104839325,
+      "learning_rate": 0.0005128683279590304,
+      "loss": 3.0905,
+      "step": 15236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5095611214637756,
+      "learning_rate": 0.0005128574788467317,
+      "loss": 3.2927,
+      "step": 15237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6200950741767883,
+      "learning_rate": 0.0005128466291738074,
+      "loss": 3.3109,
+      "step": 15238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5368150472640991,
+      "learning_rate": 0.0005128357789402861,
+      "loss": 3.004,
+      "step": 15239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49415868520736694,
+      "learning_rate": 0.0005128249281461963,
+      "loss": 3.2525,
+      "step": 15240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5034611821174622,
+      "learning_rate": 0.0005128140767915666,
+      "loss": 3.0394,
+      "step": 15241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5524153709411621,
+      "learning_rate": 0.0005128032248764256,
+      "loss": 3.0271,
+      "step": 15242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.537729024887085,
+      "learning_rate": 0.0005127923724008018,
+      "loss": 3.3173,
+      "step": 15243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5131357908248901,
+      "learning_rate": 0.0005127815193647239,
+      "loss": 2.9923,
+      "step": 15244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546363353729248,
+      "learning_rate": 0.0005127706657682205,
+      "loss": 2.9208,
+      "step": 15245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.565846860408783,
+      "learning_rate": 0.00051275981161132,
+      "loss": 3.3611,
+      "step": 15246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5199288129806519,
+      "learning_rate": 0.0005127489568940511,
+      "loss": 3.185,
+      "step": 15247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5325568318367004,
+      "learning_rate": 0.0005127381016164425,
+      "loss": 3.0524,
+      "step": 15248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5222455263137817,
+      "learning_rate": 0.0005127272457785225,
+      "loss": 3.0541,
+      "step": 15249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5106666684150696,
+      "learning_rate": 0.00051271638938032,
+      "loss": 3.2104,
+      "step": 15250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5324909090995789,
+      "learning_rate": 0.0005127055324218635,
+      "loss": 3.2489,
+      "step": 15251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5093536376953125,
+      "learning_rate": 0.0005126946749031814,
+      "loss": 3.1664,
+      "step": 15252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5190713405609131,
+      "learning_rate": 0.0005126838168243026,
+      "loss": 3.1126,
+      "step": 15253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6075634360313416,
+      "learning_rate": 0.0005126729581852556,
+      "loss": 3.4352,
+      "step": 15254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5500290393829346,
+      "learning_rate": 0.0005126620989860688,
+      "loss": 3.1548,
+      "step": 15255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5063414573669434,
+      "learning_rate": 0.000512651239226771,
+      "loss": 3.2785,
+      "step": 15256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5023425817489624,
+      "learning_rate": 0.0005126403789073909,
+      "loss": 3.2736,
+      "step": 15257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5160260796546936,
+      "learning_rate": 0.0005126295180279568,
+      "loss": 3.3132,
+      "step": 15258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.50048828125,
+      "learning_rate": 0.0005126186565884975,
+      "loss": 3.1487,
+      "step": 15259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49590444564819336,
+      "learning_rate": 0.0005126077945890417,
+      "loss": 3.2356,
+      "step": 15260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5084480047225952,
+      "learning_rate": 0.0005125969320296178,
+      "loss": 3.2633,
+      "step": 15261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5234677195549011,
+      "learning_rate": 0.0005125860689102544,
+      "loss": 3.0592,
+      "step": 15262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5146345496177673,
+      "learning_rate": 0.0005125752052309805,
+      "loss": 3.3137,
+      "step": 15263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5760210752487183,
+      "learning_rate": 0.0005125643409918242,
+      "loss": 3.3984,
+      "step": 15264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5140902996063232,
+      "learning_rate": 0.0005125534761928144,
+      "loss": 3.307,
+      "step": 15265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5068773031234741,
+      "learning_rate": 0.0005125426108339795,
+      "loss": 3.0952,
+      "step": 15266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5212466716766357,
+      "learning_rate": 0.0005125317449153484,
+      "loss": 3.1379,
+      "step": 15267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5364983677864075,
+      "learning_rate": 0.0005125208784369495,
+      "loss": 3.2761,
+      "step": 15268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4984794557094574,
+      "learning_rate": 0.0005125100113988117,
+      "loss": 3.1004,
+      "step": 15269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5525820851325989,
+      "learning_rate": 0.0005124991438009632,
+      "loss": 3.2194,
+      "step": 15270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5691623091697693,
+      "learning_rate": 0.000512488275643433,
+      "loss": 3.1136,
+      "step": 15271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5180755853652954,
+      "learning_rate": 0.0005124774069262494,
+      "loss": 3.2401,
+      "step": 15272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5265724062919617,
+      "learning_rate": 0.0005124665376494414,
+      "loss": 3.2386,
+      "step": 15273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48479267954826355,
+      "learning_rate": 0.0005124556678130374,
+      "loss": 3.1576,
+      "step": 15274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5298726558685303,
+      "learning_rate": 0.000512444797417066,
+      "loss": 2.8294,
+      "step": 15275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.509896457195282,
+      "learning_rate": 0.000512433926461556,
+      "loss": 3.0053,
+      "step": 15276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5217825174331665,
+      "learning_rate": 0.0005124230549465357,
+      "loss": 3.2738,
+      "step": 15277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4862534999847412,
+      "learning_rate": 0.0005124121828720341,
+      "loss": 3.2049,
+      "step": 15278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5458556413650513,
+      "learning_rate": 0.0005124013102380797,
+      "loss": 3.2185,
+      "step": 15279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5805124044418335,
+      "learning_rate": 0.0005123904370447011,
+      "loss": 3.0137,
+      "step": 15280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.521308958530426,
+      "learning_rate": 0.000512379563291927,
+      "loss": 3.1837,
+      "step": 15281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5270718336105347,
+      "learning_rate": 0.000512368688979786,
+      "loss": 3.4552,
+      "step": 15282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.504327654838562,
+      "learning_rate": 0.0005123578141083067,
+      "loss": 3.405,
+      "step": 15283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49246448278427124,
+      "learning_rate": 0.0005123469386775178,
+      "loss": 3.1681,
+      "step": 15284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5137251615524292,
+      "learning_rate": 0.0005123360626874479,
+      "loss": 3.4152,
+      "step": 15285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.477211058139801,
+      "learning_rate": 0.0005123251861381257,
+      "loss": 2.988,
+      "step": 15286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48780468106269836,
+      "learning_rate": 0.0005123143090295799,
+      "loss": 3.2013,
+      "step": 15287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.535379946231842,
+      "learning_rate": 0.0005123034313618389,
+      "loss": 3.1484,
+      "step": 15288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5203052163124084,
+      "learning_rate": 0.0005122925531349317,
+      "loss": 3.1849,
+      "step": 15289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5167236924171448,
+      "learning_rate": 0.0005122816743488866,
+      "loss": 3.0851,
+      "step": 15290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4870116412639618,
+      "learning_rate": 0.0005122707950037325,
+      "loss": 3.3136,
+      "step": 15291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5424923300743103,
+      "learning_rate": 0.0005122599150994981,
+      "loss": 3.1071,
+      "step": 15292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5110763311386108,
+      "learning_rate": 0.0005122490346362119,
+      "loss": 3.0956,
+      "step": 15293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5034542083740234,
+      "learning_rate": 0.0005122381536139025,
+      "loss": 3.2484,
+      "step": 15294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49909499287605286,
+      "learning_rate": 0.0005122272720325986,
+      "loss": 3.314,
+      "step": 15295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5280368328094482,
+      "learning_rate": 0.0005122163898923289,
+      "loss": 3.3027,
+      "step": 15296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4968806505203247,
+      "learning_rate": 0.0005122055071931222,
+      "loss": 3.2456,
+      "step": 15297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5014216899871826,
+      "learning_rate": 0.000512194623935007,
+      "loss": 3.1594,
+      "step": 15298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.535797655582428,
+      "learning_rate": 0.000512183740118012,
+      "loss": 3.2467,
+      "step": 15299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.52153080701828,
+      "learning_rate": 0.0005121728557421658,
+      "loss": 2.9868,
+      "step": 15300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5368970632553101,
+      "learning_rate": 0.0005121619708074972,
+      "loss": 3.1427,
+      "step": 15301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5109313726425171,
+      "learning_rate": 0.0005121510853140348,
+      "loss": 3.0168,
+      "step": 15302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5726766586303711,
+      "learning_rate": 0.0005121401992618073,
+      "loss": 3.0998,
+      "step": 15303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5439186692237854,
+      "learning_rate": 0.0005121293126508432,
+      "loss": 3.1734,
+      "step": 15304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5013107061386108,
+      "learning_rate": 0.0005121184254811714,
+      "loss": 3.3949,
+      "step": 15305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5400449633598328,
+      "learning_rate": 0.0005121075377528205,
+      "loss": 3.0781,
+      "step": 15306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5805463790893555,
+      "learning_rate": 0.0005120966494658192,
+      "loss": 3.1349,
+      "step": 15307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5074591636657715,
+      "learning_rate": 0.0005120857606201961,
+      "loss": 3.3115,
+      "step": 15308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5145214200019836,
+      "learning_rate": 0.00051207487121598,
+      "loss": 3.1443,
+      "step": 15309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5054406523704529,
+      "learning_rate": 0.0005120639812531995,
+      "loss": 3.4488,
+      "step": 15310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5463675856590271,
+      "learning_rate": 0.0005120530907318831,
+      "loss": 2.9445,
+      "step": 15311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5270100235939026,
+      "learning_rate": 0.0005120421996520597,
+      "loss": 3.2497,
+      "step": 15312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5352219939231873,
+      "learning_rate": 0.0005120313080137581,
+      "loss": 3.1157,
+      "step": 15313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.550274670124054,
+      "learning_rate": 0.0005120204158170069,
+      "loss": 3.0731,
+      "step": 15314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5238516330718994,
+      "learning_rate": 0.0005120095230618345,
+      "loss": 3.4092,
+      "step": 15315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4942419230937958,
+      "learning_rate": 0.00051199862974827,
+      "loss": 3.3665,
+      "step": 15316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.489196240901947,
+      "learning_rate": 0.0005119877358763418,
+      "loss": 3.0608,
+      "step": 15317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5123137831687927,
+      "learning_rate": 0.0005119768414460788,
+      "loss": 3.3294,
+      "step": 15318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48939448595046997,
+      "learning_rate": 0.0005119659464575096,
+      "loss": 3.3551,
+      "step": 15319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5240415930747986,
+      "learning_rate": 0.0005119550509106628,
+      "loss": 3.3512,
+      "step": 15320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5232025980949402,
+      "learning_rate": 0.0005119441548055671,
+      "loss": 3.0429,
+      "step": 15321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5226197242736816,
+      "learning_rate": 0.0005119332581422515,
+      "loss": 3.2391,
+      "step": 15322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5628093481063843,
+      "learning_rate": 0.0005119223609207443,
+      "loss": 3.3304,
+      "step": 15323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4734574854373932,
+      "learning_rate": 0.0005119114631410745,
+      "loss": 3.4568,
+      "step": 15324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.515416145324707,
+      "learning_rate": 0.0005119005648032707,
+      "loss": 3.2631,
+      "step": 15325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5349523425102234,
+      "learning_rate": 0.0005118896659073616,
+      "loss": 3.5373,
+      "step": 15326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.562615156173706,
+      "learning_rate": 0.0005118787664533757,
+      "loss": 3.1333,
+      "step": 15327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49806028604507446,
+      "learning_rate": 0.0005118678664413421,
+      "loss": 3.2441,
+      "step": 15328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49525222182273865,
+      "learning_rate": 0.0005118569658712893,
+      "loss": 3.1967,
+      "step": 15329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.548129141330719,
+      "learning_rate": 0.0005118460647432461,
+      "loss": 3.159,
+      "step": 15330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5434116721153259,
+      "learning_rate": 0.000511835163057241,
+      "loss": 3.2088,
+      "step": 15331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5034930109977722,
+      "learning_rate": 0.0005118242608133028,
+      "loss": 3.3135,
+      "step": 15332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5119454264640808,
+      "learning_rate": 0.0005118133580114604,
+      "loss": 3.1451,
+      "step": 15333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5325192213058472,
+      "learning_rate": 0.0005118024546517424,
+      "loss": 3.1097,
+      "step": 15334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5029995441436768,
+      "learning_rate": 0.0005117915507341774,
+      "loss": 3.3235,
+      "step": 15335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5167403221130371,
+      "learning_rate": 0.0005117806462587942,
+      "loss": 3.1699,
+      "step": 15336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5161690711975098,
+      "learning_rate": 0.0005117697412256216,
+      "loss": 3.3004,
+      "step": 15337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5440047383308411,
+      "learning_rate": 0.0005117588356346881,
+      "loss": 3.187,
+      "step": 15338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5046606063842773,
+      "learning_rate": 0.0005117479294860228,
+      "loss": 3.1708,
+      "step": 15339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5875895023345947,
+      "learning_rate": 0.0005117370227796542,
+      "loss": 3.21,
+      "step": 15340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5174129605293274,
+      "learning_rate": 0.0005117261155156109,
+      "loss": 3.0799,
+      "step": 15341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5070399045944214,
+      "learning_rate": 0.0005117152076939218,
+      "loss": 3.294,
+      "step": 15342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5156857967376709,
+      "learning_rate": 0.0005117042993146156,
+      "loss": 3.1872,
+      "step": 15343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5003373026847839,
+      "learning_rate": 0.000511693390377721,
+      "loss": 2.9625,
+      "step": 15344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5173194408416748,
+      "learning_rate": 0.0005116824808832668,
+      "loss": 3.0713,
+      "step": 15345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5497887134552002,
+      "learning_rate": 0.0005116715708312817,
+      "loss": 3.2665,
+      "step": 15346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5368115305900574,
+      "learning_rate": 0.0005116606602217944,
+      "loss": 3.3243,
+      "step": 15347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5352510213851929,
+      "learning_rate": 0.0005116497490548335,
+      "loss": 3.0809,
+      "step": 15348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5086609721183777,
+      "learning_rate": 0.0005116388373304281,
+      "loss": 3.2868,
+      "step": 15349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5148516297340393,
+      "learning_rate": 0.0005116279250486067,
+      "loss": 3.1258,
+      "step": 15350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5165446996688843,
+      "learning_rate": 0.000511617012209398,
+      "loss": 3.2221,
+      "step": 15351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5249726176261902,
+      "learning_rate": 0.0005116060988128308,
+      "loss": 3.2277,
+      "step": 15352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5012349486351013,
+      "learning_rate": 0.0005115951848589339,
+      "loss": 3.4042,
+      "step": 15353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5562968850135803,
+      "learning_rate": 0.0005115842703477361,
+      "loss": 3.1529,
+      "step": 15354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49537578225135803,
+      "learning_rate": 0.0005115733552792659,
+      "loss": 3.0784,
+      "step": 15355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5626443028450012,
+      "learning_rate": 0.0005115624396535522,
+      "loss": 3.1834,
+      "step": 15356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5942702889442444,
+      "learning_rate": 0.0005115515234706238,
+      "loss": 2.9109,
+      "step": 15357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5349305272102356,
+      "learning_rate": 0.0005115406067305095,
+      "loss": 3.0626,
+      "step": 15358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4917951822280884,
+      "learning_rate": 0.0005115296894332379,
+      "loss": 3.3152,
+      "step": 15359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5094746351242065,
+      "learning_rate": 0.0005115187715788377,
+      "loss": 3.1875,
+      "step": 15360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5121936798095703,
+      "learning_rate": 0.0005115078531673379,
+      "loss": 3.3627,
+      "step": 15361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5854499340057373,
+      "learning_rate": 0.000511496934198767,
+      "loss": 3.2952,
+      "step": 15362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5319592952728271,
+      "learning_rate": 0.000511486014673154,
+      "loss": 3.2366,
+      "step": 15363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.553443193435669,
+      "learning_rate": 0.0005114750945905275,
+      "loss": 3.0643,
+      "step": 15364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5256295204162598,
+      "learning_rate": 0.0005114641739509162,
+      "loss": 3.1867,
+      "step": 15365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5130583643913269,
+      "learning_rate": 0.0005114532527543492,
+      "loss": 3.2345,
+      "step": 15366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5329893231391907,
+      "learning_rate": 0.0005114423310008547,
+      "loss": 3.4116,
+      "step": 15367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5318312644958496,
+      "learning_rate": 0.000511431408690462,
+      "loss": 3.2325,
+      "step": 15368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5396196842193604,
+      "learning_rate": 0.0005114204858231997,
+      "loss": 3.1291,
+      "step": 15369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4775935113430023,
+      "learning_rate": 0.0005114095623990964,
+      "loss": 3.3107,
+      "step": 15370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.596052348613739,
+      "learning_rate": 0.000511398638418181,
+      "loss": 3.3655,
+      "step": 15371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5218362212181091,
+      "learning_rate": 0.0005113877138804824,
+      "loss": 3.0283,
+      "step": 15372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5000458359718323,
+      "learning_rate": 0.0005113767887860291,
+      "loss": 3.2683,
+      "step": 15373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5197030901908875,
+      "learning_rate": 0.0005113658631348501,
+      "loss": 3.058,
+      "step": 15374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5189664363861084,
+      "learning_rate": 0.000511354936926974,
+      "loss": 3.3382,
+      "step": 15375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.571043848991394,
+      "learning_rate": 0.0005113440101624299,
+      "loss": 3.0839,
+      "step": 15376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5379744172096252,
+      "learning_rate": 0.0005113330828412461,
+      "loss": 3.1519,
+      "step": 15377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4935210943222046,
+      "learning_rate": 0.0005113221549634517,
+      "loss": 3.1605,
+      "step": 15378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5590223670005798,
+      "learning_rate": 0.0005113112265290755,
+      "loss": 3.3283,
+      "step": 15379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5487520694732666,
+      "learning_rate": 0.0005113002975381462,
+      "loss": 3.4678,
+      "step": 15380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5301699638366699,
+      "learning_rate": 0.0005112893679906926,
+      "loss": 3.0123,
+      "step": 15381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5726925730705261,
+      "learning_rate": 0.0005112784378867432,
+      "loss": 3.1856,
+      "step": 15382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5013753771781921,
+      "learning_rate": 0.0005112675072263273,
+      "loss": 3.1004,
+      "step": 15383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5306413173675537,
+      "learning_rate": 0.0005112565760094734,
+      "loss": 3.1682,
+      "step": 15384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4865894019603729,
+      "learning_rate": 0.0005112456442362104,
+      "loss": 3.1465,
+      "step": 15385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5361476540565491,
+      "learning_rate": 0.0005112347119065669,
+      "loss": 3.5116,
+      "step": 15386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5227615833282471,
+      "learning_rate": 0.0005112237790205719,
+      "loss": 3.2517,
+      "step": 15387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5013512969017029,
+      "learning_rate": 0.000511212845578254,
+      "loss": 2.9083,
+      "step": 15388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5280064940452576,
+      "learning_rate": 0.0005112019115796424,
+      "loss": 3.3894,
+      "step": 15389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5008797645568848,
+      "learning_rate": 0.0005111909770247653,
+      "loss": 3.4333,
+      "step": 15390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5193240642547607,
+      "learning_rate": 0.000511180041913652,
+      "loss": 3.1138,
+      "step": 15391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49930593371391296,
+      "learning_rate": 0.000511169106246331,
+      "loss": 3.3978,
+      "step": 15392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48021283745765686,
+      "learning_rate": 0.0005111581700228313,
+      "loss": 3.3122,
+      "step": 15393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5200793147087097,
+      "learning_rate": 0.0005111472332431815,
+      "loss": 3.2217,
+      "step": 15394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5211922526359558,
+      "learning_rate": 0.0005111362959074106,
+      "loss": 3.1195,
+      "step": 15395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5001904368400574,
+      "learning_rate": 0.0005111253580155474,
+      "loss": 3.1291,
+      "step": 15396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5240308046340942,
+      "learning_rate": 0.0005111144195676206,
+      "loss": 3.3654,
+      "step": 15397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5145796537399292,
+      "learning_rate": 0.0005111034805636589,
+      "loss": 3.1215,
+      "step": 15398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.517951488494873,
+      "learning_rate": 0.0005110925410036914,
+      "loss": 3.1964,
+      "step": 15399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5089027285575867,
+      "learning_rate": 0.0005110816008877468,
+      "loss": 3.2889,
+      "step": 15400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5206027626991272,
+      "learning_rate": 0.0005110706602158539,
+      "loss": 3.2054,
+      "step": 15401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.522180438041687,
+      "learning_rate": 0.0005110597189880414,
+      "loss": 3.2392,
+      "step": 15402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5154605507850647,
+      "learning_rate": 0.0005110487772043383,
+      "loss": 3.1462,
+      "step": 15403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6345930695533752,
+      "learning_rate": 0.0005110378348647732,
+      "loss": 3.4421,
+      "step": 15404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5428985357284546,
+      "learning_rate": 0.0005110268919693752,
+      "loss": 3.1896,
+      "step": 15405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5568393468856812,
+      "learning_rate": 0.0005110159485181729,
+      "loss": 3.1683,
+      "step": 15406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5200351476669312,
+      "learning_rate": 0.0005110050045111953,
+      "loss": 3.0301,
+      "step": 15407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5830309391021729,
+      "learning_rate": 0.000510994059948471,
+      "loss": 3.2447,
+      "step": 15408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5186257362365723,
+      "learning_rate": 0.000510983114830029,
+      "loss": 3.4328,
+      "step": 15409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5107519626617432,
+      "learning_rate": 0.0005109721691558981,
+      "loss": 3.4096,
+      "step": 15410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5244288444519043,
+      "learning_rate": 0.0005109612229261073,
+      "loss": 3.1377,
+      "step": 15411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.47387999296188354,
+      "learning_rate": 0.000510950276140685,
+      "loss": 3.3397,
+      "step": 15412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5340439677238464,
+      "learning_rate": 0.0005109393287996602,
+      "loss": 3.1848,
+      "step": 15413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.506712794303894,
+      "learning_rate": 0.000510928380903062,
+      "loss": 3.1481,
+      "step": 15414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5220345854759216,
+      "learning_rate": 0.0005109174324509189,
+      "loss": 3.0608,
+      "step": 15415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48933979868888855,
+      "learning_rate": 0.00051090648344326,
+      "loss": 3.0937,
+      "step": 15416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5643341541290283,
+      "learning_rate": 0.0005108955338801139,
+      "loss": 2.9736,
+      "step": 15417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5401406288146973,
+      "learning_rate": 0.0005108845837615096,
+      "loss": 3.1143,
+      "step": 15418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5696349740028381,
+      "learning_rate": 0.0005108736330874759,
+      "loss": 3.4066,
+      "step": 15419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5899490118026733,
+      "learning_rate": 0.0005108626818580415,
+      "loss": 3.0104,
+      "step": 15420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.515295684337616,
+      "learning_rate": 0.0005108517300732356,
+      "loss": 3.2919,
+      "step": 15421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5573397874832153,
+      "learning_rate": 0.0005108407777330867,
+      "loss": 3.061,
+      "step": 15422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5408274531364441,
+      "learning_rate": 0.0005108298248376238,
+      "loss": 3.4797,
+      "step": 15423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5029387474060059,
+      "learning_rate": 0.0005108188713868758,
+      "loss": 3.2384,
+      "step": 15424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4932806193828583,
+      "learning_rate": 0.0005108079173808713,
+      "loss": 3.2024,
+      "step": 15425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.530828058719635,
+      "learning_rate": 0.0005107969628196393,
+      "loss": 3.1832,
+      "step": 15426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.528161883354187,
+      "learning_rate": 0.0005107860077032088,
+      "loss": 3.1924,
+      "step": 15427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5298112630844116,
+      "learning_rate": 0.0005107750520316085,
+      "loss": 3.2096,
+      "step": 15428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5262047648429871,
+      "learning_rate": 0.0005107640958048673,
+      "loss": 3.2419,
+      "step": 15429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5328362584114075,
+      "learning_rate": 0.000510753139023014,
+      "loss": 3.0983,
+      "step": 15430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5387527346611023,
+      "learning_rate": 0.0005107421816860774,
+      "loss": 3.1221,
+      "step": 15431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5194116234779358,
+      "learning_rate": 0.0005107312237940866,
+      "loss": 3.1749,
+      "step": 15432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.506995677947998,
+      "learning_rate": 0.0005107202653470703,
+      "loss": 3.1372,
+      "step": 15433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5215865969657898,
+      "learning_rate": 0.0005107093063450573,
+      "loss": 3.1794,
+      "step": 15434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5082940459251404,
+      "learning_rate": 0.0005106983467880765,
+      "loss": 3.0021,
+      "step": 15435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5554096102714539,
+      "learning_rate": 0.0005106873866761569,
+      "loss": 3.2256,
+      "step": 15436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5048863291740417,
+      "learning_rate": 0.0005106764260093273,
+      "loss": 3.3262,
+      "step": 15437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.507644534111023,
+      "learning_rate": 0.0005106654647876165,
+      "loss": 3.1868,
+      "step": 15438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5397703051567078,
+      "learning_rate": 0.0005106545030110533,
+      "loss": 3.0611,
+      "step": 15439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.524456262588501,
+      "learning_rate": 0.0005106435406796668,
+      "loss": 3.3257,
+      "step": 15440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49036774039268494,
+      "learning_rate": 0.0005106325777934857,
+      "loss": 3.5181,
+      "step": 15441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5018916726112366,
+      "learning_rate": 0.000510621614352539,
+      "loss": 3.1098,
+      "step": 15442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5243989825248718,
+      "learning_rate": 0.0005106106503568555,
+      "loss": 3.1468,
+      "step": 15443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5382325649261475,
+      "learning_rate": 0.000510599685806464,
+      "loss": 3.2289,
+      "step": 15444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5662506222724915,
+      "learning_rate": 0.0005105887207013934,
+      "loss": 3.1284,
+      "step": 15445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5015741586685181,
+      "learning_rate": 0.0005105777550416728,
+      "loss": 3.3105,
+      "step": 15446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5282851457595825,
+      "learning_rate": 0.0005105667888273309,
+      "loss": 3.2448,
+      "step": 15447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5204810500144958,
+      "learning_rate": 0.0005105558220583965,
+      "loss": 3.2382,
+      "step": 15448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5286573171615601,
+      "learning_rate": 0.0005105448547348986,
+      "loss": 3.2456,
+      "step": 15449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5265070199966431,
+      "learning_rate": 0.0005105338868568661,
+      "loss": 3.2492,
+      "step": 15450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5228598713874817,
+      "learning_rate": 0.0005105229184243279,
+      "loss": 3.1497,
+      "step": 15451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5335772633552551,
+      "learning_rate": 0.0005105119494373128,
+      "loss": 3.2094,
+      "step": 15452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4976807236671448,
+      "learning_rate": 0.0005105009798958498,
+      "loss": 3.1789,
+      "step": 15453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5059366822242737,
+      "learning_rate": 0.0005104900097999676,
+      "loss": 3.2996,
+      "step": 15454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4928973913192749,
+      "learning_rate": 0.0005104790391496953,
+      "loss": 3.2344,
+      "step": 15455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48326247930526733,
+      "learning_rate": 0.0005104680679450618,
+      "loss": 3.3739,
+      "step": 15456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5389878749847412,
+      "learning_rate": 0.0005104570961860958,
+      "loss": 3.0607,
+      "step": 15457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5126422643661499,
+      "learning_rate": 0.0005104461238728264,
+      "loss": 3.2553,
+      "step": 15458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5000913143157959,
+      "learning_rate": 0.0005104351510052823,
+      "loss": 3.156,
+      "step": 15459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4934830665588379,
+      "learning_rate": 0.0005104241775834926,
+      "loss": 3.2617,
+      "step": 15460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5197305083274841,
+      "learning_rate": 0.0005104132036074862,
+      "loss": 3.2124,
+      "step": 15461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4591943025588989,
+      "learning_rate": 0.0005104022290772918,
+      "loss": 3.1996,
+      "step": 15462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5356560349464417,
+      "learning_rate": 0.0005103912539929383,
+      "loss": 3.1692,
+      "step": 15463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5015002489089966,
+      "learning_rate": 0.0005103802783544549,
+      "loss": 3.0953,
+      "step": 15464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5325064063072205,
+      "learning_rate": 0.0005103693021618704,
+      "loss": 3.1567,
+      "step": 15465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5010794401168823,
+      "learning_rate": 0.0005103583254152134,
+      "loss": 3.0977,
+      "step": 15466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.644633948802948,
+      "learning_rate": 0.0005103473481145132,
+      "loss": 3.4053,
+      "step": 15467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5222316980361938,
+      "learning_rate": 0.0005103363702597986,
+      "loss": 3.3435,
+      "step": 15468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5279338955879211,
+      "learning_rate": 0.0005103253918510984,
+      "loss": 3.2726,
+      "step": 15469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5039032101631165,
+      "learning_rate": 0.0005103144128884416,
+      "loss": 3.1634,
+      "step": 15470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5194478631019592,
+      "learning_rate": 0.0005103034333718572,
+      "loss": 3.2469,
+      "step": 15471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5506890416145325,
+      "learning_rate": 0.000510292453301374,
+      "loss": 3.4793,
+      "step": 15472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5099464058876038,
+      "learning_rate": 0.000510281472677021,
+      "loss": 3.2517,
+      "step": 15473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5186891555786133,
+      "learning_rate": 0.0005102704914988269,
+      "loss": 3.1561,
+      "step": 15474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4984113276004791,
+      "learning_rate": 0.0005102595097668209,
+      "loss": 3.2346,
+      "step": 15475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5584531426429749,
+      "learning_rate": 0.0005102485274810319,
+      "loss": 3.1323,
+      "step": 15476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5731603503227234,
+      "learning_rate": 0.0005102375446414886,
+      "loss": 3.2928,
+      "step": 15477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4951866865158081,
+      "learning_rate": 0.0005102265612482201,
+      "loss": 3.2515,
+      "step": 15478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49377912282943726,
+      "learning_rate": 0.0005102155773012553,
+      "loss": 2.8477,
+      "step": 15479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5461580157279968,
+      "learning_rate": 0.0005102045928006231,
+      "loss": 3.1564,
+      "step": 15480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6521041393280029,
+      "learning_rate": 0.0005101936077463525,
+      "loss": 3.3003,
+      "step": 15481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48901596665382385,
+      "learning_rate": 0.0005101826221384724,
+      "loss": 3.1159,
+      "step": 15482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5265803933143616,
+      "learning_rate": 0.0005101716359770117,
+      "loss": 3.3615,
+      "step": 15483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5226854681968689,
+      "learning_rate": 0.0005101606492619994,
+      "loss": 3.0614,
+      "step": 15484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5515998601913452,
+      "learning_rate": 0.0005101496619934644,
+      "loss": 3.2238,
+      "step": 15485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.487713485956192,
+      "learning_rate": 0.0005101386741714356,
+      "loss": 3.2274,
+      "step": 15486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49414488673210144,
+      "learning_rate": 0.0005101276857959419,
+      "loss": 3.247,
+      "step": 15487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5462265610694885,
+      "learning_rate": 0.0005101166968670125,
+      "loss": 3.2225,
+      "step": 15488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5077669620513916,
+      "learning_rate": 0.000510105707384676,
+      "loss": 3.2476,
+      "step": 15489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5247009992599487,
+      "learning_rate": 0.0005100947173489615,
+      "loss": 2.8845,
+      "step": 15490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4996735751628876,
+      "learning_rate": 0.000510083726759898,
+      "loss": 3.262,
+      "step": 15491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5037980675697327,
+      "learning_rate": 0.0005100727356175145,
+      "loss": 3.072,
+      "step": 15492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.515090823173523,
+      "learning_rate": 0.0005100617439218397,
+      "loss": 3.2084,
+      "step": 15493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5135362148284912,
+      "learning_rate": 0.0005100507516729027,
+      "loss": 3.0892,
+      "step": 15494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4950462579727173,
+      "learning_rate": 0.0005100397588707325,
+      "loss": 3.0106,
+      "step": 15495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4800357520580292,
+      "learning_rate": 0.000510028765515358,
+      "loss": 3.2204,
+      "step": 15496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.534273087978363,
+      "learning_rate": 0.000510017771606808,
+      "loss": 3.1293,
+      "step": 15497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5002960562705994,
+      "learning_rate": 0.0005100067771451118,
+      "loss": 3.2213,
+      "step": 15498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5246545076370239,
+      "learning_rate": 0.0005099957821302981,
+      "loss": 3.2642,
+      "step": 15499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5540328621864319,
+      "learning_rate": 0.000509984786562396,
+      "loss": 3.1108,
+      "step": 15500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49777376651763916,
+      "learning_rate": 0.0005099737904414342,
+      "loss": 3.2185,
+      "step": 15501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5149266719818115,
+      "learning_rate": 0.0005099627937674421,
+      "loss": 3.2102,
+      "step": 15502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4814930260181427,
+      "learning_rate": 0.0005099517965404482,
+      "loss": 3.1635,
+      "step": 15503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48603329062461853,
+      "learning_rate": 0.0005099407987604818,
+      "loss": 3.2968,
+      "step": 15504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4911821484565735,
+      "learning_rate": 0.0005099298004275717,
+      "loss": 3.3795,
+      "step": 15505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5378304719924927,
+      "learning_rate": 0.0005099188015417468,
+      "loss": 3.1863,
+      "step": 15506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5142403841018677,
+      "learning_rate": 0.0005099078021030362,
+      "loss": 3.2434,
+      "step": 15507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5013147592544556,
+      "learning_rate": 0.000509896802111469,
+      "loss": 3.1608,
+      "step": 15508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4968760907649994,
+      "learning_rate": 0.000509885801567074,
+      "loss": 3.195,
+      "step": 15509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5242990255355835,
+      "learning_rate": 0.0005098748004698801,
+      "loss": 3.2617,
+      "step": 15510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5505073070526123,
+      "learning_rate": 0.0005098637988199163,
+      "loss": 3.0103,
+      "step": 15511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5673673152923584,
+      "learning_rate": 0.0005098527966172117,
+      "loss": 3.1611,
+      "step": 15512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5113774538040161,
+      "learning_rate": 0.0005098417938617952,
+      "loss": 3.4058,
+      "step": 15513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4958935081958771,
+      "learning_rate": 0.0005098307905536959,
+      "loss": 3.1639,
+      "step": 15514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5038467645645142,
+      "learning_rate": 0.0005098197866929427,
+      "loss": 3.1384,
+      "step": 15515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.493192195892334,
+      "learning_rate": 0.0005098087822795645,
+      "loss": 3.1471,
+      "step": 15516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5338720679283142,
+      "learning_rate": 0.0005097977773135903,
+      "loss": 3.2762,
+      "step": 15517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5778473615646362,
+      "learning_rate": 0.0005097867717950492,
+      "loss": 3.1879,
+      "step": 15518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5032196044921875,
+      "learning_rate": 0.0005097757657239701,
+      "loss": 3.1646,
+      "step": 15519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5262255668640137,
+      "learning_rate": 0.000509764759100382,
+      "loss": 3.2137,
+      "step": 15520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49957624077796936,
+      "learning_rate": 0.0005097537519243139,
+      "loss": 3.111,
+      "step": 15521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5052130818367004,
+      "learning_rate": 0.0005097427441957948,
+      "loss": 3.1133,
+      "step": 15522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48932185769081116,
+      "learning_rate": 0.0005097317359148536,
+      "loss": 3.3346,
+      "step": 15523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5048071146011353,
+      "learning_rate": 0.0005097207270815194,
+      "loss": 3.322,
+      "step": 15524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5462115406990051,
+      "learning_rate": 0.0005097097176958212,
+      "loss": 2.9782,
+      "step": 15525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5595006942749023,
+      "learning_rate": 0.000509698707757788,
+      "loss": 2.9495,
+      "step": 15526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5353357791900635,
+      "learning_rate": 0.0005096876972674486,
+      "loss": 3.2685,
+      "step": 15527
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4759270250797272,
+      "learning_rate": 0.0005096766862248323,
+      "loss": 3.1232,
+      "step": 15528
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5706182718276978,
+      "learning_rate": 0.000509665674629968,
+      "loss": 3.0231,
+      "step": 15529
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5052098631858826,
+      "learning_rate": 0.0005096546624828845,
+      "loss": 3.12,
+      "step": 15530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5506969690322876,
+      "learning_rate": 0.0005096436497836111,
+      "loss": 3.1312,
+      "step": 15531
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5286117792129517,
+      "learning_rate": 0.0005096326365321767,
+      "loss": 3.2991,
+      "step": 15532
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5287132859230042,
+      "learning_rate": 0.0005096216227286102,
+      "loss": 3.136,
+      "step": 15533
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.512438952922821,
+      "learning_rate": 0.0005096106083729406,
+      "loss": 3.2366,
+      "step": 15534
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5107895731925964,
+      "learning_rate": 0.0005095995934651972,
+      "loss": 3.4381,
+      "step": 15535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.523353636264801,
+      "learning_rate": 0.0005095885780054087,
+      "loss": 3.1528,
+      "step": 15536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5121175646781921,
+      "learning_rate": 0.0005095775619936042,
+      "loss": 3.2085,
+      "step": 15537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5396693348884583,
+      "learning_rate": 0.0005095665454298129,
+      "loss": 2.9958,
+      "step": 15538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49359792470932007,
+      "learning_rate": 0.0005095555283140635,
+      "loss": 3.0549,
+      "step": 15539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.584166944026947,
+      "learning_rate": 0.0005095445106463852,
+      "loss": 3.0041,
+      "step": 15540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5615528225898743,
+      "learning_rate": 0.000509533492426807,
+      "loss": 3.2767,
+      "step": 15541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5622826814651489,
+      "learning_rate": 0.0005095224736553578,
+      "loss": 3.09,
+      "step": 15542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5125862956047058,
+      "learning_rate": 0.0005095114543320668,
+      "loss": 3.2732,
+      "step": 15543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.507877767086029,
+      "learning_rate": 0.000509500434456963,
+      "loss": 2.8684,
+      "step": 15544
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5564316511154175,
+      "learning_rate": 0.0005094894140300753,
+      "loss": 3.1338,
+      "step": 15545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5386070609092712,
+      "learning_rate": 0.0005094783930514329,
+      "loss": 3.0938,
+      "step": 15546
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4967600405216217,
+      "learning_rate": 0.0005094673715210647,
+      "loss": 3.153,
+      "step": 15547
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4921207129955292,
+      "learning_rate": 0.0005094563494389997,
+      "loss": 3.1201,
+      "step": 15548
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5483155250549316,
+      "learning_rate": 0.0005094453268052671,
+      "loss": 2.9707,
+      "step": 15549
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5265492796897888,
+      "learning_rate": 0.0005094343036198956,
+      "loss": 3.1635,
+      "step": 15550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5425223708152771,
+      "learning_rate": 0.0005094232798829146,
+      "loss": 2.9743,
+      "step": 15551
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5007231831550598,
+      "learning_rate": 0.0005094122555943529,
+      "loss": 3.2458,
+      "step": 15552
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5598417520523071,
+      "learning_rate": 0.0005094012307542398,
+      "loss": 3.128,
+      "step": 15553
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5178683996200562,
+      "learning_rate": 0.000509390205362604,
+      "loss": 3.3222,
+      "step": 15554
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5192438364028931,
+      "learning_rate": 0.0005093791794194747,
+      "loss": 3.2078,
+      "step": 15555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5315037369728088,
+      "learning_rate": 0.000509368152924881,
+      "loss": 3.1406,
+      "step": 15556
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5374244451522827,
+      "learning_rate": 0.0005093571258788518,
+      "loss": 3.0132,
+      "step": 15557
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5307905077934265,
+      "learning_rate": 0.0005093460982814164,
+      "loss": 3.0444,
+      "step": 15558
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5297430157661438,
+      "learning_rate": 0.0005093350701326035,
+      "loss": 3.3374,
+      "step": 15559
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5116652846336365,
+      "learning_rate": 0.0005093240414324423,
+      "loss": 3.1824,
+      "step": 15560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5311102271080017,
+      "learning_rate": 0.000509313012180962,
+      "loss": 3.1643,
+      "step": 15561
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5218523740768433,
+      "learning_rate": 0.0005093019823781915,
+      "loss": 3.2827,
+      "step": 15562
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5624101161956787,
+      "learning_rate": 0.0005092909520241596,
+      "loss": 3.2227,
+      "step": 15563
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5099868178367615,
+      "learning_rate": 0.0005092799211188959,
+      "loss": 3.1073,
+      "step": 15564
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5127807259559631,
+      "learning_rate": 0.000509268889662429,
+      "loss": 3.2646,
+      "step": 15565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.528416097164154,
+      "learning_rate": 0.0005092578576547882,
+      "loss": 3.1796,
+      "step": 15566
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49281713366508484,
+      "learning_rate": 0.0005092468250960025,
+      "loss": 3.1264,
+      "step": 15567
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5014824271202087,
+      "learning_rate": 0.0005092357919861009,
+      "loss": 3.1486,
+      "step": 15568
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4935286045074463,
+      "learning_rate": 0.0005092247583251124,
+      "loss": 3.2687,
+      "step": 15569
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.520007312297821,
+      "learning_rate": 0.0005092137241130662,
+      "loss": 3.1096,
+      "step": 15570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5099589824676514,
+      "learning_rate": 0.0005092026893499914,
+      "loss": 3.2324,
+      "step": 15571
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5131465196609497,
+      "learning_rate": 0.0005091916540359169,
+      "loss": 3.0389,
+      "step": 15572
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49344658851623535,
+      "learning_rate": 0.0005091806181708719,
+      "loss": 3.3404,
+      "step": 15573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5237314701080322,
+      "learning_rate": 0.0005091695817548853,
+      "loss": 3.1772,
+      "step": 15574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5355901122093201,
+      "learning_rate": 0.0005091585447879864,
+      "loss": 3.2269,
+      "step": 15575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5290843844413757,
+      "learning_rate": 0.0005091475072702041,
+      "loss": 3.0839,
+      "step": 15576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5169116258621216,
+      "learning_rate": 0.0005091364692015676,
+      "loss": 3.2695,
+      "step": 15577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6524379253387451,
+      "learning_rate": 0.0005091254305821058,
+      "loss": 3.4017,
+      "step": 15578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5682939291000366,
+      "learning_rate": 0.0005091143914118478,
+      "loss": 3.2909,
+      "step": 15579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5173560976982117,
+      "learning_rate": 0.0005091033516908228,
+      "loss": 3.1853,
+      "step": 15580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5009781718254089,
+      "learning_rate": 0.00050909231141906,
+      "loss": 3.3297,
+      "step": 15581
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5405966639518738,
+      "learning_rate": 0.0005090812705965881,
+      "loss": 3.1403,
+      "step": 15582
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4971311390399933,
+      "learning_rate": 0.0005090702292234363,
+      "loss": 3.1587,
+      "step": 15583
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5876317024230957,
+      "learning_rate": 0.0005090591872996338,
+      "loss": 3.1248,
+      "step": 15584
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.665259599685669,
+      "learning_rate": 0.0005090481448252098,
+      "loss": 3.1146,
+      "step": 15585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5335424542427063,
+      "learning_rate": 0.000509037101800193,
+      "loss": 3.1938,
+      "step": 15586
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5514816045761108,
+      "learning_rate": 0.0005090260582246128,
+      "loss": 3.3065,
+      "step": 15587
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5073038339614868,
+      "learning_rate": 0.0005090150140984982,
+      "loss": 3.2372,
+      "step": 15588
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5548021793365479,
+      "learning_rate": 0.0005090039694218782,
+      "loss": 3.1944,
+      "step": 15589
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5426687598228455,
+      "learning_rate": 0.0005089929241947821,
+      "loss": 3.3712,
+      "step": 15590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5318143963813782,
+      "learning_rate": 0.0005089818784172388,
+      "loss": 3.1281,
+      "step": 15591
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5172745585441589,
+      "learning_rate": 0.0005089708320892774,
+      "loss": 3.1421,
+      "step": 15592
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5684549808502197,
+      "learning_rate": 0.0005089597852109271,
+      "loss": 3.0396,
+      "step": 15593
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5054349899291992,
+      "learning_rate": 0.000508948737782217,
+      "loss": 3.2081,
+      "step": 15594
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4890936315059662,
+      "learning_rate": 0.0005089376898031761,
+      "loss": 3.3387,
+      "step": 15595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5217352509498596,
+      "learning_rate": 0.0005089266412738334,
+      "loss": 3.1863,
+      "step": 15596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.529723048210144,
+      "learning_rate": 0.0005089155921942183,
+      "loss": 3.0914,
+      "step": 15597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5619720816612244,
+      "learning_rate": 0.0005089045425643596,
+      "loss": 3.222,
+      "step": 15598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5275721549987793,
+      "learning_rate": 0.0005088934923842866,
+      "loss": 3.2866,
+      "step": 15599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5122009515762329,
+      "learning_rate": 0.0005088824416540285,
+      "loss": 3.3297,
+      "step": 15600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5561813712120056,
+      "learning_rate": 0.000508871390373614,
+      "loss": 3.218,
+      "step": 15601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5213872194290161,
+      "learning_rate": 0.0005088603385430727,
+      "loss": 3.2862,
+      "step": 15602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.532748281955719,
+      "learning_rate": 0.0005088492861624332,
+      "loss": 3.2017,
+      "step": 15603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5274077653884888,
+      "learning_rate": 0.000508838233231725,
+      "loss": 3.2662,
+      "step": 15604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5420587062835693,
+      "learning_rate": 0.0005088271797509771,
+      "loss": 3.2567,
+      "step": 15605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5665300488471985,
+      "learning_rate": 0.0005088161257202186,
+      "loss": 3.1568,
+      "step": 15606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5209993124008179,
+      "learning_rate": 0.0005088050711394786,
+      "loss": 3.3423,
+      "step": 15607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5469440221786499,
+      "learning_rate": 0.0005087940160087862,
+      "loss": 3.3472,
+      "step": 15608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5044681429862976,
+      "learning_rate": 0.0005087829603281707,
+      "loss": 3.0458,
+      "step": 15609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5147929787635803,
+      "learning_rate": 0.0005087719040976609,
+      "loss": 3.2094,
+      "step": 15610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5366151928901672,
+      "learning_rate": 0.0005087608473172861,
+      "loss": 3.1899,
+      "step": 15611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5306175351142883,
+      "learning_rate": 0.0005087497899870754,
+      "loss": 3.0799,
+      "step": 15612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5149161219596863,
+      "learning_rate": 0.0005087387321070579,
+      "loss": 3.1933,
+      "step": 15613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5391685366630554,
+      "learning_rate": 0.0005087276736772628,
+      "loss": 3.1901,
+      "step": 15614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5293787717819214,
+      "learning_rate": 0.0005087166146977193,
+      "loss": 2.8526,
+      "step": 15615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5558761358261108,
+      "learning_rate": 0.0005087055551684562,
+      "loss": 3.1358,
+      "step": 15616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5193555951118469,
+      "learning_rate": 0.000508694495089503,
+      "loss": 3.0509,
+      "step": 15617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49987754225730896,
+      "learning_rate": 0.0005086834344608885,
+      "loss": 3.1635,
+      "step": 15618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4778132736682892,
+      "learning_rate": 0.000508672373282642,
+      "loss": 2.9455,
+      "step": 15619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5623522996902466,
+      "learning_rate": 0.0005086613115547928,
+      "loss": 3.2735,
+      "step": 15620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5368536710739136,
+      "learning_rate": 0.0005086502492773697,
+      "loss": 3.0355,
+      "step": 15621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5360807776451111,
+      "learning_rate": 0.0005086391864504022,
+      "loss": 3.2697,
+      "step": 15622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5049542188644409,
+      "learning_rate": 0.0005086281230739191,
+      "loss": 3.2567,
+      "step": 15623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5606904625892639,
+      "learning_rate": 0.0005086170591479496,
+      "loss": 3.185,
+      "step": 15624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5596327185630798,
+      "learning_rate": 0.000508605994672523,
+      "loss": 3.1354,
+      "step": 15625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.536123514175415,
+      "learning_rate": 0.0005085949296476684,
+      "loss": 3.1321,
+      "step": 15626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5361589789390564,
+      "learning_rate": 0.0005085838640734149,
+      "loss": 3.2706,
+      "step": 15627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5314226746559143,
+      "learning_rate": 0.0005085727979497915,
+      "loss": 3.1124,
+      "step": 15628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5757156610488892,
+      "learning_rate": 0.0005085617312768277,
+      "loss": 3.1707,
+      "step": 15629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5240241289138794,
+      "learning_rate": 0.0005085506640545522,
+      "loss": 3.105,
+      "step": 15630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5034173727035522,
+      "learning_rate": 0.0005085395962829946,
+      "loss": 3.1758,
+      "step": 15631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5380710363388062,
+      "learning_rate": 0.0005085285279621838,
+      "loss": 3.2823,
+      "step": 15632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5575569272041321,
+      "learning_rate": 0.000508517459092149,
+      "loss": 3.1727,
+      "step": 15633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5457764863967896,
+      "learning_rate": 0.0005085063896729192,
+      "loss": 3.0716,
+      "step": 15634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5306626558303833,
+      "learning_rate": 0.0005084953197045238,
+      "loss": 2.9403,
+      "step": 15635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5452754497528076,
+      "learning_rate": 0.0005084842491869918,
+      "loss": 3.3432,
+      "step": 15636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5277231931686401,
+      "learning_rate": 0.0005084731781203525,
+      "loss": 3.1267,
+      "step": 15637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5706171989440918,
+      "learning_rate": 0.0005084621065046349,
+      "loss": 3.2748,
+      "step": 15638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5383669137954712,
+      "learning_rate": 0.0005084510343398682,
+      "loss": 3.2757,
+      "step": 15639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5343171954154968,
+      "learning_rate": 0.0005084399616260818,
+      "loss": 3.24,
+      "step": 15640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49760371446609497,
+      "learning_rate": 0.0005084288883633044,
+      "loss": 3.049,
+      "step": 15641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5333790183067322,
+      "learning_rate": 0.0005084178145515656,
+      "loss": 3.1123,
+      "step": 15642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5920032858848572,
+      "learning_rate": 0.0005084067401908942,
+      "loss": 2.9831,
+      "step": 15643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5388069152832031,
+      "learning_rate": 0.0005083956652813196,
+      "loss": 3.0262,
+      "step": 15644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5454514622688293,
+      "learning_rate": 0.0005083845898228711,
+      "loss": 3.3217,
+      "step": 15645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5356780290603638,
+      "learning_rate": 0.0005083735138155775,
+      "loss": 3.1691,
+      "step": 15646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4733395278453827,
+      "learning_rate": 0.0005083624372594683,
+      "loss": 3.011,
+      "step": 15647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5192866921424866,
+      "learning_rate": 0.0005083513601545725,
+      "loss": 3.3161,
+      "step": 15648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4957810640335083,
+      "learning_rate": 0.0005083402825009193,
+      "loss": 3.2863,
+      "step": 15649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4861541986465454,
+      "learning_rate": 0.0005083292042985379,
+      "loss": 3.0497,
+      "step": 15650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5232390761375427,
+      "learning_rate": 0.0005083181255474575,
+      "loss": 2.973,
+      "step": 15651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5298289060592651,
+      "learning_rate": 0.0005083070462477072,
+      "loss": 3.3084,
+      "step": 15652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.511946439743042,
+      "learning_rate": 0.0005082959663993162,
+      "loss": 3.1297,
+      "step": 15653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5342845320701599,
+      "learning_rate": 0.0005082848860023137,
+      "loss": 3.1233,
+      "step": 15654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5111925601959229,
+      "learning_rate": 0.000508273805056729,
+      "loss": 3.2461,
+      "step": 15655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.553409993648529,
+      "learning_rate": 0.0005082627235625911,
+      "loss": 3.2622,
+      "step": 15656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.503697395324707,
+      "learning_rate": 0.0005082516415199293,
+      "loss": 3.0538,
+      "step": 15657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5087273716926575,
+      "learning_rate": 0.0005082405589287728,
+      "loss": 3.1954,
+      "step": 15658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5421915650367737,
+      "learning_rate": 0.0005082294757891507,
+      "loss": 3.1112,
+      "step": 15659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5316805243492126,
+      "learning_rate": 0.0005082183921010922,
+      "loss": 3.1643,
+      "step": 15660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4953380227088928,
+      "learning_rate": 0.0005082073078646266,
+      "loss": 3.2955,
+      "step": 15661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5112192034721375,
+      "learning_rate": 0.000508196223079783,
+      "loss": 3.2071,
+      "step": 15662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5044480562210083,
+      "learning_rate": 0.0005081851377465907,
+      "loss": 3.0958,
+      "step": 15663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6140612363815308,
+      "learning_rate": 0.0005081740518650787,
+      "loss": 3.3124,
+      "step": 15664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5115430951118469,
+      "learning_rate": 0.0005081629654352763,
+      "loss": 3.1975,
+      "step": 15665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5128433108329773,
+      "learning_rate": 0.0005081518784572127,
+      "loss": 3.1859,
+      "step": 15666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49921858310699463,
+      "learning_rate": 0.0005081407909309171,
+      "loss": 3.1841,
+      "step": 15667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5513691306114197,
+      "learning_rate": 0.0005081297028564189,
+      "loss": 3.032,
+      "step": 15668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5237788558006287,
+      "learning_rate": 0.0005081186142337469,
+      "loss": 3.164,
+      "step": 15669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5211585164070129,
+      "learning_rate": 0.0005081075250629306,
+      "loss": 3.2778,
+      "step": 15670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5375601053237915,
+      "learning_rate": 0.0005080964353439991,
+      "loss": 3.2492,
+      "step": 15671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5239969491958618,
+      "learning_rate": 0.0005080853450769817,
+      "loss": 2.856,
+      "step": 15672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5531800389289856,
+      "learning_rate": 0.0005080742542619075,
+      "loss": 3.125,
+      "step": 15673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4881918728351593,
+      "learning_rate": 0.0005080631628988058,
+      "loss": 3.3741,
+      "step": 15674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5813001394271851,
+      "learning_rate": 0.0005080520709877057,
+      "loss": 3.2355,
+      "step": 15675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4898727238178253,
+      "learning_rate": 0.0005080409785286366,
+      "loss": 3.0873,
+      "step": 15676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4959816038608551,
+      "learning_rate": 0.0005080298855216275,
+      "loss": 3.1912,
+      "step": 15677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5205987691879272,
+      "learning_rate": 0.0005080187919667078,
+      "loss": 3.3549,
+      "step": 15678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5168250203132629,
+      "learning_rate": 0.0005080076978639065,
+      "loss": 3.2099,
+      "step": 15679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5724286437034607,
+      "learning_rate": 0.000507996603213253,
+      "loss": 3.0972,
+      "step": 15680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5305948257446289,
+      "learning_rate": 0.0005079855080147766,
+      "loss": 3.1634,
+      "step": 15681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5316970348358154,
+      "learning_rate": 0.0005079744122685063,
+      "loss": 3.1846,
+      "step": 15682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5323175191879272,
+      "learning_rate": 0.0005079633159744715,
+      "loss": 3.1129,
+      "step": 15683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49948516488075256,
+      "learning_rate": 0.0005079522191327012,
+      "loss": 3.2589,
+      "step": 15684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5208277702331543,
+      "learning_rate": 0.0005079411217432249,
+      "loss": 3.0243,
+      "step": 15685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5961699485778809,
+      "learning_rate": 0.0005079300238060717,
+      "loss": 2.9893,
+      "step": 15686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5415461659431458,
+      "learning_rate": 0.0005079189253212708,
+      "loss": 3.1732,
+      "step": 15687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4872366487979889,
+      "learning_rate": 0.0005079078262888515,
+      "loss": 3.1972,
+      "step": 15688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.575570285320282,
+      "learning_rate": 0.0005078967267088429,
+      "loss": 2.9043,
+      "step": 15689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.520625114440918,
+      "learning_rate": 0.0005078856265812745,
+      "loss": 3.0182,
+      "step": 15690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5354496240615845,
+      "learning_rate": 0.0005078745259061752,
+      "loss": 3.2163,
+      "step": 15691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5106788873672485,
+      "learning_rate": 0.0005078634246835745,
+      "loss": 2.838,
+      "step": 15692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5290341377258301,
+      "learning_rate": 0.0005078523229135016,
+      "loss": 3.2877,
+      "step": 15693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5494564175605774,
+      "learning_rate": 0.0005078412205959856,
+      "loss": 3.204,
+      "step": 15694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5220087766647339,
+      "learning_rate": 0.0005078301177310557,
+      "loss": 3.3395,
+      "step": 15695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.573491096496582,
+      "learning_rate": 0.0005078190143187415,
+      "loss": 3.2214,
+      "step": 15696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5294093489646912,
+      "learning_rate": 0.0005078079103590719,
+      "loss": 3.1203,
+      "step": 15697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.46979430317878723,
+      "learning_rate": 0.0005077968058520762,
+      "loss": 2.9733,
+      "step": 15698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49335652589797974,
+      "learning_rate": 0.0005077857007977838,
+      "loss": 3.4289,
+      "step": 15699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.524200439453125,
+      "learning_rate": 0.0005077745951962238,
+      "loss": 3.0115,
+      "step": 15700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.522419810295105,
+      "learning_rate": 0.0005077634890474255,
+      "loss": 3.1244,
+      "step": 15701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5147000551223755,
+      "learning_rate": 0.0005077523823514183,
+      "loss": 3.2699,
+      "step": 15702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5115140676498413,
+      "learning_rate": 0.0005077412751082311,
+      "loss": 3.0467,
+      "step": 15703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5686523914337158,
+      "learning_rate": 0.0005077301673178934,
+      "loss": 3.344,
+      "step": 15704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5637875199317932,
+      "learning_rate": 0.0005077190589804346,
+      "loss": 3.1739,
+      "step": 15705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5621719360351562,
+      "learning_rate": 0.0005077079500958836,
+      "loss": 3.2666,
+      "step": 15706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5126227736473083,
+      "learning_rate": 0.0005076968406642699,
+      "loss": 3.3009,
+      "step": 15707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5188931822776794,
+      "learning_rate": 0.0005076857306856227,
+      "loss": 3.0482,
+      "step": 15708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5244320631027222,
+      "learning_rate": 0.0005076746201599712,
+      "loss": 3.2294,
+      "step": 15709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5187132954597473,
+      "learning_rate": 0.0005076635090873448,
+      "loss": 3.364,
+      "step": 15710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4979041814804077,
+      "learning_rate": 0.0005076523974677725,
+      "loss": 3.2371,
+      "step": 15711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5199201703071594,
+      "learning_rate": 0.000507641285301284,
+      "loss": 3.3204,
+      "step": 15712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5369486212730408,
+      "learning_rate": 0.0005076301725879082,
+      "loss": 3.1625,
+      "step": 15713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5636427998542786,
+      "learning_rate": 0.0005076190593276743,
+      "loss": 3.2081,
+      "step": 15714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5089654922485352,
+      "learning_rate": 0.0005076079455206121,
+      "loss": 3.3712,
+      "step": 15715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48416969180107117,
+      "learning_rate": 0.0005075968311667502,
+      "loss": 3.1725,
+      "step": 15716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5651270151138306,
+      "learning_rate": 0.0005075857162661184,
+      "loss": 3.3323,
+      "step": 15717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5310222506523132,
+      "learning_rate": 0.0005075746008187458,
+      "loss": 3.2754,
+      "step": 15718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5395601987838745,
+      "learning_rate": 0.0005075634848246616,
+      "loss": 3.352,
+      "step": 15719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.539057731628418,
+      "learning_rate": 0.0005075523682838951,
+      "loss": 3.0264,
+      "step": 15720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5499597191810608,
+      "learning_rate": 0.0005075412511964755,
+      "loss": 3.0386,
+      "step": 15721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5185397267341614,
+      "learning_rate": 0.0005075301335624323,
+      "loss": 3.2835,
+      "step": 15722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49906614422798157,
+      "learning_rate": 0.0005075190153817948,
+      "loss": 3.0069,
+      "step": 15723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5171295404434204,
+      "learning_rate": 0.000507507896654592,
+      "loss": 3.0378,
+      "step": 15724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5171312689781189,
+      "learning_rate": 0.0005074967773808534,
+      "loss": 3.1803,
+      "step": 15725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5338811874389648,
+      "learning_rate": 0.0005074856575606082,
+      "loss": 3.386,
+      "step": 15726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5010255575180054,
+      "learning_rate": 0.0005074745371938857,
+      "loss": 3.3692,
+      "step": 15727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6022095084190369,
+      "learning_rate": 0.0005074634162807152,
+      "loss": 3.1641,
+      "step": 15728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.534349262714386,
+      "learning_rate": 0.0005074522948211259,
+      "loss": 3.0681,
+      "step": 15729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5148655772209167,
+      "learning_rate": 0.0005074411728151473,
+      "loss": 3.319,
+      "step": 15730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49930626153945923,
+      "learning_rate": 0.0005074300502628085,
+      "loss": 3.251,
+      "step": 15731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4870609939098358,
+      "learning_rate": 0.0005074189271641388,
+      "loss": 3.2754,
+      "step": 15732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5289334058761597,
+      "learning_rate": 0.0005074078035191677,
+      "loss": 3.0453,
+      "step": 15733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5236629247665405,
+      "learning_rate": 0.0005073966793279243,
+      "loss": 2.9417,
+      "step": 15734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5354396104812622,
+      "learning_rate": 0.0005073855545904381,
+      "loss": 3.1531,
+      "step": 15735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4817488193511963,
+      "learning_rate": 0.0005073744293067382,
+      "loss": 3.1444,
+      "step": 15736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5554103851318359,
+      "learning_rate": 0.0005073633034768538,
+      "loss": 3.0,
+      "step": 15737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5145366787910461,
+      "learning_rate": 0.0005073521771008145,
+      "loss": 3.187,
+      "step": 15738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48250049352645874,
+      "learning_rate": 0.0005073410501786495,
+      "loss": 3.19,
+      "step": 15739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5110865831375122,
+      "learning_rate": 0.000507329922710388,
+      "loss": 3.203,
+      "step": 15740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5449311137199402,
+      "learning_rate": 0.0005073187946960594,
+      "loss": 3.1025,
+      "step": 15741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5438368916511536,
+      "learning_rate": 0.000507307666135693,
+      "loss": 3.3186,
+      "step": 15742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49835509061813354,
+      "learning_rate": 0.0005072965370293181,
+      "loss": 3.3353,
+      "step": 15743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5272834897041321,
+      "learning_rate": 0.000507285407376964,
+      "loss": 3.0211,
+      "step": 15744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5524513721466064,
+      "learning_rate": 0.0005072742771786601,
+      "loss": 3.2202,
+      "step": 15745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5341085195541382,
+      "learning_rate": 0.0005072631464344355,
+      "loss": 3.0149,
+      "step": 15746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5116642117500305,
+      "learning_rate": 0.0005072520151443197,
+      "loss": 3.5498,
+      "step": 15747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5027918815612793,
+      "learning_rate": 0.0005072408833083421,
+      "loss": 3.1924,
+      "step": 15748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.545111358165741,
+      "learning_rate": 0.0005072297509265319,
+      "loss": 3.1091,
+      "step": 15749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5268412828445435,
+      "learning_rate": 0.0005072186179989184,
+      "loss": 3.3507,
+      "step": 15750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49303969740867615,
+      "learning_rate": 0.0005072074845255309,
+      "loss": 3.1132,
+      "step": 15751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5391672849655151,
+      "learning_rate": 0.0005071963505063988,
+      "loss": 3.2345,
+      "step": 15752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4987182915210724,
+      "learning_rate": 0.0005071852159415513,
+      "loss": 3.0815,
+      "step": 15753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.538796067237854,
+      "learning_rate": 0.0005071740808310179,
+      "loss": 2.9649,
+      "step": 15754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4982840120792389,
+      "learning_rate": 0.0005071629451748277,
+      "loss": 3.2089,
+      "step": 15755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49565887451171875,
+      "learning_rate": 0.0005071518089730103,
+      "loss": 3.3456,
+      "step": 15756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5028196573257446,
+      "learning_rate": 0.0005071406722255948,
+      "loss": 3.1211,
+      "step": 15757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4904448688030243,
+      "learning_rate": 0.0005071295349326108,
+      "loss": 3.2393,
+      "step": 15758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5175606608390808,
+      "learning_rate": 0.0005071183970940874,
+      "loss": 3.4368,
+      "step": 15759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5216371417045593,
+      "learning_rate": 0.0005071072587100539,
+      "loss": 3.0616,
+      "step": 15760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5647536516189575,
+      "learning_rate": 0.0005070961197805399,
+      "loss": 3.3533,
+      "step": 15761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5448745489120483,
+      "learning_rate": 0.0005070849803055744,
+      "loss": 3.2618,
+      "step": 15762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5875873565673828,
+      "learning_rate": 0.0005070738402851871,
+      "loss": 3.2061,
+      "step": 15763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5121092200279236,
+      "learning_rate": 0.0005070626997194071,
+      "loss": 3.2049,
+      "step": 15764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5112888813018799,
+      "learning_rate": 0.0005070515586082638,
+      "loss": 2.9788,
+      "step": 15765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.537285327911377,
+      "learning_rate": 0.0005070404169517865,
+      "loss": 3.0783,
+      "step": 15766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5411680936813354,
+      "learning_rate": 0.0005070292747500045,
+      "loss": 2.9802,
+      "step": 15767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5896318554878235,
+      "learning_rate": 0.0005070181320029474,
+      "loss": 3.3174,
+      "step": 15768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5077767968177795,
+      "learning_rate": 0.0005070069887106442,
+      "loss": 3.1943,
+      "step": 15769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5382665395736694,
+      "learning_rate": 0.0005069958448731247,
+      "loss": 3.1708,
+      "step": 15770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5162548422813416,
+      "learning_rate": 0.0005069847004904178,
+      "loss": 3.2341,
+      "step": 15771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5016313195228577,
+      "learning_rate": 0.0005069735555625531,
+      "loss": 3.0787,
+      "step": 15772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5164852738380432,
+      "learning_rate": 0.0005069624100895598,
+      "loss": 3.2173,
+      "step": 15773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5445564389228821,
+      "learning_rate": 0.0005069512640714673,
+      "loss": 3.1206,
+      "step": 15774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5352805256843567,
+      "learning_rate": 0.0005069401175083053,
+      "loss": 3.3894,
+      "step": 15775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5283602476119995,
+      "learning_rate": 0.0005069289704001025,
+      "loss": 3.3639,
+      "step": 15776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5328353643417358,
+      "learning_rate": 0.0005069178227468888,
+      "loss": 3.1275,
+      "step": 15777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5417640805244446,
+      "learning_rate": 0.0005069066745486934,
+      "loss": 3.2893,
+      "step": 15778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5066238641738892,
+      "learning_rate": 0.0005068955258055455,
+      "loss": 3.1368,
+      "step": 15779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5763447284698486,
+      "learning_rate": 0.0005068843765174747,
+      "loss": 2.9799,
+      "step": 15780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5056501030921936,
+      "learning_rate": 0.0005068732266845103,
+      "loss": 3.3112,
+      "step": 15781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5100968480110168,
+      "learning_rate": 0.0005068620763066816,
+      "loss": 3.0971,
+      "step": 15782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.517602264881134,
+      "learning_rate": 0.000506850925384018,
+      "loss": 3.315,
+      "step": 15783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5092670321464539,
+      "learning_rate": 0.0005068397739165488,
+      "loss": 3.2111,
+      "step": 15784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5618929266929626,
+      "learning_rate": 0.0005068286219043035,
+      "loss": 3.1347,
+      "step": 15785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5268994569778442,
+      "learning_rate": 0.0005068174693473115,
+      "loss": 3.1182,
+      "step": 15786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5237292647361755,
+      "learning_rate": 0.000506806316245602,
+      "loss": 3.0985,
+      "step": 15787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5168699622154236,
+      "learning_rate": 0.0005067951625992044,
+      "loss": 3.1803,
+      "step": 15788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47693702578544617,
+      "learning_rate": 0.0005067840084081482,
+      "loss": 3.1477,
+      "step": 15789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5285649299621582,
+      "learning_rate": 0.0005067728536724627,
+      "loss": 3.352,
+      "step": 15790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49356603622436523,
+      "learning_rate": 0.0005067616983921774,
+      "loss": 2.9803,
+      "step": 15791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.512496829032898,
+      "learning_rate": 0.0005067505425673215,
+      "loss": 3.0556,
+      "step": 15792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5645531415939331,
+      "learning_rate": 0.0005067393861979244,
+      "loss": 3.2463,
+      "step": 15793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5089481472969055,
+      "learning_rate": 0.0005067282292840156,
+      "loss": 3.0876,
+      "step": 15794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5088877081871033,
+      "learning_rate": 0.0005067170718256246,
+      "loss": 3.4672,
+      "step": 15795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5133219957351685,
+      "learning_rate": 0.0005067059138227803,
+      "loss": 3.2869,
+      "step": 15796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5056898593902588,
+      "learning_rate": 0.0005066947552755126,
+      "loss": 3.2511,
+      "step": 15797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5130456686019897,
+      "learning_rate": 0.0005066835961838507,
+      "loss": 3.1219,
+      "step": 15798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5243682861328125,
+      "learning_rate": 0.0005066724365478239,
+      "loss": 3.0821,
+      "step": 15799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5532961487770081,
+      "learning_rate": 0.0005066612763674617,
+      "loss": 3.3234,
+      "step": 15800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5712582468986511,
+      "learning_rate": 0.0005066501156427936,
+      "loss": 3.2391,
+      "step": 15801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.531348705291748,
+      "learning_rate": 0.0005066389543738487,
+      "loss": 3.485,
+      "step": 15802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5381166338920593,
+      "learning_rate": 0.0005066277925606566,
+      "loss": 3.2193,
+      "step": 15803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5312127470970154,
+      "learning_rate": 0.0005066166302032468,
+      "loss": 3.097,
+      "step": 15804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5262129902839661,
+      "learning_rate": 0.0005066054673016484,
+      "loss": 3.1406,
+      "step": 15805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5441365838050842,
+      "learning_rate": 0.0005065943038558909,
+      "loss": 3.0766,
+      "step": 15806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4887438714504242,
+      "learning_rate": 0.0005065831398660039,
+      "loss": 3.1556,
+      "step": 15807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5036865472793579,
+      "learning_rate": 0.0005065719753320167,
+      "loss": 3.0322,
+      "step": 15808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5196355581283569,
+      "learning_rate": 0.0005065608102539586,
+      "loss": 3.0775,
+      "step": 15809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5290484428405762,
+      "learning_rate": 0.000506549644631859,
+      "loss": 3.1921,
+      "step": 15810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5046884417533875,
+      "learning_rate": 0.0005065384784657476,
+      "loss": 3.0588,
+      "step": 15811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5305639505386353,
+      "learning_rate": 0.0005065273117556534,
+      "loss": 3.1943,
+      "step": 15812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5117027163505554,
+      "learning_rate": 0.000506516144501606,
+      "loss": 3.1293,
+      "step": 15813
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52192223072052,
+      "learning_rate": 0.0005065049767036349,
+      "loss": 3.0501,
+      "step": 15814
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5212222337722778,
+      "learning_rate": 0.0005064938083617695,
+      "loss": 3.0072,
+      "step": 15815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5260198712348938,
+      "learning_rate": 0.0005064826394760391,
+      "loss": 3.2323,
+      "step": 15816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5320043563842773,
+      "learning_rate": 0.0005064714700464731,
+      "loss": 3.187,
+      "step": 15817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5128186941146851,
+      "learning_rate": 0.000506460300073101,
+      "loss": 3.4666,
+      "step": 15818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5312202572822571,
+      "learning_rate": 0.0005064491295559523,
+      "loss": 3.3288,
+      "step": 15819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5054448843002319,
+      "learning_rate": 0.0005064379584950562,
+      "loss": 3.2213,
+      "step": 15820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6765322685241699,
+      "learning_rate": 0.0005064267868904423,
+      "loss": 3.3935,
+      "step": 15821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.525917112827301,
+      "learning_rate": 0.00050641561474214,
+      "loss": 3.1906,
+      "step": 15822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5486940741539001,
+      "learning_rate": 0.0005064044420501787,
+      "loss": 3.1634,
+      "step": 15823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5462398529052734,
+      "learning_rate": 0.0005063932688145877,
+      "loss": 3.2199,
+      "step": 15824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5118207335472107,
+      "learning_rate": 0.0005063820950353966,
+      "loss": 3.217,
+      "step": 15825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5313277840614319,
+      "learning_rate": 0.000506370920712635,
+      "loss": 3.1615,
+      "step": 15826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49920400977134705,
+      "learning_rate": 0.0005063597458463319,
+      "loss": 3.3053,
+      "step": 15827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4805811047554016,
+      "learning_rate": 0.0005063485704365169,
+      "loss": 3.2122,
+      "step": 15828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5627545714378357,
+      "learning_rate": 0.0005063373944832196,
+      "loss": 3.0398,
+      "step": 15829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5166324377059937,
+      "learning_rate": 0.0005063262179864692,
+      "loss": 3.2995,
+      "step": 15830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8689070343971252,
+      "learning_rate": 0.0005063150409462954,
+      "loss": 3.2035,
+      "step": 15831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5488486289978027,
+      "learning_rate": 0.0005063038633627274,
+      "loss": 3.0854,
+      "step": 15832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5090166926383972,
+      "learning_rate": 0.0005062926852357947,
+      "loss": 3.479,
+      "step": 15833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.515069842338562,
+      "learning_rate": 0.0005062815065655269,
+      "loss": 2.9501,
+      "step": 15834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5386884212493896,
+      "learning_rate": 0.0005062703273519531,
+      "loss": 3.3355,
+      "step": 15835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5605966448783875,
+      "learning_rate": 0.0005062591475951031,
+      "loss": 3.2428,
+      "step": 15836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.580772876739502,
+      "learning_rate": 0.0005062479672950063,
+      "loss": 3.1849,
+      "step": 15837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5151135325431824,
+      "learning_rate": 0.0005062367864516919,
+      "loss": 3.139,
+      "step": 15838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5454901456832886,
+      "learning_rate": 0.0005062256050651895,
+      "loss": 3.0489,
+      "step": 15839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5578955411911011,
+      "learning_rate": 0.0005062144231355285,
+      "loss": 3.2888,
+      "step": 15840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5507684350013733,
+      "learning_rate": 0.0005062032406627384,
+      "loss": 3.2061,
+      "step": 15841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5147064924240112,
+      "learning_rate": 0.0005061920576468488,
+      "loss": 3.3796,
+      "step": 15842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5263797044754028,
+      "learning_rate": 0.0005061808740878889,
+      "loss": 3.0018,
+      "step": 15843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5236734747886658,
+      "learning_rate": 0.0005061696899858883,
+      "loss": 3.2627,
+      "step": 15844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5589058995246887,
+      "learning_rate": 0.0005061585053408764,
+      "loss": 3.3151,
+      "step": 15845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5322604179382324,
+      "learning_rate": 0.0005061473201528826,
+      "loss": 3.1628,
+      "step": 15846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5170190334320068,
+      "learning_rate": 0.0005061361344219365,
+      "loss": 3.2358,
+      "step": 15847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5249111652374268,
+      "learning_rate": 0.0005061249481480675,
+      "loss": 3.1819,
+      "step": 15848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5331235527992249,
+      "learning_rate": 0.000506113761331305,
+      "loss": 3.2316,
+      "step": 15849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5436180830001831,
+      "learning_rate": 0.0005061025739716786,
+      "loss": 2.9909,
+      "step": 15850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5154057145118713,
+      "learning_rate": 0.0005060913860692177,
+      "loss": 3.0543,
+      "step": 15851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5232006311416626,
+      "learning_rate": 0.0005060801976239516,
+      "loss": 3.1253,
+      "step": 15852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.605620801448822,
+      "learning_rate": 0.0005060690086359101,
+      "loss": 3.3018,
+      "step": 15853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4980825185775757,
+      "learning_rate": 0.0005060578191051225,
+      "loss": 3.0301,
+      "step": 15854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5214237570762634,
+      "learning_rate": 0.000506046629031618,
+      "loss": 3.4541,
+      "step": 15855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5480488538742065,
+      "learning_rate": 0.0005060354384154265,
+      "loss": 3.018,
+      "step": 15856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5750018358230591,
+      "learning_rate": 0.0005060242472565774,
+      "loss": 3.3321,
+      "step": 15857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5459998250007629,
+      "learning_rate": 0.0005060130555550999,
+      "loss": 3.1318,
+      "step": 15858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5523403882980347,
+      "learning_rate": 0.0005060018633110238,
+      "loss": 3.311,
+      "step": 15859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5640004873275757,
+      "learning_rate": 0.0005059906705243783,
+      "loss": 3.1297,
+      "step": 15860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5225830674171448,
+      "learning_rate": 0.0005059794771951931,
+      "loss": 3.2224,
+      "step": 15861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49763786792755127,
+      "learning_rate": 0.0005059682833234977,
+      "loss": 3.2197,
+      "step": 15862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.517906665802002,
+      "learning_rate": 0.0005059570889093214,
+      "loss": 3.1153,
+      "step": 15863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5416662096977234,
+      "learning_rate": 0.0005059458939526937,
+      "loss": 3.3586,
+      "step": 15864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48512333631515503,
+      "learning_rate": 0.0005059346984536442,
+      "loss": 3.1849,
+      "step": 15865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5373478531837463,
+      "learning_rate": 0.0005059235024122024,
+      "loss": 3.0875,
+      "step": 15866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4900364577770233,
+      "learning_rate": 0.0005059123058283976,
+      "loss": 3.2715,
+      "step": 15867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5320209860801697,
+      "learning_rate": 0.0005059011087022595,
+      "loss": 3.3284,
+      "step": 15868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5448189973831177,
+      "learning_rate": 0.0005058899110338175,
+      "loss": 3.1436,
+      "step": 15869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5075494647026062,
+      "learning_rate": 0.0005058787128231011,
+      "loss": 3.2013,
+      "step": 15870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5199207663536072,
+      "learning_rate": 0.0005058675140701398,
+      "loss": 3.1813,
+      "step": 15871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49894481897354126,
+      "learning_rate": 0.000505856314774963,
+      "loss": 3.1914,
+      "step": 15872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5516239404678345,
+      "learning_rate": 0.0005058451149376003,
+      "loss": 3.2814,
+      "step": 15873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5630885362625122,
+      "learning_rate": 0.0005058339145580813,
+      "loss": 3.219,
+      "step": 15874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5175797343254089,
+      "learning_rate": 0.0005058227136364353,
+      "loss": 3.3844,
+      "step": 15875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5631835460662842,
+      "learning_rate": 0.0005058115121726918,
+      "loss": 3.1561,
+      "step": 15876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5062498450279236,
+      "learning_rate": 0.0005058003101668806,
+      "loss": 3.1483,
+      "step": 15877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.531692385673523,
+      "learning_rate": 0.0005057891076190309,
+      "loss": 3.3115,
+      "step": 15878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.491470068693161,
+      "learning_rate": 0.0005057779045291723,
+      "loss": 3.1625,
+      "step": 15879
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5266138315200806,
+      "learning_rate": 0.0005057667008973341,
+      "loss": 3.2485,
+      "step": 15880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5649579763412476,
+      "learning_rate": 0.0005057554967235463,
+      "loss": 3.2064,
+      "step": 15881
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4945317804813385,
+      "learning_rate": 0.000505744292007838,
+      "loss": 3.077,
+      "step": 15882
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5235192775726318,
+      "learning_rate": 0.0005057330867502389,
+      "loss": 3.2043,
+      "step": 15883
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4946229159832001,
+      "learning_rate": 0.0005057218809507782,
+      "loss": 3.0865,
+      "step": 15884
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5265135169029236,
+      "learning_rate": 0.0005057106746094859,
+      "loss": 3.2467,
+      "step": 15885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4959494173526764,
+      "learning_rate": 0.0005056994677263913,
+      "loss": 3.212,
+      "step": 15886
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5413668155670166,
+      "learning_rate": 0.0005056882603015237,
+      "loss": 3.0447,
+      "step": 15887
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4981670081615448,
+      "learning_rate": 0.000505677052334913,
+      "loss": 3.1375,
+      "step": 15888
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5313765406608582,
+      "learning_rate": 0.0005056658438265884,
+      "loss": 3.1866,
+      "step": 15889
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5187064409255981,
+      "learning_rate": 0.0005056546347765796,
+      "loss": 3.3122,
+      "step": 15890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5525301098823547,
+      "learning_rate": 0.0005056434251849161,
+      "loss": 3.1409,
+      "step": 15891
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5722075700759888,
+      "learning_rate": 0.0005056322150516273,
+      "loss": 3.1696,
+      "step": 15892
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5104582905769348,
+      "learning_rate": 0.0005056210043767428,
+      "loss": 3.09,
+      "step": 15893
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5645208358764648,
+      "learning_rate": 0.0005056097931602923,
+      "loss": 3.2814,
+      "step": 15894
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.513430118560791,
+      "learning_rate": 0.000505598581402305,
+      "loss": 3.0936,
+      "step": 15895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5425407886505127,
+      "learning_rate": 0.0005055873691028108,
+      "loss": 3.3683,
+      "step": 15896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48319244384765625,
+      "learning_rate": 0.0005055761562618388,
+      "loss": 3.1959,
+      "step": 15897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.518714964389801,
+      "learning_rate": 0.000505564942879419,
+      "loss": 3.0281,
+      "step": 15898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5663323998451233,
+      "learning_rate": 0.0005055537289555806,
+      "loss": 3.3935,
+      "step": 15899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5486282706260681,
+      "learning_rate": 0.0005055425144903532,
+      "loss": 3.3414,
+      "step": 15900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5473208427429199,
+      "learning_rate": 0.0005055312994837664,
+      "loss": 3.3134,
+      "step": 15901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5325106382369995,
+      "learning_rate": 0.0005055200839358497,
+      "loss": 3.1135,
+      "step": 15902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5353180170059204,
+      "learning_rate": 0.0005055088678466327,
+      "loss": 3.1344,
+      "step": 15903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5123947858810425,
+      "learning_rate": 0.0005054976512161449,
+      "loss": 3.0725,
+      "step": 15904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5428218245506287,
+      "learning_rate": 0.0005054864340444158,
+      "loss": 3.244,
+      "step": 15905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5174368023872375,
+      "learning_rate": 0.000505475216331475,
+      "loss": 3.2212,
+      "step": 15906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5281883478164673,
+      "learning_rate": 0.0005054639980773519,
+      "loss": 3.1813,
+      "step": 15907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.564376175403595,
+      "learning_rate": 0.0005054527792820764,
+      "loss": 3.0156,
+      "step": 15908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5363826751708984,
+      "learning_rate": 0.0005054415599456776,
+      "loss": 3.3758,
+      "step": 15909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.514251708984375,
+      "learning_rate": 0.0005054303400681855,
+      "loss": 3.2013,
+      "step": 15910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5030025839805603,
+      "learning_rate": 0.0005054191196496293,
+      "loss": 3.2198,
+      "step": 15911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5085748434066772,
+      "learning_rate": 0.0005054078986900387,
+      "loss": 3.2473,
+      "step": 15912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5087880492210388,
+      "learning_rate": 0.0005053966771894432,
+      "loss": 3.2511,
+      "step": 15913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.550105631351471,
+      "learning_rate": 0.0005053854551478723,
+      "loss": 3.0949,
+      "step": 15914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5327011942863464,
+      "learning_rate": 0.0005053742325653557,
+      "loss": 3.2427,
+      "step": 15915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5084472894668579,
+      "learning_rate": 0.0005053630094419228,
+      "loss": 3.0718,
+      "step": 15916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5227506160736084,
+      "learning_rate": 0.0005053517857776035,
+      "loss": 3.1136,
+      "step": 15917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5210598707199097,
+      "learning_rate": 0.000505340561572427,
+      "loss": 3.266,
+      "step": 15918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5185865759849548,
+      "learning_rate": 0.000505329336826423,
+      "loss": 3.0593,
+      "step": 15919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5381935834884644,
+      "learning_rate": 0.0005053181115396209,
+      "loss": 3.1447,
+      "step": 15920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5370237231254578,
+      "learning_rate": 0.0005053068857120505,
+      "loss": 3.0868,
+      "step": 15921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5445959568023682,
+      "learning_rate": 0.0005052956593437413,
+      "loss": 3.2731,
+      "step": 15922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5113406181335449,
+      "learning_rate": 0.0005052844324347228,
+      "loss": 3.2167,
+      "step": 15923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5293589234352112,
+      "learning_rate": 0.0005052732049850246,
+      "loss": 3.3285,
+      "step": 15924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.523661196231842,
+      "learning_rate": 0.0005052619769946764,
+      "loss": 3.2882,
+      "step": 15925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5041723251342773,
+      "learning_rate": 0.0005052507484637076,
+      "loss": 3.0888,
+      "step": 15926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5257124900817871,
+      "learning_rate": 0.0005052395193921478,
+      "loss": 3.2741,
+      "step": 15927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48543840646743774,
+      "learning_rate": 0.0005052282897800266,
+      "loss": 3.3522,
+      "step": 15928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5463740825653076,
+      "learning_rate": 0.0005052170596273735,
+      "loss": 3.3716,
+      "step": 15929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5675771236419678,
+      "learning_rate": 0.0005052058289342184,
+      "loss": 3.0156,
+      "step": 15930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.511083722114563,
+      "learning_rate": 0.0005051945977005905,
+      "loss": 3.1242,
+      "step": 15931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5357405543327332,
+      "learning_rate": 0.0005051833659265195,
+      "loss": 3.2083,
+      "step": 15932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5028762221336365,
+      "learning_rate": 0.000505172133612035,
+      "loss": 3.1726,
+      "step": 15933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5303201675415039,
+      "learning_rate": 0.0005051609007571666,
+      "loss": 3.2378,
+      "step": 15934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5492925047874451,
+      "learning_rate": 0.0005051496673619439,
+      "loss": 3.3461,
+      "step": 15935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5268462896347046,
+      "learning_rate": 0.0005051384334263965,
+      "loss": 3.1529,
+      "step": 15936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5270599722862244,
+      "learning_rate": 0.0005051271989505538,
+      "loss": 3.1615,
+      "step": 15937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5161546468734741,
+      "learning_rate": 0.0005051159639344456,
+      "loss": 3.2024,
+      "step": 15938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5089823603630066,
+      "learning_rate": 0.0005051047283781015,
+      "loss": 3.1198,
+      "step": 15939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48041659593582153,
+      "learning_rate": 0.000505093492281551,
+      "loss": 3.3224,
+      "step": 15940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5333757400512695,
+      "learning_rate": 0.0005050822556448236,
+      "loss": 3.1934,
+      "step": 15941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5357717275619507,
+      "learning_rate": 0.000505071018467949,
+      "loss": 3.2549,
+      "step": 15942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5316967368125916,
+      "learning_rate": 0.0005050597807509569,
+      "loss": 3.0993,
+      "step": 15943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5037146210670471,
+      "learning_rate": 0.0005050485424938769,
+      "loss": 3.2657,
+      "step": 15944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5503755807876587,
+      "learning_rate": 0.0005050373036967384,
+      "loss": 3.2414,
+      "step": 15945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5169128179550171,
+      "learning_rate": 0.0005050260643595711,
+      "loss": 2.9953,
+      "step": 15946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5049517154693604,
+      "learning_rate": 0.0005050148244824045,
+      "loss": 3.2635,
+      "step": 15947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4886893928050995,
+      "learning_rate": 0.0005050035840652684,
+      "loss": 3.1339,
+      "step": 15948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5033433437347412,
+      "learning_rate": 0.0005049923431081924,
+      "loss": 3.0855,
+      "step": 15949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5404435396194458,
+      "learning_rate": 0.0005049811016112059,
+      "loss": 3.0479,
+      "step": 15950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5240478515625,
+      "learning_rate": 0.0005049698595743387,
+      "loss": 3.0661,
+      "step": 15951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5057730674743652,
+      "learning_rate": 0.0005049586169976202,
+      "loss": 3.3078,
+      "step": 15952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5399793982505798,
+      "learning_rate": 0.0005049473738810803,
+      "loss": 3.0763,
+      "step": 15953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4813418984413147,
+      "learning_rate": 0.0005049361302247485,
+      "loss": 3.1835,
+      "step": 15954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5555798411369324,
+      "learning_rate": 0.0005049248860286542,
+      "loss": 3.1501,
+      "step": 15955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5001389980316162,
+      "learning_rate": 0.0005049136412928273,
+      "loss": 3.143,
+      "step": 15956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5168265700340271,
+      "learning_rate": 0.0005049023960172973,
+      "loss": 3.2054,
+      "step": 15957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5343358516693115,
+      "learning_rate": 0.0005048911502020938,
+      "loss": 3.292,
+      "step": 15958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5781263113021851,
+      "learning_rate": 0.0005048799038472465,
+      "loss": 3.1776,
+      "step": 15959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5077539682388306,
+      "learning_rate": 0.0005048686569527848,
+      "loss": 3.1976,
+      "step": 15960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5165067315101624,
+      "learning_rate": 0.0005048574095187385,
+      "loss": 3.0778,
+      "step": 15961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5727941393852234,
+      "learning_rate": 0.0005048461615451374,
+      "loss": 3.2127,
+      "step": 15962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5406970381736755,
+      "learning_rate": 0.0005048349130320108,
+      "loss": 3.2311,
+      "step": 15963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5371522903442383,
+      "learning_rate": 0.0005048236639793885,
+      "loss": 3.1551,
+      "step": 15964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5899889469146729,
+      "learning_rate": 0.0005048124143873001,
+      "loss": 3.213,
+      "step": 15965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5044421553611755,
+      "learning_rate": 0.0005048011642557751,
+      "loss": 3.383,
+      "step": 15966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5018906593322754,
+      "learning_rate": 0.0005047899135848435,
+      "loss": 3.4013,
+      "step": 15967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47671008110046387,
+      "learning_rate": 0.0005047786623745345,
+      "loss": 3.3241,
+      "step": 15968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5087835788726807,
+      "learning_rate": 0.0005047674106248779,
+      "loss": 3.223,
+      "step": 15969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5542645454406738,
+      "learning_rate": 0.0005047561583359034,
+      "loss": 3.206,
+      "step": 15970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4945680499076843,
+      "learning_rate": 0.0005047449055076407,
+      "loss": 3.1473,
+      "step": 15971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4865614175796509,
+      "learning_rate": 0.0005047336521401191,
+      "loss": 3.1573,
+      "step": 15972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4855407476425171,
+      "learning_rate": 0.0005047223982333685,
+      "loss": 3.2028,
+      "step": 15973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5474462509155273,
+      "learning_rate": 0.0005047111437874186,
+      "loss": 3.4711,
+      "step": 15974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5180954933166504,
+      "learning_rate": 0.0005046998888022988,
+      "loss": 3.0933,
+      "step": 15975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5082975625991821,
+      "learning_rate": 0.0005046886332780392,
+      "loss": 3.1662,
+      "step": 15976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6138001680374146,
+      "learning_rate": 0.0005046773772146688,
+      "loss": 3.0991,
+      "step": 15977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5285327434539795,
+      "learning_rate": 0.0005046661206122178,
+      "loss": 3.1361,
+      "step": 15978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5196309685707092,
+      "learning_rate": 0.0005046548634707155,
+      "loss": 3.0654,
+      "step": 15979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5266782641410828,
+      "learning_rate": 0.0005046436057901917,
+      "loss": 3.2551,
+      "step": 15980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5191410183906555,
+      "learning_rate": 0.0005046323475706761,
+      "loss": 3.3328,
+      "step": 15981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5638286471366882,
+      "learning_rate": 0.0005046210888121982,
+      "loss": 3.324,
+      "step": 15982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5119996666908264,
+      "learning_rate": 0.0005046098295147877,
+      "loss": 3.0339,
+      "step": 15983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5803064703941345,
+      "learning_rate": 0.0005045985696784743,
+      "loss": 3.3781,
+      "step": 15984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5395102500915527,
+      "learning_rate": 0.0005045873093032878,
+      "loss": 3.4653,
+      "step": 15985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.508126437664032,
+      "learning_rate": 0.0005045760483892575,
+      "loss": 3.1083,
+      "step": 15986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48580315709114075,
+      "learning_rate": 0.0005045647869364134,
+      "loss": 3.0414,
+      "step": 15987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5052012205123901,
+      "learning_rate": 0.0005045535249447848,
+      "loss": 2.9321,
+      "step": 15988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.575672447681427,
+      "learning_rate": 0.0005045422624144019,
+      "loss": 3.2742,
+      "step": 15989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5062314867973328,
+      "learning_rate": 0.0005045309993452939,
+      "loss": 3.2034,
+      "step": 15990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5185468792915344,
+      "learning_rate": 0.0005045197357374906,
+      "loss": 3.2348,
+      "step": 15991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5151081681251526,
+      "learning_rate": 0.0005045084715910216,
+      "loss": 3.2609,
+      "step": 15992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4942055642604828,
+      "learning_rate": 0.0005044972069059167,
+      "loss": 3.2212,
+      "step": 15993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4843122065067291,
+      "learning_rate": 0.0005044859416822056,
+      "loss": 3.1712,
+      "step": 15994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5047922134399414,
+      "learning_rate": 0.0005044746759199178,
+      "loss": 3.2883,
+      "step": 15995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5109986066818237,
+      "learning_rate": 0.0005044634096190831,
+      "loss": 2.9292,
+      "step": 15996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5134650468826294,
+      "learning_rate": 0.0005044521427797311,
+      "loss": 3.1455,
+      "step": 15997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5431228280067444,
+      "learning_rate": 0.0005044408754018915,
+      "loss": 3.1955,
+      "step": 15998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5586522817611694,
+      "learning_rate": 0.000504429607485594,
+      "loss": 3.2098,
+      "step": 15999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5322924852371216,
+      "learning_rate": 0.0005044183390308682,
+      "loss": 3.3816,
+      "step": 16000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49801820516586304,
+      "learning_rate": 0.000504407070037744,
+      "loss": 3.3149,
+      "step": 16001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5596066117286682,
+      "learning_rate": 0.0005043958005062506,
+      "loss": 3.1123,
+      "step": 16002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5819718241691589,
+      "learning_rate": 0.0005043845304364183,
+      "loss": 3.3048,
+      "step": 16003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5312378406524658,
+      "learning_rate": 0.0005043732598282763,
+      "loss": 3.2235,
+      "step": 16004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4894445538520813,
+      "learning_rate": 0.0005043619886818546,
+      "loss": 3.2897,
+      "step": 16005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5514426827430725,
+      "learning_rate": 0.0005043507169971827,
+      "loss": 3.0689,
+      "step": 16006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5011432766914368,
+      "learning_rate": 0.0005043394447742903,
+      "loss": 3.2763,
+      "step": 16007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5099719762802124,
+      "learning_rate": 0.0005043281720132072,
+      "loss": 3.2354,
+      "step": 16008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5169076919555664,
+      "learning_rate": 0.000504316898713963,
+      "loss": 3.2652,
+      "step": 16009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5074393153190613,
+      "learning_rate": 0.0005043056248765875,
+      "loss": 3.3298,
+      "step": 16010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5243656635284424,
+      "learning_rate": 0.0005042943505011101,
+      "loss": 3.2304,
+      "step": 16011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5197635293006897,
+      "learning_rate": 0.0005042830755875607,
+      "loss": 3.1894,
+      "step": 16012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5310289263725281,
+      "learning_rate": 0.0005042718001359691,
+      "loss": 3.3684,
+      "step": 16013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.532052218914032,
+      "learning_rate": 0.000504260524146365,
+      "loss": 3.1898,
+      "step": 16014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5254867672920227,
+      "learning_rate": 0.0005042492476187778,
+      "loss": 3.1261,
+      "step": 16015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5229167342185974,
+      "learning_rate": 0.0005042379705532374,
+      "loss": 3.2582,
+      "step": 16016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5486670136451721,
+      "learning_rate": 0.0005042266929497736,
+      "loss": 3.1984,
+      "step": 16017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5969434380531311,
+      "learning_rate": 0.0005042154148084159,
+      "loss": 3.3827,
+      "step": 16018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4866141378879547,
+      "learning_rate": 0.0005042041361291941,
+      "loss": 3.2698,
+      "step": 16019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6796237826347351,
+      "learning_rate": 0.000504192856912138,
+      "loss": 2.9635,
+      "step": 16020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5028907656669617,
+      "learning_rate": 0.0005041815771572772,
+      "loss": 3.237,
+      "step": 16021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5195736885070801,
+      "learning_rate": 0.0005041702968646413,
+      "loss": 3.3098,
+      "step": 16022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.490695059299469,
+      "learning_rate": 0.0005041590160342603,
+      "loss": 3.2567,
+      "step": 16023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5087558031082153,
+      "learning_rate": 0.0005041477346661637,
+      "loss": 3.2146,
+      "step": 16024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5431877374649048,
+      "learning_rate": 0.0005041364527603811,
+      "loss": 3.2565,
+      "step": 16025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6024651527404785,
+      "learning_rate": 0.0005041251703169425,
+      "loss": 3.3531,
+      "step": 16026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5574399828910828,
+      "learning_rate": 0.0005041138873358776,
+      "loss": 3.1318,
+      "step": 16027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5236434936523438,
+      "learning_rate": 0.0005041026038172158,
+      "loss": 3.4343,
+      "step": 16028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5195509195327759,
+      "learning_rate": 0.0005040913197609871,
+      "loss": 3.097,
+      "step": 16029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5851946473121643,
+      "learning_rate": 0.0005040800351672211,
+      "loss": 3.3494,
+      "step": 16030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5439422130584717,
+      "learning_rate": 0.0005040687500359476,
+      "loss": 3.1376,
+      "step": 16031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5359077453613281,
+      "learning_rate": 0.0005040574643671963,
+      "loss": 3.162,
+      "step": 16032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5320112705230713,
+      "learning_rate": 0.0005040461781609969,
+      "loss": 3.3413,
+      "step": 16033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.503227174282074,
+      "learning_rate": 0.0005040348914173791,
+      "loss": 3.2933,
+      "step": 16034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5737831592559814,
+      "learning_rate": 0.0005040236041363728,
+      "loss": 3.3733,
+      "step": 16035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4747275114059448,
+      "learning_rate": 0.0005040123163180075,
+      "loss": 3.1553,
+      "step": 16036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5528791546821594,
+      "learning_rate": 0.0005040010279623129,
+      "loss": 3.1915,
+      "step": 16037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.526457667350769,
+      "learning_rate": 0.000503989739069319,
+      "loss": 3.3059,
+      "step": 16038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5216947197914124,
+      "learning_rate": 0.0005039784496390554,
+      "loss": 3.1353,
+      "step": 16039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5180486440658569,
+      "learning_rate": 0.0005039671596715517,
+      "loss": 3.0681,
+      "step": 16040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5172383785247803,
+      "learning_rate": 0.0005039558691668378,
+      "loss": 3.0283,
+      "step": 16041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5042517185211182,
+      "learning_rate": 0.0005039445781249435,
+      "loss": 3.1366,
+      "step": 16042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.546276330947876,
+      "learning_rate": 0.0005039332865458983,
+      "loss": 3.0346,
+      "step": 16043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.521732747554779,
+      "learning_rate": 0.0005039219944297321,
+      "loss": 3.2085,
+      "step": 16044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46991196274757385,
+      "learning_rate": 0.0005039107017764747,
+      "loss": 3.128,
+      "step": 16045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5190024971961975,
+      "learning_rate": 0.0005038994085861556,
+      "loss": 3.1441,
+      "step": 16046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5141576528549194,
+      "learning_rate": 0.0005038881148588048,
+      "loss": 2.9767,
+      "step": 16047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5428609848022461,
+      "learning_rate": 0.000503876820594452,
+      "loss": 3.0003,
+      "step": 16048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5093159675598145,
+      "learning_rate": 0.0005038655257931269,
+      "loss": 3.1081,
+      "step": 16049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5302721858024597,
+      "learning_rate": 0.0005038542304548591,
+      "loss": 3.2037,
+      "step": 16050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.526119589805603,
+      "learning_rate": 0.0005038429345796785,
+      "loss": 3.4543,
+      "step": 16051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5228065848350525,
+      "learning_rate": 0.0005038316381676149,
+      "loss": 3.1163,
+      "step": 16052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5292603969573975,
+      "learning_rate": 0.000503820341218698,
+      "loss": 3.2189,
+      "step": 16053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5633078813552856,
+      "learning_rate": 0.0005038090437329575,
+      "loss": 3.1288,
+      "step": 16054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5860145092010498,
+      "learning_rate": 0.0005037977457104233,
+      "loss": 3.2443,
+      "step": 16055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.614590048789978,
+      "learning_rate": 0.000503786447151125,
+      "loss": 3.0993,
+      "step": 16056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.502687394618988,
+      "learning_rate": 0.0005037751480550924,
+      "loss": 3.015,
+      "step": 16057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5312213897705078,
+      "learning_rate": 0.0005037638484223553,
+      "loss": 3.0935,
+      "step": 16058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5091737508773804,
+      "learning_rate": 0.0005037525482529435,
+      "loss": 3.1461,
+      "step": 16059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5732948780059814,
+      "learning_rate": 0.0005037412475468865,
+      "loss": 3.1839,
+      "step": 16060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5322034955024719,
+      "learning_rate": 0.0005037299463042144,
+      "loss": 3.2584,
+      "step": 16061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5124638080596924,
+      "learning_rate": 0.0005037186445249568,
+      "loss": 3.2116,
+      "step": 16062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5579886436462402,
+      "learning_rate": 0.0005037073422091436,
+      "loss": 3.1975,
+      "step": 16063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5009481310844421,
+      "learning_rate": 0.0005036960393568044,
+      "loss": 3.136,
+      "step": 16064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5488647818565369,
+      "learning_rate": 0.000503684735967969,
+      "loss": 3.0356,
+      "step": 16065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5193225145339966,
+      "learning_rate": 0.0005036734320426672,
+      "loss": 3.2853,
+      "step": 16066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5488074421882629,
+      "learning_rate": 0.0005036621275809288,
+      "loss": 3.1645,
+      "step": 16067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6234692335128784,
+      "learning_rate": 0.0005036508225827836,
+      "loss": 2.9019,
+      "step": 16068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48300352692604065,
+      "learning_rate": 0.0005036395170482613,
+      "loss": 3.0794,
+      "step": 16069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5437573194503784,
+      "learning_rate": 0.0005036282109773917,
+      "loss": 3.2105,
+      "step": 16070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6625828742980957,
+      "learning_rate": 0.0005036169043702045,
+      "loss": 3.2494,
+      "step": 16071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6107515096664429,
+      "learning_rate": 0.0005036055972267296,
+      "loss": 3.2419,
+      "step": 16072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5302785038948059,
+      "learning_rate": 0.0005035942895469967,
+      "loss": 3.1094,
+      "step": 16073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6154295802116394,
+      "learning_rate": 0.0005035829813310358,
+      "loss": 3.1699,
+      "step": 16074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6206402778625488,
+      "learning_rate": 0.0005035716725788763,
+      "loss": 3.1178,
+      "step": 16075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52013099193573,
+      "learning_rate": 0.0005035603632905484,
+      "loss": 3.0002,
+      "step": 16076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4945783019065857,
+      "learning_rate": 0.0005035490534660816,
+      "loss": 3.071,
+      "step": 16077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.509825587272644,
+      "learning_rate": 0.0005035377431055057,
+      "loss": 3.2746,
+      "step": 16078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49139195680618286,
+      "learning_rate": 0.0005035264322088505,
+      "loss": 3.2474,
+      "step": 16079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5229079127311707,
+      "learning_rate": 0.000503515120776146,
+      "loss": 3.2264,
+      "step": 16080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5331903696060181,
+      "learning_rate": 0.0005035038088074217,
+      "loss": 3.4112,
+      "step": 16081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5611699223518372,
+      "learning_rate": 0.0005034924963027077,
+      "loss": 3.0743,
+      "step": 16082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.504478394985199,
+      "learning_rate": 0.0005034811832620335,
+      "loss": 3.1197,
+      "step": 16083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5261973142623901,
+      "learning_rate": 0.0005034698696854291,
+      "loss": 3.1517,
+      "step": 16084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5199773907661438,
+      "learning_rate": 0.0005034585555729243,
+      "loss": 3.3782,
+      "step": 16085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4901329576969147,
+      "learning_rate": 0.0005034472409245485,
+      "loss": 3.1992,
+      "step": 16086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5331196784973145,
+      "learning_rate": 0.0005034359257403322,
+      "loss": 3.1904,
+      "step": 16087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5385162830352783,
+      "learning_rate": 0.0005034246100203046,
+      "loss": 2.9416,
+      "step": 16088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5102599263191223,
+      "learning_rate": 0.0005034132937644958,
+      "loss": 3.332,
+      "step": 16089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5430083274841309,
+      "learning_rate": 0.0005034019769729355,
+      "loss": 3.0467,
+      "step": 16090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.563103437423706,
+      "learning_rate": 0.0005033906596456535,
+      "loss": 3.1664,
+      "step": 16091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5284902453422546,
+      "learning_rate": 0.0005033793417826797,
+      "loss": 3.4442,
+      "step": 16092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.477546364068985,
+      "learning_rate": 0.0005033680233840439,
+      "loss": 3.2581,
+      "step": 16093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5075785517692566,
+      "learning_rate": 0.0005033567044497758,
+      "loss": 3.1257,
+      "step": 16094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5229860544204712,
+      "learning_rate": 0.0005033453849799053,
+      "loss": 3.2537,
+      "step": 16095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5341659188270569,
+      "learning_rate": 0.0005033340649744621,
+      "loss": 3.09,
+      "step": 16096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5053170323371887,
+      "learning_rate": 0.0005033227444334763,
+      "loss": 3.4498,
+      "step": 16097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5055860877037048,
+      "learning_rate": 0.0005033114233569774,
+      "loss": 2.9695,
+      "step": 16098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4993264377117157,
+      "learning_rate": 0.0005033001017449953,
+      "loss": 3.0594,
+      "step": 16099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49798890948295593,
+      "learning_rate": 0.00050328877959756,
+      "loss": 3.2009,
+      "step": 16100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5068756341934204,
+      "learning_rate": 0.000503277456914701,
+      "loss": 3.3866,
+      "step": 16101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5351049900054932,
+      "learning_rate": 0.0005032661336964484,
+      "loss": 3.0999,
+      "step": 16102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49359703063964844,
+      "learning_rate": 0.000503254809942832,
+      "loss": 3.1928,
+      "step": 16103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.526689887046814,
+      "learning_rate": 0.0005032434856538814,
+      "loss": 3.1952,
+      "step": 16104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5129401683807373,
+      "learning_rate": 0.0005032321608296266,
+      "loss": 2.8986,
+      "step": 16105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5190664529800415,
+      "learning_rate": 0.0005032208354700974,
+      "loss": 3.1921,
+      "step": 16106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5858680605888367,
+      "learning_rate": 0.0005032095095753237,
+      "loss": 3.2601,
+      "step": 16107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5072818994522095,
+      "learning_rate": 0.0005031981831453352,
+      "loss": 3.1682,
+      "step": 16108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5385854244232178,
+      "learning_rate": 0.0005031868561801618,
+      "loss": 3.2302,
+      "step": 16109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49237823486328125,
+      "learning_rate": 0.0005031755286798333,
+      "loss": 3.4822,
+      "step": 16110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5026249885559082,
+      "learning_rate": 0.0005031642006443796,
+      "loss": 3.3075,
+      "step": 16111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5177112817764282,
+      "learning_rate": 0.0005031528720738306,
+      "loss": 3.2019,
+      "step": 16112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5290037989616394,
+      "learning_rate": 0.0005031415429682159,
+      "loss": 3.0193,
+      "step": 16113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5207205414772034,
+      "learning_rate": 0.0005031302133275655,
+      "loss": 2.7481,
+      "step": 16114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5251289010047913,
+      "learning_rate": 0.0005031188831519091,
+      "loss": 3.2789,
+      "step": 16115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5037059187889099,
+      "learning_rate": 0.0005031075524412769,
+      "loss": 3.1427,
+      "step": 16116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.509303629398346,
+      "learning_rate": 0.0005030962211956982,
+      "loss": 3.4196,
+      "step": 16117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.47990882396698,
+      "learning_rate": 0.0005030848894152034,
+      "loss": 3.1945,
+      "step": 16118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5095180869102478,
+      "learning_rate": 0.0005030735570998219,
+      "loss": 3.3179,
+      "step": 16119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5448658466339111,
+      "learning_rate": 0.0005030622242495839,
+      "loss": 3.3563,
+      "step": 16120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5198306441307068,
+      "learning_rate": 0.000503050890864519,
+      "loss": 3.1658,
+      "step": 16121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5410298109054565,
+      "learning_rate": 0.0005030395569446571,
+      "loss": 3.2453,
+      "step": 16122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5249944925308228,
+      "learning_rate": 0.0005030282224900281,
+      "loss": 3.1404,
+      "step": 16123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5237621068954468,
+      "learning_rate": 0.0005030168875006618,
+      "loss": 3.1242,
+      "step": 16124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5496729016304016,
+      "learning_rate": 0.0005030055519765882,
+      "loss": 3.0199,
+      "step": 16125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5271469950675964,
+      "learning_rate": 0.000502994215917837,
+      "loss": 3.123,
+      "step": 16126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5577632188796997,
+      "learning_rate": 0.000502982879324438,
+      "loss": 3.1257,
+      "step": 16127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.498698353767395,
+      "learning_rate": 0.0005029715421964213,
+      "loss": 3.3184,
+      "step": 16128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.544201135635376,
+      "learning_rate": 0.0005029602045338166,
+      "loss": 2.9086,
+      "step": 16129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5301446914672852,
+      "learning_rate": 0.0005029488663366538,
+      "loss": 3.4906,
+      "step": 16130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5524643063545227,
+      "learning_rate": 0.0005029375276049626,
+      "loss": 3.4142,
+      "step": 16131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5376681685447693,
+      "learning_rate": 0.0005029261883387732,
+      "loss": 3.254,
+      "step": 16132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5098219513893127,
+      "learning_rate": 0.000502914848538115,
+      "loss": 3.3139,
+      "step": 16133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5753784775733948,
+      "learning_rate": 0.0005029035082030184,
+      "loss": 3.2448,
+      "step": 16134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49853768944740295,
+      "learning_rate": 0.0005028921673335129,
+      "loss": 3.2964,
+      "step": 16135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5184651017189026,
+      "learning_rate": 0.0005028808259296285,
+      "loss": 3.2924,
+      "step": 16136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5892463326454163,
+      "learning_rate": 0.000502869483991395,
+      "loss": 3.0275,
+      "step": 16137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4991600811481476,
+      "learning_rate": 0.0005028581415188425,
+      "loss": 3.1359,
+      "step": 16138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49838724732398987,
+      "learning_rate": 0.0005028467985120006,
+      "loss": 3.1147,
+      "step": 16139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5089886784553528,
+      "learning_rate": 0.0005028354549708992,
+      "loss": 2.9585,
+      "step": 16140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52250075340271,
+      "learning_rate": 0.0005028241108955683,
+      "loss": 3.3139,
+      "step": 16141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5590442419052124,
+      "learning_rate": 0.0005028127662860376,
+      "loss": 3.0677,
+      "step": 16142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5002887845039368,
+      "learning_rate": 0.0005028014211423372,
+      "loss": 3.018,
+      "step": 16143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5314575433731079,
+      "learning_rate": 0.000502790075464497,
+      "loss": 3.2391,
+      "step": 16144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4909334182739258,
+      "learning_rate": 0.0005027787292525468,
+      "loss": 3.0808,
+      "step": 16145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5140949487686157,
+      "learning_rate": 0.0005027673825065163,
+      "loss": 3.1941,
+      "step": 16146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49293190240859985,
+      "learning_rate": 0.0005027560352264355,
+      "loss": 3.0625,
+      "step": 16147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5335683822631836,
+      "learning_rate": 0.0005027446874123344,
+      "loss": 3.1745,
+      "step": 16148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5320890545845032,
+      "learning_rate": 0.0005027333390642428,
+      "loss": 3.0896,
+      "step": 16149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5180721879005432,
+      "learning_rate": 0.0005027219901821906,
+      "loss": 3.2275,
+      "step": 16150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5077602863311768,
+      "learning_rate": 0.0005027106407662077,
+      "loss": 2.9512,
+      "step": 16151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5526420474052429,
+      "learning_rate": 0.0005026992908163239,
+      "loss": 3.0982,
+      "step": 16152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5293105840682983,
+      "learning_rate": 0.0005026879403325693,
+      "loss": 3.2531,
+      "step": 16153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4970487058162689,
+      "learning_rate": 0.0005026765893149736,
+      "loss": 3.2379,
+      "step": 16154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.523263156414032,
+      "learning_rate": 0.0005026652377635669,
+      "loss": 3.3632,
+      "step": 16155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5170415639877319,
+      "learning_rate": 0.0005026538856783789,
+      "loss": 3.139,
+      "step": 16156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48945000767707825,
+      "learning_rate": 0.0005026425330594394,
+      "loss": 3.1406,
+      "step": 16157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5514994859695435,
+      "learning_rate": 0.0005026311799067786,
+      "loss": 2.9625,
+      "step": 16158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5217519402503967,
+      "learning_rate": 0.0005026198262204263,
+      "loss": 3.0109,
+      "step": 16159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5658161044120789,
+      "learning_rate": 0.0005026084720004125,
+      "loss": 2.9256,
+      "step": 16160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5193248391151428,
+      "learning_rate": 0.0005025971172467667,
+      "loss": 3.2494,
+      "step": 16161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5272179841995239,
+      "learning_rate": 0.0005025857619595193,
+      "loss": 3.312,
+      "step": 16162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4956285059452057,
+      "learning_rate": 0.0005025744061386998,
+      "loss": 3.2104,
+      "step": 16163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.521141767501831,
+      "learning_rate": 0.0005025630497843385,
+      "loss": 2.9773,
+      "step": 16164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.525787353515625,
+      "learning_rate": 0.000502551692896465,
+      "loss": 3.2225,
+      "step": 16165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5088430643081665,
+      "learning_rate": 0.0005025403354751093,
+      "loss": 3.1288,
+      "step": 16166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5418877005577087,
+      "learning_rate": 0.0005025289775203013,
+      "loss": 3.2467,
+      "step": 16167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5093033313751221,
+      "learning_rate": 0.0005025176190320711,
+      "loss": 3.4795,
+      "step": 16168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5372544527053833,
+      "learning_rate": 0.0005025062600104483,
+      "loss": 3.1049,
+      "step": 16169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5223720669746399,
+      "learning_rate": 0.0005024949004554632,
+      "loss": 3.1735,
+      "step": 16170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5225620865821838,
+      "learning_rate": 0.0005024835403671453,
+      "loss": 3.1505,
+      "step": 16171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.478676974773407,
+      "learning_rate": 0.0005024721797455249,
+      "loss": 3.1069,
+      "step": 16172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5180579423904419,
+      "learning_rate": 0.0005024608185906315,
+      "loss": 3.0396,
+      "step": 16173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5308378338813782,
+      "learning_rate": 0.0005024494569024954,
+      "loss": 3.204,
+      "step": 16174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5326209664344788,
+      "learning_rate": 0.0005024380946811464,
+      "loss": 3.0794,
+      "step": 16175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5528416037559509,
+      "learning_rate": 0.0005024267319266144,
+      "loss": 3.3221,
+      "step": 16176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5260026454925537,
+      "learning_rate": 0.0005024153686389295,
+      "loss": 2.9779,
+      "step": 16177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5304065346717834,
+      "learning_rate": 0.0005024040048181213,
+      "loss": 3.1639,
+      "step": 16178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5245885252952576,
+      "learning_rate": 0.0005023926404642199,
+      "loss": 3.2939,
+      "step": 16179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.47927168011665344,
+      "learning_rate": 0.0005023812755772553,
+      "loss": 3.2998,
+      "step": 16180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5522112250328064,
+      "learning_rate": 0.0005023699101572574,
+      "loss": 3.3079,
+      "step": 16181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.47725212574005127,
+      "learning_rate": 0.000502358544204256,
+      "loss": 3.0599,
+      "step": 16182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390349626541138,
+      "learning_rate": 0.0005023471777182813,
+      "loss": 3.2281,
+      "step": 16183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5163893699645996,
+      "learning_rate": 0.0005023358106993629,
+      "loss": 2.8327,
+      "step": 16184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5547180771827698,
+      "learning_rate": 0.000502324443147531,
+      "loss": 3.1906,
+      "step": 16185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5237986445426941,
+      "learning_rate": 0.0005023130750628154,
+      "loss": 3.336,
+      "step": 16186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.493787944316864,
+      "learning_rate": 0.0005023017064452462,
+      "loss": 3.0081,
+      "step": 16187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5115264654159546,
+      "learning_rate": 0.0005022903372948531,
+      "loss": 3.181,
+      "step": 16188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5147324204444885,
+      "learning_rate": 0.0005022789676116663,
+      "loss": 2.998,
+      "step": 16189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5190193057060242,
+      "learning_rate": 0.0005022675973957156,
+      "loss": 3.2676,
+      "step": 16190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.522584855556488,
+      "learning_rate": 0.000502256226647031,
+      "loss": 3.1966,
+      "step": 16191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5173296332359314,
+      "learning_rate": 0.0005022448553656424,
+      "loss": 3.175,
+      "step": 16192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5209254622459412,
+      "learning_rate": 0.0005022334835515797,
+      "loss": 3.113,
+      "step": 16193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5002453923225403,
+      "learning_rate": 0.000502222111204873,
+      "loss": 3.1706,
+      "step": 16194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5135509371757507,
+      "learning_rate": 0.0005022107383255522,
+      "loss": 3.2399,
+      "step": 16195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5734914541244507,
+      "learning_rate": 0.0005021993649136472,
+      "loss": 2.9791,
+      "step": 16196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5258508324623108,
+      "learning_rate": 0.000502187990969188,
+      "loss": 2.8981,
+      "step": 16197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.528937816619873,
+      "learning_rate": 0.0005021766164922044,
+      "loss": 2.9982,
+      "step": 16198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.559468686580658,
+      "learning_rate": 0.0005021652414827268,
+      "loss": 3.1467,
+      "step": 16199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5440381169319153,
+      "learning_rate": 0.0005021538659407845,
+      "loss": 3.1731,
+      "step": 16200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5404337048530579,
+      "learning_rate": 0.0005021424898664081,
+      "loss": 3.2169,
+      "step": 16201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5026121735572815,
+      "learning_rate": 0.0005021311132596272,
+      "loss": 3.211,
+      "step": 16202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5181695818901062,
+      "learning_rate": 0.0005021197361204719,
+      "loss": 3.1374,
+      "step": 16203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5274845361709595,
+      "learning_rate": 0.0005021083584489719,
+      "loss": 3.118,
+      "step": 16204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5087721943855286,
+      "learning_rate": 0.0005020969802451577,
+      "loss": 3.1625,
+      "step": 16205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4713067412376404,
+      "learning_rate": 0.0005020856015090587,
+      "loss": 3.0954,
+      "step": 16206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5224311351776123,
+      "learning_rate": 0.0005020742222407053,
+      "loss": 3.1967,
+      "step": 16207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5140992403030396,
+      "learning_rate": 0.0005020628424401272,
+      "loss": 3.3444,
+      "step": 16208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.568848729133606,
+      "learning_rate": 0.0005020514621073544,
+      "loss": 3.2186,
+      "step": 16209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5110242962837219,
+      "learning_rate": 0.000502040081242417,
+      "loss": 3.1196,
+      "step": 16210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5247518420219421,
+      "learning_rate": 0.0005020286998453449,
+      "loss": 3.2261,
+      "step": 16211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4877007007598877,
+      "learning_rate": 0.0005020173179161681,
+      "loss": 3.0439,
+      "step": 16212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6002936363220215,
+      "learning_rate": 0.0005020059354549166,
+      "loss": 3.2361,
+      "step": 16213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5274139642715454,
+      "learning_rate": 0.0005019945524616203,
+      "loss": 2.848,
+      "step": 16214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4768544137477875,
+      "learning_rate": 0.0005019831689363091,
+      "loss": 3.186,
+      "step": 16215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7357099652290344,
+      "learning_rate": 0.0005019717848790133,
+      "loss": 3.1149,
+      "step": 16216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4986879825592041,
+      "learning_rate": 0.0005019604002897626,
+      "loss": 3.2408,
+      "step": 16217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5395023822784424,
+      "learning_rate": 0.0005019490151685871,
+      "loss": 3.1948,
+      "step": 16218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5745142102241516,
+      "learning_rate": 0.0005019376295155167,
+      "loss": 3.1383,
+      "step": 16219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5301345586776733,
+      "learning_rate": 0.0005019262433305815,
+      "loss": 3.2496,
+      "step": 16220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4867192208766937,
+      "learning_rate": 0.0005019148566138115,
+      "loss": 3.2222,
+      "step": 16221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5403074622154236,
+      "learning_rate": 0.0005019034693652365,
+      "loss": 3.1839,
+      "step": 16222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5404332280158997,
+      "learning_rate": 0.0005018920815848866,
+      "loss": 3.1374,
+      "step": 16223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49627020955085754,
+      "learning_rate": 0.0005018806932727919,
+      "loss": 3.1589,
+      "step": 16224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.516038179397583,
+      "learning_rate": 0.0005018693044289823,
+      "loss": 3.1485,
+      "step": 16225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49638882279396057,
+      "learning_rate": 0.0005018579150534878,
+      "loss": 3.1552,
+      "step": 16226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4918712079524994,
+      "learning_rate": 0.0005018465251463384,
+      "loss": 3.3042,
+      "step": 16227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4965533912181854,
+      "learning_rate": 0.0005018351347075641,
+      "loss": 2.9956,
+      "step": 16228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49797412753105164,
+      "learning_rate": 0.0005018237437371949,
+      "loss": 3.0222,
+      "step": 16229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5310698747634888,
+      "learning_rate": 0.0005018123522352607,
+      "loss": 3.1327,
+      "step": 16230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4971972107887268,
+      "learning_rate": 0.0005018009602017918,
+      "loss": 3.1433,
+      "step": 16231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5360703468322754,
+      "learning_rate": 0.0005017895676368179,
+      "loss": 3.1121,
+      "step": 16232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5207992196083069,
+      "learning_rate": 0.0005017781745403692,
+      "loss": 2.9475,
+      "step": 16233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49780797958374023,
+      "learning_rate": 0.0005017667809124755,
+      "loss": 3.2944,
+      "step": 16234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5747670531272888,
+      "learning_rate": 0.0005017553867531669,
+      "loss": 3.1795,
+      "step": 16235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5032791495323181,
+      "learning_rate": 0.0005017439920624735,
+      "loss": 3.4004,
+      "step": 16236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5304082036018372,
+      "learning_rate": 0.0005017325968404252,
+      "loss": 3.2782,
+      "step": 16237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5240580439567566,
+      "learning_rate": 0.0005017212010870522,
+      "loss": 2.8734,
+      "step": 16238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4911217987537384,
+      "learning_rate": 0.0005017098048023843,
+      "loss": 3.3212,
+      "step": 16239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5378339290618896,
+      "learning_rate": 0.0005016984079864516,
+      "loss": 3.1136,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5092733502388,
+      "learning_rate": 0.0005016870106392841,
+      "loss": 3.0082,
+      "step": 16241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5283250212669373,
+      "learning_rate": 0.0005016756127609118,
+      "loss": 3.0912,
+      "step": 16242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5703310966491699,
+      "learning_rate": 0.0005016642143513649,
+      "loss": 3.0603,
+      "step": 16243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48841413855552673,
+      "learning_rate": 0.0005016528154106731,
+      "loss": 3.4069,
+      "step": 16244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.634024977684021,
+      "learning_rate": 0.0005016414159388667,
+      "loss": 3.2679,
+      "step": 16245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.497964084148407,
+      "learning_rate": 0.0005016300159359755,
+      "loss": 3.315,
+      "step": 16246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5449110865592957,
+      "learning_rate": 0.0005016186154020297,
+      "loss": 3.36,
+      "step": 16247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5338404774665833,
+      "learning_rate": 0.0005016072143370593,
+      "loss": 3.0915,
+      "step": 16248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5039170980453491,
+      "learning_rate": 0.0005015958127410942,
+      "loss": 3.1784,
+      "step": 16249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49808257818222046,
+      "learning_rate": 0.0005015844106141648,
+      "loss": 3.3481,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.555994987487793,
+      "learning_rate": 0.0005015730079563006,
+      "loss": 3.2594,
+      "step": 16251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5516330003738403,
+      "learning_rate": 0.0005015616047675319,
+      "loss": 3.1797,
+      "step": 16252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.523496150970459,
+      "learning_rate": 0.0005015502010478889,
+      "loss": 3.1193,
+      "step": 16253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5190884470939636,
+      "learning_rate": 0.0005015387967974012,
+      "loss": 3.1007,
+      "step": 16254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4918883740901947,
+      "learning_rate": 0.0005015273920160993,
+      "loss": 3.194,
+      "step": 16255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5591009855270386,
+      "learning_rate": 0.0005015159867040129,
+      "loss": 3.0611,
+      "step": 16256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4997139871120453,
+      "learning_rate": 0.0005015045808611723,
+      "loss": 3.2017,
+      "step": 16257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.63576740026474,
+      "learning_rate": 0.0005014931744876073,
+      "loss": 3.0073,
+      "step": 16258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5021123290061951,
+      "learning_rate": 0.0005014817675833481,
+      "loss": 3.3267,
+      "step": 16259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5300601720809937,
+      "learning_rate": 0.0005014703601484248,
+      "loss": 2.9822,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5239089131355286,
+      "learning_rate": 0.0005014589521828673,
+      "loss": 3.0789,
+      "step": 16261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5151042938232422,
+      "learning_rate": 0.0005014475436867056,
+      "loss": 3.1529,
+      "step": 16262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5113961696624756,
+      "learning_rate": 0.0005014361346599699,
+      "loss": 3.0487,
+      "step": 16263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5840994119644165,
+      "learning_rate": 0.0005014247251026901,
+      "loss": 3.292,
+      "step": 16264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52402663230896,
+      "learning_rate": 0.0005014133150148964,
+      "loss": 3.2253,
+      "step": 16265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5494585037231445,
+      "learning_rate": 0.0005014019043966189,
+      "loss": 2.9822,
+      "step": 16266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5364586114883423,
+      "learning_rate": 0.0005013904932478874,
+      "loss": 3.2012,
+      "step": 16267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5051799416542053,
+      "learning_rate": 0.0005013790815687322,
+      "loss": 3.228,
+      "step": 16268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5270349979400635,
+      "learning_rate": 0.0005013676693591832,
+      "loss": 3.1309,
+      "step": 16269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.523430585861206,
+      "learning_rate": 0.0005013562566192704,
+      "loss": 3.2644,
+      "step": 16270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4938635528087616,
+      "learning_rate": 0.0005013448433490241,
+      "loss": 3.3449,
+      "step": 16271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5350764989852905,
+      "learning_rate": 0.0005013334295484742,
+      "loss": 3.3254,
+      "step": 16272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5360183715820312,
+      "learning_rate": 0.0005013220152176509,
+      "loss": 2.9552,
+      "step": 16273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5047653913497925,
+      "learning_rate": 0.000501310600356584,
+      "loss": 3.172,
+      "step": 16274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5344395041465759,
+      "learning_rate": 0.0005012991849653038,
+      "loss": 3.152,
+      "step": 16275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5071520805358887,
+      "learning_rate": 0.0005012877690438403,
+      "loss": 3.2807,
+      "step": 16276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.501033365726471,
+      "learning_rate": 0.0005012763525922236,
+      "loss": 3.0155,
+      "step": 16277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5294574499130249,
+      "learning_rate": 0.0005012649356104837,
+      "loss": 3.2542,
+      "step": 16278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.565493106842041,
+      "learning_rate": 0.0005012535180986507,
+      "loss": 2.9964,
+      "step": 16279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5068030953407288,
+      "learning_rate": 0.0005012421000567545,
+      "loss": 3.2337,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078257918357849,
+      "learning_rate": 0.0005012306814848255,
+      "loss": 3.2369,
+      "step": 16281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5458744168281555,
+      "learning_rate": 0.0005012192623828935,
+      "loss": 3.4057,
+      "step": 16282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5356884002685547,
+      "learning_rate": 0.0005012078427509888,
+      "loss": 3.2353,
+      "step": 16283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5274999141693115,
+      "learning_rate": 0.0005011964225891414,
+      "loss": 3.2908,
+      "step": 16284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5029778480529785,
+      "learning_rate": 0.0005011850018973813,
+      "loss": 3.164,
+      "step": 16285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.555999219417572,
+      "learning_rate": 0.0005011735806757384,
+      "loss": 3.0256,
+      "step": 16286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48176196217536926,
+      "learning_rate": 0.0005011621589242433,
+      "loss": 3.2014,
+      "step": 16287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49973785877227783,
+      "learning_rate": 0.0005011507366429257,
+      "loss": 3.2146,
+      "step": 16288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.537611722946167,
+      "learning_rate": 0.0005011393138318157,
+      "loss": 2.9839,
+      "step": 16289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52414470911026,
+      "learning_rate": 0.0005011278904909437,
+      "loss": 3.2931,
+      "step": 16290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5327220559120178,
+      "learning_rate": 0.0005011164666203392,
+      "loss": 3.276,
+      "step": 16291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5477252006530762,
+      "learning_rate": 0.0005011050422200328,
+      "loss": 3.1397,
+      "step": 16292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5243980884552002,
+      "learning_rate": 0.0005010936172900544,
+      "loss": 3.3518,
+      "step": 16293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5618069171905518,
+      "learning_rate": 0.0005010821918304342,
+      "loss": 2.9984,
+      "step": 16294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.515941321849823,
+      "learning_rate": 0.0005010707658412021,
+      "loss": 3.1225,
+      "step": 16295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5532481670379639,
+      "learning_rate": 0.0005010593393223883,
+      "loss": 3.0783,
+      "step": 16296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4828968644142151,
+      "learning_rate": 0.0005010479122740229,
+      "loss": 3.0916,
+      "step": 16297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5126880407333374,
+      "learning_rate": 0.000501036484696136,
+      "loss": 3.2333,
+      "step": 16298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5703883171081543,
+      "learning_rate": 0.0005010250565887578,
+      "loss": 3.1723,
+      "step": 16299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5786604881286621,
+      "learning_rate": 0.0005010136279519182,
+      "loss": 3.0571,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5473567843437195,
+      "learning_rate": 0.0005010021987856474,
+      "loss": 3.1133,
+      "step": 16301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5086806416511536,
+      "learning_rate": 0.0005009907690899754,
+      "loss": 3.2605,
+      "step": 16302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5256600975990295,
+      "learning_rate": 0.0005009793388649326,
+      "loss": 3.1623,
+      "step": 16303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5034403800964355,
+      "learning_rate": 0.0005009679081105487,
+      "loss": 3.3153,
+      "step": 16304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5190585851669312,
+      "learning_rate": 0.000500956476826854,
+      "loss": 3.1154,
+      "step": 16305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.501880407333374,
+      "learning_rate": 0.0005009450450138787,
+      "loss": 3.4226,
+      "step": 16306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5461406707763672,
+      "learning_rate": 0.000500933612671653,
+      "loss": 3.1763,
+      "step": 16307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5586167573928833,
+      "learning_rate": 0.0005009221798002065,
+      "loss": 3.1583,
+      "step": 16308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5345634818077087,
+      "learning_rate": 0.0005009107463995699,
+      "loss": 3.35,
+      "step": 16309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49737313389778137,
+      "learning_rate": 0.0005008993124697729,
+      "loss": 3.158,
+      "step": 16310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5575215816497803,
+      "learning_rate": 0.0005008878780108459,
+      "loss": 3.1295,
+      "step": 16311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5068386197090149,
+      "learning_rate": 0.0005008764430228188,
+      "loss": 3.2086,
+      "step": 16312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5135799050331116,
+      "learning_rate": 0.0005008650075057218,
+      "loss": 3.4186,
+      "step": 16313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.50897216796875,
+      "learning_rate": 0.000500853571459585,
+      "loss": 3.1799,
+      "step": 16314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4997579753398895,
+      "learning_rate": 0.0005008421348844387,
+      "loss": 3.0874,
+      "step": 16315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5073021054267883,
+      "learning_rate": 0.0005008306977803128,
+      "loss": 3.1663,
+      "step": 16316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.502220630645752,
+      "learning_rate": 0.0005008192601472374,
+      "loss": 3.2532,
+      "step": 16317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5047315359115601,
+      "learning_rate": 0.0005008078219852429,
+      "loss": 3.3159,
+      "step": 16318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5216161012649536,
+      "learning_rate": 0.0005007963832943591,
+      "loss": 3.3052,
+      "step": 16319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5240123867988586,
+      "learning_rate": 0.0005007849440746163,
+      "loss": 3.3936,
+      "step": 16320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4939900040626526,
+      "learning_rate": 0.0005007735043260446,
+      "loss": 3.2465,
+      "step": 16321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5109789967536926,
+      "learning_rate": 0.0005007620640486741,
+      "loss": 3.2318,
+      "step": 16322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5235530138015747,
+      "learning_rate": 0.0005007506232425351,
+      "loss": 3.0924,
+      "step": 16323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5005676746368408,
+      "learning_rate": 0.0005007391819076574,
+      "loss": 3.3931,
+      "step": 16324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4928778409957886,
+      "learning_rate": 0.0005007277400440715,
+      "loss": 3.2683,
+      "step": 16325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5503429770469666,
+      "learning_rate": 0.0005007162976518073,
+      "loss": 3.2935,
+      "step": 16326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.526890754699707,
+      "learning_rate": 0.0005007048547308948,
+      "loss": 3.0424,
+      "step": 16327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5440188050270081,
+      "learning_rate": 0.0005006934112813646,
+      "loss": 3.2249,
+      "step": 16328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48897939920425415,
+      "learning_rate": 0.0005006819673032465,
+      "loss": 3.0303,
+      "step": 16329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4976714849472046,
+      "learning_rate": 0.0005006705227965707,
+      "loss": 3.2086,
+      "step": 16330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.519629716873169,
+      "learning_rate": 0.0005006590777613674,
+      "loss": 3.0235,
+      "step": 16331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5517348647117615,
+      "learning_rate": 0.0005006476321976667,
+      "loss": 3.133,
+      "step": 16332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5207356810569763,
+      "learning_rate": 0.0005006361861054988,
+      "loss": 3.2264,
+      "step": 16333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.521043062210083,
+      "learning_rate": 0.0005006247394848938,
+      "loss": 3.0017,
+      "step": 16334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.519502580165863,
+      "learning_rate": 0.0005006132923358818,
+      "loss": 3.2157,
+      "step": 16335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47964170575141907,
+      "learning_rate": 0.0005006018446584928,
+      "loss": 3.2735,
+      "step": 16336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.515812337398529,
+      "learning_rate": 0.0005005903964527573,
+      "loss": 3.256,
+      "step": 16337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5045713782310486,
+      "learning_rate": 0.0005005789477187055,
+      "loss": 3.1528,
+      "step": 16338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5392544865608215,
+      "learning_rate": 0.0005005674984563672,
+      "loss": 3.0644,
+      "step": 16339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5179302096366882,
+      "learning_rate": 0.0005005560486657726,
+      "loss": 3.1746,
+      "step": 16340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5171295404434204,
+      "learning_rate": 0.0005005445983469522,
+      "loss": 3.0787,
+      "step": 16341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5172526240348816,
+      "learning_rate": 0.0005005331474999357,
+      "loss": 3.0268,
+      "step": 16342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5140368938446045,
+      "learning_rate": 0.0005005216961247535,
+      "loss": 3.5288,
+      "step": 16343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.525821328163147,
+      "learning_rate": 0.0005005102442214358,
+      "loss": 3.0389,
+      "step": 16344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5332549214363098,
+      "learning_rate": 0.0005004987917900127,
+      "loss": 3.2078,
+      "step": 16345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5117021203041077,
+      "learning_rate": 0.0005004873388305145,
+      "loss": 3.1699,
+      "step": 16346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.526621401309967,
+      "learning_rate": 0.000500475885342971,
+      "loss": 3.0896,
+      "step": 16347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4984114468097687,
+      "learning_rate": 0.0005004644313274127,
+      "loss": 3.1271,
+      "step": 16348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5194051861763,
+      "learning_rate": 0.0005004529767838697,
+      "loss": 3.4425,
+      "step": 16349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.530343234539032,
+      "learning_rate": 0.0005004415217123722,
+      "loss": 2.962,
+      "step": 16350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5183805227279663,
+      "learning_rate": 0.0005004300661129502,
+      "loss": 3.2518,
+      "step": 16351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5521455407142639,
+      "learning_rate": 0.000500418609985634,
+      "loss": 3.1005,
+      "step": 16352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5434244871139526,
+      "learning_rate": 0.0005004071533304538,
+      "loss": 3.0559,
+      "step": 16353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5996847152709961,
+      "learning_rate": 0.0005003956961474397,
+      "loss": 3.0903,
+      "step": 16354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5073933005332947,
+      "learning_rate": 0.0005003842384366219,
+      "loss": 3.0769,
+      "step": 16355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5215885043144226,
+      "learning_rate": 0.0005003727801980305,
+      "loss": 3.2447,
+      "step": 16356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5436731576919556,
+      "learning_rate": 0.000500361321431696,
+      "loss": 3.1657,
+      "step": 16357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5639539361000061,
+      "learning_rate": 0.000500349862137648,
+      "loss": 3.0724,
+      "step": 16358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5657851099967957,
+      "learning_rate": 0.0005003384023159173,
+      "loss": 3.2198,
+      "step": 16359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5688062906265259,
+      "learning_rate": 0.0005003269419665336,
+      "loss": 3.0992,
+      "step": 16360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5081748366355896,
+      "learning_rate": 0.0005003154810895275,
+      "loss": 3.1934,
+      "step": 16361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5793516039848328,
+      "learning_rate": 0.0005003040196849288,
+      "loss": 3.1196,
+      "step": 16362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5692729353904724,
+      "learning_rate": 0.000500292557752768,
+      "loss": 3.0594,
+      "step": 16363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625926852226257,
+      "learning_rate": 0.0005002810952930751,
+      "loss": 3.3912,
+      "step": 16364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5317186117172241,
+      "learning_rate": 0.0005002696323058803,
+      "loss": 3.2598,
+      "step": 16365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5381388068199158,
+      "learning_rate": 0.0005002581687912138,
+      "loss": 3.1774,
+      "step": 16366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5291975736618042,
+      "learning_rate": 0.0005002467047491058,
+      "loss": 3.1837,
+      "step": 16367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5130248665809631,
+      "learning_rate": 0.0005002352401795866,
+      "loss": 3.1678,
+      "step": 16368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4932388365268707,
+      "learning_rate": 0.0005002237750826863,
+      "loss": 3.2429,
+      "step": 16369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5155643224716187,
+      "learning_rate": 0.0005002123094584351,
+      "loss": 3.3085,
+      "step": 16370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5444445013999939,
+      "learning_rate": 0.0005002008433068633,
+      "loss": 3.2845,
+      "step": 16371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5338315367698669,
+      "learning_rate": 0.0005001893766280009,
+      "loss": 3.1683,
+      "step": 16372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.535651445388794,
+      "learning_rate": 0.0005001779094218782,
+      "loss": 3.0012,
+      "step": 16373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5226747989654541,
+      "learning_rate": 0.0005001664416885254,
+      "loss": 2.9169,
+      "step": 16374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.526677131652832,
+      "learning_rate": 0.0005001549734279729,
+      "loss": 3.1332,
+      "step": 16375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.540988028049469,
+      "learning_rate": 0.0005001435046402505,
+      "loss": 3.3839,
+      "step": 16376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5214574337005615,
+      "learning_rate": 0.0005001320353253887,
+      "loss": 3.2017,
+      "step": 16377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49261197447776794,
+      "learning_rate": 0.0005001205654834176,
+      "loss": 3.2731,
+      "step": 16378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.507203996181488,
+      "learning_rate": 0.0005001090951143675,
+      "loss": 3.0643,
+      "step": 16379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5074104070663452,
+      "learning_rate": 0.0005000976242182687,
+      "loss": 3.1289,
+      "step": 16380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.533085823059082,
+      "learning_rate": 0.000500086152795151,
+      "loss": 3.1851,
+      "step": 16381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5225232243537903,
+      "learning_rate": 0.000500074680845045,
+      "loss": 3.1941,
+      "step": 16382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5232262015342712,
+      "learning_rate": 0.0005000632083679808,
+      "loss": 3.1863,
+      "step": 16383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.499962717294693,
+      "learning_rate": 0.0005000517353639886,
+      "loss": 3.2258,
+      "step": 16384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4960383176803589,
+      "learning_rate": 0.0005000402618330986,
+      "loss": 3.4133,
+      "step": 16385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5370568633079529,
+      "learning_rate": 0.0005000287877753411,
+      "loss": 3.0247,
+      "step": 16386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5205273032188416,
+      "learning_rate": 0.0005000173131907463,
+      "loss": 3.028,
+      "step": 16387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5405469536781311,
+      "learning_rate": 0.0005000058380793443,
+      "loss": 3.2769,
+      "step": 16388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5633376836776733,
+      "learning_rate": 0.0004999943624411655,
+      "loss": 3.2323,
+      "step": 16389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5104237198829651,
+      "learning_rate": 0.00049998288627624,
+      "loss": 3.2026,
+      "step": 16390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5359808206558228,
+      "learning_rate": 0.000499971409584598,
+      "loss": 2.988,
+      "step": 16391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5244869589805603,
+      "learning_rate": 0.0004999599323662699,
+      "loss": 3.309,
+      "step": 16392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5385964512825012,
+      "learning_rate": 0.0004999484546212858,
+      "loss": 3.0099,
+      "step": 16393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5380749702453613,
+      "learning_rate": 0.0004999369763496759,
+      "loss": 3.094,
+      "step": 16394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.518419086933136,
+      "learning_rate": 0.0004999254975514705,
+      "loss": 3.0281,
+      "step": 16395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5032293200492859,
+      "learning_rate": 0.0004999140182266997,
+      "loss": 3.1709,
+      "step": 16396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5485716462135315,
+      "learning_rate": 0.0004999025383753941,
+      "loss": 3.0291,
+      "step": 16397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5190295577049255,
+      "learning_rate": 0.0004998910579975835,
+      "loss": 3.1124,
+      "step": 16398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5125435590744019,
+      "learning_rate": 0.0004998795770932984,
+      "loss": 3.0552,
+      "step": 16399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5453558564186096,
+      "learning_rate": 0.000499868095662569,
+      "loss": 3.0534,
+      "step": 16400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49550220370292664,
+      "learning_rate": 0.0004998566137054255,
+      "loss": 3.3442,
+      "step": 16401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4840456247329712,
+      "learning_rate": 0.000499845131221898,
+      "loss": 3.1657,
+      "step": 16402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5119274258613586,
+      "learning_rate": 0.000499833648212017,
+      "loss": 3.4076,
+      "step": 16403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48557212948799133,
+      "learning_rate": 0.0004998221646758125,
+      "loss": 3.3897,
+      "step": 16404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.516492486000061,
+      "learning_rate": 0.000499810680613315,
+      "loss": 3.04,
+      "step": 16405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5236262083053589,
+      "learning_rate": 0.0004997991960245547,
+      "loss": 3.4405,
+      "step": 16406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5504629015922546,
+      "learning_rate": 0.0004997877109095616,
+      "loss": 3.0354,
+      "step": 16407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5525025725364685,
+      "learning_rate": 0.0004997762252683663,
+      "loss": 3.1878,
+      "step": 16408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5103166699409485,
+      "learning_rate": 0.0004997647391009986,
+      "loss": 3.0423,
+      "step": 16409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5244056582450867,
+      "learning_rate": 0.0004997532524074892,
+      "loss": 3.3401,
+      "step": 16410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5187060236930847,
+      "learning_rate": 0.0004997417651878682,
+      "loss": 3.2776,
+      "step": 16411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5291786789894104,
+      "learning_rate": 0.0004997302774421658,
+      "loss": 3.0764,
+      "step": 16412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.509762704372406,
+      "learning_rate": 0.0004997187891704123,
+      "loss": 3.2449,
+      "step": 16413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5269352793693542,
+      "learning_rate": 0.0004997073003726378,
+      "loss": 3.169,
+      "step": 16414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.526393473148346,
+      "learning_rate": 0.0004996958110488729,
+      "loss": 3.1764,
+      "step": 16415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5143744945526123,
+      "learning_rate": 0.0004996843211991477,
+      "loss": 3.1803,
+      "step": 16416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.526218056678772,
+      "learning_rate": 0.0004996728308234923,
+      "loss": 3.1629,
+      "step": 16417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.46396979689598083,
+      "learning_rate": 0.000499661339921937,
+      "loss": 3.2116,
+      "step": 16418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49755245447158813,
+      "learning_rate": 0.0004996498484945123,
+      "loss": 3.2634,
+      "step": 16419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5120879411697388,
+      "learning_rate": 0.0004996383565412483,
+      "loss": 3.1099,
+      "step": 16420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5071130394935608,
+      "learning_rate": 0.0004996268640621753,
+      "loss": 3.1576,
+      "step": 16421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5468621850013733,
+      "learning_rate": 0.0004996153710573236,
+      "loss": 3.1405,
+      "step": 16422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5439679622650146,
+      "learning_rate": 0.0004996038775267233,
+      "loss": 3.0414,
+      "step": 16423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5470306277275085,
+      "learning_rate": 0.0004995923834704049,
+      "loss": 3.2407,
+      "step": 16424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4964117109775543,
+      "learning_rate": 0.0004995808888883986,
+      "loss": 3.0354,
+      "step": 16425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5191156268119812,
+      "learning_rate": 0.0004995693937807345,
+      "loss": 3.2839,
+      "step": 16426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5142076015472412,
+      "learning_rate": 0.0004995578981474431,
+      "loss": 3.212,
+      "step": 16427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4969421923160553,
+      "learning_rate": 0.0004995464019885548,
+      "loss": 2.959,
+      "step": 16428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5990926623344421,
+      "learning_rate": 0.0004995349053040993,
+      "loss": 3.2161,
+      "step": 16429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5438251495361328,
+      "learning_rate": 0.0004995234080941075,
+      "loss": 3.2263,
+      "step": 16430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5286610722541809,
+      "learning_rate": 0.0004995119103586095,
+      "loss": 3.1011,
+      "step": 16431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5550453662872314,
+      "learning_rate": 0.0004995004120976354,
+      "loss": 3.2373,
+      "step": 16432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5107359290122986,
+      "learning_rate": 0.0004994889133112156,
+      "loss": 3.2337,
+      "step": 16433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5204411149024963,
+      "learning_rate": 0.0004994774139993805,
+      "loss": 3.3439,
+      "step": 16434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5322909355163574,
+      "learning_rate": 0.0004994659141621602,
+      "loss": 3.5071,
+      "step": 16435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5268922448158264,
+      "learning_rate": 0.0004994544137995851,
+      "loss": 3.0943,
+      "step": 16436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5423510074615479,
+      "learning_rate": 0.0004994429129116854,
+      "loss": 3.2312,
+      "step": 16437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5225160121917725,
+      "learning_rate": 0.0004994314114984915,
+      "loss": 3.1335,
+      "step": 16438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5021830797195435,
+      "learning_rate": 0.0004994199095600337,
+      "loss": 2.975,
+      "step": 16439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5276130437850952,
+      "learning_rate": 0.0004994084070963421,
+      "loss": 3.2795,
+      "step": 16440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5111821889877319,
+      "learning_rate": 0.0004993969041074473,
+      "loss": 2.9674,
+      "step": 16441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5199219584465027,
+      "learning_rate": 0.0004993854005933793,
+      "loss": 3.1355,
+      "step": 16442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5186803340911865,
+      "learning_rate": 0.0004993738965541686,
+      "loss": 3.2157,
+      "step": 16443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5261385440826416,
+      "learning_rate": 0.0004993623919898454,
+      "loss": 3.2304,
+      "step": 16444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5131444334983826,
+      "learning_rate": 0.00049935088690044,
+      "loss": 3.1116,
+      "step": 16445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5108278393745422,
+      "learning_rate": 0.0004993393812859827,
+      "loss": 3.322,
+      "step": 16446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5285989046096802,
+      "learning_rate": 0.0004993278751465039,
+      "loss": 3.3048,
+      "step": 16447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5042722225189209,
+      "learning_rate": 0.0004993163684820338,
+      "loss": 3.0386,
+      "step": 16448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5362065434455872,
+      "learning_rate": 0.0004993048612926028,
+      "loss": 3.0104,
+      "step": 16449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49889418482780457,
+      "learning_rate": 0.0004992933535782411,
+      "loss": 3.0916,
+      "step": 16450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.521359920501709,
+      "learning_rate": 0.0004992818453389791,
+      "loss": 3.1019,
+      "step": 16451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5698845386505127,
+      "learning_rate": 0.0004992703365748471,
+      "loss": 3.1863,
+      "step": 16452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45893803238868713,
+      "learning_rate": 0.0004992588272858753,
+      "loss": 3.2055,
+      "step": 16453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5086631178855896,
+      "learning_rate": 0.0004992473174720942,
+      "loss": 3.2734,
+      "step": 16454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.507537305355072,
+      "learning_rate": 0.0004992358071335338,
+      "loss": 3.327,
+      "step": 16455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5071486830711365,
+      "learning_rate": 0.0004992242962702248,
+      "loss": 3.1988,
+      "step": 16456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5107549428939819,
+      "learning_rate": 0.0004992127848821973,
+      "loss": 3.1467,
+      "step": 16457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5230646729469299,
+      "learning_rate": 0.0004992012729694817,
+      "loss": 3.1563,
+      "step": 16458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5297747850418091,
+      "learning_rate": 0.0004991897605321082,
+      "loss": 3.1184,
+      "step": 16459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5269367098808289,
+      "learning_rate": 0.0004991782475701073,
+      "loss": 2.8786,
+      "step": 16460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4930148720741272,
+      "learning_rate": 0.0004991667340835093,
+      "loss": 3.0121,
+      "step": 16461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5489005446434021,
+      "learning_rate": 0.0004991552200723443,
+      "loss": 3.0634,
+      "step": 16462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5134395360946655,
+      "learning_rate": 0.0004991437055366428,
+      "loss": 3.2313,
+      "step": 16463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5260294079780579,
+      "learning_rate": 0.0004991321904764352,
+      "loss": 3.0926,
+      "step": 16464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5649300217628479,
+      "learning_rate": 0.0004991206748917517,
+      "loss": 3.2247,
+      "step": 16465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5012965798377991,
+      "learning_rate": 0.0004991091587826225,
+      "loss": 3.1459,
+      "step": 16466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49580734968185425,
+      "learning_rate": 0.0004990976421490783,
+      "loss": 3.1876,
+      "step": 16467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54473477602005,
+      "learning_rate": 0.0004990861249911492,
+      "loss": 3.3208,
+      "step": 16468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5577808022499084,
+      "learning_rate": 0.0004990746073088655,
+      "loss": 3.0318,
+      "step": 16469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5220108032226562,
+      "learning_rate": 0.0004990630891022576,
+      "loss": 3.07,
+      "step": 16470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4759686291217804,
+      "learning_rate": 0.0004990515703713559,
+      "loss": 3.1017,
+      "step": 16471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5046111345291138,
+      "learning_rate": 0.0004990400511161907,
+      "loss": 3.1224,
+      "step": 16472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5120763182640076,
+      "learning_rate": 0.0004990285313367922,
+      "loss": 3.2141,
+      "step": 16473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5458747148513794,
+      "learning_rate": 0.0004990170110331908,
+      "loss": 3.126,
+      "step": 16474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.540588915348053,
+      "learning_rate": 0.000499005490205417,
+      "loss": 3.1728,
+      "step": 16475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5352221727371216,
+      "learning_rate": 0.000498993968853501,
+      "loss": 3.1001,
+      "step": 16476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5034423470497131,
+      "learning_rate": 0.0004989824469774732,
+      "loss": 3.1839,
+      "step": 16477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5104349255561829,
+      "learning_rate": 0.0004989709245773639,
+      "loss": 3.1802,
+      "step": 16478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5068408846855164,
+      "learning_rate": 0.0004989594016532036,
+      "loss": 3.0528,
+      "step": 16479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5584886074066162,
+      "learning_rate": 0.0004989478782050224,
+      "loss": 3.1633,
+      "step": 16480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4965101480484009,
+      "learning_rate": 0.0004989363542328508,
+      "loss": 2.9951,
+      "step": 16481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5258487462997437,
+      "learning_rate": 0.0004989248297367191,
+      "loss": 3.2203,
+      "step": 16482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49109527468681335,
+      "learning_rate": 0.0004989133047166577,
+      "loss": 3.2713,
+      "step": 16483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49905163049697876,
+      "learning_rate": 0.0004989017791726971,
+      "loss": 3.117,
+      "step": 16484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5777082443237305,
+      "learning_rate": 0.0004988902531048673,
+      "loss": 3.1115,
+      "step": 16485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5189237594604492,
+      "learning_rate": 0.0004988787265131989,
+      "loss": 3.0824,
+      "step": 16486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.569271981716156,
+      "learning_rate": 0.0004988671993977221,
+      "loss": 2.9983,
+      "step": 16487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5757372379302979,
+      "learning_rate": 0.0004988556717584676,
+      "loss": 3.0703,
+      "step": 16488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5193915963172913,
+      "learning_rate": 0.0004988441435954654,
+      "loss": 3.1659,
+      "step": 16489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5080418586730957,
+      "learning_rate": 0.000498832614908746,
+      "loss": 3.1788,
+      "step": 16490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5090472102165222,
+      "learning_rate": 0.0004988210856983398,
+      "loss": 3.1489,
+      "step": 16491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.466819167137146,
+      "learning_rate": 0.000498809555964277,
+      "loss": 3.1436,
+      "step": 16492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5463626980781555,
+      "learning_rate": 0.0004987980257065883,
+      "loss": 3.2058,
+      "step": 16493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5318609476089478,
+      "learning_rate": 0.0004987864949253037,
+      "loss": 3.333,
+      "step": 16494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5022921562194824,
+      "learning_rate": 0.0004987749636204537,
+      "loss": 3.0602,
+      "step": 16495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49121004343032837,
+      "learning_rate": 0.0004987634317920688,
+      "loss": 3.3215,
+      "step": 16496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5183935761451721,
+      "learning_rate": 0.0004987518994401792,
+      "loss": 3.2748,
+      "step": 16497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4942324757575989,
+      "learning_rate": 0.0004987403665648155,
+      "loss": 3.0571,
+      "step": 16498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5025029182434082,
+      "learning_rate": 0.0004987288331660078,
+      "loss": 3.2539,
+      "step": 16499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5152884125709534,
+      "learning_rate": 0.0004987172992437866,
+      "loss": 3.2326,
+      "step": 16500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5195552706718445,
+      "learning_rate": 0.0004987057647981824,
+      "loss": 3.0311,
+      "step": 16501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5349781513214111,
+      "learning_rate": 0.0004986942298292253,
+      "loss": 3.1752,
+      "step": 16502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5725705027580261,
+      "learning_rate": 0.0004986826943369459,
+      "loss": 3.2543,
+      "step": 16503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6165834069252014,
+      "learning_rate": 0.0004986711583213745,
+      "loss": 3.3868,
+      "step": 16504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5173270106315613,
+      "learning_rate": 0.0004986596217825415,
+      "loss": 3.1788,
+      "step": 16505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5284629464149475,
+      "learning_rate": 0.0004986480847204772,
+      "loss": 3.2897,
+      "step": 16506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5097475647926331,
+      "learning_rate": 0.0004986365471352122,
+      "loss": 2.8815,
+      "step": 16507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5338273048400879,
+      "learning_rate": 0.0004986250090267768,
+      "loss": 3.2031,
+      "step": 16508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5456317663192749,
+      "learning_rate": 0.0004986134703952013,
+      "loss": 2.9568,
+      "step": 16509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55818772315979,
+      "learning_rate": 0.0004986019312405161,
+      "loss": 3.1333,
+      "step": 16510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5008224844932556,
+      "learning_rate": 0.0004985903915627517,
+      "loss": 3.2126,
+      "step": 16511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4851469099521637,
+      "learning_rate": 0.0004985788513619384,
+      "loss": 3.0776,
+      "step": 16512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5189734697341919,
+      "learning_rate": 0.0004985673106381066,
+      "loss": 3.0294,
+      "step": 16513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5031270980834961,
+      "learning_rate": 0.0004985557693912867,
+      "loss": 3.4658,
+      "step": 16514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5297790169715881,
+      "learning_rate": 0.0004985442276215092,
+      "loss": 3.2422,
+      "step": 16515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.536777138710022,
+      "learning_rate": 0.0004985326853288044,
+      "loss": 3.1539,
+      "step": 16516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5064894556999207,
+      "learning_rate": 0.0004985211425132027,
+      "loss": 3.1133,
+      "step": 16517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.511317789554596,
+      "learning_rate": 0.0004985095991747345,
+      "loss": 3.1467,
+      "step": 16518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49673980474472046,
+      "learning_rate": 0.0004984980553134302,
+      "loss": 3.229,
+      "step": 16519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5076729655265808,
+      "learning_rate": 0.0004984865109293203,
+      "loss": 3.3867,
+      "step": 16520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5597838163375854,
+      "learning_rate": 0.0004984749660224351,
+      "loss": 3.2826,
+      "step": 16521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5371355414390564,
+      "learning_rate": 0.0004984634205928049,
+      "loss": 3.2134,
+      "step": 16522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5134449005126953,
+      "learning_rate": 0.0004984518746404604,
+      "loss": 3.2228,
+      "step": 16523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.514105498790741,
+      "learning_rate": 0.0004984403281654318,
+      "loss": 2.9934,
+      "step": 16524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5345868468284607,
+      "learning_rate": 0.0004984287811677496,
+      "loss": 3.3375,
+      "step": 16525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5686771273612976,
+      "learning_rate": 0.0004984172336474441,
+      "loss": 3.2227,
+      "step": 16526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4849824905395508,
+      "learning_rate": 0.000498405685604546,
+      "loss": 3.2113,
+      "step": 16527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5313514471054077,
+      "learning_rate": 0.0004983941370390853,
+      "loss": 3.3788,
+      "step": 16528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5120206475257874,
+      "learning_rate": 0.0004983825879510927,
+      "loss": 3.1637,
+      "step": 16529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5304714441299438,
+      "learning_rate": 0.0004983710383405984,
+      "loss": 2.988,
+      "step": 16530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5290876030921936,
+      "learning_rate": 0.0004983594882076333,
+      "loss": 3.3558,
+      "step": 16531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5857232809066772,
+      "learning_rate": 0.0004983479375522272,
+      "loss": 3.102,
+      "step": 16532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7322335839271545,
+      "learning_rate": 0.0004983363863744108,
+      "loss": 3.0765,
+      "step": 16533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5193672776222229,
+      "learning_rate": 0.0004983248346742147,
+      "loss": 3.0546,
+      "step": 16534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5496091246604919,
+      "learning_rate": 0.000498313282451669,
+      "loss": 2.9767,
+      "step": 16535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5549640655517578,
+      "learning_rate": 0.0004983017297068044,
+      "loss": 3.0501,
+      "step": 16536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49452897906303406,
+      "learning_rate": 0.0004982901764396511,
+      "loss": 3.1101,
+      "step": 16537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5188285112380981,
+      "learning_rate": 0.0004982786226502396,
+      "loss": 3.276,
+      "step": 16538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5108150243759155,
+      "learning_rate": 0.0004982670683386006,
+      "loss": 3.1788,
+      "step": 16539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5602210164070129,
+      "learning_rate": 0.0004982555135047641,
+      "loss": 3.1825,
+      "step": 16540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5105909109115601,
+      "learning_rate": 0.0004982439581487607,
+      "loss": 3.3291,
+      "step": 16541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5331498384475708,
+      "learning_rate": 0.000498232402270621,
+      "loss": 3.2307,
+      "step": 16542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5256636142730713,
+      "learning_rate": 0.0004982208458703752,
+      "loss": 2.8307,
+      "step": 16543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5503387451171875,
+      "learning_rate": 0.000498209288948054,
+      "loss": 3.2409,
+      "step": 16544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5669764280319214,
+      "learning_rate": 0.0004981977315036876,
+      "loss": 3.1763,
+      "step": 16545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5365728139877319,
+      "learning_rate": 0.0004981861735373064,
+      "loss": 3.0852,
+      "step": 16546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5224799513816833,
+      "learning_rate": 0.000498174615048941,
+      "loss": 3.1728,
+      "step": 16547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5562565922737122,
+      "learning_rate": 0.0004981630560386219,
+      "loss": 3.1805,
+      "step": 16548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5232570171356201,
+      "learning_rate": 0.0004981514965063795,
+      "loss": 3.1247,
+      "step": 16549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5539358258247375,
+      "learning_rate": 0.0004981399364522441,
+      "loss": 3.2232,
+      "step": 16550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5199190378189087,
+      "learning_rate": 0.0004981283758762462,
+      "loss": 3.1485,
+      "step": 16551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.516223132610321,
+      "learning_rate": 0.0004981168147784163,
+      "loss": 3.1427,
+      "step": 16552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.504279613494873,
+      "learning_rate": 0.0004981052531587849,
+      "loss": 3.318,
+      "step": 16553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5051873326301575,
+      "learning_rate": 0.0004980936910173823,
+      "loss": 3.2255,
+      "step": 16554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5611341595649719,
+      "learning_rate": 0.0004980821283542392,
+      "loss": 3.0024,
+      "step": 16555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5161449313163757,
+      "learning_rate": 0.0004980705651693858,
+      "loss": 3.0296,
+      "step": 16556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5371371507644653,
+      "learning_rate": 0.0004980590014628526,
+      "loss": 3.273,
+      "step": 16557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.533517599105835,
+      "learning_rate": 0.0004980474372346701,
+      "loss": 3.2346,
+      "step": 16558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5515034794807434,
+      "learning_rate": 0.0004980358724848688,
+      "loss": 3.1066,
+      "step": 16559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5124566555023193,
+      "learning_rate": 0.0004980243072134792,
+      "loss": 3.2288,
+      "step": 16560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4827638864517212,
+      "learning_rate": 0.0004980127414205317,
+      "loss": 3.3534,
+      "step": 16561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5892248749732971,
+      "learning_rate": 0.0004980011751060566,
+      "loss": 3.207,
+      "step": 16562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5171937942504883,
+      "learning_rate": 0.0004979896082700846,
+      "loss": 3.2058,
+      "step": 16563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5337836146354675,
+      "learning_rate": 0.000497978040912646,
+      "loss": 3.3217,
+      "step": 16564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5042610764503479,
+      "learning_rate": 0.0004979664730337714,
+      "loss": 3.3613,
+      "step": 16565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515651285648346,
+      "learning_rate": 0.0004979549046334913,
+      "loss": 3.0309,
+      "step": 16566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5128003358840942,
+      "learning_rate": 0.000497943335711836,
+      "loss": 3.4516,
+      "step": 16567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5067887306213379,
+      "learning_rate": 0.0004979317662688359,
+      "loss": 3.1118,
+      "step": 16568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4763834476470947,
+      "learning_rate": 0.0004979201963045218,
+      "loss": 3.3369,
+      "step": 16569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5007560849189758,
+      "learning_rate": 0.0004979086258189241,
+      "loss": 3.1067,
+      "step": 16570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.539618968963623,
+      "learning_rate": 0.0004978970548120729,
+      "loss": 3.2345,
+      "step": 16571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5771459341049194,
+      "learning_rate": 0.0004978854832839992,
+      "loss": 3.1452,
+      "step": 16572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5069820284843445,
+      "learning_rate": 0.000497873911234733,
+      "loss": 3.2058,
+      "step": 16573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5238664746284485,
+      "learning_rate": 0.000497862338664305,
+      "loss": 3.0031,
+      "step": 16574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5101388692855835,
+      "learning_rate": 0.0004978507655727458,
+      "loss": 3.123,
+      "step": 16575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5062222480773926,
+      "learning_rate": 0.0004978391919600857,
+      "loss": 3.1039,
+      "step": 16576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.571029007434845,
+      "learning_rate": 0.0004978276178263553,
+      "loss": 3.1371,
+      "step": 16577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.519842803478241,
+      "learning_rate": 0.000497816043171585,
+      "loss": 3.0524,
+      "step": 16578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6620784997940063,
+      "learning_rate": 0.0004978044679958053,
+      "loss": 3.2173,
+      "step": 16579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5876750349998474,
+      "learning_rate": 0.0004977928922990467,
+      "loss": 3.1815,
+      "step": 16580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5399341583251953,
+      "learning_rate": 0.0004977813160813397,
+      "loss": 3.3868,
+      "step": 16581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5021479725837708,
+      "learning_rate": 0.0004977697393427148,
+      "loss": 3.2335,
+      "step": 16582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5305050015449524,
+      "learning_rate": 0.0004977581620832024,
+      "loss": 3.2448,
+      "step": 16583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4928508698940277,
+      "learning_rate": 0.0004977465843028331,
+      "loss": 3.2395,
+      "step": 16584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49715203046798706,
+      "learning_rate": 0.0004977350060016374,
+      "loss": 3.1318,
+      "step": 16585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49613332748413086,
+      "learning_rate": 0.0004977234271796458,
+      "loss": 3.2116,
+      "step": 16586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5376517176628113,
+      "learning_rate": 0.0004977118478368886,
+      "loss": 3.1922,
+      "step": 16587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5719401836395264,
+      "learning_rate": 0.0004977002679733964,
+      "loss": 3.2298,
+      "step": 16588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5064448118209839,
+      "learning_rate": 0.0004976886875891998,
+      "loss": 3.2184,
+      "step": 16589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5336682200431824,
+      "learning_rate": 0.0004976771066843293,
+      "loss": 3.1029,
+      "step": 16590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5417354106903076,
+      "learning_rate": 0.0004976655252588153,
+      "loss": 3.2497,
+      "step": 16591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5155006051063538,
+      "learning_rate": 0.0004976539433126884,
+      "loss": 3.2171,
+      "step": 16592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5138181447982788,
+      "learning_rate": 0.000497642360845979,
+      "loss": 3.0924,
+      "step": 16593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5162732601165771,
+      "learning_rate": 0.0004976307778587176,
+      "loss": 3.218,
+      "step": 16594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5005708932876587,
+      "learning_rate": 0.000497619194350935,
+      "loss": 3.3464,
+      "step": 16595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5471556186676025,
+      "learning_rate": 0.0004976076103226613,
+      "loss": 3.286,
+      "step": 16596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5286422371864319,
+      "learning_rate": 0.0004975960257739272,
+      "loss": 2.9602,
+      "step": 16597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5641829967498779,
+      "learning_rate": 0.0004975844407047632,
+      "loss": 3.2977,
+      "step": 16598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5358396172523499,
+      "learning_rate": 0.0004975728551151998,
+      "loss": 3.1017,
+      "step": 16599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5861735939979553,
+      "learning_rate": 0.0004975612690052675,
+      "loss": 3.1687,
+      "step": 16600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5610910654067993,
+      "learning_rate": 0.0004975496823749969,
+      "loss": 3.1995,
+      "step": 16601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5274034738540649,
+      "learning_rate": 0.0004975380952244185,
+      "loss": 3.2817,
+      "step": 16602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5060899257659912,
+      "learning_rate": 0.0004975265075535626,
+      "loss": 3.2932,
+      "step": 16603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5002941489219666,
+      "learning_rate": 0.0004975149193624601,
+      "loss": 3.1504,
+      "step": 16604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48612910509109497,
+      "learning_rate": 0.0004975033306511412,
+      "loss": 3.1963,
+      "step": 16605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5231236815452576,
+      "learning_rate": 0.0004974917414196365,
+      "loss": 3.1359,
+      "step": 16606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5047218203544617,
+      "learning_rate": 0.0004974801516679766,
+      "loss": 3.1402,
+      "step": 16607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4852062165737152,
+      "learning_rate": 0.000497468561396192,
+      "loss": 3.2184,
+      "step": 16608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5307152271270752,
+      "learning_rate": 0.0004974569706043132,
+      "loss": 3.1907,
+      "step": 16609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5003814101219177,
+      "learning_rate": 0.0004974453792923708,
+      "loss": 3.0009,
+      "step": 16610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5295878052711487,
+      "learning_rate": 0.0004974337874603952,
+      "loss": 3.1932,
+      "step": 16611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5771824717521667,
+      "learning_rate": 0.000497422195108417,
+      "loss": 3.3406,
+      "step": 16612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5096525549888611,
+      "learning_rate": 0.0004974106022364669,
+      "loss": 3.0476,
+      "step": 16613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5477951765060425,
+      "learning_rate": 0.000497399008844575,
+      "loss": 3.2816,
+      "step": 16614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5155587196350098,
+      "learning_rate": 0.0004973874149327722,
+      "loss": 2.9737,
+      "step": 16615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5152885317802429,
+      "learning_rate": 0.0004973758205010889,
+      "loss": 3.2941,
+      "step": 16616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5501928925514221,
+      "learning_rate": 0.0004973642255495558,
+      "loss": 3.3104,
+      "step": 16617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5133585333824158,
+      "learning_rate": 0.0004973526300782032,
+      "loss": 3.1066,
+      "step": 16618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5019925832748413,
+      "learning_rate": 0.0004973410340870618,
+      "loss": 3.2807,
+      "step": 16619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5474452376365662,
+      "learning_rate": 0.0004973294375761621,
+      "loss": 3.2518,
+      "step": 16620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5307877063751221,
+      "learning_rate": 0.0004973178405455347,
+      "loss": 2.98,
+      "step": 16621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5195660591125488,
+      "learning_rate": 0.00049730624299521,
+      "loss": 3.2952,
+      "step": 16622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5276235938072205,
+      "learning_rate": 0.0004972946449252187,
+      "loss": 3.2442,
+      "step": 16623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5224148035049438,
+      "learning_rate": 0.0004972830463355912,
+      "loss": 3.0754,
+      "step": 16624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5097349286079407,
+      "learning_rate": 0.0004972714472263581,
+      "loss": 3.3027,
+      "step": 16625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.511198878288269,
+      "learning_rate": 0.00049725984759755,
+      "loss": 3.229,
+      "step": 16626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5765299797058105,
+      "learning_rate": 0.0004972482474491975,
+      "loss": 3.101,
+      "step": 16627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5171129703521729,
+      "learning_rate": 0.0004972366467813309,
+      "loss": 3.1579,
+      "step": 16628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5174678564071655,
+      "learning_rate": 0.0004972250455939811,
+      "loss": 3.2538,
+      "step": 16629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5216899514198303,
+      "learning_rate": 0.0004972134438871786,
+      "loss": 3.0923,
+      "step": 16630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5786752104759216,
+      "learning_rate": 0.0004972018416609536,
+      "loss": 3.127,
+      "step": 16631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5211129784584045,
+      "learning_rate": 0.000497190238915337,
+      "loss": 3.1011,
+      "step": 16632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5141299366950989,
+      "learning_rate": 0.0004971786356503592,
+      "loss": 2.9091,
+      "step": 16633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48368707299232483,
+      "learning_rate": 0.0004971670318660509,
+      "loss": 3.324,
+      "step": 16634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47780829668045044,
+      "learning_rate": 0.0004971554275624425,
+      "loss": 3.0679,
+      "step": 16635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5046222805976868,
+      "learning_rate": 0.0004971438227395648,
+      "loss": 3.2293,
+      "step": 16636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5113916993141174,
+      "learning_rate": 0.000497132217397448,
+      "loss": 3.0101,
+      "step": 16637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5229575037956238,
+      "learning_rate": 0.000497120611536123,
+      "loss": 3.2457,
+      "step": 16638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.512291669845581,
+      "learning_rate": 0.0004971090051556202,
+      "loss": 3.259,
+      "step": 16639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5554661750793457,
+      "learning_rate": 0.0004970973982559702,
+      "loss": 3.1555,
+      "step": 16640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5595203042030334,
+      "learning_rate": 0.0004970857908372037,
+      "loss": 3.1944,
+      "step": 16641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5318277478218079,
+      "learning_rate": 0.0004970741828993511,
+      "loss": 2.9727,
+      "step": 16642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5298823714256287,
+      "learning_rate": 0.0004970625744424431,
+      "loss": 2.9161,
+      "step": 16643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5190830230712891,
+      "learning_rate": 0.0004970509654665101,
+      "loss": 3.2831,
+      "step": 16644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5548256039619446,
+      "learning_rate": 0.0004970393559715827,
+      "loss": 3.3253,
+      "step": 16645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5534726977348328,
+      "learning_rate": 0.0004970277459576917,
+      "loss": 3.0462,
+      "step": 16646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5380357503890991,
+      "learning_rate": 0.0004970161354248675,
+      "loss": 3.2276,
+      "step": 16647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5422712564468384,
+      "learning_rate": 0.0004970045243731406,
+      "loss": 3.0344,
+      "step": 16648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5209787487983704,
+      "learning_rate": 0.0004969929128025419,
+      "loss": 3.4289,
+      "step": 16649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5309789776802063,
+      "learning_rate": 0.0004969813007131016,
+      "loss": 3.2243,
+      "step": 16650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5202519297599792,
+      "learning_rate": 0.0004969696881048507,
+      "loss": 3.3096,
+      "step": 16651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5129031538963318,
+      "learning_rate": 0.0004969580749778193,
+      "loss": 3.2491,
+      "step": 16652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5805734395980835,
+      "learning_rate": 0.0004969464613320384,
+      "loss": 3.1584,
+      "step": 16653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5051552057266235,
+      "learning_rate": 0.0004969348471675384,
+      "loss": 3.1706,
+      "step": 16654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5367906093597412,
+      "learning_rate": 0.0004969232324843497,
+      "loss": 3.2394,
+      "step": 16655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48859554529190063,
+      "learning_rate": 0.0004969116172825034,
+      "loss": 3.1854,
+      "step": 16656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5127988457679749,
+      "learning_rate": 0.0004969000015620297,
+      "loss": 3.1622,
+      "step": 16657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5549734830856323,
+      "learning_rate": 0.0004968883853229592,
+      "loss": 3.1321,
+      "step": 16658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5598774552345276,
+      "learning_rate": 0.0004968767685653226,
+      "loss": 3.4022,
+      "step": 16659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5506905913352966,
+      "learning_rate": 0.0004968651512891506,
+      "loss": 3.3344,
+      "step": 16660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5095792412757874,
+      "learning_rate": 0.0004968535334944736,
+      "loss": 3.1894,
+      "step": 16661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.526882529258728,
+      "learning_rate": 0.0004968419151813224,
+      "loss": 3.3501,
+      "step": 16662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5462402105331421,
+      "learning_rate": 0.0004968302963497273,
+      "loss": 2.8678,
+      "step": 16663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.528898298740387,
+      "learning_rate": 0.0004968186769997191,
+      "loss": 3.1944,
+      "step": 16664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5134194493293762,
+      "learning_rate": 0.0004968070571313285,
+      "loss": 3.1576,
+      "step": 16665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5237089395523071,
+      "learning_rate": 0.0004967954367445859,
+      "loss": 3.267,
+      "step": 16666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.558527946472168,
+      "learning_rate": 0.0004967838158395219,
+      "loss": 3.1597,
+      "step": 16667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5071585178375244,
+      "learning_rate": 0.0004967721944161673,
+      "loss": 3.2747,
+      "step": 16668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5414650440216064,
+      "learning_rate": 0.0004967605724745527,
+      "loss": 3.1734,
+      "step": 16669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48503783345222473,
+      "learning_rate": 0.0004967489500147086,
+      "loss": 3.1875,
+      "step": 16670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5278720855712891,
+      "learning_rate": 0.0004967373270366655,
+      "loss": 3.0154,
+      "step": 16671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5245896577835083,
+      "learning_rate": 0.0004967257035404542,
+      "loss": 2.9901,
+      "step": 16672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5041775107383728,
+      "learning_rate": 0.0004967140795261053,
+      "loss": 3.1008,
+      "step": 16673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5002780556678772,
+      "learning_rate": 0.0004967024549936493,
+      "loss": 3.3013,
+      "step": 16674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5385328531265259,
+      "learning_rate": 0.000496690829943117,
+      "loss": 3.3536,
+      "step": 16675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47223567962646484,
+      "learning_rate": 0.0004966792043745389,
+      "loss": 3.1565,
+      "step": 16676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5765127539634705,
+      "learning_rate": 0.0004966675782879455,
+      "loss": 3.1308,
+      "step": 16677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5132501125335693,
+      "learning_rate": 0.0004966559516833677,
+      "loss": 3.2131,
+      "step": 16678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5195982456207275,
+      "learning_rate": 0.0004966443245608359,
+      "loss": 3.3829,
+      "step": 16679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4910086989402771,
+      "learning_rate": 0.0004966326969203807,
+      "loss": 3.2672,
+      "step": 16680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5135030746459961,
+      "learning_rate": 0.0004966210687620329,
+      "loss": 3.3715,
+      "step": 16681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5263185501098633,
+      "learning_rate": 0.0004966094400858231,
+      "loss": 3.0218,
+      "step": 16682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5208287239074707,
+      "learning_rate": 0.0004965978108917818,
+      "loss": 3.1165,
+      "step": 16683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5360139012336731,
+      "learning_rate": 0.0004965861811799397,
+      "loss": 3.0333,
+      "step": 16684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47904497385025024,
+      "learning_rate": 0.0004965745509503275,
+      "loss": 3.2553,
+      "step": 16685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4816090166568756,
+      "learning_rate": 0.0004965629202029758,
+      "loss": 3.1955,
+      "step": 16686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5097158551216125,
+      "learning_rate": 0.0004965512889379151,
+      "loss": 3.2383,
+      "step": 16687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.509074330329895,
+      "learning_rate": 0.000496539657155176,
+      "loss": 3.2118,
+      "step": 16688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5360403060913086,
+      "learning_rate": 0.0004965280248547895,
+      "loss": 2.9973,
+      "step": 16689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5154137015342712,
+      "learning_rate": 0.0004965163920367859,
+      "loss": 3.3851,
+      "step": 16690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48965588212013245,
+      "learning_rate": 0.0004965047587011959,
+      "loss": 3.2216,
+      "step": 16691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4903731048107147,
+      "learning_rate": 0.0004964931248480503,
+      "loss": 3.2034,
+      "step": 16692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5228937268257141,
+      "learning_rate": 0.0004964814904773795,
+      "loss": 3.1598,
+      "step": 16693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5695022940635681,
+      "learning_rate": 0.0004964698555892144,
+      "loss": 3.1179,
+      "step": 16694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5156790018081665,
+      "learning_rate": 0.0004964582201835855,
+      "loss": 3.0333,
+      "step": 16695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5029388666152954,
+      "learning_rate": 0.0004964465842605234,
+      "loss": 3.1303,
+      "step": 16696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5495452880859375,
+      "learning_rate": 0.0004964349478200588,
+      "loss": 3.178,
+      "step": 16697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4993706941604614,
+      "learning_rate": 0.0004964233108622224,
+      "loss": 3.4597,
+      "step": 16698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5246551632881165,
+      "learning_rate": 0.0004964116733870448,
+      "loss": 3.1988,
+      "step": 16699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5039535164833069,
+      "learning_rate": 0.0004964000353945566,
+      "loss": 3.3829,
+      "step": 16700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5181304812431335,
+      "learning_rate": 0.0004963883968847884,
+      "loss": 3.318,
+      "step": 16701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5846447944641113,
+      "learning_rate": 0.0004963767578577712,
+      "loss": 3.1945,
+      "step": 16702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.518142580986023,
+      "learning_rate": 0.0004963651183135353,
+      "loss": 3.2293,
+      "step": 16703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5273866057395935,
+      "learning_rate": 0.0004963534782521115,
+      "loss": 3.0665,
+      "step": 16704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5090498328208923,
+      "learning_rate": 0.0004963418376735303,
+      "loss": 3.1259,
+      "step": 16705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.533167839050293,
+      "learning_rate": 0.0004963301965778226,
+      "loss": 3.3109,
+      "step": 16706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5366831421852112,
+      "learning_rate": 0.0004963185549650189,
+      "loss": 3.0486,
+      "step": 16707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5503694415092468,
+      "learning_rate": 0.00049630691283515,
+      "loss": 3.4465,
+      "step": 16708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5301731824874878,
+      "learning_rate": 0.0004962952701882463,
+      "loss": 3.219,
+      "step": 16709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4929310977458954,
+      "learning_rate": 0.0004962836270243388,
+      "loss": 3.2798,
+      "step": 16710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5285437703132629,
+      "learning_rate": 0.0004962719833434579,
+      "loss": 3.0686,
+      "step": 16711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5135146975517273,
+      "learning_rate": 0.0004962603391456345,
+      "loss": 3.1963,
+      "step": 16712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5494550466537476,
+      "learning_rate": 0.000496248694430899,
+      "loss": 3.0039,
+      "step": 16713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5301375389099121,
+      "learning_rate": 0.0004962370491992823,
+      "loss": 3.179,
+      "step": 16714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5424580574035645,
+      "learning_rate": 0.000496225403450815,
+      "loss": 3.2411,
+      "step": 16715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5308148264884949,
+      "learning_rate": 0.0004962137571855276,
+      "loss": 3.3935,
+      "step": 16716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5148587226867676,
+      "learning_rate": 0.0004962021104034511,
+      "loss": 3.0353,
+      "step": 16717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5002628564834595,
+      "learning_rate": 0.0004961904631046158,
+      "loss": 3.2585,
+      "step": 16718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5126210451126099,
+      "learning_rate": 0.0004961788152890527,
+      "loss": 3.2791,
+      "step": 16719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5162221789360046,
+      "learning_rate": 0.0004961671669567924,
+      "loss": 3.1034,
+      "step": 16720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4941016137599945,
+      "learning_rate": 0.0004961555181078655,
+      "loss": 3.2295,
+      "step": 16721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5905763506889343,
+      "learning_rate": 0.0004961438687423027,
+      "loss": 3.0365,
+      "step": 16722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49847185611724854,
+      "learning_rate": 0.0004961322188601347,
+      "loss": 3.0049,
+      "step": 16723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49443936347961426,
+      "learning_rate": 0.0004961205684613922,
+      "loss": 3.2913,
+      "step": 16724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5241524577140808,
+      "learning_rate": 0.0004961089175461059,
+      "loss": 2.907,
+      "step": 16725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5237706303596497,
+      "learning_rate": 0.0004960972661143064,
+      "loss": 3.1307,
+      "step": 16726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5108638405799866,
+      "learning_rate": 0.0004960856141660244,
+      "loss": 3.271,
+      "step": 16727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.591118335723877,
+      "learning_rate": 0.0004960739617012906,
+      "loss": 3.2943,
+      "step": 16728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5045537352561951,
+      "learning_rate": 0.0004960623087201358,
+      "loss": 3.2844,
+      "step": 16729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.569861650466919,
+      "learning_rate": 0.0004960506552225906,
+      "loss": 3.243,
+      "step": 16730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5104923844337463,
+      "learning_rate": 0.0004960390012086858,
+      "loss": 3.2714,
+      "step": 16731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5149794220924377,
+      "learning_rate": 0.0004960273466784519,
+      "loss": 2.9799,
+      "step": 16732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5185301303863525,
+      "learning_rate": 0.0004960156916319196,
+      "loss": 3.1996,
+      "step": 16733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5345672369003296,
+      "learning_rate": 0.0004960040360691199,
+      "loss": 3.1292,
+      "step": 16734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5602273941040039,
+      "learning_rate": 0.0004959923799900831,
+      "loss": 3.1398,
+      "step": 16735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5504186153411865,
+      "learning_rate": 0.0004959807233948403,
+      "loss": 3.2962,
+      "step": 16736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5613498687744141,
+      "learning_rate": 0.0004959690662834219,
+      "loss": 3.1052,
+      "step": 16737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5954709053039551,
+      "learning_rate": 0.0004959574086558586,
+      "loss": 3.1258,
+      "step": 16738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5291658639907837,
+      "learning_rate": 0.0004959457505121813,
+      "loss": 3.2572,
+      "step": 16739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5105869174003601,
+      "learning_rate": 0.0004959340918524205,
+      "loss": 3.157,
+      "step": 16740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5140098333358765,
+      "learning_rate": 0.0004959224326766071,
+      "loss": 3.2509,
+      "step": 16741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5428385734558105,
+      "learning_rate": 0.0004959107729847717,
+      "loss": 2.8749,
+      "step": 16742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5247282385826111,
+      "learning_rate": 0.000495899112776945,
+      "loss": 2.9758,
+      "step": 16743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5175803303718567,
+      "learning_rate": 0.0004958874520531578,
+      "loss": 3.2636,
+      "step": 16744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5175894498825073,
+      "learning_rate": 0.0004958757908134407,
+      "loss": 3.1949,
+      "step": 16745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49212566018104553,
+      "learning_rate": 0.0004958641290578245,
+      "loss": 3.232,
+      "step": 16746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5174520015716553,
+      "learning_rate": 0.0004958524667863399,
+      "loss": 3.2229,
+      "step": 16747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5180816054344177,
+      "learning_rate": 0.0004958408039990174,
+      "loss": 3.1441,
+      "step": 16748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5078367590904236,
+      "learning_rate": 0.0004958291406958881,
+      "loss": 3.1435,
+      "step": 16749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5164198875427246,
+      "learning_rate": 0.0004958174768769825,
+      "loss": 3.0394,
+      "step": 16750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5266910791397095,
+      "learning_rate": 0.0004958058125423315,
+      "loss": 3.048,
+      "step": 16751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.523522675037384,
+      "learning_rate": 0.0004957941476919654,
+      "loss": 3.1053,
+      "step": 16752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5101291537284851,
+      "learning_rate": 0.0004957824823259154,
+      "loss": 3.1284,
+      "step": 16753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5190765261650085,
+      "learning_rate": 0.0004957708164442119,
+      "loss": 3.1728,
+      "step": 16754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5073868632316589,
+      "learning_rate": 0.0004957591500468856,
+      "loss": 3.3285,
+      "step": 16755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5432148575782776,
+      "learning_rate": 0.0004957474831339678,
+      "loss": 2.9638,
+      "step": 16756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5055432319641113,
+      "learning_rate": 0.0004957358157054885,
+      "loss": 3.1185,
+      "step": 16757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5079781413078308,
+      "learning_rate": 0.0004957241477614787,
+      "loss": 3.0659,
+      "step": 16758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5281221866607666,
+      "learning_rate": 0.0004957124793019694,
+      "loss": 2.9227,
+      "step": 16759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5235075950622559,
+      "learning_rate": 0.0004957008103269908,
+      "loss": 3.1578,
+      "step": 16760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49377989768981934,
+      "learning_rate": 0.0004956891408365741,
+      "loss": 3.3034,
+      "step": 16761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5380136370658875,
+      "learning_rate": 0.0004956774708307499,
+      "loss": 3.1188,
+      "step": 16762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5021156072616577,
+      "learning_rate": 0.0004956658003095488,
+      "loss": 3.0115,
+      "step": 16763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5770875811576843,
+      "learning_rate": 0.0004956541292730017,
+      "loss": 3.1021,
+      "step": 16764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5153395533561707,
+      "learning_rate": 0.0004956424577211392,
+      "loss": 3.0592,
+      "step": 16765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5426536798477173,
+      "learning_rate": 0.0004956307856539922,
+      "loss": 3.2771,
+      "step": 16766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5408112406730652,
+      "learning_rate": 0.0004956191130715915,
+      "loss": 3.2663,
+      "step": 16767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5236942768096924,
+      "learning_rate": 0.0004956074399739674,
+      "loss": 3.2467,
+      "step": 16768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.523478627204895,
+      "learning_rate": 0.000495595766361151,
+      "loss": 3.16,
+      "step": 16769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5341914296150208,
+      "learning_rate": 0.0004955840922331732,
+      "loss": 3.1799,
+      "step": 16770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5106289982795715,
+      "learning_rate": 0.0004955724175900644,
+      "loss": 3.0653,
+      "step": 16771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5742872357368469,
+      "learning_rate": 0.0004955607424318555,
+      "loss": 3.1742,
+      "step": 16772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5103678107261658,
+      "learning_rate": 0.0004955490667585773,
+      "loss": 3.3389,
+      "step": 16773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5044515132904053,
+      "learning_rate": 0.0004955373905702604,
+      "loss": 2.9904,
+      "step": 16774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5162826776504517,
+      "learning_rate": 0.0004955257138669357,
+      "loss": 3.3376,
+      "step": 16775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.545150101184845,
+      "learning_rate": 0.000495514036648634,
+      "loss": 3.0011,
+      "step": 16776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5908412337303162,
+      "learning_rate": 0.0004955023589153858,
+      "loss": 3.2328,
+      "step": 16777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5039973258972168,
+      "learning_rate": 0.000495490680667222,
+      "loss": 3.0851,
+      "step": 16778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5710932016372681,
+      "learning_rate": 0.0004954790019041735,
+      "loss": 3.0769,
+      "step": 16779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48418980836868286,
+      "learning_rate": 0.0004954673226262708,
+      "loss": 3.1852,
+      "step": 16780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.540973424911499,
+      "learning_rate": 0.0004954556428335449,
+      "loss": 3.1554,
+      "step": 16781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5249459147453308,
+      "learning_rate": 0.0004954439625260264,
+      "loss": 3.1507,
+      "step": 16782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.562673807144165,
+      "learning_rate": 0.000495432281703746,
+      "loss": 3.2198,
+      "step": 16783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49689170718193054,
+      "learning_rate": 0.0004954206003667347,
+      "loss": 3.1578,
+      "step": 16784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5022982954978943,
+      "learning_rate": 0.0004954089185150233,
+      "loss": 3.2703,
+      "step": 16785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5085593461990356,
+      "learning_rate": 0.0004953972361486423,
+      "loss": 3.3836,
+      "step": 16786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5071133375167847,
+      "learning_rate": 0.0004953855532676225,
+      "loss": 3.1033,
+      "step": 16787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5436131358146667,
+      "learning_rate": 0.0004953738698719948,
+      "loss": 3.074,
+      "step": 16788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5148676037788391,
+      "learning_rate": 0.0004953621859617899,
+      "loss": 3.1817,
+      "step": 16789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4939463138580322,
+      "learning_rate": 0.0004953505015370387,
+      "loss": 3.221,
+      "step": 16790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49772271513938904,
+      "learning_rate": 0.0004953388165977717,
+      "loss": 3.0617,
+      "step": 16791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5314618945121765,
+      "learning_rate": 0.00049532713114402,
+      "loss": 3.2513,
+      "step": 16792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5048306584358215,
+      "learning_rate": 0.0004953154451758143,
+      "loss": 3.2086,
+      "step": 16793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5072234272956848,
+      "learning_rate": 0.0004953037586931851,
+      "loss": 3.1794,
+      "step": 16794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5115417838096619,
+      "learning_rate": 0.0004952920716961635,
+      "loss": 3.2774,
+      "step": 16795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5184221267700195,
+      "learning_rate": 0.0004952803841847802,
+      "loss": 3.1443,
+      "step": 16796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5140826106071472,
+      "learning_rate": 0.0004952686961590658,
+      "loss": 3.1012,
+      "step": 16797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5001091957092285,
+      "learning_rate": 0.0004952570076190514,
+      "loss": 3.2363,
+      "step": 16798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5328136682510376,
+      "learning_rate": 0.0004952453185647674,
+      "loss": 3.1932,
+      "step": 16799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5111890435218811,
+      "learning_rate": 0.000495233628996245,
+      "loss": 3.1625,
+      "step": 16800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5230638980865479,
+      "learning_rate": 0.0004952219389135149,
+      "loss": 3.1986,
+      "step": 16801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5159973502159119,
+      "learning_rate": 0.0004952102483166076,
+      "loss": 3.3281,
+      "step": 16802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5228443145751953,
+      "learning_rate": 0.000495198557205554,
+      "loss": 3.2278,
+      "step": 16803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5011665225028992,
+      "learning_rate": 0.0004951868655803851,
+      "loss": 2.9619,
+      "step": 16804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5224314332008362,
+      "learning_rate": 0.0004951751734411316,
+      "loss": 3.0738,
+      "step": 16805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5976347327232361,
+      "learning_rate": 0.0004951634807878241,
+      "loss": 3.1095,
+      "step": 16806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5195080041885376,
+      "learning_rate": 0.0004951517876204938,
+      "loss": 2.9196,
+      "step": 16807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5941303968429565,
+      "learning_rate": 0.000495140093939171,
+      "loss": 2.9227,
+      "step": 16808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48703116178512573,
+      "learning_rate": 0.0004951283997438869,
+      "loss": 3.1429,
+      "step": 16809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4839284420013428,
+      "learning_rate": 0.0004951167050346721,
+      "loss": 3.3139,
+      "step": 16810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5387030243873596,
+      "learning_rate": 0.0004951050098115574,
+      "loss": 3.2789,
+      "step": 16811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5388344526290894,
+      "learning_rate": 0.0004950933140745737,
+      "loss": 3.1641,
+      "step": 16812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5535876154899597,
+      "learning_rate": 0.0004950816178237518,
+      "loss": 3.4473,
+      "step": 16813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49097132682800293,
+      "learning_rate": 0.0004950699210591223,
+      "loss": 3.133,
+      "step": 16814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5270285606384277,
+      "learning_rate": 0.0004950582237807163,
+      "loss": 3.2307,
+      "step": 16815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48164236545562744,
+      "learning_rate": 0.0004950465259885645,
+      "loss": 3.1475,
+      "step": 16816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.531959593296051,
+      "learning_rate": 0.0004950348276826977,
+      "loss": 2.9178,
+      "step": 16817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5036345720291138,
+      "learning_rate": 0.0004950231288631466,
+      "loss": 3.4169,
+      "step": 16818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5073093175888062,
+      "learning_rate": 0.0004950114295299422,
+      "loss": 3.1668,
+      "step": 16819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5157398581504822,
+      "learning_rate": 0.0004949997296831152,
+      "loss": 3.0857,
+      "step": 16820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5297724008560181,
+      "learning_rate": 0.0004949880293226964,
+      "loss": 3.0957,
+      "step": 16821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5831009149551392,
+      "learning_rate": 0.0004949763284487166,
+      "loss": 3.3712,
+      "step": 16822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5394142866134644,
+      "learning_rate": 0.0004949646270612069,
+      "loss": 3.1525,
+      "step": 16823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234691500663757,
+      "learning_rate": 0.0004949529251601977,
+      "loss": 3.28,
+      "step": 16824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5316514372825623,
+      "learning_rate": 0.00049494122274572,
+      "loss": 3.252,
+      "step": 16825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5195567011833191,
+      "learning_rate": 0.0004949295198178048,
+      "loss": 3.2669,
+      "step": 16826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5230311751365662,
+      "learning_rate": 0.0004949178163764827,
+      "loss": 3.2243,
+      "step": 16827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5435062646865845,
+      "learning_rate": 0.0004949061124217845,
+      "loss": 3.031,
+      "step": 16828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5098403096199036,
+      "learning_rate": 0.0004948944079537413,
+      "loss": 3.1558,
+      "step": 16829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5283143520355225,
+      "learning_rate": 0.0004948827029723834,
+      "loss": 3.2979,
+      "step": 16830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5382714867591858,
+      "learning_rate": 0.0004948709974777422,
+      "loss": 3.2907,
+      "step": 16831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5517304539680481,
+      "learning_rate": 0.0004948592914698483,
+      "loss": 2.9451,
+      "step": 16832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5490243434906006,
+      "learning_rate": 0.0004948475849487325,
+      "loss": 2.9241,
+      "step": 16833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5387569665908813,
+      "learning_rate": 0.0004948358779144256,
+      "loss": 3.127,
+      "step": 16834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48919016122817993,
+      "learning_rate": 0.0004948241703669585,
+      "loss": 3.278,
+      "step": 16835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5328850746154785,
+      "learning_rate": 0.0004948124623063621,
+      "loss": 3.1679,
+      "step": 16836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5232701897621155,
+      "learning_rate": 0.0004948007537326672,
+      "loss": 3.0376,
+      "step": 16837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.49141281843185425,
+      "learning_rate": 0.0004947890446459046,
+      "loss": 3.1381,
+      "step": 16838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.508006751537323,
+      "learning_rate": 0.0004947773350461051,
+      "loss": 3.0339,
+      "step": 16839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5270553827285767,
+      "learning_rate": 0.0004947656249332995,
+      "loss": 3.0294,
+      "step": 16840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5074304342269897,
+      "learning_rate": 0.0004947539143075188,
+      "loss": 3.3718,
+      "step": 16841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5027759075164795,
+      "learning_rate": 0.0004947422031687938,
+      "loss": 3.2201,
+      "step": 16842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5462760329246521,
+      "learning_rate": 0.0004947304915171553,
+      "loss": 3.1749,
+      "step": 16843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5136876106262207,
+      "learning_rate": 0.0004947187793526341,
+      "loss": 3.087,
+      "step": 16844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5102071762084961,
+      "learning_rate": 0.0004947070666752612,
+      "loss": 3.119,
+      "step": 16845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5251736044883728,
+      "learning_rate": 0.0004946953534850672,
+      "loss": 3.0766,
+      "step": 16846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5002880692481995,
+      "learning_rate": 0.0004946836397820833,
+      "loss": 3.1961,
+      "step": 16847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5180115103721619,
+      "learning_rate": 0.0004946719255663402,
+      "loss": 3.291,
+      "step": 16848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5141701102256775,
+      "learning_rate": 0.0004946602108378685,
+      "loss": 3.2056,
+      "step": 16849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5428344011306763,
+      "learning_rate": 0.0004946484955966994,
+      "loss": 3.204,
+      "step": 16850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5158218145370483,
+      "learning_rate": 0.0004946367798428636,
+      "loss": 3.2134,
+      "step": 16851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5351789593696594,
+      "learning_rate": 0.0004946250635763919,
+      "loss": 3.0868,
+      "step": 16852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5360104441642761,
+      "learning_rate": 0.0004946133467973153,
+      "loss": 2.9659,
+      "step": 16853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5146509408950806,
+      "learning_rate": 0.0004946016295056646,
+      "loss": 3.2547,
+      "step": 16854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5186895132064819,
+      "learning_rate": 0.0004945899117014706,
+      "loss": 3.1007,
+      "step": 16855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5219805836677551,
+      "learning_rate": 0.0004945781933847644,
+      "loss": 2.9965,
+      "step": 16856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5439258813858032,
+      "learning_rate": 0.0004945664745555766,
+      "loss": 3.1165,
+      "step": 16857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5208057761192322,
+      "learning_rate": 0.0004945547552139382,
+      "loss": 3.1639,
+      "step": 16858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5104433298110962,
+      "learning_rate": 0.0004945430353598799,
+      "loss": 3.1975,
+      "step": 16859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5061667561531067,
+      "learning_rate": 0.0004945313149934327,
+      "loss": 3.0906,
+      "step": 16860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5064367651939392,
+      "learning_rate": 0.0004945195941146275,
+      "loss": 3.2914,
+      "step": 16861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5224226713180542,
+      "learning_rate": 0.0004945078727234951,
+      "loss": 3.2299,
+      "step": 16862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5368633270263672,
+      "learning_rate": 0.0004944961508200664,
+      "loss": 3.2864,
+      "step": 16863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5402868986129761,
+      "learning_rate": 0.0004944844284043723,
+      "loss": 3.0904,
+      "step": 16864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5218072533607483,
+      "learning_rate": 0.0004944727054764436,
+      "loss": 3.2323,
+      "step": 16865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5438421964645386,
+      "learning_rate": 0.0004944609820363112,
+      "loss": 3.0236,
+      "step": 16866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5093616843223572,
+      "learning_rate": 0.0004944492580840061,
+      "loss": 3.2269,
+      "step": 16867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5139485001564026,
+      "learning_rate": 0.000494437533619559,
+      "loss": 3.1943,
+      "step": 16868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5316404700279236,
+      "learning_rate": 0.0004944258086430009,
+      "loss": 3.2521,
+      "step": 16869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5201852321624756,
+      "learning_rate": 0.0004944140831543626,
+      "loss": 3.0447,
+      "step": 16870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5127416849136353,
+      "learning_rate": 0.000494402357153675,
+      "loss": 3.1414,
+      "step": 16871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5286155343055725,
+      "learning_rate": 0.0004943906306409691,
+      "loss": 2.8942,
+      "step": 16872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4988267421722412,
+      "learning_rate": 0.0004943789036162756,
+      "loss": 3.2984,
+      "step": 16873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5161953568458557,
+      "learning_rate": 0.0004943671760796255,
+      "loss": 3.3117,
+      "step": 16874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5332638025283813,
+      "learning_rate": 0.0004943554480310497,
+      "loss": 3.0845,
+      "step": 16875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5356992483139038,
+      "learning_rate": 0.000494343719470579,
+      "loss": 2.9884,
+      "step": 16876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5340079069137573,
+      "learning_rate": 0.0004943319903982444,
+      "loss": 3.1248,
+      "step": 16877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5949048399925232,
+      "learning_rate": 0.0004943202608140767,
+      "loss": 3.074,
+      "step": 16878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4927256107330322,
+      "learning_rate": 0.0004943085307181069,
+      "loss": 3.1862,
+      "step": 16879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5098362565040588,
+      "learning_rate": 0.0004942968001103656,
+      "loss": 3.249,
+      "step": 16880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.559185802936554,
+      "learning_rate": 0.0004942850689908842,
+      "loss": 3.3038,
+      "step": 16881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5128463506698608,
+      "learning_rate": 0.0004942733373596932,
+      "loss": 3.3892,
+      "step": 16882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5219104290008545,
+      "learning_rate": 0.0004942616052168236,
+      "loss": 3.3491,
+      "step": 16883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5195140838623047,
+      "learning_rate": 0.0004942498725623064,
+      "loss": 3.1738,
+      "step": 16884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.634857177734375,
+      "learning_rate": 0.0004942381393961724,
+      "loss": 3.1203,
+      "step": 16885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5101731419563293,
+      "learning_rate": 0.0004942264057184524,
+      "loss": 3.2297,
+      "step": 16886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5076460242271423,
+      "learning_rate": 0.0004942146715291775,
+      "loss": 3.2373,
+      "step": 16887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5291589498519897,
+      "learning_rate": 0.0004942029368283786,
+      "loss": 2.9861,
+      "step": 16888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5351511836051941,
+      "learning_rate": 0.0004941912016160864,
+      "loss": 3.0871,
+      "step": 16889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5280094146728516,
+      "learning_rate": 0.0004941794658923321,
+      "loss": 3.2685,
+      "step": 16890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5539615750312805,
+      "learning_rate": 0.0004941677296571463,
+      "loss": 3.3623,
+      "step": 16891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5121122002601624,
+      "learning_rate": 0.0004941559929105602,
+      "loss": 3.0476,
+      "step": 16892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.551249623298645,
+      "learning_rate": 0.0004941442556526045,
+      "loss": 2.9187,
+      "step": 16893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5004875659942627,
+      "learning_rate": 0.0004941325178833102,
+      "loss": 3.1174,
+      "step": 16894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5540305972099304,
+      "learning_rate": 0.0004941207796027084,
+      "loss": 3.2438,
+      "step": 16895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5179722905158997,
+      "learning_rate": 0.0004941090408108296,
+      "loss": 3.0293,
+      "step": 16896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5183007717132568,
+      "learning_rate": 0.000494097301507705,
+      "loss": 3.0611,
+      "step": 16897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5226312279701233,
+      "learning_rate": 0.0004940855616933654,
+      "loss": 3.2248,
+      "step": 16898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5023093819618225,
+      "learning_rate": 0.0004940738213678419,
+      "loss": 3.1283,
+      "step": 16899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5120278596878052,
+      "learning_rate": 0.0004940620805311654,
+      "loss": 3.4025,
+      "step": 16900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5045427083969116,
+      "learning_rate": 0.0004940503391833664,
+      "loss": 3.2388,
+      "step": 16901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5513672232627869,
+      "learning_rate": 0.0004940385973244765,
+      "loss": 3.2291,
+      "step": 16902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5159168243408203,
+      "learning_rate": 0.0004940268549545261,
+      "loss": 3.0269,
+      "step": 16903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5583893656730652,
+      "learning_rate": 0.0004940151120735463,
+      "loss": 3.2074,
+      "step": 16904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5445465445518494,
+      "learning_rate": 0.000494003368681568,
+      "loss": 3.1999,
+      "step": 16905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5105122327804565,
+      "learning_rate": 0.0004939916247786223,
+      "loss": 3.1474,
+      "step": 16906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5456762313842773,
+      "learning_rate": 0.0004939798803647398,
+      "loss": 3.3013,
+      "step": 16907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5078533887863159,
+      "learning_rate": 0.0004939681354399518,
+      "loss": 3.309,
+      "step": 16908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5101911425590515,
+      "learning_rate": 0.000493956390004289,
+      "loss": 2.935,
+      "step": 16909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5133745074272156,
+      "learning_rate": 0.0004939446440577823,
+      "loss": 3.2223,
+      "step": 16910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5557057857513428,
+      "learning_rate": 0.000493932897600463,
+      "loss": 3.179,
+      "step": 16911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5115715861320496,
+      "learning_rate": 0.0004939211506323615,
+      "loss": 3.0295,
+      "step": 16912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5192151069641113,
+      "learning_rate": 0.0004939094031535091,
+      "loss": 3.2356,
+      "step": 16913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5089118480682373,
+      "learning_rate": 0.0004938976551639368,
+      "loss": 3.5075,
+      "step": 16914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5137461423873901,
+      "learning_rate": 0.0004938859066636751,
+      "loss": 3.2699,
+      "step": 16915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5064414143562317,
+      "learning_rate": 0.0004938741576527555,
+      "loss": 3.0867,
+      "step": 16916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4723142683506012,
+      "learning_rate": 0.0004938624081312085,
+      "loss": 3.1705,
+      "step": 16917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5553394556045532,
+      "learning_rate": 0.0004938506580990654,
+      "loss": 3.4789,
+      "step": 16918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6030188202857971,
+      "learning_rate": 0.0004938389075563568,
+      "loss": 3.1171,
+      "step": 16919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5029276013374329,
+      "learning_rate": 0.0004938271565031139,
+      "loss": 2.9408,
+      "step": 16920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5751969814300537,
+      "learning_rate": 0.0004938154049393676,
+      "loss": 2.8694,
+      "step": 16921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5256320834159851,
+      "learning_rate": 0.0004938036528651488,
+      "loss": 3.1617,
+      "step": 16922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5677057504653931,
+      "learning_rate": 0.0004937919002804885,
+      "loss": 3.2418,
+      "step": 16923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5156370401382446,
+      "learning_rate": 0.0004937801471854176,
+      "loss": 3.0715,
+      "step": 16924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5339087843894958,
+      "learning_rate": 0.000493768393579967,
+      "loss": 3.2977,
+      "step": 16925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5484094619750977,
+      "learning_rate": 0.000493756639464168,
+      "loss": 3.0911,
+      "step": 16926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6116513013839722,
+      "learning_rate": 0.0004937448848380511,
+      "loss": 3.3455,
+      "step": 16927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.503478467464447,
+      "learning_rate": 0.0004937331297016474,
+      "loss": 3.2488,
+      "step": 16928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5339494943618774,
+      "learning_rate": 0.000493721374054988,
+      "loss": 3.0493,
+      "step": 16929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5303932428359985,
+      "learning_rate": 0.0004937096178981038,
+      "loss": 3.1785,
+      "step": 16930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48240697383880615,
+      "learning_rate": 0.0004936978612310257,
+      "loss": 3.32,
+      "step": 16931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5174941420555115,
+      "learning_rate": 0.0004936861040537848,
+      "loss": 3.2424,
+      "step": 16932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5395123362541199,
+      "learning_rate": 0.0004936743463664119,
+      "loss": 3.1445,
+      "step": 16933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5444930791854858,
+      "learning_rate": 0.0004936625881689382,
+      "loss": 3.0873,
+      "step": 16934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4984789490699768,
+      "learning_rate": 0.0004936508294613944,
+      "loss": 3.2639,
+      "step": 16935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5275869965553284,
+      "learning_rate": 0.0004936390702438115,
+      "loss": 3.1615,
+      "step": 16936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5192875266075134,
+      "learning_rate": 0.0004936273105162205,
+      "loss": 3.0108,
+      "step": 16937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5731741189956665,
+      "learning_rate": 0.0004936155502786527,
+      "loss": 3.1182,
+      "step": 16938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48318421840667725,
+      "learning_rate": 0.0004936037895311386,
+      "loss": 3.2587,
+      "step": 16939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.49875524640083313,
+      "learning_rate": 0.0004935920282737095,
+      "loss": 3.2617,
+      "step": 16940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5293728709220886,
+      "learning_rate": 0.0004935802665063962,
+      "loss": 3.3205,
+      "step": 16941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4939599335193634,
+      "learning_rate": 0.0004935685042292297,
+      "loss": 3.3076,
+      "step": 16942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5058974027633667,
+      "learning_rate": 0.0004935567414422411,
+      "loss": 3.1232,
+      "step": 16943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.517693281173706,
+      "learning_rate": 0.0004935449781454612,
+      "loss": 3.3309,
+      "step": 16944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5269827842712402,
+      "learning_rate": 0.0004935332143389212,
+      "loss": 3.2564,
+      "step": 16945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5228087306022644,
+      "learning_rate": 0.0004935214500226518,
+      "loss": 3.284,
+      "step": 16946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5041084885597229,
+      "learning_rate": 0.0004935096851966842,
+      "loss": 3.1787,
+      "step": 16947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5007483959197998,
+      "learning_rate": 0.0004934979198610493,
+      "loss": 3.0264,
+      "step": 16948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5107980370521545,
+      "learning_rate": 0.0004934861540157782,
+      "loss": 3.1986,
+      "step": 16949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5581706166267395,
+      "learning_rate": 0.0004934743876609018,
+      "loss": 3.2615,
+      "step": 16950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.555671215057373,
+      "learning_rate": 0.000493462620796451,
+      "loss": 3.0465,
+      "step": 16951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5258693695068359,
+      "learning_rate": 0.000493450853422457,
+      "loss": 3.0284,
+      "step": 16952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5174445509910583,
+      "learning_rate": 0.0004934390855389506,
+      "loss": 3.0346,
+      "step": 16953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5268626809120178,
+      "learning_rate": 0.000493427317145963,
+      "loss": 2.8945,
+      "step": 16954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.510155439376831,
+      "learning_rate": 0.0004934155482435249,
+      "loss": 3.211,
+      "step": 16955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5453447699546814,
+      "learning_rate": 0.0004934037788316676,
+      "loss": 3.0665,
+      "step": 16956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5393405556678772,
+      "learning_rate": 0.0004933920089104219,
+      "loss": 3.237,
+      "step": 16957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5359326004981995,
+      "learning_rate": 0.000493380238479819,
+      "loss": 3.0703,
+      "step": 16958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5360733270645142,
+      "learning_rate": 0.0004933684675398896,
+      "loss": 3.2203,
+      "step": 16959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5361143946647644,
+      "learning_rate": 0.000493356696090665,
+      "loss": 3.1254,
+      "step": 16960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5372664332389832,
+      "learning_rate": 0.000493344924132176,
+      "loss": 3.2338,
+      "step": 16961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.49514126777648926,
+      "learning_rate": 0.0004933331516644537,
+      "loss": 3.2379,
+      "step": 16962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5488557815551758,
+      "learning_rate": 0.0004933213786875289,
+      "loss": 3.0711,
+      "step": 16963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.533857524394989,
+      "learning_rate": 0.0004933096052014331,
+      "loss": 3.0261,
+      "step": 16964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5344963669776917,
+      "learning_rate": 0.0004932978312061969,
+      "loss": 3.234,
+      "step": 16965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5355231761932373,
+      "learning_rate": 0.0004932860567018513,
+      "loss": 3.2416,
+      "step": 16966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5180875062942505,
+      "learning_rate": 0.0004932742816884276,
+      "loss": 3.1729,
+      "step": 16967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5027824640274048,
+      "learning_rate": 0.0004932625061659564,
+      "loss": 3.1361,
+      "step": 16968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5158348083496094,
+      "learning_rate": 0.0004932507301344693,
+      "loss": 3.0643,
+      "step": 16969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5437960028648376,
+      "learning_rate": 0.0004932389535939966,
+      "loss": 3.1696,
+      "step": 16970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5309322476387024,
+      "learning_rate": 0.0004932271765445699,
+      "loss": 3.168,
+      "step": 16971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5301249027252197,
+      "learning_rate": 0.00049321539898622,
+      "loss": 2.9829,
+      "step": 16972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5049616098403931,
+      "learning_rate": 0.0004932036209189778,
+      "loss": 3.0103,
+      "step": 16973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5803033709526062,
+      "learning_rate": 0.0004931918423428746,
+      "loss": 3.1533,
+      "step": 16974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.519544243812561,
+      "learning_rate": 0.0004931800632579412,
+      "loss": 3.3152,
+      "step": 16975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5230675935745239,
+      "learning_rate": 0.0004931682836642088,
+      "loss": 3.3468,
+      "step": 16976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5146074295043945,
+      "learning_rate": 0.0004931565035617081,
+      "loss": 3.1737,
+      "step": 16977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5033275485038757,
+      "learning_rate": 0.0004931447229504705,
+      "loss": 3.1468,
+      "step": 16978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5531938076019287,
+      "learning_rate": 0.0004931329418305267,
+      "loss": 3.1175,
+      "step": 16979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5533403754234314,
+      "learning_rate": 0.0004931211602019082,
+      "loss": 3.017,
+      "step": 16980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5449845194816589,
+      "learning_rate": 0.0004931093780646455,
+      "loss": 3.333,
+      "step": 16981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4969363808631897,
+      "learning_rate": 0.0004930975954187699,
+      "loss": 3.1212,
+      "step": 16982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5029790997505188,
+      "learning_rate": 0.0004930858122643124,
+      "loss": 3.006,
+      "step": 16983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47780704498291016,
+      "learning_rate": 0.0004930740286013041,
+      "loss": 3.2491,
+      "step": 16984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.544589102268219,
+      "learning_rate": 0.0004930622444297758,
+      "loss": 3.2378,
+      "step": 16985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5521456003189087,
+      "learning_rate": 0.0004930504597497589,
+      "loss": 3.331,
+      "step": 16986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5297691822052002,
+      "learning_rate": 0.0004930386745612841,
+      "loss": 3.2069,
+      "step": 16987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5114834308624268,
+      "learning_rate": 0.0004930268888643827,
+      "loss": 3.2844,
+      "step": 16988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5340949296951294,
+      "learning_rate": 0.0004930151026590855,
+      "loss": 2.9535,
+      "step": 16989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5658579468727112,
+      "learning_rate": 0.0004930033159454237,
+      "loss": 3.1912,
+      "step": 16990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5223037004470825,
+      "learning_rate": 0.0004929915287234283,
+      "loss": 3.1753,
+      "step": 16991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5227300524711609,
+      "learning_rate": 0.0004929797409931305,
+      "loss": 3.1405,
+      "step": 16992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5639767646789551,
+      "learning_rate": 0.0004929679527545611,
+      "loss": 3.0427,
+      "step": 16993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5049862265586853,
+      "learning_rate": 0.0004929561640077512,
+      "loss": 3.2142,
+      "step": 16994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5709805488586426,
+      "learning_rate": 0.0004929443747527319,
+      "loss": 3.0873,
+      "step": 16995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5158815383911133,
+      "learning_rate": 0.0004929325849895344,
+      "loss": 3.2739,
+      "step": 16996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5977637767791748,
+      "learning_rate": 0.0004929207947181895,
+      "loss": 3.2,
+      "step": 16997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5404507517814636,
+      "learning_rate": 0.0004929090039387283,
+      "loss": 3.1116,
+      "step": 16998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5217320919036865,
+      "learning_rate": 0.0004928972126511819,
+      "loss": 3.1959,
+      "step": 16999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5257030129432678,
+      "learning_rate": 0.0004928854208555815,
+      "loss": 3.105,
+      "step": 17000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.526759922504425,
+      "learning_rate": 0.0004928736285519579,
+      "loss": 3.2357,
+      "step": 17001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5135772824287415,
+      "learning_rate": 0.0004928618357403424,
+      "loss": 3.1463,
+      "step": 17002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5537254810333252,
+      "learning_rate": 0.0004928500424207658,
+      "loss": 3.2612,
+      "step": 17003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5435295104980469,
+      "learning_rate": 0.0004928382485932594,
+      "loss": 2.9233,
+      "step": 17004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5425717830657959,
+      "learning_rate": 0.0004928264542578541,
+      "loss": 3.1752,
+      "step": 17005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5152084827423096,
+      "learning_rate": 0.000492814659414581,
+      "loss": 3.3909,
+      "step": 17006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5382000207901001,
+      "learning_rate": 0.0004928028640634714,
+      "loss": 3.092,
+      "step": 17007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5504263043403625,
+      "learning_rate": 0.000492791068204556,
+      "loss": 3.1998,
+      "step": 17008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5138139128684998,
+      "learning_rate": 0.0004927792718378661,
+      "loss": 3.2068,
+      "step": 17009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5308842062950134,
+      "learning_rate": 0.0004927674749634326,
+      "loss": 3.3046,
+      "step": 17010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5355445146560669,
+      "learning_rate": 0.0004927556775812866,
+      "loss": 3.2819,
+      "step": 17011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5167343020439148,
+      "learning_rate": 0.0004927438796914595,
+      "loss": 3.0078,
+      "step": 17012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5470657348632812,
+      "learning_rate": 0.000492732081293982,
+      "loss": 3.1608,
+      "step": 17013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5836805701255798,
+      "learning_rate": 0.0004927202823888853,
+      "loss": 3.1161,
+      "step": 17014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.558231770992279,
+      "learning_rate": 0.0004927084829762004,
+      "loss": 3.1451,
+      "step": 17015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9510738849639893,
+      "learning_rate": 0.0004926966830559585,
+      "loss": 3.305,
+      "step": 17016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5469083786010742,
+      "learning_rate": 0.0004926848826281907,
+      "loss": 3.3076,
+      "step": 17017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.530071496963501,
+      "learning_rate": 0.0004926730816929277,
+      "loss": 3.2547,
+      "step": 17018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5168865919113159,
+      "learning_rate": 0.0004926612802502011,
+      "loss": 3.1445,
+      "step": 17019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5375450849533081,
+      "learning_rate": 0.0004926494783000418,
+      "loss": 3.1627,
+      "step": 17020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5115398168563843,
+      "learning_rate": 0.0004926376758424808,
+      "loss": 3.2687,
+      "step": 17021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.571524441242218,
+      "learning_rate": 0.0004926258728775492,
+      "loss": 3.0509,
+      "step": 17022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5220298767089844,
+      "learning_rate": 0.0004926140694052782,
+      "loss": 3.0722,
+      "step": 17023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5590171217918396,
+      "learning_rate": 0.0004926022654256989,
+      "loss": 3.2197,
+      "step": 17024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5292062163352966,
+      "learning_rate": 0.0004925904609388421,
+      "loss": 3.2796,
+      "step": 17025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4746219515800476,
+      "learning_rate": 0.0004925786559447391,
+      "loss": 3.2434,
+      "step": 17026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5537968277931213,
+      "learning_rate": 0.0004925668504434211,
+      "loss": 3.0872,
+      "step": 17027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5214465260505676,
+      "learning_rate": 0.0004925550444349191,
+      "loss": 3.0574,
+      "step": 17028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.532576858997345,
+      "learning_rate": 0.0004925432379192641,
+      "loss": 3.3728,
+      "step": 17029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5130587816238403,
+      "learning_rate": 0.0004925314308964872,
+      "loss": 3.3297,
+      "step": 17030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5362662076950073,
+      "learning_rate": 0.0004925196233666196,
+      "loss": 3.3526,
+      "step": 17031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.555117666721344,
+      "learning_rate": 0.0004925078153296924,
+      "loss": 3.2804,
+      "step": 17032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5339804291725159,
+      "learning_rate": 0.0004924960067857367,
+      "loss": 3.2155,
+      "step": 17033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5521584749221802,
+      "learning_rate": 0.0004924841977347835,
+      "loss": 3.2564,
+      "step": 17034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6111563444137573,
+      "learning_rate": 0.0004924723881768639,
+      "loss": 2.9852,
+      "step": 17035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5429837107658386,
+      "learning_rate": 0.0004924605781120092,
+      "loss": 3.0712,
+      "step": 17036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5211032629013062,
+      "learning_rate": 0.0004924487675402504,
+      "loss": 3.0123,
+      "step": 17037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5365507006645203,
+      "learning_rate": 0.0004924369564616185,
+      "loss": 3.0002,
+      "step": 17038
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5042828321456909,
+      "learning_rate": 0.0004924251448761446,
+      "loss": 3.0771,
+      "step": 17039
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5612789988517761,
+      "learning_rate": 0.0004924133327838601,
+      "loss": 3.0335,
+      "step": 17040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5127493739128113,
+      "learning_rate": 0.0004924015201847958,
+      "loss": 3.1882,
+      "step": 17041
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5296911001205444,
+      "learning_rate": 0.000492389707078983,
+      "loss": 3.143,
+      "step": 17042
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5228750109672546,
+      "learning_rate": 0.0004923778934664526,
+      "loss": 3.1002,
+      "step": 17043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5364986658096313,
+      "learning_rate": 0.000492366079347236,
+      "loss": 3.0465,
+      "step": 17044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49486231803894043,
+      "learning_rate": 0.000492354264721364,
+      "loss": 3.1087,
+      "step": 17045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5611582398414612,
+      "learning_rate": 0.0004923424495888681,
+      "loss": 3.0845,
+      "step": 17046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6187002658843994,
+      "learning_rate": 0.0004923306339497791,
+      "loss": 3.0551,
+      "step": 17047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5191326141357422,
+      "learning_rate": 0.0004923188178041282,
+      "loss": 3.0557,
+      "step": 17048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5312925577163696,
+      "learning_rate": 0.0004923070011519466,
+      "loss": 3.4865,
+      "step": 17049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5266924500465393,
+      "learning_rate": 0.0004922951839932653,
+      "loss": 3.2505,
+      "step": 17050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5350109934806824,
+      "learning_rate": 0.0004922833663281156,
+      "loss": 3.3707,
+      "step": 17051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4987470507621765,
+      "learning_rate": 0.0004922715481565284,
+      "loss": 3.332,
+      "step": 17052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49026739597320557,
+      "learning_rate": 0.000492259729478535,
+      "loss": 3.1229,
+      "step": 17053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5140583515167236,
+      "learning_rate": 0.0004922479102941665,
+      "loss": 3.2348,
+      "step": 17054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5144451260566711,
+      "learning_rate": 0.0004922360906034538,
+      "loss": 3.0013,
+      "step": 17055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5055959820747375,
+      "learning_rate": 0.0004922242704064285,
+      "loss": 3.2636,
+      "step": 17056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5427704453468323,
+      "learning_rate": 0.0004922124497031214,
+      "loss": 3.36,
+      "step": 17057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5104376673698425,
+      "learning_rate": 0.0004922006284935637,
+      "loss": 3.3998,
+      "step": 17058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5442894697189331,
+      "learning_rate": 0.0004921888067777865,
+      "loss": 3.113,
+      "step": 17059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5200850963592529,
+      "learning_rate": 0.000492176984555821,
+      "loss": 3.0532,
+      "step": 17060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4936150312423706,
+      "learning_rate": 0.0004921651618276982,
+      "loss": 3.0992,
+      "step": 17061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49831634759902954,
+      "learning_rate": 0.0004921533385934495,
+      "loss": 3.2988,
+      "step": 17062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5185843706130981,
+      "learning_rate": 0.0004921415148531058,
+      "loss": 3.1793,
+      "step": 17063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5367192029953003,
+      "learning_rate": 0.0004921296906066984,
+      "loss": 3.2338,
+      "step": 17064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5617893934249878,
+      "learning_rate": 0.0004921178658542582,
+      "loss": 3.0399,
+      "step": 17065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4962318241596222,
+      "learning_rate": 0.0004921060405958167,
+      "loss": 3.1821,
+      "step": 17066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5281431674957275,
+      "learning_rate": 0.0004920942148314047,
+      "loss": 3.2091,
+      "step": 17067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5583265423774719,
+      "learning_rate": 0.0004920823885610537,
+      "loss": 3.1087,
+      "step": 17068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5328853726387024,
+      "learning_rate": 0.0004920705617847945,
+      "loss": 3.2278,
+      "step": 17069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5027572512626648,
+      "learning_rate": 0.0004920587345026585,
+      "loss": 3.1542,
+      "step": 17070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6189442873001099,
+      "learning_rate": 0.0004920469067146768,
+      "loss": 3.259,
+      "step": 17071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5087341666221619,
+      "learning_rate": 0.0004920350784208803,
+      "loss": 2.9372,
+      "step": 17072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5113468170166016,
+      "learning_rate": 0.0004920232496213005,
+      "loss": 3.241,
+      "step": 17073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5053597688674927,
+      "learning_rate": 0.0004920114203159683,
+      "loss": 3.2242,
+      "step": 17074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5626222491264343,
+      "learning_rate": 0.0004919995905049153,
+      "loss": 3.0204,
+      "step": 17075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5104367136955261,
+      "learning_rate": 0.000491987760188172,
+      "loss": 3.0962,
+      "step": 17076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.507770299911499,
+      "learning_rate": 0.00049197592936577,
+      "loss": 3.3188,
+      "step": 17077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5115891695022583,
+      "learning_rate": 0.0004919640980377404,
+      "loss": 2.9712,
+      "step": 17078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.488614946603775,
+      "learning_rate": 0.0004919522662041141,
+      "loss": 3.3266,
+      "step": 17079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47894060611724854,
+      "learning_rate": 0.0004919404338649227,
+      "loss": 3.1626,
+      "step": 17080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5004280805587769,
+      "learning_rate": 0.000491928601020197,
+      "loss": 3.016,
+      "step": 17081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5228533148765564,
+      "learning_rate": 0.0004919167676699684,
+      "loss": 3.4334,
+      "step": 17082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5518790483474731,
+      "learning_rate": 0.0004919049338142679,
+      "loss": 3.1275,
+      "step": 17083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49410319328308105,
+      "learning_rate": 0.0004918930994531268,
+      "loss": 2.9918,
+      "step": 17084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4980878233909607,
+      "learning_rate": 0.0004918812645865762,
+      "loss": 2.9434,
+      "step": 17085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5068234801292419,
+      "learning_rate": 0.0004918694292146473,
+      "loss": 3.3573,
+      "step": 17086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5471787452697754,
+      "learning_rate": 0.000491857593337371,
+      "loss": 3.1763,
+      "step": 17087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.519198477268219,
+      "learning_rate": 0.000491845756954779,
+      "loss": 3.2118,
+      "step": 17088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5100758075714111,
+      "learning_rate": 0.0004918339200669021,
+      "loss": 3.3549,
+      "step": 17089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.525219202041626,
+      "learning_rate": 0.0004918220826737717,
+      "loss": 3.1919,
+      "step": 17090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5719212889671326,
+      "learning_rate": 0.0004918102447754187,
+      "loss": 3.0999,
+      "step": 17091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5068230032920837,
+      "learning_rate": 0.0004917984063718745,
+      "loss": 3.3169,
+      "step": 17092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5205427408218384,
+      "learning_rate": 0.0004917865674631702,
+      "loss": 3.2123,
+      "step": 17093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5136102437973022,
+      "learning_rate": 0.0004917747280493369,
+      "loss": 3.4344,
+      "step": 17094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5310423374176025,
+      "learning_rate": 0.0004917628881304061,
+      "loss": 3.081,
+      "step": 17095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5061217546463013,
+      "learning_rate": 0.0004917510477064085,
+      "loss": 2.9725,
+      "step": 17096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5577815175056458,
+      "learning_rate": 0.0004917392067773756,
+      "loss": 3.2315,
+      "step": 17097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5072141289710999,
+      "learning_rate": 0.0004917273653433388,
+      "loss": 3.2273,
+      "step": 17098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5190002918243408,
+      "learning_rate": 0.0004917155234043287,
+      "loss": 3.1358,
+      "step": 17099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49393171072006226,
+      "learning_rate": 0.000491703680960377,
+      "loss": 3.1898,
+      "step": 17100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49661365151405334,
+      "learning_rate": 0.0004916918380115145,
+      "loss": 3.1599,
+      "step": 17101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.524917721748352,
+      "learning_rate": 0.0004916799945577727,
+      "loss": 3.2517,
+      "step": 17102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5378203988075256,
+      "learning_rate": 0.0004916681505991827,
+      "loss": 3.2473,
+      "step": 17103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5326823592185974,
+      "learning_rate": 0.0004916563061357756,
+      "loss": 3.1216,
+      "step": 17104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5417553782463074,
+      "learning_rate": 0.0004916444611675828,
+      "loss": 3.2,
+      "step": 17105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5240073204040527,
+      "learning_rate": 0.0004916326156946353,
+      "loss": 3.0368,
+      "step": 17106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5561971664428711,
+      "learning_rate": 0.0004916207697169643,
+      "loss": 3.2357,
+      "step": 17107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5211239457130432,
+      "learning_rate": 0.0004916089232346012,
+      "loss": 3.2528,
+      "step": 17108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4768798053264618,
+      "learning_rate": 0.000491597076247577,
+      "loss": 3.3263,
+      "step": 17109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4900779128074646,
+      "learning_rate": 0.0004915852287559231,
+      "loss": 3.2368,
+      "step": 17110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5443329811096191,
+      "learning_rate": 0.0004915733807596705,
+      "loss": 2.9507,
+      "step": 17111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5072310566902161,
+      "learning_rate": 0.0004915615322588503,
+      "loss": 3.0643,
+      "step": 17112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5106002688407898,
+      "learning_rate": 0.0004915496832534941,
+      "loss": 3.0003,
+      "step": 17113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5489236116409302,
+      "learning_rate": 0.0004915378337436328,
+      "loss": 3.0268,
+      "step": 17114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48166120052337646,
+      "learning_rate": 0.0004915259837292978,
+      "loss": 3.3047,
+      "step": 17115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5533964037895203,
+      "learning_rate": 0.0004915141332105203,
+      "loss": 3.2744,
+      "step": 17116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5538474321365356,
+      "learning_rate": 0.0004915022821873311,
+      "loss": 3.2312,
+      "step": 17117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5147866010665894,
+      "learning_rate": 0.0004914904306597621,
+      "loss": 2.9823,
+      "step": 17118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5750763416290283,
+      "learning_rate": 0.0004914785786278439,
+      "loss": 3.129,
+      "step": 17119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5425081253051758,
+      "learning_rate": 0.0004914667260916081,
+      "loss": 3.2427,
+      "step": 17120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5620341897010803,
+      "learning_rate": 0.0004914548730510859,
+      "loss": 3.1747,
+      "step": 17121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5213549137115479,
+      "learning_rate": 0.0004914430195063083,
+      "loss": 3.005,
+      "step": 17122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5401297807693481,
+      "learning_rate": 0.0004914311654573066,
+      "loss": 3.2856,
+      "step": 17123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5165397524833679,
+      "learning_rate": 0.000491419310904112,
+      "loss": 3.1327,
+      "step": 17124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5052899718284607,
+      "learning_rate": 0.0004914074558467559,
+      "loss": 3.3247,
+      "step": 17125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5615530014038086,
+      "learning_rate": 0.0004913956002852693,
+      "loss": 3.2895,
+      "step": 17126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5121963024139404,
+      "learning_rate": 0.0004913837442196836,
+      "loss": 3.0851,
+      "step": 17127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5730823278427124,
+      "learning_rate": 0.0004913718876500299,
+      "loss": 3.1933,
+      "step": 17128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5356817245483398,
+      "learning_rate": 0.0004913600305763396,
+      "loss": 3.3342,
+      "step": 17129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5603740811347961,
+      "learning_rate": 0.0004913481729986436,
+      "loss": 3.223,
+      "step": 17130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6096175312995911,
+      "learning_rate": 0.0004913363149169735,
+      "loss": 3.1724,
+      "step": 17131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5367498397827148,
+      "learning_rate": 0.0004913244563313602,
+      "loss": 3.297,
+      "step": 17132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4697509706020355,
+      "learning_rate": 0.0004913125972418352,
+      "loss": 3.3864,
+      "step": 17133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5341188907623291,
+      "learning_rate": 0.0004913007376484297,
+      "loss": 3.1566,
+      "step": 17134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5422013401985168,
+      "learning_rate": 0.0004912888775511748,
+      "loss": 2.9892,
+      "step": 17135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5414844155311584,
+      "learning_rate": 0.0004912770169501018,
+      "loss": 3.0971,
+      "step": 17136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5191177129745483,
+      "learning_rate": 0.000491265155845242,
+      "loss": 3.2724,
+      "step": 17137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.541873037815094,
+      "learning_rate": 0.0004912532942366266,
+      "loss": 3.1036,
+      "step": 17138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5198567509651184,
+      "learning_rate": 0.0004912414321242869,
+      "loss": 3.0269,
+      "step": 17139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5153297781944275,
+      "learning_rate": 0.0004912295695082539,
+      "loss": 2.9057,
+      "step": 17140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5370156764984131,
+      "learning_rate": 0.0004912177063885591,
+      "loss": 3.333,
+      "step": 17141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5156940817832947,
+      "learning_rate": 0.0004912058427652337,
+      "loss": 3.1685,
+      "step": 17142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.558294415473938,
+      "learning_rate": 0.0004911939786383089,
+      "loss": 3.0098,
+      "step": 17143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5128892064094543,
+      "learning_rate": 0.0004911821140078158,
+      "loss": 3.2013,
+      "step": 17144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5227497816085815,
+      "learning_rate": 0.0004911702488737859,
+      "loss": 3.3653,
+      "step": 17145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5522597432136536,
+      "learning_rate": 0.0004911583832362504,
+      "loss": 3.0272,
+      "step": 17146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5611497759819031,
+      "learning_rate": 0.0004911465170952405,
+      "loss": 3.3738,
+      "step": 17147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5275722742080688,
+      "learning_rate": 0.0004911346504507874,
+      "loss": 3.2767,
+      "step": 17148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5308437347412109,
+      "learning_rate": 0.0004911227833029225,
+      "loss": 3.2512,
+      "step": 17149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5144796371459961,
+      "learning_rate": 0.000491110915651677,
+      "loss": 3.2331,
+      "step": 17150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5319437384605408,
+      "learning_rate": 0.0004910990474970821,
+      "loss": 3.0634,
+      "step": 17151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5378063917160034,
+      "learning_rate": 0.000491087178839169,
+      "loss": 3.1613,
+      "step": 17152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4878802001476288,
+      "learning_rate": 0.000491075309677969,
+      "loss": 3.1845,
+      "step": 17153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5802315473556519,
+      "learning_rate": 0.0004910634400135136,
+      "loss": 3.2188,
+      "step": 17154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5172027945518494,
+      "learning_rate": 0.0004910515698458338,
+      "loss": 3.0371,
+      "step": 17155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5362008213996887,
+      "learning_rate": 0.0004910396991749608,
+      "loss": 3.1302,
+      "step": 17156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.50388503074646,
+      "learning_rate": 0.0004910278280009263,
+      "loss": 3.1382,
+      "step": 17157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5768753886222839,
+      "learning_rate": 0.0004910159563237609,
+      "loss": 3.0569,
+      "step": 17158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5347177386283875,
+      "learning_rate": 0.0004910040841434964,
+      "loss": 3.1319,
+      "step": 17159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.531885027885437,
+      "learning_rate": 0.0004909922114601639,
+      "loss": 3.4059,
+      "step": 17160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703318119049072,
+      "learning_rate": 0.0004909803382737947,
+      "loss": 3.1653,
+      "step": 17161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5395308136940002,
+      "learning_rate": 0.0004909684645844201,
+      "loss": 3.2048,
+      "step": 17162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5464442372322083,
+      "learning_rate": 0.0004909565903920711,
+      "loss": 2.963,
+      "step": 17163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5010871887207031,
+      "learning_rate": 0.0004909447156967794,
+      "loss": 3.3017,
+      "step": 17164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5098520517349243,
+      "learning_rate": 0.000490932840498576,
+      "loss": 3.076,
+      "step": 17165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6024291515350342,
+      "learning_rate": 0.0004909209647974923,
+      "loss": 3.2139,
+      "step": 17166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.521047830581665,
+      "learning_rate": 0.0004909090885935594,
+      "loss": 3.1004,
+      "step": 17167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.534773051738739,
+      "learning_rate": 0.0004908972118868088,
+      "loss": 3.2295,
+      "step": 17168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5155094265937805,
+      "learning_rate": 0.0004908853346772716,
+      "loss": 3.1153,
+      "step": 17169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.513957142829895,
+      "learning_rate": 0.0004908734569649793,
+      "loss": 3.0931,
+      "step": 17170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5340112447738647,
+      "learning_rate": 0.000490861578749963,
+      "loss": 3.0543,
+      "step": 17171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5084630250930786,
+      "learning_rate": 0.000490849700032254,
+      "loss": 3.323,
+      "step": 17172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5156233310699463,
+      "learning_rate": 0.0004908378208118836,
+      "loss": 3.2439,
+      "step": 17173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5381580591201782,
+      "learning_rate": 0.0004908259410888831,
+      "loss": 3.0843,
+      "step": 17174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5627414584159851,
+      "learning_rate": 0.0004908140608632838,
+      "loss": 2.9431,
+      "step": 17175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5171828866004944,
+      "learning_rate": 0.000490802180135117,
+      "loss": 3.2593,
+      "step": 17176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5454464554786682,
+      "learning_rate": 0.000490790298904414,
+      "loss": 3.1044,
+      "step": 17177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4904737174510956,
+      "learning_rate": 0.0004907784171712061,
+      "loss": 3.242,
+      "step": 17178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5514205694198608,
+      "learning_rate": 0.0004907665349355245,
+      "loss": 3.2227,
+      "step": 17179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49019309878349304,
+      "learning_rate": 0.0004907546521974006,
+      "loss": 3.1571,
+      "step": 17180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.552192747592926,
+      "learning_rate": 0.0004907427689568656,
+      "loss": 3.09,
+      "step": 17181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5478540658950806,
+      "learning_rate": 0.0004907308852139508,
+      "loss": 3.1038,
+      "step": 17182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5257436633110046,
+      "learning_rate": 0.0004907190009686878,
+      "loss": 3.4212,
+      "step": 17183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5158893465995789,
+      "learning_rate": 0.0004907071162211074,
+      "loss": 3.288,
+      "step": 17184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4981338083744049,
+      "learning_rate": 0.0004906952309712413,
+      "loss": 3.2846,
+      "step": 17185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5687278509140015,
+      "learning_rate": 0.0004906833452191207,
+      "loss": 3.3939,
+      "step": 17186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5252379179000854,
+      "learning_rate": 0.0004906714589647767,
+      "loss": 3.3868,
+      "step": 17187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5520195364952087,
+      "learning_rate": 0.0004906595722082409,
+      "loss": 3.2195,
+      "step": 17188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5275287628173828,
+      "learning_rate": 0.0004906476849495444,
+      "loss": 3.3285,
+      "step": 17189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5144610404968262,
+      "learning_rate": 0.0004906357971887187,
+      "loss": 3.1138,
+      "step": 17190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5572524070739746,
+      "learning_rate": 0.0004906239089257949,
+      "loss": 3.0022,
+      "step": 17191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5245261192321777,
+      "learning_rate": 0.0004906120201608044,
+      "loss": 3.1331,
+      "step": 17192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.506518542766571,
+      "learning_rate": 0.0004906001308937787,
+      "loss": 3.3921,
+      "step": 17193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4986775517463684,
+      "learning_rate": 0.0004905882411247487,
+      "loss": 3.1723,
+      "step": 17194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5022253394126892,
+      "learning_rate": 0.0004905763508537461,
+      "loss": 3.1883,
+      "step": 17195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5980327129364014,
+      "learning_rate": 0.000490564460080802,
+      "loss": 3.1204,
+      "step": 17196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5253124833106995,
+      "learning_rate": 0.0004905525688059479,
+      "loss": 3.2769,
+      "step": 17197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5250275135040283,
+      "learning_rate": 0.0004905406770292148,
+      "loss": 3.0089,
+      "step": 17198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5560047626495361,
+      "learning_rate": 0.0004905287847506343,
+      "loss": 3.1203,
+      "step": 17199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5059837102890015,
+      "learning_rate": 0.0004905168919702378,
+      "loss": 3.0903,
+      "step": 17200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.530164361000061,
+      "learning_rate": 0.0004905049986880563,
+      "loss": 3.2861,
+      "step": 17201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5831915736198425,
+      "learning_rate": 0.0004904931049041214,
+      "loss": 3.202,
+      "step": 17202
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546757698059082,
+      "learning_rate": 0.0004904812106184643,
+      "loss": 3.1316,
+      "step": 17203
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49275583028793335,
+      "learning_rate": 0.0004904693158311162,
+      "loss": 3.0704,
+      "step": 17204
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5256943106651306,
+      "learning_rate": 0.0004904574205421089,
+      "loss": 3.0593,
+      "step": 17205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.542165219783783,
+      "learning_rate": 0.0004904455247514731,
+      "loss": 3.391,
+      "step": 17206
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5653906464576721,
+      "learning_rate": 0.0004904336284592407,
+      "loss": 3.2863,
+      "step": 17207
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5180942416191101,
+      "learning_rate": 0.0004904217316654425,
+      "loss": 3.1246,
+      "step": 17208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49898773431777954,
+      "learning_rate": 0.0004904098343701102,
+      "loss": 3.1719,
+      "step": 17209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.531066358089447,
+      "learning_rate": 0.0004903979365732753,
+      "loss": 3.3398,
+      "step": 17210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4966711103916168,
+      "learning_rate": 0.0004903860382749686,
+      "loss": 3.0826,
+      "step": 17211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5284520983695984,
+      "learning_rate": 0.0004903741394752218,
+      "loss": 3.2555,
+      "step": 17212
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5519339442253113,
+      "learning_rate": 0.000490362240174066,
+      "loss": 3.0804,
+      "step": 17213
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5141843557357788,
+      "learning_rate": 0.0004903503403715329,
+      "loss": 3.2965,
+      "step": 17214
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5470436215400696,
+      "learning_rate": 0.0004903384400676535,
+      "loss": 3.0265,
+      "step": 17215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5789403319358826,
+      "learning_rate": 0.0004903265392624594,
+      "loss": 3.0417,
+      "step": 17216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.515338659286499,
+      "learning_rate": 0.0004903146379559818,
+      "loss": 3.0671,
+      "step": 17217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5383449196815491,
+      "learning_rate": 0.000490302736148252,
+      "loss": 3.0563,
+      "step": 17218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5087341666221619,
+      "learning_rate": 0.0004902908338393014,
+      "loss": 3.3356,
+      "step": 17219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5621129870414734,
+      "learning_rate": 0.0004902789310291615,
+      "loss": 3.1718,
+      "step": 17220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49984487891197205,
+      "learning_rate": 0.0004902670277178634,
+      "loss": 3.2502,
+      "step": 17221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6184155941009521,
+      "learning_rate": 0.0004902551239054386,
+      "loss": 3.1425,
+      "step": 17222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5024511218070984,
+      "learning_rate": 0.0004902432195919184,
+      "loss": 3.2221,
+      "step": 17223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5568563938140869,
+      "learning_rate": 0.0004902313147773342,
+      "loss": 2.9693,
+      "step": 17224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5303897857666016,
+      "learning_rate": 0.0004902194094617174,
+      "loss": 3.2374,
+      "step": 17225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5185246467590332,
+      "learning_rate": 0.0004902075036450992,
+      "loss": 3.2034,
+      "step": 17226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5359919667243958,
+      "learning_rate": 0.000490195597327511,
+      "loss": 2.8955,
+      "step": 17227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.530653178691864,
+      "learning_rate": 0.0004901836905089842,
+      "loss": 3.2935,
+      "step": 17228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5164836645126343,
+      "learning_rate": 0.0004901717831895503,
+      "loss": 3.229,
+      "step": 17229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5706772804260254,
+      "learning_rate": 0.0004901598753692403,
+      "loss": 3.1432,
+      "step": 17230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5535086393356323,
+      "learning_rate": 0.0004901479670480859,
+      "loss": 3.0893,
+      "step": 17231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5143651366233826,
+      "learning_rate": 0.0004901360582261185,
+      "loss": 3.1783,
+      "step": 17232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5126096606254578,
+      "learning_rate": 0.0004901241489033692,
+      "loss": 3.2156,
+      "step": 17233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5179157853126526,
+      "learning_rate": 0.0004901122390798694,
+      "loss": 3.2843,
+      "step": 17234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5556790232658386,
+      "learning_rate": 0.0004901003287556507,
+      "loss": 3.0277,
+      "step": 17235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5082927942276001,
+      "learning_rate": 0.0004900884179307441,
+      "loss": 3.1995,
+      "step": 17236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5159091353416443,
+      "learning_rate": 0.0004900765066051814,
+      "loss": 3.0237,
+      "step": 17237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49801966547966003,
+      "learning_rate": 0.0004900645947789938,
+      "loss": 3.155,
+      "step": 17238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5168977379798889,
+      "learning_rate": 0.0004900526824522125,
+      "loss": 3.1495,
+      "step": 17239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5264788269996643,
+      "learning_rate": 0.000490040769624869,
+      "loss": 2.9527,
+      "step": 17240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6261751055717468,
+      "learning_rate": 0.0004900288562969947,
+      "loss": 3.2426,
+      "step": 17241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5559073090553284,
+      "learning_rate": 0.000490016942468621,
+      "loss": 3.0032,
+      "step": 17242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5475751161575317,
+      "learning_rate": 0.0004900050281397792,
+      "loss": 3.396,
+      "step": 17243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5120872259140015,
+      "learning_rate": 0.0004899931133105007,
+      "loss": 3.1183,
+      "step": 17244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5100634694099426,
+      "learning_rate": 0.000489981197980817,
+      "loss": 3.1615,
+      "step": 17245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5490902066230774,
+      "learning_rate": 0.0004899692821507593,
+      "loss": 3.0538,
+      "step": 17246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48027950525283813,
+      "learning_rate": 0.0004899573658203591,
+      "loss": 3.1387,
+      "step": 17247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5834192633628845,
+      "learning_rate": 0.0004899454489896479,
+      "loss": 3.1026,
+      "step": 17248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5443624258041382,
+      "learning_rate": 0.0004899335316586568,
+      "loss": 3.1497,
+      "step": 17249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5452780723571777,
+      "learning_rate": 0.0004899216138274172,
+      "loss": 3.2115,
+      "step": 17250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5227807760238647,
+      "learning_rate": 0.0004899096954959608,
+      "loss": 3.2457,
+      "step": 17251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5504705905914307,
+      "learning_rate": 0.0004898977766643188,
+      "loss": 2.9626,
+      "step": 17252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5524275898933411,
+      "learning_rate": 0.0004898858573325226,
+      "loss": 3.2484,
+      "step": 17253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5162675380706787,
+      "learning_rate": 0.0004898739375006036,
+      "loss": 3.2485,
+      "step": 17254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5104801058769226,
+      "learning_rate": 0.0004898620171685932,
+      "loss": 3.2277,
+      "step": 17255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4923844635486603,
+      "learning_rate": 0.0004898500963365226,
+      "loss": 3.1075,
+      "step": 17256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5191243886947632,
+      "learning_rate": 0.0004898381750044236,
+      "loss": 3.007,
+      "step": 17257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.531520426273346,
+      "learning_rate": 0.0004898262531723273,
+      "loss": 3.1071,
+      "step": 17258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5321716070175171,
+      "learning_rate": 0.0004898143308402652,
+      "loss": 3.2239,
+      "step": 17259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.533822238445282,
+      "learning_rate": 0.0004898024080082688,
+      "loss": 3.0522,
+      "step": 17260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5310263633728027,
+      "learning_rate": 0.0004897904846763692,
+      "loss": 3.0263,
+      "step": 17261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5534727573394775,
+      "learning_rate": 0.000489778560844598,
+      "loss": 3.2554,
+      "step": 17262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5336495041847229,
+      "learning_rate": 0.0004897666365129867,
+      "loss": 3.1274,
+      "step": 17263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5278199911117554,
+      "learning_rate": 0.0004897547116815666,
+      "loss": 3.2348,
+      "step": 17264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5093748569488525,
+      "learning_rate": 0.0004897427863503691,
+      "loss": 3.1455,
+      "step": 17265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.565778374671936,
+      "learning_rate": 0.0004897308605194255,
+      "loss": 3.0988,
+      "step": 17266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5380170941352844,
+      "learning_rate": 0.0004897189341887673,
+      "loss": 3.0954,
+      "step": 17267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4972584843635559,
+      "learning_rate": 0.0004897070073584262,
+      "loss": 3.0722,
+      "step": 17268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.533125638961792,
+      "learning_rate": 0.0004896950800284332,
+      "loss": 3.1618,
+      "step": 17269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49823445081710815,
+      "learning_rate": 0.0004896831521988198,
+      "loss": 3.1106,
+      "step": 17270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5134430527687073,
+      "learning_rate": 0.0004896712238696176,
+      "loss": 2.9883,
+      "step": 17271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5448001027107239,
+      "learning_rate": 0.0004896592950408579,
+      "loss": 3.3028,
+      "step": 17272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5064616799354553,
+      "learning_rate": 0.0004896473657125719,
+      "loss": 3.4008,
+      "step": 17273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5209435820579529,
+      "learning_rate": 0.0004896354358847915,
+      "loss": 3.3331,
+      "step": 17274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5029204487800598,
+      "learning_rate": 0.0004896235055575477,
+      "loss": 3.2065,
+      "step": 17275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5218020677566528,
+      "learning_rate": 0.0004896115747308722,
+      "loss": 3.2772,
+      "step": 17276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5261409878730774,
+      "learning_rate": 0.0004895996434047962,
+      "loss": 3.357,
+      "step": 17277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5557669997215271,
+      "learning_rate": 0.0004895877115793513,
+      "loss": 3.2209,
+      "step": 17278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5135267972946167,
+      "learning_rate": 0.0004895757792545689,
+      "loss": 3.0166,
+      "step": 17279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4932655096054077,
+      "learning_rate": 0.0004895638464304802,
+      "loss": 3.1704,
+      "step": 17280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5083035230636597,
+      "learning_rate": 0.0004895519131071169,
+      "loss": 3.0698,
+      "step": 17281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4993206560611725,
+      "learning_rate": 0.0004895399792845105,
+      "loss": 3.1426,
+      "step": 17282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5377535820007324,
+      "learning_rate": 0.000489528044962692,
+      "loss": 3.2107,
+      "step": 17283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5332820415496826,
+      "learning_rate": 0.0004895161101416932,
+      "loss": 3.1717,
+      "step": 17284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6028470993041992,
+      "learning_rate": 0.0004895041748215456,
+      "loss": 3.222,
+      "step": 17285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5052365064620972,
+      "learning_rate": 0.0004894922390022803,
+      "loss": 3.1013,
+      "step": 17286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5114204287528992,
+      "learning_rate": 0.000489480302683929,
+      "loss": 3.2889,
+      "step": 17287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5291855931282043,
+      "learning_rate": 0.000489468365866523,
+      "loss": 3.0852,
+      "step": 17288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5404118895530701,
+      "learning_rate": 0.0004894564285500938,
+      "loss": 3.1578,
+      "step": 17289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5513167977333069,
+      "learning_rate": 0.0004894444907346729,
+      "loss": 3.177,
+      "step": 17290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546855628490448,
+      "learning_rate": 0.0004894325524202915,
+      "loss": 3.1454,
+      "step": 17291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5344764590263367,
+      "learning_rate": 0.0004894206136069813,
+      "loss": 3.0274,
+      "step": 17292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.531498908996582,
+      "learning_rate": 0.0004894086742947737,
+      "loss": 2.8852,
+      "step": 17293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4873696565628052,
+      "learning_rate": 0.0004893967344837,
+      "loss": 3.1002,
+      "step": 17294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5418617129325867,
+      "learning_rate": 0.0004893847941737919,
+      "loss": 3.2382,
+      "step": 17295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5241795182228088,
+      "learning_rate": 0.0004893728533650806,
+      "loss": 3.2205,
+      "step": 17296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5475279092788696,
+      "learning_rate": 0.0004893609120575976,
+      "loss": 3.2555,
+      "step": 17297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.65479975938797,
+      "learning_rate": 0.0004893489702513745,
+      "loss": 3.0701,
+      "step": 17298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5669695138931274,
+      "learning_rate": 0.0004893370279464427,
+      "loss": 3.1972,
+      "step": 17299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.517507791519165,
+      "learning_rate": 0.0004893250851428335,
+      "loss": 3.1766,
+      "step": 17300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5156394839286804,
+      "learning_rate": 0.0004893131418405786,
+      "loss": 3.2759,
+      "step": 17301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48930031061172485,
+      "learning_rate": 0.0004893011980397091,
+      "loss": 3.1122,
+      "step": 17302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.518436074256897,
+      "learning_rate": 0.0004892892537402568,
+      "loss": 3.1862,
+      "step": 17303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5759012699127197,
+      "learning_rate": 0.0004892773089422531,
+      "loss": 3.1416,
+      "step": 17304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48443731665611267,
+      "learning_rate": 0.0004892653636457293,
+      "loss": 2.9702,
+      "step": 17305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5327646732330322,
+      "learning_rate": 0.000489253417850717,
+      "loss": 3.1856,
+      "step": 17306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5462185144424438,
+      "learning_rate": 0.0004892414715572475,
+      "loss": 3.0689,
+      "step": 17307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.535132110118866,
+      "learning_rate": 0.0004892295247653526,
+      "loss": 3.1475,
+      "step": 17308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5083247423171997,
+      "learning_rate": 0.0004892175774750633,
+      "loss": 3.0753,
+      "step": 17309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5415558815002441,
+      "learning_rate": 0.0004892056296864116,
+      "loss": 3.2183,
+      "step": 17310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5436738729476929,
+      "learning_rate": 0.0004891936813994285,
+      "loss": 3.2266,
+      "step": 17311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5327355265617371,
+      "learning_rate": 0.0004891817326141457,
+      "loss": 3.2321,
+      "step": 17312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5031436681747437,
+      "learning_rate": 0.0004891697833305946,
+      "loss": 3.0962,
+      "step": 17313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5180808305740356,
+      "learning_rate": 0.0004891578335488066,
+      "loss": 3.0242,
+      "step": 17314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5372978448867798,
+      "learning_rate": 0.0004891458832688135,
+      "loss": 2.9658,
+      "step": 17315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5055201053619385,
+      "learning_rate": 0.0004891339324906464,
+      "loss": 3.0476,
+      "step": 17316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5402676463127136,
+      "learning_rate": 0.0004891219812143369,
+      "loss": 3.0622,
+      "step": 17317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4617784023284912,
+      "learning_rate": 0.0004891100294399166,
+      "loss": 3.1742,
+      "step": 17318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5083367228507996,
+      "learning_rate": 0.0004890980771674169,
+      "loss": 3.3683,
+      "step": 17319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4868304431438446,
+      "learning_rate": 0.0004890861243968691,
+      "loss": 3.3446,
+      "step": 17320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5708618760108948,
+      "learning_rate": 0.000489074171128305,
+      "loss": 3.1252,
+      "step": 17321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5267812013626099,
+      "learning_rate": 0.0004890622173617558,
+      "loss": 3.1191,
+      "step": 17322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5371107459068298,
+      "learning_rate": 0.0004890502630972532,
+      "loss": 3.1529,
+      "step": 17323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5120932459831238,
+      "learning_rate": 0.0004890383083348285,
+      "loss": 3.0703,
+      "step": 17324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5467737317085266,
+      "learning_rate": 0.0004890263530745134,
+      "loss": 3.2532,
+      "step": 17325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4981206953525543,
+      "learning_rate": 0.0004890143973163391,
+      "loss": 3.1652,
+      "step": 17326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5158091187477112,
+      "learning_rate": 0.0004890024410603372,
+      "loss": 3.0072,
+      "step": 17327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5790042281150818,
+      "learning_rate": 0.0004889904843065394,
+      "loss": 3.2988,
+      "step": 17328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5327931642532349,
+      "learning_rate": 0.0004889785270549771,
+      "loss": 3.0204,
+      "step": 17329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5174896717071533,
+      "learning_rate": 0.0004889665693056817,
+      "loss": 3.1526,
+      "step": 17330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5148389339447021,
+      "learning_rate": 0.0004889546110586847,
+      "loss": 3.1528,
+      "step": 17331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4947047829627991,
+      "learning_rate": 0.0004889426523140175,
+      "loss": 2.9137,
+      "step": 17332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5458266735076904,
+      "learning_rate": 0.0004889306930717118,
+      "loss": 3.3774,
+      "step": 17333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4884343147277832,
+      "learning_rate": 0.0004889187333317991,
+      "loss": 3.116,
+      "step": 17334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5519797801971436,
+      "learning_rate": 0.0004889067730943107,
+      "loss": 3.1202,
+      "step": 17335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5409486889839172,
+      "learning_rate": 0.0004888948123592783,
+      "loss": 3.1933,
+      "step": 17336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5440702438354492,
+      "learning_rate": 0.0004888828511267332,
+      "loss": 3.2795,
+      "step": 17337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5199924111366272,
+      "learning_rate": 0.0004888708893967071,
+      "loss": 2.9574,
+      "step": 17338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5392219424247742,
+      "learning_rate": 0.0004888589271692314,
+      "loss": 3.3653,
+      "step": 17339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49010539054870605,
+      "learning_rate": 0.0004888469644443377,
+      "loss": 3.1307,
+      "step": 17340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5155744552612305,
+      "learning_rate": 0.0004888350012220573,
+      "loss": 3.142,
+      "step": 17341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5155385732650757,
+      "learning_rate": 0.000488823037502422,
+      "loss": 3.3988,
+      "step": 17342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5531888008117676,
+      "learning_rate": 0.000488811073285463,
+      "loss": 2.9271,
+      "step": 17343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5326036214828491,
+      "learning_rate": 0.0004887991085712121,
+      "loss": 3.1794,
+      "step": 17344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5249530673027039,
+      "learning_rate": 0.0004887871433597006,
+      "loss": 3.2291,
+      "step": 17345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5138980746269226,
+      "learning_rate": 0.0004887751776509602,
+      "loss": 3.0757,
+      "step": 17346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5147232413291931,
+      "learning_rate": 0.0004887632114450222,
+      "loss": 3.0998,
+      "step": 17347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5001515746116638,
+      "learning_rate": 0.0004887512447419184,
+      "loss": 3.2291,
+      "step": 17348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5295124650001526,
+      "learning_rate": 0.00048873927754168,
+      "loss": 3.1302,
+      "step": 17349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5005471706390381,
+      "learning_rate": 0.0004887273098443388,
+      "loss": 3.0511,
+      "step": 17350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5099095702171326,
+      "learning_rate": 0.000488715341649926,
+      "loss": 3.1537,
+      "step": 17351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5452235341072083,
+      "learning_rate": 0.0004887033729584734,
+      "loss": 3.231,
+      "step": 17352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5622381567955017,
+      "learning_rate": 0.0004886914037700124,
+      "loss": 3.0763,
+      "step": 17353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5072722434997559,
+      "learning_rate": 0.0004886794340845746,
+      "loss": 3.0325,
+      "step": 17354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5430617332458496,
+      "learning_rate": 0.0004886674639021914,
+      "loss": 3.1989,
+      "step": 17355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5264281034469604,
+      "learning_rate": 0.0004886554932228945,
+      "loss": 3.2848,
+      "step": 17356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5271099209785461,
+      "learning_rate": 0.0004886435220467154,
+      "loss": 3.2634,
+      "step": 17357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578886866569519,
+      "learning_rate": 0.0004886315503736854,
+      "loss": 3.2401,
+      "step": 17358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48551952838897705,
+      "learning_rate": 0.0004886195782038364,
+      "loss": 3.2252,
+      "step": 17359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49196016788482666,
+      "learning_rate": 0.0004886076055371995,
+      "loss": 3.2997,
+      "step": 17360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5155906677246094,
+      "learning_rate": 0.0004885956323738066,
+      "loss": 3.0486,
+      "step": 17361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5028941631317139,
+      "learning_rate": 0.0004885836587136892,
+      "loss": 3.1732,
+      "step": 17362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49608007073402405,
+      "learning_rate": 0.0004885716845568786,
+      "loss": 3.0215,
+      "step": 17363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5235550999641418,
+      "learning_rate": 0.0004885597099034064,
+      "loss": 3.1015,
+      "step": 17364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5136752128601074,
+      "learning_rate": 0.0004885477347533044,
+      "loss": 3.1078,
+      "step": 17365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49417296051979065,
+      "learning_rate": 0.0004885357591066038,
+      "loss": 3.1097,
+      "step": 17366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4974791705608368,
+      "learning_rate": 0.0004885237829633363,
+      "loss": 3.176,
+      "step": 17367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4924065172672272,
+      "learning_rate": 0.0004885118063235335,
+      "loss": 3.2376,
+      "step": 17368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5056686997413635,
+      "learning_rate": 0.0004884998291872269,
+      "loss": 3.3538,
+      "step": 17369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5538915395736694,
+      "learning_rate": 0.0004884878515544481,
+      "loss": 3.1879,
+      "step": 17370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5767045617103577,
+      "learning_rate": 0.0004884758734252285,
+      "loss": 3.1592,
+      "step": 17371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5208998918533325,
+      "learning_rate": 0.0004884638947995996,
+      "loss": 3.1981,
+      "step": 17372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5236579775810242,
+      "learning_rate": 0.0004884519156775932,
+      "loss": 3.0012,
+      "step": 17373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5991045832633972,
+      "learning_rate": 0.0004884399360592407,
+      "loss": 3.1331,
+      "step": 17374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5242659449577332,
+      "learning_rate": 0.0004884279559445737,
+      "loss": 3.3605,
+      "step": 17375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5292356610298157,
+      "learning_rate": 0.0004884159753336237,
+      "loss": 3.2468,
+      "step": 17376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5420272946357727,
+      "learning_rate": 0.0004884039942264222,
+      "loss": 2.9437,
+      "step": 17377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4963341951370239,
+      "learning_rate": 0.000488392012623001,
+      "loss": 3.2331,
+      "step": 17378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.505173921585083,
+      "learning_rate": 0.0004883800305233914,
+      "loss": 3.271,
+      "step": 17379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.530937671661377,
+      "learning_rate": 0.000488368047927625,
+      "loss": 3.2967,
+      "step": 17380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5380979776382446,
+      "learning_rate": 0.0004883560648357335,
+      "loss": 3.0993,
+      "step": 17381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5381925702095032,
+      "learning_rate": 0.0004883440812477484,
+      "loss": 3.2143,
+      "step": 17382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.517504096031189,
+      "learning_rate": 0.0004883320971637012,
+      "loss": 3.2316,
+      "step": 17383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5057746767997742,
+      "learning_rate": 0.0004883201125836234,
+      "loss": 3.1969,
+      "step": 17384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5114002227783203,
+      "learning_rate": 0.0004883081275075467,
+      "loss": 3.135,
+      "step": 17385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5425366163253784,
+      "learning_rate": 0.0004882961419355028,
+      "loss": 3.0397,
+      "step": 17386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5366644859313965,
+      "learning_rate": 0.000488284155867523,
+      "loss": 3.3255,
+      "step": 17387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5261602401733398,
+      "learning_rate": 0.0004882721693036389,
+      "loss": 3.1629,
+      "step": 17388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5072944760322571,
+      "learning_rate": 0.0004882601822438823,
+      "loss": 3.1021,
+      "step": 17389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6098434329032898,
+      "learning_rate": 0.0004882481946882845,
+      "loss": 2.9911,
+      "step": 17390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.535154402256012,
+      "learning_rate": 0.0004882362066368771,
+      "loss": 3.2225,
+      "step": 17391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.508673369884491,
+      "learning_rate": 0.00048822421808969197,
+      "loss": 3.2243,
+      "step": 17392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5214126706123352,
+      "learning_rate": 0.0004882122290467603,
+      "loss": 3.1554,
+      "step": 17393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5492194294929504,
+      "learning_rate": 0.0004882002395081139,
+      "loss": 3.1623,
+      "step": 17394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.558017909526825,
+      "learning_rate": 0.0004881882494737843,
+      "loss": 3.2415,
+      "step": 17395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5010644793510437,
+      "learning_rate": 0.0004881762589438031,
+      "loss": 3.2828,
+      "step": 17396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5531977415084839,
+      "learning_rate": 0.00048816426791820175,
+      "loss": 3.0989,
+      "step": 17397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5068770051002502,
+      "learning_rate": 0.00048815227639701205,
+      "loss": 3.3572,
+      "step": 17398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6324036121368408,
+      "learning_rate": 0.00048814028438026535,
+      "loss": 3.0686,
+      "step": 17399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5348002910614014,
+      "learning_rate": 0.00048812829186799337,
+      "loss": 3.1609,
+      "step": 17400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5051308274269104,
+      "learning_rate": 0.00048811629886022777,
+      "loss": 3.0573,
+      "step": 17401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5545761585235596,
+      "learning_rate": 0.00048810430535699994,
+      "loss": 3.0548,
+      "step": 17402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.692309558391571,
+      "learning_rate": 0.0004880923113583416,
+      "loss": 3.1483,
+      "step": 17403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5532946586608887,
+      "learning_rate": 0.0004880803168642843,
+      "loss": 3.1074,
+      "step": 17404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7708467841148376,
+      "learning_rate": 0.00048806832187485964,
+      "loss": 3.3463,
+      "step": 17405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5050947666168213,
+      "learning_rate": 0.00048805632639009936,
+      "loss": 3.0551,
+      "step": 17406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49402984976768494,
+      "learning_rate": 0.0004880443304100347,
+      "loss": 3.1811,
+      "step": 17407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6107621192932129,
+      "learning_rate": 0.0004880323339346976,
+      "loss": 3.1067,
+      "step": 17408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5168527960777283,
+      "learning_rate": 0.00048802033696411946,
+      "loss": 3.1141,
+      "step": 17409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.518570601940155,
+      "learning_rate": 0.00048800833949833196,
+      "loss": 2.9312,
+      "step": 17410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5209082961082458,
+      "learning_rate": 0.0004879963415373667,
+      "loss": 2.9536,
+      "step": 17411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49775710701942444,
+      "learning_rate": 0.0004879843430812552,
+      "loss": 3.173,
+      "step": 17412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5183805823326111,
+      "learning_rate": 0.00048797234413002914,
+      "loss": 3.2114,
+      "step": 17413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5256752967834473,
+      "learning_rate": 0.00048796034468372016,
+      "loss": 3.0888,
+      "step": 17414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.47854384779930115,
+      "learning_rate": 0.0004879483447423597,
+      "loss": 3.3277,
+      "step": 17415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5341052412986755,
+      "learning_rate": 0.00048793634430597946,
+      "loss": 3.0255,
+      "step": 17416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5231382846832275,
+      "learning_rate": 0.00048792434337461107,
+      "loss": 3.367,
+      "step": 17417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5256244540214539,
+      "learning_rate": 0.0004879123419482862,
+      "loss": 3.3875,
+      "step": 17418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.529234766960144,
+      "learning_rate": 0.00048790034002703646,
+      "loss": 3.075,
+      "step": 17419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6451051831245422,
+      "learning_rate": 0.0004878883376108932,
+      "loss": 3.0641,
+      "step": 17420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5298725366592407,
+      "learning_rate": 0.00048787633469988834,
+      "loss": 3.2128,
+      "step": 17421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5713647603988647,
+      "learning_rate": 0.00048786433129405335,
+      "loss": 3.0181,
+      "step": 17422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48418837785720825,
+      "learning_rate": 0.0004878523273934198,
+      "loss": 3.0541,
+      "step": 17423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5305963754653931,
+      "learning_rate": 0.00048784032299801946,
+      "loss": 3.2004,
+      "step": 17424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5473353266716003,
+      "learning_rate": 0.00048782831810788383,
+      "loss": 3.0637,
+      "step": 17425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5448347926139832,
+      "learning_rate": 0.00048781631272304453,
+      "loss": 3.1816,
+      "step": 17426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.522763729095459,
+      "learning_rate": 0.00048780430684353327,
+      "loss": 3.0765,
+      "step": 17427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49944186210632324,
+      "learning_rate": 0.0004877923004693815,
+      "loss": 3.2881,
+      "step": 17428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.522628664970398,
+      "learning_rate": 0.0004877802936006211,
+      "loss": 3.361,
+      "step": 17429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5693197250366211,
+      "learning_rate": 0.00048776828623728346,
+      "loss": 3.3028,
+      "step": 17430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5283417701721191,
+      "learning_rate": 0.00048775627837940026,
+      "loss": 3.2618,
+      "step": 17431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5026355385780334,
+      "learning_rate": 0.0004877442700270032,
+      "loss": 3.0906,
+      "step": 17432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5224277973175049,
+      "learning_rate": 0.0004877322611801239,
+      "loss": 2.9159,
+      "step": 17433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5417836308479309,
+      "learning_rate": 0.0004877202518387939,
+      "loss": 3.1032,
+      "step": 17434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5314958691596985,
+      "learning_rate": 0.00048770824200304494,
+      "loss": 3.1461,
+      "step": 17435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5614078640937805,
+      "learning_rate": 0.0004876962316729085,
+      "loss": 3.0781,
+      "step": 17436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5217527747154236,
+      "learning_rate": 0.00048768422084841643,
+      "loss": 3.1365,
+      "step": 17437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5421162247657776,
+      "learning_rate": 0.00048767220952960013,
+      "loss": 3.2526,
+      "step": 17438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.647996187210083,
+      "learning_rate": 0.0004876601977164914,
+      "loss": 2.8944,
+      "step": 17439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5372748970985413,
+      "learning_rate": 0.0004876481854091218,
+      "loss": 2.9952,
+      "step": 17440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.541601300239563,
+      "learning_rate": 0.000487636172607523,
+      "loss": 3.25,
+      "step": 17441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5672915577888489,
+      "learning_rate": 0.0004876241593117267,
+      "loss": 2.9457,
+      "step": 17442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.530311644077301,
+      "learning_rate": 0.0004876121455217644,
+      "loss": 2.875,
+      "step": 17443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5655810236930847,
+      "learning_rate": 0.0004876001312376679,
+      "loss": 3.3092,
+      "step": 17444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5126374363899231,
+      "learning_rate": 0.0004875881164594686,
+      "loss": 3.0786,
+      "step": 17445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5114005208015442,
+      "learning_rate": 0.0004875761011871984,
+      "loss": 3.4152,
+      "step": 17446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5263176560401917,
+      "learning_rate": 0.00048756408542088885,
+      "loss": 3.2279,
+      "step": 17447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5074858069419861,
+      "learning_rate": 0.0004875520691605716,
+      "loss": 3.2816,
+      "step": 17448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5116199851036072,
+      "learning_rate": 0.00048754005240627833,
+      "loss": 3.3319,
+      "step": 17449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5231490135192871,
+      "learning_rate": 0.00048752803515804063,
+      "loss": 3.3931,
+      "step": 17450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6299848556518555,
+      "learning_rate": 0.00048751601741589014,
+      "loss": 2.8908,
+      "step": 17451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5070332288742065,
+      "learning_rate": 0.00048750399917985856,
+      "loss": 3.2715,
+      "step": 17452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6328778862953186,
+      "learning_rate": 0.0004874919804499775,
+      "loss": 3.2513,
+      "step": 17453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.52276211977005,
+      "learning_rate": 0.0004874799612262787,
+      "loss": 3.0997,
+      "step": 17454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5386821031570435,
+      "learning_rate": 0.0004874679415087938,
+      "loss": 3.2454,
+      "step": 17455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.506618320941925,
+      "learning_rate": 0.00048745592129755433,
+      "loss": 3.1048,
+      "step": 17456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5649548768997192,
+      "learning_rate": 0.0004874439005925921,
+      "loss": 2.8524,
+      "step": 17457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5274940729141235,
+      "learning_rate": 0.00048743187939393867,
+      "loss": 3.2162,
+      "step": 17458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546772301197052,
+      "learning_rate": 0.0004874198577016258,
+      "loss": 3.1506,
+      "step": 17459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5609162449836731,
+      "learning_rate": 0.00048740783551568504,
+      "loss": 3.3401,
+      "step": 17460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5612881183624268,
+      "learning_rate": 0.0004873958128361481,
+      "loss": 3.1514,
+      "step": 17461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5242128968238831,
+      "learning_rate": 0.0004873837896630467,
+      "loss": 2.9899,
+      "step": 17462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5031869411468506,
+      "learning_rate": 0.00048737176599641246,
+      "loss": 3.2779,
+      "step": 17463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5017745494842529,
+      "learning_rate": 0.00048735974183627707,
+      "loss": 3.2305,
+      "step": 17464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5221795439720154,
+      "learning_rate": 0.0004873477171826721,
+      "loss": 3.2039,
+      "step": 17465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5255293846130371,
+      "learning_rate": 0.0004873356920356293,
+      "loss": 3.1282,
+      "step": 17466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5047944784164429,
+      "learning_rate": 0.00048732366639518037,
+      "loss": 2.9352,
+      "step": 17467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5233396291732788,
+      "learning_rate": 0.000487311640261357,
+      "loss": 3.2509,
+      "step": 17468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.505703330039978,
+      "learning_rate": 0.0004872996136341908,
+      "loss": 3.2049,
+      "step": 17469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5102468729019165,
+      "learning_rate": 0.00048728758651371345,
+      "loss": 3.1221,
+      "step": 17470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5118353962898254,
+      "learning_rate": 0.00048727555889995655,
+      "loss": 3.0826,
+      "step": 17471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5210556387901306,
+      "learning_rate": 0.000487263530792952,
+      "loss": 3.0314,
+      "step": 17472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5125072598457336,
+      "learning_rate": 0.0004872515021927313,
+      "loss": 3.3292,
+      "step": 17473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5282806754112244,
+      "learning_rate": 0.00048723947309932625,
+      "loss": 3.3225,
+      "step": 17474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4971778392791748,
+      "learning_rate": 0.00048722744351276836,
+      "loss": 3.2805,
+      "step": 17475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5685766339302063,
+      "learning_rate": 0.00048721541343308946,
+      "loss": 3.1796,
+      "step": 17476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5498575568199158,
+      "learning_rate": 0.0004872033828603212,
+      "loss": 3.1008,
+      "step": 17477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5463971495628357,
+      "learning_rate": 0.0004871913517944952,
+      "loss": 3.0532,
+      "step": 17478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5287131667137146,
+      "learning_rate": 0.0004871793202356432,
+      "loss": 3.3487,
+      "step": 17479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5136821866035461,
+      "learning_rate": 0.000487167288183797,
+      "loss": 3.2506,
+      "step": 17480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5033801794052124,
+      "learning_rate": 0.0004871552556389881,
+      "loss": 3.2242,
+      "step": 17481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5138252377510071,
+      "learning_rate": 0.00048714322260124825,
+      "loss": 3.0705,
+      "step": 17482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5219220519065857,
+      "learning_rate": 0.00048713118907060923,
+      "loss": 2.9397,
+      "step": 17483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5228368043899536,
+      "learning_rate": 0.00048711915504710276,
+      "loss": 3.2545,
+      "step": 17484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5889149308204651,
+      "learning_rate": 0.00048710712053076036,
+      "loss": 3.1426,
+      "step": 17485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5183678865432739,
+      "learning_rate": 0.00048709508552161377,
+      "loss": 3.1804,
+      "step": 17486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5020313858985901,
+      "learning_rate": 0.0004870830500196947,
+      "loss": 3.4664,
+      "step": 17487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.522677481174469,
+      "learning_rate": 0.00048707101402503493,
+      "loss": 3.2824,
+      "step": 17488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5405004024505615,
+      "learning_rate": 0.0004870589775376661,
+      "loss": 3.1882,
+      "step": 17489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4743952453136444,
+      "learning_rate": 0.00048704694055762005,
+      "loss": 3.257,
+      "step": 17490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49832382798194885,
+      "learning_rate": 0.00048703490308492825,
+      "loss": 3.156,
+      "step": 17491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.553101122379303,
+      "learning_rate": 0.0004870228651196225,
+      "loss": 3.1769,
+      "step": 17492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5105452537536621,
+      "learning_rate": 0.00048701082666173447,
+      "loss": 3.228,
+      "step": 17493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4981645345687866,
+      "learning_rate": 0.0004869987877112961,
+      "loss": 3.1893,
+      "step": 17494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5161097645759583,
+      "learning_rate": 0.00048698674826833875,
+      "loss": 3.1463,
+      "step": 17495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5108583569526672,
+      "learning_rate": 0.00048697470833289434,
+      "loss": 3.1373,
+      "step": 17496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5210059881210327,
+      "learning_rate": 0.00048696266790499457,
+      "loss": 3.2682,
+      "step": 17497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5149890184402466,
+      "learning_rate": 0.000486950626984671,
+      "loss": 3.1596,
+      "step": 17498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5131500959396362,
+      "learning_rate": 0.0004869385855719555,
+      "loss": 3.0184,
+      "step": 17499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5088669061660767,
+      "learning_rate": 0.00048692654366687985,
+      "loss": 3.1619,
+      "step": 17500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5249161124229431,
+      "learning_rate": 0.0004869145012694755,
+      "loss": 3.087,
+      "step": 17501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.541541337966919,
+      "learning_rate": 0.00048690245837977445,
+      "loss": 3.0936,
+      "step": 17502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5092646479606628,
+      "learning_rate": 0.0004868904149978082,
+      "loss": 3.0542,
+      "step": 17503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5421231389045715,
+      "learning_rate": 0.0004868783711236087,
+      "loss": 3.3001,
+      "step": 17504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5062888860702515,
+      "learning_rate": 0.00048686632675720736,
+      "loss": 3.0985,
+      "step": 17505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5206896066665649,
+      "learning_rate": 0.00048685428189863616,
+      "loss": 3.0927,
+      "step": 17506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5333325862884521,
+      "learning_rate": 0.0004868422365479268,
+      "loss": 3.2301,
+      "step": 17507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5655118227005005,
+      "learning_rate": 0.0004868301907051109,
+      "loss": 3.1063,
+      "step": 17508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5310041904449463,
+      "learning_rate": 0.0004868181443702202,
+      "loss": 3.1183,
+      "step": 17509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5519410967826843,
+      "learning_rate": 0.00048680609754328644,
+      "loss": 3.0312,
+      "step": 17510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6148340106010437,
+      "learning_rate": 0.0004867940502243415,
+      "loss": 3.2461,
+      "step": 17511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.512523889541626,
+      "learning_rate": 0.00048678200241341685,
+      "loss": 3.268,
+      "step": 17512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4837605059146881,
+      "learning_rate": 0.0004867699541105444,
+      "loss": 2.947,
+      "step": 17513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5407007336616516,
+      "learning_rate": 0.00048675790531575583,
+      "loss": 3.1923,
+      "step": 17514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.495334267616272,
+      "learning_rate": 0.00048674585602908287,
+      "loss": 3.2698,
+      "step": 17515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5168710947036743,
+      "learning_rate": 0.00048673380625055735,
+      "loss": 3.2698,
+      "step": 17516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5113250613212585,
+      "learning_rate": 0.0004867217559802109,
+      "loss": 3.401,
+      "step": 17517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5109712481498718,
+      "learning_rate": 0.00048670970521807517,
+      "loss": 3.0884,
+      "step": 17518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5782906413078308,
+      "learning_rate": 0.00048669765396418214,
+      "loss": 3.0593,
+      "step": 17519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4731515347957611,
+      "learning_rate": 0.00048668560221856337,
+      "loss": 3.1659,
+      "step": 17520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5255733132362366,
+      "learning_rate": 0.0004866735499812506,
+      "loss": 3.1536,
+      "step": 17521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5351244807243347,
+      "learning_rate": 0.0004866614972522757,
+      "loss": 3.2712,
+      "step": 17522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5200554132461548,
+      "learning_rate": 0.0004866494440316703,
+      "loss": 3.1256,
+      "step": 17523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.510212242603302,
+      "learning_rate": 0.0004866373903194661,
+      "loss": 3.0483,
+      "step": 17524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5115465521812439,
+      "learning_rate": 0.0004866253361156951,
+      "loss": 2.9676,
+      "step": 17525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5391139984130859,
+      "learning_rate": 0.0004866132814203888,
+      "loss": 3.2172,
+      "step": 17526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5223973393440247,
+      "learning_rate": 0.00048660122623357904,
+      "loss": 3.2046,
+      "step": 17527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5327551364898682,
+      "learning_rate": 0.00048658917055529753,
+      "loss": 3.2311,
+      "step": 17528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5152480602264404,
+      "learning_rate": 0.0004865771143855761,
+      "loss": 3.1152,
+      "step": 17529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5064302086830139,
+      "learning_rate": 0.0004865650577244464,
+      "loss": 3.3183,
+      "step": 17530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5391542911529541,
+      "learning_rate": 0.00048655300057194036,
+      "loss": 2.9688,
+      "step": 17531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49061742424964905,
+      "learning_rate": 0.00048654094292808955,
+      "loss": 3.2061,
+      "step": 17532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5557584762573242,
+      "learning_rate": 0.00048652888479292574,
+      "loss": 3.2288,
+      "step": 17533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5739030241966248,
+      "learning_rate": 0.0004865168261664808,
+      "loss": 3.1942,
+      "step": 17534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5252779126167297,
+      "learning_rate": 0.00048650476704878643,
+      "loss": 3.271,
+      "step": 17535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234642028808594,
+      "learning_rate": 0.0004864927074398744,
+      "loss": 3.3313,
+      "step": 17536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5235569477081299,
+      "learning_rate": 0.00048648064733977653,
+      "loss": 3.38,
+      "step": 17537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5185587406158447,
+      "learning_rate": 0.00048646858674852443,
+      "loss": 3.1555,
+      "step": 17538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546230673789978,
+      "learning_rate": 0.00048645652566615,
+      "loss": 2.8245,
+      "step": 17539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5275043249130249,
+      "learning_rate": 0.0004864444640926849,
+      "loss": 3.2103,
+      "step": 17540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5376495122909546,
+      "learning_rate": 0.00048643240202816105,
+      "loss": 3.0776,
+      "step": 17541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5582915544509888,
+      "learning_rate": 0.0004864203394726101,
+      "loss": 2.9081,
+      "step": 17542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5006242990493774,
+      "learning_rate": 0.0004864082764260639,
+      "loss": 3.2375,
+      "step": 17543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5330790281295776,
+      "learning_rate": 0.0004863962128885541,
+      "loss": 3.1756,
+      "step": 17544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5119918584823608,
+      "learning_rate": 0.00048638414886011263,
+      "loss": 3.119,
+      "step": 17545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5270196199417114,
+      "learning_rate": 0.0004863720843407711,
+      "loss": 3.0816,
+      "step": 17546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.525503396987915,
+      "learning_rate": 0.0004863600193305614,
+      "loss": 3.0303,
+      "step": 17547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.528007447719574,
+      "learning_rate": 0.00048634795382951525,
+      "loss": 3.2578,
+      "step": 17548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.495852530002594,
+      "learning_rate": 0.0004863358878376645,
+      "loss": 3.1074,
+      "step": 17549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5401389598846436,
+      "learning_rate": 0.0004863238213550409,
+      "loss": 3.3623,
+      "step": 17550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5262596607208252,
+      "learning_rate": 0.00048631175438167606,
+      "loss": 3.3607,
+      "step": 17551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6294898986816406,
+      "learning_rate": 0.0004862996869176021,
+      "loss": 3.1021,
+      "step": 17552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5093241930007935,
+      "learning_rate": 0.0004862876189628505,
+      "loss": 3.1727,
+      "step": 17553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51717609167099,
+      "learning_rate": 0.0004862755505174533,
+      "loss": 3.0077,
+      "step": 17554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5517076849937439,
+      "learning_rate": 0.00048626348158144206,
+      "loss": 3.1981,
+      "step": 17555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5027273893356323,
+      "learning_rate": 0.0004862514121548486,
+      "loss": 3.0377,
+      "step": 17556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5180482268333435,
+      "learning_rate": 0.0004862393422377048,
+      "loss": 3.1539,
+      "step": 17557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5299745202064514,
+      "learning_rate": 0.00048622727183004246,
+      "loss": 3.0539,
+      "step": 17558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49521052837371826,
+      "learning_rate": 0.0004862152009318933,
+      "loss": 3.3228,
+      "step": 17559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5019090175628662,
+      "learning_rate": 0.0004862031295432892,
+      "loss": 3.1764,
+      "step": 17560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5337850451469421,
+      "learning_rate": 0.0004861910576642618,
+      "loss": 3.2292,
+      "step": 17561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5519272089004517,
+      "learning_rate": 0.000486178985294843,
+      "loss": 3.2541,
+      "step": 17562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5433281660079956,
+      "learning_rate": 0.0004861669124350646,
+      "loss": 3.2232,
+      "step": 17563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5310292840003967,
+      "learning_rate": 0.0004861548390849584,
+      "loss": 3.0135,
+      "step": 17564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.50967937707901,
+      "learning_rate": 0.00048614276524455615,
+      "loss": 3.193,
+      "step": 17565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5428502559661865,
+      "learning_rate": 0.0004861306909138897,
+      "loss": 3.2202,
+      "step": 17566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5671022534370422,
+      "learning_rate": 0.00048611861609299087,
+      "loss": 3.2069,
+      "step": 17567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.517419695854187,
+      "learning_rate": 0.0004861065407818914,
+      "loss": 3.1775,
+      "step": 17568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5266711711883545,
+      "learning_rate": 0.00048609446498062303,
+      "loss": 3.0535,
+      "step": 17569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5126215815544128,
+      "learning_rate": 0.00048608238868921773,
+      "loss": 3.2225,
+      "step": 17570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5277024507522583,
+      "learning_rate": 0.0004860703119077073,
+      "loss": 2.9624,
+      "step": 17571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5014190077781677,
+      "learning_rate": 0.00048605823463612334,
+      "loss": 3.1931,
+      "step": 17572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5363627076148987,
+      "learning_rate": 0.00048604615687449794,
+      "loss": 3.3138,
+      "step": 17573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5196545720100403,
+      "learning_rate": 0.00048603407862286266,
+      "loss": 3.0961,
+      "step": 17574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49793750047683716,
+      "learning_rate": 0.0004860219998812494,
+      "loss": 3.2758,
+      "step": 17575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5129599571228027,
+      "learning_rate": 0.00048600992064969014,
+      "loss": 3.2445,
+      "step": 17576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5278726816177368,
+      "learning_rate": 0.0004859978409282164,
+      "loss": 3.0858,
+      "step": 17577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5560405850410461,
+      "learning_rate": 0.0004859857607168602,
+      "loss": 3.0362,
+      "step": 17578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5135570764541626,
+      "learning_rate": 0.0004859736800156533,
+      "loss": 3.0105,
+      "step": 17579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5374484062194824,
+      "learning_rate": 0.0004859615988246275,
+      "loss": 3.118,
+      "step": 17580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5141043066978455,
+      "learning_rate": 0.00048594951714381465,
+      "loss": 2.999,
+      "step": 17581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5360473990440369,
+      "learning_rate": 0.0004859374349732466,
+      "loss": 3.1131,
+      "step": 17582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5004850029945374,
+      "learning_rate": 0.000485925352312955,
+      "loss": 3.1204,
+      "step": 17583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5214194655418396,
+      "learning_rate": 0.00048591326916297184,
+      "loss": 3.3562,
+      "step": 17584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5029579401016235,
+      "learning_rate": 0.000485901185523329,
+      "loss": 3.2461,
+      "step": 17585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5304317474365234,
+      "learning_rate": 0.0004858891013940582,
+      "loss": 3.2182,
+      "step": 17586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5088167786598206,
+      "learning_rate": 0.00048587701677519127,
+      "loss": 3.1036,
+      "step": 17587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5475550889968872,
+      "learning_rate": 0.00048586493166676004,
+      "loss": 3.0157,
+      "step": 17588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5142987370491028,
+      "learning_rate": 0.0004858528460687963,
+      "loss": 2.8891,
+      "step": 17589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5690994262695312,
+      "learning_rate": 0.000485840759981332,
+      "loss": 3.1212,
+      "step": 17590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5271167159080505,
+      "learning_rate": 0.00048582867340439897,
+      "loss": 3.2301,
+      "step": 17591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.512944757938385,
+      "learning_rate": 0.00048581658633802883,
+      "loss": 3.3691,
+      "step": 17592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5304507613182068,
+      "learning_rate": 0.00048580449878225367,
+      "loss": 3.3141,
+      "step": 17593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6154621839523315,
+      "learning_rate": 0.0004857924107371051,
+      "loss": 3.1346,
+      "step": 17594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.556638777256012,
+      "learning_rate": 0.0004857803222026151,
+      "loss": 3.1249,
+      "step": 17595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5296390652656555,
+      "learning_rate": 0.00048576823317881564,
+      "loss": 2.9605,
+      "step": 17596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5081044435501099,
+      "learning_rate": 0.00048575614366573827,
+      "loss": 3.182,
+      "step": 17597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5104318261146545,
+      "learning_rate": 0.0004857440536634151,
+      "loss": 3.342,
+      "step": 17598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5356442332267761,
+      "learning_rate": 0.0004857319631718777,
+      "loss": 3.0014,
+      "step": 17599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48045575618743896,
+      "learning_rate": 0.000485719872191158,
+      "loss": 3.2048,
+      "step": 17600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4831188917160034,
+      "learning_rate": 0.00048570778072128806,
+      "loss": 3.4564,
+      "step": 17601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5195653438568115,
+      "learning_rate": 0.00048569568876229944,
+      "loss": 3.4224,
+      "step": 17602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5168091654777527,
+      "learning_rate": 0.0004856835963142242,
+      "loss": 3.0934,
+      "step": 17603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49368128180503845,
+      "learning_rate": 0.0004856715033770941,
+      "loss": 2.9334,
+      "step": 17604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5106991529464722,
+      "learning_rate": 0.00048565940995094097,
+      "loss": 3.1075,
+      "step": 17605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5301958918571472,
+      "learning_rate": 0.0004856473160357967,
+      "loss": 2.885,
+      "step": 17606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5249580144882202,
+      "learning_rate": 0.00048563522163169314,
+      "loss": 3.176,
+      "step": 17607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5821053385734558,
+      "learning_rate": 0.000485623126738662,
+      "loss": 3.2309,
+      "step": 17608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5431340932846069,
+      "learning_rate": 0.00048561103135673546,
+      "loss": 3.0735,
+      "step": 17609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5106194019317627,
+      "learning_rate": 0.000485598935485945,
+      "loss": 3.1356,
+      "step": 17610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5526049733161926,
+      "learning_rate": 0.00048558683912632277,
+      "loss": 3.2272,
+      "step": 17611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5209463238716125,
+      "learning_rate": 0.00048557474227790056,
+      "loss": 3.1295,
+      "step": 17612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5306352376937866,
+      "learning_rate": 0.00048556264494071014,
+      "loss": 3.151,
+      "step": 17613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.483786016702652,
+      "learning_rate": 0.00048555054711478345,
+      "loss": 3.1528,
+      "step": 17614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5121860504150391,
+      "learning_rate": 0.00048553844880015225,
+      "loss": 3.2255,
+      "step": 17615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5123046040534973,
+      "learning_rate": 0.0004855263499968486,
+      "loss": 3.2583,
+      "step": 17616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5222660899162292,
+      "learning_rate": 0.0004855142507049042,
+      "loss": 3.3446,
+      "step": 17617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5142706632614136,
+      "learning_rate": 0.00048550215092435094,
+      "loss": 3.2527,
+      "step": 17618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5308032035827637,
+      "learning_rate": 0.00048549005065522073,
+      "loss": 3.3778,
+      "step": 17619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5231295228004456,
+      "learning_rate": 0.00048547794989754544,
+      "loss": 3.0498,
+      "step": 17620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5291686654090881,
+      "learning_rate": 0.00048546584865135684,
+      "loss": 3.1601,
+      "step": 17621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5070790648460388,
+      "learning_rate": 0.00048545374691668703,
+      "loss": 3.2343,
+      "step": 17622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5530633926391602,
+      "learning_rate": 0.00048544164469356766,
+      "loss": 3.076,
+      "step": 17623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4944988489151001,
+      "learning_rate": 0.0004854295419820307,
+      "loss": 2.9956,
+      "step": 17624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5238565802574158,
+      "learning_rate": 0.000485417438782108,
+      "loss": 2.9795,
+      "step": 17625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5411428809165955,
+      "learning_rate": 0.00048540533509383143,
+      "loss": 3.3082,
+      "step": 17626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4969344437122345,
+      "learning_rate": 0.00048539323091723296,
+      "loss": 3.0575,
+      "step": 17627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5270441174507141,
+      "learning_rate": 0.00048538112625234436,
+      "loss": 3.0456,
+      "step": 17628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5126833915710449,
+      "learning_rate": 0.00048536902109919756,
+      "loss": 3.183,
+      "step": 17629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5584427118301392,
+      "learning_rate": 0.00048535691545782445,
+      "loss": 3.2503,
+      "step": 17630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5255547761917114,
+      "learning_rate": 0.0004853448093282569,
+      "loss": 3.2817,
+      "step": 17631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5031391382217407,
+      "learning_rate": 0.0004853327027105267,
+      "loss": 3.1273,
+      "step": 17632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5100876092910767,
+      "learning_rate": 0.0004853205956046659,
+      "loss": 3.3041,
+      "step": 17633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5045557022094727,
+      "learning_rate": 0.0004853084880107064,
+      "loss": 3.3764,
+      "step": 17634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.512071430683136,
+      "learning_rate": 0.00048529637992867985,
+      "loss": 3.2025,
+      "step": 17635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.505384087562561,
+      "learning_rate": 0.00048528427135861835,
+      "loss": 3.1017,
+      "step": 17636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5387892127037048,
+      "learning_rate": 0.0004852721623005538,
+      "loss": 2.9785,
+      "step": 17637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5180559158325195,
+      "learning_rate": 0.00048526005275451804,
+      "loss": 3.3379,
+      "step": 17638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5065121650695801,
+      "learning_rate": 0.0004852479427205428,
+      "loss": 3.1839,
+      "step": 17639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5370428562164307,
+      "learning_rate": 0.00048523583219866023,
+      "loss": 3.4099,
+      "step": 17640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5340681076049805,
+      "learning_rate": 0.0004852237211889022,
+      "loss": 3.2899,
+      "step": 17641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.524978518486023,
+      "learning_rate": 0.0004852116096913004,
+      "loss": 3.0903,
+      "step": 17642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5199117660522461,
+      "learning_rate": 0.000485199497705887,
+      "loss": 3.3207,
+      "step": 17643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5268123149871826,
+      "learning_rate": 0.00048518738523269366,
+      "loss": 3.1258,
+      "step": 17644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5221728086471558,
+      "learning_rate": 0.0004851752722717524,
+      "loss": 3.0099,
+      "step": 17645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5612000226974487,
+      "learning_rate": 0.00048516315882309513,
+      "loss": 3.1997,
+      "step": 17646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4834546148777008,
+      "learning_rate": 0.00048515104488675373,
+      "loss": 3.2142,
+      "step": 17647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5152857899665833,
+      "learning_rate": 0.0004851389304627601,
+      "loss": 3.0943,
+      "step": 17648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6461136937141418,
+      "learning_rate": 0.0004851268155511462,
+      "loss": 3.0051,
+      "step": 17649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5414785146713257,
+      "learning_rate": 0.00048511470015194394,
+      "loss": 3.1216,
+      "step": 17650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5066803693771362,
+      "learning_rate": 0.000485102584265185,
+      "loss": 2.9696,
+      "step": 17651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5823258757591248,
+      "learning_rate": 0.0004850904678909016,
+      "loss": 3.2366,
+      "step": 17652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5320606231689453,
+      "learning_rate": 0.0004850783510291256,
+      "loss": 2.8672,
+      "step": 17653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5367481708526611,
+      "learning_rate": 0.0004850662336798888,
+      "loss": 3.3241,
+      "step": 17654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5208763480186462,
+      "learning_rate": 0.00048505411584322304,
+      "loss": 3.123,
+      "step": 17655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5208843946456909,
+      "learning_rate": 0.00048504199751916045,
+      "loss": 3.0314,
+      "step": 17656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5739818811416626,
+      "learning_rate": 0.00048502987870773287,
+      "loss": 3.1522,
+      "step": 17657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5213282108306885,
+      "learning_rate": 0.0004850177594089722,
+      "loss": 3.227,
+      "step": 17658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4983675479888916,
+      "learning_rate": 0.00048500563962291024,
+      "loss": 3.1158,
+      "step": 17659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5267578363418579,
+      "learning_rate": 0.00048499351934957915,
+      "loss": 3.216,
+      "step": 17660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4949793219566345,
+      "learning_rate": 0.0004849813985890107,
+      "loss": 2.8112,
+      "step": 17661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5270505547523499,
+      "learning_rate": 0.0004849692773412368,
+      "loss": 3.1422,
+      "step": 17662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.604992151260376,
+      "learning_rate": 0.00048495715560628946,
+      "loss": 3.2086,
+      "step": 17663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5165683627128601,
+      "learning_rate": 0.0004849450333842006,
+      "loss": 3.1152,
+      "step": 17664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49626970291137695,
+      "learning_rate": 0.00048493291067500214,
+      "loss": 3.0413,
+      "step": 17665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5042181611061096,
+      "learning_rate": 0.0004849207874787258,
+      "loss": 3.1962,
+      "step": 17666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5245406627655029,
+      "learning_rate": 0.0004849086637954039,
+      "loss": 3.0042,
+      "step": 17667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49068549275398254,
+      "learning_rate": 0.00048489653962506806,
+      "loss": 3.0559,
+      "step": 17668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.576702892780304,
+      "learning_rate": 0.0004848844149677504,
+      "loss": 2.9814,
+      "step": 17669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48933812975883484,
+      "learning_rate": 0.0004848722898234827,
+      "loss": 3.1889,
+      "step": 17670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5341169834136963,
+      "learning_rate": 0.00048486016419229696,
+      "loss": 3.2209,
+      "step": 17671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5466429591178894,
+      "learning_rate": 0.0004848480380742252,
+      "loss": 3.189,
+      "step": 17672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5925959348678589,
+      "learning_rate": 0.00048483591146929926,
+      "loss": 3.4804,
+      "step": 17673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.544042706489563,
+      "learning_rate": 0.00048482378437755103,
+      "loss": 3.2635,
+      "step": 17674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.518909752368927,
+      "learning_rate": 0.0004848116567990126,
+      "loss": 3.3417,
+      "step": 17675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5055917501449585,
+      "learning_rate": 0.0004847995287337158,
+      "loss": 3.0895,
+      "step": 17676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5370952486991882,
+      "learning_rate": 0.0004847874001816926,
+      "loss": 3.2929,
+      "step": 17677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5014994144439697,
+      "learning_rate": 0.00048477527114297494,
+      "loss": 3.09,
+      "step": 17678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5466054081916809,
+      "learning_rate": 0.00048476314161759486,
+      "loss": 3.032,
+      "step": 17679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.554060161113739,
+      "learning_rate": 0.0004847510116055842,
+      "loss": 3.1949,
+      "step": 17680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5110071897506714,
+      "learning_rate": 0.00048473888110697484,
+      "loss": 3.1546,
+      "step": 17681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5429432392120361,
+      "learning_rate": 0.00048472675012179887,
+      "loss": 3.1586,
+      "step": 17682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48281505703926086,
+      "learning_rate": 0.00048471461865008816,
+      "loss": 3.0572,
+      "step": 17683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5270022749900818,
+      "learning_rate": 0.00048470248669187484,
+      "loss": 3.2111,
+      "step": 17684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4998297691345215,
+      "learning_rate": 0.0004846903542471906,
+      "loss": 3.1914,
+      "step": 17685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5160195231437683,
+      "learning_rate": 0.00048467822131606747,
+      "loss": 3.1542,
+      "step": 17686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5145862698554993,
+      "learning_rate": 0.0004846660878985375,
+      "loss": 2.9667,
+      "step": 17687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5544316172599792,
+      "learning_rate": 0.00048465395399463257,
+      "loss": 3.1801,
+      "step": 17688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4879906177520752,
+      "learning_rate": 0.00048464181960438475,
+      "loss": 3.2567,
+      "step": 17689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4931853711605072,
+      "learning_rate": 0.00048462968472782586,
+      "loss": 3.2061,
+      "step": 17690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5492919683456421,
+      "learning_rate": 0.00048461754936498787,
+      "loss": 3.0187,
+      "step": 17691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5467280149459839,
+      "learning_rate": 0.00048460541351590277,
+      "loss": 3.1143,
+      "step": 17692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49284547567367554,
+      "learning_rate": 0.0004845932771806026,
+      "loss": 3.1644,
+      "step": 17693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5207919478416443,
+      "learning_rate": 0.0004845811403591193,
+      "loss": 3.1206,
+      "step": 17694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5158400535583496,
+      "learning_rate": 0.00048456900305148475,
+      "loss": 3.1989,
+      "step": 17695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4813472330570221,
+      "learning_rate": 0.00048455686525773094,
+      "loss": 2.9873,
+      "step": 17696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49759748578071594,
+      "learning_rate": 0.0004845447269778898,
+      "loss": 3.2696,
+      "step": 17697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5683988332748413,
+      "learning_rate": 0.00048453258821199347,
+      "loss": 2.8853,
+      "step": 17698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5115375518798828,
+      "learning_rate": 0.0004845204489600738,
+      "loss": 3.1641,
+      "step": 17699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5470658540725708,
+      "learning_rate": 0.0004845083092221629,
+      "loss": 3.1421,
+      "step": 17700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5110447406768799,
+      "learning_rate": 0.0004844961689982924,
+      "loss": 3.2852,
+      "step": 17701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5411603450775146,
+      "learning_rate": 0.0004844840282884947,
+      "loss": 3.1624,
+      "step": 17702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5393169522285461,
+      "learning_rate": 0.00048447188709280144,
+      "loss": 3.1563,
+      "step": 17703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.480815052986145,
+      "learning_rate": 0.00048445974541124474,
+      "loss": 3.3201,
+      "step": 17704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5160706639289856,
+      "learning_rate": 0.00048444760324385655,
+      "loss": 3.3324,
+      "step": 17705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5325861573219299,
+      "learning_rate": 0.0004844354605906689,
+      "loss": 3.0338,
+      "step": 17706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4856245517730713,
+      "learning_rate": 0.0004844233174517138,
+      "loss": 3.1634,
+      "step": 17707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5197152495384216,
+      "learning_rate": 0.00048441117382702316,
+      "loss": 3.1673,
+      "step": 17708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.537200927734375,
+      "learning_rate": 0.0004843990297166289,
+      "loss": 3.1794,
+      "step": 17709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5353251695632935,
+      "learning_rate": 0.00048438688512056317,
+      "loss": 3.0948,
+      "step": 17710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5023453831672668,
+      "learning_rate": 0.0004843747400388579,
+      "loss": 3.2727,
+      "step": 17711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5291193127632141,
+      "learning_rate": 0.00048436259447154497,
+      "loss": 2.969,
+      "step": 17712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5221428871154785,
+      "learning_rate": 0.00048435044841865646,
+      "loss": 3.1545,
+      "step": 17713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5405849814414978,
+      "learning_rate": 0.00048433830188022433,
+      "loss": 3.109,
+      "step": 17714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.518534779548645,
+      "learning_rate": 0.0004843261548562806,
+      "loss": 3.2215,
+      "step": 17715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5127702951431274,
+      "learning_rate": 0.00048431400734685724,
+      "loss": 3.1477,
+      "step": 17716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49492380023002625,
+      "learning_rate": 0.0004843018593519863,
+      "loss": 3.1372,
+      "step": 17717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5577251315116882,
+      "learning_rate": 0.0004842897108716997,
+      "loss": 3.078,
+      "step": 17718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5143783092498779,
+      "learning_rate": 0.0004842775619060295,
+      "loss": 3.0365,
+      "step": 17719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5474133491516113,
+      "learning_rate": 0.0004842654124550077,
+      "loss": 3.0492,
+      "step": 17720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5189671516418457,
+      "learning_rate": 0.0004842532625186661,
+      "loss": 3.1727,
+      "step": 17721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5471231341362,
+      "learning_rate": 0.00048424111209703706,
+      "loss": 3.169,
+      "step": 17722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4973917603492737,
+      "learning_rate": 0.00048422896119015233,
+      "loss": 3.0737,
+      "step": 17723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5816762447357178,
+      "learning_rate": 0.00048421680979804393,
+      "loss": 3.1443,
+      "step": 17724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.514470100402832,
+      "learning_rate": 0.0004842046579207439,
+      "loss": 3.111,
+      "step": 17725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5202110409736633,
+      "learning_rate": 0.0004841925055582843,
+      "loss": 3.2958,
+      "step": 17726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48983487486839294,
+      "learning_rate": 0.0004841803527106971,
+      "loss": 3.0307,
+      "step": 17727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5123018026351929,
+      "learning_rate": 0.0004841681993780142,
+      "loss": 3.0805,
+      "step": 17728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5432770252227783,
+      "learning_rate": 0.00048415604556026787,
+      "loss": 3.2386,
+      "step": 17729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5174310207366943,
+      "learning_rate": 0.00048414389125748977,
+      "loss": 3.2467,
+      "step": 17730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.503591001033783,
+      "learning_rate": 0.00048413173646971226,
+      "loss": 3.1232,
+      "step": 17731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5538753271102905,
+      "learning_rate": 0.00048411958119696716,
+      "loss": 2.9704,
+      "step": 17732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5586168766021729,
+      "learning_rate": 0.0004841074254392864,
+      "loss": 3.0653,
+      "step": 17733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5222452282905579,
+      "learning_rate": 0.0004840952691967022,
+      "loss": 3.0688,
+      "step": 17734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5168008804321289,
+      "learning_rate": 0.0004840831124692465,
+      "loss": 3.0731,
+      "step": 17735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4989527761936188,
+      "learning_rate": 0.00048407095525695125,
+      "loss": 3.0839,
+      "step": 17736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5122324824333191,
+      "learning_rate": 0.0004840587975598486,
+      "loss": 3.1072,
+      "step": 17737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5376427173614502,
+      "learning_rate": 0.0004840466393779704,
+      "loss": 3.2577,
+      "step": 17738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5357086062431335,
+      "learning_rate": 0.00048403448071134887,
+      "loss": 3.2627,
+      "step": 17739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.530827522277832,
+      "learning_rate": 0.000484022321560016,
+      "loss": 3.1058,
+      "step": 17740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5042960047721863,
+      "learning_rate": 0.0004840101619240036,
+      "loss": 3.2674,
+      "step": 17741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5273114442825317,
+      "learning_rate": 0.00048399800180334396,
+      "loss": 3.2993,
+      "step": 17742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5568901300430298,
+      "learning_rate": 0.0004839858411980689,
+      "loss": 3.0452,
+      "step": 17743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5302985906600952,
+      "learning_rate": 0.0004839736801082106,
+      "loss": 3.0976,
+      "step": 17744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5017320513725281,
+      "learning_rate": 0.00048396151853380106,
+      "loss": 3.0252,
+      "step": 17745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5245184302330017,
+      "learning_rate": 0.00048394935647487226,
+      "loss": 2.9773,
+      "step": 17746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5124706029891968,
+      "learning_rate": 0.00048393719393145617,
+      "loss": 3.0761,
+      "step": 17747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5464129447937012,
+      "learning_rate": 0.000483925030903585,
+      "loss": 3.1286,
+      "step": 17748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5206067562103271,
+      "learning_rate": 0.0004839128673912907,
+      "loss": 3.2381,
+      "step": 17749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5206215381622314,
+      "learning_rate": 0.00048390070339460526,
+      "loss": 3.0297,
+      "step": 17750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5299220085144043,
+      "learning_rate": 0.0004838885389135608,
+      "loss": 3.2162,
+      "step": 17751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5235497355461121,
+      "learning_rate": 0.00048387637394818925,
+      "loss": 3.0847,
+      "step": 17752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5170539021492004,
+      "learning_rate": 0.0004838642084985228,
+      "loss": 3.1975,
+      "step": 17753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5105252265930176,
+      "learning_rate": 0.00048385204256459334,
+      "loss": 3.2036,
+      "step": 17754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5179150700569153,
+      "learning_rate": 0.00048383987614643303,
+      "loss": 3.1076,
+      "step": 17755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5759047865867615,
+      "learning_rate": 0.0004838277092440739,
+      "loss": 3.0669,
+      "step": 17756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4824429452419281,
+      "learning_rate": 0.0004838155418575479,
+      "loss": 3.1084,
+      "step": 17757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5309851765632629,
+      "learning_rate": 0.00048380337398688713,
+      "loss": 3.0484,
+      "step": 17758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5082951188087463,
+      "learning_rate": 0.00048379120563212365,
+      "loss": 3.0945,
+      "step": 17759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5423614382743835,
+      "learning_rate": 0.0004837790367932896,
+      "loss": 3.0284,
+      "step": 17760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.503147304058075,
+      "learning_rate": 0.00048376686747041684,
+      "loss": 3.1037,
+      "step": 17761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5271511077880859,
+      "learning_rate": 0.00048375469766353754,
+      "loss": 3.2318,
+      "step": 17762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5349052548408508,
+      "learning_rate": 0.0004837425273726838,
+      "loss": 3.2816,
+      "step": 17763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5517288446426392,
+      "learning_rate": 0.0004837303565978875,
+      "loss": 3.2532,
+      "step": 17764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5327991843223572,
+      "learning_rate": 0.00048371818533918075,
+      "loss": 2.9928,
+      "step": 17765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5061706304550171,
+      "learning_rate": 0.0004837060135965958,
+      "loss": 3.2314,
+      "step": 17766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5411424040794373,
+      "learning_rate": 0.00048369384137016456,
+      "loss": 3.1747,
+      "step": 17767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4953779876232147,
+      "learning_rate": 0.000483681668659919,
+      "loss": 3.2839,
+      "step": 17768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5827047824859619,
+      "learning_rate": 0.0004836694954658913,
+      "loss": 3.2348,
+      "step": 17769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5480215549468994,
+      "learning_rate": 0.00048365732178811354,
+      "loss": 3.1482,
+      "step": 17770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4860683083534241,
+      "learning_rate": 0.00048364514762661774,
+      "loss": 3.1271,
+      "step": 17771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.552074134349823,
+      "learning_rate": 0.0004836329729814359,
+      "loss": 3.3587,
+      "step": 17772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4946533441543579,
+      "learning_rate": 0.00048362079785260027,
+      "loss": 3.0403,
+      "step": 17773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5436502695083618,
+      "learning_rate": 0.00048360862224014267,
+      "loss": 3.1044,
+      "step": 17774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5207213163375854,
+      "learning_rate": 0.00048359644614409534,
+      "loss": 3.1488,
+      "step": 17775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5504875779151917,
+      "learning_rate": 0.00048358426956449026,
+      "loss": 3.2646,
+      "step": 17776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.517497718334198,
+      "learning_rate": 0.00048357209250135964,
+      "loss": 3.1974,
+      "step": 17777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5165881514549255,
+      "learning_rate": 0.00048355991495473545,
+      "loss": 2.9446,
+      "step": 17778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5115556716918945,
+      "learning_rate": 0.0004835477369246497,
+      "loss": 3.0173,
+      "step": 17779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5526302456855774,
+      "learning_rate": 0.00048353555841113455,
+      "loss": 3.1211,
+      "step": 17780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5087662935256958,
+      "learning_rate": 0.00048352337941422207,
+      "loss": 3.1776,
+      "step": 17781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5549590587615967,
+      "learning_rate": 0.0004835111999339444,
+      "loss": 3.0331,
+      "step": 17782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5352929830551147,
+      "learning_rate": 0.00048349901997033347,
+      "loss": 3.0673,
+      "step": 17783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5016866326332092,
+      "learning_rate": 0.00048348683952342136,
+      "loss": 3.2462,
+      "step": 17784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4875168800354004,
+      "learning_rate": 0.0004834746585932404,
+      "loss": 3.2957,
+      "step": 17785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5208585858345032,
+      "learning_rate": 0.0004834624771798224,
+      "loss": 3.3014,
+      "step": 17786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5304600596427917,
+      "learning_rate": 0.00048345029528319954,
+      "loss": 3.0373,
+      "step": 17787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5236985683441162,
+      "learning_rate": 0.00048343811290340395,
+      "loss": 3.3229,
+      "step": 17788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5062572360038757,
+      "learning_rate": 0.0004834259300404676,
+      "loss": 3.0878,
+      "step": 17789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5024277567863464,
+      "learning_rate": 0.00048341374669442274,
+      "loss": 3.1605,
+      "step": 17790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5190656781196594,
+      "learning_rate": 0.0004834015628653013,
+      "loss": 3.1117,
+      "step": 17791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.514510452747345,
+      "learning_rate": 0.0004833893785531355,
+      "loss": 2.8972,
+      "step": 17792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4951634705066681,
+      "learning_rate": 0.0004833771937579574,
+      "loss": 3.242,
+      "step": 17793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5369077324867249,
+      "learning_rate": 0.000483365008479799,
+      "loss": 3.4729,
+      "step": 17794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5123760104179382,
+      "learning_rate": 0.0004833528227186925,
+      "loss": 2.9924,
+      "step": 17795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5074111223220825,
+      "learning_rate": 0.00048334063647466986,
+      "loss": 3.329,
+      "step": 17796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5059533715248108,
+      "learning_rate": 0.0004833284497477634,
+      "loss": 3.0242,
+      "step": 17797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5089158415794373,
+      "learning_rate": 0.0004833162625380049,
+      "loss": 3.2417,
+      "step": 17798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5231832265853882,
+      "learning_rate": 0.0004833040748454268,
+      "loss": 3.1011,
+      "step": 17799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5030272006988525,
+      "learning_rate": 0.00048329188667006095,
+      "loss": 3.251,
+      "step": 17800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.516279399394989,
+      "learning_rate": 0.0004832796980119396,
+      "loss": 2.9739,
+      "step": 17801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4982314705848694,
+      "learning_rate": 0.0004832675088710948,
+      "loss": 3.1481,
+      "step": 17802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5156276822090149,
+      "learning_rate": 0.00048325531924755865,
+      "loss": 3.2038,
+      "step": 17803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5175352692604065,
+      "learning_rate": 0.0004832431291413633,
+      "loss": 3.0606,
+      "step": 17804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5309943556785583,
+      "learning_rate": 0.0004832309385525407,
+      "loss": 3.1735,
+      "step": 17805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5159737467765808,
+      "learning_rate": 0.00048321874748112316,
+      "loss": 2.9259,
+      "step": 17806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5461140275001526,
+      "learning_rate": 0.00048320655592714267,
+      "loss": 3.0094,
+      "step": 17807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5539809465408325,
+      "learning_rate": 0.00048319436389063144,
+      "loss": 3.1468,
+      "step": 17808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5527653694152832,
+      "learning_rate": 0.00048318217137162145,
+      "loss": 3.0324,
+      "step": 17809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5274596810340881,
+      "learning_rate": 0.00048316997837014486,
+      "loss": 3.2017,
+      "step": 17810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5476181507110596,
+      "learning_rate": 0.00048315778488623376,
+      "loss": 3.2789,
+      "step": 17811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5081568956375122,
+      "learning_rate": 0.0004831455909199204,
+      "loss": 3.1682,
+      "step": 17812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5192145109176636,
+      "learning_rate": 0.00048313339647123677,
+      "loss": 3.3376,
+      "step": 17813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5384969115257263,
+      "learning_rate": 0.00048312120154021495,
+      "loss": 3.1973,
+      "step": 17814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.543453574180603,
+      "learning_rate": 0.00048310900612688726,
+      "loss": 3.2764,
+      "step": 17815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53953617811203,
+      "learning_rate": 0.00048309681023128557,
+      "loss": 2.9113,
+      "step": 17816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5175008177757263,
+      "learning_rate": 0.00048308461385344214,
+      "loss": 3.3165,
+      "step": 17817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5362492799758911,
+      "learning_rate": 0.0004830724169933891,
+      "loss": 3.1817,
+      "step": 17818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5375868082046509,
+      "learning_rate": 0.0004830602196511586,
+      "loss": 3.4691,
+      "step": 17819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.486717164516449,
+      "learning_rate": 0.0004830480218267826,
+      "loss": 3.0631,
+      "step": 17820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5145992040634155,
+      "learning_rate": 0.00048303582352029345,
+      "loss": 3.1301,
+      "step": 17821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5075737833976746,
+      "learning_rate": 0.00048302362473172307,
+      "loss": 3.2556,
+      "step": 17822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5056702494621277,
+      "learning_rate": 0.0004830114254611037,
+      "loss": 3.0136,
+      "step": 17823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5289957523345947,
+      "learning_rate": 0.00048299922570846756,
+      "loss": 3.0611,
+      "step": 17824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5203523635864258,
+      "learning_rate": 0.00048298702547384655,
+      "loss": 3.1454,
+      "step": 17825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5633178949356079,
+      "learning_rate": 0.00048297482475727295,
+      "loss": 2.9976,
+      "step": 17826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5474057197570801,
+      "learning_rate": 0.00048296262355877897,
+      "loss": 3.1389,
+      "step": 17827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5257229208946228,
+      "learning_rate": 0.0004829504218783966,
+      "loss": 3.1458,
+      "step": 17828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.519266664981842,
+      "learning_rate": 0.000482938219716158,
+      "loss": 3.0224,
+      "step": 17829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5056769251823425,
+      "learning_rate": 0.0004829260170720953,
+      "loss": 3.2842,
+      "step": 17830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5659570693969727,
+      "learning_rate": 0.0004829138139462408,
+      "loss": 2.9025,
+      "step": 17831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4957529306411743,
+      "learning_rate": 0.00048290161033862636,
+      "loss": 3.181,
+      "step": 17832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5560140609741211,
+      "learning_rate": 0.0004828894062492844,
+      "loss": 3.1902,
+      "step": 17833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5515144467353821,
+      "learning_rate": 0.00048287720167824696,
+      "loss": 2.9447,
+      "step": 17834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5933049917221069,
+      "learning_rate": 0.00048286499662554604,
+      "loss": 3.0005,
+      "step": 17835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5094730854034424,
+      "learning_rate": 0.000482852791091214,
+      "loss": 3.3128,
+      "step": 17836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4971897304058075,
+      "learning_rate": 0.0004828405850752829,
+      "loss": 3.2185,
+      "step": 17837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5200973749160767,
+      "learning_rate": 0.0004828283785777848,
+      "loss": 3.1677,
+      "step": 17838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5161073207855225,
+      "learning_rate": 0.00048281617159875203,
+      "loss": 3.3305,
+      "step": 17839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5352292656898499,
+      "learning_rate": 0.0004828039641382167,
+      "loss": 3.2175,
+      "step": 17840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49615222215652466,
+      "learning_rate": 0.00048279175619621073,
+      "loss": 3.2629,
+      "step": 17841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5312261581420898,
+      "learning_rate": 0.0004827795477727666,
+      "loss": 3.0147,
+      "step": 17842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5360621809959412,
+      "learning_rate": 0.0004827673388679163,
+      "loss": 3.0288,
+      "step": 17843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5203635096549988,
+      "learning_rate": 0.000482755129481692,
+      "loss": 3.1102,
+      "step": 17844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.522244930267334,
+      "learning_rate": 0.0004827429196141259,
+      "loss": 3.4478,
+      "step": 17845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5168775916099548,
+      "learning_rate": 0.00048273070926525,
+      "loss": 3.2777,
+      "step": 17846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5495973229408264,
+      "learning_rate": 0.0004827184984350966,
+      "loss": 3.2903,
+      "step": 17847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.537128746509552,
+      "learning_rate": 0.0004827062871236979,
+      "loss": 3.3552,
+      "step": 17848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5331090092658997,
+      "learning_rate": 0.00048269407533108597,
+      "loss": 3.1168,
+      "step": 17849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4784494936466217,
+      "learning_rate": 0.00048268186305729305,
+      "loss": 3.2552,
+      "step": 17850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5230741500854492,
+      "learning_rate": 0.00048266965030235116,
+      "loss": 3.164,
+      "step": 17851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4925912320613861,
+      "learning_rate": 0.0004826574370662927,
+      "loss": 3.0437,
+      "step": 17852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48744553327560425,
+      "learning_rate": 0.00048264522334914964,
+      "loss": 3.0296,
+      "step": 17853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4946111738681793,
+      "learning_rate": 0.0004826330091509542,
+      "loss": 2.9261,
+      "step": 17854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.523500919342041,
+      "learning_rate": 0.0004826207944717386,
+      "loss": 3.0822,
+      "step": 17855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5203326344490051,
+      "learning_rate": 0.00048260857931153487,
+      "loss": 3.2367,
+      "step": 17856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4974472224712372,
+      "learning_rate": 0.00048259636367037535,
+      "loss": 3.3098,
+      "step": 17857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4856162667274475,
+      "learning_rate": 0.00048258414754829226,
+      "loss": 3.4445,
+      "step": 17858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5150359869003296,
+      "learning_rate": 0.0004825719309453175,
+      "loss": 3.2439,
+      "step": 17859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5459046959877014,
+      "learning_rate": 0.00048255971386148346,
+      "loss": 3.2596,
+      "step": 17860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515584409236908,
+      "learning_rate": 0.0004825474962968223,
+      "loss": 3.2124,
+      "step": 17861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4975724518299103,
+      "learning_rate": 0.00048253527825136615,
+      "loss": 3.0945,
+      "step": 17862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5150234699249268,
+      "learning_rate": 0.00048252305972514725,
+      "loss": 2.9796,
+      "step": 17863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5973613262176514,
+      "learning_rate": 0.0004825108407181977,
+      "loss": 3.1476,
+      "step": 17864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.514706015586853,
+      "learning_rate": 0.0004824986212305497,
+      "loss": 2.8594,
+      "step": 17865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5519055128097534,
+      "learning_rate": 0.0004824864012622355,
+      "loss": 3.0926,
+      "step": 17866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5024924278259277,
+      "learning_rate": 0.00048247418081328724,
+      "loss": 3.3669,
+      "step": 17867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5211602449417114,
+      "learning_rate": 0.000482461959883737,
+      "loss": 3.0907,
+      "step": 17868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49125707149505615,
+      "learning_rate": 0.0004824497384736171,
+      "loss": 3.233,
+      "step": 17869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5268430709838867,
+      "learning_rate": 0.00048243751658295984,
+      "loss": 2.996,
+      "step": 17870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5411224961280823,
+      "learning_rate": 0.00048242529421179715,
+      "loss": 3.2701,
+      "step": 17871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.534550666809082,
+      "learning_rate": 0.00048241307136016133,
+      "loss": 3.0247,
+      "step": 17872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5058745741844177,
+      "learning_rate": 0.0004824008480280847,
+      "loss": 3.3005,
+      "step": 17873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5441546440124512,
+      "learning_rate": 0.00048238862421559923,
+      "loss": 3.0935,
+      "step": 17874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5148599147796631,
+      "learning_rate": 0.0004823763999227373,
+      "loss": 3.2805,
+      "step": 17875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5039719939231873,
+      "learning_rate": 0.00048236417514953094,
+      "loss": 3.2335,
+      "step": 17876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5283924341201782,
+      "learning_rate": 0.0004823519498960125,
+      "loss": 3.0868,
+      "step": 17877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5511729717254639,
+      "learning_rate": 0.00048233972416221417,
+      "loss": 3.0867,
+      "step": 17878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5116291642189026,
+      "learning_rate": 0.00048232749794816806,
+      "loss": 3.2757,
+      "step": 17879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5327066779136658,
+      "learning_rate": 0.00048231527125390636,
+      "loss": 3.0075,
+      "step": 17880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5154629945755005,
+      "learning_rate": 0.00048230304407946144,
+      "loss": 3.2752,
+      "step": 17881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5452733635902405,
+      "learning_rate": 0.00048229081642486523,
+      "loss": 3.2965,
+      "step": 17882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48976922035217285,
+      "learning_rate": 0.0004822785882901502,
+      "loss": 3.1155,
+      "step": 17883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5699204206466675,
+      "learning_rate": 0.00048226635967534834,
+      "loss": 3.2289,
+      "step": 17884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5284189581871033,
+      "learning_rate": 0.0004822541305804921,
+      "loss": 3.3774,
+      "step": 17885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48988762497901917,
+      "learning_rate": 0.00048224190100561355,
+      "loss": 3.0301,
+      "step": 17886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.496111124753952,
+      "learning_rate": 0.00048222967095074476,
+      "loss": 2.9562,
+      "step": 17887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.510610044002533,
+      "learning_rate": 0.0004822174404159182,
+      "loss": 3.4054,
+      "step": 17888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5512583255767822,
+      "learning_rate": 0.00048220520940116593,
+      "loss": 3.1477,
+      "step": 17889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4915429651737213,
+      "learning_rate": 0.00048219297790652024,
+      "loss": 3.1844,
+      "step": 17890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5592296719551086,
+      "learning_rate": 0.0004821807459320134,
+      "loss": 3.2548,
+      "step": 17891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49113866686820984,
+      "learning_rate": 0.0004821685134776773,
+      "loss": 3.1582,
+      "step": 17892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5161522626876831,
+      "learning_rate": 0.0004821562805435446,
+      "loss": 3.1451,
+      "step": 17893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5243754982948303,
+      "learning_rate": 0.00048214404712964713,
+      "loss": 3.3366,
+      "step": 17894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5705099701881409,
+      "learning_rate": 0.00048213181323601754,
+      "loss": 3.019,
+      "step": 17895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48698729276657104,
+      "learning_rate": 0.00048211957886268764,
+      "loss": 2.9851,
+      "step": 17896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5281487703323364,
+      "learning_rate": 0.0004821073440096898,
+      "loss": 3.1592,
+      "step": 17897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.524651288986206,
+      "learning_rate": 0.0004820951086770563,
+      "loss": 3.1948,
+      "step": 17898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5448842644691467,
+      "learning_rate": 0.0004820828728648194,
+      "loss": 3.1414,
+      "step": 17899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5146068930625916,
+      "learning_rate": 0.0004820706365730112,
+      "loss": 3.0062,
+      "step": 17900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49320217967033386,
+      "learning_rate": 0.000482058399801664,
+      "loss": 3.3239,
+      "step": 17901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5205990076065063,
+      "learning_rate": 0.00048204616255080997,
+      "loss": 3.2077,
+      "step": 17902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5180717706680298,
+      "learning_rate": 0.00048203392482048136,
+      "loss": 3.2312,
+      "step": 17903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5263290405273438,
+      "learning_rate": 0.0004820216866107105,
+      "loss": 3.2392,
+      "step": 17904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5504626631736755,
+      "learning_rate": 0.00048200944792152955,
+      "loss": 3.0838,
+      "step": 17905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5280824303627014,
+      "learning_rate": 0.0004819972087529707,
+      "loss": 3.2164,
+      "step": 17906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5233747959136963,
+      "learning_rate": 0.00048198496910506624,
+      "loss": 3.1294,
+      "step": 17907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5285931825637817,
+      "learning_rate": 0.00048197272897784835,
+      "loss": 3.0499,
+      "step": 17908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5670228600502014,
+      "learning_rate": 0.0004819604883713494,
+      "loss": 3.1822,
+      "step": 17909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5432270765304565,
+      "learning_rate": 0.0004819482472856015,
+      "loss": 3.041,
+      "step": 17910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5179724097251892,
+      "learning_rate": 0.0004819360057206369,
+      "loss": 3.327,
+      "step": 17911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5055398344993591,
+      "learning_rate": 0.0004819237636764879,
+      "loss": 3.2491,
+      "step": 17912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.531039834022522,
+      "learning_rate": 0.0004819115211531867,
+      "loss": 3.1858,
+      "step": 17913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5148627161979675,
+      "learning_rate": 0.00048189927815076565,
+      "loss": 3.1547,
+      "step": 17914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5544278621673584,
+      "learning_rate": 0.0004818870346692569,
+      "loss": 3.2301,
+      "step": 17915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.564547598361969,
+      "learning_rate": 0.00048187479070869267,
+      "loss": 3.0203,
+      "step": 17916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5364139080047607,
+      "learning_rate": 0.0004818625462691052,
+      "loss": 2.8439,
+      "step": 17917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5530861020088196,
+      "learning_rate": 0.00048185030135052676,
+      "loss": 3.1894,
+      "step": 17918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5109603404998779,
+      "learning_rate": 0.00048183805595298975,
+      "loss": 3.2271,
+      "step": 17919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49450230598449707,
+      "learning_rate": 0.0004818258100765262,
+      "loss": 3.1592,
+      "step": 17920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5180266499519348,
+      "learning_rate": 0.0004818135637211685,
+      "loss": 3.0048,
+      "step": 17921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5050731301307678,
+      "learning_rate": 0.00048180131688694883,
+      "loss": 3.2961,
+      "step": 17922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5055239200592041,
+      "learning_rate": 0.0004817890695738994,
+      "loss": 3.1692,
+      "step": 17923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5030365586280823,
+      "learning_rate": 0.00048177682178205273,
+      "loss": 3.1633,
+      "step": 17924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6658396124839783,
+      "learning_rate": 0.00048176457351144084,
+      "loss": 3.2588,
+      "step": 17925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5380383133888245,
+      "learning_rate": 0.000481752324762096,
+      "loss": 3.2116,
+      "step": 17926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.507267951965332,
+      "learning_rate": 0.00048174007553405056,
+      "loss": 3.279,
+      "step": 17927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5685015320777893,
+      "learning_rate": 0.0004817278258273366,
+      "loss": 3.0932,
+      "step": 17928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5036908984184265,
+      "learning_rate": 0.0004817155756419866,
+      "loss": 3.1301,
+      "step": 17929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.516451895236969,
+      "learning_rate": 0.0004817033249780328,
+      "loss": 2.9769,
+      "step": 17930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5260326862335205,
+      "learning_rate": 0.00048169107383550744,
+      "loss": 3.3973,
+      "step": 17931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5365659594535828,
+      "learning_rate": 0.0004816788222144427,
+      "loss": 3.1838,
+      "step": 17932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5538583397865295,
+      "learning_rate": 0.0004816665701148709,
+      "loss": 3.1503,
+      "step": 17933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5433288812637329,
+      "learning_rate": 0.00048165431753682434,
+      "loss": 3.3311,
+      "step": 17934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5256428718566895,
+      "learning_rate": 0.0004816420644803352,
+      "loss": 3.1285,
+      "step": 17935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5579695701599121,
+      "learning_rate": 0.0004816298109454359,
+      "loss": 3.0131,
+      "step": 17936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49737823009490967,
+      "learning_rate": 0.00048161755693215864,
+      "loss": 3.287,
+      "step": 17937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5029211640357971,
+      "learning_rate": 0.00048160530244053564,
+      "loss": 3.2576,
+      "step": 17938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5466650724411011,
+      "learning_rate": 0.0004815930474705992,
+      "loss": 2.9216,
+      "step": 17939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5182828307151794,
+      "learning_rate": 0.0004815807920223816,
+      "loss": 3.2743,
+      "step": 17940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5201794505119324,
+      "learning_rate": 0.00048156853609591525,
+      "loss": 3.2135,
+      "step": 17941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5152579545974731,
+      "learning_rate": 0.0004815562796912323,
+      "loss": 3.1489,
+      "step": 17942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5244562029838562,
+      "learning_rate": 0.00048154402280836504,
+      "loss": 3.016,
+      "step": 17943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5533185601234436,
+      "learning_rate": 0.00048153176544734575,
+      "loss": 3.2095,
+      "step": 17944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5288299918174744,
+      "learning_rate": 0.0004815195076082067,
+      "loss": 3.0927,
+      "step": 17945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5070801377296448,
+      "learning_rate": 0.00048150724929098027,
+      "loss": 3.1879,
+      "step": 17946
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5345847606658936,
+      "learning_rate": 0.0004814949904956986,
+      "loss": 3.2283,
+      "step": 17947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5151923298835754,
+      "learning_rate": 0.0004814827312223941,
+      "loss": 3.1727,
+      "step": 17948
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5357820987701416,
+      "learning_rate": 0.000481470471471099,
+      "loss": 3.2602,
+      "step": 17949
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4863514304161072,
+      "learning_rate": 0.00048145821124184556,
+      "loss": 3.1904,
+      "step": 17950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5032357573509216,
+      "learning_rate": 0.00048144595053466616,
+      "loss": 3.1701,
+      "step": 17951
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5471490025520325,
+      "learning_rate": 0.00048143368934959306,
+      "loss": 3.2434,
+      "step": 17952
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4980376362800598,
+      "learning_rate": 0.00048142142768665844,
+      "loss": 3.0547,
+      "step": 17953
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4976464509963989,
+      "learning_rate": 0.0004814091655458947,
+      "loss": 3.0471,
+      "step": 17954
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5372846722602844,
+      "learning_rate": 0.0004813969029273343,
+      "loss": 3.2855,
+      "step": 17955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5928505659103394,
+      "learning_rate": 0.00048138463983100926,
+      "loss": 3.3892,
+      "step": 17956
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5085715055465698,
+      "learning_rate": 0.00048137237625695207,
+      "loss": 3.1189,
+      "step": 17957
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5404731631278992,
+      "learning_rate": 0.00048136011220519486,
+      "loss": 3.259,
+      "step": 17958
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5489948987960815,
+      "learning_rate": 0.00048134784767577,
+      "loss": 3.0618,
+      "step": 17959
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5055735111236572,
+      "learning_rate": 0.0004813355826687099,
+      "loss": 3.149,
+      "step": 17960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5085407495498657,
+      "learning_rate": 0.00048132331718404663,
+      "loss": 3.2623,
+      "step": 17961
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.509974479675293,
+      "learning_rate": 0.0004813110512218127,
+      "loss": 3.0828,
+      "step": 17962
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5331966876983643,
+      "learning_rate": 0.00048129878478204047,
+      "loss": 3.1375,
+      "step": 17963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4968498945236206,
+      "learning_rate": 0.000481286517864762,
+      "loss": 3.2832,
+      "step": 17964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5317128896713257,
+      "learning_rate": 0.0004812742504700097,
+      "loss": 3.1265,
+      "step": 17965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5396853685379028,
+      "learning_rate": 0.000481261982597816,
+      "loss": 3.3518,
+      "step": 17966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5501961708068848,
+      "learning_rate": 0.00048124971424821315,
+      "loss": 3.2126,
+      "step": 17967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5294440388679504,
+      "learning_rate": 0.00048123744542123345,
+      "loss": 3.1612,
+      "step": 17968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5000289678573608,
+      "learning_rate": 0.0004812251761169091,
+      "loss": 3.4135,
+      "step": 17969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49727967381477356,
+      "learning_rate": 0.00048121290633527247,
+      "loss": 2.9709,
+      "step": 17970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5257933735847473,
+      "learning_rate": 0.0004812006360763561,
+      "loss": 3.5203,
+      "step": 17971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5392364263534546,
+      "learning_rate": 0.000481188365340192,
+      "loss": 3.163,
+      "step": 17972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5661436319351196,
+      "learning_rate": 0.0004811760941268127,
+      "loss": 3.0577,
+      "step": 17973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5415921211242676,
+      "learning_rate": 0.0004811638224362503,
+      "loss": 2.9468,
+      "step": 17974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5405935645103455,
+      "learning_rate": 0.0004811515502685374,
+      "loss": 3.1172,
+      "step": 17975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5455482006072998,
+      "learning_rate": 0.00048113927762370614,
+      "loss": 3.1385,
+      "step": 17976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5725964307785034,
+      "learning_rate": 0.00048112700450178884,
+      "loss": 3.0701,
+      "step": 17977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5180301070213318,
+      "learning_rate": 0.00048111473090281797,
+      "loss": 3.0,
+      "step": 17978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5263218879699707,
+      "learning_rate": 0.00048110245682682557,
+      "loss": 3.2949,
+      "step": 17979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5030066967010498,
+      "learning_rate": 0.00048109018227384434,
+      "loss": 3.3081,
+      "step": 17980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5428774952888489,
+      "learning_rate": 0.0004810779072439063,
+      "loss": 3.2514,
+      "step": 17981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5697459578514099,
+      "learning_rate": 0.0004810656317370439,
+      "loss": 3.3227,
+      "step": 17982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.522610068321228,
+      "learning_rate": 0.00048105335575328957,
+      "loss": 3.135,
+      "step": 17983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.529964804649353,
+      "learning_rate": 0.0004810410792926755,
+      "loss": 3.2755,
+      "step": 17984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5211076736450195,
+      "learning_rate": 0.00048102880235523405,
+      "loss": 3.3943,
+      "step": 17985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5373325347900391,
+      "learning_rate": 0.0004810165249409976,
+      "loss": 3.0734,
+      "step": 17986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.537948727607727,
+      "learning_rate": 0.00048100424704999845,
+      "loss": 3.2213,
+      "step": 17987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49827033281326294,
+      "learning_rate": 0.00048099196868226895,
+      "loss": 3.0853,
+      "step": 17988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6188706755638123,
+      "learning_rate": 0.0004809796898378414,
+      "loss": 3.1362,
+      "step": 17989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5543064475059509,
+      "learning_rate": 0.00048096741051674826,
+      "loss": 3.0382,
+      "step": 17990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49779966473579407,
+      "learning_rate": 0.00048095513071902174,
+      "loss": 3.1029,
+      "step": 17991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5499799251556396,
+      "learning_rate": 0.00048094285044469415,
+      "loss": 3.3016,
+      "step": 17992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5513309836387634,
+      "learning_rate": 0.00048093056969379807,
+      "loss": 3.2205,
+      "step": 17993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49735820293426514,
+      "learning_rate": 0.0004809182884663656,
+      "loss": 3.2093,
+      "step": 17994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5192501544952393,
+      "learning_rate": 0.00048090600676242923,
+      "loss": 3.1286,
+      "step": 17995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5403336882591248,
+      "learning_rate": 0.00048089372458202115,
+      "loss": 3.014,
+      "step": 17996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4965198338031769,
+      "learning_rate": 0.00048088144192517387,
+      "loss": 3.115,
+      "step": 17997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.957526683807373,
+      "learning_rate": 0.0004808691587919197,
+      "loss": 3.324,
+      "step": 17998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5187360048294067,
+      "learning_rate": 0.00048085687518229105,
+      "loss": 3.2005,
+      "step": 17999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4975668489933014,
+      "learning_rate": 0.0004808445910963201,
+      "loss": 3.3057,
+      "step": 18000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5819920897483826,
+      "learning_rate": 0.00048083230653403925,
+      "loss": 3.0514,
+      "step": 18001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546259880065918,
+      "learning_rate": 0.000480820021495481,
+      "loss": 2.7754,
+      "step": 18002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49003690481185913,
+      "learning_rate": 0.00048080773598067753,
+      "loss": 3.2775,
+      "step": 18003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515501081943512,
+      "learning_rate": 0.00048079544998966137,
+      "loss": 3.2949,
+      "step": 18004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5453561544418335,
+      "learning_rate": 0.0004807831635224647,
+      "loss": 3.0814,
+      "step": 18005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5018565058708191,
+      "learning_rate": 0.00048077087657912005,
+      "loss": 3.4001,
+      "step": 18006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5057216882705688,
+      "learning_rate": 0.00048075858915965966,
+      "loss": 3.1185,
+      "step": 18007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5177599191665649,
+      "learning_rate": 0.00048074630126411597,
+      "loss": 2.9411,
+      "step": 18008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5200800895690918,
+      "learning_rate": 0.00048073401289252133,
+      "loss": 3.33,
+      "step": 18009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5048688054084778,
+      "learning_rate": 0.000480721724044908,
+      "loss": 3.3512,
+      "step": 18010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5326023697853088,
+      "learning_rate": 0.0004807094347213085,
+      "loss": 3.1505,
+      "step": 18011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.536878228187561,
+      "learning_rate": 0.0004806971449217551,
+      "loss": 3.0033,
+      "step": 18012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5812287926673889,
+      "learning_rate": 0.0004806848546462802,
+      "loss": 2.9029,
+      "step": 18013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5413682460784912,
+      "learning_rate": 0.00048067256389491613,
+      "loss": 3.137,
+      "step": 18014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47277653217315674,
+      "learning_rate": 0.00048066027266769533,
+      "loss": 3.1877,
+      "step": 18015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5533841848373413,
+      "learning_rate": 0.0004806479809646501,
+      "loss": 3.2071,
+      "step": 18016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5669878721237183,
+      "learning_rate": 0.0004806356887858129,
+      "loss": 3.3474,
+      "step": 18017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5068527460098267,
+      "learning_rate": 0.0004806233961312161,
+      "loss": 3.3926,
+      "step": 18018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.706555187702179,
+      "learning_rate": 0.00048061110300089203,
+      "loss": 3.1562,
+      "step": 18019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5449730157852173,
+      "learning_rate": 0.00048059880939487295,
+      "loss": 3.3523,
+      "step": 18020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5296577215194702,
+      "learning_rate": 0.0004805865153131915,
+      "loss": 3.0907,
+      "step": 18021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5225890278816223,
+      "learning_rate": 0.0004805742207558799,
+      "loss": 3.0886,
+      "step": 18022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6007885336875916,
+      "learning_rate": 0.00048056192572297046,
+      "loss": 3.399,
+      "step": 18023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5413122177124023,
+      "learning_rate": 0.00048054963021449575,
+      "loss": 3.1492,
+      "step": 18024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5432742238044739,
+      "learning_rate": 0.00048053733423048797,
+      "loss": 3.166,
+      "step": 18025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5184715986251831,
+      "learning_rate": 0.0004805250377709797,
+      "loss": 2.9957,
+      "step": 18026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5569450259208679,
+      "learning_rate": 0.0004805127408360032,
+      "loss": 3.1505,
+      "step": 18027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5036641359329224,
+      "learning_rate": 0.00048050044342559087,
+      "loss": 3.0816,
+      "step": 18028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6360530257225037,
+      "learning_rate": 0.0004804881455397751,
+      "loss": 3.154,
+      "step": 18029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5100707411766052,
+      "learning_rate": 0.00048047584717858825,
+      "loss": 3.2289,
+      "step": 18030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5365529656410217,
+      "learning_rate": 0.00048046354834206277,
+      "loss": 3.1688,
+      "step": 18031
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5900627970695496,
+      "learning_rate": 0.0004804512490302311,
+      "loss": 3.2375,
+      "step": 18032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.566939651966095,
+      "learning_rate": 0.0004804389492431255,
+      "loss": 2.9624,
+      "step": 18033
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5332266688346863,
+      "learning_rate": 0.0004804266489807785,
+      "loss": 3.0906,
+      "step": 18034
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5751016139984131,
+      "learning_rate": 0.0004804143482432224,
+      "loss": 3.2817,
+      "step": 18035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6242155432701111,
+      "learning_rate": 0.0004804020470304896,
+      "loss": 3.0636,
+      "step": 18036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5122875571250916,
+      "learning_rate": 0.00048038974534261256,
+      "loss": 3.0294,
+      "step": 18037
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5618741512298584,
+      "learning_rate": 0.00048037744317962357,
+      "loss": 3.0718,
+      "step": 18038
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5436500310897827,
+      "learning_rate": 0.0004803651405415553,
+      "loss": 3.3318,
+      "step": 18039
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078718662261963,
+      "learning_rate": 0.0004803528374284398,
+      "loss": 3.3323,
+      "step": 18040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5190596580505371,
+      "learning_rate": 0.00048034053384030963,
+      "loss": 3.4991,
+      "step": 18041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5513627529144287,
+      "learning_rate": 0.0004803282297771973,
+      "loss": 3.1212,
+      "step": 18042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5310901403427124,
+      "learning_rate": 0.000480315925239135,
+      "loss": 3.0746,
+      "step": 18043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5999692678451538,
+      "learning_rate": 0.00048030362022615533,
+      "loss": 3.2624,
+      "step": 18044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49754953384399414,
+      "learning_rate": 0.00048029131473829065,
+      "loss": 3.2649,
+      "step": 18045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5372548699378967,
+      "learning_rate": 0.00048027900877557327,
+      "loss": 3.2262,
+      "step": 18046
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.528826117515564,
+      "learning_rate": 0.00048026670233803574,
+      "loss": 3.2068,
+      "step": 18047
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5041863322257996,
+      "learning_rate": 0.00048025439542571035,
+      "loss": 3.0409,
+      "step": 18048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5768441557884216,
+      "learning_rate": 0.00048024208803862964,
+      "loss": 3.2841,
+      "step": 18049
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5213032960891724,
+      "learning_rate": 0.00048022978017682596,
+      "loss": 3.2703,
+      "step": 18050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5481829643249512,
+      "learning_rate": 0.00048021747184033163,
+      "loss": 3.1394,
+      "step": 18051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5505427122116089,
+      "learning_rate": 0.00048020516302917923,
+      "loss": 3.4499,
+      "step": 18052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5292855501174927,
+      "learning_rate": 0.00048019285374340106,
+      "loss": 3.2116,
+      "step": 18053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5407589077949524,
+      "learning_rate": 0.00048018054398302966,
+      "loss": 3.3331,
+      "step": 18054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5229291915893555,
+      "learning_rate": 0.0004801682337480974,
+      "loss": 3.3401,
+      "step": 18055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47680187225341797,
+      "learning_rate": 0.00048015592303863653,
+      "loss": 2.9878,
+      "step": 18056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5591782331466675,
+      "learning_rate": 0.00048014361185467986,
+      "loss": 3.3692,
+      "step": 18057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5464686155319214,
+      "learning_rate": 0.0004801313001962594,
+      "loss": 3.2056,
+      "step": 18058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5384355187416077,
+      "learning_rate": 0.00048011898806340787,
+      "loss": 3.3496,
+      "step": 18059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5165776610374451,
+      "learning_rate": 0.00048010667545615753,
+      "loss": 3.3195,
+      "step": 18060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.489169716835022,
+      "learning_rate": 0.00048009436237454083,
+      "loss": 3.109,
+      "step": 18061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5217012763023376,
+      "learning_rate": 0.00048008204881859034,
+      "loss": 3.3279,
+      "step": 18062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5407010912895203,
+      "learning_rate": 0.00048006973478833837,
+      "loss": 3.1728,
+      "step": 18063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5479187369346619,
+      "learning_rate": 0.0004800574202838174,
+      "loss": 2.9849,
+      "step": 18064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5429320335388184,
+      "learning_rate": 0.00048004510530505977,
+      "loss": 3.1398,
+      "step": 18065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5046254992485046,
+      "learning_rate": 0.000480032789852098,
+      "loss": 3.1535,
+      "step": 18066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5253599882125854,
+      "learning_rate": 0.00048002047392496443,
+      "loss": 3.0859,
+      "step": 18067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5937367081642151,
+      "learning_rate": 0.0004800081575236917,
+      "loss": 3.4202,
+      "step": 18068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.541233479976654,
+      "learning_rate": 0.0004799958406483121,
+      "loss": 3.3945,
+      "step": 18069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5336980819702148,
+      "learning_rate": 0.00047998352329885815,
+      "loss": 3.2491,
+      "step": 18070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5044065713882446,
+      "learning_rate": 0.00047997120547536214,
+      "loss": 3.2804,
+      "step": 18071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4936935305595398,
+      "learning_rate": 0.0004799588871778566,
+      "loss": 3.3098,
+      "step": 18072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55002760887146,
+      "learning_rate": 0.0004799465684063741,
+      "loss": 2.9614,
+      "step": 18073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5033633708953857,
+      "learning_rate": 0.00047993424916094687,
+      "loss": 3.0059,
+      "step": 18074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5659985542297363,
+      "learning_rate": 0.00047992192944160746,
+      "loss": 3.0396,
+      "step": 18075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5208994150161743,
+      "learning_rate": 0.0004799096092483884,
+      "loss": 3.1537,
+      "step": 18076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5257906913757324,
+      "learning_rate": 0.00047989728858132194,
+      "loss": 3.0355,
+      "step": 18077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53220534324646,
+      "learning_rate": 0.0004798849674404407,
+      "loss": 3.109,
+      "step": 18078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49276986718177795,
+      "learning_rate": 0.0004798726458257771,
+      "loss": 3.0571,
+      "step": 18079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5161966681480408,
+      "learning_rate": 0.0004798603237373636,
+      "loss": 3.2301,
+      "step": 18080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5360630750656128,
+      "learning_rate": 0.0004798480011752325,
+      "loss": 2.9056,
+      "step": 18081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5495896339416504,
+      "learning_rate": 0.00047983567813941644,
+      "loss": 3.0623,
+      "step": 18082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5134875178337097,
+      "learning_rate": 0.00047982335462994785,
+      "loss": 3.2153,
+      "step": 18083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49886998534202576,
+      "learning_rate": 0.00047981103064685904,
+      "loss": 3.1122,
+      "step": 18084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390711426734924,
+      "learning_rate": 0.00047979870619018275,
+      "loss": 3.1859,
+      "step": 18085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5308481454849243,
+      "learning_rate": 0.00047978638125995113,
+      "loss": 3.2953,
+      "step": 18086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5197398662567139,
+      "learning_rate": 0.0004797740558561968,
+      "loss": 3.2586,
+      "step": 18087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5043891668319702,
+      "learning_rate": 0.0004797617299789522,
+      "loss": 3.2689,
+      "step": 18088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.573706865310669,
+      "learning_rate": 0.00047974940362824987,
+      "loss": 3.2213,
+      "step": 18089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5655125379562378,
+      "learning_rate": 0.00047973707680412224,
+      "loss": 3.0872,
+      "step": 18090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5690857768058777,
+      "learning_rate": 0.00047972474950660164,
+      "loss": 3.0915,
+      "step": 18091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5151222348213196,
+      "learning_rate": 0.00047971242173572065,
+      "loss": 3.0679,
+      "step": 18092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6133147478103638,
+      "learning_rate": 0.00047970009349151174,
+      "loss": 3.2749,
+      "step": 18093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5890400409698486,
+      "learning_rate": 0.0004796877647740074,
+      "loss": 3.0752,
+      "step": 18094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5378167629241943,
+      "learning_rate": 0.00047967543558324,
+      "loss": 3.2817,
+      "step": 18095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5156680941581726,
+      "learning_rate": 0.0004796631059192422,
+      "loss": 3.079,
+      "step": 18096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5856471657752991,
+      "learning_rate": 0.0004796507757820462,
+      "loss": 3.1464,
+      "step": 18097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5163522362709045,
+      "learning_rate": 0.00047963844517168473,
+      "loss": 3.284,
+      "step": 18098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546302199363708,
+      "learning_rate": 0.00047962611408819015,
+      "loss": 3.0596,
+      "step": 18099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5673893690109253,
+      "learning_rate": 0.00047961378253159496,
+      "loss": 3.1698,
+      "step": 18100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5053473711013794,
+      "learning_rate": 0.0004796014505019317,
+      "loss": 3.0218,
+      "step": 18101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5111680626869202,
+      "learning_rate": 0.0004795891179992326,
+      "loss": 3.1531,
+      "step": 18102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5182185769081116,
+      "learning_rate": 0.00047957678502353045,
+      "loss": 3.0609,
+      "step": 18103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5161014199256897,
+      "learning_rate": 0.0004795644515748576,
+      "loss": 2.9817,
+      "step": 18104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5604485273361206,
+      "learning_rate": 0.0004795521176532466,
+      "loss": 3.2267,
+      "step": 18105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5659258961677551,
+      "learning_rate": 0.00047953978325872976,
+      "loss": 3.256,
+      "step": 18106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5269966721534729,
+      "learning_rate": 0.00047952744839133973,
+      "loss": 3.1339,
+      "step": 18107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5653190612792969,
+      "learning_rate": 0.000479515113051109,
+      "loss": 3.0896,
+      "step": 18108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5687468647956848,
+      "learning_rate": 0.00047950277723806994,
+      "loss": 3.2164,
+      "step": 18109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48306307196617126,
+      "learning_rate": 0.00047949044095225524,
+      "loss": 3.1984,
+      "step": 18110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5953868627548218,
+      "learning_rate": 0.00047947810419369716,
+      "loss": 3.0912,
+      "step": 18111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5070611238479614,
+      "learning_rate": 0.0004794657669624283,
+      "loss": 3.1685,
+      "step": 18112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5517260432243347,
+      "learning_rate": 0.00047945342925848116,
+      "loss": 3.4973,
+      "step": 18113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5137361884117126,
+      "learning_rate": 0.00047944109108188817,
+      "loss": 3.3537,
+      "step": 18114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5287407636642456,
+      "learning_rate": 0.00047942875243268187,
+      "loss": 3.182,
+      "step": 18115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5217429399490356,
+      "learning_rate": 0.0004794164133108949,
+      "loss": 3.2714,
+      "step": 18116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48636454343795776,
+      "learning_rate": 0.00047940407371655956,
+      "loss": 3.1138,
+      "step": 18117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5578451156616211,
+      "learning_rate": 0.00047939173364970833,
+      "loss": 3.1668,
+      "step": 18118
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5369355082511902,
+      "learning_rate": 0.0004793793931103739,
+      "loss": 3.0913,
+      "step": 18119
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5531823635101318,
+      "learning_rate": 0.0004793670520985886,
+      "loss": 2.9706,
+      "step": 18120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5396066904067993,
+      "learning_rate": 0.0004793547106143851,
+      "loss": 3.2062,
+      "step": 18121
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5657452344894409,
+      "learning_rate": 0.00047934236865779576,
+      "loss": 3.1972,
+      "step": 18122
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5130107998847961,
+      "learning_rate": 0.0004793300262288531,
+      "loss": 3.2101,
+      "step": 18123
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5585821270942688,
+      "learning_rate": 0.00047931768332758976,
+      "loss": 3.2184,
+      "step": 18124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5192179679870605,
+      "learning_rate": 0.0004793053399540381,
+      "loss": 3.1912,
+      "step": 18125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5495747327804565,
+      "learning_rate": 0.00047929299610823065,
+      "loss": 3.2991,
+      "step": 18126
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5496625304222107,
+      "learning_rate": 0.0004792806517902,
+      "loss": 3.2042,
+      "step": 18127
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.531826913356781,
+      "learning_rate": 0.00047926830699997853,
+      "loss": 3.0819,
+      "step": 18128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5123338103294373,
+      "learning_rate": 0.00047925596173759895,
+      "loss": 2.9806,
+      "step": 18129
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.561798095703125,
+      "learning_rate": 0.0004792436160030936,
+      "loss": 3.012,
+      "step": 18130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5473758578300476,
+      "learning_rate": 0.0004792312697964951,
+      "loss": 3.167,
+      "step": 18131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5174449682235718,
+      "learning_rate": 0.000479218923117836,
+      "loss": 3.0154,
+      "step": 18132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5553672909736633,
+      "learning_rate": 0.0004792065759671486,
+      "loss": 3.0273,
+      "step": 18133
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5386341214179993,
+      "learning_rate": 0.0004791942283444656,
+      "loss": 3.0787,
+      "step": 18134
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4873228669166565,
+      "learning_rate": 0.0004791818802498195,
+      "loss": 3.0802,
+      "step": 18135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5535769462585449,
+      "learning_rate": 0.00047916953168324284,
+      "loss": 3.0712,
+      "step": 18136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5669405460357666,
+      "learning_rate": 0.0004791571826447681,
+      "loss": 2.7861,
+      "step": 18137
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5906660556793213,
+      "learning_rate": 0.0004791448331344278,
+      "loss": 3.2052,
+      "step": 18138
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5243983268737793,
+      "learning_rate": 0.0004791324831522545,
+      "loss": 3.0515,
+      "step": 18139
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5486578345298767,
+      "learning_rate": 0.00047912013269828073,
+      "loss": 3.2827,
+      "step": 18140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5488114953041077,
+      "learning_rate": 0.00047910778177253906,
+      "loss": 3.2188,
+      "step": 18141
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5279528498649597,
+      "learning_rate": 0.00047909543037506183,
+      "loss": 3.2853,
+      "step": 18142
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5320385694503784,
+      "learning_rate": 0.00047908307850588175,
+      "loss": 3.0498,
+      "step": 18143
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5765833258628845,
+      "learning_rate": 0.0004790707261650313,
+      "loss": 3.1843,
+      "step": 18144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5396232604980469,
+      "learning_rate": 0.0004790583733525431,
+      "loss": 3.358,
+      "step": 18145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5210521817207336,
+      "learning_rate": 0.00047904602006844957,
+      "loss": 3.2112,
+      "step": 18146
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5123770833015442,
+      "learning_rate": 0.00047903366631278323,
+      "loss": 2.9842,
+      "step": 18147
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5506312847137451,
+      "learning_rate": 0.00047902131208557667,
+      "loss": 3.1214,
+      "step": 18148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.511390745639801,
+      "learning_rate": 0.00047900895738686245,
+      "loss": 3.147,
+      "step": 18149
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.561761736869812,
+      "learning_rate": 0.0004789966022166732,
+      "loss": 3.2444,
+      "step": 18150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5431965589523315,
+      "learning_rate": 0.00047898424657504126,
+      "loss": 3.0698,
+      "step": 18151
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5159698128700256,
+      "learning_rate": 0.00047897189046199924,
+      "loss": 3.238,
+      "step": 18152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5606123208999634,
+      "learning_rate": 0.0004789595338775797,
+      "loss": 3.0589,
+      "step": 18153
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5363898873329163,
+      "learning_rate": 0.0004789471768218152,
+      "loss": 3.1029,
+      "step": 18154
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5742982029914856,
+      "learning_rate": 0.00047893481929473826,
+      "loss": 3.1311,
+      "step": 18155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5812978744506836,
+      "learning_rate": 0.00047892246129638147,
+      "loss": 3.1121,
+      "step": 18156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5430973768234253,
+      "learning_rate": 0.00047891010282677735,
+      "loss": 3.005,
+      "step": 18157
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.524570107460022,
+      "learning_rate": 0.00047889774388595847,
+      "loss": 3.1939,
+      "step": 18158
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5263261198997498,
+      "learning_rate": 0.0004788853844739573,
+      "loss": 3.114,
+      "step": 18159
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5268939137458801,
+      "learning_rate": 0.00047887302459080644,
+      "loss": 3.271,
+      "step": 18160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5581185817718506,
+      "learning_rate": 0.00047886066423653855,
+      "loss": 3.1335,
+      "step": 18161
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5777899026870728,
+      "learning_rate": 0.0004788483034111861,
+      "loss": 3.2539,
+      "step": 18162
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5424737334251404,
+      "learning_rate": 0.0004788359421147816,
+      "loss": 3.352,
+      "step": 18163
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5313791632652283,
+      "learning_rate": 0.0004788235803473576,
+      "loss": 3.2132,
+      "step": 18164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5410112738609314,
+      "learning_rate": 0.0004788112181089467,
+      "loss": 3.1526,
+      "step": 18165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5331111550331116,
+      "learning_rate": 0.0004787988553995815,
+      "loss": 3.1902,
+      "step": 18166
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5786098837852478,
+      "learning_rate": 0.00047878649221929455,
+      "loss": 3.2107,
+      "step": 18167
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5680441856384277,
+      "learning_rate": 0.00047877412856811834,
+      "loss": 3.223,
+      "step": 18168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5643728971481323,
+      "learning_rate": 0.0004787617644460855,
+      "loss": 3.0881,
+      "step": 18169
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.526891827583313,
+      "learning_rate": 0.0004787493998532286,
+      "loss": 2.929,
+      "step": 18170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5148626565933228,
+      "learning_rate": 0.00047873703478958015,
+      "loss": 3.1915,
+      "step": 18171
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.00047872466925517274,
+      "loss": 3.0245,
+      "step": 18172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5895242094993591,
+      "learning_rate": 0.000478712303250039,
+      "loss": 2.8557,
+      "step": 18173
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4982737898826599,
+      "learning_rate": 0.0004786999367742114,
+      "loss": 3.1632,
+      "step": 18174
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5337832570075989,
+      "learning_rate": 0.00047868756982772265,
+      "loss": 3.2289,
+      "step": 18175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5422765612602234,
+      "learning_rate": 0.0004786752024106051,
+      "loss": 3.305,
+      "step": 18176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5369231700897217,
+      "learning_rate": 0.0004786628345228915,
+      "loss": 3.1646,
+      "step": 18177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49044135212898254,
+      "learning_rate": 0.00047865046616461446,
+      "loss": 3.1306,
+      "step": 18178
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.520637571811676,
+      "learning_rate": 0.00047863809733580633,
+      "loss": 3.1229,
+      "step": 18179
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5776728987693787,
+      "learning_rate": 0.00047862572803649995,
+      "loss": 3.1429,
+      "step": 18180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5988162755966187,
+      "learning_rate": 0.0004786133582667277,
+      "loss": 3.1521,
+      "step": 18181
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5374999642372131,
+      "learning_rate": 0.00047860098802652234,
+      "loss": 3.1968,
+      "step": 18182
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5586443543434143,
+      "learning_rate": 0.0004785886173159163,
+      "loss": 3.1005,
+      "step": 18183
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.516861081123352,
+      "learning_rate": 0.00047857624613494216,
+      "loss": 3.184,
+      "step": 18184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49465465545654297,
+      "learning_rate": 0.0004785638744836326,
+      "loss": 3.3298,
+      "step": 18185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5254295468330383,
+      "learning_rate": 0.00047855150236202006,
+      "loss": 3.2069,
+      "step": 18186
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5072837471961975,
+      "learning_rate": 0.0004785391297701374,
+      "loss": 3.2564,
+      "step": 18187
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5069559812545776,
+      "learning_rate": 0.00047852675670801694,
+      "loss": 3.0861,
+      "step": 18188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.527495801448822,
+      "learning_rate": 0.0004785143831756913,
+      "loss": 3.0807,
+      "step": 18189
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4928951561450958,
+      "learning_rate": 0.00047850200917319325,
+      "loss": 3.0653,
+      "step": 18190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5371759533882141,
+      "learning_rate": 0.0004784896347005552,
+      "loss": 3.1952,
+      "step": 18191
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5376279354095459,
+      "learning_rate": 0.0004784772597578098,
+      "loss": 3.0298,
+      "step": 18192
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5507842302322388,
+      "learning_rate": 0.00047846488434498966,
+      "loss": 3.3163,
+      "step": 18193
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5731067657470703,
+      "learning_rate": 0.0004784525084621274,
+      "loss": 3.1126,
+      "step": 18194
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5198032855987549,
+      "learning_rate": 0.00047844013210925556,
+      "loss": 3.0856,
+      "step": 18195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5343044400215149,
+      "learning_rate": 0.0004784277552864067,
+      "loss": 3.3659,
+      "step": 18196
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49065908789634705,
+      "learning_rate": 0.00047841537799361345,
+      "loss": 3.1321,
+      "step": 18197
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.521992027759552,
+      "learning_rate": 0.0004784030002309085,
+      "loss": 3.2767,
+      "step": 18198
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5364126563072205,
+      "learning_rate": 0.0004783906219983244,
+      "loss": 3.3079,
+      "step": 18199
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5045154690742493,
+      "learning_rate": 0.0004783782432958937,
+      "loss": 3.2046,
+      "step": 18200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5102838277816772,
+      "learning_rate": 0.000478365864123649,
+      "loss": 3.0836,
+      "step": 18201
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5909935235977173,
+      "learning_rate": 0.000478353484481623,
+      "loss": 2.9596,
+      "step": 18202
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5541250705718994,
+      "learning_rate": 0.0004783411043698483,
+      "loss": 3.1793,
+      "step": 18203
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5354369878768921,
+      "learning_rate": 0.00047832872378835735,
+      "loss": 3.4151,
+      "step": 18204
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5579158067703247,
+      "learning_rate": 0.0004783163427371828,
+      "loss": 3.2288,
+      "step": 18205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5801526308059692,
+      "learning_rate": 0.0004783039612163575,
+      "loss": 3.1398,
+      "step": 18206
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5363543629646301,
+      "learning_rate": 0.00047829157922591375,
+      "loss": 3.3956,
+      "step": 18207
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5684135556221008,
+      "learning_rate": 0.00047827919676588437,
+      "loss": 3.2425,
+      "step": 18208
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5314901471138,
+      "learning_rate": 0.0004782668138363019,
+      "loss": 3.2665,
+      "step": 18209
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5224202871322632,
+      "learning_rate": 0.0004782544304371988,
+      "loss": 3.0523,
+      "step": 18210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5397300124168396,
+      "learning_rate": 0.00047824204656860794,
+      "loss": 3.1839,
+      "step": 18211
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5352901220321655,
+      "learning_rate": 0.0004782296622305619,
+      "loss": 3.2884,
+      "step": 18212
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5520905256271362,
+      "learning_rate": 0.0004782172774230932,
+      "loss": 3.0375,
+      "step": 18213
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5324362516403198,
+      "learning_rate": 0.0004782048921462344,
+      "loss": 3.2757,
+      "step": 18214
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5429298281669617,
+      "learning_rate": 0.00047819250640001833,
+      "loss": 3.2702,
+      "step": 18215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5223371386528015,
+      "learning_rate": 0.00047818012018447737,
+      "loss": 3.135,
+      "step": 18216
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5844689607620239,
+      "learning_rate": 0.0004781677334996443,
+      "loss": 3.1293,
+      "step": 18217
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.513873279094696,
+      "learning_rate": 0.00047815534634555187,
+      "loss": 3.1109,
+      "step": 18218
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.522002637386322,
+      "learning_rate": 0.00047814295872223233,
+      "loss": 3.249,
+      "step": 18219
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4845925569534302,
+      "learning_rate": 0.0004781305706297186,
+      "loss": 3.1046,
+      "step": 18220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5436228513717651,
+      "learning_rate": 0.00047811818206804324,
+      "loss": 3.1851,
+      "step": 18221
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4941715598106384,
+      "learning_rate": 0.0004781057930372389,
+      "loss": 3.3141,
+      "step": 18222
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5171759128570557,
+      "learning_rate": 0.0004780934035373381,
+      "loss": 3.3435,
+      "step": 18223
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5393630862236023,
+      "learning_rate": 0.0004780810135683736,
+      "loss": 3.1683,
+      "step": 18224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5018099546432495,
+      "learning_rate": 0.0004780686231303779,
+      "loss": 3.2249,
+      "step": 18225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5250770449638367,
+      "learning_rate": 0.0004780562322233839,
+      "loss": 3.3423,
+      "step": 18226
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5291233658790588,
+      "learning_rate": 0.0004780438408474239,
+      "loss": 3.191,
+      "step": 18227
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.521236002445221,
+      "learning_rate": 0.00047803144900253076,
+      "loss": 2.9488,
+      "step": 18228
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5012958645820618,
+      "learning_rate": 0.000478019056688737,
+      "loss": 3.3888,
+      "step": 18229
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5145010352134705,
+      "learning_rate": 0.0004780066639060753,
+      "loss": 3.2967,
+      "step": 18230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5077473521232605,
+      "learning_rate": 0.0004779942706545783,
+      "loss": 3.1262,
+      "step": 18231
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4998171627521515,
+      "learning_rate": 0.00047798187693427876,
+      "loss": 3.258,
+      "step": 18232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5533021688461304,
+      "learning_rate": 0.000477969482745209,
+      "loss": 3.0468,
+      "step": 18233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5420917868614197,
+      "learning_rate": 0.0004779570880874021,
+      "loss": 3.2048,
+      "step": 18234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5190141201019287,
+      "learning_rate": 0.0004779446929608904,
+      "loss": 3.2259,
+      "step": 18235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5503576993942261,
+      "learning_rate": 0.00047793229736570647,
+      "loss": 3.3452,
+      "step": 18236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4898211359977722,
+      "learning_rate": 0.0004779199013018833,
+      "loss": 3.0858,
+      "step": 18237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5298957824707031,
+      "learning_rate": 0.0004779075047694533,
+      "loss": 3.1858,
+      "step": 18238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5395136475563049,
+      "learning_rate": 0.00047789510776844916,
+      "loss": 3.1886,
+      "step": 18239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5132995247840881,
+      "learning_rate": 0.0004778827102989035,
+      "loss": 2.9574,
+      "step": 18240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5644649267196655,
+      "learning_rate": 0.0004778703123608491,
+      "loss": 3.2534,
+      "step": 18241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5382485389709473,
+      "learning_rate": 0.0004778579139543184,
+      "loss": 3.4383,
+      "step": 18242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.485053688287735,
+      "learning_rate": 0.0004778455150793444,
+      "loss": 3.2485,
+      "step": 18243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49913448095321655,
+      "learning_rate": 0.0004778331157359594,
+      "loss": 3.0871,
+      "step": 18244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.550160825252533,
+      "learning_rate": 0.0004778207159241962,
+      "loss": 3.0416,
+      "step": 18245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.507668137550354,
+      "learning_rate": 0.0004778083156440874,
+      "loss": 3.0399,
+      "step": 18246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5110635757446289,
+      "learning_rate": 0.00047779591489566583,
+      "loss": 2.992,
+      "step": 18247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5091555714607239,
+      "learning_rate": 0.000477783513678964,
+      "loss": 3.0346,
+      "step": 18248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5132473111152649,
+      "learning_rate": 0.00047777111199401464,
+      "loss": 3.1061,
+      "step": 18249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5340614914894104,
+      "learning_rate": 0.0004777587098408503,
+      "loss": 3.0863,
+      "step": 18250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5357064008712769,
+      "learning_rate": 0.0004777463072195037,
+      "loss": 3.3543,
+      "step": 18251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5139802098274231,
+      "learning_rate": 0.0004777339041300077,
+      "loss": 3.2554,
+      "step": 18252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5598965287208557,
+      "learning_rate": 0.0004777215005723947,
+      "loss": 2.9029,
+      "step": 18253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5134537220001221,
+      "learning_rate": 0.0004777090965466975,
+      "loss": 3.0175,
+      "step": 18254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6165127754211426,
+      "learning_rate": 0.0004776966920529487,
+      "loss": 3.15,
+      "step": 18255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5549294352531433,
+      "learning_rate": 0.00047768428709118094,
+      "loss": 3.2095,
+      "step": 18256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5079606771469116,
+      "learning_rate": 0.0004776718816614271,
+      "loss": 3.1343,
+      "step": 18257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5567631125450134,
+      "learning_rate": 0.0004776594757637197,
+      "loss": 3.0041,
+      "step": 18258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.528123140335083,
+      "learning_rate": 0.00047764706939809143,
+      "loss": 3.1465,
+      "step": 18259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5787830352783203,
+      "learning_rate": 0.000477634662564575,
+      "loss": 3.06,
+      "step": 18260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5511603355407715,
+      "learning_rate": 0.00047762225526320297,
+      "loss": 3.1724,
+      "step": 18261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5570658445358276,
+      "learning_rate": 0.00047760984749400806,
+      "loss": 3.219,
+      "step": 18262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5181900858879089,
+      "learning_rate": 0.00047759743925702313,
+      "loss": 3.1698,
+      "step": 18263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5297556519508362,
+      "learning_rate": 0.00047758503055228064,
+      "loss": 3.0683,
+      "step": 18264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5381805896759033,
+      "learning_rate": 0.0004775726213798134,
+      "loss": 3.1135,
+      "step": 18265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5791769027709961,
+      "learning_rate": 0.000477560211739654,
+      "loss": 3.2524,
+      "step": 18266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5343177914619446,
+      "learning_rate": 0.0004775478016318352,
+      "loss": 3.055,
+      "step": 18267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5313341021537781,
+      "learning_rate": 0.00047753539105638965,
+      "loss": 3.1583,
+      "step": 18268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5037796497344971,
+      "learning_rate": 0.00047752298001335007,
+      "loss": 3.2608,
+      "step": 18269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5314697027206421,
+      "learning_rate": 0.0004775105685027491,
+      "loss": 3.0916,
+      "step": 18270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5548476576805115,
+      "learning_rate": 0.0004774981565246195,
+      "loss": 3.1843,
+      "step": 18271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5024945735931396,
+      "learning_rate": 0.00047748574407899394,
+      "loss": 3.1704,
+      "step": 18272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5327395796775818,
+      "learning_rate": 0.000477473331165905,
+      "loss": 3.1438,
+      "step": 18273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5295423269271851,
+      "learning_rate": 0.00047746091778538553,
+      "loss": 3.0959,
+      "step": 18274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5354325771331787,
+      "learning_rate": 0.0004774485039374681,
+      "loss": 3.1872,
+      "step": 18275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5765413641929626,
+      "learning_rate": 0.0004774360896221855,
+      "loss": 3.0475,
+      "step": 18276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5822243690490723,
+      "learning_rate": 0.0004774236748395704,
+      "loss": 3.0958,
+      "step": 18277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5113519430160522,
+      "learning_rate": 0.0004774112595896554,
+      "loss": 3.2856,
+      "step": 18278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5153071284294128,
+      "learning_rate": 0.00047739884387247334,
+      "loss": 3.3148,
+      "step": 18279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5023435354232788,
+      "learning_rate": 0.0004773864276880569,
+      "loss": 3.1991,
+      "step": 18280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5883546471595764,
+      "learning_rate": 0.00047737401103643866,
+      "loss": 3.0173,
+      "step": 18281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51036536693573,
+      "learning_rate": 0.0004773615939176515,
+      "loss": 3.1579,
+      "step": 18282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5233753323554993,
+      "learning_rate": 0.00047734917633172804,
+      "loss": 3.2266,
+      "step": 18283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5153245329856873,
+      "learning_rate": 0.00047733675827870087,
+      "loss": 3.3655,
+      "step": 18284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5131426453590393,
+      "learning_rate": 0.000477324339758603,
+      "loss": 3.2084,
+      "step": 18285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.532879650592804,
+      "learning_rate": 0.00047731192077146673,
+      "loss": 3.2726,
+      "step": 18286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.547607421875,
+      "learning_rate": 0.00047729950131732515,
+      "loss": 3.2178,
+      "step": 18287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5815809965133667,
+      "learning_rate": 0.0004772870813962107,
+      "loss": 3.0909,
+      "step": 18288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5434556603431702,
+      "learning_rate": 0.00047727466100815617,
+      "loss": 3.1988,
+      "step": 18289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5472313761711121,
+      "learning_rate": 0.0004772622401531944,
+      "loss": 3.2903,
+      "step": 18290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5081930160522461,
+      "learning_rate": 0.000477249818831358,
+      "loss": 3.1709,
+      "step": 18291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5110074281692505,
+      "learning_rate": 0.00047723739704267964,
+      "loss": 3.4036,
+      "step": 18292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390797257423401,
+      "learning_rate": 0.0004772249747871921,
+      "loss": 3.2653,
+      "step": 18293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5095894932746887,
+      "learning_rate": 0.00047721255206492814,
+      "loss": 3.3201,
+      "step": 18294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5310311913490295,
+      "learning_rate": 0.00047720012887592035,
+      "loss": 3.1775,
+      "step": 18295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5089913606643677,
+      "learning_rate": 0.0004771877052202016,
+      "loss": 3.095,
+      "step": 18296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5221785306930542,
+      "learning_rate": 0.0004771752810978044,
+      "loss": 2.9333,
+      "step": 18297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5876374244689941,
+      "learning_rate": 0.0004771628565087617,
+      "loss": 3.179,
+      "step": 18298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5047369599342346,
+      "learning_rate": 0.0004771504314531061,
+      "loss": 3.1365,
+      "step": 18299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5240013003349304,
+      "learning_rate": 0.0004771380059308705,
+      "loss": 3.3203,
+      "step": 18300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5100141763687134,
+      "learning_rate": 0.0004771255799420873,
+      "loss": 3.1165,
+      "step": 18301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5566077828407288,
+      "learning_rate": 0.0004771131534867894,
+      "loss": 3.1252,
+      "step": 18302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5329804420471191,
+      "learning_rate": 0.00047710072656500965,
+      "loss": 3.0706,
+      "step": 18303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5710445642471313,
+      "learning_rate": 0.00047708829917678065,
+      "loss": 3.086,
+      "step": 18304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5329696536064148,
+      "learning_rate": 0.00047707587132213514,
+      "loss": 3.0563,
+      "step": 18305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5192325115203857,
+      "learning_rate": 0.00047706344300110586,
+      "loss": 3.245,
+      "step": 18306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5296958088874817,
+      "learning_rate": 0.00047705101421372556,
+      "loss": 3.0319,
+      "step": 18307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5326006412506104,
+      "learning_rate": 0.00047703858496002697,
+      "loss": 3.1762,
+      "step": 18308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5101357698440552,
+      "learning_rate": 0.0004770261552400428,
+      "loss": 3.0583,
+      "step": 18309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5171705484390259,
+      "learning_rate": 0.0004770137250538058,
+      "loss": 3.0426,
+      "step": 18310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5427629351615906,
+      "learning_rate": 0.0004770012944013487,
+      "loss": 3.1091,
+      "step": 18311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5223405957221985,
+      "learning_rate": 0.0004769888632827043,
+      "loss": 3.0479,
+      "step": 18312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5293853282928467,
+      "learning_rate": 0.0004769764316979052,
+      "loss": 3.3125,
+      "step": 18313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5013025999069214,
+      "learning_rate": 0.00047696399964698434,
+      "loss": 3.0253,
+      "step": 18314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5177825093269348,
+      "learning_rate": 0.0004769515671299743,
+      "loss": 3.0815,
+      "step": 18315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5152139663696289,
+      "learning_rate": 0.00047693913414690795,
+      "loss": 3.1761,
+      "step": 18316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5001190304756165,
+      "learning_rate": 0.000476926700697818,
+      "loss": 3.0974,
+      "step": 18317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5169618129730225,
+      "learning_rate": 0.00047691426678273706,
+      "loss": 3.0948,
+      "step": 18318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49277180433273315,
+      "learning_rate": 0.00047690183240169803,
+      "loss": 2.9982,
+      "step": 18319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.536400318145752,
+      "learning_rate": 0.00047688939755473363,
+      "loss": 3.2277,
+      "step": 18320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4991290271282196,
+      "learning_rate": 0.0004768769622418766,
+      "loss": 3.0983,
+      "step": 18321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6011770963668823,
+      "learning_rate": 0.0004768645264631597,
+      "loss": 3.1713,
+      "step": 18322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4987938404083252,
+      "learning_rate": 0.00047685209021861567,
+      "loss": 3.1428,
+      "step": 18323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.507752537727356,
+      "learning_rate": 0.0004768396535082773,
+      "loss": 3.2872,
+      "step": 18324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5121263265609741,
+      "learning_rate": 0.0004768272163321772,
+      "loss": 2.9893,
+      "step": 18325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5051730275154114,
+      "learning_rate": 0.00047681477869034836,
+      "loss": 3.1656,
+      "step": 18326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5293095111846924,
+      "learning_rate": 0.0004768023405828235,
+      "loss": 2.9227,
+      "step": 18327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5156005620956421,
+      "learning_rate": 0.0004767899020096351,
+      "loss": 3.2264,
+      "step": 18328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5265857577323914,
+      "learning_rate": 0.0004767774629708162,
+      "loss": 3.1721,
+      "step": 18329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5211267471313477,
+      "learning_rate": 0.0004767650234663995,
+      "loss": 3.0617,
+      "step": 18330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5150598883628845,
+      "learning_rate": 0.00047675258349641773,
+      "loss": 3.2196,
+      "step": 18331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5152638554573059,
+      "learning_rate": 0.00047674014306090375,
+      "loss": 3.3045,
+      "step": 18332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5148764252662659,
+      "learning_rate": 0.0004767277021598901,
+      "loss": 3.0967,
+      "step": 18333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5162348747253418,
+      "learning_rate": 0.0004767152607934098,
+      "loss": 2.8442,
+      "step": 18334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5254623889923096,
+      "learning_rate": 0.00047670281896149553,
+      "loss": 3.1644,
+      "step": 18335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5040985345840454,
+      "learning_rate": 0.00047669037666418,
+      "loss": 2.972,
+      "step": 18336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5320616960525513,
+      "learning_rate": 0.00047667793390149607,
+      "loss": 3.3935,
+      "step": 18337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5272945761680603,
+      "learning_rate": 0.0004766654906734764,
+      "loss": 3.1817,
+      "step": 18338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5453786253929138,
+      "learning_rate": 0.0004766530469801538,
+      "loss": 3.1222,
+      "step": 18339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5060163140296936,
+      "learning_rate": 0.00047664060282156113,
+      "loss": 3.1181,
+      "step": 18340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5192004442214966,
+      "learning_rate": 0.00047662815819773106,
+      "loss": 3.4064,
+      "step": 18341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5070635080337524,
+      "learning_rate": 0.00047661571310869655,
+      "loss": 3.1392,
+      "step": 18342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5369440317153931,
+      "learning_rate": 0.0004766032675544901,
+      "loss": 3.0245,
+      "step": 18343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5121726393699646,
+      "learning_rate": 0.0004765908215351446,
+      "loss": 3.2891,
+      "step": 18344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5277031064033508,
+      "learning_rate": 0.000476578375050693,
+      "loss": 3.0794,
+      "step": 18345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5574108958244324,
+      "learning_rate": 0.00047656592810116794,
+      "loss": 2.8832,
+      "step": 18346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5462555885314941,
+      "learning_rate": 0.0004765534806866021,
+      "loss": 3.0332,
+      "step": 18347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5030328035354614,
+      "learning_rate": 0.0004765410328070284,
+      "loss": 3.347,
+      "step": 18348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5085557103157043,
+      "learning_rate": 0.00047652858446247954,
+      "loss": 3.1421,
+      "step": 18349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5109598636627197,
+      "learning_rate": 0.00047651613565298853,
+      "loss": 3.2073,
+      "step": 18350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5085839033126831,
+      "learning_rate": 0.00047650368637858784,
+      "loss": 3.1515,
+      "step": 18351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.524103581905365,
+      "learning_rate": 0.00047649123663931047,
+      "loss": 3.1561,
+      "step": 18352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5291694402694702,
+      "learning_rate": 0.0004764787864351892,
+      "loss": 3.1722,
+      "step": 18353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.590645432472229,
+      "learning_rate": 0.0004764663357662567,
+      "loss": 3.1482,
+      "step": 18354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5344750285148621,
+      "learning_rate": 0.0004764538846325458,
+      "loss": 3.1071,
+      "step": 18355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5120199918746948,
+      "learning_rate": 0.00047644143303408935,
+      "loss": 3.0998,
+      "step": 18356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5367302894592285,
+      "learning_rate": 0.00047642898097092015,
+      "loss": 3.2283,
+      "step": 18357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5890613794326782,
+      "learning_rate": 0.000476416528443071,
+      "loss": 3.1638,
+      "step": 18358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5009247660636902,
+      "learning_rate": 0.0004764040754505746,
+      "loss": 3.2533,
+      "step": 18359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.528739333152771,
+      "learning_rate": 0.00047639162199346384,
+      "loss": 3.0933,
+      "step": 18360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5243030786514282,
+      "learning_rate": 0.00047637916807177153,
+      "loss": 3.1643,
+      "step": 18361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5276967287063599,
+      "learning_rate": 0.0004763667136855304,
+      "loss": 3.2262,
+      "step": 18362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5526394844055176,
+      "learning_rate": 0.0004763542588347733,
+      "loss": 3.0935,
+      "step": 18363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5326481461524963,
+      "learning_rate": 0.00047634180351953295,
+      "loss": 3.1979,
+      "step": 18364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5490707159042358,
+      "learning_rate": 0.0004763293477398423,
+      "loss": 3.137,
+      "step": 18365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5224721431732178,
+      "learning_rate": 0.0004763168914957341,
+      "loss": 3.255,
+      "step": 18366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5512878894805908,
+      "learning_rate": 0.0004763044347872412,
+      "loss": 2.9361,
+      "step": 18367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5557450652122498,
+      "learning_rate": 0.00047629197761439617,
+      "loss": 3.1725,
+      "step": 18368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5146492719650269,
+      "learning_rate": 0.0004762795199772321,
+      "loss": 3.3767,
+      "step": 18369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6718878746032715,
+      "learning_rate": 0.00047626706187578173,
+      "loss": 2.9582,
+      "step": 18370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5548587441444397,
+      "learning_rate": 0.0004762546033100778,
+      "loss": 3.1481,
+      "step": 18371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5026013255119324,
+      "learning_rate": 0.0004762421442801532,
+      "loss": 3.3895,
+      "step": 18372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5194470286369324,
+      "learning_rate": 0.00047622968478604064,
+      "loss": 3.5001,
+      "step": 18373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5245919823646545,
+      "learning_rate": 0.00047621722482777307,
+      "loss": 3.1301,
+      "step": 18374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5411098599433899,
+      "learning_rate": 0.00047620476440538316,
+      "loss": 3.0287,
+      "step": 18375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5193166732788086,
+      "learning_rate": 0.000476192303518904,
+      "loss": 3.1614,
+      "step": 18376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5470539927482605,
+      "learning_rate": 0.000476179842168368,
+      "loss": 3.184,
+      "step": 18377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5132575631141663,
+      "learning_rate": 0.0004761673803538084,
+      "loss": 3.0766,
+      "step": 18378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5734747648239136,
+      "learning_rate": 0.00047615491807525764,
+      "loss": 3.2,
+      "step": 18379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.536043107509613,
+      "learning_rate": 0.0004761424553327488,
+      "loss": 3.2023,
+      "step": 18380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5419288277626038,
+      "learning_rate": 0.00047612999212631464,
+      "loss": 3.2887,
+      "step": 18381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5045163631439209,
+      "learning_rate": 0.00047611752845598803,
+      "loss": 3.1443,
+      "step": 18382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5081565976142883,
+      "learning_rate": 0.0004761050643218017,
+      "loss": 3.0771,
+      "step": 18383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5471091866493225,
+      "learning_rate": 0.00047609259972378843,
+      "loss": 3.1064,
+      "step": 18384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5175535082817078,
+      "learning_rate": 0.00047608013466198125,
+      "loss": 3.1124,
+      "step": 18385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5380799174308777,
+      "learning_rate": 0.0004760676691364129,
+      "loss": 2.9918,
+      "step": 18386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5299926996231079,
+      "learning_rate": 0.0004760552031471162,
+      "loss": 3.3698,
+      "step": 18387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5114349722862244,
+      "learning_rate": 0.00047604273669412387,
+      "loss": 2.9142,
+      "step": 18388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5319020748138428,
+      "learning_rate": 0.000476030269777469,
+      "loss": 3.3346,
+      "step": 18389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5370768904685974,
+      "learning_rate": 0.0004760178023971841,
+      "loss": 3.076,
+      "step": 18390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5257225036621094,
+      "learning_rate": 0.0004760053345533023,
+      "loss": 3.0894,
+      "step": 18391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5191155076026917,
+      "learning_rate": 0.0004759928662458562,
+      "loss": 3.0241,
+      "step": 18392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5163692831993103,
+      "learning_rate": 0.0004759803974748789,
+      "loss": 3.0367,
+      "step": 18393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5202220678329468,
+      "learning_rate": 0.00047596792824040315,
+      "loss": 3.1828,
+      "step": 18394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5226185321807861,
+      "learning_rate": 0.0004759554585424616,
+      "loss": 3.1161,
+      "step": 18395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5085725784301758,
+      "learning_rate": 0.0004759429883810873,
+      "loss": 3.1653,
+      "step": 18396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5269706845283508,
+      "learning_rate": 0.000475930517756313,
+      "loss": 3.3381,
+      "step": 18397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5295675992965698,
+      "learning_rate": 0.00047591804666817164,
+      "loss": 3.1354,
+      "step": 18398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5254941582679749,
+      "learning_rate": 0.00047590557511669596,
+      "loss": 3.4621,
+      "step": 18399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4832818806171417,
+      "learning_rate": 0.00047589310310191873,
+      "loss": 3.0591,
+      "step": 18400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5161293745040894,
+      "learning_rate": 0.0004758806306238731,
+      "loss": 3.2809,
+      "step": 18401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5065767168998718,
+      "learning_rate": 0.0004758681576825916,
+      "loss": 3.2523,
+      "step": 18402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48637884855270386,
+      "learning_rate": 0.0004758556842781074,
+      "loss": 3.2953,
+      "step": 18403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5562406778335571,
+      "learning_rate": 0.000475843210410453,
+      "loss": 3.0296,
+      "step": 18404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.553834855556488,
+      "learning_rate": 0.0004758307360796615,
+      "loss": 3.3277,
+      "step": 18405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5108713507652283,
+      "learning_rate": 0.00047581826128576557,
+      "loss": 3.1895,
+      "step": 18406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4815329909324646,
+      "learning_rate": 0.0004758057860287984,
+      "loss": 3.2556,
+      "step": 18407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5623984336853027,
+      "learning_rate": 0.00047579331030879246,
+      "loss": 3.0558,
+      "step": 18408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5089913010597229,
+      "learning_rate": 0.00047578083412578085,
+      "loss": 3.1773,
+      "step": 18409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5331538319587708,
+      "learning_rate": 0.00047576835747979626,
+      "loss": 3.1828,
+      "step": 18410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.519395112991333,
+      "learning_rate": 0.0004757558803708717,
+      "loss": 3.0746,
+      "step": 18411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.538934051990509,
+      "learning_rate": 0.00047574340279903993,
+      "loss": 3.1324,
+      "step": 18412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48608413338661194,
+      "learning_rate": 0.0004757309247643339,
+      "loss": 3.0911,
+      "step": 18413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5462820529937744,
+      "learning_rate": 0.0004757184462667865,
+      "loss": 3.2101,
+      "step": 18414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5184861421585083,
+      "learning_rate": 0.00047570596730643036,
+      "loss": 3.2584,
+      "step": 18415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5258368253707886,
+      "learning_rate": 0.0004756934878832987,
+      "loss": 3.1552,
+      "step": 18416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5133891701698303,
+      "learning_rate": 0.0004756810079974241,
+      "loss": 3.2167,
+      "step": 18417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49524182081222534,
+      "learning_rate": 0.0004756685276488396,
+      "loss": 2.9747,
+      "step": 18418
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5075833797454834,
+      "learning_rate": 0.000475656046837578,
+      "loss": 3.0269,
+      "step": 18419
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4994834065437317,
+      "learning_rate": 0.0004756435655636721,
+      "loss": 3.2545,
+      "step": 18420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5248388648033142,
+      "learning_rate": 0.00047563108382715487,
+      "loss": 3.1916,
+      "step": 18421
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5069358944892883,
+      "learning_rate": 0.00047561860162805925,
+      "loss": 3.0174,
+      "step": 18422
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5556639432907104,
+      "learning_rate": 0.0004756061189664179,
+      "loss": 3.0031,
+      "step": 18423
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49170437455177307,
+      "learning_rate": 0.00047559363584226394,
+      "loss": 3.2966,
+      "step": 18424
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5303202867507935,
+      "learning_rate": 0.0004755811522556301,
+      "loss": 2.9769,
+      "step": 18425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49278274178504944,
+      "learning_rate": 0.0004755686682065493,
+      "loss": 3.1306,
+      "step": 18426
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5985496044158936,
+      "learning_rate": 0.0004755561836950544,
+      "loss": 3.0919,
+      "step": 18427
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5184206962585449,
+      "learning_rate": 0.00047554369872117834,
+      "loss": 3.1546,
+      "step": 18428
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5128380656242371,
+      "learning_rate": 0.000475531213284954,
+      "loss": 3.2689,
+      "step": 18429
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5515463948249817,
+      "learning_rate": 0.0004755187273864141,
+      "loss": 3.2732,
+      "step": 18430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5561820864677429,
+      "learning_rate": 0.00047550624102559173,
+      "loss": 3.1093,
+      "step": 18431
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.610331654548645,
+      "learning_rate": 0.0004754937542025197,
+      "loss": 3.0545,
+      "step": 18432
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5147985219955444,
+      "learning_rate": 0.0004754812669172308,
+      "loss": 3.0679,
+      "step": 18433
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.553709864616394,
+      "learning_rate": 0.0004754687791697581,
+      "loss": 3.4146,
+      "step": 18434
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5066872239112854,
+      "learning_rate": 0.0004754562909601344,
+      "loss": 3.041,
+      "step": 18435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5329242944717407,
+      "learning_rate": 0.0004754438022883926,
+      "loss": 3.0274,
+      "step": 18436
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.570603609085083,
+      "learning_rate": 0.00047543131315456566,
+      "loss": 3.26,
+      "step": 18437
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5356221795082092,
+      "learning_rate": 0.0004754188235586863,
+      "loss": 3.054,
+      "step": 18438
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5116856694221497,
+      "learning_rate": 0.00047540633350078753,
+      "loss": 3.0638,
+      "step": 18439
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5043549537658691,
+      "learning_rate": 0.00047539384298090227,
+      "loss": 3.1825,
+      "step": 18440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4952833950519562,
+      "learning_rate": 0.00047538135199906334,
+      "loss": 3.3498,
+      "step": 18441
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5181792378425598,
+      "learning_rate": 0.0004753688605553037,
+      "loss": 3.1791,
+      "step": 18442
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5022110939025879,
+      "learning_rate": 0.0004753563686496562,
+      "loss": 3.0999,
+      "step": 18443
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5453217625617981,
+      "learning_rate": 0.0004753438762821539,
+      "loss": 3.1889,
+      "step": 18444
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5252522826194763,
+      "learning_rate": 0.0004753313834528294,
+      "loss": 3.2477,
+      "step": 18445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5274051427841187,
+      "learning_rate": 0.0004753188901617159,
+      "loss": 3.0772,
+      "step": 18446
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5190703272819519,
+      "learning_rate": 0.00047530639640884617,
+      "loss": 3.3103,
+      "step": 18447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5288125872612,
+      "learning_rate": 0.0004752939021942531,
+      "loss": 2.9752,
+      "step": 18448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5266156792640686,
+      "learning_rate": 0.0004752814075179696,
+      "loss": 3.1599,
+      "step": 18449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.506328284740448,
+      "learning_rate": 0.00047526891238002867,
+      "loss": 3.2425,
+      "step": 18450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5088960528373718,
+      "learning_rate": 0.00047525641678046313,
+      "loss": 3.3905,
+      "step": 18451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5209529995918274,
+      "learning_rate": 0.00047524392071930595,
+      "loss": 3.1773,
+      "step": 18452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.524338960647583,
+      "learning_rate": 0.0004752314241965899,
+      "loss": 3.1845,
+      "step": 18453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5339444279670715,
+      "learning_rate": 0.00047521892721234804,
+      "loss": 2.802,
+      "step": 18454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5499926805496216,
+      "learning_rate": 0.0004752064297666134,
+      "loss": 2.874,
+      "step": 18455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5348438620567322,
+      "learning_rate": 0.00047519393185941857,
+      "loss": 3.202,
+      "step": 18456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5232212543487549,
+      "learning_rate": 0.00047518143349079673,
+      "loss": 3.4002,
+      "step": 18457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4834090769290924,
+      "learning_rate": 0.00047516893466078067,
+      "loss": 3.2367,
+      "step": 18458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5394415855407715,
+      "learning_rate": 0.00047515643536940336,
+      "loss": 2.8562,
+      "step": 18459
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5132725238800049,
+      "learning_rate": 0.0004751439356166977,
+      "loss": 3.3076,
+      "step": 18460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5205427408218384,
+      "learning_rate": 0.00047513143540269665,
+      "loss": 3.3082,
+      "step": 18461
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4787410497665405,
+      "learning_rate": 0.00047511893472743306,
+      "loss": 3.0763,
+      "step": 18462
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5411839485168457,
+      "learning_rate": 0.00047510643359093983,
+      "loss": 3.0592,
+      "step": 18463
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5531505346298218,
+      "learning_rate": 0.0004750939319932501,
+      "loss": 3.062,
+      "step": 18464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52873295545578,
+      "learning_rate": 0.0004750814299343966,
+      "loss": 3.3097,
+      "step": 18465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5306869149208069,
+      "learning_rate": 0.00047506892741441234,
+      "loss": 3.046,
+      "step": 18466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5929498076438904,
+      "learning_rate": 0.0004750564244333302,
+      "loss": 2.9961,
+      "step": 18467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5276499390602112,
+      "learning_rate": 0.0004750439209911831,
+      "loss": 3.1753,
+      "step": 18468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5233253836631775,
+      "learning_rate": 0.00047503141708800394,
+      "loss": 3.1066,
+      "step": 18469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.527217447757721,
+      "learning_rate": 0.00047501891272382584,
+      "loss": 3.0618,
+      "step": 18470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5036908388137817,
+      "learning_rate": 0.00047500640789868156,
+      "loss": 2.9057,
+      "step": 18471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5078242421150208,
+      "learning_rate": 0.000474993902612604,
+      "loss": 3.4062,
+      "step": 18472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5183581113815308,
+      "learning_rate": 0.0004749813968656262,
+      "loss": 3.1168,
+      "step": 18473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5693612098693848,
+      "learning_rate": 0.0004749688906577812,
+      "loss": 2.7811,
+      "step": 18474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5193041563034058,
+      "learning_rate": 0.0004749563839891017,
+      "loss": 3.0456,
+      "step": 18475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5137484669685364,
+      "learning_rate": 0.0004749438768596208,
+      "loss": 3.1382,
+      "step": 18476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.536248505115509,
+      "learning_rate": 0.00047493136926937137,
+      "loss": 2.9656,
+      "step": 18477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5261755585670471,
+      "learning_rate": 0.0004749188612183865,
+      "loss": 3.4115,
+      "step": 18478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5276913642883301,
+      "learning_rate": 0.00047490635270669885,
+      "loss": 3.3476,
+      "step": 18479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5933910012245178,
+      "learning_rate": 0.00047489384373434156,
+      "loss": 3.1409,
+      "step": 18480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5715004205703735,
+      "learning_rate": 0.00047488133430134764,
+      "loss": 3.1825,
+      "step": 18481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5107777118682861,
+      "learning_rate": 0.0004748688244077499,
+      "loss": 3.1458,
+      "step": 18482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49393430352211,
+      "learning_rate": 0.00047485631405358127,
+      "loss": 3.0858,
+      "step": 18483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.514667272567749,
+      "learning_rate": 0.0004748438032388748,
+      "loss": 3.2898,
+      "step": 18484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5319417715072632,
+      "learning_rate": 0.00047483129196366336,
+      "loss": 3.3375,
+      "step": 18485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5158092975616455,
+      "learning_rate": 0.00047481878022798005,
+      "loss": 3.162,
+      "step": 18486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5055559277534485,
+      "learning_rate": 0.0004748062680318576,
+      "loss": 3.204,
+      "step": 18487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5113157033920288,
+      "learning_rate": 0.0004747937553753291,
+      "loss": 3.156,
+      "step": 18488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5367483496665955,
+      "learning_rate": 0.00047478124225842764,
+      "loss": 3.0957,
+      "step": 18489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5241609811782837,
+      "learning_rate": 0.0004747687286811859,
+      "loss": 3.2377,
+      "step": 18490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49339574575424194,
+      "learning_rate": 0.0004747562146436369,
+      "loss": 3.1479,
+      "step": 18491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.512546956539154,
+      "learning_rate": 0.0004747437001458138,
+      "loss": 3.2149,
+      "step": 18492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5302311182022095,
+      "learning_rate": 0.0004747311851877493,
+      "loss": 3.1386,
+      "step": 18493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5048165321350098,
+      "learning_rate": 0.0004747186697694766,
+      "loss": 3.132,
+      "step": 18494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5217496156692505,
+      "learning_rate": 0.0004747061538910285,
+      "loss": 3.2443,
+      "step": 18495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5357189774513245,
+      "learning_rate": 0.00047469363755243797,
+      "loss": 3.0927,
+      "step": 18496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5418717265129089,
+      "learning_rate": 0.00047468112075373797,
+      "loss": 3.3158,
+      "step": 18497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5489016771316528,
+      "learning_rate": 0.0004746686034949617,
+      "loss": 3.3727,
+      "step": 18498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5446333885192871,
+      "learning_rate": 0.0004746560857761418,
+      "loss": 3.3276,
+      "step": 18499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5174359679222107,
+      "learning_rate": 0.0004746435675973115,
+      "loss": 2.9588,
+      "step": 18500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5231800079345703,
+      "learning_rate": 0.0004746310489585036,
+      "loss": 3.1466,
+      "step": 18501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5035195350646973,
+      "learning_rate": 0.000474618529859751,
+      "loss": 2.9377,
+      "step": 18502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5154784917831421,
+      "learning_rate": 0.000474606010301087,
+      "loss": 3.1582,
+      "step": 18503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5330885052680969,
+      "learning_rate": 0.0004745934902825442,
+      "loss": 3.1804,
+      "step": 18504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4971103370189667,
+      "learning_rate": 0.0004745809698041558,
+      "loss": 3.2581,
+      "step": 18505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.521308422088623,
+      "learning_rate": 0.00047456844886595485,
+      "loss": 3.0899,
+      "step": 18506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5021144151687622,
+      "learning_rate": 0.00047455592746797404,
+      "loss": 3.2276,
+      "step": 18507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5008601546287537,
+      "learning_rate": 0.0004745434056102466,
+      "loss": 3.2421,
+      "step": 18508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5506516695022583,
+      "learning_rate": 0.0004745308832928054,
+      "loss": 3.3397,
+      "step": 18509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5164962410926819,
+      "learning_rate": 0.0004745183605156835,
+      "loss": 2.8647,
+      "step": 18510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6326226592063904,
+      "learning_rate": 0.00047450583727891375,
+      "loss": 3.2928,
+      "step": 18511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.507206380367279,
+      "learning_rate": 0.0004744933135825292,
+      "loss": 3.1631,
+      "step": 18512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5537012815475464,
+      "learning_rate": 0.00047448078942656285,
+      "loss": 3.2021,
+      "step": 18513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5219477415084839,
+      "learning_rate": 0.00047446826481104774,
+      "loss": 3.1808,
+      "step": 18514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5431067943572998,
+      "learning_rate": 0.00047445573973601673,
+      "loss": 3.2918,
+      "step": 18515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5809006690979004,
+      "learning_rate": 0.0004744432142015029,
+      "loss": 3.0758,
+      "step": 18516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5215701460838318,
+      "learning_rate": 0.0004744306882075392,
+      "loss": 3.1764,
+      "step": 18517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4926910102367401,
+      "learning_rate": 0.0004744181617541587,
+      "loss": 3.2759,
+      "step": 18518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390565991401672,
+      "learning_rate": 0.0004744056348413942,
+      "loss": 3.1481,
+      "step": 18519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.523172914981842,
+      "learning_rate": 0.00047439310746927896,
+      "loss": 3.0277,
+      "step": 18520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5034345984458923,
+      "learning_rate": 0.0004743805796378458,
+      "loss": 3.2691,
+      "step": 18521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5197492837905884,
+      "learning_rate": 0.00047436805134712774,
+      "loss": 3.1503,
+      "step": 18522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5116248726844788,
+      "learning_rate": 0.00047435552259715775,
+      "loss": 3.1957,
+      "step": 18523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5436865091323853,
+      "learning_rate": 0.0004743429933879689,
+      "loss": 3.2398,
+      "step": 18524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5367627739906311,
+      "learning_rate": 0.0004743304637195942,
+      "loss": 3.0797,
+      "step": 18525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49198198318481445,
+      "learning_rate": 0.0004743179335920666,
+      "loss": 3.2193,
+      "step": 18526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5412742495536804,
+      "learning_rate": 0.00047430540300541903,
+      "loss": 3.1761,
+      "step": 18527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5427865386009216,
+      "learning_rate": 0.0004742928719596846,
+      "loss": 3.1722,
+      "step": 18528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5101169347763062,
+      "learning_rate": 0.00047428034045489636,
+      "loss": 3.0649,
+      "step": 18529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5221566557884216,
+      "learning_rate": 0.00047426780849108723,
+      "loss": 3.0671,
+      "step": 18530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5130900144577026,
+      "learning_rate": 0.00047425527606829017,
+      "loss": 3.2507,
+      "step": 18531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.522251307964325,
+      "learning_rate": 0.0004742427431865383,
+      "loss": 3.1376,
+      "step": 18532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5170092582702637,
+      "learning_rate": 0.00047423020984586455,
+      "loss": 3.3573,
+      "step": 18533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5282794833183289,
+      "learning_rate": 0.0004742176760463019,
+      "loss": 3.0622,
+      "step": 18534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5373873114585876,
+      "learning_rate": 0.0004742051417878836,
+      "loss": 2.9846,
+      "step": 18535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5340383052825928,
+      "learning_rate": 0.0004741926070706423,
+      "loss": 3.3364,
+      "step": 18536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5768246650695801,
+      "learning_rate": 0.0004741800718946113,
+      "loss": 3.0228,
+      "step": 18537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.508990466594696,
+      "learning_rate": 0.0004741675362598235,
+      "loss": 3.1434,
+      "step": 18538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5513434410095215,
+      "learning_rate": 0.00047415500016631187,
+      "loss": 3.1098,
+      "step": 18539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5139069557189941,
+      "learning_rate": 0.00047414246361410955,
+      "loss": 3.2203,
+      "step": 18540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5511561632156372,
+      "learning_rate": 0.0004741299266032495,
+      "loss": 3.1658,
+      "step": 18541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5869661569595337,
+      "learning_rate": 0.0004741173891337647,
+      "loss": 2.964,
+      "step": 18542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5423870086669922,
+      "learning_rate": 0.0004741048512056882,
+      "loss": 3.1912,
+      "step": 18543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5338352918624878,
+      "learning_rate": 0.00047409231281905296,
+      "loss": 2.918,
+      "step": 18544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5400632619857788,
+      "learning_rate": 0.0004740797739738922,
+      "loss": 3.3985,
+      "step": 18545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515967607498169,
+      "learning_rate": 0.00047406723467023886,
+      "loss": 3.3357,
+      "step": 18546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5285216569900513,
+      "learning_rate": 0.0004740546949081258,
+      "loss": 3.2011,
+      "step": 18547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5364177227020264,
+      "learning_rate": 0.0004740421546875862,
+      "loss": 3.1159,
+      "step": 18548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5312500596046448,
+      "learning_rate": 0.000474029614008653,
+      "loss": 3.2883,
+      "step": 18549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5263493657112122,
+      "learning_rate": 0.0004740170728713594,
+      "loss": 3.2674,
+      "step": 18550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49179357290267944,
+      "learning_rate": 0.00047400453127573834,
+      "loss": 3.2257,
+      "step": 18551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.527242124080658,
+      "learning_rate": 0.0004739919892218228,
+      "loss": 3.1437,
+      "step": 18552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5284623503684998,
+      "learning_rate": 0.0004739794467096458,
+      "loss": 3.1903,
+      "step": 18553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5200530886650085,
+      "learning_rate": 0.0004739669037392404,
+      "loss": 3.1414,
+      "step": 18554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5152423977851868,
+      "learning_rate": 0.0004739543603106397,
+      "loss": 3.1332,
+      "step": 18555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5122321844100952,
+      "learning_rate": 0.0004739418164238766,
+      "loss": 3.2215,
+      "step": 18556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5178201198577881,
+      "learning_rate": 0.0004739292720789844,
+      "loss": 3.16,
+      "step": 18557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5422595143318176,
+      "learning_rate": 0.00047391672727599584,
+      "loss": 3.1908,
+      "step": 18558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5113765597343445,
+      "learning_rate": 0.0004739041820149441,
+      "loss": 3.1155,
+      "step": 18559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5253166556358337,
+      "learning_rate": 0.00047389163629586235,
+      "loss": 2.9544,
+      "step": 18560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5218266248703003,
+      "learning_rate": 0.00047387909011878333,
+      "loss": 2.8984,
+      "step": 18561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5512539744377136,
+      "learning_rate": 0.00047386654348374036,
+      "loss": 3.0539,
+      "step": 18562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5330885052680969,
+      "learning_rate": 0.0004738539963907664,
+      "loss": 3.091,
+      "step": 18563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5763328075408936,
+      "learning_rate": 0.0004738414488398943,
+      "loss": 3.1968,
+      "step": 18564
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5141280889511108,
+      "learning_rate": 0.0004738289008311575,
+      "loss": 3.1497,
+      "step": 18565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5118435025215149,
+      "learning_rate": 0.0004738163523645887,
+      "loss": 2.9679,
+      "step": 18566
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5282021760940552,
+      "learning_rate": 0.00047380380344022105,
+      "loss": 3.0492,
+      "step": 18567
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6659740805625916,
+      "learning_rate": 0.0004737912540580877,
+      "loss": 3.1675,
+      "step": 18568
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.520656168460846,
+      "learning_rate": 0.0004737787042182216,
+      "loss": 3.2612,
+      "step": 18569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.550905168056488,
+      "learning_rate": 0.0004737661539206558,
+      "loss": 3.0811,
+      "step": 18570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.532606840133667,
+      "learning_rate": 0.0004737536031654234,
+      "loss": 3.0482,
+      "step": 18571
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5544597506523132,
+      "learning_rate": 0.00047374105195255756,
+      "loss": 3.0428,
+      "step": 18572
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.534093976020813,
+      "learning_rate": 0.0004737285002820912,
+      "loss": 3.176,
+      "step": 18573
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5330844521522522,
+      "learning_rate": 0.00047371594815405726,
+      "loss": 3.0923,
+      "step": 18574
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.46944698691368103,
+      "learning_rate": 0.00047370339556848906,
+      "loss": 3.0567,
+      "step": 18575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5329000949859619,
+      "learning_rate": 0.0004736908425254195,
+      "loss": 3.0776,
+      "step": 18576
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5113318562507629,
+      "learning_rate": 0.0004736782890248818,
+      "loss": 3.0746,
+      "step": 18577
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.507371723651886,
+      "learning_rate": 0.0004736657350669088,
+      "loss": 3.1737,
+      "step": 18578
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5299968719482422,
+      "learning_rate": 0.0004736531806515337,
+      "loss": 3.0233,
+      "step": 18579
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5169973373413086,
+      "learning_rate": 0.0004736406257787895,
+      "loss": 3.1592,
+      "step": 18580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5072317719459534,
+      "learning_rate": 0.0004736280704487094,
+      "loss": 3.24,
+      "step": 18581
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.592707097530365,
+      "learning_rate": 0.00047361551466132634,
+      "loss": 3.1434,
+      "step": 18582
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5373207926750183,
+      "learning_rate": 0.00047360295841667344,
+      "loss": 3.1949,
+      "step": 18583
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.555252194404602,
+      "learning_rate": 0.0004735904017147838,
+      "loss": 3.2368,
+      "step": 18584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5238771438598633,
+      "learning_rate": 0.00047357784455569034,
+      "loss": 3.2168,
+      "step": 18585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5633851289749146,
+      "learning_rate": 0.0004735652869394263,
+      "loss": 3.4173,
+      "step": 18586
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5264574289321899,
+      "learning_rate": 0.0004735527288660247,
+      "loss": 3.1726,
+      "step": 18587
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5027996301651001,
+      "learning_rate": 0.00047354017033551864,
+      "loss": 3.466,
+      "step": 18588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5182791948318481,
+      "learning_rate": 0.00047352761134794114,
+      "loss": 3.0939,
+      "step": 18589
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4975305497646332,
+      "learning_rate": 0.00047351505190332526,
+      "loss": 3.3298,
+      "step": 18590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5365167856216431,
+      "learning_rate": 0.0004735024920017042,
+      "loss": 3.0189,
+      "step": 18591
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5220871567726135,
+      "learning_rate": 0.00047348993164311086,
+      "loss": 3.3163,
+      "step": 18592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5201936960220337,
+      "learning_rate": 0.00047347737082757853,
+      "loss": 3.1742,
+      "step": 18593
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5495591163635254,
+      "learning_rate": 0.0004734648095551402,
+      "loss": 3.2221,
+      "step": 18594
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.504267156124115,
+      "learning_rate": 0.0004734522478258289,
+      "loss": 3.4931,
+      "step": 18595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5782771706581116,
+      "learning_rate": 0.0004734396856396778,
+      "loss": 3.1074,
+      "step": 18596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.523160457611084,
+      "learning_rate": 0.0004734271229967199,
+      "loss": 3.0729,
+      "step": 18597
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5083732008934021,
+      "learning_rate": 0.0004734145598969884,
+      "loss": 3.2266,
+      "step": 18598
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.538865864276886,
+      "learning_rate": 0.00047340199634051625,
+      "loss": 3.0961,
+      "step": 18599
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.528105616569519,
+      "learning_rate": 0.00047338943232733664,
+      "loss": 3.0448,
+      "step": 18600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5887966156005859,
+      "learning_rate": 0.00047337686785748256,
+      "loss": 3.0425,
+      "step": 18601
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49562567472457886,
+      "learning_rate": 0.0004733643029309873,
+      "loss": 3.3061,
+      "step": 18602
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5125460028648376,
+      "learning_rate": 0.0004733517375478838,
+      "loss": 2.8714,
+      "step": 18603
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5145890712738037,
+      "learning_rate": 0.0004733391717082052,
+      "loss": 3.0679,
+      "step": 18604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5396844148635864,
+      "learning_rate": 0.00047332660541198453,
+      "loss": 3.1425,
+      "step": 18605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5691478252410889,
+      "learning_rate": 0.00047331403865925496,
+      "loss": 3.3351,
+      "step": 18606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5248494148254395,
+      "learning_rate": 0.0004733014714500496,
+      "loss": 3.2248,
+      "step": 18607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49655720591545105,
+      "learning_rate": 0.0004732889037844014,
+      "loss": 3.0864,
+      "step": 18608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5340755581855774,
+      "learning_rate": 0.0004732763356623437,
+      "loss": 3.1839,
+      "step": 18609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5292792916297913,
+      "learning_rate": 0.0004732637670839094,
+      "loss": 3.3367,
+      "step": 18610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48931747674942017,
+      "learning_rate": 0.0004732511980491318,
+      "loss": 3.4235,
+      "step": 18611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5485087037086487,
+      "learning_rate": 0.0004732386285580438,
+      "loss": 3.0873,
+      "step": 18612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5549004077911377,
+      "learning_rate": 0.0004732260586106787,
+      "loss": 3.109,
+      "step": 18613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5397618412971497,
+      "learning_rate": 0.00047321348820706933,
+      "loss": 3.1704,
+      "step": 18614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5148725509643555,
+      "learning_rate": 0.000473200917347249,
+      "loss": 3.1237,
+      "step": 18615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49203985929489136,
+      "learning_rate": 0.00047318834603125084,
+      "loss": 3.1485,
+      "step": 18616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5316979289054871,
+      "learning_rate": 0.00047317577425910794,
+      "loss": 3.2129,
+      "step": 18617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5986040234565735,
+      "learning_rate": 0.00047316320203085333,
+      "loss": 3.1176,
+      "step": 18618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5387392044067383,
+      "learning_rate": 0.0004731506293465202,
+      "loss": 3.1098,
+      "step": 18619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5144277811050415,
+      "learning_rate": 0.0004731380562061416,
+      "loss": 3.1374,
+      "step": 18620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5290771126747131,
+      "learning_rate": 0.0004731254826097507,
+      "loss": 3.1971,
+      "step": 18621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5087745785713196,
+      "learning_rate": 0.00047311290855738057,
+      "loss": 3.3343,
+      "step": 18622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5245369076728821,
+      "learning_rate": 0.00047310033404906435,
+      "loss": 3.2556,
+      "step": 18623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5257542729377747,
+      "learning_rate": 0.00047308775908483526,
+      "loss": 3.2636,
+      "step": 18624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.50067138671875,
+      "learning_rate": 0.00047307518366472627,
+      "loss": 3.2939,
+      "step": 18625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.550284743309021,
+      "learning_rate": 0.0004730626077887705,
+      "loss": 3.1894,
+      "step": 18626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5422661900520325,
+      "learning_rate": 0.00047305003145700115,
+      "loss": 3.2083,
+      "step": 18627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5483496785163879,
+      "learning_rate": 0.00047303745466945133,
+      "loss": 3.029,
+      "step": 18628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5778025388717651,
+      "learning_rate": 0.0004730248774261542,
+      "loss": 3.0593,
+      "step": 18629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5487713813781738,
+      "learning_rate": 0.0004730122997271428,
+      "loss": 3.2694,
+      "step": 18630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5207985043525696,
+      "learning_rate": 0.00047299972157245027,
+      "loss": 3.1083,
+      "step": 18631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5357333421707153,
+      "learning_rate": 0.00047298714296210983,
+      "loss": 3.2432,
+      "step": 18632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5432522296905518,
+      "learning_rate": 0.00047297456389615446,
+      "loss": 3.461,
+      "step": 18633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4886643886566162,
+      "learning_rate": 0.0004729619843746175,
+      "loss": 3.2724,
+      "step": 18634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5570962429046631,
+      "learning_rate": 0.00047294940439753197,
+      "loss": 3.1144,
+      "step": 18635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.559655487537384,
+      "learning_rate": 0.00047293682396493087,
+      "loss": 3.2365,
+      "step": 18636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5164363384246826,
+      "learning_rate": 0.0004729242430768475,
+      "loss": 3.0816,
+      "step": 18637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5455911159515381,
+      "learning_rate": 0.00047291166173331497,
+      "loss": 3.2254,
+      "step": 18638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5552606582641602,
+      "learning_rate": 0.0004728990799343664,
+      "loss": 3.1659,
+      "step": 18639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5630795955657959,
+      "learning_rate": 0.00047288649768003493,
+      "loss": 3.2478,
+      "step": 18640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5146662592887878,
+      "learning_rate": 0.00047287391497035367,
+      "loss": 3.1473,
+      "step": 18641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5582475066184998,
+      "learning_rate": 0.00047286133180535585,
+      "loss": 3.0386,
+      "step": 18642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5350977778434753,
+      "learning_rate": 0.00047284874818507444,
+      "loss": 3.1054,
+      "step": 18643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5022348165512085,
+      "learning_rate": 0.0004728361641095428,
+      "loss": 3.1659,
+      "step": 18644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5127667188644409,
+      "learning_rate": 0.00047282357957879396,
+      "loss": 3.0755,
+      "step": 18645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5345514416694641,
+      "learning_rate": 0.000472810994592861,
+      "loss": 3.3378,
+      "step": 18646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5057428479194641,
+      "learning_rate": 0.0004727984091517772,
+      "loss": 3.0245,
+      "step": 18647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5135453343391418,
+      "learning_rate": 0.0004727858232555756,
+      "loss": 3.1026,
+      "step": 18648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.539706289768219,
+      "learning_rate": 0.00047277323690428943,
+      "loss": 3.1584,
+      "step": 18649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5512105822563171,
+      "learning_rate": 0.00047276065009795184,
+      "loss": 3.2228,
+      "step": 18650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4987412095069885,
+      "learning_rate": 0.00047274806283659584,
+      "loss": 3.3146,
+      "step": 18651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5345869064331055,
+      "learning_rate": 0.00047273547512025483,
+      "loss": 3.2434,
+      "step": 18652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5088310837745667,
+      "learning_rate": 0.0004727228869489617,
+      "loss": 3.2025,
+      "step": 18653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5335360765457153,
+      "learning_rate": 0.00047271029832274974,
+      "loss": 3.2062,
+      "step": 18654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5184034705162048,
+      "learning_rate": 0.0004726977092416522,
+      "loss": 3.2948,
+      "step": 18655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5256913304328918,
+      "learning_rate": 0.00047268511970570203,
+      "loss": 3.186,
+      "step": 18656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5344150066375732,
+      "learning_rate": 0.0004726725297149325,
+      "loss": 3.1916,
+      "step": 18657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6184183955192566,
+      "learning_rate": 0.0004726599392693768,
+      "loss": 3.3629,
+      "step": 18658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49395012855529785,
+      "learning_rate": 0.00047264734836906805,
+      "loss": 3.1883,
+      "step": 18659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.532318651676178,
+      "learning_rate": 0.00047263475701403933,
+      "loss": 2.9763,
+      "step": 18660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6311165690422058,
+      "learning_rate": 0.000472622165204324,
+      "loss": 3.2059,
+      "step": 18661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5295420289039612,
+      "learning_rate": 0.000472609572939955,
+      "loss": 3.1236,
+      "step": 18662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5523192882537842,
+      "learning_rate": 0.0004725969802209656,
+      "loss": 3.2315,
+      "step": 18663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52613365650177,
+      "learning_rate": 0.0004725843870473891,
+      "loss": 3.1807,
+      "step": 18664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5176199078559875,
+      "learning_rate": 0.0004725717934192584,
+      "loss": 3.0597,
+      "step": 18665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5190128684043884,
+      "learning_rate": 0.00047255919933660695,
+      "loss": 2.7741,
+      "step": 18666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5189964175224304,
+      "learning_rate": 0.0004725466047994676,
+      "loss": 3.3417,
+      "step": 18667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5286535024642944,
+      "learning_rate": 0.0004725340098078738,
+      "loss": 3.1332,
+      "step": 18668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5291597247123718,
+      "learning_rate": 0.00047252141436185857,
+      "loss": 3.0903,
+      "step": 18669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5236563682556152,
+      "learning_rate": 0.00047250881846145525,
+      "loss": 3.16,
+      "step": 18670
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5021608471870422,
+      "learning_rate": 0.0004724962221066969,
+      "loss": 3.2304,
+      "step": 18671
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5602541565895081,
+      "learning_rate": 0.0004724836252976165,
+      "loss": 3.2892,
+      "step": 18672
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5403422117233276,
+      "learning_rate": 0.0004724710280342477,
+      "loss": 3.4172,
+      "step": 18673
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5357560515403748,
+      "learning_rate": 0.0004724584303166232,
+      "loss": 3.4125,
+      "step": 18674
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5377302765846252,
+      "learning_rate": 0.0004724458321447765,
+      "loss": 3.1649,
+      "step": 18675
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5543449521064758,
+      "learning_rate": 0.0004724332335187406,
+      "loss": 3.0616,
+      "step": 18676
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5366581082344055,
+      "learning_rate": 0.0004724206344385487,
+      "loss": 3.1542,
+      "step": 18677
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5297412276268005,
+      "learning_rate": 0.00047240803490423413,
+      "loss": 3.3427,
+      "step": 18678
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6091488599777222,
+      "learning_rate": 0.00047239543491582993,
+      "loss": 3.0589,
+      "step": 18679
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5463923215866089,
+      "learning_rate": 0.0004723828344733694,
+      "loss": 3.1833,
+      "step": 18680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5743600726127625,
+      "learning_rate": 0.0004723702335768856,
+      "loss": 3.0632,
+      "step": 18681
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5193010568618774,
+      "learning_rate": 0.00047235763222641176,
+      "loss": 3.3067,
+      "step": 18682
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5466593503952026,
+      "learning_rate": 0.0004723450304219811,
+      "loss": 3.1118,
+      "step": 18683
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5027567148208618,
+      "learning_rate": 0.0004723324281636268,
+      "loss": 3.3245,
+      "step": 18684
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5612360239028931,
+      "learning_rate": 0.00047231982545138207,
+      "loss": 3.0984,
+      "step": 18685
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49265792965888977,
+      "learning_rate": 0.0004723072222852801,
+      "loss": 3.2577,
+      "step": 18686
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5681557059288025,
+      "learning_rate": 0.00047229461866535403,
+      "loss": 3.1238,
+      "step": 18687
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5673431158065796,
+      "learning_rate": 0.00047228201459163706,
+      "loss": 3.1261,
+      "step": 18688
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5525531768798828,
+      "learning_rate": 0.00047226941006416246,
+      "loss": 3.3082,
+      "step": 18689
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5107969641685486,
+      "learning_rate": 0.00047225680508296346,
+      "loss": 3.0155,
+      "step": 18690
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5525048971176147,
+      "learning_rate": 0.00047224419964807306,
+      "loss": 3.0787,
+      "step": 18691
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5003828406333923,
+      "learning_rate": 0.00047223159375952466,
+      "loss": 3.253,
+      "step": 18692
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49926337599754333,
+      "learning_rate": 0.0004722189874173514,
+      "loss": 3.1447,
+      "step": 18693
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49048912525177,
+      "learning_rate": 0.00047220638062158647,
+      "loss": 3.1531,
+      "step": 18694
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5253885984420776,
+      "learning_rate": 0.000472193773372263,
+      "loss": 3.2023,
+      "step": 18695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5272942185401917,
+      "learning_rate": 0.0004721811656694144,
+      "loss": 3.1818,
+      "step": 18696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5017642378807068,
+      "learning_rate": 0.00047216855751307365,
+      "loss": 3.174,
+      "step": 18697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5158599615097046,
+      "learning_rate": 0.000472155948903274,
+      "loss": 3.2798,
+      "step": 18698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48646584153175354,
+      "learning_rate": 0.00047214333984004885,
+      "loss": 3.1985,
+      "step": 18699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5171990990638733,
+      "learning_rate": 0.0004721307303234312,
+      "loss": 3.128,
+      "step": 18700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5293672680854797,
+      "learning_rate": 0.0004721181203534543,
+      "loss": 3.1083,
+      "step": 18701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5768073797225952,
+      "learning_rate": 0.00047210550993015155,
+      "loss": 3.1399,
+      "step": 18702
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5409702658653259,
+      "learning_rate": 0.0004720928990535558,
+      "loss": 3.3084,
+      "step": 18703
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5259953737258911,
+      "learning_rate": 0.0004720802877237006,
+      "loss": 3.1811,
+      "step": 18704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5130147337913513,
+      "learning_rate": 0.0004720676759406191,
+      "loss": 2.9815,
+      "step": 18705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4972241520881653,
+      "learning_rate": 0.0004720550637043443,
+      "loss": 3.1822,
+      "step": 18706
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5117876529693604,
+      "learning_rate": 0.00047204245101490966,
+      "loss": 3.1758,
+      "step": 18707
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5374881625175476,
+      "learning_rate": 0.0004720298378723482,
+      "loss": 3.3106,
+      "step": 18708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5357831716537476,
+      "learning_rate": 0.0004720172242766934,
+      "loss": 3.1024,
+      "step": 18709
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5341154932975769,
+      "learning_rate": 0.0004720046102279783,
+      "loss": 3.0437,
+      "step": 18710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.541704535484314,
+      "learning_rate": 0.00047199199572623617,
+      "loss": 3.3368,
+      "step": 18711
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5049653053283691,
+      "learning_rate": 0.0004719793807715002,
+      "loss": 3.1156,
+      "step": 18712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.517701268196106,
+      "learning_rate": 0.0004719667653638036,
+      "loss": 3.1901,
+      "step": 18713
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.547834038734436,
+      "learning_rate": 0.00047195414950317965,
+      "loss": 3.1269,
+      "step": 18714
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5490109920501709,
+      "learning_rate": 0.0004719415331896616,
+      "loss": 3.1857,
+      "step": 18715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5507968068122864,
+      "learning_rate": 0.0004719289164232827,
+      "loss": 3.3192,
+      "step": 18716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5235647559165955,
+      "learning_rate": 0.000471916299204076,
+      "loss": 3.0702,
+      "step": 18717
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.524595320224762,
+      "learning_rate": 0.0004719036815320749,
+      "loss": 3.0391,
+      "step": 18718
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5237332582473755,
+      "learning_rate": 0.0004718910634073126,
+      "loss": 3.1249,
+      "step": 18719
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.541515052318573,
+      "learning_rate": 0.0004718784448298223,
+      "loss": 3.0022,
+      "step": 18720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5793763995170593,
+      "learning_rate": 0.0004718658257996373,
+      "loss": 3.1743,
+      "step": 18721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5258360505104065,
+      "learning_rate": 0.00047185320631679074,
+      "loss": 3.2977,
+      "step": 18722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5656243562698364,
+      "learning_rate": 0.0004718405863813159,
+      "loss": 3.304,
+      "step": 18723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5465549826622009,
+      "learning_rate": 0.0004718279659932461,
+      "loss": 3.3521,
+      "step": 18724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5304923057556152,
+      "learning_rate": 0.0004718153451526144,
+      "loss": 3.1136,
+      "step": 18725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5266689658164978,
+      "learning_rate": 0.00047180272385945426,
+      "loss": 3.258,
+      "step": 18726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5561476945877075,
+      "learning_rate": 0.0004717901021137988,
+      "loss": 3.1804,
+      "step": 18727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5388331413269043,
+      "learning_rate": 0.0004717774799156812,
+      "loss": 3.2172,
+      "step": 18728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5558831691741943,
+      "learning_rate": 0.00047176485726513486,
+      "loss": 3.0524,
+      "step": 18729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5594724416732788,
+      "learning_rate": 0.00047175223416219287,
+      "loss": 3.1281,
+      "step": 18730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5364587306976318,
+      "learning_rate": 0.00047173961060688856,
+      "loss": 3.0472,
+      "step": 18731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4832528531551361,
+      "learning_rate": 0.0004717269865992552,
+      "loss": 3.1427,
+      "step": 18732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5160096287727356,
+      "learning_rate": 0.00047171436213932594,
+      "loss": 3.2637,
+      "step": 18733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5533026456832886,
+      "learning_rate": 0.0004717017372271341,
+      "loss": 3.3568,
+      "step": 18734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6159023642539978,
+      "learning_rate": 0.000471689111862713,
+      "loss": 3.4341,
+      "step": 18735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5422484874725342,
+      "learning_rate": 0.00047167648604609587,
+      "loss": 3.0425,
+      "step": 18736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5426680445671082,
+      "learning_rate": 0.00047166385977731584,
+      "loss": 3.0185,
+      "step": 18737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5291516780853271,
+      "learning_rate": 0.0004716512330564062,
+      "loss": 3.0756,
+      "step": 18738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5165324211120605,
+      "learning_rate": 0.0004716386058834003,
+      "loss": 3.1437,
+      "step": 18739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5441447496414185,
+      "learning_rate": 0.00047162597825833126,
+      "loss": 3.1802,
+      "step": 18740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5122418403625488,
+      "learning_rate": 0.00047161335018123255,
+      "loss": 3.0825,
+      "step": 18741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5164358019828796,
+      "learning_rate": 0.0004716007216521372,
+      "loss": 3.1459,
+      "step": 18742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5052106976509094,
+      "learning_rate": 0.0004715880926710787,
+      "loss": 3.1541,
+      "step": 18743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51518315076828,
+      "learning_rate": 0.0004715754632380901,
+      "loss": 3.1712,
+      "step": 18744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5447625517845154,
+      "learning_rate": 0.0004715628333532047,
+      "loss": 3.2673,
+      "step": 18745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5232028961181641,
+      "learning_rate": 0.00047155020301645596,
+      "loss": 3.2926,
+      "step": 18746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5608052015304565,
+      "learning_rate": 0.0004715375722278769,
+      "loss": 3.2625,
+      "step": 18747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5659303665161133,
+      "learning_rate": 0.0004715249409875009,
+      "loss": 3.1281,
+      "step": 18748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49152451753616333,
+      "learning_rate": 0.0004715123092953612,
+      "loss": 3.1931,
+      "step": 18749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5154252052307129,
+      "learning_rate": 0.00047149967715149114,
+      "loss": 3.0923,
+      "step": 18750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5097779035568237,
+      "learning_rate": 0.00047148704455592387,
+      "loss": 2.9026,
+      "step": 18751
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5418404936790466,
+      "learning_rate": 0.00047147441150869273,
+      "loss": 2.9388,
+      "step": 18752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5441518425941467,
+      "learning_rate": 0.00047146177800983104,
+      "loss": 3.1708,
+      "step": 18753
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5178873538970947,
+      "learning_rate": 0.00047144914405937194,
+      "loss": 3.2744,
+      "step": 18754
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5090039968490601,
+      "learning_rate": 0.000471436509657349,
+      "loss": 3.1552,
+      "step": 18755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5149325728416443,
+      "learning_rate": 0.000471423874803795,
+      "loss": 3.2349,
+      "step": 18756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5354869961738586,
+      "learning_rate": 0.00047141123949874373,
+      "loss": 3.3082,
+      "step": 18757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.562247633934021,
+      "learning_rate": 0.00047139860374222813,
+      "loss": 3.2232,
+      "step": 18758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5432979464530945,
+      "learning_rate": 0.00047138596753428164,
+      "loss": 3.2849,
+      "step": 18759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5252031683921814,
+      "learning_rate": 0.00047137333087493745,
+      "loss": 3.0496,
+      "step": 18760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5139029622077942,
+      "learning_rate": 0.00047136069376422894,
+      "loss": 3.0417,
+      "step": 18761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49422696232795715,
+      "learning_rate": 0.00047134805620218935,
+      "loss": 3.0737,
+      "step": 18762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.548893928527832,
+      "learning_rate": 0.0004713354181888519,
+      "loss": 3.2133,
+      "step": 18763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.534731924533844,
+      "learning_rate": 0.00047132277972425,
+      "loss": 3.1941,
+      "step": 18764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5220165252685547,
+      "learning_rate": 0.00047131014080841675,
+      "loss": 3.1918,
+      "step": 18765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5755146145820618,
+      "learning_rate": 0.00047129750144138567,
+      "loss": 3.1908,
+      "step": 18766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.508004903793335,
+      "learning_rate": 0.0004712848616231899,
+      "loss": 3.2737,
+      "step": 18767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5089250802993774,
+      "learning_rate": 0.0004712722213538628,
+      "loss": 2.9431,
+      "step": 18768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.573453962802887,
+      "learning_rate": 0.0004712595806334376,
+      "loss": 3.064,
+      "step": 18769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5167307257652283,
+      "learning_rate": 0.0004712469394619476,
+      "loss": 3.0751,
+      "step": 18770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5273975729942322,
+      "learning_rate": 0.00047123429783942614,
+      "loss": 3.1842,
+      "step": 18771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5206418633460999,
+      "learning_rate": 0.00047122165576590655,
+      "loss": 3.0887,
+      "step": 18772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5117706656455994,
+      "learning_rate": 0.00047120901324142206,
+      "loss": 3.3396,
+      "step": 18773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5400931239128113,
+      "learning_rate": 0.0004711963702660059,
+      "loss": 3.1191,
+      "step": 18774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5695016384124756,
+      "learning_rate": 0.0004711837268396915,
+      "loss": 3.0463,
+      "step": 18775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5119388699531555,
+      "learning_rate": 0.000471171082962512,
+      "loss": 3.3504,
+      "step": 18776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5598418712615967,
+      "learning_rate": 0.000471158438634501,
+      "loss": 3.1825,
+      "step": 18777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5664933919906616,
+      "learning_rate": 0.0004711457938556915,
+      "loss": 3.1576,
+      "step": 18778
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5691736340522766,
+      "learning_rate": 0.000471133148626117,
+      "loss": 3.0413,
+      "step": 18779
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5121724009513855,
+      "learning_rate": 0.00047112050294581065,
+      "loss": 3.4054,
+      "step": 18780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5062749981880188,
+      "learning_rate": 0.0004711078568148058,
+      "loss": 3.1947,
+      "step": 18781
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5275105237960815,
+      "learning_rate": 0.00047109521023313585,
+      "loss": 3.1643,
+      "step": 18782
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5385158061981201,
+      "learning_rate": 0.000471082563200834,
+      "loss": 3.0541,
+      "step": 18783
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5302549600601196,
+      "learning_rate": 0.00047106991571793357,
+      "loss": 2.8629,
+      "step": 18784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5567852258682251,
+      "learning_rate": 0.0004710572677844679,
+      "loss": 3.1464,
+      "step": 18785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5577290058135986,
+      "learning_rate": 0.0004710446194004704,
+      "loss": 3.2132,
+      "step": 18786
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.516573965549469,
+      "learning_rate": 0.0004710319705659744,
+      "loss": 3.3098,
+      "step": 18787
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5409716367721558,
+      "learning_rate": 0.0004710193212810129,
+      "loss": 3.0126,
+      "step": 18788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5368973612785339,
+      "learning_rate": 0.00047100667154561946,
+      "loss": 3.2313,
+      "step": 18789
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48436570167541504,
+      "learning_rate": 0.0004709940213598274,
+      "loss": 3.1132,
+      "step": 18790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5413197875022888,
+      "learning_rate": 0.00047098137072366995,
+      "loss": 3.1343,
+      "step": 18791
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.530865490436554,
+      "learning_rate": 0.00047096871963718046,
+      "loss": 3.0911,
+      "step": 18792
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5218861699104309,
+      "learning_rate": 0.00047095606810039237,
+      "loss": 3.2357,
+      "step": 18793
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5184247493743896,
+      "learning_rate": 0.00047094341611333886,
+      "loss": 3.1619,
+      "step": 18794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5146009922027588,
+      "learning_rate": 0.00047093076367605316,
+      "loss": 3.2411,
+      "step": 18795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5616743564605713,
+      "learning_rate": 0.0004709181107885689,
+      "loss": 3.233,
+      "step": 18796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5262179970741272,
+      "learning_rate": 0.0004709054574509192,
+      "loss": 3.3252,
+      "step": 18797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5056144595146179,
+      "learning_rate": 0.0004708928036631374,
+      "loss": 3.1111,
+      "step": 18798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5371179580688477,
+      "learning_rate": 0.00047088014942525675,
+      "loss": 3.1897,
+      "step": 18799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5067994594573975,
+      "learning_rate": 0.0004708674947373108,
+      "loss": 3.3786,
+      "step": 18800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5345014929771423,
+      "learning_rate": 0.0004708548395993326,
+      "loss": 3.2104,
+      "step": 18801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5636339783668518,
+      "learning_rate": 0.00047084218401135576,
+      "loss": 3.3303,
+      "step": 18802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4965422749519348,
+      "learning_rate": 0.0004708295279734135,
+      "loss": 3.1805,
+      "step": 18803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.533908486366272,
+      "learning_rate": 0.0004708168714855391,
+      "loss": 3.1346,
+      "step": 18804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.512013852596283,
+      "learning_rate": 0.0004708042145477659,
+      "loss": 3.1826,
+      "step": 18805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5449269413948059,
+      "learning_rate": 0.0004707915571601273,
+      "loss": 3.1899,
+      "step": 18806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5012300610542297,
+      "learning_rate": 0.0004707788993226567,
+      "loss": 3.4398,
+      "step": 18807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5492137670516968,
+      "learning_rate": 0.00047076624103538725,
+      "loss": 3.1645,
+      "step": 18808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5013618469238281,
+      "learning_rate": 0.00047075358229835236,
+      "loss": 3.1924,
+      "step": 18809
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5345459580421448,
+      "learning_rate": 0.00047074092311158536,
+      "loss": 3.0099,
+      "step": 18810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5092979669570923,
+      "learning_rate": 0.0004707282634751197,
+      "loss": 3.1141,
+      "step": 18811
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5068957805633545,
+      "learning_rate": 0.0004707156033889887,
+      "loss": 3.3472,
+      "step": 18812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5265266299247742,
+      "learning_rate": 0.0004707029428532256,
+      "loss": 2.9948,
+      "step": 18813
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5371838808059692,
+      "learning_rate": 0.00047069028186786386,
+      "loss": 3.2557,
+      "step": 18814
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5402524471282959,
+      "learning_rate": 0.0004706776204329367,
+      "loss": 3.1617,
+      "step": 18815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5330641269683838,
+      "learning_rate": 0.0004706649585484775,
+      "loss": 3.1771,
+      "step": 18816
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5558748841285706,
+      "learning_rate": 0.0004706522962145197,
+      "loss": 3.1897,
+      "step": 18817
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5087716579437256,
+      "learning_rate": 0.0004706396334310966,
+      "loss": 3.1177,
+      "step": 18818
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5257171392440796,
+      "learning_rate": 0.00047062697019824155,
+      "loss": 3.0529,
+      "step": 18819
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5224782228469849,
+      "learning_rate": 0.00047061430651598783,
+      "loss": 3.3477,
+      "step": 18820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5466150641441345,
+      "learning_rate": 0.0004706016423843689,
+      "loss": 2.9722,
+      "step": 18821
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5087209939956665,
+      "learning_rate": 0.0004705889778034181,
+      "loss": 3.4086,
+      "step": 18822
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5442378520965576,
+      "learning_rate": 0.00047057631277316875,
+      "loss": 3.4454,
+      "step": 18823
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5326034426689148,
+      "learning_rate": 0.00047056364729365424,
+      "loss": 3.2417,
+      "step": 18824
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4734261929988861,
+      "learning_rate": 0.0004705509813649079,
+      "loss": 3.2332,
+      "step": 18825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.543708324432373,
+      "learning_rate": 0.0004705383149869631,
+      "loss": 3.2303,
+      "step": 18826
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5420598983764648,
+      "learning_rate": 0.00047052564815985317,
+      "loss": 3.2229,
+      "step": 18827
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.543842077255249,
+      "learning_rate": 0.00047051298088361155,
+      "loss": 3.2,
+      "step": 18828
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.511565089225769,
+      "learning_rate": 0.00047050031315827156,
+      "loss": 3.1562,
+      "step": 18829
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5324139595031738,
+      "learning_rate": 0.0004704876449838665,
+      "loss": 3.0808,
+      "step": 18830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5224776268005371,
+      "learning_rate": 0.00047047497636042976,
+      "loss": 3.131,
+      "step": 18831
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.525739312171936,
+      "learning_rate": 0.00047046230728799485,
+      "loss": 3.2305,
+      "step": 18832
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5212183594703674,
+      "learning_rate": 0.00047044963776659497,
+      "loss": 3.0373,
+      "step": 18833
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5085528492927551,
+      "learning_rate": 0.0004704369677962636,
+      "loss": 3.235,
+      "step": 18834
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5560317635536194,
+      "learning_rate": 0.0004704242973770339,
+      "loss": 3.2723,
+      "step": 18835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48996028304100037,
+      "learning_rate": 0.00047041162650893957,
+      "loss": 3.139,
+      "step": 18836
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5113476514816284,
+      "learning_rate": 0.0004703989551920138,
+      "loss": 3.1393,
+      "step": 18837
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5235118865966797,
+      "learning_rate": 0.0004703862834262899,
+      "loss": 3.1353,
+      "step": 18838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.523135244846344,
+      "learning_rate": 0.0004703736112118014,
+      "loss": 3.1062,
+      "step": 18839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5716493725776672,
+      "learning_rate": 0.0004703609385485815,
+      "loss": 3.2806,
+      "step": 18840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4963917136192322,
+      "learning_rate": 0.00047034826543666374,
+      "loss": 3.0589,
+      "step": 18841
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54476398229599,
+      "learning_rate": 0.0004703355918760814,
+      "loss": 3.0997,
+      "step": 18842
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48611146211624146,
+      "learning_rate": 0.0004703229178668679,
+      "loss": 3.0625,
+      "step": 18843
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5003273487091064,
+      "learning_rate": 0.0004703102434090566,
+      "loss": 3.0927,
+      "step": 18844
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5381352305412292,
+      "learning_rate": 0.00047029756850268097,
+      "loss": 2.7574,
+      "step": 18845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5133535265922546,
+      "learning_rate": 0.00047028489314777423,
+      "loss": 3.2066,
+      "step": 18846
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4886670708656311,
+      "learning_rate": 0.0004702722173443699,
+      "loss": 2.9163,
+      "step": 18847
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.513353705406189,
+      "learning_rate": 0.0004702595410925013,
+      "loss": 3.2403,
+      "step": 18848
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5126110911369324,
+      "learning_rate": 0.00047024686439220187,
+      "loss": 3.1207,
+      "step": 18849
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.525732696056366,
+      "learning_rate": 0.0004702341872435049,
+      "loss": 3.2541,
+      "step": 18850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5350430607795715,
+      "learning_rate": 0.0004702215096464439,
+      "loss": 3.1009,
+      "step": 18851
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5231349468231201,
+      "learning_rate": 0.0004702088316010522,
+      "loss": 2.9764,
+      "step": 18852
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5163970589637756,
+      "learning_rate": 0.00047019615310736314,
+      "loss": 2.9218,
+      "step": 18853
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5017314553260803,
+      "learning_rate": 0.00047018347416541027,
+      "loss": 3.249,
+      "step": 18854
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5135806798934937,
+      "learning_rate": 0.00047017079477522676,
+      "loss": 3.0667,
+      "step": 18855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4848870635032654,
+      "learning_rate": 0.0004701581149368463,
+      "loss": 3.2183,
+      "step": 18856
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5242249369621277,
+      "learning_rate": 0.0004701454346503019,
+      "loss": 3.0505,
+      "step": 18857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5194023847579956,
+      "learning_rate": 0.00047013275391562736,
+      "loss": 3.2237,
+      "step": 18858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5167834162712097,
+      "learning_rate": 0.00047012007273285574,
+      "loss": 3.0208,
+      "step": 18859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5210264921188354,
+      "learning_rate": 0.0004701073911020207,
+      "loss": 3.0097,
+      "step": 18860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5134527087211609,
+      "learning_rate": 0.0004700947090231555,
+      "loss": 3.0335,
+      "step": 18861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5310779213905334,
+      "learning_rate": 0.00047008202649629347,
+      "loss": 3.2074,
+      "step": 18862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5398174524307251,
+      "learning_rate": 0.0004700693435214682,
+      "loss": 2.9155,
+      "step": 18863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5127217173576355,
+      "learning_rate": 0.000470056660098713,
+      "loss": 3.0767,
+      "step": 18864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4852171838283539,
+      "learning_rate": 0.00047004397622806135,
+      "loss": 3.2109,
+      "step": 18865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.514201283454895,
+      "learning_rate": 0.0004700312919095465,
+      "loss": 3.2189,
+      "step": 18866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4986250698566437,
+      "learning_rate": 0.000470018607143202,
+      "loss": 3.2653,
+      "step": 18867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.531745195388794,
+      "learning_rate": 0.00047000592192906115,
+      "loss": 3.1656,
+      "step": 18868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5156833529472351,
+      "learning_rate": 0.00046999323626715746,
+      "loss": 3.1795,
+      "step": 18869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5258499383926392,
+      "learning_rate": 0.00046998055015752435,
+      "loss": 3.031,
+      "step": 18870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4953664541244507,
+      "learning_rate": 0.00046996786360019516,
+      "loss": 3.1802,
+      "step": 18871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5456046462059021,
+      "learning_rate": 0.00046995517659520323,
+      "loss": 3.224,
+      "step": 18872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5381906628608704,
+      "learning_rate": 0.00046994248914258217,
+      "loss": 3.1333,
+      "step": 18873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5225964784622192,
+      "learning_rate": 0.00046992980124236527,
+      "loss": 3.2528,
+      "step": 18874
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.492789089679718,
+      "learning_rate": 0.00046991711289458597,
+      "loss": 2.9492,
+      "step": 18875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5383958220481873,
+      "learning_rate": 0.00046990442409927777,
+      "loss": 3.2909,
+      "step": 18876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5192603468894958,
+      "learning_rate": 0.0004698917348564739,
+      "loss": 3.021,
+      "step": 18877
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5470327138900757,
+      "learning_rate": 0.000469879045166208,
+      "loss": 3.0586,
+      "step": 18878
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5224716067314148,
+      "learning_rate": 0.0004698663550285133,
+      "loss": 2.977,
+      "step": 18879
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5188048481941223,
+      "learning_rate": 0.00046985366444342335,
+      "loss": 3.1992,
+      "step": 18880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5498374700546265,
+      "learning_rate": 0.0004698409734109716,
+      "loss": 2.9874,
+      "step": 18881
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5041355490684509,
+      "learning_rate": 0.00046982828193119127,
+      "loss": 3.0512,
+      "step": 18882
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5342227220535278,
+      "learning_rate": 0.00046981559000411604,
+      "loss": 3.1102,
+      "step": 18883
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5024885535240173,
+      "learning_rate": 0.0004698028976297792,
+      "loss": 3.1677,
+      "step": 18884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5300989151000977,
+      "learning_rate": 0.0004697902048082142,
+      "loss": 3.1714,
+      "step": 18885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5051508545875549,
+      "learning_rate": 0.00046977751153945457,
+      "loss": 2.9421,
+      "step": 18886
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5393108129501343,
+      "learning_rate": 0.0004697648178235335,
+      "loss": 3.0567,
+      "step": 18887
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5067991614341736,
+      "learning_rate": 0.00046975212366048467,
+      "loss": 3.2442,
+      "step": 18888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5369142889976501,
+      "learning_rate": 0.00046973942905034137,
+      "loss": 3.0943,
+      "step": 18889
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5115844011306763,
+      "learning_rate": 0.0004697267339931371,
+      "loss": 3.1095,
+      "step": 18890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48783326148986816,
+      "learning_rate": 0.0004697140384889053,
+      "loss": 3.1938,
+      "step": 18891
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5423147678375244,
+      "learning_rate": 0.00046970134253767937,
+      "loss": 3.1264,
+      "step": 18892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5428636074066162,
+      "learning_rate": 0.00046968864613949267,
+      "loss": 3.3557,
+      "step": 18893
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5630121827125549,
+      "learning_rate": 0.00046967594929437885,
+      "loss": 2.9878,
+      "step": 18894
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5453979969024658,
+      "learning_rate": 0.00046966325200237117,
+      "loss": 3.0636,
+      "step": 18895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5532796382904053,
+      "learning_rate": 0.0004696505542635032,
+      "loss": 3.1966,
+      "step": 18896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5100991725921631,
+      "learning_rate": 0.00046963785607780827,
+      "loss": 3.0472,
+      "step": 18897
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5884232521057129,
+      "learning_rate": 0.0004696251574453198,
+      "loss": 2.9916,
+      "step": 18898
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5338824391365051,
+      "learning_rate": 0.0004696124583660715,
+      "loss": 3.2273,
+      "step": 18899
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5424486398696899,
+      "learning_rate": 0.00046959975884009646,
+      "loss": 3.3073,
+      "step": 18900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5141465067863464,
+      "learning_rate": 0.0004695870588674284,
+      "loss": 3.2295,
+      "step": 18901
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5177417993545532,
+      "learning_rate": 0.00046957435844810063,
+      "loss": 3.2054,
+      "step": 18902
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5106486082077026,
+      "learning_rate": 0.0004695616575821466,
+      "loss": 3.1774,
+      "step": 18903
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5100563168525696,
+      "learning_rate": 0.0004695489562695998,
+      "loss": 3.0875,
+      "step": 18904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.524345338344574,
+      "learning_rate": 0.0004695362545104938,
+      "loss": 3.2862,
+      "step": 18905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5443475246429443,
+      "learning_rate": 0.0004695235523048618,
+      "loss": 2.8944,
+      "step": 18906
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5217635631561279,
+      "learning_rate": 0.00046951084965273746,
+      "loss": 3.171,
+      "step": 18907
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5589973330497742,
+      "learning_rate": 0.00046949814655415406,
+      "loss": 3.0554,
+      "step": 18908
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49502140283584595,
+      "learning_rate": 0.0004694854430091452,
+      "loss": 3.2162,
+      "step": 18909
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5253203511238098,
+      "learning_rate": 0.00046947273901774443,
+      "loss": 3.2409,
+      "step": 18910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.508833646774292,
+      "learning_rate": 0.000469460034579985,
+      "loss": 3.0954,
+      "step": 18911
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5880780816078186,
+      "learning_rate": 0.0004694473296959004,
+      "loss": 3.1979,
+      "step": 18912
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5805985331535339,
+      "learning_rate": 0.0004694346243655242,
+      "loss": 3.0149,
+      "step": 18913
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5318664908409119,
+      "learning_rate": 0.00046942191858888977,
+      "loss": 3.1204,
+      "step": 18914
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5232014060020447,
+      "learning_rate": 0.00046940921236603064,
+      "loss": 3.1569,
+      "step": 18915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5750177502632141,
+      "learning_rate": 0.00046939650569698024,
+      "loss": 2.9964,
+      "step": 18916
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5180473327636719,
+      "learning_rate": 0.00046938379858177203,
+      "loss": 3.063,
+      "step": 18917
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5332753658294678,
+      "learning_rate": 0.0004693710910204395,
+      "loss": 3.2211,
+      "step": 18918
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5309212803840637,
+      "learning_rate": 0.0004693583830130162,
+      "loss": 3.2455,
+      "step": 18919
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.530160129070282,
+      "learning_rate": 0.0004693456745595354,
+      "loss": 2.9723,
+      "step": 18920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5082368850708008,
+      "learning_rate": 0.00046933296566003076,
+      "loss": 3.102,
+      "step": 18921
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5547425746917725,
+      "learning_rate": 0.00046932025631453564,
+      "loss": 2.9199,
+      "step": 18922
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5218966603279114,
+      "learning_rate": 0.0004693075465230835,
+      "loss": 3.0292,
+      "step": 18923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.551936149597168,
+      "learning_rate": 0.00046929483628570793,
+      "loss": 3.1246,
+      "step": 18924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.516350269317627,
+      "learning_rate": 0.0004692821256024423,
+      "loss": 3.2375,
+      "step": 18925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5745588541030884,
+      "learning_rate": 0.00046926941447332016,
+      "loss": 2.9728,
+      "step": 18926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5034775733947754,
+      "learning_rate": 0.00046925670289837496,
+      "loss": 3.3807,
+      "step": 18927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6255304217338562,
+      "learning_rate": 0.0004692439908776402,
+      "loss": 3.2879,
+      "step": 18928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49725767970085144,
+      "learning_rate": 0.00046923127841114924,
+      "loss": 3.0999,
+      "step": 18929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5189871788024902,
+      "learning_rate": 0.0004692185654989358,
+      "loss": 3.1402,
+      "step": 18930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5304571390151978,
+      "learning_rate": 0.00046920585214103324,
+      "loss": 3.2725,
+      "step": 18931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5378173589706421,
+      "learning_rate": 0.00046919313833747485,
+      "loss": 3.1548,
+      "step": 18932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5394854545593262,
+      "learning_rate": 0.00046918042408829446,
+      "loss": 3.1416,
+      "step": 18933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5187913179397583,
+      "learning_rate": 0.0004691677093935253,
+      "loss": 2.9619,
+      "step": 18934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5153508186340332,
+      "learning_rate": 0.000469154994253201,
+      "loss": 3.1343,
+      "step": 18935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5249007344245911,
+      "learning_rate": 0.0004691422786673549,
+      "loss": 3.223,
+      "step": 18936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.514267086982727,
+      "learning_rate": 0.00046912956263602077,
+      "loss": 3.1555,
+      "step": 18937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.515212893486023,
+      "learning_rate": 0.0004691168461592318,
+      "loss": 3.1127,
+      "step": 18938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48960253596305847,
+      "learning_rate": 0.0004691041292370216,
+      "loss": 3.1147,
+      "step": 18939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5144261717796326,
+      "learning_rate": 0.0004690914118694237,
+      "loss": 3.2216,
+      "step": 18940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.523517906665802,
+      "learning_rate": 0.0004690786940564716,
+      "loss": 3.0786,
+      "step": 18941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5452386140823364,
+      "learning_rate": 0.00046906597579819865,
+      "loss": 3.1958,
+      "step": 18942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5395174026489258,
+      "learning_rate": 0.0004690532570946386,
+      "loss": 2.9997,
+      "step": 18943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5560466051101685,
+      "learning_rate": 0.00046904053794582465,
+      "loss": 3.3784,
+      "step": 18944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.525627076625824,
+      "learning_rate": 0.0004690278183517906,
+      "loss": 2.9482,
+      "step": 18945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5005621314048767,
+      "learning_rate": 0.0004690150983125697,
+      "loss": 3.0905,
+      "step": 18946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5510935187339783,
+      "learning_rate": 0.0004690023778281957,
+      "loss": 3.3415,
+      "step": 18947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5250145196914673,
+      "learning_rate": 0.0004689896568987018,
+      "loss": 3.1714,
+      "step": 18948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.521665096282959,
+      "learning_rate": 0.0004689769355241217,
+      "loss": 3.2226,
+      "step": 18949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6473313570022583,
+      "learning_rate": 0.00046896421370448894,
+      "loss": 3.2003,
+      "step": 18950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5071868300437927,
+      "learning_rate": 0.0004689514914398369,
+      "loss": 2.9599,
+      "step": 18951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5094021558761597,
+      "learning_rate": 0.0004689387687301992,
+      "loss": 2.9249,
+      "step": 18952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5259853601455688,
+      "learning_rate": 0.0004689260455756092,
+      "loss": 3.2138,
+      "step": 18953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4910939335823059,
+      "learning_rate": 0.00046891332197610057,
+      "loss": 3.1469,
+      "step": 18954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5359907746315002,
+      "learning_rate": 0.00046890059793170676,
+      "loss": 3.2667,
+      "step": 18955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5295614004135132,
+      "learning_rate": 0.00046888787344246134,
+      "loss": 2.998,
+      "step": 18956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5557469129562378,
+      "learning_rate": 0.0004688751485083977,
+      "loss": 3.1919,
+      "step": 18957
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5100527405738831,
+      "learning_rate": 0.0004688624231295494,
+      "loss": 3.078,
+      "step": 18958
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5492949485778809,
+      "learning_rate": 0.00046884969730595,
+      "loss": 3.2749,
+      "step": 18959
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5385236144065857,
+      "learning_rate": 0.00046883697103763293,
+      "loss": 3.0871,
+      "step": 18960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5516214966773987,
+      "learning_rate": 0.0004688242443246319,
+      "loss": 2.9884,
+      "step": 18961
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5254327654838562,
+      "learning_rate": 0.00046881151716698027,
+      "loss": 3.0269,
+      "step": 18962
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.50152587890625,
+      "learning_rate": 0.0004687987895647115,
+      "loss": 3.1315,
+      "step": 18963
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5856966376304626,
+      "learning_rate": 0.0004687860615178593,
+      "loss": 2.9434,
+      "step": 18964
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5326898694038391,
+      "learning_rate": 0.00046877333302645707,
+      "loss": 3.2463,
+      "step": 18965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5047300457954407,
+      "learning_rate": 0.00046876060409053826,
+      "loss": 3.0858,
+      "step": 18966
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.601098358631134,
+      "learning_rate": 0.00046874787471013667,
+      "loss": 3.0427,
+      "step": 18967
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5478784441947937,
+      "learning_rate": 0.0004687351448852855,
+      "loss": 3.1414,
+      "step": 18968
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5420882701873779,
+      "learning_rate": 0.0004687224146160186,
+      "loss": 3.1592,
+      "step": 18969
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5226287245750427,
+      "learning_rate": 0.0004687096839023692,
+      "loss": 3.1503,
+      "step": 18970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5345956087112427,
+      "learning_rate": 0.000468696952744371,
+      "loss": 3.3743,
+      "step": 18971
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5423299074172974,
+      "learning_rate": 0.00046868422114205755,
+      "loss": 3.0927,
+      "step": 18972
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.532418966293335,
+      "learning_rate": 0.00046867148909546225,
+      "loss": 3.2014,
+      "step": 18973
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5317066311836243,
+      "learning_rate": 0.0004686587566046187,
+      "loss": 3.1707,
+      "step": 18974
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5457773208618164,
+      "learning_rate": 0.00046864602366956037,
+      "loss": 2.9554,
+      "step": 18975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5188515186309814,
+      "learning_rate": 0.00046863329029032095,
+      "loss": 3.2574,
+      "step": 18976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5296633839607239,
+      "learning_rate": 0.000468620556466934,
+      "loss": 3.1508,
+      "step": 18977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5568720102310181,
+      "learning_rate": 0.0004686078221994329,
+      "loss": 3.2111,
+      "step": 18978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5664288997650146,
+      "learning_rate": 0.00046859508748785126,
+      "loss": 3.0099,
+      "step": 18979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5301545262336731,
+      "learning_rate": 0.0004685823523322225,
+      "loss": 3.2176,
+      "step": 18980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5368116497993469,
+      "learning_rate": 0.0004685696167325803,
+      "loss": 3.2631,
+      "step": 18981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.534150242805481,
+      "learning_rate": 0.0004685568806889582,
+      "loss": 3.0123,
+      "step": 18982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4915403723716736,
+      "learning_rate": 0.00046854414420138975,
+      "loss": 3.3539,
+      "step": 18983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5354322195053101,
+      "learning_rate": 0.00046853140726990845,
+      "loss": 3.245,
+      "step": 18984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4962729811668396,
+      "learning_rate": 0.0004685186698945478,
+      "loss": 3.1999,
+      "step": 18985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5085036158561707,
+      "learning_rate": 0.0004685059320753414,
+      "loss": 3.2156,
+      "step": 18986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5053709745407104,
+      "learning_rate": 0.00046849319381232285,
+      "loss": 3.1658,
+      "step": 18987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5463716983795166,
+      "learning_rate": 0.0004684804551055257,
+      "loss": 3.1854,
+      "step": 18988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5284098386764526,
+      "learning_rate": 0.0004684677159549834,
+      "loss": 3.3753,
+      "step": 18989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5338199734687805,
+      "learning_rate": 0.0004684549763607296,
+      "loss": 3.1831,
+      "step": 18990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5269510746002197,
+      "learning_rate": 0.0004684422363227977,
+      "loss": 3.2542,
+      "step": 18991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5175192356109619,
+      "learning_rate": 0.0004684294958412215,
+      "loss": 3.0125,
+      "step": 18992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5156930088996887,
+      "learning_rate": 0.0004684167549160344,
+      "loss": 3.3845,
+      "step": 18993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5119640231132507,
+      "learning_rate": 0.0004684040135472699,
+      "loss": 3.0625,
+      "step": 18994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390982031822205,
+      "learning_rate": 0.00046839127173496176,
+      "loss": 3.0787,
+      "step": 18995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5263301134109497,
+      "learning_rate": 0.00046837852947914324,
+      "loss": 3.0068,
+      "step": 18996
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5323955416679382,
+      "learning_rate": 0.00046836578677984824,
+      "loss": 3.0541,
+      "step": 18997
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5111451745033264,
+      "learning_rate": 0.00046835304363711006,
+      "loss": 3.058,
+      "step": 18998
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5534102916717529,
+      "learning_rate": 0.0004683403000509624,
+      "loss": 3.2107,
+      "step": 18999
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5770662426948547,
+      "learning_rate": 0.0004683275560214388,
+      "loss": 3.053,
+      "step": 19000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6135122179985046,
+      "learning_rate": 0.00046831481154857287,
+      "loss": 3.2861,
+      "step": 19001
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5752828121185303,
+      "learning_rate": 0.00046830206663239803,
+      "loss": 2.9293,
+      "step": 19002
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5392802953720093,
+      "learning_rate": 0.00046828932127294796,
+      "loss": 3.1803,
+      "step": 19003
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5212385654449463,
+      "learning_rate": 0.00046827657547025625,
+      "loss": 3.2022,
+      "step": 19004
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5026683807373047,
+      "learning_rate": 0.00046826382922435635,
+      "loss": 3.3249,
+      "step": 19005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.536699652671814,
+      "learning_rate": 0.000468251082535282,
+      "loss": 3.1748,
+      "step": 19006
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5301597714424133,
+      "learning_rate": 0.0004682383354030666,
+      "loss": 3.1808,
+      "step": 19007
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5335184335708618,
+      "learning_rate": 0.0004682255878277438,
+      "loss": 3.2511,
+      "step": 19008
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5118494033813477,
+      "learning_rate": 0.00046821283980934725,
+      "loss": 3.0333,
+      "step": 19009
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5088841319084167,
+      "learning_rate": 0.0004682000913479104,
+      "loss": 3.1185,
+      "step": 19010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5405777096748352,
+      "learning_rate": 0.00046818734244346677,
+      "loss": 3.4098,
+      "step": 19011
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5063764452934265,
+      "learning_rate": 0.0004681745930960502,
+      "loss": 3.2377,
+      "step": 19012
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5210672616958618,
+      "learning_rate": 0.0004681618433056941,
+      "loss": 3.2403,
+      "step": 19013
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5515353679656982,
+      "learning_rate": 0.00046814909307243204,
+      "loss": 3.2621,
+      "step": 19014
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5045962929725647,
+      "learning_rate": 0.00046813634239629764,
+      "loss": 3.2523,
+      "step": 19015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4997437596321106,
+      "learning_rate": 0.0004681235912773245,
+      "loss": 3.0961,
+      "step": 19016
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5317462682723999,
+      "learning_rate": 0.0004681108397155461,
+      "loss": 3.061,
+      "step": 19017
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5284067988395691,
+      "learning_rate": 0.00046809808771099614,
+      "loss": 3.2071,
+      "step": 19018
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5324392914772034,
+      "learning_rate": 0.00046808533526370826,
+      "loss": 3.0611,
+      "step": 19019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5281850695610046,
+      "learning_rate": 0.0004680725823737158,
+      "loss": 3.0396,
+      "step": 19020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5365699529647827,
+      "learning_rate": 0.00046805982904105255,
+      "loss": 3.0534,
+      "step": 19021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5103819966316223,
+      "learning_rate": 0.00046804707526575215,
+      "loss": 2.9962,
+      "step": 19022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.517741322517395,
+      "learning_rate": 0.000468034321047848,
+      "loss": 3.0247,
+      "step": 19023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5281891822814941,
+      "learning_rate": 0.00046802156638737385,
+      "loss": 3.0705,
+      "step": 19024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5412678718566895,
+      "learning_rate": 0.00046800881128436316,
+      "loss": 2.9022,
+      "step": 19025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5153235793113708,
+      "learning_rate": 0.0004679960557388496,
+      "loss": 3.1095,
+      "step": 19026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5097496509552002,
+      "learning_rate": 0.0004679832997508668,
+      "loss": 3.0689,
+      "step": 19027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5528456568717957,
+      "learning_rate": 0.0004679705433204483,
+      "loss": 3.1618,
+      "step": 19028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5313249230384827,
+      "learning_rate": 0.0004679577864476278,
+      "loss": 3.1174,
+      "step": 19029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5198782086372375,
+      "learning_rate": 0.00046794502913243875,
+      "loss": 3.1929,
+      "step": 19030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5220719575881958,
+      "learning_rate": 0.00046793227137491473,
+      "loss": 3.0891,
+      "step": 19031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5011788606643677,
+      "learning_rate": 0.0004679195131750895,
+      "loss": 3.106,
+      "step": 19032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.525837242603302,
+      "learning_rate": 0.00046790675453299666,
+      "loss": 3.1896,
+      "step": 19033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5511639714241028,
+      "learning_rate": 0.0004678939954486696,
+      "loss": 3.213,
+      "step": 19034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5156654715538025,
+      "learning_rate": 0.00046788123592214224,
+      "loss": 3.1804,
+      "step": 19035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49573853611946106,
+      "learning_rate": 0.00046786847595344774,
+      "loss": 3.1656,
+      "step": 19036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5024093389511108,
+      "learning_rate": 0.00046785571554262026,
+      "loss": 3.1114,
+      "step": 19037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5238023996353149,
+      "learning_rate": 0.000467842954689693,
+      "loss": 3.1527,
+      "step": 19038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5388356447219849,
+      "learning_rate": 0.0004678301933946997,
+      "loss": 3.2028,
+      "step": 19039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.512860119342804,
+      "learning_rate": 0.00046781743165767405,
+      "loss": 3.0169,
+      "step": 19040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5536538362503052,
+      "learning_rate": 0.0004678046694786495,
+      "loss": 2.8985,
+      "step": 19041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4811255931854248,
+      "learning_rate": 0.0004677919068576597,
+      "loss": 3.3228,
+      "step": 19042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5000419616699219,
+      "learning_rate": 0.00046777914379473847,
+      "loss": 3.0492,
+      "step": 19043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5147316455841064,
+      "learning_rate": 0.0004677663802899192,
+      "loss": 3.2764,
+      "step": 19044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5300585031509399,
+      "learning_rate": 0.0004677536163432355,
+      "loss": 2.9444,
+      "step": 19045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.515038788318634,
+      "learning_rate": 0.0004677408519547211,
+      "loss": 3.0745,
+      "step": 19046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5364577174186707,
+      "learning_rate": 0.0004677280871244096,
+      "loss": 3.026,
+      "step": 19047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5051068067550659,
+      "learning_rate": 0.00046771532185233456,
+      "loss": 3.2398,
+      "step": 19048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5928849577903748,
+      "learning_rate": 0.00046770255613852967,
+      "loss": 3.2081,
+      "step": 19049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5345249772071838,
+      "learning_rate": 0.0004676897899830285,
+      "loss": 3.0596,
+      "step": 19050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.513729453086853,
+      "learning_rate": 0.00046767702338586475,
+      "loss": 3.086,
+      "step": 19051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5252007246017456,
+      "learning_rate": 0.0004676642563470719,
+      "loss": 3.015,
+      "step": 19052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5086860060691833,
+      "learning_rate": 0.00046765148886668376,
+      "loss": 3.1461,
+      "step": 19053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.535895824432373,
+      "learning_rate": 0.0004676387209447338,
+      "loss": 3.1293,
+      "step": 19054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5222206711769104,
+      "learning_rate": 0.0004676259525812558,
+      "loss": 3.0691,
+      "step": 19055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5708796381950378,
+      "learning_rate": 0.0004676131837762832,
+      "loss": 3.1535,
+      "step": 19056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49365147948265076,
+      "learning_rate": 0.0004676004145298497,
+      "loss": 3.268,
+      "step": 19057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49543845653533936,
+      "learning_rate": 0.0004675876448419891,
+      "loss": 3.2324,
+      "step": 19058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5274909138679504,
+      "learning_rate": 0.00046757487471273476,
+      "loss": 3.0694,
+      "step": 19059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5357721447944641,
+      "learning_rate": 0.0004675621041421206,
+      "loss": 3.2226,
+      "step": 19060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5113617777824402,
+      "learning_rate": 0.00046754933313018,
+      "loss": 3.2211,
+      "step": 19061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.500162661075592,
+      "learning_rate": 0.0004675365616769467,
+      "loss": 3.2316,
+      "step": 19062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5388630628585815,
+      "learning_rate": 0.00046752378978245435,
+      "loss": 3.1944,
+      "step": 19063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5528714656829834,
+      "learning_rate": 0.00046751101744673654,
+      "loss": 3.3388,
+      "step": 19064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5113033056259155,
+      "learning_rate": 0.0004674982446698271,
+      "loss": 3.0037,
+      "step": 19065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5158981084823608,
+      "learning_rate": 0.00046748547145175943,
+      "loss": 3.2846,
+      "step": 19066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5108698010444641,
+      "learning_rate": 0.0004674726977925672,
+      "loss": 3.2234,
+      "step": 19067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5266377329826355,
+      "learning_rate": 0.00046745992369228416,
+      "loss": 3.0511,
+      "step": 19068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5855154991149902,
+      "learning_rate": 0.00046744714915094394,
+      "loss": 3.1555,
+      "step": 19069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5538336038589478,
+      "learning_rate": 0.0004674343741685801,
+      "loss": 3.1339,
+      "step": 19070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5286270976066589,
+      "learning_rate": 0.0004674215987452264,
+      "loss": 3.314,
+      "step": 19071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5068775415420532,
+      "learning_rate": 0.00046740882288091634,
+      "loss": 3.113,
+      "step": 19072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5175677537918091,
+      "learning_rate": 0.0004673960465756837,
+      "loss": 3.2558,
+      "step": 19073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5037120580673218,
+      "learning_rate": 0.00046738326982956216,
+      "loss": 3.232,
+      "step": 19074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5219966769218445,
+      "learning_rate": 0.00046737049264258525,
+      "loss": 3.0916,
+      "step": 19075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5453975796699524,
+      "learning_rate": 0.00046735771501478675,
+      "loss": 3.1197,
+      "step": 19076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5948343873023987,
+      "learning_rate": 0.00046734493694620006,
+      "loss": 2.7689,
+      "step": 19077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5255204439163208,
+      "learning_rate": 0.0004673321584368591,
+      "loss": 3.3097,
+      "step": 19078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.508298933506012,
+      "learning_rate": 0.0004673193794867975,
+      "loss": 3.3616,
+      "step": 19079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5323010683059692,
+      "learning_rate": 0.0004673066000960488,
+      "loss": 3.034,
+      "step": 19080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5434712767601013,
+      "learning_rate": 0.00046729382026464676,
+      "loss": 3.1091,
+      "step": 19081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5189756155014038,
+      "learning_rate": 0.0004672810399926249,
+      "loss": 3.1893,
+      "step": 19082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5350856184959412,
+      "learning_rate": 0.000467268259280017,
+      "loss": 3.1141,
+      "step": 19083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6006237864494324,
+      "learning_rate": 0.0004672554781268568,
+      "loss": 3.2461,
+      "step": 19084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5676541328430176,
+      "learning_rate": 0.00046724269653317774,
+      "loss": 3.1233,
+      "step": 19085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5647549629211426,
+      "learning_rate": 0.00046722991449901373,
+      "loss": 3.1782,
+      "step": 19086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5188087821006775,
+      "learning_rate": 0.00046721713202439816,
+      "loss": 3.049,
+      "step": 19087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4982798099517822,
+      "learning_rate": 0.00046720434910936493,
+      "loss": 3.1946,
+      "step": 19088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.549485445022583,
+      "learning_rate": 0.00046719156575394754,
+      "loss": 3.2946,
+      "step": 19089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5249114632606506,
+      "learning_rate": 0.00046717878195817985,
+      "loss": 3.2452,
+      "step": 19090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5417079925537109,
+      "learning_rate": 0.00046716599772209544,
+      "loss": 3.4411,
+      "step": 19091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5425086617469788,
+      "learning_rate": 0.00046715321304572786,
+      "loss": 3.1668,
+      "step": 19092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5171887278556824,
+      "learning_rate": 0.0004671404279291109,
+      "loss": 3.158,
+      "step": 19093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5016870498657227,
+      "learning_rate": 0.00046712764237227827,
+      "loss": 3.2172,
+      "step": 19094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4983656406402588,
+      "learning_rate": 0.0004671148563752636,
+      "loss": 3.0947,
+      "step": 19095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5325701832771301,
+      "learning_rate": 0.0004671020699381005,
+      "loss": 3.0837,
+      "step": 19096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5779774785041809,
+      "learning_rate": 0.0004670892830608228,
+      "loss": 3.0161,
+      "step": 19097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53163081407547,
+      "learning_rate": 0.000467076495743464,
+      "loss": 3.1027,
+      "step": 19098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5103086233139038,
+      "learning_rate": 0.0004670637079860579,
+      "loss": 3.242,
+      "step": 19099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5351077914237976,
+      "learning_rate": 0.00046705091978863815,
+      "loss": 3.0781,
+      "step": 19100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5188274383544922,
+      "learning_rate": 0.0004670381311512384,
+      "loss": 3.1909,
+      "step": 19101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.514674186706543,
+      "learning_rate": 0.0004670253420738924,
+      "loss": 3.1387,
+      "step": 19102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4893554449081421,
+      "learning_rate": 0.00046701255255663374,
+      "loss": 3.322,
+      "step": 19103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49983033537864685,
+      "learning_rate": 0.00046699976259949614,
+      "loss": 3.0131,
+      "step": 19104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5099627375602722,
+      "learning_rate": 0.00046698697220251344,
+      "loss": 2.979,
+      "step": 19105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5100765228271484,
+      "learning_rate": 0.0004669741813657191,
+      "loss": 2.9599,
+      "step": 19106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5214683413505554,
+      "learning_rate": 0.00046696139008914697,
+      "loss": 3.074,
+      "step": 19107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.532818078994751,
+      "learning_rate": 0.0004669485983728305,
+      "loss": 3.1915,
+      "step": 19108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48344460129737854,
+      "learning_rate": 0.00046693580621680363,
+      "loss": 3.2483,
+      "step": 19109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5342273116111755,
+      "learning_rate": 0.0004669230136211,
+      "loss": 2.9076,
+      "step": 19110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5045246481895447,
+      "learning_rate": 0.0004669102205857533,
+      "loss": 3.2139,
+      "step": 19111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5883970856666565,
+      "learning_rate": 0.0004668974271107972,
+      "loss": 2.8011,
+      "step": 19112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49823927879333496,
+      "learning_rate": 0.0004668846331962654,
+      "loss": 3.0764,
+      "step": 19113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5665649175643921,
+      "learning_rate": 0.00046687183884219156,
+      "loss": 3.1699,
+      "step": 19114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5797495245933533,
+      "learning_rate": 0.0004668590440486094,
+      "loss": 3.0427,
+      "step": 19115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5581980347633362,
+      "learning_rate": 0.0004668462488155527,
+      "loss": 3.2031,
+      "step": 19116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5199931263923645,
+      "learning_rate": 0.00046683345314305503,
+      "loss": 3.1927,
+      "step": 19117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5535823702812195,
+      "learning_rate": 0.00046682065703115014,
+      "loss": 3.1794,
+      "step": 19118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.539772629737854,
+      "learning_rate": 0.0004668078604798718,
+      "loss": 3.3153,
+      "step": 19119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5513725876808167,
+      "learning_rate": 0.00046679506348925367,
+      "loss": 3.0796,
+      "step": 19120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5442554950714111,
+      "learning_rate": 0.00046678226605932936,
+      "loss": 3.2118,
+      "step": 19121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6911001205444336,
+      "learning_rate": 0.0004667694681901327,
+      "loss": 3.2077,
+      "step": 19122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5045164227485657,
+      "learning_rate": 0.0004667566698816974,
+      "loss": 3.2937,
+      "step": 19123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5363736152648926,
+      "learning_rate": 0.000466743871134057,
+      "loss": 3.0034,
+      "step": 19124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.530031681060791,
+      "learning_rate": 0.0004667310719472455,
+      "loss": 3.0441,
+      "step": 19125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5397589206695557,
+      "learning_rate": 0.00046671827232129634,
+      "loss": 3.2575,
+      "step": 19126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5312854051589966,
+      "learning_rate": 0.0004667054722562433,
+      "loss": 3.1987,
+      "step": 19127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5177079439163208,
+      "learning_rate": 0.0004666926717521203,
+      "loss": 3.2154,
+      "step": 19128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.525133490562439,
+      "learning_rate": 0.00046667987080896065,
+      "loss": 3.2139,
+      "step": 19129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5109131932258606,
+      "learning_rate": 0.0004666670694267985,
+      "loss": 2.9454,
+      "step": 19130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.557773232460022,
+      "learning_rate": 0.00046665426760566733,
+      "loss": 3.3336,
+      "step": 19131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.534517228603363,
+      "learning_rate": 0.00046664146534560076,
+      "loss": 3.0634,
+      "step": 19132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5470923781394958,
+      "learning_rate": 0.0004666286626466328,
+      "loss": 3.1499,
+      "step": 19133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5293968319892883,
+      "learning_rate": 0.0004666158595087969,
+      "loss": 3.2722,
+      "step": 19134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5513948202133179,
+      "learning_rate": 0.00046660305593212694,
+      "loss": 2.9834,
+      "step": 19135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5250282883644104,
+      "learning_rate": 0.00046659025191665655,
+      "loss": 3.2274,
+      "step": 19136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5674799680709839,
+      "learning_rate": 0.0004665774474624196,
+      "loss": 3.1483,
+      "step": 19137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49620136618614197,
+      "learning_rate": 0.0004665646425694496,
+      "loss": 3.1663,
+      "step": 19138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5123617053031921,
+      "learning_rate": 0.0004665518372377804,
+      "loss": 3.02,
+      "step": 19139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5374035835266113,
+      "learning_rate": 0.00046653903146744576,
+      "loss": 3.227,
+      "step": 19140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5367292165756226,
+      "learning_rate": 0.0004665262252584794,
+      "loss": 3.1528,
+      "step": 19141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5225591659545898,
+      "learning_rate": 0.00046651341861091497,
+      "loss": 3.145,
+      "step": 19142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5538541674613953,
+      "learning_rate": 0.00046650061152478617,
+      "loss": 3.1249,
+      "step": 19143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5019550323486328,
+      "learning_rate": 0.00046648780400012686,
+      "loss": 3.125,
+      "step": 19144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.487470418214798,
+      "learning_rate": 0.00046647499603697076,
+      "loss": 3.0214,
+      "step": 19145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5381044149398804,
+      "learning_rate": 0.0004664621876353515,
+      "loss": 3.0006,
+      "step": 19146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.563580334186554,
+      "learning_rate": 0.0004664493787953029,
+      "loss": 3.2355,
+      "step": 19147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5690339207649231,
+      "learning_rate": 0.00046643656951685867,
+      "loss": 3.1099,
+      "step": 19148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5722571015357971,
+      "learning_rate": 0.0004664237598000525,
+      "loss": 3.2627,
+      "step": 19149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5227568745613098,
+      "learning_rate": 0.00046641094964491826,
+      "loss": 3.2356,
+      "step": 19150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5287798047065735,
+      "learning_rate": 0.00046639813905148954,
+      "loss": 3.0889,
+      "step": 19151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5487927198410034,
+      "learning_rate": 0.00046638532801980017,
+      "loss": 3.235,
+      "step": 19152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5198444724082947,
+      "learning_rate": 0.0004663725165498839,
+      "loss": 3.1417,
+      "step": 19153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5517452955245972,
+      "learning_rate": 0.00046635970464177436,
+      "loss": 3.3419,
+      "step": 19154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5759121179580688,
+      "learning_rate": 0.0004663468922955054,
+      "loss": 3.3514,
+      "step": 19155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5133375525474548,
+      "learning_rate": 0.00046633407951111075,
+      "loss": 3.1282,
+      "step": 19156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5205239653587341,
+      "learning_rate": 0.0004663212662886242,
+      "loss": 3.0449,
+      "step": 19157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5108022093772888,
+      "learning_rate": 0.00046630845262807935,
+      "loss": 3.2283,
+      "step": 19158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5807484984397888,
+      "learning_rate": 0.00046629563852951006,
+      "loss": 3.1741,
+      "step": 19159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5120785236358643,
+      "learning_rate": 0.0004662828239929502,
+      "loss": 3.148,
+      "step": 19160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5248308181762695,
+      "learning_rate": 0.00046627000901843316,
+      "loss": 3.2335,
+      "step": 19161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5054056644439697,
+      "learning_rate": 0.00046625719360599314,
+      "loss": 3.0538,
+      "step": 19162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5169258117675781,
+      "learning_rate": 0.0004662443777556635,
+      "loss": 3.2335,
+      "step": 19163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5012264847755432,
+      "learning_rate": 0.0004662315614674782,
+      "loss": 3.0468,
+      "step": 19164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5255052447319031,
+      "learning_rate": 0.00046621874474147104,
+      "loss": 3.2484,
+      "step": 19165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5075822472572327,
+      "learning_rate": 0.0004662059275776756,
+      "loss": 3.2547,
+      "step": 19166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703867077827454,
+      "learning_rate": 0.0004661931099761258,
+      "loss": 3.1126,
+      "step": 19167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5306580066680908,
+      "learning_rate": 0.0004661802919368553,
+      "loss": 3.0482,
+      "step": 19168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5228316187858582,
+      "learning_rate": 0.0004661674734598979,
+      "loss": 3.1605,
+      "step": 19169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5195432901382446,
+      "learning_rate": 0.0004661546545452873,
+      "loss": 3.1001,
+      "step": 19170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5943834781646729,
+      "learning_rate": 0.00046614183519305745,
+      "loss": 3.0419,
+      "step": 19171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5026077628135681,
+      "learning_rate": 0.00046612901540324186,
+      "loss": 3.1679,
+      "step": 19172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5464365482330322,
+      "learning_rate": 0.00046611619517587447,
+      "loss": 3.3324,
+      "step": 19173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5255415439605713,
+      "learning_rate": 0.00046610337451098895,
+      "loss": 3.0416,
+      "step": 19174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5834014415740967,
+      "learning_rate": 0.0004660905534086192,
+      "loss": 3.2447,
+      "step": 19175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5191861391067505,
+      "learning_rate": 0.0004660777318687988,
+      "loss": 3.0245,
+      "step": 19176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.523767352104187,
+      "learning_rate": 0.00046606490989156165,
+      "loss": 3.0654,
+      "step": 19177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5461781620979309,
+      "learning_rate": 0.00046605208747694155,
+      "loss": 3.1267,
+      "step": 19178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.507882297039032,
+      "learning_rate": 0.0004660392646249721,
+      "loss": 3.1003,
+      "step": 19179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.543165385723114,
+      "learning_rate": 0.00046602644133568715,
+      "loss": 3.0974,
+      "step": 19180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5282258987426758,
+      "learning_rate": 0.0004660136176091207,
+      "loss": 3.1831,
+      "step": 19181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5365543961524963,
+      "learning_rate": 0.0004660007934453062,
+      "loss": 3.1649,
+      "step": 19182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5323196649551392,
+      "learning_rate": 0.0004659879688442776,
+      "loss": 3.0073,
+      "step": 19183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5398780107498169,
+      "learning_rate": 0.00046597514380606854,
+      "loss": 3.0684,
+      "step": 19184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5349022746086121,
+      "learning_rate": 0.0004659623183307129,
+      "loss": 3.065,
+      "step": 19185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5340750217437744,
+      "learning_rate": 0.0004659494924182446,
+      "loss": 3.074,
+      "step": 19186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5252206921577454,
+      "learning_rate": 0.0004659366660686972,
+      "loss": 3.1038,
+      "step": 19187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5100142359733582,
+      "learning_rate": 0.00046592383928210457,
+      "loss": 3.0467,
+      "step": 19188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5399733185768127,
+      "learning_rate": 0.00046591101205850047,
+      "loss": 3.0333,
+      "step": 19189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5943720936775208,
+      "learning_rate": 0.0004658981843979186,
+      "loss": 3.072,
+      "step": 19190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5421499609947205,
+      "learning_rate": 0.0004658853563003929,
+      "loss": 3.3646,
+      "step": 19191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.532548189163208,
+      "learning_rate": 0.00046587252776595717,
+      "loss": 3.0745,
+      "step": 19192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5318267345428467,
+      "learning_rate": 0.0004658596987946451,
+      "loss": 3.0672,
+      "step": 19193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5340205430984497,
+      "learning_rate": 0.00046584686938649044,
+      "loss": 3.3863,
+      "step": 19194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5387397408485413,
+      "learning_rate": 0.00046583403954152705,
+      "loss": 3.1411,
+      "step": 19195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5424886345863342,
+      "learning_rate": 0.0004658212092597888,
+      "loss": 3.3,
+      "step": 19196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5942268967628479,
+      "learning_rate": 0.0004658083785413093,
+      "loss": 3.0225,
+      "step": 19197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5251699090003967,
+      "learning_rate": 0.00046579554738612245,
+      "loss": 3.1073,
+      "step": 19198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5253727436065674,
+      "learning_rate": 0.0004657827157942621,
+      "loss": 3.3228,
+      "step": 19199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.521644651889801,
+      "learning_rate": 0.0004657698837657619,
+      "loss": 3.2844,
+      "step": 19200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5412613749504089,
+      "learning_rate": 0.00046575705130065585,
+      "loss": 3.2638,
+      "step": 19201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5078155994415283,
+      "learning_rate": 0.00046574421839897754,
+      "loss": 3.1013,
+      "step": 19202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5175430178642273,
+      "learning_rate": 0.0004657313850607609,
+      "loss": 2.9286,
+      "step": 19203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5150618553161621,
+      "learning_rate": 0.0004657185512860397,
+      "loss": 3.1024,
+      "step": 19204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5550238490104675,
+      "learning_rate": 0.0004657057170748477,
+      "loss": 2.8348,
+      "step": 19205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4976348280906677,
+      "learning_rate": 0.00046569288242721867,
+      "loss": 3.1287,
+      "step": 19206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5535556674003601,
+      "learning_rate": 0.00046568004734318655,
+      "loss": 3.095,
+      "step": 19207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.525455892086029,
+      "learning_rate": 0.0004656672118227851,
+      "loss": 3.2602,
+      "step": 19208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5400662422180176,
+      "learning_rate": 0.00046565437586604805,
+      "loss": 3.2045,
+      "step": 19209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5572945475578308,
+      "learning_rate": 0.0004656415394730092,
+      "loss": 3.3045,
+      "step": 19210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5140613913536072,
+      "learning_rate": 0.00046562870264370244,
+      "loss": 3.2455,
+      "step": 19211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5592833757400513,
+      "learning_rate": 0.0004656158653781616,
+      "loss": 3.1826,
+      "step": 19212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5442313551902771,
+      "learning_rate": 0.0004656030276764205,
+      "loss": 3.1839,
+      "step": 19213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5157129168510437,
+      "learning_rate": 0.0004655901895385128,
+      "loss": 3.1051,
+      "step": 19214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5172207355499268,
+      "learning_rate": 0.00046557735096447244,
+      "loss": 3.331,
+      "step": 19215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5096434354782104,
+      "learning_rate": 0.0004655645119543331,
+      "loss": 2.9974,
+      "step": 19216
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5327327847480774,
+      "learning_rate": 0.00046555167250812886,
+      "loss": 3.1229,
+      "step": 19217
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5447360873222351,
+      "learning_rate": 0.00046553883262589324,
+      "loss": 3.0939,
+      "step": 19218
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5559058785438538,
+      "learning_rate": 0.0004655259923076603,
+      "loss": 3.1901,
+      "step": 19219
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49087318778038025,
+      "learning_rate": 0.0004655131515534637,
+      "loss": 3.2312,
+      "step": 19220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5363554954528809,
+      "learning_rate": 0.00046550031036333734,
+      "loss": 3.0807,
+      "step": 19221
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5305533409118652,
+      "learning_rate": 0.000465487468737315,
+      "loss": 3.0706,
+      "step": 19222
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5256800651550293,
+      "learning_rate": 0.0004654746266754306,
+      "loss": 3.1917,
+      "step": 19223
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5493904948234558,
+      "learning_rate": 0.00046546178417771774,
+      "loss": 3.0993,
+      "step": 19224
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5530810356140137,
+      "learning_rate": 0.0004654489412442105,
+      "loss": 3.1231,
+      "step": 19225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5260566473007202,
+      "learning_rate": 0.0004654360978749424,
+      "loss": 3.1445,
+      "step": 19226
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5247520804405212,
+      "learning_rate": 0.0004654232540699476,
+      "loss": 3.3109,
+      "step": 19227
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5059221386909485,
+      "learning_rate": 0.0004654104098292598,
+      "loss": 3.1187,
+      "step": 19228
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5202158093452454,
+      "learning_rate": 0.00046539756515291285,
+      "loss": 3.2282,
+      "step": 19229
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.542550802230835,
+      "learning_rate": 0.0004653847200409405,
+      "loss": 3.134,
+      "step": 19230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5520852208137512,
+      "learning_rate": 0.00046537187449337654,
+      "loss": 3.2093,
+      "step": 19231
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5276000499725342,
+      "learning_rate": 0.00046535902851025496,
+      "loss": 3.0294,
+      "step": 19232
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5380484461784363,
+      "learning_rate": 0.0004653461820916096,
+      "loss": 3.097,
+      "step": 19233
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5322332382202148,
+      "learning_rate": 0.0004653333352374741,
+      "loss": 3.2955,
+      "step": 19234
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5831676125526428,
+      "learning_rate": 0.00046532048794788243,
+      "loss": 3.1426,
+      "step": 19235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5310155153274536,
+      "learning_rate": 0.00046530764022286835,
+      "loss": 3.2932,
+      "step": 19236
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48738566040992737,
+      "learning_rate": 0.00046529479206246585,
+      "loss": 3.0159,
+      "step": 19237
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5302239060401917,
+      "learning_rate": 0.0004652819434667088,
+      "loss": 3.221,
+      "step": 19238
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5241603851318359,
+      "learning_rate": 0.00046526909443563074,
+      "loss": 3.0768,
+      "step": 19239
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5549540519714355,
+      "learning_rate": 0.0004652562449692658,
+      "loss": 3.2177,
+      "step": 19240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.521250307559967,
+      "learning_rate": 0.00046524339506764755,
+      "loss": 3.2072,
+      "step": 19241
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5026115775108337,
+      "learning_rate": 0.0004652305447308101,
+      "loss": 3.1384,
+      "step": 19242
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5054687261581421,
+      "learning_rate": 0.0004652176939587872,
+      "loss": 2.9898,
+      "step": 19243
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5394466519355774,
+      "learning_rate": 0.00046520484275161273,
+      "loss": 3.0466,
+      "step": 19244
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.534470796585083,
+      "learning_rate": 0.0004651919911093204,
+      "loss": 3.0333,
+      "step": 19245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5293780565261841,
+      "learning_rate": 0.0004651791390319443,
+      "loss": 3.3452,
+      "step": 19246
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5452441573143005,
+      "learning_rate": 0.00046516628651951806,
+      "loss": 3.1538,
+      "step": 19247
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49200424551963806,
+      "learning_rate": 0.00046515343357207554,
+      "loss": 3.3007,
+      "step": 19248
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5668293833732605,
+      "learning_rate": 0.0004651405801896507,
+      "loss": 3.2227,
+      "step": 19249
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5335699915885925,
+      "learning_rate": 0.00046512772637227745,
+      "loss": 3.2257,
+      "step": 19250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5185949206352234,
+      "learning_rate": 0.00046511487211998954,
+      "loss": 3.0215,
+      "step": 19251
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5684577822685242,
+      "learning_rate": 0.0004651020174328207,
+      "loss": 3.1563,
+      "step": 19252
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5220710635185242,
+      "learning_rate": 0.000465089162310805,
+      "loss": 3.2089,
+      "step": 19253
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5233851075172424,
+      "learning_rate": 0.0004650763067539762,
+      "loss": 3.2768,
+      "step": 19254
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5065687298774719,
+      "learning_rate": 0.00046506345076236823,
+      "loss": 3.1146,
+      "step": 19255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5225540399551392,
+      "learning_rate": 0.0004650505943360148,
+      "loss": 3.1571,
+      "step": 19256
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5644252300262451,
+      "learning_rate": 0.00046503773747494994,
+      "loss": 3.2514,
+      "step": 19257
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5171868205070496,
+      "learning_rate": 0.00046502488017920743,
+      "loss": 3.1537,
+      "step": 19258
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5381227731704712,
+      "learning_rate": 0.0004650120224488212,
+      "loss": 3.3608,
+      "step": 19259
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5154421329498291,
+      "learning_rate": 0.000464999164283825,
+      "loss": 3.0479,
+      "step": 19260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5348089933395386,
+      "learning_rate": 0.00046498630568425273,
+      "loss": 3.2822,
+      "step": 19261
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.518226683139801,
+      "learning_rate": 0.0004649734466501383,
+      "loss": 3.1494,
+      "step": 19262
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5043905973434448,
+      "learning_rate": 0.0004649605871815156,
+      "loss": 3.3799,
+      "step": 19263
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5411002039909363,
+      "learning_rate": 0.0004649477272784184,
+      "loss": 3.1399,
+      "step": 19264
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55306476354599,
+      "learning_rate": 0.0004649348669408807,
+      "loss": 3.0225,
+      "step": 19265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5231731534004211,
+      "learning_rate": 0.00046492200616893623,
+      "loss": 3.0566,
+      "step": 19266
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5047503113746643,
+      "learning_rate": 0.00046490914496261895,
+      "loss": 3.0925,
+      "step": 19267
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4931429922580719,
+      "learning_rate": 0.0004648962833219627,
+      "loss": 3.3032,
+      "step": 19268
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5375720858573914,
+      "learning_rate": 0.0004648834212470015,
+      "loss": 2.9951,
+      "step": 19269
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5181878209114075,
+      "learning_rate": 0.000464870558737769,
+      "loss": 3.1347,
+      "step": 19270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5091689825057983,
+      "learning_rate": 0.00046485769579429924,
+      "loss": 3.1832,
+      "step": 19271
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5010989308357239,
+      "learning_rate": 0.000464844832416626,
+      "loss": 3.3125,
+      "step": 19272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5260189771652222,
+      "learning_rate": 0.0004648319686047831,
+      "loss": 3.1826,
+      "step": 19273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5514676570892334,
+      "learning_rate": 0.0004648191043588046,
+      "loss": 3.2043,
+      "step": 19274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5793471932411194,
+      "learning_rate": 0.0004648062396787243,
+      "loss": 3.106,
+      "step": 19275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5154271125793457,
+      "learning_rate": 0.00046479337456457615,
+      "loss": 3.215,
+      "step": 19276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5268782377243042,
+      "learning_rate": 0.0004647805090163939,
+      "loss": 3.0782,
+      "step": 19277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5476327538490295,
+      "learning_rate": 0.0004647676430342115,
+      "loss": 3.0688,
+      "step": 19278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5455402135848999,
+      "learning_rate": 0.00046475477661806283,
+      "loss": 3.1699,
+      "step": 19279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5583153963088989,
+      "learning_rate": 0.00046474190976798183,
+      "loss": 3.0606,
+      "step": 19280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5432180762290955,
+      "learning_rate": 0.0004647290424840023,
+      "loss": 3.3071,
+      "step": 19281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5387990474700928,
+      "learning_rate": 0.00046471617476615825,
+      "loss": 3.2044,
+      "step": 19282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5935296416282654,
+      "learning_rate": 0.0004647033066144834,
+      "loss": 3.0468,
+      "step": 19283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5138740539550781,
+      "learning_rate": 0.00046469043802901174,
+      "loss": 2.9805,
+      "step": 19284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5473163723945618,
+      "learning_rate": 0.0004646775690097772,
+      "loss": 3.0915,
+      "step": 19285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5593499541282654,
+      "learning_rate": 0.0004646646995568136,
+      "loss": 3.2594,
+      "step": 19286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5125278234481812,
+      "learning_rate": 0.0004646518296701549,
+      "loss": 2.9646,
+      "step": 19287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5372989177703857,
+      "learning_rate": 0.000464638959349835,
+      "loss": 3.0846,
+      "step": 19288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5440700650215149,
+      "learning_rate": 0.0004646260885958877,
+      "loss": 3.2548,
+      "step": 19289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5197234749794006,
+      "learning_rate": 0.000464613217408347,
+      "loss": 2.9965,
+      "step": 19290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5141440629959106,
+      "learning_rate": 0.0004646003457872468,
+      "loss": 3.0835,
+      "step": 19291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5057777762413025,
+      "learning_rate": 0.000464587473732621,
+      "loss": 3.2426,
+      "step": 19292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5304477214813232,
+      "learning_rate": 0.0004645746012445033,
+      "loss": 3.1578,
+      "step": 19293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49843043088912964,
+      "learning_rate": 0.00046456172832292795,
+      "loss": 3.2066,
+      "step": 19294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5783942341804504,
+      "learning_rate": 0.0004645488549679286,
+      "loss": 3.3959,
+      "step": 19295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.529236376285553,
+      "learning_rate": 0.0004645359811795393,
+      "loss": 3.1483,
+      "step": 19296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5238991379737854,
+      "learning_rate": 0.0004645231069577938,
+      "loss": 3.0694,
+      "step": 19297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5467185974121094,
+      "learning_rate": 0.00046451023230272606,
+      "loss": 3.1344,
+      "step": 19298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5267390608787537,
+      "learning_rate": 0.00046449735721437014,
+      "loss": 2.9927,
+      "step": 19299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5365429520606995,
+      "learning_rate": 0.0004644844816927598,
+      "loss": 3.0123,
+      "step": 19300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5720606446266174,
+      "learning_rate": 0.000464471605737929,
+      "loss": 3.1872,
+      "step": 19301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5571929216384888,
+      "learning_rate": 0.00046445872934991163,
+      "loss": 3.2165,
+      "step": 19302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6748465299606323,
+      "learning_rate": 0.00046444585252874163,
+      "loss": 2.8737,
+      "step": 19303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5558235049247742,
+      "learning_rate": 0.00046443297527445286,
+      "loss": 3.1197,
+      "step": 19304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5244531631469727,
+      "learning_rate": 0.0004644200975870793,
+      "loss": 3.2594,
+      "step": 19305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5120171904563904,
+      "learning_rate": 0.0004644072194666549,
+      "loss": 3.2905,
+      "step": 19306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.502840518951416,
+      "learning_rate": 0.0004643943409132135,
+      "loss": 2.9992,
+      "step": 19307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5086363554000854,
+      "learning_rate": 0.000464381461926789,
+      "loss": 2.906,
+      "step": 19308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5207031965255737,
+      "learning_rate": 0.0004643685825074154,
+      "loss": 3.3385,
+      "step": 19309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.512380838394165,
+      "learning_rate": 0.0004643557026551266,
+      "loss": 3.2487,
+      "step": 19310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.544256865978241,
+      "learning_rate": 0.00046434282236995655,
+      "loss": 3.1312,
+      "step": 19311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5640758872032166,
+      "learning_rate": 0.000464329941651939,
+      "loss": 3.1362,
+      "step": 19312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5477793216705322,
+      "learning_rate": 0.0004643170605011081,
+      "loss": 2.9544,
+      "step": 19313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5070680975914001,
+      "learning_rate": 0.00046430417891749764,
+      "loss": 3.1279,
+      "step": 19314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5301992893218994,
+      "learning_rate": 0.00046429129690114167,
+      "loss": 3.2035,
+      "step": 19315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5810301899909973,
+      "learning_rate": 0.00046427841445207394,
+      "loss": 3.1866,
+      "step": 19316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.544684112071991,
+      "learning_rate": 0.00046426553157032855,
+      "loss": 3.2024,
+      "step": 19317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5401524305343628,
+      "learning_rate": 0.0004642526482559394,
+      "loss": 2.8806,
+      "step": 19318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6001310348510742,
+      "learning_rate": 0.0004642397645089403,
+      "loss": 3.1245,
+      "step": 19319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5786098837852478,
+      "learning_rate": 0.0004642268803293653,
+      "loss": 3.1757,
+      "step": 19320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.579869270324707,
+      "learning_rate": 0.00046421399571724834,
+      "loss": 3.1806,
+      "step": 19321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5166104435920715,
+      "learning_rate": 0.0004642011106726233,
+      "loss": 3.0663,
+      "step": 19322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5852053165435791,
+      "learning_rate": 0.00046418822519552416,
+      "loss": 3.1294,
+      "step": 19323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4941536784172058,
+      "learning_rate": 0.0004641753392859847,
+      "loss": 3.0502,
+      "step": 19324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5297553539276123,
+      "learning_rate": 0.00046416245294403916,
+      "loss": 3.356,
+      "step": 19325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5243952870368958,
+      "learning_rate": 0.00046414956616972126,
+      "loss": 3.1456,
+      "step": 19326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5288766026496887,
+      "learning_rate": 0.00046413667896306495,
+      "loss": 2.9972,
+      "step": 19327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5393479466438293,
+      "learning_rate": 0.00046412379132410427,
+      "loss": 2.8842,
+      "step": 19328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49066945910453796,
+      "learning_rate": 0.0004641109032528731,
+      "loss": 2.9538,
+      "step": 19329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.509368360042572,
+      "learning_rate": 0.0004640980147494053,
+      "loss": 3.1996,
+      "step": 19330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5085881948471069,
+      "learning_rate": 0.00046408512581373507,
+      "loss": 3.1171,
+      "step": 19331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5356560349464417,
+      "learning_rate": 0.00046407223644589606,
+      "loss": 3.0997,
+      "step": 19332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5444095730781555,
+      "learning_rate": 0.0004640593466459225,
+      "loss": 2.9612,
+      "step": 19333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5353986620903015,
+      "learning_rate": 0.00046404645641384804,
+      "loss": 2.9857,
+      "step": 19334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5451989769935608,
+      "learning_rate": 0.00046403356574970683,
+      "loss": 3.2518,
+      "step": 19335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5122106671333313,
+      "learning_rate": 0.0004640206746535328,
+      "loss": 2.9274,
+      "step": 19336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.512254536151886,
+      "learning_rate": 0.00046400778312536,
+      "loss": 3.2984,
+      "step": 19337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5183076858520508,
+      "learning_rate": 0.00046399489116522204,
+      "loss": 3.2875,
+      "step": 19338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5587219595909119,
+      "learning_rate": 0.00046398199877315313,
+      "loss": 3.0897,
+      "step": 19339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5293834209442139,
+      "learning_rate": 0.00046396910594918736,
+      "loss": 3.1424,
+      "step": 19340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5190085172653198,
+      "learning_rate": 0.0004639562126933584,
+      "loss": 3.1517,
+      "step": 19341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5054040551185608,
+      "learning_rate": 0.0004639433190057004,
+      "loss": 3.1675,
+      "step": 19342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5302183628082275,
+      "learning_rate": 0.0004639304248862472,
+      "loss": 3.322,
+      "step": 19343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5174694657325745,
+      "learning_rate": 0.0004639175303350328,
+      "loss": 3.1216,
+      "step": 19344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5287314057350159,
+      "learning_rate": 0.00046390463535209115,
+      "loss": 2.9587,
+      "step": 19345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49835264682769775,
+      "learning_rate": 0.0004638917399374563,
+      "loss": 3.1747,
+      "step": 19346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5517767071723938,
+      "learning_rate": 0.0004638788440911621,
+      "loss": 3.1357,
+      "step": 19347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5469534993171692,
+      "learning_rate": 0.0004638659478132426,
+      "loss": 3.1799,
+      "step": 19348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5432090163230896,
+      "learning_rate": 0.0004638530511037317,
+      "loss": 3.2207,
+      "step": 19349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6552304029464722,
+      "learning_rate": 0.00046384015396266345,
+      "loss": 3.1993,
+      "step": 19350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5897645354270935,
+      "learning_rate": 0.00046382725639007165,
+      "loss": 3.1047,
+      "step": 19351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5444059371948242,
+      "learning_rate": 0.0004638143583859905,
+      "loss": 3.164,
+      "step": 19352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5374630093574524,
+      "learning_rate": 0.0004638014599504539,
+      "loss": 2.9453,
+      "step": 19353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5767490267753601,
+      "learning_rate": 0.0004637885610834956,
+      "loss": 3.1528,
+      "step": 19354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5264842510223389,
+      "learning_rate": 0.0004637756617851499,
+      "loss": 2.9574,
+      "step": 19355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5502093434333801,
+      "learning_rate": 0.00046376276205545053,
+      "loss": 3.3082,
+      "step": 19356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5488525629043579,
+      "learning_rate": 0.00046374986189443165,
+      "loss": 2.9616,
+      "step": 19357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5013023018836975,
+      "learning_rate": 0.0004637369613021271,
+      "loss": 3.1773,
+      "step": 19358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5348243713378906,
+      "learning_rate": 0.0004637240602785709,
+      "loss": 3.1887,
+      "step": 19359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5497328042984009,
+      "learning_rate": 0.00046371115882379706,
+      "loss": 3.0564,
+      "step": 19360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.545487105846405,
+      "learning_rate": 0.0004636982569378395,
+      "loss": 3.0005,
+      "step": 19361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5591291785240173,
+      "learning_rate": 0.0004636853546207323,
+      "loss": 3.0755,
+      "step": 19362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5495262145996094,
+      "learning_rate": 0.0004636724518725093,
+      "loss": 3.4243,
+      "step": 19363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5503337979316711,
+      "learning_rate": 0.00046365954869320464,
+      "loss": 2.9641,
+      "step": 19364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.508242130279541,
+      "learning_rate": 0.00046364664508285217,
+      "loss": 3.2054,
+      "step": 19365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5202712416648865,
+      "learning_rate": 0.0004636337410414859,
+      "loss": 2.9891,
+      "step": 19366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.57612144947052,
+      "learning_rate": 0.0004636208365691399,
+      "loss": 3.0345,
+      "step": 19367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5074007511138916,
+      "learning_rate": 0.0004636079316658481,
+      "loss": 3.0326,
+      "step": 19368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5342861413955688,
+      "learning_rate": 0.0004635950263316445,
+      "loss": 3.0882,
+      "step": 19369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.531694769859314,
+      "learning_rate": 0.00046358212056656306,
+      "loss": 3.249,
+      "step": 19370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5404502749443054,
+      "learning_rate": 0.00046356921437063777,
+      "loss": 3.159,
+      "step": 19371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5267776250839233,
+      "learning_rate": 0.00046355630774390274,
+      "loss": 3.1012,
+      "step": 19372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390035510063171,
+      "learning_rate": 0.0004635434006863919,
+      "loss": 3.223,
+      "step": 19373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49688008427619934,
+      "learning_rate": 0.00046353049319813904,
+      "loss": 3.1074,
+      "step": 19374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.557471752166748,
+      "learning_rate": 0.00046351758527917846,
+      "loss": 2.9832,
+      "step": 19375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4971594214439392,
+      "learning_rate": 0.000463504676929544,
+      "loss": 3.2285,
+      "step": 19376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5553884506225586,
+      "learning_rate": 0.0004634917681492697,
+      "loss": 3.1043,
+      "step": 19377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5280371308326721,
+      "learning_rate": 0.00046347885893838957,
+      "loss": 3.2346,
+      "step": 19378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5034427046775818,
+      "learning_rate": 0.0004634659492969376,
+      "loss": 3.1326,
+      "step": 19379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5268436670303345,
+      "learning_rate": 0.0004634530392249476,
+      "loss": 3.2807,
+      "step": 19380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234237909317017,
+      "learning_rate": 0.000463440128722454,
+      "loss": 3.1346,
+      "step": 19381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.539435863494873,
+      "learning_rate": 0.0004634272177894904,
+      "loss": 3.1197,
+      "step": 19382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5603162050247192,
+      "learning_rate": 0.000463414306426091,
+      "loss": 3.0302,
+      "step": 19383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5288932919502258,
+      "learning_rate": 0.0004634013946322898,
+      "loss": 3.2148,
+      "step": 19384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5117092728614807,
+      "learning_rate": 0.00046338848240812063,
+      "loss": 3.3284,
+      "step": 19385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5178090929985046,
+      "learning_rate": 0.0004633755697536178,
+      "loss": 3.2783,
+      "step": 19386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5456027388572693,
+      "learning_rate": 0.0004633626566688152,
+      "loss": 2.9413,
+      "step": 19387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4823963940143585,
+      "learning_rate": 0.0004633497431537467,
+      "loss": 3.0158,
+      "step": 19388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5608025193214417,
+      "learning_rate": 0.0004633368292084464,
+      "loss": 3.2938,
+      "step": 19389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5916173458099365,
+      "learning_rate": 0.0004633239148329483,
+      "loss": 3.0889,
+      "step": 19390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5329937934875488,
+      "learning_rate": 0.0004633110000272866,
+      "loss": 2.8621,
+      "step": 19391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5109493732452393,
+      "learning_rate": 0.00046329808479149496,
+      "loss": 2.8752,
+      "step": 19392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5147404074668884,
+      "learning_rate": 0.0004632851691256078,
+      "loss": 3.2002,
+      "step": 19393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5089914798736572,
+      "learning_rate": 0.00046327225302965873,
+      "loss": 3.2138,
+      "step": 19394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5555095672607422,
+      "learning_rate": 0.00046325933650368206,
+      "loss": 3.2143,
+      "step": 19395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6058218479156494,
+      "learning_rate": 0.0004632464195477118,
+      "loss": 3.1112,
+      "step": 19396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5387574434280396,
+      "learning_rate": 0.00046323350216178174,
+      "loss": 3.0377,
+      "step": 19397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5902401804924011,
+      "learning_rate": 0.00046322058434592617,
+      "loss": 3.0203,
+      "step": 19398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5146337747573853,
+      "learning_rate": 0.0004632076661001789,
+      "loss": 2.8675,
+      "step": 19399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.508232057094574,
+      "learning_rate": 0.0004631947474245741,
+      "loss": 3.2896,
+      "step": 19400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5544096827507019,
+      "learning_rate": 0.00046318182831914565,
+      "loss": 3.0256,
+      "step": 19401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5086759328842163,
+      "learning_rate": 0.0004631689087839278,
+      "loss": 3.1559,
+      "step": 19402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5778783559799194,
+      "learning_rate": 0.0004631559888189544,
+      "loss": 3.1122,
+      "step": 19403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5708818435668945,
+      "learning_rate": 0.00046314306842425954,
+      "loss": 3.1534,
+      "step": 19404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5399811863899231,
+      "learning_rate": 0.00046313014759987716,
+      "loss": 3.0594,
+      "step": 19405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5436029434204102,
+      "learning_rate": 0.0004631172263458414,
+      "loss": 3.0062,
+      "step": 19406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5369762182235718,
+      "learning_rate": 0.0004631043046621863,
+      "loss": 3.0197,
+      "step": 19407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5159459710121155,
+      "learning_rate": 0.00046309138254894586,
+      "loss": 3.0661,
+      "step": 19408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5201807618141174,
+      "learning_rate": 0.00046307846000615406,
+      "loss": 3.1795,
+      "step": 19409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5022982954978943,
+      "learning_rate": 0.000463065537033845,
+      "loss": 3.2928,
+      "step": 19410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5162597298622131,
+      "learning_rate": 0.0004630526136320527,
+      "loss": 3.4038,
+      "step": 19411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5503665804862976,
+      "learning_rate": 0.0004630396898008112,
+      "loss": 3.1428,
+      "step": 19412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5348483920097351,
+      "learning_rate": 0.0004630267655401545,
+      "loss": 2.9746,
+      "step": 19413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5585376024246216,
+      "learning_rate": 0.00046301384085011666,
+      "loss": 3.4144,
+      "step": 19414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5287220478057861,
+      "learning_rate": 0.0004630009157307319,
+      "loss": 3.0886,
+      "step": 19415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5285464525222778,
+      "learning_rate": 0.00046298799018203385,
+      "loss": 3.0322,
+      "step": 19416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5362486839294434,
+      "learning_rate": 0.00046297506420405697,
+      "loss": 3.1066,
+      "step": 19417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5312342643737793,
+      "learning_rate": 0.00046296213779683506,
+      "loss": 3.1128,
+      "step": 19418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5306200981140137,
+      "learning_rate": 0.0004629492109604023,
+      "loss": 3.0913,
+      "step": 19419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5508071780204773,
+      "learning_rate": 0.0004629362836947927,
+      "loss": 3.2167,
+      "step": 19420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5401845574378967,
+      "learning_rate": 0.0004629233560000401,
+      "loss": 2.9779,
+      "step": 19421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5715287327766418,
+      "learning_rate": 0.00046291042787617896,
+      "loss": 3.215,
+      "step": 19422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5305150747299194,
+      "learning_rate": 0.000462897499323243,
+      "loss": 3.1558,
+      "step": 19423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5180661082267761,
+      "learning_rate": 0.00046288457034126645,
+      "loss": 3.02,
+      "step": 19424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5526053309440613,
+      "learning_rate": 0.0004628716409302832,
+      "loss": 3.0368,
+      "step": 19425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5305153727531433,
+      "learning_rate": 0.00046285871109032743,
+      "loss": 3.0802,
+      "step": 19426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5597370266914368,
+      "learning_rate": 0.00046284578082143315,
+      "loss": 3.0115,
+      "step": 19427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5298964977264404,
+      "learning_rate": 0.00046283285012363446,
+      "loss": 3.121,
+      "step": 19428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4885500371456146,
+      "learning_rate": 0.00046281991899696533,
+      "loss": 3.2538,
+      "step": 19429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5146791934967041,
+      "learning_rate": 0.00046280698744145994,
+      "loss": 3.0759,
+      "step": 19430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5360423922538757,
+      "learning_rate": 0.0004627940554571522,
+      "loss": 3.0549,
+      "step": 19431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5529478192329407,
+      "learning_rate": 0.0004627811230440763,
+      "loss": 3.1961,
+      "step": 19432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5056607723236084,
+      "learning_rate": 0.00046276819020226617,
+      "loss": 2.9044,
+      "step": 19433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5211524367332458,
+      "learning_rate": 0.000462755256931756,
+      "loss": 3.2076,
+      "step": 19434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5306402444839478,
+      "learning_rate": 0.00046274232323257984,
+      "loss": 3.2183,
+      "step": 19435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5615695714950562,
+      "learning_rate": 0.00046272938910477167,
+      "loss": 3.0429,
+      "step": 19436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5487105250358582,
+      "learning_rate": 0.0004627164545483656,
+      "loss": 3.0273,
+      "step": 19437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5336460471153259,
+      "learning_rate": 0.0004627035195633958,
+      "loss": 3.3118,
+      "step": 19438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5271636247634888,
+      "learning_rate": 0.00046269058414989613,
+      "loss": 3.1076,
+      "step": 19439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5462597012519836,
+      "learning_rate": 0.00046267764830790084,
+      "loss": 3.1968,
+      "step": 19440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5113808512687683,
+      "learning_rate": 0.000462664712037444,
+      "loss": 3.2844,
+      "step": 19441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5376443266868591,
+      "learning_rate": 0.0004626517753385595,
+      "loss": 3.1548,
+      "step": 19442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49754318594932556,
+      "learning_rate": 0.0004626388382112815,
+      "loss": 3.083,
+      "step": 19443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4905608594417572,
+      "learning_rate": 0.0004626259006556441,
+      "loss": 3.3342,
+      "step": 19444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5019567012786865,
+      "learning_rate": 0.00046261296267168146,
+      "loss": 3.1041,
+      "step": 19445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5304019451141357,
+      "learning_rate": 0.00046260002425942755,
+      "loss": 3.3963,
+      "step": 19446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5429536700248718,
+      "learning_rate": 0.0004625870854189165,
+      "loss": 3.04,
+      "step": 19447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5398834943771362,
+      "learning_rate": 0.0004625741461501823,
+      "loss": 3.1989,
+      "step": 19448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5834211707115173,
+      "learning_rate": 0.00046256120645325906,
+      "loss": 3.161,
+      "step": 19449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5296370387077332,
+      "learning_rate": 0.000462548266328181,
+      "loss": 3.0605,
+      "step": 19450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5220499634742737,
+      "learning_rate": 0.000462535325774982,
+      "loss": 3.2008,
+      "step": 19451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5129631757736206,
+      "learning_rate": 0.00046252238479369624,
+      "loss": 3.1372,
+      "step": 19452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5435758233070374,
+      "learning_rate": 0.0004625094433843579,
+      "loss": 3.0829,
+      "step": 19453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.57452392578125,
+      "learning_rate": 0.0004624965015470008,
+      "loss": 3.2124,
+      "step": 19454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5123562216758728,
+      "learning_rate": 0.0004624835592816593,
+      "loss": 3.2627,
+      "step": 19455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5587196350097656,
+      "learning_rate": 0.00046247061658836726,
+      "loss": 3.0844,
+      "step": 19456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5435320138931274,
+      "learning_rate": 0.000462457673467159,
+      "loss": 3.2731,
+      "step": 19457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5726426243782043,
+      "learning_rate": 0.0004624447299180685,
+      "loss": 3.237,
+      "step": 19458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5328337550163269,
+      "learning_rate": 0.0004624317859411298,
+      "loss": 3.0805,
+      "step": 19459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5529270172119141,
+      "learning_rate": 0.00046241884153637696,
+      "loss": 2.9795,
+      "step": 19460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5545976161956787,
+      "learning_rate": 0.00046240589670384425,
+      "loss": 3.0586,
+      "step": 19461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48107704520225525,
+      "learning_rate": 0.0004623929514435656,
+      "loss": 3.1515,
+      "step": 19462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.576178789138794,
+      "learning_rate": 0.0004623800057555752,
+      "loss": 3.2103,
+      "step": 19463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5457341074943542,
+      "learning_rate": 0.00046236705963990715,
+      "loss": 3.2126,
+      "step": 19464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5450583696365356,
+      "learning_rate": 0.0004623541130965955,
+      "loss": 3.0782,
+      "step": 19465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5224042534828186,
+      "learning_rate": 0.00046234116612567437,
+      "loss": 3.1618,
+      "step": 19466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5329810380935669,
+      "learning_rate": 0.00046232821872717783,
+      "loss": 3.2908,
+      "step": 19467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5755662322044373,
+      "learning_rate": 0.00046231527090113993,
+      "loss": 3.0585,
+      "step": 19468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5130234956741333,
+      "learning_rate": 0.000462302322647595,
+      "loss": 3.0353,
+      "step": 19469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5430505275726318,
+      "learning_rate": 0.0004622893739665769,
+      "loss": 3.3614,
+      "step": 19470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5734212398529053,
+      "learning_rate": 0.0004622764248581198,
+      "loss": 3.1137,
+      "step": 19471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5490341782569885,
+      "learning_rate": 0.00046226347532225777,
+      "loss": 2.9489,
+      "step": 19472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.506134569644928,
+      "learning_rate": 0.0004622505253590251,
+      "loss": 3.2452,
+      "step": 19473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5133436322212219,
+      "learning_rate": 0.00046223757496845577,
+      "loss": 3.1175,
+      "step": 19474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4948771595954895,
+      "learning_rate": 0.0004622246241505839,
+      "loss": 3.2184,
+      "step": 19475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.663033664226532,
+      "learning_rate": 0.0004622116729054435,
+      "loss": 3.2785,
+      "step": 19476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5038350224494934,
+      "learning_rate": 0.0004621987212330688,
+      "loss": 3.1118,
+      "step": 19477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5808413624763489,
+      "learning_rate": 0.00046218576913349387,
+      "loss": 3.2685,
+      "step": 19478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5481523275375366,
+      "learning_rate": 0.00046217281660675293,
+      "loss": 3.1188,
+      "step": 19479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5153646469116211,
+      "learning_rate": 0.0004621598636528799,
+      "loss": 3.2296,
+      "step": 19480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.539476215839386,
+      "learning_rate": 0.00046214691027190904,
+      "loss": 3.3285,
+      "step": 19481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4928472638130188,
+      "learning_rate": 0.0004621339564638744,
+      "loss": 3.1049,
+      "step": 19482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4804633557796478,
+      "learning_rate": 0.00046212100222881014,
+      "loss": 3.2119,
+      "step": 19483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.524301290512085,
+      "learning_rate": 0.0004621080475667504,
+      "loss": 3.103,
+      "step": 19484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50786954164505,
+      "learning_rate": 0.0004620950924777292,
+      "loss": 3.1123,
+      "step": 19485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5153059363365173,
+      "learning_rate": 0.0004620821369617808,
+      "loss": 3.3104,
+      "step": 19486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5356342196464539,
+      "learning_rate": 0.0004620691810189392,
+      "loss": 3.2367,
+      "step": 19487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5034014582633972,
+      "learning_rate": 0.00046205622464923847,
+      "loss": 3.2318,
+      "step": 19488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5269924402236938,
+      "learning_rate": 0.0004620432678527129,
+      "loss": 3.0648,
+      "step": 19489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5155195593833923,
+      "learning_rate": 0.00046203031062939666,
+      "loss": 3.0857,
+      "step": 19490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48004335165023804,
+      "learning_rate": 0.00046201735297932356,
+      "loss": 3.1868,
+      "step": 19491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5057433843612671,
+      "learning_rate": 0.0004620043949025281,
+      "loss": 3.1481,
+      "step": 19492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5076611638069153,
+      "learning_rate": 0.00046199143639904416,
+      "loss": 3.0926,
+      "step": 19493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5741182565689087,
+      "learning_rate": 0.000461978477468906,
+      "loss": 3.1103,
+      "step": 19494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5225886106491089,
+      "learning_rate": 0.00046196551811214775,
+      "loss": 3.1027,
+      "step": 19495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5760247111320496,
+      "learning_rate": 0.0004619525583288034,
+      "loss": 3.2087,
+      "step": 19496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5268136858940125,
+      "learning_rate": 0.0004619395981189072,
+      "loss": 3.1337,
+      "step": 19497
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5018905401229858,
+      "learning_rate": 0.0004619266374824932,
+      "loss": 3.2123,
+      "step": 19498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.509712815284729,
+      "learning_rate": 0.00046191367641959573,
+      "loss": 3.0303,
+      "step": 19499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5171404480934143,
+      "learning_rate": 0.00046190071493024874,
+      "loss": 3.0558,
+      "step": 19500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5230562686920166,
+      "learning_rate": 0.00046188775301448645,
+      "loss": 3.175,
+      "step": 19501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5551393032073975,
+      "learning_rate": 0.00046187479067234295,
+      "loss": 3.0652,
+      "step": 19502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5329045653343201,
+      "learning_rate": 0.0004618618279038524,
+      "loss": 3.0641,
+      "step": 19503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4875706732273102,
+      "learning_rate": 0.00046184886470904894,
+      "loss": 3.0987,
+      "step": 19504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5308302640914917,
+      "learning_rate": 0.00046183590108796673,
+      "loss": 3.156,
+      "step": 19505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5178402662277222,
+      "learning_rate": 0.0004618229370406399,
+      "loss": 3.0371,
+      "step": 19506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5128605365753174,
+      "learning_rate": 0.00046180997256710263,
+      "loss": 3.2408,
+      "step": 19507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5387104749679565,
+      "learning_rate": 0.000461797007667389,
+      "loss": 3.2067,
+      "step": 19508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5741128325462341,
+      "learning_rate": 0.0004617840423415332,
+      "loss": 3.2154,
+      "step": 19509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5508107542991638,
+      "learning_rate": 0.0004617710765895693,
+      "loss": 3.1757,
+      "step": 19510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5864234566688538,
+      "learning_rate": 0.00046175811041153164,
+      "loss": 3.1171,
+      "step": 19511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5283697247505188,
+      "learning_rate": 0.00046174514380745423,
+      "loss": 3.4572,
+      "step": 19512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4930495619773865,
+      "learning_rate": 0.0004617321767773711,
+      "loss": 2.9718,
+      "step": 19513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49955621361732483,
+      "learning_rate": 0.00046171920932131666,
+      "loss": 3.215,
+      "step": 19514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5142083764076233,
+      "learning_rate": 0.00046170624143932495,
+      "loss": 3.2683,
+      "step": 19515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4969509243965149,
+      "learning_rate": 0.0004616932731314301,
+      "loss": 3.0456,
+      "step": 19516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5548650622367859,
+      "learning_rate": 0.00046168030439766626,
+      "loss": 3.0429,
+      "step": 19517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5828065276145935,
+      "learning_rate": 0.00046166733523806774,
+      "loss": 3.3183,
+      "step": 19518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5464107990264893,
+      "learning_rate": 0.0004616543656526684,
+      "loss": 3.0603,
+      "step": 19519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5539250373840332,
+      "learning_rate": 0.0004616413956415026,
+      "loss": 3.2492,
+      "step": 19520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5596001744270325,
+      "learning_rate": 0.00046162842520460455,
+      "loss": 3.1842,
+      "step": 19521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5353409647941589,
+      "learning_rate": 0.00046161545434200837,
+      "loss": 3.1103,
+      "step": 19522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5830059051513672,
+      "learning_rate": 0.0004616024830537481,
+      "loss": 3.2124,
+      "step": 19523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.542768120765686,
+      "learning_rate": 0.0004615895113398579,
+      "loss": 3.1671,
+      "step": 19524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5151305198669434,
+      "learning_rate": 0.00046157653920037223,
+      "loss": 3.0929,
+      "step": 19525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5473406910896301,
+      "learning_rate": 0.000461563566635325,
+      "loss": 3.0294,
+      "step": 19526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5182105302810669,
+      "learning_rate": 0.00046155059364475035,
+      "loss": 2.9818,
+      "step": 19527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5524483323097229,
+      "learning_rate": 0.0004615376202286825,
+      "loss": 3.2716,
+      "step": 19528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5527673363685608,
+      "learning_rate": 0.0004615246463871557,
+      "loss": 3.0887,
+      "step": 19529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5192067623138428,
+      "learning_rate": 0.0004615116721202041,
+      "loss": 3.1964,
+      "step": 19530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5294162631034851,
+      "learning_rate": 0.0004614986974278618,
+      "loss": 3.1753,
+      "step": 19531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5797945261001587,
+      "learning_rate": 0.0004614857223101631,
+      "loss": 3.1273,
+      "step": 19532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5336905717849731,
+      "learning_rate": 0.000461472746767142,
+      "loss": 3.2099,
+      "step": 19533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5431013703346252,
+      "learning_rate": 0.0004614597707988327,
+      "loss": 3.2577,
+      "step": 19534
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5563946962356567,
+      "learning_rate": 0.0004614467944052695,
+      "loss": 2.8293,
+      "step": 19535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5071994662284851,
+      "learning_rate": 0.00046143381758648666,
+      "loss": 3.0538,
+      "step": 19536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5291922092437744,
+      "learning_rate": 0.000461420840342518,
+      "loss": 3.0829,
+      "step": 19537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5239132642745972,
+      "learning_rate": 0.000461407862673398,
+      "loss": 3.1648,
+      "step": 19538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5256131887435913,
+      "learning_rate": 0.00046139488457916074,
+      "loss": 3.1761,
+      "step": 19539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5487732887268066,
+      "learning_rate": 0.0004613819060598404,
+      "loss": 3.2024,
+      "step": 19540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5254712700843811,
+      "learning_rate": 0.00046136892711547123,
+      "loss": 3.1627,
+      "step": 19541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5246246457099915,
+      "learning_rate": 0.00046135594774608733,
+      "loss": 3.0529,
+      "step": 19542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5375425219535828,
+      "learning_rate": 0.0004613429679517229,
+      "loss": 3.3443,
+      "step": 19543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5527751445770264,
+      "learning_rate": 0.00046132998773241215,
+      "loss": 2.9413,
+      "step": 19544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5222285389900208,
+      "learning_rate": 0.00046131700708818924,
+      "loss": 3.3199,
+      "step": 19545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.577925443649292,
+      "learning_rate": 0.0004613040260190884,
+      "loss": 2.9484,
+      "step": 19546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5509508848190308,
+      "learning_rate": 0.0004612910445251439,
+      "loss": 3.2603,
+      "step": 19547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5331478714942932,
+      "learning_rate": 0.0004612780626063897,
+      "loss": 3.0465,
+      "step": 19548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5440930128097534,
+      "learning_rate": 0.0004612650802628602,
+      "loss": 2.9711,
+      "step": 19549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5174834728240967,
+      "learning_rate": 0.0004612520974945895,
+      "loss": 3.2217,
+      "step": 19550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5301517844200134,
+      "learning_rate": 0.00046123911430161175,
+      "loss": 3.1833,
+      "step": 19551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.556764543056488,
+      "learning_rate": 0.00046122613068396123,
+      "loss": 2.9512,
+      "step": 19552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5274459719657898,
+      "learning_rate": 0.00046121314664167214,
+      "loss": 3.0559,
+      "step": 19553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5233287811279297,
+      "learning_rate": 0.0004612001621747786,
+      "loss": 3.1821,
+      "step": 19554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49667537212371826,
+      "learning_rate": 0.00046118717728331493,
+      "loss": 3.1545,
+      "step": 19555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5035794377326965,
+      "learning_rate": 0.00046117419196731516,
+      "loss": 2.9625,
+      "step": 19556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.536008894443512,
+      "learning_rate": 0.00046116120622681365,
+      "loss": 3.3597,
+      "step": 19557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5389508008956909,
+      "learning_rate": 0.0004611482200618446,
+      "loss": 3.0717,
+      "step": 19558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4923076629638672,
+      "learning_rate": 0.00046113523347244206,
+      "loss": 3.066,
+      "step": 19559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4959174394607544,
+      "learning_rate": 0.00046112224645864023,
+      "loss": 3.1596,
+      "step": 19560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5146404504776001,
+      "learning_rate": 0.0004611092590204736,
+      "loss": 3.1807,
+      "step": 19561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5296205282211304,
+      "learning_rate": 0.0004610962711579761,
+      "loss": 3.0542,
+      "step": 19562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5231249928474426,
+      "learning_rate": 0.00046108328287118203,
+      "loss": 3.3756,
+      "step": 19563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5085944533348083,
+      "learning_rate": 0.0004610702941601256,
+      "loss": 3.316,
+      "step": 19564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5056644082069397,
+      "learning_rate": 0.00046105730502484107,
+      "loss": 3.0114,
+      "step": 19565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5329414010047913,
+      "learning_rate": 0.00046104431546536246,
+      "loss": 3.3503,
+      "step": 19566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5404815077781677,
+      "learning_rate": 0.00046103132548172424,
+      "loss": 3.1329,
+      "step": 19567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5342490673065186,
+      "learning_rate": 0.00046101833507396044,
+      "loss": 3.1882,
+      "step": 19568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5149338245391846,
+      "learning_rate": 0.0004610053442421054,
+      "loss": 3.2602,
+      "step": 19569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5321595072746277,
+      "learning_rate": 0.0004609923529861932,
+      "loss": 3.0032,
+      "step": 19570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5394263863563538,
+      "learning_rate": 0.0004609793613062581,
+      "loss": 3.2131,
+      "step": 19571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5631637573242188,
+      "learning_rate": 0.00046096636920233444,
+      "loss": 3.0242,
+      "step": 19572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5770866870880127,
+      "learning_rate": 0.00046095337667445633,
+      "loss": 3.1954,
+      "step": 19573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5115048289299011,
+      "learning_rate": 0.00046094038372265794,
+      "loss": 3.2293,
+      "step": 19574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5524685978889465,
+      "learning_rate": 0.0004609273903469735,
+      "loss": 3.208,
+      "step": 19575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5510011315345764,
+      "learning_rate": 0.00046091439654743745,
+      "loss": 3.228,
+      "step": 19576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5279592275619507,
+      "learning_rate": 0.00046090140232408377,
+      "loss": 2.9963,
+      "step": 19577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5411564111709595,
+      "learning_rate": 0.00046088840767694674,
+      "loss": 3.2272,
+      "step": 19578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5248085260391235,
+      "learning_rate": 0.00046087541260606056,
+      "loss": 3.2172,
+      "step": 19579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5476329922676086,
+      "learning_rate": 0.0004608624171114596,
+      "loss": 3.3069,
+      "step": 19580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5507305264472961,
+      "learning_rate": 0.000460849421193178,
+      "loss": 3.096,
+      "step": 19581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5406851172447205,
+      "learning_rate": 0.00046083642485124983,
+      "loss": 3.0172,
+      "step": 19582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5372514724731445,
+      "learning_rate": 0.0004608234280857096,
+      "loss": 3.1018,
+      "step": 19583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5436272621154785,
+      "learning_rate": 0.00046081043089659144,
+      "loss": 3.1314,
+      "step": 19584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5231220126152039,
+      "learning_rate": 0.0004607974332839295,
+      "loss": 3.1906,
+      "step": 19585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.553106963634491,
+      "learning_rate": 0.000460784435247758,
+      "loss": 3.1702,
+      "step": 19586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5628478527069092,
+      "learning_rate": 0.0004607714367881114,
+      "loss": 3.1288,
+      "step": 19587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5031068921089172,
+      "learning_rate": 0.0004607584379050237,
+      "loss": 3.193,
+      "step": 19588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5432695746421814,
+      "learning_rate": 0.00046074543859852917,
+      "loss": 3.0016,
+      "step": 19589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5564255118370056,
+      "learning_rate": 0.00046073243886866216,
+      "loss": 3.1017,
+      "step": 19590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7009369134902954,
+      "learning_rate": 0.0004607194387154567,
+      "loss": 3.213,
+      "step": 19591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5528953671455383,
+      "learning_rate": 0.0004607064381389473,
+      "loss": 3.1717,
+      "step": 19592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5072028636932373,
+      "learning_rate": 0.00046069343713916805,
+      "loss": 3.1961,
+      "step": 19593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5909320712089539,
+      "learning_rate": 0.0004606804357161532,
+      "loss": 3.2262,
+      "step": 19594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5035519003868103,
+      "learning_rate": 0.0004606674338699371,
+      "loss": 2.93,
+      "step": 19595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5433776378631592,
+      "learning_rate": 0.0004606544316005537,
+      "loss": 3.5309,
+      "step": 19596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.550793468952179,
+      "learning_rate": 0.00046064142890803764,
+      "loss": 3.2649,
+      "step": 19597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5635651350021362,
+      "learning_rate": 0.00046062842579242284,
+      "loss": 3.0402,
+      "step": 19598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5558289289474487,
+      "learning_rate": 0.0004606154222537438,
+      "loss": 3.2313,
+      "step": 19599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5349392294883728,
+      "learning_rate": 0.0004606024182920345,
+      "loss": 3.2041,
+      "step": 19600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48463213443756104,
+      "learning_rate": 0.0004605894139073294,
+      "loss": 3.0712,
+      "step": 19601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5352935194969177,
+      "learning_rate": 0.00046057640909966276,
+      "loss": 3.1331,
+      "step": 19602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5178282856941223,
+      "learning_rate": 0.00046056340386906866,
+      "loss": 3.1748,
+      "step": 19603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5154528617858887,
+      "learning_rate": 0.0004605503982155815,
+      "loss": 3.3167,
+      "step": 19604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.525182843208313,
+      "learning_rate": 0.0004605373921392355,
+      "loss": 3.2507,
+      "step": 19605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.549490213394165,
+      "learning_rate": 0.0004605243856400649,
+      "loss": 3.0423,
+      "step": 19606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546172022819519,
+      "learning_rate": 0.00046051137871810395,
+      "loss": 3.1048,
+      "step": 19607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5274287462234497,
+      "learning_rate": 0.00046049837137338695,
+      "loss": 3.0897,
+      "step": 19608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5080083012580872,
+      "learning_rate": 0.00046048536360594815,
+      "loss": 3.1981,
+      "step": 19609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5662698745727539,
+      "learning_rate": 0.00046047235541582174,
+      "loss": 3.2204,
+      "step": 19610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49622294306755066,
+      "learning_rate": 0.00046045934680304194,
+      "loss": 2.9424,
+      "step": 19611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5338951945304871,
+      "learning_rate": 0.0004604463377676432,
+      "loss": 3.0564,
+      "step": 19612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5102505683898926,
+      "learning_rate": 0.00046043332830965973,
+      "loss": 3.1175,
+      "step": 19613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5088198781013489,
+      "learning_rate": 0.00046042031842912575,
+      "loss": 3.0823,
+      "step": 19614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4949013590812683,
+      "learning_rate": 0.0004604073081260754,
+      "loss": 3.1718,
+      "step": 19615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.529629647731781,
+      "learning_rate": 0.00046039429740054314,
+      "loss": 3.2467,
+      "step": 19616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5418373942375183,
+      "learning_rate": 0.0004603812862525632,
+      "loss": 3.0991,
+      "step": 19617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5319532155990601,
+      "learning_rate": 0.00046036827468216976,
+      "loss": 3.222,
+      "step": 19618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5246682167053223,
+      "learning_rate": 0.0004603552626893972,
+      "loss": 3.1567,
+      "step": 19619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5561453700065613,
+      "learning_rate": 0.00046034225027427974,
+      "loss": 3.1118,
+      "step": 19620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.554530143737793,
+      "learning_rate": 0.0004603292374368516,
+      "loss": 3.1151,
+      "step": 19621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5948189496994019,
+      "learning_rate": 0.0004603162241771471,
+      "loss": 3.2595,
+      "step": 19622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5682088136672974,
+      "learning_rate": 0.0004603032104952006,
+      "loss": 3.2056,
+      "step": 19623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5129702091217041,
+      "learning_rate": 0.0004602901963910463,
+      "loss": 3.1712,
+      "step": 19624
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5358684659004211,
+      "learning_rate": 0.0004602771818647184,
+      "loss": 3.0202,
+      "step": 19625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5510245561599731,
+      "learning_rate": 0.0004602641669162512,
+      "loss": 3.1402,
+      "step": 19626
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5696225762367249,
+      "learning_rate": 0.00046025115154567917,
+      "loss": 3.0832,
+      "step": 19627
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5569515824317932,
+      "learning_rate": 0.0004602381357530364,
+      "loss": 2.9884,
+      "step": 19628
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5148624181747437,
+      "learning_rate": 0.00046022511953835716,
+      "loss": 3.1615,
+      "step": 19629
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5084801316261292,
+      "learning_rate": 0.00046021210290167583,
+      "loss": 3.1071,
+      "step": 19630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5485509634017944,
+      "learning_rate": 0.00046019908584302665,
+      "loss": 3.0196,
+      "step": 19631
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5287085175514221,
+      "learning_rate": 0.0004601860683624439,
+      "loss": 3.1482,
+      "step": 19632
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4873697757720947,
+      "learning_rate": 0.00046017305045996187,
+      "loss": 3.2477,
+      "step": 19633
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5111532807350159,
+      "learning_rate": 0.0004601600321356149,
+      "loss": 3.2722,
+      "step": 19634
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5960693359375,
+      "learning_rate": 0.00046014701338943723,
+      "loss": 2.9375,
+      "step": 19635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5306110382080078,
+      "learning_rate": 0.00046013399422146307,
+      "loss": 3.0017,
+      "step": 19636
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5201436877250671,
+      "learning_rate": 0.0004601209746317268,
+      "loss": 3.2602,
+      "step": 19637
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5063363909721375,
+      "learning_rate": 0.0004601079546202628,
+      "loss": 3.1268,
+      "step": 19638
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5271068215370178,
+      "learning_rate": 0.00046009493418710514,
+      "loss": 3.2246,
+      "step": 19639
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.527073860168457,
+      "learning_rate": 0.00046008191333228826,
+      "loss": 3.0653,
+      "step": 19640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5272981524467468,
+      "learning_rate": 0.0004600688920558465,
+      "loss": 3.215,
+      "step": 19641
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5382295846939087,
+      "learning_rate": 0.000460055870357814,
+      "loss": 3.0829,
+      "step": 19642
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5190293192863464,
+      "learning_rate": 0.0004600428482382252,
+      "loss": 3.1779,
+      "step": 19643
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.508520781993866,
+      "learning_rate": 0.0004600298256971143,
+      "loss": 3.2162,
+      "step": 19644
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5364028215408325,
+      "learning_rate": 0.0004600168027345156,
+      "loss": 3.2912,
+      "step": 19645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.541970431804657,
+      "learning_rate": 0.00046000377935046357,
+      "loss": 3.2242,
+      "step": 19646
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.522739052772522,
+      "learning_rate": 0.0004599907555449922,
+      "loss": 3.1251,
+      "step": 19647
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5367051959037781,
+      "learning_rate": 0.00045997773131813607,
+      "loss": 3.1206,
+      "step": 19648
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5582072734832764,
+      "learning_rate": 0.0004599647066699295,
+      "loss": 3.1991,
+      "step": 19649
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5519962310791016,
+      "learning_rate": 0.00045995168160040654,
+      "loss": 3.0101,
+      "step": 19650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.529492199420929,
+      "learning_rate": 0.0004599386561096016,
+      "loss": 3.188,
+      "step": 19651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5170547366142273,
+      "learning_rate": 0.0004599256301975491,
+      "loss": 3.1679,
+      "step": 19652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5805275440216064,
+      "learning_rate": 0.00045991260386428327,
+      "loss": 3.0967,
+      "step": 19653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5890424847602844,
+      "learning_rate": 0.0004598995771098384,
+      "loss": 3.3168,
+      "step": 19654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5336859226226807,
+      "learning_rate": 0.0004598865499342488,
+      "loss": 3.0317,
+      "step": 19655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6045253872871399,
+      "learning_rate": 0.0004598735223375488,
+      "loss": 3.1161,
+      "step": 19656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5241043567657471,
+      "learning_rate": 0.00045986049431977265,
+      "loss": 3.141,
+      "step": 19657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515809178352356,
+      "learning_rate": 0.0004598474658809548,
+      "loss": 3.1444,
+      "step": 19658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5346136093139648,
+      "learning_rate": 0.00045983443702112946,
+      "loss": 3.181,
+      "step": 19659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5317142605781555,
+      "learning_rate": 0.00045982140774033104,
+      "loss": 3.4307,
+      "step": 19660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.558459997177124,
+      "learning_rate": 0.0004598083780385938,
+      "loss": 3.2656,
+      "step": 19661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5433655381202698,
+      "learning_rate": 0.0004597953479159519,
+      "loss": 3.1691,
+      "step": 19662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5130866765975952,
+      "learning_rate": 0.0004597823173724399,
+      "loss": 3.2532,
+      "step": 19663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5559121966362,
+      "learning_rate": 0.0004597692864080921,
+      "loss": 3.3065,
+      "step": 19664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4829885959625244,
+      "learning_rate": 0.0004597562550229426,
+      "loss": 3.2244,
+      "step": 19665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5515817403793335,
+      "learning_rate": 0.00045974322321702595,
+      "loss": 3.0803,
+      "step": 19666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5156745314598083,
+      "learning_rate": 0.0004597301909903764,
+      "loss": 3.254,
+      "step": 19667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.574837863445282,
+      "learning_rate": 0.0004597171583430282,
+      "loss": 3.2968,
+      "step": 19668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5619298219680786,
+      "learning_rate": 0.0004597041252750158,
+      "loss": 3.1675,
+      "step": 19669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5266140103340149,
+      "learning_rate": 0.00045969109178637345,
+      "loss": 3.226,
+      "step": 19670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5088499784469604,
+      "learning_rate": 0.00045967805787713563,
+      "loss": 3.0859,
+      "step": 19671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5350180268287659,
+      "learning_rate": 0.0004596650235473364,
+      "loss": 3.1795,
+      "step": 19672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5337379574775696,
+      "learning_rate": 0.0004596519887970102,
+      "loss": 3.135,
+      "step": 19673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5811973810195923,
+      "learning_rate": 0.0004596389536261914,
+      "loss": 3.3238,
+      "step": 19674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5261113047599792,
+      "learning_rate": 0.00045962591803491444,
+      "loss": 3.3253,
+      "step": 19675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5248528718948364,
+      "learning_rate": 0.00045961288202321345,
+      "loss": 3.1994,
+      "step": 19676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.510138750076294,
+      "learning_rate": 0.0004595998455911228,
+      "loss": 3.3404,
+      "step": 19677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5522067546844482,
+      "learning_rate": 0.0004595868087386769,
+      "loss": 2.9776,
+      "step": 19678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7141591906547546,
+      "learning_rate": 0.0004595737714659101,
+      "loss": 3.2153,
+      "step": 19679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49402374029159546,
+      "learning_rate": 0.00045956073377285663,
+      "loss": 3.3181,
+      "step": 19680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5236515402793884,
+      "learning_rate": 0.00045954769565955094,
+      "loss": 3.3036,
+      "step": 19681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.525768518447876,
+      "learning_rate": 0.0004595346571260274,
+      "loss": 3.1389,
+      "step": 19682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6157646179199219,
+      "learning_rate": 0.00045952161817232014,
+      "loss": 3.1411,
+      "step": 19683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5027342438697815,
+      "learning_rate": 0.00045950857879846366,
+      "loss": 3.3379,
+      "step": 19684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5160840153694153,
+      "learning_rate": 0.00045949553900449233,
+      "loss": 3.1588,
+      "step": 19685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5430676341056824,
+      "learning_rate": 0.00045948249879044047,
+      "loss": 3.1516,
+      "step": 19686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5123657584190369,
+      "learning_rate": 0.00045946945815634235,
+      "loss": 2.9488,
+      "step": 19687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4994308054447174,
+      "learning_rate": 0.0004594564171022323,
+      "loss": 3.1864,
+      "step": 19688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5245607495307922,
+      "learning_rate": 0.0004594433756281449,
+      "loss": 3.1557,
+      "step": 19689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.508391261100769,
+      "learning_rate": 0.0004594303337341143,
+      "loss": 3.1624,
+      "step": 19690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5643361806869507,
+      "learning_rate": 0.00045941729142017477,
+      "loss": 3.0144,
+      "step": 19691
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5318419933319092,
+      "learning_rate": 0.00045940424868636085,
+      "loss": 2.9979,
+      "step": 19692
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5172003507614136,
+      "learning_rate": 0.00045939120553270676,
+      "loss": 3.1654,
+      "step": 19693
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5160081386566162,
+      "learning_rate": 0.000459378161959247,
+      "loss": 3.1003,
+      "step": 19694
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5427367687225342,
+      "learning_rate": 0.0004593651179660158,
+      "loss": 3.2305,
+      "step": 19695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.523469865322113,
+      "learning_rate": 0.00045935207355304744,
+      "loss": 3.2241,
+      "step": 19696
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5474386215209961,
+      "learning_rate": 0.0004593390287203766,
+      "loss": 3.1163,
+      "step": 19697
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5134205222129822,
+      "learning_rate": 0.0004593259834680372,
+      "loss": 3.2631,
+      "step": 19698
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5353065729141235,
+      "learning_rate": 0.00045931293779606397,
+      "loss": 3.2246,
+      "step": 19699
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5378754138946533,
+      "learning_rate": 0.00045929989170449115,
+      "loss": 3.3742,
+      "step": 19700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078233480453491,
+      "learning_rate": 0.00045928684519335296,
+      "loss": 3.0678,
+      "step": 19701
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.534570038318634,
+      "learning_rate": 0.00045927379826268393,
+      "loss": 3.2471,
+      "step": 19702
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5478757619857788,
+      "learning_rate": 0.0004592607509125184,
+      "loss": 2.8799,
+      "step": 19703
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5131859183311462,
+      "learning_rate": 0.0004592477031428906,
+      "loss": 3.0923,
+      "step": 19704
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5837509036064148,
+      "learning_rate": 0.00045923465495383513,
+      "loss": 3.4114,
+      "step": 19705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5130158066749573,
+      "learning_rate": 0.0004592216063453861,
+      "loss": 3.1598,
+      "step": 19706
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5604761242866516,
+      "learning_rate": 0.000459208557317578,
+      "loss": 3.2225,
+      "step": 19707
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5324618220329285,
+      "learning_rate": 0.00045919550787044527,
+      "loss": 3.1212,
+      "step": 19708
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5190209150314331,
+      "learning_rate": 0.0004591824580040222,
+      "loss": 3.0561,
+      "step": 19709
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5445383191108704,
+      "learning_rate": 0.0004591694077183432,
+      "loss": 3.1508,
+      "step": 19710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5293034315109253,
+      "learning_rate": 0.0004591563570134427,
+      "loss": 3.187,
+      "step": 19711
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5551683902740479,
+      "learning_rate": 0.0004591433058893548,
+      "loss": 3.1458,
+      "step": 19712
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5138912796974182,
+      "learning_rate": 0.00045913025434611413,
+      "loss": 2.9319,
+      "step": 19713
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5497405529022217,
+      "learning_rate": 0.000459117202383755,
+      "loss": 3.1789,
+      "step": 19714
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.572364091873169,
+      "learning_rate": 0.00045910415000231173,
+      "loss": 3.2243,
+      "step": 19715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5154417753219604,
+      "learning_rate": 0.0004590910972018188,
+      "loss": 2.9992,
+      "step": 19716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5528554320335388,
+      "learning_rate": 0.00045907804398231055,
+      "loss": 3.0243,
+      "step": 19717
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5160757303237915,
+      "learning_rate": 0.0004590649903438213,
+      "loss": 3.1319,
+      "step": 19718
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4939347803592682,
+      "learning_rate": 0.0004590519362863855,
+      "loss": 2.9028,
+      "step": 19719
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5522775650024414,
+      "learning_rate": 0.0004590388818100375,
+      "loss": 3.0864,
+      "step": 19720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5493162870407104,
+      "learning_rate": 0.00045902582691481175,
+      "loss": 3.3321,
+      "step": 19721
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5284614562988281,
+      "learning_rate": 0.00045901277160074257,
+      "loss": 3.1432,
+      "step": 19722
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5101198554039001,
+      "learning_rate": 0.00045899971586786436,
+      "loss": 3.3082,
+      "step": 19723
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5157438516616821,
+      "learning_rate": 0.00045898665971621133,
+      "loss": 3.1203,
+      "step": 19724
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.538258969783783,
+      "learning_rate": 0.00045897360314581823,
+      "loss": 2.9963,
+      "step": 19725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5072723627090454,
+      "learning_rate": 0.0004589605461567192,
+      "loss": 3.2203,
+      "step": 19726
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5226778388023376,
+      "learning_rate": 0.0004589474887489486,
+      "loss": 3.2614,
+      "step": 19727
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5579911470413208,
+      "learning_rate": 0.000458934430922541,
+      "loss": 3.1676,
+      "step": 19728
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.514929473400116,
+      "learning_rate": 0.0004589213726775307,
+      "loss": 3.1178,
+      "step": 19729
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5149416923522949,
+      "learning_rate": 0.000458908314013952,
+      "loss": 3.1532,
+      "step": 19730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5170868039131165,
+      "learning_rate": 0.00045889525493183954,
+      "loss": 3.1691,
+      "step": 19731
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5064755082130432,
+      "learning_rate": 0.0004588821954312274,
+      "loss": 2.9093,
+      "step": 19732
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5192744135856628,
+      "learning_rate": 0.00045886913551215026,
+      "loss": 3.0864,
+      "step": 19733
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5007292032241821,
+      "learning_rate": 0.0004588560751746423,
+      "loss": 3.2911,
+      "step": 19734
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5587050914764404,
+      "learning_rate": 0.000458843014418738,
+      "loss": 3.0655,
+      "step": 19735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5086398720741272,
+      "learning_rate": 0.0004588299532444718,
+      "loss": 3.1727,
+      "step": 19736
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5102616548538208,
+      "learning_rate": 0.0004588168916518781,
+      "loss": 3.1426,
+      "step": 19737
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5597113966941833,
+      "learning_rate": 0.0004588038296409913,
+      "loss": 3.0143,
+      "step": 19738
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5260500311851501,
+      "learning_rate": 0.00045879076721184564,
+      "loss": 2.9833,
+      "step": 19739
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5252350568771362,
+      "learning_rate": 0.00045877770436447587,
+      "loss": 3.433,
+      "step": 19740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5167698264122009,
+      "learning_rate": 0.00045876464109891605,
+      "loss": 3.3007,
+      "step": 19741
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.512118935585022,
+      "learning_rate": 0.0004587515774152007,
+      "loss": 3.1823,
+      "step": 19742
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53267502784729,
+      "learning_rate": 0.00045873851331336424,
+      "loss": 3.4128,
+      "step": 19743
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5323781371116638,
+      "learning_rate": 0.00045872544879344115,
+      "loss": 3.2691,
+      "step": 19744
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5187110900878906,
+      "learning_rate": 0.0004587123838554657,
+      "loss": 2.9825,
+      "step": 19745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5188029408454895,
+      "learning_rate": 0.0004586993184994725,
+      "loss": 3.2929,
+      "step": 19746
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5603750348091125,
+      "learning_rate": 0.00045868625272549573,
+      "loss": 3.1024,
+      "step": 19747
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5371494889259338,
+      "learning_rate": 0.00045867318653356994,
+      "loss": 3.2192,
+      "step": 19748
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5349953770637512,
+      "learning_rate": 0.00045866011992372953,
+      "loss": 3.1364,
+      "step": 19749
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5063318610191345,
+      "learning_rate": 0.0004586470528960088,
+      "loss": 3.1449,
+      "step": 19750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5154860019683838,
+      "learning_rate": 0.00045863398545044245,
+      "loss": 3.0492,
+      "step": 19751
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4861948490142822,
+      "learning_rate": 0.0004586209175870646,
+      "loss": 3.3389,
+      "step": 19752
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5171583294868469,
+      "learning_rate": 0.0004586078493059098,
+      "loss": 3.1478,
+      "step": 19753
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5161516666412354,
+      "learning_rate": 0.0004585947806070124,
+      "loss": 2.9258,
+      "step": 19754
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5537519454956055,
+      "learning_rate": 0.00045858171149040696,
+      "loss": 3.1637,
+      "step": 19755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5510892271995544,
+      "learning_rate": 0.0004585686419561277,
+      "loss": 3.1517,
+      "step": 19756
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5581437349319458,
+      "learning_rate": 0.0004585555720042093,
+      "loss": 3.4651,
+      "step": 19757
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5405652523040771,
+      "learning_rate": 0.00045854250163468587,
+      "loss": 3.2142,
+      "step": 19758
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.536334753036499,
+      "learning_rate": 0.00045852943084759214,
+      "loss": 3.2114,
+      "step": 19759
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5953757166862488,
+      "learning_rate": 0.00045851635964296234,
+      "loss": 3.0759,
+      "step": 19760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4927489161491394,
+      "learning_rate": 0.00045850328802083095,
+      "loss": 3.3361,
+      "step": 19761
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5365741848945618,
+      "learning_rate": 0.00045849021598123245,
+      "loss": 3.0068,
+      "step": 19762
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5352941751480103,
+      "learning_rate": 0.0004584771435242012,
+      "loss": 3.2842,
+      "step": 19763
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5236901640892029,
+      "learning_rate": 0.00045846407064977163,
+      "loss": 3.0848,
+      "step": 19764
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5306557416915894,
+      "learning_rate": 0.0004584509973579782,
+      "loss": 3.1199,
+      "step": 19765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5267027616500854,
+      "learning_rate": 0.00045843792364885533,
+      "loss": 3.346,
+      "step": 19766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5062419176101685,
+      "learning_rate": 0.0004584248495224375,
+      "loss": 3.1383,
+      "step": 19767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5400051474571228,
+      "learning_rate": 0.00045841177497875905,
+      "loss": 3.0832,
+      "step": 19768
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5341521501541138,
+      "learning_rate": 0.00045839870001785455,
+      "loss": 3.1805,
+      "step": 19769
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5922198295593262,
+      "learning_rate": 0.00045838562463975833,
+      "loss": 3.1083,
+      "step": 19770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49147966504096985,
+      "learning_rate": 0.0004583725488445048,
+      "loss": 3.2987,
+      "step": 19771
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49468183517456055,
+      "learning_rate": 0.00045835947263212846,
+      "loss": 3.3835,
+      "step": 19772
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5233425498008728,
+      "learning_rate": 0.00045834639600266387,
+      "loss": 3.1561,
+      "step": 19773
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5618370175361633,
+      "learning_rate": 0.0004583333189561453,
+      "loss": 3.0311,
+      "step": 19774
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.526765763759613,
+      "learning_rate": 0.00045832024149260707,
+      "loss": 3.2129,
+      "step": 19775
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5225813388824463,
+      "learning_rate": 0.0004583071636120839,
+      "loss": 3.273,
+      "step": 19776
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.514604389667511,
+      "learning_rate": 0.00045829408531461023,
+      "loss": 3.0254,
+      "step": 19777
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5545125603675842,
+      "learning_rate": 0.00045828100660022037,
+      "loss": 3.1412,
+      "step": 19778
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.544165849685669,
+      "learning_rate": 0.00045826792746894875,
+      "loss": 3.1968,
+      "step": 19779
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5067391395568848,
+      "learning_rate": 0.0004582548479208298,
+      "loss": 3.0627,
+      "step": 19780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4904179275035858,
+      "learning_rate": 0.0004582417679558981,
+      "loss": 3.0005,
+      "step": 19781
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5631465315818787,
+      "learning_rate": 0.0004582286875741881,
+      "loss": 3.186,
+      "step": 19782
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5135632753372192,
+      "learning_rate": 0.00045821560677573414,
+      "loss": 3.1411,
+      "step": 19783
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5340128540992737,
+      "learning_rate": 0.0004582025255605708,
+      "loss": 2.9759,
+      "step": 19784
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625054836273193,
+      "learning_rate": 0.0004581894439287322,
+      "loss": 3.1578,
+      "step": 19785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5040884017944336,
+      "learning_rate": 0.00045817636188025333,
+      "loss": 3.1584,
+      "step": 19786
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5131189823150635,
+      "learning_rate": 0.00045816327941516823,
+      "loss": 3.3082,
+      "step": 19787
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5243518948554993,
+      "learning_rate": 0.0004581501965335115,
+      "loss": 3.0006,
+      "step": 19788
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5257127285003662,
+      "learning_rate": 0.0004581371132353176,
+      "loss": 3.2521,
+      "step": 19789
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5097827911376953,
+      "learning_rate": 0.00045812402952062105,
+      "loss": 3.0542,
+      "step": 19790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.572334349155426,
+      "learning_rate": 0.00045811094538945614,
+      "loss": 2.9641,
+      "step": 19791
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.529404878616333,
+      "learning_rate": 0.00045809786084185746,
+      "loss": 3.0158,
+      "step": 19792
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5602056980133057,
+      "learning_rate": 0.00045808477587785945,
+      "loss": 3.2361,
+      "step": 19793
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5243390202522278,
+      "learning_rate": 0.00045807169049749653,
+      "loss": 3.2635,
+      "step": 19794
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5974134802818298,
+      "learning_rate": 0.0004580586047008033,
+      "loss": 3.0509,
+      "step": 19795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4903809428215027,
+      "learning_rate": 0.0004580455184878139,
+      "loss": 3.148,
+      "step": 19796
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5298634767532349,
+      "learning_rate": 0.00045803243185856327,
+      "loss": 3.2435,
+      "step": 19797
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5195876955986023,
+      "learning_rate": 0.00045801934481308547,
+      "loss": 3.1942,
+      "step": 19798
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.581200361251831,
+      "learning_rate": 0.0004580062573514153,
+      "loss": 3.017,
+      "step": 19799
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5249642729759216,
+      "learning_rate": 0.0004579931694735869,
+      "loss": 3.0202,
+      "step": 19800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5444178581237793,
+      "learning_rate": 0.0004579800811796349,
+      "loss": 3.1979,
+      "step": 19801
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5659769177436829,
+      "learning_rate": 0.00045796699246959384,
+      "loss": 3.2618,
+      "step": 19802
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5352531671524048,
+      "learning_rate": 0.00045795390334349813,
+      "loss": 3.0874,
+      "step": 19803
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5184085965156555,
+      "learning_rate": 0.0004579408138013822,
+      "loss": 3.1453,
+      "step": 19804
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5398754477500916,
+      "learning_rate": 0.0004579277238432805,
+      "loss": 3.0704,
+      "step": 19805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5079792737960815,
+      "learning_rate": 0.00045791463346922756,
+      "loss": 3.1852,
+      "step": 19806
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.521885871887207,
+      "learning_rate": 0.00045790154267925795,
+      "loss": 3.2204,
+      "step": 19807
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5815854072570801,
+      "learning_rate": 0.00045788845147340607,
+      "loss": 2.9648,
+      "step": 19808
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5417388081550598,
+      "learning_rate": 0.00045787535985170643,
+      "loss": 3.3381,
+      "step": 19809
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5544579029083252,
+      "learning_rate": 0.00045786226781419347,
+      "loss": 3.0114,
+      "step": 19810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5227230191230774,
+      "learning_rate": 0.00045784917536090155,
+      "loss": 3.1927,
+      "step": 19811
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5432718992233276,
+      "learning_rate": 0.0004578360824918653,
+      "loss": 2.9235,
+      "step": 19812
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5189009308815002,
+      "learning_rate": 0.0004578229892071193,
+      "loss": 3.1517,
+      "step": 19813
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5389320850372314,
+      "learning_rate": 0.000457809895506698,
+      "loss": 3.2118,
+      "step": 19814
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5334956049919128,
+      "learning_rate": 0.0004577968013906356,
+      "loss": 2.9926,
+      "step": 19815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5629310011863708,
+      "learning_rate": 0.0004577837068589669,
+      "loss": 3.139,
+      "step": 19816
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5584556460380554,
+      "learning_rate": 0.0004577706119117262,
+      "loss": 3.049,
+      "step": 19817
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4945135712623596,
+      "learning_rate": 0.0004577575165489482,
+      "loss": 3.1779,
+      "step": 19818
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5308895111083984,
+      "learning_rate": 0.0004577444207706671,
+      "loss": 3.3229,
+      "step": 19819
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5382962226867676,
+      "learning_rate": 0.00045773132457691766,
+      "loss": 3.2655,
+      "step": 19820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5207489728927612,
+      "learning_rate": 0.0004577182279677342,
+      "loss": 3.2315,
+      "step": 19821
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5437517166137695,
+      "learning_rate": 0.0004577051309431513,
+      "loss": 3.1617,
+      "step": 19822
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5084508657455444,
+      "learning_rate": 0.0004576920335032035,
+      "loss": 3.1338,
+      "step": 19823
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48307299613952637,
+      "learning_rate": 0.0004576789356479253,
+      "loss": 3.0642,
+      "step": 19824
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.520545244216919,
+      "learning_rate": 0.000457665837377351,
+      "loss": 2.956,
+      "step": 19825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5704953074455261,
+      "learning_rate": 0.00045765273869151514,
+      "loss": 3.1909,
+      "step": 19826
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5432178974151611,
+      "learning_rate": 0.00045763963959045246,
+      "loss": 3.1516,
+      "step": 19827
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5819136500358582,
+      "learning_rate": 0.0004576265400741973,
+      "loss": 3.1398,
+      "step": 19828
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51753169298172,
+      "learning_rate": 0.00045761344014278414,
+      "loss": 2.8872,
+      "step": 19829
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5441774725914001,
+      "learning_rate": 0.0004576003397962475,
+      "loss": 3.1017,
+      "step": 19830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5453305840492249,
+      "learning_rate": 0.0004575872390346219,
+      "loss": 3.1247,
+      "step": 19831
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5290966033935547,
+      "learning_rate": 0.0004575741378579419,
+      "loss": 3.0589,
+      "step": 19832
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5535355806350708,
+      "learning_rate": 0.0004575610362662418,
+      "loss": 3.0727,
+      "step": 19833
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5289100408554077,
+      "learning_rate": 0.00045754793425955637,
+      "loss": 3.0748,
+      "step": 19834
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.681056559085846,
+      "learning_rate": 0.00045753483183792005,
+      "loss": 3.1107,
+      "step": 19835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.514905571937561,
+      "learning_rate": 0.0004575217290013672,
+      "loss": 3.1522,
+      "step": 19836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5799508094787598,
+      "learning_rate": 0.00045750862574993236,
+      "loss": 3.1752,
+      "step": 19837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5410858988761902,
+      "learning_rate": 0.0004574955220836503,
+      "loss": 3.1188,
+      "step": 19838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.541213870048523,
+      "learning_rate": 0.0004574824180025553,
+      "loss": 3.1487,
+      "step": 19839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5377730131149292,
+      "learning_rate": 0.00045746931350668183,
+      "loss": 3.2846,
+      "step": 19840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5022572875022888,
+      "learning_rate": 0.0004574562085960646,
+      "loss": 3.1793,
+      "step": 19841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.535351574420929,
+      "learning_rate": 0.0004574431032707379,
+      "loss": 2.9624,
+      "step": 19842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5488033890724182,
+      "learning_rate": 0.0004574299975307364,
+      "loss": 3.1208,
+      "step": 19843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5779278874397278,
+      "learning_rate": 0.00045741689137609467,
+      "loss": 3.1375,
+      "step": 19844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5179200768470764,
+      "learning_rate": 0.00045740378480684706,
+      "loss": 2.9844,
+      "step": 19845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5455824732780457,
+      "learning_rate": 0.00045739067782302824,
+      "loss": 2.9835,
+      "step": 19846
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5416254997253418,
+      "learning_rate": 0.00045737757042467263,
+      "loss": 3.4106,
+      "step": 19847
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5338101387023926,
+      "learning_rate": 0.0004573644626118147,
+      "loss": 3.2009,
+      "step": 19848
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5255735516548157,
+      "learning_rate": 0.0004573513543844892,
+      "loss": 3.203,
+      "step": 19849
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5638295412063599,
+      "learning_rate": 0.0004573382457427305,
+      "loss": 3.2313,
+      "step": 19850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5044254660606384,
+      "learning_rate": 0.0004573251366865731,
+      "loss": 3.1181,
+      "step": 19851
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.675537109375,
+      "learning_rate": 0.00045731202721605144,
+      "loss": 3.3099,
+      "step": 19852
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.517725944519043,
+      "learning_rate": 0.0004572989173312004,
+      "loss": 3.1187,
+      "step": 19853
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5297645330429077,
+      "learning_rate": 0.0004572858070320542,
+      "loss": 3.2745,
+      "step": 19854
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5531717538833618,
+      "learning_rate": 0.00045727269631864743,
+      "loss": 3.0869,
+      "step": 19855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5511997938156128,
+      "learning_rate": 0.0004572595851910147,
+      "loss": 3.2544,
+      "step": 19856
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5770429968833923,
+      "learning_rate": 0.0004572464736491904,
+      "loss": 3.058,
+      "step": 19857
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5271993279457092,
+      "learning_rate": 0.00045723336169320916,
+      "loss": 3.221,
+      "step": 19858
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5206511616706848,
+      "learning_rate": 0.0004572202493231056,
+      "loss": 3.3138,
+      "step": 19859
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.537216067314148,
+      "learning_rate": 0.0004572071365389141,
+      "loss": 3.0961,
+      "step": 19860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5249066948890686,
+      "learning_rate": 0.00045719402334066926,
+      "loss": 3.1989,
+      "step": 19861
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49584659934043884,
+      "learning_rate": 0.0004571809097284055,
+      "loss": 3.3242,
+      "step": 19862
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4906226694583893,
+      "learning_rate": 0.00045716779570215767,
+      "loss": 3.2433,
+      "step": 19863
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5297616720199585,
+      "learning_rate": 0.00045715468126196,
+      "loss": 3.0477,
+      "step": 19864
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5272489786148071,
+      "learning_rate": 0.00045714156640784727,
+      "loss": 3.1792,
+      "step": 19865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5244301557540894,
+      "learning_rate": 0.0004571284511398538,
+      "loss": 3.3991,
+      "step": 19866
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5065206289291382,
+      "learning_rate": 0.0004571153354580142,
+      "loss": 3.0658,
+      "step": 19867
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5413605570793152,
+      "learning_rate": 0.00045710221936236305,
+      "loss": 3.2466,
+      "step": 19868
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5623077154159546,
+      "learning_rate": 0.00045708910285293487,
+      "loss": 3.2348,
+      "step": 19869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5285463333129883,
+      "learning_rate": 0.0004570759859297643,
+      "loss": 2.9883,
+      "step": 19870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.530189573764801,
+      "learning_rate": 0.0004570628685928858,
+      "loss": 3.1729,
+      "step": 19871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5128258466720581,
+      "learning_rate": 0.00045704975084233395,
+      "loss": 3.0372,
+      "step": 19872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5443863868713379,
+      "learning_rate": 0.0004570366326781432,
+      "loss": 3.1384,
+      "step": 19873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.547491729259491,
+      "learning_rate": 0.0004570235141003482,
+      "loss": 3.2092,
+      "step": 19874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5074962377548218,
+      "learning_rate": 0.0004570103951089836,
+      "loss": 3.0942,
+      "step": 19875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.574950635433197,
+      "learning_rate": 0.00045699727570408375,
+      "loss": 3.1655,
+      "step": 19876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5343877077102661,
+      "learning_rate": 0.00045698415588568334,
+      "loss": 3.1291,
+      "step": 19877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5411051511764526,
+      "learning_rate": 0.0004569710356538167,
+      "loss": 3.2585,
+      "step": 19878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5204007625579834,
+      "learning_rate": 0.0004569579150085188,
+      "loss": 3.2438,
+      "step": 19879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.518437385559082,
+      "learning_rate": 0.0004569447939498238,
+      "loss": 3.0511,
+      "step": 19880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5438820123672485,
+      "learning_rate": 0.0004569316724777665,
+      "loss": 3.1059,
+      "step": 19881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5277252793312073,
+      "learning_rate": 0.00045691855059238126,
+      "loss": 3.1655,
+      "step": 19882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4945885241031647,
+      "learning_rate": 0.0004569054282937029,
+      "loss": 3.0627,
+      "step": 19883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5790326595306396,
+      "learning_rate": 0.0004568923055817657,
+      "loss": 3.1567,
+      "step": 19884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5103229880332947,
+      "learning_rate": 0.0004568791824566045,
+      "loss": 3.2317,
+      "step": 19885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5120726823806763,
+      "learning_rate": 0.00045686605891825363,
+      "loss": 3.2693,
+      "step": 19886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078148245811462,
+      "learning_rate": 0.00045685293496674784,
+      "loss": 3.2933,
+      "step": 19887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5411629676818848,
+      "learning_rate": 0.0004568398106021215,
+      "loss": 2.9652,
+      "step": 19888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5563063621520996,
+      "learning_rate": 0.00045682668582440933,
+      "loss": 3.2037,
+      "step": 19889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5265916585922241,
+      "learning_rate": 0.0004568135606336459,
+      "loss": 3.0914,
+      "step": 19890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5466210842132568,
+      "learning_rate": 0.0004568004350298657,
+      "loss": 3.1264,
+      "step": 19891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5217623710632324,
+      "learning_rate": 0.0004567873090131033,
+      "loss": 3.1154,
+      "step": 19892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.514886736869812,
+      "learning_rate": 0.0004567741825833933,
+      "loss": 3.2016,
+      "step": 19893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6545686721801758,
+      "learning_rate": 0.00045676105574077026,
+      "loss": 3.2672,
+      "step": 19894
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.527996838092804,
+      "learning_rate": 0.0004567479284852689,
+      "loss": 3.0232,
+      "step": 19895
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5357192754745483,
+      "learning_rate": 0.0004567348008169235,
+      "loss": 3.1696,
+      "step": 19896
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5141521692276001,
+      "learning_rate": 0.00045672167273576894,
+      "loss": 3.205,
+      "step": 19897
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5258877277374268,
+      "learning_rate": 0.00045670854424183953,
+      "loss": 3.2861,
+      "step": 19898
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.574230432510376,
+      "learning_rate": 0.00045669541533517,
+      "loss": 3.245,
+      "step": 19899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5562745928764343,
+      "learning_rate": 0.00045668228601579495,
+      "loss": 2.9874,
+      "step": 19900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5093985199928284,
+      "learning_rate": 0.0004566691562837489,
+      "loss": 3.2173,
+      "step": 19901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5798713564872742,
+      "learning_rate": 0.0004566560261390664,
+      "loss": 3.2883,
+      "step": 19902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5926944017410278,
+      "learning_rate": 0.0004566428955817821,
+      "loss": 3.1588,
+      "step": 19903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5162215828895569,
+      "learning_rate": 0.0004566297646119306,
+      "loss": 3.1708,
+      "step": 19904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5461186766624451,
+      "learning_rate": 0.0004566166332295464,
+      "loss": 2.9684,
+      "step": 19905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5802727341651917,
+      "learning_rate": 0.0004566035014346641,
+      "loss": 3.1052,
+      "step": 19906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5327712297439575,
+      "learning_rate": 0.0004565903692273184,
+      "loss": 3.114,
+      "step": 19907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5120739936828613,
+      "learning_rate": 0.00045657723660754384,
+      "loss": 3.0222,
+      "step": 19908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5327271819114685,
+      "learning_rate": 0.0004565641035753748,
+      "loss": 3.0759,
+      "step": 19909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.524299144744873,
+      "learning_rate": 0.00045655097013084616,
+      "loss": 3.2712,
+      "step": 19910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5683066248893738,
+      "learning_rate": 0.00045653783627399235,
+      "loss": 3.2524,
+      "step": 19911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5321347117424011,
+      "learning_rate": 0.00045652470200484806,
+      "loss": 3.205,
+      "step": 19912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390219688415527,
+      "learning_rate": 0.0004565115673234478,
+      "loss": 3.1628,
+      "step": 19913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5145443677902222,
+      "learning_rate": 0.0004564984322298261,
+      "loss": 3.2744,
+      "step": 19914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5551137924194336,
+      "learning_rate": 0.0004564852967240178,
+      "loss": 3.1102,
+      "step": 19915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.524955153465271,
+      "learning_rate": 0.00045647216080605725,
+      "loss": 3.3276,
+      "step": 19916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5320925712585449,
+      "learning_rate": 0.00045645902447597923,
+      "loss": 3.3742,
+      "step": 19917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.518169105052948,
+      "learning_rate": 0.00045644588773381813,
+      "loss": 3.1312,
+      "step": 19918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5469488501548767,
+      "learning_rate": 0.00045643275057960877,
+      "loss": 3.137,
+      "step": 19919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5299994945526123,
+      "learning_rate": 0.0004564196130133856,
+      "loss": 3.2838,
+      "step": 19920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5019270181655884,
+      "learning_rate": 0.0004564064750351833,
+      "loss": 3.0652,
+      "step": 19921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5183849930763245,
+      "learning_rate": 0.00045639333664503637,
+      "loss": 2.9966,
+      "step": 19922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5174096822738647,
+      "learning_rate": 0.0004563801978429797,
+      "loss": 2.9913,
+      "step": 19923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5443447828292847,
+      "learning_rate": 0.0004563670586290475,
+      "loss": 3.057,
+      "step": 19924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.564255952835083,
+      "learning_rate": 0.0004563539190032746,
+      "loss": 3.0481,
+      "step": 19925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5157086253166199,
+      "learning_rate": 0.0004563407789656956,
+      "loss": 3.0848,
+      "step": 19926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5265598893165588,
+      "learning_rate": 0.00045632763851634496,
+      "loss": 3.1655,
+      "step": 19927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5157798528671265,
+      "learning_rate": 0.00045631449765525753,
+      "loss": 3.2964,
+      "step": 19928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5361979603767395,
+      "learning_rate": 0.00045630135638246775,
+      "loss": 3.1274,
+      "step": 19929
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5458618998527527,
+      "learning_rate": 0.0004562882146980103,
+      "loss": 3.1837,
+      "step": 19930
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49934300780296326,
+      "learning_rate": 0.0004562750726019197,
+      "loss": 3.0736,
+      "step": 19931
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5094740390777588,
+      "learning_rate": 0.00045626193009423076,
+      "loss": 2.8927,
+      "step": 19932
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5375152826309204,
+      "learning_rate": 0.00045624878717497784,
+      "loss": 3.2418,
+      "step": 19933
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5470656156539917,
+      "learning_rate": 0.0004562356438441957,
+      "loss": 3.2453,
+      "step": 19934
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5325232148170471,
+      "learning_rate": 0.000456222500101919,
+      "loss": 3.1305,
+      "step": 19935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5318748950958252,
+      "learning_rate": 0.00045620935594818234,
+      "loss": 3.0646,
+      "step": 19936
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5223355889320374,
+      "learning_rate": 0.0004561962113830203,
+      "loss": 3.2699,
+      "step": 19937
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5093883872032166,
+      "learning_rate": 0.00045618306640646744,
+      "loss": 3.2268,
+      "step": 19938
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5158838033676147,
+      "learning_rate": 0.0004561699210185584,
+      "loss": 3.0998,
+      "step": 19939
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5459488034248352,
+      "learning_rate": 0.00045615677521932785,
+      "loss": 3.142,
+      "step": 19940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.574743390083313,
+      "learning_rate": 0.00045614362900881053,
+      "loss": 3.1598,
+      "step": 19941
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5598559379577637,
+      "learning_rate": 0.00045613048238704086,
+      "loss": 3.0253,
+      "step": 19942
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.498945027589798,
+      "learning_rate": 0.0004561173353540535,
+      "loss": 3.3936,
+      "step": 19943
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5641855597496033,
+      "learning_rate": 0.0004561041879098832,
+      "loss": 3.4009,
+      "step": 19944
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5260211825370789,
+      "learning_rate": 0.0004560910400545645,
+      "loss": 3.2477,
+      "step": 19945
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5271660089492798,
+      "learning_rate": 0.000456077891788132,
+      "loss": 3.2298,
+      "step": 19946
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.499926894903183,
+      "learning_rate": 0.00045606474311062035,
+      "loss": 3.2045,
+      "step": 19947
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5305925011634827,
+      "learning_rate": 0.00045605159402206435,
+      "loss": 3.1442,
+      "step": 19948
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5455648303031921,
+      "learning_rate": 0.0004560384445224984,
+      "loss": 3.0783,
+      "step": 19949
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5389901995658875,
+      "learning_rate": 0.0004560252946119571,
+      "loss": 3.148,
+      "step": 19950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5463103652000427,
+      "learning_rate": 0.00045601214429047534,
+      "loss": 2.9844,
+      "step": 19951
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5185086131095886,
+      "learning_rate": 0.00045599899355808763,
+      "loss": 3.1557,
+      "step": 19952
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5186192393302917,
+      "learning_rate": 0.0004559858424148285,
+      "loss": 3.2037,
+      "step": 19953
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5496533513069153,
+      "learning_rate": 0.00045597269086073273,
+      "loss": 3.1717,
+      "step": 19954
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5932478308677673,
+      "learning_rate": 0.00045595953889583483,
+      "loss": 2.9012,
+      "step": 19955
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5475051403045654,
+      "learning_rate": 0.0004559463865201696,
+      "loss": 3.2653,
+      "step": 19956
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5264762043952942,
+      "learning_rate": 0.0004559332337337716,
+      "loss": 3.2941,
+      "step": 19957
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.503323495388031,
+      "learning_rate": 0.0004559200805366755,
+      "loss": 3.1422,
+      "step": 19958
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5633005499839783,
+      "learning_rate": 0.0004559069269289159,
+      "loss": 3.2524,
+      "step": 19959
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5244492888450623,
+      "learning_rate": 0.0004558937729105274,
+      "loss": 3.1133,
+      "step": 19960
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5386800765991211,
+      "learning_rate": 0.00045588061848154477,
+      "loss": 3.2379,
+      "step": 19961
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5170870423316956,
+      "learning_rate": 0.00045586746364200255,
+      "loss": 3.2358,
+      "step": 19962
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.561007559299469,
+      "learning_rate": 0.00045585430839193547,
+      "loss": 2.9753,
+      "step": 19963
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4810415208339691,
+      "learning_rate": 0.00045584115273137807,
+      "loss": 3.1482,
+      "step": 19964
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6551200747489929,
+      "learning_rate": 0.0004558279966603651,
+      "loss": 2.9541,
+      "step": 19965
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5145920515060425,
+      "learning_rate": 0.00045581484017893117,
+      "loss": 3.125,
+      "step": 19966
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4966809153556824,
+      "learning_rate": 0.000455801683287111,
+      "loss": 3.098,
+      "step": 19967
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5068411827087402,
+      "learning_rate": 0.00045578852598493914,
+      "loss": 2.9166,
+      "step": 19968
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5475143790245056,
+      "learning_rate": 0.00045577536827245025,
+      "loss": 2.9523,
+      "step": 19969
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5503715872764587,
+      "learning_rate": 0.000455762210149679,
+      "loss": 3.1146,
+      "step": 19970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5356963276863098,
+      "learning_rate": 0.00045574905161666014,
+      "loss": 2.8835,
+      "step": 19971
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5304535627365112,
+      "learning_rate": 0.0004557358926734282,
+      "loss": 3.2342,
+      "step": 19972
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.511310875415802,
+      "learning_rate": 0.0004557227333200179,
+      "loss": 3.2055,
+      "step": 19973
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4973483085632324,
+      "learning_rate": 0.000455709573556464,
+      "loss": 3.2209,
+      "step": 19974
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5432186126708984,
+      "learning_rate": 0.00045569641338280087,
+      "loss": 2.9227,
+      "step": 19975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5128586292266846,
+      "learning_rate": 0.00045568325279906344,
+      "loss": 3.4699,
+      "step": 19976
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5524463057518005,
+      "learning_rate": 0.0004556700918052864,
+      "loss": 3.1689,
+      "step": 19977
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5236881375312805,
+      "learning_rate": 0.0004556569304015041,
+      "loss": 3.2393,
+      "step": 19978
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5629315972328186,
+      "learning_rate": 0.0004556437685877515,
+      "loss": 3.2885,
+      "step": 19979
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5505418181419373,
+      "learning_rate": 0.0004556306063640632,
+      "loss": 3.1929,
+      "step": 19980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5632915496826172,
+      "learning_rate": 0.0004556174437304737,
+      "loss": 3.0301,
+      "step": 19981
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5991239547729492,
+      "learning_rate": 0.00045560428068701787,
+      "loss": 3.2244,
+      "step": 19982
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5407932996749878,
+      "learning_rate": 0.00045559111723373036,
+      "loss": 3.239,
+      "step": 19983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5531562566757202,
+      "learning_rate": 0.00045557795337064575,
+      "loss": 3.3026,
+      "step": 19984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5307109951972961,
+      "learning_rate": 0.00045556478909779876,
+      "loss": 3.2586,
+      "step": 19985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.515917181968689,
+      "learning_rate": 0.000455551624415224,
+      "loss": 3.0974,
+      "step": 19986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5192380547523499,
+      "learning_rate": 0.0004555384593229562,
+      "loss": 3.2065,
+      "step": 19987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.523467481136322,
+      "learning_rate": 0.00045552529382103025,
+      "loss": 3.0622,
+      "step": 19988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5278224945068359,
+      "learning_rate": 0.00045551212790948036,
+      "loss": 3.1597,
+      "step": 19989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5236921310424805,
+      "learning_rate": 0.00045549896158834154,
+      "loss": 3.0244,
+      "step": 19990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5542269945144653,
+      "learning_rate": 0.0004554857948576483,
+      "loss": 3.2861,
+      "step": 19991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5261695981025696,
+      "learning_rate": 0.00045547262771743555,
+      "loss": 3.0632,
+      "step": 19992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.530767023563385,
+      "learning_rate": 0.00045545946016773775,
+      "loss": 3.2265,
+      "step": 19993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5161399841308594,
+      "learning_rate": 0.00045544629220858966,
+      "loss": 3.2562,
+      "step": 19994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4969814717769623,
+      "learning_rate": 0.00045543312384002595,
+      "loss": 3.0832,
+      "step": 19995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5116978287696838,
+      "learning_rate": 0.0004554199550620812,
+      "loss": 3.15,
+      "step": 19996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5162671208381653,
+      "learning_rate": 0.00045540678587479037,
+      "loss": 3.2172,
+      "step": 19997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5306782722473145,
+      "learning_rate": 0.0004553936162781879,
+      "loss": 3.1452,
+      "step": 19998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.535821259021759,
+      "learning_rate": 0.0004553804462723086,
+      "loss": 3.2178,
+      "step": 19999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5315077900886536,
+      "learning_rate": 0.00045536727585718706,
+      "loss": 3.2123,
+      "step": 20000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.522333025932312,
+      "learning_rate": 0.0004553541050328579,
+      "loss": 3.0348,
+      "step": 20001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5301592350006104,
+      "learning_rate": 0.000455340933799356,
+      "loss": 3.1113,
+      "step": 20002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.584934651851654,
+      "learning_rate": 0.0004553277621567161,
+      "loss": 3.18,
+      "step": 20003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6135393977165222,
+      "learning_rate": 0.0004553145901049727,
+      "loss": 3.1347,
+      "step": 20004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5235838890075684,
+      "learning_rate": 0.0004553014176441605,
+      "loss": 2.9696,
+      "step": 20005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5571325421333313,
+      "learning_rate": 0.0004552882447743143,
+      "loss": 3.215,
+      "step": 20006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5705804824829102,
+      "learning_rate": 0.0004552750714954688,
+      "loss": 3.2719,
+      "step": 20007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5331918597221375,
+      "learning_rate": 0.00045526189780765856,
+      "loss": 3.2195,
+      "step": 20008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5457714200019836,
+      "learning_rate": 0.00045524872371091834,
+      "loss": 3.2391,
+      "step": 20009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5504556894302368,
+      "learning_rate": 0.000455235549205283,
+      "loss": 3.2419,
+      "step": 20010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49661940336227417,
+      "learning_rate": 0.000455222374290787,
+      "loss": 3.1479,
+      "step": 20011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5197951197624207,
+      "learning_rate": 0.00045520919896746516,
+      "loss": 3.0056,
+      "step": 20012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5557920932769775,
+      "learning_rate": 0.00045519602323535206,
+      "loss": 3.2958,
+      "step": 20013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53912353515625,
+      "learning_rate": 0.0004551828470944827,
+      "loss": 2.8655,
+      "step": 20014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5061671137809753,
+      "learning_rate": 0.0004551696705448915,
+      "loss": 3.2487,
+      "step": 20015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5309056639671326,
+      "learning_rate": 0.00045515649358661317,
+      "loss": 3.1987,
+      "step": 20016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5490389466285706,
+      "learning_rate": 0.0004551433162196826,
+      "loss": 3.084,
+      "step": 20017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5171396136283875,
+      "learning_rate": 0.00045513013844413435,
+      "loss": 3.2162,
+      "step": 20018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5325402617454529,
+      "learning_rate": 0.00045511696026000317,
+      "loss": 3.0539,
+      "step": 20019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5274950861930847,
+      "learning_rate": 0.00045510378166732375,
+      "loss": 3.3672,
+      "step": 20020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5318315625190735,
+      "learning_rate": 0.0004550906026661309,
+      "loss": 3.0185,
+      "step": 20021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5088624358177185,
+      "learning_rate": 0.00045507742325645914,
+      "loss": 3.1083,
+      "step": 20022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5656629800796509,
+      "learning_rate": 0.0004550642434383433,
+      "loss": 2.9348,
+      "step": 20023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5793680548667908,
+      "learning_rate": 0.0004550510632118182,
+      "loss": 3.0635,
+      "step": 20024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5440924167633057,
+      "learning_rate": 0.0004550378825769184,
+      "loss": 2.9255,
+      "step": 20025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49890273809432983,
+      "learning_rate": 0.0004550247015336786,
+      "loss": 3.2322,
+      "step": 20026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.541521430015564,
+      "learning_rate": 0.0004550115200821335,
+      "loss": 2.9845,
+      "step": 20027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5471722483634949,
+      "learning_rate": 0.00045499833822231807,
+      "loss": 3.002,
+      "step": 20028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5366112589836121,
+      "learning_rate": 0.0004549851559542668,
+      "loss": 3.1757,
+      "step": 20029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5042491555213928,
+      "learning_rate": 0.0004549719732780143,
+      "loss": 3.3144,
+      "step": 20030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5647759437561035,
+      "learning_rate": 0.0004549587901935957,
+      "loss": 2.917,
+      "step": 20031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5422675013542175,
+      "learning_rate": 0.00045494560670104525,
+      "loss": 3.0393,
+      "step": 20032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.532295823097229,
+      "learning_rate": 0.000454932422800398,
+      "loss": 3.0517,
+      "step": 20033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5531377792358398,
+      "learning_rate": 0.00045491923849168856,
+      "loss": 3.1039,
+      "step": 20034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4993976056575775,
+      "learning_rate": 0.0004549060537749516,
+      "loss": 2.9075,
+      "step": 20035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5766929984092712,
+      "learning_rate": 0.00045489286865022205,
+      "loss": 3.448,
+      "step": 20036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.525661289691925,
+      "learning_rate": 0.00045487968311753427,
+      "loss": 3.204,
+      "step": 20037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5298983454704285,
+      "learning_rate": 0.0004548664971769234,
+      "loss": 3.1754,
+      "step": 20038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5623906254768372,
+      "learning_rate": 0.000454853310828424,
+      "loss": 3.2299,
+      "step": 20039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5601446032524109,
+      "learning_rate": 0.0004548401240720706,
+      "loss": 3.135,
+      "step": 20040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5393292307853699,
+      "learning_rate": 0.0004548269369078982,
+      "loss": 3.1141,
+      "step": 20041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5087615847587585,
+      "learning_rate": 0.0004548137493359414,
+      "loss": 2.9242,
+      "step": 20042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5590735673904419,
+      "learning_rate": 0.00045480056135623515,
+      "loss": 3.1836,
+      "step": 20043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5152249932289124,
+      "learning_rate": 0.00045478737296881383,
+      "loss": 3.0764,
+      "step": 20044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.551261842250824,
+      "learning_rate": 0.0004547741841737125,
+      "loss": 3.1649,
+      "step": 20045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5181664824485779,
+      "learning_rate": 0.0004547609949709656,
+      "loss": 3.231,
+      "step": 20046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5038676857948303,
+      "learning_rate": 0.00045474780536060815,
+      "loss": 3.141,
+      "step": 20047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5354689359664917,
+      "learning_rate": 0.00045473461534267475,
+      "loss": 3.2463,
+      "step": 20048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5693992972373962,
+      "learning_rate": 0.00045472142491720006,
+      "loss": 3.0521,
+      "step": 20049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.532825767993927,
+      "learning_rate": 0.00045470823408421903,
+      "loss": 3.1193,
+      "step": 20050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5376405715942383,
+      "learning_rate": 0.0004546950428437662,
+      "loss": 3.2254,
+      "step": 20051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5095952749252319,
+      "learning_rate": 0.00045468185119587644,
+      "loss": 2.8989,
+      "step": 20052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5750041604042053,
+      "learning_rate": 0.00045466865914058443,
+      "loss": 3.045,
+      "step": 20053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5071499347686768,
+      "learning_rate": 0.00045465546667792497,
+      "loss": 3.0452,
+      "step": 20054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49465328454971313,
+      "learning_rate": 0.0004546422738079327,
+      "loss": 3.1946,
+      "step": 20055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5186102986335754,
+      "learning_rate": 0.0004546290805306426,
+      "loss": 3.2024,
+      "step": 20056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5584270358085632,
+      "learning_rate": 0.00045461588684608914,
+      "loss": 3.0319,
+      "step": 20057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5318349003791809,
+      "learning_rate": 0.0004546026927543072,
+      "loss": 2.9715,
+      "step": 20058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5026640295982361,
+      "learning_rate": 0.0004545894982553315,
+      "loss": 3.2427,
+      "step": 20059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5623165965080261,
+      "learning_rate": 0.0004545763033491968,
+      "loss": 3.2415,
+      "step": 20060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5373814702033997,
+      "learning_rate": 0.000454563108035938,
+      "loss": 3.3984,
+      "step": 20061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5032324194908142,
+      "learning_rate": 0.00045454991231558967,
+      "loss": 3.3189,
+      "step": 20062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5357590913772583,
+      "learning_rate": 0.00045453671618818646,
+      "loss": 3.0099,
+      "step": 20063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5333057641983032,
+      "learning_rate": 0.00045452351965376335,
+      "loss": 3.1873,
+      "step": 20064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5362182855606079,
+      "learning_rate": 0.0004545103227123552,
+      "loss": 3.1468,
+      "step": 20065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5074638724327087,
+      "learning_rate": 0.0004544971253639964,
+      "loss": 3.0841,
+      "step": 20066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5193195343017578,
+      "learning_rate": 0.000454483927608722,
+      "loss": 2.9173,
+      "step": 20067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5239328742027283,
+      "learning_rate": 0.0004544707294465667,
+      "loss": 3.2981,
+      "step": 20068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5404732823371887,
+      "learning_rate": 0.00045445753087756507,
+      "loss": 3.056,
+      "step": 20069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6814656257629395,
+      "learning_rate": 0.0004544443319017521,
+      "loss": 3.009,
+      "step": 20070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5465784668922424,
+      "learning_rate": 0.0004544311325191625,
+      "loss": 3.3474,
+      "step": 20071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5160285234451294,
+      "learning_rate": 0.00045441793272983107,
+      "loss": 2.8708,
+      "step": 20072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4963335692882538,
+      "learning_rate": 0.00045440473253379246,
+      "loss": 3.2226,
+      "step": 20073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5127677917480469,
+      "learning_rate": 0.00045439153193108155,
+      "loss": 3.2884,
+      "step": 20074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5165018439292908,
+      "learning_rate": 0.000454378330921733,
+      "loss": 3.0943,
+      "step": 20075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5334299206733704,
+      "learning_rate": 0.0004543651295057817,
+      "loss": 3.1644,
+      "step": 20076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5283639430999756,
+      "learning_rate": 0.0004543519276832624,
+      "loss": 3.1569,
+      "step": 20077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4838026762008667,
+      "learning_rate": 0.00045433872545420966,
+      "loss": 3.2105,
+      "step": 20078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5402780175209045,
+      "learning_rate": 0.00045432552281865854,
+      "loss": 2.9989,
+      "step": 20079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49882394075393677,
+      "learning_rate": 0.00045431231977664365,
+      "loss": 3.1154,
+      "step": 20080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5234292149543762,
+      "learning_rate": 0.00045429911632819986,
+      "loss": 2.9596,
+      "step": 20081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5079872608184814,
+      "learning_rate": 0.00045428591247336186,
+      "loss": 3.0581,
+      "step": 20082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6135797500610352,
+      "learning_rate": 0.0004542727082121644,
+      "loss": 3.2326,
+      "step": 20083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5245426893234253,
+      "learning_rate": 0.0004542595035446424,
+      "loss": 3.0909,
+      "step": 20084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5490965247154236,
+      "learning_rate": 0.0004542462984708305,
+      "loss": 3.227,
+      "step": 20085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5142471790313721,
+      "learning_rate": 0.0004542330929907636,
+      "loss": 3.2628,
+      "step": 20086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5162678360939026,
+      "learning_rate": 0.0004542198871044764,
+      "loss": 2.9272,
+      "step": 20087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.507563591003418,
+      "learning_rate": 0.00045420668081200364,
+      "loss": 3.107,
+      "step": 20088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49381065368652344,
+      "learning_rate": 0.00045419347411338015,
+      "loss": 3.141,
+      "step": 20089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5393127202987671,
+      "learning_rate": 0.00045418026700864083,
+      "loss": 3.4015,
+      "step": 20090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5078210830688477,
+      "learning_rate": 0.00045416705949782036,
+      "loss": 3.0461,
+      "step": 20091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5738405585289001,
+      "learning_rate": 0.00045415385158095343,
+      "loss": 3.1799,
+      "step": 20092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5566117167472839,
+      "learning_rate": 0.00045414064325807497,
+      "loss": 3.0548,
+      "step": 20093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5047556161880493,
+      "learning_rate": 0.0004541274345292197,
+      "loss": 3.0273,
+      "step": 20094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5348913073539734,
+      "learning_rate": 0.00045411422539442246,
+      "loss": 3.15,
+      "step": 20095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5044894218444824,
+      "learning_rate": 0.0004541010158537179,
+      "loss": 3.1082,
+      "step": 20096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5232664942741394,
+      "learning_rate": 0.00045408780590714103,
+      "loss": 3.1608,
+      "step": 20097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5661266446113586,
+      "learning_rate": 0.00045407459555472647,
+      "loss": 3.3411,
+      "step": 20098
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234664082527161,
+      "learning_rate": 0.0004540613847965092,
+      "loss": 3.2362,
+      "step": 20099
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5506570935249329,
+      "learning_rate": 0.0004540481736325238,
+      "loss": 3.1079,
+      "step": 20100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5332557559013367,
+      "learning_rate": 0.00045403496206280515,
+      "loss": 3.0901,
+      "step": 20101
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5116790533065796,
+      "learning_rate": 0.0004540217500873881,
+      "loss": 3.0163,
+      "step": 20102
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4963262677192688,
+      "learning_rate": 0.00045400853770630737,
+      "loss": 3.1048,
+      "step": 20103
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5719655752182007,
+      "learning_rate": 0.0004539953249195978,
+      "loss": 3.1882,
+      "step": 20104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6064972877502441,
+      "learning_rate": 0.0004539821117272942,
+      "loss": 3.1947,
+      "step": 20105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5541272163391113,
+      "learning_rate": 0.00045396889812943136,
+      "loss": 3.1424,
+      "step": 20106
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4937479496002197,
+      "learning_rate": 0.000453955684126044,
+      "loss": 3.3325,
+      "step": 20107
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.540358304977417,
+      "learning_rate": 0.0004539424697171671,
+      "loss": 3.3724,
+      "step": 20108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5559720396995544,
+      "learning_rate": 0.0004539292549028352,
+      "loss": 2.9808,
+      "step": 20109
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5821501612663269,
+      "learning_rate": 0.0004539160396830834,
+      "loss": 3.2358,
+      "step": 20110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5225769281387329,
+      "learning_rate": 0.00045390282405794634,
+      "loss": 3.2651,
+      "step": 20111
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.50220787525177,
+      "learning_rate": 0.0004538896080274589,
+      "loss": 3.1423,
+      "step": 20112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5386639833450317,
+      "learning_rate": 0.0004538763915916559,
+      "loss": 3.2382,
+      "step": 20113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5450523495674133,
+      "learning_rate": 0.0004538631747505719,
+      "loss": 3.2255,
+      "step": 20114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52878737449646,
+      "learning_rate": 0.000453849957504242,
+      "loss": 2.9539,
+      "step": 20115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5469775199890137,
+      "learning_rate": 0.000453836739852701,
+      "loss": 2.9554,
+      "step": 20116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5183457136154175,
+      "learning_rate": 0.0004538235217959836,
+      "loss": 3.0908,
+      "step": 20117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5041838884353638,
+      "learning_rate": 0.0004538103033341246,
+      "loss": 3.1683,
+      "step": 20118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5320984125137329,
+      "learning_rate": 0.00045379708446715894,
+      "loss": 3.1186,
+      "step": 20119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5572245717048645,
+      "learning_rate": 0.00045378386519512125,
+      "loss": 3.2272,
+      "step": 20120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6518499255180359,
+      "learning_rate": 0.0004537706455180465,
+      "loss": 3.1933,
+      "step": 20121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5835242867469788,
+      "learning_rate": 0.0004537574254359695,
+      "loss": 3.2654,
+      "step": 20122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5642088055610657,
+      "learning_rate": 0.0004537442049489251,
+      "loss": 3.2556,
+      "step": 20123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5181591510772705,
+      "learning_rate": 0.00045373098405694786,
+      "loss": 3.1488,
+      "step": 20124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5602161884307861,
+      "learning_rate": 0.0004537177627600729,
+      "loss": 3.0757,
+      "step": 20125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5879859328269958,
+      "learning_rate": 0.00045370454105833494,
+      "loss": 3.2037,
+      "step": 20126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5683193802833557,
+      "learning_rate": 0.00045369131895176885,
+      "loss": 2.9962,
+      "step": 20127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5079681873321533,
+      "learning_rate": 0.0004536780964404093,
+      "loss": 3.0535,
+      "step": 20128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5396766662597656,
+      "learning_rate": 0.0004536648735242912,
+      "loss": 3.1192,
+      "step": 20129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5396043658256531,
+      "learning_rate": 0.0004536516502034495,
+      "loss": 3.2537,
+      "step": 20130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5514666438102722,
+      "learning_rate": 0.0004536384264779189,
+      "loss": 2.9611,
+      "step": 20131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5507171750068665,
+      "learning_rate": 0.0004536252023477343,
+      "loss": 3.2089,
+      "step": 20132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5136435627937317,
+      "learning_rate": 0.0004536119778129304,
+      "loss": 3.0577,
+      "step": 20133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5072475671768188,
+      "learning_rate": 0.00045359875287354214,
+      "loss": 3.2917,
+      "step": 20134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5063516497612,
+      "learning_rate": 0.00045358552752960426,
+      "loss": 3.2465,
+      "step": 20135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.545147716999054,
+      "learning_rate": 0.00045357230178115173,
+      "loss": 3.042,
+      "step": 20136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5409039855003357,
+      "learning_rate": 0.0004535590756282193,
+      "loss": 3.3088,
+      "step": 20137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5154100656509399,
+      "learning_rate": 0.0004535458490708419,
+      "loss": 3.2658,
+      "step": 20138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5560367703437805,
+      "learning_rate": 0.0004535326221090542,
+      "loss": 3.0783,
+      "step": 20139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5108457207679749,
+      "learning_rate": 0.0004535193947428911,
+      "loss": 3.1656,
+      "step": 20140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5491657257080078,
+      "learning_rate": 0.0004535061669723875,
+      "loss": 3.3704,
+      "step": 20141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5991771221160889,
+      "learning_rate": 0.00045349293879757814,
+      "loss": 3.3599,
+      "step": 20142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.530377209186554,
+      "learning_rate": 0.00045347971021849796,
+      "loss": 3.3084,
+      "step": 20143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546180009841919,
+      "learning_rate": 0.0004534664812351818,
+      "loss": 3.4132,
+      "step": 20144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5626497864723206,
+      "learning_rate": 0.0004534532518476644,
+      "loss": 3.2105,
+      "step": 20145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.508690357208252,
+      "learning_rate": 0.00045344002205598074,
+      "loss": 3.274,
+      "step": 20146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5522834062576294,
+      "learning_rate": 0.00045342679186016554,
+      "loss": 3.0697,
+      "step": 20147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5134068727493286,
+      "learning_rate": 0.00045341356126025366,
+      "loss": 3.2479,
+      "step": 20148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5544729232788086,
+      "learning_rate": 0.00045340033025628013,
+      "loss": 3.0962,
+      "step": 20149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5248307585716248,
+      "learning_rate": 0.00045338709884827947,
+      "loss": 3.262,
+      "step": 20150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5262455344200134,
+      "learning_rate": 0.00045337386703628676,
+      "loss": 3.1155,
+      "step": 20151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5151076912879944,
+      "learning_rate": 0.00045336063482033696,
+      "loss": 3.0329,
+      "step": 20152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5224588513374329,
+      "learning_rate": 0.00045334740220046455,
+      "loss": 3.1338,
+      "step": 20153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5186708569526672,
+      "learning_rate": 0.0004533341691767047,
+      "loss": 3.0215,
+      "step": 20154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5381702780723572,
+      "learning_rate": 0.0004533209357490921,
+      "loss": 3.2761,
+      "step": 20155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5323552489280701,
+      "learning_rate": 0.00045330770191766176,
+      "loss": 3.1807,
+      "step": 20156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5015830993652344,
+      "learning_rate": 0.0004532944676824484,
+      "loss": 3.0101,
+      "step": 20157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5259556770324707,
+      "learning_rate": 0.0004532812330434869,
+      "loss": 3.1135,
+      "step": 20158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.544632613658905,
+      "learning_rate": 0.00045326799800081213,
+      "loss": 3.2611,
+      "step": 20159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5256531834602356,
+      "learning_rate": 0.00045325476255445886,
+      "loss": 3.0899,
+      "step": 20160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5340846180915833,
+      "learning_rate": 0.00045324152670446217,
+      "loss": 3.2513,
+      "step": 20161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5545790791511536,
+      "learning_rate": 0.00045322829045085674,
+      "loss": 2.9697,
+      "step": 20162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5166133642196655,
+      "learning_rate": 0.00045321505379367755,
+      "loss": 3.2083,
+      "step": 20163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5389953851699829,
+      "learning_rate": 0.00045320181673295934,
+      "loss": 2.8237,
+      "step": 20164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5397850871086121,
+      "learning_rate": 0.00045318857926873697,
+      "loss": 3.0537,
+      "step": 20165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5113707780838013,
+      "learning_rate": 0.0004531753414010454,
+      "loss": 3.1937,
+      "step": 20166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5229274034500122,
+      "learning_rate": 0.00045316210312991954,
+      "loss": 3.3204,
+      "step": 20167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5383738279342651,
+      "learning_rate": 0.00045314886445539415,
+      "loss": 3.3542,
+      "step": 20168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5492693185806274,
+      "learning_rate": 0.00045313562537750403,
+      "loss": 3.1827,
+      "step": 20169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5596176385879517,
+      "learning_rate": 0.00045312238589628425,
+      "loss": 3.1018,
+      "step": 20170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5374123454093933,
+      "learning_rate": 0.00045310914601176956,
+      "loss": 3.2086,
+      "step": 20171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5204119086265564,
+      "learning_rate": 0.0004530959057239947,
+      "loss": 3.0035,
+      "step": 20172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5454662442207336,
+      "learning_rate": 0.0004530826650329948,
+      "loss": 3.2789,
+      "step": 20173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5484021902084351,
+      "learning_rate": 0.00045306942393880475,
+      "loss": 3.1308,
+      "step": 20174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5319493412971497,
+      "learning_rate": 0.00045305618244145915,
+      "loss": 3.2038,
+      "step": 20175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.529231071472168,
+      "learning_rate": 0.000453042940540993,
+      "loss": 3.1151,
+      "step": 20176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5511825084686279,
+      "learning_rate": 0.0004530296982374412,
+      "loss": 3.2654,
+      "step": 20177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5340495109558105,
+      "learning_rate": 0.0004530164555308388,
+      "loss": 3.2052,
+      "step": 20178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5212454795837402,
+      "learning_rate": 0.0004530032124212203,
+      "loss": 3.1776,
+      "step": 20179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4973335564136505,
+      "learning_rate": 0.00045298996890862087,
+      "loss": 3.2903,
+      "step": 20180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5556290149688721,
+      "learning_rate": 0.0004529767249930753,
+      "loss": 3.0608,
+      "step": 20181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5058223009109497,
+      "learning_rate": 0.00045296348067461846,
+      "loss": 3.2073,
+      "step": 20182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5586454272270203,
+      "learning_rate": 0.00045295023595328526,
+      "loss": 3.0764,
+      "step": 20183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5111760497093201,
+      "learning_rate": 0.0004529369908291106,
+      "loss": 2.9447,
+      "step": 20184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5140455961227417,
+      "learning_rate": 0.00045292374530212935,
+      "loss": 3.1685,
+      "step": 20185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5379027128219604,
+      "learning_rate": 0.0004529104993723763,
+      "loss": 3.1918,
+      "step": 20186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4842080771923065,
+      "learning_rate": 0.0004528972530398864,
+      "loss": 3.2543,
+      "step": 20187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5368223786354065,
+      "learning_rate": 0.0004528840063046947,
+      "loss": 3.1263,
+      "step": 20188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5022522807121277,
+      "learning_rate": 0.0004528707591668359,
+      "loss": 3.0206,
+      "step": 20189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5098612904548645,
+      "learning_rate": 0.0004528575116263449,
+      "loss": 2.9806,
+      "step": 20190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5354186296463013,
+      "learning_rate": 0.00045284426368325664,
+      "loss": 3.1973,
+      "step": 20191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5448538661003113,
+      "learning_rate": 0.00045283101533760613,
+      "loss": 3.1923,
+      "step": 20192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5186355710029602,
+      "learning_rate": 0.00045281776658942795,
+      "loss": 3.1251,
+      "step": 20193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5228528380393982,
+      "learning_rate": 0.00045280451743875727,
+      "loss": 3.1935,
+      "step": 20194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5764790177345276,
+      "learning_rate": 0.00045279126788562885,
+      "loss": 3.1752,
+      "step": 20195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5042545199394226,
+      "learning_rate": 0.0004527780179300777,
+      "loss": 3.021,
+      "step": 20196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5068015456199646,
+      "learning_rate": 0.00045276476757213864,
+      "loss": 3.1387,
+      "step": 20197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5311840176582336,
+      "learning_rate": 0.00045275151681184656,
+      "loss": 2.9341,
+      "step": 20198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57582688331604,
+      "learning_rate": 0.0004527382656492364,
+      "loss": 2.6663,
+      "step": 20199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5275920033454895,
+      "learning_rate": 0.00045272501408434313,
+      "loss": 2.8333,
+      "step": 20200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5140002965927124,
+      "learning_rate": 0.00045271176211720133,
+      "loss": 3.1819,
+      "step": 20201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5207107067108154,
+      "learning_rate": 0.0004526985097478464,
+      "loss": 3.1758,
+      "step": 20202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5530972480773926,
+      "learning_rate": 0.0004526852569763129,
+      "loss": 3.134,
+      "step": 20203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5513622164726257,
+      "learning_rate": 0.00045267200380263577,
+      "loss": 3.0947,
+      "step": 20204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5309062004089355,
+      "learning_rate": 0.00045265875022685,
+      "loss": 3.1286,
+      "step": 20205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5550915002822876,
+      "learning_rate": 0.0004526454962489904,
+      "loss": 3.1141,
+      "step": 20206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5073148608207703,
+      "learning_rate": 0.000452632241869092,
+      "loss": 3.1703,
+      "step": 20207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.544578492641449,
+      "learning_rate": 0.00045261898708718966,
+      "loss": 3.1061,
+      "step": 20208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5766841173171997,
+      "learning_rate": 0.0004526057319033182,
+      "loss": 3.0463,
+      "step": 20209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5074905753135681,
+      "learning_rate": 0.00045259247631751265,
+      "loss": 2.9534,
+      "step": 20210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5443254709243774,
+      "learning_rate": 0.00045257922032980794,
+      "loss": 3.3291,
+      "step": 20211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.505168080329895,
+      "learning_rate": 0.0004525659639402388,
+      "loss": 3.2416,
+      "step": 20212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5314830541610718,
+      "learning_rate": 0.00045255270714884035,
+      "loss": 3.1842,
+      "step": 20213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5724264979362488,
+      "learning_rate": 0.00045253944995564746,
+      "loss": 3.0492,
+      "step": 20214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5013821721076965,
+      "learning_rate": 0.0004525261923606951,
+      "loss": 2.9629,
+      "step": 20215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5013492703437805,
+      "learning_rate": 0.0004525129343640179,
+      "loss": 3.0972,
+      "step": 20216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5238487124443054,
+      "learning_rate": 0.00045249967596565105,
+      "loss": 3.1335,
+      "step": 20217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5291327238082886,
+      "learning_rate": 0.0004524864171656295,
+      "loss": 3.0083,
+      "step": 20218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5021530985832214,
+      "learning_rate": 0.000452473157963988,
+      "loss": 3.1272,
+      "step": 20219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5439477562904358,
+      "learning_rate": 0.00045245989836076154,
+      "loss": 3.0997,
+      "step": 20220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.528767466545105,
+      "learning_rate": 0.00045244663835598505,
+      "loss": 3.1914,
+      "step": 20221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5145512223243713,
+      "learning_rate": 0.00045243337794969343,
+      "loss": 3.1779,
+      "step": 20222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5444108247756958,
+      "learning_rate": 0.00045242011714192174,
+      "loss": 3.1191,
+      "step": 20223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.537213146686554,
+      "learning_rate": 0.00045240685593270474,
+      "loss": 3.0256,
+      "step": 20224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4993702471256256,
+      "learning_rate": 0.00045239359432207733,
+      "loss": 3.2517,
+      "step": 20225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.547257125377655,
+      "learning_rate": 0.00045238033231007464,
+      "loss": 3.1549,
+      "step": 20226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5551975965499878,
+      "learning_rate": 0.0004523670698967314,
+      "loss": 3.1891,
+      "step": 20227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5113047361373901,
+      "learning_rate": 0.00045235380708208255,
+      "loss": 3.192,
+      "step": 20228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5420333743095398,
+      "learning_rate": 0.0004523405438661633,
+      "loss": 3.0101,
+      "step": 20229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5321565866470337,
+      "learning_rate": 0.0004523272802490083,
+      "loss": 2.9095,
+      "step": 20230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5126756429672241,
+      "learning_rate": 0.0004523140162306525,
+      "loss": 3.0235,
+      "step": 20231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4914495646953583,
+      "learning_rate": 0.0004523007518111309,
+      "loss": 3.1374,
+      "step": 20232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.501433253288269,
+      "learning_rate": 0.00045228748699047844,
+      "loss": 2.8911,
+      "step": 20233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5397343635559082,
+      "learning_rate": 0.00045227422176872997,
+      "loss": 3.2692,
+      "step": 20234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5279703736305237,
+      "learning_rate": 0.0004522609561459207,
+      "loss": 3.2089,
+      "step": 20235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.533789336681366,
+      "learning_rate": 0.00045224769012208526,
+      "loss": 3.0578,
+      "step": 20236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5652415752410889,
+      "learning_rate": 0.0004522344236972587,
+      "loss": 3.2415,
+      "step": 20237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4909104108810425,
+      "learning_rate": 0.0004522211568714759,
+      "loss": 3.1395,
+      "step": 20238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5707433819770813,
+      "learning_rate": 0.000452207889644772,
+      "loss": 3.1975,
+      "step": 20239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6347876191139221,
+      "learning_rate": 0.0004521946220171818,
+      "loss": 3.0497,
+      "step": 20240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5172315835952759,
+      "learning_rate": 0.0004521813539887402,
+      "loss": 3.335,
+      "step": 20241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.563426673412323,
+      "learning_rate": 0.00045216808555948216,
+      "loss": 3.099,
+      "step": 20242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5538248419761658,
+      "learning_rate": 0.0004521548167294428,
+      "loss": 3.2189,
+      "step": 20243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5178148150444031,
+      "learning_rate": 0.0004521415474986568,
+      "loss": 3.0827,
+      "step": 20244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5169044733047485,
+      "learning_rate": 0.00045212827786715934,
+      "loss": 3.1631,
+      "step": 20245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5301280617713928,
+      "learning_rate": 0.0004521150078349852,
+      "loss": 3.2562,
+      "step": 20246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5236168503761292,
+      "learning_rate": 0.00045210173740216944,
+      "loss": 3.1808,
+      "step": 20247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6142783164978027,
+      "learning_rate": 0.00045208846656874703,
+      "loss": 3.0495,
+      "step": 20248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5414308309555054,
+      "learning_rate": 0.00045207519533475274,
+      "loss": 2.9324,
+      "step": 20249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.485738068819046,
+      "learning_rate": 0.0004520619237002218,
+      "loss": 3.1258,
+      "step": 20250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5361891388893127,
+      "learning_rate": 0.0004520486516651889,
+      "loss": 3.0218,
+      "step": 20251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5489019155502319,
+      "learning_rate": 0.00045203537922968915,
+      "loss": 3.201,
+      "step": 20252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.519248902797699,
+      "learning_rate": 0.00045202210639375747,
+      "loss": 3.2805,
+      "step": 20253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4982036352157593,
+      "learning_rate": 0.0004520088331574289,
+      "loss": 3.2192,
+      "step": 20254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5829838514328003,
+      "learning_rate": 0.00045199555952073824,
+      "loss": 3.2538,
+      "step": 20255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5190660953521729,
+      "learning_rate": 0.00045198228548372056,
+      "loss": 3.0256,
+      "step": 20256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5045883655548096,
+      "learning_rate": 0.00045196901104641073,
+      "loss": 3.0908,
+      "step": 20257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5397174954414368,
+      "learning_rate": 0.0004519557362088438,
+      "loss": 3.3517,
+      "step": 20258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5239180326461792,
+      "learning_rate": 0.00045194246097105467,
+      "loss": 3.358,
+      "step": 20259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52662593126297,
+      "learning_rate": 0.0004519291853330784,
+      "loss": 3.168,
+      "step": 20260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5203768014907837,
+      "learning_rate": 0.00045191590929494994,
+      "loss": 3.0684,
+      "step": 20261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5362476706504822,
+      "learning_rate": 0.0004519026328567041,
+      "loss": 3.1779,
+      "step": 20262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5580528974533081,
+      "learning_rate": 0.00045188935601837604,
+      "loss": 3.0697,
+      "step": 20263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5004633069038391,
+      "learning_rate": 0.00045187607878000057,
+      "loss": 3.2735,
+      "step": 20264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5355501174926758,
+      "learning_rate": 0.0004518628011416128,
+      "loss": 3.1514,
+      "step": 20265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5439483523368835,
+      "learning_rate": 0.0004518495231032477,
+      "loss": 3.2839,
+      "step": 20266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5279213190078735,
+      "learning_rate": 0.00045183624466494006,
+      "loss": 3.1213,
+      "step": 20267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5979596376419067,
+      "learning_rate": 0.00045182296582672496,
+      "loss": 3.1,
+      "step": 20268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5285710096359253,
+      "learning_rate": 0.0004518096865886375,
+      "loss": 3.3007,
+      "step": 20269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.571806788444519,
+      "learning_rate": 0.00045179640695071246,
+      "loss": 3.225,
+      "step": 20270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5280594229698181,
+      "learning_rate": 0.000451783126912985,
+      "loss": 3.1499,
+      "step": 20271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5388377904891968,
+      "learning_rate": 0.00045176984647548994,
+      "loss": 3.1795,
+      "step": 20272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5278087258338928,
+      "learning_rate": 0.00045175656563826224,
+      "loss": 3.2452,
+      "step": 20273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5274179577827454,
+      "learning_rate": 0.000451743284401337,
+      "loss": 3.1447,
+      "step": 20274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5059541463851929,
+      "learning_rate": 0.0004517300027647491,
+      "loss": 3.0891,
+      "step": 20275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5273914337158203,
+      "learning_rate": 0.0004517167207285337,
+      "loss": 3.1935,
+      "step": 20276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5308375954627991,
+      "learning_rate": 0.00045170343829272565,
+      "loss": 3.1559,
+      "step": 20277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.530507504940033,
+      "learning_rate": 0.00045169015545735975,
+      "loss": 2.9621,
+      "step": 20278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5165071487426758,
+      "learning_rate": 0.0004516768722224713,
+      "loss": 3.1421,
+      "step": 20279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5435473322868347,
+      "learning_rate": 0.00045166358858809524,
+      "loss": 3.0609,
+      "step": 20280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5258364081382751,
+      "learning_rate": 0.0004516503045542664,
+      "loss": 2.9946,
+      "step": 20281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.49890580773353577,
+      "learning_rate": 0.0004516370201210198,
+      "loss": 3.0339,
+      "step": 20282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5879883170127869,
+      "learning_rate": 0.0004516237352883905,
+      "loss": 3.0634,
+      "step": 20283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5395330786705017,
+      "learning_rate": 0.00045161045005641344,
+      "loss": 3.2598,
+      "step": 20284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5076286196708679,
+      "learning_rate": 0.00045159716442512367,
+      "loss": 3.1879,
+      "step": 20285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.508956253528595,
+      "learning_rate": 0.0004515838783945562,
+      "loss": 3.0942,
+      "step": 20286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5356499552726746,
+      "learning_rate": 0.00045157059196474593,
+      "loss": 3.1142,
+      "step": 20287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.535220742225647,
+      "learning_rate": 0.0004515573051357278,
+      "loss": 2.9021,
+      "step": 20288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5434659123420715,
+      "learning_rate": 0.000451544017907537,
+      "loss": 3.0314,
+      "step": 20289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5554776191711426,
+      "learning_rate": 0.0004515307302802084,
+      "loss": 2.9867,
+      "step": 20290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5213914513587952,
+      "learning_rate": 0.00045151744225377697,
+      "loss": 3.2232,
+      "step": 20291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5432573556900024,
+      "learning_rate": 0.0004515041538282778,
+      "loss": 3.1003,
+      "step": 20292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4953993558883667,
+      "learning_rate": 0.00045149086500374585,
+      "loss": 3.2513,
+      "step": 20293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5025012493133545,
+      "learning_rate": 0.0004514775757802161,
+      "loss": 3.2112,
+      "step": 20294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5353289246559143,
+      "learning_rate": 0.0004514642861577236,
+      "loss": 3.0389,
+      "step": 20295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5243365168571472,
+      "learning_rate": 0.0004514509961363033,
+      "loss": 3.0729,
+      "step": 20296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5028917789459229,
+      "learning_rate": 0.0004514377057159902,
+      "loss": 3.1696,
+      "step": 20297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5022268891334534,
+      "learning_rate": 0.0004514244148968194,
+      "loss": 2.985,
+      "step": 20298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5356721878051758,
+      "learning_rate": 0.00045141112367882573,
+      "loss": 3.148,
+      "step": 20299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5337976217269897,
+      "learning_rate": 0.0004513978320620443,
+      "loss": 2.9342,
+      "step": 20300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5110748410224915,
+      "learning_rate": 0.00045138454004651016,
+      "loss": 3.2471,
+      "step": 20301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5441080331802368,
+      "learning_rate": 0.00045137124763225834,
+      "loss": 3.0728,
+      "step": 20302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5328307747840881,
+      "learning_rate": 0.00045135795481932375,
+      "loss": 3.1303,
+      "step": 20303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5662921071052551,
+      "learning_rate": 0.00045134466160774136,
+      "loss": 3.2292,
+      "step": 20304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5180608034133911,
+      "learning_rate": 0.0004513313679975463,
+      "loss": 3.303,
+      "step": 20305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5455940961837769,
+      "learning_rate": 0.0004513180739887736,
+      "loss": 3.3352,
+      "step": 20306
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5291252732276917,
+      "learning_rate": 0.00045130477958145817,
+      "loss": 3.2022,
+      "step": 20307
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5131491422653198,
+      "learning_rate": 0.00045129148477563504,
+      "loss": 3.1474,
+      "step": 20308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5317510962486267,
+      "learning_rate": 0.0004512781895713393,
+      "loss": 3.1086,
+      "step": 20309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5623073577880859,
+      "learning_rate": 0.0004512648939686059,
+      "loss": 3.0699,
+      "step": 20310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5236507654190063,
+      "learning_rate": 0.0004512515979674698,
+      "loss": 2.855,
+      "step": 20311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5638948082923889,
+      "learning_rate": 0.00045123830156796626,
+      "loss": 2.9894,
+      "step": 20312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515745222568512,
+      "learning_rate": 0.0004512250047701301,
+      "loss": 3.3086,
+      "step": 20313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5643942952156067,
+      "learning_rate": 0.0004512117075739963,
+      "loss": 3.1348,
+      "step": 20314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5207350850105286,
+      "learning_rate": 0.0004511984099796,
+      "loss": 3.3207,
+      "step": 20315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5619992017745972,
+      "learning_rate": 0.0004511851119869762,
+      "loss": 3.3062,
+      "step": 20316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5149810910224915,
+      "learning_rate": 0.00045117181359615995,
+      "loss": 3.1298,
+      "step": 20317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.528748095035553,
+      "learning_rate": 0.00045115851480718625,
+      "loss": 3.1774,
+      "step": 20318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5054760575294495,
+      "learning_rate": 0.00045114521562008994,
+      "loss": 3.1187,
+      "step": 20319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5329738259315491,
+      "learning_rate": 0.00045113191603490636,
+      "loss": 3.0029,
+      "step": 20320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5501445531845093,
+      "learning_rate": 0.0004511186160516704,
+      "loss": 3.3807,
+      "step": 20321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5410068035125732,
+      "learning_rate": 0.00045110531567041715,
+      "loss": 3.0491,
+      "step": 20322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5346465110778809,
+      "learning_rate": 0.0004510920148911814,
+      "loss": 3.1324,
+      "step": 20323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5548122525215149,
+      "learning_rate": 0.00045107871371399854,
+      "loss": 3.0366,
+      "step": 20324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5406346917152405,
+      "learning_rate": 0.0004510654121389033,
+      "loss": 3.353,
+      "step": 20325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5303011536598206,
+      "learning_rate": 0.0004510521101659308,
+      "loss": 3.1889,
+      "step": 20326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5354381203651428,
+      "learning_rate": 0.0004510388077951163,
+      "loss": 3.0421,
+      "step": 20327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5239824056625366,
+      "learning_rate": 0.0004510255050264945,
+      "loss": 3.2569,
+      "step": 20328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5124732255935669,
+      "learning_rate": 0.00045101220186010056,
+      "loss": 3.2477,
+      "step": 20329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.528676450252533,
+      "learning_rate": 0.00045099889829596965,
+      "loss": 3.027,
+      "step": 20330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5336599349975586,
+      "learning_rate": 0.00045098559433413675,
+      "loss": 2.9263,
+      "step": 20331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5987780094146729,
+      "learning_rate": 0.00045097228997463676,
+      "loss": 3.0757,
+      "step": 20332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5460469722747803,
+      "learning_rate": 0.00045095898521750475,
+      "loss": 2.9283,
+      "step": 20333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5135546326637268,
+      "learning_rate": 0.0004509456800627759,
+      "loss": 3.0009,
+      "step": 20334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5272367596626282,
+      "learning_rate": 0.0004509323745104852,
+      "loss": 3.1307,
+      "step": 20335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5386490225791931,
+      "learning_rate": 0.00045091906856066765,
+      "loss": 3.2579,
+      "step": 20336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54600989818573,
+      "learning_rate": 0.00045090576221335833,
+      "loss": 3.3436,
+      "step": 20337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5243917107582092,
+      "learning_rate": 0.00045089245546859225,
+      "loss": 3.1057,
+      "step": 20338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5199042558670044,
+      "learning_rate": 0.0004508791483264045,
+      "loss": 3.1192,
+      "step": 20339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5423737168312073,
+      "learning_rate": 0.0004508658407868301,
+      "loss": 3.2666,
+      "step": 20340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.592741847038269,
+      "learning_rate": 0.000450852532849904,
+      "loss": 2.9049,
+      "step": 20341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5735978484153748,
+      "learning_rate": 0.0004508392245156615,
+      "loss": 3.1306,
+      "step": 20342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5209268927574158,
+      "learning_rate": 0.0004508259157841375,
+      "loss": 3.3385,
+      "step": 20343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5497926473617554,
+      "learning_rate": 0.0004508126066553671,
+      "loss": 3.0599,
+      "step": 20344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5306968092918396,
+      "learning_rate": 0.0004507992971293852,
+      "loss": 3.067,
+      "step": 20345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5316476821899414,
+      "learning_rate": 0.00045078598720622707,
+      "loss": 3.1147,
+      "step": 20346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.528810977935791,
+      "learning_rate": 0.0004507726768859277,
+      "loss": 3.0775,
+      "step": 20347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5107991099357605,
+      "learning_rate": 0.00045075936616852206,
+      "loss": 3.1226,
+      "step": 20348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.513142466545105,
+      "learning_rate": 0.0004507460550540452,
+      "loss": 3.1823,
+      "step": 20349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6255738735198975,
+      "learning_rate": 0.0004507327435425323,
+      "loss": 3.0949,
+      "step": 20350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5634238719940186,
+      "learning_rate": 0.00045071943163401833,
+      "loss": 2.9638,
+      "step": 20351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5344851016998291,
+      "learning_rate": 0.0004507061193285384,
+      "loss": 3.1896,
+      "step": 20352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5088251233100891,
+      "learning_rate": 0.00045069280662612764,
+      "loss": 3.0654,
+      "step": 20353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5598111748695374,
+      "learning_rate": 0.00045067949352682097,
+      "loss": 3.1859,
+      "step": 20354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5614855289459229,
+      "learning_rate": 0.00045066618003065343,
+      "loss": 3.1514,
+      "step": 20355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5441474318504333,
+      "learning_rate": 0.0004506528661376603,
+      "loss": 3.2191,
+      "step": 20356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5195251107215881,
+      "learning_rate": 0.0004506395518478765,
+      "loss": 3.0811,
+      "step": 20357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5556178092956543,
+      "learning_rate": 0.00045062623716133704,
+      "loss": 3.1629,
+      "step": 20358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5472621917724609,
+      "learning_rate": 0.00045061292207807706,
+      "loss": 3.049,
+      "step": 20359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5794954299926758,
+      "learning_rate": 0.00045059960659813163,
+      "loss": 3.0257,
+      "step": 20360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5246169567108154,
+      "learning_rate": 0.0004505862907215359,
+      "loss": 3.2008,
+      "step": 20361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6028974056243896,
+      "learning_rate": 0.0004505729744483248,
+      "loss": 3.1549,
+      "step": 20362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5344709753990173,
+      "learning_rate": 0.0004505596577785334,
+      "loss": 3.1225,
+      "step": 20363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49651411175727844,
+      "learning_rate": 0.000450546340712197,
+      "loss": 3.1713,
+      "step": 20364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5115389823913574,
+      "learning_rate": 0.0004505330232493504,
+      "loss": 3.1379,
+      "step": 20365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5067347288131714,
+      "learning_rate": 0.00045051970539002875,
+      "loss": 3.1478,
+      "step": 20366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.583716630935669,
+      "learning_rate": 0.00045050638713426723,
+      "loss": 3.0329,
+      "step": 20367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5057085156440735,
+      "learning_rate": 0.0004504930684821009,
+      "loss": 3.1614,
+      "step": 20368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5527999401092529,
+      "learning_rate": 0.00045047974943356473,
+      "loss": 3.1671,
+      "step": 20369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.549720287322998,
+      "learning_rate": 0.00045046642998869385,
+      "loss": 3.1481,
+      "step": 20370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5233119130134583,
+      "learning_rate": 0.0004504531101475234,
+      "loss": 3.2703,
+      "step": 20371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5394256711006165,
+      "learning_rate": 0.0004504397899100883,
+      "loss": 3.0763,
+      "step": 20372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5158621668815613,
+      "learning_rate": 0.0004504264692764239,
+      "loss": 3.3617,
+      "step": 20373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5600789189338684,
+      "learning_rate": 0.000450413148246565,
+      "loss": 3.1524,
+      "step": 20374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5463529825210571,
+      "learning_rate": 0.00045039982682054696,
+      "loss": 3.378,
+      "step": 20375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48953327536582947,
+      "learning_rate": 0.0004503865049984045,
+      "loss": 3.0364,
+      "step": 20376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5219492316246033,
+      "learning_rate": 0.0004503731827801731,
+      "loss": 3.0698,
+      "step": 20377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.557974636554718,
+      "learning_rate": 0.0004503598601658877,
+      "loss": 3.2128,
+      "step": 20378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5604296922683716,
+      "learning_rate": 0.0004503465371555833,
+      "loss": 2.9745,
+      "step": 20379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5162230730056763,
+      "learning_rate": 0.0004503332137492951,
+      "loss": 3.0853,
+      "step": 20380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5778909921646118,
+      "learning_rate": 0.00045031988994705796,
+      "loss": 3.2235,
+      "step": 20381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5140716433525085,
+      "learning_rate": 0.00045030656574890745,
+      "loss": 3.2436,
+      "step": 20382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5335753560066223,
+      "learning_rate": 0.0004502932411548781,
+      "loss": 3.1531,
+      "step": 20383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5693366527557373,
+      "learning_rate": 0.00045027991616500545,
+      "loss": 3.3253,
+      "step": 20384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5139354467391968,
+      "learning_rate": 0.0004502665907793244,
+      "loss": 2.9069,
+      "step": 20385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49556657671928406,
+      "learning_rate": 0.00045025326499787007,
+      "loss": 3.2041,
+      "step": 20386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.513113796710968,
+      "learning_rate": 0.0004502399388206775,
+      "loss": 3.2202,
+      "step": 20387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5639035701751709,
+      "learning_rate": 0.0004502266122477819,
+      "loss": 3.2123,
+      "step": 20388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5301312208175659,
+      "learning_rate": 0.00045021328527921825,
+      "loss": 3.1671,
+      "step": 20389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5456419587135315,
+      "learning_rate": 0.00045019995791502175,
+      "loss": 3.1206,
+      "step": 20390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5344980359077454,
+      "learning_rate": 0.00045018663015522747,
+      "loss": 3.1698,
+      "step": 20391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5286158919334412,
+      "learning_rate": 0.00045017330199987053,
+      "loss": 3.1031,
+      "step": 20392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5467382073402405,
+      "learning_rate": 0.000450159973448986,
+      "loss": 3.0543,
+      "step": 20393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5618236064910889,
+      "learning_rate": 0.000450146644502609,
+      "loss": 3.0087,
+      "step": 20394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5205358266830444,
+      "learning_rate": 0.0004501333151607747,
+      "loss": 3.2562,
+      "step": 20395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49830591678619385,
+      "learning_rate": 0.0004501199854235181,
+      "loss": 3.0565,
+      "step": 20396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5011833310127258,
+      "learning_rate": 0.0004501066552908743,
+      "loss": 3.174,
+      "step": 20397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5207628607749939,
+      "learning_rate": 0.00045009332476287847,
+      "loss": 3.16,
+      "step": 20398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5745342969894409,
+      "learning_rate": 0.00045007999383956564,
+      "loss": 3.434,
+      "step": 20399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5692901015281677,
+      "learning_rate": 0.00045006666252097113,
+      "loss": 3.283,
+      "step": 20400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.47566521167755127,
+      "learning_rate": 0.00045005333080712985,
+      "loss": 3.0213,
+      "step": 20401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5313223004341125,
+      "learning_rate": 0.000450039998698077,
+      "loss": 3.1536,
+      "step": 20402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5201346278190613,
+      "learning_rate": 0.0004500266661938476,
+      "loss": 3.1187,
+      "step": 20403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5419844388961792,
+      "learning_rate": 0.000450013333294477,
+      "loss": 3.0795,
+      "step": 20404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6561049222946167,
+      "learning_rate": 0.00045,
+      "loss": 3.1308,
+      "step": 20405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5238326191902161,
+      "learning_rate": 0.00044998666631045184,
+      "loss": 3.3698,
+      "step": 20406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5477263927459717,
+      "learning_rate": 0.0004499733322258678,
+      "loss": 3.1224,
+      "step": 20407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5472010374069214,
+      "learning_rate": 0.0004499599977462828,
+      "loss": 2.8647,
+      "step": 20408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5201596021652222,
+      "learning_rate": 0.00044994666287173196,
+      "loss": 3.1365,
+      "step": 20409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5760728716850281,
+      "learning_rate": 0.0004499333276022506,
+      "loss": 2.9512,
+      "step": 20410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5273072719573975,
+      "learning_rate": 0.0004499199919378736,
+      "loss": 2.9243,
+      "step": 20411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5336431860923767,
+      "learning_rate": 0.0004499066558786362,
+      "loss": 3.16,
+      "step": 20412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5312026739120483,
+      "learning_rate": 0.0004498933194245735,
+      "loss": 3.1535,
+      "step": 20413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5205845832824707,
+      "learning_rate": 0.00044987998257572075,
+      "loss": 3.1798,
+      "step": 20414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5186797380447388,
+      "learning_rate": 0.000449866645332113,
+      "loss": 2.9815,
+      "step": 20415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5282081365585327,
+      "learning_rate": 0.0004498533076937852,
+      "loss": 3.1629,
+      "step": 20416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.501063346862793,
+      "learning_rate": 0.00044983996966077263,
+      "loss": 3.1123,
+      "step": 20417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5373258590698242,
+      "learning_rate": 0.0004498266312331106,
+      "loss": 3.0061,
+      "step": 20418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5097174048423767,
+      "learning_rate": 0.0004498132924108339,
+      "loss": 3.1618,
+      "step": 20419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4925655126571655,
+      "learning_rate": 0.00044979995319397787,
+      "loss": 3.2096,
+      "step": 20420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5566980242729187,
+      "learning_rate": 0.0004497866135825776,
+      "loss": 2.9665,
+      "step": 20421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.62166428565979,
+      "learning_rate": 0.00044977327357666815,
+      "loss": 3.0511,
+      "step": 20422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49404507875442505,
+      "learning_rate": 0.00044975993317628477,
+      "loss": 3.1312,
+      "step": 20423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5426779985427856,
+      "learning_rate": 0.00044974659238146257,
+      "loss": 3.3202,
+      "step": 20424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.549683690071106,
+      "learning_rate": 0.0004497332511922366,
+      "loss": 3.1595,
+      "step": 20425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5156567692756653,
+      "learning_rate": 0.0004497199096086421,
+      "loss": 3.1343,
+      "step": 20426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5068504810333252,
+      "learning_rate": 0.0004497065676307142,
+      "loss": 3.0144,
+      "step": 20427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5446821451187134,
+      "learning_rate": 0.00044969322525848795,
+      "loss": 3.1761,
+      "step": 20428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5488018989562988,
+      "learning_rate": 0.00044967988249199867,
+      "loss": 3.261,
+      "step": 20429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.513907790184021,
+      "learning_rate": 0.0004496665393312813,
+      "loss": 3.1929,
+      "step": 20430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5646592974662781,
+      "learning_rate": 0.0004496531957763711,
+      "loss": 3.0872,
+      "step": 20431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.487352192401886,
+      "learning_rate": 0.0004496398518273031,
+      "loss": 3.044,
+      "step": 20432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48989337682724,
+      "learning_rate": 0.00044962650748411263,
+      "loss": 3.0768,
+      "step": 20433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8140915036201477,
+      "learning_rate": 0.0004496131627468347,
+      "loss": 3.0274,
+      "step": 20434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5698384046554565,
+      "learning_rate": 0.0004495998176155045,
+      "loss": 3.1436,
+      "step": 20435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5074122548103333,
+      "learning_rate": 0.00044958647209015714,
+      "loss": 3.1786,
+      "step": 20436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5354686379432678,
+      "learning_rate": 0.0004495731261708278,
+      "loss": 3.1297,
+      "step": 20437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5391170382499695,
+      "learning_rate": 0.0004495597798575517,
+      "loss": 3.0141,
+      "step": 20438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5479790568351746,
+      "learning_rate": 0.0004495464331503638,
+      "loss": 2.9269,
+      "step": 20439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5683173537254333,
+      "learning_rate": 0.00044953308604929953,
+      "loss": 3.09,
+      "step": 20440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5525175333023071,
+      "learning_rate": 0.0004495197385543938,
+      "loss": 3.1034,
+      "step": 20441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5185175538063049,
+      "learning_rate": 0.0004495063906656818,
+      "loss": 3.2782,
+      "step": 20442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5282908082008362,
+      "learning_rate": 0.0004494930423831988,
+      "loss": 3.0233,
+      "step": 20443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5449439883232117,
+      "learning_rate": 0.00044947969370698,
+      "loss": 3.0911,
+      "step": 20444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5355467796325684,
+      "learning_rate": 0.00044946634463706035,
+      "loss": 3.019,
+      "step": 20445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.514240562915802,
+      "learning_rate": 0.00044945299517347507,
+      "loss": 3.3083,
+      "step": 20446
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5259093642234802,
+      "learning_rate": 0.00044943964531625946,
+      "loss": 3.0219,
+      "step": 20447
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5237060189247131,
+      "learning_rate": 0.00044942629506544853,
+      "loss": 3.2387,
+      "step": 20448
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.518451988697052,
+      "learning_rate": 0.00044941294442107753,
+      "loss": 3.0307,
+      "step": 20449
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.518038272857666,
+      "learning_rate": 0.0004493995933831815,
+      "loss": 3.093,
+      "step": 20450
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5178418159484863,
+      "learning_rate": 0.00044938624195179586,
+      "loss": 3.059,
+      "step": 20451
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5047763586044312,
+      "learning_rate": 0.00044937289012695545,
+      "loss": 3.1335,
+      "step": 20452
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5251728892326355,
+      "learning_rate": 0.0004493595379086956,
+      "loss": 3.0702,
+      "step": 20453
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5640724301338196,
+      "learning_rate": 0.0004493461852970515,
+      "loss": 3.1382,
+      "step": 20454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5112975835800171,
+      "learning_rate": 0.0004493328322920584,
+      "loss": 3.2491,
+      "step": 20455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5223793983459473,
+      "learning_rate": 0.00044931947889375126,
+      "loss": 3.0409,
+      "step": 20456
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5379955172538757,
+      "learning_rate": 0.00044930612510216535,
+      "loss": 3.2273,
+      "step": 20457
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49986040592193604,
+      "learning_rate": 0.0004492927709173359,
+      "loss": 3.0213,
+      "step": 20458
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5016982555389404,
+      "learning_rate": 0.00044927941633929796,
+      "loss": 3.1779,
+      "step": 20459
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5460448265075684,
+      "learning_rate": 0.00044926606136808675,
+      "loss": 3.0865,
+      "step": 20460
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5995565056800842,
+      "learning_rate": 0.00044925270600373754,
+      "loss": 3.0536,
+      "step": 20461
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5643872618675232,
+      "learning_rate": 0.0004492393502462854,
+      "loss": 3.1441,
+      "step": 20462
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5247363448143005,
+      "learning_rate": 0.00044922599409576543,
+      "loss": 3.1199,
+      "step": 20463
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5810442566871643,
+      "learning_rate": 0.000449212637552213,
+      "loss": 3.1335,
+      "step": 20464
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5455927848815918,
+      "learning_rate": 0.0004491992806156632,
+      "loss": 2.9627,
+      "step": 20465
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5521914958953857,
+      "learning_rate": 0.00044918592328615126,
+      "loss": 3.1621,
+      "step": 20466
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5023887753486633,
+      "learning_rate": 0.00044917256556371225,
+      "loss": 2.9914,
+      "step": 20467
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5143914222717285,
+      "learning_rate": 0.00044915920744838136,
+      "loss": 3.181,
+      "step": 20468
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5196571350097656,
+      "learning_rate": 0.000449145848940194,
+      "loss": 3.1984,
+      "step": 20469
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5121498107910156,
+      "learning_rate": 0.00044913249003918505,
+      "loss": 3.0403,
+      "step": 20470
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5418211817741394,
+      "learning_rate": 0.00044911913074538977,
+      "loss": 3.0662,
+      "step": 20471
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5008609890937805,
+      "learning_rate": 0.00044910577105884345,
+      "loss": 3.0714,
+      "step": 20472
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4887140393257141,
+      "learning_rate": 0.00044909241097958126,
+      "loss": 3.1298,
+      "step": 20473
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5122173428535461,
+      "learning_rate": 0.00044907905050763834,
+      "loss": 3.1986,
+      "step": 20474
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5146946907043457,
+      "learning_rate": 0.0004490656896430498,
+      "loss": 3.0118,
+      "step": 20475
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5460460782051086,
+      "learning_rate": 0.00044905232838585103,
+      "loss": 3.177,
+      "step": 20476
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5463166832923889,
+      "learning_rate": 0.0004490389667360771,
+      "loss": 3.2155,
+      "step": 20477
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5182934403419495,
+      "learning_rate": 0.00044902560469376314,
+      "loss": 3.1258,
+      "step": 20478
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5243579149246216,
+      "learning_rate": 0.0004490122422589445,
+      "loss": 3.1593,
+      "step": 20479
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5343524813652039,
+      "learning_rate": 0.00044899887943165634,
+      "loss": 3.1276,
+      "step": 20480
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4999130368232727,
+      "learning_rate": 0.0004489855162119337,
+      "loss": 3.2389,
+      "step": 20481
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5230382680892944,
+      "learning_rate": 0.0004489721525998119,
+      "loss": 3.0518,
+      "step": 20482
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.516069769859314,
+      "learning_rate": 0.0004489587885953261,
+      "loss": 3.1985,
+      "step": 20483
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5170885324478149,
+      "learning_rate": 0.0004489454241985116,
+      "loss": 3.0926,
+      "step": 20484
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5558387041091919,
+      "learning_rate": 0.0004489320594094035,
+      "loss": 3.0827,
+      "step": 20485
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5676001906394958,
+      "learning_rate": 0.000448918694228037,
+      "loss": 3.0072,
+      "step": 20486
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.552216112613678,
+      "learning_rate": 0.00044890532865444723,
+      "loss": 3.1179,
+      "step": 20487
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5196042656898499,
+      "learning_rate": 0.0004488919626886696,
+      "loss": 3.1181,
+      "step": 20488
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5262749791145325,
+      "learning_rate": 0.0004488785963307391,
+      "loss": 3.0952,
+      "step": 20489
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5152573585510254,
+      "learning_rate": 0.0004488652295806911,
+      "loss": 2.9795,
+      "step": 20490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5612766742706299,
+      "learning_rate": 0.0004488518624385608,
+      "loss": 3.1665,
+      "step": 20491
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5625437498092651,
+      "learning_rate": 0.0004488384949043833,
+      "loss": 3.3981,
+      "step": 20492
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5344268679618835,
+      "learning_rate": 0.00044882512697819383,
+      "loss": 3.0059,
+      "step": 20493
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5187119841575623,
+      "learning_rate": 0.0004488117586600275,
+      "loss": 3.1045,
+      "step": 20494
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5119081139564514,
+      "learning_rate": 0.0004487983899499198,
+      "loss": 3.1185,
+      "step": 20495
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5100210905075073,
+      "learning_rate": 0.00044878502084790564,
+      "loss": 3.0568,
+      "step": 20496
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5492945313453674,
+      "learning_rate": 0.0004487716513540205,
+      "loss": 3.2403,
+      "step": 20497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5489396452903748,
+      "learning_rate": 0.0004487582814682994,
+      "loss": 2.8496,
+      "step": 20498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5205833315849304,
+      "learning_rate": 0.0004487449111907776,
+      "loss": 2.8851,
+      "step": 20499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5621615648269653,
+      "learning_rate": 0.0004487315405214903,
+      "loss": 2.9772,
+      "step": 20500
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5375982522964478,
+      "learning_rate": 0.00044871816946047286,
+      "loss": 3.1913,
+      "step": 20501
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5279413461685181,
+      "learning_rate": 0.0004487047980077604,
+      "loss": 3.0989,
+      "step": 20502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5290868282318115,
+      "learning_rate": 0.00044869142616338803,
+      "loss": 3.0419,
+      "step": 20503
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5295721292495728,
+      "learning_rate": 0.00044867805392739097,
+      "loss": 3.1698,
+      "step": 20504
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.531470000743866,
+      "learning_rate": 0.00044866468129980464,
+      "loss": 2.8806,
+      "step": 20505
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5461409091949463,
+      "learning_rate": 0.0004486513082806642,
+      "loss": 3.332,
+      "step": 20506
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5531258583068848,
+      "learning_rate": 0.00044863793487000475,
+      "loss": 3.2661,
+      "step": 20507
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5420560240745544,
+      "learning_rate": 0.00044862456106786166,
+      "loss": 3.0748,
+      "step": 20508
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5236428380012512,
+      "learning_rate": 0.00044861118687427,
+      "loss": 3.0693,
+      "step": 20509
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5320883393287659,
+      "learning_rate": 0.00044859781228926505,
+      "loss": 3.2626,
+      "step": 20510
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5048290491104126,
+      "learning_rate": 0.0004485844373128821,
+      "loss": 3.0668,
+      "step": 20511
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5460326075553894,
+      "learning_rate": 0.00044857106194515635,
+      "loss": 3.1908,
+      "step": 20512
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4952123463153839,
+      "learning_rate": 0.0004485576861861231,
+      "loss": 3.0871,
+      "step": 20513
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5370456576347351,
+      "learning_rate": 0.0004485443100358173,
+      "loss": 3.0149,
+      "step": 20514
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5347251892089844,
+      "learning_rate": 0.0004485309334942745,
+      "loss": 3.1556,
+      "step": 20515
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5180676579475403,
+      "learning_rate": 0.0004485175565615298,
+      "loss": 3.0769,
+      "step": 20516
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5209394097328186,
+      "learning_rate": 0.0004485041792376184,
+      "loss": 3.2223,
+      "step": 20517
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5076389908790588,
+      "learning_rate": 0.00044849080152257564,
+      "loss": 3.1164,
+      "step": 20518
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5367461442947388,
+      "learning_rate": 0.00044847742341643654,
+      "loss": 3.1111,
+      "step": 20519
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5419523119926453,
+      "learning_rate": 0.0004484640449192367,
+      "loss": 2.9594,
+      "step": 20520
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5312870144844055,
+      "learning_rate": 0.00044845066603101103,
+      "loss": 3.0924,
+      "step": 20521
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5425320863723755,
+      "learning_rate": 0.0004484372867517948,
+      "loss": 3.1313,
+      "step": 20522
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5226745009422302,
+      "learning_rate": 0.00044842390708162345,
+      "loss": 3.1729,
+      "step": 20523
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5791988968849182,
+      "learning_rate": 0.000448410527020532,
+      "loss": 2.9848,
+      "step": 20524
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5693748593330383,
+      "learning_rate": 0.0004483971465685558,
+      "loss": 3.1733,
+      "step": 20525
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5362666249275208,
+      "learning_rate": 0.0004483837657257301,
+      "loss": 3.0081,
+      "step": 20526
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5347479581832886,
+      "learning_rate": 0.0004483703844920901,
+      "loss": 3.0189,
+      "step": 20527
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.529528796672821,
+      "learning_rate": 0.00044835700286767114,
+      "loss": 2.9859,
+      "step": 20528
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5108277797698975,
+      "learning_rate": 0.0004483436208525083,
+      "loss": 3.1131,
+      "step": 20529
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5495550632476807,
+      "learning_rate": 0.00044833023844663693,
+      "loss": 3.2628,
+      "step": 20530
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5666603446006775,
+      "learning_rate": 0.0004483168556500922,
+      "loss": 3.1046,
+      "step": 20531
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5286827683448792,
+      "learning_rate": 0.00044830347246290956,
+      "loss": 2.8698,
+      "step": 20532
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5373377203941345,
+      "learning_rate": 0.000448290088885124,
+      "loss": 2.9937,
+      "step": 20533
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5299882292747498,
+      "learning_rate": 0.00044827670491677095,
+      "loss": 3.0146,
+      "step": 20534
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5432134866714478,
+      "learning_rate": 0.00044826332055788553,
+      "loss": 3.3426,
+      "step": 20535
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5193071961402893,
+      "learning_rate": 0.00044824993580850313,
+      "loss": 2.8983,
+      "step": 20536
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5308823585510254,
+      "learning_rate": 0.00044823655066865886,
+      "loss": 3.3115,
+      "step": 20537
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5253787040710449,
+      "learning_rate": 0.0004482231651383881,
+      "loss": 3.2114,
+      "step": 20538
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.583215594291687,
+      "learning_rate": 0.000448209779217726,
+      "loss": 3.1361,
+      "step": 20539
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5469831824302673,
+      "learning_rate": 0.0004481963929067078,
+      "loss": 3.1252,
+      "step": 20540
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5351081490516663,
+      "learning_rate": 0.000448183006205369,
+      "loss": 3.0161,
+      "step": 20541
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4953567385673523,
+      "learning_rate": 0.00044816961911374464,
+      "loss": 3.3067,
+      "step": 20542
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5496410727500916,
+      "learning_rate": 0.00044815623163186994,
+      "loss": 3.1639,
+      "step": 20543
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5172861218452454,
+      "learning_rate": 0.0004481428437597803,
+      "loss": 2.8968,
+      "step": 20544
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5500335097312927,
+      "learning_rate": 0.0004481294554975108,
+      "loss": 3.1042,
+      "step": 20545
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5287953019142151,
+      "learning_rate": 0.000448116066845097,
+      "loss": 3.0327,
+      "step": 20546
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5284257531166077,
+      "learning_rate": 0.00044810267780257386,
+      "loss": 3.2541,
+      "step": 20547
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5908827781677246,
+      "learning_rate": 0.0004480892883699768,
+      "loss": 3.0347,
+      "step": 20548
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5372423529624939,
+      "learning_rate": 0.00044807589854734106,
+      "loss": 2.9296,
+      "step": 20549
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5452413558959961,
+      "learning_rate": 0.0004480625083347019,
+      "loss": 2.8282,
+      "step": 20550
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5139778256416321,
+      "learning_rate": 0.0004480491177320946,
+      "loss": 2.9481,
+      "step": 20551
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5498486757278442,
+      "learning_rate": 0.0004480357267395544,
+      "loss": 2.954,
+      "step": 20552
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.550351083278656,
+      "learning_rate": 0.00044802233535711666,
+      "loss": 3.1411,
+      "step": 20553
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5372947454452515,
+      "learning_rate": 0.0004480089435848165,
+      "loss": 3.0283,
+      "step": 20554
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5474216341972351,
+      "learning_rate": 0.0004479955514226892,
+      "loss": 3.0082,
+      "step": 20555
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.578901469707489,
+      "learning_rate": 0.0004479821588707702,
+      "loss": 3.0129,
+      "step": 20556
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5624405145645142,
+      "learning_rate": 0.0004479687659290947,
+      "loss": 2.9875,
+      "step": 20557
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5033628940582275,
+      "learning_rate": 0.0004479553725976979,
+      "loss": 3.0755,
+      "step": 20558
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5456181764602661,
+      "learning_rate": 0.0004479419788766151,
+      "loss": 2.937,
+      "step": 20559
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.595481276512146,
+      "learning_rate": 0.0004479285847658816,
+      "loss": 3.1074,
+      "step": 20560
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5183221101760864,
+      "learning_rate": 0.00044791519026553267,
+      "loss": 2.8591,
+      "step": 20561
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.537140965461731,
+      "learning_rate": 0.0004479017953756037,
+      "loss": 3.0136,
+      "step": 20562
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5604956746101379,
+      "learning_rate": 0.00044788840009612975,
+      "loss": 3.0379,
+      "step": 20563
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5302656292915344,
+      "learning_rate": 0.0004478750044271463,
+      "loss": 3.1423,
+      "step": 20564
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5702511072158813,
+      "learning_rate": 0.0004478616083686884,
+      "loss": 3.0888,
+      "step": 20565
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5529466271400452,
+      "learning_rate": 0.0004478482119207916,
+      "loss": 3.0283,
+      "step": 20566
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.49684593081474304,
+      "learning_rate": 0.0004478348150834911,
+      "loss": 3.3656,
+      "step": 20567
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5289061665534973,
+      "learning_rate": 0.00044782141785682216,
+      "loss": 3.0752,
+      "step": 20568
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.527275025844574,
+      "learning_rate": 0.00044780802024081993,
+      "loss": 3.1675,
+      "step": 20569
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5364570617675781,
+      "learning_rate": 0.00044779462223551995,
+      "loss": 2.9237,
+      "step": 20570
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5187329053878784,
+      "learning_rate": 0.00044778122384095723,
+      "loss": 2.9779,
+      "step": 20571
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5524115562438965,
+      "learning_rate": 0.0004477678250571673,
+      "loss": 3.1814,
+      "step": 20572
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5237755179405212,
+      "learning_rate": 0.00044775442588418536,
+      "loss": 3.1823,
+      "step": 20573
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5263662338256836,
+      "learning_rate": 0.0004477410263220467,
+      "loss": 3.1545,
+      "step": 20574
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5157889127731323,
+      "learning_rate": 0.00044772762637078665,
+      "loss": 2.9089,
+      "step": 20575
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5321006178855896,
+      "learning_rate": 0.0004477142260304403,
+      "loss": 3.334,
+      "step": 20576
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5617817044258118,
+      "learning_rate": 0.0004477008253010432,
+      "loss": 2.9518,
+      "step": 20577
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5136963725090027,
+      "learning_rate": 0.00044768742418263053,
+      "loss": 2.9438,
+      "step": 20578
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5390790104866028,
+      "learning_rate": 0.00044767402267523773,
+      "loss": 3.1914,
+      "step": 20579
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5621577501296997,
+      "learning_rate": 0.0004476606207788999,
+      "loss": 3.0937,
+      "step": 20580
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5234653949737549,
+      "learning_rate": 0.0004476472184936523,
+      "loss": 3.087,
+      "step": 20581
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5514974594116211,
+      "learning_rate": 0.0004476338158195305,
+      "loss": 2.9886,
+      "step": 20582
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5614942908287048,
+      "learning_rate": 0.0004476204127565696,
+      "loss": 3.0167,
+      "step": 20583
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5162169933319092,
+      "learning_rate": 0.0004476070093048049,
+      "loss": 3.1614,
+      "step": 20584
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5244741439819336,
+      "learning_rate": 0.00044759360546427175,
+      "loss": 3.1057,
+      "step": 20585
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5119442343711853,
+      "learning_rate": 0.0004475802012350055,
+      "loss": 2.9263,
+      "step": 20586
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48682332038879395,
+      "learning_rate": 0.0004475667966170414,
+      "loss": 3.0508,
+      "step": 20587
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5387108325958252,
+      "learning_rate": 0.00044755339161041467,
+      "loss": 2.8952,
+      "step": 20588
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5450228452682495,
+      "learning_rate": 0.00044753998621516075,
+      "loss": 3.332,
+      "step": 20589
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5503659844398499,
+      "learning_rate": 0.000447526580431315,
+      "loss": 2.9963,
+      "step": 20590
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5000738501548767,
+      "learning_rate": 0.0004475131742589125,
+      "loss": 2.9757,
+      "step": 20591
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5334386229515076,
+      "learning_rate": 0.00044749976769798875,
+      "loss": 3.3276,
+      "step": 20592
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5666833519935608,
+      "learning_rate": 0.00044748636074857904,
+      "loss": 3.1172,
+      "step": 20593
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5799906253814697,
+      "learning_rate": 0.00044747295341071857,
+      "loss": 3.0151,
+      "step": 20594
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5235270857810974,
+      "learning_rate": 0.00044745954568444266,
+      "loss": 2.9707,
+      "step": 20595
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5744423866271973,
+      "learning_rate": 0.0004474461375697867,
+      "loss": 2.9403,
+      "step": 20596
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.571638286113739,
+      "learning_rate": 0.00044743272906678616,
+      "loss": 3.1176,
+      "step": 20597
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5430690050125122,
+      "learning_rate": 0.00044741932017547604,
+      "loss": 3.1179,
+      "step": 20598
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5648912191390991,
+      "learning_rate": 0.0004474059108958918,
+      "loss": 3.1133,
+      "step": 20599
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5373355746269226,
+      "learning_rate": 0.00044739250122806883,
+      "loss": 3.0403,
+      "step": 20600
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5279290676116943,
+      "learning_rate": 0.0004473790911720423,
+      "loss": 3.1451,
+      "step": 20601
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.565265953540802,
+      "learning_rate": 0.0004473656807278477,
+      "loss": 3.1434,
+      "step": 20602
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5297691226005554,
+      "learning_rate": 0.00044735226989552014,
+      "loss": 3.2683,
+      "step": 20603
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5779011249542236,
+      "learning_rate": 0.0004473388586750952,
+      "loss": 3.1056,
+      "step": 20604
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5088967084884644,
+      "learning_rate": 0.000447325447066608,
+      "loss": 3.2698,
+      "step": 20605
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5209295153617859,
+      "learning_rate": 0.00044731203507009386,
+      "loss": 3.1012,
+      "step": 20606
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.517296552658081,
+      "learning_rate": 0.0004472986226855882,
+      "loss": 3.1231,
+      "step": 20607
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5545491576194763,
+      "learning_rate": 0.0004472852099131264,
+      "loss": 3.1827,
+      "step": 20608
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5324487686157227,
+      "learning_rate": 0.00044727179675274365,
+      "loss": 2.9018,
+      "step": 20609
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5557654500007629,
+      "learning_rate": 0.00044725838320447533,
+      "loss": 2.9244,
+      "step": 20610
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5294561982154846,
+      "learning_rate": 0.00044724496926835673,
+      "loss": 3.2031,
+      "step": 20611
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4875947833061218,
+      "learning_rate": 0.0004472315549444233,
+      "loss": 2.9372,
+      "step": 20612
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5307518243789673,
+      "learning_rate": 0.00044721814023271025,
+      "loss": 3.1983,
+      "step": 20613
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5280675888061523,
+      "learning_rate": 0.00044720472513325296,
+      "loss": 3.1898,
+      "step": 20614
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5537136793136597,
+      "learning_rate": 0.0004471913096460867,
+      "loss": 3.2588,
+      "step": 20615
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5422837138175964,
+      "learning_rate": 0.00044717789377124695,
+      "loss": 3.0336,
+      "step": 20616
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5528018474578857,
+      "learning_rate": 0.0004471644775087688,
+      "loss": 3.0397,
+      "step": 20617
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5226984620094299,
+      "learning_rate": 0.00044715106085868784,
+      "loss": 3.2272,
+      "step": 20618
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5283844470977783,
+      "learning_rate": 0.0004471376438210394,
+      "loss": 3.2111,
+      "step": 20619
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5217439532279968,
+      "learning_rate": 0.00044712422639585863,
+      "loss": 3.0164,
+      "step": 20620
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5724291205406189,
+      "learning_rate": 0.0004471108085831809,
+      "loss": 3.1398,
+      "step": 20621
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5415415167808533,
+      "learning_rate": 0.0004470973903830417,
+      "loss": 3.0994,
+      "step": 20622
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5410470366477966,
+      "learning_rate": 0.00044708397179547626,
+      "loss": 3.1288,
+      "step": 20623
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5544204711914062,
+      "learning_rate": 0.0004470705528205199,
+      "loss": 3.2814,
+      "step": 20624
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5386510491371155,
+      "learning_rate": 0.0004470571334582081,
+      "loss": 3.041,
+      "step": 20625
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5222130417823792,
+      "learning_rate": 0.00044704371370857607,
+      "loss": 3.1223,
+      "step": 20626
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5678313374519348,
+      "learning_rate": 0.0004470302935716591,
+      "loss": 2.9597,
+      "step": 20627
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5152552127838135,
+      "learning_rate": 0.00044701687304749276,
+      "loss": 2.9862,
+      "step": 20628
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5379000902175903,
+      "learning_rate": 0.0004470034521361122,
+      "loss": 3.1142,
+      "step": 20629
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5807501673698425,
+      "learning_rate": 0.0004469900308375529,
+      "loss": 3.2559,
+      "step": 20630
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5624402165412903,
+      "learning_rate": 0.0004469766091518502,
+      "loss": 3.1534,
+      "step": 20631
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5430057048797607,
+      "learning_rate": 0.0004469631870790392,
+      "loss": 3.1118,
+      "step": 20632
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5057716369628906,
+      "learning_rate": 0.0004469497646191556,
+      "loss": 2.9988,
+      "step": 20633
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5503453612327576,
+      "learning_rate": 0.0004469363417722346,
+      "loss": 3.0555,
+      "step": 20634
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5180730223655701,
+      "learning_rate": 0.0004469229185383115,
+      "loss": 3.1024,
+      "step": 20635
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5402533411979675,
+      "learning_rate": 0.0004469094949174217,
+      "loss": 3.1781,
+      "step": 20636
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6356052160263062,
+      "learning_rate": 0.0004468960709096006,
+      "loss": 3.0357,
+      "step": 20637
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5212626457214355,
+      "learning_rate": 0.0004468826465148835,
+      "loss": 3.0721,
+      "step": 20638
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5012721419334412,
+      "learning_rate": 0.00044686922173330584,
+      "loss": 3.0777,
+      "step": 20639
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5270209908485413,
+      "learning_rate": 0.00044685579656490287,
+      "loss": 3.2153,
+      "step": 20640
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5531798005104065,
+      "learning_rate": 0.00044684237100971,
+      "loss": 3.1648,
+      "step": 20641
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5050953030586243,
+      "learning_rate": 0.00044682894506776246,
+      "loss": 3.2221,
+      "step": 20642
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5169261693954468,
+      "learning_rate": 0.0004468155187390959,
+      "loss": 3.1671,
+      "step": 20643
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5264015197753906,
+      "learning_rate": 0.0004468020920237455,
+      "loss": 3.0668,
+      "step": 20644
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5398049354553223,
+      "learning_rate": 0.0004467886649217466,
+      "loss": 3.1813,
+      "step": 20645
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5570565462112427,
+      "learning_rate": 0.0004467752374331346,
+      "loss": 2.8079,
+      "step": 20646
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5235486030578613,
+      "learning_rate": 0.0004467618095579449,
+      "loss": 3.1429,
+      "step": 20647
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5539608001708984,
+      "learning_rate": 0.0004467483812962128,
+      "loss": 3.1071,
+      "step": 20648
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5413884520530701,
+      "learning_rate": 0.0004467349526479738,
+      "loss": 2.7632,
+      "step": 20649
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5679776072502136,
+      "learning_rate": 0.00044672152361326307,
+      "loss": 3.0299,
+      "step": 20650
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5429620742797852,
+      "learning_rate": 0.0004467080941921161,
+      "loss": 3.1816,
+      "step": 20651
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5164234042167664,
+      "learning_rate": 0.00044669466438456833,
+      "loss": 3.2438,
+      "step": 20652
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5403413772583008,
+      "learning_rate": 0.0004466812341906549,
+      "loss": 3.1561,
+      "step": 20653
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5405500531196594,
+      "learning_rate": 0.00044666780361041143,
+      "loss": 3.229,
+      "step": 20654
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5156397819519043,
+      "learning_rate": 0.0004466543726438733,
+      "loss": 3.0178,
+      "step": 20655
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5206366777420044,
+      "learning_rate": 0.00044664094129107557,
+      "loss": 2.7869,
+      "step": 20656
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5250067710876465,
+      "learning_rate": 0.00044662750955205393,
+      "loss": 3.0707,
+      "step": 20657
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5365163087844849,
+      "learning_rate": 0.00044661407742684355,
+      "loss": 2.8563,
+      "step": 20658
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5403412580490112,
+      "learning_rate": 0.00044660064491548003,
+      "loss": 3.037,
+      "step": 20659
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5710760354995728,
+      "learning_rate": 0.00044658721201799856,
+      "loss": 2.9855,
+      "step": 20660
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5208877921104431,
+      "learning_rate": 0.00044657377873443454,
+      "loss": 3.0927,
+      "step": 20661
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5228715538978577,
+      "learning_rate": 0.00044656034506482354,
+      "loss": 3.0964,
+      "step": 20662
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5221062302589417,
+      "learning_rate": 0.00044654691100920067,
+      "loss": 3.1844,
+      "step": 20663
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5395482778549194,
+      "learning_rate": 0.00044653347656760145,
+      "loss": 3.2539,
+      "step": 20664
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5604360103607178,
+      "learning_rate": 0.00044652004174006133,
+      "loss": 3.3412,
+      "step": 20665
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5532729029655457,
+      "learning_rate": 0.0004465066065266156,
+      "loss": 3.1004,
+      "step": 20666
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5858287811279297,
+      "learning_rate": 0.0004464931709272996,
+      "loss": 3.0845,
+      "step": 20667
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.506764829158783,
+      "learning_rate": 0.0004464797349421488,
+      "loss": 3.3258,
+      "step": 20668
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.974445641040802,
+      "learning_rate": 0.00044646629857119854,
+      "loss": 2.916,
+      "step": 20669
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5399503111839294,
+      "learning_rate": 0.0004464528618144843,
+      "loss": 2.9863,
+      "step": 20670
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5604903101921082,
+      "learning_rate": 0.0004464394246720415,
+      "loss": 3.0492,
+      "step": 20671
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5462265014648438,
+      "learning_rate": 0.00044642598714390527,
+      "loss": 3.0305,
+      "step": 20672
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.512052595615387,
+      "learning_rate": 0.00044641254923011124,
+      "loss": 3.0729,
+      "step": 20673
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.521513044834137,
+      "learning_rate": 0.0004463991109306947,
+      "loss": 3.0554,
+      "step": 20674
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5272539258003235,
+      "learning_rate": 0.0004463856722456911,
+      "loss": 3.171,
+      "step": 20675
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5902304649353027,
+      "learning_rate": 0.00044637223317513583,
+      "loss": 3.1536,
+      "step": 20676
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5469979643821716,
+      "learning_rate": 0.00044635879371906427,
+      "loss": 2.9833,
+      "step": 20677
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5268833041191101,
+      "learning_rate": 0.0004463453538775118,
+      "loss": 3.2544,
+      "step": 20678
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5354952812194824,
+      "learning_rate": 0.0004463319136505138,
+      "loss": 2.8549,
+      "step": 20679
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6132659316062927,
+      "learning_rate": 0.0004463184730381057,
+      "loss": 2.9428,
+      "step": 20680
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5660962462425232,
+      "learning_rate": 0.000446305032040323,
+      "loss": 2.9525,
+      "step": 20681
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5585618019104004,
+      "learning_rate": 0.0004462915906572009,
+      "loss": 3.1878,
+      "step": 20682
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6061045527458191,
+      "learning_rate": 0.0004462781488887749,
+      "loss": 3.2639,
+      "step": 20683
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.550226628780365,
+      "learning_rate": 0.00044626470673508043,
+      "loss": 3.0404,
+      "step": 20684
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5399933457374573,
+      "learning_rate": 0.00044625126419615296,
+      "loss": 3.0385,
+      "step": 20685
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5338366031646729,
+      "learning_rate": 0.0004462378212720277,
+      "loss": 3.1391,
+      "step": 20686
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5670849084854126,
+      "learning_rate": 0.00044622437796274016,
+      "loss": 3.1828,
+      "step": 20687
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5496315360069275,
+      "learning_rate": 0.0004462109342683259,
+      "loss": 2.9966,
+      "step": 20688
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5669832229614258,
+      "learning_rate": 0.00044619749018881994,
+      "loss": 2.9435,
+      "step": 20689
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5618574023246765,
+      "learning_rate": 0.0004461840457242581,
+      "loss": 3.1487,
+      "step": 20690
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5448206067085266,
+      "learning_rate": 0.00044617060087467556,
+      "loss": 3.1906,
+      "step": 20691
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5339007377624512,
+      "learning_rate": 0.0004461571556401078,
+      "loss": 3.164,
+      "step": 20692
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5732952952384949,
+      "learning_rate": 0.00044614371002059026,
+      "loss": 3.197,
+      "step": 20693
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5226903557777405,
+      "learning_rate": 0.0004461302640161582,
+      "loss": 3.1404,
+      "step": 20694
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.547990620136261,
+      "learning_rate": 0.00044611681762684723,
+      "loss": 3.2876,
+      "step": 20695
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5328800082206726,
+      "learning_rate": 0.0004461033708526927,
+      "loss": 3.0255,
+      "step": 20696
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5429891347885132,
+      "learning_rate": 0.00044608992369372995,
+      "loss": 3.4003,
+      "step": 20697
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.561837375164032,
+      "learning_rate": 0.00044607647614999454,
+      "loss": 3.0505,
+      "step": 20698
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5248981714248657,
+      "learning_rate": 0.00044606302822152176,
+      "loss": 3.3295,
+      "step": 20699
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5702486038208008,
+      "learning_rate": 0.000446049579908347,
+      "loss": 3.072,
+      "step": 20700
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5280510187149048,
+      "learning_rate": 0.0004460361312105058,
+      "loss": 3.1817,
+      "step": 20701
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5336136221885681,
+      "learning_rate": 0.0004460226821280336,
+      "loss": 3.1157,
+      "step": 20702
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.523258626461029,
+      "learning_rate": 0.0004460092326609658,
+      "loss": 3.1815,
+      "step": 20703
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5246843099594116,
+      "learning_rate": 0.00044599578280933756,
+      "loss": 3.0741,
+      "step": 20704
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5236319303512573,
+      "learning_rate": 0.00044598233257318474,
+      "loss": 2.9568,
+      "step": 20705
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5435445308685303,
+      "learning_rate": 0.0004459688819525425,
+      "loss": 2.9715,
+      "step": 20706
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5246100425720215,
+      "learning_rate": 0.00044595543094744623,
+      "loss": 3.0166,
+      "step": 20707
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5761013627052307,
+      "learning_rate": 0.00044594197955793156,
+      "loss": 3.0693,
+      "step": 20708
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5138135552406311,
+      "learning_rate": 0.00044592852778403366,
+      "loss": 2.9992,
+      "step": 20709
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5575985312461853,
+      "learning_rate": 0.00044591507562578825,
+      "loss": 2.9294,
+      "step": 20710
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5381579995155334,
+      "learning_rate": 0.0004459016230832306,
+      "loss": 3.0036,
+      "step": 20711
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5545154213905334,
+      "learning_rate": 0.00044588817015639605,
+      "loss": 3.0897,
+      "step": 20712
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.507051408290863,
+      "learning_rate": 0.00044587471684532016,
+      "loss": 3.0543,
+      "step": 20713
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5791917443275452,
+      "learning_rate": 0.00044586126315003836,
+      "loss": 3.2333,
+      "step": 20714
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5103766322135925,
+      "learning_rate": 0.0004458478090705861,
+      "loss": 3.193,
+      "step": 20715
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5501724481582642,
+      "learning_rate": 0.00044583435460699875,
+      "loss": 3.0267,
+      "step": 20716
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5263850092887878,
+      "learning_rate": 0.00044582089975931185,
+      "loss": 3.2196,
+      "step": 20717
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5362185835838318,
+      "learning_rate": 0.0004458074445275607,
+      "loss": 2.9511,
+      "step": 20718
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.535193920135498,
+      "learning_rate": 0.0004457939889117807,
+      "loss": 3.2362,
+      "step": 20719
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5330086350440979,
+      "learning_rate": 0.0004457805329120075,
+      "loss": 2.9531,
+      "step": 20720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5201607942581177,
+      "learning_rate": 0.00044576707652827646,
+      "loss": 3.2588,
+      "step": 20721
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5220875144004822,
+      "learning_rate": 0.000445753619760623,
+      "loss": 3.0818,
+      "step": 20722
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5611007213592529,
+      "learning_rate": 0.0004457401626090825,
+      "loss": 3.1012,
+      "step": 20723
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5594449639320374,
+      "learning_rate": 0.0004457267050736904,
+      "loss": 3.3061,
+      "step": 20724
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5162807106971741,
+      "learning_rate": 0.00044571324715448235,
+      "loss": 3.0391,
+      "step": 20725
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5748878717422485,
+      "learning_rate": 0.00044569978885149354,
+      "loss": 3.1104,
+      "step": 20726
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5533029437065125,
+      "learning_rate": 0.0004456863301647596,
+      "loss": 3.0481,
+      "step": 20727
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5461177229881287,
+      "learning_rate": 0.00044567287109431586,
+      "loss": 3.0085,
+      "step": 20728
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5752653479576111,
+      "learning_rate": 0.00044565941164019784,
+      "loss": 3.1422,
+      "step": 20729
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5168489813804626,
+      "learning_rate": 0.00044564595180244095,
+      "loss": 3.0288,
+      "step": 20730
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5368278622627258,
+      "learning_rate": 0.00044563249158108064,
+      "loss": 3.215,
+      "step": 20731
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5282604098320007,
+      "learning_rate": 0.00044561903097615243,
+      "loss": 3.0692,
+      "step": 20732
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.568565309047699,
+      "learning_rate": 0.00044560556998769166,
+      "loss": 3.1969,
+      "step": 20733
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5004165172576904,
+      "learning_rate": 0.0004455921086157338,
+      "loss": 2.8507,
+      "step": 20734
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5740750432014465,
+      "learning_rate": 0.0004455786468603144,
+      "loss": 3.0852,
+      "step": 20735
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5363320112228394,
+      "learning_rate": 0.0004455651847214689,
+      "loss": 2.942,
+      "step": 20736
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5179160237312317,
+      "learning_rate": 0.00044555172219923263,
+      "loss": 3.0065,
+      "step": 20737
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5659471154212952,
+      "learning_rate": 0.0004455382592936412,
+      "loss": 3.1902,
+      "step": 20738
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.514021635055542,
+      "learning_rate": 0.00044552479600473,
+      "loss": 2.9964,
+      "step": 20739
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5664601922035217,
+      "learning_rate": 0.00044551133233253443,
+      "loss": 2.9117,
+      "step": 20740
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5283511877059937,
+      "learning_rate": 0.0004454978682770901,
+      "loss": 3.3048,
+      "step": 20741
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5478894710540771,
+      "learning_rate": 0.0004454844038384323,
+      "loss": 3.247,
+      "step": 20742
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5319153666496277,
+      "learning_rate": 0.0004454709390165967,
+      "loss": 3.2392,
+      "step": 20743
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5409319400787354,
+      "learning_rate": 0.0004454574738116186,
+      "loss": 3.0977,
+      "step": 20744
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5745616555213928,
+      "learning_rate": 0.0004454440082235334,
+      "loss": 3.072,
+      "step": 20745
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6125838756561279,
+      "learning_rate": 0.0004454305422523768,
+      "loss": 3.2891,
+      "step": 20746
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5295475721359253,
+      "learning_rate": 0.00044541707589818404,
+      "loss": 3.3818,
+      "step": 20747
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.588213324546814,
+      "learning_rate": 0.0004454036091609908,
+      "loss": 3.2082,
+      "step": 20748
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5516581535339355,
+      "learning_rate": 0.0004453901420408324,
+      "loss": 3.3228,
+      "step": 20749
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5377952456474304,
+      "learning_rate": 0.0004453766745377442,
+      "loss": 3.0391,
+      "step": 20750
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5352370738983154,
+      "learning_rate": 0.000445363206651762,
+      "loss": 3.2461,
+      "step": 20751
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5557248592376709,
+      "learning_rate": 0.0004453497383829211,
+      "loss": 2.8931,
+      "step": 20752
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5509526133537292,
+      "learning_rate": 0.00044533626973125687,
+      "loss": 3.1886,
+      "step": 20753
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5012487173080444,
+      "learning_rate": 0.0004453228006968049,
+      "loss": 2.9703,
+      "step": 20754
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5457666516304016,
+      "learning_rate": 0.0004453093312796006,
+      "loss": 2.8753,
+      "step": 20755
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5392427444458008,
+      "learning_rate": 0.0004452958614796795,
+      "loss": 3.1255,
+      "step": 20756
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5236079096794128,
+      "learning_rate": 0.0004452823912970772,
+      "loss": 2.8528,
+      "step": 20757
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5426657795906067,
+      "learning_rate": 0.0004452689207318289,
+      "loss": 3.0241,
+      "step": 20758
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5343982577323914,
+      "learning_rate": 0.00044525544978397025,
+      "loss": 3.1399,
+      "step": 20759
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.535350501537323,
+      "learning_rate": 0.0004452419784535367,
+      "loss": 3.1378,
+      "step": 20760
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5450648665428162,
+      "learning_rate": 0.0004452285067405638,
+      "loss": 3.1517,
+      "step": 20761
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5389657020568848,
+      "learning_rate": 0.0004452150346450869,
+      "loss": 3.3121,
+      "step": 20762
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5154426693916321,
+      "learning_rate": 0.00044520156216714145,
+      "loss": 3.1708,
+      "step": 20763
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5560336112976074,
+      "learning_rate": 0.0004451880893067632,
+      "loss": 3.0882,
+      "step": 20764
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5327739119529724,
+      "learning_rate": 0.0004451746160639874,
+      "loss": 3.1679,
+      "step": 20765
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5418607592582703,
+      "learning_rate": 0.0004451611424388496,
+      "loss": 3.1208,
+      "step": 20766
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.57447749376297,
+      "learning_rate": 0.0004451476684313852,
+      "loss": 2.9681,
+      "step": 20767
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5182052850723267,
+      "learning_rate": 0.0004451341940416299,
+      "loss": 3.2641,
+      "step": 20768
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5607951283454895,
+      "learning_rate": 0.00044512071926961904,
+      "loss": 3.117,
+      "step": 20769
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5186867713928223,
+      "learning_rate": 0.0004451072441153881,
+      "loss": 3.0579,
+      "step": 20770
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5786254405975342,
+      "learning_rate": 0.0004450937685789725,
+      "loss": 3.1559,
+      "step": 20771
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5199046730995178,
+      "learning_rate": 0.0004450802926604081,
+      "loss": 3.0105,
+      "step": 20772
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5281519293785095,
+      "learning_rate": 0.00044506681635972996,
+      "loss": 3.0246,
+      "step": 20773
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4995059072971344,
+      "learning_rate": 0.0004450533396769737,
+      "loss": 3.2023,
+      "step": 20774
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.526586651802063,
+      "learning_rate": 0.000445039862612175,
+      "loss": 2.8645,
+      "step": 20775
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5115561485290527,
+      "learning_rate": 0.00044502638516536906,
+      "loss": 2.8865,
+      "step": 20776
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5567725896835327,
+      "learning_rate": 0.00044501290733659166,
+      "loss": 3.3244,
+      "step": 20777
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5572482943534851,
+      "learning_rate": 0.00044499942912587813,
+      "loss": 3.0157,
+      "step": 20778
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5388816595077515,
+      "learning_rate": 0.00044498595053326403,
+      "loss": 2.9588,
+      "step": 20779
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5826587677001953,
+      "learning_rate": 0.0004449724715587849,
+      "loss": 2.875,
+      "step": 20780
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5563898682594299,
+      "learning_rate": 0.000444958992202476,
+      "loss": 2.9278,
+      "step": 20781
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5438665747642517,
+      "learning_rate": 0.0004449455124643731,
+      "loss": 3.1942,
+      "step": 20782
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5232053995132446,
+      "learning_rate": 0.00044493203234451166,
+      "loss": 3.0504,
+      "step": 20783
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5295712947845459,
+      "learning_rate": 0.00044491855184292713,
+      "loss": 3.0075,
+      "step": 20784
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5307983756065369,
+      "learning_rate": 0.000444905070959655,
+      "loss": 3.025,
+      "step": 20785
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.619365394115448,
+      "learning_rate": 0.00044489158969473076,
+      "loss": 2.8898,
+      "step": 20786
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5883886218070984,
+      "learning_rate": 0.00044487810804819,
+      "loss": 2.9573,
+      "step": 20787
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5858182907104492,
+      "learning_rate": 0.0004448646260200682,
+      "loss": 3.2317,
+      "step": 20788
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5257620215415955,
+      "learning_rate": 0.0004448511436104008,
+      "loss": 3.1254,
+      "step": 20789
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5104060173034668,
+      "learning_rate": 0.00044483766081922346,
+      "loss": 3.1547,
+      "step": 20790
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5185402035713196,
+      "learning_rate": 0.00044482417764657147,
+      "loss": 3.099,
+      "step": 20791
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5819932818412781,
+      "learning_rate": 0.00044481069409248056,
+      "loss": 3.0686,
+      "step": 20792
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5629081726074219,
+      "learning_rate": 0.00044479721015698614,
+      "loss": 3.1719,
+      "step": 20793
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5439802408218384,
+      "learning_rate": 0.00044478372584012376,
+      "loss": 3.1272,
+      "step": 20794
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5107980370521545,
+      "learning_rate": 0.0004447702411419289,
+      "loss": 3.1131,
+      "step": 20795
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5272974967956543,
+      "learning_rate": 0.000444756756062437,
+      "loss": 3.1204,
+      "step": 20796
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5283223986625671,
+      "learning_rate": 0.00044474327060168374,
+      "loss": 3.1007,
+      "step": 20797
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5268027186393738,
+      "learning_rate": 0.00044472978475970453,
+      "loss": 3.1367,
+      "step": 20798
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5282880663871765,
+      "learning_rate": 0.00044471629853653496,
+      "loss": 3.08,
+      "step": 20799
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6580334901809692,
+      "learning_rate": 0.0004447028119322105,
+      "loss": 3.0333,
+      "step": 20800
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5694079995155334,
+      "learning_rate": 0.0004446893249467666,
+      "loss": 3.1917,
+      "step": 20801
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5710819959640503,
+      "learning_rate": 0.00044467583758023895,
+      "loss": 3.1855,
+      "step": 20802
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5371394157409668,
+      "learning_rate": 0.000444662349832663,
+      "loss": 3.1907,
+      "step": 20803
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5172489285469055,
+      "learning_rate": 0.0004446488617040742,
+      "loss": 3.2504,
+      "step": 20804
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.531115710735321,
+      "learning_rate": 0.0004446353731945082,
+      "loss": 2.9735,
+      "step": 20805
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49821171164512634,
+      "learning_rate": 0.00044462188430400044,
+      "loss": 3.0562,
+      "step": 20806
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6454533338546753,
+      "learning_rate": 0.00044460839503258643,
+      "loss": 3.1137,
+      "step": 20807
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.64560866355896,
+      "learning_rate": 0.0004445949053803018,
+      "loss": 3.0143,
+      "step": 20808
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6050903797149658,
+      "learning_rate": 0.000444581415347182,
+      "loss": 3.0204,
+      "step": 20809
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5002428293228149,
+      "learning_rate": 0.00044456792493326256,
+      "loss": 2.9459,
+      "step": 20810
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.534598708152771,
+      "learning_rate": 0.000444554434138579,
+      "loss": 3.1241,
+      "step": 20811
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6012741923332214,
+      "learning_rate": 0.0004445409429631669,
+      "loss": 3.0102,
+      "step": 20812
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5511965155601501,
+      "learning_rate": 0.0004445274514070618,
+      "loss": 3.1177,
+      "step": 20813
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5788381099700928,
+      "learning_rate": 0.0004445139594702992,
+      "loss": 3.2391,
+      "step": 20814
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6373004913330078,
+      "learning_rate": 0.0004445004671529147,
+      "loss": 3.0853,
+      "step": 20815
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5785526633262634,
+      "learning_rate": 0.00044448697445494367,
+      "loss": 3.1423,
+      "step": 20816
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5426591634750366,
+      "learning_rate": 0.00044447348137642177,
+      "loss": 3.3273,
+      "step": 20817
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5850669145584106,
+      "learning_rate": 0.00044445998791738453,
+      "loss": 2.9099,
+      "step": 20818
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5589839816093445,
+      "learning_rate": 0.0004444464940778676,
+      "loss": 3.0122,
+      "step": 20819
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5280645489692688,
+      "learning_rate": 0.0004444329998579063,
+      "loss": 3.1035,
+      "step": 20820
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.593413233757019,
+      "learning_rate": 0.00044441950525753626,
+      "loss": 3.1984,
+      "step": 20821
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5713319182395935,
+      "learning_rate": 0.00044440601027679303,
+      "loss": 3.2343,
+      "step": 20822
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5799878239631653,
+      "learning_rate": 0.0004443925149157123,
+      "loss": 2.8615,
+      "step": 20823
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5746982097625732,
+      "learning_rate": 0.0004443790191743293,
+      "loss": 3.1915,
+      "step": 20824
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5620375275611877,
+      "learning_rate": 0.00044436552305267984,
+      "loss": 2.9491,
+      "step": 20825
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5882107615470886,
+      "learning_rate": 0.00044435202655079934,
+      "loss": 3.0635,
+      "step": 20826
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5356667637825012,
+      "learning_rate": 0.0004443385296687234,
+      "loss": 2.9083,
+      "step": 20827
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5147374868392944,
+      "learning_rate": 0.00044432503240648757,
+      "loss": 3.1242,
+      "step": 20828
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.550109326839447,
+      "learning_rate": 0.00044431153476412737,
+      "loss": 2.9514,
+      "step": 20829
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5489082336425781,
+      "learning_rate": 0.0004442980367416784,
+      "loss": 3.1057,
+      "step": 20830
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49882984161376953,
+      "learning_rate": 0.0004442845383391761,
+      "loss": 3.1469,
+      "step": 20831
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6269406080245972,
+      "learning_rate": 0.00044427103955665606,
+      "loss": 3.0736,
+      "step": 20832
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5686376690864563,
+      "learning_rate": 0.00044425754039415394,
+      "loss": 3.1878,
+      "step": 20833
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5298441052436829,
+      "learning_rate": 0.00044424404085170526,
+      "loss": 3.2007,
+      "step": 20834
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5424293279647827,
+      "learning_rate": 0.00044423054092934547,
+      "loss": 3.182,
+      "step": 20835
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5576657652854919,
+      "learning_rate": 0.0004442170406271102,
+      "loss": 3.2624,
+      "step": 20836
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5367722511291504,
+      "learning_rate": 0.00044420353994503503,
+      "loss": 3.0781,
+      "step": 20837
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.57781982421875,
+      "learning_rate": 0.00044419003888315544,
+      "loss": 3.0836,
+      "step": 20838
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5427062511444092,
+      "learning_rate": 0.00044417653744150705,
+      "loss": 3.1441,
+      "step": 20839
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5554705858230591,
+      "learning_rate": 0.0004441630356201254,
+      "loss": 2.918,
+      "step": 20840
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5354554653167725,
+      "learning_rate": 0.00044414953341904615,
+      "loss": 2.9882,
+      "step": 20841
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5082659721374512,
+      "learning_rate": 0.0004441360308383047,
+      "loss": 3.0436,
+      "step": 20842
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5301640033721924,
+      "learning_rate": 0.00044412252787793665,
+      "loss": 2.9847,
+      "step": 20843
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5346314907073975,
+      "learning_rate": 0.0004441090245379776,
+      "loss": 3.1013,
+      "step": 20844
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5592143535614014,
+      "learning_rate": 0.0004440955208184632,
+      "loss": 3.0265,
+      "step": 20845
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5529156923294067,
+      "learning_rate": 0.00044408201671942884,
+      "loss": 2.9318,
+      "step": 20846
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5625426769256592,
+      "learning_rate": 0.0004440685122409102,
+      "loss": 3.1834,
+      "step": 20847
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5384853482246399,
+      "learning_rate": 0.00044405500738294284,
+      "loss": 3.1567,
+      "step": 20848
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5495803952217102,
+      "learning_rate": 0.0004440415021455624,
+      "loss": 3.1289,
+      "step": 20849
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5577335357666016,
+      "learning_rate": 0.0004440279965288042,
+      "loss": 3.0094,
+      "step": 20850
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6182305216789246,
+      "learning_rate": 0.0004440144905327041,
+      "loss": 3.1001,
+      "step": 20851
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5557514429092407,
+      "learning_rate": 0.00044400098415729754,
+      "loss": 3.1761,
+      "step": 20852
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5622000694274902,
+      "learning_rate": 0.00044398747740261995,
+      "loss": 2.9326,
+      "step": 20853
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5277696251869202,
+      "learning_rate": 0.00044397397026870724,
+      "loss": 3.0355,
+      "step": 20854
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5236798524856567,
+      "learning_rate": 0.0004439604627555947,
+      "loss": 3.1484,
+      "step": 20855
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5746625661849976,
+      "learning_rate": 0.0004439469548633181,
+      "loss": 3.1309,
+      "step": 20856
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5382353663444519,
+      "learning_rate": 0.00044393344659191284,
+      "loss": 3.0546,
+      "step": 20857
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5459259748458862,
+      "learning_rate": 0.00044391993794141456,
+      "loss": 3.1935,
+      "step": 20858
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5256115794181824,
+      "learning_rate": 0.000443906428911859,
+      "loss": 3.2811,
+      "step": 20859
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5325765013694763,
+      "learning_rate": 0.00044389291950328144,
+      "loss": 3.2127,
+      "step": 20860
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5285824537277222,
+      "learning_rate": 0.00044387940971571773,
+      "loss": 3.1977,
+      "step": 20861
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5795076489448547,
+      "learning_rate": 0.00044386589954920324,
+      "loss": 3.1272,
+      "step": 20862
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.522125780582428,
+      "learning_rate": 0.0004438523890037738,
+      "loss": 3.1669,
+      "step": 20863
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5119063258171082,
+      "learning_rate": 0.0004438388780794647,
+      "loss": 3.2521,
+      "step": 20864
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5453227758407593,
+      "learning_rate": 0.00044382536677631176,
+      "loss": 3.1975,
+      "step": 20865
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5346236228942871,
+      "learning_rate": 0.0004438118550943504,
+      "loss": 3.1095,
+      "step": 20866
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5401219129562378,
+      "learning_rate": 0.00044379834303361645,
+      "loss": 3.1477,
+      "step": 20867
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5600337386131287,
+      "learning_rate": 0.0004437848305941452,
+      "loss": 3.135,
+      "step": 20868
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5327202081680298,
+      "learning_rate": 0.00044377131777597245,
+      "loss": 2.9688,
+      "step": 20869
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.557734489440918,
+      "learning_rate": 0.0004437578045791338,
+      "loss": 3.2244,
+      "step": 20870
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5608131885528564,
+      "learning_rate": 0.0004437442910036646,
+      "loss": 3.2067,
+      "step": 20871
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5300242304801941,
+      "learning_rate": 0.0004437307770496007,
+      "loss": 3.3376,
+      "step": 20872
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5177008509635925,
+      "learning_rate": 0.00044371726271697745,
+      "loss": 3.0261,
+      "step": 20873
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5148898959159851,
+      "learning_rate": 0.0004437037480058308,
+      "loss": 3.2279,
+      "step": 20874
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5264756679534912,
+      "learning_rate": 0.000443690232916196,
+      "loss": 3.0305,
+      "step": 20875
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5689302682876587,
+      "learning_rate": 0.00044367671744810884,
+      "loss": 3.263,
+      "step": 20876
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5246378779411316,
+      "learning_rate": 0.00044366320160160483,
+      "loss": 3.2429,
+      "step": 20877
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5417928695678711,
+      "learning_rate": 0.0004436496853767196,
+      "loss": 3.1085,
+      "step": 20878
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5359458327293396,
+      "learning_rate": 0.0004436361687734887,
+      "loss": 3.1174,
+      "step": 20879
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5729898810386658,
+      "learning_rate": 0.00044362265179194785,
+      "loss": 2.9681,
+      "step": 20880
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5482082366943359,
+      "learning_rate": 0.00044360913443213255,
+      "loss": 3.1622,
+      "step": 20881
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5981318354606628,
+      "learning_rate": 0.0004435956166940784,
+      "loss": 3.0729,
+      "step": 20882
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5297590494155884,
+      "learning_rate": 0.00044358209857782107,
+      "loss": 2.9377,
+      "step": 20883
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5387971997261047,
+      "learning_rate": 0.00044356858008339606,
+      "loss": 3.1172,
+      "step": 20884
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5694130063056946,
+      "learning_rate": 0.0004435550612108391,
+      "loss": 3.0498,
+      "step": 20885
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5102159380912781,
+      "learning_rate": 0.00044354154196018576,
+      "loss": 3.0899,
+      "step": 20886
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5234137773513794,
+      "learning_rate": 0.0004435280223314715,
+      "loss": 3.0329,
+      "step": 20887
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5225757956504822,
+      "learning_rate": 0.00044351450232473215,
+      "loss": 3.3125,
+      "step": 20888
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5035845041275024,
+      "learning_rate": 0.00044350098194000326,
+      "loss": 3.1396,
+      "step": 20889
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4975343942642212,
+      "learning_rate": 0.00044348746117732035,
+      "loss": 3.2713,
+      "step": 20890
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49224379658699036,
+      "learning_rate": 0.00044347394003671904,
+      "loss": 3.1967,
+      "step": 20891
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5111834406852722,
+      "learning_rate": 0.0004434604185182351,
+      "loss": 3.0417,
+      "step": 20892
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5302641987800598,
+      "learning_rate": 0.00044344689662190396,
+      "loss": 2.9933,
+      "step": 20893
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5250119566917419,
+      "learning_rate": 0.00044343337434776125,
+      "loss": 3.0811,
+      "step": 20894
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5314276218414307,
+      "learning_rate": 0.00044341985169584267,
+      "loss": 3.2476,
+      "step": 20895
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5408056974411011,
+      "learning_rate": 0.0004434063286661838,
+      "loss": 3.0161,
+      "step": 20896
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5244573354721069,
+      "learning_rate": 0.00044339280525882026,
+      "loss": 3.2421,
+      "step": 20897
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5429604053497314,
+      "learning_rate": 0.0004433792814737877,
+      "loss": 3.1625,
+      "step": 20898
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5206112265586853,
+      "learning_rate": 0.00044336575731112167,
+      "loss": 3.0023,
+      "step": 20899
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5225122570991516,
+      "learning_rate": 0.0004433522327708579,
+      "loss": 3.059,
+      "step": 20900
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5603109002113342,
+      "learning_rate": 0.0004433387078530318,
+      "loss": 2.9529,
+      "step": 20901
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.536223828792572,
+      "learning_rate": 0.00044332518255767926,
+      "loss": 2.9272,
+      "step": 20902
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5303646922111511,
+      "learning_rate": 0.00044331165688483575,
+      "loss": 3.1247,
+      "step": 20903
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5381811261177063,
+      "learning_rate": 0.00044329813083453685,
+      "loss": 3.1472,
+      "step": 20904
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5109459757804871,
+      "learning_rate": 0.0004432846044068183,
+      "loss": 3.186,
+      "step": 20905
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5285735726356506,
+      "learning_rate": 0.00044327107760171565,
+      "loss": 3.1352,
+      "step": 20906
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5155451893806458,
+      "learning_rate": 0.0004432575504192646,
+      "loss": 2.8308,
+      "step": 20907
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5433021187782288,
+      "learning_rate": 0.00044324402285950067,
+      "loss": 3.295,
+      "step": 20908
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5422189235687256,
+      "learning_rate": 0.00044323049492245954,
+      "loss": 3.0311,
+      "step": 20909
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5412802696228027,
+      "learning_rate": 0.000443216966608177,
+      "loss": 2.9925,
+      "step": 20910
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5609269142150879,
+      "learning_rate": 0.00044320343791668835,
+      "loss": 3.1073,
+      "step": 20911
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5714790225028992,
+      "learning_rate": 0.0004431899088480295,
+      "loss": 2.9488,
+      "step": 20912
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5238274931907654,
+      "learning_rate": 0.00044317637940223596,
+      "loss": 3.0765,
+      "step": 20913
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5142273902893066,
+      "learning_rate": 0.0004431628495793434,
+      "loss": 2.9384,
+      "step": 20914
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5264378786087036,
+      "learning_rate": 0.00044314931937938746,
+      "loss": 3.1992,
+      "step": 20915
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5493513345718384,
+      "learning_rate": 0.00044313578880240376,
+      "loss": 3.225,
+      "step": 20916
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5445050597190857,
+      "learning_rate": 0.00044312225784842794,
+      "loss": 3.0925,
+      "step": 20917
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5303078889846802,
+      "learning_rate": 0.0004431087265174957,
+      "loss": 3.0849,
+      "step": 20918
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5611590147018433,
+      "learning_rate": 0.0004430951948096425,
+      "loss": 3.0506,
+      "step": 20919
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.52338707447052,
+      "learning_rate": 0.0004430816627249041,
+      "loss": 3.1669,
+      "step": 20920
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5278040170669556,
+      "learning_rate": 0.00044306813026331627,
+      "loss": 3.0816,
+      "step": 20921
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5498781800270081,
+      "learning_rate": 0.00044305459742491435,
+      "loss": 3.0202,
+      "step": 20922
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5435249209403992,
+      "learning_rate": 0.00044304106420973424,
+      "loss": 3.0002,
+      "step": 20923
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5514103174209595,
+      "learning_rate": 0.00044302753061781154,
+      "loss": 3.0938,
+      "step": 20924
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5727975368499756,
+      "learning_rate": 0.0004430139966491817,
+      "loss": 2.9635,
+      "step": 20925
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4973636567592621,
+      "learning_rate": 0.00044300046230388065,
+      "loss": 3.0122,
+      "step": 20926
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5237885117530823,
+      "learning_rate": 0.0004429869275819439,
+      "loss": 3.0556,
+      "step": 20927
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5372980237007141,
+      "learning_rate": 0.000442973392483407,
+      "loss": 3.109,
+      "step": 20928
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5056158900260925,
+      "learning_rate": 0.00044295985700830583,
+      "loss": 2.9878,
+      "step": 20929
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5174768567085266,
+      "learning_rate": 0.00044294632115667574,
+      "loss": 3.0221,
+      "step": 20930
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5897946953773499,
+      "learning_rate": 0.0004429327849285527,
+      "loss": 3.0459,
+      "step": 20931
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5413453578948975,
+      "learning_rate": 0.00044291924832397223,
+      "loss": 3.4315,
+      "step": 20932
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5361354947090149,
+      "learning_rate": 0.0004429057113429699,
+      "loss": 3.1492,
+      "step": 20933
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5186129808425903,
+      "learning_rate": 0.00044289217398558144,
+      "loss": 3.0495,
+      "step": 20934
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6256024241447449,
+      "learning_rate": 0.0004428786362518424,
+      "loss": 3.0102,
+      "step": 20935
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5609905123710632,
+      "learning_rate": 0.00044286509814178866,
+      "loss": 3.0142,
+      "step": 20936
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5775830149650574,
+      "learning_rate": 0.00044285155965545573,
+      "loss": 3.0254,
+      "step": 20937
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5974602699279785,
+      "learning_rate": 0.0004428380207928792,
+      "loss": 3.2062,
+      "step": 20938
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5600636005401611,
+      "learning_rate": 0.0004428244815540949,
+      "loss": 3.0143,
+      "step": 20939
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5269203782081604,
+      "learning_rate": 0.00044281094193913837,
+      "loss": 3.2125,
+      "step": 20940
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5174791812896729,
+      "learning_rate": 0.0004427974019480453,
+      "loss": 3.2041,
+      "step": 20941
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5623819231987,
+      "learning_rate": 0.00044278386158085135,
+      "loss": 2.9925,
+      "step": 20942
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5486443638801575,
+      "learning_rate": 0.00044277032083759227,
+      "loss": 2.9681,
+      "step": 20943
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.536963939666748,
+      "learning_rate": 0.0004427567797183036,
+      "loss": 3.0695,
+      "step": 20944
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5224885940551758,
+      "learning_rate": 0.00044274323822302095,
+      "loss": 3.2491,
+      "step": 20945
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5229834914207458,
+      "learning_rate": 0.0004427296963517801,
+      "loss": 3.227,
+      "step": 20946
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.49516236782073975,
+      "learning_rate": 0.0004427161541046169,
+      "loss": 3.0456,
+      "step": 20947
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4853455126285553,
+      "learning_rate": 0.00044270261148156656,
+      "loss": 3.1123,
+      "step": 20948
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.547874927520752,
+      "learning_rate": 0.00044268906848266514,
+      "loss": 3.1893,
+      "step": 20949
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5608683228492737,
+      "learning_rate": 0.00044267552510794813,
+      "loss": 3.2441,
+      "step": 20950
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5529409646987915,
+      "learning_rate": 0.00044266198135745126,
+      "loss": 3.0731,
+      "step": 20951
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5526501536369324,
+      "learning_rate": 0.0004426484372312102,
+      "loss": 2.9746,
+      "step": 20952
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5399184823036194,
+      "learning_rate": 0.0004426348927292606,
+      "loss": 3.0924,
+      "step": 20953
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5350891351699829,
+      "learning_rate": 0.00044262134785163815,
+      "loss": 3.1553,
+      "step": 20954
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.523499071598053,
+      "learning_rate": 0.00044260780259837846,
+      "loss": 2.9259,
+      "step": 20955
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.519354522228241,
+      "learning_rate": 0.0004425942569695173,
+      "loss": 3.0434,
+      "step": 20956
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5334795117378235,
+      "learning_rate": 0.00044258071096509033,
+      "loss": 3.1886,
+      "step": 20957
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5632496476173401,
+      "learning_rate": 0.0004425671645851333,
+      "loss": 2.9461,
+      "step": 20958
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5479925870895386,
+      "learning_rate": 0.0004425536178296816,
+      "loss": 3.0682,
+      "step": 20959
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5662195086479187,
+      "learning_rate": 0.0004425400706987712,
+      "loss": 3.1047,
+      "step": 20960
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5479673147201538,
+      "learning_rate": 0.0004425265231924377,
+      "loss": 3.1159,
+      "step": 20961
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5581120848655701,
+      "learning_rate": 0.0004425129753107168,
+      "loss": 3.1118,
+      "step": 20962
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.562798261642456,
+      "learning_rate": 0.00044249942705364403,
+      "loss": 3.1258,
+      "step": 20963
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5359207987785339,
+      "learning_rate": 0.0004424858784212553,
+      "loss": 3.1239,
+      "step": 20964
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.502883791923523,
+      "learning_rate": 0.0004424723294135862,
+      "loss": 3.0139,
+      "step": 20965
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5590471625328064,
+      "learning_rate": 0.0004424587800306723,
+      "loss": 3.0502,
+      "step": 20966
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5840969085693359,
+      "learning_rate": 0.0004424452302725495,
+      "loss": 3.2399,
+      "step": 20967
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5094526410102844,
+      "learning_rate": 0.00044243168013925326,
+      "loss": 3.0227,
+      "step": 20968
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5850446224212646,
+      "learning_rate": 0.0004424181296308195,
+      "loss": 3.0581,
+      "step": 20969
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5337172746658325,
+      "learning_rate": 0.0004424045787472838,
+      "loss": 3.1475,
+      "step": 20970
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5515691041946411,
+      "learning_rate": 0.0004423910274886817,
+      "loss": 2.8972,
+      "step": 20971
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5271448493003845,
+      "learning_rate": 0.0004423774758550492,
+      "loss": 3.1464,
+      "step": 20972
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5401154160499573,
+      "learning_rate": 0.0004423639238464217,
+      "loss": 3.0886,
+      "step": 20973
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5540905594825745,
+      "learning_rate": 0.0004423503714628351,
+      "loss": 3.0209,
+      "step": 20974
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5254117250442505,
+      "learning_rate": 0.00044233681870432497,
+      "loss": 3.0147,
+      "step": 20975
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5366216897964478,
+      "learning_rate": 0.0004423232655709271,
+      "loss": 3.191,
+      "step": 20976
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.56840580701828,
+      "learning_rate": 0.00044230971206267716,
+      "loss": 3.1895,
+      "step": 20977
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5454578995704651,
+      "learning_rate": 0.0004422961581796108,
+      "loss": 3.1494,
+      "step": 20978
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5930941104888916,
+      "learning_rate": 0.0004422826039217637,
+      "loss": 3.036,
+      "step": 20979
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5133326649665833,
+      "learning_rate": 0.00044226904928917167,
+      "loss": 3.2738,
+      "step": 20980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5534525513648987,
+      "learning_rate": 0.0004422554942818703,
+      "loss": 3.0809,
+      "step": 20981
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5347831845283508,
+      "learning_rate": 0.00044224193889989534,
+      "loss": 3.062,
+      "step": 20982
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5829472541809082,
+      "learning_rate": 0.00044222838314328253,
+      "loss": 3.0576,
+      "step": 20983
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5392614603042603,
+      "learning_rate": 0.00044221482701206746,
+      "loss": 3.0753,
+      "step": 20984
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5648199319839478,
+      "learning_rate": 0.000442201270506286,
+      "loss": 2.8233,
+      "step": 20985
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5383753776550293,
+      "learning_rate": 0.00044218771362597366,
+      "loss": 3.1036,
+      "step": 20986
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5280416011810303,
+      "learning_rate": 0.0004421741563711663,
+      "loss": 3.2258,
+      "step": 20987
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5613174438476562,
+      "learning_rate": 0.0004421605987418996,
+      "loss": 3.0589,
+      "step": 20988
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.551788866519928,
+      "learning_rate": 0.00044214704073820913,
+      "loss": 3.3632,
+      "step": 20989
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5428652763366699,
+      "learning_rate": 0.00044213348236013083,
+      "loss": 2.9893,
+      "step": 20990
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5227196216583252,
+      "learning_rate": 0.00044211992360770024,
+      "loss": 3.0317,
+      "step": 20991
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4978506863117218,
+      "learning_rate": 0.0004421063644809532,
+      "loss": 2.8453,
+      "step": 20992
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5322362780570984,
+      "learning_rate": 0.0004420928049799252,
+      "loss": 3.1358,
+      "step": 20993
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5089728236198425,
+      "learning_rate": 0.0004420792451046523,
+      "loss": 3.2766,
+      "step": 20994
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5534874200820923,
+      "learning_rate": 0.00044206568485516987,
+      "loss": 3.0792,
+      "step": 20995
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.530881941318512,
+      "learning_rate": 0.00044205212423151366,
+      "loss": 3.2128,
+      "step": 20996
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5450581312179565,
+      "learning_rate": 0.0004420385632337196,
+      "loss": 3.2734,
+      "step": 20997
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5764946937561035,
+      "learning_rate": 0.00044202500186182344,
+      "loss": 2.963,
+      "step": 20998
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5664291381835938,
+      "learning_rate": 0.0004420114401158606,
+      "loss": 3.1355,
+      "step": 20999
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5113416314125061,
+      "learning_rate": 0.00044199787799586695,
+      "loss": 3.1817,
+      "step": 21000
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5444121360778809,
+      "learning_rate": 0.0004419843155018783,
+      "loss": 2.9416,
+      "step": 21001
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5358783602714539,
+      "learning_rate": 0.00044197075263393025,
+      "loss": 3.1008,
+      "step": 21002
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5368015170097351,
+      "learning_rate": 0.0004419571893920586,
+      "loss": 2.9991,
+      "step": 21003
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5103992819786072,
+      "learning_rate": 0.000441943625776299,
+      "loss": 3.1975,
+      "step": 21004
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5424513816833496,
+      "learning_rate": 0.0004419300617866872,
+      "loss": 3.0472,
+      "step": 21005
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5450505614280701,
+      "learning_rate": 0.00044191649742325893,
+      "loss": 3.0328,
+      "step": 21006
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5383841395378113,
+      "learning_rate": 0.00044190293268604985,
+      "loss": 3.1623,
+      "step": 21007
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5024534463882446,
+      "learning_rate": 0.0004418893675750958,
+      "loss": 3.2498,
+      "step": 21008
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5342257618904114,
+      "learning_rate": 0.0004418758020904326,
+      "loss": 3.198,
+      "step": 21009
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5287760496139526,
+      "learning_rate": 0.0004418622362320957,
+      "loss": 3.22,
+      "step": 21010
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5340474247932434,
+      "learning_rate": 0.00044184867000012103,
+      "loss": 3.2134,
+      "step": 21011
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6150773167610168,
+      "learning_rate": 0.00044183510339454424,
+      "loss": 3.0173,
+      "step": 21012
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5500164031982422,
+      "learning_rate": 0.00044182153641540104,
+      "loss": 3.0495,
+      "step": 21013
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5410817861557007,
+      "learning_rate": 0.00044180796906272726,
+      "loss": 3.0827,
+      "step": 21014
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5385421514511108,
+      "learning_rate": 0.00044179440133655857,
+      "loss": 2.9338,
+      "step": 21015
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5334212779998779,
+      "learning_rate": 0.0004417808332369307,
+      "loss": 2.9631,
+      "step": 21016
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5273452401161194,
+      "learning_rate": 0.0004417672647638794,
+      "loss": 3.2463,
+      "step": 21017
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5127688050270081,
+      "learning_rate": 0.00044175369591744045,
+      "loss": 3.2323,
+      "step": 21018
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5606804490089417,
+      "learning_rate": 0.00044174012669764953,
+      "loss": 3.0001,
+      "step": 21019
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5266181230545044,
+      "learning_rate": 0.00044172655710454244,
+      "loss": 3.1192,
+      "step": 21020
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5243688821792603,
+      "learning_rate": 0.0004417129871381548,
+      "loss": 3.2963,
+      "step": 21021
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5265859365463257,
+      "learning_rate": 0.0004416994167985224,
+      "loss": 3.1358,
+      "step": 21022
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5116739273071289,
+      "learning_rate": 0.0004416858460856811,
+      "loss": 2.9943,
+      "step": 21023
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.543095052242279,
+      "learning_rate": 0.0004416722749996665,
+      "loss": 2.9124,
+      "step": 21024
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.550938069820404,
+      "learning_rate": 0.0004416587035405143,
+      "loss": 3.2173,
+      "step": 21025
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5226496458053589,
+      "learning_rate": 0.00044164513170826046,
+      "loss": 3.1445,
+      "step": 21026
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5465294122695923,
+      "learning_rate": 0.0004416315595029406,
+      "loss": 3.1399,
+      "step": 21027
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5636183619499207,
+      "learning_rate": 0.00044161798692459044,
+      "loss": 3.2226,
+      "step": 21028
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5646752119064331,
+      "learning_rate": 0.0004416044139732457,
+      "loss": 3.0557,
+      "step": 21029
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5204924941062927,
+      "learning_rate": 0.0004415908406489423,
+      "loss": 3.1327,
+      "step": 21030
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5620864033699036,
+      "learning_rate": 0.00044157726695171585,
+      "loss": 3.0268,
+      "step": 21031
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5253093838691711,
+      "learning_rate": 0.0004415636928816021,
+      "loss": 3.0451,
+      "step": 21032
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5312303900718689,
+      "learning_rate": 0.0004415501184386368,
+      "loss": 3.1457,
+      "step": 21033
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5458439588546753,
+      "learning_rate": 0.0004415365436228558,
+      "loss": 3.0072,
+      "step": 21034
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5164788961410522,
+      "learning_rate": 0.00044152296843429474,
+      "loss": 3.0088,
+      "step": 21035
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5227516293525696,
+      "learning_rate": 0.0004415093928729894,
+      "loss": 2.7882,
+      "step": 21036
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5490872859954834,
+      "learning_rate": 0.0004414958169389756,
+      "loss": 3.146,
+      "step": 21037
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5249037742614746,
+      "learning_rate": 0.000441482240632289,
+      "loss": 3.1291,
+      "step": 21038
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.590631902217865,
+      "learning_rate": 0.00044146866395296545,
+      "loss": 3.1117,
+      "step": 21039
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5558299422264099,
+      "learning_rate": 0.00044145508690104056,
+      "loss": 2.882,
+      "step": 21040
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5602800846099854,
+      "learning_rate": 0.00044144150947655035,
+      "loss": 2.8093,
+      "step": 21041
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5326724648475647,
+      "learning_rate": 0.00044142793167953033,
+      "loss": 3.3363,
+      "step": 21042
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5459589958190918,
+      "learning_rate": 0.0004414143535100164,
+      "loss": 3.1367,
+      "step": 21043
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5694371461868286,
+      "learning_rate": 0.00044140077496804427,
+      "loss": 3.1642,
+      "step": 21044
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.566260039806366,
+      "learning_rate": 0.00044138719605364967,
+      "loss": 3.2619,
+      "step": 21045
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5829527974128723,
+      "learning_rate": 0.00044137361676686843,
+      "loss": 3.019,
+      "step": 21046
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5660526752471924,
+      "learning_rate": 0.00044136003710773633,
+      "loss": 3.0725,
+      "step": 21047
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5259793996810913,
+      "learning_rate": 0.000441346457076289,
+      "loss": 3.1078,
+      "step": 21048
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5162500143051147,
+      "learning_rate": 0.0004413328766725624,
+      "loss": 2.8862,
+      "step": 21049
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.560583233833313,
+      "learning_rate": 0.00044131929589659213,
+      "loss": 3.1804,
+      "step": 21050
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5528599619865417,
+      "learning_rate": 0.00044130571474841406,
+      "loss": 2.9594,
+      "step": 21051
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5608258843421936,
+      "learning_rate": 0.000441292133228064,
+      "loss": 3.2056,
+      "step": 21052
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.546389102935791,
+      "learning_rate": 0.0004412785513355776,
+      "loss": 3.0145,
+      "step": 21053
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5085633397102356,
+      "learning_rate": 0.0004412649690709907,
+      "loss": 3.1211,
+      "step": 21054
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.543613851070404,
+      "learning_rate": 0.000441251386434339,
+      "loss": 2.996,
+      "step": 21055
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5142641067504883,
+      "learning_rate": 0.00044123780342565844,
+      "loss": 3.1159,
+      "step": 21056
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5283182859420776,
+      "learning_rate": 0.00044122422004498466,
+      "loss": 3.142,
+      "step": 21057
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5592846274375916,
+      "learning_rate": 0.0004412106362923533,
+      "loss": 3.3198,
+      "step": 21058
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5456278324127197,
+      "learning_rate": 0.00044119705216780046,
+      "loss": 3.1421,
+      "step": 21059
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5718361139297485,
+      "learning_rate": 0.0004411834676713618,
+      "loss": 2.7871,
+      "step": 21060
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5030151605606079,
+      "learning_rate": 0.000441169882803073,
+      "loss": 3.1338,
+      "step": 21061
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5248968005180359,
+      "learning_rate": 0.0004411562975629698,
+      "loss": 3.273,
+      "step": 21062
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5286002159118652,
+      "learning_rate": 0.00044114271195108814,
+      "loss": 3.218,
+      "step": 21063
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.511260986328125,
+      "learning_rate": 0.0004411291259674638,
+      "loss": 3.1222,
+      "step": 21064
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5192838907241821,
+      "learning_rate": 0.00044111553961213244,
+      "loss": 3.0487,
+      "step": 21065
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.49542543292045593,
+      "learning_rate": 0.00044110195288512996,
+      "loss": 2.9655,
+      "step": 21066
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5238027572631836,
+      "learning_rate": 0.0004410883657864921,
+      "loss": 3.0834,
+      "step": 21067
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5153197050094604,
+      "learning_rate": 0.00044107477831625456,
+      "loss": 2.9757,
+      "step": 21068
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5175606608390808,
+      "learning_rate": 0.00044106119047445324,
+      "loss": 3.0309,
+      "step": 21069
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5227696299552917,
+      "learning_rate": 0.00044104760226112386,
+      "loss": 3.0677,
+      "step": 21070
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5260235071182251,
+      "learning_rate": 0.0004410340136763023,
+      "loss": 3.1779,
+      "step": 21071
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.567876398563385,
+      "learning_rate": 0.0004410204247200243,
+      "loss": 2.9599,
+      "step": 21072
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5913751721382141,
+      "learning_rate": 0.0004410068353923255,
+      "loss": 3.0299,
+      "step": 21073
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5325940847396851,
+      "learning_rate": 0.000440993245693242,
+      "loss": 3.0769,
+      "step": 21074
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5309601426124573,
+      "learning_rate": 0.0004409796556228094,
+      "loss": 3.043,
+      "step": 21075
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5216854810714722,
+      "learning_rate": 0.0004409660651810635,
+      "loss": 3.264,
+      "step": 21076
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5479691624641418,
+      "learning_rate": 0.00044095247436804006,
+      "loss": 2.9584,
+      "step": 21077
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5204125642776489,
+      "learning_rate": 0.00044093888318377507,
+      "loss": 3.1291,
+      "step": 21078
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5596486330032349,
+      "learning_rate": 0.00044092529162830397,
+      "loss": 3.1667,
+      "step": 21079
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5593887567520142,
+      "learning_rate": 0.0004409116997016629,
+      "loss": 2.9974,
+      "step": 21080
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5333544611930847,
+      "learning_rate": 0.00044089810740388755,
+      "loss": 2.9289,
+      "step": 21081
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5625095367431641,
+      "learning_rate": 0.0004408845147350137,
+      "loss": 3.0538,
+      "step": 21082
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5155503153800964,
+      "learning_rate": 0.0004408709216950771,
+      "loss": 2.9528,
+      "step": 21083
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5525764226913452,
+      "learning_rate": 0.00044085732828411355,
+      "loss": 3.1537,
+      "step": 21084
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5102589726448059,
+      "learning_rate": 0.000440843734502159,
+      "loss": 3.3095,
+      "step": 21085
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5470672845840454,
+      "learning_rate": 0.00044083014034924917,
+      "loss": 3.2635,
+      "step": 21086
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5264706015586853,
+      "learning_rate": 0.0004408165458254198,
+      "loss": 3.5163,
+      "step": 21087
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5716114044189453,
+      "learning_rate": 0.00044080295093070675,
+      "loss": 3.1289,
+      "step": 21088
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.54709792137146,
+      "learning_rate": 0.0004407893556651459,
+      "loss": 3.2208,
+      "step": 21089
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6111564636230469,
+      "learning_rate": 0.0004407757600287729,
+      "loss": 3.1601,
+      "step": 21090
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5455335974693298,
+      "learning_rate": 0.0004407621640216237,
+      "loss": 3.0098,
+      "step": 21091
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5645591020584106,
+      "learning_rate": 0.00044074856764373397,
+      "loss": 3.1535,
+      "step": 21092
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.53360915184021,
+      "learning_rate": 0.00044073497089513976,
+      "loss": 3.0372,
+      "step": 21093
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5450435876846313,
+      "learning_rate": 0.00044072137377587655,
+      "loss": 2.9299,
+      "step": 21094
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5204038023948669,
+      "learning_rate": 0.0004407077762859804,
+      "loss": 2.9236,
+      "step": 21095
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5424490571022034,
+      "learning_rate": 0.00044069417842548705,
+      "loss": 3.2089,
+      "step": 21096
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5224910378456116,
+      "learning_rate": 0.0004406805801944323,
+      "loss": 3.0936,
+      "step": 21097
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5337421894073486,
+      "learning_rate": 0.00044066698159285196,
+      "loss": 3.069,
+      "step": 21098
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5657770037651062,
+      "learning_rate": 0.00044065338262078184,
+      "loss": 3.0047,
+      "step": 21099
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5155889987945557,
+      "learning_rate": 0.00044063978327825784,
+      "loss": 3.1635,
+      "step": 21100
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.560525119304657,
+      "learning_rate": 0.00044062618356531573,
+      "loss": 3.1346,
+      "step": 21101
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5268765687942505,
+      "learning_rate": 0.00044061258348199124,
+      "loss": 2.9914,
+      "step": 21102
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5468574166297913,
+      "learning_rate": 0.0004405989830283203,
+      "loss": 3.3028,
+      "step": 21103
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5861000418663025,
+      "learning_rate": 0.00044058538220433866,
+      "loss": 3.2922,
+      "step": 21104
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5265964269638062,
+      "learning_rate": 0.0004405717810100823,
+      "loss": 3.2175,
+      "step": 21105
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5208206176757812,
+      "learning_rate": 0.00044055817944558683,
+      "loss": 3.2776,
+      "step": 21106
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5527623891830444,
+      "learning_rate": 0.0004405445775108882,
+      "loss": 3.129,
+      "step": 21107
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5665170550346375,
+      "learning_rate": 0.00044053097520602223,
+      "loss": 2.96,
+      "step": 21108
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5453583002090454,
+      "learning_rate": 0.0004405173725310246,
+      "loss": 3.0435,
+      "step": 21109
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5542186498641968,
+      "learning_rate": 0.0004405037694859313,
+      "loss": 3.1687,
+      "step": 21110
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5576282143592834,
+      "learning_rate": 0.0004404901660707782,
+      "loss": 3.0296,
+      "step": 21111
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5522076487541199,
+      "learning_rate": 0.0004404765622856009,
+      "loss": 2.9013,
+      "step": 21112
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5171756744384766,
+      "learning_rate": 0.0004404629581304355,
+      "loss": 3.1172,
+      "step": 21113
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5567202568054199,
+      "learning_rate": 0.00044044935360531765,
+      "loss": 2.9008,
+      "step": 21114
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5733157992362976,
+      "learning_rate": 0.0004404357487102832,
+      "loss": 2.8591,
+      "step": 21115
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5684636831283569,
+      "learning_rate": 0.0004404221434453681,
+      "loss": 3.2353,
+      "step": 21116
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5617053508758545,
+      "learning_rate": 0.000440408537810608,
+      "loss": 3.1757,
+      "step": 21117
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5794061422348022,
+      "learning_rate": 0.0004403949318060389,
+      "loss": 3.0882,
+      "step": 21118
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5320546627044678,
+      "learning_rate": 0.00044038132543169656,
+      "loss": 3.0382,
+      "step": 21119
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5199233293533325,
+      "learning_rate": 0.00044036771868761674,
+      "loss": 2.9563,
+      "step": 21120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5452735424041748,
+      "learning_rate": 0.0004403541115738354,
+      "loss": 3.1861,
+      "step": 21121
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5553098917007446,
+      "learning_rate": 0.00044034050409038847,
+      "loss": 3.1264,
+      "step": 21122
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5302225947380066,
+      "learning_rate": 0.0004403268962373116,
+      "loss": 3.0457,
+      "step": 21123
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5287278890609741,
+      "learning_rate": 0.00044031328801464063,
+      "loss": 3.0917,
+      "step": 21124
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5537627935409546,
+      "learning_rate": 0.0004402996794224115,
+      "loss": 3.0715,
+      "step": 21125
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5674768686294556,
+      "learning_rate": 0.00044028607046066,
+      "loss": 3.1355,
+      "step": 21126
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.505564272403717,
+      "learning_rate": 0.00044027246112942204,
+      "loss": 2.8738,
+      "step": 21127
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.537092924118042,
+      "learning_rate": 0.00044025885142873336,
+      "loss": 3.241,
+      "step": 21128
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.514145016670227,
+      "learning_rate": 0.0004402452413586299,
+      "loss": 3.0004,
+      "step": 21129
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5343130230903625,
+      "learning_rate": 0.00044023163091914735,
+      "loss": 3.1431,
+      "step": 21130
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5057582855224609,
+      "learning_rate": 0.00044021802011032175,
+      "loss": 3.0871,
+      "step": 21131
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5197306871414185,
+      "learning_rate": 0.0004402044089321889,
+      "loss": 3.0718,
+      "step": 21132
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.594575047492981,
+      "learning_rate": 0.00044019079738478464,
+      "loss": 3.019,
+      "step": 21133
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5036117434501648,
+      "learning_rate": 0.00044017718546814473,
+      "loss": 2.9993,
+      "step": 21134
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5579177737236023,
+      "learning_rate": 0.0004401635731823051,
+      "loss": 3.3196,
+      "step": 21135
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5319854021072388,
+      "learning_rate": 0.0004401499605273016,
+      "loss": 3.0564,
+      "step": 21136
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5116011500358582,
+      "learning_rate": 0.0004401363475031701,
+      "loss": 3.1638,
+      "step": 21137
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5688881278038025,
+      "learning_rate": 0.00044012273410994643,
+      "loss": 3.2267,
+      "step": 21138
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5254831314086914,
+      "learning_rate": 0.0004401091203476664,
+      "loss": 2.9605,
+      "step": 21139
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5896934270858765,
+      "learning_rate": 0.00044009550621636597,
+      "loss": 3.0805,
+      "step": 21140
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5865271091461182,
+      "learning_rate": 0.0004400818917160809,
+      "loss": 3.0489,
+      "step": 21141
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5466059446334839,
+      "learning_rate": 0.0004400682768468471,
+      "loss": 3.198,
+      "step": 21142
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5474870800971985,
+      "learning_rate": 0.00044005466160870044,
+      "loss": 3.1467,
+      "step": 21143
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5172299742698669,
+      "learning_rate": 0.0004400410460016767,
+      "loss": 3.1705,
+      "step": 21144
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5260311365127563,
+      "learning_rate": 0.0004400274300258118,
+      "loss": 3.1582,
+      "step": 21145
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5976084470748901,
+      "learning_rate": 0.0004400138136811417,
+      "loss": 3.1107,
+      "step": 21146
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5335716009140015,
+      "learning_rate": 0.0004400001969677021,
+      "loss": 3.1361,
+      "step": 21147
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5297675728797913,
+      "learning_rate": 0.0004399865798855289,
+      "loss": 3.0556,
+      "step": 21148
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5369957685470581,
+      "learning_rate": 0.0004399729624346581,
+      "loss": 3.162,
+      "step": 21149
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5566016435623169,
+      "learning_rate": 0.00043995934461512525,
+      "loss": 3.0682,
+      "step": 21150
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5417798757553101,
+      "learning_rate": 0.00043994572642696656,
+      "loss": 3.1494,
+      "step": 21151
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5253806114196777,
+      "learning_rate": 0.0004399321078702177,
+      "loss": 2.9937,
+      "step": 21152
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5396881699562073,
+      "learning_rate": 0.00043991848894491464,
+      "loss": 2.9225,
+      "step": 21153
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5461945533752441,
+      "learning_rate": 0.0004399048696510932,
+      "loss": 3.1803,
+      "step": 21154
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5187393426895142,
+      "learning_rate": 0.0004398912499887893,
+      "loss": 3.0976,
+      "step": 21155
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5625729560852051,
+      "learning_rate": 0.00043987762995803867,
+      "loss": 2.9497,
+      "step": 21156
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5398131012916565,
+      "learning_rate": 0.00043986400955887733,
+      "loss": 3.0063,
+      "step": 21157
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5377815961837769,
+      "learning_rate": 0.0004398503887913411,
+      "loss": 3.0048,
+      "step": 21158
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5535577535629272,
+      "learning_rate": 0.0004398367676554659,
+      "loss": 3.0225,
+      "step": 21159
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5389450192451477,
+      "learning_rate": 0.0004398231461512875,
+      "loss": 3.0952,
+      "step": 21160
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5288699865341187,
+      "learning_rate": 0.0004398095242788419,
+      "loss": 3.0028,
+      "step": 21161
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5519115924835205,
+      "learning_rate": 0.00043979590203816496,
+      "loss": 3.0892,
+      "step": 21162
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5267183780670166,
+      "learning_rate": 0.0004397822794292924,
+      "loss": 3.1845,
+      "step": 21163
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5286177396774292,
+      "learning_rate": 0.0004397686564522603,
+      "loss": 3.182,
+      "step": 21164
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5310975909233093,
+      "learning_rate": 0.00043975503310710436,
+      "loss": 3.0952,
+      "step": 21165
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5732306838035583,
+      "learning_rate": 0.0004397414093938607,
+      "loss": 3.0944,
+      "step": 21166
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5607469081878662,
+      "learning_rate": 0.000439727785312565,
+      "loss": 2.9405,
+      "step": 21167
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5709458589553833,
+      "learning_rate": 0.0004397141608632532,
+      "loss": 3.1021,
+      "step": 21168
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5287217497825623,
+      "learning_rate": 0.00043970053604596126,
+      "loss": 3.0768,
+      "step": 21169
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5781598091125488,
+      "learning_rate": 0.0004396869108607249,
+      "loss": 3.214,
+      "step": 21170
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5571904182434082,
+      "learning_rate": 0.00043967328530758004,
+      "loss": 3.0581,
+      "step": 21171
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5215210914611816,
+      "learning_rate": 0.0004396596593865627,
+      "loss": 2.9378,
+      "step": 21172
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6075809597969055,
+      "learning_rate": 0.0004396460330977088,
+      "loss": 2.9941,
+      "step": 21173
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5475901365280151,
+      "learning_rate": 0.000439632406441054,
+      "loss": 3.277,
+      "step": 21174
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5757737159729004,
+      "learning_rate": 0.00043961877941663433,
+      "loss": 3.2825,
+      "step": 21175
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5772517919540405,
+      "learning_rate": 0.0004396051520244857,
+      "loss": 3.1746,
+      "step": 21176
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5674006938934326,
+      "learning_rate": 0.00043959152426464393,
+      "loss": 3.1398,
+      "step": 21177
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5191792845726013,
+      "learning_rate": 0.00043957789613714493,
+      "loss": 3.1744,
+      "step": 21178
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5092511177062988,
+      "learning_rate": 0.0004395642676420247,
+      "loss": 3.0467,
+      "step": 21179
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5251277089118958,
+      "learning_rate": 0.000439550638779319,
+      "loss": 2.8873,
+      "step": 21180
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5171329975128174,
+      "learning_rate": 0.0004395370095490637,
+      "loss": 2.8804,
+      "step": 21181
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5684592127799988,
+      "learning_rate": 0.0004395233799512949,
+      "loss": 3.1189,
+      "step": 21182
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5765049457550049,
+      "learning_rate": 0.0004395097499860483,
+      "loss": 3.0344,
+      "step": 21183
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5399144291877747,
+      "learning_rate": 0.0004394961196533599,
+      "loss": 3.1513,
+      "step": 21184
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5205578207969666,
+      "learning_rate": 0.00043948248895326556,
+      "loss": 3.0598,
+      "step": 21185
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5342963933944702,
+      "learning_rate": 0.0004394688578858011,
+      "loss": 3.099,
+      "step": 21186
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5252129435539246,
+      "learning_rate": 0.00043945522645100264,
+      "loss": 3.1557,
+      "step": 21187
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5322774052619934,
+      "learning_rate": 0.0004394415946489059,
+      "loss": 3.1597,
+      "step": 21188
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5448768138885498,
+      "learning_rate": 0.00043942796247954685,
+      "loss": 3.07,
+      "step": 21189
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5391474962234497,
+      "learning_rate": 0.00043941432994296136,
+      "loss": 3.2065,
+      "step": 21190
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5174856185913086,
+      "learning_rate": 0.0004394006970391854,
+      "loss": 3.3146,
+      "step": 21191
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5522015690803528,
+      "learning_rate": 0.00043938706376825474,
+      "loss": 2.9034,
+      "step": 21192
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5845143795013428,
+      "learning_rate": 0.0004393734301302054,
+      "loss": 3.2134,
+      "step": 21193
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5457573533058167,
+      "learning_rate": 0.0004393597961250733,
+      "loss": 2.9945,
+      "step": 21194
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5486933588981628,
+      "learning_rate": 0.0004393461617528944,
+      "loss": 3.1322,
+      "step": 21195
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5403074622154236,
+      "learning_rate": 0.00043933252701370445,
+      "loss": 3.1812,
+      "step": 21196
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5348062515258789,
+      "learning_rate": 0.00043931889190753937,
+      "loss": 3.1889,
+      "step": 21197
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5630962252616882,
+      "learning_rate": 0.00043930525643443523,
+      "loss": 3.1408,
+      "step": 21198
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5177236795425415,
+      "learning_rate": 0.00043929162059442786,
+      "loss": 3.0108,
+      "step": 21199
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6250644326210022,
+      "learning_rate": 0.00043927798438755315,
+      "loss": 2.906,
+      "step": 21200
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.596113920211792,
+      "learning_rate": 0.00043926434781384703,
+      "loss": 3.0855,
+      "step": 21201
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5471876859664917,
+      "learning_rate": 0.0004392507108733454,
+      "loss": 2.9947,
+      "step": 21202
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.529731273651123,
+      "learning_rate": 0.00043923707356608414,
+      "loss": 3.0602,
+      "step": 21203
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5773414969444275,
+      "learning_rate": 0.0004392234358920993,
+      "loss": 3.1828,
+      "step": 21204
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5588471293449402,
+      "learning_rate": 0.0004392097978514267,
+      "loss": 3.2139,
+      "step": 21205
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5582773089408875,
+      "learning_rate": 0.0004391961594441024,
+      "loss": 3.0678,
+      "step": 21206
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5097718834877014,
+      "learning_rate": 0.000439182520670162,
+      "loss": 3.1309,
+      "step": 21207
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6590524911880493,
+      "learning_rate": 0.00043916888152964174,
+      "loss": 2.9753,
+      "step": 21208
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.545197606086731,
+      "learning_rate": 0.0004391552420225775,
+      "loss": 3.1811,
+      "step": 21209
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.557826042175293,
+      "learning_rate": 0.000439141602149005,
+      "loss": 3.1237,
+      "step": 21210
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5410374999046326,
+      "learning_rate": 0.0004391279619089604,
+      "loss": 3.0101,
+      "step": 21211
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5548087358474731,
+      "learning_rate": 0.0004391143213024794,
+      "loss": 3.0225,
+      "step": 21212
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5963765382766724,
+      "learning_rate": 0.0004391006803295982,
+      "loss": 3.2314,
+      "step": 21213
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5618867874145508,
+      "learning_rate": 0.0004390870389903525,
+      "loss": 3.0364,
+      "step": 21214
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5385682582855225,
+      "learning_rate": 0.0004390733972847783,
+      "loss": 3.1931,
+      "step": 21215
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5348148941993713,
+      "learning_rate": 0.0004390597552129116,
+      "loss": 3.0025,
+      "step": 21216
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5246903896331787,
+      "learning_rate": 0.00043904611277478816,
+      "loss": 3.2069,
+      "step": 21217
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5948673486709595,
+      "learning_rate": 0.0004390324699704441,
+      "loss": 2.9454,
+      "step": 21218
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5328468084335327,
+      "learning_rate": 0.0004390188267999152,
+      "loss": 3.3611,
+      "step": 21219
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5216394066810608,
+      "learning_rate": 0.0004390051832632376,
+      "loss": 3.0009,
+      "step": 21220
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7438657283782959,
+      "learning_rate": 0.000438991539360447,
+      "loss": 3.1814,
+      "step": 21221
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5399791598320007,
+      "learning_rate": 0.0004389778950915794,
+      "loss": 2.9652,
+      "step": 21222
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6201183795928955,
+      "learning_rate": 0.00043896425045667083,
+      "loss": 3.1625,
+      "step": 21223
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6501622200012207,
+      "learning_rate": 0.00043895060545575716,
+      "loss": 2.9417,
+      "step": 21224
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5394759178161621,
+      "learning_rate": 0.00043893696008887436,
+      "loss": 3.2379,
+      "step": 21225
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5677136182785034,
+      "learning_rate": 0.0004389233143560583,
+      "loss": 3.1612,
+      "step": 21226
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5787396430969238,
+      "learning_rate": 0.00043890966825734504,
+      "loss": 3.1594,
+      "step": 21227
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5350275635719299,
+      "learning_rate": 0.00043889602179277037,
+      "loss": 3.0324,
+      "step": 21228
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5961716175079346,
+      "learning_rate": 0.0004388823749623704,
+      "loss": 3.1801,
+      "step": 21229
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5254611372947693,
+      "learning_rate": 0.00043886872776618084,
+      "loss": 2.9515,
+      "step": 21230
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5171568393707275,
+      "learning_rate": 0.0004388550802042379,
+      "loss": 2.9371,
+      "step": 21231
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5672535300254822,
+      "learning_rate": 0.0004388414322765773,
+      "loss": 2.9656,
+      "step": 21232
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.551892101764679,
+      "learning_rate": 0.0004388277839832352,
+      "loss": 3.0332,
+      "step": 21233
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.571571946144104,
+      "learning_rate": 0.00043881413532424735,
+      "loss": 3.3032,
+      "step": 21234
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5441715717315674,
+      "learning_rate": 0.0004388004862996499,
+      "loss": 3.0702,
+      "step": 21235
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5653895735740662,
+      "learning_rate": 0.00043878683690947855,
+      "loss": 3.2735,
+      "step": 21236
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.548560380935669,
+      "learning_rate": 0.00043877318715376937,
+      "loss": 3.2087,
+      "step": 21237
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5536999702453613,
+      "learning_rate": 0.00043875953703255844,
+      "loss": 3.2161,
+      "step": 21238
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5580829977989197,
+      "learning_rate": 0.00043874588654588145,
+      "loss": 3.2814,
+      "step": 21239
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5434279441833496,
+      "learning_rate": 0.00043873223569377456,
+      "loss": 3.1529,
+      "step": 21240
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5689913630485535,
+      "learning_rate": 0.00043871858447627375,
+      "loss": 3.1142,
+      "step": 21241
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.554728090763092,
+      "learning_rate": 0.0004387049328934148,
+      "loss": 3.0535,
+      "step": 21242
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5019456148147583,
+      "learning_rate": 0.00043869128094523377,
+      "loss": 3.2623,
+      "step": 21243
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5831273794174194,
+      "learning_rate": 0.00043867762863176654,
+      "loss": 3.2874,
+      "step": 21244
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5366065502166748,
+      "learning_rate": 0.0004386639759530492,
+      "loss": 2.943,
+      "step": 21245
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5083133578300476,
+      "learning_rate": 0.00043865032290911764,
+      "loss": 3.0425,
+      "step": 21246
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5319649577140808,
+      "learning_rate": 0.0004386366695000078,
+      "loss": 3.1767,
+      "step": 21247
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5332556366920471,
+      "learning_rate": 0.0004386230157257556,
+      "loss": 3.1384,
+      "step": 21248
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.511458694934845,
+      "learning_rate": 0.0004386093615863972,
+      "loss": 3.0138,
+      "step": 21249
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5145825147628784,
+      "learning_rate": 0.0004385957070819683,
+      "loss": 3.0111,
+      "step": 21250
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5127465128898621,
+      "learning_rate": 0.00043858205221250496,
+      "loss": 2.944,
+      "step": 21251
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5706564784049988,
+      "learning_rate": 0.0004385683969780432,
+      "loss": 3.2836,
+      "step": 21252
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5455026030540466,
+      "learning_rate": 0.00043855474137861894,
+      "loss": 3.0848,
+      "step": 21253
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5656867623329163,
+      "learning_rate": 0.00043854108541426813,
+      "loss": 3.077,
+      "step": 21254
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5728113651275635,
+      "learning_rate": 0.00043852742908502676,
+      "loss": 3.1265,
+      "step": 21255
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5125572085380554,
+      "learning_rate": 0.0004385137723909309,
+      "loss": 3.0643,
+      "step": 21256
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5141593217849731,
+      "learning_rate": 0.00043850011533201643,
+      "loss": 2.931,
+      "step": 21257
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5243428945541382,
+      "learning_rate": 0.0004384864579083192,
+      "loss": 2.8332,
+      "step": 21258
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5448452234268188,
+      "learning_rate": 0.0004384728001198753,
+      "loss": 3.0665,
+      "step": 21259
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5277436375617981,
+      "learning_rate": 0.0004384591419667208,
+      "loss": 2.9827,
+      "step": 21260
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5299006700515747,
+      "learning_rate": 0.00043844548344889147,
+      "loss": 2.9862,
+      "step": 21261
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5536235570907593,
+      "learning_rate": 0.00043843182456642336,
+      "loss": 3.3202,
+      "step": 21262
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5166927576065063,
+      "learning_rate": 0.0004384181653193525,
+      "loss": 3.0759,
+      "step": 21263
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5285334587097168,
+      "learning_rate": 0.0004384045057077149,
+      "loss": 3.2137,
+      "step": 21264
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5186501145362854,
+      "learning_rate": 0.00043839084573154635,
+      "loss": 2.869,
+      "step": 21265
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.49396657943725586,
+      "learning_rate": 0.0004383771853908831,
+      "loss": 2.9673,
+      "step": 21266
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.559752345085144,
+      "learning_rate": 0.0004383635246857608,
+      "loss": 3.2458,
+      "step": 21267
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5300299525260925,
+      "learning_rate": 0.0004383498636162157,
+      "loss": 3.0761,
+      "step": 21268
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5181698203086853,
+      "learning_rate": 0.0004383362021822837,
+      "loss": 3.0975,
+      "step": 21269
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5253779888153076,
+      "learning_rate": 0.0004383225403840007,
+      "loss": 3.0754,
+      "step": 21270
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.519812822341919,
+      "learning_rate": 0.0004383088782214029,
+      "loss": 3.1472,
+      "step": 21271
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5578188300132751,
+      "learning_rate": 0.000438295215694526,
+      "loss": 3.1391,
+      "step": 21272
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5534501075744629,
+      "learning_rate": 0.00043828155280340613,
+      "loss": 2.9352,
+      "step": 21273
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5486505031585693,
+      "learning_rate": 0.00043826788954807924,
+      "loss": 3.2344,
+      "step": 21274
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5528342127799988,
+      "learning_rate": 0.0004382542259285814,
+      "loss": 3.014,
+      "step": 21275
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5291114449501038,
+      "learning_rate": 0.0004382405619449485,
+      "loss": 2.9962,
+      "step": 21276
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5442724227905273,
+      "learning_rate": 0.0004382268975972166,
+      "loss": 2.9954,
+      "step": 21277
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5385951995849609,
+      "learning_rate": 0.00043821323288542163,
+      "loss": 3.3394,
+      "step": 21278
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5655089616775513,
+      "learning_rate": 0.00043819956780959953,
+      "loss": 3.0762,
+      "step": 21279
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5629735589027405,
+      "learning_rate": 0.00043818590236978643,
+      "loss": 3.1988,
+      "step": 21280
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5002853274345398,
+      "learning_rate": 0.0004381722365660183,
+      "loss": 3.2544,
+      "step": 21281
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6046683192253113,
+      "learning_rate": 0.00043815857039833107,
+      "loss": 3.0964,
+      "step": 21282
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5325875282287598,
+      "learning_rate": 0.0004381449038667608,
+      "loss": 3.0506,
+      "step": 21283
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.568646252155304,
+      "learning_rate": 0.0004381312369713433,
+      "loss": 3.3053,
+      "step": 21284
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5482074022293091,
+      "learning_rate": 0.0004381175697121148,
+      "loss": 3.158,
+      "step": 21285
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5522348284721375,
+      "learning_rate": 0.0004381039020891112,
+      "loss": 3.0225,
+      "step": 21286
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5628721714019775,
+      "learning_rate": 0.0004380902341023685,
+      "loss": 3.2838,
+      "step": 21287
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5041287541389465,
+      "learning_rate": 0.0004380765657519227,
+      "loss": 3.1902,
+      "step": 21288
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5522623062133789,
+      "learning_rate": 0.0004380628970378098,
+      "loss": 3.1488,
+      "step": 21289
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5240798592567444,
+      "learning_rate": 0.0004380492279600658,
+      "loss": 2.9737,
+      "step": 21290
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.532329261302948,
+      "learning_rate": 0.00043803555851872663,
+      "loss": 3.1439,
+      "step": 21291
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5527257919311523,
+      "learning_rate": 0.0004380218887138284,
+      "loss": 3.1124,
+      "step": 21292
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5227271318435669,
+      "learning_rate": 0.0004380082185454072,
+      "loss": 3.1057,
+      "step": 21293
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5554690957069397,
+      "learning_rate": 0.00043799454801349876,
+      "loss": 3.0577,
+      "step": 21294
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5670045614242554,
+      "learning_rate": 0.0004379808771181393,
+      "loss": 3.0609,
+      "step": 21295
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5955886840820312,
+      "learning_rate": 0.00043796720585936475,
+      "loss": 3.13,
+      "step": 21296
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.546093761920929,
+      "learning_rate": 0.0004379535342372112,
+      "loss": 3.1067,
+      "step": 21297
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5156376361846924,
+      "learning_rate": 0.0004379398622517145,
+      "loss": 3.1343,
+      "step": 21298
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5549337267875671,
+      "learning_rate": 0.0004379261899029107,
+      "loss": 3.1189,
+      "step": 21299
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5297946929931641,
+      "learning_rate": 0.000437912517190836,
+      "loss": 2.8991,
+      "step": 21300
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.566015362739563,
+      "learning_rate": 0.0004378988441155262,
+      "loss": 2.8466,
+      "step": 21301
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5280096530914307,
+      "learning_rate": 0.0004378851706770174,
+      "loss": 3.1138,
+      "step": 21302
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5393192768096924,
+      "learning_rate": 0.00043787149687534555,
+      "loss": 3.0668,
+      "step": 21303
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5300957560539246,
+      "learning_rate": 0.00043785782271054676,
+      "loss": 3.1409,
+      "step": 21304
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5181828737258911,
+      "learning_rate": 0.000437844148182657,
+      "loss": 3.191,
+      "step": 21305
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5715096592903137,
+      "learning_rate": 0.00043783047329171225,
+      "loss": 3.0793,
+      "step": 21306
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5914278626441956,
+      "learning_rate": 0.0004378167980377486,
+      "loss": 3.0991,
+      "step": 21307
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.520348310470581,
+      "learning_rate": 0.0004378031224208021,
+      "loss": 2.8398,
+      "step": 21308
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5296949744224548,
+      "learning_rate": 0.00043778944644090855,
+      "loss": 3.1946,
+      "step": 21309
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5429492592811584,
+      "learning_rate": 0.00043777577009810416,
+      "loss": 3.0502,
+      "step": 21310
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.524493396282196,
+      "learning_rate": 0.00043776209339242494,
+      "loss": 2.9544,
+      "step": 21311
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5563609004020691,
+      "learning_rate": 0.00043774841632390684,
+      "loss": 3.0386,
+      "step": 21312
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5135151147842407,
+      "learning_rate": 0.0004377347388925859,
+      "loss": 3.131,
+      "step": 21313
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5470807552337646,
+      "learning_rate": 0.00043772106109849824,
+      "loss": 3.1214,
+      "step": 21314
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5473372340202332,
+      "learning_rate": 0.0004377073829416798,
+      "loss": 3.019,
+      "step": 21315
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5544005632400513,
+      "learning_rate": 0.00043769370442216655,
+      "loss": 3.2527,
+      "step": 21316
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5615202188491821,
+      "learning_rate": 0.0004376800255399946,
+      "loss": 3.1916,
+      "step": 21317
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5153510570526123,
+      "learning_rate": 0.0004376663462952,
+      "loss": 3.1229,
+      "step": 21318
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5381267666816711,
+      "learning_rate": 0.00043765266668781874,
+      "loss": 3.0508,
+      "step": 21319
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5629384517669678,
+      "learning_rate": 0.00043763898671788677,
+      "loss": 2.9926,
+      "step": 21320
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5090012550354004,
+      "learning_rate": 0.00043762530638544026,
+      "loss": 3.157,
+      "step": 21321
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5678680539131165,
+      "learning_rate": 0.00043761162569051526,
+      "loss": 2.9729,
+      "step": 21322
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5650219917297363,
+      "learning_rate": 0.0004375979446331476,
+      "loss": 3.3654,
+      "step": 21323
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5620492100715637,
+      "learning_rate": 0.00043758426321337355,
+      "loss": 3.1908,
+      "step": 21324
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5956307053565979,
+      "learning_rate": 0.0004375705814312288,
+      "loss": 2.8768,
+      "step": 21325
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5379623174667358,
+      "learning_rate": 0.00043755689928674985,
+      "loss": 3.0727,
+      "step": 21326
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5664665699005127,
+      "learning_rate": 0.00043754321677997237,
+      "loss": 2.7609,
+      "step": 21327
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.527374267578125,
+      "learning_rate": 0.0004375295339109326,
+      "loss": 3.1016,
+      "step": 21328
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5893529653549194,
+      "learning_rate": 0.0004375158506796665,
+      "loss": 3.1118,
+      "step": 21329
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5682183504104614,
+      "learning_rate": 0.0004375021670862101,
+      "loss": 3.1754,
+      "step": 21330
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5709665417671204,
+      "learning_rate": 0.0004374884831305995,
+      "loss": 3.0515,
+      "step": 21331
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5187592506408691,
+      "learning_rate": 0.00043747479881287054,
+      "loss": 3.2883,
+      "step": 21332
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5648815035820007,
+      "learning_rate": 0.0004374611141330596,
+      "loss": 3.1431,
+      "step": 21333
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.500318169593811,
+      "learning_rate": 0.0004374474290912025,
+      "loss": 3.0056,
+      "step": 21334
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5123018622398376,
+      "learning_rate": 0.00043743374368733517,
+      "loss": 3.0267,
+      "step": 21335
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5368598103523254,
+      "learning_rate": 0.00043742005792149393,
+      "loss": 3.1508,
+      "step": 21336
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5632544159889221,
+      "learning_rate": 0.00043740637179371473,
+      "loss": 3.1209,
+      "step": 21337
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5279106497764587,
+      "learning_rate": 0.0004373926853040336,
+      "loss": 3.3057,
+      "step": 21338
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5026139616966248,
+      "learning_rate": 0.0004373789984524865,
+      "loss": 3.1144,
+      "step": 21339
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5417594909667969,
+      "learning_rate": 0.0004373653112391096,
+      "loss": 3.1992,
+      "step": 21340
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5107925534248352,
+      "learning_rate": 0.0004373516236639389,
+      "loss": 3.1297,
+      "step": 21341
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.532360315322876,
+      "learning_rate": 0.00043733793572701046,
+      "loss": 3.104,
+      "step": 21342
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5431941151618958,
+      "learning_rate": 0.0004373242474283603,
+      "loss": 3.0766,
+      "step": 21343
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5421182513237,
+      "learning_rate": 0.00043731055876802463,
+      "loss": 3.2512,
+      "step": 21344
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5257422924041748,
+      "learning_rate": 0.0004372968697460392,
+      "loss": 3.1045,
+      "step": 21345
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4982752203941345,
+      "learning_rate": 0.0004372831803624403,
+      "loss": 3.182,
+      "step": 21346
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6206594705581665,
+      "learning_rate": 0.000437269490617264,
+      "loss": 3.0784,
+      "step": 21347
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5198937058448792,
+      "learning_rate": 0.00043725580051054625,
+      "loss": 3.357,
+      "step": 21348
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5259478092193604,
+      "learning_rate": 0.00043724211004232303,
+      "loss": 3.0271,
+      "step": 21349
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.523259699344635,
+      "learning_rate": 0.00043722841921263055,
+      "loss": 3.1932,
+      "step": 21350
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5149825215339661,
+      "learning_rate": 0.0004372147280215049,
+      "loss": 3.171,
+      "step": 21351
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5463266968727112,
+      "learning_rate": 0.00043720103646898204,
+      "loss": 2.9304,
+      "step": 21352
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5451768636703491,
+      "learning_rate": 0.000437187344555098,
+      "loss": 3.0214,
+      "step": 21353
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5087425112724304,
+      "learning_rate": 0.0004371736522798889,
+      "loss": 2.9616,
+      "step": 21354
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5477582216262817,
+      "learning_rate": 0.0004371599596433909,
+      "loss": 3.0898,
+      "step": 21355
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5459511280059814,
+      "learning_rate": 0.0004371462666456398,
+      "loss": 3.2466,
+      "step": 21356
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5397904515266418,
+      "learning_rate": 0.00043713257328667195,
+      "loss": 2.9689,
+      "step": 21357
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.528396487236023,
+      "learning_rate": 0.00043711887956652325,
+      "loss": 2.9729,
+      "step": 21358
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5352534651756287,
+      "learning_rate": 0.00043710518548522985,
+      "loss": 3.2941,
+      "step": 21359
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.529255747795105,
+      "learning_rate": 0.0004370914910428278,
+      "loss": 3.1262,
+      "step": 21360
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5711490511894226,
+      "learning_rate": 0.000437077796239353,
+      "loss": 2.9798,
+      "step": 21361
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5521866083145142,
+      "learning_rate": 0.00043706410107484183,
+      "loss": 3.0056,
+      "step": 21362
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5383753776550293,
+      "learning_rate": 0.0004370504055493301,
+      "loss": 2.9838,
+      "step": 21363
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5527660846710205,
+      "learning_rate": 0.000437036709662854,
+      "loss": 3.1455,
+      "step": 21364
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5740596652030945,
+      "learning_rate": 0.0004370230134154496,
+      "loss": 3.1023,
+      "step": 21365
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5765985250473022,
+      "learning_rate": 0.0004370093168071529,
+      "loss": 2.9812,
+      "step": 21366
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5971035957336426,
+      "learning_rate": 0.00043699561983800007,
+      "loss": 3.1561,
+      "step": 21367
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5382588505744934,
+      "learning_rate": 0.0004369819225080272,
+      "loss": 3.1575,
+      "step": 21368
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5307689905166626,
+      "learning_rate": 0.0004369682248172702,
+      "loss": 3.0418,
+      "step": 21369
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.551903486251831,
+      "learning_rate": 0.00043695452676576537,
+      "loss": 3.1686,
+      "step": 21370
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5409027934074402,
+      "learning_rate": 0.00043694082835354854,
+      "loss": 2.9421,
+      "step": 21371
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5229660868644714,
+      "learning_rate": 0.000436927129580656,
+      "loss": 2.9775,
+      "step": 21372
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5019228458404541,
+      "learning_rate": 0.0004369134304471238,
+      "loss": 3.0205,
+      "step": 21373
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5376384854316711,
+      "learning_rate": 0.0004368997309529879,
+      "loss": 3.1344,
+      "step": 21374
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5281094908714294,
+      "learning_rate": 0.0004368860310982845,
+      "loss": 2.9118,
+      "step": 21375
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5503444075584412,
+      "learning_rate": 0.0004368723308830495,
+      "loss": 2.9234,
+      "step": 21376
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5356557369232178,
+      "learning_rate": 0.00043685863030731934,
+      "loss": 3.2536,
+      "step": 21377
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5426025390625,
+      "learning_rate": 0.0004368449293711298,
+      "loss": 3.1201,
+      "step": 21378
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5253592133522034,
+      "learning_rate": 0.00043683122807451695,
+      "loss": 3.1486,
+      "step": 21379
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5157790184020996,
+      "learning_rate": 0.0004368175264175171,
+      "loss": 2.9994,
+      "step": 21380
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5356464385986328,
+      "learning_rate": 0.0004368038244001661,
+      "loss": 2.9935,
+      "step": 21381
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5198023915290833,
+      "learning_rate": 0.00043679012202250023,
+      "loss": 2.9333,
+      "step": 21382
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6020377278327942,
+      "learning_rate": 0.00043677641928455553,
+      "loss": 3.1164,
+      "step": 21383
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.540126383304596,
+      "learning_rate": 0.0004367627161863681,
+      "loss": 3.0763,
+      "step": 21384
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5544701814651489,
+      "learning_rate": 0.00043674901272797395,
+      "loss": 2.9656,
+      "step": 21385
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.57041335105896,
+      "learning_rate": 0.0004367353089094092,
+      "loss": 3.096,
+      "step": 21386
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5330620408058167,
+      "learning_rate": 0.0004367216047307099,
+      "loss": 3.0942,
+      "step": 21387
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5197447538375854,
+      "learning_rate": 0.0004367079001919124,
+      "loss": 2.9982,
+      "step": 21388
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5680919885635376,
+      "learning_rate": 0.0004366941952930524,
+      "loss": 3.1648,
+      "step": 21389
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6198605895042419,
+      "learning_rate": 0.0004366804900341663,
+      "loss": 3.0439,
+      "step": 21390
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5180275440216064,
+      "learning_rate": 0.00043666678441529004,
+      "loss": 2.7934,
+      "step": 21391
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5544524192810059,
+      "learning_rate": 0.00043665307843645976,
+      "loss": 2.8057,
+      "step": 21392
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6640998125076294,
+      "learning_rate": 0.0004366393720977116,
+      "loss": 3.081,
+      "step": 21393
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5651010870933533,
+      "learning_rate": 0.00043662566539908165,
+      "loss": 2.9706,
+      "step": 21394
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5061413049697876,
+      "learning_rate": 0.00043661195834060603,
+      "loss": 3.0914,
+      "step": 21395
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5113970637321472,
+      "learning_rate": 0.00043659825092232075,
+      "loss": 3.0878,
+      "step": 21396
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.521929919719696,
+      "learning_rate": 0.0004365845431442619,
+      "loss": 3.1446,
+      "step": 21397
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5344848036766052,
+      "learning_rate": 0.0004365708350064657,
+      "loss": 3.2237,
+      "step": 21398
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5206398963928223,
+      "learning_rate": 0.0004365571265089683,
+      "loss": 3.0752,
+      "step": 21399
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.583916187286377,
+      "learning_rate": 0.0004365434176518057,
+      "loss": 3.0648,
+      "step": 21400
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5128787159919739,
+      "learning_rate": 0.0004365297084350139,
+      "loss": 3.3052,
+      "step": 21401
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5522475242614746,
+      "learning_rate": 0.00043651599885862917,
+      "loss": 3.155,
+      "step": 21402
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5218787789344788,
+      "learning_rate": 0.00043650228892268755,
+      "loss": 2.8891,
+      "step": 21403
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5376594662666321,
+      "learning_rate": 0.0004364885786272252,
+      "loss": 3.1309,
+      "step": 21404
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5462920069694519,
+      "learning_rate": 0.0004364748679722782,
+      "loss": 3.139,
+      "step": 21405
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5011506080627441,
+      "learning_rate": 0.0004364611569578827,
+      "loss": 3.2214,
+      "step": 21406
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5488290190696716,
+      "learning_rate": 0.00043644744558407465,
+      "loss": 2.852,
+      "step": 21407
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5682697296142578,
+      "learning_rate": 0.0004364337338508904,
+      "loss": 3.0328,
+      "step": 21408
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5012774467468262,
+      "learning_rate": 0.0004364200217583659,
+      "loss": 2.9092,
+      "step": 21409
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5332022309303284,
+      "learning_rate": 0.00043640630930653744,
+      "loss": 3.1763,
+      "step": 21410
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.49493035674095154,
+      "learning_rate": 0.00043639259649544084,
+      "loss": 3.2316,
+      "step": 21411
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5405813455581665,
+      "learning_rate": 0.00043637888332511245,
+      "loss": 3.1027,
+      "step": 21412
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.531501829624176,
+      "learning_rate": 0.0004363651697955884,
+      "loss": 3.249,
+      "step": 21413
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5270764827728271,
+      "learning_rate": 0.0004363514559069047,
+      "loss": 2.9355,
+      "step": 21414
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5820385217666626,
+      "learning_rate": 0.00043633774165909744,
+      "loss": 3.1851,
+      "step": 21415
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5525915026664734,
+      "learning_rate": 0.0004363240270522029,
+      "loss": 2.9714,
+      "step": 21416
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5313616991043091,
+      "learning_rate": 0.00043631031208625704,
+      "loss": 3.1641,
+      "step": 21417
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5130735635757446,
+      "learning_rate": 0.0004362965967612961,
+      "loss": 3.1911,
+      "step": 21418
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5332739353179932,
+      "learning_rate": 0.0004362828810773561,
+      "loss": 3.2662,
+      "step": 21419
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4927477538585663,
+      "learning_rate": 0.0004362691650344732,
+      "loss": 3.204,
+      "step": 21420
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5364248752593994,
+      "learning_rate": 0.00043625544863268366,
+      "loss": 3.1127,
+      "step": 21421
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5149003267288208,
+      "learning_rate": 0.0004362417318720234,
+      "loss": 3.1497,
+      "step": 21422
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5644574761390686,
+      "learning_rate": 0.00043622801475252866,
+      "loss": 3.0328,
+      "step": 21423
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5710899829864502,
+      "learning_rate": 0.0004362142972742356,
+      "loss": 2.9797,
+      "step": 21424
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5768548846244812,
+      "learning_rate": 0.00043620057943718023,
+      "loss": 3.2249,
+      "step": 21425
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5565029382705688,
+      "learning_rate": 0.0004361868612413988,
+      "loss": 3.0481,
+      "step": 21426
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5066186785697937,
+      "learning_rate": 0.0004361731426869273,
+      "loss": 3.2739,
+      "step": 21427
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5121492743492126,
+      "learning_rate": 0.000436159423773802,
+      "loss": 3.0838,
+      "step": 21428
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5542187094688416,
+      "learning_rate": 0.000436145704502059,
+      "loss": 3.2834,
+      "step": 21429
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5797318816184998,
+      "learning_rate": 0.0004361319848717344,
+      "loss": 2.9999,
+      "step": 21430
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5291181802749634,
+      "learning_rate": 0.00043611826488286433,
+      "loss": 3.0283,
+      "step": 21431
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5589236617088318,
+      "learning_rate": 0.00043610454453548505,
+      "loss": 3.2367,
+      "step": 21432
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5176942348480225,
+      "learning_rate": 0.0004360908238296324,
+      "loss": 3.0578,
+      "step": 21433
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5708388686180115,
+      "learning_rate": 0.0004360771027653428,
+      "loss": 2.9649,
+      "step": 21434
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5405948758125305,
+      "learning_rate": 0.0004360633813426524,
+      "loss": 3.1171,
+      "step": 21435
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5287649631500244,
+      "learning_rate": 0.00043604965956159716,
+      "loss": 3.1744,
+      "step": 21436
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5235514640808105,
+      "learning_rate": 0.0004360359374222133,
+      "loss": 2.9239,
+      "step": 21437
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5278797745704651,
+      "learning_rate": 0.0004360222149245369,
+      "loss": 2.9237,
+      "step": 21438
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5604250431060791,
+      "learning_rate": 0.0004360084920686042,
+      "loss": 3.0579,
+      "step": 21439
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5143141150474548,
+      "learning_rate": 0.0004359947688544514,
+      "loss": 3.105,
+      "step": 21440
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5379894971847534,
+      "learning_rate": 0.0004359810452821145,
+      "loss": 2.9211,
+      "step": 21441
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5371367931365967,
+      "learning_rate": 0.00043596732135162965,
+      "loss": 3.3108,
+      "step": 21442
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5497921109199524,
+      "learning_rate": 0.00043595359706303307,
+      "loss": 3.1547,
+      "step": 21443
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5384805798530579,
+      "learning_rate": 0.0004359398724163609,
+      "loss": 3.2494,
+      "step": 21444
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5575272440910339,
+      "learning_rate": 0.00043592614741164924,
+      "loss": 3.0403,
+      "step": 21445
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5456324219703674,
+      "learning_rate": 0.0004359124220489343,
+      "loss": 3.1509,
+      "step": 21446
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5280311107635498,
+      "learning_rate": 0.00043589869632825216,
+      "loss": 3.0561,
+      "step": 21447
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.542693555355072,
+      "learning_rate": 0.000435884970249639,
+      "loss": 2.9996,
+      "step": 21448
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.531836748123169,
+      "learning_rate": 0.00043587124381313104,
+      "loss": 3.1081,
+      "step": 21449
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7221289873123169,
+      "learning_rate": 0.0004358575170187644,
+      "loss": 2.9935,
+      "step": 21450
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5565702319145203,
+      "learning_rate": 0.0004358437898665751,
+      "loss": 2.9451,
+      "step": 21451
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5191698670387268,
+      "learning_rate": 0.0004358300623565994,
+      "loss": 2.9686,
+      "step": 21452
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5412079691886902,
+      "learning_rate": 0.0004358163344888735,
+      "loss": 3.1743,
+      "step": 21453
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5316622257232666,
+      "learning_rate": 0.00043580260626343355,
+      "loss": 3.1001,
+      "step": 21454
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5519059896469116,
+      "learning_rate": 0.00043578887768031563,
+      "loss": 3.2956,
+      "step": 21455
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.507394552230835,
+      "learning_rate": 0.0004357751487395559,
+      "loss": 2.9144,
+      "step": 21456
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5262680649757385,
+      "learning_rate": 0.00043576141944119066,
+      "loss": 3.1354,
+      "step": 21457
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5504863858222961,
+      "learning_rate": 0.0004357476897852558,
+      "loss": 3.2812,
+      "step": 21458
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5362016558647156,
+      "learning_rate": 0.0004357339597717878,
+      "loss": 3.0618,
+      "step": 21459
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5501962900161743,
+      "learning_rate": 0.0004357202294008227,
+      "loss": 3.2196,
+      "step": 21460
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5307948589324951,
+      "learning_rate": 0.00043570649867239657,
+      "loss": 3.2588,
+      "step": 21461
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5310550332069397,
+      "learning_rate": 0.00043569276758654565,
+      "loss": 3.2253,
+      "step": 21462
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6042519211769104,
+      "learning_rate": 0.00043567903614330603,
+      "loss": 3.2064,
+      "step": 21463
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5459402799606323,
+      "learning_rate": 0.00043566530434271407,
+      "loss": 2.8162,
+      "step": 21464
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5372798442840576,
+      "learning_rate": 0.0004356515721848057,
+      "loss": 3.1366,
+      "step": 21465
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5367051959037781,
+      "learning_rate": 0.0004356378396696172,
+      "loss": 3.0794,
+      "step": 21466
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5590572953224182,
+      "learning_rate": 0.0004356241067971848,
+      "loss": 3.1426,
+      "step": 21467
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5460636615753174,
+      "learning_rate": 0.0004356103735675446,
+      "loss": 3.1592,
+      "step": 21468
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5461768507957458,
+      "learning_rate": 0.00043559663998073264,
+      "loss": 2.9046,
+      "step": 21469
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5281131267547607,
+      "learning_rate": 0.00043558290603678537,
+      "loss": 2.9527,
+      "step": 21470
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5545151829719543,
+      "learning_rate": 0.00043556917173573877,
+      "loss": 2.9797,
+      "step": 21471
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5535064935684204,
+      "learning_rate": 0.00043555543707762915,
+      "loss": 3.0521,
+      "step": 21472
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5981447100639343,
+      "learning_rate": 0.0004355417020624925,
+      "loss": 3.0078,
+      "step": 21473
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5304649472236633,
+      "learning_rate": 0.000435527966690365,
+      "loss": 3.0455,
+      "step": 21474
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5150699615478516,
+      "learning_rate": 0.0004355142309612831,
+      "loss": 3.1424,
+      "step": 21475
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5696814060211182,
+      "learning_rate": 0.0004355004948752827,
+      "loss": 3.0476,
+      "step": 21476
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6048765778541565,
+      "learning_rate": 0.0004354867584324002,
+      "loss": 3.0301,
+      "step": 21477
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5578874349594116,
+      "learning_rate": 0.0004354730216326714,
+      "loss": 2.9297,
+      "step": 21478
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5642783641815186,
+      "learning_rate": 0.00043545928447613294,
+      "loss": 3.1453,
+      "step": 21479
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5326113104820251,
+      "learning_rate": 0.00043544554696282075,
+      "loss": 3.2642,
+      "step": 21480
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.54203861951828,
+      "learning_rate": 0.00043543180909277104,
+      "loss": 2.9685,
+      "step": 21481
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5783830881118774,
+      "learning_rate": 0.00043541807086602,
+      "loss": 3.046,
+      "step": 21482
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5656664371490479,
+      "learning_rate": 0.0004354043322826039,
+      "loss": 3.1215,
+      "step": 21483
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5210521221160889,
+      "learning_rate": 0.00043539059334255875,
+      "loss": 3.3077,
+      "step": 21484
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5449597239494324,
+      "learning_rate": 0.00043537685404592086,
+      "loss": 3.0667,
+      "step": 21485
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5515075325965881,
+      "learning_rate": 0.00043536311439272644,
+      "loss": 3.0646,
+      "step": 21486
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5687845945358276,
+      "learning_rate": 0.00043534937438301154,
+      "loss": 3.2882,
+      "step": 21487
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6585469245910645,
+      "learning_rate": 0.00043533563401681254,
+      "loss": 3.0092,
+      "step": 21488
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5366838574409485,
+      "learning_rate": 0.00043532189329416543,
+      "loss": 3.0872,
+      "step": 21489
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5649664998054504,
+      "learning_rate": 0.00043530815221510654,
+      "loss": 2.9941,
+      "step": 21490
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5187537670135498,
+      "learning_rate": 0.000435294410779672,
+      "loss": 3.248,
+      "step": 21491
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5714203119277954,
+      "learning_rate": 0.0004352806689878981,
+      "loss": 3.077,
+      "step": 21492
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5206428170204163,
+      "learning_rate": 0.0004352669268398208,
+      "loss": 2.8386,
+      "step": 21493
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5927132368087769,
+      "learning_rate": 0.00043525318433547654,
+      "loss": 2.9447,
+      "step": 21494
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5436407923698425,
+      "learning_rate": 0.0004352394414749014,
+      "loss": 3.2219,
+      "step": 21495
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5328101515769958,
+      "learning_rate": 0.0004352256982581316,
+      "loss": 3.0678,
+      "step": 21496
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7768601179122925,
+      "learning_rate": 0.0004352119546852034,
+      "loss": 3.0912,
+      "step": 21497
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.531445324420929,
+      "learning_rate": 0.0004351982107561529,
+      "loss": 3.0633,
+      "step": 21498
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.565925657749176,
+      "learning_rate": 0.00043518446647101625,
+      "loss": 3.1066,
+      "step": 21499
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5696445107460022,
+      "learning_rate": 0.0004351707218298298,
+      "loss": 3.1754,
+      "step": 21500
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5589838624000549,
+      "learning_rate": 0.0004351569768326298,
+      "loss": 2.9757,
+      "step": 21501
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.539966881275177,
+      "learning_rate": 0.0004351432314794522,
+      "loss": 3.2406,
+      "step": 21502
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5462685823440552,
+      "learning_rate": 0.0004351294857703333,
+      "loss": 3.1382,
+      "step": 21503
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5470060110092163,
+      "learning_rate": 0.0004351157397053094,
+      "loss": 3.0518,
+      "step": 21504
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5174694657325745,
+      "learning_rate": 0.0004351019932844166,
+      "loss": 3.1323,
+      "step": 21505
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.533613920211792,
+      "learning_rate": 0.00043508824650769125,
+      "loss": 3.0013,
+      "step": 21506
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5810818076133728,
+      "learning_rate": 0.0004350744993751694,
+      "loss": 3.3575,
+      "step": 21507
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5429716110229492,
+      "learning_rate": 0.0004350607518868873,
+      "loss": 3.0456,
+      "step": 21508
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5317225456237793,
+      "learning_rate": 0.0004350470040428812,
+      "loss": 3.0374,
+      "step": 21509
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5231348276138306,
+      "learning_rate": 0.0004350332558431872,
+      "loss": 3.0947,
+      "step": 21510
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5566467642784119,
+      "learning_rate": 0.00043501950728784165,
+      "loss": 3.2049,
+      "step": 21511
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5202820301055908,
+      "learning_rate": 0.0004350057583768808,
+      "loss": 3.0884,
+      "step": 21512
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5291932821273804,
+      "learning_rate": 0.00043499200911034067,
+      "loss": 2.998,
+      "step": 21513
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5607914924621582,
+      "learning_rate": 0.0004349782594882576,
+      "loss": 3.1113,
+      "step": 21514
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.557773768901825,
+      "learning_rate": 0.0004349645095106677,
+      "loss": 3.0285,
+      "step": 21515
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5368457436561584,
+      "learning_rate": 0.00043495075917760726,
+      "loss": 3.0034,
+      "step": 21516
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.526913583278656,
+      "learning_rate": 0.00043493700848911256,
+      "loss": 3.1358,
+      "step": 21517
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5346729755401611,
+      "learning_rate": 0.0004349232574452197,
+      "loss": 3.1478,
+      "step": 21518
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.554854154586792,
+      "learning_rate": 0.00043490950604596504,
+      "loss": 3.1444,
+      "step": 21519
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5558192133903503,
+      "learning_rate": 0.00043489575429138454,
+      "loss": 2.9225,
+      "step": 21520
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5165724158287048,
+      "learning_rate": 0.00043488200218151467,
+      "loss": 3.0812,
+      "step": 21521
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5703836679458618,
+      "learning_rate": 0.0004348682497163916,
+      "loss": 3.0469,
+      "step": 21522
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5449686050415039,
+      "learning_rate": 0.0004348544968960515,
+      "loss": 3.1636,
+      "step": 21523
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6131134033203125,
+      "learning_rate": 0.0004348407437205306,
+      "loss": 3.1791,
+      "step": 21524
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.563544750213623,
+      "learning_rate": 0.000434826990189865,
+      "loss": 3.2026,
+      "step": 21525
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5371457934379578,
+      "learning_rate": 0.0004348132363040913,
+      "loss": 3.04,
+      "step": 21526
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5195092558860779,
+      "learning_rate": 0.00043479948206324537,
+      "loss": 3.2698,
+      "step": 21527
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5485871434211731,
+      "learning_rate": 0.0004347857274673635,
+      "loss": 3.0969,
+      "step": 21528
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5748703479766846,
+      "learning_rate": 0.000434771972516482,
+      "loss": 3.1038,
+      "step": 21529
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5171026587486267,
+      "learning_rate": 0.0004347582172106371,
+      "loss": 3.1868,
+      "step": 21530
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5326929092407227,
+      "learning_rate": 0.00043474446154986493,
+      "loss": 2.8315,
+      "step": 21531
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5850439667701721,
+      "learning_rate": 0.00043473070553420184,
+      "loss": 3.2327,
+      "step": 21532
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5027980804443359,
+      "learning_rate": 0.00043471694916368393,
+      "loss": 3.0613,
+      "step": 21533
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5234944224357605,
+      "learning_rate": 0.0004347031924383476,
+      "loss": 3.2077,
+      "step": 21534
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5578376650810242,
+      "learning_rate": 0.0004346894353582289,
+      "loss": 3.0004,
+      "step": 21535
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.49966779351234436,
+      "learning_rate": 0.00043467567792336413,
+      "loss": 2.9163,
+      "step": 21536
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5220000147819519,
+      "learning_rate": 0.0004346619201337896,
+      "loss": 3.1377,
+      "step": 21537
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5169272422790527,
+      "learning_rate": 0.0004346481619895415,
+      "loss": 3.0862,
+      "step": 21538
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5571988821029663,
+      "learning_rate": 0.000434634403490656,
+      "loss": 2.9343,
+      "step": 21539
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5117565989494324,
+      "learning_rate": 0.0004346206446371694,
+      "loss": 3.1412,
+      "step": 21540
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6265973448753357,
+      "learning_rate": 0.00043460688542911797,
+      "loss": 3.1845,
+      "step": 21541
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.526631772518158,
+      "learning_rate": 0.00043459312586653794,
+      "loss": 3.0802,
+      "step": 21542
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6042065620422363,
+      "learning_rate": 0.00043457936594946543,
+      "loss": 3.0749,
+      "step": 21543
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5184115171432495,
+      "learning_rate": 0.0004345656056779368,
+      "loss": 2.9749,
+      "step": 21544
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.560426652431488,
+      "learning_rate": 0.0004345518450519884,
+      "loss": 3.2708,
+      "step": 21545
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5563343167304993,
+      "learning_rate": 0.0004345380840716561,
+      "loss": 3.0454,
+      "step": 21546
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5319254398345947,
+      "learning_rate": 0.0004345243227369765,
+      "loss": 3.0709,
+      "step": 21547
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5544840693473816,
+      "learning_rate": 0.00043451056104798583,
+      "loss": 3.2259,
+      "step": 21548
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5176193714141846,
+      "learning_rate": 0.00043449679900472,
+      "loss": 3.1498,
+      "step": 21549
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5405734777450562,
+      "learning_rate": 0.0004344830366072156,
+      "loss": 2.9556,
+      "step": 21550
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5425923466682434,
+      "learning_rate": 0.00043446927385550874,
+      "loss": 3.234,
+      "step": 21551
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5294197201728821,
+      "learning_rate": 0.00043445551074963577,
+      "loss": 3.239,
+      "step": 21552
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6730620861053467,
+      "learning_rate": 0.00043444174728963277,
+      "loss": 3.1966,
+      "step": 21553
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.538125216960907,
+      "learning_rate": 0.0004344279834755361,
+      "loss": 3.1703,
+      "step": 21554
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5608587861061096,
+      "learning_rate": 0.00043441421930738203,
+      "loss": 2.9973,
+      "step": 21555
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5333779454231262,
+      "learning_rate": 0.00043440045478520677,
+      "loss": 3.3017,
+      "step": 21556
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5401734113693237,
+      "learning_rate": 0.00043438668990904654,
+      "loss": 3.0712,
+      "step": 21557
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5446183085441589,
+      "learning_rate": 0.0004343729246789377,
+      "loss": 3.1535,
+      "step": 21558
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5327209234237671,
+      "learning_rate": 0.0004343591590949164,
+      "loss": 2.9274,
+      "step": 21559
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6195545196533203,
+      "learning_rate": 0.0004343453931570189,
+      "loss": 3.0559,
+      "step": 21560
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5407031178474426,
+      "learning_rate": 0.0004343316268652815,
+      "loss": 3.1942,
+      "step": 21561
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5744057893753052,
+      "learning_rate": 0.0004343178602197404,
+      "loss": 3.0987,
+      "step": 21562
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5579788684844971,
+      "learning_rate": 0.0004343040932204321,
+      "loss": 3.0874,
+      "step": 21563
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5619020462036133,
+      "learning_rate": 0.0004342903258673925,
+      "loss": 3.2928,
+      "step": 21564
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5300469398498535,
+      "learning_rate": 0.00043427655816065803,
+      "loss": 3.3128,
+      "step": 21565
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5448369979858398,
+      "learning_rate": 0.000434262790100265,
+      "loss": 3.2438,
+      "step": 21566
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5507218241691589,
+      "learning_rate": 0.00043424902168624963,
+      "loss": 3.1717,
+      "step": 21567
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5300918221473694,
+      "learning_rate": 0.0004342352529186481,
+      "loss": 2.9758,
+      "step": 21568
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.537909746170044,
+      "learning_rate": 0.0004342214837974968,
+      "loss": 3.0509,
+      "step": 21569
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5577003955841064,
+      "learning_rate": 0.000434207714322832,
+      "loss": 3.1697,
+      "step": 21570
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5539290308952332,
+      "learning_rate": 0.00043419394449468974,
+      "loss": 2.8774,
+      "step": 21571
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5374821424484253,
+      "learning_rate": 0.0004341801743131066,
+      "loss": 2.8361,
+      "step": 21572
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5573104023933411,
+      "learning_rate": 0.0004341664037781186,
+      "loss": 3.016,
+      "step": 21573
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5746645331382751,
+      "learning_rate": 0.00043415263288976223,
+      "loss": 3.2311,
+      "step": 21574
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.524361789226532,
+      "learning_rate": 0.00043413886164807357,
+      "loss": 3.2525,
+      "step": 21575
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5610430836677551,
+      "learning_rate": 0.00043412509005308895,
+      "loss": 3.2134,
+      "step": 21576
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5629414319992065,
+      "learning_rate": 0.0004341113181048447,
+      "loss": 3.0197,
+      "step": 21577
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5457921624183655,
+      "learning_rate": 0.00043409754580337704,
+      "loss": 3.1535,
+      "step": 21578
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5242516398429871,
+      "learning_rate": 0.0004340837731487223,
+      "loss": 3.0792,
+      "step": 21579
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5492205023765564,
+      "learning_rate": 0.0004340700001409167,
+      "loss": 2.9799,
+      "step": 21580
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5539306402206421,
+      "learning_rate": 0.0004340562267799964,
+      "loss": 2.9942,
+      "step": 21581
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5252792835235596,
+      "learning_rate": 0.00043404245306599785,
+      "loss": 2.9845,
+      "step": 21582
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5377252697944641,
+      "learning_rate": 0.0004340286789989573,
+      "loss": 3.0228,
+      "step": 21583
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5324434638023376,
+      "learning_rate": 0.000434014904578911,
+      "loss": 3.0127,
+      "step": 21584
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5117992162704468,
+      "learning_rate": 0.0004340011298058953,
+      "loss": 3.0829,
+      "step": 21585
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5125967860221863,
+      "learning_rate": 0.0004339873546799464,
+      "loss": 3.1856,
+      "step": 21586
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.502491295337677,
+      "learning_rate": 0.0004339735792011004,
+      "loss": 3.1148,
+      "step": 21587
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5541898608207703,
+      "learning_rate": 0.000433959803369394,
+      "loss": 3.0582,
+      "step": 21588
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5374077558517456,
+      "learning_rate": 0.0004339460271848632,
+      "loss": 3.0841,
+      "step": 21589
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5276811122894287,
+      "learning_rate": 0.00043393225064754427,
+      "loss": 3.1326,
+      "step": 21590
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5653324723243713,
+      "learning_rate": 0.0004339184737574736,
+      "loss": 3.1166,
+      "step": 21591
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.570012092590332,
+      "learning_rate": 0.0004339046965146875,
+      "loss": 2.9614,
+      "step": 21592
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.030922770500183,
+      "learning_rate": 0.00043389091891922216,
+      "loss": 3.1976,
+      "step": 21593
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5297044515609741,
+      "learning_rate": 0.0004338771409711139,
+      "loss": 3.275,
+      "step": 21594
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5701044201850891,
+      "learning_rate": 0.00043386336267039907,
+      "loss": 2.9492,
+      "step": 21595
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5452165007591248,
+      "learning_rate": 0.0004338495840171139,
+      "loss": 3.2421,
+      "step": 21596
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5350081324577332,
+      "learning_rate": 0.00043383580501129453,
+      "loss": 2.9024,
+      "step": 21597
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5603721737861633,
+      "learning_rate": 0.00043382202565297757,
+      "loss": 2.9177,
+      "step": 21598
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.559532642364502,
+      "learning_rate": 0.00043380824594219916,
+      "loss": 3.1842,
+      "step": 21599
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5277261137962341,
+      "learning_rate": 0.00043379446587899547,
+      "loss": 3.056,
+      "step": 21600
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5619999766349792,
+      "learning_rate": 0.000433780685463403,
+      "loss": 3.2529,
+      "step": 21601
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5646684169769287,
+      "learning_rate": 0.0004337669046954578,
+      "loss": 3.2436,
+      "step": 21602
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5427494645118713,
+      "learning_rate": 0.0004337531235751965,
+      "loss": 3.1102,
+      "step": 21603
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5302757620811462,
+      "learning_rate": 0.00043373934210265517,
+      "loss": 3.152,
+      "step": 21604
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5829156637191772,
+      "learning_rate": 0.00043372556027787014,
+      "loss": 2.9137,
+      "step": 21605
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5077518820762634,
+      "learning_rate": 0.0004337117781008777,
+      "loss": 2.888,
+      "step": 21606
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5359871983528137,
+      "learning_rate": 0.00043369799557171413,
+      "loss": 3.1001,
+      "step": 21607
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5384306311607361,
+      "learning_rate": 0.0004336842126904159,
+      "loss": 3.2559,
+      "step": 21608
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5284044742584229,
+      "learning_rate": 0.00043367042945701903,
+      "loss": 3.4134,
+      "step": 21609
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5183429718017578,
+      "learning_rate": 0.0004336566458715601,
+      "loss": 2.9548,
+      "step": 21610
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5918934345245361,
+      "learning_rate": 0.0004336428619340752,
+      "loss": 3.1684,
+      "step": 21611
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.555814266204834,
+      "learning_rate": 0.0004336290776446007,
+      "loss": 2.91,
+      "step": 21612
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.551300048828125,
+      "learning_rate": 0.000433615293003173,
+      "loss": 3.1364,
+      "step": 21613
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5664514899253845,
+      "learning_rate": 0.00043360150800982836,
+      "loss": 3.1111,
+      "step": 21614
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5319811105728149,
+      "learning_rate": 0.000433587722664603,
+      "loss": 2.9241,
+      "step": 21615
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.560508131980896,
+      "learning_rate": 0.00043357393696753327,
+      "loss": 3.0344,
+      "step": 21616
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5378097891807556,
+      "learning_rate": 0.00043356015091865555,
+      "loss": 3.2021,
+      "step": 21617
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5653480887413025,
+      "learning_rate": 0.00043354636451800604,
+      "loss": 3.1775,
+      "step": 21618
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5392255187034607,
+      "learning_rate": 0.00043353257776562114,
+      "loss": 3.4296,
+      "step": 21619
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5405376553535461,
+      "learning_rate": 0.00043351879066153714,
+      "loss": 3.2874,
+      "step": 21620
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5263589024543762,
+      "learning_rate": 0.0004335050032057904,
+      "loss": 3.0824,
+      "step": 21621
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5315642356872559,
+      "learning_rate": 0.0004334912153984171,
+      "loss": 3.1399,
+      "step": 21622
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5322493314743042,
+      "learning_rate": 0.0004334774272394536,
+      "loss": 3.1131,
+      "step": 21623
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5302774310112,
+      "learning_rate": 0.0004334636387289362,
+      "loss": 3.124,
+      "step": 21624
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5446813702583313,
+      "learning_rate": 0.00043344984986690144,
+      "loss": 3.0245,
+      "step": 21625
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5131939649581909,
+      "learning_rate": 0.00043343606065338535,
+      "loss": 2.8756,
+      "step": 21626
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5474516153335571,
+      "learning_rate": 0.0004334222710884243,
+      "loss": 3.3199,
+      "step": 21627
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5619019269943237,
+      "learning_rate": 0.00043340848117205476,
+      "loss": 3.1003,
+      "step": 21628
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5801265239715576,
+      "learning_rate": 0.0004333946909043129,
+      "loss": 2.9848,
+      "step": 21629
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5769397616386414,
+      "learning_rate": 0.00043338090028523517,
+      "loss": 3.037,
+      "step": 21630
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5513659119606018,
+      "learning_rate": 0.0004333671093148577,
+      "loss": 2.9997,
+      "step": 21631
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5560690760612488,
+      "learning_rate": 0.0004333533179932171,
+      "loss": 2.9288,
+      "step": 21632
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5233303308486938,
+      "learning_rate": 0.00043333952632034937,
+      "loss": 3.0383,
+      "step": 21633
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5534685850143433,
+      "learning_rate": 0.00043332573429629106,
+      "loss": 3.0307,
+      "step": 21634
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5539918541908264,
+      "learning_rate": 0.00043331194192107843,
+      "loss": 3.0601,
+      "step": 21635
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5524195432662964,
+      "learning_rate": 0.0004332981491947478,
+      "loss": 3.0906,
+      "step": 21636
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5301859974861145,
+      "learning_rate": 0.00043328435611733547,
+      "loss": 3.0146,
+      "step": 21637
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5274763107299805,
+      "learning_rate": 0.0004332705626888778,
+      "loss": 3.1862,
+      "step": 21638
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6033293604850769,
+      "learning_rate": 0.0004332567689094111,
+      "loss": 2.8906,
+      "step": 21639
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5655192732810974,
+      "learning_rate": 0.00043324297477897176,
+      "loss": 3.1,
+      "step": 21640
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.578748345375061,
+      "learning_rate": 0.00043322918029759606,
+      "loss": 3.1724,
+      "step": 21641
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5158365368843079,
+      "learning_rate": 0.0004332153854653203,
+      "loss": 3.1176,
+      "step": 21642
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5151905417442322,
+      "learning_rate": 0.0004332015902821809,
+      "loss": 3.1752,
+      "step": 21643
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5332306027412415,
+      "learning_rate": 0.0004331877947482141,
+      "loss": 3.0743,
+      "step": 21644
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.576889157295227,
+      "learning_rate": 0.0004331739988634563,
+      "loss": 2.9053,
+      "step": 21645
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5512627363204956,
+      "learning_rate": 0.00043316020262794386,
+      "loss": 3.171,
+      "step": 21646
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5933674573898315,
+      "learning_rate": 0.00043314640604171303,
+      "loss": 3.1488,
+      "step": 21647
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5496378540992737,
+      "learning_rate": 0.00043313260910480015,
+      "loss": 3.0251,
+      "step": 21648
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5148006677627563,
+      "learning_rate": 0.0004331188118172417,
+      "loss": 3.2106,
+      "step": 21649
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5798931121826172,
+      "learning_rate": 0.0004331050141790739,
+      "loss": 2.9724,
+      "step": 21650
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.567121684551239,
+      "learning_rate": 0.0004330912161903331,
+      "loss": 3.1009,
+      "step": 21651
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5588269233703613,
+      "learning_rate": 0.00043307741785105554,
+      "loss": 2.9866,
+      "step": 21652
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5300847291946411,
+      "learning_rate": 0.0004330636191612778,
+      "loss": 2.9866,
+      "step": 21653
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5314716100692749,
+      "learning_rate": 0.00043304982012103604,
+      "loss": 3.0428,
+      "step": 21654
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5971154570579529,
+      "learning_rate": 0.0004330360207303667,
+      "loss": 3.1492,
+      "step": 21655
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7022563219070435,
+      "learning_rate": 0.00043302222098930604,
+      "loss": 3.0283,
+      "step": 21656
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5879517793655396,
+      "learning_rate": 0.0004330084208978905,
+      "loss": 3.0705,
+      "step": 21657
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5577943921089172,
+      "learning_rate": 0.0004329946204561563,
+      "loss": 3.0612,
+      "step": 21658
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5404565930366516,
+      "learning_rate": 0.0004329808196641399,
+      "loss": 3.1377,
+      "step": 21659
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5410740971565247,
+      "learning_rate": 0.00043296701852187764,
+      "loss": 3.3028,
+      "step": 21660
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5340285897254944,
+      "learning_rate": 0.0004329532170294059,
+      "loss": 3.3272,
+      "step": 21661
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5251579284667969,
+      "learning_rate": 0.0004329394151867609,
+      "loss": 3.6786,
+      "step": 21662
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5060845613479614,
+      "learning_rate": 0.00043292561299397906,
+      "loss": 3.2364,
+      "step": 21663
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.568199872970581,
+      "learning_rate": 0.0004329118104510967,
+      "loss": 2.9753,
+      "step": 21664
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5439720749855042,
+      "learning_rate": 0.0004328980075581503,
+      "loss": 3.0583,
+      "step": 21665
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5244399309158325,
+      "learning_rate": 0.00043288420431517613,
+      "loss": 3.0426,
+      "step": 21666
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5180443525314331,
+      "learning_rate": 0.0004328704007222105,
+      "loss": 3.1875,
+      "step": 21667
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.48462343215942383,
+      "learning_rate": 0.00043285659677928974,
+      "loss": 3.0099,
+      "step": 21668
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5198360681533813,
+      "learning_rate": 0.00043284279248645034,
+      "loss": 2.9479,
+      "step": 21669
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5512001514434814,
+      "learning_rate": 0.0004328289878437286,
+      "loss": 3.1317,
+      "step": 21670
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5167283415794373,
+      "learning_rate": 0.00043281518285116083,
+      "loss": 3.2503,
+      "step": 21671
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5513035655021667,
+      "learning_rate": 0.0004328013775087835,
+      "loss": 3.2889,
+      "step": 21672
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5547505617141724,
+      "learning_rate": 0.0004327875718166328,
+      "loss": 3.1017,
+      "step": 21673
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5487305521965027,
+      "learning_rate": 0.0004327737657747452,
+      "loss": 2.9976,
+      "step": 21674
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5412231087684631,
+      "learning_rate": 0.00043275995938315713,
+      "loss": 3.1704,
+      "step": 21675
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5449442863464355,
+      "learning_rate": 0.0004327461526419049,
+      "loss": 3.0803,
+      "step": 21676
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.566912829875946,
+      "learning_rate": 0.0004327323455510248,
+      "loss": 3.2645,
+      "step": 21677
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.551240861415863,
+      "learning_rate": 0.00043271853811055315,
+      "loss": 2.908,
+      "step": 21678
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5297688245773315,
+      "learning_rate": 0.00043270473032052656,
+      "loss": 3.0444,
+      "step": 21679
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5144250392913818,
+      "learning_rate": 0.00043269092218098116,
+      "loss": 2.945,
+      "step": 21680
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5448917746543884,
+      "learning_rate": 0.00043267711369195344,
+      "loss": 3.1495,
+      "step": 21681
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5812903642654419,
+      "learning_rate": 0.00043266330485347977,
+      "loss": 3.1761,
+      "step": 21682
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5192112922668457,
+      "learning_rate": 0.0004326494956655965,
+      "loss": 3.2557,
+      "step": 21683
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5966460704803467,
+      "learning_rate": 0.00043263568612833986,
+      "loss": 3.0654,
+      "step": 21684
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5527370572090149,
+      "learning_rate": 0.00043262187624174643,
+      "loss": 3.0416,
+      "step": 21685
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5254831910133362,
+      "learning_rate": 0.0004326080660058525,
+      "loss": 3.1767,
+      "step": 21686
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5413762927055359,
+      "learning_rate": 0.0004325942554206945,
+      "loss": 3.2666,
+      "step": 21687
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5316290855407715,
+      "learning_rate": 0.0004325804444863087,
+      "loss": 2.9736,
+      "step": 21688
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5323042273521423,
+      "learning_rate": 0.00043256663320273146,
+      "loss": 2.9594,
+      "step": 21689
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5886972546577454,
+      "learning_rate": 0.0004325528215699993,
+      "loss": 3.1795,
+      "step": 21690
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5144338607788086,
+      "learning_rate": 0.00043253900958814854,
+      "loss": 2.8915,
+      "step": 21691
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5302631258964539,
+      "learning_rate": 0.0004325251972572155,
+      "loss": 3.2102,
+      "step": 21692
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.534359335899353,
+      "learning_rate": 0.00043251138457723655,
+      "loss": 3.0445,
+      "step": 21693
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5207918882369995,
+      "learning_rate": 0.0004324975715482482,
+      "loss": 3.1168,
+      "step": 21694
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5734944343566895,
+      "learning_rate": 0.00043248375817028664,
+      "loss": 2.9265,
+      "step": 21695
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5305697321891785,
+      "learning_rate": 0.00043246994444338845,
+      "loss": 3.1291,
+      "step": 21696
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5317648649215698,
+      "learning_rate": 0.00043245613036758994,
+      "loss": 2.9974,
+      "step": 21697
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5591028332710266,
+      "learning_rate": 0.0004324423159429275,
+      "loss": 3.0407,
+      "step": 21698
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5748549103736877,
+      "learning_rate": 0.0004324285011694374,
+      "loss": 3.2274,
+      "step": 21699
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5331709980964661,
+      "learning_rate": 0.000432414686047156,
+      "loss": 3.1211,
+      "step": 21700
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5106818079948425,
+      "learning_rate": 0.00043240087057612003,
+      "loss": 2.9146,
+      "step": 21701
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5366308689117432,
+      "learning_rate": 0.0004323870547563655,
+      "loss": 3.3333,
+      "step": 21702
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5502612590789795,
+      "learning_rate": 0.000432373238587929,
+      "loss": 3.2715,
+      "step": 21703
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5137237906455994,
+      "learning_rate": 0.00043235942207084686,
+      "loss": 3.1082,
+      "step": 21704
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.506195068359375,
+      "learning_rate": 0.00043234560520515546,
+      "loss": 2.997,
+      "step": 21705
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5691025257110596,
+      "learning_rate": 0.0004323317879908913,
+      "loss": 3.1141,
+      "step": 21706
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5307044386863708,
+      "learning_rate": 0.0004323179704280905,
+      "loss": 2.9933,
+      "step": 21707
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5617866516113281,
+      "learning_rate": 0.0004323041525167898,
+      "loss": 2.9266,
+      "step": 21708
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5357836484909058,
+      "learning_rate": 0.0004322903342570254,
+      "loss": 2.938,
+      "step": 21709
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5504767894744873,
+      "learning_rate": 0.0004322765156488336,
+      "loss": 3.1159,
+      "step": 21710
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.537348747253418,
+      "learning_rate": 0.00043226269669225097,
+      "loss": 3.1031,
+      "step": 21711
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5378528237342834,
+      "learning_rate": 0.0004322488773873139,
+      "loss": 3.1137,
+      "step": 21712
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5131245255470276,
+      "learning_rate": 0.00043223505773405874,
+      "loss": 3.0508,
+      "step": 21713
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.533130943775177,
+      "learning_rate": 0.00043222123773252187,
+      "loss": 3.1552,
+      "step": 21714
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5789387822151184,
+      "learning_rate": 0.0004322074173827396,
+      "loss": 3.1302,
+      "step": 21715
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5390761494636536,
+      "learning_rate": 0.0004321935966847486,
+      "loss": 2.9585,
+      "step": 21716
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5794828534126282,
+      "learning_rate": 0.00043217977563858503,
+      "loss": 3.1416,
+      "step": 21717
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5323156118392944,
+      "learning_rate": 0.00043216595424428535,
+      "loss": 3.0928,
+      "step": 21718
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5307153463363647,
+      "learning_rate": 0.00043215213250188603,
+      "loss": 3.1511,
+      "step": 21719
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5392312407493591,
+      "learning_rate": 0.00043213831041142344,
+      "loss": 3.0964,
+      "step": 21720
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5451873540878296,
+      "learning_rate": 0.0004321244879729339,
+      "loss": 3.0374,
+      "step": 21721
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5187518000602722,
+      "learning_rate": 0.0004321106651864539,
+      "loss": 3.1617,
+      "step": 21722
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5146177411079407,
+      "learning_rate": 0.0004320968420520199,
+      "loss": 2.9917,
+      "step": 21723
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5389916300773621,
+      "learning_rate": 0.00043208301856966825,
+      "loss": 3.11,
+      "step": 21724
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6462126970291138,
+      "learning_rate": 0.0004320691947394352,
+      "loss": 3.0173,
+      "step": 21725
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5558220148086548,
+      "learning_rate": 0.00043205537056135735,
+      "loss": 3.0579,
+      "step": 21726
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5146065354347229,
+      "learning_rate": 0.0004320415460354712,
+      "loss": 3.0825,
+      "step": 21727
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5261913537979126,
+      "learning_rate": 0.00043202772116181297,
+      "loss": 3.142,
+      "step": 21728
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5271784663200378,
+      "learning_rate": 0.0004320138959404191,
+      "loss": 3.0003,
+      "step": 21729
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5543608665466309,
+      "learning_rate": 0.00043200007037132605,
+      "loss": 3.0188,
+      "step": 21730
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5403302311897278,
+      "learning_rate": 0.00043198624445457023,
+      "loss": 3.0804,
+      "step": 21731
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5381479263305664,
+      "learning_rate": 0.000431972418190188,
+      "loss": 3.0915,
+      "step": 21732
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5359466075897217,
+      "learning_rate": 0.0004319585915782159,
+      "loss": 3.1204,
+      "step": 21733
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5506048202514648,
+      "learning_rate": 0.0004319447646186902,
+      "loss": 3.1377,
+      "step": 21734
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5816706418991089,
+      "learning_rate": 0.00043193093731164737,
+      "loss": 3.3217,
+      "step": 21735
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.543936014175415,
+      "learning_rate": 0.00043191710965712385,
+      "loss": 3.3028,
+      "step": 21736
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5475443601608276,
+      "learning_rate": 0.0004319032816551561,
+      "loss": 3.1762,
+      "step": 21737
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5085486173629761,
+      "learning_rate": 0.0004318894533057805,
+      "loss": 3.2277,
+      "step": 21738
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5390211343765259,
+      "learning_rate": 0.0004318756246090334,
+      "loss": 3.0807,
+      "step": 21739
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5398746132850647,
+      "learning_rate": 0.00043186179556495134,
+      "loss": 2.9588,
+      "step": 21740
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5575569272041321,
+      "learning_rate": 0.00043184796617357063,
+      "loss": 2.9674,
+      "step": 21741
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5330104231834412,
+      "learning_rate": 0.00043183413643492784,
+      "loss": 3.0524,
+      "step": 21742
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5363730788230896,
+      "learning_rate": 0.0004318203063490592,
+      "loss": 3.0613,
+      "step": 21743
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5895825028419495,
+      "learning_rate": 0.0004318064759160013,
+      "loss": 2.9534,
+      "step": 21744
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.560642659664154,
+      "learning_rate": 0.00043179264513579056,
+      "loss": 2.9325,
+      "step": 21745
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.588516354560852,
+      "learning_rate": 0.0004317788140084632,
+      "loss": 2.9729,
+      "step": 21746
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5563256144523621,
+      "learning_rate": 0.0004317649825340559,
+      "loss": 3.0784,
+      "step": 21747
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.51703280210495,
+      "learning_rate": 0.00043175115071260496,
+      "loss": 3.132,
+      "step": 21748
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6014596819877625,
+      "learning_rate": 0.00043173731854414697,
+      "loss": 3.1507,
+      "step": 21749
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5737568736076355,
+      "learning_rate": 0.00043172348602871815,
+      "loss": 3.1788,
+      "step": 21750
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5655453205108643,
+      "learning_rate": 0.0004317096531663549,
+      "loss": 3.0277,
+      "step": 21751
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5275116562843323,
+      "learning_rate": 0.000431695819957094,
+      "loss": 3.0195,
+      "step": 21752
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5256880521774292,
+      "learning_rate": 0.00043168198640097154,
+      "loss": 3.193,
+      "step": 21753
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.547167181968689,
+      "learning_rate": 0.00043166815249802404,
+      "loss": 2.8198,
+      "step": 21754
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.522355854511261,
+      "learning_rate": 0.000431654318248288,
+      "loss": 3.2742,
+      "step": 21755
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5362539887428284,
+      "learning_rate": 0.00043164048365179977,
+      "loss": 3.1321,
+      "step": 21756
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5473201274871826,
+      "learning_rate": 0.00043162664870859596,
+      "loss": 3.2113,
+      "step": 21757
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5578786134719849,
+      "learning_rate": 0.0004316128134187128,
+      "loss": 3.3162,
+      "step": 21758
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5824030041694641,
+      "learning_rate": 0.0004315989777821868,
+      "loss": 3.1726,
+      "step": 21759
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5133621096611023,
+      "learning_rate": 0.00043158514179905455,
+      "loss": 3.0582,
+      "step": 21760
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5308951139450073,
+      "learning_rate": 0.00043157130546935224,
+      "loss": 3.3032,
+      "step": 21761
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5566686391830444,
+      "learning_rate": 0.0004315574687931164,
+      "loss": 3.3004,
+      "step": 21762
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6115220189094543,
+      "learning_rate": 0.0004315436317703836,
+      "loss": 2.962,
+      "step": 21763
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.52565997838974,
+      "learning_rate": 0.0004315297944011902,
+      "loss": 3.1985,
+      "step": 21764
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5332111716270447,
+      "learning_rate": 0.0004315159566855726,
+      "loss": 3.0903,
+      "step": 21765
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5685824155807495,
+      "learning_rate": 0.00043150211862356724,
+      "loss": 2.9799,
+      "step": 21766
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5400283932685852,
+      "learning_rate": 0.0004314882802152107,
+      "loss": 3.1849,
+      "step": 21767
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5025684833526611,
+      "learning_rate": 0.00043147444146053924,
+      "loss": 3.1382,
+      "step": 21768
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5145972371101379,
+      "learning_rate": 0.0004314606023595894,
+      "loss": 3.0659,
+      "step": 21769
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5135764479637146,
+      "learning_rate": 0.0004314467629123977,
+      "loss": 3.0465,
+      "step": 21770
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5721059441566467,
+      "learning_rate": 0.00043143292311900055,
+      "loss": 3.1543,
+      "step": 21771
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5446774959564209,
+      "learning_rate": 0.00043141908297943425,
+      "loss": 3.3055,
+      "step": 21772
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5134919285774231,
+      "learning_rate": 0.0004314052424937355,
+      "loss": 2.8909,
+      "step": 21773
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.518467903137207,
+      "learning_rate": 0.00043139140166194057,
+      "loss": 3.2775,
+      "step": 21774
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5316968560218811,
+      "learning_rate": 0.000431377560484086,
+      "loss": 3.2102,
+      "step": 21775
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5495380163192749,
+      "learning_rate": 0.0004313637189602082,
+      "loss": 3.1133,
+      "step": 21776
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5162259340286255,
+      "learning_rate": 0.00043134987709034364,
+      "loss": 3.1323,
+      "step": 21777
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5471155047416687,
+      "learning_rate": 0.0004313360348745289,
+      "loss": 3.0562,
+      "step": 21778
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5574952960014343,
+      "learning_rate": 0.0004313221923128001,
+      "loss": 3.027,
+      "step": 21779
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5334025025367737,
+      "learning_rate": 0.00043130834940519405,
+      "loss": 3.2028,
+      "step": 21780
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5593041181564331,
+      "learning_rate": 0.0004312945061517471,
+      "loss": 3.1218,
+      "step": 21781
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5666584968566895,
+      "learning_rate": 0.00043128066255249565,
+      "loss": 3.173,
+      "step": 21782
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5271990299224854,
+      "learning_rate": 0.0004312668186074762,
+      "loss": 3.0105,
+      "step": 21783
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5458244681358337,
+      "learning_rate": 0.0004312529743167252,
+      "loss": 2.9528,
+      "step": 21784
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5140916705131531,
+      "learning_rate": 0.0004312391296802792,
+      "loss": 3.1281,
+      "step": 21785
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5619597434997559,
+      "learning_rate": 0.0004312252846981745,
+      "loss": 2.9598,
+      "step": 21786
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5158681273460388,
+      "learning_rate": 0.0004312114393704476,
+      "loss": 3.0605,
+      "step": 21787
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6134312152862549,
+      "learning_rate": 0.00043119759369713515,
+      "loss": 3.008,
+      "step": 21788
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5468149781227112,
+      "learning_rate": 0.0004311837476782735,
+      "loss": 2.905,
+      "step": 21789
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5272074341773987,
+      "learning_rate": 0.000431169901313899,
+      "loss": 3.1477,
+      "step": 21790
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5446023344993591,
+      "learning_rate": 0.0004311560546040483,
+      "loss": 3.1807,
+      "step": 21791
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5380364656448364,
+      "learning_rate": 0.0004311422075487578,
+      "loss": 2.9937,
+      "step": 21792
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5410694479942322,
+      "learning_rate": 0.00043112836014806385,
+      "loss": 3.0857,
+      "step": 21793
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5312390327453613,
+      "learning_rate": 0.00043111451240200317,
+      "loss": 2.923,
+      "step": 21794
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5283505320549011,
+      "learning_rate": 0.000431100664310612,
+      "loss": 3.2019,
+      "step": 21795
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5480759739875793,
+      "learning_rate": 0.000431086815873927,
+      "loss": 3.091,
+      "step": 21796
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5413346290588379,
+      "learning_rate": 0.00043107296709198447,
+      "loss": 3.1228,
+      "step": 21797
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6117550134658813,
+      "learning_rate": 0.00043105911796482104,
+      "loss": 3.184,
+      "step": 21798
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.514468252658844,
+      "learning_rate": 0.00043104526849247306,
+      "loss": 3.165,
+      "step": 21799
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.542933464050293,
+      "learning_rate": 0.00043103141867497713,
+      "loss": 3.1681,
+      "step": 21800
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5387585163116455,
+      "learning_rate": 0.0004310175685123696,
+      "loss": 3.1421,
+      "step": 21801
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5503361821174622,
+      "learning_rate": 0.00043100371800468696,
+      "loss": 3.2455,
+      "step": 21802
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5273892879486084,
+      "learning_rate": 0.00043098986715196596,
+      "loss": 3.1728,
+      "step": 21803
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5189810395240784,
+      "learning_rate": 0.0004309760159542426,
+      "loss": 3.1855,
+      "step": 21804
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5295484066009521,
+      "learning_rate": 0.0004309621644115538,
+      "loss": 3.0618,
+      "step": 21805
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5467668175697327,
+      "learning_rate": 0.0004309483125239358,
+      "loss": 2.8678,
+      "step": 21806
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5472779273986816,
+      "learning_rate": 0.00043093446029142515,
+      "loss": 3.0829,
+      "step": 21807
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5611221194267273,
+      "learning_rate": 0.0004309206077140583,
+      "loss": 3.1447,
+      "step": 21808
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5621599555015564,
+      "learning_rate": 0.0004309067547918718,
+      "loss": 3.067,
+      "step": 21809
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5594969391822815,
+      "learning_rate": 0.0004308929015249021,
+      "loss": 3.1116,
+      "step": 21810
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5223482847213745,
+      "learning_rate": 0.0004308790479131857,
+      "loss": 3.0363,
+      "step": 21811
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5609418749809265,
+      "learning_rate": 0.00043086519395675897,
+      "loss": 3.0169,
+      "step": 21812
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5535745024681091,
+      "learning_rate": 0.0004308513396556585,
+      "loss": 2.7776,
+      "step": 21813
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5150181651115417,
+      "learning_rate": 0.000430837485009921,
+      "loss": 2.8866,
+      "step": 21814
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5247113108634949,
+      "learning_rate": 0.0004308236300195826,
+      "loss": 3.1063,
+      "step": 21815
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5388785600662231,
+      "learning_rate": 0.00043080977468467995,
+      "loss": 3.3039,
+      "step": 21816
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6435933709144592,
+      "learning_rate": 0.00043079591900524954,
+      "loss": 2.9531,
+      "step": 21817
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5959275960922241,
+      "learning_rate": 0.00043078206298132783,
+      "loss": 3.0319,
+      "step": 21818
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5229818820953369,
+      "learning_rate": 0.00043076820661295135,
+      "loss": 3.1529,
+      "step": 21819
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5054709911346436,
+      "learning_rate": 0.0004307543499001566,
+      "loss": 3.1894,
+      "step": 21820
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5688246488571167,
+      "learning_rate": 0.00043074049284297995,
+      "loss": 3.086,
+      "step": 21821
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.598905622959137,
+      "learning_rate": 0.00043072663544145817,
+      "loss": 3.2127,
+      "step": 21822
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5313297510147095,
+      "learning_rate": 0.00043071277769562744,
+      "loss": 3.0613,
+      "step": 21823
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5411408543586731,
+      "learning_rate": 0.00043069891960552446,
+      "loss": 3.1642,
+      "step": 21824
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5270294547080994,
+      "learning_rate": 0.0004306850611711858,
+      "loss": 3.2479,
+      "step": 21825
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5298737287521362,
+      "learning_rate": 0.00043067120239264776,
+      "loss": 3.2503,
+      "step": 21826
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.544009804725647,
+      "learning_rate": 0.0004306573432699468,
+      "loss": 3.013,
+      "step": 21827
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.51810622215271,
+      "learning_rate": 0.0004306434838031196,
+      "loss": 3.04,
+      "step": 21828
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5270453095436096,
+      "learning_rate": 0.0004306296239922027,
+      "loss": 2.9395,
+      "step": 21829
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5867906212806702,
+      "learning_rate": 0.0004306157638372325,
+      "loss": 3.0299,
+      "step": 21830
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5443738698959351,
+      "learning_rate": 0.0004306019033382454,
+      "loss": 3.091,
+      "step": 21831
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5243510007858276,
+      "learning_rate": 0.0004305880424952781,
+      "loss": 3.0763,
+      "step": 21832
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5453609824180603,
+      "learning_rate": 0.00043057418130836706,
+      "loss": 3.0139,
+      "step": 21833
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5322545766830444,
+      "learning_rate": 0.0004305603197775487,
+      "loss": 2.9306,
+      "step": 21834
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5745664834976196,
+      "learning_rate": 0.00043054645790285964,
+      "loss": 3.1827,
+      "step": 21835
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5085702538490295,
+      "learning_rate": 0.0004305325956843363,
+      "loss": 3.0335,
+      "step": 21836
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5018842816352844,
+      "learning_rate": 0.00043051873312201526,
+      "loss": 3.0121,
+      "step": 21837
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5426105260848999,
+      "learning_rate": 0.00043050487021593285,
+      "loss": 3.0908,
+      "step": 21838
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.529456615447998,
+      "learning_rate": 0.0004304910069661259,
+      "loss": 3.108,
+      "step": 21839
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5271927714347839,
+      "learning_rate": 0.0004304771433726307,
+      "loss": 3.0775,
+      "step": 21840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5279380679130554,
+      "learning_rate": 0.00043046327943548383,
+      "loss": 3.059,
+      "step": 21841
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5312852263450623,
+      "learning_rate": 0.0004304494151547217,
+      "loss": 3.2413,
+      "step": 21842
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.525822103023529,
+      "learning_rate": 0.00043043555053038095,
+      "loss": 3.0476,
+      "step": 21843
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5275469422340393,
+      "learning_rate": 0.0004304216855624981,
+      "loss": 3.1864,
+      "step": 21844
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5422474145889282,
+      "learning_rate": 0.00043040782025110966,
+      "loss": 3.1227,
+      "step": 21845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5537338256835938,
+      "learning_rate": 0.00043039395459625203,
+      "loss": 3.2381,
+      "step": 21846
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5338166356086731,
+      "learning_rate": 0.000430380088597962,
+      "loss": 3.2902,
+      "step": 21847
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5591501593589783,
+      "learning_rate": 0.00043036622225627567,
+      "loss": 3.2828,
+      "step": 21848
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5365943312644958,
+      "learning_rate": 0.00043035235557122985,
+      "loss": 2.9454,
+      "step": 21849
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5561801195144653,
+      "learning_rate": 0.00043033848854286106,
+      "loss": 3.1468,
+      "step": 21850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5465043187141418,
+      "learning_rate": 0.0004303246211712059,
+      "loss": 3.0577,
+      "step": 21851
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5761997699737549,
+      "learning_rate": 0.0004303107534563006,
+      "loss": 2.8597,
+      "step": 21852
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5243141055107117,
+      "learning_rate": 0.0004302968853981819,
+      "loss": 3.114,
+      "step": 21853
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5474295020103455,
+      "learning_rate": 0.0004302830169968863,
+      "loss": 3.0841,
+      "step": 21854
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5314984321594238,
+      "learning_rate": 0.0004302691482524503,
+      "loss": 3.25,
+      "step": 21855
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5473430156707764,
+      "learning_rate": 0.0004302552791649105,
+      "loss": 3.2211,
+      "step": 21856
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6076376438140869,
+      "learning_rate": 0.0004302414097343032,
+      "loss": 3.1258,
+      "step": 21857
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.555774986743927,
+      "learning_rate": 0.0004302275399606653,
+      "loss": 3.2512,
+      "step": 21858
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5176228880882263,
+      "learning_rate": 0.0004302136698440329,
+      "loss": 3.0976,
+      "step": 21859
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5607179403305054,
+      "learning_rate": 0.0004301997993844429,
+      "loss": 3.1035,
+      "step": 21860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5279338955879211,
+      "learning_rate": 0.00043018592858193165,
+      "loss": 3.0925,
+      "step": 21861
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.506940484046936,
+      "learning_rate": 0.00043017205743653577,
+      "loss": 3.1532,
+      "step": 21862
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5521911978721619,
+      "learning_rate": 0.0004301581859482917,
+      "loss": 3.0362,
+      "step": 21863
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5607286095619202,
+      "learning_rate": 0.0004301443141172359,
+      "loss": 3.0133,
+      "step": 21864
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5357416272163391,
+      "learning_rate": 0.00043013044194340513,
+      "loss": 3.1389,
+      "step": 21865
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5648325681686401,
+      "learning_rate": 0.0004301165694268359,
+      "loss": 3.0022,
+      "step": 21866
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5390356779098511,
+      "learning_rate": 0.0004301026965675646,
+      "loss": 3.0165,
+      "step": 21867
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5444007515907288,
+      "learning_rate": 0.00043008882336562776,
+      "loss": 3.1716,
+      "step": 21868
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5538380742073059,
+      "learning_rate": 0.00043007494982106204,
+      "loss": 2.798,
+      "step": 21869
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5302233099937439,
+      "learning_rate": 0.000430061075933904,
+      "loss": 3.2093,
+      "step": 21870
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5354451537132263,
+      "learning_rate": 0.00043004720170419005,
+      "loss": 3.1233,
+      "step": 21871
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.553068995475769,
+      "learning_rate": 0.00043003332713195685,
+      "loss": 2.9238,
+      "step": 21872
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5181149840354919,
+      "learning_rate": 0.0004300194522172409,
+      "loss": 3.185,
+      "step": 21873
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.53992760181427,
+      "learning_rate": 0.0004300055769600786,
+      "loss": 3.1029,
+      "step": 21874
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5325241088867188,
+      "learning_rate": 0.00042999170136050674,
+      "loss": 3.1176,
+      "step": 21875
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5475149750709534,
+      "learning_rate": 0.0004299778254185618,
+      "loss": 3.044,
+      "step": 21876
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5398291945457458,
+      "learning_rate": 0.00042996394913428023,
+      "loss": 3.1001,
+      "step": 21877
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5641384124755859,
+      "learning_rate": 0.0004299500725076986,
+      "loss": 2.9772,
+      "step": 21878
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5371559262275696,
+      "learning_rate": 0.00042993619553885346,
+      "loss": 2.9344,
+      "step": 21879
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5246857404708862,
+      "learning_rate": 0.00042992231822778154,
+      "loss": 3.0261,
+      "step": 21880
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5782071948051453,
+      "learning_rate": 0.0004299084405745191,
+      "loss": 3.1893,
+      "step": 21881
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5363188982009888,
+      "learning_rate": 0.0004298945625791028,
+      "loss": 2.9895,
+      "step": 21882
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5487208366394043,
+      "learning_rate": 0.0004298806842415693,
+      "loss": 2.9313,
+      "step": 21883
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5662634968757629,
+      "learning_rate": 0.0004298668055619551,
+      "loss": 3.0855,
+      "step": 21884
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5252002477645874,
+      "learning_rate": 0.0004298529265402967,
+      "loss": 2.8349,
+      "step": 21885
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7216542363166809,
+      "learning_rate": 0.00042983904717663066,
+      "loss": 3.1169,
+      "step": 21886
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5242555141448975,
+      "learning_rate": 0.0004298251674709936,
+      "loss": 3.0732,
+      "step": 21887
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5485615730285645,
+      "learning_rate": 0.00042981128742342197,
+      "loss": 3.0352,
+      "step": 21888
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5380790829658508,
+      "learning_rate": 0.00042979740703395243,
+      "loss": 3.0147,
+      "step": 21889
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5466700792312622,
+      "learning_rate": 0.0004297835263026215,
+      "loss": 3.0625,
+      "step": 21890
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6295219659805298,
+      "learning_rate": 0.00042976964522946576,
+      "loss": 2.8748,
+      "step": 21891
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5317986011505127,
+      "learning_rate": 0.0004297557638145217,
+      "loss": 2.9312,
+      "step": 21892
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5600071549415588,
+      "learning_rate": 0.000429741882057826,
+      "loss": 3.1082,
+      "step": 21893
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5254099369049072,
+      "learning_rate": 0.0004297279999594151,
+      "loss": 2.9492,
+      "step": 21894
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.557667076587677,
+      "learning_rate": 0.00042971411751932564,
+      "loss": 2.9501,
+      "step": 21895
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5373759865760803,
+      "learning_rate": 0.0004297002347375941,
+      "loss": 2.9849,
+      "step": 21896
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5822132229804993,
+      "learning_rate": 0.00042968635161425717,
+      "loss": 3.1294,
+      "step": 21897
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5314226150512695,
+      "learning_rate": 0.0004296724681493513,
+      "loss": 3.0397,
+      "step": 21898
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5426365733146667,
+      "learning_rate": 0.00042965858434291317,
+      "loss": 2.9715,
+      "step": 21899
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5367540121078491,
+      "learning_rate": 0.00042964470019497917,
+      "loss": 3.0716,
+      "step": 21900
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5575674176216125,
+      "learning_rate": 0.000429630815705586,
+      "loss": 3.2273,
+      "step": 21901
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5367324948310852,
+      "learning_rate": 0.0004296169308747704,
+      "loss": 2.9208,
+      "step": 21902
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5453097224235535,
+      "learning_rate": 0.00042960304570256855,
+      "loss": 3.211,
+      "step": 21903
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5521195530891418,
+      "learning_rate": 0.00042958916018901724,
+      "loss": 3.0296,
+      "step": 21904
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5742548108100891,
+      "learning_rate": 0.0004295752743341531,
+      "loss": 2.9801,
+      "step": 21905
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5651740431785583,
+      "learning_rate": 0.00042956138813801263,
+      "loss": 3.216,
+      "step": 21906
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5569771528244019,
+      "learning_rate": 0.0004295475016006323,
+      "loss": 3.0902,
+      "step": 21907
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5515370965003967,
+      "learning_rate": 0.0004295336147220488,
+      "loss": 3.2262,
+      "step": 21908
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5465577244758606,
+      "learning_rate": 0.0004295197275022988,
+      "loss": 3.0305,
+      "step": 21909
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5600085854530334,
+      "learning_rate": 0.00042950583994141867,
+      "loss": 3.2192,
+      "step": 21910
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5199048519134521,
+      "learning_rate": 0.00042949195203944505,
+      "loss": 3.2271,
+      "step": 21911
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5712819695472717,
+      "learning_rate": 0.0004294780637964146,
+      "loss": 3.1676,
+      "step": 21912
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5674663782119751,
+      "learning_rate": 0.00042946417521236394,
+      "loss": 3.0938,
+      "step": 21913
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5764085054397583,
+      "learning_rate": 0.0004294502862873294,
+      "loss": 3.1466,
+      "step": 21914
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5495882630348206,
+      "learning_rate": 0.0004294363970213477,
+      "loss": 3.1355,
+      "step": 21915
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5135552287101746,
+      "learning_rate": 0.0004294225074144556,
+      "loss": 2.932,
+      "step": 21916
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5441713929176331,
+      "learning_rate": 0.0004294086174666895,
+      "loss": 3.0714,
+      "step": 21917
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.539913535118103,
+      "learning_rate": 0.00042939472717808596,
+      "loss": 3.0711,
+      "step": 21918
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5363076329231262,
+      "learning_rate": 0.00042938083654868155,
+      "loss": 3.0245,
+      "step": 21919
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5468325614929199,
+      "learning_rate": 0.000429366945578513,
+      "loss": 2.9802,
+      "step": 21920
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5180367827415466,
+      "learning_rate": 0.0004293530542676168,
+      "loss": 3.2203,
+      "step": 21921
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5398200154304504,
+      "learning_rate": 0.00042933916261602946,
+      "loss": 2.9909,
+      "step": 21922
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5421903729438782,
+      "learning_rate": 0.0004293252706237878,
+      "loss": 2.8939,
+      "step": 21923
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6031635999679565,
+      "learning_rate": 0.00042931137829092824,
+      "loss": 2.8989,
+      "step": 21924
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5310101509094238,
+      "learning_rate": 0.00042929748561748723,
+      "loss": 2.9027,
+      "step": 21925
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5500092506408691,
+      "learning_rate": 0.0004292835926035017,
+      "loss": 3.0154,
+      "step": 21926
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5398533940315247,
+      "learning_rate": 0.00042926969924900806,
+      "loss": 3.1928,
+      "step": 21927
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5290772914886475,
+      "learning_rate": 0.0004292558055540429,
+      "loss": 3.0907,
+      "step": 21928
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.553503692150116,
+      "learning_rate": 0.00042924191151864276,
+      "loss": 3.2309,
+      "step": 21929
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5868078470230103,
+      "learning_rate": 0.00042922801714284425,
+      "loss": 3.244,
+      "step": 21930
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.535077691078186,
+      "learning_rate": 0.0004292141224266842,
+      "loss": 3.3127,
+      "step": 21931
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5569257736206055,
+      "learning_rate": 0.0004292002273701989,
+      "loss": 3.0853,
+      "step": 21932
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5465898513793945,
+      "learning_rate": 0.00042918633197342506,
+      "loss": 3.0265,
+      "step": 21933
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5579783916473389,
+      "learning_rate": 0.00042917243623639934,
+      "loss": 3.087,
+      "step": 21934
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5138366222381592,
+      "learning_rate": 0.00042915854015915825,
+      "loss": 3.2086,
+      "step": 21935
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5494992136955261,
+      "learning_rate": 0.0004291446437417383,
+      "loss": 3.0049,
+      "step": 21936
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5276719927787781,
+      "learning_rate": 0.00042913074698417645,
+      "loss": 3.0952,
+      "step": 21937
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6109719276428223,
+      "learning_rate": 0.00042911684988650894,
+      "loss": 3.0782,
+      "step": 21938
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5517315864562988,
+      "learning_rate": 0.0004291029524487725,
+      "loss": 2.9928,
+      "step": 21939
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5608121752738953,
+      "learning_rate": 0.0004290890546710038,
+      "loss": 3.0408,
+      "step": 21940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5328395366668701,
+      "learning_rate": 0.0004290751565532392,
+      "loss": 3.0053,
+      "step": 21941
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.519293487071991,
+      "learning_rate": 0.0004290612580955156,
+      "loss": 3.1365,
+      "step": 21942
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5312771797180176,
+      "learning_rate": 0.0004290473592978695,
+      "loss": 3.1626,
+      "step": 21943
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5550572276115417,
+      "learning_rate": 0.00042903346016033746,
+      "loss": 2.9873,
+      "step": 21944
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5349480509757996,
+      "learning_rate": 0.0004290195606829561,
+      "loss": 3.0404,
+      "step": 21945
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.53115314245224,
+      "learning_rate": 0.0004290056608657621,
+      "loss": 3.1882,
+      "step": 21946
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5174546837806702,
+      "learning_rate": 0.00042899176070879195,
+      "loss": 3.1557,
+      "step": 21947
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5151444673538208,
+      "learning_rate": 0.0004289778602120824,
+      "loss": 2.9821,
+      "step": 21948
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5512619614601135,
+      "learning_rate": 0.00042896395937567003,
+      "loss": 3.2606,
+      "step": 21949
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5739850997924805,
+      "learning_rate": 0.00042895005819959124,
+      "loss": 3.0911,
+      "step": 21950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5571601390838623,
+      "learning_rate": 0.00042893615668388287,
+      "loss": 3.0645,
+      "step": 21951
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5094681978225708,
+      "learning_rate": 0.00042892225482858145,
+      "loss": 3.1872,
+      "step": 21952
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.531054675579071,
+      "learning_rate": 0.00042890835263372374,
+      "loss": 2.8496,
+      "step": 21953
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5505704283714294,
+      "learning_rate": 0.0004288944500993462,
+      "loss": 3.0071,
+      "step": 21954
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5394859313964844,
+      "learning_rate": 0.0004288805472254855,
+      "loss": 3.2745,
+      "step": 21955
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5463074445724487,
+      "learning_rate": 0.0004288666440121782,
+      "loss": 3.0487,
+      "step": 21956
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5753787159919739,
+      "learning_rate": 0.00042885274045946096,
+      "loss": 3.0229,
+      "step": 21957
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5389484763145447,
+      "learning_rate": 0.0004288388365673704,
+      "loss": 2.9471,
+      "step": 21958
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5314404964447021,
+      "learning_rate": 0.00042882493233594316,
+      "loss": 3.1627,
+      "step": 21959
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5405857563018799,
+      "learning_rate": 0.0004288110277652159,
+      "loss": 3.0673,
+      "step": 21960
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5381196141242981,
+      "learning_rate": 0.0004287971228552251,
+      "loss": 3.1375,
+      "step": 21961
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5330754518508911,
+      "learning_rate": 0.0004287832176060074,
+      "loss": 2.9481,
+      "step": 21962
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5497124791145325,
+      "learning_rate": 0.0004287693120175996,
+      "loss": 3.0511,
+      "step": 21963
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5864923596382141,
+      "learning_rate": 0.0004287554060900382,
+      "loss": 3.1814,
+      "step": 21964
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5531643629074097,
+      "learning_rate": 0.0004287414998233598,
+      "loss": 3.0753,
+      "step": 21965
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5571635961532593,
+      "learning_rate": 0.0004287275932176011,
+      "loss": 2.926,
+      "step": 21966
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5513104796409607,
+      "learning_rate": 0.00042871368627279873,
+      "loss": 3.1882,
+      "step": 21967
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.548090934753418,
+      "learning_rate": 0.0004286997789889892,
+      "loss": 3.265,
+      "step": 21968
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5597220063209534,
+      "learning_rate": 0.00042868587136620927,
+      "loss": 2.9922,
+      "step": 21969
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5412937998771667,
+      "learning_rate": 0.0004286719634044955,
+      "loss": 3.0208,
+      "step": 21970
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5296668410301208,
+      "learning_rate": 0.00042865805510388456,
+      "loss": 3.1406,
+      "step": 21971
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5780396461486816,
+      "learning_rate": 0.000428644146464413,
+      "loss": 3.0601,
+      "step": 21972
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5567464828491211,
+      "learning_rate": 0.00042863023748611757,
+      "loss": 3.2744,
+      "step": 21973
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5450098514556885,
+      "learning_rate": 0.00042861632816903485,
+      "loss": 3.0649,
+      "step": 21974
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5599391460418701,
+      "learning_rate": 0.00042860241851320157,
+      "loss": 3.1379,
+      "step": 21975
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5682699680328369,
+      "learning_rate": 0.00042858850851865414,
+      "loss": 3.0891,
+      "step": 21976
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.578801155090332,
+      "learning_rate": 0.0004285745981854293,
+      "loss": 2.6889,
+      "step": 21977
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5470324754714966,
+      "learning_rate": 0.0004285606875135638,
+      "loss": 3.2862,
+      "step": 21978
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5256801843643188,
+      "learning_rate": 0.00042854677650309416,
+      "loss": 3.3307,
+      "step": 21979
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5400407910346985,
+      "learning_rate": 0.0004285328651540571,
+      "loss": 2.9132,
+      "step": 21980
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5310783982276917,
+      "learning_rate": 0.00042851895346648916,
+      "loss": 3.0173,
+      "step": 21981
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5397918820381165,
+      "learning_rate": 0.00042850504144042697,
+      "loss": 3.1176,
+      "step": 21982
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5340597033500671,
+      "learning_rate": 0.0004284911290759073,
+      "loss": 3.2694,
+      "step": 21983
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5765253901481628,
+      "learning_rate": 0.0004284772163729667,
+      "loss": 2.9625,
+      "step": 21984
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5977330803871155,
+      "learning_rate": 0.0004284633033316419,
+      "loss": 2.8798,
+      "step": 21985
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5429593324661255,
+      "learning_rate": 0.0004284493899519695,
+      "loss": 3.2659,
+      "step": 21986
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5388059020042419,
+      "learning_rate": 0.000428435476233986,
+      "loss": 2.9844,
+      "step": 21987
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5572206377983093,
+      "learning_rate": 0.0004284215621777282,
+      "loss": 3.0551,
+      "step": 21988
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5941150784492493,
+      "learning_rate": 0.0004284076477832328,
+      "loss": 2.8342,
+      "step": 21989
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5501981973648071,
+      "learning_rate": 0.00042839373305053635,
+      "loss": 3.0811,
+      "step": 21990
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5432038903236389,
+      "learning_rate": 0.00042837981797967553,
+      "loss": 3.0867,
+      "step": 21991
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5538283586502075,
+      "learning_rate": 0.00042836590257068686,
+      "loss": 3.3209,
+      "step": 21992
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5830565094947815,
+      "learning_rate": 0.00042835198682360725,
+      "loss": 2.9133,
+      "step": 21993
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5611438155174255,
+      "learning_rate": 0.00042833807073847313,
+      "loss": 2.9194,
+      "step": 21994
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5672524571418762,
+      "learning_rate": 0.0004283241543153212,
+      "loss": 3.0636,
+      "step": 21995
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5321059226989746,
+      "learning_rate": 0.00042831023755418823,
+      "loss": 3.1639,
+      "step": 21996
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.574716329574585,
+      "learning_rate": 0.0004282963204551108,
+      "loss": 3.1508,
+      "step": 21997
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.540168046951294,
+      "learning_rate": 0.0004282824030181254,
+      "loss": 2.825,
+      "step": 21998
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5404049158096313,
+      "learning_rate": 0.00042826848524326906,
+      "loss": 3.2393,
+      "step": 21999
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5294651389122009,
+      "learning_rate": 0.0004282545671305781,
+      "loss": 3.2042,
+      "step": 22000
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5121399760246277,
+      "learning_rate": 0.0004282406486800893,
+      "loss": 3.3916,
+      "step": 22001
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5601375699043274,
+      "learning_rate": 0.0004282267298918392,
+      "loss": 2.9623,
+      "step": 22002
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5114780068397522,
+      "learning_rate": 0.0004282128107658647,
+      "loss": 2.9027,
+      "step": 22003
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5449838638305664,
+      "learning_rate": 0.0004281988913022024,
+      "loss": 3.0159,
+      "step": 22004
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5693271160125732,
+      "learning_rate": 0.0004281849715008888,
+      "loss": 3.162,
+      "step": 22005
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5240155458450317,
+      "learning_rate": 0.0004281710513619606,
+      "loss": 3.1996,
+      "step": 22006
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5236963629722595,
+      "learning_rate": 0.0004281571308854546,
+      "loss": 3.3674,
+      "step": 22007
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.581671953201294,
+      "learning_rate": 0.00042814321007140734,
+      "loss": 3.1306,
+      "step": 22008
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5509883761405945,
+      "learning_rate": 0.00042812928891985556,
+      "loss": 3.1849,
+      "step": 22009
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5407490134239197,
+      "learning_rate": 0.00042811536743083583,
+      "loss": 3.2232,
+      "step": 22010
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5536962151527405,
+      "learning_rate": 0.00042810144560438495,
+      "loss": 2.9887,
+      "step": 22011
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5936238765716553,
+      "learning_rate": 0.0004280875234405395,
+      "loss": 2.943,
+      "step": 22012
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5407906174659729,
+      "learning_rate": 0.0004280736009393361,
+      "loss": 3.271,
+      "step": 22013
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.569972813129425,
+      "learning_rate": 0.00042805967810081156,
+      "loss": 3.0759,
+      "step": 22014
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5183060765266418,
+      "learning_rate": 0.00042804575492500245,
+      "loss": 3.2847,
+      "step": 22015
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5623323917388916,
+      "learning_rate": 0.00042803183141194547,
+      "loss": 3.2323,
+      "step": 22016
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5643849968910217,
+      "learning_rate": 0.00042801790756167733,
+      "loss": 3.0499,
+      "step": 22017
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5582993030548096,
+      "learning_rate": 0.00042800398337423457,
+      "loss": 3.051,
+      "step": 22018
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5692091584205627,
+      "learning_rate": 0.000427990058849654,
+      "loss": 3.0984,
+      "step": 22019
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5723261833190918,
+      "learning_rate": 0.0004279761339879722,
+      "loss": 3.0136,
+      "step": 22020
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5091471672058105,
+      "learning_rate": 0.0004279622087892259,
+      "loss": 3.2192,
+      "step": 22021
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5590563416481018,
+      "learning_rate": 0.0004279482832534519,
+      "loss": 3.0695,
+      "step": 22022
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5491331815719604,
+      "learning_rate": 0.0004279343573806866,
+      "loss": 3.1763,
+      "step": 22023
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5999693274497986,
+      "learning_rate": 0.00042792043117096686,
+      "loss": 2.8919,
+      "step": 22024
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5291171669960022,
+      "learning_rate": 0.0004279065046243293,
+      "loss": 3.404,
+      "step": 22025
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5340095162391663,
+      "learning_rate": 0.00042789257774081073,
+      "loss": 3.2515,
+      "step": 22026
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5588778853416443,
+      "learning_rate": 0.00042787865052044765,
+      "loss": 3.3579,
+      "step": 22027
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5633630156517029,
+      "learning_rate": 0.0004278647229632767,
+      "loss": 3.0803,
+      "step": 22028
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5125625133514404,
+      "learning_rate": 0.0004278507950693348,
+      "loss": 2.9785,
+      "step": 22029
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6125033497810364,
+      "learning_rate": 0.00042783686683865854,
+      "loss": 3.1076,
+      "step": 22030
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5360808372497559,
+      "learning_rate": 0.00042782293827128457,
+      "loss": 3.1738,
+      "step": 22031
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5409636497497559,
+      "learning_rate": 0.00042780900936724957,
+      "loss": 3.0977,
+      "step": 22032
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5472274422645569,
+      "learning_rate": 0.0004277950801265902,
+      "loss": 2.994,
+      "step": 22033
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5469160079956055,
+      "learning_rate": 0.00042778115054934314,
+      "loss": 3.0537,
+      "step": 22034
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5636223554611206,
+      "learning_rate": 0.0004277672206355452,
+      "loss": 3.01,
+      "step": 22035
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6423721313476562,
+      "learning_rate": 0.000427753290385233,
+      "loss": 2.9419,
+      "step": 22036
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.53389573097229,
+      "learning_rate": 0.00042773935979844327,
+      "loss": 3.0792,
+      "step": 22037
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5442113280296326,
+      "learning_rate": 0.00042772542887521247,
+      "loss": 3.0606,
+      "step": 22038
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5777226686477661,
+      "learning_rate": 0.0004277114976155776,
+      "loss": 3.2876,
+      "step": 22039
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5454306602478027,
+      "learning_rate": 0.0004276975660195752,
+      "loss": 2.9339,
+      "step": 22040
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.544465959072113,
+      "learning_rate": 0.000427683634087242,
+      "loss": 3.0272,
+      "step": 22041
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5971094369888306,
+      "learning_rate": 0.00042766970181861466,
+      "loss": 3.0422,
+      "step": 22042
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5495122075080872,
+      "learning_rate": 0.00042765576921372986,
+      "loss": 3.0553,
+      "step": 22043
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5605582594871521,
+      "learning_rate": 0.00042764183627262437,
+      "loss": 3.3398,
+      "step": 22044
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.564033031463623,
+      "learning_rate": 0.00042762790299533484,
+      "loss": 3.0244,
+      "step": 22045
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5470286011695862,
+      "learning_rate": 0.000427613969381898,
+      "loss": 2.9283,
+      "step": 22046
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5006475448608398,
+      "learning_rate": 0.00042760003543235056,
+      "loss": 2.9596,
+      "step": 22047
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5973907709121704,
+      "learning_rate": 0.00042758610114672905,
+      "loss": 2.888,
+      "step": 22048
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5585212111473083,
+      "learning_rate": 0.00042757216652507046,
+      "loss": 3.1172,
+      "step": 22049
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5310104489326477,
+      "learning_rate": 0.00042755823156741127,
+      "loss": 3.0854,
+      "step": 22050
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5287449359893799,
+      "learning_rate": 0.00042754429627378824,
+      "loss": 3.1736,
+      "step": 22051
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5601946711540222,
+      "learning_rate": 0.00042753036064423813,
+      "loss": 3.1034,
+      "step": 22052
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5369007587432861,
+      "learning_rate": 0.00042751642467879746,
+      "loss": 2.9668,
+      "step": 22053
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5678867101669312,
+      "learning_rate": 0.00042750248837750317,
+      "loss": 3.331,
+      "step": 22054
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5435790419578552,
+      "learning_rate": 0.0004274885517403919,
+      "loss": 2.823,
+      "step": 22055
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4990702271461487,
+      "learning_rate": 0.00042747461476750026,
+      "loss": 2.8055,
+      "step": 22056
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.551609218120575,
+      "learning_rate": 0.000427460677458865,
+      "loss": 3.0637,
+      "step": 22057
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5618946552276611,
+      "learning_rate": 0.00042744673981452285,
+      "loss": 3.0707,
+      "step": 22058
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5400345921516418,
+      "learning_rate": 0.0004274328018345105,
+      "loss": 2.8357,
+      "step": 22059
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5235921740531921,
+      "learning_rate": 0.0004274188635188647,
+      "loss": 3.0765,
+      "step": 22060
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5762670636177063,
+      "learning_rate": 0.00042740492486762215,
+      "loss": 3.2314,
+      "step": 22061
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5849509239196777,
+      "learning_rate": 0.0004273909858808196,
+      "loss": 2.9629,
+      "step": 22062
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5751528143882751,
+      "learning_rate": 0.0004273770465584937,
+      "loss": 3.1594,
+      "step": 22063
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5581267476081848,
+      "learning_rate": 0.00042736310690068106,
+      "loss": 3.0604,
+      "step": 22064
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6129317283630371,
+      "learning_rate": 0.00042734916690741854,
+      "loss": 3.1213,
+      "step": 22065
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5418512225151062,
+      "learning_rate": 0.00042733522657874295,
+      "loss": 3.0227,
+      "step": 22066
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5475884079933167,
+      "learning_rate": 0.0004273212859146908,
+      "loss": 3.254,
+      "step": 22067
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5630029439926147,
+      "learning_rate": 0.00042730734491529883,
+      "loss": 3.1402,
+      "step": 22068
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5417295098304749,
+      "learning_rate": 0.00042729340358060386,
+      "loss": 3.2479,
+      "step": 22069
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5166475772857666,
+      "learning_rate": 0.0004272794619106426,
+      "loss": 3.123,
+      "step": 22070
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.554018497467041,
+      "learning_rate": 0.00042726551990545167,
+      "loss": 3.0511,
+      "step": 22071
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.520998477935791,
+      "learning_rate": 0.00042725157756506784,
+      "loss": 2.9895,
+      "step": 22072
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5566690564155579,
+      "learning_rate": 0.00042723763488952796,
+      "loss": 3.1612,
+      "step": 22073
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5547829270362854,
+      "learning_rate": 0.0004272236918788685,
+      "loss": 3.1395,
+      "step": 22074
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6837734580039978,
+      "learning_rate": 0.00042720974853312633,
+      "loss": 3.0537,
+      "step": 22075
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6061183214187622,
+      "learning_rate": 0.00042719580485233826,
+      "loss": 2.9883,
+      "step": 22076
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5670740604400635,
+      "learning_rate": 0.00042718186083654094,
+      "loss": 3.0004,
+      "step": 22077
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5454275608062744,
+      "learning_rate": 0.000427167916485771,
+      "loss": 3.0537,
+      "step": 22078
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5352046489715576,
+      "learning_rate": 0.00042715397180006516,
+      "loss": 3.1883,
+      "step": 22079
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5673153400421143,
+      "learning_rate": 0.0004271400267794604,
+      "loss": 3.106,
+      "step": 22080
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5358964800834656,
+      "learning_rate": 0.0004271260814239932,
+      "loss": 3.0141,
+      "step": 22081
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6573508977890015,
+      "learning_rate": 0.0004271121357337003,
+      "loss": 3.0543,
+      "step": 22082
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5617774128913879,
+      "learning_rate": 0.0004270981897086186,
+      "loss": 3.0414,
+      "step": 22083
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5869306921958923,
+      "learning_rate": 0.0004270842433487847,
+      "loss": 2.9109,
+      "step": 22084
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5464381575584412,
+      "learning_rate": 0.00042707029665423535,
+      "loss": 2.9973,
+      "step": 22085
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5536450147628784,
+      "learning_rate": 0.00042705634962500726,
+      "loss": 3.1901,
+      "step": 22086
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5760237574577332,
+      "learning_rate": 0.00042704240226113725,
+      "loss": 3.2899,
+      "step": 22087
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5971628427505493,
+      "learning_rate": 0.00042702845456266207,
+      "loss": 2.9888,
+      "step": 22088
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5124607682228088,
+      "learning_rate": 0.0004270145065296183,
+      "loss": 3.1622,
+      "step": 22089
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5447020530700684,
+      "learning_rate": 0.0004270005581620427,
+      "loss": 3.1413,
+      "step": 22090
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5350099802017212,
+      "learning_rate": 0.00042698660945997215,
+      "loss": 3.0142,
+      "step": 22091
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5177395343780518,
+      "learning_rate": 0.0004269726604234433,
+      "loss": 3.2439,
+      "step": 22092
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5653977990150452,
+      "learning_rate": 0.0004269587110524929,
+      "loss": 3.0023,
+      "step": 22093
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5047891736030579,
+      "learning_rate": 0.0004269447613471577,
+      "loss": 2.987,
+      "step": 22094
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.540993869304657,
+      "learning_rate": 0.00042693081130747444,
+      "loss": 3.1142,
+      "step": 22095
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5562851428985596,
+      "learning_rate": 0.0004269168609334798,
+      "loss": 2.9607,
+      "step": 22096
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5496192574501038,
+      "learning_rate": 0.0004269029102252106,
+      "loss": 3.0992,
+      "step": 22097
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5273379683494568,
+      "learning_rate": 0.0004268889591827036,
+      "loss": 2.926,
+      "step": 22098
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5540835857391357,
+      "learning_rate": 0.00042687500780599544,
+      "loss": 3.0501,
+      "step": 22099
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5160466432571411,
+      "learning_rate": 0.0004268610560951229,
+      "loss": 3.152,
+      "step": 22100
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5677379965782166,
+      "learning_rate": 0.0004268471040501228,
+      "loss": 3.0895,
+      "step": 22101
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5393813848495483,
+      "learning_rate": 0.00042683315167103196,
+      "loss": 3.0357,
+      "step": 22102
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5417753458023071,
+      "learning_rate": 0.00042681919895788683,
+      "loss": 3.1901,
+      "step": 22103
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4963914155960083,
+      "learning_rate": 0.0004268052459107244,
+      "loss": 3.3258,
+      "step": 22104
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5311129689216614,
+      "learning_rate": 0.0004267912925295813,
+      "loss": 3.1505,
+      "step": 22105
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5351983904838562,
+      "learning_rate": 0.00042677733881449446,
+      "loss": 3.0273,
+      "step": 22106
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5361194610595703,
+      "learning_rate": 0.0004267633847655004,
+      "loss": 2.9613,
+      "step": 22107
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5353830456733704,
+      "learning_rate": 0.000426749430382636,
+      "loss": 3.164,
+      "step": 22108
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5087656378746033,
+      "learning_rate": 0.0004267354756659381,
+      "loss": 3.2214,
+      "step": 22109
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5443871021270752,
+      "learning_rate": 0.00042672152061544314,
+      "loss": 3.043,
+      "step": 22110
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.567270815372467,
+      "learning_rate": 0.0004267075652311882,
+      "loss": 3.1497,
+      "step": 22111
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.520046055316925,
+      "learning_rate": 0.00042669360951321,
+      "loss": 3.1484,
+      "step": 22112
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5559588670730591,
+      "learning_rate": 0.00042667965346154514,
+      "loss": 3.0016,
+      "step": 22113
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5984733700752258,
+      "learning_rate": 0.0004266656970762305,
+      "loss": 3.0313,
+      "step": 22114
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5269164443016052,
+      "learning_rate": 0.0004266517403573026,
+      "loss": 3.2197,
+      "step": 22115
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6935906410217285,
+      "learning_rate": 0.00042663778330479863,
+      "loss": 3.2497,
+      "step": 22116
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5183548927307129,
+      "learning_rate": 0.000426623825918755,
+      "loss": 3.3231,
+      "step": 22117
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5844008326530457,
+      "learning_rate": 0.0004266098681992086,
+      "loss": 3.2325,
+      "step": 22118
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4959857165813446,
+      "learning_rate": 0.0004265959101461962,
+      "loss": 3.2061,
+      "step": 22119
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5572559833526611,
+      "learning_rate": 0.0004265819517597545,
+      "loss": 2.9519,
+      "step": 22120
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5319798588752747,
+      "learning_rate": 0.0004265679930399203,
+      "loss": 3.1005,
+      "step": 22121
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5380662679672241,
+      "learning_rate": 0.00042655403398673036,
+      "loss": 3.1476,
+      "step": 22122
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5653159618377686,
+      "learning_rate": 0.0004265400746002214,
+      "loss": 3.0577,
+      "step": 22123
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5704825520515442,
+      "learning_rate": 0.0004265261148804304,
+      "loss": 3.1806,
+      "step": 22124
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5176539421081543,
+      "learning_rate": 0.0004265121548273938,
+      "loss": 3.1776,
+      "step": 22125
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5719426870346069,
+      "learning_rate": 0.0004264981944411486,
+      "loss": 3.0773,
+      "step": 22126
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.48471060395240784,
+      "learning_rate": 0.0004264842337217315,
+      "loss": 3.0145,
+      "step": 22127
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.515109121799469,
+      "learning_rate": 0.00042647027266917926,
+      "loss": 3.2385,
+      "step": 22128
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5453347563743591,
+      "learning_rate": 0.0004264563112835287,
+      "loss": 3.0087,
+      "step": 22129
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5508375763893127,
+      "learning_rate": 0.00042644234956481646,
+      "loss": 3.0716,
+      "step": 22130
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5232502818107605,
+      "learning_rate": 0.00042642838751307953,
+      "loss": 3.2418,
+      "step": 22131
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5501592755317688,
+      "learning_rate": 0.00042641442512835446,
+      "loss": 3.0997,
+      "step": 22132
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.54051673412323,
+      "learning_rate": 0.00042640046241067817,
+      "loss": 3.012,
+      "step": 22133
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.52208411693573,
+      "learning_rate": 0.00042638649936008736,
+      "loss": 3.0246,
+      "step": 22134
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6216174364089966,
+      "learning_rate": 0.0004263725359766189,
+      "loss": 3.0755,
+      "step": 22135
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5394618511199951,
+      "learning_rate": 0.00042635857226030934,
+      "loss": 3.1135,
+      "step": 22136
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5192962288856506,
+      "learning_rate": 0.00042634460821119576,
+      "loss": 3.096,
+      "step": 22137
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5286705493927002,
+      "learning_rate": 0.0004263306438293148,
+      "loss": 3.0381,
+      "step": 22138
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5518622398376465,
+      "learning_rate": 0.0004263166791147032,
+      "loss": 3.0308,
+      "step": 22139
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5467196702957153,
+      "learning_rate": 0.0004263027140673978,
+      "loss": 3.172,
+      "step": 22140
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5387907028198242,
+      "learning_rate": 0.00042628874868743527,
+      "loss": 3.3111,
+      "step": 22141
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5448318719863892,
+      "learning_rate": 0.00042627478297485263,
+      "loss": 2.9398,
+      "step": 22142
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.57187819480896,
+      "learning_rate": 0.0004262608169296864,
+      "loss": 3.2116,
+      "step": 22143
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6088796257972717,
+      "learning_rate": 0.00042624685055197355,
+      "loss": 3.0892,
+      "step": 22144
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5294945240020752,
+      "learning_rate": 0.00042623288384175073,
+      "loss": 3.0999,
+      "step": 22145
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5282583236694336,
+      "learning_rate": 0.00042621891679905477,
+      "loss": 3.2845,
+      "step": 22146
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.556050181388855,
+      "learning_rate": 0.00042620494942392247,
+      "loss": 3.1564,
+      "step": 22147
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5315901041030884,
+      "learning_rate": 0.0004261909817163907,
+      "loss": 3.0353,
+      "step": 22148
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5163993239402771,
+      "learning_rate": 0.00042617701367649616,
+      "loss": 3.0004,
+      "step": 22149
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5579254031181335,
+      "learning_rate": 0.00042616304530427565,
+      "loss": 3.0205,
+      "step": 22150
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5503444671630859,
+      "learning_rate": 0.00042614907659976593,
+      "loss": 3.187,
+      "step": 22151
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5766018033027649,
+      "learning_rate": 0.0004261351075630038,
+      "loss": 3.2087,
+      "step": 22152
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5432386994361877,
+      "learning_rate": 0.00042612113819402616,
+      "loss": 3.0861,
+      "step": 22153
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5107462406158447,
+      "learning_rate": 0.0004261071684928696,
+      "loss": 2.9946,
+      "step": 22154
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5693354606628418,
+      "learning_rate": 0.00042609319845957115,
+      "loss": 2.9644,
+      "step": 22155
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6248636841773987,
+      "learning_rate": 0.00042607922809416736,
+      "loss": 3.1401,
+      "step": 22156
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5272868871688843,
+      "learning_rate": 0.0004260652573966952,
+      "loss": 2.98,
+      "step": 22157
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5685650706291199,
+      "learning_rate": 0.00042605128636719145,
+      "loss": 2.7644,
+      "step": 22158
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5144070982933044,
+      "learning_rate": 0.0004260373150056928,
+      "loss": 3.078,
+      "step": 22159
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5530477166175842,
+      "learning_rate": 0.00042602334331223615,
+      "loss": 3.1147,
+      "step": 22160
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5493810772895813,
+      "learning_rate": 0.0004260093712868583,
+      "loss": 3.2218,
+      "step": 22161
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5624595284461975,
+      "learning_rate": 0.00042599539892959593,
+      "loss": 3.3188,
+      "step": 22162
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5464377999305725,
+      "learning_rate": 0.00042598142624048597,
+      "loss": 3.1378,
+      "step": 22163
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5302308201789856,
+      "learning_rate": 0.00042596745321956525,
+      "loss": 3.4019,
+      "step": 22164
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5592415928840637,
+      "learning_rate": 0.0004259534798668705,
+      "loss": 2.9761,
+      "step": 22165
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6183668971061707,
+      "learning_rate": 0.0004259395061824384,
+      "loss": 2.8968,
+      "step": 22166
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5616676807403564,
+      "learning_rate": 0.00042592553216630593,
+      "loss": 3.13,
+      "step": 22167
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5477641820907593,
+      "learning_rate": 0.0004259115578185099,
+      "loss": 3.1877,
+      "step": 22168
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5532886385917664,
+      "learning_rate": 0.000425897583139087,
+      "loss": 2.7935,
+      "step": 22169
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5679529309272766,
+      "learning_rate": 0.00042588360812807416,
+      "loss": 3.2103,
+      "step": 22170
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5274649262428284,
+      "learning_rate": 0.00042586963278550807,
+      "loss": 2.974,
+      "step": 22171
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5478577613830566,
+      "learning_rate": 0.0004258556571114256,
+      "loss": 3.0606,
+      "step": 22172
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5519079566001892,
+      "learning_rate": 0.00042584168110586354,
+      "loss": 3.1437,
+      "step": 22173
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5189817547798157,
+      "learning_rate": 0.00042582770476885877,
+      "loss": 3.0276,
+      "step": 22174
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.544884979724884,
+      "learning_rate": 0.00042581372810044797,
+      "loss": 3.0585,
+      "step": 22175
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5363135933876038,
+      "learning_rate": 0.0004257997511006681,
+      "loss": 3.2155,
+      "step": 22176
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5269807577133179,
+      "learning_rate": 0.0004257857737695558,
+      "loss": 3.0673,
+      "step": 22177
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.535571813583374,
+      "learning_rate": 0.00042577179610714795,
+      "loss": 3.3365,
+      "step": 22178
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6153188943862915,
+      "learning_rate": 0.00042575781811348155,
+      "loss": 2.9663,
+      "step": 22179
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5156885385513306,
+      "learning_rate": 0.0004257438397885932,
+      "loss": 3.134,
+      "step": 22180
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5604032874107361,
+      "learning_rate": 0.00042572986113251975,
+      "loss": 3.2063,
+      "step": 22181
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5438750386238098,
+      "learning_rate": 0.000425715882145298,
+      "loss": 3.0467,
+      "step": 22182
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5508151054382324,
+      "learning_rate": 0.00042570190282696484,
+      "loss": 3.2381,
+      "step": 22183
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7027482986450195,
+      "learning_rate": 0.0004256879231775571,
+      "loss": 3.0696,
+      "step": 22184
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5488677620887756,
+      "learning_rate": 0.00042567394319711156,
+      "loss": 2.7966,
+      "step": 22185
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6230414509773254,
+      "learning_rate": 0.000425659962885665,
+      "loss": 3.157,
+      "step": 22186
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5432460308074951,
+      "learning_rate": 0.0004256459822432543,
+      "loss": 3.2616,
+      "step": 22187
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.520793616771698,
+      "learning_rate": 0.00042563200126991625,
+      "loss": 3.0267,
+      "step": 22188
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.533034086227417,
+      "learning_rate": 0.0004256180199656877,
+      "loss": 3.0244,
+      "step": 22189
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5230249762535095,
+      "learning_rate": 0.0004256040383306055,
+      "loss": 3.1633,
+      "step": 22190
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5286110639572144,
+      "learning_rate": 0.00042559005636470636,
+      "loss": 3.0445,
+      "step": 22191
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.541115403175354,
+      "learning_rate": 0.00042557607406802715,
+      "loss": 3.1721,
+      "step": 22192
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5243132710456848,
+      "learning_rate": 0.0004255620914406048,
+      "loss": 3.2148,
+      "step": 22193
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5529786944389343,
+      "learning_rate": 0.00042554810848247613,
+      "loss": 3.1431,
+      "step": 22194
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5576296448707581,
+      "learning_rate": 0.00042553412519367783,
+      "loss": 3.0609,
+      "step": 22195
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5397801995277405,
+      "learning_rate": 0.00042552014157424677,
+      "loss": 2.8357,
+      "step": 22196
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5517771244049072,
+      "learning_rate": 0.00042550615762421986,
+      "loss": 3.21,
+      "step": 22197
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5241683721542358,
+      "learning_rate": 0.0004254921733436339,
+      "loss": 3.1942,
+      "step": 22198
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5698009729385376,
+      "learning_rate": 0.00042547818873252567,
+      "loss": 3.1291,
+      "step": 22199
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5370516777038574,
+      "learning_rate": 0.000425464203790932,
+      "loss": 3.3494,
+      "step": 22200
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6023409962654114,
+      "learning_rate": 0.0004254502185188899,
+      "loss": 3.1343,
+      "step": 22201
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5534052848815918,
+      "learning_rate": 0.000425436232916436,
+      "loss": 3.1015,
+      "step": 22202
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5142847299575806,
+      "learning_rate": 0.00042542224698360713,
+      "loss": 2.9254,
+      "step": 22203
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5432417392730713,
+      "learning_rate": 0.00042540826072044036,
+      "loss": 3.1449,
+      "step": 22204
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5595294833183289,
+      "learning_rate": 0.0004253942741269722,
+      "loss": 3.3928,
+      "step": 22205
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5585522651672363,
+      "learning_rate": 0.00042538028720323976,
+      "loss": 3.2472,
+      "step": 22206
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6130068302154541,
+      "learning_rate": 0.0004253662999492797,
+      "loss": 3.0106,
+      "step": 22207
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5520464777946472,
+      "learning_rate": 0.000425352312365129,
+      "loss": 3.2156,
+      "step": 22208
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5398738384246826,
+      "learning_rate": 0.0004253383244508243,
+      "loss": 3.2174,
+      "step": 22209
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5600001215934753,
+      "learning_rate": 0.00042532433620640276,
+      "loss": 3.0729,
+      "step": 22210
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5530575513839722,
+      "learning_rate": 0.00042531034763190095,
+      "loss": 3.0908,
+      "step": 22211
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.550061047077179,
+      "learning_rate": 0.0004252963587273558,
+      "loss": 3.0893,
+      "step": 22212
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5559527277946472,
+      "learning_rate": 0.0004252823694928041,
+      "loss": 2.8335,
+      "step": 22213
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7743620276451111,
+      "learning_rate": 0.00042526837992828287,
+      "loss": 3.1083,
+      "step": 22214
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5912230610847473,
+      "learning_rate": 0.0004252543900338288,
+      "loss": 3.0625,
+      "step": 22215
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5378463864326477,
+      "learning_rate": 0.0004252403998094787,
+      "loss": 3.0251,
+      "step": 22216
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.563388466835022,
+      "learning_rate": 0.0004252264092552696,
+      "loss": 3.153,
+      "step": 22217
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5485290288925171,
+      "learning_rate": 0.0004252124183712381,
+      "loss": 3.3472,
+      "step": 22218
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5503448247909546,
+      "learning_rate": 0.0004251984271574213,
+      "loss": 3.1022,
+      "step": 22219
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5638839602470398,
+      "learning_rate": 0.0004251844356138559,
+      "loss": 3.016,
+      "step": 22220
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5732755064964294,
+      "learning_rate": 0.0004251704437405788,
+      "loss": 2.9927,
+      "step": 22221
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5192015171051025,
+      "learning_rate": 0.00042515645153762683,
+      "loss": 3.2112,
+      "step": 22222
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.49721595644950867,
+      "learning_rate": 0.0004251424590050368,
+      "loss": 3.0738,
+      "step": 22223
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5540822148323059,
+      "learning_rate": 0.00042512846614284566,
+      "loss": 3.1562,
+      "step": 22224
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5434120893478394,
+      "learning_rate": 0.0004251144729510903,
+      "loss": 3.0941,
+      "step": 22225
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5510311722755432,
+      "learning_rate": 0.0004251004794298074,
+      "loss": 2.973,
+      "step": 22226
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5258462429046631,
+      "learning_rate": 0.000425086485579034,
+      "loss": 2.8038,
+      "step": 22227
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5398142337799072,
+      "learning_rate": 0.0004250724913988067,
+      "loss": 3.1084,
+      "step": 22228
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5346091985702515,
+      "learning_rate": 0.00042505849688916265,
+      "loss": 2.9255,
+      "step": 22229
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5627778172492981,
+      "learning_rate": 0.0004250445020501387,
+      "loss": 3.1123,
+      "step": 22230
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5590901970863342,
+      "learning_rate": 0.0004250305068817715,
+      "loss": 3.1776,
+      "step": 22231
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5717828869819641,
+      "learning_rate": 0.0004250165113840979,
+      "loss": 3.1254,
+      "step": 22232
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5412927269935608,
+      "learning_rate": 0.00042500251555715497,
+      "loss": 3.1597,
+      "step": 22233
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5460436344146729,
+      "learning_rate": 0.0004249885194009794,
+      "loss": 3.062,
+      "step": 22234
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5982441306114197,
+      "learning_rate": 0.0004249745229156082,
+      "loss": 3.0677,
+      "step": 22235
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6243758201599121,
+      "learning_rate": 0.00042496052610107817,
+      "loss": 3.0814,
+      "step": 22236
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5570179224014282,
+      "learning_rate": 0.00042494652895742614,
+      "loss": 3.1024,
+      "step": 22237
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5606381297111511,
+      "learning_rate": 0.00042493253148468893,
+      "loss": 3.1431,
+      "step": 22238
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5624200701713562,
+      "learning_rate": 0.0004249185336829035,
+      "loss": 3.0207,
+      "step": 22239
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5409414768218994,
+      "learning_rate": 0.00042490453555210677,
+      "loss": 3.0627,
+      "step": 22240
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5465520024299622,
+      "learning_rate": 0.00042489053709233554,
+      "loss": 3.1536,
+      "step": 22241
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5893270373344421,
+      "learning_rate": 0.00042487653830362664,
+      "loss": 3.2021,
+      "step": 22242
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5436384081840515,
+      "learning_rate": 0.0004248625391860168,
+      "loss": 3.0846,
+      "step": 22243
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5279334187507629,
+      "learning_rate": 0.00042484853973954334,
+      "loss": 2.9912,
+      "step": 22244
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5567160844802856,
+      "learning_rate": 0.00042483453996424266,
+      "loss": 3.0722,
+      "step": 22245
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5820968151092529,
+      "learning_rate": 0.00042482053986015193,
+      "loss": 3.2603,
+      "step": 22246
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5296043753623962,
+      "learning_rate": 0.00042480653942730785,
+      "loss": 2.8998,
+      "step": 22247
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5903947353363037,
+      "learning_rate": 0.0004247925386657475,
+      "loss": 2.9915,
+      "step": 22248
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5724281072616577,
+      "learning_rate": 0.00042477853757550744,
+      "loss": 3.0855,
+      "step": 22249
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5451366305351257,
+      "learning_rate": 0.0004247645361566248,
+      "loss": 3.2043,
+      "step": 22250
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5273775458335876,
+      "learning_rate": 0.0004247505344091364,
+      "loss": 3.0935,
+      "step": 22251
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5182363390922546,
+      "learning_rate": 0.00042473653233307914,
+      "loss": 3.0011,
+      "step": 22252
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5897952914237976,
+      "learning_rate": 0.0004247225299284898,
+      "loss": 3.0319,
+      "step": 22253
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5534531474113464,
+      "learning_rate": 0.00042470852719540526,
+      "loss": 3.067,
+      "step": 22254
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6196662187576294,
+      "learning_rate": 0.0004246945241338626,
+      "loss": 3.1267,
+      "step": 22255
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5049815773963928,
+      "learning_rate": 0.0004246805207438985,
+      "loss": 3.0543,
+      "step": 22256
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5456374287605286,
+      "learning_rate": 0.00042466651702554984,
+      "loss": 2.9115,
+      "step": 22257
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5522916913032532,
+      "learning_rate": 0.00042465251297885365,
+      "loss": 3.1055,
+      "step": 22258
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5380954146385193,
+      "learning_rate": 0.0004246385086038467,
+      "loss": 2.957,
+      "step": 22259
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5323505997657776,
+      "learning_rate": 0.00042462450390056593,
+      "loss": 3.18,
+      "step": 22260
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5911219120025635,
+      "learning_rate": 0.0004246104988690481,
+      "loss": 3.0147,
+      "step": 22261
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5536802411079407,
+      "learning_rate": 0.00042459649350933026,
+      "loss": 2.9749,
+      "step": 22262
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5558472275733948,
+      "learning_rate": 0.00042458248782144937,
+      "loss": 3.0357,
+      "step": 22263
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5627835988998413,
+      "learning_rate": 0.00042456848180544196,
+      "loss": 3.1438,
+      "step": 22264
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5418185591697693,
+      "learning_rate": 0.00042455447546134526,
+      "loss": 3.2752,
+      "step": 22265
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5398939251899719,
+      "learning_rate": 0.0004245404687891961,
+      "loss": 3.2987,
+      "step": 22266
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5812855362892151,
+      "learning_rate": 0.0004245264617890312,
+      "loss": 3.2357,
+      "step": 22267
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5603014826774597,
+      "learning_rate": 0.00042451245446088764,
+      "loss": 3.1168,
+      "step": 22268
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5525497794151306,
+      "learning_rate": 0.00042449844680480213,
+      "loss": 3.0883,
+      "step": 22269
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5205212235450745,
+      "learning_rate": 0.0004244844388208118,
+      "loss": 2.9792,
+      "step": 22270
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5575414896011353,
+      "learning_rate": 0.00042447043050895334,
+      "loss": 3.0944,
+      "step": 22271
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5364728569984436,
+      "learning_rate": 0.00042445642186926373,
+      "loss": 2.9596,
+      "step": 22272
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5745815634727478,
+      "learning_rate": 0.0004244424129017799,
+      "loss": 3.0151,
+      "step": 22273
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5751864314079285,
+      "learning_rate": 0.0004244284036065387,
+      "loss": 3.1432,
+      "step": 22274
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5519656538963318,
+      "learning_rate": 0.00042441439398357696,
+      "loss": 3.0842,
+      "step": 22275
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5192199945449829,
+      "learning_rate": 0.0004244003840329317,
+      "loss": 3.051,
+      "step": 22276
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5178653597831726,
+      "learning_rate": 0.00042438637375463977,
+      "loss": 2.9998,
+      "step": 22277
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5251026153564453,
+      "learning_rate": 0.00042437236314873814,
+      "loss": 3.2947,
+      "step": 22278
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5604255199432373,
+      "learning_rate": 0.00042435835221526347,
+      "loss": 3.0916,
+      "step": 22279
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5459524393081665,
+      "learning_rate": 0.000424344340954253,
+      "loss": 2.81,
+      "step": 22280
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5570917725563049,
+      "learning_rate": 0.0004243303293657434,
+      "loss": 2.95,
+      "step": 22281
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5491553544998169,
+      "learning_rate": 0.00042431631744977165,
+      "loss": 2.8636,
+      "step": 22282
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5401395559310913,
+      "learning_rate": 0.0004243023052063746,
+      "loss": 3.0923,
+      "step": 22283
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.619458019733429,
+      "learning_rate": 0.00042428829263558925,
+      "loss": 3.1343,
+      "step": 22284
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6060642600059509,
+      "learning_rate": 0.0004242742797374524,
+      "loss": 2.9652,
+      "step": 22285
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5287819504737854,
+      "learning_rate": 0.0004242602665120011,
+      "loss": 2.9648,
+      "step": 22286
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5329878330230713,
+      "learning_rate": 0.0004242462529592721,
+      "loss": 3.0809,
+      "step": 22287
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5269094705581665,
+      "learning_rate": 0.00042423223907930244,
+      "loss": 3.151,
+      "step": 22288
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5310917496681213,
+      "learning_rate": 0.00042421822487212897,
+      "loss": 3.0105,
+      "step": 22289
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5504202246665955,
+      "learning_rate": 0.0004242042103377885,
+      "loss": 3.1283,
+      "step": 22290
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5809419751167297,
+      "learning_rate": 0.00042419019547631815,
+      "loss": 3.0451,
+      "step": 22291
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5401477217674255,
+      "learning_rate": 0.0004241761802877547,
+      "loss": 2.8967,
+      "step": 22292
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5379841327667236,
+      "learning_rate": 0.00042416216477213507,
+      "loss": 3.0813,
+      "step": 22293
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5749803781509399,
+      "learning_rate": 0.0004241481489294962,
+      "loss": 3.0934,
+      "step": 22294
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5080057978630066,
+      "learning_rate": 0.000424134132759875,
+      "loss": 3.249,
+      "step": 22295
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5740684270858765,
+      "learning_rate": 0.0004241201162633084,
+      "loss": 3.2469,
+      "step": 22296
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6056188344955444,
+      "learning_rate": 0.0004241060994398333,
+      "loss": 3.118,
+      "step": 22297
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5314207673072815,
+      "learning_rate": 0.0004240920822894866,
+      "loss": 3.1676,
+      "step": 22298
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5285966992378235,
+      "learning_rate": 0.00042407806481230525,
+      "loss": 3.1967,
+      "step": 22299
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5155075192451477,
+      "learning_rate": 0.0004240640470083261,
+      "loss": 3.0571,
+      "step": 22300
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5423807501792908,
+      "learning_rate": 0.0004240500288775861,
+      "loss": 2.8794,
+      "step": 22301
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5311214327812195,
+      "learning_rate": 0.00042403601042012237,
+      "loss": 3.1135,
+      "step": 22302
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.530709981918335,
+      "learning_rate": 0.00042402199163597164,
+      "loss": 3.0584,
+      "step": 22303
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5539402961730957,
+      "learning_rate": 0.00042400797252517075,
+      "loss": 3.2242,
+      "step": 22304
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5307897925376892,
+      "learning_rate": 0.0004239939530877567,
+      "loss": 2.8867,
+      "step": 22305
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5629311800003052,
+      "learning_rate": 0.0004239799333237666,
+      "loss": 3.2001,
+      "step": 22306
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.533023476600647,
+      "learning_rate": 0.000423965913233237,
+      "loss": 2.9562,
+      "step": 22307
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5523614287376404,
+      "learning_rate": 0.0004239518928162052,
+      "loss": 3.0158,
+      "step": 22308
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5242599844932556,
+      "learning_rate": 0.00042393787207270793,
+      "loss": 3.0386,
+      "step": 22309
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5767312049865723,
+      "learning_rate": 0.0004239238510027821,
+      "loss": 3.1031,
+      "step": 22310
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5401198863983154,
+      "learning_rate": 0.0004239098296064648,
+      "loss": 3.0174,
+      "step": 22311
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5361244082450867,
+      "learning_rate": 0.0004238958078837927,
+      "loss": 3.1997,
+      "step": 22312
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5413596630096436,
+      "learning_rate": 0.000423881785834803,
+      "loss": 3.0465,
+      "step": 22313
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5148143172264099,
+      "learning_rate": 0.0004238677634595326,
+      "loss": 3.1441,
+      "step": 22314
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5160139203071594,
+      "learning_rate": 0.0004238537407580182,
+      "loss": 3.0543,
+      "step": 22315
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5179514288902283,
+      "learning_rate": 0.0004238397177302969,
+      "loss": 3.2671,
+      "step": 22316
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5649704337120056,
+      "learning_rate": 0.00042382569437640573,
+      "loss": 3.1858,
+      "step": 22317
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5431568622589111,
+      "learning_rate": 0.0004238116706963814,
+      "loss": 3.3245,
+      "step": 22318
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6058350205421448,
+      "learning_rate": 0.00042379764669026103,
+      "loss": 3.1804,
+      "step": 22319
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5485496520996094,
+      "learning_rate": 0.0004237836223580814,
+      "loss": 3.1041,
+      "step": 22320
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5472593903541565,
+      "learning_rate": 0.00042376959769987966,
+      "loss": 3.1254,
+      "step": 22321
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5176267027854919,
+      "learning_rate": 0.00042375557271569247,
+      "loss": 3.2074,
+      "step": 22322
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5529015064239502,
+      "learning_rate": 0.000423741547405557,
+      "loss": 3.1134,
+      "step": 22323
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5462213158607483,
+      "learning_rate": 0.00042372752176951013,
+      "loss": 3.1059,
+      "step": 22324
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5348901152610779,
+      "learning_rate": 0.0004237134958075887,
+      "loss": 3.1225,
+      "step": 22325
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6045992374420166,
+      "learning_rate": 0.0004236994695198298,
+      "loss": 3.0944,
+      "step": 22326
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5294902920722961,
+      "learning_rate": 0.0004236854429062703,
+      "loss": 3.1684,
+      "step": 22327
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5004157423973083,
+      "learning_rate": 0.00042367141596694717,
+      "loss": 2.921,
+      "step": 22328
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5467502474784851,
+      "learning_rate": 0.00042365738870189725,
+      "loss": 3.0567,
+      "step": 22329
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5614712238311768,
+      "learning_rate": 0.0004236433611111576,
+      "loss": 3.165,
+      "step": 22330
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5636160373687744,
+      "learning_rate": 0.0004236293331947651,
+      "loss": 2.931,
+      "step": 22331
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.510614275932312,
+      "learning_rate": 0.0004236153049527568,
+      "loss": 2.879,
+      "step": 22332
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5611421465873718,
+      "learning_rate": 0.00042360127638516954,
+      "loss": 3.0598,
+      "step": 22333
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5256950855255127,
+      "learning_rate": 0.0004235872474920403,
+      "loss": 2.9983,
+      "step": 22334
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.528252899646759,
+      "learning_rate": 0.00042357321827340603,
+      "loss": 3.2655,
+      "step": 22335
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.528097927570343,
+      "learning_rate": 0.0004235591887293037,
+      "loss": 2.9888,
+      "step": 22336
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4826788008213043,
+      "learning_rate": 0.0004235451588597702,
+      "loss": 3.2171,
+      "step": 22337
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5288520455360413,
+      "learning_rate": 0.00042353112866484266,
+      "loss": 3.1955,
+      "step": 22338
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5336189866065979,
+      "learning_rate": 0.0004235170981445579,
+      "loss": 3.0057,
+      "step": 22339
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5776178240776062,
+      "learning_rate": 0.0004235030672989527,
+      "loss": 2.89,
+      "step": 22340
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.533409059047699,
+      "learning_rate": 0.00042348903612806423,
+      "loss": 3.0783,
+      "step": 22341
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.547615647315979,
+      "learning_rate": 0.0004234750046319294,
+      "loss": 3.1035,
+      "step": 22342
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5494769215583801,
+      "learning_rate": 0.00042346097281058536,
+      "loss": 3.252,
+      "step": 22343
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.570833146572113,
+      "learning_rate": 0.0004234469406640687,
+      "loss": 3.049,
+      "step": 22344
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5642995834350586,
+      "learning_rate": 0.0004234329081924166,
+      "loss": 3.1304,
+      "step": 22345
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5693486332893372,
+      "learning_rate": 0.00042341887539566595,
+      "loss": 2.8819,
+      "step": 22346
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5467201471328735,
+      "learning_rate": 0.0004234048422738538,
+      "loss": 3.0296,
+      "step": 22347
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5357093811035156,
+      "learning_rate": 0.00042339080882701697,
+      "loss": 3.0394,
+      "step": 22348
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5732297301292419,
+      "learning_rate": 0.00042337677505519254,
+      "loss": 3.1511,
+      "step": 22349
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5539453029632568,
+      "learning_rate": 0.00042336274095841744,
+      "loss": 3.0462,
+      "step": 22350
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5515852570533752,
+      "learning_rate": 0.00042334870653672863,
+      "loss": 2.9138,
+      "step": 22351
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5160804986953735,
+      "learning_rate": 0.00042333467179016303,
+      "loss": 3.0142,
+      "step": 22352
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5109854936599731,
+      "learning_rate": 0.0004233206367187576,
+      "loss": 3.1363,
+      "step": 22353
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5832191705703735,
+      "learning_rate": 0.0004233066013225495,
+      "loss": 3.0025,
+      "step": 22354
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5390937924385071,
+      "learning_rate": 0.00042329256560157537,
+      "loss": 3.1417,
+      "step": 22355
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5653042793273926,
+      "learning_rate": 0.0004232785295558724,
+      "loss": 3.1148,
+      "step": 22356
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5726116895675659,
+      "learning_rate": 0.0004232644931854775,
+      "loss": 3.2039,
+      "step": 22357
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5899278521537781,
+      "learning_rate": 0.0004232504564904278,
+      "loss": 3.2208,
+      "step": 22358
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5757837295532227,
+      "learning_rate": 0.00042323641947075995,
+      "loss": 3.1871,
+      "step": 22359
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5611856579780579,
+      "learning_rate": 0.0004232223821265111,
+      "loss": 3.1254,
+      "step": 22360
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.568060576915741,
+      "learning_rate": 0.00042320834445771827,
+      "loss": 3.2821,
+      "step": 22361
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5462589859962463,
+      "learning_rate": 0.0004231943064644183,
+      "loss": 3.0981,
+      "step": 22362
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5303399562835693,
+      "learning_rate": 0.0004231802681466483,
+      "loss": 3.2174,
+      "step": 22363
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5866662263870239,
+      "learning_rate": 0.0004231662295044451,
+      "loss": 2.9066,
+      "step": 22364
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5583893060684204,
+      "learning_rate": 0.0004231521905378459,
+      "loss": 2.9209,
+      "step": 22365
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5403869152069092,
+      "learning_rate": 0.00042313815124688745,
+      "loss": 3.0106,
+      "step": 22366
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5992947220802307,
+      "learning_rate": 0.0004231241116316068,
+      "loss": 3.2221,
+      "step": 22367
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5342088937759399,
+      "learning_rate": 0.0004231100716920409,
+      "loss": 2.932,
+      "step": 22368
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5689389705657959,
+      "learning_rate": 0.00042309603142822686,
+      "loss": 3.0559,
+      "step": 22369
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5494974851608276,
+      "learning_rate": 0.0004230819908402015,
+      "loss": 3.1754,
+      "step": 22370
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5982123613357544,
+      "learning_rate": 0.0004230679499280018,
+      "loss": 3.2388,
+      "step": 22371
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5255215167999268,
+      "learning_rate": 0.00042305390869166494,
+      "loss": 3.2487,
+      "step": 22372
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5233165621757507,
+      "learning_rate": 0.00042303986713122766,
+      "loss": 3.0138,
+      "step": 22373
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5595539808273315,
+      "learning_rate": 0.0004230258252467271,
+      "loss": 3.1675,
+      "step": 22374
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5262620449066162,
+      "learning_rate": 0.00042301178303820025,
+      "loss": 3.3422,
+      "step": 22375
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5396249890327454,
+      "learning_rate": 0.00042299774050568394,
+      "loss": 3.0504,
+      "step": 22376
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5464133620262146,
+      "learning_rate": 0.00042298369764921524,
+      "loss": 3.2645,
+      "step": 22377
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.561158299446106,
+      "learning_rate": 0.0004229696544688312,
+      "loss": 3.1811,
+      "step": 22378
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5148482918739319,
+      "learning_rate": 0.0004229556109645688,
+      "loss": 3.2374,
+      "step": 22379
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5801580548286438,
+      "learning_rate": 0.00042294156713646494,
+      "loss": 2.9333,
+      "step": 22380
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5597841739654541,
+      "learning_rate": 0.0004229275229845567,
+      "loss": 3.2094,
+      "step": 22381
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5296369791030884,
+      "learning_rate": 0.00042291347850888087,
+      "loss": 2.8841,
+      "step": 22382
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.553082287311554,
+      "learning_rate": 0.00042289943370947476,
+      "loss": 3.1016,
+      "step": 22383
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6039657592773438,
+      "learning_rate": 0.0004228853885863751,
+      "loss": 3.0928,
+      "step": 22384
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5544348359107971,
+      "learning_rate": 0.000422871343139619,
+      "loss": 3.093,
+      "step": 22385
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5134222507476807,
+      "learning_rate": 0.00042285729736924344,
+      "loss": 3.134,
+      "step": 22386
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5257523059844971,
+      "learning_rate": 0.00042284325127528536,
+      "loss": 3.1072,
+      "step": 22387
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5168401598930359,
+      "learning_rate": 0.00042282920485778183,
+      "loss": 3.0503,
+      "step": 22388
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5573122501373291,
+      "learning_rate": 0.0004228151581167698,
+      "loss": 3.0215,
+      "step": 22389
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5416275858879089,
+      "learning_rate": 0.00042280111105228634,
+      "loss": 3.1959,
+      "step": 22390
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5634099841117859,
+      "learning_rate": 0.00042278706366436836,
+      "loss": 3.1155,
+      "step": 22391
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.58818519115448,
+      "learning_rate": 0.0004227730159530528,
+      "loss": 3.0849,
+      "step": 22392
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5335586071014404,
+      "learning_rate": 0.0004227589679183768,
+      "loss": 2.8114,
+      "step": 22393
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5638708472251892,
+      "learning_rate": 0.00042274491956037735,
+      "loss": 2.9018,
+      "step": 22394
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5199167132377625,
+      "learning_rate": 0.0004227308708790914,
+      "loss": 2.95,
+      "step": 22395
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5466358661651611,
+      "learning_rate": 0.00042271682187455594,
+      "loss": 3.0032,
+      "step": 22396
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5106999278068542,
+      "learning_rate": 0.000422702772546808,
+      "loss": 3.1893,
+      "step": 22397
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5572851896286011,
+      "learning_rate": 0.00042268872289588454,
+      "loss": 2.9937,
+      "step": 22398
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5566379427909851,
+      "learning_rate": 0.0004226746729218226,
+      "loss": 2.8665,
+      "step": 22399
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5310401916503906,
+      "learning_rate": 0.00042266062262465913,
+      "loss": 3.1665,
+      "step": 22400
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5580877065658569,
+      "learning_rate": 0.0004226465720044313,
+      "loss": 3.0617,
+      "step": 22401
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5510329604148865,
+      "learning_rate": 0.0004226325210611759,
+      "loss": 3.1422,
+      "step": 22402
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6322705149650574,
+      "learning_rate": 0.00042261846979493005,
+      "loss": 3.0221,
+      "step": 22403
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.56231290102005,
+      "learning_rate": 0.00042260441820573077,
+      "loss": 3.0854,
+      "step": 22404
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5650330781936646,
+      "learning_rate": 0.00042259036629361515,
+      "loss": 3.1679,
+      "step": 22405
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5603417754173279,
+      "learning_rate": 0.00042257631405862,
+      "loss": 3.1113,
+      "step": 22406
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5578677654266357,
+      "learning_rate": 0.0004225622615007824,
+      "loss": 3.2132,
+      "step": 22407
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5346105694770813,
+      "learning_rate": 0.00042254820862013947,
+      "loss": 3.1515,
+      "step": 22408
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5724475979804993,
+      "learning_rate": 0.0004225341554167281,
+      "loss": 3.0174,
+      "step": 22409
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6160552501678467,
+      "learning_rate": 0.00042252010189058534,
+      "loss": 3.0215,
+      "step": 22410
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5466728806495667,
+      "learning_rate": 0.0004225060480417482,
+      "loss": 3.0644,
+      "step": 22411
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5517079830169678,
+      "learning_rate": 0.00042249199387025383,
+      "loss": 3.0097,
+      "step": 22412
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5601084232330322,
+      "learning_rate": 0.00042247793937613893,
+      "loss": 3.0367,
+      "step": 22413
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5142292976379395,
+      "learning_rate": 0.00042246388455944076,
+      "loss": 3.0302,
+      "step": 22414
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5414049625396729,
+      "learning_rate": 0.0004224498294201964,
+      "loss": 3.0613,
+      "step": 22415
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5946900844573975,
+      "learning_rate": 0.0004224357739584427,
+      "loss": 2.8341,
+      "step": 22416
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5414168238639832,
+      "learning_rate": 0.0004224217181742167,
+      "loss": 3.109,
+      "step": 22417
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5585944056510925,
+      "learning_rate": 0.0004224076620675554,
+      "loss": 3.1668,
+      "step": 22418
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5433940887451172,
+      "learning_rate": 0.000422393605638496,
+      "loss": 2.8695,
+      "step": 22419
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5730898976325989,
+      "learning_rate": 0.00042237954888707535,
+      "loss": 3.1879,
+      "step": 22420
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5441948771476746,
+      "learning_rate": 0.0004223654918133305,
+      "loss": 3.0039,
+      "step": 22421
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5263099074363708,
+      "learning_rate": 0.00042235143441729853,
+      "loss": 3.0273,
+      "step": 22422
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5462498068809509,
+      "learning_rate": 0.0004223373766990164,
+      "loss": 3.1222,
+      "step": 22423
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5350505709648132,
+      "learning_rate": 0.0004223233186585211,
+      "loss": 3.0815,
+      "step": 22424
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5752918720245361,
+      "learning_rate": 0.0004223092602958498,
+      "loss": 3.2427,
+      "step": 22425
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5051820278167725,
+      "learning_rate": 0.0004222952016110395,
+      "loss": 3.1282,
+      "step": 22426
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.568838894367218,
+      "learning_rate": 0.0004222811426041271,
+      "loss": 3.133,
+      "step": 22427
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5199463367462158,
+      "learning_rate": 0.00042226708327514973,
+      "loss": 3.1213,
+      "step": 22428
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5748323202133179,
+      "learning_rate": 0.00042225302362414435,
+      "loss": 3.2312,
+      "step": 22429
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5747201442718506,
+      "learning_rate": 0.00042223896365114815,
+      "loss": 2.9831,
+      "step": 22430
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5636919140815735,
+      "learning_rate": 0.00042222490335619797,
+      "loss": 3.0873,
+      "step": 22431
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5946205854415894,
+      "learning_rate": 0.00042221084273933086,
+      "loss": 2.8779,
+      "step": 22432
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.545413613319397,
+      "learning_rate": 0.0004221967818005839,
+      "loss": 3.1714,
+      "step": 22433
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5125501751899719,
+      "learning_rate": 0.00042218272053999427,
+      "loss": 3.2075,
+      "step": 22434
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5402535796165466,
+      "learning_rate": 0.0004221686589575988,
+      "loss": 3.0146,
+      "step": 22435
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5473110675811768,
+      "learning_rate": 0.0004221545970534346,
+      "loss": 3.0171,
+      "step": 22436
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5473494529724121,
+      "learning_rate": 0.00042214053482753864,
+      "loss": 2.9442,
+      "step": 22437
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5252751708030701,
+      "learning_rate": 0.00042212647227994805,
+      "loss": 3.2078,
+      "step": 22438
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5198621153831482,
+      "learning_rate": 0.0004221124094106999,
+      "loss": 2.9712,
+      "step": 22439
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5305339694023132,
+      "learning_rate": 0.00042209834621983105,
+      "loss": 3.0418,
+      "step": 22440
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5615828037261963,
+      "learning_rate": 0.00042208428270737875,
+      "loss": 3.0449,
+      "step": 22441
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.554909884929657,
+      "learning_rate": 0.00042207021887337995,
+      "loss": 3.2129,
+      "step": 22442
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.578995406627655,
+      "learning_rate": 0.0004220561547178717,
+      "loss": 3.0968,
+      "step": 22443
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5198182463645935,
+      "learning_rate": 0.0004220420902408908,
+      "loss": 3.035,
+      "step": 22444
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5391382575035095,
+      "learning_rate": 0.0004220280254424748,
+      "loss": 3.1567,
+      "step": 22445
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5328619480133057,
+      "learning_rate": 0.0004220139603226603,
+      "loss": 3.2176,
+      "step": 22446
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5532649159431458,
+      "learning_rate": 0.0004219998948814846,
+      "loss": 3.105,
+      "step": 22447
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5565139055252075,
+      "learning_rate": 0.00042198582911898467,
+      "loss": 3.1877,
+      "step": 22448
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.522406280040741,
+      "learning_rate": 0.00042197176303519746,
+      "loss": 3.0828,
+      "step": 22449
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5588423609733582,
+      "learning_rate": 0.00042195769663016017,
+      "loss": 3.1538,
+      "step": 22450
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5381976962089539,
+      "learning_rate": 0.0004219436299039097,
+      "loss": 3.2286,
+      "step": 22451
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5260021090507507,
+      "learning_rate": 0.00042192956285648335,
+      "loss": 3.2763,
+      "step": 22452
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5152723789215088,
+      "learning_rate": 0.00042191549548791787,
+      "loss": 3.0711,
+      "step": 22453
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5516050457954407,
+      "learning_rate": 0.00042190142779825035,
+      "loss": 2.9179,
+      "step": 22454
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.539625883102417,
+      "learning_rate": 0.000421887359787518,
+      "loss": 2.9954,
+      "step": 22455
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5438560843467712,
+      "learning_rate": 0.0004218732914557579,
+      "loss": 3.1334,
+      "step": 22456
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5455101132392883,
+      "learning_rate": 0.00042185922280300696,
+      "loss": 3.2678,
+      "step": 22457
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5365552306175232,
+      "learning_rate": 0.0004218451538293023,
+      "loss": 3.0484,
+      "step": 22458
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.549579918384552,
+      "learning_rate": 0.0004218310845346808,
+      "loss": 3.3398,
+      "step": 22459
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.537279486656189,
+      "learning_rate": 0.00042181701491917983,
+      "loss": 3.0408,
+      "step": 22460
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5590521693229675,
+      "learning_rate": 0.0004218029449828362,
+      "loss": 3.1195,
+      "step": 22461
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5591493844985962,
+      "learning_rate": 0.0004217888747256871,
+      "loss": 3.0398,
+      "step": 22462
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5219199061393738,
+      "learning_rate": 0.00042177480414776956,
+      "loss": 2.9367,
+      "step": 22463
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5333996415138245,
+      "learning_rate": 0.0004217607332491206,
+      "loss": 3.2085,
+      "step": 22464
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5313884615898132,
+      "learning_rate": 0.00042174666202977727,
+      "loss": 2.8805,
+      "step": 22465
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5648304224014282,
+      "learning_rate": 0.0004217325904897766,
+      "loss": 3.2723,
+      "step": 22466
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5362172722816467,
+      "learning_rate": 0.0004217185186291559,
+      "loss": 3.132,
+      "step": 22467
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5308043956756592,
+      "learning_rate": 0.00042170444644795197,
+      "loss": 3.1991,
+      "step": 22468
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5285980701446533,
+      "learning_rate": 0.0004216903739462018,
+      "loss": 2.998,
+      "step": 22469
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.529820442199707,
+      "learning_rate": 0.00042167630112394284,
+      "loss": 2.9558,
+      "step": 22470
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5706838965415955,
+      "learning_rate": 0.0004216622279812118,
+      "loss": 3.1321,
+      "step": 22471
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5605747103691101,
+      "learning_rate": 0.0004216481545180459,
+      "loss": 3.1693,
+      "step": 22472
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.52308189868927,
+      "learning_rate": 0.00042163408073448214,
+      "loss": 2.8613,
+      "step": 22473
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5640896558761597,
+      "learning_rate": 0.0004216200066305576,
+      "loss": 3.1425,
+      "step": 22474
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5801900029182434,
+      "learning_rate": 0.0004216059322063094,
+      "loss": 3.1555,
+      "step": 22475
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5530162453651428,
+      "learning_rate": 0.0004215918574617746,
+      "loss": 3.1143,
+      "step": 22476
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5521156191825867,
+      "learning_rate": 0.0004215777823969902,
+      "loss": 3.103,
+      "step": 22477
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5479637980461121,
+      "learning_rate": 0.0004215637070119934,
+      "loss": 3.0354,
+      "step": 22478
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.551988422870636,
+      "learning_rate": 0.0004215496313068212,
+      "loss": 3.1629,
+      "step": 22479
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5706598162651062,
+      "learning_rate": 0.0004215355552815105,
+      "loss": 3.1129,
+      "step": 22480
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5351764559745789,
+      "learning_rate": 0.0004215214789360987,
+      "loss": 3.071,
+      "step": 22481
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5388553738594055,
+      "learning_rate": 0.00042150740227062263,
+      "loss": 3.0048,
+      "step": 22482
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5683891773223877,
+      "learning_rate": 0.0004214933252851195,
+      "loss": 3.0454,
+      "step": 22483
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5642884969711304,
+      "learning_rate": 0.00042147924797962625,
+      "loss": 3.0447,
+      "step": 22484
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5272451043128967,
+      "learning_rate": 0.0004214651703541801,
+      "loss": 2.9475,
+      "step": 22485
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5619935989379883,
+      "learning_rate": 0.00042145109240881805,
+      "loss": 2.9711,
+      "step": 22486
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6104692220687866,
+      "learning_rate": 0.0004214370141435772,
+      "loss": 3.2122,
+      "step": 22487
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5447784066200256,
+      "learning_rate": 0.0004214229355584946,
+      "loss": 3.0263,
+      "step": 22488
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5399470329284668,
+      "learning_rate": 0.0004214088566536075,
+      "loss": 3.0962,
+      "step": 22489
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5518352389335632,
+      "learning_rate": 0.0004213947774289526,
+      "loss": 2.8093,
+      "step": 22490
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5711850523948669,
+      "learning_rate": 0.00042138069788456736,
+      "loss": 2.9572,
+      "step": 22491
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5700511932373047,
+      "learning_rate": 0.00042136661802048874,
+      "loss": 3.058,
+      "step": 22492
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5496143102645874,
+      "learning_rate": 0.0004213525378367538,
+      "loss": 3.0367,
+      "step": 22493
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5528961420059204,
+      "learning_rate": 0.0004213384573333996,
+      "loss": 3.0467,
+      "step": 22494
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5584120750427246,
+      "learning_rate": 0.00042132437651046315,
+      "loss": 3.1026,
+      "step": 22495
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5595943927764893,
+      "learning_rate": 0.0004213102953679818,
+      "loss": 2.9348,
+      "step": 22496
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5819007158279419,
+      "learning_rate": 0.0004212962139059924,
+      "loss": 3.1339,
+      "step": 22497
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5432302951812744,
+      "learning_rate": 0.00042128213212453216,
+      "loss": 3.1587,
+      "step": 22498
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5066870450973511,
+      "learning_rate": 0.00042126805002363815,
+      "loss": 3.1825,
+      "step": 22499
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.531493604183197,
+      "learning_rate": 0.0004212539676033474,
+      "loss": 3.3026,
+      "step": 22500
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5524630546569824,
+      "learning_rate": 0.00042123988486369703,
+      "loss": 3.0143,
+      "step": 22501
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6150621771812439,
+      "learning_rate": 0.00042122580180472413,
+      "loss": 2.8606,
+      "step": 22502
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.55769282579422,
+      "learning_rate": 0.00042121171842646584,
+      "loss": 3.0798,
+      "step": 22503
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5654763579368591,
+      "learning_rate": 0.00042119763472895917,
+      "loss": 3.1062,
+      "step": 22504
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5415908098220825,
+      "learning_rate": 0.00042118355071224116,
+      "loss": 3.2484,
+      "step": 22505
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5293490290641785,
+      "learning_rate": 0.00042116946637634915,
+      "loss": 3.3383,
+      "step": 22506
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5430842041969299,
+      "learning_rate": 0.00042115538172132007,
+      "loss": 2.9673,
+      "step": 22507
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9073577523231506,
+      "learning_rate": 0.000421141296747191,
+      "loss": 3.183,
+      "step": 22508
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5296067595481873,
+      "learning_rate": 0.00042112721145399904,
+      "loss": 3.0694,
+      "step": 22509
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5582892894744873,
+      "learning_rate": 0.00042111312584178136,
+      "loss": 2.915,
+      "step": 22510
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5482213497161865,
+      "learning_rate": 0.000421099039910575,
+      "loss": 3.0819,
+      "step": 22511
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.534598708152771,
+      "learning_rate": 0.00042108495366041703,
+      "loss": 3.0956,
+      "step": 22512
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5279784202575684,
+      "learning_rate": 0.0004210708670913446,
+      "loss": 3.136,
+      "step": 22513
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5435617566108704,
+      "learning_rate": 0.00042105678020339495,
+      "loss": 3.3483,
+      "step": 22514
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5815554857254028,
+      "learning_rate": 0.00042104269299660487,
+      "loss": 3.2077,
+      "step": 22515
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5558618903160095,
+      "learning_rate": 0.00042102860547101165,
+      "loss": 2.9707,
+      "step": 22516
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5012040734291077,
+      "learning_rate": 0.00042101451762665247,
+      "loss": 3.1028,
+      "step": 22517
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5293975472450256,
+      "learning_rate": 0.00042100042946356436,
+      "loss": 2.9891,
+      "step": 22518
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5717222690582275,
+      "learning_rate": 0.0004209863409817843,
+      "loss": 3.0784,
+      "step": 22519
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5145463347434998,
+      "learning_rate": 0.0004209722521813495,
+      "loss": 2.9634,
+      "step": 22520
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5181236267089844,
+      "learning_rate": 0.0004209581630622971,
+      "loss": 3.2429,
+      "step": 22521
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5378340482711792,
+      "learning_rate": 0.0004209440736246642,
+      "loss": 3.152,
+      "step": 22522
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7082079648971558,
+      "learning_rate": 0.0004209299838684879,
+      "loss": 2.9937,
+      "step": 22523
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5510277152061462,
+      "learning_rate": 0.00042091589379380525,
+      "loss": 3.4617,
+      "step": 22524
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5319778919219971,
+      "learning_rate": 0.00042090180340065346,
+      "loss": 3.0375,
+      "step": 22525
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5088480710983276,
+      "learning_rate": 0.0004208877126890695,
+      "loss": 3.2293,
+      "step": 22526
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5296842455863953,
+      "learning_rate": 0.00042087362165909057,
+      "loss": 3.0811,
+      "step": 22527
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5537300109863281,
+      "learning_rate": 0.0004208595303107539,
+      "loss": 3.1295,
+      "step": 22528
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5461785197257996,
+      "learning_rate": 0.00042084543864409646,
+      "loss": 3.2753,
+      "step": 22529
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.571995735168457,
+      "learning_rate": 0.0004208313466591553,
+      "loss": 3.087,
+      "step": 22530
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5290436148643494,
+      "learning_rate": 0.00042081725435596764,
+      "loss": 3.0314,
+      "step": 22531
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5442591905593872,
+      "learning_rate": 0.00042080316173457064,
+      "loss": 3.2783,
+      "step": 22532
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5432357788085938,
+      "learning_rate": 0.00042078906879500135,
+      "loss": 3.2025,
+      "step": 22533
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.53834468126297,
+      "learning_rate": 0.0004207749755372969,
+      "loss": 3.2905,
+      "step": 22534
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5716333389282227,
+      "learning_rate": 0.0004207608819614944,
+      "loss": 3.1693,
+      "step": 22535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5577285885810852,
+      "learning_rate": 0.00042074678806763095,
+      "loss": 3.0463,
+      "step": 22536
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5382303595542908,
+      "learning_rate": 0.00042073269385574374,
+      "loss": 3.0666,
+      "step": 22537
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5647816061973572,
+      "learning_rate": 0.00042071859932586985,
+      "loss": 3.0805,
+      "step": 22538
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5214908123016357,
+      "learning_rate": 0.0004207045044780464,
+      "loss": 3.1665,
+      "step": 22539
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5525656938552856,
+      "learning_rate": 0.0004206904093123106,
+      "loss": 3.0327,
+      "step": 22540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5515208840370178,
+      "learning_rate": 0.0004206763138286993,
+      "loss": 3.1713,
+      "step": 22541
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5453616380691528,
+      "learning_rate": 0.0004206622180272499,
+      "loss": 3.0146,
+      "step": 22542
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5157879590988159,
+      "learning_rate": 0.00042064812190799946,
+      "loss": 3.1077,
+      "step": 22543
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5753068327903748,
+      "learning_rate": 0.0004206340254709851,
+      "loss": 2.9845,
+      "step": 22544
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5232889652252197,
+      "learning_rate": 0.0004206199287162439,
+      "loss": 3.1151,
+      "step": 22545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.527743399143219,
+      "learning_rate": 0.000420605831643813,
+      "loss": 3.25,
+      "step": 22546
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5670763254165649,
+      "learning_rate": 0.00042059173425372965,
+      "loss": 3.3105,
+      "step": 22547
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5618078708648682,
+      "learning_rate": 0.0004205776365460307,
+      "loss": 3.1793,
+      "step": 22548
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5352552533149719,
+      "learning_rate": 0.00042056353852075364,
+      "loss": 3.1404,
+      "step": 22549
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5104782581329346,
+      "learning_rate": 0.0004205494401779354,
+      "loss": 3.1323,
+      "step": 22550
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5447371602058411,
+      "learning_rate": 0.00042053534151761306,
+      "loss": 3.3586,
+      "step": 22551
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5591816902160645,
+      "learning_rate": 0.0004205212425398238,
+      "loss": 3.0942,
+      "step": 22552
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5504888892173767,
+      "learning_rate": 0.00042050714324460485,
+      "loss": 3.0393,
+      "step": 22553
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5453820824623108,
+      "learning_rate": 0.00042049304363199334,
+      "loss": 3.0541,
+      "step": 22554
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5488752722740173,
+      "learning_rate": 0.00042047894370202625,
+      "loss": 3.0418,
+      "step": 22555
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5417929887771606,
+      "learning_rate": 0.00042046484345474084,
+      "loss": 3.0206,
+      "step": 22556
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5546737909317017,
+      "learning_rate": 0.00042045074289017413,
+      "loss": 3.0535,
+      "step": 22557
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5946574807167053,
+      "learning_rate": 0.00042043664200836346,
+      "loss": 3.1673,
+      "step": 22558
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5759459137916565,
+      "learning_rate": 0.0004204225408093458,
+      "loss": 3.0653,
+      "step": 22559
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5452854037284851,
+      "learning_rate": 0.0004204084392931583,
+      "loss": 3.092,
+      "step": 22560
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5289528369903564,
+      "learning_rate": 0.0004203943374598382,
+      "loss": 3.2014,
+      "step": 22561
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5537521243095398,
+      "learning_rate": 0.00042038023530942265,
+      "loss": 3.1951,
+      "step": 22562
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5194694399833679,
+      "learning_rate": 0.0004203661328419486,
+      "loss": 2.9367,
+      "step": 22563
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5450275540351868,
+      "learning_rate": 0.00042035203005745336,
+      "loss": 2.9492,
+      "step": 22564
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.527119517326355,
+      "learning_rate": 0.0004203379269559741,
+      "loss": 3.1056,
+      "step": 22565
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5397108197212219,
+      "learning_rate": 0.0004203238235375479,
+      "loss": 3.1839,
+      "step": 22566
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5303124189376831,
+      "learning_rate": 0.0004203097198022118,
+      "loss": 3.2767,
+      "step": 22567
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5147061347961426,
+      "learning_rate": 0.0004202956157500031,
+      "loss": 3.1816,
+      "step": 22568
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5415562391281128,
+      "learning_rate": 0.00042028151138095895,
+      "loss": 3.3555,
+      "step": 22569
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5352077484130859,
+      "learning_rate": 0.0004202674066951165,
+      "loss": 3.0523,
+      "step": 22570
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5709248781204224,
+      "learning_rate": 0.0004202533016925127,
+      "loss": 3.0357,
+      "step": 22571
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5660057663917542,
+      "learning_rate": 0.0004202391963731849,
+      "loss": 3.1059,
+      "step": 22572
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5669631958007812,
+      "learning_rate": 0.0004202250907371702,
+      "loss": 3.0434,
+      "step": 22573
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5338017344474792,
+      "learning_rate": 0.0004202109847845057,
+      "loss": 3.0701,
+      "step": 22574
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5378772020339966,
+      "learning_rate": 0.00042019687851522873,
+      "loss": 3.3332,
+      "step": 22575
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5645471215248108,
+      "learning_rate": 0.0004201827719293762,
+      "loss": 3.2192,
+      "step": 22576
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5384810566902161,
+      "learning_rate": 0.00042016866502698536,
+      "loss": 2.9694,
+      "step": 22577
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5898525714874268,
+      "learning_rate": 0.00042015455780809345,
+      "loss": 2.8896,
+      "step": 22578
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5633717179298401,
+      "learning_rate": 0.00042014045027273755,
+      "loss": 2.9377,
+      "step": 22579
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.551085352897644,
+      "learning_rate": 0.00042012634242095487,
+      "loss": 3.0136,
+      "step": 22580
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5334330201148987,
+      "learning_rate": 0.00042011223425278253,
+      "loss": 3.0485,
+      "step": 22581
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5643008947372437,
+      "learning_rate": 0.0004200981257682575,
+      "loss": 3.142,
+      "step": 22582
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5641648769378662,
+      "learning_rate": 0.0004200840169674173,
+      "loss": 3.2212,
+      "step": 22583
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5251689553260803,
+      "learning_rate": 0.00042006990785029886,
+      "loss": 3.2124,
+      "step": 22584
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5536843538284302,
+      "learning_rate": 0.0004200557984169394,
+      "loss": 3.3359,
+      "step": 22585
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5234001874923706,
+      "learning_rate": 0.00042004168866737607,
+      "loss": 3.098,
+      "step": 22586
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5152735114097595,
+      "learning_rate": 0.000420027578601646,
+      "loss": 3.1305,
+      "step": 22587
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5633012056350708,
+      "learning_rate": 0.0004200134682197864,
+      "loss": 3.0323,
+      "step": 22588
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5583577752113342,
+      "learning_rate": 0.00041999935752183446,
+      "loss": 3.1308,
+      "step": 22589
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5771940350532532,
+      "learning_rate": 0.0004199852465078273,
+      "loss": 3.1042,
+      "step": 22590
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5792399644851685,
+      "learning_rate": 0.0004199711351778021,
+      "loss": 2.8656,
+      "step": 22591
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5672153234481812,
+      "learning_rate": 0.0004199570235317959,
+      "loss": 2.8594,
+      "step": 22592
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5597004890441895,
+      "learning_rate": 0.0004199429115698461,
+      "loss": 3.1582,
+      "step": 22593
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5668061971664429,
+      "learning_rate": 0.00041992879929198977,
+      "loss": 2.9705,
+      "step": 22594
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5375760793685913,
+      "learning_rate": 0.000419914686698264,
+      "loss": 2.9946,
+      "step": 22595
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5492685437202454,
+      "learning_rate": 0.00041990057378870607,
+      "loss": 2.9779,
+      "step": 22596
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5831897854804993,
+      "learning_rate": 0.00041988646056335304,
+      "loss": 3.0992,
+      "step": 22597
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5664513111114502,
+      "learning_rate": 0.00041987234702224217,
+      "loss": 3.0218,
+      "step": 22598
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5811390280723572,
+      "learning_rate": 0.00041985823316541065,
+      "loss": 2.9748,
+      "step": 22599
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5502529740333557,
+      "learning_rate": 0.00041984411899289556,
+      "loss": 3.1049,
+      "step": 22600
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.57771897315979,
+      "learning_rate": 0.0004198300045047342,
+      "loss": 3.1521,
+      "step": 22601
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5387015342712402,
+      "learning_rate": 0.00041981588970096365,
+      "loss": 3.0113,
+      "step": 22602
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.551127016544342,
+      "learning_rate": 0.000419801774581621,
+      "loss": 3.27,
+      "step": 22603
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5647044777870178,
+      "learning_rate": 0.0004197876591467436,
+      "loss": 3.3264,
+      "step": 22604
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5634638071060181,
+      "learning_rate": 0.0004197735433963686,
+      "loss": 2.9904,
+      "step": 22605
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5413936972618103,
+      "learning_rate": 0.0004197594273305332,
+      "loss": 3.1296,
+      "step": 22606
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.526145875453949,
+      "learning_rate": 0.0004197453109492743,
+      "loss": 3.0424,
+      "step": 22607
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5350415706634521,
+      "learning_rate": 0.0004197311942526294,
+      "loss": 3.137,
+      "step": 22608
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5814176797866821,
+      "learning_rate": 0.0004197170772406357,
+      "loss": 3.0361,
+      "step": 22609
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5808823108673096,
+      "learning_rate": 0.0004197029599133301,
+      "loss": 3.1731,
+      "step": 22610
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5619063973426819,
+      "learning_rate": 0.00041968884227075,
+      "loss": 2.9452,
+      "step": 22611
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5297744870185852,
+      "learning_rate": 0.0004196747243129325,
+      "loss": 3.2803,
+      "step": 22612
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.527192234992981,
+      "learning_rate": 0.00041966060603991484,
+      "loss": 3.133,
+      "step": 22613
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5389567017555237,
+      "learning_rate": 0.00041964648745173417,
+      "loss": 3.2495,
+      "step": 22614
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5459379553794861,
+      "learning_rate": 0.0004196323685484277,
+      "loss": 3.1859,
+      "step": 22615
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.558315634727478,
+      "learning_rate": 0.0004196182493300326,
+      "loss": 3.0052,
+      "step": 22616
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5612319707870483,
+      "learning_rate": 0.00041960412979658604,
+      "loss": 3.2109,
+      "step": 22617
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5649484992027283,
+      "learning_rate": 0.00041959000994812513,
+      "loss": 3.0352,
+      "step": 22618
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5512779951095581,
+      "learning_rate": 0.00041957588978468717,
+      "loss": 3.1786,
+      "step": 22619
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5309266448020935,
+      "learning_rate": 0.00041956176930630946,
+      "loss": 3.0735,
+      "step": 22620
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5767373442649841,
+      "learning_rate": 0.0004195476485130289,
+      "loss": 2.8406,
+      "step": 22621
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5628970861434937,
+      "learning_rate": 0.00041953352740488293,
+      "loss": 3.0715,
+      "step": 22622
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5341333150863647,
+      "learning_rate": 0.00041951940598190866,
+      "loss": 3.0542,
+      "step": 22623
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5453346967697144,
+      "learning_rate": 0.0004195052842441432,
+      "loss": 3.0502,
+      "step": 22624
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5791659951210022,
+      "learning_rate": 0.00041949116219162393,
+      "loss": 3.192,
+      "step": 22625
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5513650178909302,
+      "learning_rate": 0.0004194770398243879,
+      "loss": 3.1476,
+      "step": 22626
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5383404493331909,
+      "learning_rate": 0.0004194629171424724,
+      "loss": 2.8348,
+      "step": 22627
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5772968530654907,
+      "learning_rate": 0.00041944879414591434,
+      "loss": 3.2311,
+      "step": 22628
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5837339758872986,
+      "learning_rate": 0.0004194346708347513,
+      "loss": 2.9275,
+      "step": 22629
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5292240381240845,
+      "learning_rate": 0.0004194205472090204,
+      "loss": 3.0504,
+      "step": 22630
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5393617153167725,
+      "learning_rate": 0.0004194064232687587,
+      "loss": 2.9578,
+      "step": 22631
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5465915203094482,
+      "learning_rate": 0.0004193922990140034,
+      "loss": 3.2505,
+      "step": 22632
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5606839656829834,
+      "learning_rate": 0.0004193781744447917,
+      "loss": 2.9551,
+      "step": 22633
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5263482928276062,
+      "learning_rate": 0.0004193640495611611,
+      "loss": 3.2808,
+      "step": 22634
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5450267195701599,
+      "learning_rate": 0.0004193499243631484,
+      "loss": 3.1876,
+      "step": 22635
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5732899904251099,
+      "learning_rate": 0.000419335798850791,
+      "loss": 2.9856,
+      "step": 22636
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.55254727602005,
+      "learning_rate": 0.0004193216730241261,
+      "loss": 3.1878,
+      "step": 22637
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5654364824295044,
+      "learning_rate": 0.00041930754688319086,
+      "loss": 3.154,
+      "step": 22638
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5142787098884583,
+      "learning_rate": 0.0004192934204280224,
+      "loss": 3.4391,
+      "step": 22639
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5899552702903748,
+      "learning_rate": 0.00041927929365865815,
+      "loss": 3.126,
+      "step": 22640
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5144187808036804,
+      "learning_rate": 0.00041926516657513513,
+      "loss": 3.1848,
+      "step": 22641
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5327566266059875,
+      "learning_rate": 0.0004192510391774907,
+      "loss": 3.2201,
+      "step": 22642
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5743151307106018,
+      "learning_rate": 0.00041923691146576194,
+      "loss": 3.0937,
+      "step": 22643
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.538710355758667,
+      "learning_rate": 0.000419222783439986,
+      "loss": 3.1129,
+      "step": 22644
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5390024781227112,
+      "learning_rate": 0.00041920865510020045,
+      "loss": 3.0571,
+      "step": 22645
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5429919958114624,
+      "learning_rate": 0.000419194526446442,
+      "loss": 3.0848,
+      "step": 22646
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.532172679901123,
+      "learning_rate": 0.0004191803974787482,
+      "loss": 2.8347,
+      "step": 22647
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5515339970588684,
+      "learning_rate": 0.00041916626819715617,
+      "loss": 3.0664,
+      "step": 22648
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5160419940948486,
+      "learning_rate": 0.0004191521386017031,
+      "loss": 3.019,
+      "step": 22649
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5607364177703857,
+      "learning_rate": 0.0004191380086924263,
+      "loss": 2.895,
+      "step": 22650
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5337927937507629,
+      "learning_rate": 0.00041912387846936277,
+      "loss": 3.1408,
+      "step": 22651
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5739155411720276,
+      "learning_rate": 0.00041910974793255,
+      "loss": 3.2168,
+      "step": 22652
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.579247236251831,
+      "learning_rate": 0.0004190956170820251,
+      "loss": 2.9161,
+      "step": 22653
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5730478167533875,
+      "learning_rate": 0.0004190814859178251,
+      "loss": 3.0244,
+      "step": 22654
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5330377817153931,
+      "learning_rate": 0.0004190673544399875,
+      "loss": 2.9202,
+      "step": 22655
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5503820180892944,
+      "learning_rate": 0.00041905322264854946,
+      "loss": 3.1697,
+      "step": 22656
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5626627206802368,
+      "learning_rate": 0.00041903909054354804,
+      "loss": 3.1492,
+      "step": 22657
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5176795721054077,
+      "learning_rate": 0.0004190249581250206,
+      "loss": 3.0295,
+      "step": 22658
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5228406190872192,
+      "learning_rate": 0.00041901082539300426,
+      "loss": 3.0384,
+      "step": 22659
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5561695694923401,
+      "learning_rate": 0.0004189966923475364,
+      "loss": 3.0576,
+      "step": 22660
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5149094462394714,
+      "learning_rate": 0.0004189825589886542,
+      "loss": 3.0316,
+      "step": 22661
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5652263164520264,
+      "learning_rate": 0.00041896842531639476,
+      "loss": 3.0514,
+      "step": 22662
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5679181218147278,
+      "learning_rate": 0.00041895429133079534,
+      "loss": 3.0378,
+      "step": 22663
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5677348375320435,
+      "learning_rate": 0.00041894015703189326,
+      "loss": 3.1728,
+      "step": 22664
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5508793592453003,
+      "learning_rate": 0.00041892602241972566,
+      "loss": 3.1805,
+      "step": 22665
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5380650162696838,
+      "learning_rate": 0.0004189118874943298,
+      "loss": 2.9287,
+      "step": 22666
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5400630235671997,
+      "learning_rate": 0.00041889775225574305,
+      "loss": 3.0148,
+      "step": 22667
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5459088683128357,
+      "learning_rate": 0.0004188836167040024,
+      "loss": 3.2834,
+      "step": 22668
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5284379720687866,
+      "learning_rate": 0.00041886948083914513,
+      "loss": 3.0427,
+      "step": 22669
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5413404107093811,
+      "learning_rate": 0.0004188553446612086,
+      "loss": 3.0797,
+      "step": 22670
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6029809713363647,
+      "learning_rate": 0.00041884120817023,
+      "loss": 3.0449,
+      "step": 22671
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5510432720184326,
+      "learning_rate": 0.0004188270713662465,
+      "loss": 3.0182,
+      "step": 22672
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5862864255905151,
+      "learning_rate": 0.0004188129342492953,
+      "loss": 3.2882,
+      "step": 22673
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5308249592781067,
+      "learning_rate": 0.0004187987968194137,
+      "loss": 3.0142,
+      "step": 22674
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5954753756523132,
+      "learning_rate": 0.000418784659076639,
+      "loss": 3.2163,
+      "step": 22675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5317316651344299,
+      "learning_rate": 0.00041877052102100835,
+      "loss": 2.9223,
+      "step": 22676
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5436576008796692,
+      "learning_rate": 0.00041875638265255894,
+      "loss": 3.0907,
+      "step": 22677
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5222603678703308,
+      "learning_rate": 0.0004187422439713282,
+      "loss": 2.9988,
+      "step": 22678
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5745731592178345,
+      "learning_rate": 0.0004187281049773532,
+      "loss": 2.8509,
+      "step": 22679
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5338032841682434,
+      "learning_rate": 0.0004187139656706711,
+      "loss": 3.1961,
+      "step": 22680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5111261010169983,
+      "learning_rate": 0.00041869982605131933,
+      "loss": 3.2963,
+      "step": 22681
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5470437407493591,
+      "learning_rate": 0.0004186856861193352,
+      "loss": 3.1833,
+      "step": 22682
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5256366729736328,
+      "learning_rate": 0.0004186715458747557,
+      "loss": 3.2079,
+      "step": 22683
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5694595575332642,
+      "learning_rate": 0.00041865740531761813,
+      "loss": 2.9047,
+      "step": 22684
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5024998188018799,
+      "learning_rate": 0.0004186432644479599,
+      "loss": 3.2674,
+      "step": 22685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5744755268096924,
+      "learning_rate": 0.000418629123265818,
+      "loss": 3.1185,
+      "step": 22686
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5288079380989075,
+      "learning_rate": 0.00041861498177122995,
+      "loss": 3.1339,
+      "step": 22687
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5845724940299988,
+      "learning_rate": 0.0004186008399642328,
+      "loss": 2.8917,
+      "step": 22688
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5915581583976746,
+      "learning_rate": 0.00041858669784486396,
+      "loss": 3.0969,
+      "step": 22689
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5309663414955139,
+      "learning_rate": 0.0004185725554131604,
+      "loss": 3.0139,
+      "step": 22690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5378329753875732,
+      "learning_rate": 0.0004185584126691597,
+      "loss": 3.2019,
+      "step": 22691
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5664974451065063,
+      "learning_rate": 0.0004185442696128989,
+      "loss": 3.0974,
+      "step": 22692
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5150554776191711,
+      "learning_rate": 0.00041853012624441544,
+      "loss": 3.0179,
+      "step": 22693
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5422317981719971,
+      "learning_rate": 0.0004185159825637463,
+      "loss": 3.0485,
+      "step": 22694
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5424741506576538,
+      "learning_rate": 0.0004185018385709288,
+      "loss": 3.1692,
+      "step": 22695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5382916927337646,
+      "learning_rate": 0.0004184876942660004,
+      "loss": 3.027,
+      "step": 22696
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5341999530792236,
+      "learning_rate": 0.00041847354964899823,
+      "loss": 3.0156,
+      "step": 22697
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5673853754997253,
+      "learning_rate": 0.00041845940471995946,
+      "loss": 3.045,
+      "step": 22698
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5665754079818726,
+      "learning_rate": 0.0004184452594789214,
+      "loss": 3.0651,
+      "step": 22699
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5702019333839417,
+      "learning_rate": 0.0004184311139259214,
+      "loss": 2.9393,
+      "step": 22700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5696156620979309,
+      "learning_rate": 0.00041841696806099656,
+      "loss": 2.9479,
+      "step": 22701
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.577709436416626,
+      "learning_rate": 0.00041840282188418426,
+      "loss": 3.0455,
+      "step": 22702
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5778018832206726,
+      "learning_rate": 0.0004183886753955217,
+      "loss": 3.0691,
+      "step": 22703
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5543364882469177,
+      "learning_rate": 0.0004183745285950462,
+      "loss": 3.0442,
+      "step": 22704
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5635522603988647,
+      "learning_rate": 0.0004183603814827949,
+      "loss": 3.0389,
+      "step": 22705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5350740551948547,
+      "learning_rate": 0.0004183462340588051,
+      "loss": 3.1186,
+      "step": 22706
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5741170048713684,
+      "learning_rate": 0.00041833208632311423,
+      "loss": 3.017,
+      "step": 22707
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5641077756881714,
+      "learning_rate": 0.0004183179382757593,
+      "loss": 3.2578,
+      "step": 22708
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5855833888053894,
+      "learning_rate": 0.0004183037899167778,
+      "loss": 2.9378,
+      "step": 22709
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5699617266654968,
+      "learning_rate": 0.0004182896412462067,
+      "loss": 3.2279,
+      "step": 22710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5196274518966675,
+      "learning_rate": 0.0004182754922640837,
+      "loss": 3.0284,
+      "step": 22711
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5257676243782043,
+      "learning_rate": 0.0004182613429704456,
+      "loss": 3.1445,
+      "step": 22712
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5339217782020569,
+      "learning_rate": 0.00041824719336533005,
+      "loss": 3.0042,
+      "step": 22713
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5554669499397278,
+      "learning_rate": 0.00041823304344877404,
+      "loss": 3.0607,
+      "step": 22714
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5328848958015442,
+      "learning_rate": 0.000418218893220815,
+      "loss": 3.1213,
+      "step": 22715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5345682501792908,
+      "learning_rate": 0.0004182047426814901,
+      "loss": 3.122,
+      "step": 22716
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5637184381484985,
+      "learning_rate": 0.00041819059183083665,
+      "loss": 3.186,
+      "step": 22717
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5496988892555237,
+      "learning_rate": 0.00041817644066889203,
+      "loss": 3.1549,
+      "step": 22718
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5595161318778992,
+      "learning_rate": 0.00041816228919569335,
+      "loss": 3.1797,
+      "step": 22719
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5612072944641113,
+      "learning_rate": 0.00041814813741127796,
+      "loss": 3.1245,
+      "step": 22720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5443363785743713,
+      "learning_rate": 0.000418133985315683,
+      "loss": 3.2956,
+      "step": 22721
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5439444780349731,
+      "learning_rate": 0.000418119832908946,
+      "loss": 3.0141,
+      "step": 22722
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5160937309265137,
+      "learning_rate": 0.000418105680191104,
+      "loss": 3.2615,
+      "step": 22723
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5264231562614441,
+      "learning_rate": 0.0004180915271621945,
+      "loss": 3.0696,
+      "step": 22724
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5493642687797546,
+      "learning_rate": 0.0004180773738222545,
+      "loss": 3.0094,
+      "step": 22725
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.535229504108429,
+      "learning_rate": 0.0004180632201713215,
+      "loss": 2.9877,
+      "step": 22726
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5449292063713074,
+      "learning_rate": 0.0004180490662094326,
+      "loss": 3.1424,
+      "step": 22727
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5288578271865845,
+      "learning_rate": 0.00041803491193662524,
+      "loss": 3.1605,
+      "step": 22728
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5462258458137512,
+      "learning_rate": 0.0004180207573529367,
+      "loss": 3.204,
+      "step": 22729
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.507749080657959,
+      "learning_rate": 0.00041800660245840414,
+      "loss": 3.1402,
+      "step": 22730
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.554622232913971,
+      "learning_rate": 0.00041799244725306483,
+      "loss": 3.1245,
+      "step": 22731
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5544594526290894,
+      "learning_rate": 0.0004179782917369562,
+      "loss": 3.0031,
+      "step": 22732
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5762028098106384,
+      "learning_rate": 0.00041796413591011557,
+      "loss": 2.9274,
+      "step": 22733
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5185121893882751,
+      "learning_rate": 0.00041794997977257994,
+      "loss": 2.9783,
+      "step": 22734
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5906330943107605,
+      "learning_rate": 0.0004179358233243868,
+      "loss": 3.1314,
+      "step": 22735
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5670438408851624,
+      "learning_rate": 0.00041792166656557343,
+      "loss": 3.1062,
+      "step": 22736
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6015538573265076,
+      "learning_rate": 0.00041790750949617704,
+      "loss": 2.9635,
+      "step": 22737
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5294987559318542,
+      "learning_rate": 0.000417893352116235,
+      "loss": 3.0218,
+      "step": 22738
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5639286041259766,
+      "learning_rate": 0.00041787919442578457,
+      "loss": 3.1125,
+      "step": 22739
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5568161010742188,
+      "learning_rate": 0.0004178650364248631,
+      "loss": 3.2293,
+      "step": 22740
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5709725022315979,
+      "learning_rate": 0.0004178508781135076,
+      "loss": 3.1711,
+      "step": 22741
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5765569806098938,
+      "learning_rate": 0.00041783671949175566,
+      "loss": 3.2865,
+      "step": 22742
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5151780247688293,
+      "learning_rate": 0.0004178225605596445,
+      "loss": 3.2819,
+      "step": 22743
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5453171730041504,
+      "learning_rate": 0.00041780840131721143,
+      "loss": 3.089,
+      "step": 22744
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5520200729370117,
+      "learning_rate": 0.0004177942417644937,
+      "loss": 3.0535,
+      "step": 22745
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5933725237846375,
+      "learning_rate": 0.00041778008190152856,
+      "loss": 3.0728,
+      "step": 22746
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5700779557228088,
+      "learning_rate": 0.0004177659217283533,
+      "loss": 3.0971,
+      "step": 22747
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5266256928443909,
+      "learning_rate": 0.0004177517612450054,
+      "loss": 3.1489,
+      "step": 22748
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5342414975166321,
+      "learning_rate": 0.00041773760045152195,
+      "loss": 3.1828,
+      "step": 22749
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5465877056121826,
+      "learning_rate": 0.0004177234393479403,
+      "loss": 3.0301,
+      "step": 22750
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5596908330917358,
+      "learning_rate": 0.0004177092779342978,
+      "loss": 2.8841,
+      "step": 22751
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5438765287399292,
+      "learning_rate": 0.00041769511621063166,
+      "loss": 3.1112,
+      "step": 22752
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5041543245315552,
+      "learning_rate": 0.00041768095417697937,
+      "loss": 3.0129,
+      "step": 22753
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.527295708656311,
+      "learning_rate": 0.00041766679183337797,
+      "loss": 3.2012,
+      "step": 22754
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5652879476547241,
+      "learning_rate": 0.00041765262917986497,
+      "loss": 3.2307,
+      "step": 22755
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5673266053199768,
+      "learning_rate": 0.00041763846621647755,
+      "loss": 3.135,
+      "step": 22756
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5326421856880188,
+      "learning_rate": 0.00041762430294325296,
+      "loss": 3.0118,
+      "step": 22757
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5760022401809692,
+      "learning_rate": 0.0004176101393602287,
+      "loss": 3.2049,
+      "step": 22758
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5360777974128723,
+      "learning_rate": 0.00041759597546744194,
+      "loss": 3.1222,
+      "step": 22759
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5167760848999023,
+      "learning_rate": 0.00041758181126493006,
+      "loss": 2.9691,
+      "step": 22760
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8578059077262878,
+      "learning_rate": 0.00041756764675273024,
+      "loss": 3.2545,
+      "step": 22761
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6276934742927551,
+      "learning_rate": 0.0004175534819308799,
+      "loss": 3.0123,
+      "step": 22762
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5215904712677002,
+      "learning_rate": 0.0004175393167994163,
+      "loss": 3.1768,
+      "step": 22763
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.51545250415802,
+      "learning_rate": 0.00041752515135837676,
+      "loss": 2.9006,
+      "step": 22764
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.535327672958374,
+      "learning_rate": 0.00041751098560779856,
+      "loss": 3.1213,
+      "step": 22765
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.54879230260849,
+      "learning_rate": 0.00041749681954771913,
+      "loss": 3.2119,
+      "step": 22766
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5459108352661133,
+      "learning_rate": 0.00041748265317817566,
+      "loss": 3.0567,
+      "step": 22767
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5210204720497131,
+      "learning_rate": 0.0004174684864992054,
+      "loss": 3.2307,
+      "step": 22768
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5359011292457581,
+      "learning_rate": 0.0004174543195108459,
+      "loss": 2.9444,
+      "step": 22769
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5488296747207642,
+      "learning_rate": 0.00041744015221313423,
+      "loss": 3.263,
+      "step": 22770
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5671960115432739,
+      "learning_rate": 0.0004174259846061078,
+      "loss": 2.9674,
+      "step": 22771
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5221547484397888,
+      "learning_rate": 0.00041741181668980386,
+      "loss": 2.937,
+      "step": 22772
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5802128911018372,
+      "learning_rate": 0.00041739764846425995,
+      "loss": 3.0267,
+      "step": 22773
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5907021760940552,
+      "learning_rate": 0.0004173834799295132,
+      "loss": 3.1479,
+      "step": 22774
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5466511845588684,
+      "learning_rate": 0.0004173693110856009,
+      "loss": 2.9301,
+      "step": 22775
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5275335311889648,
+      "learning_rate": 0.00041735514193256044,
+      "loss": 3.1554,
+      "step": 22776
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5344946980476379,
+      "learning_rate": 0.0004173409724704291,
+      "loss": 3.2402,
+      "step": 22777
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5361987948417664,
+      "learning_rate": 0.0004173268026992442,
+      "loss": 3.0482,
+      "step": 22778
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5685098171234131,
+      "learning_rate": 0.0004173126326190431,
+      "loss": 3.1086,
+      "step": 22779
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5359070897102356,
+      "learning_rate": 0.0004172984622298632,
+      "loss": 3.1898,
+      "step": 22780
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5278506875038147,
+      "learning_rate": 0.0004172842915317417,
+      "loss": 3.0062,
+      "step": 22781
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5743756294250488,
+      "learning_rate": 0.00041727012052471577,
+      "loss": 3.1249,
+      "step": 22782
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5638933777809143,
+      "learning_rate": 0.00041725594920882304,
+      "loss": 2.8144,
+      "step": 22783
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5870307683944702,
+      "learning_rate": 0.00041724177758410073,
+      "loss": 3.1692,
+      "step": 22784
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.530704140663147,
+      "learning_rate": 0.0004172276056505861,
+      "loss": 2.8009,
+      "step": 22785
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5861652493476868,
+      "learning_rate": 0.00041721343340831655,
+      "loss": 3.2099,
+      "step": 22786
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5355147123336792,
+      "learning_rate": 0.0004171992608573293,
+      "loss": 2.9651,
+      "step": 22787
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5698812007904053,
+      "learning_rate": 0.0004171850879976618,
+      "loss": 3.0561,
+      "step": 22788
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5511211156845093,
+      "learning_rate": 0.0004171709148293514,
+      "loss": 3.2349,
+      "step": 22789
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5315529108047485,
+      "learning_rate": 0.00041715674135243524,
+      "loss": 2.962,
+      "step": 22790
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5924056768417358,
+      "learning_rate": 0.0004171425675669509,
+      "loss": 3.0684,
+      "step": 22791
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5384601950645447,
+      "learning_rate": 0.00041712839347293556,
+      "loss": 3.1687,
+      "step": 22792
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5141586065292358,
+      "learning_rate": 0.0004171142190704264,
+      "loss": 3.0498,
+      "step": 22793
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5134872794151306,
+      "learning_rate": 0.00041710004435946104,
+      "loss": 3.0604,
+      "step": 22794
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5418192744255066,
+      "learning_rate": 0.0004170858693400768,
+      "loss": 3.0924,
+      "step": 22795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5468102097511292,
+      "learning_rate": 0.0004170716940123108,
+      "loss": 2.9778,
+      "step": 22796
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.519443929195404,
+      "learning_rate": 0.00041705751837620055,
+      "loss": 3.0669,
+      "step": 22797
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.534935712814331,
+      "learning_rate": 0.0004170433424317832,
+      "loss": 3.0799,
+      "step": 22798
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5599324703216553,
+      "learning_rate": 0.00041702916617909633,
+      "loss": 3.2856,
+      "step": 22799
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5286010503768921,
+      "learning_rate": 0.0004170149896181771,
+      "loss": 2.9179,
+      "step": 22800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5498127341270447,
+      "learning_rate": 0.000417000812749063,
+      "loss": 2.9982,
+      "step": 22801
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5402305126190186,
+      "learning_rate": 0.00041698663557179126,
+      "loss": 2.922,
+      "step": 22802
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5140150785446167,
+      "learning_rate": 0.0004169724580863991,
+      "loss": 3.1083,
+      "step": 22803
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.518584132194519,
+      "learning_rate": 0.00041695828029292406,
+      "loss": 3.1583,
+      "step": 22804
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5516182780265808,
+      "learning_rate": 0.00041694410219140344,
+      "loss": 3.2764,
+      "step": 22805
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5740772485733032,
+      "learning_rate": 0.0004169299237818747,
+      "loss": 3.0374,
+      "step": 22806
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5997231602668762,
+      "learning_rate": 0.00041691574506437485,
+      "loss": 3.056,
+      "step": 22807
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.566772997379303,
+      "learning_rate": 0.0004169015660389414,
+      "loss": 3.1544,
+      "step": 22808
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5957667827606201,
+      "learning_rate": 0.0004168873867056119,
+      "loss": 3.0864,
+      "step": 22809
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5558214783668518,
+      "learning_rate": 0.00041687320706442345,
+      "loss": 3.0475,
+      "step": 22810
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5514797568321228,
+      "learning_rate": 0.0004168590271154134,
+      "loss": 3.1521,
+      "step": 22811
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5616558194160461,
+      "learning_rate": 0.0004168448468586192,
+      "loss": 2.9716,
+      "step": 22812
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.555022120475769,
+      "learning_rate": 0.00041683066629407816,
+      "loss": 2.985,
+      "step": 22813
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5655428767204285,
+      "learning_rate": 0.00041681648542182765,
+      "loss": 3.1157,
+      "step": 22814
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5333276391029358,
+      "learning_rate": 0.00041680230424190497,
+      "loss": 3.1813,
+      "step": 22815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5510712265968323,
+      "learning_rate": 0.0004167881227543475,
+      "loss": 2.9741,
+      "step": 22816
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5697930455207825,
+      "learning_rate": 0.0004167739409591927,
+      "loss": 3.3855,
+      "step": 22817
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5547995567321777,
+      "learning_rate": 0.0004167597588564776,
+      "loss": 3.1625,
+      "step": 22818
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5366219878196716,
+      "learning_rate": 0.00041674557644623986,
+      "loss": 3.1218,
+      "step": 22819
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5644329786300659,
+      "learning_rate": 0.0004167313937285169,
+      "loss": 3.2227,
+      "step": 22820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5153218507766724,
+      "learning_rate": 0.0004167172107033457,
+      "loss": 3.1479,
+      "step": 22821
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5740698575973511,
+      "learning_rate": 0.00041670302737076395,
+      "loss": 3.0745,
+      "step": 22822
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5266560316085815,
+      "learning_rate": 0.00041668884373080876,
+      "loss": 3.0566,
+      "step": 22823
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5471863150596619,
+      "learning_rate": 0.0004166746597835178,
+      "loss": 3.0059,
+      "step": 22824
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.53224116563797,
+      "learning_rate": 0.00041666047552892815,
+      "loss": 3.1979,
+      "step": 22825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5300314426422119,
+      "learning_rate": 0.00041664629096707716,
+      "loss": 3.0605,
+      "step": 22826
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.536897599697113,
+      "learning_rate": 0.00041663210609800246,
+      "loss": 3.075,
+      "step": 22827
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5496629476547241,
+      "learning_rate": 0.0004166179209217411,
+      "loss": 3.1034,
+      "step": 22828
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5242640376091003,
+      "learning_rate": 0.00041660373543833066,
+      "loss": 3.0961,
+      "step": 22829
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5705289840698242,
+      "learning_rate": 0.00041658954964780837,
+      "loss": 3.0197,
+      "step": 22830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5280457139015198,
+      "learning_rate": 0.0004165753635502118,
+      "loss": 3.0366,
+      "step": 22831
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5403161644935608,
+      "learning_rate": 0.000416561177145578,
+      "loss": 3.0601,
+      "step": 22832
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5531915426254272,
+      "learning_rate": 0.0004165469904339445,
+      "loss": 3.1521,
+      "step": 22833
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5562331080436707,
+      "learning_rate": 0.0004165328034153486,
+      "loss": 3.1089,
+      "step": 22834
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5130084156990051,
+      "learning_rate": 0.00041651861608982786,
+      "loss": 3.2384,
+      "step": 22835
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5995485186576843,
+      "learning_rate": 0.00041650442845741944,
+      "loss": 2.9602,
+      "step": 22836
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5520250201225281,
+      "learning_rate": 0.00041649024051816077,
+      "loss": 2.9841,
+      "step": 22837
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5484051704406738,
+      "learning_rate": 0.0004164760522720892,
+      "loss": 3.0843,
+      "step": 22838
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5702111124992371,
+      "learning_rate": 0.0004164618637192422,
+      "loss": 3.1287,
+      "step": 22839
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5361772179603577,
+      "learning_rate": 0.00041644767485965705,
+      "loss": 2.8899,
+      "step": 22840
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5582685470581055,
+      "learning_rate": 0.0004164334856933711,
+      "loss": 3.2705,
+      "step": 22841
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5814189910888672,
+      "learning_rate": 0.00041641929622042185,
+      "loss": 2.9787,
+      "step": 22842
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.543645977973938,
+      "learning_rate": 0.0004164051064408465,
+      "loss": 2.9174,
+      "step": 22843
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5395931601524353,
+      "learning_rate": 0.0004163909163546824,
+      "loss": 2.9608,
+      "step": 22844
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5268449783325195,
+      "learning_rate": 0.0004163767259619671,
+      "loss": 3.1027,
+      "step": 22845
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5362355709075928,
+      "learning_rate": 0.0004163625352627381,
+      "loss": 2.9239,
+      "step": 22846
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5752016305923462,
+      "learning_rate": 0.00041634834425703236,
+      "loss": 3.0805,
+      "step": 22847
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5360502004623413,
+      "learning_rate": 0.0004163341529448875,
+      "loss": 2.909,
+      "step": 22848
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5390860438346863,
+      "learning_rate": 0.00041631996132634086,
+      "loss": 3.0375,
+      "step": 22849
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5611980557441711,
+      "learning_rate": 0.0004163057694014299,
+      "loss": 3.1001,
+      "step": 22850
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5395224094390869,
+      "learning_rate": 0.00041629157717019185,
+      "loss": 3.2519,
+      "step": 22851
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5417009592056274,
+      "learning_rate": 0.00041627738463266426,
+      "loss": 2.8745,
+      "step": 22852
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5183812379837036,
+      "learning_rate": 0.0004162631917888844,
+      "loss": 3.1916,
+      "step": 22853
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5822927355766296,
+      "learning_rate": 0.0004162489986388896,
+      "loss": 3.0131,
+      "step": 22854
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5340526103973389,
+      "learning_rate": 0.0004162348051827173,
+      "loss": 3.151,
+      "step": 22855
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5964843034744263,
+      "learning_rate": 0.00041622061142040494,
+      "loss": 3.1859,
+      "step": 22856
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5237846970558167,
+      "learning_rate": 0.0004162064173519899,
+      "loss": 2.9817,
+      "step": 22857
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5963344573974609,
+      "learning_rate": 0.0004161922229775094,
+      "loss": 3.1043,
+      "step": 22858
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5289567708969116,
+      "learning_rate": 0.00041617802829700104,
+      "loss": 3.0147,
+      "step": 22859
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.545599102973938,
+      "learning_rate": 0.00041616383331050214,
+      "loss": 3.104,
+      "step": 22860
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5368244647979736,
+      "learning_rate": 0.00041614963801804995,
+      "loss": 2.9397,
+      "step": 22861
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5587390661239624,
+      "learning_rate": 0.00041613544241968204,
+      "loss": 3.1038,
+      "step": 22862
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5761477947235107,
+      "learning_rate": 0.0004161212465154357,
+      "loss": 3.1405,
+      "step": 22863
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5684610605239868,
+      "learning_rate": 0.00041610705030534835,
+      "loss": 3.0149,
+      "step": 22864
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5535710453987122,
+      "learning_rate": 0.0004160928537894574,
+      "loss": 3.1474,
+      "step": 22865
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5354270339012146,
+      "learning_rate": 0.00041607865696780016,
+      "loss": 2.9943,
+      "step": 22866
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5734537839889526,
+      "learning_rate": 0.0004160644598404141,
+      "loss": 3.0333,
+      "step": 22867
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5363245010375977,
+      "learning_rate": 0.0004160502624073367,
+      "loss": 3.1273,
+      "step": 22868
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5657275319099426,
+      "learning_rate": 0.0004160360646686051,
+      "loss": 3.1475,
+      "step": 22869
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5877406001091003,
+      "learning_rate": 0.0004160218666242569,
+      "loss": 3.1525,
+      "step": 22870
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5854254961013794,
+      "learning_rate": 0.0004160076682743294,
+      "loss": 3.343,
+      "step": 22871
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4987305700778961,
+      "learning_rate": 0.00041599346961886005,
+      "loss": 3.1971,
+      "step": 22872
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5421817302703857,
+      "learning_rate": 0.0004159792706578862,
+      "loss": 3.1557,
+      "step": 22873
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5647212862968445,
+      "learning_rate": 0.0004159650713914453,
+      "loss": 3.059,
+      "step": 22874
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5287631154060364,
+      "learning_rate": 0.0004159508718195747,
+      "loss": 3.1896,
+      "step": 22875
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5936976671218872,
+      "learning_rate": 0.0004159366719423118,
+      "loss": 3.188,
+      "step": 22876
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5288457870483398,
+      "learning_rate": 0.00041592247175969413,
+      "loss": 3.1293,
+      "step": 22877
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5637390613555908,
+      "learning_rate": 0.00041590827127175884,
+      "loss": 3.1525,
+      "step": 22878
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5691234469413757,
+      "learning_rate": 0.00041589407047854356,
+      "loss": 3.1165,
+      "step": 22879
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5463374853134155,
+      "learning_rate": 0.0004158798693800855,
+      "loss": 3.185,
+      "step": 22880
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5565144419670105,
+      "learning_rate": 0.0004158656679764223,
+      "loss": 3.179,
+      "step": 22881
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5509166717529297,
+      "learning_rate": 0.0004158514662675911,
+      "loss": 3.1439,
+      "step": 22882
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5458000302314758,
+      "learning_rate": 0.00041583726425362957,
+      "loss": 2.9293,
+      "step": 22883
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.536400318145752,
+      "learning_rate": 0.0004158230619345749,
+      "loss": 2.9197,
+      "step": 22884
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5528376698493958,
+      "learning_rate": 0.0004158088593104645,
+      "loss": 3.1646,
+      "step": 22885
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5160115361213684,
+      "learning_rate": 0.000415794656381336,
+      "loss": 3.0597,
+      "step": 22886
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5468571782112122,
+      "learning_rate": 0.0004157804531472266,
+      "loss": 3.257,
+      "step": 22887
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5390594601631165,
+      "learning_rate": 0.0004157662496081738,
+      "loss": 3.2257,
+      "step": 22888
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5498918890953064,
+      "learning_rate": 0.00041575204576421493,
+      "loss": 3.1676,
+      "step": 22889
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5376027822494507,
+      "learning_rate": 0.0004157378416153874,
+      "loss": 3.087,
+      "step": 22890
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5271984934806824,
+      "learning_rate": 0.00041572363716172876,
+      "loss": 3.092,
+      "step": 22891
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5507243275642395,
+      "learning_rate": 0.00041570943240327627,
+      "loss": 3.0022,
+      "step": 22892
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5260735154151917,
+      "learning_rate": 0.0004156952273400675,
+      "loss": 3.0662,
+      "step": 22893
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5336798429489136,
+      "learning_rate": 0.0004156810219721397,
+      "loss": 3.1374,
+      "step": 22894
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5313898324966431,
+      "learning_rate": 0.0004156668162995303,
+      "loss": 3.1071,
+      "step": 22895
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5449457764625549,
+      "learning_rate": 0.0004156526103222768,
+      "loss": 3.2074,
+      "step": 22896
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5850553512573242,
+      "learning_rate": 0.00041563840404041665,
+      "loss": 2.977,
+      "step": 22897
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5299965143203735,
+      "learning_rate": 0.00041562419745398715,
+      "loss": 3.154,
+      "step": 22898
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5133044719696045,
+      "learning_rate": 0.00041560999056302575,
+      "loss": 3.0494,
+      "step": 22899
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6028819680213928,
+      "learning_rate": 0.0004155957833675699,
+      "loss": 2.9978,
+      "step": 22900
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5400010347366333,
+      "learning_rate": 0.00041558157586765704,
+      "loss": 3.2693,
+      "step": 22901
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5579959750175476,
+      "learning_rate": 0.0004155673680633245,
+      "loss": 3.1617,
+      "step": 22902
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5807639360427856,
+      "learning_rate": 0.00041555315995460976,
+      "loss": 3.0796,
+      "step": 22903
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.547631025314331,
+      "learning_rate": 0.0004155389515415503,
+      "loss": 3.0593,
+      "step": 22904
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5586033463478088,
+      "learning_rate": 0.00041552474282418336,
+      "loss": 3.0915,
+      "step": 22905
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5403106212615967,
+      "learning_rate": 0.00041551053380254656,
+      "loss": 3.0695,
+      "step": 22906
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5964490175247192,
+      "learning_rate": 0.0004154963244766772,
+      "loss": 2.9246,
+      "step": 22907
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6330263018608093,
+      "learning_rate": 0.00041548211484661286,
+      "loss": 3.1909,
+      "step": 22908
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5445305109024048,
+      "learning_rate": 0.0004154679049123908,
+      "loss": 3.2143,
+      "step": 22909
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.49582967162132263,
+      "learning_rate": 0.0004154536946740484,
+      "loss": 3.0744,
+      "step": 22910
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6021095514297485,
+      "learning_rate": 0.0004154394841316233,
+      "loss": 3.1623,
+      "step": 22911
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5545721650123596,
+      "learning_rate": 0.00041542527328515273,
+      "loss": 3.0835,
+      "step": 22912
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5633933544158936,
+      "learning_rate": 0.0004154110621346743,
+      "loss": 3.0356,
+      "step": 22913
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5791369676589966,
+      "learning_rate": 0.00041539685068022527,
+      "loss": 3.1277,
+      "step": 22914
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.588784396648407,
+      "learning_rate": 0.0004153826389218432,
+      "loss": 3.1858,
+      "step": 22915
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5584728121757507,
+      "learning_rate": 0.0004153684268595654,
+      "loss": 2.9936,
+      "step": 22916
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6015306711196899,
+      "learning_rate": 0.00041535421449342936,
+      "loss": 3.0412,
+      "step": 22917
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6056309342384338,
+      "learning_rate": 0.00041534000182347257,
+      "loss": 2.8666,
+      "step": 22918
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5673311948776245,
+      "learning_rate": 0.00041532578884973247,
+      "loss": 3.0582,
+      "step": 22919
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5334855318069458,
+      "learning_rate": 0.0004153115755722464,
+      "loss": 3.0588,
+      "step": 22920
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.589484453201294,
+      "learning_rate": 0.0004152973619910517,
+      "loss": 3.1469,
+      "step": 22921
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5717774629592896,
+      "learning_rate": 0.0004152831481061861,
+      "loss": 2.9427,
+      "step": 22922
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6408734321594238,
+      "learning_rate": 0.00041526893391768683,
+      "loss": 3.0572,
+      "step": 22923
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5661062002182007,
+      "learning_rate": 0.0004152547194255913,
+      "loss": 2.8823,
+      "step": 22924
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5968921780586243,
+      "learning_rate": 0.0004152405046299371,
+      "loss": 3.0135,
+      "step": 22925
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5666815638542175,
+      "learning_rate": 0.0004152262895307616,
+      "loss": 3.0244,
+      "step": 22926
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5514220595359802,
+      "learning_rate": 0.00041521207412810213,
+      "loss": 2.9396,
+      "step": 22927
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5360330939292908,
+      "learning_rate": 0.0004151978584219963,
+      "loss": 3.1101,
+      "step": 22928
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5596000552177429,
+      "learning_rate": 0.0004151836424124815,
+      "loss": 3.379,
+      "step": 22929
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5417352318763733,
+      "learning_rate": 0.0004151694260995952,
+      "loss": 3.0873,
+      "step": 22930
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4918670952320099,
+      "learning_rate": 0.0004151552094833746,
+      "loss": 3.1007,
+      "step": 22931
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5756099224090576,
+      "learning_rate": 0.00041514099256385753,
+      "loss": 3.0216,
+      "step": 22932
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5461629033088684,
+      "learning_rate": 0.0004151267753410812,
+      "loss": 3.107,
+      "step": 22933
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5471493005752563,
+      "learning_rate": 0.00041511255781508306,
+      "loss": 3.1213,
+      "step": 22934
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.600498616695404,
+      "learning_rate": 0.00041509833998590063,
+      "loss": 3.1274,
+      "step": 22935
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5208067297935486,
+      "learning_rate": 0.00041508412185357126,
+      "loss": 3.1673,
+      "step": 22936
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5667810440063477,
+      "learning_rate": 0.00041506990341813257,
+      "loss": 3.0739,
+      "step": 22937
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5090360045433044,
+      "learning_rate": 0.00041505568467962187,
+      "loss": 3.2307,
+      "step": 22938
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5476797819137573,
+      "learning_rate": 0.0004150414656380766,
+      "loss": 3.0299,
+      "step": 22939
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5608437657356262,
+      "learning_rate": 0.00041502724629353423,
+      "loss": 3.0124,
+      "step": 22940
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5343563556671143,
+      "learning_rate": 0.0004150130266460322,
+      "loss": 3.0477,
+      "step": 22941
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5381127595901489,
+      "learning_rate": 0.0004149988066956081,
+      "loss": 3.1718,
+      "step": 22942
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5361109375953674,
+      "learning_rate": 0.00041498458644229926,
+      "loss": 3.0919,
+      "step": 22943
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5653386116027832,
+      "learning_rate": 0.0004149703658861431,
+      "loss": 3.1964,
+      "step": 22944
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5344790816307068,
+      "learning_rate": 0.0004149561450271772,
+      "loss": 2.9937,
+      "step": 22945
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5467535257339478,
+      "learning_rate": 0.00041494192386543885,
+      "loss": 3.1078,
+      "step": 22946
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6098953485488892,
+      "learning_rate": 0.0004149277024009655,
+      "loss": 3.178,
+      "step": 22947
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5292100310325623,
+      "learning_rate": 0.00041491348063379496,
+      "loss": 3.0706,
+      "step": 22948
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5830525159835815,
+      "learning_rate": 0.0004148992585639643,
+      "loss": 3.1763,
+      "step": 22949
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5466551184654236,
+      "learning_rate": 0.00041488503619151103,
+      "loss": 2.918,
+      "step": 22950
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5626237392425537,
+      "learning_rate": 0.00041487081351647274,
+      "loss": 2.8359,
+      "step": 22951
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6049579977989197,
+      "learning_rate": 0.00041485659053888696,
+      "loss": 3.1318,
+      "step": 22952
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5381505489349365,
+      "learning_rate": 0.0004148423672587908,
+      "loss": 3.1775,
+      "step": 22953
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5498600006103516,
+      "learning_rate": 0.0004148281436762221,
+      "loss": 3.1253,
+      "step": 22954
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5618240237236023,
+      "learning_rate": 0.00041481391979121817,
+      "loss": 2.8616,
+      "step": 22955
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5557640194892883,
+      "learning_rate": 0.00041479969560381645,
+      "loss": 3.1876,
+      "step": 22956
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5765591859817505,
+      "learning_rate": 0.0004147854711140543,
+      "loss": 3.2525,
+      "step": 22957
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5476953983306885,
+      "learning_rate": 0.0004147712463219694,
+      "loss": 2.9851,
+      "step": 22958
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5657317042350769,
+      "learning_rate": 0.00041475702122759924,
+      "loss": 2.9882,
+      "step": 22959
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5439834594726562,
+      "learning_rate": 0.00041474279583098104,
+      "loss": 3.3056,
+      "step": 22960
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5226112604141235,
+      "learning_rate": 0.0004147285701321525,
+      "loss": 3.0917,
+      "step": 22961
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5537917613983154,
+      "learning_rate": 0.0004147143441311509,
+      "loss": 3.0329,
+      "step": 22962
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5705544948577881,
+      "learning_rate": 0.00041470011782801373,
+      "loss": 3.3239,
+      "step": 22963
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5459992289543152,
+      "learning_rate": 0.0004146858912227786,
+      "loss": 3.0937,
+      "step": 22964
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6103513240814209,
+      "learning_rate": 0.00041467166431548297,
+      "loss": 3.2497,
+      "step": 22965
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.591571033000946,
+      "learning_rate": 0.00041465743710616417,
+      "loss": 3.1043,
+      "step": 22966
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5303111672401428,
+      "learning_rate": 0.00041464320959485977,
+      "loss": 3.111,
+      "step": 22967
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5746419429779053,
+      "learning_rate": 0.00041462898178160713,
+      "loss": 2.8603,
+      "step": 22968
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5438621044158936,
+      "learning_rate": 0.00041461475366644387,
+      "loss": 3.2288,
+      "step": 22969
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6318126320838928,
+      "learning_rate": 0.00041460052524940755,
+      "loss": 3.101,
+      "step": 22970
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.549228847026825,
+      "learning_rate": 0.00041458629653053536,
+      "loss": 3.118,
+      "step": 22971
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5160179138183594,
+      "learning_rate": 0.0004145720675098649,
+      "loss": 2.8717,
+      "step": 22972
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5238551497459412,
+      "learning_rate": 0.0004145578381874337,
+      "loss": 3.0346,
+      "step": 22973
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5568090677261353,
+      "learning_rate": 0.00041454360856327927,
+      "loss": 3.0845,
+      "step": 22974
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5128166079521179,
+      "learning_rate": 0.0004145293786374389,
+      "loss": 3.0882,
+      "step": 22975
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5460500717163086,
+      "learning_rate": 0.00041451514840995025,
+      "loss": 3.252,
+      "step": 22976
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5138880610466003,
+      "learning_rate": 0.00041450091788085075,
+      "loss": 3.0992,
+      "step": 22977
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5311721563339233,
+      "learning_rate": 0.0004144866870501778,
+      "loss": 3.0771,
+      "step": 22978
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5450507402420044,
+      "learning_rate": 0.000414472455917969,
+      "loss": 3.0126,
+      "step": 22979
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5414109230041504,
+      "learning_rate": 0.0004144582244842618,
+      "loss": 3.1002,
+      "step": 22980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5404634475708008,
+      "learning_rate": 0.0004144439927490937,
+      "loss": 2.9594,
+      "step": 22981
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.564774751663208,
+      "learning_rate": 0.000414429760712502,
+      "loss": 3.2406,
+      "step": 22982
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5353189706802368,
+      "learning_rate": 0.0004144155283745244,
+      "loss": 2.8896,
+      "step": 22983
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5611631274223328,
+      "learning_rate": 0.00041440129573519843,
+      "loss": 2.9384,
+      "step": 22984
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5437825918197632,
+      "learning_rate": 0.00041438706279456133,
+      "loss": 3.0938,
+      "step": 22985
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5270655155181885,
+      "learning_rate": 0.0004143728295526508,
+      "loss": 3.125,
+      "step": 22986
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5163417458534241,
+      "learning_rate": 0.0004143585960095042,
+      "loss": 3.1141,
+      "step": 22987
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5559436678886414,
+      "learning_rate": 0.0004143443621651591,
+      "loss": 3.045,
+      "step": 22988
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5473134517669678,
+      "learning_rate": 0.00041433012801965296,
+      "loss": 3.099,
+      "step": 22989
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.567561686038971,
+      "learning_rate": 0.00041431589357302325,
+      "loss": 3.1221,
+      "step": 22990
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5573049783706665,
+      "learning_rate": 0.0004143016588253074,
+      "loss": 3.0608,
+      "step": 22991
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5784956812858582,
+      "learning_rate": 0.00041428742377654315,
+      "loss": 3.14,
+      "step": 22992
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5378555655479431,
+      "learning_rate": 0.0004142731884267676,
+      "loss": 3.107,
+      "step": 22993
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5544320344924927,
+      "learning_rate": 0.00041425895277601865,
+      "loss": 2.9306,
+      "step": 22994
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.536886990070343,
+      "learning_rate": 0.00041424471682433353,
+      "loss": 3.1313,
+      "step": 22995
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5307275056838989,
+      "learning_rate": 0.00041423048057174984,
+      "loss": 2.9803,
+      "step": 22996
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5753107666969299,
+      "learning_rate": 0.00041421624401830504,
+      "loss": 3.0398,
+      "step": 22997
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5424846410751343,
+      "learning_rate": 0.00041420200716403663,
+      "loss": 3.042,
+      "step": 22998
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.549351692199707,
+      "learning_rate": 0.00041418777000898214,
+      "loss": 3.035,
+      "step": 22999
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5418702960014343,
+      "learning_rate": 0.000414173532553179,
+      "loss": 2.835,
+      "step": 23000
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5061715841293335,
+      "learning_rate": 0.0004141592947966647,
+      "loss": 3.0573,
+      "step": 23001
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5275315046310425,
+      "learning_rate": 0.00041414505673947687,
+      "loss": 3.2917,
+      "step": 23002
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5400534272193909,
+      "learning_rate": 0.00041413081838165294,
+      "loss": 2.8802,
+      "step": 23003
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5462265014648438,
+      "learning_rate": 0.0004141165797232304,
+      "loss": 3.0331,
+      "step": 23004
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.515167772769928,
+      "learning_rate": 0.00041410234076424664,
+      "loss": 3.0803,
+      "step": 23005
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.557414174079895,
+      "learning_rate": 0.0004140881015047394,
+      "loss": 3.1524,
+      "step": 23006
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5607942342758179,
+      "learning_rate": 0.000414073861944746,
+      "loss": 3.1207,
+      "step": 23007
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5383069515228271,
+      "learning_rate": 0.000414059622084304,
+      "loss": 3.0531,
+      "step": 23008
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5618544816970825,
+      "learning_rate": 0.0004140453819234509,
+      "loss": 3.1155,
+      "step": 23009
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5300397276878357,
+      "learning_rate": 0.0004140311414622243,
+      "loss": 3.2335,
+      "step": 23010
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5628443360328674,
+      "learning_rate": 0.0004140169007006615,
+      "loss": 3.3326,
+      "step": 23011
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5267999768257141,
+      "learning_rate": 0.0004140026596388002,
+      "loss": 3.1844,
+      "step": 23012
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5517027378082275,
+      "learning_rate": 0.0004139884182766778,
+      "loss": 3.1886,
+      "step": 23013
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5515194535255432,
+      "learning_rate": 0.00041397417661433183,
+      "loss": 3.0972,
+      "step": 23014
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5435203313827515,
+      "learning_rate": 0.00041395993465179984,
+      "loss": 3.0578,
+      "step": 23015
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5524570941925049,
+      "learning_rate": 0.0004139456923891193,
+      "loss": 3.0469,
+      "step": 23016
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5539202094078064,
+      "learning_rate": 0.0004139314498263278,
+      "loss": 3.0198,
+      "step": 23017
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5375218987464905,
+      "learning_rate": 0.00041391720696346267,
+      "loss": 3.1131,
+      "step": 23018
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6372843384742737,
+      "learning_rate": 0.00041390296380056156,
+      "loss": 3.2209,
+      "step": 23019
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5595908164978027,
+      "learning_rate": 0.00041388872033766203,
+      "loss": 3.2019,
+      "step": 23020
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5718174576759338,
+      "learning_rate": 0.00041387447657480154,
+      "loss": 3.0502,
+      "step": 23021
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5092378854751587,
+      "learning_rate": 0.0004138602325120176,
+      "loss": 3.2092,
+      "step": 23022
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.507887601852417,
+      "learning_rate": 0.0004138459881493476,
+      "loss": 3.2074,
+      "step": 23023
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.502585232257843,
+      "learning_rate": 0.0004138317434868293,
+      "loss": 3.1789,
+      "step": 23024
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5779663920402527,
+      "learning_rate": 0.0004138174985245001,
+      "loss": 3.0216,
+      "step": 23025
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.580353856086731,
+      "learning_rate": 0.0004138032532623974,
+      "loss": 3.0584,
+      "step": 23026
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5854467153549194,
+      "learning_rate": 0.00041378900770055896,
+      "loss": 3.0347,
+      "step": 23027
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5568118095397949,
+      "learning_rate": 0.0004137747618390221,
+      "loss": 2.9267,
+      "step": 23028
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5543918609619141,
+      "learning_rate": 0.0004137605156778244,
+      "loss": 3.0005,
+      "step": 23029
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5414887070655823,
+      "learning_rate": 0.00041374626921700345,
+      "loss": 3.0285,
+      "step": 23030
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5289766788482666,
+      "learning_rate": 0.00041373202245659663,
+      "loss": 2.9147,
+      "step": 23031
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5012800097465515,
+      "learning_rate": 0.0004137177753966417,
+      "loss": 2.962,
+      "step": 23032
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5366766452789307,
+      "learning_rate": 0.0004137035280371759,
+      "loss": 2.9534,
+      "step": 23033
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.524847149848938,
+      "learning_rate": 0.0004136892803782369,
+      "loss": 3.2415,
+      "step": 23034
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5566405057907104,
+      "learning_rate": 0.0004136750324198623,
+      "loss": 3.0081,
+      "step": 23035
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5404036045074463,
+      "learning_rate": 0.0004136607841620895,
+      "loss": 3.1826,
+      "step": 23036
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5708770155906677,
+      "learning_rate": 0.00041364653560495605,
+      "loss": 2.9988,
+      "step": 23037
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.557307779788971,
+      "learning_rate": 0.00041363228674849955,
+      "loss": 2.9934,
+      "step": 23038
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5527191758155823,
+      "learning_rate": 0.0004136180375927574,
+      "loss": 3.2225,
+      "step": 23039
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5346764922142029,
+      "learning_rate": 0.0004136037881377672,
+      "loss": 3.1241,
+      "step": 23040
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5354068875312805,
+      "learning_rate": 0.0004135895383835666,
+      "loss": 3.0392,
+      "step": 23041
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5253552198410034,
+      "learning_rate": 0.00041357528833019285,
+      "loss": 3.235,
+      "step": 23042
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5762072205543518,
+      "learning_rate": 0.0004135610379776838,
+      "loss": 3.0428,
+      "step": 23043
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.554805338382721,
+      "learning_rate": 0.00041354678732607677,
+      "loss": 3.0625,
+      "step": 23044
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5384538173675537,
+      "learning_rate": 0.00041353253637540935,
+      "loss": 2.9583,
+      "step": 23045
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5378395915031433,
+      "learning_rate": 0.00041351828512571913,
+      "loss": 3.2187,
+      "step": 23046
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5567187070846558,
+      "learning_rate": 0.0004135040335770436,
+      "loss": 3.2139,
+      "step": 23047
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.504665732383728,
+      "learning_rate": 0.0004134897817294202,
+      "loss": 3.2191,
+      "step": 23048
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5528210997581482,
+      "learning_rate": 0.0004134755295828865,
+      "loss": 3.1749,
+      "step": 23049
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5319205522537231,
+      "learning_rate": 0.00041346127713748023,
+      "loss": 3.1229,
+      "step": 23050
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5349425077438354,
+      "learning_rate": 0.00041344702439323873,
+      "loss": 2.9538,
+      "step": 23051
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5903375148773193,
+      "learning_rate": 0.0004134327713501997,
+      "loss": 3.0624,
+      "step": 23052
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5711473226547241,
+      "learning_rate": 0.0004134185180084004,
+      "loss": 3.2639,
+      "step": 23053
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5508195161819458,
+      "learning_rate": 0.0004134042643678787,
+      "loss": 2.9987,
+      "step": 23054
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5824445486068726,
+      "learning_rate": 0.00041339001042867195,
+      "loss": 3.1495,
+      "step": 23055
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5956960320472717,
+      "learning_rate": 0.0004133757561908177,
+      "loss": 3.384,
+      "step": 23056
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5386295318603516,
+      "learning_rate": 0.0004133615016543536,
+      "loss": 3.0116,
+      "step": 23057
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5671989321708679,
+      "learning_rate": 0.0004133472468193171,
+      "loss": 2.8844,
+      "step": 23058
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5901870727539062,
+      "learning_rate": 0.00041333299168574565,
+      "loss": 3.076,
+      "step": 23059
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.566315233707428,
+      "learning_rate": 0.00041331873625367703,
+      "loss": 3.1643,
+      "step": 23060
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5477898716926575,
+      "learning_rate": 0.0004133044805231487,
+      "loss": 3.3274,
+      "step": 23061
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.54547119140625,
+      "learning_rate": 0.0004132902244941981,
+      "loss": 3.2076,
+      "step": 23062
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5427873730659485,
+      "learning_rate": 0.0004132759681668629,
+      "loss": 3.1031,
+      "step": 23063
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6867091655731201,
+      "learning_rate": 0.0004132617115411805,
+      "loss": 2.9959,
+      "step": 23064
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6565173864364624,
+      "learning_rate": 0.00041324745461718867,
+      "loss": 3.1724,
+      "step": 23065
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5303226709365845,
+      "learning_rate": 0.0004132331973949248,
+      "loss": 3.0475,
+      "step": 23066
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.537520170211792,
+      "learning_rate": 0.00041321893987442647,
+      "loss": 3.1931,
+      "step": 23067
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5558544397354126,
+      "learning_rate": 0.00041320468205573125,
+      "loss": 2.9065,
+      "step": 23068
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5704731941223145,
+      "learning_rate": 0.00041319042393887675,
+      "loss": 3.0102,
+      "step": 23069
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5413227081298828,
+      "learning_rate": 0.0004131761655239003,
+      "loss": 2.8069,
+      "step": 23070
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5648303627967834,
+      "learning_rate": 0.00041316190681083963,
+      "loss": 3.0567,
+      "step": 23071
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5538491010665894,
+      "learning_rate": 0.00041314764779973247,
+      "loss": 3.1438,
+      "step": 23072
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5491966605186462,
+      "learning_rate": 0.0004131333884906161,
+      "loss": 2.889,
+      "step": 23073
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.572340190410614,
+      "learning_rate": 0.000413119128883528,
+      "loss": 3.1123,
+      "step": 23074
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5277241468429565,
+      "learning_rate": 0.00041310486897850604,
+      "loss": 3.0489,
+      "step": 23075
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5485084652900696,
+      "learning_rate": 0.0004130906087755876,
+      "loss": 3.134,
+      "step": 23076
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.510722815990448,
+      "learning_rate": 0.0004130763482748102,
+      "loss": 3.1496,
+      "step": 23077
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5820092558860779,
+      "learning_rate": 0.0004130620874762115,
+      "loss": 2.9881,
+      "step": 23078
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5545257329940796,
+      "learning_rate": 0.00041304782637982903,
+      "loss": 3.1989,
+      "step": 23079
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5428977012634277,
+      "learning_rate": 0.00041303356498570036,
+      "loss": 2.969,
+      "step": 23080
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5307221412658691,
+      "learning_rate": 0.000413019303293863,
+      "loss": 3.1404,
+      "step": 23081
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5607012510299683,
+      "learning_rate": 0.0004130050413043545,
+      "loss": 3.0869,
+      "step": 23082
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5869443416595459,
+      "learning_rate": 0.00041299077901721257,
+      "loss": 3.0999,
+      "step": 23083
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5176264047622681,
+      "learning_rate": 0.00041297651643247465,
+      "loss": 2.937,
+      "step": 23084
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5371460318565369,
+      "learning_rate": 0.00041296225355017826,
+      "loss": 2.9182,
+      "step": 23085
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5200726985931396,
+      "learning_rate": 0.00041294799037036107,
+      "loss": 3.0559,
+      "step": 23086
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5590261220932007,
+      "learning_rate": 0.0004129337268930607,
+      "loss": 3.2351,
+      "step": 23087
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5306074023246765,
+      "learning_rate": 0.00041291946311831446,
+      "loss": 3.2695,
+      "step": 23088
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5653502345085144,
+      "learning_rate": 0.0004129051990461602,
+      "loss": 3.1316,
+      "step": 23089
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5234277248382568,
+      "learning_rate": 0.00041289093467663536,
+      "loss": 3.009,
+      "step": 23090
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5736457109451294,
+      "learning_rate": 0.0004128766700097775,
+      "loss": 3.1699,
+      "step": 23091
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.607955276966095,
+      "learning_rate": 0.00041286240504562416,
+      "loss": 2.9498,
+      "step": 23092
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5673606395721436,
+      "learning_rate": 0.00041284813978421307,
+      "loss": 3.0839,
+      "step": 23093
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5501077175140381,
+      "learning_rate": 0.0004128338742255817,
+      "loss": 3.0898,
+      "step": 23094
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5964276790618896,
+      "learning_rate": 0.0004128196083697676,
+      "loss": 3.0892,
+      "step": 23095
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5602797865867615,
+      "learning_rate": 0.00041280534221680825,
+      "loss": 3.0006,
+      "step": 23096
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5748618841171265,
+      "learning_rate": 0.00041279107576674154,
+      "loss": 2.9523,
+      "step": 23097
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5489750504493713,
+      "learning_rate": 0.0004127768090196047,
+      "loss": 3.1644,
+      "step": 23098
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5585864186286926,
+      "learning_rate": 0.00041276254197543545,
+      "loss": 3.1808,
+      "step": 23099
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6432881951332092,
+      "learning_rate": 0.0004127482746342714,
+      "loss": 2.9174,
+      "step": 23100
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5519260764122009,
+      "learning_rate": 0.0004127340069961501,
+      "loss": 3.0454,
+      "step": 23101
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5701799988746643,
+      "learning_rate": 0.00041271973906110916,
+      "loss": 3.0101,
+      "step": 23102
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5463408827781677,
+      "learning_rate": 0.000412705470829186,
+      "loss": 3.3575,
+      "step": 23103
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5591546297073364,
+      "learning_rate": 0.0004126912023004184,
+      "loss": 2.722,
+      "step": 23104
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6106502413749695,
+      "learning_rate": 0.00041267693347484403,
+      "loss": 3.2153,
+      "step": 23105
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5445353984832764,
+      "learning_rate": 0.00041266266435250004,
+      "loss": 3.1634,
+      "step": 23106
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5435615181922913,
+      "learning_rate": 0.00041264839493342434,
+      "loss": 3.0936,
+      "step": 23107
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5400312542915344,
+      "learning_rate": 0.00041263412521765454,
+      "loss": 3.1408,
+      "step": 23108
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5756552219390869,
+      "learning_rate": 0.00041261985520522806,
+      "loss": 3.2401,
+      "step": 23109
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.573711633682251,
+      "learning_rate": 0.00041260558489618256,
+      "loss": 3.0456,
+      "step": 23110
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5202263593673706,
+      "learning_rate": 0.0004125913142905555,
+      "loss": 2.958,
+      "step": 23111
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5493197441101074,
+      "learning_rate": 0.0004125770433883848,
+      "loss": 3.1262,
+      "step": 23112
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5797801613807678,
+      "learning_rate": 0.00041256277218970774,
+      "loss": 3.0246,
+      "step": 23113
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5684946179389954,
+      "learning_rate": 0.000412548500694562,
+      "loss": 3.1813,
+      "step": 23114
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.54192054271698,
+      "learning_rate": 0.00041253422890298515,
+      "loss": 2.8924,
+      "step": 23115
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5381801724433899,
+      "learning_rate": 0.00041251995681501483,
+      "loss": 3.0613,
+      "step": 23116
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5283708572387695,
+      "learning_rate": 0.0004125056844306886,
+      "loss": 3.0999,
+      "step": 23117
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5617417693138123,
+      "learning_rate": 0.000412491411750044,
+      "loss": 2.9397,
+      "step": 23118
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6005975604057312,
+      "learning_rate": 0.0004124771387731188,
+      "loss": 3.3285,
+      "step": 23119
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5726503729820251,
+      "learning_rate": 0.00041246286549995035,
+      "loss": 3.1522,
+      "step": 23120
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5579819679260254,
+      "learning_rate": 0.0004124485919305763,
+      "loss": 3.278,
+      "step": 23121
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5398461818695068,
+      "learning_rate": 0.00041243431806503437,
+      "loss": 2.8825,
+      "step": 23122
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5483943819999695,
+      "learning_rate": 0.00041242004390336216,
+      "loss": 3.1911,
+      "step": 23123
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5343546867370605,
+      "learning_rate": 0.0004124057694455971,
+      "loss": 3.0709,
+      "step": 23124
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5518394708633423,
+      "learning_rate": 0.0004123914946917769,
+      "loss": 3.088,
+      "step": 23125
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.541649580001831,
+      "learning_rate": 0.00041237721964193914,
+      "loss": 2.9646,
+      "step": 23126
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5012128949165344,
+      "learning_rate": 0.00041236294429612143,
+      "loss": 3.0232,
+      "step": 23127
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5622835755348206,
+      "learning_rate": 0.0004123486686543613,
+      "loss": 3.1136,
+      "step": 23128
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5492423176765442,
+      "learning_rate": 0.00041233439271669644,
+      "loss": 3.2057,
+      "step": 23129
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5439629554748535,
+      "learning_rate": 0.0004123201164831644,
+      "loss": 2.9353,
+      "step": 23130
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5032997727394104,
+      "learning_rate": 0.0004123058399538027,
+      "loss": 3.1584,
+      "step": 23131
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5795646905899048,
+      "learning_rate": 0.0004122915631286492,
+      "loss": 3.1761,
+      "step": 23132
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5473034381866455,
+      "learning_rate": 0.0004122772860077412,
+      "loss": 2.8715,
+      "step": 23133
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5456127524375916,
+      "learning_rate": 0.0004122630085911165,
+      "loss": 2.9562,
+      "step": 23134
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5090837478637695,
+      "learning_rate": 0.00041224873087881265,
+      "loss": 3.0014,
+      "step": 23135
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5396968126296997,
+      "learning_rate": 0.00041223445287086715,
+      "loss": 3.173,
+      "step": 23136
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5504181981086731,
+      "learning_rate": 0.0004122201745673178,
+      "loss": 3.0763,
+      "step": 23137
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.551391065120697,
+      "learning_rate": 0.00041220589596820204,
+      "loss": 3.0412,
+      "step": 23138
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5569359064102173,
+      "learning_rate": 0.0004121916170735577,
+      "loss": 3.136,
+      "step": 23139
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5515356063842773,
+      "learning_rate": 0.00041217733788342204,
+      "loss": 3.1659,
+      "step": 23140
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5313219428062439,
+      "learning_rate": 0.0004121630583978329,
+      "loss": 2.951,
+      "step": 23141
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5136998295783997,
+      "learning_rate": 0.00041214877861682795,
+      "loss": 2.9565,
+      "step": 23142
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.531518816947937,
+      "learning_rate": 0.00041213449854044464,
+      "loss": 3.2081,
+      "step": 23143
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5740218758583069,
+      "learning_rate": 0.0004121202181687206,
+      "loss": 3.0908,
+      "step": 23144
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5602694153785706,
+      "learning_rate": 0.0004121059375016936,
+      "loss": 3.0052,
+      "step": 23145
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5783454775810242,
+      "learning_rate": 0.0004120916565394011,
+      "loss": 3.056,
+      "step": 23146
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5084561705589294,
+      "learning_rate": 0.00041207737528188056,
+      "loss": 3.152,
+      "step": 23147
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5585759878158569,
+      "learning_rate": 0.00041206309372917,
+      "loss": 3.1573,
+      "step": 23148
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5565611720085144,
+      "learning_rate": 0.00041204881188130674,
+      "loss": 3.1032,
+      "step": 23149
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5540475249290466,
+      "learning_rate": 0.00041203452973832843,
+      "loss": 2.8712,
+      "step": 23150
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5279893279075623,
+      "learning_rate": 0.0004120202473002728,
+      "loss": 3.0523,
+      "step": 23151
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5814106464385986,
+      "learning_rate": 0.0004120059645671774,
+      "loss": 3.2728,
+      "step": 23152
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5316987633705139,
+      "learning_rate": 0.0004119916815390798,
+      "loss": 3.185,
+      "step": 23153
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5352542996406555,
+      "learning_rate": 0.00041197739821601767,
+      "loss": 2.9784,
+      "step": 23154
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.587202250957489,
+      "learning_rate": 0.00041196311459802866,
+      "loss": 3.1149,
+      "step": 23155
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5219794511795044,
+      "learning_rate": 0.0004119488306851504,
+      "loss": 3.1565,
+      "step": 23156
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5088773965835571,
+      "learning_rate": 0.0004119345464774203,
+      "loss": 2.8584,
+      "step": 23157
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5774076581001282,
+      "learning_rate": 0.0004119202619748763,
+      "loss": 3.02,
+      "step": 23158
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5148767232894897,
+      "learning_rate": 0.0004119059771775559,
+      "loss": 3.0098,
+      "step": 23159
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5569112300872803,
+      "learning_rate": 0.00041189169208549655,
+      "loss": 3.0286,
+      "step": 23160
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5477631688117981,
+      "learning_rate": 0.00041187740669873605,
+      "loss": 3.2834,
+      "step": 23161
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5244240164756775,
+      "learning_rate": 0.00041186312101731195,
+      "loss": 3.1129,
+      "step": 23162
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6075512170791626,
+      "learning_rate": 0.0004118488350412621,
+      "loss": 3.1451,
+      "step": 23163
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5244618058204651,
+      "learning_rate": 0.0004118345487706238,
+      "loss": 2.9186,
+      "step": 23164
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5191154479980469,
+      "learning_rate": 0.00041182026220543485,
+      "loss": 3.2372,
+      "step": 23165
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.540847897529602,
+      "learning_rate": 0.00041180597534573284,
+      "loss": 2.9988,
+      "step": 23166
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5709373950958252,
+      "learning_rate": 0.0004117916881915554,
+      "loss": 3.1161,
+      "step": 23167
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6198864579200745,
+      "learning_rate": 0.0004117774007429402,
+      "loss": 3.0857,
+      "step": 23168
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5441683530807495,
+      "learning_rate": 0.00041176311299992484,
+      "loss": 3.1707,
+      "step": 23169
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5470789670944214,
+      "learning_rate": 0.000411748824962547,
+      "loss": 3.0281,
+      "step": 23170
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5256059169769287,
+      "learning_rate": 0.00041173453663084417,
+      "loss": 3.274,
+      "step": 23171
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5547487735748291,
+      "learning_rate": 0.00041172024800485403,
+      "loss": 3.1233,
+      "step": 23172
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5724942684173584,
+      "learning_rate": 0.00041170595908461436,
+      "loss": 3.2013,
+      "step": 23173
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5241184830665588,
+      "learning_rate": 0.00041169166987016265,
+      "loss": 3.3218,
+      "step": 23174
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5703988671302795,
+      "learning_rate": 0.00041167738036153664,
+      "loss": 3.1718,
+      "step": 23175
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5422783493995667,
+      "learning_rate": 0.0004116630905587738,
+      "loss": 3.011,
+      "step": 23176
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.581802248954773,
+      "learning_rate": 0.00041164880046191195,
+      "loss": 2.9262,
+      "step": 23177
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5663591027259827,
+      "learning_rate": 0.0004116345100709886,
+      "loss": 3.3378,
+      "step": 23178
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5546205639839172,
+      "learning_rate": 0.00041162021938604147,
+      "loss": 3.0222,
+      "step": 23179
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.583781361579895,
+      "learning_rate": 0.0004116059284071081,
+      "loss": 3.0932,
+      "step": 23180
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5724231004714966,
+      "learning_rate": 0.0004115916371342263,
+      "loss": 3.3044,
+      "step": 23181
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5571467876434326,
+      "learning_rate": 0.0004115773455674336,
+      "loss": 3.1352,
+      "step": 23182
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5352392792701721,
+      "learning_rate": 0.00041156305370676754,
+      "loss": 3.0393,
+      "step": 23183
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5406389236450195,
+      "learning_rate": 0.0004115487615522658,
+      "loss": 3.0062,
+      "step": 23184
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6502759456634521,
+      "learning_rate": 0.0004115344691039663,
+      "loss": 2.9669,
+      "step": 23185
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5307773351669312,
+      "learning_rate": 0.00041152017636190643,
+      "loss": 3.3796,
+      "step": 23186
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5802614092826843,
+      "learning_rate": 0.00041150588332612383,
+      "loss": 3.2169,
+      "step": 23187
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5349951386451721,
+      "learning_rate": 0.0004114915899966561,
+      "loss": 3.144,
+      "step": 23188
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5569307804107666,
+      "learning_rate": 0.0004114772963735411,
+      "loss": 2.9595,
+      "step": 23189
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5304500460624695,
+      "learning_rate": 0.0004114630024568163,
+      "loss": 3.0931,
+      "step": 23190
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5246667861938477,
+      "learning_rate": 0.00041144870824651945,
+      "loss": 3.2074,
+      "step": 23191
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.524143397808075,
+      "learning_rate": 0.00041143441374268824,
+      "loss": 2.7597,
+      "step": 23192
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5903804898262024,
+      "learning_rate": 0.00041142011894536003,
+      "loss": 3.0905,
+      "step": 23193
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5838131308555603,
+      "learning_rate": 0.0004114058238545728,
+      "loss": 3.0597,
+      "step": 23194
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5308125615119934,
+      "learning_rate": 0.000411391528470364,
+      "loss": 3.0877,
+      "step": 23195
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5069236755371094,
+      "learning_rate": 0.0004113772327927714,
+      "loss": 3.1319,
+      "step": 23196
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.533857524394989,
+      "learning_rate": 0.0004113629368218326,
+      "loss": 2.9676,
+      "step": 23197
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5685359835624695,
+      "learning_rate": 0.00041134864055758513,
+      "loss": 3.0839,
+      "step": 23198
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6156092286109924,
+      "learning_rate": 0.00041133434400006695,
+      "loss": 3.2202,
+      "step": 23199
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5481438636779785,
+      "learning_rate": 0.0004113200471493155,
+      "loss": 3.1862,
+      "step": 23200
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5097211599349976,
+      "learning_rate": 0.0004113057500053684,
+      "loss": 3.0117,
+      "step": 23201
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5516743063926697,
+      "learning_rate": 0.00041129145256826345,
+      "loss": 3.0283,
+      "step": 23202
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5222983360290527,
+      "learning_rate": 0.0004112771548380382,
+      "loss": 2.9669,
+      "step": 23203
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5839381217956543,
+      "learning_rate": 0.0004112628568147303,
+      "loss": 3.015,
+      "step": 23204
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5553185939788818,
+      "learning_rate": 0.0004112485584983775,
+      "loss": 2.915,
+      "step": 23205
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5800415873527527,
+      "learning_rate": 0.0004112342598890174,
+      "loss": 3.1186,
+      "step": 23206
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.569153904914856,
+      "learning_rate": 0.0004112199609866877,
+      "loss": 2.9612,
+      "step": 23207
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5515949130058289,
+      "learning_rate": 0.00041120566179142596,
+      "loss": 3.1817,
+      "step": 23208
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5314306020736694,
+      "learning_rate": 0.00041119136230326984,
+      "loss": 3.0635,
+      "step": 23209
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5693814754486084,
+      "learning_rate": 0.00041117706252225727,
+      "loss": 3.0755,
+      "step": 23210
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5441717505455017,
+      "learning_rate": 0.0004111627624484256,
+      "loss": 3.1078,
+      "step": 23211
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5225312113761902,
+      "learning_rate": 0.0004111484620818126,
+      "loss": 2.992,
+      "step": 23212
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5491369962692261,
+      "learning_rate": 0.00041113416142245587,
+      "loss": 2.9803,
+      "step": 23213
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.530925989151001,
+      "learning_rate": 0.00041111986047039333,
+      "loss": 3.0763,
+      "step": 23214
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5920950174331665,
+      "learning_rate": 0.00041110555922566235,
+      "loss": 3.195,
+      "step": 23215
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.544621467590332,
+      "learning_rate": 0.0004110912576883007,
+      "loss": 3.1061,
+      "step": 23216
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.526262640953064,
+      "learning_rate": 0.0004110769558583461,
+      "loss": 3.167,
+      "step": 23217
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5441073179244995,
+      "learning_rate": 0.00041106265373583615,
+      "loss": 3.0507,
+      "step": 23218
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.529232382774353,
+      "learning_rate": 0.00041104835132080856,
+      "loss": 3.1984,
+      "step": 23219
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5716642141342163,
+      "learning_rate": 0.00041103404861330095,
+      "loss": 3.1493,
+      "step": 23220
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5743255019187927,
+      "learning_rate": 0.0004110197456133511,
+      "loss": 3.0924,
+      "step": 23221
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6008023023605347,
+      "learning_rate": 0.0004110054423209966,
+      "loss": 3.0093,
+      "step": 23222
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5539536476135254,
+      "learning_rate": 0.00041099113873627505,
+      "loss": 3.0168,
+      "step": 23223
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5438982248306274,
+      "learning_rate": 0.0004109768348592242,
+      "loss": 2.9509,
+      "step": 23224
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5372403860092163,
+      "learning_rate": 0.0004109625306898818,
+      "loss": 3.1003,
+      "step": 23225
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5633687376976013,
+      "learning_rate": 0.0004109482262282854,
+      "loss": 3.1442,
+      "step": 23226
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6077851057052612,
+      "learning_rate": 0.0004109339214744727,
+      "loss": 3.112,
+      "step": 23227
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5495371222496033,
+      "learning_rate": 0.00041091961642848143,
+      "loss": 3.1602,
+      "step": 23228
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5324305295944214,
+      "learning_rate": 0.0004109053110903493,
+      "loss": 3.0703,
+      "step": 23229
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5331787467002869,
+      "learning_rate": 0.00041089100546011384,
+      "loss": 3.1745,
+      "step": 23230
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5928345918655396,
+      "learning_rate": 0.00041087669953781284,
+      "loss": 3.0949,
+      "step": 23231
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5761058330535889,
+      "learning_rate": 0.0004108623933234841,
+      "loss": 3.163,
+      "step": 23232
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5908313989639282,
+      "learning_rate": 0.0004108480868171649,
+      "loss": 3.2201,
+      "step": 23233
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5895100235939026,
+      "learning_rate": 0.00041083378001889327,
+      "loss": 2.8646,
+      "step": 23234
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5367791652679443,
+      "learning_rate": 0.0004108194729287068,
+      "loss": 3.2775,
+      "step": 23235
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5391991138458252,
+      "learning_rate": 0.00041080516554664327,
+      "loss": 3.1389,
+      "step": 23236
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5764256119728088,
+      "learning_rate": 0.00041079085787274015,
+      "loss": 2.9448,
+      "step": 23237
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5493453145027161,
+      "learning_rate": 0.0004107765499070352,
+      "loss": 3.2057,
+      "step": 23238
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5182337760925293,
+      "learning_rate": 0.00041076224164956614,
+      "loss": 3.0028,
+      "step": 23239
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5463001132011414,
+      "learning_rate": 0.0004107479331003707,
+      "loss": 3.3298,
+      "step": 23240
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5629000067710876,
+      "learning_rate": 0.0004107336242594865,
+      "loss": 3.131,
+      "step": 23241
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5420428514480591,
+      "learning_rate": 0.0004107193151269512,
+      "loss": 3.1846,
+      "step": 23242
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5327157974243164,
+      "learning_rate": 0.0004107050057028027,
+      "loss": 3.1262,
+      "step": 23243
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5222551822662354,
+      "learning_rate": 0.0004106906959870783,
+      "loss": 3.1124,
+      "step": 23244
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5623447895050049,
+      "learning_rate": 0.00041067638597981604,
+      "loss": 3.1196,
+      "step": 23245
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5250970721244812,
+      "learning_rate": 0.00041066207568105344,
+      "loss": 3.1107,
+      "step": 23246
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5632120966911316,
+      "learning_rate": 0.0004106477650908283,
+      "loss": 3.0244,
+      "step": 23247
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5749978423118591,
+      "learning_rate": 0.0004106334542091782,
+      "loss": 3.0074,
+      "step": 23248
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.548491358757019,
+      "learning_rate": 0.00041061914303614085,
+      "loss": 2.8617,
+      "step": 23249
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5380204319953918,
+      "learning_rate": 0.00041060483157175396,
+      "loss": 3.0328,
+      "step": 23250
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5522347092628479,
+      "learning_rate": 0.00041059051981605523,
+      "loss": 3.1683,
+      "step": 23251
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5788419842720032,
+      "learning_rate": 0.00041057620776908244,
+      "loss": 3.1031,
+      "step": 23252
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5642889738082886,
+      "learning_rate": 0.0004105618954308731,
+      "loss": 2.9625,
+      "step": 23253
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5192211866378784,
+      "learning_rate": 0.00041054758280146506,
+      "loss": 3.1176,
+      "step": 23254
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5606915950775146,
+      "learning_rate": 0.0004105332698808959,
+      "loss": 2.9903,
+      "step": 23255
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5417342782020569,
+      "learning_rate": 0.0004105189566692034,
+      "loss": 2.9431,
+      "step": 23256
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5401815176010132,
+      "learning_rate": 0.00041050464316642524,
+      "loss": 2.9124,
+      "step": 23257
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5581971406936646,
+      "learning_rate": 0.0004104903293725992,
+      "loss": 3.1402,
+      "step": 23258
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5499171018600464,
+      "learning_rate": 0.0004104760152877628,
+      "loss": 3.1808,
+      "step": 23259
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5453315377235413,
+      "learning_rate": 0.0004104617009119539,
+      "loss": 3.0836,
+      "step": 23260
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5436408519744873,
+      "learning_rate": 0.0004104473862452101,
+      "loss": 3.2032,
+      "step": 23261
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.595960795879364,
+      "learning_rate": 0.00041043307128756914,
+      "loss": 3.0531,
+      "step": 23262
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5662114024162292,
+      "learning_rate": 0.0004104187560390687,
+      "loss": 2.9397,
+      "step": 23263
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.548102080821991,
+      "learning_rate": 0.00041040444049974655,
+      "loss": 2.9091,
+      "step": 23264
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5476301312446594,
+      "learning_rate": 0.00041039012466964034,
+      "loss": 2.8001,
+      "step": 23265
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5653447508811951,
+      "learning_rate": 0.00041037580854878775,
+      "loss": 3.2257,
+      "step": 23266
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5923088788986206,
+      "learning_rate": 0.00041036149213722657,
+      "loss": 3.1438,
+      "step": 23267
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5683751702308655,
+      "learning_rate": 0.0004103471754349945,
+      "loss": 3.0826,
+      "step": 23268
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5276745557785034,
+      "learning_rate": 0.00041033285844212914,
+      "loss": 3.1649,
+      "step": 23269
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5372428894042969,
+      "learning_rate": 0.0004103185411586682,
+      "loss": 3.1504,
+      "step": 23270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5391978621482849,
+      "learning_rate": 0.00041030422358464955,
+      "loss": 3.1403,
+      "step": 23271
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5556966066360474,
+      "learning_rate": 0.0004102899057201108,
+      "loss": 3.1582,
+      "step": 23272
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5490073561668396,
+      "learning_rate": 0.00041027558756508967,
+      "loss": 3.1393,
+      "step": 23273
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5323960185050964,
+      "learning_rate": 0.00041026126911962386,
+      "loss": 3.0192,
+      "step": 23274
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5622329711914062,
+      "learning_rate": 0.00041024695038375104,
+      "loss": 3.0085,
+      "step": 23275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5242308378219604,
+      "learning_rate": 0.00041023263135750904,
+      "loss": 3.1784,
+      "step": 23276
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5954573154449463,
+      "learning_rate": 0.0004102183120409354,
+      "loss": 3.0906,
+      "step": 23277
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.542190670967102,
+      "learning_rate": 0.000410203992434068,
+      "loss": 3.102,
+      "step": 23278
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5081149339675903,
+      "learning_rate": 0.0004101896725369445,
+      "loss": 3.1372,
+      "step": 23279
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5064743161201477,
+      "learning_rate": 0.0004101753523496026,
+      "loss": 3.1823,
+      "step": 23280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6618224382400513,
+      "learning_rate": 0.0004101610318720801,
+      "loss": 3.0695,
+      "step": 23281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5303663015365601,
+      "learning_rate": 0.00041014671110441457,
+      "loss": 3.1745,
+      "step": 23282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5263445973396301,
+      "learning_rate": 0.0004101323900466438,
+      "loss": 3.022,
+      "step": 23283
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5309288501739502,
+      "learning_rate": 0.00041011806869880555,
+      "loss": 2.762,
+      "step": 23284
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5699345469474792,
+      "learning_rate": 0.00041010374706093735,
+      "loss": 2.9958,
+      "step": 23285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5486957430839539,
+      "learning_rate": 0.0004100894251330772,
+      "loss": 3.0072,
+      "step": 23286
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5336732864379883,
+      "learning_rate": 0.00041007510291526277,
+      "loss": 3.1703,
+      "step": 23287
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5408821702003479,
+      "learning_rate": 0.00041006078040753163,
+      "loss": 3.0636,
+      "step": 23288
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5387353897094727,
+      "learning_rate": 0.0004100464576099215,
+      "loss": 3.0886,
+      "step": 23289
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5669566988945007,
+      "learning_rate": 0.00041003213452247026,
+      "loss": 3.0922,
+      "step": 23290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5251073241233826,
+      "learning_rate": 0.0004100178111452155,
+      "loss": 3.3268,
+      "step": 23291
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.536972165107727,
+      "learning_rate": 0.000410003487478195,
+      "loss": 2.9341,
+      "step": 23292
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5253141522407532,
+      "learning_rate": 0.0004099891635214465,
+      "loss": 3.0662,
+      "step": 23293
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6211135983467102,
+      "learning_rate": 0.00040997483927500785,
+      "loss": 3.1949,
+      "step": 23294
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5622311234474182,
+      "learning_rate": 0.0004099605147389164,
+      "loss": 3.218,
+      "step": 23295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5925061702728271,
+      "learning_rate": 0.0004099461899132102,
+      "loss": 3.2212,
+      "step": 23296
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5344102382659912,
+      "learning_rate": 0.00040993186479792694,
+      "loss": 2.9887,
+      "step": 23297
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6124759316444397,
+      "learning_rate": 0.0004099175393931044,
+      "loss": 3.0576,
+      "step": 23298
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5534654259681702,
+      "learning_rate": 0.00040990321369878006,
+      "loss": 3.1516,
+      "step": 23299
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5637291073799133,
+      "learning_rate": 0.0004098888877149918,
+      "loss": 2.9373,
+      "step": 23300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5289487242698669,
+      "learning_rate": 0.00040987456144177747,
+      "loss": 3.0108,
+      "step": 23301
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5784006118774414,
+      "learning_rate": 0.0004098602348791746,
+      "loss": 2.9401,
+      "step": 23302
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5589538216590881,
+      "learning_rate": 0.00040984590802722106,
+      "loss": 2.9519,
+      "step": 23303
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.573984682559967,
+      "learning_rate": 0.00040983158088595456,
+      "loss": 2.9528,
+      "step": 23304
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5732395648956299,
+      "learning_rate": 0.0004098172534554128,
+      "loss": 3.081,
+      "step": 23305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5892529487609863,
+      "learning_rate": 0.0004098029257356334,
+      "loss": 3.1453,
+      "step": 23306
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5673011541366577,
+      "learning_rate": 0.0004097885977266544,
+      "loss": 3.0574,
+      "step": 23307
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5459287166595459,
+      "learning_rate": 0.00040977426942851326,
+      "loss": 3.0539,
+      "step": 23308
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.545254647731781,
+      "learning_rate": 0.0004097599408412479,
+      "loss": 2.968,
+      "step": 23309
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5447995066642761,
+      "learning_rate": 0.0004097456119648959,
+      "loss": 2.9872,
+      "step": 23310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5695222020149231,
+      "learning_rate": 0.0004097312827994951,
+      "loss": 3.0334,
+      "step": 23311
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5799598693847656,
+      "learning_rate": 0.00040971695334508323,
+      "loss": 3.1335,
+      "step": 23312
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5451822876930237,
+      "learning_rate": 0.000409702623601698,
+      "loss": 3.1662,
+      "step": 23313
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5687471628189087,
+      "learning_rate": 0.00040968829356937726,
+      "loss": 3.1632,
+      "step": 23314
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5137524604797363,
+      "learning_rate": 0.00040967396324815853,
+      "loss": 3.1517,
+      "step": 23315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.535132110118866,
+      "learning_rate": 0.00040965963263807977,
+      "loss": 3.0008,
+      "step": 23316
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5807247161865234,
+      "learning_rate": 0.0004096453017391786,
+      "loss": 3.0634,
+      "step": 23317
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.585669994354248,
+      "learning_rate": 0.00040963097055149287,
+      "loss": 3.0915,
+      "step": 23318
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5763066411018372,
+      "learning_rate": 0.00040961663907506025,
+      "loss": 3.1631,
+      "step": 23319
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.563368558883667,
+      "learning_rate": 0.0004096023073099185,
+      "loss": 3.3913,
+      "step": 23320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5762029886245728,
+      "learning_rate": 0.00040958797525610525,
+      "loss": 3.0807,
+      "step": 23321
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5520102977752686,
+      "learning_rate": 0.0004095736429136585,
+      "loss": 3.1066,
+      "step": 23322
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.537263035774231,
+      "learning_rate": 0.0004095593102826159,
+      "loss": 3.1889,
+      "step": 23323
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5755485892295837,
+      "learning_rate": 0.00040954497736301503,
+      "loss": 3.0597,
+      "step": 23324
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5270916819572449,
+      "learning_rate": 0.00040953064415489383,
+      "loss": 3.1009,
+      "step": 23325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5502188801765442,
+      "learning_rate": 0.00040951631065828996,
+      "loss": 2.9506,
+      "step": 23326
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5408546328544617,
+      "learning_rate": 0.00040950197687324126,
+      "loss": 2.9593,
+      "step": 23327
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5552592873573303,
+      "learning_rate": 0.0004094876427997854,
+      "loss": 2.8134,
+      "step": 23328
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5457807779312134,
+      "learning_rate": 0.00040947330843796016,
+      "loss": 3.2202,
+      "step": 23329
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6020430326461792,
+      "learning_rate": 0.0004094589737878033,
+      "loss": 3.1423,
+      "step": 23330
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5497642159461975,
+      "learning_rate": 0.00040944463884935256,
+      "loss": 3.1488,
+      "step": 23331
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5718896389007568,
+      "learning_rate": 0.00040943030362264565,
+      "loss": 3.162,
+      "step": 23332
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5680598616600037,
+      "learning_rate": 0.00040941596810772045,
+      "loss": 3.1336,
+      "step": 23333
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5549580454826355,
+      "learning_rate": 0.0004094016323046147,
+      "loss": 3.2387,
+      "step": 23334
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.586732029914856,
+      "learning_rate": 0.00040938729621336605,
+      "loss": 3.0624,
+      "step": 23335
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5963259339332581,
+      "learning_rate": 0.00040937295983401226,
+      "loss": 2.9214,
+      "step": 23336
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5552912354469299,
+      "learning_rate": 0.00040935862316659116,
+      "loss": 3.0501,
+      "step": 23337
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.509372889995575,
+      "learning_rate": 0.0004093442862111406,
+      "loss": 3.0564,
+      "step": 23338
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5622632503509521,
+      "learning_rate": 0.00040932994896769815,
+      "loss": 3.0737,
+      "step": 23339
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5708377361297607,
+      "learning_rate": 0.0004093156114363016,
+      "loss": 3.2455,
+      "step": 23340
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5660071969032288,
+      "learning_rate": 0.00040930127361698887,
+      "loss": 3.1272,
+      "step": 23341
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5480991005897522,
+      "learning_rate": 0.00040928693550979753,
+      "loss": 3.2418,
+      "step": 23342
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.560892641544342,
+      "learning_rate": 0.0004092725971147655,
+      "loss": 2.9972,
+      "step": 23343
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5563073754310608,
+      "learning_rate": 0.00040925825843193044,
+      "loss": 3.082,
+      "step": 23344
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5521158576011658,
+      "learning_rate": 0.0004092439194613302,
+      "loss": 3.0532,
+      "step": 23345
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5784248113632202,
+      "learning_rate": 0.0004092295802030025,
+      "loss": 2.9597,
+      "step": 23346
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.551121175289154,
+      "learning_rate": 0.00040921524065698505,
+      "loss": 3.1059,
+      "step": 23347
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5605080723762512,
+      "learning_rate": 0.00040920090082331565,
+      "loss": 3.1398,
+      "step": 23348
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5794740319252014,
+      "learning_rate": 0.00040918656070203224,
+      "loss": 2.8233,
+      "step": 23349
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5346746444702148,
+      "learning_rate": 0.00040917222029317234,
+      "loss": 3.1589,
+      "step": 23350
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.540032684803009,
+      "learning_rate": 0.0004091578795967739,
+      "loss": 3.0918,
+      "step": 23351
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5193694233894348,
+      "learning_rate": 0.00040914353861287446,
+      "loss": 3.3181,
+      "step": 23352
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5316674113273621,
+      "learning_rate": 0.00040912919734151205,
+      "loss": 3.1497,
+      "step": 23353
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5241429805755615,
+      "learning_rate": 0.00040911485578272433,
+      "loss": 3.0704,
+      "step": 23354
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5313575863838196,
+      "learning_rate": 0.00040910051393654905,
+      "loss": 3.2197,
+      "step": 23355
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5644849538803101,
+      "learning_rate": 0.00040908617180302403,
+      "loss": 3.0536,
+      "step": 23356
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5238354802131653,
+      "learning_rate": 0.000409071829382187,
+      "loss": 3.1429,
+      "step": 23357
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5016031265258789,
+      "learning_rate": 0.00040905748667407576,
+      "loss": 3.2018,
+      "step": 23358
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5598127245903015,
+      "learning_rate": 0.00040904314367872814,
+      "loss": 3.0267,
+      "step": 23359
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.557971715927124,
+      "learning_rate": 0.0004090288003961819,
+      "loss": 3.3368,
+      "step": 23360
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5256807208061218,
+      "learning_rate": 0.00040901445682647473,
+      "loss": 3.044,
+      "step": 23361
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5249261260032654,
+      "learning_rate": 0.0004090001129696444,
+      "loss": 3.1656,
+      "step": 23362
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5444071888923645,
+      "learning_rate": 0.0004089857688257289,
+      "loss": 3.0243,
+      "step": 23363
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5587050318717957,
+      "learning_rate": 0.00040897142439476575,
+      "loss": 3.3505,
+      "step": 23364
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5856272578239441,
+      "learning_rate": 0.00040895707967679283,
+      "loss": 3.0021,
+      "step": 23365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5790854096412659,
+      "learning_rate": 0.000408942734671848,
+      "loss": 3.1403,
+      "step": 23366
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.53217613697052,
+      "learning_rate": 0.00040892838937996894,
+      "loss": 3.2079,
+      "step": 23367
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5617000460624695,
+      "learning_rate": 0.0004089140438011934,
+      "loss": 2.9651,
+      "step": 23368
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5272381901741028,
+      "learning_rate": 0.0004088996979355593,
+      "loss": 3.0486,
+      "step": 23369
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.543768048286438,
+      "learning_rate": 0.0004088853517831044,
+      "loss": 3.0786,
+      "step": 23370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5788832902908325,
+      "learning_rate": 0.0004088710053438664,
+      "loss": 3.0349,
+      "step": 23371
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5837016105651855,
+      "learning_rate": 0.000408856658617883,
+      "loss": 3.0429,
+      "step": 23372
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5341154336929321,
+      "learning_rate": 0.00040884231160519225,
+      "loss": 2.9318,
+      "step": 23373
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5163646936416626,
+      "learning_rate": 0.0004088279643058318,
+      "loss": 3.2756,
+      "step": 23374
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5284332036972046,
+      "learning_rate": 0.00040881361671983937,
+      "loss": 3.1314,
+      "step": 23375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5456615686416626,
+      "learning_rate": 0.0004087992688472529,
+      "loss": 2.9555,
+      "step": 23376
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5189909338951111,
+      "learning_rate": 0.00040878492068811004,
+      "loss": 2.9664,
+      "step": 23377
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5077756643295288,
+      "learning_rate": 0.0004087705722424486,
+      "loss": 3.0454,
+      "step": 23378
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5315677523612976,
+      "learning_rate": 0.0004087562235103065,
+      "loss": 3.1714,
+      "step": 23379
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5543997883796692,
+      "learning_rate": 0.00040874187449172134,
+      "loss": 3.0133,
+      "step": 23380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.560238242149353,
+      "learning_rate": 0.0004087275251867311,
+      "loss": 3.0748,
+      "step": 23381
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5396565198898315,
+      "learning_rate": 0.0004087131755953734,
+      "loss": 3.253,
+      "step": 23382
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5551555156707764,
+      "learning_rate": 0.0004086988257176861,
+      "loss": 2.9754,
+      "step": 23383
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5715532302856445,
+      "learning_rate": 0.00040868447555370707,
+      "loss": 3.0113,
+      "step": 23384
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5877143144607544,
+      "learning_rate": 0.0004086701251034741,
+      "loss": 3.0628,
+      "step": 23385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.547389566898346,
+      "learning_rate": 0.00040865577436702483,
+      "loss": 3.0458,
+      "step": 23386
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5390537977218628,
+      "learning_rate": 0.00040864142334439725,
+      "loss": 3.17,
+      "step": 23387
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5335311889648438,
+      "learning_rate": 0.000408627072035629,
+      "loss": 2.9859,
+      "step": 23388
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5414642691612244,
+      "learning_rate": 0.00040861272044075803,
+      "loss": 3.0252,
+      "step": 23389
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5587677359580994,
+      "learning_rate": 0.00040859836855982196,
+      "loss": 3.0681,
+      "step": 23390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6065161824226379,
+      "learning_rate": 0.0004085840163928588,
+      "loss": 2.8998,
+      "step": 23391
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5559492111206055,
+      "learning_rate": 0.0004085696639399061,
+      "loss": 3.1246,
+      "step": 23392
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5463036894798279,
+      "learning_rate": 0.0004085553112010019,
+      "loss": 3.0655,
+      "step": 23393
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5573052167892456,
+      "learning_rate": 0.00040854095817618384,
+      "loss": 2.826,
+      "step": 23394
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5249746441841125,
+      "learning_rate": 0.0004085266048654899,
+      "loss": 3.0882,
+      "step": 23395
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5521349906921387,
+      "learning_rate": 0.00040851225126895766,
+      "loss": 2.9564,
+      "step": 23396
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.550536572933197,
+      "learning_rate": 0.00040849789738662504,
+      "loss": 2.9643,
+      "step": 23397
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5547955632209778,
+      "learning_rate": 0.00040848354321852985,
+      "loss": 3.2019,
+      "step": 23398
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5499030351638794,
+      "learning_rate": 0.00040846918876470987,
+      "loss": 3.1381,
+      "step": 23399
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.548309326171875,
+      "learning_rate": 0.000408454834025203,
+      "loss": 3.0595,
+      "step": 23400
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5371249914169312,
+      "learning_rate": 0.0004084404790000469,
+      "loss": 2.9364,
+      "step": 23401
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5594949722290039,
+      "learning_rate": 0.00040842612368927945,
+      "loss": 3.1443,
+      "step": 23402
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5433090329170227,
+      "learning_rate": 0.0004084117680929385,
+      "loss": 2.9709,
+      "step": 23403
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5581390261650085,
+      "learning_rate": 0.0004083974122110618,
+      "loss": 2.9891,
+      "step": 23404
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.540836751461029,
+      "learning_rate": 0.0004083830560436871,
+      "loss": 3.2713,
+      "step": 23405
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5747362971305847,
+      "learning_rate": 0.00040836869959085236,
+      "loss": 3.0827,
+      "step": 23406
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5611171126365662,
+      "learning_rate": 0.0004083543428525954,
+      "loss": 3.1278,
+      "step": 23407
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5559061765670776,
+      "learning_rate": 0.0004083399858289538,
+      "loss": 3.0086,
+      "step": 23408
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5641838908195496,
+      "learning_rate": 0.0004083256285199656,
+      "loss": 3.1977,
+      "step": 23409
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5294587016105652,
+      "learning_rate": 0.0004083112709256685,
+      "loss": 3.0881,
+      "step": 23410
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5804036259651184,
+      "learning_rate": 0.0004082969130461005,
+      "loss": 2.8935,
+      "step": 23411
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5449517369270325,
+      "learning_rate": 0.00040828255488129916,
+      "loss": 3.2973,
+      "step": 23412
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5502049326896667,
+      "learning_rate": 0.00040826819643130236,
+      "loss": 3.2441,
+      "step": 23413
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5578358769416809,
+      "learning_rate": 0.0004082538376961481,
+      "loss": 3.2078,
+      "step": 23414
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5473148822784424,
+      "learning_rate": 0.00040823947867587397,
+      "loss": 3.1273,
+      "step": 23415
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6118950843811035,
+      "learning_rate": 0.00040822511937051793,
+      "loss": 3.0009,
+      "step": 23416
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5265421867370605,
+      "learning_rate": 0.0004082107597801177,
+      "loss": 3.1703,
+      "step": 23417
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5736425518989563,
+      "learning_rate": 0.0004081963999047112,
+      "loss": 3.0401,
+      "step": 23418
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.572830080986023,
+      "learning_rate": 0.00040818203974433623,
+      "loss": 3.2007,
+      "step": 23419
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5175371170043945,
+      "learning_rate": 0.0004081676792990305,
+      "loss": 3.3056,
+      "step": 23420
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5525279641151428,
+      "learning_rate": 0.000408153318568832,
+      "loss": 2.9136,
+      "step": 23421
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5556487441062927,
+      "learning_rate": 0.00040813895755377846,
+      "loss": 3.2804,
+      "step": 23422
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5376157164573669,
+      "learning_rate": 0.00040812459625390774,
+      "loss": 3.0081,
+      "step": 23423
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5336998701095581,
+      "learning_rate": 0.0004081102346692575,
+      "loss": 3.0859,
+      "step": 23424
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5490345358848572,
+      "learning_rate": 0.0004080958727998659,
+      "loss": 3.1922,
+      "step": 23425
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5638200044631958,
+      "learning_rate": 0.0004080815106457705,
+      "loss": 3.2863,
+      "step": 23426
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5599801540374756,
+      "learning_rate": 0.0004080671482070092,
+      "loss": 3.0831,
+      "step": 23427
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5768230557441711,
+      "learning_rate": 0.0004080527854836198,
+      "loss": 3.0359,
+      "step": 23428
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.591766357421875,
+      "learning_rate": 0.00040803842247564014,
+      "loss": 2.9248,
+      "step": 23429
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5740898847579956,
+      "learning_rate": 0.0004080240591831081,
+      "loss": 3.3461,
+      "step": 23430
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6085920929908752,
+      "learning_rate": 0.0004080096956060615,
+      "loss": 2.907,
+      "step": 23431
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5598239898681641,
+      "learning_rate": 0.00040799533174453806,
+      "loss": 3.2174,
+      "step": 23432
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5733321905136108,
+      "learning_rate": 0.00040798096759857587,
+      "loss": 3.2491,
+      "step": 23433
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5727006793022156,
+      "learning_rate": 0.00040796660316821243,
+      "loss": 3.1616,
+      "step": 23434
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5333359241485596,
+      "learning_rate": 0.00040795223845348574,
+      "loss": 3.3976,
+      "step": 23435
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.57224440574646,
+      "learning_rate": 0.00040793787345443376,
+      "loss": 3.1264,
+      "step": 23436
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5239753723144531,
+      "learning_rate": 0.0004079235081710941,
+      "loss": 2.9847,
+      "step": 23437
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5496913194656372,
+      "learning_rate": 0.0004079091426035047,
+      "loss": 2.9158,
+      "step": 23438
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5474773049354553,
+      "learning_rate": 0.0004078947767517034,
+      "loss": 2.9988,
+      "step": 23439
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5493378639221191,
+      "learning_rate": 0.000407880410615728,
+      "loss": 3.0624,
+      "step": 23440
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5190552473068237,
+      "learning_rate": 0.0004078660441956164,
+      "loss": 3.2048,
+      "step": 23441
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5590404272079468,
+      "learning_rate": 0.0004078516774914064,
+      "loss": 3.277,
+      "step": 23442
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5510579347610474,
+      "learning_rate": 0.00040783731050313577,
+      "loss": 3.0241,
+      "step": 23443
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.534829318523407,
+      "learning_rate": 0.0004078229432308425,
+      "loss": 3.1232,
+      "step": 23444
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5541685223579407,
+      "learning_rate": 0.0004078085756745643,
+      "loss": 3.0511,
+      "step": 23445
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5376542806625366,
+      "learning_rate": 0.0004077942078343391,
+      "loss": 3.2308,
+      "step": 23446
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5467217564582825,
+      "learning_rate": 0.00040777983971020473,
+      "loss": 3.0171,
+      "step": 23447
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5329676866531372,
+      "learning_rate": 0.0004077654713021989,
+      "loss": 3.0905,
+      "step": 23448
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5383985638618469,
+      "learning_rate": 0.0004077511026103596,
+      "loss": 2.947,
+      "step": 23449
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5557945966720581,
+      "learning_rate": 0.00040773673363472465,
+      "loss": 3.0836,
+      "step": 23450
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5461722612380981,
+      "learning_rate": 0.0004077223643753319,
+      "loss": 3.061,
+      "step": 23451
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5956500172615051,
+      "learning_rate": 0.00040770799483221914,
+      "loss": 3.1172,
+      "step": 23452
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5400736331939697,
+      "learning_rate": 0.00040769362500542425,
+      "loss": 3.1167,
+      "step": 23453
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5344555377960205,
+      "learning_rate": 0.0004076792548949851,
+      "loss": 3.1462,
+      "step": 23454
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5609763860702515,
+      "learning_rate": 0.00040766488450093947,
+      "loss": 3.0945,
+      "step": 23455
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5006883144378662,
+      "learning_rate": 0.00040765051382332527,
+      "loss": 3.1908,
+      "step": 23456
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.566444456577301,
+      "learning_rate": 0.0004076361428621804,
+      "loss": 3.2382,
+      "step": 23457
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5385657548904419,
+      "learning_rate": 0.00040762177161754264,
+      "loss": 2.8872,
+      "step": 23458
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5454058647155762,
+      "learning_rate": 0.00040760740008944973,
+      "loss": 2.94,
+      "step": 23459
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5491828322410583,
+      "learning_rate": 0.00040759302827793965,
+      "loss": 3.0851,
+      "step": 23460
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5193360447883606,
+      "learning_rate": 0.0004075786561830503,
+      "loss": 2.9652,
+      "step": 23461
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.534925103187561,
+      "learning_rate": 0.0004075642838048195,
+      "loss": 3.0946,
+      "step": 23462
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5688135027885437,
+      "learning_rate": 0.00040754991114328506,
+      "loss": 2.9249,
+      "step": 23463
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5361714959144592,
+      "learning_rate": 0.00040753553819848485,
+      "loss": 3.122,
+      "step": 23464
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5302151441574097,
+      "learning_rate": 0.0004075211649704568,
+      "loss": 2.9271,
+      "step": 23465
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5387076735496521,
+      "learning_rate": 0.0004075067914592385,
+      "loss": 3.0661,
+      "step": 23466
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5546697974205017,
+      "learning_rate": 0.0004074924176648681,
+      "loss": 3.1147,
+      "step": 23467
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5648727416992188,
+      "learning_rate": 0.0004074780435873834,
+      "loss": 2.842,
+      "step": 23468
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5225715637207031,
+      "learning_rate": 0.00040746366922682223,
+      "loss": 3.1203,
+      "step": 23469
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5498008728027344,
+      "learning_rate": 0.00040744929458322234,
+      "loss": 3.1043,
+      "step": 23470
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5694062113761902,
+      "learning_rate": 0.0004074349196566217,
+      "loss": 3.1744,
+      "step": 23471
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5407145023345947,
+      "learning_rate": 0.0004074205444470582,
+      "loss": 3.1199,
+      "step": 23472
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5782432556152344,
+      "learning_rate": 0.00040740616895456976,
+      "loss": 3.1877,
+      "step": 23473
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5868061184883118,
+      "learning_rate": 0.00040739179317919404,
+      "loss": 3.298,
+      "step": 23474
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5635443329811096,
+      "learning_rate": 0.00040737741712096895,
+      "loss": 3.4142,
+      "step": 23475
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5290628671646118,
+      "learning_rate": 0.00040736304077993254,
+      "loss": 3.034,
+      "step": 23476
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5918185710906982,
+      "learning_rate": 0.00040734866415612244,
+      "loss": 3.0493,
+      "step": 23477
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5599088668823242,
+      "learning_rate": 0.00040733428724957664,
+      "loss": 3.263,
+      "step": 23478
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5094558596611023,
+      "learning_rate": 0.000407319910060333,
+      "loss": 3.1001,
+      "step": 23479
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6165416836738586,
+      "learning_rate": 0.0004073055325884293,
+      "loss": 3.0571,
+      "step": 23480
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5464115142822266,
+      "learning_rate": 0.00040729115483390357,
+      "loss": 3.1606,
+      "step": 23481
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5529763698577881,
+      "learning_rate": 0.00040727677679679354,
+      "loss": 3.0281,
+      "step": 23482
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5393357276916504,
+      "learning_rate": 0.00040726239847713717,
+      "loss": 3.0014,
+      "step": 23483
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5649782419204712,
+      "learning_rate": 0.00040724801987497236,
+      "loss": 2.9159,
+      "step": 23484
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5705831050872803,
+      "learning_rate": 0.0004072336409903367,
+      "loss": 3.0978,
+      "step": 23485
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5222892165184021,
+      "learning_rate": 0.0004072192618232683,
+      "loss": 3.0826,
+      "step": 23486
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5631487965583801,
+      "learning_rate": 0.0004072048823738052,
+      "loss": 3.0558,
+      "step": 23487
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5665938258171082,
+      "learning_rate": 0.00040719050264198493,
+      "loss": 3.2,
+      "step": 23488
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5198308825492859,
+      "learning_rate": 0.0004071761226278455,
+      "loss": 2.922,
+      "step": 23489
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5553904175758362,
+      "learning_rate": 0.00040716174233142475,
+      "loss": 3.1187,
+      "step": 23490
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5327962636947632,
+      "learning_rate": 0.00040714736175276073,
+      "loss": 3.0355,
+      "step": 23491
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5585110187530518,
+      "learning_rate": 0.0004071329808918911,
+      "loss": 3.0713,
+      "step": 23492
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5731109976768494,
+      "learning_rate": 0.0004071185997488538,
+      "loss": 3.1877,
+      "step": 23493
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5556461215019226,
+      "learning_rate": 0.0004071042183236867,
+      "loss": 2.8029,
+      "step": 23494
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5197128653526306,
+      "learning_rate": 0.00040708983661642785,
+      "loss": 3.1096,
+      "step": 23495
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5375272035598755,
+      "learning_rate": 0.00040707545462711483,
+      "loss": 3.1094,
+      "step": 23496
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5286567807197571,
+      "learning_rate": 0.00040706107235578565,
+      "loss": 2.863,
+      "step": 23497
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5974603891372681,
+      "learning_rate": 0.00040704668980247837,
+      "loss": 2.7441,
+      "step": 23498
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5531563758850098,
+      "learning_rate": 0.0004070323069672306,
+      "loss": 2.9755,
+      "step": 23499
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.518351137638092,
+      "learning_rate": 0.00040701792385008034,
+      "loss": 3.214,
+      "step": 23500
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5526236891746521,
+      "learning_rate": 0.00040700354045106543,
+      "loss": 2.9767,
+      "step": 23501
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5635164976119995,
+      "learning_rate": 0.0004069891567702238,
+      "loss": 3.2102,
+      "step": 23502
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5762778520584106,
+      "learning_rate": 0.00040697477280759336,
+      "loss": 3.1312,
+      "step": 23503
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5641893148422241,
+      "learning_rate": 0.000406960388563212,
+      "loss": 3.2119,
+      "step": 23504
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5612379908561707,
+      "learning_rate": 0.0004069460040371175,
+      "loss": 3.1178,
+      "step": 23505
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5747913718223572,
+      "learning_rate": 0.0004069316192293478,
+      "loss": 3.3066,
+      "step": 23506
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5277937650680542,
+      "learning_rate": 0.0004069172341399408,
+      "loss": 3.2138,
+      "step": 23507
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5456346869468689,
+      "learning_rate": 0.0004069028487689344,
+      "loss": 3.0695,
+      "step": 23508
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5213927626609802,
+      "learning_rate": 0.00040688846311636654,
+      "loss": 3.2051,
+      "step": 23509
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5669615268707275,
+      "learning_rate": 0.00040687407718227494,
+      "loss": 2.951,
+      "step": 23510
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5207639336585999,
+      "learning_rate": 0.0004068596909666975,
+      "loss": 3.2213,
+      "step": 23511
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5501463413238525,
+      "learning_rate": 0.0004068453044696724,
+      "loss": 3.1305,
+      "step": 23512
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5365391969680786,
+      "learning_rate": 0.00040683091769123724,
+      "loss": 2.9962,
+      "step": 23513
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.514644980430603,
+      "learning_rate": 0.00040681653063143,
+      "loss": 3.0177,
+      "step": 23514
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5442690849304199,
+      "learning_rate": 0.0004068021432902886,
+      "loss": 3.1254,
+      "step": 23515
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.508587121963501,
+      "learning_rate": 0.00040678775566785086,
+      "loss": 3.0467,
+      "step": 23516
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5336624383926392,
+      "learning_rate": 0.00040677336776415477,
+      "loss": 3.0867,
+      "step": 23517
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5563424229621887,
+      "learning_rate": 0.00040675897957923816,
+      "loss": 3.1045,
+      "step": 23518
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5636367797851562,
+      "learning_rate": 0.0004067445911131389,
+      "loss": 3.4081,
+      "step": 23519
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5157569050788879,
+      "learning_rate": 0.00040673020236589504,
+      "loss": 3.1305,
+      "step": 23520
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5343283414840698,
+      "learning_rate": 0.00040671581333754427,
+      "loss": 3.1285,
+      "step": 23521
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5524337291717529,
+      "learning_rate": 0.00040670142402812455,
+      "loss": 3.1464,
+      "step": 23522
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.638904333114624,
+      "learning_rate": 0.0004066870344376739,
+      "loss": 2.9966,
+      "step": 23523
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5390263199806213,
+      "learning_rate": 0.00040667264456623016,
+      "loss": 3.2796,
+      "step": 23524
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.588818371295929,
+      "learning_rate": 0.0004066582544138312,
+      "loss": 3.1716,
+      "step": 23525
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6020216345787048,
+      "learning_rate": 0.00040664386398051483,
+      "loss": 3.2594,
+      "step": 23526
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5486242771148682,
+      "learning_rate": 0.00040662947326631913,
+      "loss": 3.0998,
+      "step": 23527
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5452220439910889,
+      "learning_rate": 0.0004066150822712819,
+      "loss": 2.853,
+      "step": 23528
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5223864912986755,
+      "learning_rate": 0.000406600690995441,
+      "loss": 3.1148,
+      "step": 23529
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5381213426589966,
+      "learning_rate": 0.00040658629943883447,
+      "loss": 3.2326,
+      "step": 23530
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5543666481971741,
+      "learning_rate": 0.00040657190760150013,
+      "loss": 3.2087,
+      "step": 23531
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5769178867340088,
+      "learning_rate": 0.00040655751548347583,
+      "loss": 3.337,
+      "step": 23532
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5365214347839355,
+      "learning_rate": 0.0004065431230847996,
+      "loss": 3.0142,
+      "step": 23533
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5313071012496948,
+      "learning_rate": 0.00040652873040550926,
+      "loss": 3.0712,
+      "step": 23534
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6116557121276855,
+      "learning_rate": 0.0004065143374456429,
+      "loss": 3.0255,
+      "step": 23535
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5586936473846436,
+      "learning_rate": 0.0004064999442052381,
+      "loss": 2.9973,
+      "step": 23536
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5303226709365845,
+      "learning_rate": 0.00040648555068433293,
+      "loss": 3.2365,
+      "step": 23537
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5624071359634399,
+      "learning_rate": 0.0004064711568829654,
+      "loss": 3.0904,
+      "step": 23538
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5486844778060913,
+      "learning_rate": 0.00040645676280117327,
+      "loss": 3.0888,
+      "step": 23539
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5096597671508789,
+      "learning_rate": 0.00040644236843899456,
+      "loss": 3.0044,
+      "step": 23540
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5384427905082703,
+      "learning_rate": 0.00040642797379646713,
+      "loss": 3.0459,
+      "step": 23541
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.551108181476593,
+      "learning_rate": 0.0004064135788736289,
+      "loss": 3.2747,
+      "step": 23542
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5435267090797424,
+      "learning_rate": 0.00040639918367051776,
+      "loss": 3.2177,
+      "step": 23543
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5716750621795654,
+      "learning_rate": 0.00040638478818717165,
+      "loss": 3.2905,
+      "step": 23544
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5734542608261108,
+      "learning_rate": 0.0004063703924236285,
+      "loss": 3.4277,
+      "step": 23545
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5500821471214294,
+      "learning_rate": 0.0004063559963799262,
+      "loss": 3.1241,
+      "step": 23546
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6079295873641968,
+      "learning_rate": 0.00040634160005610263,
+      "loss": 2.9869,
+      "step": 23547
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5442874431610107,
+      "learning_rate": 0.00040632720345219585,
+      "loss": 3.1278,
+      "step": 23548
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5300388336181641,
+      "learning_rate": 0.00040631280656824365,
+      "loss": 3.0279,
+      "step": 23549
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5295524597167969,
+      "learning_rate": 0.000406298409404284,
+      "loss": 3.2271,
+      "step": 23550
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5851762890815735,
+      "learning_rate": 0.0004062840119603547,
+      "loss": 3.0275,
+      "step": 23551
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5763968825340271,
+      "learning_rate": 0.00040626961423649376,
+      "loss": 3.0724,
+      "step": 23552
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.534521758556366,
+      "learning_rate": 0.0004062552162327393,
+      "loss": 3.0856,
+      "step": 23553
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5817490220069885,
+      "learning_rate": 0.0004062408179491288,
+      "loss": 3.1984,
+      "step": 23554
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.548794686794281,
+      "learning_rate": 0.0004062264193857006,
+      "loss": 3.0361,
+      "step": 23555
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5278934240341187,
+      "learning_rate": 0.0004062120205424924,
+      "loss": 2.9397,
+      "step": 23556
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5557737350463867,
+      "learning_rate": 0.0004061976214195422,
+      "loss": 3.0761,
+      "step": 23557
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6352978944778442,
+      "learning_rate": 0.0004061832220168879,
+      "loss": 3.2706,
+      "step": 23558
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5525439381599426,
+      "learning_rate": 0.00040616882233456735,
+      "loss": 3.1981,
+      "step": 23559
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5778921246528625,
+      "learning_rate": 0.0004061544223726187,
+      "loss": 3.0622,
+      "step": 23560
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5746831297874451,
+      "learning_rate": 0.0004061400221310796,
+      "loss": 3.335,
+      "step": 23561
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5181114077568054,
+      "learning_rate": 0.00040612562160998817,
+      "loss": 3.0855,
+      "step": 23562
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5968653559684753,
+      "learning_rate": 0.00040611122080938224,
+      "loss": 3.1416,
+      "step": 23563
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5655339956283569,
+      "learning_rate": 0.0004060968197292999,
+      "loss": 3.0138,
+      "step": 23564
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5550764203071594,
+      "learning_rate": 0.00040608241836977886,
+      "loss": 3.0035,
+      "step": 23565
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5673668384552002,
+      "learning_rate": 0.0004060680167308571,
+      "loss": 3.0248,
+      "step": 23566
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5502573251724243,
+      "learning_rate": 0.0004060536148125726,
+      "loss": 3.3946,
+      "step": 23567
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5763288140296936,
+      "learning_rate": 0.00040603921261496336,
+      "loss": 3.1626,
+      "step": 23568
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5315199494361877,
+      "learning_rate": 0.0004060248101380672,
+      "loss": 2.9473,
+      "step": 23569
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5365606546401978,
+      "learning_rate": 0.00040601040738192214,
+      "loss": 3.0127,
+      "step": 23570
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5493980050086975,
+      "learning_rate": 0.000405996004346566,
+      "loss": 2.8699,
+      "step": 23571
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5779330730438232,
+      "learning_rate": 0.0004059816010320369,
+      "loss": 3.0912,
+      "step": 23572
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5333728790283203,
+      "learning_rate": 0.00040596719743837253,
+      "loss": 3.2728,
+      "step": 23573
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5707440376281738,
+      "learning_rate": 0.0004059527935656109,
+      "loss": 3.0194,
+      "step": 23574
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5592421293258667,
+      "learning_rate": 0.0004059383894137902,
+      "loss": 2.8304,
+      "step": 23575
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5238040089607239,
+      "learning_rate": 0.0004059239849829481,
+      "loss": 2.9382,
+      "step": 23576
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5463867783546448,
+      "learning_rate": 0.00040590958027312255,
+      "loss": 3.3477,
+      "step": 23577
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5499131083488464,
+      "learning_rate": 0.0004058951752843516,
+      "loss": 3.274,
+      "step": 23578
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5574745535850525,
+      "learning_rate": 0.0004058807700166731,
+      "loss": 2.9782,
+      "step": 23579
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5117610692977905,
+      "learning_rate": 0.0004058663644701251,
+      "loss": 3.0166,
+      "step": 23580
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5790355205535889,
+      "learning_rate": 0.00040585195864474544,
+      "loss": 3.2326,
+      "step": 23581
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5722032785415649,
+      "learning_rate": 0.00040583755254057204,
+      "loss": 3.116,
+      "step": 23582
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5745276212692261,
+      "learning_rate": 0.00040582314615764293,
+      "loss": 3.0388,
+      "step": 23583
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5759443640708923,
+      "learning_rate": 0.000405808739495996,
+      "loss": 2.9188,
+      "step": 23584
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5403763651847839,
+      "learning_rate": 0.00040579433255566927,
+      "loss": 3.0228,
+      "step": 23585
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5597376227378845,
+      "learning_rate": 0.00040577992533670065,
+      "loss": 3.3179,
+      "step": 23586
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.565265417098999,
+      "learning_rate": 0.000405765517839128,
+      "loss": 3.2375,
+      "step": 23587
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.584064781665802,
+      "learning_rate": 0.00040575111006298925,
+      "loss": 2.8922,
+      "step": 23588
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5477665066719055,
+      "learning_rate": 0.00040573670200832253,
+      "loss": 3.0098,
+      "step": 23589
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5629181265830994,
+      "learning_rate": 0.0004057222936751657,
+      "loss": 3.0527,
+      "step": 23590
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5321118235588074,
+      "learning_rate": 0.0004057078850635567,
+      "loss": 3.1828,
+      "step": 23591
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5479146242141724,
+      "learning_rate": 0.00040569347617353343,
+      "loss": 3.1211,
+      "step": 23592
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5773611068725586,
+      "learning_rate": 0.0004056790670051339,
+      "loss": 2.9645,
+      "step": 23593
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5333780646324158,
+      "learning_rate": 0.000405664657558396,
+      "loss": 3.1218,
+      "step": 23594
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5267741680145264,
+      "learning_rate": 0.00040565024783335775,
+      "loss": 3.0897,
+      "step": 23595
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5395077466964722,
+      "learning_rate": 0.00040563583783005707,
+      "loss": 3.0424,
+      "step": 23596
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.562576949596405,
+      "learning_rate": 0.00040562142754853206,
+      "loss": 3.3082,
+      "step": 23597
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5555576682090759,
+      "learning_rate": 0.00040560701698882034,
+      "loss": 3.0102,
+      "step": 23598
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5478713512420654,
+      "learning_rate": 0.0004055926061509602,
+      "loss": 2.9769,
+      "step": 23599
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5155161619186401,
+      "learning_rate": 0.00040557819503498943,
+      "loss": 3.0548,
+      "step": 23600
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.563854455947876,
+      "learning_rate": 0.0004055637836409459,
+      "loss": 2.9703,
+      "step": 23601
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5164005756378174,
+      "learning_rate": 0.0004055493719688678,
+      "loss": 3.032,
+      "step": 23602
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6590232849121094,
+      "learning_rate": 0.00040553496001879296,
+      "loss": 3.1402,
+      "step": 23603
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5741829872131348,
+      "learning_rate": 0.00040552054779075935,
+      "loss": 2.7843,
+      "step": 23604
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5617377758026123,
+      "learning_rate": 0.00040550613528480493,
+      "loss": 3.1812,
+      "step": 23605
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.52630615234375,
+      "learning_rate": 0.0004054917225009676,
+      "loss": 3.1588,
+      "step": 23606
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5488269329071045,
+      "learning_rate": 0.0004054773094392854,
+      "loss": 3.2574,
+      "step": 23607
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7178289294242859,
+      "learning_rate": 0.00040546289609979624,
+      "loss": 2.9536,
+      "step": 23608
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5843186974525452,
+      "learning_rate": 0.0004054484824825382,
+      "loss": 3.1758,
+      "step": 23609
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6291219592094421,
+      "learning_rate": 0.00040543406858754903,
+      "loss": 3.0335,
+      "step": 23610
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5768383741378784,
+      "learning_rate": 0.0004054196544148669,
+      "loss": 2.8571,
+      "step": 23611
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5430237054824829,
+      "learning_rate": 0.0004054052399645297,
+      "loss": 2.9117,
+      "step": 23612
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5219388008117676,
+      "learning_rate": 0.00040539082523657536,
+      "loss": 3.1468,
+      "step": 23613
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.547167181968689,
+      "learning_rate": 0.0004053764102310417,
+      "loss": 3.237,
+      "step": 23614
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5904198288917542,
+      "learning_rate": 0.0004053619949479671,
+      "loss": 3.1433,
+      "step": 23615
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5035914778709412,
+      "learning_rate": 0.0004053475793873892,
+      "loss": 2.966,
+      "step": 23616
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5650410652160645,
+      "learning_rate": 0.00040533316354934607,
+      "loss": 3.1302,
+      "step": 23617
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5530869960784912,
+      "learning_rate": 0.0004053187474338756,
+      "loss": 3.0593,
+      "step": 23618
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.560907781124115,
+      "learning_rate": 0.00040530433104101583,
+      "loss": 3.1948,
+      "step": 23619
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5566398501396179,
+      "learning_rate": 0.00040528991437080474,
+      "loss": 3.2553,
+      "step": 23620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5609525442123413,
+      "learning_rate": 0.00040527549742328027,
+      "loss": 2.9138,
+      "step": 23621
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5385319590568542,
+      "learning_rate": 0.0004052610801984805,
+      "loss": 3.1255,
+      "step": 23622
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5298342108726501,
+      "learning_rate": 0.0004052466626964432,
+      "loss": 3.3531,
+      "step": 23623
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5590463280677795,
+      "learning_rate": 0.00040523224491720636,
+      "loss": 3.0695,
+      "step": 23624
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5449240803718567,
+      "learning_rate": 0.00040521782686080816,
+      "loss": 3.0979,
+      "step": 23625
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5203918218612671,
+      "learning_rate": 0.00040520340852728647,
+      "loss": 3.023,
+      "step": 23626
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6454471945762634,
+      "learning_rate": 0.0004051889899166792,
+      "loss": 3.0371,
+      "step": 23627
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.546628475189209,
+      "learning_rate": 0.0004051745710290244,
+      "loss": 2.9801,
+      "step": 23628
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5471029877662659,
+      "learning_rate": 0.00040516015186436,
+      "loss": 3.0225,
+      "step": 23629
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.57635498046875,
+      "learning_rate": 0.00040514573242272396,
+      "loss": 3.1095,
+      "step": 23630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5368325710296631,
+      "learning_rate": 0.0004051313127041544,
+      "loss": 3.0904,
+      "step": 23631
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5363153219223022,
+      "learning_rate": 0.00040511689270868905,
+      "loss": 2.986,
+      "step": 23632
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5831332802772522,
+      "learning_rate": 0.00040510247243636614,
+      "loss": 3.0365,
+      "step": 23633
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5615931749343872,
+      "learning_rate": 0.0004050880518872235,
+      "loss": 3.142,
+      "step": 23634
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5161574482917786,
+      "learning_rate": 0.0004050736310612992,
+      "loss": 3.2197,
+      "step": 23635
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5491262078285217,
+      "learning_rate": 0.00040505920995863114,
+      "loss": 2.9822,
+      "step": 23636
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5797311067581177,
+      "learning_rate": 0.00040504478857925736,
+      "loss": 3.1336,
+      "step": 23637
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5718839168548584,
+      "learning_rate": 0.00040503036692321584,
+      "loss": 3.0789,
+      "step": 23638
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.529486358165741,
+      "learning_rate": 0.0004050159449905445,
+      "loss": 3.0697,
+      "step": 23639
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5350091457366943,
+      "learning_rate": 0.0004050015227812815,
+      "loss": 3.1539,
+      "step": 23640
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5986202359199524,
+      "learning_rate": 0.0004049871002954645,
+      "loss": 3.2035,
+      "step": 23641
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5608084797859192,
+      "learning_rate": 0.0004049726775331318,
+      "loss": 3.0752,
+      "step": 23642
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5415424108505249,
+      "learning_rate": 0.00040495825449432125,
+      "loss": 3.2252,
+      "step": 23643
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5681462287902832,
+      "learning_rate": 0.00040494383117907083,
+      "loss": 3.228,
+      "step": 23644
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5260352492332458,
+      "learning_rate": 0.0004049294075874186,
+      "loss": 3.1951,
+      "step": 23645
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5569955706596375,
+      "learning_rate": 0.0004049149837194024,
+      "loss": 3.1433,
+      "step": 23646
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5537160634994507,
+      "learning_rate": 0.00040490055957506047,
+      "loss": 2.7933,
+      "step": 23647
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5751855373382568,
+      "learning_rate": 0.00040488613515443064,
+      "loss": 2.9887,
+      "step": 23648
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6018997430801392,
+      "learning_rate": 0.00040487171045755087,
+      "loss": 2.9387,
+      "step": 23649
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5773023366928101,
+      "learning_rate": 0.0004048572854844591,
+      "loss": 3.0559,
+      "step": 23650
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5616927146911621,
+      "learning_rate": 0.0004048428602351936,
+      "loss": 3.2197,
+      "step": 23651
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5932993292808533,
+      "learning_rate": 0.00040482843470979207,
+      "loss": 2.9324,
+      "step": 23652
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5337039232254028,
+      "learning_rate": 0.0004048140089082927,
+      "loss": 3.0964,
+      "step": 23653
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5320896506309509,
+      "learning_rate": 0.00040479958283073334,
+      "loss": 3.0553,
+      "step": 23654
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5578566193580627,
+      "learning_rate": 0.0004047851564771521,
+      "loss": 2.9863,
+      "step": 23655
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5269665718078613,
+      "learning_rate": 0.00040477072984758687,
+      "loss": 3.0494,
+      "step": 23656
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6187220811843872,
+      "learning_rate": 0.0004047563029420757,
+      "loss": 3.0293,
+      "step": 23657
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5536328554153442,
+      "learning_rate": 0.00040474187576065656,
+      "loss": 2.991,
+      "step": 23658
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5319032669067383,
+      "learning_rate": 0.0004047274483033676,
+      "loss": 3.0669,
+      "step": 23659
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5195010304450989,
+      "learning_rate": 0.00040471302057024653,
+      "loss": 2.8542,
+      "step": 23660
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5610055327415466,
+      "learning_rate": 0.0004046985925613316,
+      "loss": 2.9366,
+      "step": 23661
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5877775549888611,
+      "learning_rate": 0.0004046841642766608,
+      "loss": 3.2743,
+      "step": 23662
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5556247234344482,
+      "learning_rate": 0.000404669735716272,
+      "loss": 3.2213,
+      "step": 23663
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5412140488624573,
+      "learning_rate": 0.00040465530688020324,
+      "loss": 3.2489,
+      "step": 23664
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5714347958564758,
+      "learning_rate": 0.00040464087776849244,
+      "loss": 3.0755,
+      "step": 23665
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5732872486114502,
+      "learning_rate": 0.00040462644838117783,
+      "loss": 3.0111,
+      "step": 23666
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5459180474281311,
+      "learning_rate": 0.00040461201871829736,
+      "loss": 2.8323,
+      "step": 23667
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5625903606414795,
+      "learning_rate": 0.00040459758877988886,
+      "loss": 3.0188,
+      "step": 23668
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5748194456100464,
+      "learning_rate": 0.00040458315856599044,
+      "loss": 2.6947,
+      "step": 23669
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5342017412185669,
+      "learning_rate": 0.00040456872807664016,
+      "loss": 3.1381,
+      "step": 23670
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5473812818527222,
+      "learning_rate": 0.0004045542973118759,
+      "loss": 3.1439,
+      "step": 23671
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.517485499382019,
+      "learning_rate": 0.00040453986627173586,
+      "loss": 3.0633,
+      "step": 23672
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5450983047485352,
+      "learning_rate": 0.0004045254349562579,
+      "loss": 3.056,
+      "step": 23673
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6088648438453674,
+      "learning_rate": 0.00040451100336548,
+      "loss": 3.1054,
+      "step": 23674
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5295372009277344,
+      "learning_rate": 0.0004044965714994402,
+      "loss": 3.1122,
+      "step": 23675
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5838629007339478,
+      "learning_rate": 0.00040448213935817654,
+      "loss": 3.1979,
+      "step": 23676
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5640658736228943,
+      "learning_rate": 0.0004044677069417272,
+      "loss": 3.163,
+      "step": 23677
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5375956296920776,
+      "learning_rate": 0.00040445327425012986,
+      "loss": 3.2325,
+      "step": 23678
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5397927761077881,
+      "learning_rate": 0.0004044388412834228,
+      "loss": 3.1302,
+      "step": 23679
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5797949433326721,
+      "learning_rate": 0.00040442440804164384,
+      "loss": 3.0378,
+      "step": 23680
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5403143763542175,
+      "learning_rate": 0.00040440997452483116,
+      "loss": 3.1009,
+      "step": 23681
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5413429141044617,
+      "learning_rate": 0.00040439554073302264,
+      "loss": 2.7718,
+      "step": 23682
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5402602553367615,
+      "learning_rate": 0.0004043811066662564,
+      "loss": 3.1233,
+      "step": 23683
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5142017006874084,
+      "learning_rate": 0.00040436667232457044,
+      "loss": 3.0242,
+      "step": 23684
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5384440422058105,
+      "learning_rate": 0.0004043522377080026,
+      "loss": 3.1141,
+      "step": 23685
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.539061963558197,
+      "learning_rate": 0.0004043378028165912,
+      "loss": 2.9895,
+      "step": 23686
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5121895670890808,
+      "learning_rate": 0.0004043233676503741,
+      "loss": 3.0622,
+      "step": 23687
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5728363990783691,
+      "learning_rate": 0.00040430893220938925,
+      "loss": 3.0116,
+      "step": 23688
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5647080540657043,
+      "learning_rate": 0.00040429449649367487,
+      "loss": 3.0383,
+      "step": 23689
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5658575296401978,
+      "learning_rate": 0.0004042800605032687,
+      "loss": 3.0062,
+      "step": 23690
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5728873014450073,
+      "learning_rate": 0.00040426562423820904,
+      "loss": 2.9859,
+      "step": 23691
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5607736110687256,
+      "learning_rate": 0.0004042511876985338,
+      "loss": 3.1462,
+      "step": 23692
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5391986966133118,
+      "learning_rate": 0.00040423675088428095,
+      "loss": 3.165,
+      "step": 23693
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5368701815605164,
+      "learning_rate": 0.0004042223137954885,
+      "loss": 3.1621,
+      "step": 23694
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.578046441078186,
+      "learning_rate": 0.0004042078764321945,
+      "loss": 3.2745,
+      "step": 23695
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5303065776824951,
+      "learning_rate": 0.0004041934387944372,
+      "loss": 3.2873,
+      "step": 23696
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5496455430984497,
+      "learning_rate": 0.00040417900088225435,
+      "loss": 2.9276,
+      "step": 23697
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6175884008407593,
+      "learning_rate": 0.00040416456269568404,
+      "loss": 2.8928,
+      "step": 23698
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5311169028282166,
+      "learning_rate": 0.0004041501242347644,
+      "loss": 3.1252,
+      "step": 23699
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5432295203208923,
+      "learning_rate": 0.0004041356854995332,
+      "loss": 3.2053,
+      "step": 23700
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5909796357154846,
+      "learning_rate": 0.00040412124649002876,
+      "loss": 3.0875,
+      "step": 23701
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5231812000274658,
+      "learning_rate": 0.000404106807206289,
+      "loss": 2.9765,
+      "step": 23702
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5160446763038635,
+      "learning_rate": 0.0004040923676483519,
+      "loss": 3.0671,
+      "step": 23703
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5727381706237793,
+      "learning_rate": 0.00040407792781625555,
+      "loss": 2.9777,
+      "step": 23704
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5669409036636353,
+      "learning_rate": 0.000404063487710038,
+      "loss": 3.185,
+      "step": 23705
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5885549783706665,
+      "learning_rate": 0.0004040490473297372,
+      "loss": 3.3162,
+      "step": 23706
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5521764159202576,
+      "learning_rate": 0.0004040346066753913,
+      "loss": 3.0025,
+      "step": 23707
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5233154892921448,
+      "learning_rate": 0.0004040201657470382,
+      "loss": 3.0222,
+      "step": 23708
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6025112867355347,
+      "learning_rate": 0.00040400572454471603,
+      "loss": 3.0527,
+      "step": 23709
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5110952258110046,
+      "learning_rate": 0.00040399128306846285,
+      "loss": 3.1567,
+      "step": 23710
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5763827562332153,
+      "learning_rate": 0.00040397684131831656,
+      "loss": 3.1964,
+      "step": 23711
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5659586191177368,
+      "learning_rate": 0.00040396239929431534,
+      "loss": 3.1401,
+      "step": 23712
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5608625411987305,
+      "learning_rate": 0.0004039479569964971,
+      "loss": 3.0601,
+      "step": 23713
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5368182063102722,
+      "learning_rate": 0.00040393351442490007,
+      "loss": 2.9722,
+      "step": 23714
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5225261449813843,
+      "learning_rate": 0.00040391907157956214,
+      "loss": 3.0737,
+      "step": 23715
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5581330060958862,
+      "learning_rate": 0.0004039046284605212,
+      "loss": 3.1152,
+      "step": 23716
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5440344214439392,
+      "learning_rate": 0.00040389018506781563,
+      "loss": 3.2068,
+      "step": 23717
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5621079802513123,
+      "learning_rate": 0.0004038757414014833,
+      "loss": 3.0555,
+      "step": 23718
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.566859781742096,
+      "learning_rate": 0.00040386129746156215,
+      "loss": 2.9883,
+      "step": 23719
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5255882143974304,
+      "learning_rate": 0.0004038468532480905,
+      "loss": 2.8779,
+      "step": 23720
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5612319111824036,
+      "learning_rate": 0.0004038324087611061,
+      "loss": 3.115,
+      "step": 23721
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5677599906921387,
+      "learning_rate": 0.00040381796400064716,
+      "loss": 3.0859,
+      "step": 23722
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5437223315238953,
+      "learning_rate": 0.00040380351896675157,
+      "loss": 3.0952,
+      "step": 23723
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5329674482345581,
+      "learning_rate": 0.0004037890736594577,
+      "loss": 3.0249,
+      "step": 23724
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5678834319114685,
+      "learning_rate": 0.00040377462807880324,
+      "loss": 3.0096,
+      "step": 23725
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5391994118690491,
+      "learning_rate": 0.0004037601822248264,
+      "loss": 3.2036,
+      "step": 23726
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5704066753387451,
+      "learning_rate": 0.0004037457360975652,
+      "loss": 3.2307,
+      "step": 23727
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.528113842010498,
+      "learning_rate": 0.00040373128969705774,
+      "loss": 3.3054,
+      "step": 23728
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5105913877487183,
+      "learning_rate": 0.00040371684302334203,
+      "loss": 3.173,
+      "step": 23729
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5552327036857605,
+      "learning_rate": 0.0004037023960764561,
+      "loss": 3.1766,
+      "step": 23730
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5556386113166809,
+      "learning_rate": 0.000403687948856438,
+      "loss": 2.9731,
+      "step": 23731
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5652289986610413,
+      "learning_rate": 0.0004036735013633258,
+      "loss": 3.2652,
+      "step": 23732
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4932308495044708,
+      "learning_rate": 0.0004036590535971576,
+      "loss": 3.1269,
+      "step": 23733
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5384199023246765,
+      "learning_rate": 0.0004036446055579713,
+      "loss": 3.1946,
+      "step": 23734
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5091381669044495,
+      "learning_rate": 0.00040363015724580517,
+      "loss": 2.9825,
+      "step": 23735
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5702914595603943,
+      "learning_rate": 0.0004036157086606971,
+      "loss": 3.1606,
+      "step": 23736
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5650493502616882,
+      "learning_rate": 0.00040360125980268513,
+      "loss": 3.0586,
+      "step": 23737
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5145954489707947,
+      "learning_rate": 0.00040358681067180746,
+      "loss": 3.2997,
+      "step": 23738
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5607908368110657,
+      "learning_rate": 0.00040357236126810207,
+      "loss": 3.0453,
+      "step": 23739
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5422621965408325,
+      "learning_rate": 0.000403557911591607,
+      "loss": 2.933,
+      "step": 23740
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5554590225219727,
+      "learning_rate": 0.0004035434616423603,
+      "loss": 2.9494,
+      "step": 23741
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6001891493797302,
+      "learning_rate": 0.0004035290114204,
+      "loss": 3.1366,
+      "step": 23742
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5242499113082886,
+      "learning_rate": 0.00040351456092576433,
+      "loss": 3.277,
+      "step": 23743
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5295645594596863,
+      "learning_rate": 0.00040350011015849116,
+      "loss": 3.2324,
+      "step": 23744
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5429471135139465,
+      "learning_rate": 0.00040348565911861866,
+      "loss": 3.0578,
+      "step": 23745
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5438092350959778,
+      "learning_rate": 0.0004034712078061849,
+      "loss": 3.0985,
+      "step": 23746
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5752182006835938,
+      "learning_rate": 0.0004034567562212277,
+      "loss": 3.2343,
+      "step": 23747
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5302133560180664,
+      "learning_rate": 0.0004034423043637854,
+      "loss": 3.0236,
+      "step": 23748
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5672379732131958,
+      "learning_rate": 0.000403427852233896,
+      "loss": 3.2175,
+      "step": 23749
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5564298033714294,
+      "learning_rate": 0.0004034133998315976,
+      "loss": 3.2707,
+      "step": 23750
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5528357028961182,
+      "learning_rate": 0.00040339894715692815,
+      "loss": 3.1907,
+      "step": 23751
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5306649208068848,
+      "learning_rate": 0.0004033844942099257,
+      "loss": 3.0427,
+      "step": 23752
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5104833841323853,
+      "learning_rate": 0.00040337004099062854,
+      "loss": 3.1894,
+      "step": 23753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5718398094177246,
+      "learning_rate": 0.00040335558749907455,
+      "loss": 3.0523,
+      "step": 23754
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.568389892578125,
+      "learning_rate": 0.00040334113373530174,
+      "loss": 3.0214,
+      "step": 23755
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.567775547504425,
+      "learning_rate": 0.00040332667969934837,
+      "loss": 3.217,
+      "step": 23756
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5069112777709961,
+      "learning_rate": 0.00040331222539125234,
+      "loss": 2.9524,
+      "step": 23757
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5586796402931213,
+      "learning_rate": 0.0004032977708110518,
+      "loss": 3.1551,
+      "step": 23758
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5643654465675354,
+      "learning_rate": 0.0004032833159587848,
+      "loss": 2.9142,
+      "step": 23759
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5675867199897766,
+      "learning_rate": 0.00040326886083448944,
+      "loss": 2.9242,
+      "step": 23760
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5585340261459351,
+      "learning_rate": 0.00040325440543820387,
+      "loss": 3.1171,
+      "step": 23761
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5579453706741333,
+      "learning_rate": 0.0004032399497699659,
+      "loss": 3.082,
+      "step": 23762
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5434767603874207,
+      "learning_rate": 0.00040322549382981385,
+      "loss": 2.9839,
+      "step": 23763
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5664116740226746,
+      "learning_rate": 0.00040321103761778584,
+      "loss": 2.8925,
+      "step": 23764
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5497961640357971,
+      "learning_rate": 0.00040319658113391967,
+      "loss": 3.0979,
+      "step": 23765
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5843026041984558,
+      "learning_rate": 0.00040318212437825355,
+      "loss": 3.1466,
+      "step": 23766
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5227053761482239,
+      "learning_rate": 0.00040316766735082565,
+      "loss": 2.9814,
+      "step": 23767
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5491824150085449,
+      "learning_rate": 0.0004031532100516739,
+      "loss": 3.0808,
+      "step": 23768
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5716967582702637,
+      "learning_rate": 0.00040313875248083647,
+      "loss": 3.2243,
+      "step": 23769
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5618823170661926,
+      "learning_rate": 0.00040312429463835146,
+      "loss": 3.162,
+      "step": 23770
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5274955630302429,
+      "learning_rate": 0.00040310983652425695,
+      "loss": 2.892,
+      "step": 23771
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6322970390319824,
+      "learning_rate": 0.0004030953781385909,
+      "loss": 3.0052,
+      "step": 23772
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5498408079147339,
+      "learning_rate": 0.00040308091948139136,
+      "loss": 2.9249,
+      "step": 23773
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5550096035003662,
+      "learning_rate": 0.0004030664605526966,
+      "loss": 3.0928,
+      "step": 23774
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5641299486160278,
+      "learning_rate": 0.00040305200135254475,
+      "loss": 3.1518,
+      "step": 23775
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5485759973526001,
+      "learning_rate": 0.0004030375418809736,
+      "loss": 2.9378,
+      "step": 23776
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5412816405296326,
+      "learning_rate": 0.0004030230821380215,
+      "loss": 2.9547,
+      "step": 23777
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5825008749961853,
+      "learning_rate": 0.0004030086221237263,
+      "loss": 3.1217,
+      "step": 23778
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.533901035785675,
+      "learning_rate": 0.00040299416183812634,
+      "loss": 3.1133,
+      "step": 23779
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5950905084609985,
+      "learning_rate": 0.00040297970128125955,
+      "loss": 3.0327,
+      "step": 23780
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5376090407371521,
+      "learning_rate": 0.00040296524045316403,
+      "loss": 3.1263,
+      "step": 23781
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.552521288394928,
+      "learning_rate": 0.00040295077935387783,
+      "loss": 3.0581,
+      "step": 23782
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5569019913673401,
+      "learning_rate": 0.0004029363179834391,
+      "loss": 3.0234,
+      "step": 23783
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5935654044151306,
+      "learning_rate": 0.000402921856341886,
+      "loss": 3.1942,
+      "step": 23784
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5477899312973022,
+      "learning_rate": 0.00040290739442925644,
+      "loss": 3.1383,
+      "step": 23785
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5384113788604736,
+      "learning_rate": 0.00040289293224558874,
+      "loss": 3.1697,
+      "step": 23786
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5377801656723022,
+      "learning_rate": 0.00040287846979092075,
+      "loss": 3.1197,
+      "step": 23787
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.560335636138916,
+      "learning_rate": 0.0004028640070652907,
+      "loss": 2.9627,
+      "step": 23788
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5258608460426331,
+      "learning_rate": 0.0004028495440687366,
+      "loss": 3.1526,
+      "step": 23789
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5460545420646667,
+      "learning_rate": 0.00040283508080129674,
+      "loss": 3.1596,
+      "step": 23790
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5956621766090393,
+      "learning_rate": 0.00040282061726300896,
+      "loss": 3.0147,
+      "step": 23791
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5216905474662781,
+      "learning_rate": 0.00040280615345391144,
+      "loss": 2.9739,
+      "step": 23792
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.508186399936676,
+      "learning_rate": 0.00040279168937404233,
+      "loss": 3.0885,
+      "step": 23793
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5671628713607788,
+      "learning_rate": 0.0004027772250234397,
+      "loss": 3.0457,
+      "step": 23794
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.582146406173706,
+      "learning_rate": 0.0004027627604021416,
+      "loss": 2.9687,
+      "step": 23795
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6142930388450623,
+      "learning_rate": 0.0004027482955101863,
+      "loss": 3.0002,
+      "step": 23796
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5688886642456055,
+      "learning_rate": 0.00040273383034761164,
+      "loss": 3.0811,
+      "step": 23797
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5424383878707886,
+      "learning_rate": 0.0004027193649144558,
+      "loss": 3.1528,
+      "step": 23798
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6007323861122131,
+      "learning_rate": 0.00040270489921075706,
+      "loss": 2.7161,
+      "step": 23799
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5330042839050293,
+      "learning_rate": 0.0004026904332365533,
+      "loss": 2.7069,
+      "step": 23800
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5643306970596313,
+      "learning_rate": 0.0004026759669918828,
+      "loss": 2.8007,
+      "step": 23801
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5669317245483398,
+      "learning_rate": 0.00040266150047678346,
+      "loss": 3.0615,
+      "step": 23802
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6755366325378418,
+      "learning_rate": 0.00040264703369129336,
+      "loss": 3.0581,
+      "step": 23803
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5715072751045227,
+      "learning_rate": 0.000402632566635451,
+      "loss": 3.1604,
+      "step": 23804
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5770458579063416,
+      "learning_rate": 0.0004026180993092941,
+      "loss": 2.9243,
+      "step": 23805
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5324617028236389,
+      "learning_rate": 0.0004026036317128608,
+      "loss": 3.3259,
+      "step": 23806
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5772245526313782,
+      "learning_rate": 0.00040258916384618935,
+      "loss": 3.2526,
+      "step": 23807
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5446373224258423,
+      "learning_rate": 0.0004025746957093178,
+      "loss": 3.2472,
+      "step": 23808
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5985844135284424,
+      "learning_rate": 0.0004025602273022842,
+      "loss": 3.1847,
+      "step": 23809
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5797475576400757,
+      "learning_rate": 0.00040254575862512667,
+      "loss": 2.9985,
+      "step": 23810
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5373989939689636,
+      "learning_rate": 0.00040253128967788343,
+      "loss": 2.9025,
+      "step": 23811
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5533153414726257,
+      "learning_rate": 0.0004025168204605925,
+      "loss": 3.2097,
+      "step": 23812
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5629292726516724,
+      "learning_rate": 0.000402502350973292,
+      "loss": 3.1,
+      "step": 23813
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5873868465423584,
+      "learning_rate": 0.00040248788121601995,
+      "loss": 3.1942,
+      "step": 23814
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5775611996650696,
+      "learning_rate": 0.00040247341118881464,
+      "loss": 2.9935,
+      "step": 23815
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5191925168037415,
+      "learning_rate": 0.000402458940891714,
+      "loss": 3.1487,
+      "step": 23816
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5707910656929016,
+      "learning_rate": 0.00040244447032475626,
+      "loss": 3.0536,
+      "step": 23817
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.555317223072052,
+      "learning_rate": 0.0004024299994879795,
+      "loss": 2.9648,
+      "step": 23818
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5581421852111816,
+      "learning_rate": 0.0004024155283814219,
+      "loss": 3.1438,
+      "step": 23819
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.607385516166687,
+      "learning_rate": 0.0004024010570051214,
+      "loss": 2.9115,
+      "step": 23820
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5479592084884644,
+      "learning_rate": 0.0004023865853591163,
+      "loss": 2.9924,
+      "step": 23821
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5671947598457336,
+      "learning_rate": 0.0004023721134434446,
+      "loss": 2.9375,
+      "step": 23822
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5418531894683838,
+      "learning_rate": 0.0004023576412581445,
+      "loss": 3.1671,
+      "step": 23823
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5851473212242126,
+      "learning_rate": 0.00040234316880325403,
+      "loss": 3.1394,
+      "step": 23824
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5450239181518555,
+      "learning_rate": 0.0004023286960788113,
+      "loss": 3.149,
+      "step": 23825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5457543730735779,
+      "learning_rate": 0.00040231422308485465,
+      "loss": 2.8019,
+      "step": 23826
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5215305685997009,
+      "learning_rate": 0.0004022997498214219,
+      "loss": 3.36,
+      "step": 23827
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5583608746528625,
+      "learning_rate": 0.0004022852762885513,
+      "loss": 2.9352,
+      "step": 23828
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5622576475143433,
+      "learning_rate": 0.0004022708024862809,
+      "loss": 3.029,
+      "step": 23829
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5386072993278503,
+      "learning_rate": 0.0004022563284146491,
+      "loss": 3.2786,
+      "step": 23830
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6399205327033997,
+      "learning_rate": 0.0004022418540736936,
+      "loss": 2.9981,
+      "step": 23831
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5961595177650452,
+      "learning_rate": 0.0004022273794634528,
+      "loss": 3.1205,
+      "step": 23832
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5393549203872681,
+      "learning_rate": 0.0004022129045839648,
+      "loss": 2.9914,
+      "step": 23833
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5417840480804443,
+      "learning_rate": 0.0004021984294352676,
+      "loss": 3.0758,
+      "step": 23834
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.582919180393219,
+      "learning_rate": 0.00040218395401739944,
+      "loss": 3.1598,
+      "step": 23835
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5843506455421448,
+      "learning_rate": 0.0004021694783303985,
+      "loss": 2.9717,
+      "step": 23836
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5359125733375549,
+      "learning_rate": 0.00040215500237430275,
+      "loss": 3.0764,
+      "step": 23837
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.608666181564331,
+      "learning_rate": 0.00040214052614915045,
+      "loss": 2.9321,
+      "step": 23838
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5720089673995972,
+      "learning_rate": 0.0004021260496549795,
+      "loss": 3.0585,
+      "step": 23839
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.528830885887146,
+      "learning_rate": 0.00040211157289182824,
+      "loss": 3.2521,
+      "step": 23840
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.57254958152771,
+      "learning_rate": 0.00040209709585973483,
+      "loss": 3.148,
+      "step": 23841
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.529454231262207,
+      "learning_rate": 0.0004020826185587373,
+      "loss": 3.0549,
+      "step": 23842
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5671842694282532,
+      "learning_rate": 0.00040206814098887373,
+      "loss": 3.1676,
+      "step": 23843
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5969575047492981,
+      "learning_rate": 0.00040205366315018237,
+      "loss": 3.3495,
+      "step": 23844
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5233942270278931,
+      "learning_rate": 0.00040203918504270135,
+      "loss": 2.9989,
+      "step": 23845
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5281428694725037,
+      "learning_rate": 0.0004020247066664686,
+      "loss": 3.1292,
+      "step": 23846
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5774059891700745,
+      "learning_rate": 0.00040201022802152257,
+      "loss": 2.922,
+      "step": 23847
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5433593392372131,
+      "learning_rate": 0.0004019957491079012,
+      "loss": 3.0951,
+      "step": 23848
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5464013814926147,
+      "learning_rate": 0.00040198126992564264,
+      "loss": 3.2079,
+      "step": 23849
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5496659874916077,
+      "learning_rate": 0.00040196679047478493,
+      "loss": 3.1016,
+      "step": 23850
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.542181134223938,
+      "learning_rate": 0.00040195231075536644,
+      "loss": 3.0924,
+      "step": 23851
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.596310019493103,
+      "learning_rate": 0.00040193783076742523,
+      "loss": 3.1904,
+      "step": 23852
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5427301526069641,
+      "learning_rate": 0.0004019233505109993,
+      "loss": 3.1372,
+      "step": 23853
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5772640705108643,
+      "learning_rate": 0.00040190886998612695,
+      "loss": 2.9875,
+      "step": 23854
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5249664187431335,
+      "learning_rate": 0.00040189438919284616,
+      "loss": 3.1938,
+      "step": 23855
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5812081694602966,
+      "learning_rate": 0.00040187990813119525,
+      "loss": 3.1606,
+      "step": 23856
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5547779202461243,
+      "learning_rate": 0.0004018654268012122,
+      "loss": 3.2556,
+      "step": 23857
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5426790714263916,
+      "learning_rate": 0.0004018509452029352,
+      "loss": 3.1825,
+      "step": 23858
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5476883053779602,
+      "learning_rate": 0.0004018364633364025,
+      "loss": 3.0241,
+      "step": 23859
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5173876285552979,
+      "learning_rate": 0.00040182198120165206,
+      "loss": 3.0905,
+      "step": 23860
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5504484176635742,
+      "learning_rate": 0.00040180749879872217,
+      "loss": 3.212,
+      "step": 23861
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5541347861289978,
+      "learning_rate": 0.0004017930161276509,
+      "loss": 2.9228,
+      "step": 23862
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.580261766910553,
+      "learning_rate": 0.0004017785331884765,
+      "loss": 2.9678,
+      "step": 23863
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6364923715591431,
+      "learning_rate": 0.000401764049981237,
+      "loss": 3.1864,
+      "step": 23864
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5156949758529663,
+      "learning_rate": 0.0004017495665059704,
+      "loss": 3.3035,
+      "step": 23865
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5521565675735474,
+      "learning_rate": 0.00040173508276271526,
+      "loss": 2.9317,
+      "step": 23866
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5523860454559326,
+      "learning_rate": 0.0004017205987515094,
+      "loss": 3.1082,
+      "step": 23867
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5502112507820129,
+      "learning_rate": 0.00040170611447239103,
+      "loss": 3.0595,
+      "step": 23868
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5495514273643494,
+      "learning_rate": 0.00040169162992539833,
+      "loss": 3.05,
+      "step": 23869
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5663180947303772,
+      "learning_rate": 0.00040167714511056947,
+      "loss": 3.133,
+      "step": 23870
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5522775650024414,
+      "learning_rate": 0.0004016626600279426,
+      "loss": 3.1148,
+      "step": 23871
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5634680390357971,
+      "learning_rate": 0.00040164817467755584,
+      "loss": 3.1422,
+      "step": 23872
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5490415096282959,
+      "learning_rate": 0.0004016336890594473,
+      "loss": 2.8798,
+      "step": 23873
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5496166348457336,
+      "learning_rate": 0.00040161920317365534,
+      "loss": 2.8576,
+      "step": 23874
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5630446076393127,
+      "learning_rate": 0.00040160471702021773,
+      "loss": 3.266,
+      "step": 23875
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5701159238815308,
+      "learning_rate": 0.0004015902305991729,
+      "loss": 3.0536,
+      "step": 23876
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5700026154518127,
+      "learning_rate": 0.00040157574391055907,
+      "loss": 2.979,
+      "step": 23877
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5726680159568787,
+      "learning_rate": 0.0004015612569544143,
+      "loss": 3.0092,
+      "step": 23878
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5549166202545166,
+      "learning_rate": 0.0004015467697307766,
+      "loss": 3.2318,
+      "step": 23879
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5379598736763,
+      "learning_rate": 0.0004015322822396842,
+      "loss": 2.9661,
+      "step": 23880
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5367391705513,
+      "learning_rate": 0.00040151779448117545,
+      "loss": 3.1219,
+      "step": 23881
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5859577655792236,
+      "learning_rate": 0.00040150330645528833,
+      "loss": 3.0839,
+      "step": 23882
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5608367323875427,
+      "learning_rate": 0.0004014888181620611,
+      "loss": 3.1321,
+      "step": 23883
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5417865514755249,
+      "learning_rate": 0.00040147432960153175,
+      "loss": 2.9928,
+      "step": 23884
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5564191341400146,
+      "learning_rate": 0.0004014598407737386,
+      "loss": 3.0766,
+      "step": 23885
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5441717505455017,
+      "learning_rate": 0.0004014453516787197,
+      "loss": 3.1433,
+      "step": 23886
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5383797287940979,
+      "learning_rate": 0.0004014308623165133,
+      "loss": 3.194,
+      "step": 23887
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6364131569862366,
+      "learning_rate": 0.00040141637268715754,
+      "loss": 3.347,
+      "step": 23888
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.565334677696228,
+      "learning_rate": 0.0004014018827906906,
+      "loss": 3.252,
+      "step": 23889
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5537137985229492,
+      "learning_rate": 0.0004013873926271506,
+      "loss": 3.0982,
+      "step": 23890
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5591994524002075,
+      "learning_rate": 0.0004013729021965756,
+      "loss": 3.0882,
+      "step": 23891
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5560016632080078,
+      "learning_rate": 0.0004013584114990041,
+      "loss": 3.1173,
+      "step": 23892
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5662415623664856,
+      "learning_rate": 0.00040134392053447387,
+      "loss": 2.995,
+      "step": 23893
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5805153846740723,
+      "learning_rate": 0.0004013294293030233,
+      "loss": 3.3028,
+      "step": 23894
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5564658045768738,
+      "learning_rate": 0.0004013149378046905,
+      "loss": 2.9882,
+      "step": 23895
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5353381633758545,
+      "learning_rate": 0.0004013004460395137,
+      "loss": 2.9248,
+      "step": 23896
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5824044346809387,
+      "learning_rate": 0.000401285954007531,
+      "loss": 2.8985,
+      "step": 23897
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5630552768707275,
+      "learning_rate": 0.0004012714617087806,
+      "loss": 3.1425,
+      "step": 23898
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5532223582267761,
+      "learning_rate": 0.0004012569691433008,
+      "loss": 3.1631,
+      "step": 23899
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5480971932411194,
+      "learning_rate": 0.00040124247631112943,
+      "loss": 2.8964,
+      "step": 23900
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5849064588546753,
+      "learning_rate": 0.00040122798321230485,
+      "loss": 2.9088,
+      "step": 23901
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5372636318206787,
+      "learning_rate": 0.0004012134898468653,
+      "loss": 2.7437,
+      "step": 23902
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5332382321357727,
+      "learning_rate": 0.000401198996214849,
+      "loss": 3.2017,
+      "step": 23903
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5774133801460266,
+      "learning_rate": 0.00040118450231629393,
+      "loss": 3.2153,
+      "step": 23904
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5455067157745361,
+      "learning_rate": 0.0004011700081512384,
+      "loss": 3.1119,
+      "step": 23905
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5268059968948364,
+      "learning_rate": 0.00040115551371972045,
+      "loss": 3.0404,
+      "step": 23906
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5563852190971375,
+      "learning_rate": 0.0004011410190217784,
+      "loss": 2.8807,
+      "step": 23907
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5379848480224609,
+      "learning_rate": 0.0004011265240574503,
+      "loss": 3.1135,
+      "step": 23908
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5759731531143188,
+      "learning_rate": 0.0004011120288267745,
+      "loss": 3.054,
+      "step": 23909
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.571288526058197,
+      "learning_rate": 0.0004010975333297891,
+      "loss": 3.1158,
+      "step": 23910
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5381522178649902,
+      "learning_rate": 0.00040108303756653213,
+      "loss": 3.1559,
+      "step": 23911
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6180474162101746,
+      "learning_rate": 0.00040106854153704196,
+      "loss": 3.0476,
+      "step": 23912
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5325801372528076,
+      "learning_rate": 0.00040105404524135664,
+      "loss": 3.0364,
+      "step": 23913
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5483402013778687,
+      "learning_rate": 0.00040103954867951454,
+      "loss": 2.9758,
+      "step": 23914
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5332612991333008,
+      "learning_rate": 0.00040102505185155365,
+      "loss": 3.0309,
+      "step": 23915
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5399067997932434,
+      "learning_rate": 0.00040101055475751216,
+      "loss": 3.2472,
+      "step": 23916
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5286165475845337,
+      "learning_rate": 0.0004009960573974284,
+      "loss": 3.3073,
+      "step": 23917
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5932464003562927,
+      "learning_rate": 0.0004009815597713404,
+      "loss": 3.0836,
+      "step": 23918
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5477786660194397,
+      "learning_rate": 0.00040096706187928643,
+      "loss": 3.1094,
+      "step": 23919
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5101897716522217,
+      "learning_rate": 0.0004009525637213046,
+      "loss": 3.043,
+      "step": 23920
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5300383567810059,
+      "learning_rate": 0.0004009380652974332,
+      "loss": 3.2401,
+      "step": 23921
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5190649628639221,
+      "learning_rate": 0.0004009235666077103,
+      "loss": 3.224,
+      "step": 23922
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5531300902366638,
+      "learning_rate": 0.00040090906765217425,
+      "loss": 2.9455,
+      "step": 23923
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5615624189376831,
+      "learning_rate": 0.00040089456843086304,
+      "loss": 2.9886,
+      "step": 23924
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.562551736831665,
+      "learning_rate": 0.0004008800689438151,
+      "loss": 3.0263,
+      "step": 23925
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.541335940361023,
+      "learning_rate": 0.00040086556919106833,
+      "loss": 3.0767,
+      "step": 23926
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5761546492576599,
+      "learning_rate": 0.00040085106917266106,
+      "loss": 2.9538,
+      "step": 23927
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5472856163978577,
+      "learning_rate": 0.00040083656888863163,
+      "loss": 2.8641,
+      "step": 23928
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5612488389015198,
+      "learning_rate": 0.000400822068339018,
+      "loss": 2.9537,
+      "step": 23929
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5841686725616455,
+      "learning_rate": 0.0004008075675238584,
+      "loss": 3.1257,
+      "step": 23930
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5874391794204712,
+      "learning_rate": 0.0004007930664431911,
+      "loss": 3.0393,
+      "step": 23931
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5234120488166809,
+      "learning_rate": 0.00040077856509705424,
+      "loss": 3.1151,
+      "step": 23932
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5505731105804443,
+      "learning_rate": 0.00040076406348548603,
+      "loss": 3.1803,
+      "step": 23933
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5326248407363892,
+      "learning_rate": 0.00040074956160852476,
+      "loss": 3.2542,
+      "step": 23934
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5707079768180847,
+      "learning_rate": 0.00040073505946620837,
+      "loss": 3.028,
+      "step": 23935
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5336579084396362,
+      "learning_rate": 0.0004007205570585754,
+      "loss": 3.2687,
+      "step": 23936
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5472836494445801,
+      "learning_rate": 0.00040070605438566375,
+      "loss": 2.8165,
+      "step": 23937
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6249715089797974,
+      "learning_rate": 0.00040069155144751173,
+      "loss": 3.1156,
+      "step": 23938
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.533698320388794,
+      "learning_rate": 0.00040067704824415764,
+      "loss": 2.9911,
+      "step": 23939
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5471363663673401,
+      "learning_rate": 0.00040066254477563956,
+      "loss": 2.9763,
+      "step": 23940
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5757012963294983,
+      "learning_rate": 0.0004006480410419957,
+      "loss": 3.2284,
+      "step": 23941
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5495263934135437,
+      "learning_rate": 0.0004006335370432641,
+      "loss": 2.9926,
+      "step": 23942
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5950400233268738,
+      "learning_rate": 0.00040061903277948335,
+      "loss": 2.9854,
+      "step": 23943
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5375027060508728,
+      "learning_rate": 0.00040060452825069133,
+      "loss": 3.3018,
+      "step": 23944
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5809231400489807,
+      "learning_rate": 0.00040059002345692644,
+      "loss": 3.1755,
+      "step": 23945
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5961464643478394,
+      "learning_rate": 0.00040057551839822665,
+      "loss": 3.0865,
+      "step": 23946
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.574310839176178,
+      "learning_rate": 0.0004005610130746303,
+      "loss": 3.0654,
+      "step": 23947
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5345257520675659,
+      "learning_rate": 0.00040054650748617575,
+      "loss": 3.1405,
+      "step": 23948
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.565226674079895,
+      "learning_rate": 0.0004005320016329009,
+      "loss": 2.7882,
+      "step": 23949
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5397844314575195,
+      "learning_rate": 0.00040051749551484425,
+      "loss": 2.9382,
+      "step": 23950
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5760013461112976,
+      "learning_rate": 0.00040050298913204375,
+      "loss": 3.017,
+      "step": 23951
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5508855581283569,
+      "learning_rate": 0.0004004884824845377,
+      "loss": 3.2093,
+      "step": 23952
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5632848739624023,
+      "learning_rate": 0.0004004739755723644,
+      "loss": 3.0376,
+      "step": 23953
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.536578893661499,
+      "learning_rate": 0.00040045946839556196,
+      "loss": 3.0994,
+      "step": 23954
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5494163632392883,
+      "learning_rate": 0.00040044496095416863,
+      "loss": 2.9971,
+      "step": 23955
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5547284483909607,
+      "learning_rate": 0.0004004304532482226,
+      "loss": 3.268,
+      "step": 23956
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.551442563533783,
+      "learning_rate": 0.00040041594527776204,
+      "loss": 3.0469,
+      "step": 23957
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5382516384124756,
+      "learning_rate": 0.00040040143704282526,
+      "loss": 3.4146,
+      "step": 23958
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5559409856796265,
+      "learning_rate": 0.0004003869285434504,
+      "loss": 3.0229,
+      "step": 23959
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5598931312561035,
+      "learning_rate": 0.0004003724197796756,
+      "loss": 3.0695,
+      "step": 23960
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.59652179479599,
+      "learning_rate": 0.00040035791075153935,
+      "loss": 3.1012,
+      "step": 23961
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5521093606948853,
+      "learning_rate": 0.0004003434014590796,
+      "loss": 3.1512,
+      "step": 23962
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5462397933006287,
+      "learning_rate": 0.0004003288919023346,
+      "loss": 3.1358,
+      "step": 23963
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5487635731697083,
+      "learning_rate": 0.0004003143820813425,
+      "loss": 3.373,
+      "step": 23964
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6183003783226013,
+      "learning_rate": 0.00040029987199614186,
+      "loss": 2.8504,
+      "step": 23965
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5639381408691406,
+      "learning_rate": 0.00040028536164677053,
+      "loss": 2.9285,
+      "step": 23966
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5435758829116821,
+      "learning_rate": 0.00040027085103326685,
+      "loss": 3.1571,
+      "step": 23967
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5409099459648132,
+      "learning_rate": 0.0004002563401556691,
+      "loss": 3.0643,
+      "step": 23968
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.564572811126709,
+      "learning_rate": 0.00040024182901401544,
+      "loss": 3.0944,
+      "step": 23969
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5847697257995605,
+      "learning_rate": 0.00040022731760834405,
+      "loss": 3.1069,
+      "step": 23970
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5416810512542725,
+      "learning_rate": 0.00040021280593869317,
+      "loss": 3.1393,
+      "step": 23971
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5735564231872559,
+      "learning_rate": 0.00040019829400510113,
+      "loss": 3.278,
+      "step": 23972
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5522201657295227,
+      "learning_rate": 0.000400183781807606,
+      "loss": 3.1469,
+      "step": 23973
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5339809060096741,
+      "learning_rate": 0.00040016926934624607,
+      "loss": 3.2685,
+      "step": 23974
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5736157894134521,
+      "learning_rate": 0.00040015475662105965,
+      "loss": 3.1981,
+      "step": 23975
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5715289115905762,
+      "learning_rate": 0.00040014024363208483,
+      "loss": 3.1167,
+      "step": 23976
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5391354560852051,
+      "learning_rate": 0.00040012573037935984,
+      "loss": 3.2589,
+      "step": 23977
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5813712477684021,
+      "learning_rate": 0.0004001112168629229,
+      "loss": 3.0653,
+      "step": 23978
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6210728287696838,
+      "learning_rate": 0.0004000967030828124,
+      "loss": 3.1392,
+      "step": 23979
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5406283736228943,
+      "learning_rate": 0.0004000821890390664,
+      "loss": 3.2409,
+      "step": 23980
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5507621765136719,
+      "learning_rate": 0.00040006767473172323,
+      "loss": 3.1331,
+      "step": 23981
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5478275418281555,
+      "learning_rate": 0.00040005316016082095,
+      "loss": 2.9733,
+      "step": 23982
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6610849499702454,
+      "learning_rate": 0.00040003864532639797,
+      "loss": 2.9103,
+      "step": 23983
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5360865592956543,
+      "learning_rate": 0.00040002413022849245,
+      "loss": 3.118,
+      "step": 23984
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5579555034637451,
+      "learning_rate": 0.0004000096148671426,
+      "loss": 3.1183,
+      "step": 23985
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5361294746398926,
+      "learning_rate": 0.0003999950992423867,
+      "loss": 3.1518,
+      "step": 23986
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5184491872787476,
+      "learning_rate": 0.000399980583354263,
+      "loss": 2.8963,
+      "step": 23987
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.551647961139679,
+      "learning_rate": 0.0003999660672028095,
+      "loss": 3.1424,
+      "step": 23988
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5427066087722778,
+      "learning_rate": 0.0003999515507880648,
+      "loss": 3.0262,
+      "step": 23989
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5327326059341431,
+      "learning_rate": 0.0003999370341100669,
+      "loss": 2.976,
+      "step": 23990
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5397793650627136,
+      "learning_rate": 0.0003999225171688541,
+      "loss": 2.9345,
+      "step": 23991
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5645992159843445,
+      "learning_rate": 0.00039990799996446466,
+      "loss": 2.9815,
+      "step": 23992
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5220346450805664,
+      "learning_rate": 0.00039989348249693657,
+      "loss": 3.2895,
+      "step": 23993
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5440986156463623,
+      "learning_rate": 0.00039987896476630845,
+      "loss": 2.9243,
+      "step": 23994
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6222359538078308,
+      "learning_rate": 0.00039986444677261835,
+      "loss": 3.1331,
+      "step": 23995
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5543932914733887,
+      "learning_rate": 0.0003998499285159045,
+      "loss": 3.0349,
+      "step": 23996
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5490472316741943,
+      "learning_rate": 0.0003998354099962051,
+      "loss": 3.0956,
+      "step": 23997
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5755308866500854,
+      "learning_rate": 0.0003998208912135585,
+      "loss": 3.0513,
+      "step": 23998
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5586222410202026,
+      "learning_rate": 0.0003998063721680029,
+      "loss": 3.3977,
+      "step": 23999
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5579812526702881,
+      "learning_rate": 0.00039979185285957645,
+      "loss": 2.8466,
+      "step": 24000
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5372399091720581,
+      "learning_rate": 0.0003997773332883176,
+      "loss": 3.0433,
+      "step": 24001
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5638691186904907,
+      "learning_rate": 0.0003997628134542644,
+      "loss": 2.8779,
+      "step": 24002
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5711449384689331,
+      "learning_rate": 0.0003997482933574551,
+      "loss": 3.1942,
+      "step": 24003
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5793060660362244,
+      "learning_rate": 0.0003997337729979279,
+      "loss": 2.9375,
+      "step": 24004
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5681477189064026,
+      "learning_rate": 0.00039971925237572127,
+      "loss": 3.0421,
+      "step": 24005
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5376327037811279,
+      "learning_rate": 0.0003997047314908733,
+      "loss": 3.0035,
+      "step": 24006
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5498843789100647,
+      "learning_rate": 0.00039969021034342233,
+      "loss": 3.1599,
+      "step": 24007
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5775700807571411,
+      "learning_rate": 0.0003996756889334064,
+      "loss": 2.8919,
+      "step": 24008
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5586049556732178,
+      "learning_rate": 0.00039966116726086395,
+      "loss": 3.3297,
+      "step": 24009
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5521610379219055,
+      "learning_rate": 0.0003996466453258331,
+      "loss": 2.9728,
+      "step": 24010
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5638512969017029,
+      "learning_rate": 0.0003996321231283523,
+      "loss": 3.262,
+      "step": 24011
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7236627340316772,
+      "learning_rate": 0.0003996176006684596,
+      "loss": 3.0747,
+      "step": 24012
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.555939793586731,
+      "learning_rate": 0.0003996030779461933,
+      "loss": 3.1563,
+      "step": 24013
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5715100169181824,
+      "learning_rate": 0.0003995885549615916,
+      "loss": 3.1999,
+      "step": 24014
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5524293780326843,
+      "learning_rate": 0.00039957403171469286,
+      "loss": 2.9745,
+      "step": 24015
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.545418381690979,
+      "learning_rate": 0.0003995595082055354,
+      "loss": 2.8698,
+      "step": 24016
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5677865743637085,
+      "learning_rate": 0.00039954498443415723,
+      "loss": 2.9115,
+      "step": 24017
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5774267911911011,
+      "learning_rate": 0.0003995304604005967,
+      "loss": 3.2582,
+      "step": 24018
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5392197966575623,
+      "learning_rate": 0.00039951593610489214,
+      "loss": 3.1144,
+      "step": 24019
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6039441823959351,
+      "learning_rate": 0.00039950141154708176,
+      "loss": 2.9836,
+      "step": 24020
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5584951639175415,
+      "learning_rate": 0.0003994868867272038,
+      "loss": 3.0626,
+      "step": 24021
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5053369402885437,
+      "learning_rate": 0.0003994723616452965,
+      "loss": 2.9783,
+      "step": 24022
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5872316956520081,
+      "learning_rate": 0.0003994578363013982,
+      "loss": 3.0212,
+      "step": 24023
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6137145161628723,
+      "learning_rate": 0.0003994433106955471,
+      "loss": 3.0148,
+      "step": 24024
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8941653370857239,
+      "learning_rate": 0.00039942878482778144,
+      "loss": 2.8822,
+      "step": 24025
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5612619519233704,
+      "learning_rate": 0.00039941425869813947,
+      "loss": 3.1952,
+      "step": 24026
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.553356409072876,
+      "learning_rate": 0.0003993997323066596,
+      "loss": 3.2152,
+      "step": 24027
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5677022337913513,
+      "learning_rate": 0.00039938520565337984,
+      "loss": 3.0083,
+      "step": 24028
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5404044985771179,
+      "learning_rate": 0.0003993706787383385,
+      "loss": 3.0666,
+      "step": 24029
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5190367698669434,
+      "learning_rate": 0.00039935615156157407,
+      "loss": 3.021,
+      "step": 24030
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5348750948905945,
+      "learning_rate": 0.0003993416241231246,
+      "loss": 3.0933,
+      "step": 24031
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5640134811401367,
+      "learning_rate": 0.0003993270964230284,
+      "loss": 3.0944,
+      "step": 24032
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.571087658405304,
+      "learning_rate": 0.0003993125684613238,
+      "loss": 2.8386,
+      "step": 24033
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5981314778327942,
+      "learning_rate": 0.000399298040238049,
+      "loss": 3.3129,
+      "step": 24034
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5638099908828735,
+      "learning_rate": 0.00039928351175324216,
+      "loss": 3.1856,
+      "step": 24035
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5106322169303894,
+      "learning_rate": 0.00039926898300694173,
+      "loss": 2.9335,
+      "step": 24036
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.545299232006073,
+      "learning_rate": 0.0003992544539991859,
+      "loss": 3.3001,
+      "step": 24037
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5997403264045715,
+      "learning_rate": 0.00039923992473001307,
+      "loss": 3.2128,
+      "step": 24038
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5346583127975464,
+      "learning_rate": 0.0003992253951994612,
+      "loss": 2.8758,
+      "step": 24039
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.560383141040802,
+      "learning_rate": 0.0003992108654075687,
+      "loss": 2.9834,
+      "step": 24040
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5270285606384277,
+      "learning_rate": 0.0003991963353543741,
+      "loss": 3.0306,
+      "step": 24041
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5616056323051453,
+      "learning_rate": 0.00039918180503991524,
+      "loss": 3.1098,
+      "step": 24042
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5329706072807312,
+      "learning_rate": 0.0003991672744642307,
+      "loss": 3.1355,
+      "step": 24043
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7826784253120422,
+      "learning_rate": 0.0003991527436273586,
+      "loss": 3.0251,
+      "step": 24044
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5575398206710815,
+      "learning_rate": 0.0003991382125293373,
+      "loss": 3.0598,
+      "step": 24045
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5318834185600281,
+      "learning_rate": 0.0003991236811702049,
+      "loss": 3.1486,
+      "step": 24046
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5418036580085754,
+      "learning_rate": 0.0003991091495499999,
+      "loss": 2.9798,
+      "step": 24047
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5505343675613403,
+      "learning_rate": 0.0003990946176687605,
+      "loss": 3.2627,
+      "step": 24048
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5388018488883972,
+      "learning_rate": 0.000399080085526525,
+      "loss": 3.0686,
+      "step": 24049
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6034098267555237,
+      "learning_rate": 0.0003990655531233314,
+      "loss": 3.0145,
+      "step": 24050
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5737993717193604,
+      "learning_rate": 0.0003990510204592184,
+      "loss": 3.0209,
+      "step": 24051
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7313047051429749,
+      "learning_rate": 0.000399036487534224,
+      "loss": 3.1884,
+      "step": 24052
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5465160012245178,
+      "learning_rate": 0.00039902195434838656,
+      "loss": 3.2552,
+      "step": 24053
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5565720200538635,
+      "learning_rate": 0.00039900742090174434,
+      "loss": 3.2226,
+      "step": 24054
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.525080144405365,
+      "learning_rate": 0.0003989928871943356,
+      "loss": 3.1065,
+      "step": 24055
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5119069814682007,
+      "learning_rate": 0.00039897835322619875,
+      "loss": 3.3655,
+      "step": 24056
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5813723206520081,
+      "learning_rate": 0.00039896381899737194,
+      "loss": 2.8918,
+      "step": 24057
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5397539138793945,
+      "learning_rate": 0.0003989492845078934,
+      "loss": 3.1229,
+      "step": 24058
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.571702778339386,
+      "learning_rate": 0.0003989347497578015,
+      "loss": 3.1404,
+      "step": 24059
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5537548065185547,
+      "learning_rate": 0.00039892021474713455,
+      "loss": 3.0636,
+      "step": 24060
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5488508343696594,
+      "learning_rate": 0.0003989056794759307,
+      "loss": 3.1616,
+      "step": 24061
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5703139901161194,
+      "learning_rate": 0.00039889114394422846,
+      "loss": 3.2048,
+      "step": 24062
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5412448644638062,
+      "learning_rate": 0.000398876608152066,
+      "loss": 3.161,
+      "step": 24063
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5422495007514954,
+      "learning_rate": 0.0003988620720994815,
+      "loss": 3.114,
+      "step": 24064
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5373824834823608,
+      "learning_rate": 0.00039884753578651327,
+      "loss": 3.222,
+      "step": 24065
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5398744344711304,
+      "learning_rate": 0.0003988329992131997,
+      "loss": 2.9718,
+      "step": 24066
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5511226058006287,
+      "learning_rate": 0.00039881846237957907,
+      "loss": 3.0787,
+      "step": 24067
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.548173725605011,
+      "learning_rate": 0.00039880392528568963,
+      "loss": 2.9258,
+      "step": 24068
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5512843728065491,
+      "learning_rate": 0.0003987893879315696,
+      "loss": 2.9631,
+      "step": 24069
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.58067387342453,
+      "learning_rate": 0.0003987748503172574,
+      "loss": 3.2054,
+      "step": 24070
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5880147814750671,
+      "learning_rate": 0.00039876031244279127,
+      "loss": 3.1545,
+      "step": 24071
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5471587777137756,
+      "learning_rate": 0.0003987457743082094,
+      "loss": 3.0934,
+      "step": 24072
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5884271264076233,
+      "learning_rate": 0.00039873123591355024,
+      "loss": 3.1801,
+      "step": 24073
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5782662630081177,
+      "learning_rate": 0.00039871669725885197,
+      "loss": 2.9995,
+      "step": 24074
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5610849261283875,
+      "learning_rate": 0.0003987021583441529,
+      "loss": 2.916,
+      "step": 24075
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5222210884094238,
+      "learning_rate": 0.00039868761916949134,
+      "loss": 3.0684,
+      "step": 24076
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5449653267860413,
+      "learning_rate": 0.0003986730797349056,
+      "loss": 3.2768,
+      "step": 24077
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5474014282226562,
+      "learning_rate": 0.00039865854004043406,
+      "loss": 3.3026,
+      "step": 24078
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.564146876335144,
+      "learning_rate": 0.0003986440000861148,
+      "loss": 3.0649,
+      "step": 24079
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5155508518218994,
+      "learning_rate": 0.00039862945987198626,
+      "loss": 2.9618,
+      "step": 24080
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5813475251197815,
+      "learning_rate": 0.0003986149193980867,
+      "loss": 3.1608,
+      "step": 24081
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5683935284614563,
+      "learning_rate": 0.0003986003786644544,
+      "loss": 3.1103,
+      "step": 24082
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1067701578140259,
+      "learning_rate": 0.0003985858376711277,
+      "loss": 3.224,
+      "step": 24083
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5283617377281189,
+      "learning_rate": 0.0003985712964181448,
+      "loss": 3.2039,
+      "step": 24084
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5957078337669373,
+      "learning_rate": 0.00039855675490554416,
+      "loss": 3.2682,
+      "step": 24085
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.549714207649231,
+      "learning_rate": 0.0003985422131333639,
+      "loss": 3.0082,
+      "step": 24086
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5549113154411316,
+      "learning_rate": 0.00039852767110164255,
+      "loss": 3.0617,
+      "step": 24087
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5381019115447998,
+      "learning_rate": 0.0003985131288104182,
+      "loss": 3.1637,
+      "step": 24088
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5498014092445374,
+      "learning_rate": 0.0003984985862597293,
+      "loss": 3.1576,
+      "step": 24089
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5414360761642456,
+      "learning_rate": 0.000398484043449614,
+      "loss": 3.2084,
+      "step": 24090
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5614427924156189,
+      "learning_rate": 0.0003984695003801107,
+      "loss": 3.0681,
+      "step": 24091
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5734679698944092,
+      "learning_rate": 0.00039845495705125776,
+      "loss": 3.0153,
+      "step": 24092
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5509762763977051,
+      "learning_rate": 0.00039844041346309337,
+      "loss": 2.9354,
+      "step": 24093
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5663211941719055,
+      "learning_rate": 0.00039842586961565585,
+      "loss": 3.1139,
+      "step": 24094
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5385051965713501,
+      "learning_rate": 0.0003984113255089835,
+      "loss": 3.0458,
+      "step": 24095
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.543887197971344,
+      "learning_rate": 0.0003983967811431147,
+      "loss": 3.1119,
+      "step": 24096
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5807494521141052,
+      "learning_rate": 0.0003983822365180877,
+      "loss": 3.1381,
+      "step": 24097
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5525929927825928,
+      "learning_rate": 0.0003983676916339408,
+      "loss": 3.1725,
+      "step": 24098
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5865118503570557,
+      "learning_rate": 0.00039835314649071235,
+      "loss": 3.1903,
+      "step": 24099
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5506489276885986,
+      "learning_rate": 0.0003983386010884408,
+      "loss": 2.9352,
+      "step": 24100
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.561968207359314,
+      "learning_rate": 0.00039832405542716403,
+      "loss": 2.9563,
+      "step": 24101
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5727139115333557,
+      "learning_rate": 0.00039830950950692077,
+      "loss": 3.2105,
+      "step": 24102
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6121776103973389,
+      "learning_rate": 0.0003982949633277492,
+      "loss": 3.0155,
+      "step": 24103
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5447525382041931,
+      "learning_rate": 0.0003982804168896876,
+      "loss": 3.2316,
+      "step": 24104
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.540285587310791,
+      "learning_rate": 0.0003982658701927742,
+      "loss": 3.2454,
+      "step": 24105
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5771687030792236,
+      "learning_rate": 0.00039825132323704744,
+      "loss": 3.3046,
+      "step": 24106
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5980179905891418,
+      "learning_rate": 0.00039823677602254576,
+      "loss": 3.088,
+      "step": 24107
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5670889019966125,
+      "learning_rate": 0.0003982222285493072,
+      "loss": 3.3893,
+      "step": 24108
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5559228658676147,
+      "learning_rate": 0.0003982076808173701,
+      "loss": 3.106,
+      "step": 24109
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5641289353370667,
+      "learning_rate": 0.000398193132826773,
+      "loss": 3.0903,
+      "step": 24110
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5894806981086731,
+      "learning_rate": 0.000398178584577554,
+      "loss": 3.1442,
+      "step": 24111
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.557578980922699,
+      "learning_rate": 0.0003981640360697516,
+      "loss": 2.9691,
+      "step": 24112
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5747301578521729,
+      "learning_rate": 0.000398149487303404,
+      "loss": 2.961,
+      "step": 24113
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5464869737625122,
+      "learning_rate": 0.00039813493827854955,
+      "loss": 3.1954,
+      "step": 24114
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5992234945297241,
+      "learning_rate": 0.0003981203889952265,
+      "loss": 3.0543,
+      "step": 24115
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5721291899681091,
+      "learning_rate": 0.00039810583945347326,
+      "loss": 3.1144,
+      "step": 24116
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6171442270278931,
+      "learning_rate": 0.000398091289653328,
+      "loss": 3.1549,
+      "step": 24117
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5718544125556946,
+      "learning_rate": 0.00039807673959482936,
+      "loss": 3.1521,
+      "step": 24118
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5414995551109314,
+      "learning_rate": 0.0003980621892780154,
+      "loss": 3.1951,
+      "step": 24119
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5666779279708862,
+      "learning_rate": 0.00039804763870292443,
+      "loss": 3.0715,
+      "step": 24120
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5580045580863953,
+      "learning_rate": 0.0003980330878695949,
+      "loss": 3.2015,
+      "step": 24121
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5530674457550049,
+      "learning_rate": 0.00039801853677806507,
+      "loss": 3.3237,
+      "step": 24122
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5504000186920166,
+      "learning_rate": 0.0003980039854283733,
+      "loss": 3.0533,
+      "step": 24123
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5976004600524902,
+      "learning_rate": 0.0003979894338205579,
+      "loss": 3.1773,
+      "step": 24124
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5514919757843018,
+      "learning_rate": 0.00039797488195465725,
+      "loss": 3.1461,
+      "step": 24125
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.572727620601654,
+      "learning_rate": 0.0003979603298307095,
+      "loss": 3.2894,
+      "step": 24126
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5436162948608398,
+      "learning_rate": 0.00039794577744875313,
+      "loss": 3.0208,
+      "step": 24127
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5952852368354797,
+      "learning_rate": 0.0003979312248088264,
+      "loss": 3.1083,
+      "step": 24128
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5643694400787354,
+      "learning_rate": 0.00039791667191096773,
+      "loss": 3.1304,
+      "step": 24129
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.55324786901474,
+      "learning_rate": 0.0003979021187552154,
+      "loss": 2.9863,
+      "step": 24130
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6007174253463745,
+      "learning_rate": 0.0003978875653416077,
+      "loss": 2.975,
+      "step": 24131
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5391753315925598,
+      "learning_rate": 0.00039787301167018297,
+      "loss": 2.843,
+      "step": 24132
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5486133098602295,
+      "learning_rate": 0.00039785845774097957,
+      "loss": 2.8428,
+      "step": 24133
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5005714297294617,
+      "learning_rate": 0.00039784390355403584,
+      "loss": 3.3195,
+      "step": 24134
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5840128064155579,
+      "learning_rate": 0.0003978293491093901,
+      "loss": 3.1903,
+      "step": 24135
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5641036033630371,
+      "learning_rate": 0.00039781479440708067,
+      "loss": 3.0371,
+      "step": 24136
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5386980772018433,
+      "learning_rate": 0.0003978002394471459,
+      "loss": 3.1271,
+      "step": 24137
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5703660249710083,
+      "learning_rate": 0.00039778568422962413,
+      "loss": 3.1415,
+      "step": 24138
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5552119016647339,
+      "learning_rate": 0.0003977711287545537,
+      "loss": 3.0826,
+      "step": 24139
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5309668779373169,
+      "learning_rate": 0.00039775657302197295,
+      "loss": 3.1237,
+      "step": 24140
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5294352173805237,
+      "learning_rate": 0.0003977420170319201,
+      "loss": 3.1383,
+      "step": 24141
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.522890031337738,
+      "learning_rate": 0.0003977274607844337,
+      "loss": 3.101,
+      "step": 24142
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6287959218025208,
+      "learning_rate": 0.000397712904279552,
+      "loss": 3.0709,
+      "step": 24143
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5275660157203674,
+      "learning_rate": 0.0003976983475173132,
+      "loss": 3.1218,
+      "step": 24144
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5518277287483215,
+      "learning_rate": 0.0003976837904977558,
+      "loss": 3.0445,
+      "step": 24145
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5437619090080261,
+      "learning_rate": 0.0003976692332209181,
+      "loss": 3.0587,
+      "step": 24146
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6017915606498718,
+      "learning_rate": 0.00039765467568683846,
+      "loss": 3.0058,
+      "step": 24147
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5371351838111877,
+      "learning_rate": 0.00039764011789555513,
+      "loss": 2.9967,
+      "step": 24148
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5650933980941772,
+      "learning_rate": 0.0003976255598471066,
+      "loss": 2.981,
+      "step": 24149
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5488359332084656,
+      "learning_rate": 0.00039761100154153103,
+      "loss": 3.1748,
+      "step": 24150
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5488278269767761,
+      "learning_rate": 0.00039759644297886703,
+      "loss": 2.9815,
+      "step": 24151
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.562656819820404,
+      "learning_rate": 0.0003975818841591526,
+      "loss": 3.2104,
+      "step": 24152
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.530597448348999,
+      "learning_rate": 0.00039756732508242635,
+      "loss": 3.1563,
+      "step": 24153
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.520431399345398,
+      "learning_rate": 0.00039755276574872665,
+      "loss": 3.0995,
+      "step": 24154
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5567206740379333,
+      "learning_rate": 0.0003975382061580916,
+      "loss": 3.1326,
+      "step": 24155
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5744289755821228,
+      "learning_rate": 0.0003975236463105597,
+      "loss": 3.183,
+      "step": 24156
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5486968755722046,
+      "learning_rate": 0.00039750908620616936,
+      "loss": 3.2061,
+      "step": 24157
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5852683186531067,
+      "learning_rate": 0.00039749452584495875,
+      "loss": 3.0303,
+      "step": 24158
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5728834271430969,
+      "learning_rate": 0.0003974799652269664,
+      "loss": 3.0113,
+      "step": 24159
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5454309582710266,
+      "learning_rate": 0.0003974654043522306,
+      "loss": 3.1408,
+      "step": 24160
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5599636435508728,
+      "learning_rate": 0.00039745084322078956,
+      "loss": 3.2156,
+      "step": 24161
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.540389358997345,
+      "learning_rate": 0.0003974362818326819,
+      "loss": 3.0243,
+      "step": 24162
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5856047868728638,
+      "learning_rate": 0.0003974217201879457,
+      "loss": 2.8006,
+      "step": 24163
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5460880398750305,
+      "learning_rate": 0.0003974071582866195,
+      "loss": 3.036,
+      "step": 24164
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5401052832603455,
+      "learning_rate": 0.00039739259612874163,
+      "loss": 3.0225,
+      "step": 24165
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6093460917472839,
+      "learning_rate": 0.0003973780337143504,
+      "loss": 3.0323,
+      "step": 24166
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5864385366439819,
+      "learning_rate": 0.0003973634710434841,
+      "loss": 3.0327,
+      "step": 24167
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5695944428443909,
+      "learning_rate": 0.00039734890811618116,
+      "loss": 3.1903,
+      "step": 24168
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5272200107574463,
+      "learning_rate": 0.00039733434493248,
+      "loss": 3.1964,
+      "step": 24169
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5584858655929565,
+      "learning_rate": 0.00039731978149241883,
+      "loss": 2.8314,
+      "step": 24170
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.542620837688446,
+      "learning_rate": 0.0003973052177960361,
+      "loss": 3.1823,
+      "step": 24171
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5688046813011169,
+      "learning_rate": 0.00039729065384337025,
+      "loss": 2.9456,
+      "step": 24172
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5506514310836792,
+      "learning_rate": 0.0003972760896344594,
+      "loss": 3.1059,
+      "step": 24173
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5505015254020691,
+      "learning_rate": 0.00039726152516934214,
+      "loss": 3.2437,
+      "step": 24174
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5867189764976501,
+      "learning_rate": 0.0003972469604480567,
+      "loss": 2.8984,
+      "step": 24175
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5614295601844788,
+      "learning_rate": 0.00039723239547064155,
+      "loss": 3.1634,
+      "step": 24176
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5461905002593994,
+      "learning_rate": 0.00039721783023713496,
+      "loss": 3.1337,
+      "step": 24177
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5388670563697815,
+      "learning_rate": 0.0003972032647475752,
+      "loss": 3.1324,
+      "step": 24178
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5529229640960693,
+      "learning_rate": 0.00039718869900200085,
+      "loss": 3.2731,
+      "step": 24179
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5644211173057556,
+      "learning_rate": 0.0003971741330004502,
+      "loss": 3.0189,
+      "step": 24180
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5382275581359863,
+      "learning_rate": 0.00039715956674296154,
+      "loss": 3.1831,
+      "step": 24181
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5577836036682129,
+      "learning_rate": 0.00039714500022957326,
+      "loss": 3.1967,
+      "step": 24182
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5363103151321411,
+      "learning_rate": 0.0003971304334603238,
+      "loss": 2.9104,
+      "step": 24183
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.538547694683075,
+      "learning_rate": 0.0003971158664352514,
+      "loss": 3.1109,
+      "step": 24184
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5949356555938721,
+      "learning_rate": 0.0003971012991543945,
+      "loss": 3.0122,
+      "step": 24185
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5695586204528809,
+      "learning_rate": 0.0003970867316177915,
+      "loss": 3.1775,
+      "step": 24186
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6203005313873291,
+      "learning_rate": 0.0003970721638254808,
+      "loss": 3.1296,
+      "step": 24187
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5122324228286743,
+      "learning_rate": 0.0003970575957775006,
+      "loss": 3.079,
+      "step": 24188
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5441964864730835,
+      "learning_rate": 0.0003970430274738894,
+      "loss": 3.109,
+      "step": 24189
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5421319007873535,
+      "learning_rate": 0.00039702845891468556,
+      "loss": 3.1183,
+      "step": 24190
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5920052528381348,
+      "learning_rate": 0.0003970138900999275,
+      "loss": 3.0434,
+      "step": 24191
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5872803926467896,
+      "learning_rate": 0.0003969993210296535,
+      "loss": 2.8057,
+      "step": 24192
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.557303786277771,
+      "learning_rate": 0.0003969847517039018,
+      "loss": 3.0588,
+      "step": 24193
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5577976107597351,
+      "learning_rate": 0.0003969701821227111,
+      "loss": 3.2499,
+      "step": 24194
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5627845525741577,
+      "learning_rate": 0.00039695561228611953,
+      "loss": 3.059,
+      "step": 24195
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.607268750667572,
+      "learning_rate": 0.0003969410421941655,
+      "loss": 3.0055,
+      "step": 24196
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5258342027664185,
+      "learning_rate": 0.0003969264718468875,
+      "loss": 2.8618,
+      "step": 24197
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5296414494514465,
+      "learning_rate": 0.0003969119012443238,
+      "loss": 3.2552,
+      "step": 24198
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5506784915924072,
+      "learning_rate": 0.00039689733038651275,
+      "loss": 2.9617,
+      "step": 24199
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5685448050498962,
+      "learning_rate": 0.00039688275927349286,
+      "loss": 3.12,
+      "step": 24200
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5335084199905396,
+      "learning_rate": 0.0003968681879053024,
+      "loss": 3.0428,
+      "step": 24201
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6116160750389099,
+      "learning_rate": 0.0003968536162819798,
+      "loss": 2.9557,
+      "step": 24202
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5441951751708984,
+      "learning_rate": 0.00039683904440356336,
+      "loss": 3.272,
+      "step": 24203
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5496557950973511,
+      "learning_rate": 0.00039682447227009146,
+      "loss": 3.1211,
+      "step": 24204
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6027690768241882,
+      "learning_rate": 0.00039680989988160275,
+      "loss": 3.2041,
+      "step": 24205
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7103535532951355,
+      "learning_rate": 0.00039679532723813516,
+      "loss": 3.384,
+      "step": 24206
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.570144772529602,
+      "learning_rate": 0.00039678075433972746,
+      "loss": 3.0117,
+      "step": 24207
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5552254915237427,
+      "learning_rate": 0.0003967661811864178,
+      "loss": 3.0736,
+      "step": 24208
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5385003685951233,
+      "learning_rate": 0.00039675160777824465,
+      "loss": 3.0785,
+      "step": 24209
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5606467127799988,
+      "learning_rate": 0.00039673703411524647,
+      "loss": 3.0978,
+      "step": 24210
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5405035614967346,
+      "learning_rate": 0.0003967224601974615,
+      "loss": 3.1803,
+      "step": 24211
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5441907048225403,
+      "learning_rate": 0.0003967078860249281,
+      "loss": 3.0588,
+      "step": 24212
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5538133382797241,
+      "learning_rate": 0.0003966933115976849,
+      "loss": 3.1926,
+      "step": 24213
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.528517484664917,
+      "learning_rate": 0.00039667873691576997,
+      "loss": 3.4385,
+      "step": 24214
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5590735077857971,
+      "learning_rate": 0.0003966641619792219,
+      "loss": 3.067,
+      "step": 24215
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.591706395149231,
+      "learning_rate": 0.0003966495867880791,
+      "loss": 3.0503,
+      "step": 24216
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5574644207954407,
+      "learning_rate": 0.00039663501134237983,
+      "loss": 3.0304,
+      "step": 24217
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5569961667060852,
+      "learning_rate": 0.00039662043564216257,
+      "loss": 2.9025,
+      "step": 24218
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5450246930122375,
+      "learning_rate": 0.00039660585968746554,
+      "loss": 3.1877,
+      "step": 24219
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.562933623790741,
+      "learning_rate": 0.00039659128347832747,
+      "loss": 3.0375,
+      "step": 24220
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5538458824157715,
+      "learning_rate": 0.0003965767070147865,
+      "loss": 3.0816,
+      "step": 24221
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.60169917345047,
+      "learning_rate": 0.000396562130296881,
+      "loss": 3.1415,
+      "step": 24222
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5521277785301208,
+      "learning_rate": 0.00039654755332464947,
+      "loss": 3.0604,
+      "step": 24223
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5698322653770447,
+      "learning_rate": 0.0003965329760981303,
+      "loss": 3.283,
+      "step": 24224
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5327075123786926,
+      "learning_rate": 0.0003965183986173618,
+      "loss": 3.0547,
+      "step": 24225
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5273543000221252,
+      "learning_rate": 0.0003965038208823824,
+      "loss": 3.2407,
+      "step": 24226
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7036231756210327,
+      "learning_rate": 0.00039648924289323057,
+      "loss": 3.1334,
+      "step": 24227
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5902255773544312,
+      "learning_rate": 0.00039647466464994463,
+      "loss": 2.9559,
+      "step": 24228
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5623027086257935,
+      "learning_rate": 0.0003964600861525629,
+      "loss": 2.899,
+      "step": 24229
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5198655128479004,
+      "learning_rate": 0.00039644550740112393,
+      "loss": 3.029,
+      "step": 24230
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5730830430984497,
+      "learning_rate": 0.00039643092839566614,
+      "loss": 2.9395,
+      "step": 24231
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5397255420684814,
+      "learning_rate": 0.00039641634913622776,
+      "loss": 3.1824,
+      "step": 24232
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5603347420692444,
+      "learning_rate": 0.0003964017696228473,
+      "loss": 3.0076,
+      "step": 24233
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.556710422039032,
+      "learning_rate": 0.00039638718985556307,
+      "loss": 3.1231,
+      "step": 24234
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5215965509414673,
+      "learning_rate": 0.0003963726098344136,
+      "loss": 3.0744,
+      "step": 24235
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5522467494010925,
+      "learning_rate": 0.0003963580295594372,
+      "loss": 3.2905,
+      "step": 24236
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5642509460449219,
+      "learning_rate": 0.0003963434490306722,
+      "loss": 3.1552,
+      "step": 24237
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5451124310493469,
+      "learning_rate": 0.0003963288682481573,
+      "loss": 3.0713,
+      "step": 24238
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5440728664398193,
+      "learning_rate": 0.00039631428721193055,
+      "loss": 3.1808,
+      "step": 24239
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5344854593276978,
+      "learning_rate": 0.00039629970592203043,
+      "loss": 3.118,
+      "step": 24240
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5784305930137634,
+      "learning_rate": 0.00039628512437849553,
+      "loss": 3.2412,
+      "step": 24241
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5519080758094788,
+      "learning_rate": 0.00039627054258136417,
+      "loss": 3.3491,
+      "step": 24242
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5608418583869934,
+      "learning_rate": 0.0003962559605306747,
+      "loss": 3.0723,
+      "step": 24243
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5935826897621155,
+      "learning_rate": 0.0003962413782264656,
+      "loss": 3.1175,
+      "step": 24244
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5591660141944885,
+      "learning_rate": 0.0003962267956687752,
+      "loss": 3.0419,
+      "step": 24245
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6037750244140625,
+      "learning_rate": 0.0003962122128576419,
+      "loss": 3.0272,
+      "step": 24246
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5254233479499817,
+      "learning_rate": 0.00039619762979310416,
+      "loss": 3.1549,
+      "step": 24247
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5338966846466064,
+      "learning_rate": 0.0003961830464752004,
+      "loss": 3.102,
+      "step": 24248
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5736029744148254,
+      "learning_rate": 0.000396168462903969,
+      "loss": 3.2333,
+      "step": 24249
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5698594450950623,
+      "learning_rate": 0.00039615387907944834,
+      "loss": 3.2735,
+      "step": 24250
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5339264869689941,
+      "learning_rate": 0.0003961392950016769,
+      "loss": 2.9759,
+      "step": 24251
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5492169857025146,
+      "learning_rate": 0.000396124710670693,
+      "loss": 3.0123,
+      "step": 24252
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5674329996109009,
+      "learning_rate": 0.00039611012608653523,
+      "loss": 3.2794,
+      "step": 24253
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5592032670974731,
+      "learning_rate": 0.00039609554124924183,
+      "loss": 3.1204,
+      "step": 24254
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5712992548942566,
+      "learning_rate": 0.0003960809561588512,
+      "loss": 2.8761,
+      "step": 24255
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5755642652511597,
+      "learning_rate": 0.0003960663708154019,
+      "loss": 2.9666,
+      "step": 24256
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5444374084472656,
+      "learning_rate": 0.0003960517852189323,
+      "loss": 3.0622,
+      "step": 24257
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5554080605506897,
+      "learning_rate": 0.00039603719936948074,
+      "loss": 3.2019,
+      "step": 24258
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6017959713935852,
+      "learning_rate": 0.0003960226132670857,
+      "loss": 2.9427,
+      "step": 24259
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5440652966499329,
+      "learning_rate": 0.00039600802691178556,
+      "loss": 3.1756,
+      "step": 24260
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5530151128768921,
+      "learning_rate": 0.0003959934403036187,
+      "loss": 2.9358,
+      "step": 24261
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5590417385101318,
+      "learning_rate": 0.00039597885344262366,
+      "loss": 3.0388,
+      "step": 24262
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5701658725738525,
+      "learning_rate": 0.0003959642663288387,
+      "loss": 3.0352,
+      "step": 24263
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5556230545043945,
+      "learning_rate": 0.0003959496789623025,
+      "loss": 3.1208,
+      "step": 24264
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5601057410240173,
+      "learning_rate": 0.00039593509134305317,
+      "loss": 3.0421,
+      "step": 24265
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5562484860420227,
+      "learning_rate": 0.00039592050347112933,
+      "loss": 3.1872,
+      "step": 24266
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5837891101837158,
+      "learning_rate": 0.00039590591534656937,
+      "loss": 3.0225,
+      "step": 24267
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5743361115455627,
+      "learning_rate": 0.0003958913269694116,
+      "loss": 3.1041,
+      "step": 24268
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6154342889785767,
+      "learning_rate": 0.00039587673833969464,
+      "loss": 2.9143,
+      "step": 24269
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5403562188148499,
+      "learning_rate": 0.00039586214945745664,
+      "loss": 2.9925,
+      "step": 24270
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5781658291816711,
+      "learning_rate": 0.00039584756032273635,
+      "loss": 3.1171,
+      "step": 24271
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5835286378860474,
+      "learning_rate": 0.00039583297093557195,
+      "loss": 3.1155,
+      "step": 24272
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5412124395370483,
+      "learning_rate": 0.00039581838129600197,
+      "loss": 3.056,
+      "step": 24273
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5261394381523132,
+      "learning_rate": 0.00039580379140406475,
+      "loss": 3.0973,
+      "step": 24274
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5671351552009583,
+      "learning_rate": 0.00039578920125979893,
+      "loss": 3.0282,
+      "step": 24275
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5514994263648987,
+      "learning_rate": 0.0003957746108632427,
+      "loss": 3.1963,
+      "step": 24276
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.576802134513855,
+      "learning_rate": 0.00039576002021443456,
+      "loss": 3.0313,
+      "step": 24277
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5313069224357605,
+      "learning_rate": 0.00039574542931341297,
+      "loss": 3.2151,
+      "step": 24278
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.536454439163208,
+      "learning_rate": 0.00039573083816021637,
+      "loss": 3.1187,
+      "step": 24279
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5508918762207031,
+      "learning_rate": 0.00039571624675488313,
+      "loss": 2.9256,
+      "step": 24280
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5629120469093323,
+      "learning_rate": 0.0003957016550974517,
+      "loss": 3.0875,
+      "step": 24281
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5517382025718689,
+      "learning_rate": 0.0003956870631879606,
+      "loss": 2.9645,
+      "step": 24282
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5510847568511963,
+      "learning_rate": 0.00039567247102644817,
+      "loss": 3.0028,
+      "step": 24283
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5683895945549011,
+      "learning_rate": 0.0003956578786129528,
+      "loss": 3.2715,
+      "step": 24284
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5647633671760559,
+      "learning_rate": 0.00039564328594751306,
+      "loss": 2.9625,
+      "step": 24285
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5621564388275146,
+      "learning_rate": 0.00039562869303016724,
+      "loss": 3.2011,
+      "step": 24286
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5985468029975891,
+      "learning_rate": 0.0003956140998609539,
+      "loss": 3.3408,
+      "step": 24287
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5399235486984253,
+      "learning_rate": 0.00039559950643991144,
+      "loss": 3.1029,
+      "step": 24288
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5459802150726318,
+      "learning_rate": 0.0003955849127670783,
+      "loss": 2.9857,
+      "step": 24289
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5875890254974365,
+      "learning_rate": 0.00039557031884249286,
+      "loss": 2.938,
+      "step": 24290
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5480020642280579,
+      "learning_rate": 0.0003955557246661935,
+      "loss": 3.3471,
+      "step": 24291
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5252436995506287,
+      "learning_rate": 0.0003955411302382189,
+      "loss": 3.2947,
+      "step": 24292
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5304093956947327,
+      "learning_rate": 0.0003955265355586073,
+      "loss": 3.3092,
+      "step": 24293
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5480414628982544,
+      "learning_rate": 0.00039551194062739713,
+      "loss": 3.0931,
+      "step": 24294
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5469361543655396,
+      "learning_rate": 0.0003954973454446269,
+      "loss": 3.1805,
+      "step": 24295
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5798367857933044,
+      "learning_rate": 0.0003954827500103351,
+      "loss": 3.0148,
+      "step": 24296
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5936307907104492,
+      "learning_rate": 0.0003954681543245601,
+      "loss": 2.8564,
+      "step": 24297
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5227268934249878,
+      "learning_rate": 0.0003954535583873403,
+      "loss": 3.0408,
+      "step": 24298
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5441100001335144,
+      "learning_rate": 0.0003954389621987143,
+      "loss": 3.0124,
+      "step": 24299
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5319622159004211,
+      "learning_rate": 0.0003954243657587203,
+      "loss": 2.8448,
+      "step": 24300
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.640273928642273,
+      "learning_rate": 0.000395409769067397,
+      "loss": 3.1077,
+      "step": 24301
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5414076447486877,
+      "learning_rate": 0.0003953951721247826,
+      "loss": 3.1144,
+      "step": 24302
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5242058038711548,
+      "learning_rate": 0.00039538057493091584,
+      "loss": 2.9945,
+      "step": 24303
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5769972205162048,
+      "learning_rate": 0.00039536597748583496,
+      "loss": 2.9874,
+      "step": 24304
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5243396759033203,
+      "learning_rate": 0.0003953513797895784,
+      "loss": 3.0878,
+      "step": 24305
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5818905234336853,
+      "learning_rate": 0.0003953367818421847,
+      "loss": 3.1227,
+      "step": 24306
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5517255067825317,
+      "learning_rate": 0.00039532218364369225,
+      "loss": 3.006,
+      "step": 24307
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5664587020874023,
+      "learning_rate": 0.0003953075851941395,
+      "loss": 3.0768,
+      "step": 24308
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.547131359577179,
+      "learning_rate": 0.0003952929864935649,
+      "loss": 3.054,
+      "step": 24309
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5453808307647705,
+      "learning_rate": 0.00039527838754200694,
+      "loss": 3.1298,
+      "step": 24310
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.549038827419281,
+      "learning_rate": 0.00039526378833950406,
+      "loss": 2.9323,
+      "step": 24311
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5720995664596558,
+      "learning_rate": 0.00039524918888609467,
+      "loss": 3.0093,
+      "step": 24312
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5207635164260864,
+      "learning_rate": 0.0003952345891818172,
+      "loss": 2.853,
+      "step": 24313
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5525292754173279,
+      "learning_rate": 0.0003952199892267102,
+      "loss": 3.2314,
+      "step": 24314
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5358814001083374,
+      "learning_rate": 0.0003952053890208121,
+      "loss": 2.8513,
+      "step": 24315
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5728432536125183,
+      "learning_rate": 0.00039519078856416136,
+      "loss": 3.1612,
+      "step": 24316
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5785890817642212,
+      "learning_rate": 0.0003951761878567962,
+      "loss": 3.1774,
+      "step": 24317
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6545047760009766,
+      "learning_rate": 0.00039516158689875555,
+      "loss": 2.9229,
+      "step": 24318
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5325422883033752,
+      "learning_rate": 0.0003951469856900775,
+      "loss": 2.9676,
+      "step": 24319
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5706045627593994,
+      "learning_rate": 0.0003951323842308005,
+      "loss": 3.0287,
+      "step": 24320
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5675446391105652,
+      "learning_rate": 0.00039511778252096323,
+      "loss": 3.1198,
+      "step": 24321
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5522854924201965,
+      "learning_rate": 0.00039510318056060396,
+      "loss": 2.9583,
+      "step": 24322
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5343487858772278,
+      "learning_rate": 0.00039508857834976124,
+      "loss": 3.3867,
+      "step": 24323
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5691496133804321,
+      "learning_rate": 0.00039507397588847346,
+      "loss": 3.257,
+      "step": 24324
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5750447511672974,
+      "learning_rate": 0.00039505937317677915,
+      "loss": 2.8305,
+      "step": 24325
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6205912828445435,
+      "learning_rate": 0.0003950447702147168,
+      "loss": 2.6935,
+      "step": 24326
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5219376087188721,
+      "learning_rate": 0.0003950301670023247,
+      "loss": 3.142,
+      "step": 24327
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5368987917900085,
+      "learning_rate": 0.00039501556353964146,
+      "loss": 3.1002,
+      "step": 24328
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5906136631965637,
+      "learning_rate": 0.0003950009598267057,
+      "loss": 2.9932,
+      "step": 24329
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5572717785835266,
+      "learning_rate": 0.0003949863558635555,
+      "loss": 3.0393,
+      "step": 24330
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5366165041923523,
+      "learning_rate": 0.00039497175165022956,
+      "loss": 2.838,
+      "step": 24331
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5345458388328552,
+      "learning_rate": 0.0003949571471867663,
+      "loss": 2.8754,
+      "step": 24332
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5713475942611694,
+      "learning_rate": 0.00039494254247320423,
+      "loss": 2.9894,
+      "step": 24333
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5291685461997986,
+      "learning_rate": 0.00039492793750958173,
+      "loss": 3.1748,
+      "step": 24334
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5592206120491028,
+      "learning_rate": 0.0003949133322959373,
+      "loss": 3.2341,
+      "step": 24335
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5599371194839478,
+      "learning_rate": 0.0003948987268323094,
+      "loss": 3.2312,
+      "step": 24336
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5585187077522278,
+      "learning_rate": 0.0003948841211187366,
+      "loss": 3.0695,
+      "step": 24337
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5566004514694214,
+      "learning_rate": 0.0003948695151552572,
+      "loss": 3.202,
+      "step": 24338
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5448996424674988,
+      "learning_rate": 0.0003948549089419098,
+      "loss": 3.0852,
+      "step": 24339
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5520145893096924,
+      "learning_rate": 0.0003948403024787329,
+      "loss": 3.1257,
+      "step": 24340
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5336360931396484,
+      "learning_rate": 0.00039482569576576484,
+      "loss": 3.1598,
+      "step": 24341
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5652227997779846,
+      "learning_rate": 0.00039481108880304403,
+      "loss": 3.0888,
+      "step": 24342
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5527975559234619,
+      "learning_rate": 0.0003947964815906091,
+      "loss": 3.1309,
+      "step": 24343
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5577079057693481,
+      "learning_rate": 0.0003947818741284987,
+      "loss": 3.1479,
+      "step": 24344
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5621340274810791,
+      "learning_rate": 0.0003947672664167509,
+      "loss": 3.0016,
+      "step": 24345
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5487366914749146,
+      "learning_rate": 0.0003947526584554044,
+      "loss": 3.0614,
+      "step": 24346
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5261529088020325,
+      "learning_rate": 0.0003947380502444976,
+      "loss": 3.2059,
+      "step": 24347
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5683314204216003,
+      "learning_rate": 0.000394723441784069,
+      "loss": 3.2642,
+      "step": 24348
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5974188446998596,
+      "learning_rate": 0.0003947088330741571,
+      "loss": 3.0305,
+      "step": 24349
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5211678147315979,
+      "learning_rate": 0.0003946942241148004,
+      "loss": 3.1625,
+      "step": 24350
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.553632915019989,
+      "learning_rate": 0.0003946796149060373,
+      "loss": 3.1093,
+      "step": 24351
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5715633034706116,
+      "learning_rate": 0.00039466500544790633,
+      "loss": 3.187,
+      "step": 24352
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5198462009429932,
+      "learning_rate": 0.00039465039574044597,
+      "loss": 3.118,
+      "step": 24353
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5220559239387512,
+      "learning_rate": 0.00039463578578369463,
+      "loss": 3.1978,
+      "step": 24354
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5109367370605469,
+      "learning_rate": 0.00039462117557769095,
+      "loss": 3.1694,
+      "step": 24355
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5752742886543274,
+      "learning_rate": 0.0003946065651224732,
+      "loss": 3.0348,
+      "step": 24356
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5721966028213501,
+      "learning_rate": 0.00039459195441807996,
+      "loss": 3.1092,
+      "step": 24357
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5311843156814575,
+      "learning_rate": 0.0003945773434645498,
+      "loss": 3.1743,
+      "step": 24358
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5497158765792847,
+      "learning_rate": 0.00039456273226192107,
+      "loss": 2.7434,
+      "step": 24359
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5891440510749817,
+      "learning_rate": 0.0003945481208102323,
+      "loss": 3.17,
+      "step": 24360
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5738333463668823,
+      "learning_rate": 0.0003945335091095219,
+      "loss": 3.0605,
+      "step": 24361
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5178869366645813,
+      "learning_rate": 0.0003945188971598285,
+      "loss": 3.113,
+      "step": 24362
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5586382746696472,
+      "learning_rate": 0.00039450428496119055,
+      "loss": 3.1943,
+      "step": 24363
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5722424983978271,
+      "learning_rate": 0.0003944896725136464,
+      "loss": 2.919,
+      "step": 24364
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.547747015953064,
+      "learning_rate": 0.0003944750598172348,
+      "loss": 3.0477,
+      "step": 24365
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5872486233711243,
+      "learning_rate": 0.000394460446871994,
+      "loss": 3.2328,
+      "step": 24366
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5272241234779358,
+      "learning_rate": 0.0003944458336779626,
+      "loss": 3.1448,
+      "step": 24367
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5975600481033325,
+      "learning_rate": 0.00039443122023517894,
+      "loss": 3.0531,
+      "step": 24368
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5917435884475708,
+      "learning_rate": 0.0003944166065436817,
+      "loss": 3.1291,
+      "step": 24369
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5495269894599915,
+      "learning_rate": 0.00039440199260350927,
+      "loss": 3.056,
+      "step": 24370
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5593695640563965,
+      "learning_rate": 0.00039438737841470013,
+      "loss": 2.9955,
+      "step": 24371
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5663504600524902,
+      "learning_rate": 0.0003943727639772928,
+      "loss": 3.0223,
+      "step": 24372
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5404443144798279,
+      "learning_rate": 0.00039435814929132586,
+      "loss": 3.2137,
+      "step": 24373
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.52837735414505,
+      "learning_rate": 0.00039434353435683766,
+      "loss": 3.2715,
+      "step": 24374
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5688294768333435,
+      "learning_rate": 0.0003943289191738667,
+      "loss": 3.1562,
+      "step": 24375
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5669829249382019,
+      "learning_rate": 0.00039431430374245154,
+      "loss": 3.2254,
+      "step": 24376
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5809857249259949,
+      "learning_rate": 0.00039429968806263077,
+      "loss": 3.1399,
+      "step": 24377
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5638022422790527,
+      "learning_rate": 0.00039428507213444254,
+      "loss": 3.0601,
+      "step": 24378
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5520922541618347,
+      "learning_rate": 0.0003942704559579257,
+      "loss": 3.0555,
+      "step": 24379
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5883176922798157,
+      "learning_rate": 0.0003942558395331188,
+      "loss": 3.0359,
+      "step": 24380
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5609521269798279,
+      "learning_rate": 0.0003942412228600599,
+      "loss": 3.101,
+      "step": 24381
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5642240047454834,
+      "learning_rate": 0.0003942266059387879,
+      "loss": 2.9432,
+      "step": 24382
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5660053491592407,
+      "learning_rate": 0.000394211988769341,
+      "loss": 3.0687,
+      "step": 24383
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5571756362915039,
+      "learning_rate": 0.00039419737135175804,
+      "loss": 2.9455,
+      "step": 24384
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6327832937240601,
+      "learning_rate": 0.0003941827536860773,
+      "loss": 3.3466,
+      "step": 24385
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5416423082351685,
+      "learning_rate": 0.0003941681357723372,
+      "loss": 3.1423,
+      "step": 24386
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.56650310754776,
+      "learning_rate": 0.0003941535176105764,
+      "loss": 3.2203,
+      "step": 24387
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.516886830329895,
+      "learning_rate": 0.00039413889920083335,
+      "loss": 2.9785,
+      "step": 24388
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5625539422035217,
+      "learning_rate": 0.0003941242805431466,
+      "loss": 3.0304,
+      "step": 24389
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5716773867607117,
+      "learning_rate": 0.0003941096616375546,
+      "loss": 3.2899,
+      "step": 24390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.580208420753479,
+      "learning_rate": 0.00039409504248409595,
+      "loss": 3.1046,
+      "step": 24391
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5867831707000732,
+      "learning_rate": 0.0003940804230828089,
+      "loss": 3.0618,
+      "step": 24392
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5907902717590332,
+      "learning_rate": 0.00039406580343373217,
+      "loss": 3.3336,
+      "step": 24393
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5453541278839111,
+      "learning_rate": 0.0003940511835369041,
+      "loss": 2.8692,
+      "step": 24394
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5520537495613098,
+      "learning_rate": 0.0003940365633923636,
+      "loss": 3.0061,
+      "step": 24395
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5857133865356445,
+      "learning_rate": 0.0003940219430001487,
+      "loss": 3.0555,
+      "step": 24396
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5415214896202087,
+      "learning_rate": 0.00039400732236029816,
+      "loss": 3.0087,
+      "step": 24397
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5039085745811462,
+      "learning_rate": 0.0003939927014728504,
+      "loss": 3.37,
+      "step": 24398
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5987560153007507,
+      "learning_rate": 0.000393978080337844,
+      "loss": 3.3081,
+      "step": 24399
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5762282013893127,
+      "learning_rate": 0.0003939634589553173,
+      "loss": 3.048,
+      "step": 24400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5720338225364685,
+      "learning_rate": 0.000393948837325309,
+      "loss": 3.0585,
+      "step": 24401
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5712810754776001,
+      "learning_rate": 0.00039393421544785766,
+      "loss": 3.0319,
+      "step": 24402
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5783328413963318,
+      "learning_rate": 0.0003939195933230015,
+      "loss": 3.0169,
+      "step": 24403
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5550515055656433,
+      "learning_rate": 0.00039390497095077924,
+      "loss": 3.2278,
+      "step": 24404
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5710436105728149,
+      "learning_rate": 0.0003938903483312294,
+      "loss": 3.0629,
+      "step": 24405
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.2239809036254883,
+      "learning_rate": 0.00039387572546439046,
+      "loss": 3.3184,
+      "step": 24406
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5772742033004761,
+      "learning_rate": 0.00039386110235030094,
+      "loss": 3.2588,
+      "step": 24407
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.576033353805542,
+      "learning_rate": 0.00039384647898899934,
+      "loss": 2.8866,
+      "step": 24408
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5748345851898193,
+      "learning_rate": 0.0003938318553805241,
+      "loss": 3.0798,
+      "step": 24409
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5759305953979492,
+      "learning_rate": 0.00039381723152491385,
+      "loss": 3.0736,
+      "step": 24410
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5629940032958984,
+      "learning_rate": 0.0003938026074222071,
+      "loss": 3.3334,
+      "step": 24411
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5594444274902344,
+      "learning_rate": 0.00039378798307244234,
+      "loss": 2.9158,
+      "step": 24412
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5800905823707581,
+      "learning_rate": 0.000393773358475658,
+      "loss": 2.9424,
+      "step": 24413
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5819612741470337,
+      "learning_rate": 0.0003937587336318927,
+      "loss": 3.0931,
+      "step": 24414
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5524073243141174,
+      "learning_rate": 0.00039374410854118495,
+      "loss": 3.1325,
+      "step": 24415
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5730950832366943,
+      "learning_rate": 0.0003937294832035733,
+      "loss": 2.9164,
+      "step": 24416
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5506508350372314,
+      "learning_rate": 0.00039371485761909627,
+      "loss": 3.0187,
+      "step": 24417
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5478151440620422,
+      "learning_rate": 0.00039370023178779233,
+      "loss": 2.788,
+      "step": 24418
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5452743768692017,
+      "learning_rate": 0.00039368560570969985,
+      "loss": 2.9242,
+      "step": 24419
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.49868834018707275,
+      "learning_rate": 0.0003936709793848577,
+      "loss": 2.9301,
+      "step": 24420
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.597121000289917,
+      "learning_rate": 0.00039365635281330417,
+      "loss": 2.7868,
+      "step": 24421
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5634990334510803,
+      "learning_rate": 0.00039364172599507777,
+      "loss": 2.8906,
+      "step": 24422
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5244052410125732,
+      "learning_rate": 0.0003936270989302171,
+      "loss": 3.1885,
+      "step": 24423
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5397709012031555,
+      "learning_rate": 0.0003936124716187607,
+      "loss": 3.1282,
+      "step": 24424
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5438342690467834,
+      "learning_rate": 0.00039359784406074706,
+      "loss": 3.1376,
+      "step": 24425
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5476157665252686,
+      "learning_rate": 0.00039358321625621473,
+      "loss": 3.1603,
+      "step": 24426
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5249218940734863,
+      "learning_rate": 0.00039356858820520217,
+      "loss": 3.1193,
+      "step": 24427
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.548941969871521,
+      "learning_rate": 0.0003935539599077481,
+      "loss": 3.2686,
+      "step": 24428
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5758370757102966,
+      "learning_rate": 0.0003935393313638908,
+      "loss": 3.1205,
+      "step": 24429
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5937607884407043,
+      "learning_rate": 0.0003935247025736688,
+      "loss": 2.9228,
+      "step": 24430
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5786687731742859,
+      "learning_rate": 0.00039351007353712086,
+      "loss": 2.9846,
+      "step": 24431
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5394309163093567,
+      "learning_rate": 0.0003934954442542854,
+      "loss": 3.2228,
+      "step": 24432
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5585785508155823,
+      "learning_rate": 0.00039348081472520086,
+      "loss": 3.1961,
+      "step": 24433
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5906318426132202,
+      "learning_rate": 0.00039346618494990584,
+      "loss": 3.2208,
+      "step": 24434
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5899059176445007,
+      "learning_rate": 0.00039345155492843893,
+      "loss": 3.0491,
+      "step": 24435
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5428707003593445,
+      "learning_rate": 0.00039343692466083856,
+      "loss": 2.9778,
+      "step": 24436
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5404731631278992,
+      "learning_rate": 0.0003934222941471433,
+      "loss": 2.9239,
+      "step": 24437
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5213099718093872,
+      "learning_rate": 0.00039340766338739175,
+      "loss": 3.2136,
+      "step": 24438
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5822370052337646,
+      "learning_rate": 0.00039339303238162247,
+      "loss": 3.0366,
+      "step": 24439
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.581151008605957,
+      "learning_rate": 0.00039337840112987373,
+      "loss": 3.1906,
+      "step": 24440
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5343998670578003,
+      "learning_rate": 0.00039336376963218435,
+      "loss": 3.202,
+      "step": 24441
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5333302617073059,
+      "learning_rate": 0.00039334913788859283,
+      "loss": 3.2351,
+      "step": 24442
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5405131578445435,
+      "learning_rate": 0.00039333450589913754,
+      "loss": 3.0949,
+      "step": 24443
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5513492822647095,
+      "learning_rate": 0.0003933198736638572,
+      "loss": 3.023,
+      "step": 24444
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.587897539138794,
+      "learning_rate": 0.00039330524118279015,
+      "loss": 3.1852,
+      "step": 24445
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5661784410476685,
+      "learning_rate": 0.00039329060845597524,
+      "loss": 2.9803,
+      "step": 24446
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5629352927207947,
+      "learning_rate": 0.0003932759754834507,
+      "loss": 3.0325,
+      "step": 24447
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5622941851615906,
+      "learning_rate": 0.00039326134226525515,
+      "loss": 2.8962,
+      "step": 24448
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5289006233215332,
+      "learning_rate": 0.00039324670880142726,
+      "loss": 3.1545,
+      "step": 24449
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5608207583427429,
+      "learning_rate": 0.00039323207509200545,
+      "loss": 3.1841,
+      "step": 24450
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5729897618293762,
+      "learning_rate": 0.0003932174411370283,
+      "loss": 3.1255,
+      "step": 24451
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5416507720947266,
+      "learning_rate": 0.00039320280693653435,
+      "loss": 3.0346,
+      "step": 24452
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5522789359092712,
+      "learning_rate": 0.00039318817249056224,
+      "loss": 3.1717,
+      "step": 24453
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6044777631759644,
+      "learning_rate": 0.00039317353779915034,
+      "loss": 3.0349,
+      "step": 24454
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5705912709236145,
+      "learning_rate": 0.0003931589028623371,
+      "loss": 3.0584,
+      "step": 24455
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5330730676651001,
+      "learning_rate": 0.0003931442676801614,
+      "loss": 3.2936,
+      "step": 24456
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.526887059211731,
+      "learning_rate": 0.0003931296322526617,
+      "loss": 2.9784,
+      "step": 24457
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5354217886924744,
+      "learning_rate": 0.0003931149965798764,
+      "loss": 3.1643,
+      "step": 24458
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5469911098480225,
+      "learning_rate": 0.0003931003606618441,
+      "loss": 3.0493,
+      "step": 24459
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5541418790817261,
+      "learning_rate": 0.00039308572449860336,
+      "loss": 2.8854,
+      "step": 24460
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.568221390247345,
+      "learning_rate": 0.0003930710880901928,
+      "loss": 2.958,
+      "step": 24461
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5707933902740479,
+      "learning_rate": 0.0003930564514366509,
+      "loss": 3.1631,
+      "step": 24462
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5753080248832703,
+      "learning_rate": 0.00039304181453801614,
+      "loss": 3.1559,
+      "step": 24463
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5365299582481384,
+      "learning_rate": 0.0003930271773943272,
+      "loss": 3.1203,
+      "step": 24464
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5476408004760742,
+      "learning_rate": 0.00039301254000562256,
+      "loss": 3.0281,
+      "step": 24465
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5769860744476318,
+      "learning_rate": 0.0003929979023719408,
+      "loss": 3.121,
+      "step": 24466
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6829390525817871,
+      "learning_rate": 0.00039298326449332044,
+      "loss": 2.9079,
+      "step": 24467
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5319774746894836,
+      "learning_rate": 0.00039296862636980015,
+      "loss": 2.9785,
+      "step": 24468
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5541231632232666,
+      "learning_rate": 0.0003929539880014183,
+      "loss": 3.0751,
+      "step": 24469
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5718669295310974,
+      "learning_rate": 0.00039293934938821354,
+      "loss": 2.9471,
+      "step": 24470
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5447272658348083,
+      "learning_rate": 0.0003929247105302244,
+      "loss": 2.8957,
+      "step": 24471
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5534474849700928,
+      "learning_rate": 0.0003929100714274895,
+      "loss": 3.0906,
+      "step": 24472
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9372397661209106,
+      "learning_rate": 0.00039289543208004734,
+      "loss": 3.2436,
+      "step": 24473
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5349345207214355,
+      "learning_rate": 0.00039288079248793646,
+      "loss": 3.0896,
+      "step": 24474
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5702376365661621,
+      "learning_rate": 0.0003928661526511955,
+      "loss": 2.9394,
+      "step": 24475
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5644256472587585,
+      "learning_rate": 0.00039285151256986296,
+      "loss": 3.2093,
+      "step": 24476
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5325481295585632,
+      "learning_rate": 0.00039283687224397744,
+      "loss": 2.844,
+      "step": 24477
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5378096103668213,
+      "learning_rate": 0.0003928222316735773,
+      "loss": 3.0989,
+      "step": 24478
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5697956085205078,
+      "learning_rate": 0.0003928075908587015,
+      "loss": 3.1788,
+      "step": 24479
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5447378754615784,
+      "learning_rate": 0.0003927929497993882,
+      "loss": 3.0596,
+      "step": 24480
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6714164614677429,
+      "learning_rate": 0.00039277830849567615,
+      "loss": 2.8876,
+      "step": 24481
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5856902003288269,
+      "learning_rate": 0.000392763666947604,
+      "loss": 2.8871,
+      "step": 24482
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.58744215965271,
+      "learning_rate": 0.0003927490251552101,
+      "loss": 3.1451,
+      "step": 24483
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.563101053237915,
+      "learning_rate": 0.00039273438311853315,
+      "loss": 3.0435,
+      "step": 24484
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5481135845184326,
+      "learning_rate": 0.00039271974083761167,
+      "loss": 3.0859,
+      "step": 24485
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5843150019645691,
+      "learning_rate": 0.0003927050983124842,
+      "loss": 3.0686,
+      "step": 24486
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5351256728172302,
+      "learning_rate": 0.0003926904555431894,
+      "loss": 3.0086,
+      "step": 24487
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6021181344985962,
+      "learning_rate": 0.00039267581252976574,
+      "loss": 2.9542,
+      "step": 24488
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5292304754257202,
+      "learning_rate": 0.00039266116927225186,
+      "loss": 3.0605,
+      "step": 24489
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5263028144836426,
+      "learning_rate": 0.00039264652577068634,
+      "loss": 2.9425,
+      "step": 24490
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5306416749954224,
+      "learning_rate": 0.0003926318820251076,
+      "loss": 3.0063,
+      "step": 24491
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5751937031745911,
+      "learning_rate": 0.00039261723803555427,
+      "loss": 3.029,
+      "step": 24492
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5904483795166016,
+      "learning_rate": 0.00039260259380206517,
+      "loss": 3.3765,
+      "step": 24493
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5914291739463806,
+      "learning_rate": 0.00039258794932467845,
+      "loss": 2.9694,
+      "step": 24494
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5873374938964844,
+      "learning_rate": 0.000392573304603433,
+      "loss": 3.0701,
+      "step": 24495
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6778043508529663,
+      "learning_rate": 0.00039255865963836714,
+      "loss": 3.0645,
+      "step": 24496
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5494279265403748,
+      "learning_rate": 0.0003925440144295198,
+      "loss": 3.1383,
+      "step": 24497
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5600928068161011,
+      "learning_rate": 0.0003925293689769292,
+      "loss": 3.2732,
+      "step": 24498
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5523821711540222,
+      "learning_rate": 0.0003925147232806341,
+      "loss": 2.9552,
+      "step": 24499
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5339621305465698,
+      "learning_rate": 0.000392500077340673,
+      "loss": 3.0629,
+      "step": 24500
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5230332016944885,
+      "learning_rate": 0.0003924854311570845,
+      "loss": 3.3095,
+      "step": 24501
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5600953698158264,
+      "learning_rate": 0.00039247078472990716,
+      "loss": 3.1273,
+      "step": 24502
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5858113765716553,
+      "learning_rate": 0.00039245613805917955,
+      "loss": 2.9782,
+      "step": 24503
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5632593631744385,
+      "learning_rate": 0.00039244149114494036,
+      "loss": 3.1121,
+      "step": 24504
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5598071813583374,
+      "learning_rate": 0.00039242684398722806,
+      "loss": 3.083,
+      "step": 24505
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.598629891872406,
+      "learning_rate": 0.00039241219658608115,
+      "loss": 3.0239,
+      "step": 24506
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6259543895721436,
+      "learning_rate": 0.00039239754894153833,
+      "loss": 3.2252,
+      "step": 24507
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5557901263237,
+      "learning_rate": 0.00039238290105363815,
+      "loss": 3.1079,
+      "step": 24508
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5629055500030518,
+      "learning_rate": 0.00039236825292241926,
+      "loss": 3.1228,
+      "step": 24509
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.578502357006073,
+      "learning_rate": 0.00039235360454792015,
+      "loss": 3.042,
+      "step": 24510
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.591256320476532,
+      "learning_rate": 0.0003923389559301793,
+      "loss": 3.0652,
+      "step": 24511
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.566704273223877,
+      "learning_rate": 0.00039232430706923554,
+      "loss": 3.0541,
+      "step": 24512
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5784484148025513,
+      "learning_rate": 0.00039230965796512723,
+      "loss": 2.751,
+      "step": 24513
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5382969975471497,
+      "learning_rate": 0.0003922950086178931,
+      "loss": 2.9646,
+      "step": 24514
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5786687731742859,
+      "learning_rate": 0.00039228035902757173,
+      "loss": 3.1632,
+      "step": 24515
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6045399308204651,
+      "learning_rate": 0.0003922657091942016,
+      "loss": 3.0142,
+      "step": 24516
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5659675598144531,
+      "learning_rate": 0.00039225105911782124,
+      "loss": 3.183,
+      "step": 24517
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5671629905700684,
+      "learning_rate": 0.00039223640879846943,
+      "loss": 3.1611,
+      "step": 24518
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5662316083908081,
+      "learning_rate": 0.0003922217582361848,
+      "loss": 3.2395,
+      "step": 24519
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5527083873748779,
+      "learning_rate": 0.00039220710743100565,
+      "loss": 3.312,
+      "step": 24520
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.553294837474823,
+      "learning_rate": 0.0003921924563829708,
+      "loss": 3.0211,
+      "step": 24521
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.658713161945343,
+      "learning_rate": 0.0003921778050921187,
+      "loss": 3.2227,
+      "step": 24522
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6088311076164246,
+      "learning_rate": 0.00039216315355848803,
+      "loss": 3.2464,
+      "step": 24523
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6018931269645691,
+      "learning_rate": 0.0003921485017821173,
+      "loss": 3.1967,
+      "step": 24524
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5140994787216187,
+      "learning_rate": 0.00039213384976304527,
+      "loss": 3.0007,
+      "step": 24525
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5533731579780579,
+      "learning_rate": 0.0003921191975013103,
+      "loss": 2.992,
+      "step": 24526
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5808982849121094,
+      "learning_rate": 0.00039210454499695116,
+      "loss": 3.2767,
+      "step": 24527
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5596871972084045,
+      "learning_rate": 0.00039208989225000633,
+      "loss": 3.1751,
+      "step": 24528
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.542251467704773,
+      "learning_rate": 0.00039207523926051453,
+      "loss": 3.0595,
+      "step": 24529
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6116825342178345,
+      "learning_rate": 0.0003920605860285142,
+      "loss": 3.1456,
+      "step": 24530
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5442792177200317,
+      "learning_rate": 0.000392045932554044,
+      "loss": 3.2253,
+      "step": 24531
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5724294781684875,
+      "learning_rate": 0.0003920312788371425,
+      "loss": 3.3303,
+      "step": 24532
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5326899290084839,
+      "learning_rate": 0.00039201662487784844,
+      "loss": 2.9463,
+      "step": 24533
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5549022555351257,
+      "learning_rate": 0.0003920019706762002,
+      "loss": 3.0103,
+      "step": 24534
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.533281147480011,
+      "learning_rate": 0.0003919873162322365,
+      "loss": 3.0965,
+      "step": 24535
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5557292699813843,
+      "learning_rate": 0.0003919726615459959,
+      "loss": 3.0286,
+      "step": 24536
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.558367133140564,
+      "learning_rate": 0.000391958006617517,
+      "loss": 3.2902,
+      "step": 24537
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5451139211654663,
+      "learning_rate": 0.00039194335144683844,
+      "loss": 3.2517,
+      "step": 24538
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5590549111366272,
+      "learning_rate": 0.00039192869603399877,
+      "loss": 2.977,
+      "step": 24539
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5913295745849609,
+      "learning_rate": 0.00039191404037903664,
+      "loss": 2.984,
+      "step": 24540
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5299440622329712,
+      "learning_rate": 0.0003918993844819906,
+      "loss": 3.129,
+      "step": 24541
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5578473210334778,
+      "learning_rate": 0.0003918847283428992,
+      "loss": 2.9886,
+      "step": 24542
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5618113875389099,
+      "learning_rate": 0.0003918700719618012,
+      "loss": 3.1438,
+      "step": 24543
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5523414015769958,
+      "learning_rate": 0.0003918554153387351,
+      "loss": 3.1398,
+      "step": 24544
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5433165431022644,
+      "learning_rate": 0.0003918407584737395,
+      "loss": 3.0228,
+      "step": 24545
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5516573786735535,
+      "learning_rate": 0.0003918261013668531,
+      "loss": 3.2176,
+      "step": 24546
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.548460841178894,
+      "learning_rate": 0.00039181144401811426,
+      "loss": 3.0904,
+      "step": 24547
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.57688307762146,
+      "learning_rate": 0.0003917967864275618,
+      "loss": 2.8368,
+      "step": 24548
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5394658446311951,
+      "learning_rate": 0.0003917821285952343,
+      "loss": 3.1838,
+      "step": 24549
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5557076334953308,
+      "learning_rate": 0.0003917674705211703,
+      "loss": 3.0597,
+      "step": 24550
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5342375636100769,
+      "learning_rate": 0.00039175281220540844,
+      "loss": 2.9482,
+      "step": 24551
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5965703725814819,
+      "learning_rate": 0.00039173815364798744,
+      "loss": 2.9069,
+      "step": 24552
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5493952631950378,
+      "learning_rate": 0.0003917234948489457,
+      "loss": 3.0136,
+      "step": 24553
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.563545286655426,
+      "learning_rate": 0.0003917088358083219,
+      "loss": 2.9065,
+      "step": 24554
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.575928807258606,
+      "learning_rate": 0.0003916941765261548,
+      "loss": 3.1358,
+      "step": 24555
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5435425043106079,
+      "learning_rate": 0.0003916795170024828,
+      "loss": 2.9169,
+      "step": 24556
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5062286853790283,
+      "learning_rate": 0.0003916648572373446,
+      "loss": 2.9257,
+      "step": 24557
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5492591857910156,
+      "learning_rate": 0.0003916501972307787,
+      "loss": 3.0671,
+      "step": 24558
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5551416873931885,
+      "learning_rate": 0.000391635536982824,
+      "loss": 3.117,
+      "step": 24559
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5714171528816223,
+      "learning_rate": 0.0003916208764935189,
+      "loss": 3.0568,
+      "step": 24560
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5315268635749817,
+      "learning_rate": 0.000391606215762902,
+      "loss": 3.2503,
+      "step": 24561
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5530765056610107,
+      "learning_rate": 0.00039159155479101196,
+      "loss": 3.0153,
+      "step": 24562
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5797913074493408,
+      "learning_rate": 0.0003915768935778874,
+      "loss": 3.2337,
+      "step": 24563
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5785074830055237,
+      "learning_rate": 0.0003915622321235669,
+      "loss": 3.169,
+      "step": 24564
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5695971846580505,
+      "learning_rate": 0.0003915475704280891,
+      "loss": 3.0674,
+      "step": 24565
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5791469216346741,
+      "learning_rate": 0.00039153290849149275,
+      "loss": 3.1011,
+      "step": 24566
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5918307304382324,
+      "learning_rate": 0.0003915182463138161,
+      "loss": 3.1235,
+      "step": 24567
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5434072613716125,
+      "learning_rate": 0.00039150358389509815,
+      "loss": 2.9738,
+      "step": 24568
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6053929328918457,
+      "learning_rate": 0.0003914889212353773,
+      "loss": 3.0853,
+      "step": 24569
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5387932062149048,
+      "learning_rate": 0.0003914742583346924,
+      "loss": 3.0236,
+      "step": 24570
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.572083592414856,
+      "learning_rate": 0.00039145959519308176,
+      "loss": 2.891,
+      "step": 24571
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5593575835227966,
+      "learning_rate": 0.0003914449318105842,
+      "loss": 3.3136,
+      "step": 24572
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5914036631584167,
+      "learning_rate": 0.0003914302681872382,
+      "loss": 3.0979,
+      "step": 24573
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5585025548934937,
+      "learning_rate": 0.0003914156043230826,
+      "loss": 3.1267,
+      "step": 24574
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5593438148498535,
+      "learning_rate": 0.0003914009402181558,
+      "loss": 2.9214,
+      "step": 24575
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5880452394485474,
+      "learning_rate": 0.0003913862758724966,
+      "loss": 3.0177,
+      "step": 24576
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5511958003044128,
+      "learning_rate": 0.00039137161128614345,
+      "loss": 3.0525,
+      "step": 24577
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5612953305244446,
+      "learning_rate": 0.00039135694645913506,
+      "loss": 3.189,
+      "step": 24578
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5851930379867554,
+      "learning_rate": 0.0003913422813915101,
+      "loss": 2.9018,
+      "step": 24579
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5985074043273926,
+      "learning_rate": 0.00039132761608330716,
+      "loss": 3.0104,
+      "step": 24580
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.545453667640686,
+      "learning_rate": 0.0003913129505345649,
+      "loss": 3.1573,
+      "step": 24581
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5423109531402588,
+      "learning_rate": 0.0003912982847453218,
+      "loss": 2.9684,
+      "step": 24582
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5626406073570251,
+      "learning_rate": 0.0003912836187156166,
+      "loss": 3.0915,
+      "step": 24583
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5251243710517883,
+      "learning_rate": 0.000391268952445488,
+      "loss": 2.8087,
+      "step": 24584
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5345422625541687,
+      "learning_rate": 0.0003912542859349745,
+      "loss": 3.0146,
+      "step": 24585
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.563607394695282,
+      "learning_rate": 0.0003912396191841148,
+      "loss": 3.1094,
+      "step": 24586
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5774620175361633,
+      "learning_rate": 0.00039122495219294734,
+      "loss": 3.1227,
+      "step": 24587
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6058965921401978,
+      "learning_rate": 0.0003912102849615111,
+      "loss": 3.0073,
+      "step": 24588
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5583047866821289,
+      "learning_rate": 0.00039119561748984446,
+      "loss": 3.2368,
+      "step": 24589
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5507198572158813,
+      "learning_rate": 0.00039118094977798605,
+      "loss": 3.062,
+      "step": 24590
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5293483734130859,
+      "learning_rate": 0.00039116628182597464,
+      "loss": 3.1221,
+      "step": 24591
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5906999111175537,
+      "learning_rate": 0.0003911516136338489,
+      "loss": 3.1421,
+      "step": 24592
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5505334734916687,
+      "learning_rate": 0.0003911369452016472,
+      "loss": 3.2336,
+      "step": 24593
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5691468119621277,
+      "learning_rate": 0.00039112227652940825,
+      "loss": 2.9671,
+      "step": 24594
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6026682257652283,
+      "learning_rate": 0.00039110760761717095,
+      "loss": 3.0389,
+      "step": 24595
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5420871376991272,
+      "learning_rate": 0.00039109293846497365,
+      "loss": 3.0556,
+      "step": 24596
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5781927704811096,
+      "learning_rate": 0.0003910782690728551,
+      "loss": 3.016,
+      "step": 24597
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5528817772865295,
+      "learning_rate": 0.0003910635994408539,
+      "loss": 3.1946,
+      "step": 24598
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5814532041549683,
+      "learning_rate": 0.0003910489295690087,
+      "loss": 3.0579,
+      "step": 24599
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6129553914070129,
+      "learning_rate": 0.00039103425945735817,
+      "loss": 3.1454,
+      "step": 24600
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5644974708557129,
+      "learning_rate": 0.0003910195891059409,
+      "loss": 2.8995,
+      "step": 24601
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5495239496231079,
+      "learning_rate": 0.0003910049185147955,
+      "loss": 3.2219,
+      "step": 24602
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5390620231628418,
+      "learning_rate": 0.0003909902476839608,
+      "loss": 3.3411,
+      "step": 24603
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5480315685272217,
+      "learning_rate": 0.0003909755766134751,
+      "loss": 3.2259,
+      "step": 24604
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5507662892341614,
+      "learning_rate": 0.0003909609053033774,
+      "loss": 3.0374,
+      "step": 24605
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5539723038673401,
+      "learning_rate": 0.00039094623375370626,
+      "loss": 3.1604,
+      "step": 24606
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5907084345817566,
+      "learning_rate": 0.00039093156196450007,
+      "loss": 3.0563,
+      "step": 24607
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.569409191608429,
+      "learning_rate": 0.0003909168899357977,
+      "loss": 3.0649,
+      "step": 24608
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.529839277267456,
+      "learning_rate": 0.00039090221766763765,
+      "loss": 3.0733,
+      "step": 24609
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5536385774612427,
+      "learning_rate": 0.00039088754516005883,
+      "loss": 3.1285,
+      "step": 24610
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5590646266937256,
+      "learning_rate": 0.0003908728724130996,
+      "loss": 2.8481,
+      "step": 24611
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5736700296401978,
+      "learning_rate": 0.00039085819942679876,
+      "loss": 3.0309,
+      "step": 24612
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5361476540565491,
+      "learning_rate": 0.00039084352620119485,
+      "loss": 3.094,
+      "step": 24613
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5522602796554565,
+      "learning_rate": 0.0003908288527363266,
+      "loss": 3.1455,
+      "step": 24614
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5784001350402832,
+      "learning_rate": 0.00039081417903223263,
+      "loss": 3.0122,
+      "step": 24615
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5663528442382812,
+      "learning_rate": 0.00039079950508895156,
+      "loss": 2.9354,
+      "step": 24616
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5776238441467285,
+      "learning_rate": 0.00039078483090652217,
+      "loss": 3.0843,
+      "step": 24617
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5722895860671997,
+      "learning_rate": 0.00039077015648498293,
+      "loss": 3.0982,
+      "step": 24618
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5736603140830994,
+      "learning_rate": 0.0003907554818243725,
+      "loss": 2.9995,
+      "step": 24619
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5491389632225037,
+      "learning_rate": 0.00039074080692472966,
+      "loss": 2.9911,
+      "step": 24620
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5545860528945923,
+      "learning_rate": 0.0003907261317860931,
+      "loss": 3.074,
+      "step": 24621
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6800806522369385,
+      "learning_rate": 0.0003907114564085013,
+      "loss": 3.2136,
+      "step": 24622
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5676558017730713,
+      "learning_rate": 0.0003906967807919929,
+      "loss": 3.1701,
+      "step": 24623
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.540444552898407,
+      "learning_rate": 0.0003906821049366067,
+      "loss": 3.037,
+      "step": 24624
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.583018958568573,
+      "learning_rate": 0.00039066742884238134,
+      "loss": 3.1316,
+      "step": 24625
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5733615159988403,
+      "learning_rate": 0.00039065275250935535,
+      "loss": 2.998,
+      "step": 24626
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5599751472473145,
+      "learning_rate": 0.00039063807593756744,
+      "loss": 3.0973,
+      "step": 24627
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5724146962165833,
+      "learning_rate": 0.0003906233991270563,
+      "loss": 2.935,
+      "step": 24628
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5427190065383911,
+      "learning_rate": 0.0003906087220778607,
+      "loss": 3.0934,
+      "step": 24629
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5561226010322571,
+      "learning_rate": 0.000390594044790019,
+      "loss": 3.2373,
+      "step": 24630
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5783647894859314,
+      "learning_rate": 0.0003905793672635701,
+      "loss": 3.0405,
+      "step": 24631
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5754805207252502,
+      "learning_rate": 0.00039056468949855253,
+      "loss": 3.2403,
+      "step": 24632
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.558912992477417,
+      "learning_rate": 0.000390550011495005,
+      "loss": 3.0081,
+      "step": 24633
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5601596236228943,
+      "learning_rate": 0.0003905353332529662,
+      "loss": 3.0755,
+      "step": 24634
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5683655142784119,
+      "learning_rate": 0.0003905206547724748,
+      "loss": 3.0046,
+      "step": 24635
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.556339681148529,
+      "learning_rate": 0.00039050597605356927,
+      "loss": 3.1934,
+      "step": 24636
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5865704417228699,
+      "learning_rate": 0.0003904912970962885,
+      "loss": 2.9044,
+      "step": 24637
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5483426451683044,
+      "learning_rate": 0.00039047661790067107,
+      "loss": 2.9773,
+      "step": 24638
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5963693857192993,
+      "learning_rate": 0.00039046193846675563,
+      "loss": 3.1899,
+      "step": 24639
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5514311790466309,
+      "learning_rate": 0.00039044725879458086,
+      "loss": 3.0625,
+      "step": 24640
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5385047793388367,
+      "learning_rate": 0.0003904325788841854,
+      "loss": 3.1098,
+      "step": 24641
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5517145991325378,
+      "learning_rate": 0.00039041789873560796,
+      "loss": 2.91,
+      "step": 24642
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5656097531318665,
+      "learning_rate": 0.00039040321834888725,
+      "loss": 3.1927,
+      "step": 24643
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5733511447906494,
+      "learning_rate": 0.00039038853772406176,
+      "loss": 3.0975,
+      "step": 24644
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5602204203605652,
+      "learning_rate": 0.00039037385686117024,
+      "loss": 3.1281,
+      "step": 24645
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.551032304763794,
+      "learning_rate": 0.0003903591757602514,
+      "loss": 3.3527,
+      "step": 24646
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5796739459037781,
+      "learning_rate": 0.0003903444944213439,
+      "loss": 2.9157,
+      "step": 24647
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5496836304664612,
+      "learning_rate": 0.0003903298128444864,
+      "loss": 3.0158,
+      "step": 24648
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5698907375335693,
+      "learning_rate": 0.0003903151310297176,
+      "loss": 3.2225,
+      "step": 24649
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5394782423973083,
+      "learning_rate": 0.000390300448977076,
+      "loss": 3.0648,
+      "step": 24650
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5592007040977478,
+      "learning_rate": 0.00039028576668660047,
+      "loss": 3.0059,
+      "step": 24651
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5389769077301025,
+      "learning_rate": 0.0003902710841583296,
+      "loss": 3.0591,
+      "step": 24652
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5288125872612,
+      "learning_rate": 0.00039025640139230206,
+      "loss": 3.061,
+      "step": 24653
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5310102105140686,
+      "learning_rate": 0.0003902417183885566,
+      "loss": 3.0722,
+      "step": 24654
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5603654980659485,
+      "learning_rate": 0.00039022703514713167,
+      "loss": 2.9501,
+      "step": 24655
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6040814518928528,
+      "learning_rate": 0.00039021235166806617,
+      "loss": 2.9324,
+      "step": 24656
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5468594431877136,
+      "learning_rate": 0.00039019766795139873,
+      "loss": 3.1048,
+      "step": 24657
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5571081042289734,
+      "learning_rate": 0.00039018298399716797,
+      "loss": 3.1354,
+      "step": 24658
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5484535694122314,
+      "learning_rate": 0.0003901682998054126,
+      "loss": 3.2044,
+      "step": 24659
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5328761339187622,
+      "learning_rate": 0.0003901536153761712,
+      "loss": 2.9363,
+      "step": 24660
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5752051472663879,
+      "learning_rate": 0.0003901389307094827,
+      "loss": 3.0491,
+      "step": 24661
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7036685943603516,
+      "learning_rate": 0.0003901242458053854,
+      "loss": 2.9426,
+      "step": 24662
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5637723803520203,
+      "learning_rate": 0.0003901095606639183,
+      "loss": 2.9939,
+      "step": 24663
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5546841025352478,
+      "learning_rate": 0.00039009487528512,
+      "loss": 2.8504,
+      "step": 24664
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5602949857711792,
+      "learning_rate": 0.00039008018966902913,
+      "loss": 3.3448,
+      "step": 24665
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5492205023765564,
+      "learning_rate": 0.00039006550381568425,
+      "loss": 3.0376,
+      "step": 24666
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6059777140617371,
+      "learning_rate": 0.0003900508177251242,
+      "loss": 3.0945,
+      "step": 24667
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5843932628631592,
+      "learning_rate": 0.0003900361313973877,
+      "loss": 3.0452,
+      "step": 24668
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.59035325050354,
+      "learning_rate": 0.0003900214448325133,
+      "loss": 3.145,
+      "step": 24669
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5399641394615173,
+      "learning_rate": 0.00039000675803053985,
+      "loss": 3.2487,
+      "step": 24670
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6116369366645813,
+      "learning_rate": 0.0003899920709915057,
+      "loss": 3.2259,
+      "step": 24671
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5800071954727173,
+      "learning_rate": 0.00038997738371545,
+      "loss": 2.9216,
+      "step": 24672
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.608817458152771,
+      "learning_rate": 0.00038996269620241103,
+      "loss": 3.0501,
+      "step": 24673
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5628402829170227,
+      "learning_rate": 0.00038994800845242766,
+      "loss": 3.0112,
+      "step": 24674
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5566816329956055,
+      "learning_rate": 0.00038993332046553857,
+      "loss": 3.2272,
+      "step": 24675
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5373614430427551,
+      "learning_rate": 0.0003899186322417824,
+      "loss": 2.8742,
+      "step": 24676
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5301047563552856,
+      "learning_rate": 0.0003899039437811979,
+      "loss": 3.0577,
+      "step": 24677
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5556225776672363,
+      "learning_rate": 0.00038988925508382367,
+      "loss": 3.0731,
+      "step": 24678
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5261009931564331,
+      "learning_rate": 0.00038987456614969853,
+      "loss": 2.9431,
+      "step": 24679
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6140367388725281,
+      "learning_rate": 0.000389859876978861,
+      "loss": 3.1981,
+      "step": 24680
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5956748127937317,
+      "learning_rate": 0.00038984518757134983,
+      "loss": 3.1423,
+      "step": 24681
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5351552963256836,
+      "learning_rate": 0.00038983049792720373,
+      "loss": 3.0685,
+      "step": 24682
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5489239692687988,
+      "learning_rate": 0.00038981580804646146,
+      "loss": 3.2112,
+      "step": 24683
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5213435292243958,
+      "learning_rate": 0.00038980111792916156,
+      "loss": 3.0596,
+      "step": 24684
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5946751236915588,
+      "learning_rate": 0.00038978642757534285,
+      "loss": 3.0055,
+      "step": 24685
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5385335087776184,
+      "learning_rate": 0.0003897717369850439,
+      "loss": 3.146,
+      "step": 24686
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5397436022758484,
+      "learning_rate": 0.00038975704615830354,
+      "loss": 3.0557,
+      "step": 24687
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5667961239814758,
+      "learning_rate": 0.00038974235509516036,
+      "loss": 3.1658,
+      "step": 24688
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5461593866348267,
+      "learning_rate": 0.0003897276637956531,
+      "loss": 3.0923,
+      "step": 24689
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5522091388702393,
+      "learning_rate": 0.0003897129722598204,
+      "loss": 3.0271,
+      "step": 24690
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5801872611045837,
+      "learning_rate": 0.00038969828048770105,
+      "loss": 3.3063,
+      "step": 24691
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5313283801078796,
+      "learning_rate": 0.0003896835884793337,
+      "loss": 2.9629,
+      "step": 24692
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5575485229492188,
+      "learning_rate": 0.000389668896234757,
+      "loss": 3.1138,
+      "step": 24693
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5732911825180054,
+      "learning_rate": 0.0003896542037540098,
+      "loss": 3.1065,
+      "step": 24694
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5623870491981506,
+      "learning_rate": 0.0003896395110371306,
+      "loss": 3.2062,
+      "step": 24695
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5679678320884705,
+      "learning_rate": 0.00038962481808415807,
+      "loss": 3.0646,
+      "step": 24696
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.578198254108429,
+      "learning_rate": 0.0003896101248951312,
+      "loss": 3.3072,
+      "step": 24697
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5689905881881714,
+      "learning_rate": 0.0003895954314700884,
+      "loss": 3.1302,
+      "step": 24698
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5264217853546143,
+      "learning_rate": 0.00038958073780906845,
+      "loss": 3.0872,
+      "step": 24699
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5618980526924133,
+      "learning_rate": 0.00038956604391211016,
+      "loss": 3.2415,
+      "step": 24700
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5655535459518433,
+      "learning_rate": 0.0003895513497792521,
+      "loss": 3.0696,
+      "step": 24701
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5163285136222839,
+      "learning_rate": 0.000389536655410533,
+      "loss": 3.1711,
+      "step": 24702
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7728413343429565,
+      "learning_rate": 0.0003895219608059916,
+      "loss": 3.008,
+      "step": 24703
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5822486877441406,
+      "learning_rate": 0.0003895072659656666,
+      "loss": 3.1006,
+      "step": 24704
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5410815477371216,
+      "learning_rate": 0.00038949257088959674,
+      "loss": 2.9528,
+      "step": 24705
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5558587312698364,
+      "learning_rate": 0.0003894778755778206,
+      "loss": 3.0161,
+      "step": 24706
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5925754904747009,
+      "learning_rate": 0.0003894631800303769,
+      "loss": 3.1687,
+      "step": 24707
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5678203701972961,
+      "learning_rate": 0.00038944848424730456,
+      "loss": 3.1855,
+      "step": 24708
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5488570928573608,
+      "learning_rate": 0.00038943378822864203,
+      "loss": 3.2203,
+      "step": 24709
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5649476051330566,
+      "learning_rate": 0.00038941909197442813,
+      "loss": 3.1933,
+      "step": 24710
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.599998950958252,
+      "learning_rate": 0.0003894043954847015,
+      "loss": 3.1783,
+      "step": 24711
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5275146961212158,
+      "learning_rate": 0.00038938969875950097,
+      "loss": 3.058,
+      "step": 24712
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5559340119361877,
+      "learning_rate": 0.0003893750017988651,
+      "loss": 3.2087,
+      "step": 24713
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5753278732299805,
+      "learning_rate": 0.00038936030460283276,
+      "loss": 3.2299,
+      "step": 24714
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5743871927261353,
+      "learning_rate": 0.0003893456071714425,
+      "loss": 3.1301,
+      "step": 24715
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5973168611526489,
+      "learning_rate": 0.0003893309095047332,
+      "loss": 2.9648,
+      "step": 24716
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5473564863204956,
+      "learning_rate": 0.00038931621160274337,
+      "loss": 3.1843,
+      "step": 24717
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5533649325370789,
+      "learning_rate": 0.00038930151346551185,
+      "loss": 2.9377,
+      "step": 24718
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.551533043384552,
+      "learning_rate": 0.0003892868150930775,
+      "loss": 3.2214,
+      "step": 24719
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5771843791007996,
+      "learning_rate": 0.0003892721164854787,
+      "loss": 3.0803,
+      "step": 24720
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5571311712265015,
+      "learning_rate": 0.0003892574176427543,
+      "loss": 2.997,
+      "step": 24721
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5600820183753967,
+      "learning_rate": 0.00038924271856494305,
+      "loss": 3.082,
+      "step": 24722
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5592990517616272,
+      "learning_rate": 0.0003892280192520837,
+      "loss": 3.0899,
+      "step": 24723
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5576978921890259,
+      "learning_rate": 0.00038921331970421494,
+      "loss": 3.3335,
+      "step": 24724
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5841004848480225,
+      "learning_rate": 0.0003891986199213754,
+      "loss": 3.0562,
+      "step": 24725
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5277174115180969,
+      "learning_rate": 0.0003891839199036039,
+      "loss": 2.9582,
+      "step": 24726
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5508502125740051,
+      "learning_rate": 0.0003891692196509391,
+      "loss": 3.1022,
+      "step": 24727
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5699372887611389,
+      "learning_rate": 0.00038915451916341973,
+      "loss": 3.0745,
+      "step": 24728
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5580192804336548,
+      "learning_rate": 0.0003891398184410846,
+      "loss": 2.8848,
+      "step": 24729
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6069357395172119,
+      "learning_rate": 0.00038912511748397235,
+      "loss": 2.7991,
+      "step": 24730
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5295454859733582,
+      "learning_rate": 0.00038911041629212156,
+      "loss": 2.9931,
+      "step": 24731
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5806009769439697,
+      "learning_rate": 0.0003890957148655711,
+      "loss": 3.0751,
+      "step": 24732
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5367207527160645,
+      "learning_rate": 0.0003890810132043597,
+      "loss": 2.8636,
+      "step": 24733
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5449971556663513,
+      "learning_rate": 0.00038906631130852614,
+      "loss": 2.8036,
+      "step": 24734
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5441697835922241,
+      "learning_rate": 0.00038905160917810896,
+      "loss": 3.0157,
+      "step": 24735
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5573702454566956,
+      "learning_rate": 0.000389036906813147,
+      "loss": 3.2127,
+      "step": 24736
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5617653131484985,
+      "learning_rate": 0.000389022204213679,
+      "loss": 3.0465,
+      "step": 24737
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5416889786720276,
+      "learning_rate": 0.0003890075013797437,
+      "loss": 3.1156,
+      "step": 24738
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5474164485931396,
+      "learning_rate": 0.0003889927983113796,
+      "loss": 3.0115,
+      "step": 24739
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5465503931045532,
+      "learning_rate": 0.0003889780950086257,
+      "loss": 3.2573,
+      "step": 24740
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5438640117645264,
+      "learning_rate": 0.00038896339147152066,
+      "loss": 3.0766,
+      "step": 24741
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5305720567703247,
+      "learning_rate": 0.00038894868770010315,
+      "loss": 2.8584,
+      "step": 24742
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5442739129066467,
+      "learning_rate": 0.0003889339836944118,
+      "loss": 3.0769,
+      "step": 24743
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6273128390312195,
+      "learning_rate": 0.0003889192794544856,
+      "loss": 2.9629,
+      "step": 24744
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5607988834381104,
+      "learning_rate": 0.0003889045749803631,
+      "loss": 3.2894,
+      "step": 24745
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5226300954818726,
+      "learning_rate": 0.00038888987027208303,
+      "loss": 3.0109,
+      "step": 24746
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5920791625976562,
+      "learning_rate": 0.0003888751653296841,
+      "loss": 3.0861,
+      "step": 24747
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.552906334400177,
+      "learning_rate": 0.0003888604601532052,
+      "loss": 3.0653,
+      "step": 24748
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.565069317817688,
+      "learning_rate": 0.0003888457547426848,
+      "loss": 3.0598,
+      "step": 24749
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5915144681930542,
+      "learning_rate": 0.0003888310490981619,
+      "loss": 3.0202,
+      "step": 24750
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5616911053657532,
+      "learning_rate": 0.0003888163432196751,
+      "loss": 2.9658,
+      "step": 24751
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5911539793014526,
+      "learning_rate": 0.0003888016371072631,
+      "loss": 2.8024,
+      "step": 24752
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5526560544967651,
+      "learning_rate": 0.0003887869307609647,
+      "loss": 3.0992,
+      "step": 24753
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5728417634963989,
+      "learning_rate": 0.00038877222418081856,
+      "loss": 3.0105,
+      "step": 24754
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5776088833808899,
+      "learning_rate": 0.00038875751736686353,
+      "loss": 3.1478,
+      "step": 24755
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5825873613357544,
+      "learning_rate": 0.00038874281031913837,
+      "loss": 3.2437,
+      "step": 24756
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5393372774124146,
+      "learning_rate": 0.0003887281030376816,
+      "loss": 3.0675,
+      "step": 24757
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5632210969924927,
+      "learning_rate": 0.00038871339552253205,
+      "loss": 2.968,
+      "step": 24758
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.553977370262146,
+      "learning_rate": 0.00038869868777372863,
+      "loss": 3.0479,
+      "step": 24759
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.588313639163971,
+      "learning_rate": 0.00038868397979130984,
+      "loss": 3.0204,
+      "step": 24760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5613222718238831,
+      "learning_rate": 0.0003886692715753145,
+      "loss": 2.8621,
+      "step": 24761
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5414893627166748,
+      "learning_rate": 0.0003886545631257814,
+      "loss": 2.8851,
+      "step": 24762
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5642477869987488,
+      "learning_rate": 0.0003886398544427493,
+      "loss": 2.9733,
+      "step": 24763
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5491669178009033,
+      "learning_rate": 0.00038862514552625676,
+      "loss": 3.1974,
+      "step": 24764
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5662064552307129,
+      "learning_rate": 0.0003886104363763427,
+      "loss": 2.9075,
+      "step": 24765
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5613273978233337,
+      "learning_rate": 0.0003885957269930458,
+      "loss": 3.1875,
+      "step": 24766
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5412261486053467,
+      "learning_rate": 0.0003885810173764049,
+      "loss": 3.193,
+      "step": 24767
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5381949543952942,
+      "learning_rate": 0.0003885663075264585,
+      "loss": 3.0033,
+      "step": 24768
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5455308556556702,
+      "learning_rate": 0.0003885515974432455,
+      "loss": 3.0217,
+      "step": 24769
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5200797915458679,
+      "learning_rate": 0.0003885368871268047,
+      "loss": 2.9717,
+      "step": 24770
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5431541800498962,
+      "learning_rate": 0.00038852217657717484,
+      "loss": 2.9984,
+      "step": 24771
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5343523621559143,
+      "learning_rate": 0.0003885074657943945,
+      "loss": 3.1071,
+      "step": 24772
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5573854446411133,
+      "learning_rate": 0.00038849275477850255,
+      "loss": 3.0734,
+      "step": 24773
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5465708374977112,
+      "learning_rate": 0.0003884780435295377,
+      "loss": 3.0892,
+      "step": 24774
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5360824465751648,
+      "learning_rate": 0.00038846333204753874,
+      "loss": 2.935,
+      "step": 24775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5567554831504822,
+      "learning_rate": 0.0003884486203325444,
+      "loss": 3.0637,
+      "step": 24776
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5666375756263733,
+      "learning_rate": 0.0003884339083845934,
+      "loss": 2.9533,
+      "step": 24777
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5433690547943115,
+      "learning_rate": 0.0003884191962037245,
+      "loss": 3.0508,
+      "step": 24778
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5279322266578674,
+      "learning_rate": 0.0003884044837899764,
+      "loss": 3.0479,
+      "step": 24779
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.524019718170166,
+      "learning_rate": 0.00038838977114338797,
+      "loss": 3.0864,
+      "step": 24780
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5386550426483154,
+      "learning_rate": 0.00038837505826399795,
+      "loss": 3.109,
+      "step": 24781
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5483960509300232,
+      "learning_rate": 0.0003883603451518449,
+      "loss": 2.943,
+      "step": 24782
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5979861617088318,
+      "learning_rate": 0.0003883456318069678,
+      "loss": 3.1003,
+      "step": 24783
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5268682837486267,
+      "learning_rate": 0.0003883309182294052,
+      "loss": 3.2098,
+      "step": 24784
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5503677129745483,
+      "learning_rate": 0.0003883162044191962,
+      "loss": 3.1854,
+      "step": 24785
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5744519829750061,
+      "learning_rate": 0.00038830149037637906,
+      "loss": 3.1521,
+      "step": 24786
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5381206274032593,
+      "learning_rate": 0.00038828677610099294,
+      "loss": 3.1749,
+      "step": 24787
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.569991946220398,
+      "learning_rate": 0.00038827206159307634,
+      "loss": 3.2375,
+      "step": 24788
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5447730422019958,
+      "learning_rate": 0.0003882573468526682,
+      "loss": 3.1548,
+      "step": 24789
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5430207252502441,
+      "learning_rate": 0.0003882426318798071,
+      "loss": 3.0477,
+      "step": 24790
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5270898342132568,
+      "learning_rate": 0.00038822791667453196,
+      "loss": 3.1821,
+      "step": 24791
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5762162804603577,
+      "learning_rate": 0.0003882132012368815,
+      "loss": 2.8535,
+      "step": 24792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5520655512809753,
+      "learning_rate": 0.00038819848556689444,
+      "loss": 3.1805,
+      "step": 24793
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.557744026184082,
+      "learning_rate": 0.00038818376966460945,
+      "loss": 3.0716,
+      "step": 24794
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5885498523712158,
+      "learning_rate": 0.0003881690535300654,
+      "loss": 3.1626,
+      "step": 24795
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.559755802154541,
+      "learning_rate": 0.00038815433716330117,
+      "loss": 3.0249,
+      "step": 24796
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5420626401901245,
+      "learning_rate": 0.0003881396205643553,
+      "loss": 3.1339,
+      "step": 24797
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5564322471618652,
+      "learning_rate": 0.00038812490373326655,
+      "loss": 2.9808,
+      "step": 24798
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5733435750007629,
+      "learning_rate": 0.00038811018667007387,
+      "loss": 2.9819,
+      "step": 24799
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5582762956619263,
+      "learning_rate": 0.00038809546937481585,
+      "loss": 2.9805,
+      "step": 24800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.577571451663971,
+      "learning_rate": 0.00038808075184753135,
+      "loss": 3.1034,
+      "step": 24801
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5833460688591003,
+      "learning_rate": 0.000388066034088259,
+      "loss": 3.1709,
+      "step": 24802
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5358278155326843,
+      "learning_rate": 0.0003880513160970378,
+      "loss": 2.8111,
+      "step": 24803
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5528193116188049,
+      "learning_rate": 0.00038803659787390633,
+      "loss": 3.4981,
+      "step": 24804
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5586181879043579,
+      "learning_rate": 0.0003880218794189034,
+      "loss": 2.8715,
+      "step": 24805
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5519729256629944,
+      "learning_rate": 0.00038800716073206785,
+      "loss": 3.1437,
+      "step": 24806
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5765265226364136,
+      "learning_rate": 0.0003879924418134383,
+      "loss": 3.204,
+      "step": 24807
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5296533107757568,
+      "learning_rate": 0.00038797772266305365,
+      "loss": 3.1702,
+      "step": 24808
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5647825598716736,
+      "learning_rate": 0.00038796300328095245,
+      "loss": 3.1688,
+      "step": 24809
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5361009836196899,
+      "learning_rate": 0.00038794828366717384,
+      "loss": 3.139,
+      "step": 24810
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5457804799079895,
+      "learning_rate": 0.0003879335638217563,
+      "loss": 3.1071,
+      "step": 24811
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5646703839302063,
+      "learning_rate": 0.0003879188437447386,
+      "loss": 3.1663,
+      "step": 24812
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5620137453079224,
+      "learning_rate": 0.0003879041234361597,
+      "loss": 3.2054,
+      "step": 24813
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5515795350074768,
+      "learning_rate": 0.0003878894028960582,
+      "loss": 3.1787,
+      "step": 24814
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5928753614425659,
+      "learning_rate": 0.00038787468212447287,
+      "loss": 3.2087,
+      "step": 24815
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5143396854400635,
+      "learning_rate": 0.00038785996112144263,
+      "loss": 3.0081,
+      "step": 24816
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5518531799316406,
+      "learning_rate": 0.0003878452398870061,
+      "loss": 3.0035,
+      "step": 24817
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5611719489097595,
+      "learning_rate": 0.00038783051842120227,
+      "loss": 3.1042,
+      "step": 24818
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5301064252853394,
+      "learning_rate": 0.0003878157967240696,
+      "loss": 3.0097,
+      "step": 24819
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.575943648815155,
+      "learning_rate": 0.00038780107479564695,
+      "loss": 2.8759,
+      "step": 24820
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5593007802963257,
+      "learning_rate": 0.00038778635263597336,
+      "loss": 3.2868,
+      "step": 24821
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5593023300170898,
+      "learning_rate": 0.00038777163024508726,
+      "loss": 3.0437,
+      "step": 24822
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5502452254295349,
+      "learning_rate": 0.0003877569076230277,
+      "loss": 3.179,
+      "step": 24823
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5816333889961243,
+      "learning_rate": 0.0003877421847698333,
+      "loss": 2.8658,
+      "step": 24824
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5567662119865417,
+      "learning_rate": 0.0003877274616855428,
+      "loss": 2.9821,
+      "step": 24825
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5488287210464478,
+      "learning_rate": 0.0003877127383701951,
+      "loss": 3.1091,
+      "step": 24826
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.534261167049408,
+      "learning_rate": 0.0003876980148238289,
+      "loss": 2.8979,
+      "step": 24827
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5407263040542603,
+      "learning_rate": 0.000387683291046483,
+      "loss": 3.2039,
+      "step": 24828
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5979218482971191,
+      "learning_rate": 0.0003876685670381964,
+      "loss": 3.146,
+      "step": 24829
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5862070322036743,
+      "learning_rate": 0.0003876538427990073,
+      "loss": 3.0431,
+      "step": 24830
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5690160393714905,
+      "learning_rate": 0.0003876391183289551,
+      "loss": 3.1891,
+      "step": 24831
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5591949224472046,
+      "learning_rate": 0.00038762439362807825,
+      "loss": 3.1933,
+      "step": 24832
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5995550751686096,
+      "learning_rate": 0.0003876096686964157,
+      "loss": 3.0018,
+      "step": 24833
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5738639235496521,
+      "learning_rate": 0.000387594943534006,
+      "loss": 3.0424,
+      "step": 24834
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5617707371711731,
+      "learning_rate": 0.00038758021814088806,
+      "loss": 2.9083,
+      "step": 24835
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5648507475852966,
+      "learning_rate": 0.00038756549251710087,
+      "loss": 2.8961,
+      "step": 24836
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6076490879058838,
+      "learning_rate": 0.00038755076666268284,
+      "loss": 3.1156,
+      "step": 24837
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5720618963241577,
+      "learning_rate": 0.00038753604057767307,
+      "loss": 2.8891,
+      "step": 24838
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5429185628890991,
+      "learning_rate": 0.00038752131426211014,
+      "loss": 3.1625,
+      "step": 24839
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5565420389175415,
+      "learning_rate": 0.00038750658771603293,
+      "loss": 3.0722,
+      "step": 24840
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5758388042449951,
+      "learning_rate": 0.00038749186093948025,
+      "loss": 3.2304,
+      "step": 24841
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5630769729614258,
+      "learning_rate": 0.00038747713393249075,
+      "loss": 3.1224,
+      "step": 24842
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5871720910072327,
+      "learning_rate": 0.0003874624066951034,
+      "loss": 2.9868,
+      "step": 24843
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5743727684020996,
+      "learning_rate": 0.00038744767922735694,
+      "loss": 3.1502,
+      "step": 24844
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5455135107040405,
+      "learning_rate": 0.00038743295152929,
+      "loss": 3.2048,
+      "step": 24845
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5863217115402222,
+      "learning_rate": 0.0003874182236009415,
+      "loss": 3.1062,
+      "step": 24846
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5657460689544678,
+      "learning_rate": 0.00038740349544235037,
+      "loss": 3.026,
+      "step": 24847
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5640048980712891,
+      "learning_rate": 0.0003873887670535551,
+      "loss": 2.9006,
+      "step": 24848
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5563377737998962,
+      "learning_rate": 0.00038737403843459476,
+      "loss": 3.159,
+      "step": 24849
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5812902450561523,
+      "learning_rate": 0.0003873593095855079,
+      "loss": 3.2999,
+      "step": 24850
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5502186417579651,
+      "learning_rate": 0.0003873445805063335,
+      "loss": 3.1485,
+      "step": 24851
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6045628786087036,
+      "learning_rate": 0.00038732985119711025,
+      "loss": 3.1333,
+      "step": 24852
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5175624489784241,
+      "learning_rate": 0.000387315121657877,
+      "loss": 3.1418,
+      "step": 24853
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5742688179016113,
+      "learning_rate": 0.00038730039188867254,
+      "loss": 3.2371,
+      "step": 24854
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5657257437705994,
+      "learning_rate": 0.0003872856618895356,
+      "loss": 3.2323,
+      "step": 24855
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5566824674606323,
+      "learning_rate": 0.00038727093166050503,
+      "loss": 2.8398,
+      "step": 24856
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5399627089500427,
+      "learning_rate": 0.0003872562012016197,
+      "loss": 2.7679,
+      "step": 24857
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5554389357566833,
+      "learning_rate": 0.0003872414705129183,
+      "loss": 2.9823,
+      "step": 24858
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5519767999649048,
+      "learning_rate": 0.00038722673959443963,
+      "loss": 3.1616,
+      "step": 24859
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5414425134658813,
+      "learning_rate": 0.00038721200844622257,
+      "loss": 3.1378,
+      "step": 24860
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5529886484146118,
+      "learning_rate": 0.0003871972770683058,
+      "loss": 3.2881,
+      "step": 24861
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5492392778396606,
+      "learning_rate": 0.00038718254546072815,
+      "loss": 3.008,
+      "step": 24862
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5687651634216309,
+      "learning_rate": 0.00038716781362352847,
+      "loss": 3.2633,
+      "step": 24863
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5575495958328247,
+      "learning_rate": 0.0003871530815567456,
+      "loss": 2.9893,
+      "step": 24864
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6896212100982666,
+      "learning_rate": 0.00038713834926041825,
+      "loss": 3.0161,
+      "step": 24865
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5157473087310791,
+      "learning_rate": 0.00038712361673458524,
+      "loss": 3.2208,
+      "step": 24866
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5381258130073547,
+      "learning_rate": 0.00038710888397928544,
+      "loss": 2.9892,
+      "step": 24867
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.577904999256134,
+      "learning_rate": 0.00038709415099455756,
+      "loss": 3.0439,
+      "step": 24868
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5449947118759155,
+      "learning_rate": 0.00038707941778044053,
+      "loss": 2.9669,
+      "step": 24869
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.557723879814148,
+      "learning_rate": 0.000387064684336973,
+      "loss": 3.1256,
+      "step": 24870
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.557279109954834,
+      "learning_rate": 0.0003870499506641938,
+      "loss": 3.046,
+      "step": 24871
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5531541705131531,
+      "learning_rate": 0.0003870352167621419,
+      "loss": 2.9408,
+      "step": 24872
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5873280167579651,
+      "learning_rate": 0.0003870204826308559,
+      "loss": 3.2446,
+      "step": 24873
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5433573722839355,
+      "learning_rate": 0.0003870057482703748,
+      "loss": 3.1398,
+      "step": 24874
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.574711263179779,
+      "learning_rate": 0.0003869910136807372,
+      "loss": 3.261,
+      "step": 24875
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5614281296730042,
+      "learning_rate": 0.00038697627886198204,
+      "loss": 3.2698,
+      "step": 24876
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5823600888252258,
+      "learning_rate": 0.0003869615438141481,
+      "loss": 2.9666,
+      "step": 24877
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5654398798942566,
+      "learning_rate": 0.00038694680853727423,
+      "loss": 3.2506,
+      "step": 24878
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6435055136680603,
+      "learning_rate": 0.0003869320730313991,
+      "loss": 3.0697,
+      "step": 24879
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5971924662590027,
+      "learning_rate": 0.0003869173372965618,
+      "loss": 3.1637,
+      "step": 24880
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5890087485313416,
+      "learning_rate": 0.0003869026013328008,
+      "loss": 3.1743,
+      "step": 24881
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5819125771522522,
+      "learning_rate": 0.0003868878651401551,
+      "loss": 3.0368,
+      "step": 24882
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5607448816299438,
+      "learning_rate": 0.00038687312871866353,
+      "loss": 3.2866,
+      "step": 24883
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5538012981414795,
+      "learning_rate": 0.0003868583920683648,
+      "loss": 2.9722,
+      "step": 24884
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5519068241119385,
+      "learning_rate": 0.00038684365518929783,
+      "loss": 3.1197,
+      "step": 24885
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5657366514205933,
+      "learning_rate": 0.00038682891808150126,
+      "loss": 3.3442,
+      "step": 24886
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5565481781959534,
+      "learning_rate": 0.00038681418074501425,
+      "loss": 3.1246,
+      "step": 24887
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5680059194564819,
+      "learning_rate": 0.00038679944317987524,
+      "loss": 2.9612,
+      "step": 24888
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5660066604614258,
+      "learning_rate": 0.00038678470538612325,
+      "loss": 3.0581,
+      "step": 24889
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.575242280960083,
+      "learning_rate": 0.00038676996736379705,
+      "loss": 3.0885,
+      "step": 24890
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5943603515625,
+      "learning_rate": 0.00038675522911293546,
+      "loss": 3.0541,
+      "step": 24891
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5383366942405701,
+      "learning_rate": 0.0003867404906335773,
+      "loss": 2.9924,
+      "step": 24892
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5978488326072693,
+      "learning_rate": 0.0003867257519257614,
+      "loss": 3.0037,
+      "step": 24893
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5685155987739563,
+      "learning_rate": 0.00038671101298952653,
+      "loss": 3.1197,
+      "step": 24894
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5263768434524536,
+      "learning_rate": 0.00038669627382491154,
+      "loss": 3.3945,
+      "step": 24895
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5680824518203735,
+      "learning_rate": 0.0003866815344319552,
+      "loss": 3.2136,
+      "step": 24896
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5891380906105042,
+      "learning_rate": 0.0003866667948106964,
+      "loss": 3.1585,
+      "step": 24897
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5633382797241211,
+      "learning_rate": 0.000386652054961174,
+      "loss": 3.1851,
+      "step": 24898
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5629472732543945,
+      "learning_rate": 0.00038663731488342674,
+      "loss": 3.1504,
+      "step": 24899
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5522317290306091,
+      "learning_rate": 0.0003866225745774935,
+      "loss": 3.1356,
+      "step": 24900
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5409548282623291,
+      "learning_rate": 0.0003866078340434129,
+      "loss": 3.1164,
+      "step": 24901
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5608346462249756,
+      "learning_rate": 0.0003865930932812241,
+      "loss": 2.968,
+      "step": 24902
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5439483523368835,
+      "learning_rate": 0.00038657835229096577,
+      "loss": 3.13,
+      "step": 24903
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5601171255111694,
+      "learning_rate": 0.00038656361107267665,
+      "loss": 3.0996,
+      "step": 24904
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5870618224143982,
+      "learning_rate": 0.00038654886962639565,
+      "loss": 3.0531,
+      "step": 24905
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6319909691810608,
+      "learning_rate": 0.0003865341279521617,
+      "loss": 3.1479,
+      "step": 24906
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6008051037788391,
+      "learning_rate": 0.00038651938605001324,
+      "loss": 2.8563,
+      "step": 24907
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.549253523349762,
+      "learning_rate": 0.0003865046439199895,
+      "loss": 3.1851,
+      "step": 24908
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5713079571723938,
+      "learning_rate": 0.0003864899015621293,
+      "loss": 3.2438,
+      "step": 24909
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.55234694480896,
+      "learning_rate": 0.00038647515897647126,
+      "loss": 3.1086,
+      "step": 24910
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5803124308586121,
+      "learning_rate": 0.00038646041616305425,
+      "loss": 2.9661,
+      "step": 24911
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5373254418373108,
+      "learning_rate": 0.0003864456731219172,
+      "loss": 3.077,
+      "step": 24912
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5681374669075012,
+      "learning_rate": 0.00038643092985309887,
+      "loss": 3.1995,
+      "step": 24913
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.586463451385498,
+      "learning_rate": 0.00038641618635663807,
+      "loss": 2.939,
+      "step": 24914
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5377286076545715,
+      "learning_rate": 0.0003864014426325737,
+      "loss": 3.2162,
+      "step": 24915
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.583235502243042,
+      "learning_rate": 0.00038638669868094454,
+      "loss": 3.0064,
+      "step": 24916
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5170654654502869,
+      "learning_rate": 0.00038637195450178945,
+      "loss": 3.0976,
+      "step": 24917
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5743283033370972,
+      "learning_rate": 0.00038635721009514727,
+      "loss": 3.1247,
+      "step": 24918
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5659081935882568,
+      "learning_rate": 0.00038634246546105676,
+      "loss": 3.0991,
+      "step": 24919
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5626477599143982,
+      "learning_rate": 0.0003863277205995569,
+      "loss": 2.9735,
+      "step": 24920
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.525495707988739,
+      "learning_rate": 0.00038631297551068643,
+      "loss": 2.8968,
+      "step": 24921
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5388118624687195,
+      "learning_rate": 0.00038629823019448406,
+      "loss": 2.9915,
+      "step": 24922
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5696624517440796,
+      "learning_rate": 0.00038628348465098895,
+      "loss": 3.2016,
+      "step": 24923
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5766279697418213,
+      "learning_rate": 0.00038626873888023965,
+      "loss": 3.1201,
+      "step": 24924
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5821237564086914,
+      "learning_rate": 0.00038625399288227507,
+      "loss": 2.9969,
+      "step": 24925
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.570102870464325,
+      "learning_rate": 0.0003862392466571341,
+      "loss": 2.9724,
+      "step": 24926
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5588696599006653,
+      "learning_rate": 0.00038622450020485557,
+      "loss": 2.8742,
+      "step": 24927
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5280874967575073,
+      "learning_rate": 0.0003862097535254783,
+      "loss": 3.2865,
+      "step": 24928
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5691091418266296,
+      "learning_rate": 0.0003861950066190411,
+      "loss": 3.1516,
+      "step": 24929
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5672391057014465,
+      "learning_rate": 0.00038618025948558283,
+      "loss": 3.0566,
+      "step": 24930
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5959036350250244,
+      "learning_rate": 0.0003861655121251425,
+      "loss": 2.8303,
+      "step": 24931
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.55399090051651,
+      "learning_rate": 0.00038615076453775855,
+      "loss": 3.0734,
+      "step": 24932
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5398240089416504,
+      "learning_rate": 0.0003861360167234702,
+      "loss": 2.8201,
+      "step": 24933
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5640655159950256,
+      "learning_rate": 0.00038612126868231615,
+      "loss": 3.159,
+      "step": 24934
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5160124897956848,
+      "learning_rate": 0.00038610652041433526,
+      "loss": 3.1671,
+      "step": 24935
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5614923238754272,
+      "learning_rate": 0.0003860917719195663,
+      "loss": 3.0478,
+      "step": 24936
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5773776173591614,
+      "learning_rate": 0.00038607702319804827,
+      "loss": 3.1768,
+      "step": 24937
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5507655143737793,
+      "learning_rate": 0.0003860622742498199,
+      "loss": 3.0052,
+      "step": 24938
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6294071078300476,
+      "learning_rate": 0.00038604752507492,
+      "loss": 2.8901,
+      "step": 24939
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5447244644165039,
+      "learning_rate": 0.0003860327756733876,
+      "loss": 3.0518,
+      "step": 24940
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5702586770057678,
+      "learning_rate": 0.0003860180260452613,
+      "loss": 2.9634,
+      "step": 24941
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5664628148078918,
+      "learning_rate": 0.00038600327619058017,
+      "loss": 2.7702,
+      "step": 24942
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6090885400772095,
+      "learning_rate": 0.00038598852610938287,
+      "loss": 2.9708,
+      "step": 24943
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5620157718658447,
+      "learning_rate": 0.00038597377580170837,
+      "loss": 3.2652,
+      "step": 24944
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5198007225990295,
+      "learning_rate": 0.0003859590252675956,
+      "loss": 3.0837,
+      "step": 24945
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5496712327003479,
+      "learning_rate": 0.0003859442745070832,
+      "loss": 3.1429,
+      "step": 24946
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5467877984046936,
+      "learning_rate": 0.0003859295235202101,
+      "loss": 3.1486,
+      "step": 24947
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7828726172447205,
+      "learning_rate": 0.0003859147723070151,
+      "loss": 3.0038,
+      "step": 24948
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5720941424369812,
+      "learning_rate": 0.00038590002086753735,
+      "loss": 3.0948,
+      "step": 24949
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6058550477027893,
+      "learning_rate": 0.0003858852692018154,
+      "loss": 3.1602,
+      "step": 24950
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5809235572814941,
+      "learning_rate": 0.0003858705173098881,
+      "loss": 3.0933,
+      "step": 24951
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5285111665725708,
+      "learning_rate": 0.0003858557651917944,
+      "loss": 3.0971,
+      "step": 24952
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5728558301925659,
+      "learning_rate": 0.0003858410128475731,
+      "loss": 3.2456,
+      "step": 24953
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.55547696352005,
+      "learning_rate": 0.0003858262602772632,
+      "loss": 3.1905,
+      "step": 24954
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5570668578147888,
+      "learning_rate": 0.0003858115074809034,
+      "loss": 3.0247,
+      "step": 24955
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5741897225379944,
+      "learning_rate": 0.00038579675445853264,
+      "loss": 3.1321,
+      "step": 24956
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.538557231426239,
+      "learning_rate": 0.00038578200121018976,
+      "loss": 3.133,
+      "step": 24957
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.576298713684082,
+      "learning_rate": 0.0003857672477359134,
+      "loss": 3.1654,
+      "step": 24958
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5457455515861511,
+      "learning_rate": 0.00038575249403574276,
+      "loss": 3.1689,
+      "step": 24959
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5845022201538086,
+      "learning_rate": 0.00038573774010971666,
+      "loss": 2.9617,
+      "step": 24960
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5689767599105835,
+      "learning_rate": 0.00038572298595787374,
+      "loss": 2.9591,
+      "step": 24961
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5592098832130432,
+      "learning_rate": 0.0003857082315802529,
+      "loss": 3.0949,
+      "step": 24962
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5083266496658325,
+      "learning_rate": 0.0003856934769768932,
+      "loss": 2.9243,
+      "step": 24963
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5481277108192444,
+      "learning_rate": 0.0003856787221478333,
+      "loss": 3.1659,
+      "step": 24964
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5553989410400391,
+      "learning_rate": 0.00038566396709311215,
+      "loss": 3.1719,
+      "step": 24965
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5781741738319397,
+      "learning_rate": 0.00038564921181276867,
+      "loss": 3.1653,
+      "step": 24966
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5346400141716003,
+      "learning_rate": 0.00038563445630684153,
+      "loss": 2.9946,
+      "step": 24967
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5499003529548645,
+      "learning_rate": 0.00038561970057536977,
+      "loss": 2.9757,
+      "step": 24968
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5649663209915161,
+      "learning_rate": 0.0003856049446183922,
+      "loss": 3.0766,
+      "step": 24969
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5463762283325195,
+      "learning_rate": 0.0003855901884359476,
+      "loss": 2.9893,
+      "step": 24970
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5486137270927429,
+      "learning_rate": 0.00038557543202807513,
+      "loss": 3.1587,
+      "step": 24971
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5812304019927979,
+      "learning_rate": 0.0003855606753948133,
+      "loss": 3.0608,
+      "step": 24972
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6184636354446411,
+      "learning_rate": 0.000385545918536201,
+      "loss": 3.0295,
+      "step": 24973
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5484544038772583,
+      "learning_rate": 0.00038553116145227743,
+      "loss": 3.0903,
+      "step": 24974
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5901756882667542,
+      "learning_rate": 0.0003855164041430811,
+      "loss": 3.1048,
+      "step": 24975
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5762054920196533,
+      "learning_rate": 0.0003855016466086511,
+      "loss": 3.1637,
+      "step": 24976
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5577733516693115,
+      "learning_rate": 0.00038548688884902616,
+      "loss": 3.1275,
+      "step": 24977
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5851019620895386,
+      "learning_rate": 0.00038547213086424526,
+      "loss": 3.3804,
+      "step": 24978
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5264205932617188,
+      "learning_rate": 0.0003854573726543472,
+      "loss": 2.9034,
+      "step": 24979
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.529139518737793,
+      "learning_rate": 0.0003854426142193708,
+      "loss": 3.13,
+      "step": 24980
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5476992726325989,
+      "learning_rate": 0.000385427855559355,
+      "loss": 3.1176,
+      "step": 24981
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5486899614334106,
+      "learning_rate": 0.0003854130966743388,
+      "loss": 3.0411,
+      "step": 24982
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5495012998580933,
+      "learning_rate": 0.00038539833756436086,
+      "loss": 3.0098,
+      "step": 24983
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5303811430931091,
+      "learning_rate": 0.0003853835782294601,
+      "loss": 3.1033,
+      "step": 24984
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5448635220527649,
+      "learning_rate": 0.00038536881866967554,
+      "loss": 2.9309,
+      "step": 24985
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5653622150421143,
+      "learning_rate": 0.0003853540588850459,
+      "loss": 3.0957,
+      "step": 24986
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6244167685508728,
+      "learning_rate": 0.00038533929887561003,
+      "loss": 3.134,
+      "step": 24987
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5666537284851074,
+      "learning_rate": 0.00038532453864140685,
+      "loss": 3.1218,
+      "step": 24988
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5365763306617737,
+      "learning_rate": 0.00038530977818247535,
+      "loss": 2.8336,
+      "step": 24989
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5451518297195435,
+      "learning_rate": 0.0003852950174988542,
+      "loss": 2.8832,
+      "step": 24990
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5833927989006042,
+      "learning_rate": 0.00038528025659058255,
+      "loss": 2.9466,
+      "step": 24991
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5745109915733337,
+      "learning_rate": 0.000385265495457699,
+      "loss": 3.0929,
+      "step": 24992
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5827248096466064,
+      "learning_rate": 0.00038525073410024263,
+      "loss": 3.0634,
+      "step": 24993
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.544745922088623,
+      "learning_rate": 0.00038523597251825207,
+      "loss": 3.1738,
+      "step": 24994
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5738770961761475,
+      "learning_rate": 0.0003852212107117665,
+      "loss": 2.9558,
+      "step": 24995
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.631645143032074,
+      "learning_rate": 0.0003852064486808247,
+      "loss": 2.8622,
+      "step": 24996
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5620419979095459,
+      "learning_rate": 0.0003851916864254654,
+      "loss": 2.6171,
+      "step": 24997
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5576983094215393,
+      "learning_rate": 0.0003851769239457277,
+      "loss": 3.1946,
+      "step": 24998
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6105443835258484,
+      "learning_rate": 0.0003851621612416502,
+      "loss": 3.0174,
+      "step": 24999
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5353231430053711,
+      "learning_rate": 0.00038514739831327214,
+      "loss": 2.901,
+      "step": 25000
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.600609540939331,
+      "learning_rate": 0.0003851326351606321,
+      "loss": 3.0875,
+      "step": 25001
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6092653274536133,
+      "learning_rate": 0.0003851178717837691,
+      "loss": 3.0929,
+      "step": 25002
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5666999816894531,
+      "learning_rate": 0.00038510310818272203,
+      "loss": 2.9879,
+      "step": 25003
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5675250291824341,
+      "learning_rate": 0.0003850883443575298,
+      "loss": 3.2092,
+      "step": 25004
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5591527223587036,
+      "learning_rate": 0.0003850735803082312,
+      "loss": 3.0273,
+      "step": 25005
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5454131364822388,
+      "learning_rate": 0.0003850588160348652,
+      "loss": 3.1456,
+      "step": 25006
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5496153235435486,
+      "learning_rate": 0.00038504405153747066,
+      "loss": 2.9497,
+      "step": 25007
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5893572568893433,
+      "learning_rate": 0.0003850292868160864,
+      "loss": 3.2013,
+      "step": 25008
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6619960069656372,
+      "learning_rate": 0.0003850145218707513,
+      "loss": 3.0674,
+      "step": 25009
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5621336102485657,
+      "learning_rate": 0.00038499975670150437,
+      "loss": 3.1975,
+      "step": 25010
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5568299889564514,
+      "learning_rate": 0.00038498499130838455,
+      "loss": 3.166,
+      "step": 25011
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5435270071029663,
+      "learning_rate": 0.0003849702256914305,
+      "loss": 3.067,
+      "step": 25012
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5628631114959717,
+      "learning_rate": 0.0003849554598506812,
+      "loss": 3.017,
+      "step": 25013
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5889310240745544,
+      "learning_rate": 0.00038494069378617566,
+      "loss": 3.1861,
+      "step": 25014
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5779412984848022,
+      "learning_rate": 0.0003849259274979526,
+      "loss": 3.0177,
+      "step": 25015
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5566220283508301,
+      "learning_rate": 0.00038491116098605103,
+      "loss": 3.0443,
+      "step": 25016
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5883911848068237,
+      "learning_rate": 0.0003848963942505098,
+      "loss": 2.8931,
+      "step": 25017
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5884630084037781,
+      "learning_rate": 0.0003848816272913678,
+      "loss": 2.978,
+      "step": 25018
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5841085910797119,
+      "learning_rate": 0.000384866860108664,
+      "loss": 2.8837,
+      "step": 25019
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5884820818901062,
+      "learning_rate": 0.00038485209270243707,
+      "loss": 3.0159,
+      "step": 25020
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6222401857376099,
+      "learning_rate": 0.0003848373250727262,
+      "loss": 3.1475,
+      "step": 25021
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6234578490257263,
+      "learning_rate": 0.0003848225572195702,
+      "loss": 3.2096,
+      "step": 25022
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6008965373039246,
+      "learning_rate": 0.00038480778914300775,
+      "loss": 3.2338,
+      "step": 25023
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5736437439918518,
+      "learning_rate": 0.000384793020843078,
+      "loss": 3.1751,
+      "step": 25024
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5647448301315308,
+      "learning_rate": 0.00038477825231981967,
+      "loss": 3.1197,
+      "step": 25025
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5257458686828613,
+      "learning_rate": 0.00038476348357327175,
+      "loss": 3.2591,
+      "step": 25026
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.553209662437439,
+      "learning_rate": 0.00038474871460347315,
+      "loss": 3.3708,
+      "step": 25027
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6161184310913086,
+      "learning_rate": 0.00038473394541046276,
+      "loss": 2.9489,
+      "step": 25028
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5613665580749512,
+      "learning_rate": 0.0003847191759942794,
+      "loss": 2.9919,
+      "step": 25029
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5895677804946899,
+      "learning_rate": 0.00038470440635496214,
+      "loss": 2.8271,
+      "step": 25030
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5605174899101257,
+      "learning_rate": 0.00038468963649254976,
+      "loss": 3.0974,
+      "step": 25031
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5734592080116272,
+      "learning_rate": 0.00038467486640708106,
+      "loss": 2.9672,
+      "step": 25032
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6092544794082642,
+      "learning_rate": 0.0003846600960985953,
+      "loss": 3.1621,
+      "step": 25033
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5776329636573792,
+      "learning_rate": 0.0003846453255671309,
+      "loss": 3.0484,
+      "step": 25034
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.555668294429779,
+      "learning_rate": 0.000384630554812727,
+      "loss": 2.8431,
+      "step": 25035
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5371507406234741,
+      "learning_rate": 0.00038461578383542266,
+      "loss": 3.2272,
+      "step": 25036
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5646740198135376,
+      "learning_rate": 0.00038460101263525654,
+      "loss": 3.1364,
+      "step": 25037
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5981430411338806,
+      "learning_rate": 0.0003845862412122677,
+      "loss": 3.0595,
+      "step": 25038
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5709407329559326,
+      "learning_rate": 0.00038457146956649496,
+      "loss": 3.3635,
+      "step": 25039
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5592368245124817,
+      "learning_rate": 0.00038455669769797723,
+      "loss": 3.0549,
+      "step": 25040
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5950730443000793,
+      "learning_rate": 0.00038454192560675337,
+      "loss": 3.0904,
+      "step": 25041
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5664187073707581,
+      "learning_rate": 0.00038452715329286236,
+      "loss": 3.2247,
+      "step": 25042
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5902990102767944,
+      "learning_rate": 0.00038451238075634317,
+      "loss": 2.9901,
+      "step": 25043
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5837813019752502,
+      "learning_rate": 0.00038449760799723467,
+      "loss": 3.281,
+      "step": 25044
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5957124829292297,
+      "learning_rate": 0.0003844828350155756,
+      "loss": 3.1995,
+      "step": 25045
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5741605758666992,
+      "learning_rate": 0.000384468061811405,
+      "loss": 3.0371,
+      "step": 25046
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5534799098968506,
+      "learning_rate": 0.00038445328838476197,
+      "loss": 3.2169,
+      "step": 25047
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5551855564117432,
+      "learning_rate": 0.00038443851473568503,
+      "loss": 2.9418,
+      "step": 25048
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5379355549812317,
+      "learning_rate": 0.0003844237408642134,
+      "loss": 3.0535,
+      "step": 25049
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5475854873657227,
+      "learning_rate": 0.0003844089667703858,
+      "loss": 2.9308,
+      "step": 25050
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5689952969551086,
+      "learning_rate": 0.0003843941924542413,
+      "loss": 2.939,
+      "step": 25051
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5918108224868774,
+      "learning_rate": 0.0003843794179158187,
+      "loss": 2.8464,
+      "step": 25052
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5874655842781067,
+      "learning_rate": 0.000384364643155157,
+      "loss": 3.1436,
+      "step": 25053
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.535483181476593,
+      "learning_rate": 0.00038434986817229504,
+      "loss": 2.867,
+      "step": 25054
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6092949509620667,
+      "learning_rate": 0.0003843350929672719,
+      "loss": 3.1436,
+      "step": 25055
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5654955506324768,
+      "learning_rate": 0.0003843203175401261,
+      "loss": 3.0727,
+      "step": 25056
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5241906642913818,
+      "learning_rate": 0.0003843055418908969,
+      "loss": 2.7895,
+      "step": 25057
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.577303409576416,
+      "learning_rate": 0.0003842907660196232,
+      "loss": 2.9957,
+      "step": 25058
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5571273565292358,
+      "learning_rate": 0.0003842759899263438,
+      "loss": 3.1266,
+      "step": 25059
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5683521628379822,
+      "learning_rate": 0.0003842612136110976,
+      "loss": 2.9504,
+      "step": 25060
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5134350657463074,
+      "learning_rate": 0.00038424643707392364,
+      "loss": 3.295,
+      "step": 25061
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6152227520942688,
+      "learning_rate": 0.0003842316603148608,
+      "loss": 2.764,
+      "step": 25062
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5723081231117249,
+      "learning_rate": 0.000384216883333948,
+      "loss": 3.0838,
+      "step": 25063
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5563600659370422,
+      "learning_rate": 0.000384202106131224,
+      "loss": 3.2504,
+      "step": 25064
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5456048250198364,
+      "learning_rate": 0.00038418732870672804,
+      "loss": 2.9962,
+      "step": 25065
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5585772395133972,
+      "learning_rate": 0.0003841725510604987,
+      "loss": 2.8871,
+      "step": 25066
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5854873061180115,
+      "learning_rate": 0.00038415777319257517,
+      "loss": 2.9557,
+      "step": 25067
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5794108510017395,
+      "learning_rate": 0.0003841429951029962,
+      "loss": 3.1969,
+      "step": 25068
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.688174307346344,
+      "learning_rate": 0.0003841282167918008,
+      "loss": 3.1212,
+      "step": 25069
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5816777944564819,
+      "learning_rate": 0.0003841134382590279,
+      "loss": 3.0208,
+      "step": 25070
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5570030808448792,
+      "learning_rate": 0.00038409865950471634,
+      "loss": 3.1805,
+      "step": 25071
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5498480796813965,
+      "learning_rate": 0.00038408388052890507,
+      "loss": 3.0436,
+      "step": 25072
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5754635334014893,
+      "learning_rate": 0.0003840691013316331,
+      "loss": 3.1856,
+      "step": 25073
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5430830121040344,
+      "learning_rate": 0.0003840543219129393,
+      "loss": 2.9776,
+      "step": 25074
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5669942498207092,
+      "learning_rate": 0.0003840395422728626,
+      "loss": 3.1792,
+      "step": 25075
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5448963046073914,
+      "learning_rate": 0.0003840247624114418,
+      "loss": 3.0217,
+      "step": 25076
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5780636072158813,
+      "learning_rate": 0.0003840099823287161,
+      "loss": 3.0021,
+      "step": 25077
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5799130201339722,
+      "learning_rate": 0.00038399520202472414,
+      "loss": 2.9277,
+      "step": 25078
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5543425679206848,
+      "learning_rate": 0.0003839804214995051,
+      "loss": 3.0757,
+      "step": 25079
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5854793787002563,
+      "learning_rate": 0.0003839656407530977,
+      "loss": 3.0283,
+      "step": 25080
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5901440978050232,
+      "learning_rate": 0.00038395085978554103,
+      "loss": 3.0345,
+      "step": 25081
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6366936564445496,
+      "learning_rate": 0.00038393607859687394,
+      "loss": 3.1309,
+      "step": 25082
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.554716169834137,
+      "learning_rate": 0.0003839212971871354,
+      "loss": 3.1066,
+      "step": 25083
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5704500675201416,
+      "learning_rate": 0.0003839065155563643,
+      "loss": 2.9534,
+      "step": 25084
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5622331500053406,
+      "learning_rate": 0.00038389173370459955,
+      "loss": 3.0545,
+      "step": 25085
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5637291073799133,
+      "learning_rate": 0.0003838769516318801,
+      "loss": 3.018,
+      "step": 25086
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5698617696762085,
+      "learning_rate": 0.000383862169338245,
+      "loss": 3.1951,
+      "step": 25087
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5883769392967224,
+      "learning_rate": 0.00038384738682373297,
+      "loss": 3.2073,
+      "step": 25088
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.54131680727005,
+      "learning_rate": 0.0003838326040883831,
+      "loss": 3.1775,
+      "step": 25089
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5682424306869507,
+      "learning_rate": 0.00038381782113223426,
+      "loss": 3.1067,
+      "step": 25090
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5473423004150391,
+      "learning_rate": 0.0003838030379553255,
+      "loss": 2.8472,
+      "step": 25091
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5303641557693481,
+      "learning_rate": 0.0003837882545576956,
+      "loss": 3.1275,
+      "step": 25092
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5696659684181213,
+      "learning_rate": 0.00038377347093938353,
+      "loss": 3.0899,
+      "step": 25093
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5956259965896606,
+      "learning_rate": 0.00038375868710042836,
+      "loss": 3.0071,
+      "step": 25094
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5872520208358765,
+      "learning_rate": 0.00038374390304086884,
+      "loss": 3.1197,
+      "step": 25095
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5226383805274963,
+      "learning_rate": 0.00038372911876074405,
+      "loss": 3.1719,
+      "step": 25096
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5661742687225342,
+      "learning_rate": 0.00038371433426009276,
+      "loss": 2.9626,
+      "step": 25097
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5806180834770203,
+      "learning_rate": 0.0003836995495389542,
+      "loss": 3.1586,
+      "step": 25098
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5616850256919861,
+      "learning_rate": 0.00038368476459736705,
+      "loss": 3.0818,
+      "step": 25099
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5661494731903076,
+      "learning_rate": 0.0003836699794353703,
+      "loss": 3.0206,
+      "step": 25100
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5832769274711609,
+      "learning_rate": 0.000383655194053003,
+      "loss": 2.8287,
+      "step": 25101
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5530614256858826,
+      "learning_rate": 0.000383640408450304,
+      "loss": 3.0402,
+      "step": 25102
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5663602352142334,
+      "learning_rate": 0.00038362562262731224,
+      "loss": 3.2647,
+      "step": 25103
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5610358119010925,
+      "learning_rate": 0.00038361083658406666,
+      "loss": 3.0342,
+      "step": 25104
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5962334275245667,
+      "learning_rate": 0.0003835960503206063,
+      "loss": 3.039,
+      "step": 25105
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5851008296012878,
+      "learning_rate": 0.00038358126383697,
+      "loss": 2.7736,
+      "step": 25106
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6189122796058655,
+      "learning_rate": 0.0003835664771331967,
+      "loss": 2.9924,
+      "step": 25107
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5865572094917297,
+      "learning_rate": 0.00038355169020932534,
+      "loss": 3.1454,
+      "step": 25108
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5793099403381348,
+      "learning_rate": 0.00038353690306539506,
+      "loss": 3.0783,
+      "step": 25109
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.577268660068512,
+      "learning_rate": 0.00038352211570144454,
+      "loss": 3.1842,
+      "step": 25110
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.524253249168396,
+      "learning_rate": 0.0003835073281175129,
+      "loss": 3.1753,
+      "step": 25111
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.553202748298645,
+      "learning_rate": 0.0003834925403136389,
+      "loss": 2.9092,
+      "step": 25112
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5760665535926819,
+      "learning_rate": 0.0003834777522898618,
+      "loss": 3.1152,
+      "step": 25113
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5376695990562439,
+      "learning_rate": 0.00038346296404622024,
+      "loss": 3.0326,
+      "step": 25114
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5973801612854004,
+      "learning_rate": 0.00038344817558275335,
+      "loss": 3.1555,
+      "step": 25115
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5529061555862427,
+      "learning_rate": 0.00038343338689950004,
+      "loss": 3.1566,
+      "step": 25116
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6026521325111389,
+      "learning_rate": 0.0003834185979964992,
+      "loss": 3.1871,
+      "step": 25117
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5608975887298584,
+      "learning_rate": 0.0003834038088737898,
+      "loss": 2.9394,
+      "step": 25118
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.530158519744873,
+      "learning_rate": 0.00038338901953141087,
+      "loss": 3.0237,
+      "step": 25119
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5642078518867493,
+      "learning_rate": 0.0003833742299694014,
+      "loss": 3.2089,
+      "step": 25120
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.526268720626831,
+      "learning_rate": 0.0003833594401878001,
+      "loss": 3.1074,
+      "step": 25121
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5271223187446594,
+      "learning_rate": 0.00038334465018664615,
+      "loss": 3.1286,
+      "step": 25122
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5779740214347839,
+      "learning_rate": 0.0003833298599659783,
+      "loss": 3.0723,
+      "step": 25123
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5797901749610901,
+      "learning_rate": 0.00038331506952583584,
+      "loss": 3.1721,
+      "step": 25124
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5432953834533691,
+      "learning_rate": 0.00038330027886625744,
+      "loss": 2.9935,
+      "step": 25125
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5544005632400513,
+      "learning_rate": 0.0003832854879872821,
+      "loss": 2.8895,
+      "step": 25126
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5620940327644348,
+      "learning_rate": 0.00038327069688894885,
+      "loss": 3.053,
+      "step": 25127
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5733068585395813,
+      "learning_rate": 0.00038325590557129665,
+      "loss": 2.9839,
+      "step": 25128
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5551736354827881,
+      "learning_rate": 0.00038324111403436435,
+      "loss": 3.268,
+      "step": 25129
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.584686279296875,
+      "learning_rate": 0.00038322632227819097,
+      "loss": 3.2248,
+      "step": 25130
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6471843123435974,
+      "learning_rate": 0.00038321153030281555,
+      "loss": 3.0796,
+      "step": 25131
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5832377672195435,
+      "learning_rate": 0.000383196738108277,
+      "loss": 3.0214,
+      "step": 25132
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5637081861495972,
+      "learning_rate": 0.0003831819456946141,
+      "loss": 3.1815,
+      "step": 25133
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5314174890518188,
+      "learning_rate": 0.00038316715306186604,
+      "loss": 2.947,
+      "step": 25134
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.564139187335968,
+      "learning_rate": 0.0003831523602100717,
+      "loss": 3.0108,
+      "step": 25135
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5853064656257629,
+      "learning_rate": 0.00038313756713927004,
+      "loss": 2.963,
+      "step": 25136
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5824783444404602,
+      "learning_rate": 0.0003831227738495001,
+      "loss": 3.0401,
+      "step": 25137
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.590248703956604,
+      "learning_rate": 0.0003831079803408007,
+      "loss": 3.1766,
+      "step": 25138
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5577486157417297,
+      "learning_rate": 0.0003830931866132109,
+      "loss": 2.9757,
+      "step": 25139
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5475689768791199,
+      "learning_rate": 0.0003830783926667696,
+      "loss": 3.2508,
+      "step": 25140
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5604707598686218,
+      "learning_rate": 0.0003830635985015158,
+      "loss": 2.9826,
+      "step": 25141
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5635944604873657,
+      "learning_rate": 0.0003830488041174885,
+      "loss": 3.0229,
+      "step": 25142
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5407227873802185,
+      "learning_rate": 0.00038303400951472666,
+      "loss": 2.8838,
+      "step": 25143
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5590699315071106,
+      "learning_rate": 0.0003830192146932692,
+      "loss": 3.2534,
+      "step": 25144
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5566086769104004,
+      "learning_rate": 0.00038300441965315505,
+      "loss": 2.9712,
+      "step": 25145
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5474108457565308,
+      "learning_rate": 0.0003829896243944234,
+      "loss": 3.1384,
+      "step": 25146
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5965452790260315,
+      "learning_rate": 0.00038297482891711286,
+      "loss": 3.0701,
+      "step": 25147
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5965765714645386,
+      "learning_rate": 0.0003829600332212626,
+      "loss": 3.038,
+      "step": 25148
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.547390878200531,
+      "learning_rate": 0.00038294523730691174,
+      "loss": 2.8804,
+      "step": 25149
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5904244184494019,
+      "learning_rate": 0.00038293044117409896,
+      "loss": 2.9532,
+      "step": 25150
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5980238914489746,
+      "learning_rate": 0.00038291564482286335,
+      "loss": 3.1317,
+      "step": 25151
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6068405508995056,
+      "learning_rate": 0.00038290084825324394,
+      "loss": 3.1838,
+      "step": 25152
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5584993362426758,
+      "learning_rate": 0.00038288605146527966,
+      "loss": 3.1354,
+      "step": 25153
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5200023651123047,
+      "learning_rate": 0.00038287125445900943,
+      "loss": 3.1367,
+      "step": 25154
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5400229096412659,
+      "learning_rate": 0.00038285645723447223,
+      "loss": 3.1147,
+      "step": 25155
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6204379200935364,
+      "learning_rate": 0.0003828416597917071,
+      "loss": 3.0594,
+      "step": 25156
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5439574122428894,
+      "learning_rate": 0.0003828268621307531,
+      "loss": 3.294,
+      "step": 25157
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5771875381469727,
+      "learning_rate": 0.00038281206425164893,
+      "loss": 3.033,
+      "step": 25158
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5748353004455566,
+      "learning_rate": 0.0003827972661544337,
+      "loss": 3.3069,
+      "step": 25159
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5352070331573486,
+      "learning_rate": 0.00038278246783914657,
+      "loss": 3.2755,
+      "step": 25160
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.533769965171814,
+      "learning_rate": 0.00038276766930582626,
+      "loss": 3.2693,
+      "step": 25161
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5342288017272949,
+      "learning_rate": 0.0003827528705545118,
+      "loss": 3.1622,
+      "step": 25162
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5443651676177979,
+      "learning_rate": 0.00038273807158524216,
+      "loss": 3.1329,
+      "step": 25163
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5513403415679932,
+      "learning_rate": 0.0003827232723980566,
+      "loss": 3.0414,
+      "step": 25164
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5552654266357422,
+      "learning_rate": 0.00038270847299299367,
+      "loss": 3.0813,
+      "step": 25165
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5734013319015503,
+      "learning_rate": 0.00038269367337009257,
+      "loss": 3.0957,
+      "step": 25166
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6037915349006653,
+      "learning_rate": 0.0003826788735293923,
+      "loss": 3.2297,
+      "step": 25167
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5523087382316589,
+      "learning_rate": 0.0003826640734709317,
+      "loss": 2.9195,
+      "step": 25168
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5460664629936218,
+      "learning_rate": 0.00038264927319474985,
+      "loss": 3.202,
+      "step": 25169
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5905874967575073,
+      "learning_rate": 0.0003826344727008858,
+      "loss": 2.9142,
+      "step": 25170
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5918222665786743,
+      "learning_rate": 0.0003826196719893785,
+      "loss": 3.0206,
+      "step": 25171
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.539842426776886,
+      "learning_rate": 0.00038260487106026676,
+      "loss": 3.134,
+      "step": 25172
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5324265956878662,
+      "learning_rate": 0.00038259006991358966,
+      "loss": 3.1243,
+      "step": 25173
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5232551097869873,
+      "learning_rate": 0.00038257526854938625,
+      "loss": 3.4041,
+      "step": 25174
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5780461430549622,
+      "learning_rate": 0.0003825604669676955,
+      "loss": 3.1711,
+      "step": 25175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5784001350402832,
+      "learning_rate": 0.00038254566516855633,
+      "loss": 3.0586,
+      "step": 25176
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5395734310150146,
+      "learning_rate": 0.0003825308631520079,
+      "loss": 3.1729,
+      "step": 25177
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.57415372133255,
+      "learning_rate": 0.0003825160609180889,
+      "loss": 2.9213,
+      "step": 25178
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5726572275161743,
+      "learning_rate": 0.0003825012584668385,
+      "loss": 3.1291,
+      "step": 25179
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.571429431438446,
+      "learning_rate": 0.00038248645579829567,
+      "loss": 2.9315,
+      "step": 25180
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5924422144889832,
+      "learning_rate": 0.0003824716529124994,
+      "loss": 3.3614,
+      "step": 25181
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5415179133415222,
+      "learning_rate": 0.0003824568498094887,
+      "loss": 3.1108,
+      "step": 25182
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5803284645080566,
+      "learning_rate": 0.0003824420464893026,
+      "loss": 3.2433,
+      "step": 25183
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5539078712463379,
+      "learning_rate": 0.00038242724295197984,
+      "loss": 2.8731,
+      "step": 25184
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.549834668636322,
+      "learning_rate": 0.00038241243919755967,
+      "loss": 3.0925,
+      "step": 25185
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5797627568244934,
+      "learning_rate": 0.0003823976352260811,
+      "loss": 3.2575,
+      "step": 25186
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.57608562707901,
+      "learning_rate": 0.0003823828310375829,
+      "loss": 2.9751,
+      "step": 25187
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.587173342704773,
+      "learning_rate": 0.0003823680266321042,
+      "loss": 3.1131,
+      "step": 25188
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5533391833305359,
+      "learning_rate": 0.00038235322200968396,
+      "loss": 3.2133,
+      "step": 25189
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5643752813339233,
+      "learning_rate": 0.00038233841717036116,
+      "loss": 3.2175,
+      "step": 25190
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5208031535148621,
+      "learning_rate": 0.00038232361211417484,
+      "loss": 3.1639,
+      "step": 25191
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5685301423072815,
+      "learning_rate": 0.00038230880684116396,
+      "loss": 3.0885,
+      "step": 25192
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5917596817016602,
+      "learning_rate": 0.0003822940013513675,
+      "loss": 2.8533,
+      "step": 25193
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5756223797798157,
+      "learning_rate": 0.0003822791956448246,
+      "loss": 3.0704,
+      "step": 25194
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5527355074882507,
+      "learning_rate": 0.0003822643897215739,
+      "loss": 2.929,
+      "step": 25195
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6033982634544373,
+      "learning_rate": 0.0003822495835816548,
+      "loss": 3.028,
+      "step": 25196
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5483729243278503,
+      "learning_rate": 0.00038223477722510623,
+      "loss": 3.1289,
+      "step": 25197
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5902836918830872,
+      "learning_rate": 0.0003822199706519669,
+      "loss": 3.0741,
+      "step": 25198
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5841925740242004,
+      "learning_rate": 0.000382205163862276,
+      "loss": 3.0422,
+      "step": 25199
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5686092972755432,
+      "learning_rate": 0.0003821903568560726,
+      "loss": 3.0314,
+      "step": 25200
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5432037115097046,
+      "learning_rate": 0.0003821755496333956,
+      "loss": 3.2488,
+      "step": 25201
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.546054482460022,
+      "learning_rate": 0.000382160742194284,
+      "loss": 2.9305,
+      "step": 25202
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5392909049987793,
+      "learning_rate": 0.0003821459345387768,
+      "loss": 3.0723,
+      "step": 25203
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5378334522247314,
+      "learning_rate": 0.00038213112666691303,
+      "loss": 2.9682,
+      "step": 25204
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5470960736274719,
+      "learning_rate": 0.0003821163185787317,
+      "loss": 3.1647,
+      "step": 25205
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5794514417648315,
+      "learning_rate": 0.00038210151027427176,
+      "loss": 3.0332,
+      "step": 25206
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.611207127571106,
+      "learning_rate": 0.00038208670175357226,
+      "loss": 3.2109,
+      "step": 25207
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5396631956100464,
+      "learning_rate": 0.00038207189301667227,
+      "loss": 3.1659,
+      "step": 25208
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5904266238212585,
+      "learning_rate": 0.00038205708406361056,
+      "loss": 3.0525,
+      "step": 25209
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5775328278541565,
+      "learning_rate": 0.0003820422748944263,
+      "loss": 3.0498,
+      "step": 25210
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5564374327659607,
+      "learning_rate": 0.00038202746550915857,
+      "loss": 3.1151,
+      "step": 25211
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5537264943122864,
+      "learning_rate": 0.00038201265590784625,
+      "loss": 3.2185,
+      "step": 25212
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5900477170944214,
+      "learning_rate": 0.00038199784609052834,
+      "loss": 3.1797,
+      "step": 25213
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5341221690177917,
+      "learning_rate": 0.00038198303605724385,
+      "loss": 3.1518,
+      "step": 25214
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5363330245018005,
+      "learning_rate": 0.00038196822580803194,
+      "loss": 2.8538,
+      "step": 25215
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5329729914665222,
+      "learning_rate": 0.0003819534153429314,
+      "loss": 3.1315,
+      "step": 25216
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5687403082847595,
+      "learning_rate": 0.0003819386046619814,
+      "loss": 3.1472,
+      "step": 25217
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5683347582817078,
+      "learning_rate": 0.00038192379376522087,
+      "loss": 3.2782,
+      "step": 25218
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5241825580596924,
+      "learning_rate": 0.0003819089826526889,
+      "loss": 2.9968,
+      "step": 25219
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5936827659606934,
+      "learning_rate": 0.00038189417132442424,
+      "loss": 2.9825,
+      "step": 25220
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5642827749252319,
+      "learning_rate": 0.0003818793597804662,
+      "loss": 2.9358,
+      "step": 25221
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5551975965499878,
+      "learning_rate": 0.00038186454802085373,
+      "loss": 3.0648,
+      "step": 25222
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5875920057296753,
+      "learning_rate": 0.00038184973604562573,
+      "loss": 3.0493,
+      "step": 25223
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5492988228797913,
+      "learning_rate": 0.00038183492385482137,
+      "loss": 3.1779,
+      "step": 25224
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5400311350822449,
+      "learning_rate": 0.0003818201114484794,
+      "loss": 3.2273,
+      "step": 25225
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5957015156745911,
+      "learning_rate": 0.00038180529882663914,
+      "loss": 3.1921,
+      "step": 25226
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5321521162986755,
+      "learning_rate": 0.00038179048598933944,
+      "loss": 3.0101,
+      "step": 25227
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5768882036209106,
+      "learning_rate": 0.00038177567293661935,
+      "loss": 2.9978,
+      "step": 25228
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5608060359954834,
+      "learning_rate": 0.0003817608596685179,
+      "loss": 3.0694,
+      "step": 25229
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5481438040733337,
+      "learning_rate": 0.00038174604618507397,
+      "loss": 3.0578,
+      "step": 25230
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.570463240146637,
+      "learning_rate": 0.00038173123248632675,
+      "loss": 3.1923,
+      "step": 25231
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6016967296600342,
+      "learning_rate": 0.0003817164185723152,
+      "loss": 3.0681,
+      "step": 25232
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5828872919082642,
+      "learning_rate": 0.00038170160444307835,
+      "loss": 3.0874,
+      "step": 25233
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5865387916564941,
+      "learning_rate": 0.00038168679009865523,
+      "loss": 3.0659,
+      "step": 25234
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5488494038581848,
+      "learning_rate": 0.00038167197553908467,
+      "loss": 2.983,
+      "step": 25235
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5348222255706787,
+      "learning_rate": 0.00038165716076440595,
+      "loss": 3.0415,
+      "step": 25236
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5851247906684875,
+      "learning_rate": 0.0003816423457746581,
+      "loss": 3.0879,
+      "step": 25237
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6187082529067993,
+      "learning_rate": 0.00038162753056987986,
+      "loss": 3.1523,
+      "step": 25238
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5365039110183716,
+      "learning_rate": 0.00038161271515011044,
+      "loss": 3.1376,
+      "step": 25239
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5412645936012268,
+      "learning_rate": 0.00038159789951538884,
+      "loss": 2.9421,
+      "step": 25240
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5671802759170532,
+      "learning_rate": 0.00038158308366575413,
+      "loss": 3.0663,
+      "step": 25241
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5550203919410706,
+      "learning_rate": 0.00038156826760124524,
+      "loss": 2.914,
+      "step": 25242
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5574257969856262,
+      "learning_rate": 0.0003815534513219012,
+      "loss": 2.9547,
+      "step": 25243
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5445705652236938,
+      "learning_rate": 0.0003815386348277611,
+      "loss": 3.0675,
+      "step": 25244
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5782180428504944,
+      "learning_rate": 0.00038152381811886393,
+      "loss": 3.0991,
+      "step": 25245
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5452681183815002,
+      "learning_rate": 0.0003815090011952487,
+      "loss": 2.943,
+      "step": 25246
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5488862991333008,
+      "learning_rate": 0.00038149418405695443,
+      "loss": 2.8228,
+      "step": 25247
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5082352161407471,
+      "learning_rate": 0.0003814793667040203,
+      "loss": 3.0863,
+      "step": 25248
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5326624512672424,
+      "learning_rate": 0.000381464549136485,
+      "loss": 3.1076,
+      "step": 25249
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.568250834941864,
+      "learning_rate": 0.0003814497313543879,
+      "loss": 3.1315,
+      "step": 25250
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5677963495254517,
+      "learning_rate": 0.00038143491335776783,
+      "loss": 3.137,
+      "step": 25251
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5616076588630676,
+      "learning_rate": 0.0003814200951466638,
+      "loss": 3.0442,
+      "step": 25252
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5570437908172607,
+      "learning_rate": 0.00038140527672111496,
+      "loss": 3.0987,
+      "step": 25253
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5529975891113281,
+      "learning_rate": 0.00038139045808116036,
+      "loss": 3.2037,
+      "step": 25254
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6022285223007202,
+      "learning_rate": 0.0003813756392268389,
+      "loss": 3.0878,
+      "step": 25255
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5654017925262451,
+      "learning_rate": 0.00038136082015818965,
+      "loss": 3.1599,
+      "step": 25256
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5939410924911499,
+      "learning_rate": 0.00038134600087525166,
+      "loss": 3.2432,
+      "step": 25257
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5679937601089478,
+      "learning_rate": 0.000381331181378064,
+      "loss": 2.9138,
+      "step": 25258
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.584583580493927,
+      "learning_rate": 0.00038131636166666567,
+      "loss": 3.2667,
+      "step": 25259
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5996798872947693,
+      "learning_rate": 0.0003813015417410957,
+      "loss": 3.2236,
+      "step": 25260
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5446107387542725,
+      "learning_rate": 0.00038128672160139295,
+      "loss": 3.1285,
+      "step": 25261
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5301439762115479,
+      "learning_rate": 0.0003812719012475968,
+      "loss": 3.0362,
+      "step": 25262
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5626558661460876,
+      "learning_rate": 0.000381257080679746,
+      "loss": 3.0798,
+      "step": 25263
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5594674944877625,
+      "learning_rate": 0.00038124225989787975,
+      "loss": 2.9337,
+      "step": 25264
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5297481417655945,
+      "learning_rate": 0.00038122743890203704,
+      "loss": 3.1576,
+      "step": 25265
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5789366960525513,
+      "learning_rate": 0.00038121261769225685,
+      "loss": 3.0484,
+      "step": 25266
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5611054301261902,
+      "learning_rate": 0.0003811977962685782,
+      "loss": 3.1415,
+      "step": 25267
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5735259056091309,
+      "learning_rate": 0.0003811829746310403,
+      "loss": 3.061,
+      "step": 25268
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5358028411865234,
+      "learning_rate": 0.00038116815277968195,
+      "loss": 3.0565,
+      "step": 25269
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5914501547813416,
+      "learning_rate": 0.00038115333071454246,
+      "loss": 3.1501,
+      "step": 25270
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5669057369232178,
+      "learning_rate": 0.0003811385084356606,
+      "loss": 3.2603,
+      "step": 25271
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.549128532409668,
+      "learning_rate": 0.00038112368594307554,
+      "loss": 3.0044,
+      "step": 25272
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5490759015083313,
+      "learning_rate": 0.0003811088632368263,
+      "loss": 3.2027,
+      "step": 25273
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5359374284744263,
+      "learning_rate": 0.000381094040316952,
+      "loss": 3.2197,
+      "step": 25274
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5579591393470764,
+      "learning_rate": 0.0003810792171834915,
+      "loss": 3.1138,
+      "step": 25275
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5682498216629028,
+      "learning_rate": 0.00038106439383648396,
+      "loss": 3.0713,
+      "step": 25276
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5187976360321045,
+      "learning_rate": 0.0003810495702759685,
+      "loss": 2.9462,
+      "step": 25277
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5464273691177368,
+      "learning_rate": 0.00038103474650198396,
+      "loss": 2.9093,
+      "step": 25278
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5833659768104553,
+      "learning_rate": 0.0003810199225145695,
+      "loss": 3.0675,
+      "step": 25279
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5375162959098816,
+      "learning_rate": 0.0003810050983137642,
+      "loss": 2.9478,
+      "step": 25280
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5666972994804382,
+      "learning_rate": 0.00038099027389960704,
+      "loss": 2.9796,
+      "step": 25281
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.526598334312439,
+      "learning_rate": 0.0003809754492721371,
+      "loss": 3.0346,
+      "step": 25282
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5676760077476501,
+      "learning_rate": 0.0003809606244313934,
+      "loss": 3.2404,
+      "step": 25283
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5630086064338684,
+      "learning_rate": 0.00038094579937741506,
+      "loss": 2.936,
+      "step": 25284
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5473533272743225,
+      "learning_rate": 0.00038093097411024106,
+      "loss": 3.0267,
+      "step": 25285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6100202202796936,
+      "learning_rate": 0.0003809161486299103,
+      "loss": 3.1214,
+      "step": 25286
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5748114585876465,
+      "learning_rate": 0.0003809013229364621,
+      "loss": 3.1075,
+      "step": 25287
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.567639172077179,
+      "learning_rate": 0.00038088649702993537,
+      "loss": 3.0915,
+      "step": 25288
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5674895644187927,
+      "learning_rate": 0.00038087167091036916,
+      "loss": 3.0159,
+      "step": 25289
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5389357209205627,
+      "learning_rate": 0.0003808568445778025,
+      "loss": 3.1525,
+      "step": 25290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5887845754623413,
+      "learning_rate": 0.00038084201803227456,
+      "loss": 3.0113,
+      "step": 25291
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5666884183883667,
+      "learning_rate": 0.00038082719127382423,
+      "loss": 3.0803,
+      "step": 25292
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5322736501693726,
+      "learning_rate": 0.00038081236430249064,
+      "loss": 3.0178,
+      "step": 25293
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.519856333732605,
+      "learning_rate": 0.0003807975371183129,
+      "loss": 3.0541,
+      "step": 25294
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5431399345397949,
+      "learning_rate": 0.0003807827097213299,
+      "loss": 3.0365,
+      "step": 25295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5920628905296326,
+      "learning_rate": 0.0003807678821115809,
+      "loss": 3.0214,
+      "step": 25296
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5433076024055481,
+      "learning_rate": 0.0003807530542891047,
+      "loss": 2.8857,
+      "step": 25297
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5601280331611633,
+      "learning_rate": 0.0003807382262539406,
+      "loss": 2.8714,
+      "step": 25298
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5662012696266174,
+      "learning_rate": 0.00038072339800612753,
+      "loss": 2.8085,
+      "step": 25299
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5706889629364014,
+      "learning_rate": 0.0003807085695457046,
+      "loss": 3.1877,
+      "step": 25300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5297551155090332,
+      "learning_rate": 0.0003806937408727108,
+      "loss": 3.0017,
+      "step": 25301
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5498740673065186,
+      "learning_rate": 0.0003806789119871852,
+      "loss": 3.1668,
+      "step": 25302
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6031630635261536,
+      "learning_rate": 0.0003806640828891669,
+      "loss": 3.1125,
+      "step": 25303
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5248180627822876,
+      "learning_rate": 0.00038064925357869493,
+      "loss": 2.9471,
+      "step": 25304
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5380234718322754,
+      "learning_rate": 0.0003806344240558083,
+      "loss": 3.0616,
+      "step": 25305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.544638991355896,
+      "learning_rate": 0.00038061959432054613,
+      "loss": 3.1263,
+      "step": 25306
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5599265694618225,
+      "learning_rate": 0.00038060476437294747,
+      "loss": 3.083,
+      "step": 25307
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5477656126022339,
+      "learning_rate": 0.0003805899342130514,
+      "loss": 3.1702,
+      "step": 25308
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5713855028152466,
+      "learning_rate": 0.00038057510384089695,
+      "loss": 3.1395,
+      "step": 25309
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5825148820877075,
+      "learning_rate": 0.00038056027325652315,
+      "loss": 3.1251,
+      "step": 25310
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5400784015655518,
+      "learning_rate": 0.00038054544245996915,
+      "loss": 3.0086,
+      "step": 25311
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5549777150154114,
+      "learning_rate": 0.00038053061145127385,
+      "loss": 3.0571,
+      "step": 25312
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5992767810821533,
+      "learning_rate": 0.0003805157802304766,
+      "loss": 3.0949,
+      "step": 25313
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5606101155281067,
+      "learning_rate": 0.0003805009487976161,
+      "loss": 3.0668,
+      "step": 25314
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6076589822769165,
+      "learning_rate": 0.00038048611715273166,
+      "loss": 3.0241,
+      "step": 25315
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5783843994140625,
+      "learning_rate": 0.00038047128529586225,
+      "loss": 3.0586,
+      "step": 25316
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5715208053588867,
+      "learning_rate": 0.000380456453227047,
+      "loss": 3.1446,
+      "step": 25317
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5512316823005676,
+      "learning_rate": 0.00038044162094632493,
+      "loss": 3.0362,
+      "step": 25318
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.539901852607727,
+      "learning_rate": 0.0003804267884537351,
+      "loss": 3.0684,
+      "step": 25319
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5319052338600159,
+      "learning_rate": 0.0003804119557493166,
+      "loss": 3.0007,
+      "step": 25320
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5402718782424927,
+      "learning_rate": 0.0003803971228331085,
+      "loss": 2.9589,
+      "step": 25321
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6263967156410217,
+      "learning_rate": 0.0003803822897051498,
+      "loss": 3.0552,
+      "step": 25322
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5595197081565857,
+      "learning_rate": 0.00038036745636547964,
+      "loss": 2.8282,
+      "step": 25323
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5648815035820007,
+      "learning_rate": 0.0003803526228141371,
+      "loss": 2.9941,
+      "step": 25324
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6085686683654785,
+      "learning_rate": 0.00038033778905116123,
+      "loss": 3.0553,
+      "step": 25325
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5713744163513184,
+      "learning_rate": 0.000380322955076591,
+      "loss": 2.9707,
+      "step": 25326
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5762509107589722,
+      "learning_rate": 0.0003803081208904656,
+      "loss": 3.0925,
+      "step": 25327
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5664671063423157,
+      "learning_rate": 0.0003802932864928241,
+      "loss": 2.9792,
+      "step": 25328
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5856859087944031,
+      "learning_rate": 0.00038027845188370554,
+      "loss": 2.8393,
+      "step": 25329
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5801443457603455,
+      "learning_rate": 0.00038026361706314894,
+      "loss": 3.1843,
+      "step": 25330
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5778241157531738,
+      "learning_rate": 0.00038024878203119346,
+      "loss": 3.1812,
+      "step": 25331
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5924252271652222,
+      "learning_rate": 0.0003802339467878781,
+      "loss": 3.0418,
+      "step": 25332
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5770880579948425,
+      "learning_rate": 0.00038021911133324194,
+      "loss": 2.8706,
+      "step": 25333
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.547100841999054,
+      "learning_rate": 0.0003802042756673241,
+      "loss": 3.2091,
+      "step": 25334
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6044512391090393,
+      "learning_rate": 0.0003801894397901637,
+      "loss": 3.153,
+      "step": 25335
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5438991189002991,
+      "learning_rate": 0.00038017460370179976,
+      "loss": 3.1416,
+      "step": 25336
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5397824645042419,
+      "learning_rate": 0.00038015976740227126,
+      "loss": 3.065,
+      "step": 25337
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6003246903419495,
+      "learning_rate": 0.0003801449308916173,
+      "loss": 2.9869,
+      "step": 25338
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5817223191261292,
+      "learning_rate": 0.00038013009416987714,
+      "loss": 3.0956,
+      "step": 25339
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5436729192733765,
+      "learning_rate": 0.00038011525723708974,
+      "loss": 3.1657,
+      "step": 25340
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5461451411247253,
+      "learning_rate": 0.0003801004200932941,
+      "loss": 2.9516,
+      "step": 25341
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5592874884605408,
+      "learning_rate": 0.0003800855827385294,
+      "loss": 3.01,
+      "step": 25342
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5277195572853088,
+      "learning_rate": 0.00038007074517283466,
+      "loss": 2.9353,
+      "step": 25343
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5816713571548462,
+      "learning_rate": 0.000380055907396249,
+      "loss": 3.2137,
+      "step": 25344
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5725088715553284,
+      "learning_rate": 0.00038004106940881145,
+      "loss": 2.9513,
+      "step": 25345
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5226502418518066,
+      "learning_rate": 0.0003800262312105612,
+      "loss": 2.9282,
+      "step": 25346
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5807681679725647,
+      "learning_rate": 0.0003800113928015373,
+      "loss": 3.0696,
+      "step": 25347
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5598049163818359,
+      "learning_rate": 0.00037999655418177866,
+      "loss": 3.0249,
+      "step": 25348
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5281652212142944,
+      "learning_rate": 0.00037998171535132456,
+      "loss": 3.0612,
+      "step": 25349
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5705523490905762,
+      "learning_rate": 0.00037996687631021404,
+      "loss": 3.0364,
+      "step": 25350
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.599563717842102,
+      "learning_rate": 0.00037995203705848613,
+      "loss": 3.0559,
+      "step": 25351
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5675721168518066,
+      "learning_rate": 0.0003799371975961799,
+      "loss": 2.8777,
+      "step": 25352
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5328501462936401,
+      "learning_rate": 0.00037992235792333457,
+      "loss": 3.0203,
+      "step": 25353
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.554772138595581,
+      "learning_rate": 0.000379907518039989,
+      "loss": 3.1119,
+      "step": 25354
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6034314632415771,
+      "learning_rate": 0.0003798926779461825,
+      "loss": 2.9382,
+      "step": 25355
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.582343578338623,
+      "learning_rate": 0.000379877837641954,
+      "loss": 3.0815,
+      "step": 25356
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5476296544075012,
+      "learning_rate": 0.00037986299712734273,
+      "loss": 3.1573,
+      "step": 25357
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5945506691932678,
+      "learning_rate": 0.00037984815640238764,
+      "loss": 3.1786,
+      "step": 25358
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5869128704071045,
+      "learning_rate": 0.0003798333154671279,
+      "loss": 3.2192,
+      "step": 25359
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.585900604724884,
+      "learning_rate": 0.00037981847432160253,
+      "loss": 2.7552,
+      "step": 25360
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6065345406532288,
+      "learning_rate": 0.0003798036329658508,
+      "loss": 3.0549,
+      "step": 25361
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6202573180198669,
+      "learning_rate": 0.0003797887913999115,
+      "loss": 3.2176,
+      "step": 25362
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5414901971817017,
+      "learning_rate": 0.00037977394962382383,
+      "loss": 3.2003,
+      "step": 25363
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5944116115570068,
+      "learning_rate": 0.00037975910763762713,
+      "loss": 3.0485,
+      "step": 25364
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5542477369308472,
+      "learning_rate": 0.00037974426544136014,
+      "loss": 3.0894,
+      "step": 25365
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5856495499610901,
+      "learning_rate": 0.00037972942303506215,
+      "loss": 3.2326,
+      "step": 25366
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5559060573577881,
+      "learning_rate": 0.0003797145804187722,
+      "loss": 3.1288,
+      "step": 25367
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6178712248802185,
+      "learning_rate": 0.0003796997375925294,
+      "loss": 2.901,
+      "step": 25368
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.534890353679657,
+      "learning_rate": 0.00037968489455637286,
+      "loss": 3.0637,
+      "step": 25369
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5386385917663574,
+      "learning_rate": 0.0003796700513103416,
+      "loss": 3.1557,
+      "step": 25370
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.574946403503418,
+      "learning_rate": 0.00037965520785447474,
+      "loss": 3.0145,
+      "step": 25371
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5754439830780029,
+      "learning_rate": 0.0003796403641888114,
+      "loss": 3.0277,
+      "step": 25372
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6202484369277954,
+      "learning_rate": 0.0003796255203133907,
+      "loss": 3.1161,
+      "step": 25373
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5231525897979736,
+      "learning_rate": 0.0003796106762282517,
+      "loss": 3.1765,
+      "step": 25374
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5547069311141968,
+      "learning_rate": 0.0003795958319334334,
+      "loss": 3.0762,
+      "step": 25375
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5507140159606934,
+      "learning_rate": 0.0003795809874289751,
+      "loss": 2.9072,
+      "step": 25376
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5551223754882812,
+      "learning_rate": 0.0003795661427149158,
+      "loss": 2.9539,
+      "step": 25377
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5440731048583984,
+      "learning_rate": 0.0003795512977912946,
+      "loss": 3.1145,
+      "step": 25378
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5543565154075623,
+      "learning_rate": 0.0003795364526581505,
+      "loss": 3.1151,
+      "step": 25379
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5475043654441833,
+      "learning_rate": 0.0003795216073155228,
+      "loss": 3.1024,
+      "step": 25380
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5892181396484375,
+      "learning_rate": 0.00037950676176345044,
+      "loss": 2.9669,
+      "step": 25381
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5114701390266418,
+      "learning_rate": 0.00037949191600197254,
+      "loss": 2.8436,
+      "step": 25382
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5685053467750549,
+      "learning_rate": 0.00037947707003112833,
+      "loss": 3.0272,
+      "step": 25383
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.579219400882721,
+      "learning_rate": 0.0003794622238509567,
+      "loss": 3.0499,
+      "step": 25384
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5205180644989014,
+      "learning_rate": 0.00037944737746149694,
+      "loss": 3.0484,
+      "step": 25385
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.561149001121521,
+      "learning_rate": 0.0003794325308627881,
+      "loss": 3.0581,
+      "step": 25386
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5951744318008423,
+      "learning_rate": 0.0003794176840548691,
+      "loss": 3.1024,
+      "step": 25387
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6370974779129028,
+      "learning_rate": 0.0003794028370377794,
+      "loss": 3.1705,
+      "step": 25388
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5966722965240479,
+      "learning_rate": 0.0003793879898115577,
+      "loss": 2.9605,
+      "step": 25389
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5954158306121826,
+      "learning_rate": 0.00037937314237624357,
+      "loss": 3.0356,
+      "step": 25390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5606188774108887,
+      "learning_rate": 0.00037935829473187573,
+      "loss": 2.9174,
+      "step": 25391
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5629304051399231,
+      "learning_rate": 0.0003793434468784934,
+      "loss": 2.9253,
+      "step": 25392
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5472045540809631,
+      "learning_rate": 0.0003793285988161357,
+      "loss": 3.3048,
+      "step": 25393
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5473329424858093,
+      "learning_rate": 0.0003793137505448417,
+      "loss": 3.1923,
+      "step": 25394
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5582367777824402,
+      "learning_rate": 0.0003792989020646506,
+      "loss": 3.0302,
+      "step": 25395
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2111254930496216,
+      "learning_rate": 0.00037928405337560143,
+      "loss": 2.9217,
+      "step": 25396
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5781081318855286,
+      "learning_rate": 0.0003792692044777334,
+      "loss": 3.1969,
+      "step": 25397
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5536330938339233,
+      "learning_rate": 0.00037925435537108546,
+      "loss": 3.2146,
+      "step": 25398
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5391330122947693,
+      "learning_rate": 0.00037923950605569677,
+      "loss": 2.8489,
+      "step": 25399
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5480332374572754,
+      "learning_rate": 0.00037922465653160653,
+      "loss": 2.9888,
+      "step": 25400
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5818240642547607,
+      "learning_rate": 0.00037920980679885383,
+      "loss": 2.974,
+      "step": 25401
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5594255924224854,
+      "learning_rate": 0.0003791949568574777,
+      "loss": 3.0717,
+      "step": 25402
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5563430786132812,
+      "learning_rate": 0.00037918010670751735,
+      "loss": 3.3801,
+      "step": 25403
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5488848090171814,
+      "learning_rate": 0.0003791652563490118,
+      "loss": 2.8365,
+      "step": 25404
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5589650869369507,
+      "learning_rate": 0.0003791504057820001,
+      "loss": 2.8259,
+      "step": 25405
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5604705214500427,
+      "learning_rate": 0.00037913555500652154,
+      "loss": 3.1208,
+      "step": 25406
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5571016073226929,
+      "learning_rate": 0.0003791207040226152,
+      "loss": 3.1,
+      "step": 25407
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.559099555015564,
+      "learning_rate": 0.00037910585283032006,
+      "loss": 3.0902,
+      "step": 25408
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5342800617218018,
+      "learning_rate": 0.0003790910014296754,
+      "loss": 3.0075,
+      "step": 25409
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.555751621723175,
+      "learning_rate": 0.00037907614982072023,
+      "loss": 3.2053,
+      "step": 25410
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6236014366149902,
+      "learning_rate": 0.0003790612980034937,
+      "loss": 3.0104,
+      "step": 25411
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5512464642524719,
+      "learning_rate": 0.000379046445978035,
+      "loss": 3.1207,
+      "step": 25412
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5613242387771606,
+      "learning_rate": 0.00037903159374438314,
+      "loss": 3.1059,
+      "step": 25413
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5795272588729858,
+      "learning_rate": 0.00037901674130257724,
+      "loss": 3.1214,
+      "step": 25414
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5552756190299988,
+      "learning_rate": 0.00037900188865265646,
+      "loss": 3.2082,
+      "step": 25415
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5399115681648254,
+      "learning_rate": 0.0003789870357946599,
+      "loss": 3.035,
+      "step": 25416
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5368036031723022,
+      "learning_rate": 0.0003789721827286267,
+      "loss": 2.8561,
+      "step": 25417
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5200400352478027,
+      "learning_rate": 0.000378957329454596,
+      "loss": 2.9214,
+      "step": 25418
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.569953978061676,
+      "learning_rate": 0.0003789424759726068,
+      "loss": 3.2006,
+      "step": 25419
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5677704215049744,
+      "learning_rate": 0.0003789276222826984,
+      "loss": 2.8954,
+      "step": 25420
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5435131788253784,
+      "learning_rate": 0.00037891276838490976,
+      "loss": 3.0026,
+      "step": 25421
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5513231754302979,
+      "learning_rate": 0.00037889791427928016,
+      "loss": 3.173,
+      "step": 25422
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5995567440986633,
+      "learning_rate": 0.00037888305996584864,
+      "loss": 2.996,
+      "step": 25423
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5517016649246216,
+      "learning_rate": 0.00037886820544465427,
+      "loss": 2.964,
+      "step": 25424
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.572759747505188,
+      "learning_rate": 0.00037885335071573614,
+      "loss": 2.8945,
+      "step": 25425
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6202085018157959,
+      "learning_rate": 0.0003788384957791336,
+      "loss": 3.1124,
+      "step": 25426
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5376129746437073,
+      "learning_rate": 0.00037882364063488566,
+      "loss": 3.0313,
+      "step": 25427
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5526015758514404,
+      "learning_rate": 0.00037880878528303126,
+      "loss": 3.0498,
+      "step": 25428
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.553801417350769,
+      "learning_rate": 0.0003787939297236098,
+      "loss": 3.0979,
+      "step": 25429
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5524164438247681,
+      "learning_rate": 0.0003787790739566603,
+      "loss": 3.1177,
+      "step": 25430
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5112387537956238,
+      "learning_rate": 0.00037876421798222176,
+      "loss": 3.2164,
+      "step": 25431
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5780650973320007,
+      "learning_rate": 0.00037874936180033356,
+      "loss": 3.0476,
+      "step": 25432
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5552143454551697,
+      "learning_rate": 0.0003787345054110347,
+      "loss": 3.1861,
+      "step": 25433
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5469130277633667,
+      "learning_rate": 0.00037871964881436423,
+      "loss": 2.8781,
+      "step": 25434
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5303400158882141,
+      "learning_rate": 0.00037870479201036137,
+      "loss": 2.987,
+      "step": 25435
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5632544755935669,
+      "learning_rate": 0.00037868993499906524,
+      "loss": 3.0197,
+      "step": 25436
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5879452228546143,
+      "learning_rate": 0.000378675077780515,
+      "loss": 3.1763,
+      "step": 25437
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6507343649864197,
+      "learning_rate": 0.00037866022035474976,
+      "loss": 2.9622,
+      "step": 25438
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5414554476737976,
+      "learning_rate": 0.0003786453627218086,
+      "loss": 3.0865,
+      "step": 25439
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0003786305048817306,
+      "loss": 3.1138,
+      "step": 25440
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5658540725708008,
+      "learning_rate": 0.0003786156468345551,
+      "loss": 2.9164,
+      "step": 25441
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5722022652626038,
+      "learning_rate": 0.0003786007885803211,
+      "loss": 3.1499,
+      "step": 25442
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5444322228431702,
+      "learning_rate": 0.0003785859301190677,
+      "loss": 3.0966,
+      "step": 25443
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5430464148521423,
+      "learning_rate": 0.00037857107145083415,
+      "loss": 2.9632,
+      "step": 25444
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5783938765525818,
+      "learning_rate": 0.00037855621257565955,
+      "loss": 2.8965,
+      "step": 25445
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5244552493095398,
+      "learning_rate": 0.00037854135349358285,
+      "loss": 3.1105,
+      "step": 25446
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5750003457069397,
+      "learning_rate": 0.0003785264942046434,
+      "loss": 3.1679,
+      "step": 25447
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5536020398139954,
+      "learning_rate": 0.00037851163470888045,
+      "loss": 3.4111,
+      "step": 25448
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7663441896438599,
+      "learning_rate": 0.0003784967750063328,
+      "loss": 2.8687,
+      "step": 25449
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5692669153213501,
+      "learning_rate": 0.0003784819150970397,
+      "loss": 2.9943,
+      "step": 25450
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5337068438529968,
+      "learning_rate": 0.0003784670549810404,
+      "loss": 3.1486,
+      "step": 25451
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5633428692817688,
+      "learning_rate": 0.0003784521946583741,
+      "loss": 3.135,
+      "step": 25452
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5698487758636475,
+      "learning_rate": 0.0003784373341290797,
+      "loss": 3.0194,
+      "step": 25453
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5451970100402832,
+      "learning_rate": 0.00037842247339319645,
+      "loss": 3.0151,
+      "step": 25454
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5770736336708069,
+      "learning_rate": 0.0003784076124507635,
+      "loss": 3.2033,
+      "step": 25455
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5762639045715332,
+      "learning_rate": 0.00037839275130182004,
+      "loss": 2.977,
+      "step": 25456
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5343278646469116,
+      "learning_rate": 0.0003783778899464051,
+      "loss": 3.0204,
+      "step": 25457
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5425517559051514,
+      "learning_rate": 0.00037836302838455794,
+      "loss": 3.0833,
+      "step": 25458
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5324981212615967,
+      "learning_rate": 0.0003783481666163176,
+      "loss": 3.2069,
+      "step": 25459
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5692557096481323,
+      "learning_rate": 0.0003783333046417233,
+      "loss": 3.0091,
+      "step": 25460
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5556652545928955,
+      "learning_rate": 0.00037831844246081413,
+      "loss": 2.867,
+      "step": 25461
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5014756321907043,
+      "learning_rate": 0.00037830358007362924,
+      "loss": 3.1228,
+      "step": 25462
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5904489159584045,
+      "learning_rate": 0.00037828871748020785,
+      "loss": 3.0725,
+      "step": 25463
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5578276515007019,
+      "learning_rate": 0.000378273854680589,
+      "loss": 3.1744,
+      "step": 25464
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5439172387123108,
+      "learning_rate": 0.00037825899167481184,
+      "loss": 3.1216,
+      "step": 25465
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5510421395301819,
+      "learning_rate": 0.0003782441284629156,
+      "loss": 3.1522,
+      "step": 25466
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.581859827041626,
+      "learning_rate": 0.0003782292650449394,
+      "loss": 3.0195,
+      "step": 25467
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5343931913375854,
+      "learning_rate": 0.00037821440142092236,
+      "loss": 3.1396,
+      "step": 25468
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5974150896072388,
+      "learning_rate": 0.0003781995375909036,
+      "loss": 3.2128,
+      "step": 25469
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5799618363380432,
+      "learning_rate": 0.00037818467355492234,
+      "loss": 3.0855,
+      "step": 25470
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5386278629302979,
+      "learning_rate": 0.00037816980931301773,
+      "loss": 3.1758,
+      "step": 25471
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6230940818786621,
+      "learning_rate": 0.0003781549448652288,
+      "loss": 3.2251,
+      "step": 25472
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5379871129989624,
+      "learning_rate": 0.00037814008021159486,
+      "loss": 3.0945,
+      "step": 25473
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5654126405715942,
+      "learning_rate": 0.00037812521535215504,
+      "loss": 3.1067,
+      "step": 25474
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5493576526641846,
+      "learning_rate": 0.00037811035028694834,
+      "loss": 3.1932,
+      "step": 25475
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5735651850700378,
+      "learning_rate": 0.000378095485016014,
+      "loss": 3.2,
+      "step": 25476
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5522319674491882,
+      "learning_rate": 0.0003780806195393913,
+      "loss": 3.0353,
+      "step": 25477
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.575622022151947,
+      "learning_rate": 0.00037806575385711916,
+      "loss": 2.8894,
+      "step": 25478
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6093813180923462,
+      "learning_rate": 0.0003780508879692369,
+      "loss": 2.8872,
+      "step": 25479
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5739896297454834,
+      "learning_rate": 0.00037803602187578357,
+      "loss": 2.9564,
+      "step": 25480
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.560309648513794,
+      "learning_rate": 0.00037802115557679844,
+      "loss": 3.3268,
+      "step": 25481
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5892646908760071,
+      "learning_rate": 0.00037800628907232056,
+      "loss": 3.1495,
+      "step": 25482
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5857096910476685,
+      "learning_rate": 0.0003779914223623892,
+      "loss": 2.9523,
+      "step": 25483
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5739219784736633,
+      "learning_rate": 0.0003779765554470433,
+      "loss": 2.9484,
+      "step": 25484
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5851458311080933,
+      "learning_rate": 0.0003779616883263223,
+      "loss": 3.0733,
+      "step": 25485
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5476492047309875,
+      "learning_rate": 0.0003779468210002652,
+      "loss": 3.0483,
+      "step": 25486
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5560165047645569,
+      "learning_rate": 0.000377931953468911,
+      "loss": 3.0811,
+      "step": 25487
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5552588105201721,
+      "learning_rate": 0.00037791708573229926,
+      "loss": 3.3325,
+      "step": 25488
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5610097646713257,
+      "learning_rate": 0.0003779022177904689,
+      "loss": 3.1599,
+      "step": 25489
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6263189911842346,
+      "learning_rate": 0.00037788734964345897,
+      "loss": 3.2049,
+      "step": 25490
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5741820931434631,
+      "learning_rate": 0.0003778724812913088,
+      "loss": 3.0006,
+      "step": 25491
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.579188346862793,
+      "learning_rate": 0.00037785761273405745,
+      "loss": 3.0864,
+      "step": 25492
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5346486568450928,
+      "learning_rate": 0.00037784274397174416,
+      "loss": 2.8912,
+      "step": 25493
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5457817912101746,
+      "learning_rate": 0.0003778278750044081,
+      "loss": 2.8513,
+      "step": 25494
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5900013446807861,
+      "learning_rate": 0.00037781300583208835,
+      "loss": 3.0625,
+      "step": 25495
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5467310547828674,
+      "learning_rate": 0.0003777981364548242,
+      "loss": 3.1045,
+      "step": 25496
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5489881038665771,
+      "learning_rate": 0.0003777832668726546,
+      "loss": 2.9645,
+      "step": 25497
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5877793431282043,
+      "learning_rate": 0.0003777683970856189,
+      "loss": 3.1978,
+      "step": 25498
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5987591743469238,
+      "learning_rate": 0.0003777535270937563,
+      "loss": 2.9118,
+      "step": 25499
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5459617972373962,
+      "learning_rate": 0.00037773865689710585,
+      "loss": 3.2444,
+      "step": 25500
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5510004162788391,
+      "learning_rate": 0.00037772378649570667,
+      "loss": 3.0586,
+      "step": 25501
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5648934245109558,
+      "learning_rate": 0.000377708915889598,
+      "loss": 3.0798,
+      "step": 25502
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5991185307502747,
+      "learning_rate": 0.00037769404507881897,
+      "loss": 3.095,
+      "step": 25503
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5326690673828125,
+      "learning_rate": 0.00037767917406340883,
+      "loss": 3.0806,
+      "step": 25504
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5778882503509521,
+      "learning_rate": 0.0003776643028434067,
+      "loss": 3.0523,
+      "step": 25505
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5468420386314392,
+      "learning_rate": 0.00037764943141885174,
+      "loss": 3.038,
+      "step": 25506
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.617231547832489,
+      "learning_rate": 0.0003776345597897831,
+      "loss": 2.9654,
+      "step": 25507
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5568265914916992,
+      "learning_rate": 0.00037761968795624,
+      "loss": 2.8395,
+      "step": 25508
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5809617638587952,
+      "learning_rate": 0.00037760481591826153,
+      "loss": 3.1054,
+      "step": 25509
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5677468180656433,
+      "learning_rate": 0.00037758994367588695,
+      "loss": 2.9851,
+      "step": 25510
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5191762447357178,
+      "learning_rate": 0.00037757507122915544,
+      "loss": 3.2637,
+      "step": 25511
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5888081192970276,
+      "learning_rate": 0.00037756019857810604,
+      "loss": 3.06,
+      "step": 25512
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5667383074760437,
+      "learning_rate": 0.000377545325722778,
+      "loss": 2.9374,
+      "step": 25513
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5688818693161011,
+      "learning_rate": 0.0003775304526632106,
+      "loss": 3.2734,
+      "step": 25514
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5567358136177063,
+      "learning_rate": 0.0003775155793994428,
+      "loss": 3.0426,
+      "step": 25515
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5471105575561523,
+      "learning_rate": 0.00037750070593151394,
+      "loss": 3.25,
+      "step": 25516
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5713381767272949,
+      "learning_rate": 0.00037748583225946306,
+      "loss": 3.0582,
+      "step": 25517
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5663639903068542,
+      "learning_rate": 0.00037747095838332954,
+      "loss": 3.2855,
+      "step": 25518
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5420747399330139,
+      "learning_rate": 0.00037745608430315227,
+      "loss": 3.3252,
+      "step": 25519
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5536601543426514,
+      "learning_rate": 0.0003774412100189707,
+      "loss": 3.0771,
+      "step": 25520
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5411950945854187,
+      "learning_rate": 0.0003774263355308238,
+      "loss": 3.0431,
+      "step": 25521
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.571185827255249,
+      "learning_rate": 0.0003774114608387509,
+      "loss": 3.2206,
+      "step": 25522
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5684292316436768,
+      "learning_rate": 0.0003773965859427911,
+      "loss": 3.0063,
+      "step": 25523
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5304699540138245,
+      "learning_rate": 0.0003773817108429836,
+      "loss": 3.0202,
+      "step": 25524
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5694625973701477,
+      "learning_rate": 0.0003773668355393676,
+      "loss": 3.0962,
+      "step": 25525
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.545238733291626,
+      "learning_rate": 0.00037735196003198213,
+      "loss": 2.9854,
+      "step": 25526
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.578179657459259,
+      "learning_rate": 0.0003773370843208666,
+      "loss": 3.029,
+      "step": 25527
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5803759694099426,
+      "learning_rate": 0.00037732220840606,
+      "loss": 3.2408,
+      "step": 25528
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5503095388412476,
+      "learning_rate": 0.0003773073322876016,
+      "loss": 3.4273,
+      "step": 25529
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.532310426235199,
+      "learning_rate": 0.0003772924559655305,
+      "loss": 3.0394,
+      "step": 25530
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5836457014083862,
+      "learning_rate": 0.000377277579439886,
+      "loss": 2.9752,
+      "step": 25531
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5723191499710083,
+      "learning_rate": 0.0003772627027107073,
+      "loss": 3.0069,
+      "step": 25532
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5557374358177185,
+      "learning_rate": 0.00037724782577803345,
+      "loss": 3.2241,
+      "step": 25533
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5663606524467468,
+      "learning_rate": 0.0003772329486419036,
+      "loss": 3.2456,
+      "step": 25534
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5517666339874268,
+      "learning_rate": 0.0003772180713023572,
+      "loss": 2.9475,
+      "step": 25535
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5494801998138428,
+      "learning_rate": 0.0003772031937594332,
+      "loss": 3.1934,
+      "step": 25536
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5216752886772156,
+      "learning_rate": 0.0003771883160131708,
+      "loss": 3.0643,
+      "step": 25537
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5742051005363464,
+      "learning_rate": 0.0003771734380636092,
+      "loss": 2.9488,
+      "step": 25538
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.53158038854599,
+      "learning_rate": 0.00037715855991078776,
+      "loss": 2.9859,
+      "step": 25539
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5515921711921692,
+      "learning_rate": 0.00037714368155474545,
+      "loss": 3.0378,
+      "step": 25540
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5819881558418274,
+      "learning_rate": 0.0003771288029955215,
+      "loss": 3.0099,
+      "step": 25541
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5356078743934631,
+      "learning_rate": 0.00037711392423315517,
+      "loss": 3.1849,
+      "step": 25542
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5517314076423645,
+      "learning_rate": 0.00037709904526768553,
+      "loss": 3.141,
+      "step": 25543
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5887091159820557,
+      "learning_rate": 0.0003770841660991519,
+      "loss": 3.2121,
+      "step": 25544
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6048860549926758,
+      "learning_rate": 0.0003770692867275934,
+      "loss": 3.0597,
+      "step": 25545
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5582966804504395,
+      "learning_rate": 0.0003770544071530492,
+      "loss": 2.9708,
+      "step": 25546
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.588004469871521,
+      "learning_rate": 0.0003770395273755586,
+      "loss": 3.0733,
+      "step": 25547
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6418083906173706,
+      "learning_rate": 0.00037702464739516065,
+      "loss": 2.9607,
+      "step": 25548
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5486046075820923,
+      "learning_rate": 0.00037700976721189464,
+      "loss": 3.0641,
+      "step": 25549
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5865910649299622,
+      "learning_rate": 0.0003769948868257997,
+      "loss": 3.2268,
+      "step": 25550
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.571828305721283,
+      "learning_rate": 0.0003769800062369151,
+      "loss": 3.1411,
+      "step": 25551
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6386631727218628,
+      "learning_rate": 0.00037696512544527986,
+      "loss": 3.2847,
+      "step": 25552
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5370537042617798,
+      "learning_rate": 0.0003769502444509333,
+      "loss": 2.9321,
+      "step": 25553
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5921683311462402,
+      "learning_rate": 0.00037693536325391475,
+      "loss": 3.0641,
+      "step": 25554
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.533803403377533,
+      "learning_rate": 0.0003769204818542632,
+      "loss": 2.8298,
+      "step": 25555
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5238431096076965,
+      "learning_rate": 0.0003769056002520178,
+      "loss": 3.0571,
+      "step": 25556
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6707302331924438,
+      "learning_rate": 0.00037689071844721796,
+      "loss": 3.0363,
+      "step": 25557
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5460362434387207,
+      "learning_rate": 0.00037687583643990273,
+      "loss": 3.3561,
+      "step": 25558
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5882094502449036,
+      "learning_rate": 0.00037686095423011134,
+      "loss": 3.268,
+      "step": 25559
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5774903297424316,
+      "learning_rate": 0.000376846071817883,
+      "loss": 3.0655,
+      "step": 25560
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5424748659133911,
+      "learning_rate": 0.0003768311892032569,
+      "loss": 3.0053,
+      "step": 25561
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5767691731452942,
+      "learning_rate": 0.00037681630638627223,
+      "loss": 2.964,
+      "step": 25562
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5361145734786987,
+      "learning_rate": 0.0003768014233669682,
+      "loss": 3.1929,
+      "step": 25563
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.582098126411438,
+      "learning_rate": 0.000376786540145384,
+      "loss": 2.9265,
+      "step": 25564
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5639384984970093,
+      "learning_rate": 0.00037677165672155884,
+      "loss": 2.9726,
+      "step": 25565
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.608085036277771,
+      "learning_rate": 0.0003767567730955319,
+      "loss": 3.1035,
+      "step": 25566
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5574251413345337,
+      "learning_rate": 0.00037674188926734237,
+      "loss": 2.9313,
+      "step": 25567
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5575839281082153,
+      "learning_rate": 0.00037672700523702944,
+      "loss": 3.0273,
+      "step": 25568
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6037725806236267,
+      "learning_rate": 0.0003767121210046324,
+      "loss": 3.2235,
+      "step": 25569
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5843100547790527,
+      "learning_rate": 0.0003766972365701904,
+      "loss": 3.0442,
+      "step": 25570
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5654159188270569,
+      "learning_rate": 0.0003766823519337426,
+      "loss": 3.1427,
+      "step": 25571
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5915213227272034,
+      "learning_rate": 0.00037666746709532825,
+      "loss": 3.533,
+      "step": 25572
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5641651153564453,
+      "learning_rate": 0.0003766525820549866,
+      "loss": 3.0753,
+      "step": 25573
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5694716572761536,
+      "learning_rate": 0.0003766376968127567,
+      "loss": 3.1201,
+      "step": 25574
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5624406933784485,
+      "learning_rate": 0.00037662281136867793,
+      "loss": 3.0598,
+      "step": 25575
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.560565710067749,
+      "learning_rate": 0.00037660792572278943,
+      "loss": 3.111,
+      "step": 25576
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5786372423171997,
+      "learning_rate": 0.0003765930398751304,
+      "loss": 3.0341,
+      "step": 25577
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.535211980342865,
+      "learning_rate": 0.00037657815382574004,
+      "loss": 2.8081,
+      "step": 25578
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5386492609977722,
+      "learning_rate": 0.0003765632675746576,
+      "loss": 3.2728,
+      "step": 25579
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5663537979125977,
+      "learning_rate": 0.00037654838112192215,
+      "loss": 3.2322,
+      "step": 25580
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6106393337249756,
+      "learning_rate": 0.00037653349446757303,
+      "loss": 2.9779,
+      "step": 25581
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5462380051612854,
+      "learning_rate": 0.0003765186076116495,
+      "loss": 3.2093,
+      "step": 25582
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5549441576004028,
+      "learning_rate": 0.0003765037205541905,
+      "loss": 3.0289,
+      "step": 25583
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5625503063201904,
+      "learning_rate": 0.0003764888332952355,
+      "loss": 3.1362,
+      "step": 25584
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.582071840763092,
+      "learning_rate": 0.0003764739458348237,
+      "loss": 3.035,
+      "step": 25585
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5360458493232727,
+      "learning_rate": 0.0003764590581729942,
+      "loss": 2.8962,
+      "step": 25586
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5536467432975769,
+      "learning_rate": 0.0003764441703097864,
+      "loss": 3.0696,
+      "step": 25587
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5753385424613953,
+      "learning_rate": 0.0003764292822452392,
+      "loss": 3.1238,
+      "step": 25588
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5996004343032837,
+      "learning_rate": 0.00037641439397939196,
+      "loss": 3.0104,
+      "step": 25589
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5788471698760986,
+      "learning_rate": 0.00037639950551228397,
+      "loss": 3.1478,
+      "step": 25590
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5919115543365479,
+      "learning_rate": 0.00037638461684395445,
+      "loss": 3.2666,
+      "step": 25591
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5502909421920776,
+      "learning_rate": 0.0003763697279744424,
+      "loss": 3.3325,
+      "step": 25592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5541229844093323,
+      "learning_rate": 0.0003763548389037873,
+      "loss": 3.0699,
+      "step": 25593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5316629409790039,
+      "learning_rate": 0.0003763399496320282,
+      "loss": 3.2613,
+      "step": 25594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5380682349205017,
+      "learning_rate": 0.00037632506015920433,
+      "loss": 3.2379,
+      "step": 25595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6165295243263245,
+      "learning_rate": 0.00037631017048535503,
+      "loss": 3.3058,
+      "step": 25596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5723467469215393,
+      "learning_rate": 0.0003762952806105193,
+      "loss": 3.1968,
+      "step": 25597
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5385229587554932,
+      "learning_rate": 0.0003762803905347367,
+      "loss": 3.1022,
+      "step": 25598
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5821444392204285,
+      "learning_rate": 0.0003762655002580461,
+      "loss": 3.036,
+      "step": 25599
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5753803253173828,
+      "learning_rate": 0.0003762506097804868,
+      "loss": 3.0051,
+      "step": 25600
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5555531978607178,
+      "learning_rate": 0.00037623571910209817,
+      "loss": 3.2474,
+      "step": 25601
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5928922295570374,
+      "learning_rate": 0.00037622082822291926,
+      "loss": 3.2314,
+      "step": 25602
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5480112433433533,
+      "learning_rate": 0.00037620593714298936,
+      "loss": 3.2276,
+      "step": 25603
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5796836018562317,
+      "learning_rate": 0.00037619104586234767,
+      "loss": 2.9367,
+      "step": 25604
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.535346508026123,
+      "learning_rate": 0.0003761761543810335,
+      "loss": 3.2006,
+      "step": 25605
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6117957830429077,
+      "learning_rate": 0.0003761612626990859,
+      "loss": 3.0288,
+      "step": 25606
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5398089289665222,
+      "learning_rate": 0.0003761463708165443,
+      "loss": 3.1357,
+      "step": 25607
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5904059410095215,
+      "learning_rate": 0.0003761314787334477,
+      "loss": 3.1091,
+      "step": 25608
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5292810797691345,
+      "learning_rate": 0.00037611658644983565,
+      "loss": 2.9714,
+      "step": 25609
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5337926149368286,
+      "learning_rate": 0.0003761016939657469,
+      "loss": 3.1214,
+      "step": 25610
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5683236122131348,
+      "learning_rate": 0.00037608680128122104,
+      "loss": 2.8445,
+      "step": 25611
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5837772488594055,
+      "learning_rate": 0.0003760719083962973,
+      "loss": 3.0071,
+      "step": 25612
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5842117071151733,
+      "learning_rate": 0.00037605701531101467,
+      "loss": 2.9642,
+      "step": 25613
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5550433993339539,
+      "learning_rate": 0.0003760421220254125,
+      "loss": 3.0106,
+      "step": 25614
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5435743927955627,
+      "learning_rate": 0.00037602722853953,
+      "loss": 3.0388,
+      "step": 25615
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6093657612800598,
+      "learning_rate": 0.00037601233485340645,
+      "loss": 3.0343,
+      "step": 25616
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5637142062187195,
+      "learning_rate": 0.0003759974409670811,
+      "loss": 3.1616,
+      "step": 25617
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5223868489265442,
+      "learning_rate": 0.000375982546880593,
+      "loss": 3.189,
+      "step": 25618
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.578644871711731,
+      "learning_rate": 0.00037596765259398157,
+      "loss": 2.9092,
+      "step": 25619
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5757935643196106,
+      "learning_rate": 0.00037595275810728597,
+      "loss": 3.1162,
+      "step": 25620
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6163303852081299,
+      "learning_rate": 0.00037593786342054536,
+      "loss": 3.0835,
+      "step": 25621
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5282571315765381,
+      "learning_rate": 0.00037592296853379906,
+      "loss": 3.0402,
+      "step": 25622
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5777333378791809,
+      "learning_rate": 0.00037590807344708624,
+      "loss": 3.1,
+      "step": 25623
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5467130541801453,
+      "learning_rate": 0.0003758931781604463,
+      "loss": 3.2504,
+      "step": 25624
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5582188367843628,
+      "learning_rate": 0.00037587828267391814,
+      "loss": 3.0181,
+      "step": 25625
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5501718521118164,
+      "learning_rate": 0.0003758633869875413,
+      "loss": 2.9322,
+      "step": 25626
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5445879101753235,
+      "learning_rate": 0.0003758484911013549,
+      "loss": 3.2152,
+      "step": 25627
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5596219301223755,
+      "learning_rate": 0.0003758335950153982,
+      "loss": 3.0867,
+      "step": 25628
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5973536968231201,
+      "learning_rate": 0.00037581869872971036,
+      "loss": 2.9504,
+      "step": 25629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5466140508651733,
+      "learning_rate": 0.0003758038022443306,
+      "loss": 2.9407,
+      "step": 25630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6324753165245056,
+      "learning_rate": 0.0003757889055592983,
+      "loss": 3.0794,
+      "step": 25631
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6060929298400879,
+      "learning_rate": 0.0003757740086746526,
+      "loss": 3.026,
+      "step": 25632
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5800380110740662,
+      "learning_rate": 0.00037575911159043273,
+      "loss": 3.027,
+      "step": 25633
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5752260684967041,
+      "learning_rate": 0.00037574421430667794,
+      "loss": 3.1041,
+      "step": 25634
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5729960799217224,
+      "learning_rate": 0.00037572931682342735,
+      "loss": 3.0167,
+      "step": 25635
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5333654284477234,
+      "learning_rate": 0.0003757144191407204,
+      "loss": 3.3717,
+      "step": 25636
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5531308650970459,
+      "learning_rate": 0.0003756995212585963,
+      "loss": 3.0053,
+      "step": 25637
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.563683271408081,
+      "learning_rate": 0.00037568462317709423,
+      "loss": 3.0348,
+      "step": 25638
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5568251609802246,
+      "learning_rate": 0.00037566972489625345,
+      "loss": 2.9534,
+      "step": 25639
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5349379181861877,
+      "learning_rate": 0.0003756548264161131,
+      "loss": 3.0158,
+      "step": 25640
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5677779316902161,
+      "learning_rate": 0.0003756399277367125,
+      "loss": 3.0896,
+      "step": 25641
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5582015514373779,
+      "learning_rate": 0.0003756250288580909,
+      "loss": 3.0478,
+      "step": 25642
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.539047360420227,
+      "learning_rate": 0.00037561012978028755,
+      "loss": 2.9591,
+      "step": 25643
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5575973391532898,
+      "learning_rate": 0.0003755952305033416,
+      "loss": 3.1518,
+      "step": 25644
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5632666349411011,
+      "learning_rate": 0.00037558033102729247,
+      "loss": 3.1166,
+      "step": 25645
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5250673890113831,
+      "learning_rate": 0.00037556543135217925,
+      "loss": 3.0891,
+      "step": 25646
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.571751594543457,
+      "learning_rate": 0.00037555053147804115,
+      "loss": 2.9595,
+      "step": 25647
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5709319114685059,
+      "learning_rate": 0.0003755356314049176,
+      "loss": 2.9572,
+      "step": 25648
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6075392961502075,
+      "learning_rate": 0.0003755207311328478,
+      "loss": 3.1889,
+      "step": 25649
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5540075898170471,
+      "learning_rate": 0.00037550583066187077,
+      "loss": 3.046,
+      "step": 25650
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5717571973800659,
+      "learning_rate": 0.0003754909299920259,
+      "loss": 3.1389,
+      "step": 25651
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5485336780548096,
+      "learning_rate": 0.00037547602912335257,
+      "loss": 2.9172,
+      "step": 25652
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6030722260475159,
+      "learning_rate": 0.00037546112805588984,
+      "loss": 2.8894,
+      "step": 25653
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5882648825645447,
+      "learning_rate": 0.000375446226789677,
+      "loss": 2.9703,
+      "step": 25654
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5773414969444275,
+      "learning_rate": 0.0003754313253247533,
+      "loss": 3.2545,
+      "step": 25655
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5464721322059631,
+      "learning_rate": 0.00037541642366115813,
+      "loss": 3.2103,
+      "step": 25656
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.59247225522995,
+      "learning_rate": 0.00037540152179893054,
+      "loss": 3.0535,
+      "step": 25657
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5408132076263428,
+      "learning_rate": 0.0003753866197381098,
+      "loss": 2.9782,
+      "step": 25658
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5488524436950684,
+      "learning_rate": 0.0003753717174787352,
+      "loss": 2.9827,
+      "step": 25659
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5907993316650391,
+      "learning_rate": 0.0003753568150208462,
+      "loss": 3.2103,
+      "step": 25660
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.568075954914093,
+      "learning_rate": 0.0003753419123644816,
+      "loss": 2.89,
+      "step": 25661
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6084296703338623,
+      "learning_rate": 0.000375327009509681,
+      "loss": 3.1351,
+      "step": 25662
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5559454560279846,
+      "learning_rate": 0.00037531210645648364,
+      "loss": 3.2423,
+      "step": 25663
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5514500141143799,
+      "learning_rate": 0.00037529720320492865,
+      "loss": 3.065,
+      "step": 25664
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5497595071792603,
+      "learning_rate": 0.00037528229975505525,
+      "loss": 3.0918,
+      "step": 25665
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5739080905914307,
+      "learning_rate": 0.00037526739610690274,
+      "loss": 3.0684,
+      "step": 25666
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5830264687538147,
+      "learning_rate": 0.00037525249226051053,
+      "loss": 3.192,
+      "step": 25667
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5546814203262329,
+      "learning_rate": 0.00037523758821591765,
+      "loss": 3.0591,
+      "step": 25668
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5485287308692932,
+      "learning_rate": 0.00037522268397316346,
+      "loss": 3.1005,
+      "step": 25669
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5739904642105103,
+      "learning_rate": 0.0003752077795322872,
+      "loss": 3.0651,
+      "step": 25670
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6011346578598022,
+      "learning_rate": 0.0003751928748933281,
+      "loss": 3.1442,
+      "step": 25671
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5489895939826965,
+      "learning_rate": 0.0003751779700563254,
+      "loss": 2.8407,
+      "step": 25672
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5272802710533142,
+      "learning_rate": 0.00037516306502131843,
+      "loss": 3.075,
+      "step": 25673
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6106321811676025,
+      "learning_rate": 0.0003751481597883464,
+      "loss": 3.0871,
+      "step": 25674
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5586559772491455,
+      "learning_rate": 0.0003751332543574487,
+      "loss": 2.8499,
+      "step": 25675
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5668689012527466,
+      "learning_rate": 0.0003751183487286644,
+      "loss": 3.2104,
+      "step": 25676
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5656947493553162,
+      "learning_rate": 0.0003751034429020327,
+      "loss": 3.0308,
+      "step": 25677
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5509669184684753,
+      "learning_rate": 0.0003750885368775932,
+      "loss": 3.2247,
+      "step": 25678
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.581092894077301,
+      "learning_rate": 0.00037507363065538476,
+      "loss": 3.0334,
+      "step": 25679
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5440135598182678,
+      "learning_rate": 0.0003750587242354469,
+      "loss": 3.162,
+      "step": 25680
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5443293452262878,
+      "learning_rate": 0.0003750438176178188,
+      "loss": 3.1321,
+      "step": 25681
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5595262050628662,
+      "learning_rate": 0.00037502891080253974,
+      "loss": 3.1387,
+      "step": 25682
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5496951937675476,
+      "learning_rate": 0.00037501400378964894,
+      "loss": 2.9888,
+      "step": 25683
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5384321212768555,
+      "learning_rate": 0.00037499909657918567,
+      "loss": 3.1049,
+      "step": 25684
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6079148650169373,
+      "learning_rate": 0.0003749841891711893,
+      "loss": 3.1243,
+      "step": 25685
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5618904829025269,
+      "learning_rate": 0.000374969281565699,
+      "loss": 2.9805,
+      "step": 25686
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6069911122322083,
+      "learning_rate": 0.00037495437376275397,
+      "loss": 3.343,
+      "step": 25687
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7043254971504211,
+      "learning_rate": 0.0003749394657623936,
+      "loss": 3.1197,
+      "step": 25688
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5782732367515564,
+      "learning_rate": 0.00037492455756465705,
+      "loss": 2.9482,
+      "step": 25689
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5924282670021057,
+      "learning_rate": 0.00037490964916958363,
+      "loss": 3.2733,
+      "step": 25690
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5641574859619141,
+      "learning_rate": 0.0003748947405772127,
+      "loss": 3.0016,
+      "step": 25691
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.596458375453949,
+      "learning_rate": 0.0003748798317875833,
+      "loss": 3.0319,
+      "step": 25692
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6097240447998047,
+      "learning_rate": 0.00037486492280073495,
+      "loss": 3.3717,
+      "step": 25693
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5604461431503296,
+      "learning_rate": 0.00037485001361670675,
+      "loss": 2.8643,
+      "step": 25694
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5882611274719238,
+      "learning_rate": 0.000374835104235538,
+      "loss": 2.9761,
+      "step": 25695
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5667465925216675,
+      "learning_rate": 0.000374820194657268,
+      "loss": 3.1456,
+      "step": 25696
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5910031199455261,
+      "learning_rate": 0.00037480528488193605,
+      "loss": 2.9162,
+      "step": 25697
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5716198682785034,
+      "learning_rate": 0.00037479037490958136,
+      "loss": 3.0564,
+      "step": 25698
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5668255090713501,
+      "learning_rate": 0.0003747754647402432,
+      "loss": 3.2102,
+      "step": 25699
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5733356475830078,
+      "learning_rate": 0.00037476055437396093,
+      "loss": 3.3008,
+      "step": 25700
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5971511006355286,
+      "learning_rate": 0.00037474564381077363,
+      "loss": 3.0688,
+      "step": 25701
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5802672505378723,
+      "learning_rate": 0.0003747307330507207,
+      "loss": 2.9328,
+      "step": 25702
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6406810879707336,
+      "learning_rate": 0.0003747158220938415,
+      "loss": 3.0136,
+      "step": 25703
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5840638279914856,
+      "learning_rate": 0.0003747009109401752,
+      "loss": 3.0113,
+      "step": 25704
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.55064457654953,
+      "learning_rate": 0.000374685999589761,
+      "loss": 3.0928,
+      "step": 25705
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5626662373542786,
+      "learning_rate": 0.00037467108804263826,
+      "loss": 2.9046,
+      "step": 25706
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6290438175201416,
+      "learning_rate": 0.00037465617629884625,
+      "loss": 3.0341,
+      "step": 25707
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5759810209274292,
+      "learning_rate": 0.00037464126435842425,
+      "loss": 3.1417,
+      "step": 25708
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6134321689605713,
+      "learning_rate": 0.00037462635222141146,
+      "loss": 3.2309,
+      "step": 25709
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.578365683555603,
+      "learning_rate": 0.0003746114398878473,
+      "loss": 3.1481,
+      "step": 25710
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5462852716445923,
+      "learning_rate": 0.00037459652735777105,
+      "loss": 3.1731,
+      "step": 25711
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.636318027973175,
+      "learning_rate": 0.00037458161463122165,
+      "loss": 2.9947,
+      "step": 25712
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5679839849472046,
+      "learning_rate": 0.00037456670170823883,
+      "loss": 3.122,
+      "step": 25713
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5561462640762329,
+      "learning_rate": 0.0003745517885888617,
+      "loss": 3.2713,
+      "step": 25714
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6204407811164856,
+      "learning_rate": 0.00037453687527312944,
+      "loss": 3.0994,
+      "step": 25715
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.52779620885849,
+      "learning_rate": 0.0003745219617610813,
+      "loss": 3.1557,
+      "step": 25716
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5464897155761719,
+      "learning_rate": 0.00037450704805275676,
+      "loss": 3.1137,
+      "step": 25717
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6134060621261597,
+      "learning_rate": 0.00037449213414819497,
+      "loss": 2.9496,
+      "step": 25718
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5954126119613647,
+      "learning_rate": 0.00037447722004743526,
+      "loss": 3.1065,
+      "step": 25719
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5473684668540955,
+      "learning_rate": 0.00037446230575051686,
+      "loss": 3.0049,
+      "step": 25720
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5591050386428833,
+      "learning_rate": 0.0003744473912574791,
+      "loss": 3.114,
+      "step": 25721
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5836315751075745,
+      "learning_rate": 0.0003744324765683613,
+      "loss": 3.046,
+      "step": 25722
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6122407913208008,
+      "learning_rate": 0.0003744175616832025,
+      "loss": 3.0378,
+      "step": 25723
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.570861279964447,
+      "learning_rate": 0.0003744026466020422,
+      "loss": 3.1949,
+      "step": 25724
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.534397542476654,
+      "learning_rate": 0.00037438773132491984,
+      "loss": 3.0312,
+      "step": 25725
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.536069929599762,
+      "learning_rate": 0.0003743728158518744,
+      "loss": 3.0838,
+      "step": 25726
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5722454190254211,
+      "learning_rate": 0.0003743579001829452,
+      "loss": 3.2043,
+      "step": 25727
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5977911949157715,
+      "learning_rate": 0.0003743429843181716,
+      "loss": 3.0963,
+      "step": 25728
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.566826343536377,
+      "learning_rate": 0.000374328068257593,
+      "loss": 2.9931,
+      "step": 25729
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5475797057151794,
+      "learning_rate": 0.00037431315200124853,
+      "loss": 3.1475,
+      "step": 25730
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5495274662971497,
+      "learning_rate": 0.0003742982355491775,
+      "loss": 3.1989,
+      "step": 25731
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.559955358505249,
+      "learning_rate": 0.00037428331890141926,
+      "loss": 3.015,
+      "step": 25732
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6134682893753052,
+      "learning_rate": 0.000374268402058013,
+      "loss": 2.9699,
+      "step": 25733
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5740057826042175,
+      "learning_rate": 0.0003742534850189981,
+      "loss": 3.2211,
+      "step": 25734
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.610024094581604,
+      "learning_rate": 0.00037423856778441377,
+      "loss": 2.9174,
+      "step": 25735
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6208667755126953,
+      "learning_rate": 0.00037422365035429936,
+      "loss": 3.155,
+      "step": 25736
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5483558773994446,
+      "learning_rate": 0.00037420873272869424,
+      "loss": 3.1108,
+      "step": 25737
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5884615182876587,
+      "learning_rate": 0.00037419381490763745,
+      "loss": 3.1444,
+      "step": 25738
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5598413348197937,
+      "learning_rate": 0.00037417889689116845,
+      "loss": 3.1624,
+      "step": 25739
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5523557066917419,
+      "learning_rate": 0.0003741639786793266,
+      "loss": 3.1961,
+      "step": 25740
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6840699911117554,
+      "learning_rate": 0.00037414906027215104,
+      "loss": 3.0235,
+      "step": 25741
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5679307579994202,
+      "learning_rate": 0.0003741341416696811,
+      "loss": 3.0551,
+      "step": 25742
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5635999441146851,
+      "learning_rate": 0.00037411922287195615,
+      "loss": 3.016,
+      "step": 25743
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.561026394367218,
+      "learning_rate": 0.0003741043038790154,
+      "loss": 2.9945,
+      "step": 25744
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5478890538215637,
+      "learning_rate": 0.0003740893846908982,
+      "loss": 3.12,
+      "step": 25745
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5661895871162415,
+      "learning_rate": 0.0003740744653076438,
+      "loss": 3.1501,
+      "step": 25746
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5741684436798096,
+      "learning_rate": 0.00037405954572929155,
+      "loss": 2.9514,
+      "step": 25747
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5624836087226868,
+      "learning_rate": 0.0003740446259558807,
+      "loss": 3.0239,
+      "step": 25748
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5481595993041992,
+      "learning_rate": 0.0003740297059874505,
+      "loss": 3.1957,
+      "step": 25749
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5887608528137207,
+      "learning_rate": 0.0003740147858240403,
+      "loss": 3.2924,
+      "step": 25750
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5847222805023193,
+      "learning_rate": 0.0003739998654656896,
+      "loss": 3.0791,
+      "step": 25751
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6101052761077881,
+      "learning_rate": 0.00037398494491243736,
+      "loss": 2.9752,
+      "step": 25752
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5880925059318542,
+      "learning_rate": 0.00037397002416432283,
+      "loss": 2.9945,
+      "step": 25753
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5594073534011841,
+      "learning_rate": 0.00037395510322138574,
+      "loss": 3.0858,
+      "step": 25754
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5806262493133545,
+      "learning_rate": 0.00037394018208366505,
+      "loss": 3.1636,
+      "step": 25755
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6017138361930847,
+      "learning_rate": 0.00037392526075120014,
+      "loss": 3.1412,
+      "step": 25756
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5765290260314941,
+      "learning_rate": 0.00037391033922403035,
+      "loss": 3.1801,
+      "step": 25757
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6031667590141296,
+      "learning_rate": 0.00037389541750219494,
+      "loss": 3.0756,
+      "step": 25758
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5250658988952637,
+      "learning_rate": 0.0003738804955857332,
+      "loss": 3.0219,
+      "step": 25759
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5491588115692139,
+      "learning_rate": 0.0003738655734746845,
+      "loss": 2.9849,
+      "step": 25760
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5519552826881409,
+      "learning_rate": 0.000373850651169088,
+      "loss": 3.1823,
+      "step": 25761
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5483062267303467,
+      "learning_rate": 0.0003738357286689832,
+      "loss": 3.0518,
+      "step": 25762
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5620872378349304,
+      "learning_rate": 0.0003738208059744093,
+      "loss": 3.1545,
+      "step": 25763
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5591562986373901,
+      "learning_rate": 0.00037380588308540553,
+      "loss": 3.0748,
+      "step": 25764
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6279521584510803,
+      "learning_rate": 0.0003737909600020113,
+      "loss": 3.0854,
+      "step": 25765
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6239844560623169,
+      "learning_rate": 0.0003737760367242659,
+      "loss": 3.1755,
+      "step": 25766
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.536332905292511,
+      "learning_rate": 0.0003737611132522087,
+      "loss": 3.1022,
+      "step": 25767
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5568017363548279,
+      "learning_rate": 0.00037374618958587873,
+      "loss": 3.2276,
+      "step": 25768
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.546877920627594,
+      "learning_rate": 0.0003737312657253156,
+      "loss": 2.9514,
+      "step": 25769
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6245750784873962,
+      "learning_rate": 0.0003737163416705585,
+      "loss": 3.0925,
+      "step": 25770
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.609423816204071,
+      "learning_rate": 0.0003737014174216467,
+      "loss": 3.0713,
+      "step": 25771
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5926220417022705,
+      "learning_rate": 0.0003736864929786196,
+      "loss": 3.1992,
+      "step": 25772
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5451133251190186,
+      "learning_rate": 0.0003736715683415165,
+      "loss": 3.0744,
+      "step": 25773
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5710421204566956,
+      "learning_rate": 0.00037365664351037654,
+      "loss": 3.0953,
+      "step": 25774
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5600199699401855,
+      "learning_rate": 0.0003736417184852393,
+      "loss": 3.1269,
+      "step": 25775
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5701009035110474,
+      "learning_rate": 0.00037362679326614386,
+      "loss": 3.1154,
+      "step": 25776
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6089168787002563,
+      "learning_rate": 0.0003736118678531297,
+      "loss": 3.0837,
+      "step": 25777
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5491774082183838,
+      "learning_rate": 0.0003735969422462359,
+      "loss": 3.3128,
+      "step": 25778
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5762932300567627,
+      "learning_rate": 0.000373582016445502,
+      "loss": 3.0711,
+      "step": 25779
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5476071834564209,
+      "learning_rate": 0.0003735670904509673,
+      "loss": 2.887,
+      "step": 25780
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6149681806564331,
+      "learning_rate": 0.00037355216426267093,
+      "loss": 3.0442,
+      "step": 25781
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5968160033226013,
+      "learning_rate": 0.00037353723788065236,
+      "loss": 2.8897,
+      "step": 25782
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5533415675163269,
+      "learning_rate": 0.00037352231130495083,
+      "loss": 3.1161,
+      "step": 25783
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5641621351242065,
+      "learning_rate": 0.00037350738453560575,
+      "loss": 3.1067,
+      "step": 25784
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5823823809623718,
+      "learning_rate": 0.00037349245757265626,
+      "loss": 3.1385,
+      "step": 25785
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6093672513961792,
+      "learning_rate": 0.0003734775304161419,
+      "loss": 3.1447,
+      "step": 25786
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5846396684646606,
+      "learning_rate": 0.00037346260306610175,
+      "loss": 2.9545,
+      "step": 25787
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5794227719306946,
+      "learning_rate": 0.0003734476755225754,
+      "loss": 2.8661,
+      "step": 25788
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.566815972328186,
+      "learning_rate": 0.0003734327477856018,
+      "loss": 3.0093,
+      "step": 25789
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5866063833236694,
+      "learning_rate": 0.00037341781985522057,
+      "loss": 2.9509,
+      "step": 25790
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.552588939666748,
+      "learning_rate": 0.000373402891731471,
+      "loss": 3.1778,
+      "step": 25791
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5920195579528809,
+      "learning_rate": 0.0003733879634143923,
+      "loss": 3.0382,
+      "step": 25792
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5480687618255615,
+      "learning_rate": 0.0003733730349040238,
+      "loss": 3.0309,
+      "step": 25793
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5250555276870728,
+      "learning_rate": 0.0003733581062004049,
+      "loss": 3.2171,
+      "step": 25794
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6039732694625854,
+      "learning_rate": 0.0003733431773035748,
+      "loss": 3.1888,
+      "step": 25795
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5604811310768127,
+      "learning_rate": 0.0003733282482135729,
+      "loss": 3.1069,
+      "step": 25796
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.597496509552002,
+      "learning_rate": 0.0003733133189304385,
+      "loss": 3.1567,
+      "step": 25797
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5957289338111877,
+      "learning_rate": 0.00037329838945421095,
+      "loss": 3.0694,
+      "step": 25798
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5216500163078308,
+      "learning_rate": 0.00037328345978492966,
+      "loss": 3.0539,
+      "step": 25799
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5505014061927795,
+      "learning_rate": 0.0003732685299226336,
+      "loss": 3.1195,
+      "step": 25800
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5500777363777161,
+      "learning_rate": 0.0003732535998673624,
+      "loss": 3.007,
+      "step": 25801
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6095644235610962,
+      "learning_rate": 0.00037323866961915545,
+      "loss": 2.8892,
+      "step": 25802
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5533205270767212,
+      "learning_rate": 0.0003732237391780519,
+      "loss": 3.0392,
+      "step": 25803
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.576029360294342,
+      "learning_rate": 0.0003732088085440911,
+      "loss": 2.8924,
+      "step": 25804
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5615506172180176,
+      "learning_rate": 0.00037319387771731237,
+      "loss": 3.0801,
+      "step": 25805
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5568882822990417,
+      "learning_rate": 0.000373178946697755,
+      "loss": 3.1054,
+      "step": 25806
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.554295003414154,
+      "learning_rate": 0.00037316401548545845,
+      "loss": 3.1786,
+      "step": 25807
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5566840767860413,
+      "learning_rate": 0.0003731490840804619,
+      "loss": 3.0353,
+      "step": 25808
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5562843084335327,
+      "learning_rate": 0.00037313415248280477,
+      "loss": 3.3003,
+      "step": 25809
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.569489061832428,
+      "learning_rate": 0.00037311922069252636,
+      "loss": 3.155,
+      "step": 25810
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.537324070930481,
+      "learning_rate": 0.00037310428870966595,
+      "loss": 3.0376,
+      "step": 25811
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5831682085990906,
+      "learning_rate": 0.00037308935653426295,
+      "loss": 3.1946,
+      "step": 25812
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5841866135597229,
+      "learning_rate": 0.0003730744241663567,
+      "loss": 3.2204,
+      "step": 25813
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5594460368156433,
+      "learning_rate": 0.0003730594916059864,
+      "loss": 3.1202,
+      "step": 25814
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.635952353477478,
+      "learning_rate": 0.0003730445588531915,
+      "loss": 3.0469,
+      "step": 25815
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5496728420257568,
+      "learning_rate": 0.00037302962590801133,
+      "loss": 3.0797,
+      "step": 25816
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5622497797012329,
+      "learning_rate": 0.00037301469277048515,
+      "loss": 3.0121,
+      "step": 25817
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5610560774803162,
+      "learning_rate": 0.0003729997594406523,
+      "loss": 2.9417,
+      "step": 25818
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5938122272491455,
+      "learning_rate": 0.0003729848259185521,
+      "loss": 2.8371,
+      "step": 25819
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5797512531280518,
+      "learning_rate": 0.0003729698922042239,
+      "loss": 2.8959,
+      "step": 25820
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5663214921951294,
+      "learning_rate": 0.00037295495829770713,
+      "loss": 3.0954,
+      "step": 25821
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5849252343177795,
+      "learning_rate": 0.00037294002419904104,
+      "loss": 2.9451,
+      "step": 25822
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5354498028755188,
+      "learning_rate": 0.00037292508990826494,
+      "loss": 3.0579,
+      "step": 25823
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5267759561538696,
+      "learning_rate": 0.0003729101554254182,
+      "loss": 3.1529,
+      "step": 25824
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5743757486343384,
+      "learning_rate": 0.00037289522075054007,
+      "loss": 2.9971,
+      "step": 25825
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5666912198066711,
+      "learning_rate": 0.00037288028588367,
+      "loss": 3.1979,
+      "step": 25826
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5688309669494629,
+      "learning_rate": 0.0003728653508248474,
+      "loss": 3.0861,
+      "step": 25827
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5575060248374939,
+      "learning_rate": 0.00037285041557411135,
+      "loss": 3.1098,
+      "step": 25828
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5437538623809814,
+      "learning_rate": 0.0003728354801315014,
+      "loss": 3.1594,
+      "step": 25829
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6106091141700745,
+      "learning_rate": 0.00037282054449705665,
+      "loss": 3.0998,
+      "step": 25830
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6044281125068665,
+      "learning_rate": 0.0003728056086708168,
+      "loss": 3.0176,
+      "step": 25831
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5630094408988953,
+      "learning_rate": 0.00037279067265282094,
+      "loss": 3.1904,
+      "step": 25832
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5635753273963928,
+      "learning_rate": 0.0003727757364431084,
+      "loss": 3.1233,
+      "step": 25833
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5579873919487,
+      "learning_rate": 0.0003727608000417186,
+      "loss": 2.8207,
+      "step": 25834
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5992753505706787,
+      "learning_rate": 0.0003727458634486909,
+      "loss": 3.0802,
+      "step": 25835
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5542998313903809,
+      "learning_rate": 0.0003727309266640645,
+      "loss": 3.1656,
+      "step": 25836
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5979121327400208,
+      "learning_rate": 0.0003727159896878789,
+      "loss": 3.2221,
+      "step": 25837
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5464156270027161,
+      "learning_rate": 0.00037270105252017335,
+      "loss": 3.0004,
+      "step": 25838
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5646561980247498,
+      "learning_rate": 0.00037268611516098725,
+      "loss": 3.3144,
+      "step": 25839
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.60554438829422,
+      "learning_rate": 0.00037267117761036,
+      "loss": 3.0871,
+      "step": 25840
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5654262900352478,
+      "learning_rate": 0.0003726562398683307,
+      "loss": 2.938,
+      "step": 25841
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5625606179237366,
+      "learning_rate": 0.0003726413019349389,
+      "loss": 3.161,
+      "step": 25842
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5684981346130371,
+      "learning_rate": 0.0003726263638102239,
+      "loss": 3.0923,
+      "step": 25843
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5840631723403931,
+      "learning_rate": 0.00037261142549422505,
+      "loss": 3.1465,
+      "step": 25844
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5702567100524902,
+      "learning_rate": 0.0003725964869869816,
+      "loss": 3.1471,
+      "step": 25845
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.61046302318573,
+      "learning_rate": 0.0003725815482885331,
+      "loss": 3.1633,
+      "step": 25846
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.557167649269104,
+      "learning_rate": 0.00037256660939891865,
+      "loss": 3.0677,
+      "step": 25847
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5500303506851196,
+      "learning_rate": 0.00037255167031817775,
+      "loss": 3.3199,
+      "step": 25848
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5514801144599915,
+      "learning_rate": 0.00037253673104634973,
+      "loss": 3.1673,
+      "step": 25849
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5589605569839478,
+      "learning_rate": 0.000372521791583474,
+      "loss": 3.0712,
+      "step": 25850
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5684011578559875,
+      "learning_rate": 0.0003725068519295897,
+      "loss": 3.0508,
+      "step": 25851
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.559701144695282,
+      "learning_rate": 0.0003724919120847363,
+      "loss": 3.0181,
+      "step": 25852
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5717242360115051,
+      "learning_rate": 0.00037247697204895327,
+      "loss": 2.9951,
+      "step": 25853
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5555175542831421,
+      "learning_rate": 0.0003724620318222798,
+      "loss": 2.9823,
+      "step": 25854
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5793154239654541,
+      "learning_rate": 0.00037244709140475527,
+      "loss": 3.1666,
+      "step": 25855
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.552769660949707,
+      "learning_rate": 0.0003724321507964191,
+      "loss": 3.2035,
+      "step": 25856
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5471254587173462,
+      "learning_rate": 0.0003724172099973105,
+      "loss": 3.1123,
+      "step": 25857
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5685672760009766,
+      "learning_rate": 0.0003724022690074689,
+      "loss": 3.0484,
+      "step": 25858
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5822762250900269,
+      "learning_rate": 0.0003723873278269337,
+      "loss": 3.0311,
+      "step": 25859
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5946375131607056,
+      "learning_rate": 0.0003723723864557442,
+      "loss": 3.2406,
+      "step": 25860
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5624514818191528,
+      "learning_rate": 0.0003723574448939398,
+      "loss": 3.0958,
+      "step": 25861
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6144577264785767,
+      "learning_rate": 0.0003723425031415598,
+      "loss": 3.117,
+      "step": 25862
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5530250668525696,
+      "learning_rate": 0.0003723275611986435,
+      "loss": 3.0762,
+      "step": 25863
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5647308230400085,
+      "learning_rate": 0.0003723126190652304,
+      "loss": 3.2921,
+      "step": 25864
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5888532996177673,
+      "learning_rate": 0.0003722976767413598,
+      "loss": 3.1137,
+      "step": 25865
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5185219645500183,
+      "learning_rate": 0.0003722827342270709,
+      "loss": 3.0245,
+      "step": 25866
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5749824643135071,
+      "learning_rate": 0.00037226779152240336,
+      "loss": 3.186,
+      "step": 25867
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5528282523155212,
+      "learning_rate": 0.0003722528486273962,
+      "loss": 3.329,
+      "step": 25868
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5749999284744263,
+      "learning_rate": 0.00037223790554208905,
+      "loss": 3.1513,
+      "step": 25869
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5693772435188293,
+      "learning_rate": 0.00037222296226652115,
+      "loss": 3.1707,
+      "step": 25870
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5503202080726624,
+      "learning_rate": 0.0003722080188007318,
+      "loss": 3.2248,
+      "step": 25871
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.639018714427948,
+      "learning_rate": 0.0003721930751447605,
+      "loss": 3.225,
+      "step": 25872
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5662673711776733,
+      "learning_rate": 0.00037217813129864646,
+      "loss": 3.1343,
+      "step": 25873
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6173918843269348,
+      "learning_rate": 0.0003721631872624292,
+      "loss": 3.1776,
+      "step": 25874
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5484563112258911,
+      "learning_rate": 0.000372148243036148,
+      "loss": 3.0044,
+      "step": 25875
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.55157870054245,
+      "learning_rate": 0.00037213329861984215,
+      "loss": 3.0107,
+      "step": 25876
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5684902667999268,
+      "learning_rate": 0.000372118354013551,
+      "loss": 2.8688,
+      "step": 25877
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5629575252532959,
+      "learning_rate": 0.00037210340921731415,
+      "loss": 3.3597,
+      "step": 25878
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5630050897598267,
+      "learning_rate": 0.0003720884642311707,
+      "loss": 2.9911,
+      "step": 25879
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5761358141899109,
+      "learning_rate": 0.0003720735190551602,
+      "loss": 3.1811,
+      "step": 25880
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5436105132102966,
+      "learning_rate": 0.0003720585736893218,
+      "loss": 2.9532,
+      "step": 25881
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6045496463775635,
+      "learning_rate": 0.00037204362813369503,
+      "loss": 2.8204,
+      "step": 25882
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5377562642097473,
+      "learning_rate": 0.0003720286823883192,
+      "loss": 3.1401,
+      "step": 25883
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5733555555343628,
+      "learning_rate": 0.0003720137364532337,
+      "loss": 3.1384,
+      "step": 25884
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5425565242767334,
+      "learning_rate": 0.00037199879032847786,
+      "loss": 2.9473,
+      "step": 25885
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5795601606369019,
+      "learning_rate": 0.00037198384401409114,
+      "loss": 2.8055,
+      "step": 25886
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5669113993644714,
+      "learning_rate": 0.0003719688975101127,
+      "loss": 3.1459,
+      "step": 25887
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5572779178619385,
+      "learning_rate": 0.0003719539508165821,
+      "loss": 3.1748,
+      "step": 25888
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5794333219528198,
+      "learning_rate": 0.0003719390039335387,
+      "loss": 3.1461,
+      "step": 25889
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5872516632080078,
+      "learning_rate": 0.00037192405686102174,
+      "loss": 2.8606,
+      "step": 25890
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5817949175834656,
+      "learning_rate": 0.0003719091095990707,
+      "loss": 3.0884,
+      "step": 25891
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.590277910232544,
+      "learning_rate": 0.00037189416214772477,
+      "loss": 3.0184,
+      "step": 25892
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5216969847679138,
+      "learning_rate": 0.00037187921450702367,
+      "loss": 3.1245,
+      "step": 25893
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.545471727848053,
+      "learning_rate": 0.0003718642666770064,
+      "loss": 3.1712,
+      "step": 25894
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5600536465644836,
+      "learning_rate": 0.0003718493186577125,
+      "loss": 2.9244,
+      "step": 25895
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5645468831062317,
+      "learning_rate": 0.00037183437044918134,
+      "loss": 3.2514,
+      "step": 25896
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5973244309425354,
+      "learning_rate": 0.0003718194220514523,
+      "loss": 2.9894,
+      "step": 25897
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5746738314628601,
+      "learning_rate": 0.0003718044734645647,
+      "loss": 2.8643,
+      "step": 25898
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5265095829963684,
+      "learning_rate": 0.00037178952468855793,
+      "loss": 3.2349,
+      "step": 25899
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5884544849395752,
+      "learning_rate": 0.0003717745757234714,
+      "loss": 3.1012,
+      "step": 25900
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5789063572883606,
+      "learning_rate": 0.00037175962656934447,
+      "loss": 3.0908,
+      "step": 25901
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.568000316619873,
+      "learning_rate": 0.0003717446772262163,
+      "loss": 2.9355,
+      "step": 25902
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5491638779640198,
+      "learning_rate": 0.0003717297276941267,
+      "loss": 3.2674,
+      "step": 25903
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5411843061447144,
+      "learning_rate": 0.00037171477797311473,
+      "loss": 3.0928,
+      "step": 25904
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6108878254890442,
+      "learning_rate": 0.0003716998280632198,
+      "loss": 3.2653,
+      "step": 25905
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5373321771621704,
+      "learning_rate": 0.00037168487796448134,
+      "loss": 3.0299,
+      "step": 25906
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5621278882026672,
+      "learning_rate": 0.0003716699276769387,
+      "loss": 3.019,
+      "step": 25907
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5679013133049011,
+      "learning_rate": 0.00037165497720063125,
+      "loss": 3.3508,
+      "step": 25908
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5926734209060669,
+      "learning_rate": 0.0003716400265355984,
+      "loss": 3.0921,
+      "step": 25909
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5435253381729126,
+      "learning_rate": 0.00037162507568187954,
+      "loss": 2.8687,
+      "step": 25910
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5271333456039429,
+      "learning_rate": 0.0003716101246395139,
+      "loss": 2.9884,
+      "step": 25911
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5629345774650574,
+      "learning_rate": 0.00037159517340854117,
+      "loss": 2.8821,
+      "step": 25912
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5234237313270569,
+      "learning_rate": 0.0003715802219890003,
+      "loss": 3.2772,
+      "step": 25913
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5610054135322571,
+      "learning_rate": 0.00037156527038093105,
+      "loss": 3.1372,
+      "step": 25914
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5486628413200378,
+      "learning_rate": 0.0003715503185843727,
+      "loss": 3.1335,
+      "step": 25915
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5305410027503967,
+      "learning_rate": 0.00037153536659936446,
+      "loss": 3.053,
+      "step": 25916
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6576041579246521,
+      "learning_rate": 0.00037152041442594584,
+      "loss": 3.1786,
+      "step": 25917
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5794118046760559,
+      "learning_rate": 0.00037150546206415625,
+      "loss": 3.1182,
+      "step": 25918
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5535377264022827,
+      "learning_rate": 0.00037149050951403506,
+      "loss": 2.9718,
+      "step": 25919
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5769936442375183,
+      "learning_rate": 0.0003714755567756216,
+      "loss": 2.987,
+      "step": 25920
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.59618079662323,
+      "learning_rate": 0.00037146060384895527,
+      "loss": 3.2961,
+      "step": 25921
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5642500519752502,
+      "learning_rate": 0.00037144565073407544,
+      "loss": 2.9942,
+      "step": 25922
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5690189003944397,
+      "learning_rate": 0.0003714306974310215,
+      "loss": 3.0254,
+      "step": 25923
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5622614026069641,
+      "learning_rate": 0.0003714157439398329,
+      "loss": 3.1844,
+      "step": 25924
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5633580684661865,
+      "learning_rate": 0.000371400790260549,
+      "loss": 3.1334,
+      "step": 25925
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5582607984542847,
+      "learning_rate": 0.00037138583639320915,
+      "loss": 3.026,
+      "step": 25926
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5949597954750061,
+      "learning_rate": 0.00037137088233785273,
+      "loss": 3.076,
+      "step": 25927
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5441601872444153,
+      "learning_rate": 0.000371355928094519,
+      "loss": 3.0003,
+      "step": 25928
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.606289803981781,
+      "learning_rate": 0.00037134097366324774,
+      "loss": 3.1652,
+      "step": 25929
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.551216185092926,
+      "learning_rate": 0.000371326019044078,
+      "loss": 3.121,
+      "step": 25930
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5833144783973694,
+      "learning_rate": 0.0003713110642370492,
+      "loss": 3.4873,
+      "step": 25931
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5816163420677185,
+      "learning_rate": 0.00037129610924220075,
+      "loss": 3.1136,
+      "step": 25932
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5476531386375427,
+      "learning_rate": 0.00037128115405957217,
+      "loss": 3.0812,
+      "step": 25933
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5457040071487427,
+      "learning_rate": 0.0003712661986892027,
+      "loss": 3.187,
+      "step": 25934
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5884016156196594,
+      "learning_rate": 0.0003712512431311317,
+      "loss": 3.158,
+      "step": 25935
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5727925300598145,
+      "learning_rate": 0.0003712362873853988,
+      "loss": 3.2241,
+      "step": 25936
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5885441303253174,
+      "learning_rate": 0.00037122133145204314,
+      "loss": 3.1813,
+      "step": 25937
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6078060865402222,
+      "learning_rate": 0.00037120637533110415,
+      "loss": 3.0894,
+      "step": 25938
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.51329106092453,
+      "learning_rate": 0.00037119141902262133,
+      "loss": 3.0433,
+      "step": 25939
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6121600866317749,
+      "learning_rate": 0.0003711764625266341,
+      "loss": 2.934,
+      "step": 25940
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5866429805755615,
+      "learning_rate": 0.0003711615058431816,
+      "loss": 2.975,
+      "step": 25941
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5531361103057861,
+      "learning_rate": 0.00037114654897230353,
+      "loss": 3.0505,
+      "step": 25942
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7733150124549866,
+      "learning_rate": 0.00037113159191403896,
+      "loss": 3.2246,
+      "step": 25943
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5927426218986511,
+      "learning_rate": 0.0003711166346684276,
+      "loss": 2.9347,
+      "step": 25944
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5862705707550049,
+      "learning_rate": 0.00037110167723550876,
+      "loss": 2.817,
+      "step": 25945
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6004965901374817,
+      "learning_rate": 0.0003710867196153217,
+      "loss": 3.1906,
+      "step": 25946
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.563130259513855,
+      "learning_rate": 0.0003710717618079059,
+      "loss": 3.124,
+      "step": 25947
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5575785636901855,
+      "learning_rate": 0.0003710568038133008,
+      "loss": 3.0767,
+      "step": 25948
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5379173159599304,
+      "learning_rate": 0.0003710418456315457,
+      "loss": 3.207,
+      "step": 25949
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6190086603164673,
+      "learning_rate": 0.0003710268872626801,
+      "loss": 2.9019,
+      "step": 25950
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5316742062568665,
+      "learning_rate": 0.0003710119287067433,
+      "loss": 2.996,
+      "step": 25951
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5887004137039185,
+      "learning_rate": 0.0003709969699637749,
+      "loss": 2.9726,
+      "step": 25952
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.560197651386261,
+      "learning_rate": 0.00037098201103381395,
+      "loss": 3.1608,
+      "step": 25953
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5498967170715332,
+      "learning_rate": 0.0003709670519169,
+      "loss": 3.1144,
+      "step": 25954
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.571072518825531,
+      "learning_rate": 0.0003709520926130727,
+      "loss": 3.1361,
+      "step": 25955
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5782363414764404,
+      "learning_rate": 0.0003709371331223712,
+      "loss": 2.9036,
+      "step": 25956
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5756257772445679,
+      "learning_rate": 0.00037092217344483487,
+      "loss": 2.8669,
+      "step": 25957
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5726673603057861,
+      "learning_rate": 0.00037090721358050324,
+      "loss": 2.84,
+      "step": 25958
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5259421467781067,
+      "learning_rate": 0.00037089225352941556,
+      "loss": 3.0836,
+      "step": 25959
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5834009051322937,
+      "learning_rate": 0.0003708772932916114,
+      "loss": 2.9831,
+      "step": 25960
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5248730778694153,
+      "learning_rate": 0.0003708623328671301,
+      "loss": 3.1426,
+      "step": 25961
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5834124088287354,
+      "learning_rate": 0.000370847372256011,
+      "loss": 3.0896,
+      "step": 25962
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5581058859825134,
+      "learning_rate": 0.0003708324114582937,
+      "loss": 2.8448,
+      "step": 25963
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6077955961227417,
+      "learning_rate": 0.0003708174504740173,
+      "loss": 2.7767,
+      "step": 25964
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5812661647796631,
+      "learning_rate": 0.00037080248930322136,
+      "loss": 3.1861,
+      "step": 25965
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5405663847923279,
+      "learning_rate": 0.00037078752794594545,
+      "loss": 3.198,
+      "step": 25966
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5473437905311584,
+      "learning_rate": 0.0003707725664022287,
+      "loss": 2.858,
+      "step": 25967
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.539084255695343,
+      "learning_rate": 0.0003707576046721106,
+      "loss": 3.262,
+      "step": 25968
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5603315830230713,
+      "learning_rate": 0.00037074264275563064,
+      "loss": 3.0577,
+      "step": 25969
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5609519481658936,
+      "learning_rate": 0.0003707276806528282,
+      "loss": 3.1609,
+      "step": 25970
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5762900710105896,
+      "learning_rate": 0.0003707127183637426,
+      "loss": 3.103,
+      "step": 25971
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5924959778785706,
+      "learning_rate": 0.0003706977558884133,
+      "loss": 2.9045,
+      "step": 25972
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5308851003646851,
+      "learning_rate": 0.0003706827932268798,
+      "loss": 2.995,
+      "step": 25973
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.605690062046051,
+      "learning_rate": 0.0003706678303791813,
+      "loss": 3.0956,
+      "step": 25974
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5422868728637695,
+      "learning_rate": 0.0003706528673453574,
+      "loss": 2.8659,
+      "step": 25975
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5357705950737,
+      "learning_rate": 0.00037063790412544747,
+      "loss": 2.9898,
+      "step": 25976
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5521494150161743,
+      "learning_rate": 0.00037062294071949094,
+      "loss": 2.9851,
+      "step": 25977
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5768835544586182,
+      "learning_rate": 0.0003706079771275271,
+      "loss": 3.1727,
+      "step": 25978
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5737690329551697,
+      "learning_rate": 0.00037059301334959536,
+      "loss": 2.8939,
+      "step": 25979
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5862882733345032,
+      "learning_rate": 0.0003705780493857354,
+      "loss": 2.9907,
+      "step": 25980
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5231944918632507,
+      "learning_rate": 0.00037056308523598637,
+      "loss": 2.9625,
+      "step": 25981
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5753133893013,
+      "learning_rate": 0.0003705481209003877,
+      "loss": 2.9762,
+      "step": 25982
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5374119281768799,
+      "learning_rate": 0.00037053315637897887,
+      "loss": 3.159,
+      "step": 25983
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5538880228996277,
+      "learning_rate": 0.0003705181916717993,
+      "loss": 2.9437,
+      "step": 25984
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5667445063591003,
+      "learning_rate": 0.00037050322677888837,
+      "loss": 3.2595,
+      "step": 25985
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5335412621498108,
+      "learning_rate": 0.0003704882617002855,
+      "loss": 3.1344,
+      "step": 25986
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5830770134925842,
+      "learning_rate": 0.00037047329643603014,
+      "loss": 3.1744,
+      "step": 25987
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.549442708492279,
+      "learning_rate": 0.00037045833098616176,
+      "loss": 2.9406,
+      "step": 25988
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5874332785606384,
+      "learning_rate": 0.00037044336535071954,
+      "loss": 3.1014,
+      "step": 25989
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5202533006668091,
+      "learning_rate": 0.00037042839952974307,
+      "loss": 3.0192,
+      "step": 25990
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5560673475265503,
+      "learning_rate": 0.00037041343352327185,
+      "loss": 3.1054,
+      "step": 25991
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5724155306816101,
+      "learning_rate": 0.00037039846733134514,
+      "loss": 3.2477,
+      "step": 25992
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.528751790523529,
+      "learning_rate": 0.00037038350095400246,
+      "loss": 3.2029,
+      "step": 25993
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.55881667137146,
+      "learning_rate": 0.0003703685343912831,
+      "loss": 3.1877,
+      "step": 25994
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5411615967750549,
+      "learning_rate": 0.0003703535676432266,
+      "loss": 3.1792,
+      "step": 25995
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5560112595558167,
+      "learning_rate": 0.00037033860070987237,
+      "loss": 3.1054,
+      "step": 25996
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5627215504646301,
+      "learning_rate": 0.00037032363359125974,
+      "loss": 3.0732,
+      "step": 25997
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.600377082824707,
+      "learning_rate": 0.0003703086662874283,
+      "loss": 3.0793,
+      "step": 25998
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5853224992752075,
+      "learning_rate": 0.00037029369879841735,
+      "loss": 2.9752,
+      "step": 25999
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5395614504814148,
+      "learning_rate": 0.0003702787311242662,
+      "loss": 3.0602,
+      "step": 26000
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.555023729801178,
+      "learning_rate": 0.0003702637632650145,
+      "loss": 3.2359,
+      "step": 26001
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5492096543312073,
+      "learning_rate": 0.00037024879522070157,
+      "loss": 2.8831,
+      "step": 26002
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5381262302398682,
+      "learning_rate": 0.00037023382699136686,
+      "loss": 3.0899,
+      "step": 26003
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5479117631912231,
+      "learning_rate": 0.0003702188585770497,
+      "loss": 3.2155,
+      "step": 26004
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5779570937156677,
+      "learning_rate": 0.00037020388997778954,
+      "loss": 3.0202,
+      "step": 26005
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5237216949462891,
+      "learning_rate": 0.0003701889211936259,
+      "loss": 3.0036,
+      "step": 26006
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5618526339530945,
+      "learning_rate": 0.00037017395222459817,
+      "loss": 3.1664,
+      "step": 26007
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5620414614677429,
+      "learning_rate": 0.0003701589830707457,
+      "loss": 3.166,
+      "step": 26008
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5533866286277771,
+      "learning_rate": 0.00037014401373210805,
+      "loss": 3.2417,
+      "step": 26009
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6008995771408081,
+      "learning_rate": 0.00037012904420872454,
+      "loss": 2.9938,
+      "step": 26010
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5710824728012085,
+      "learning_rate": 0.0003701140745006345,
+      "loss": 3.0668,
+      "step": 26011
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5625746846199036,
+      "learning_rate": 0.0003700991046078777,
+      "loss": 3.1091,
+      "step": 26012
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5556397438049316,
+      "learning_rate": 0.00037008413453049315,
+      "loss": 3.1491,
+      "step": 26013
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5745502710342407,
+      "learning_rate": 0.0003700691642685207,
+      "loss": 3.0732,
+      "step": 26014
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5311285257339478,
+      "learning_rate": 0.00037005419382199933,
+      "loss": 2.9755,
+      "step": 26015
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.573462963104248,
+      "learning_rate": 0.00037003922319096876,
+      "loss": 3.2097,
+      "step": 26016
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6004371047019958,
+      "learning_rate": 0.00037002425237546844,
+      "loss": 2.9479,
+      "step": 26017
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5266828536987305,
+      "learning_rate": 0.00037000928137553764,
+      "loss": 3.1431,
+      "step": 26018
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5740906000137329,
+      "learning_rate": 0.00036999431019121593,
+      "loss": 2.9836,
+      "step": 26019
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5576544404029846,
+      "learning_rate": 0.0003699793388225426,
+      "loss": 2.9814,
+      "step": 26020
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.568863570690155,
+      "learning_rate": 0.0003699643672695572,
+      "loss": 3.1698,
+      "step": 26021
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5414212942123413,
+      "learning_rate": 0.0003699493955322991,
+      "loss": 2.9529,
+      "step": 26022
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5581563711166382,
+      "learning_rate": 0.00036993442361080777,
+      "loss": 3.2253,
+      "step": 26023
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.578456461429596,
+      "learning_rate": 0.0003699194515051226,
+      "loss": 3.0586,
+      "step": 26024
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5558168888092041,
+      "learning_rate": 0.00036990447921528307,
+      "loss": 2.9423,
+      "step": 26025
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5423651337623596,
+      "learning_rate": 0.0003698895067413286,
+      "loss": 3.0314,
+      "step": 26026
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6390382647514343,
+      "learning_rate": 0.0003698745340832986,
+      "loss": 3.0495,
+      "step": 26027
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5630806684494019,
+      "learning_rate": 0.0003698595612412327,
+      "loss": 3.0398,
+      "step": 26028
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5756039023399353,
+      "learning_rate": 0.0003698445882151699,
+      "loss": 3.2904,
+      "step": 26029
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5660070180892944,
+      "learning_rate": 0.00036982961500515006,
+      "loss": 3.0634,
+      "step": 26030
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6267983913421631,
+      "learning_rate": 0.0003698146416112124,
+      "loss": 2.9549,
+      "step": 26031
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5327684283256531,
+      "learning_rate": 0.0003697996680333964,
+      "loss": 2.9134,
+      "step": 26032
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.543506383895874,
+      "learning_rate": 0.00036978469427174145,
+      "loss": 3.131,
+      "step": 26033
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5573419332504272,
+      "learning_rate": 0.00036976972032628716,
+      "loss": 3.2857,
+      "step": 26034
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5342174768447876,
+      "learning_rate": 0.0003697547461970728,
+      "loss": 3.2678,
+      "step": 26035
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5931220650672913,
+      "learning_rate": 0.0003697397718841379,
+      "loss": 3.0568,
+      "step": 26036
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5429893136024475,
+      "learning_rate": 0.0003697247973875218,
+      "loss": 3.1089,
+      "step": 26037
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5769212245941162,
+      "learning_rate": 0.000369709822707264,
+      "loss": 3.0458,
+      "step": 26038
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.549676239490509,
+      "learning_rate": 0.00036969484784340405,
+      "loss": 3.0389,
+      "step": 26039
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6053200364112854,
+      "learning_rate": 0.00036967987279598125,
+      "loss": 3.218,
+      "step": 26040
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6391343474388123,
+      "learning_rate": 0.00036966489756503495,
+      "loss": 3.1639,
+      "step": 26041
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.632569432258606,
+      "learning_rate": 0.00036964992215060484,
+      "loss": 3.2523,
+      "step": 26042
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5485760569572449,
+      "learning_rate": 0.00036963494655273016,
+      "loss": 3.1035,
+      "step": 26043
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6080588698387146,
+      "learning_rate": 0.0003696199707714505,
+      "loss": 3.2853,
+      "step": 26044
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5590080618858337,
+      "learning_rate": 0.0003696049948068052,
+      "loss": 2.9434,
+      "step": 26045
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5305129885673523,
+      "learning_rate": 0.0003695900186588338,
+      "loss": 3.0217,
+      "step": 26046
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5904659032821655,
+      "learning_rate": 0.00036957504232757555,
+      "loss": 3.0828,
+      "step": 26047
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5809376239776611,
+      "learning_rate": 0.00036956006581307017,
+      "loss": 3.1483,
+      "step": 26048
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5855849385261536,
+      "learning_rate": 0.0003695450891153569,
+      "loss": 3.1708,
+      "step": 26049
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.55824214220047,
+      "learning_rate": 0.00036953011223447526,
+      "loss": 3.0215,
+      "step": 26050
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5497469305992126,
+      "learning_rate": 0.0003695151351704646,
+      "loss": 3.1189,
+      "step": 26051
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.558712899684906,
+      "learning_rate": 0.0003695001579233646,
+      "loss": 3.0199,
+      "step": 26052
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5559713840484619,
+      "learning_rate": 0.0003694851804932145,
+      "loss": 3.1461,
+      "step": 26053
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5822705030441284,
+      "learning_rate": 0.0003694702028800538,
+      "loss": 3.0182,
+      "step": 26054
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.579784631729126,
+      "learning_rate": 0.00036945522508392197,
+      "loss": 3.0663,
+      "step": 26055
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5558045506477356,
+      "learning_rate": 0.00036944024710485834,
+      "loss": 3.012,
+      "step": 26056
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5622739195823669,
+      "learning_rate": 0.0003694252689429026,
+      "loss": 3.1491,
+      "step": 26057
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5217506289482117,
+      "learning_rate": 0.00036941029059809407,
+      "loss": 3.0809,
+      "step": 26058
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5513173341751099,
+      "learning_rate": 0.00036939531207047207,
+      "loss": 2.9512,
+      "step": 26059
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5600497722625732,
+      "learning_rate": 0.00036938033336007624,
+      "loss": 3.2733,
+      "step": 26060
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5448271036148071,
+      "learning_rate": 0.00036936535446694595,
+      "loss": 2.8906,
+      "step": 26061
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.564911961555481,
+      "learning_rate": 0.0003693503753911207,
+      "loss": 2.9206,
+      "step": 26062
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5715450644493103,
+      "learning_rate": 0.0003693353961326399,
+      "loss": 2.9661,
+      "step": 26063
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5593858361244202,
+      "learning_rate": 0.000369320416691543,
+      "loss": 3.2893,
+      "step": 26064
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5523359179496765,
+      "learning_rate": 0.00036930543706786953,
+      "loss": 3.0601,
+      "step": 26065
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5331974029541016,
+      "learning_rate": 0.0003692904572616588,
+      "loss": 3.0984,
+      "step": 26066
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5936278700828552,
+      "learning_rate": 0.00036927547727295024,
+      "loss": 2.9378,
+      "step": 26067
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5834710001945496,
+      "learning_rate": 0.0003692604971017836,
+      "loss": 3.0566,
+      "step": 26068
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6169381737709045,
+      "learning_rate": 0.0003692455167481981,
+      "loss": 3.1942,
+      "step": 26069
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5709614753723145,
+      "learning_rate": 0.00036923053621223316,
+      "loss": 3.2825,
+      "step": 26070
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5453691482543945,
+      "learning_rate": 0.0003692155554939283,
+      "loss": 3.0132,
+      "step": 26071
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5543729662895203,
+      "learning_rate": 0.00036920057459332304,
+      "loss": 3.0603,
+      "step": 26072
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.572906494140625,
+      "learning_rate": 0.0003691855935104568,
+      "loss": 3.2762,
+      "step": 26073
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5718940496444702,
+      "learning_rate": 0.0003691706122453689,
+      "loss": 3.1996,
+      "step": 26074
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.534610390663147,
+      "learning_rate": 0.00036915563079809906,
+      "loss": 3.0727,
+      "step": 26075
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6224848628044128,
+      "learning_rate": 0.00036914064916868664,
+      "loss": 2.8755,
+      "step": 26076
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5702244639396667,
+      "learning_rate": 0.0003691256673571709,
+      "loss": 3.2086,
+      "step": 26077
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.567895770072937,
+      "learning_rate": 0.0003691106853635915,
+      "loss": 3.1703,
+      "step": 26078
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5513347387313843,
+      "learning_rate": 0.00036909570318798793,
+      "loss": 2.9534,
+      "step": 26079
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5436404943466187,
+      "learning_rate": 0.00036908072083039953,
+      "loss": 2.9507,
+      "step": 26080
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5403690338134766,
+      "learning_rate": 0.00036906573829086585,
+      "loss": 3.1926,
+      "step": 26081
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5600476264953613,
+      "learning_rate": 0.00036905075556942625,
+      "loss": 3.1376,
+      "step": 26082
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6057079434394836,
+      "learning_rate": 0.0003690357726661202,
+      "loss": 3.2067,
+      "step": 26083
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.575212299823761,
+      "learning_rate": 0.0003690207895809873,
+      "loss": 3.1779,
+      "step": 26084
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5557694435119629,
+      "learning_rate": 0.00036900580631406683,
+      "loss": 2.9421,
+      "step": 26085
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5446481108665466,
+      "learning_rate": 0.00036899082286539846,
+      "loss": 3.1631,
+      "step": 26086
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5733548998832703,
+      "learning_rate": 0.0003689758392350215,
+      "loss": 3.1699,
+      "step": 26087
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5587981343269348,
+      "learning_rate": 0.00036896085542297545,
+      "loss": 3.245,
+      "step": 26088
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5554731488227844,
+      "learning_rate": 0.00036894587142929973,
+      "loss": 3.1064,
+      "step": 26089
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.583678662776947,
+      "learning_rate": 0.0003689308872540339,
+      "loss": 3.456,
+      "step": 26090
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5885918140411377,
+      "learning_rate": 0.0003689159028972175,
+      "loss": 3.3753,
+      "step": 26091
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5499343872070312,
+      "learning_rate": 0.00036890091835888966,
+      "loss": 3.1146,
+      "step": 26092
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5878946781158447,
+      "learning_rate": 0.0003688859336390902,
+      "loss": 2.9873,
+      "step": 26093
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5701496601104736,
+      "learning_rate": 0.0003688709487378584,
+      "loss": 3.1183,
+      "step": 26094
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5646578669548035,
+      "learning_rate": 0.00036885596365523375,
+      "loss": 3.3045,
+      "step": 26095
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5535494089126587,
+      "learning_rate": 0.0003688409783912557,
+      "loss": 3.0198,
+      "step": 26096
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5509173274040222,
+      "learning_rate": 0.0003688259929459639,
+      "loss": 3.1636,
+      "step": 26097
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5773724317550659,
+      "learning_rate": 0.00036881100731939756,
+      "loss": 3.0848,
+      "step": 26098
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5771932601928711,
+      "learning_rate": 0.00036879602151159633,
+      "loss": 2.8706,
+      "step": 26099
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5937799215316772,
+      "learning_rate": 0.0003687810355225996,
+      "loss": 3.1804,
+      "step": 26100
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5623434782028198,
+      "learning_rate": 0.00036876604935244696,
+      "loss": 3.0081,
+      "step": 26101
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5697646737098694,
+      "learning_rate": 0.0003687510630011776,
+      "loss": 3.1045,
+      "step": 26102
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.562432050704956,
+      "learning_rate": 0.00036873607646883125,
+      "loss": 3.0006,
+      "step": 26103
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5757910013198853,
+      "learning_rate": 0.0003687210897554474,
+      "loss": 2.8835,
+      "step": 26104
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5502630472183228,
+      "learning_rate": 0.0003687061028610653,
+      "loss": 3.1017,
+      "step": 26105
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5945281386375427,
+      "learning_rate": 0.00036869111578572463,
+      "loss": 2.9427,
+      "step": 26106
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.562147855758667,
+      "learning_rate": 0.0003686761285294647,
+      "loss": 2.8959,
+      "step": 26107
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5568060874938965,
+      "learning_rate": 0.0003686611410923251,
+      "loss": 3.1515,
+      "step": 26108
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5851083993911743,
+      "learning_rate": 0.0003686461534743453,
+      "loss": 2.9902,
+      "step": 26109
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5441685318946838,
+      "learning_rate": 0.0003686311656755647,
+      "loss": 3.1402,
+      "step": 26110
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5639269351959229,
+      "learning_rate": 0.0003686161776960228,
+      "loss": 2.9541,
+      "step": 26111
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5413896441459656,
+      "learning_rate": 0.0003686011895357592,
+      "loss": 3.0014,
+      "step": 26112
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5627785921096802,
+      "learning_rate": 0.0003685862011948131,
+      "loss": 3.0785,
+      "step": 26113
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6175768375396729,
+      "learning_rate": 0.00036857121267322424,
+      "loss": 3.2395,
+      "step": 26114
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5553895235061646,
+      "learning_rate": 0.000368556223971032,
+      "loss": 3.1911,
+      "step": 26115
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5802327394485474,
+      "learning_rate": 0.00036854123508827595,
+      "loss": 3.1287,
+      "step": 26116
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.561752200126648,
+      "learning_rate": 0.00036852624602499534,
+      "loss": 3.2061,
+      "step": 26117
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6642087697982788,
+      "learning_rate": 0.0003685112567812298,
+      "loss": 3.0804,
+      "step": 26118
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5541870594024658,
+      "learning_rate": 0.0003684962673570189,
+      "loss": 3.0745,
+      "step": 26119
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5653018355369568,
+      "learning_rate": 0.0003684812777524019,
+      "loss": 3.1752,
+      "step": 26120
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5748471021652222,
+      "learning_rate": 0.0003684662879674185,
+      "loss": 3.1672,
+      "step": 26121
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6177536845207214,
+      "learning_rate": 0.00036845129800210796,
+      "loss": 3.0762,
+      "step": 26122
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5813118815422058,
+      "learning_rate": 0.00036843630785650996,
+      "loss": 3.0871,
+      "step": 26123
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5994959473609924,
+      "learning_rate": 0.0003684213175306639,
+      "loss": 2.9619,
+      "step": 26124
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6316370368003845,
+      "learning_rate": 0.0003684063270246092,
+      "loss": 3.1457,
+      "step": 26125
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5947257876396179,
+      "learning_rate": 0.0003683913363383854,
+      "loss": 3.196,
+      "step": 26126
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5939607620239258,
+      "learning_rate": 0.00036837634547203207,
+      "loss": 3.2313,
+      "step": 26127
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5786470770835876,
+      "learning_rate": 0.0003683613544255884,
+      "loss": 3.1698,
+      "step": 26128
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5446995496749878,
+      "learning_rate": 0.00036834636319909425,
+      "loss": 3.4354,
+      "step": 26129
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5662088394165039,
+      "learning_rate": 0.00036833137179258897,
+      "loss": 3.104,
+      "step": 26130
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5539671182632446,
+      "learning_rate": 0.0003683163802061119,
+      "loss": 3.2375,
+      "step": 26131
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.576744019985199,
+      "learning_rate": 0.00036830138843970274,
+      "loss": 3.222,
+      "step": 26132
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5656651258468628,
+      "learning_rate": 0.0003682863964934007,
+      "loss": 3.09,
+      "step": 26133
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5607897043228149,
+      "learning_rate": 0.0003682714043672456,
+      "loss": 3.0891,
+      "step": 26134
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5817605257034302,
+      "learning_rate": 0.00036825641206127667,
+      "loss": 3.1378,
+      "step": 26135
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5782478451728821,
+      "learning_rate": 0.0003682414195755335,
+      "loss": 3.0419,
+      "step": 26136
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.575994610786438,
+      "learning_rate": 0.0003682264269100556,
+      "loss": 3.3339,
+      "step": 26137
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6501098871231079,
+      "learning_rate": 0.00036821143406488236,
+      "loss": 3.114,
+      "step": 26138
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5928285121917725,
+      "learning_rate": 0.0003681964410400534,
+      "loss": 3.1105,
+      "step": 26139
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5721243619918823,
+      "learning_rate": 0.00036818144783560804,
+      "loss": 2.8842,
+      "step": 26140
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5766890645027161,
+      "learning_rate": 0.0003681664544515861,
+      "loss": 2.9824,
+      "step": 26141
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5364489555358887,
+      "learning_rate": 0.00036815146088802663,
+      "loss": 3.034,
+      "step": 26142
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5811851620674133,
+      "learning_rate": 0.0003681364671449693,
+      "loss": 3.1483,
+      "step": 26143
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6257297992706299,
+      "learning_rate": 0.0003681214732224537,
+      "loss": 3.1882,
+      "step": 26144
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5299636721611023,
+      "learning_rate": 0.00036810647912051924,
+      "loss": 3.008,
+      "step": 26145
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5441427826881409,
+      "learning_rate": 0.00036809148483920543,
+      "loss": 3.0218,
+      "step": 26146
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.573083758354187,
+      "learning_rate": 0.00036807649037855173,
+      "loss": 3.064,
+      "step": 26147
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5455029010772705,
+      "learning_rate": 0.0003680614957385977,
+      "loss": 3.1727,
+      "step": 26148
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5636841654777527,
+      "learning_rate": 0.00036804650091938276,
+      "loss": 3.0886,
+      "step": 26149
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5653840899467468,
+      "learning_rate": 0.0003680315059209464,
+      "loss": 3.1209,
+      "step": 26150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5494506359100342,
+      "learning_rate": 0.0003680165107433282,
+      "loss": 2.749,
+      "step": 26151
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6142790913581848,
+      "learning_rate": 0.00036800151538656764,
+      "loss": 3.0748,
+      "step": 26152
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5461570620536804,
+      "learning_rate": 0.00036798651985070415,
+      "loss": 3.0659,
+      "step": 26153
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5468064546585083,
+      "learning_rate": 0.00036797152413577713,
+      "loss": 2.7883,
+      "step": 26154
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5630825757980347,
+      "learning_rate": 0.0003679565282418263,
+      "loss": 3.0646,
+      "step": 26155
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5849050879478455,
+      "learning_rate": 0.0003679415321688911,
+      "loss": 3.1669,
+      "step": 26156
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5444782972335815,
+      "learning_rate": 0.0003679265359170109,
+      "loss": 3.1053,
+      "step": 26157
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5830182433128357,
+      "learning_rate": 0.00036791153948622534,
+      "loss": 2.9533,
+      "step": 26158
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5592882633209229,
+      "learning_rate": 0.00036789654287657376,
+      "loss": 3.0239,
+      "step": 26159
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5635514855384827,
+      "learning_rate": 0.00036788154608809583,
+      "loss": 3.0102,
+      "step": 26160
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5419289469718933,
+      "learning_rate": 0.00036786654912083095,
+      "loss": 3.061,
+      "step": 26161
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5628889799118042,
+      "learning_rate": 0.0003678515519748186,
+      "loss": 2.9982,
+      "step": 26162
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5445288419723511,
+      "learning_rate": 0.00036783655465009844,
+      "loss": 3.154,
+      "step": 26163
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5867927074432373,
+      "learning_rate": 0.00036782155714670974,
+      "loss": 2.873,
+      "step": 26164
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5903654098510742,
+      "learning_rate": 0.0003678065594646921,
+      "loss": 3.2046,
+      "step": 26165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5572836995124817,
+      "learning_rate": 0.00036779156160408514,
+      "loss": 3.1028,
+      "step": 26166
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5365197658538818,
+      "learning_rate": 0.0003677765635649282,
+      "loss": 2.9957,
+      "step": 26167
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.567305862903595,
+      "learning_rate": 0.0003677615653472608,
+      "loss": 3.0076,
+      "step": 26168
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5919384956359863,
+      "learning_rate": 0.00036774656695112246,
+      "loss": 2.7972,
+      "step": 26169
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5833246111869812,
+      "learning_rate": 0.0003677315683765528,
+      "loss": 2.9785,
+      "step": 26170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5576356053352356,
+      "learning_rate": 0.00036771656962359116,
+      "loss": 3.0201,
+      "step": 26171
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5311302542686462,
+      "learning_rate": 0.0003677015706922771,
+      "loss": 3.1483,
+      "step": 26172
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6967976689338684,
+      "learning_rate": 0.0003676865715826502,
+      "loss": 3.0607,
+      "step": 26173
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5725787281990051,
+      "learning_rate": 0.00036767157229474975,
+      "loss": 3.0772,
+      "step": 26174
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5482276082038879,
+      "learning_rate": 0.0003676565728286155,
+      "loss": 3.0737,
+      "step": 26175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5845776796340942,
+      "learning_rate": 0.0003676415731842868,
+      "loss": 3.0155,
+      "step": 26176
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.56215900182724,
+      "learning_rate": 0.0003676265733618033,
+      "loss": 3.1768,
+      "step": 26177
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.638214111328125,
+      "learning_rate": 0.00036761157336120444,
+      "loss": 3.0996,
+      "step": 26178
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5645318627357483,
+      "learning_rate": 0.0003675965731825296,
+      "loss": 3.1095,
+      "step": 26179
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5821544528007507,
+      "learning_rate": 0.0003675815728258184,
+      "loss": 3.1556,
+      "step": 26180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.564754843711853,
+      "learning_rate": 0.00036756657229111055,
+      "loss": 3.1484,
+      "step": 26181
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.559941291809082,
+      "learning_rate": 0.0003675515715784452,
+      "loss": 3.2373,
+      "step": 26182
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5572832226753235,
+      "learning_rate": 0.00036753657068786194,
+      "loss": 3.0946,
+      "step": 26183
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5751796364784241,
+      "learning_rate": 0.00036752156961940047,
+      "loss": 3.1057,
+      "step": 26184
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7486311793327332,
+      "learning_rate": 0.0003675065683731001,
+      "loss": 2.961,
+      "step": 26185
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.593322217464447,
+      "learning_rate": 0.0003674915669490004,
+      "loss": 3.2956,
+      "step": 26186
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5414199233055115,
+      "learning_rate": 0.00036747656534714094,
+      "loss": 2.9403,
+      "step": 26187
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.613577127456665,
+      "learning_rate": 0.00036746156356756124,
+      "loss": 3.0939,
+      "step": 26188
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6109552979469299,
+      "learning_rate": 0.00036744656161030075,
+      "loss": 3.0483,
+      "step": 26189
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6177679300308228,
+      "learning_rate": 0.00036743155947539894,
+      "loss": 3.0387,
+      "step": 26190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.557323157787323,
+      "learning_rate": 0.0003674165571628954,
+      "loss": 3.1379,
+      "step": 26191
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6148905754089355,
+      "learning_rate": 0.0003674015546728297,
+      "loss": 3.0182,
+      "step": 26192
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5857757329940796,
+      "learning_rate": 0.00036738655200524125,
+      "loss": 3.0732,
+      "step": 26193
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5747817754745483,
+      "learning_rate": 0.00036737154916016956,
+      "loss": 2.8798,
+      "step": 26194
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5815746188163757,
+      "learning_rate": 0.00036735654613765414,
+      "loss": 3.2067,
+      "step": 26195
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5818893313407898,
+      "learning_rate": 0.0003673415429377345,
+      "loss": 3.1208,
+      "step": 26196
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5754139423370361,
+      "learning_rate": 0.0003673265395604503,
+      "loss": 3.2264,
+      "step": 26197
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5612207055091858,
+      "learning_rate": 0.00036731153600584094,
+      "loss": 3.0593,
+      "step": 26198
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5592952966690063,
+      "learning_rate": 0.0003672965322739459,
+      "loss": 3.3381,
+      "step": 26199
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5539454817771912,
+      "learning_rate": 0.0003672815283648048,
+      "loss": 3.2051,
+      "step": 26200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5505897998809814,
+      "learning_rate": 0.00036726652427845706,
+      "loss": 3.0047,
+      "step": 26201
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6153740286827087,
+      "learning_rate": 0.00036725152001494226,
+      "loss": 3.0584,
+      "step": 26202
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5606221556663513,
+      "learning_rate": 0.0003672365155742999,
+      "loss": 3.0653,
+      "step": 26203
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5304719805717468,
+      "learning_rate": 0.0003672215109565696,
+      "loss": 3.1395,
+      "step": 26204
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5453681349754333,
+      "learning_rate": 0.00036720650616179056,
+      "loss": 3.0008,
+      "step": 26205
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5311553478240967,
+      "learning_rate": 0.00036719150119000264,
+      "loss": 3.022,
+      "step": 26206
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5572003126144409,
+      "learning_rate": 0.0003671764960412453,
+      "loss": 3.121,
+      "step": 26207
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6088648438453674,
+      "learning_rate": 0.0003671614907155579,
+      "loss": 3.1107,
+      "step": 26208
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6267878413200378,
+      "learning_rate": 0.0003671464852129801,
+      "loss": 3.201,
+      "step": 26209
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5654227137565613,
+      "learning_rate": 0.0003671314795335513,
+      "loss": 3.1772,
+      "step": 26210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5888288617134094,
+      "learning_rate": 0.0003671164736773112,
+      "loss": 2.9714,
+      "step": 26211
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.542107343673706,
+      "learning_rate": 0.00036710146764429915,
+      "loss": 3.125,
+      "step": 26212
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5171507596969604,
+      "learning_rate": 0.00036708646143455477,
+      "loss": 3.0741,
+      "step": 26213
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5658581852912903,
+      "learning_rate": 0.00036707145504811763,
+      "loss": 3.1359,
+      "step": 26214
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.583036482334137,
+      "learning_rate": 0.00036705644848502716,
+      "loss": 3.0781,
+      "step": 26215
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.553354024887085,
+      "learning_rate": 0.00036704144174532286,
+      "loss": 3.1118,
+      "step": 26216
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5677482485771179,
+      "learning_rate": 0.0003670264348290443,
+      "loss": 3.0024,
+      "step": 26217
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6651977896690369,
+      "learning_rate": 0.0003670114277362311,
+      "loss": 2.9855,
+      "step": 26218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6104764342308044,
+      "learning_rate": 0.00036699642046692264,
+      "loss": 2.775,
+      "step": 26219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5855001211166382,
+      "learning_rate": 0.0003669814130211584,
+      "loss": 2.8599,
+      "step": 26220
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5708563923835754,
+      "learning_rate": 0.0003669664053989782,
+      "loss": 3.1819,
+      "step": 26221
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5730376243591309,
+      "learning_rate": 0.00036695139760042125,
+      "loss": 3.2359,
+      "step": 26222
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5522173643112183,
+      "learning_rate": 0.0003669363896255272,
+      "loss": 2.9391,
+      "step": 26223
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6628093123435974,
+      "learning_rate": 0.00036692138147433566,
+      "loss": 2.9406,
+      "step": 26224
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5647956132888794,
+      "learning_rate": 0.00036690637314688607,
+      "loss": 3.0561,
+      "step": 26225
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.541031002998352,
+      "learning_rate": 0.0003668913646432179,
+      "loss": 3.0519,
+      "step": 26226
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5448402166366577,
+      "learning_rate": 0.00036687635596337084,
+      "loss": 3.3572,
+      "step": 26227
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5486478209495544,
+      "learning_rate": 0.00036686134710738424,
+      "loss": 2.9945,
+      "step": 26228
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5686172842979431,
+      "learning_rate": 0.0003668463380752978,
+      "loss": 2.9606,
+      "step": 26229
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.54524165391922,
+      "learning_rate": 0.00036683132886715093,
+      "loss": 2.9665,
+      "step": 26230
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5635949969291687,
+      "learning_rate": 0.0003668163194829831,
+      "loss": 2.8544,
+      "step": 26231
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6484821438789368,
+      "learning_rate": 0.00036680130992283413,
+      "loss": 3.0592,
+      "step": 26232
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5802894830703735,
+      "learning_rate": 0.00036678630018674327,
+      "loss": 3.1788,
+      "step": 26233
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5594797730445862,
+      "learning_rate": 0.0003667712902747501,
+      "loss": 3.0831,
+      "step": 26234
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5902174115180969,
+      "learning_rate": 0.0003667562801868943,
+      "loss": 3.0271,
+      "step": 26235
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5711631178855896,
+      "learning_rate": 0.00036674126992321525,
+      "loss": 2.6302,
+      "step": 26236
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.527863621711731,
+      "learning_rate": 0.0003667262594837525,
+      "loss": 3.146,
+      "step": 26237
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5474887490272522,
+      "learning_rate": 0.0003667112488685457,
+      "loss": 3.1675,
+      "step": 26238
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.570087730884552,
+      "learning_rate": 0.00036669623807763434,
+      "loss": 3.0728,
+      "step": 26239
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.582175612449646,
+      "learning_rate": 0.00036668122711105793,
+      "loss": 3.0308,
+      "step": 26240
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5625713467597961,
+      "learning_rate": 0.00036666621596885584,
+      "loss": 3.1616,
+      "step": 26241
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5727894306182861,
+      "learning_rate": 0.00036665120465106786,
+      "loss": 3.1067,
+      "step": 26242
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5584968328475952,
+      "learning_rate": 0.00036663619315773356,
+      "loss": 2.8448,
+      "step": 26243
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5649468898773193,
+      "learning_rate": 0.0003666211814888922,
+      "loss": 3.0906,
+      "step": 26244
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5201927423477173,
+      "learning_rate": 0.0003666061696445835,
+      "loss": 3.1189,
+      "step": 26245
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5626629590988159,
+      "learning_rate": 0.0003665911576248469,
+      "loss": 3.1336,
+      "step": 26246
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.580909013748169,
+      "learning_rate": 0.00036657614542972216,
+      "loss": 2.9194,
+      "step": 26247
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5470664501190186,
+      "learning_rate": 0.00036656113305924857,
+      "loss": 2.9906,
+      "step": 26248
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5362159013748169,
+      "learning_rate": 0.0003665461205134657,
+      "loss": 3.0446,
+      "step": 26249
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5313224792480469,
+      "learning_rate": 0.0003665311077924133,
+      "loss": 3.1602,
+      "step": 26250
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5694699287414551,
+      "learning_rate": 0.0003665160948961307,
+      "loss": 2.9436,
+      "step": 26251
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5497816205024719,
+      "learning_rate": 0.0003665010818246575,
+      "loss": 3.0351,
+      "step": 26252
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5596659779548645,
+      "learning_rate": 0.00036648606857803324,
+      "loss": 3.0668,
+      "step": 26253
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6068937182426453,
+      "learning_rate": 0.0003664710551562976,
+      "loss": 3.0934,
+      "step": 26254
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5531221032142639,
+      "learning_rate": 0.0003664560415594899,
+      "loss": 2.95,
+      "step": 26255
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5915762782096863,
+      "learning_rate": 0.00036644102778764963,
+      "loss": 2.8952,
+      "step": 26256
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5725966095924377,
+      "learning_rate": 0.0003664260138408167,
+      "loss": 3.1125,
+      "step": 26257
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5639798045158386,
+      "learning_rate": 0.00036641099971903033,
+      "loss": 3.226,
+      "step": 26258
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5750287771224976,
+      "learning_rate": 0.0003663959854223302,
+      "loss": 3.0015,
+      "step": 26259
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5510179400444031,
+      "learning_rate": 0.0003663809709507558,
+      "loss": 2.9119,
+      "step": 26260
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6045522093772888,
+      "learning_rate": 0.00036636595630434664,
+      "loss": 3.2426,
+      "step": 26261
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5941025018692017,
+      "learning_rate": 0.0003663509414831424,
+      "loss": 3.1572,
+      "step": 26262
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5708783268928528,
+      "learning_rate": 0.0003663359264871825,
+      "loss": 3.0711,
+      "step": 26263
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5933129191398621,
+      "learning_rate": 0.00036632091131650664,
+      "loss": 2.9989,
+      "step": 26264
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5684006214141846,
+      "learning_rate": 0.00036630589597115424,
+      "loss": 3.0895,
+      "step": 26265
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.555566132068634,
+      "learning_rate": 0.00036629088045116485,
+      "loss": 3.0108,
+      "step": 26266
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5590725541114807,
+      "learning_rate": 0.0003662758647565779,
+      "loss": 3.1664,
+      "step": 26267
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5853378176689148,
+      "learning_rate": 0.00036626084888743325,
+      "loss": 3.2277,
+      "step": 26268
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5699875354766846,
+      "learning_rate": 0.00036624583284377025,
+      "loss": 2.9595,
+      "step": 26269
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5524995923042297,
+      "learning_rate": 0.0003662308166256284,
+      "loss": 3.1834,
+      "step": 26270
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5816988945007324,
+      "learning_rate": 0.0003662158002330474,
+      "loss": 3.0219,
+      "step": 26271
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5661150813102722,
+      "learning_rate": 0.0003662007836660666,
+      "loss": 2.9393,
+      "step": 26272
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5578607320785522,
+      "learning_rate": 0.0003661857669247258,
+      "loss": 2.9993,
+      "step": 26273
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5460517406463623,
+      "learning_rate": 0.0003661707500090644,
+      "loss": 2.9927,
+      "step": 26274
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5846410393714905,
+      "learning_rate": 0.000366155732919122,
+      "loss": 3.0969,
+      "step": 26275
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5628123879432678,
+      "learning_rate": 0.0003661407156549382,
+      "loss": 3.1571,
+      "step": 26276
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5363184809684753,
+      "learning_rate": 0.0003661256982165523,
+      "loss": 2.9979,
+      "step": 26277
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5130451321601868,
+      "learning_rate": 0.0003661106806040041,
+      "loss": 3.1275,
+      "step": 26278
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5658857226371765,
+      "learning_rate": 0.0003660956628173331,
+      "loss": 3.0147,
+      "step": 26279
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.56007319688797,
+      "learning_rate": 0.00036608064485657896,
+      "loss": 3.1505,
+      "step": 26280
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6172868013381958,
+      "learning_rate": 0.00036606562672178097,
+      "loss": 3.1135,
+      "step": 26281
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5785728693008423,
+      "learning_rate": 0.0003660506084129789,
+      "loss": 3.201,
+      "step": 26282
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.60005784034729,
+      "learning_rate": 0.0003660355899302123,
+      "loss": 3.0202,
+      "step": 26283
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5991455912590027,
+      "learning_rate": 0.0003660205712735205,
+      "loss": 3.0963,
+      "step": 26284
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6216191649436951,
+      "learning_rate": 0.0003660055524429433,
+      "loss": 3.1045,
+      "step": 26285
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5496450662612915,
+      "learning_rate": 0.00036599053343852026,
+      "loss": 2.9522,
+      "step": 26286
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5841752886772156,
+      "learning_rate": 0.00036597551426029076,
+      "loss": 3.2331,
+      "step": 26287
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5961534380912781,
+      "learning_rate": 0.00036596049490829444,
+      "loss": 3.0853,
+      "step": 26288
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5837684273719788,
+      "learning_rate": 0.0003659454753825709,
+      "loss": 2.9664,
+      "step": 26289
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5820373892784119,
+      "learning_rate": 0.0003659304556831597,
+      "loss": 2.9441,
+      "step": 26290
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5966834425926208,
+      "learning_rate": 0.00036591543581010043,
+      "loss": 2.9539,
+      "step": 26291
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5738745331764221,
+      "learning_rate": 0.00036590041576343237,
+      "loss": 3.2159,
+      "step": 26292
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5458084344863892,
+      "learning_rate": 0.00036588539554319546,
+      "loss": 3.111,
+      "step": 26293
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5425732731819153,
+      "learning_rate": 0.00036587037514942917,
+      "loss": 3.3392,
+      "step": 26294
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6192220449447632,
+      "learning_rate": 0.0003658553545821728,
+      "loss": 3.0393,
+      "step": 26295
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5837422013282776,
+      "learning_rate": 0.0003658403338414662,
+      "loss": 3.2242,
+      "step": 26296
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5665894746780396,
+      "learning_rate": 0.00036582531292734887,
+      "loss": 3.3003,
+      "step": 26297
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5463784337043762,
+      "learning_rate": 0.00036581029183986027,
+      "loss": 3.0505,
+      "step": 26298
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5693387389183044,
+      "learning_rate": 0.00036579527057904005,
+      "loss": 3.14,
+      "step": 26299
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5675256848335266,
+      "learning_rate": 0.00036578024914492774,
+      "loss": 3.2636,
+      "step": 26300
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5448147654533386,
+      "learning_rate": 0.0003657652275375629,
+      "loss": 3.1507,
+      "step": 26301
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5830021500587463,
+      "learning_rate": 0.00036575020575698515,
+      "loss": 3.1082,
+      "step": 26302
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5744504332542419,
+      "learning_rate": 0.0003657351838032339,
+      "loss": 3.057,
+      "step": 26303
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5693845152854919,
+      "learning_rate": 0.00036572016167634895,
+      "loss": 2.9967,
+      "step": 26304
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5699017643928528,
+      "learning_rate": 0.00036570513937636974,
+      "loss": 3.131,
+      "step": 26305
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6183264851570129,
+      "learning_rate": 0.0003656901169033358,
+      "loss": 3.0087,
+      "step": 26306
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5810492634773254,
+      "learning_rate": 0.00036567509425728675,
+      "loss": 3.1105,
+      "step": 26307
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5565503239631653,
+      "learning_rate": 0.0003656600714382621,
+      "loss": 2.939,
+      "step": 26308
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5506647825241089,
+      "learning_rate": 0.00036564504844630155,
+      "loss": 3.0549,
+      "step": 26309
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.53928142786026,
+      "learning_rate": 0.00036563002528144445,
+      "loss": 3.0072,
+      "step": 26310
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5917065143585205,
+      "learning_rate": 0.00036561500194373057,
+      "loss": 3.0097,
+      "step": 26311
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.584202766418457,
+      "learning_rate": 0.0003655999784331994,
+      "loss": 3.2125,
+      "step": 26312
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5942773222923279,
+      "learning_rate": 0.00036558495474989044,
+      "loss": 2.9767,
+      "step": 26313
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5491356253623962,
+      "learning_rate": 0.0003655699308938434,
+      "loss": 3.1626,
+      "step": 26314
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5383910536766052,
+      "learning_rate": 0.0003655549068650978,
+      "loss": 3.2015,
+      "step": 26315
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5347186326980591,
+      "learning_rate": 0.00036553988266369323,
+      "loss": 3.1515,
+      "step": 26316
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5822970867156982,
+      "learning_rate": 0.0003655248582896692,
+      "loss": 3.1044,
+      "step": 26317
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5540775060653687,
+      "learning_rate": 0.0003655098337430651,
+      "loss": 3.1069,
+      "step": 26318
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5978838205337524,
+      "learning_rate": 0.00036549480902392094,
+      "loss": 2.8344,
+      "step": 26319
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.546452522277832,
+      "learning_rate": 0.00036547978413227597,
+      "loss": 2.8552,
+      "step": 26320
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5403779149055481,
+      "learning_rate": 0.0003654647590681699,
+      "loss": 3.1634,
+      "step": 26321
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5638641119003296,
+      "learning_rate": 0.0003654497338316422,
+      "loss": 3.0284,
+      "step": 26322
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5681201219558716,
+      "learning_rate": 0.0003654347084227325,
+      "loss": 3.1705,
+      "step": 26323
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5366926193237305,
+      "learning_rate": 0.0003654196828414804,
+      "loss": 3.0462,
+      "step": 26324
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5784863829612732,
+      "learning_rate": 0.0003654046570879254,
+      "loss": 3.0464,
+      "step": 26325
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5681374073028564,
+      "learning_rate": 0.0003653896311621072,
+      "loss": 2.7868,
+      "step": 26326
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5848317742347717,
+      "learning_rate": 0.00036537460506406533,
+      "loss": 3.2213,
+      "step": 26327
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5783429741859436,
+      "learning_rate": 0.00036535957879383913,
+      "loss": 2.9188,
+      "step": 26328
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5719509124755859,
+      "learning_rate": 0.00036534455235146847,
+      "loss": 3.0323,
+      "step": 26329
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5757635235786438,
+      "learning_rate": 0.00036532952573699296,
+      "loss": 3.2186,
+      "step": 26330
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.56982421875,
+      "learning_rate": 0.00036531449895045197,
+      "loss": 2.9804,
+      "step": 26331
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6023058295249939,
+      "learning_rate": 0.00036529947199188516,
+      "loss": 2.9169,
+      "step": 26332
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5599899888038635,
+      "learning_rate": 0.00036528444486133197,
+      "loss": 3.1561,
+      "step": 26333
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5620229840278625,
+      "learning_rate": 0.00036526941755883233,
+      "loss": 2.9379,
+      "step": 26334
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6028508543968201,
+      "learning_rate": 0.0003652543900844255,
+      "loss": 3.1183,
+      "step": 26335
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5630659461021423,
+      "learning_rate": 0.0003652393624381512,
+      "loss": 3.0555,
+      "step": 26336
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5554866790771484,
+      "learning_rate": 0.00036522433462004894,
+      "loss": 3.0993,
+      "step": 26337
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.595605194568634,
+      "learning_rate": 0.0003652093066301583,
+      "loss": 2.8999,
+      "step": 26338
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5355490446090698,
+      "learning_rate": 0.00036519427846851897,
+      "loss": 2.9579,
+      "step": 26339
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5914220213890076,
+      "learning_rate": 0.0003651792501351704,
+      "loss": 2.9879,
+      "step": 26340
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5520111322402954,
+      "learning_rate": 0.0003651642216301523,
+      "loss": 2.8358,
+      "step": 26341
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5735923051834106,
+      "learning_rate": 0.00036514919295350414,
+      "loss": 3.1905,
+      "step": 26342
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.583451509475708,
+      "learning_rate": 0.0003651341641052656,
+      "loss": 3.2077,
+      "step": 26343
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5587623715400696,
+      "learning_rate": 0.0003651191350854761,
+      "loss": 3.1237,
+      "step": 26344
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5999316573143005,
+      "learning_rate": 0.0003651041058941755,
+      "loss": 2.9136,
+      "step": 26345
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5469268560409546,
+      "learning_rate": 0.000365089076531403,
+      "loss": 3.1107,
+      "step": 26346
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5756465792655945,
+      "learning_rate": 0.0003650740469971985,
+      "loss": 3.0659,
+      "step": 26347
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5584861040115356,
+      "learning_rate": 0.00036505901729160146,
+      "loss": 3.095,
+      "step": 26348
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5867658853530884,
+      "learning_rate": 0.0003650439874146515,
+      "loss": 2.9141,
+      "step": 26349
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5676002502441406,
+      "learning_rate": 0.0003650289573663882,
+      "loss": 3.2514,
+      "step": 26350
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5888804793357849,
+      "learning_rate": 0.0003650139271468512,
+      "loss": 3.1704,
+      "step": 26351
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5709172487258911,
+      "learning_rate": 0.00036499889675608,
+      "loss": 3.2181,
+      "step": 26352
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.560498058795929,
+      "learning_rate": 0.0003649838661941142,
+      "loss": 3.1361,
+      "step": 26353
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5983741283416748,
+      "learning_rate": 0.0003649688354609934,
+      "loss": 2.8439,
+      "step": 26354
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5880367159843445,
+      "learning_rate": 0.0003649538045567571,
+      "loss": 2.862,
+      "step": 26355
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5435540080070496,
+      "learning_rate": 0.00036493877348144516,
+      "loss": 3.0035,
+      "step": 26356
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5692511796951294,
+      "learning_rate": 0.00036492374223509686,
+      "loss": 3.3313,
+      "step": 26357
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5610879063606262,
+      "learning_rate": 0.000364908710817752,
+      "loss": 3.1846,
+      "step": 26358
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5607772469520569,
+      "learning_rate": 0.00036489367922945,
+      "loss": 3.1864,
+      "step": 26359
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5416234731674194,
+      "learning_rate": 0.00036487864747023055,
+      "loss": 3.1087,
+      "step": 26360
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5746535062789917,
+      "learning_rate": 0.0003648636155401333,
+      "loss": 2.9298,
+      "step": 26361
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5644704699516296,
+      "learning_rate": 0.0003648485834391977,
+      "loss": 3.1022,
+      "step": 26362
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5640776753425598,
+      "learning_rate": 0.00036483355116746346,
+      "loss": 3.1836,
+      "step": 26363
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6316478848457336,
+      "learning_rate": 0.00036481851872497003,
+      "loss": 2.929,
+      "step": 26364
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5855584740638733,
+      "learning_rate": 0.0003648034861117572,
+      "loss": 2.9929,
+      "step": 26365
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5345319509506226,
+      "learning_rate": 0.00036478845332786446,
+      "loss": 3.098,
+      "step": 26366
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6064936518669128,
+      "learning_rate": 0.0003647734203733314,
+      "loss": 3.1827,
+      "step": 26367
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5909245610237122,
+      "learning_rate": 0.0003647583872481976,
+      "loss": 2.9642,
+      "step": 26368
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5904161334037781,
+      "learning_rate": 0.00036474335395250256,
+      "loss": 2.8956,
+      "step": 26369
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6184599995613098,
+      "learning_rate": 0.00036472832048628615,
+      "loss": 3.0058,
+      "step": 26370
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5821686387062073,
+      "learning_rate": 0.0003647132868495877,
+      "loss": 3.1638,
+      "step": 26371
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5509762167930603,
+      "learning_rate": 0.00036469825304244696,
+      "loss": 3.0594,
+      "step": 26372
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5704058408737183,
+      "learning_rate": 0.00036468321906490346,
+      "loss": 3.077,
+      "step": 26373
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5785650014877319,
+      "learning_rate": 0.0003646681849169968,
+      "loss": 3.288,
+      "step": 26374
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5538581609725952,
+      "learning_rate": 0.0003646531505987666,
+      "loss": 3.2065,
+      "step": 26375
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5718187093734741,
+      "learning_rate": 0.00036463811611025246,
+      "loss": 3.0738,
+      "step": 26376
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5524046421051025,
+      "learning_rate": 0.0003646230814514939,
+      "loss": 2.9332,
+      "step": 26377
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.528167188167572,
+      "learning_rate": 0.0003646080466225307,
+      "loss": 3.0779,
+      "step": 26378
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5965219736099243,
+      "learning_rate": 0.00036459301162340224,
+      "loss": 3.0625,
+      "step": 26379
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5643555521965027,
+      "learning_rate": 0.00036457797645414816,
+      "loss": 3.1285,
+      "step": 26380
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5772526264190674,
+      "learning_rate": 0.00036456294111480823,
+      "loss": 2.9713,
+      "step": 26381
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5830170512199402,
+      "learning_rate": 0.0003645479056054219,
+      "loss": 3.3178,
+      "step": 26382
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5554019808769226,
+      "learning_rate": 0.0003645328699260288,
+      "loss": 3.0084,
+      "step": 26383
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6931495070457458,
+      "learning_rate": 0.0003645178340766686,
+      "loss": 3.0225,
+      "step": 26384
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5578175187110901,
+      "learning_rate": 0.0003645027980573807,
+      "loss": 2.7523,
+      "step": 26385
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5534636378288269,
+      "learning_rate": 0.00036448776186820495,
+      "loss": 3.299,
+      "step": 26386
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5296128392219543,
+      "learning_rate": 0.00036447272550918085,
+      "loss": 3.2003,
+      "step": 26387
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.550498902797699,
+      "learning_rate": 0.00036445768898034803,
+      "loss": 2.945,
+      "step": 26388
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5512216687202454,
+      "learning_rate": 0.000364442652281746,
+      "loss": 3.1243,
+      "step": 26389
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.552977979183197,
+      "learning_rate": 0.0003644276154134144,
+      "loss": 3.0263,
+      "step": 26390
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5711373090744019,
+      "learning_rate": 0.00036441257837539293,
+      "loss": 3.0337,
+      "step": 26391
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6151379942893982,
+      "learning_rate": 0.00036439754116772107,
+      "loss": 3.0109,
+      "step": 26392
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5529595017433167,
+      "learning_rate": 0.0003643825037904385,
+      "loss": 3.1506,
+      "step": 26393
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5650982856750488,
+      "learning_rate": 0.00036436746624358485,
+      "loss": 3.193,
+      "step": 26394
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5503296852111816,
+      "learning_rate": 0.0003643524285271996,
+      "loss": 3.1614,
+      "step": 26395
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5685902237892151,
+      "learning_rate": 0.00036433739064132257,
+      "loss": 3.0464,
+      "step": 26396
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5520438551902771,
+      "learning_rate": 0.0003643223525859931,
+      "loss": 3.0988,
+      "step": 26397
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5292802453041077,
+      "learning_rate": 0.00036430731436125097,
+      "loss": 3.1851,
+      "step": 26398
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5346623659133911,
+      "learning_rate": 0.00036429227596713575,
+      "loss": 3.0767,
+      "step": 26399
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5210029482841492,
+      "learning_rate": 0.000364277237403687,
+      "loss": 3.0894,
+      "step": 26400
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5822991728782654,
+      "learning_rate": 0.00036426219867094446,
+      "loss": 3.1194,
+      "step": 26401
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5511487126350403,
+      "learning_rate": 0.0003642471597689477,
+      "loss": 2.9108,
+      "step": 26402
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.575802743434906,
+      "learning_rate": 0.0003642321206977362,
+      "loss": 3.0977,
+      "step": 26403
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5361688137054443,
+      "learning_rate": 0.00036421708145734967,
+      "loss": 3.2185,
+      "step": 26404
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5705873966217041,
+      "learning_rate": 0.00036420204204782767,
+      "loss": 3.0483,
+      "step": 26405
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5684397220611572,
+      "learning_rate": 0.00036418700246920987,
+      "loss": 2.9966,
+      "step": 26406
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6203626394271851,
+      "learning_rate": 0.00036417196272153595,
+      "loss": 3.0679,
+      "step": 26407
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.55103999376297,
+      "learning_rate": 0.0003641569228048453,
+      "loss": 3.128,
+      "step": 26408
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5689854621887207,
+      "learning_rate": 0.0003641418827191777,
+      "loss": 2.9817,
+      "step": 26409
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5948728322982788,
+      "learning_rate": 0.00036412684246457276,
+      "loss": 3.2106,
+      "step": 26410
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5416500568389893,
+      "learning_rate": 0.0003641118020410701,
+      "loss": 2.9918,
+      "step": 26411
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5819432139396667,
+      "learning_rate": 0.00036409676144870916,
+      "loss": 2.9693,
+      "step": 26412
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5323217511177063,
+      "learning_rate": 0.00036408172068752977,
+      "loss": 3.0868,
+      "step": 26413
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5703701972961426,
+      "learning_rate": 0.00036406667975757143,
+      "loss": 3.0828,
+      "step": 26414
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5979413390159607,
+      "learning_rate": 0.00036405163865887383,
+      "loss": 2.9349,
+      "step": 26415
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6079196929931641,
+      "learning_rate": 0.0003640365973914765,
+      "loss": 3.2151,
+      "step": 26416
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0003640215559554191,
+      "loss": 3.2673,
+      "step": 26417
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.576112687587738,
+      "learning_rate": 0.00036400651435074136,
+      "loss": 3.2403,
+      "step": 26418
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6245038509368896,
+      "learning_rate": 0.0003639914725774826,
+      "loss": 3.04,
+      "step": 26419
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5649415254592896,
+      "learning_rate": 0.0003639764306356827,
+      "loss": 2.7593,
+      "step": 26420
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6045398116111755,
+      "learning_rate": 0.0003639613885253812,
+      "loss": 2.9559,
+      "step": 26421
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5652473568916321,
+      "learning_rate": 0.00036394634624661767,
+      "loss": 2.9299,
+      "step": 26422
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5572347640991211,
+      "learning_rate": 0.0003639313037994318,
+      "loss": 3.1491,
+      "step": 26423
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7249118685722351,
+      "learning_rate": 0.00036391626118386316,
+      "loss": 3.1411,
+      "step": 26424
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.581577718257904,
+      "learning_rate": 0.00036390121839995147,
+      "loss": 3.0565,
+      "step": 26425
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5638243556022644,
+      "learning_rate": 0.00036388617544773615,
+      "loss": 3.0026,
+      "step": 26426
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.585462749004364,
+      "learning_rate": 0.000363871132327257,
+      "loss": 3.1258,
+      "step": 26427
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6039126515388489,
+      "learning_rate": 0.00036385608903855357,
+      "loss": 3.2572,
+      "step": 26428
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5659081339836121,
+      "learning_rate": 0.0003638410455816655,
+      "loss": 3.0501,
+      "step": 26429
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.610337495803833,
+      "learning_rate": 0.00036382600195663243,
+      "loss": 2.9333,
+      "step": 26430
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5635059475898743,
+      "learning_rate": 0.00036381095816349386,
+      "loss": 3.2617,
+      "step": 26431
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5821727514266968,
+      "learning_rate": 0.00036379591420228964,
+      "loss": 3.0316,
+      "step": 26432
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6468302607536316,
+      "learning_rate": 0.0003637808700730592,
+      "loss": 3.1508,
+      "step": 26433
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6181358098983765,
+      "learning_rate": 0.0003637658257758421,
+      "loss": 3.0418,
+      "step": 26434
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5608850717544556,
+      "learning_rate": 0.0003637507813106783,
+      "loss": 3.0204,
+      "step": 26435
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.584475576877594,
+      "learning_rate": 0.0003637357366776071,
+      "loss": 3.158,
+      "step": 26436
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5593780875205994,
+      "learning_rate": 0.00036372069187666826,
+      "loss": 3.0006,
+      "step": 26437
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5664603114128113,
+      "learning_rate": 0.0003637056469079013,
+      "loss": 3.2513,
+      "step": 26438
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5719146132469177,
+      "learning_rate": 0.00036369060177134603,
+      "loss": 3.076,
+      "step": 26439
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5447546243667603,
+      "learning_rate": 0.000363675556467042,
+      "loss": 2.7886,
+      "step": 26440
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.607576310634613,
+      "learning_rate": 0.00036366051099502865,
+      "loss": 3.0216,
+      "step": 26441
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5661342740058899,
+      "learning_rate": 0.00036364546535534584,
+      "loss": 3.0811,
+      "step": 26442
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5804358124732971,
+      "learning_rate": 0.0003636304195480332,
+      "loss": 3.0749,
+      "step": 26443
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5507537126541138,
+      "learning_rate": 0.00036361537357313024,
+      "loss": 3.028,
+      "step": 26444
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6021802425384521,
+      "learning_rate": 0.0003636003274306766,
+      "loss": 3.0707,
+      "step": 26445
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6064923405647278,
+      "learning_rate": 0.0003635852811207119,
+      "loss": 3.1915,
+      "step": 26446
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5652244687080383,
+      "learning_rate": 0.0003635702346432759,
+      "loss": 2.9867,
+      "step": 26447
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5297040939331055,
+      "learning_rate": 0.0003635551879984081,
+      "loss": 3.2154,
+      "step": 26448
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6315213441848755,
+      "learning_rate": 0.00036354014118614815,
+      "loss": 2.971,
+      "step": 26449
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5433605313301086,
+      "learning_rate": 0.0003635250942065357,
+      "loss": 3.0269,
+      "step": 26450
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5568627119064331,
+      "learning_rate": 0.0003635100470596104,
+      "loss": 3.047,
+      "step": 26451
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.563654363155365,
+      "learning_rate": 0.0003634949997454118,
+      "loss": 2.9285,
+      "step": 26452
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5413941740989685,
+      "learning_rate": 0.00036347995226397963,
+      "loss": 3.1223,
+      "step": 26453
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5784571766853333,
+      "learning_rate": 0.00036346490461535347,
+      "loss": 3.0467,
+      "step": 26454
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5411424040794373,
+      "learning_rate": 0.00036344985679957303,
+      "loss": 3.1296,
+      "step": 26455
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5692468285560608,
+      "learning_rate": 0.00036343480881667785,
+      "loss": 2.8744,
+      "step": 26456
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6368603706359863,
+      "learning_rate": 0.00036341976066670754,
+      "loss": 3.0441,
+      "step": 26457
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.546215832233429,
+      "learning_rate": 0.00036340471234970186,
+      "loss": 3.0811,
+      "step": 26458
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.582275390625,
+      "learning_rate": 0.0003633896638657003,
+      "loss": 3.0124,
+      "step": 26459
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5756465792655945,
+      "learning_rate": 0.00036337461521474256,
+      "loss": 2.997,
+      "step": 26460
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6010496020317078,
+      "learning_rate": 0.0003633595663968683,
+      "loss": 3.0752,
+      "step": 26461
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5619287490844727,
+      "learning_rate": 0.00036334451741211715,
+      "loss": 2.9749,
+      "step": 26462
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5364715456962585,
+      "learning_rate": 0.00036332946826052873,
+      "loss": 3.3402,
+      "step": 26463
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5843341946601868,
+      "learning_rate": 0.00036331441894214275,
+      "loss": 3.1867,
+      "step": 26464
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5982407331466675,
+      "learning_rate": 0.00036329936945699866,
+      "loss": 3.0285,
+      "step": 26465
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5931727290153503,
+      "learning_rate": 0.0003632843198051363,
+      "loss": 2.9798,
+      "step": 26466
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5525808930397034,
+      "learning_rate": 0.00036326926998659514,
+      "loss": 3.1638,
+      "step": 26467
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5755184888839722,
+      "learning_rate": 0.0003632542200014149,
+      "loss": 3.1532,
+      "step": 26468
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5768822431564331,
+      "learning_rate": 0.00036323916984963534,
+      "loss": 3.1317,
+      "step": 26469
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5717555284500122,
+      "learning_rate": 0.0003632241195312959,
+      "loss": 3.1596,
+      "step": 26470
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5730190873146057,
+      "learning_rate": 0.00036320906904643624,
+      "loss": 2.9991,
+      "step": 26471
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5641260147094727,
+      "learning_rate": 0.00036319401839509616,
+      "loss": 3.223,
+      "step": 26472
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5625524520874023,
+      "learning_rate": 0.0003631789675773151,
+      "loss": 3.0021,
+      "step": 26473
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5481633543968201,
+      "learning_rate": 0.00036316391659313283,
+      "loss": 3.018,
+      "step": 26474
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6047376990318298,
+      "learning_rate": 0.000363148865442589,
+      "loss": 2.9506,
+      "step": 26475
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5548542737960815,
+      "learning_rate": 0.00036313381412572313,
+      "loss": 3.0886,
+      "step": 26476
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5693079829216003,
+      "learning_rate": 0.000363118762642575,
+      "loss": 3.0023,
+      "step": 26477
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5523504614830017,
+      "learning_rate": 0.00036310371099318423,
+      "loss": 3.0461,
+      "step": 26478
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5627286434173584,
+      "learning_rate": 0.0003630886591775904,
+      "loss": 3.0872,
+      "step": 26479
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5450271368026733,
+      "learning_rate": 0.00036307360719583316,
+      "loss": 2.9879,
+      "step": 26480
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5695874094963074,
+      "learning_rate": 0.0003630585550479522,
+      "loss": 2.8817,
+      "step": 26481
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5733633041381836,
+      "learning_rate": 0.0003630435027339871,
+      "loss": 3.1564,
+      "step": 26482
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6155542731285095,
+      "learning_rate": 0.00036302845025397765,
+      "loss": 3.1808,
+      "step": 26483
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5391807556152344,
+      "learning_rate": 0.00036301339760796335,
+      "loss": 3.1328,
+      "step": 26484
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5551477670669556,
+      "learning_rate": 0.0003629983447959838,
+      "loss": 3.0805,
+      "step": 26485
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5700846910476685,
+      "learning_rate": 0.0003629832918180788,
+      "loss": 3.0817,
+      "step": 26486
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5457755327224731,
+      "learning_rate": 0.00036296823867428793,
+      "loss": 3.0501,
+      "step": 26487
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5421336889266968,
+      "learning_rate": 0.0003629531853646508,
+      "loss": 3.0883,
+      "step": 26488
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5504266023635864,
+      "learning_rate": 0.00036293813188920717,
+      "loss": 3.1587,
+      "step": 26489
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5366515517234802,
+      "learning_rate": 0.00036292307824799654,
+      "loss": 3.1722,
+      "step": 26490
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5298593044281006,
+      "learning_rate": 0.0003629080244410587,
+      "loss": 3.1069,
+      "step": 26491
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5564733147621155,
+      "learning_rate": 0.0003628929704684332,
+      "loss": 3.0798,
+      "step": 26492
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5588745474815369,
+      "learning_rate": 0.0003628779163301597,
+      "loss": 3.2701,
+      "step": 26493
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6034541130065918,
+      "learning_rate": 0.0003628628620262779,
+      "loss": 2.9638,
+      "step": 26494
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.54830402135849,
+      "learning_rate": 0.0003628478075568274,
+      "loss": 3.1212,
+      "step": 26495
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6392161250114441,
+      "learning_rate": 0.0003628327529218479,
+      "loss": 3.3524,
+      "step": 26496
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5647044777870178,
+      "learning_rate": 0.000362817698121379,
+      "loss": 3.0019,
+      "step": 26497
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5975145697593689,
+      "learning_rate": 0.0003628026431554603,
+      "loss": 3.0822,
+      "step": 26498
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5746291279792786,
+      "learning_rate": 0.00036278758802413166,
+      "loss": 3.2055,
+      "step": 26499
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6121372580528259,
+      "learning_rate": 0.0003627725327274325,
+      "loss": 3.0877,
+      "step": 26500
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.550165593624115,
+      "learning_rate": 0.0003627574772654026,
+      "loss": 3.1267,
+      "step": 26501
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6445377469062805,
+      "learning_rate": 0.00036274242163808164,
+      "loss": 3.2958,
+      "step": 26502
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5646840333938599,
+      "learning_rate": 0.00036272736584550913,
+      "loss": 2.9063,
+      "step": 26503
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6248049736022949,
+      "learning_rate": 0.0003627123098877248,
+      "loss": 3.142,
+      "step": 26504
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5855923295021057,
+      "learning_rate": 0.00036269725376476835,
+      "loss": 2.976,
+      "step": 26505
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5521533489227295,
+      "learning_rate": 0.0003626821974766795,
+      "loss": 3.2744,
+      "step": 26506
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6442424058914185,
+      "learning_rate": 0.00036266714102349773,
+      "loss": 3.0419,
+      "step": 26507
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5635976791381836,
+      "learning_rate": 0.00036265208440526266,
+      "loss": 3.0997,
+      "step": 26508
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.671798825263977,
+      "learning_rate": 0.0003626370276220142,
+      "loss": 3.037,
+      "step": 26509
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5717962384223938,
+      "learning_rate": 0.0003626219706737918,
+      "loss": 3.3547,
+      "step": 26510
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5673326253890991,
+      "learning_rate": 0.0003626069135606352,
+      "loss": 3.074,
+      "step": 26511
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5579215884208679,
+      "learning_rate": 0.00036259185628258406,
+      "loss": 3.1001,
+      "step": 26512
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6117144227027893,
+      "learning_rate": 0.000362576798839678,
+      "loss": 2.9077,
+      "step": 26513
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5998569130897522,
+      "learning_rate": 0.00036256174123195663,
+      "loss": 3.0331,
+      "step": 26514
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.618823230266571,
+      "learning_rate": 0.00036254668345945976,
+      "loss": 3.2402,
+      "step": 26515
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6104410290718079,
+      "learning_rate": 0.0003625316255222269,
+      "loss": 3.0506,
+      "step": 26516
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6050528883934021,
+      "learning_rate": 0.0003625165674202979,
+      "loss": 3.0046,
+      "step": 26517
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5506466627120972,
+      "learning_rate": 0.0003625015091537121,
+      "loss": 2.9101,
+      "step": 26518
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5498501658439636,
+      "learning_rate": 0.0003624864507225094,
+      "loss": 3.0459,
+      "step": 26519
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5689669251441956,
+      "learning_rate": 0.0003624713921267296,
+      "loss": 3.1916,
+      "step": 26520
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5426681637763977,
+      "learning_rate": 0.000362456333366412,
+      "loss": 2.9685,
+      "step": 26521
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.604785680770874,
+      "learning_rate": 0.0003624412744415965,
+      "loss": 3.0663,
+      "step": 26522
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5710480213165283,
+      "learning_rate": 0.0003624262153523226,
+      "loss": 2.9204,
+      "step": 26523
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6017956733703613,
+      "learning_rate": 0.0003624111560986302,
+      "loss": 3.011,
+      "step": 26524
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5829985737800598,
+      "learning_rate": 0.00036239609668055876,
+      "loss": 2.9643,
+      "step": 26525
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6299350261688232,
+      "learning_rate": 0.000362381037098148,
+      "loss": 2.9682,
+      "step": 26526
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.613353967666626,
+      "learning_rate": 0.0003623659773514376,
+      "loss": 3.0985,
+      "step": 26527
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5568817257881165,
+      "learning_rate": 0.00036235091744046724,
+      "loss": 2.8491,
+      "step": 26528
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5802782773971558,
+      "learning_rate": 0.0003623358573652765,
+      "loss": 3.1953,
+      "step": 26529
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5671431422233582,
+      "learning_rate": 0.0003623207971259052,
+      "loss": 2.9758,
+      "step": 26530
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5735278725624084,
+      "learning_rate": 0.0003623057367223929,
+      "loss": 3.119,
+      "step": 26531
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6418932676315308,
+      "learning_rate": 0.0003622906761547793,
+      "loss": 2.8465,
+      "step": 26532
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6021503210067749,
+      "learning_rate": 0.0003622756154231039,
+      "loss": 2.9617,
+      "step": 26533
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5770435929298401,
+      "learning_rate": 0.0003622605545274067,
+      "loss": 3.1281,
+      "step": 26534
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6095218062400818,
+      "learning_rate": 0.0003622454934677271,
+      "loss": 2.9292,
+      "step": 26535
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6107041239738464,
+      "learning_rate": 0.00036223043224410483,
+      "loss": 3.124,
+      "step": 26536
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5658743977546692,
+      "learning_rate": 0.0003622153708565796,
+      "loss": 3.0487,
+      "step": 26537
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5829092264175415,
+      "learning_rate": 0.000362200309305191,
+      "loss": 3.0127,
+      "step": 26538
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5712616443634033,
+      "learning_rate": 0.00036218524758997885,
+      "loss": 3.0086,
+      "step": 26539
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5344133377075195,
+      "learning_rate": 0.0003621701857109827,
+      "loss": 3.1907,
+      "step": 26540
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5803573727607727,
+      "learning_rate": 0.0003621551236682422,
+      "loss": 3.1347,
+      "step": 26541
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.566353976726532,
+      "learning_rate": 0.0003621400614617972,
+      "loss": 3.1542,
+      "step": 26542
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5749666094779968,
+      "learning_rate": 0.00036212499909168706,
+      "loss": 3.1458,
+      "step": 26543
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6262149214744568,
+      "learning_rate": 0.00036210993655795163,
+      "loss": 3.1817,
+      "step": 26544
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.629041314125061,
+      "learning_rate": 0.00036209487386063075,
+      "loss": 3.0455,
+      "step": 26545
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5451223850250244,
+      "learning_rate": 0.0003620798109997638,
+      "loss": 3.1538,
+      "step": 26546
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5455161929130554,
+      "learning_rate": 0.00036206474797539066,
+      "loss": 3.1579,
+      "step": 26547
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5440385341644287,
+      "learning_rate": 0.0003620496847875508,
+      "loss": 3.1674,
+      "step": 26548
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5767881274223328,
+      "learning_rate": 0.00036203462143628406,
+      "loss": 3.1339,
+      "step": 26549
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5350778698921204,
+      "learning_rate": 0.0003620195579216301,
+      "loss": 3.0594,
+      "step": 26550
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5580967664718628,
+      "learning_rate": 0.0003620044942436285,
+      "loss": 3.04,
+      "step": 26551
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5478173494338989,
+      "learning_rate": 0.00036198943040231904,
+      "loss": 3.1905,
+      "step": 26552
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6238259077072144,
+      "learning_rate": 0.00036197436639774145,
+      "loss": 3.0206,
+      "step": 26553
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6044278144836426,
+      "learning_rate": 0.00036195930222993506,
+      "loss": 3.0768,
+      "step": 26554
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5655761957168579,
+      "learning_rate": 0.00036194423789893997,
+      "loss": 3.1671,
+      "step": 26555
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5538684725761414,
+      "learning_rate": 0.00036192917340479566,
+      "loss": 3.0836,
+      "step": 26556
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5838094353675842,
+      "learning_rate": 0.0003619141087475418,
+      "loss": 3.0374,
+      "step": 26557
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5795071721076965,
+      "learning_rate": 0.00036189904392721814,
+      "loss": 2.9839,
+      "step": 26558
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5634998083114624,
+      "learning_rate": 0.0003618839789438642,
+      "loss": 3.0211,
+      "step": 26559
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5954675078392029,
+      "learning_rate": 0.0003618689137975199,
+      "loss": 3.102,
+      "step": 26560
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5686203241348267,
+      "learning_rate": 0.0003618538484882247,
+      "loss": 3.0318,
+      "step": 26561
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5589979887008667,
+      "learning_rate": 0.0003618387830160184,
+      "loss": 2.9281,
+      "step": 26562
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5522468686103821,
+      "learning_rate": 0.0003618237173809406,
+      "loss": 2.823,
+      "step": 26563
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5559152364730835,
+      "learning_rate": 0.00036180865158303116,
+      "loss": 3.1592,
+      "step": 26564
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5371105074882507,
+      "learning_rate": 0.00036179358562232945,
+      "loss": 3.0656,
+      "step": 26565
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5746232271194458,
+      "learning_rate": 0.00036177851949887544,
+      "loss": 3.042,
+      "step": 26566
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5857375264167786,
+      "learning_rate": 0.00036176345321270865,
+      "loss": 3.1788,
+      "step": 26567
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5602015852928162,
+      "learning_rate": 0.00036174838676386893,
+      "loss": 2.8983,
+      "step": 26568
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5846152305603027,
+      "learning_rate": 0.00036173332015239563,
+      "loss": 2.7467,
+      "step": 26569
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.541037917137146,
+      "learning_rate": 0.0003617182533783288,
+      "loss": 2.9718,
+      "step": 26570
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5495973825454712,
+      "learning_rate": 0.000361703186441708,
+      "loss": 3.0222,
+      "step": 26571
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5735442042350769,
+      "learning_rate": 0.00036168811934257275,
+      "loss": 3.2686,
+      "step": 26572
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5584540367126465,
+      "learning_rate": 0.00036167305208096293,
+      "loss": 2.8147,
+      "step": 26573
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5874664187431335,
+      "learning_rate": 0.00036165798465691816,
+      "loss": 3.1831,
+      "step": 26574
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5908390879631042,
+      "learning_rate": 0.00036164291707047817,
+      "loss": 2.9889,
+      "step": 26575
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5895627737045288,
+      "learning_rate": 0.00036162784932168254,
+      "loss": 2.9451,
+      "step": 26576
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5433921813964844,
+      "learning_rate": 0.000361612781410571,
+      "loss": 3.2065,
+      "step": 26577
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5549830794334412,
+      "learning_rate": 0.0003615977133371833,
+      "loss": 3.1892,
+      "step": 26578
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6039674282073975,
+      "learning_rate": 0.0003615826451015591,
+      "loss": 2.8504,
+      "step": 26579
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.526862382888794,
+      "learning_rate": 0.0003615675767037379,
+      "loss": 3.1544,
+      "step": 26580
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5542131662368774,
+      "learning_rate": 0.0003615525081437597,
+      "loss": 3.1678,
+      "step": 26581
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5980327129364014,
+      "learning_rate": 0.0003615374394216641,
+      "loss": 3.1879,
+      "step": 26582
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5924147367477417,
+      "learning_rate": 0.00036152237053749064,
+      "loss": 3.0792,
+      "step": 26583
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5593456029891968,
+      "learning_rate": 0.00036150730149127905,
+      "loss": 2.8554,
+      "step": 26584
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5941464304924011,
+      "learning_rate": 0.0003614922322830691,
+      "loss": 3.0402,
+      "step": 26585
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5819092988967896,
+      "learning_rate": 0.0003614771629129005,
+      "loss": 3.1216,
+      "step": 26586
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5705400705337524,
+      "learning_rate": 0.0003614620933808128,
+      "loss": 3.1876,
+      "step": 26587
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5921162366867065,
+      "learning_rate": 0.00036144702368684587,
+      "loss": 2.9728,
+      "step": 26588
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5853647589683533,
+      "learning_rate": 0.0003614319538310392,
+      "loss": 3.2357,
+      "step": 26589
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5218216180801392,
+      "learning_rate": 0.0003614168838134327,
+      "loss": 3.1811,
+      "step": 26590
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5547449588775635,
+      "learning_rate": 0.00036140181363406585,
+      "loss": 3.2543,
+      "step": 26591
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5389057993888855,
+      "learning_rate": 0.0003613867432929785,
+      "loss": 3.062,
+      "step": 26592
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5760816931724548,
+      "learning_rate": 0.0003613716727902103,
+      "loss": 3.0541,
+      "step": 26593
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5420430302619934,
+      "learning_rate": 0.00036135660212580084,
+      "loss": 3.1731,
+      "step": 26594
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5819205045700073,
+      "learning_rate": 0.0003613415312997899,
+      "loss": 3.0585,
+      "step": 26595
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5555254817008972,
+      "learning_rate": 0.00036132646031221725,
+      "loss": 3.1077,
+      "step": 26596
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.580630362033844,
+      "learning_rate": 0.0003613113891631225,
+      "loss": 2.9911,
+      "step": 26597
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5484400391578674,
+      "learning_rate": 0.0003612963178525453,
+      "loss": 3.0661,
+      "step": 26598
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5431928634643555,
+      "learning_rate": 0.0003612812463805254,
+      "loss": 3.32,
+      "step": 26599
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5618100166320801,
+      "learning_rate": 0.00036126617474710254,
+      "loss": 3.0773,
+      "step": 26600
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5543720722198486,
+      "learning_rate": 0.0003612511029523163,
+      "loss": 2.8697,
+      "step": 26601
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6032953858375549,
+      "learning_rate": 0.00036123603099620644,
+      "loss": 3.1716,
+      "step": 26602
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5478854179382324,
+      "learning_rate": 0.0003612209588788127,
+      "loss": 3.1367,
+      "step": 26603
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5998310446739197,
+      "learning_rate": 0.0003612058866001748,
+      "loss": 3.1668,
+      "step": 26604
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5461510419845581,
+      "learning_rate": 0.00036119081416033225,
+      "loss": 3.213,
+      "step": 26605
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5615707635879517,
+      "learning_rate": 0.0003611757415593249,
+      "loss": 3.0274,
+      "step": 26606
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5446403622627258,
+      "learning_rate": 0.0003611606687971925,
+      "loss": 3.1591,
+      "step": 26607
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5432937145233154,
+      "learning_rate": 0.0003611455958739746,
+      "loss": 2.9783,
+      "step": 26608
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5421168208122253,
+      "learning_rate": 0.000361130522789711,
+      "loss": 3.0148,
+      "step": 26609
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5991039872169495,
+      "learning_rate": 0.0003611154495444413,
+      "loss": 2.9486,
+      "step": 26610
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5641329884529114,
+      "learning_rate": 0.0003611003761382054,
+      "loss": 2.8076,
+      "step": 26611
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5584830045700073,
+      "learning_rate": 0.00036108530257104274,
+      "loss": 2.9863,
+      "step": 26612
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.591867983341217,
+      "learning_rate": 0.0003610702288429932,
+      "loss": 2.9895,
+      "step": 26613
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5379605293273926,
+      "learning_rate": 0.0003610551549540964,
+      "loss": 3.1134,
+      "step": 26614
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.570203959941864,
+      "learning_rate": 0.00036104008090439215,
+      "loss": 3.0738,
+      "step": 26615
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.531862735748291,
+      "learning_rate": 0.00036102500669392,
+      "loss": 2.9935,
+      "step": 26616
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5866312980651855,
+      "learning_rate": 0.0003610099323227197,
+      "loss": 3.0712,
+      "step": 26617
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6026278734207153,
+      "learning_rate": 0.00036099485779083105,
+      "loss": 3.1248,
+      "step": 26618
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5626994967460632,
+      "learning_rate": 0.0003609797830982937,
+      "loss": 3.2064,
+      "step": 26619
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5667855143547058,
+      "learning_rate": 0.00036096470824514734,
+      "loss": 3.1069,
+      "step": 26620
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5892822742462158,
+      "learning_rate": 0.00036094963323143154,
+      "loss": 3.106,
+      "step": 26621
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5754657983779907,
+      "learning_rate": 0.00036093455805718633,
+      "loss": 2.9988,
+      "step": 26622
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5949243903160095,
+      "learning_rate": 0.0003609194827224511,
+      "loss": 2.8595,
+      "step": 26623
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5688416957855225,
+      "learning_rate": 0.00036090440722726574,
+      "loss": 3.1287,
+      "step": 26624
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.563495934009552,
+      "learning_rate": 0.0003608893315716698,
+      "loss": 3.0331,
+      "step": 26625
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5768406987190247,
+      "learning_rate": 0.0003608742557557031,
+      "loss": 3.1165,
+      "step": 26626
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6202779412269592,
+      "learning_rate": 0.00036085917977940533,
+      "loss": 3.0652,
+      "step": 26627
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.556969940662384,
+      "learning_rate": 0.0003608441036428163,
+      "loss": 3.163,
+      "step": 26628
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5989406704902649,
+      "learning_rate": 0.0003608290273459755,
+      "loss": 2.933,
+      "step": 26629
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5388544201850891,
+      "learning_rate": 0.0003608139508889228,
+      "loss": 3.2661,
+      "step": 26630
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5508213639259338,
+      "learning_rate": 0.00036079887427169775,
+      "loss": 3.278,
+      "step": 26631
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5620518326759338,
+      "learning_rate": 0.00036078379749434025,
+      "loss": 2.8503,
+      "step": 26632
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5819762349128723,
+      "learning_rate": 0.00036076872055689,
+      "loss": 3.1962,
+      "step": 26633
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5653870701789856,
+      "learning_rate": 0.0003607536434593866,
+      "loss": 3.0897,
+      "step": 26634
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6262655258178711,
+      "learning_rate": 0.0003607385662018697,
+      "loss": 3.0569,
+      "step": 26635
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5709400773048401,
+      "learning_rate": 0.0003607234887843792,
+      "loss": 3.0048,
+      "step": 26636
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5970860719680786,
+      "learning_rate": 0.00036070841120695467,
+      "loss": 3.1965,
+      "step": 26637
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5762056112289429,
+      "learning_rate": 0.0003606933334696359,
+      "loss": 3.0631,
+      "step": 26638
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5508270263671875,
+      "learning_rate": 0.0003606782555724625,
+      "loss": 3.2962,
+      "step": 26639
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5481507778167725,
+      "learning_rate": 0.00036066317751547427,
+      "loss": 3.0597,
+      "step": 26640
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5892797708511353,
+      "learning_rate": 0.000360648099298711,
+      "loss": 3.0084,
+      "step": 26641
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5414859056472778,
+      "learning_rate": 0.0003606330209222122,
+      "loss": 3.1328,
+      "step": 26642
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5723474621772766,
+      "learning_rate": 0.0003606179423860178,
+      "loss": 3.0681,
+      "step": 26643
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5581324100494385,
+      "learning_rate": 0.00036060286369016736,
+      "loss": 3.0845,
+      "step": 26644
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5598316788673401,
+      "learning_rate": 0.0003605877848347006,
+      "loss": 3.1011,
+      "step": 26645
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.55232173204422,
+      "learning_rate": 0.0003605727058196572,
+      "loss": 3.3967,
+      "step": 26646
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6095204949378967,
+      "learning_rate": 0.0003605576266450771,
+      "loss": 3.0187,
+      "step": 26647
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6039015650749207,
+      "learning_rate": 0.00036054254731099984,
+      "loss": 2.9391,
+      "step": 26648
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5620260238647461,
+      "learning_rate": 0.00036052746781746513,
+      "loss": 3.067,
+      "step": 26649
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5486672520637512,
+      "learning_rate": 0.0003605123881645127,
+      "loss": 3.137,
+      "step": 26650
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5771086812019348,
+      "learning_rate": 0.0003604973083521823,
+      "loss": 3.0855,
+      "step": 26651
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5530979037284851,
+      "learning_rate": 0.0003604822283805137,
+      "loss": 3.3067,
+      "step": 26652
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5996308922767639,
+      "learning_rate": 0.00036046714824954643,
+      "loss": 2.8549,
+      "step": 26653
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5549845099449158,
+      "learning_rate": 0.00036045206795932045,
+      "loss": 2.9425,
+      "step": 26654
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5953495502471924,
+      "learning_rate": 0.00036043698750987536,
+      "loss": 3.1898,
+      "step": 26655
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5502009987831116,
+      "learning_rate": 0.00036042190690125075,
+      "loss": 3.0692,
+      "step": 26656
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5558758974075317,
+      "learning_rate": 0.0003604068261334865,
+      "loss": 3.0813,
+      "step": 26657
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5885339975357056,
+      "learning_rate": 0.0003603917452066224,
+      "loss": 3.0579,
+      "step": 26658
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5545154809951782,
+      "learning_rate": 0.000360376664120698,
+      "loss": 3.0188,
+      "step": 26659
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6164775490760803,
+      "learning_rate": 0.00036036158287575305,
+      "loss": 3.0655,
+      "step": 26660
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6128799915313721,
+      "learning_rate": 0.0003603465014718273,
+      "loss": 3.0116,
+      "step": 26661
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5506448149681091,
+      "learning_rate": 0.00036033141990896055,
+      "loss": 3.1085,
+      "step": 26662
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5668236613273621,
+      "learning_rate": 0.0003603163381871924,
+      "loss": 3.0768,
+      "step": 26663
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5560147166252136,
+      "learning_rate": 0.00036030125630656266,
+      "loss": 3.0992,
+      "step": 26664
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5709168910980225,
+      "learning_rate": 0.00036028617426711097,
+      "loss": 3.2251,
+      "step": 26665
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5365732908248901,
+      "learning_rate": 0.00036027109206887716,
+      "loss": 2.8935,
+      "step": 26666
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.559626042842865,
+      "learning_rate": 0.00036025600971190076,
+      "loss": 3.0365,
+      "step": 26667
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6036146879196167,
+      "learning_rate": 0.00036024092719622177,
+      "loss": 2.8695,
+      "step": 26668
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5651452541351318,
+      "learning_rate": 0.0003602258445218797,
+      "loss": 3.1308,
+      "step": 26669
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5566524267196655,
+      "learning_rate": 0.0003602107616889145,
+      "loss": 2.8533,
+      "step": 26670
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5793358683586121,
+      "learning_rate": 0.00036019567869736556,
+      "loss": 3.0315,
+      "step": 26671
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5406820774078369,
+      "learning_rate": 0.0003601805955472728,
+      "loss": 3.1241,
+      "step": 26672
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.590872585773468,
+      "learning_rate": 0.000360165512238676,
+      "loss": 3.0732,
+      "step": 26673
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6023873090744019,
+      "learning_rate": 0.00036015042877161484,
+      "loss": 3.323,
+      "step": 26674
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.575550377368927,
+      "learning_rate": 0.00036013534514612894,
+      "loss": 3.0127,
+      "step": 26675
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.583389163017273,
+      "learning_rate": 0.00036012026136225813,
+      "loss": 3.1352,
+      "step": 26676
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5580568909645081,
+      "learning_rate": 0.0003601051774200422,
+      "loss": 3.1552,
+      "step": 26677
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5758119225502014,
+      "learning_rate": 0.0003600900933195207,
+      "loss": 2.7847,
+      "step": 26678
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5357880592346191,
+      "learning_rate": 0.00036007500906073346,
+      "loss": 3.0774,
+      "step": 26679
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5659300684928894,
+      "learning_rate": 0.0003600599246437203,
+      "loss": 3.1311,
+      "step": 26680
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5540302395820618,
+      "learning_rate": 0.00036004484006852083,
+      "loss": 3.2338,
+      "step": 26681
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5465559363365173,
+      "learning_rate": 0.00036002975533517477,
+      "loss": 3.0121,
+      "step": 26682
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6347934007644653,
+      "learning_rate": 0.00036001467044372183,
+      "loss": 3.2469,
+      "step": 26683
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6012324690818787,
+      "learning_rate": 0.000359999585394202,
+      "loss": 2.9777,
+      "step": 26684
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5452451705932617,
+      "learning_rate": 0.00035998450018665467,
+      "loss": 3.0931,
+      "step": 26685
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6085728406906128,
+      "learning_rate": 0.0003599694148211197,
+      "loss": 3.0797,
+      "step": 26686
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5629120469093323,
+      "learning_rate": 0.0003599543292976369,
+      "loss": 3.1281,
+      "step": 26687
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5537635087966919,
+      "learning_rate": 0.00035993924361624587,
+      "loss": 3.176,
+      "step": 26688
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5594296455383301,
+      "learning_rate": 0.0003599241577769864,
+      "loss": 3.1268,
+      "step": 26689
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5878404974937439,
+      "learning_rate": 0.00035990907177989827,
+      "loss": 2.9992,
+      "step": 26690
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5586697459220886,
+      "learning_rate": 0.0003598939856250212,
+      "loss": 3.011,
+      "step": 26691
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5551723837852478,
+      "learning_rate": 0.00035987889931239484,
+      "loss": 2.9541,
+      "step": 26692
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5686328411102295,
+      "learning_rate": 0.000359863812842059,
+      "loss": 3.3019,
+      "step": 26693
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5753780603408813,
+      "learning_rate": 0.00035984872621405337,
+      "loss": 3.2523,
+      "step": 26694
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5661664009094238,
+      "learning_rate": 0.0003598336394284178,
+      "loss": 3.0049,
+      "step": 26695
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5784562826156616,
+      "learning_rate": 0.0003598185524851919,
+      "loss": 2.8481,
+      "step": 26696
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6021720767021179,
+      "learning_rate": 0.0003598034653844154,
+      "loss": 3.2881,
+      "step": 26697
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5414631366729736,
+      "learning_rate": 0.0003597883781261281,
+      "loss": 3.1237,
+      "step": 26698
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5559752583503723,
+      "learning_rate": 0.00035977329071036976,
+      "loss": 3.177,
+      "step": 26699
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6183595061302185,
+      "learning_rate": 0.00035975820313718004,
+      "loss": 2.9357,
+      "step": 26700
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5673224925994873,
+      "learning_rate": 0.00035974311540659876,
+      "loss": 3.0607,
+      "step": 26701
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6092291474342346,
+      "learning_rate": 0.00035972802751866554,
+      "loss": 2.9099,
+      "step": 26702
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.541416347026825,
+      "learning_rate": 0.0003597129394734202,
+      "loss": 3.0918,
+      "step": 26703
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6167910695075989,
+      "learning_rate": 0.0003596978512709025,
+      "loss": 2.8992,
+      "step": 26704
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6436028480529785,
+      "learning_rate": 0.0003596827629111521,
+      "loss": 2.8388,
+      "step": 26705
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5907847285270691,
+      "learning_rate": 0.0003596676743942089,
+      "loss": 3.1507,
+      "step": 26706
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5741157531738281,
+      "learning_rate": 0.00035965258572011246,
+      "loss": 3.16,
+      "step": 26707
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.545328676700592,
+      "learning_rate": 0.00035963749688890253,
+      "loss": 3.1713,
+      "step": 26708
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5556525588035583,
+      "learning_rate": 0.000359622407900619,
+      "loss": 3.1528,
+      "step": 26709
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5546360015869141,
+      "learning_rate": 0.0003596073187553015,
+      "loss": 2.8478,
+      "step": 26710
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6153631806373596,
+      "learning_rate": 0.0003595922294529898,
+      "loss": 3.0284,
+      "step": 26711
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5715863704681396,
+      "learning_rate": 0.0003595771399937236,
+      "loss": 3.2872,
+      "step": 26712
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5800357460975647,
+      "learning_rate": 0.00035956205037754275,
+      "loss": 3.2561,
+      "step": 26713
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.557467520236969,
+      "learning_rate": 0.0003595469606044869,
+      "loss": 3.136,
+      "step": 26714
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5892956852912903,
+      "learning_rate": 0.00035953187067459575,
+      "loss": 3.1928,
+      "step": 26715
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5511610507965088,
+      "learning_rate": 0.00035951678058790913,
+      "loss": 2.9516,
+      "step": 26716
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5597865581512451,
+      "learning_rate": 0.0003595016903444669,
+      "loss": 3.0769,
+      "step": 26717
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5834679007530212,
+      "learning_rate": 0.00035948659994430843,
+      "loss": 3.0175,
+      "step": 26718
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5704622268676758,
+      "learning_rate": 0.0003594715093874738,
+      "loss": 3.0997,
+      "step": 26719
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5578005909919739,
+      "learning_rate": 0.0003594564186740027,
+      "loss": 3.0305,
+      "step": 26720
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6064382195472717,
+      "learning_rate": 0.0003594413278039349,
+      "loss": 3.0124,
+      "step": 26721
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5383819341659546,
+      "learning_rate": 0.00035942623677731004,
+      "loss": 3.1193,
+      "step": 26722
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5984358191490173,
+      "learning_rate": 0.00035941114559416776,
+      "loss": 3.0482,
+      "step": 26723
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5805661082267761,
+      "learning_rate": 0.0003593960542545481,
+      "loss": 2.9751,
+      "step": 26724
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5724626779556274,
+      "learning_rate": 0.0003593809627584906,
+      "loss": 3.0252,
+      "step": 26725
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.545838475227356,
+      "learning_rate": 0.0003593658711060352,
+      "loss": 3.1077,
+      "step": 26726
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5773608088493347,
+      "learning_rate": 0.0003593507792972214,
+      "loss": 3.4112,
+      "step": 26727
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5865770578384399,
+      "learning_rate": 0.0003593356873320891,
+      "loss": 3.1387,
+      "step": 26728
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5342745184898376,
+      "learning_rate": 0.000359320595210678,
+      "loss": 2.9596,
+      "step": 26729
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.567023754119873,
+      "learning_rate": 0.0003593055029330279,
+      "loss": 3.2155,
+      "step": 26730
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5480973124504089,
+      "learning_rate": 0.0003592904104991785,
+      "loss": 3.1711,
+      "step": 26731
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5474660396575928,
+      "learning_rate": 0.0003592753179091696,
+      "loss": 2.9138,
+      "step": 26732
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5918500423431396,
+      "learning_rate": 0.00035926022516304085,
+      "loss": 3.0106,
+      "step": 26733
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5381353497505188,
+      "learning_rate": 0.00035924513226083207,
+      "loss": 2.9064,
+      "step": 26734
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5770012140274048,
+      "learning_rate": 0.00035923003920258306,
+      "loss": 2.8773,
+      "step": 26735
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6201854348182678,
+      "learning_rate": 0.0003592149459883336,
+      "loss": 3.1339,
+      "step": 26736
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.559162437915802,
+      "learning_rate": 0.0003591998526181233,
+      "loss": 3.0548,
+      "step": 26737
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5843203067779541,
+      "learning_rate": 0.0003591847590919919,
+      "loss": 3.2486,
+      "step": 26738
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6074045896530151,
+      "learning_rate": 0.00035916966540997934,
+      "loss": 3.2939,
+      "step": 26739
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.605433464050293,
+      "learning_rate": 0.0003591545715721252,
+      "loss": 3.2019,
+      "step": 26740
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6092764735221863,
+      "learning_rate": 0.00035913947757846936,
+      "loss": 3.0387,
+      "step": 26741
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5730652213096619,
+      "learning_rate": 0.00035912438342905153,
+      "loss": 3.214,
+      "step": 26742
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5732026696205139,
+      "learning_rate": 0.0003591092891239115,
+      "loss": 2.8984,
+      "step": 26743
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5797938704490662,
+      "learning_rate": 0.0003590941946630888,
+      "loss": 2.9541,
+      "step": 26744
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.555505096912384,
+      "learning_rate": 0.00035907910004662343,
+      "loss": 2.9514,
+      "step": 26745
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6078484058380127,
+      "learning_rate": 0.00035906400527455524,
+      "loss": 3.2675,
+      "step": 26746
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5532565116882324,
+      "learning_rate": 0.0003590489103469236,
+      "loss": 3.0698,
+      "step": 26747
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5619004964828491,
+      "learning_rate": 0.00035903381526376863,
+      "loss": 2.9572,
+      "step": 26748
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5572454929351807,
+      "learning_rate": 0.0003590187200251299,
+      "loss": 3.2266,
+      "step": 26749
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5924436450004578,
+      "learning_rate": 0.00035900362463104724,
+      "loss": 2.8406,
+      "step": 26750
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6133246421813965,
+      "learning_rate": 0.00035898852908156044,
+      "loss": 2.9873,
+      "step": 26751
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5816695094108582,
+      "learning_rate": 0.00035897343337670915,
+      "loss": 2.9221,
+      "step": 26752
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5877933502197266,
+      "learning_rate": 0.0003589583375165331,
+      "loss": 3.0776,
+      "step": 26753
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5489791631698608,
+      "learning_rate": 0.00035894324150107225,
+      "loss": 2.8286,
+      "step": 26754
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5626378059387207,
+      "learning_rate": 0.0003589281453303662,
+      "loss": 3.1682,
+      "step": 26755
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5878295302391052,
+      "learning_rate": 0.0003589130490044548,
+      "loss": 3.0382,
+      "step": 26756
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5876728892326355,
+      "learning_rate": 0.00035889795252337783,
+      "loss": 2.8998,
+      "step": 26757
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5746318101882935,
+      "learning_rate": 0.00035888285588717486,
+      "loss": 2.9987,
+      "step": 26758
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5840819478034973,
+      "learning_rate": 0.0003588677590958857,
+      "loss": 3.0352,
+      "step": 26759
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5611565709114075,
+      "learning_rate": 0.0003588526621495504,
+      "loss": 3.0672,
+      "step": 26760
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.595608115196228,
+      "learning_rate": 0.0003588375650482085,
+      "loss": 2.868,
+      "step": 26761
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5854583978652954,
+      "learning_rate": 0.00035882246779189963,
+      "loss": 3.467,
+      "step": 26762
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5326132774353027,
+      "learning_rate": 0.0003588073703806638,
+      "loss": 3.0428,
+      "step": 26763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5872798562049866,
+      "learning_rate": 0.0003587922728145407,
+      "loss": 2.8781,
+      "step": 26764
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5464463829994202,
+      "learning_rate": 0.00035877717509356997,
+      "loss": 3.3402,
+      "step": 26765
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6281166672706604,
+      "learning_rate": 0.0003587620772177915,
+      "loss": 3.1979,
+      "step": 26766
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5617860555648804,
+      "learning_rate": 0.000358746979187245,
+      "loss": 3.0359,
+      "step": 26767
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6079573035240173,
+      "learning_rate": 0.0003587318810019704,
+      "loss": 3.0718,
+      "step": 26768
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5587426424026489,
+      "learning_rate": 0.0003587167826620073,
+      "loss": 2.9656,
+      "step": 26769
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5827581882476807,
+      "learning_rate": 0.0003587016841673953,
+      "loss": 3.1422,
+      "step": 26770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6292572021484375,
+      "learning_rate": 0.00035868658551817457,
+      "loss": 3.1551,
+      "step": 26771
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6260858774185181,
+      "learning_rate": 0.0003586714867143846,
+      "loss": 3.1601,
+      "step": 26772
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5667866468429565,
+      "learning_rate": 0.0003586563877560652,
+      "loss": 3.1435,
+      "step": 26773
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6008673310279846,
+      "learning_rate": 0.0003586412886432562,
+      "loss": 2.9808,
+      "step": 26774
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5359012484550476,
+      "learning_rate": 0.0003586261893759973,
+      "loss": 3.068,
+      "step": 26775
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5466083288192749,
+      "learning_rate": 0.00035861108995432833,
+      "loss": 3.0155,
+      "step": 26776
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5517134666442871,
+      "learning_rate": 0.000358595990378289,
+      "loss": 2.9651,
+      "step": 26777
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.55260169506073,
+      "learning_rate": 0.00035858089064791913,
+      "loss": 2.9182,
+      "step": 26778
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5760247111320496,
+      "learning_rate": 0.0003585657907632585,
+      "loss": 2.8463,
+      "step": 26779
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.596767246723175,
+      "learning_rate": 0.0003585506907243467,
+      "loss": 3.3759,
+      "step": 26780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5780153274536133,
+      "learning_rate": 0.0003585355905312238,
+      "loss": 3.0084,
+      "step": 26781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5638839602470398,
+      "learning_rate": 0.00035852049018392934,
+      "loss": 3.1286,
+      "step": 26782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5510829091072083,
+      "learning_rate": 0.0003585053896825033,
+      "loss": 2.9903,
+      "step": 26783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6048457026481628,
+      "learning_rate": 0.0003584902890269852,
+      "loss": 3.044,
+      "step": 26784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5909347534179688,
+      "learning_rate": 0.0003584751882174149,
+      "loss": 3.0262,
+      "step": 26785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5883007645606995,
+      "learning_rate": 0.00035846008725383237,
+      "loss": 3.2037,
+      "step": 26786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6115594506263733,
+      "learning_rate": 0.0003584449861362771,
+      "loss": 3.1592,
+      "step": 26787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5681421756744385,
+      "learning_rate": 0.00035842988486478903,
+      "loss": 2.844,
+      "step": 26788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6011251211166382,
+      "learning_rate": 0.00035841478343940784,
+      "loss": 3.0141,
+      "step": 26789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5426762700080872,
+      "learning_rate": 0.0003583996818601734,
+      "loss": 3.2084,
+      "step": 26790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6285281181335449,
+      "learning_rate": 0.0003583845801271254,
+      "loss": 3.0548,
+      "step": 26791
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6543646454811096,
+      "learning_rate": 0.0003583694782403036,
+      "loss": 2.9277,
+      "step": 26792
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5835058093070984,
+      "learning_rate": 0.0003583543761997479,
+      "loss": 2.9471,
+      "step": 26793
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5639581084251404,
+      "learning_rate": 0.0003583392740054981,
+      "loss": 2.9748,
+      "step": 26794
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5835383534431458,
+      "learning_rate": 0.0003583241716575937,
+      "loss": 3.1496,
+      "step": 26795
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5585667490959167,
+      "learning_rate": 0.00035830906915607476,
+      "loss": 2.8306,
+      "step": 26796
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6079902052879333,
+      "learning_rate": 0.000358293966500981,
+      "loss": 3.0118,
+      "step": 26797
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6015511155128479,
+      "learning_rate": 0.00035827886369235203,
+      "loss": 3.0294,
+      "step": 26798
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5599113702774048,
+      "learning_rate": 0.00035826376073022783,
+      "loss": 3.0174,
+      "step": 26799
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5897526741027832,
+      "learning_rate": 0.0003582486576146481,
+      "loss": 3.1255,
+      "step": 26800
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.570527195930481,
+      "learning_rate": 0.00035823355434565256,
+      "loss": 2.9353,
+      "step": 26801
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5881776809692383,
+      "learning_rate": 0.0003582184509232811,
+      "loss": 3.1936,
+      "step": 26802
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5746738314628601,
+      "learning_rate": 0.00035820334734757336,
+      "loss": 2.9701,
+      "step": 26803
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5680601000785828,
+      "learning_rate": 0.0003581882436185693,
+      "loss": 3.3507,
+      "step": 26804
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5403523445129395,
+      "learning_rate": 0.0003581731397363085,
+      "loss": 3.082,
+      "step": 26805
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5722833275794983,
+      "learning_rate": 0.0003581580357008309,
+      "loss": 3.041,
+      "step": 26806
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5738394260406494,
+      "learning_rate": 0.00035814293151217625,
+      "loss": 2.9287,
+      "step": 26807
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6539166569709778,
+      "learning_rate": 0.00035812782717038435,
+      "loss": 3.1007,
+      "step": 26808
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5289335250854492,
+      "learning_rate": 0.0003581127226754949,
+      "loss": 2.8963,
+      "step": 26809
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5261499285697937,
+      "learning_rate": 0.0003580976180275477,
+      "loss": 3.2443,
+      "step": 26810
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5962910652160645,
+      "learning_rate": 0.00035808251322658256,
+      "loss": 2.9977,
+      "step": 26811
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6628442406654358,
+      "learning_rate": 0.0003580674082726392,
+      "loss": 3.1296,
+      "step": 26812
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5591052770614624,
+      "learning_rate": 0.00035805230316575756,
+      "loss": 3.0394,
+      "step": 26813
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5713579058647156,
+      "learning_rate": 0.00035803719790597727,
+      "loss": 3.0677,
+      "step": 26814
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5537461042404175,
+      "learning_rate": 0.0003580220924933382,
+      "loss": 2.9197,
+      "step": 26815
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5648530721664429,
+      "learning_rate": 0.0003580069869278801,
+      "loss": 3.0549,
+      "step": 26816
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.595908522605896,
+      "learning_rate": 0.0003579918812096428,
+      "loss": 2.9516,
+      "step": 26817
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5620924830436707,
+      "learning_rate": 0.00035797677533866597,
+      "loss": 2.9171,
+      "step": 26818
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5861213803291321,
+      "learning_rate": 0.00035796166931498956,
+      "loss": 2.938,
+      "step": 26819
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6200598478317261,
+      "learning_rate": 0.0003579465631386532,
+      "loss": 3.1466,
+      "step": 26820
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5952265858650208,
+      "learning_rate": 0.00035793145680969665,
+      "loss": 3.194,
+      "step": 26821
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5975573062896729,
+      "learning_rate": 0.00035791635032816,
+      "loss": 3.1285,
+      "step": 26822
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5594826340675354,
+      "learning_rate": 0.0003579012436940827,
+      "loss": 3.0251,
+      "step": 26823
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5513445734977722,
+      "learning_rate": 0.00035788613690750467,
+      "loss": 3.175,
+      "step": 26824
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5082058310508728,
+      "learning_rate": 0.0003578710299684657,
+      "loss": 3.244,
+      "step": 26825
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.560768723487854,
+      "learning_rate": 0.00035785592287700563,
+      "loss": 3.1634,
+      "step": 26826
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6477553844451904,
+      "learning_rate": 0.00035784081563316415,
+      "loss": 2.7827,
+      "step": 26827
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5783835053443909,
+      "learning_rate": 0.0003578257082369811,
+      "loss": 3.0997,
+      "step": 26828
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6416663527488708,
+      "learning_rate": 0.0003578106006884962,
+      "loss": 3.2326,
+      "step": 26829
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5797775983810425,
+      "learning_rate": 0.00035779549298774947,
+      "loss": 3.1772,
+      "step": 26830
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5553704500198364,
+      "learning_rate": 0.00035778038513478033,
+      "loss": 2.9769,
+      "step": 26831
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5943940281867981,
+      "learning_rate": 0.0003577652771296289,
+      "loss": 3.226,
+      "step": 26832
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5958367586135864,
+      "learning_rate": 0.00035775016897233487,
+      "loss": 3.1038,
+      "step": 26833
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5723438858985901,
+      "learning_rate": 0.000357735060662938,
+      "loss": 3.147,
+      "step": 26834
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5667865872383118,
+      "learning_rate": 0.0003577199522014781,
+      "loss": 3.0796,
+      "step": 26835
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.585544228553772,
+      "learning_rate": 0.00035770484358799474,
+      "loss": 3.0049,
+      "step": 26836
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.586850106716156,
+      "learning_rate": 0.00035768973482252814,
+      "loss": 3.0851,
+      "step": 26837
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6466086506843567,
+      "learning_rate": 0.00035767462590511793,
+      "loss": 2.9241,
+      "step": 26838
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5506917834281921,
+      "learning_rate": 0.00035765951683580367,
+      "loss": 3.1222,
+      "step": 26839
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.545690655708313,
+      "learning_rate": 0.0003576444076146255,
+      "loss": 3.0191,
+      "step": 26840
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6054341793060303,
+      "learning_rate": 0.00035762929824162295,
+      "loss": 2.9853,
+      "step": 26841
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5546567440032959,
+      "learning_rate": 0.00035761418871683596,
+      "loss": 2.9322,
+      "step": 26842
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5472365617752075,
+      "learning_rate": 0.00035759907904030425,
+      "loss": 3.009,
+      "step": 26843
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5797613263130188,
+      "learning_rate": 0.0003575839692120677,
+      "loss": 3.1111,
+      "step": 26844
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5437989234924316,
+      "learning_rate": 0.0003575688592321661,
+      "loss": 3.1502,
+      "step": 26845
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.568385124206543,
+      "learning_rate": 0.00035755374910063914,
+      "loss": 2.8418,
+      "step": 26846
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5577401518821716,
+      "learning_rate": 0.0003575386388175266,
+      "loss": 2.9497,
+      "step": 26847
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5725609064102173,
+      "learning_rate": 0.0003575235283828685,
+      "loss": 3.0207,
+      "step": 26848
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6471617817878723,
+      "learning_rate": 0.00035750841779670444,
+      "loss": 2.9642,
+      "step": 26849
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5361818075180054,
+      "learning_rate": 0.0003574933070590743,
+      "loss": 2.933,
+      "step": 26850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5480607748031616,
+      "learning_rate": 0.00035747819617001783,
+      "loss": 3.1263,
+      "step": 26851
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5243863463401794,
+      "learning_rate": 0.00035746308512957486,
+      "loss": 3.0354,
+      "step": 26852
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5858286023139954,
+      "learning_rate": 0.0003574479739377852,
+      "loss": 3.1605,
+      "step": 26853
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5662160515785217,
+      "learning_rate": 0.0003574328625946885,
+      "loss": 3.2539,
+      "step": 26854
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5930647253990173,
+      "learning_rate": 0.0003574177511003248,
+      "loss": 2.7858,
+      "step": 26855
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5709060430526733,
+      "learning_rate": 0.0003574026394547339,
+      "loss": 3.0939,
+      "step": 26856
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6525852680206299,
+      "learning_rate": 0.00035738752765795533,
+      "loss": 3.0615,
+      "step": 26857
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5705323815345764,
+      "learning_rate": 0.0003573724157100291,
+      "loss": 2.8539,
+      "step": 26858
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5645226836204529,
+      "learning_rate": 0.00035735730361099506,
+      "loss": 3.1802,
+      "step": 26859
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5563384294509888,
+      "learning_rate": 0.00035734219136089287,
+      "loss": 3.1764,
+      "step": 26860
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5454398989677429,
+      "learning_rate": 0.0003573270789597623,
+      "loss": 3.2813,
+      "step": 26861
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5893054008483887,
+      "learning_rate": 0.0003573119664076433,
+      "loss": 2.9479,
+      "step": 26862
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6017407178878784,
+      "learning_rate": 0.00035729685370457557,
+      "loss": 2.981,
+      "step": 26863
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5848926901817322,
+      "learning_rate": 0.00035728174085059906,
+      "loss": 2.9967,
+      "step": 26864
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5659344792366028,
+      "learning_rate": 0.0003572666278457533,
+      "loss": 3.0336,
+      "step": 26865
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5689499378204346,
+      "learning_rate": 0.0003572515146900784,
+      "loss": 3.1079,
+      "step": 26866
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5527841448783875,
+      "learning_rate": 0.00035723640138361404,
+      "loss": 2.9584,
+      "step": 26867
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5790805220603943,
+      "learning_rate": 0.0003572212879263999,
+      "loss": 3.1476,
+      "step": 26868
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5489739775657654,
+      "learning_rate": 0.00035720617431847604,
+      "loss": 3.0897,
+      "step": 26869
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5729896426200867,
+      "learning_rate": 0.00035719106055988206,
+      "loss": 3.0103,
+      "step": 26870
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6073732376098633,
+      "learning_rate": 0.0003571759466506578,
+      "loss": 2.9835,
+      "step": 26871
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.63726407289505,
+      "learning_rate": 0.0003571608325908431,
+      "loss": 2.8864,
+      "step": 26872
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5688294172286987,
+      "learning_rate": 0.0003571457183804779,
+      "loss": 2.7645,
+      "step": 26873
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5426622033119202,
+      "learning_rate": 0.00035713060401960174,
+      "loss": 3.0762,
+      "step": 26874
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.57330322265625,
+      "learning_rate": 0.0003571154895082546,
+      "loss": 3.0234,
+      "step": 26875
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5708470940589905,
+      "learning_rate": 0.0003571003748464762,
+      "loss": 2.8983,
+      "step": 26876
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5946699976921082,
+      "learning_rate": 0.00035708526003430647,
+      "loss": 3.0277,
+      "step": 26877
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5799338817596436,
+      "learning_rate": 0.0003570701450717851,
+      "loss": 3.0412,
+      "step": 26878
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.558404803276062,
+      "learning_rate": 0.000357055029958952,
+      "loss": 3.0555,
+      "step": 26879
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7721039056777954,
+      "learning_rate": 0.00035703991469584686,
+      "loss": 3.0843,
+      "step": 26880
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5947189927101135,
+      "learning_rate": 0.0003570247992825097,
+      "loss": 3.2052,
+      "step": 26881
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5546451807022095,
+      "learning_rate": 0.00035700968371898007,
+      "loss": 3.1527,
+      "step": 26882
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5192341208457947,
+      "learning_rate": 0.0003569945680052979,
+      "loss": 3.0922,
+      "step": 26883
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.570080041885376,
+      "learning_rate": 0.00035697945214150306,
+      "loss": 3.1406,
+      "step": 26884
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5520020723342896,
+      "learning_rate": 0.00035696433612763536,
+      "loss": 3.2305,
+      "step": 26885
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5862802267074585,
+      "learning_rate": 0.0003569492199637345,
+      "loss": 3.0221,
+      "step": 26886
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5665404200553894,
+      "learning_rate": 0.0003569341036498404,
+      "loss": 3.0495,
+      "step": 26887
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5680643320083618,
+      "learning_rate": 0.00035691898718599274,
+      "loss": 3.2575,
+      "step": 26888
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5598427057266235,
+      "learning_rate": 0.0003569038705722315,
+      "loss": 2.9607,
+      "step": 26889
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5373253226280212,
+      "learning_rate": 0.00035688875380859636,
+      "loss": 2.9572,
+      "step": 26890
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5592238306999207,
+      "learning_rate": 0.00035687363689512714,
+      "loss": 2.8467,
+      "step": 26891
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7021776437759399,
+      "learning_rate": 0.0003568585198318639,
+      "loss": 2.9761,
+      "step": 26892
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6009718775749207,
+      "learning_rate": 0.00035684340261884606,
+      "loss": 3.3808,
+      "step": 26893
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6137123107910156,
+      "learning_rate": 0.0003568282852561137,
+      "loss": 3.1077,
+      "step": 26894
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5503768920898438,
+      "learning_rate": 0.0003568131677437066,
+      "loss": 3.0737,
+      "step": 26895
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6017237901687622,
+      "learning_rate": 0.0003567980500816646,
+      "loss": 3.045,
+      "step": 26896
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5975255966186523,
+      "learning_rate": 0.00035678293227002744,
+      "loss": 3.138,
+      "step": 26897
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5789278745651245,
+      "learning_rate": 0.0003567678143088348,
+      "loss": 2.9976,
+      "step": 26898
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6184183359146118,
+      "learning_rate": 0.0003567526961981269,
+      "loss": 2.8672,
+      "step": 26899
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5539001822471619,
+      "learning_rate": 0.0003567375779379432,
+      "loss": 3.2786,
+      "step": 26900
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5661340951919556,
+      "learning_rate": 0.00035672245952832367,
+      "loss": 3.013,
+      "step": 26901
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5806612968444824,
+      "learning_rate": 0.000356707340969308,
+      "loss": 2.8103,
+      "step": 26902
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6162651181221008,
+      "learning_rate": 0.00035669222226093625,
+      "loss": 3.3411,
+      "step": 26903
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.615796685218811,
+      "learning_rate": 0.000356677103403248,
+      "loss": 2.9888,
+      "step": 26904
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5535275340080261,
+      "learning_rate": 0.0003566619843962832,
+      "loss": 3.151,
+      "step": 26905
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5614963173866272,
+      "learning_rate": 0.0003566468652400817,
+      "loss": 3.1728,
+      "step": 26906
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5885602235794067,
+      "learning_rate": 0.0003566317459346832,
+      "loss": 2.9835,
+      "step": 26907
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5614680647850037,
+      "learning_rate": 0.00035661662648012757,
+      "loss": 3.2124,
+      "step": 26908
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6076555848121643,
+      "learning_rate": 0.0003566015068764546,
+      "loss": 2.9445,
+      "step": 26909
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5833714604377747,
+      "learning_rate": 0.0003565863871237043,
+      "loss": 3.1111,
+      "step": 26910
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5717360973358154,
+      "learning_rate": 0.0003565712672219162,
+      "loss": 2.9425,
+      "step": 26911
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5546928644180298,
+      "learning_rate": 0.0003565561471711303,
+      "loss": 2.9382,
+      "step": 26912
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6168055534362793,
+      "learning_rate": 0.0003565410269713865,
+      "loss": 2.9419,
+      "step": 26913
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6003530025482178,
+      "learning_rate": 0.0003565259066227244,
+      "loss": 3.2024,
+      "step": 26914
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5346034169197083,
+      "learning_rate": 0.000356510786125184,
+      "loss": 3.1215,
+      "step": 26915
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5668526291847229,
+      "learning_rate": 0.00035649566547880495,
+      "loss": 3.1087,
+      "step": 26916
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7049530148506165,
+      "learning_rate": 0.0003564805446836273,
+      "loss": 3.0214,
+      "step": 26917
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5999130010604858,
+      "learning_rate": 0.0003564654237396907,
+      "loss": 2.9735,
+      "step": 26918
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5857048034667969,
+      "learning_rate": 0.0003564503026470351,
+      "loss": 2.8567,
+      "step": 26919
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5745254158973694,
+      "learning_rate": 0.0003564351814057003,
+      "loss": 3.0944,
+      "step": 26920
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5591810941696167,
+      "learning_rate": 0.0003564200600157261,
+      "loss": 3.124,
+      "step": 26921
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5647045373916626,
+      "learning_rate": 0.0003564049384771522,
+      "loss": 2.9827,
+      "step": 26922
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5440300107002258,
+      "learning_rate": 0.00035638981679001863,
+      "loss": 3.0062,
+      "step": 26923
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5504894852638245,
+      "learning_rate": 0.0003563746949543651,
+      "loss": 3.1177,
+      "step": 26924
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5864576697349548,
+      "learning_rate": 0.00035635957297023153,
+      "loss": 3.037,
+      "step": 26925
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5730857849121094,
+      "learning_rate": 0.00035634445083765765,
+      "loss": 2.8243,
+      "step": 26926
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5582221746444702,
+      "learning_rate": 0.0003563293285566833,
+      "loss": 3.1537,
+      "step": 26927
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5400102138519287,
+      "learning_rate": 0.00035631420612734836,
+      "loss": 2.955,
+      "step": 26928
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5986092686653137,
+      "learning_rate": 0.00035629908354969263,
+      "loss": 3.1307,
+      "step": 26929
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5586465001106262,
+      "learning_rate": 0.000356283960823756,
+      "loss": 3.022,
+      "step": 26930
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6682041883468628,
+      "learning_rate": 0.0003562688379495782,
+      "loss": 2.9636,
+      "step": 26931
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5432518124580383,
+      "learning_rate": 0.0003562537149271992,
+      "loss": 3.2366,
+      "step": 26932
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6639944314956665,
+      "learning_rate": 0.0003562385917566587,
+      "loss": 3.053,
+      "step": 26933
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5308532118797302,
+      "learning_rate": 0.0003562234684379965,
+      "loss": 3.0926,
+      "step": 26934
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5658736824989319,
+      "learning_rate": 0.0003562083449712525,
+      "loss": 3.069,
+      "step": 26935
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5990812182426453,
+      "learning_rate": 0.0003561932213564666,
+      "loss": 3.0671,
+      "step": 26936
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5363149046897888,
+      "learning_rate": 0.00035617809759367854,
+      "loss": 2.8551,
+      "step": 26937
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5616311430931091,
+      "learning_rate": 0.0003561629736829282,
+      "loss": 3.2784,
+      "step": 26938
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5642018914222717,
+      "learning_rate": 0.00035614784962425546,
+      "loss": 3.0111,
+      "step": 26939
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.558562695980072,
+      "learning_rate": 0.0003561327254177,
+      "loss": 3.0165,
+      "step": 26940
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5792576670646667,
+      "learning_rate": 0.00035611760106330176,
+      "loss": 3.0326,
+      "step": 26941
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5646781921386719,
+      "learning_rate": 0.0003561024765611005,
+      "loss": 3.0448,
+      "step": 26942
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5456598997116089,
+      "learning_rate": 0.00035608735191113624,
+      "loss": 2.9734,
+      "step": 26943
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.601516842842102,
+      "learning_rate": 0.0003560722271134486,
+      "loss": 3.1814,
+      "step": 26944
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5825067758560181,
+      "learning_rate": 0.0003560571021680775,
+      "loss": 2.9143,
+      "step": 26945
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5741273164749146,
+      "learning_rate": 0.00035604197707506287,
+      "loss": 2.8824,
+      "step": 26946
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6200869083404541,
+      "learning_rate": 0.0003560268518344444,
+      "loss": 3.2152,
+      "step": 26947
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6041608452796936,
+      "learning_rate": 0.000356011726446262,
+      "loss": 3.0224,
+      "step": 26948
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5746281147003174,
+      "learning_rate": 0.00035599660091055544,
+      "loss": 3.0115,
+      "step": 26949
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5583726763725281,
+      "learning_rate": 0.0003559814752273647,
+      "loss": 3.2274,
+      "step": 26950
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5739026665687561,
+      "learning_rate": 0.00035596634939672947,
+      "loss": 3.0633,
+      "step": 26951
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.551236629486084,
+      "learning_rate": 0.0003559512234186896,
+      "loss": 3.0881,
+      "step": 26952
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5410869717597961,
+      "learning_rate": 0.0003559360972932851,
+      "loss": 3.0438,
+      "step": 26953
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6108107566833496,
+      "learning_rate": 0.00035592097102055554,
+      "loss": 3.1604,
+      "step": 26954
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5781041979789734,
+      "learning_rate": 0.00035590584460054104,
+      "loss": 3.1562,
+      "step": 26955
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5774297118186951,
+      "learning_rate": 0.00035589071803328125,
+      "loss": 2.885,
+      "step": 26956
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5566204786300659,
+      "learning_rate": 0.00035587559131881603,
+      "loss": 3.1957,
+      "step": 26957
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5503354668617249,
+      "learning_rate": 0.00035586046445718533,
+      "loss": 3.0237,
+      "step": 26958
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5675086975097656,
+      "learning_rate": 0.00035584533744842877,
+      "loss": 3.0587,
+      "step": 26959
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5769832730293274,
+      "learning_rate": 0.00035583021029258647,
+      "loss": 3.1677,
+      "step": 26960
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.614290714263916,
+      "learning_rate": 0.00035581508298969817,
+      "loss": 2.9065,
+      "step": 26961
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5537709593772888,
+      "learning_rate": 0.00035579995553980365,
+      "loss": 3.1548,
+      "step": 26962
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6154881119728088,
+      "learning_rate": 0.0003557848279429427,
+      "loss": 3.3156,
+      "step": 26963
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5546302795410156,
+      "learning_rate": 0.00035576970019915536,
+      "loss": 2.939,
+      "step": 26964
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6049430966377258,
+      "learning_rate": 0.00035575457230848127,
+      "loss": 2.7861,
+      "step": 26965
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5868650078773499,
+      "learning_rate": 0.00035573944427096045,
+      "loss": 3.0257,
+      "step": 26966
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5958408713340759,
+      "learning_rate": 0.0003557243160866326,
+      "loss": 2.7296,
+      "step": 26967
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5856796503067017,
+      "learning_rate": 0.00035570918775553763,
+      "loss": 3.0672,
+      "step": 26968
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5468848347663879,
+      "learning_rate": 0.0003556940592777155,
+      "loss": 2.9349,
+      "step": 26969
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5882099270820618,
+      "learning_rate": 0.0003556789306532057,
+      "loss": 3.1575,
+      "step": 26970
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5558639168739319,
+      "learning_rate": 0.0003556638018820485,
+      "loss": 2.948,
+      "step": 26971
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.582410454750061,
+      "learning_rate": 0.0003556486729642835,
+      "loss": 3.0795,
+      "step": 26972
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5897334218025208,
+      "learning_rate": 0.0003556335438999506,
+      "loss": 2.867,
+      "step": 26973
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5801978707313538,
+      "learning_rate": 0.0003556184146890897,
+      "loss": 3.112,
+      "step": 26974
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5620837807655334,
+      "learning_rate": 0.00035560328533174057,
+      "loss": 3.1498,
+      "step": 26975
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5551138520240784,
+      "learning_rate": 0.00035558815582794305,
+      "loss": 3.1267,
+      "step": 26976
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5612112283706665,
+      "learning_rate": 0.00035557302617773704,
+      "loss": 3.1203,
+      "step": 26977
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6193846464157104,
+      "learning_rate": 0.0003555578963811624,
+      "loss": 3.0384,
+      "step": 26978
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6214466691017151,
+      "learning_rate": 0.0003555427664382589,
+      "loss": 3.0276,
+      "step": 26979
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5702659487724304,
+      "learning_rate": 0.00035552763634906655,
+      "loss": 3.2472,
+      "step": 26980
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5576087832450867,
+      "learning_rate": 0.000355512506113625,
+      "loss": 2.9014,
+      "step": 26981
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.578071653842926,
+      "learning_rate": 0.00035549737573197415,
+      "loss": 3.1932,
+      "step": 26982
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5625930428504944,
+      "learning_rate": 0.00035548224520415403,
+      "loss": 3.0508,
+      "step": 26983
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6400315761566162,
+      "learning_rate": 0.0003554671145302044,
+      "loss": 3.0855,
+      "step": 26984
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5485404133796692,
+      "learning_rate": 0.00035545198371016477,
+      "loss": 3.1335,
+      "step": 26985
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5850667357444763,
+      "learning_rate": 0.0003554368527440756,
+      "loss": 2.8681,
+      "step": 26986
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5740143656730652,
+      "learning_rate": 0.00035542172163197626,
+      "loss": 3.0152,
+      "step": 26987
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5631483793258667,
+      "learning_rate": 0.0003554065903739068,
+      "loss": 3.1364,
+      "step": 26988
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.604009211063385,
+      "learning_rate": 0.0003553914589699071,
+      "loss": 3.162,
+      "step": 26989
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5691906213760376,
+      "learning_rate": 0.00035537632742001685,
+      "loss": 3.1606,
+      "step": 26990
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.629859983921051,
+      "learning_rate": 0.00035536119572427603,
+      "loss": 2.9259,
+      "step": 26991
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5835464000701904,
+      "learning_rate": 0.00035534606388272456,
+      "loss": 3.2523,
+      "step": 26992
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5969851613044739,
+      "learning_rate": 0.00035533093189540216,
+      "loss": 3.1687,
+      "step": 26993
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.541480302810669,
+      "learning_rate": 0.0003553157997623488,
+      "loss": 2.9884,
+      "step": 26994
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5837870240211487,
+      "learning_rate": 0.0003553006674836041,
+      "loss": 2.7295,
+      "step": 26995
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.609303891658783,
+      "learning_rate": 0.00035528553505920825,
+      "loss": 3.0732,
+      "step": 26996
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5507553815841675,
+      "learning_rate": 0.00035527040248920085,
+      "loss": 3.0481,
+      "step": 26997
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5484434366226196,
+      "learning_rate": 0.00035525526977362197,
+      "loss": 3.0371,
+      "step": 26998
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5981265902519226,
+      "learning_rate": 0.0003552401369125112,
+      "loss": 3.0613,
+      "step": 26999
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5397427678108215,
+      "learning_rate": 0.00035522500390590856,
+      "loss": 3.1179,
+      "step": 27000
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5383358001708984,
+      "learning_rate": 0.00035520987075385394,
+      "loss": 3.0821,
+      "step": 27001
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.564812183380127,
+      "learning_rate": 0.0003551947374563871,
+      "loss": 2.9963,
+      "step": 27002
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5476768016815186,
+      "learning_rate": 0.00035517960401354806,
+      "loss": 3.0847,
+      "step": 27003
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5901896357536316,
+      "learning_rate": 0.00035516447042537643,
+      "loss": 3.2513,
+      "step": 27004
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5683808326721191,
+      "learning_rate": 0.0003551493366919123,
+      "loss": 3.2693,
+      "step": 27005
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.563849151134491,
+      "learning_rate": 0.00035513420281319533,
+      "loss": 3.1826,
+      "step": 27006
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5702254772186279,
+      "learning_rate": 0.0003551190687892655,
+      "loss": 3.1978,
+      "step": 27007
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5525578856468201,
+      "learning_rate": 0.00035510393462016273,
+      "loss": 3.1693,
+      "step": 27008
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5483326315879822,
+      "learning_rate": 0.0003550888003059268,
+      "loss": 3.0161,
+      "step": 27009
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6483874320983887,
+      "learning_rate": 0.0003550736658465975,
+      "loss": 3.1485,
+      "step": 27010
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5594171285629272,
+      "learning_rate": 0.0003550585312422147,
+      "loss": 2.9653,
+      "step": 27011
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5625665187835693,
+      "learning_rate": 0.0003550433964928185,
+      "loss": 3.1067,
+      "step": 27012
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5742067694664001,
+      "learning_rate": 0.0003550282615984485,
+      "loss": 3.0806,
+      "step": 27013
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6136051416397095,
+      "learning_rate": 0.0003550131265591446,
+      "loss": 3.2865,
+      "step": 27014
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7125194072723389,
+      "learning_rate": 0.0003549979913749468,
+      "loss": 3.0679,
+      "step": 27015
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5937182307243347,
+      "learning_rate": 0.0003549828560458948,
+      "loss": 2.9668,
+      "step": 27016
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5569949150085449,
+      "learning_rate": 0.0003549677205720286,
+      "loss": 3.0069,
+      "step": 27017
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5426252484321594,
+      "learning_rate": 0.000354952584953388,
+      "loss": 3.1013,
+      "step": 27018
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5773874521255493,
+      "learning_rate": 0.0003549374491900128,
+      "loss": 2.9514,
+      "step": 27019
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5283383131027222,
+      "learning_rate": 0.00035492231328194306,
+      "loss": 2.845,
+      "step": 27020
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5627959966659546,
+      "learning_rate": 0.00035490717722921834,
+      "loss": 3.0731,
+      "step": 27021
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.584900975227356,
+      "learning_rate": 0.0003548920410318788,
+      "loss": 2.9836,
+      "step": 27022
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5437769889831543,
+      "learning_rate": 0.0003548769046899642,
+      "loss": 3.0244,
+      "step": 27023
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5971786975860596,
+      "learning_rate": 0.00035486176820351435,
+      "loss": 3.2019,
+      "step": 27024
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5955939888954163,
+      "learning_rate": 0.0003548466315725691,
+      "loss": 2.9108,
+      "step": 27025
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5769616961479187,
+      "learning_rate": 0.0003548314947971685,
+      "loss": 2.9552,
+      "step": 27026
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5936218500137329,
+      "learning_rate": 0.0003548163578773522,
+      "loss": 3.0934,
+      "step": 27027
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5732731819152832,
+      "learning_rate": 0.0003548012208131602,
+      "loss": 3.1205,
+      "step": 27028
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5782310962677002,
+      "learning_rate": 0.00035478608360463233,
+      "loss": 2.8778,
+      "step": 27029
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5631444454193115,
+      "learning_rate": 0.0003547709462518085,
+      "loss": 2.9783,
+      "step": 27030
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5475609302520752,
+      "learning_rate": 0.0003547558087547285,
+      "loss": 2.9648,
+      "step": 27031
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5806057453155518,
+      "learning_rate": 0.0003547406711134322,
+      "loss": 2.8895,
+      "step": 27032
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5406964421272278,
+      "learning_rate": 0.00035472553332795956,
+      "loss": 3.2604,
+      "step": 27033
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5376538634300232,
+      "learning_rate": 0.00035471039539835045,
+      "loss": 3.0621,
+      "step": 27034
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5387920141220093,
+      "learning_rate": 0.00035469525732464464,
+      "loss": 3.0724,
+      "step": 27035
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6255925297737122,
+      "learning_rate": 0.0003546801191068819,
+      "loss": 3.0135,
+      "step": 27036
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5752521753311157,
+      "learning_rate": 0.00035466498074510246,
+      "loss": 3.1237,
+      "step": 27037
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5635229349136353,
+      "learning_rate": 0.0003546498422393459,
+      "loss": 3.0906,
+      "step": 27038
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5960707664489746,
+      "learning_rate": 0.0003546347035896522,
+      "loss": 2.9693,
+      "step": 27039
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5936806201934814,
+      "learning_rate": 0.00035461956479606113,
+      "loss": 3.1782,
+      "step": 27040
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5770202875137329,
+      "learning_rate": 0.0003546044258586127,
+      "loss": 3.1503,
+      "step": 27041
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.607388973236084,
+      "learning_rate": 0.0003545892867773467,
+      "loss": 3.1449,
+      "step": 27042
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5307580232620239,
+      "learning_rate": 0.00035457414755230303,
+      "loss": 3.0214,
+      "step": 27043
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5784158110618591,
+      "learning_rate": 0.0003545590081835215,
+      "loss": 3.0766,
+      "step": 27044
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5791460275650024,
+      "learning_rate": 0.0003545438686710422,
+      "loss": 2.999,
+      "step": 27045
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5575026869773865,
+      "learning_rate": 0.00035452872901490476,
+      "loss": 3.0514,
+      "step": 27046
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.54404217004776,
+      "learning_rate": 0.0003545135892151491,
+      "loss": 2.9649,
+      "step": 27047
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5947194695472717,
+      "learning_rate": 0.00035449844927181525,
+      "loss": 3.0767,
+      "step": 27048
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5407034158706665,
+      "learning_rate": 0.00035448330918494284,
+      "loss": 2.9749,
+      "step": 27049
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5737541317939758,
+      "learning_rate": 0.00035446816895457194,
+      "loss": 3.2021,
+      "step": 27050
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5305442214012146,
+      "learning_rate": 0.00035445302858074234,
+      "loss": 3.2367,
+      "step": 27051
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5993185639381409,
+      "learning_rate": 0.000354437888063494,
+      "loss": 3.2942,
+      "step": 27052
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5892103314399719,
+      "learning_rate": 0.00035442274740286666,
+      "loss": 3.1939,
+      "step": 27053
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6093845963478088,
+      "learning_rate": 0.00035440760659890034,
+      "loss": 2.9334,
+      "step": 27054
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5915141701698303,
+      "learning_rate": 0.00035439246565163485,
+      "loss": 3.1818,
+      "step": 27055
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5777159333229065,
+      "learning_rate": 0.0003543773245611101,
+      "loss": 3.1014,
+      "step": 27056
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6174264550209045,
+      "learning_rate": 0.0003543621833273658,
+      "loss": 3.036,
+      "step": 27057
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.584246814250946,
+      "learning_rate": 0.00035434704195044206,
+      "loss": 3.0767,
+      "step": 27058
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5757298469543457,
+      "learning_rate": 0.0003543319004303787,
+      "loss": 3.1766,
+      "step": 27059
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5471251606941223,
+      "learning_rate": 0.0003543167587672156,
+      "loss": 3.087,
+      "step": 27060
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6035965085029602,
+      "learning_rate": 0.00035430161696099256,
+      "loss": 3.0348,
+      "step": 27061
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5603316426277161,
+      "learning_rate": 0.00035428647501174943,
+      "loss": 3.1404,
+      "step": 27062
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6384663581848145,
+      "learning_rate": 0.00035427133291952633,
+      "loss": 2.9896,
+      "step": 27063
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5779756903648376,
+      "learning_rate": 0.00035425619068436294,
+      "loss": 3.0965,
+      "step": 27064
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5718583464622498,
+      "learning_rate": 0.00035424104830629916,
+      "loss": 3.1002,
+      "step": 27065
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5677472352981567,
+      "learning_rate": 0.0003542259057853749,
+      "loss": 3.1221,
+      "step": 27066
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.575522243976593,
+      "learning_rate": 0.00035421076312163,
+      "loss": 3.0776,
+      "step": 27067
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5618709325790405,
+      "learning_rate": 0.0003541956203151044,
+      "loss": 2.8956,
+      "step": 27068
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5451009273529053,
+      "learning_rate": 0.000354180477365838,
+      "loss": 3.0641,
+      "step": 27069
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.618395984172821,
+      "learning_rate": 0.0003541653342738706,
+      "loss": 2.8972,
+      "step": 27070
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5467285513877869,
+      "learning_rate": 0.0003541501910392422,
+      "loss": 3.0502,
+      "step": 27071
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6058321595191956,
+      "learning_rate": 0.00035413504766199257,
+      "loss": 2.9549,
+      "step": 27072
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5389471054077148,
+      "learning_rate": 0.00035411990414216165,
+      "loss": 3.2793,
+      "step": 27073
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6167011857032776,
+      "learning_rate": 0.0003541047604797894,
+      "loss": 3.0513,
+      "step": 27074
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6449909210205078,
+      "learning_rate": 0.00035408961667491554,
+      "loss": 2.965,
+      "step": 27075
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5643405914306641,
+      "learning_rate": 0.0003540744727275801,
+      "loss": 2.8066,
+      "step": 27076
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5791833400726318,
+      "learning_rate": 0.0003540593286378228,
+      "loss": 3.0858,
+      "step": 27077
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5517277717590332,
+      "learning_rate": 0.0003540441844056837,
+      "loss": 3.2508,
+      "step": 27078
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5787827372550964,
+      "learning_rate": 0.00035402904003120254,
+      "loss": 3.0538,
+      "step": 27079
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5583292841911316,
+      "learning_rate": 0.0003540138955144194,
+      "loss": 3.1336,
+      "step": 27080
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5447515249252319,
+      "learning_rate": 0.00035399875085537403,
+      "loss": 2.9675,
+      "step": 27081
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5473519563674927,
+      "learning_rate": 0.00035398360605410636,
+      "loss": 2.6685,
+      "step": 27082
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6127763986587524,
+      "learning_rate": 0.00035396846111065614,
+      "loss": 2.9615,
+      "step": 27083
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5656737685203552,
+      "learning_rate": 0.00035395331602506344,
+      "loss": 3.0083,
+      "step": 27084
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6185324192047119,
+      "learning_rate": 0.0003539381707973682,
+      "loss": 3.1667,
+      "step": 27085
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5483812689781189,
+      "learning_rate": 0.00035392302542761007,
+      "loss": 3.0815,
+      "step": 27086
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5831971764564514,
+      "learning_rate": 0.0003539078799158291,
+      "loss": 3.103,
+      "step": 27087
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5833414196968079,
+      "learning_rate": 0.0003538927342620651,
+      "loss": 2.9695,
+      "step": 27088
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.543148934841156,
+      "learning_rate": 0.0003538775884663581,
+      "loss": 3.0304,
+      "step": 27089
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6388452649116516,
+      "learning_rate": 0.0003538624425287479,
+      "loss": 3.0587,
+      "step": 27090
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.601694643497467,
+      "learning_rate": 0.0003538472964492743,
+      "loss": 2.9898,
+      "step": 27091
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6530476212501526,
+      "learning_rate": 0.00035383215022797735,
+      "loss": 3.0919,
+      "step": 27092
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5525394678115845,
+      "learning_rate": 0.0003538170038648969,
+      "loss": 3.1727,
+      "step": 27093
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.562701404094696,
+      "learning_rate": 0.0003538018573600728,
+      "loss": 3.1354,
+      "step": 27094
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6188597679138184,
+      "learning_rate": 0.0003537867107135449,
+      "loss": 3.1935,
+      "step": 27095
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5348824858665466,
+      "learning_rate": 0.0003537715639253533,
+      "loss": 2.9653,
+      "step": 27096
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5821238160133362,
+      "learning_rate": 0.00035375641699553765,
+      "loss": 2.973,
+      "step": 27097
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5549263954162598,
+      "learning_rate": 0.0003537412699241379,
+      "loss": 3.0837,
+      "step": 27098
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5607396960258484,
+      "learning_rate": 0.0003537261227111941,
+      "loss": 3.1077,
+      "step": 27099
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5585238337516785,
+      "learning_rate": 0.00035371097535674597,
+      "loss": 2.7311,
+      "step": 27100
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5680395364761353,
+      "learning_rate": 0.00035369582786083344,
+      "loss": 2.9498,
+      "step": 27101
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.582237184047699,
+      "learning_rate": 0.00035368068022349644,
+      "loss": 2.9487,
+      "step": 27102
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6273513436317444,
+      "learning_rate": 0.00035366553244477487,
+      "loss": 3.2227,
+      "step": 27103
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5687019228935242,
+      "learning_rate": 0.0003536503845247087,
+      "loss": 3.0682,
+      "step": 27104
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5843102931976318,
+      "learning_rate": 0.0003536352364633376,
+      "loss": 3.2752,
+      "step": 27105
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5721175670623779,
+      "learning_rate": 0.00035362008826070166,
+      "loss": 3.2501,
+      "step": 27106
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5818900465965271,
+      "learning_rate": 0.00035360493991684085,
+      "loss": 3.0691,
+      "step": 27107
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5466283559799194,
+      "learning_rate": 0.00035358979143179474,
+      "loss": 2.9073,
+      "step": 27108
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5692821741104126,
+      "learning_rate": 0.00035357464280560356,
+      "loss": 2.9878,
+      "step": 27109
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6057048439979553,
+      "learning_rate": 0.000353559494038307,
+      "loss": 3.1523,
+      "step": 27110
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7662555575370789,
+      "learning_rate": 0.0003535443451299452,
+      "loss": 3.0899,
+      "step": 27111
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6070552468299866,
+      "learning_rate": 0.00035352919608055777,
+      "loss": 3.0291,
+      "step": 27112
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5470950603485107,
+      "learning_rate": 0.0003535140468901847,
+      "loss": 2.9255,
+      "step": 27113
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5878978967666626,
+      "learning_rate": 0.00035349889755886604,
+      "loss": 2.828,
+      "step": 27114
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5929617881774902,
+      "learning_rate": 0.00035348374808664155,
+      "loss": 2.9915,
+      "step": 27115
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5788769125938416,
+      "learning_rate": 0.0003534685984735511,
+      "loss": 3.2226,
+      "step": 27116
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5542086362838745,
+      "learning_rate": 0.0003534534487196347,
+      "loss": 3.269,
+      "step": 27117
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5548458099365234,
+      "learning_rate": 0.0003534382988249323,
+      "loss": 3.044,
+      "step": 27118
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5794603228569031,
+      "learning_rate": 0.0003534231487894835,
+      "loss": 3.151,
+      "step": 27119
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6125290989875793,
+      "learning_rate": 0.0003534079986133285,
+      "loss": 3.0768,
+      "step": 27120
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5400205254554749,
+      "learning_rate": 0.00035339284829650716,
+      "loss": 3.0074,
+      "step": 27121
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5447748303413391,
+      "learning_rate": 0.0003533776978390593,
+      "loss": 3.031,
+      "step": 27122
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5757708549499512,
+      "learning_rate": 0.00035336254724102487,
+      "loss": 3.0801,
+      "step": 27123
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5604739189147949,
+      "learning_rate": 0.0003533473965024437,
+      "loss": 3.0351,
+      "step": 27124
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5703689455986023,
+      "learning_rate": 0.00035333224562335584,
+      "loss": 3.2484,
+      "step": 27125
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5888387560844421,
+      "learning_rate": 0.00035331709460380105,
+      "loss": 2.9558,
+      "step": 27126
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5701423287391663,
+      "learning_rate": 0.0003533019434438193,
+      "loss": 3.1372,
+      "step": 27127
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5726101398468018,
+      "learning_rate": 0.0003532867921434505,
+      "loss": 2.931,
+      "step": 27128
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5894249677658081,
+      "learning_rate": 0.00035327164070273457,
+      "loss": 3.1105,
+      "step": 27129
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5488255620002747,
+      "learning_rate": 0.0003532564891217113,
+      "loss": 3.0723,
+      "step": 27130
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5957520604133606,
+      "learning_rate": 0.00035324133740042077,
+      "loss": 3.1297,
+      "step": 27131
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5768669843673706,
+      "learning_rate": 0.00035322618553890276,
+      "loss": 2.9614,
+      "step": 27132
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5305699110031128,
+      "learning_rate": 0.00035321103353719736,
+      "loss": 2.8638,
+      "step": 27133
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5721878409385681,
+      "learning_rate": 0.00035319588139534413,
+      "loss": 3.203,
+      "step": 27134
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5690099596977234,
+      "learning_rate": 0.0003531807291133833,
+      "loss": 3.0033,
+      "step": 27135
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6488524079322815,
+      "learning_rate": 0.0003531655766913547,
+      "loss": 3.1853,
+      "step": 27136
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6288285255432129,
+      "learning_rate": 0.00035315042412929813,
+      "loss": 3.1391,
+      "step": 27137
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.586223840713501,
+      "learning_rate": 0.00035313527142725364,
+      "loss": 2.9698,
+      "step": 27138
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5799707174301147,
+      "learning_rate": 0.000353120118585261,
+      "loss": 3.1249,
+      "step": 27139
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.555907666683197,
+      "learning_rate": 0.0003531049656033602,
+      "loss": 3.1875,
+      "step": 27140
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5725418329238892,
+      "learning_rate": 0.00035308981248159113,
+      "loss": 3.1464,
+      "step": 27141
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.542593240737915,
+      "learning_rate": 0.00035307465921999374,
+      "loss": 3.1674,
+      "step": 27142
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5610252022743225,
+      "learning_rate": 0.0003530595058186079,
+      "loss": 3.0766,
+      "step": 27143
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6028214693069458,
+      "learning_rate": 0.0003530443522774735,
+      "loss": 3.3676,
+      "step": 27144
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5492804646492004,
+      "learning_rate": 0.0003530291985966305,
+      "loss": 3.0786,
+      "step": 27145
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5824100375175476,
+      "learning_rate": 0.0003530140447761188,
+      "loss": 3.1429,
+      "step": 27146
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5860824584960938,
+      "learning_rate": 0.0003529988908159784,
+      "loss": 3.1449,
+      "step": 27147
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6098974347114563,
+      "learning_rate": 0.0003529837367162491,
+      "loss": 3.0493,
+      "step": 27148
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5788512825965881,
+      "learning_rate": 0.0003529685824769707,
+      "loss": 2.9447,
+      "step": 27149
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5958915948867798,
+      "learning_rate": 0.0003529534280981834,
+      "loss": 3.1485,
+      "step": 27150
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.552696704864502,
+      "learning_rate": 0.0003529382735799269,
+      "loss": 3.0101,
+      "step": 27151
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.571831226348877,
+      "learning_rate": 0.00035292311892224114,
+      "loss": 2.9421,
+      "step": 27152
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5760900378227234,
+      "learning_rate": 0.0003529079641251661,
+      "loss": 3.0002,
+      "step": 27153
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6153770089149475,
+      "learning_rate": 0.0003528928091887417,
+      "loss": 2.9689,
+      "step": 27154
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5628465414047241,
+      "learning_rate": 0.00035287765411300777,
+      "loss": 2.8312,
+      "step": 27155
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5893851518630981,
+      "learning_rate": 0.00035286249889800425,
+      "loss": 2.9367,
+      "step": 27156
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5659940242767334,
+      "learning_rate": 0.00035284734354377117,
+      "loss": 3.078,
+      "step": 27157
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5788155198097229,
+      "learning_rate": 0.0003528321880503483,
+      "loss": 2.9413,
+      "step": 27158
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6268181204795837,
+      "learning_rate": 0.00035281703241777566,
+      "loss": 3.0774,
+      "step": 27159
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5367437601089478,
+      "learning_rate": 0.000352801876646093,
+      "loss": 3.0378,
+      "step": 27160
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5380226969718933,
+      "learning_rate": 0.00035278672073534045,
+      "loss": 2.9433,
+      "step": 27161
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5570831894874573,
+      "learning_rate": 0.00035277156468555785,
+      "loss": 3.0755,
+      "step": 27162
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5446562767028809,
+      "learning_rate": 0.00035275640849678514,
+      "loss": 2.9051,
+      "step": 27163
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5816878080368042,
+      "learning_rate": 0.0003527412521690621,
+      "loss": 3.0586,
+      "step": 27164
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6048847436904907,
+      "learning_rate": 0.0003527260957024288,
+      "loss": 3.2571,
+      "step": 27165
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5430023670196533,
+      "learning_rate": 0.0003527109390969251,
+      "loss": 3.1818,
+      "step": 27166
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5993807911872864,
+      "learning_rate": 0.00035269578235259095,
+      "loss": 3.0313,
+      "step": 27167
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.638937771320343,
+      "learning_rate": 0.0003526806254694662,
+      "loss": 3.304,
+      "step": 27168
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5941781997680664,
+      "learning_rate": 0.000352665468447591,
+      "loss": 2.8478,
+      "step": 27169
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5695825815200806,
+      "learning_rate": 0.0003526503112870049,
+      "loss": 3.1401,
+      "step": 27170
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.563214898109436,
+      "learning_rate": 0.000352635153987748,
+      "loss": 3.2235,
+      "step": 27171
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5385993719100952,
+      "learning_rate": 0.00035261999654986035,
+      "loss": 3.0607,
+      "step": 27172
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.612762987613678,
+      "learning_rate": 0.0003526048389733818,
+      "loss": 3.0266,
+      "step": 27173
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5819412469863892,
+      "learning_rate": 0.0003525896812583521,
+      "loss": 2.93,
+      "step": 27174
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5829300880432129,
+      "learning_rate": 0.0003525745234048112,
+      "loss": 2.9441,
+      "step": 27175
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5720100998878479,
+      "learning_rate": 0.00035255936541279937,
+      "loss": 3.02,
+      "step": 27176
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5454930663108826,
+      "learning_rate": 0.0003525442072823562,
+      "loss": 2.9635,
+      "step": 27177
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5697073936462402,
+      "learning_rate": 0.0003525290490135217,
+      "loss": 3.1289,
+      "step": 27178
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5639901161193848,
+      "learning_rate": 0.00035251389060633574,
+      "loss": 3.077,
+      "step": 27179
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5947508811950684,
+      "learning_rate": 0.0003524987320608383,
+      "loss": 3.0138,
+      "step": 27180
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.555055558681488,
+      "learning_rate": 0.00035248357337706935,
+      "loss": 3.1569,
+      "step": 27181
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5873993039131165,
+      "learning_rate": 0.00035246841455506875,
+      "loss": 3.0096,
+      "step": 27182
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5661149621009827,
+      "learning_rate": 0.00035245325559487645,
+      "loss": 2.9492,
+      "step": 27183
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.55522221326828,
+      "learning_rate": 0.0003524380964965324,
+      "loss": 3.1046,
+      "step": 27184
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5756954550743103,
+      "learning_rate": 0.00035242293726007643,
+      "loss": 3.1467,
+      "step": 27185
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5577954649925232,
+      "learning_rate": 0.00035240777788554857,
+      "loss": 3.1245,
+      "step": 27186
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5636711120605469,
+      "learning_rate": 0.00035239261837298876,
+      "loss": 3.1186,
+      "step": 27187
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5825818181037903,
+      "learning_rate": 0.0003523774587224368,
+      "loss": 3.1649,
+      "step": 27188
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.570296585559845,
+      "learning_rate": 0.0003523622989339327,
+      "loss": 2.9089,
+      "step": 27189
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5361173152923584,
+      "learning_rate": 0.0003523471390075164,
+      "loss": 2.944,
+      "step": 27190
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5616334080696106,
+      "learning_rate": 0.0003523319789432278,
+      "loss": 3.2198,
+      "step": 27191
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6106418371200562,
+      "learning_rate": 0.00035231681874110685,
+      "loss": 3.0541,
+      "step": 27192
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6105589270591736,
+      "learning_rate": 0.00035230165840119344,
+      "loss": 3.1066,
+      "step": 27193
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5929547548294067,
+      "learning_rate": 0.0003522864979235276,
+      "loss": 2.9546,
+      "step": 27194
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5807821154594421,
+      "learning_rate": 0.0003522713373081491,
+      "loss": 3.1256,
+      "step": 27195
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5757489204406738,
+      "learning_rate": 0.00035225617655509797,
+      "loss": 3.0397,
+      "step": 27196
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5699480772018433,
+      "learning_rate": 0.0003522410156644142,
+      "loss": 3.3366,
+      "step": 27197
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6023076176643372,
+      "learning_rate": 0.00035222585463613764,
+      "loss": 3.103,
+      "step": 27198
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5710323452949524,
+      "learning_rate": 0.0003522106934703082,
+      "loss": 3.1119,
+      "step": 27199
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5762554407119751,
+      "learning_rate": 0.0003521955321669658,
+      "loss": 3.0356,
+      "step": 27200
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5505509972572327,
+      "learning_rate": 0.00035218037072615047,
+      "loss": 3.0826,
+      "step": 27201
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5784319639205933,
+      "learning_rate": 0.00035216520914790205,
+      "loss": 3.0583,
+      "step": 27202
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5438815951347351,
+      "learning_rate": 0.00035215004743226056,
+      "loss": 2.8998,
+      "step": 27203
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5637693405151367,
+      "learning_rate": 0.0003521348855792658,
+      "loss": 3.0585,
+      "step": 27204
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5552592277526855,
+      "learning_rate": 0.00035211972358895776,
+      "loss": 2.8598,
+      "step": 27205
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6333317756652832,
+      "learning_rate": 0.0003521045614613765,
+      "loss": 3.0529,
+      "step": 27206
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6301619410514832,
+      "learning_rate": 0.0003520893991965618,
+      "loss": 2.963,
+      "step": 27207
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5590476989746094,
+      "learning_rate": 0.0003520742367945537,
+      "loss": 2.9396,
+      "step": 27208
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5794091820716858,
+      "learning_rate": 0.00035205907425539214,
+      "loss": 3.0028,
+      "step": 27209
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5639303922653198,
+      "learning_rate": 0.00035204391157911684,
+      "loss": 3.0661,
+      "step": 27210
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5517282485961914,
+      "learning_rate": 0.0003520287487657679,
+      "loss": 2.9745,
+      "step": 27211
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6066323518753052,
+      "learning_rate": 0.0003520135858153854,
+      "loss": 3.1238,
+      "step": 27212
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5425803661346436,
+      "learning_rate": 0.000351998422728009,
+      "loss": 3.1991,
+      "step": 27213
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5908896923065186,
+      "learning_rate": 0.0003519832595036788,
+      "loss": 3.1234,
+      "step": 27214
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5824942588806152,
+      "learning_rate": 0.0003519680961424347,
+      "loss": 3.0476,
+      "step": 27215
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5757591724395752,
+      "learning_rate": 0.00035195293264431664,
+      "loss": 3.0747,
+      "step": 27216
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5482543110847473,
+      "learning_rate": 0.00035193776900936457,
+      "loss": 3.2661,
+      "step": 27217
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6184133887290955,
+      "learning_rate": 0.00035192260523761833,
+      "loss": 3.3704,
+      "step": 27218
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5994088053703308,
+      "learning_rate": 0.000351907441329118,
+      "loss": 2.8227,
+      "step": 27219
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5714132785797119,
+      "learning_rate": 0.0003518922772839035,
+      "loss": 3.102,
+      "step": 27220
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5737739205360413,
+      "learning_rate": 0.00035187711310201463,
+      "loss": 3.0617,
+      "step": 27221
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5780137777328491,
+      "learning_rate": 0.0003518619487834915,
+      "loss": 3.0814,
+      "step": 27222
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5979353785514832,
+      "learning_rate": 0.00035184678432837396,
+      "loss": 3.1507,
+      "step": 27223
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5943918824195862,
+      "learning_rate": 0.000351831619736702,
+      "loss": 2.9943,
+      "step": 27224
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5988668203353882,
+      "learning_rate": 0.0003518164550085155,
+      "loss": 3.1677,
+      "step": 27225
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5408164262771606,
+      "learning_rate": 0.00035180129014385437,
+      "loss": 3.1091,
+      "step": 27226
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5693389177322388,
+      "learning_rate": 0.0003517861251427587,
+      "loss": 3.2419,
+      "step": 27227
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5514744520187378,
+      "learning_rate": 0.00035177096000526837,
+      "loss": 3.0412,
+      "step": 27228
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5758994221687317,
+      "learning_rate": 0.0003517557947314232,
+      "loss": 3.0927,
+      "step": 27229
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5742641687393188,
+      "learning_rate": 0.00035174062932126325,
+      "loss": 3.0443,
+      "step": 27230
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.587285041809082,
+      "learning_rate": 0.00035172546377482843,
+      "loss": 3.2015,
+      "step": 27231
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5484117865562439,
+      "learning_rate": 0.0003517102980921587,
+      "loss": 3.2877,
+      "step": 27232
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5856320261955261,
+      "learning_rate": 0.000351695132273294,
+      "loss": 3.1225,
+      "step": 27233
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5759357810020447,
+      "learning_rate": 0.00035167996631827435,
+      "loss": 3.2482,
+      "step": 27234
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.556387722492218,
+      "learning_rate": 0.0003516648002271396,
+      "loss": 2.9464,
+      "step": 27235
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5765631198883057,
+      "learning_rate": 0.00035164963399992964,
+      "loss": 3.0678,
+      "step": 27236
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5664901733398438,
+      "learning_rate": 0.0003516344676366844,
+      "loss": 3.1126,
+      "step": 27237
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5461036562919617,
+      "learning_rate": 0.00035161930113744404,
+      "loss": 3.123,
+      "step": 27238
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5736780762672424,
+      "learning_rate": 0.00035160413450224836,
+      "loss": 3.1065,
+      "step": 27239
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5623251795768738,
+      "learning_rate": 0.00035158896773113737,
+      "loss": 3.1294,
+      "step": 27240
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5517589449882507,
+      "learning_rate": 0.00035157380082415085,
+      "loss": 3.1211,
+      "step": 27241
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6113983988761902,
+      "learning_rate": 0.000351558633781329,
+      "loss": 2.9765,
+      "step": 27242
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5473376512527466,
+      "learning_rate": 0.00035154346660271144,
+      "loss": 3.0213,
+      "step": 27243
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5686004161834717,
+      "learning_rate": 0.00035152829928833846,
+      "loss": 3.1543,
+      "step": 27244
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5545783042907715,
+      "learning_rate": 0.0003515131318382498,
+      "loss": 3.1051,
+      "step": 27245
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5631744861602783,
+      "learning_rate": 0.0003514979642524855,
+      "loss": 3.134,
+      "step": 27246
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5703157186508179,
+      "learning_rate": 0.0003514827965310854,
+      "loss": 3.1174,
+      "step": 27247
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5550886392593384,
+      "learning_rate": 0.0003514676286740896,
+      "loss": 3.1043,
+      "step": 27248
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5608128309249878,
+      "learning_rate": 0.00035145246068153804,
+      "loss": 3.0714,
+      "step": 27249
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6270610690116882,
+      "learning_rate": 0.00035143729255347045,
+      "loss": 3.0144,
+      "step": 27250
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6169301867485046,
+      "learning_rate": 0.000351422124289927,
+      "loss": 2.9064,
+      "step": 27251
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6073354482650757,
+      "learning_rate": 0.0003514069558909475,
+      "loss": 3.1812,
+      "step": 27252
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5473766326904297,
+      "learning_rate": 0.00035139178735657214,
+      "loss": 3.0509,
+      "step": 27253
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5443681478500366,
+      "learning_rate": 0.00035137661868684056,
+      "loss": 3.1905,
+      "step": 27254
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5778075456619263,
+      "learning_rate": 0.00035136144988179286,
+      "loss": 2.8105,
+      "step": 27255
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.581585168838501,
+      "learning_rate": 0.00035134628094146903,
+      "loss": 3.0035,
+      "step": 27256
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6521691679954529,
+      "learning_rate": 0.0003513311118659089,
+      "loss": 2.892,
+      "step": 27257
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6202303171157837,
+      "learning_rate": 0.0003513159426551526,
+      "loss": 2.8636,
+      "step": 27258
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.573544979095459,
+      "learning_rate": 0.00035130077330923997,
+      "loss": 3.2428,
+      "step": 27259
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6818689107894897,
+      "learning_rate": 0.00035128560382821097,
+      "loss": 2.8362,
+      "step": 27260
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5987098813056946,
+      "learning_rate": 0.0003512704342121055,
+      "loss": 2.9345,
+      "step": 27261
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6563383936882019,
+      "learning_rate": 0.0003512552644609636,
+      "loss": 2.9929,
+      "step": 27262
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5845065712928772,
+      "learning_rate": 0.00035124009457482524,
+      "loss": 3.0538,
+      "step": 27263
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5674929618835449,
+      "learning_rate": 0.0003512249245537303,
+      "loss": 3.1348,
+      "step": 27264
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5974243879318237,
+      "learning_rate": 0.00035120975439771883,
+      "loss": 3.188,
+      "step": 27265
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5698199272155762,
+      "learning_rate": 0.0003511945841068306,
+      "loss": 2.9841,
+      "step": 27266
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5573409795761108,
+      "learning_rate": 0.0003511794136811058,
+      "loss": 2.9835,
+      "step": 27267
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5393723249435425,
+      "learning_rate": 0.00035116424312058416,
+      "loss": 3.1279,
+      "step": 27268
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5727754235267639,
+      "learning_rate": 0.0003511490724253058,
+      "loss": 3.0733,
+      "step": 27269
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5624639391899109,
+      "learning_rate": 0.00035113390159531067,
+      "loss": 3.0563,
+      "step": 27270
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5651065707206726,
+      "learning_rate": 0.0003511187306306387,
+      "loss": 2.9617,
+      "step": 27271
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5551247000694275,
+      "learning_rate": 0.00035110355953132976,
+      "loss": 3.096,
+      "step": 27272
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5362761616706848,
+      "learning_rate": 0.0003510883882974239,
+      "loss": 3.133,
+      "step": 27273
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6459551453590393,
+      "learning_rate": 0.00035107321692896105,
+      "loss": 2.9986,
+      "step": 27274
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5925512313842773,
+      "learning_rate": 0.00035105804542598124,
+      "loss": 3.1903,
+      "step": 27275
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5428948998451233,
+      "learning_rate": 0.0003510428737885243,
+      "loss": 2.8865,
+      "step": 27276
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5987960696220398,
+      "learning_rate": 0.0003510277020166303,
+      "loss": 3.1261,
+      "step": 27277
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5686396956443787,
+      "learning_rate": 0.00035101253011033914,
+      "loss": 2.9944,
+      "step": 27278
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5981069207191467,
+      "learning_rate": 0.00035099735806969073,
+      "loss": 3.1263,
+      "step": 27279
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5610902309417725,
+      "learning_rate": 0.0003509821858947252,
+      "loss": 2.8438,
+      "step": 27280
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5656661987304688,
+      "learning_rate": 0.0003509670135854823,
+      "loss": 3.1531,
+      "step": 27281
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5572855472564697,
+      "learning_rate": 0.0003509518411420022,
+      "loss": 3.1005,
+      "step": 27282
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5681395530700684,
+      "learning_rate": 0.0003509366685643246,
+      "loss": 3.1331,
+      "step": 27283
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.603127121925354,
+      "learning_rate": 0.0003509214958524897,
+      "loss": 3.265,
+      "step": 27284
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5976519584655762,
+      "learning_rate": 0.0003509063230065374,
+      "loss": 2.9344,
+      "step": 27285
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5389209985733032,
+      "learning_rate": 0.0003508911500265077,
+      "loss": 3.0831,
+      "step": 27286
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5420101881027222,
+      "learning_rate": 0.0003508759769124405,
+      "loss": 3.0395,
+      "step": 27287
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5472182035446167,
+      "learning_rate": 0.00035086080366437554,
+      "loss": 2.9825,
+      "step": 27288
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5901409983634949,
+      "learning_rate": 0.0003508456302823533,
+      "loss": 3.0206,
+      "step": 27289
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5679577589035034,
+      "learning_rate": 0.0003508304567664133,
+      "loss": 3.061,
+      "step": 27290
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5943844318389893,
+      "learning_rate": 0.0003508152831165957,
+      "loss": 2.9039,
+      "step": 27291
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.619647741317749,
+      "learning_rate": 0.0003508001093329404,
+      "loss": 3.2114,
+      "step": 27292
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5737244486808777,
+      "learning_rate": 0.0003507849354154874,
+      "loss": 2.9268,
+      "step": 27293
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.574532687664032,
+      "learning_rate": 0.00035076976136427665,
+      "loss": 3.3231,
+      "step": 27294
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6363462805747986,
+      "learning_rate": 0.00035075458717934816,
+      "loss": 3.227,
+      "step": 27295
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5762098431587219,
+      "learning_rate": 0.00035073941286074183,
+      "loss": 3.2721,
+      "step": 27296
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6113587021827698,
+      "learning_rate": 0.0003507242384084977,
+      "loss": 2.9124,
+      "step": 27297
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5779880881309509,
+      "learning_rate": 0.00035070906382265554,
+      "loss": 3.2056,
+      "step": 27298
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5307174921035767,
+      "learning_rate": 0.00035069388910325555,
+      "loss": 2.8949,
+      "step": 27299
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5478222966194153,
+      "learning_rate": 0.0003506787142503377,
+      "loss": 3.1332,
+      "step": 27300
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5192190408706665,
+      "learning_rate": 0.0003506635392639418,
+      "loss": 3.0535,
+      "step": 27301
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5842856168746948,
+      "learning_rate": 0.00035064836414410783,
+      "loss": 2.8508,
+      "step": 27302
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5742753148078918,
+      "learning_rate": 0.00035063318889087586,
+      "loss": 3.1244,
+      "step": 27303
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.576598048210144,
+      "learning_rate": 0.0003506180135042859,
+      "loss": 3.0379,
+      "step": 27304
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5685365796089172,
+      "learning_rate": 0.0003506028379843777,
+      "loss": 3.1901,
+      "step": 27305
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6361297965049744,
+      "learning_rate": 0.0003505876623311914,
+      "loss": 3.3425,
+      "step": 27306
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6099052429199219,
+      "learning_rate": 0.00035057248654476694,
+      "loss": 3.1613,
+      "step": 27307
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5758441686630249,
+      "learning_rate": 0.0003505573106251443,
+      "loss": 3.2276,
+      "step": 27308
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5512416958808899,
+      "learning_rate": 0.00035054213457236347,
+      "loss": 3.0823,
+      "step": 27309
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5952394008636475,
+      "learning_rate": 0.0003505269583864643,
+      "loss": 3.3958,
+      "step": 27310
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.569972813129425,
+      "learning_rate": 0.00035051178206748696,
+      "loss": 3.049,
+      "step": 27311
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5571011304855347,
+      "learning_rate": 0.00035049660561547124,
+      "loss": 3.2249,
+      "step": 27312
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6160045862197876,
+      "learning_rate": 0.0003504814290304572,
+      "loss": 3.3971,
+      "step": 27313
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6157644987106323,
+      "learning_rate": 0.0003504662523124848,
+      "loss": 2.9801,
+      "step": 27314
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5774762034416199,
+      "learning_rate": 0.00035045107546159395,
+      "loss": 3.0074,
+      "step": 27315
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6000568270683289,
+      "learning_rate": 0.00035043589847782465,
+      "loss": 2.9885,
+      "step": 27316
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5656048655509949,
+      "learning_rate": 0.00035042072136121696,
+      "loss": 3.1387,
+      "step": 27317
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.613006055355072,
+      "learning_rate": 0.0003504055441118108,
+      "loss": 3.0847,
+      "step": 27318
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5949397683143616,
+      "learning_rate": 0.00035039036672964607,
+      "loss": 3.1484,
+      "step": 27319
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5421934127807617,
+      "learning_rate": 0.00035037518921476283,
+      "loss": 3.214,
+      "step": 27320
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.574578046798706,
+      "learning_rate": 0.0003503600115672011,
+      "loss": 3.0972,
+      "step": 27321
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5909102559089661,
+      "learning_rate": 0.0003503448337870008,
+      "loss": 3.0916,
+      "step": 27322
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5599302649497986,
+      "learning_rate": 0.00035032965587420187,
+      "loss": 2.9846,
+      "step": 27323
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.558342695236206,
+      "learning_rate": 0.0003503144778288442,
+      "loss": 3.1657,
+      "step": 27324
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6576507091522217,
+      "learning_rate": 0.000350299299650968,
+      "loss": 2.9349,
+      "step": 27325
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5906968712806702,
+      "learning_rate": 0.00035028412134061314,
+      "loss": 2.8856,
+      "step": 27326
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.53647381067276,
+      "learning_rate": 0.0003502689428978196,
+      "loss": 3.2723,
+      "step": 27327
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8482022285461426,
+      "learning_rate": 0.00035025376432262724,
+      "loss": 3.0158,
+      "step": 27328
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.603201687335968,
+      "learning_rate": 0.0003502385856150762,
+      "loss": 3.1107,
+      "step": 27329
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.551382303237915,
+      "learning_rate": 0.0003502234067752064,
+      "loss": 3.1786,
+      "step": 27330
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5987643003463745,
+      "learning_rate": 0.0003502082278030577,
+      "loss": 3.0718,
+      "step": 27331
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5653188228607178,
+      "learning_rate": 0.0003501930486986703,
+      "loss": 2.962,
+      "step": 27332
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5514883399009705,
+      "learning_rate": 0.0003501778694620841,
+      "loss": 3.1602,
+      "step": 27333
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.571922242641449,
+      "learning_rate": 0.0003501626900933389,
+      "loss": 3.0968,
+      "step": 27334
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5911169648170471,
+      "learning_rate": 0.0003501475105924749,
+      "loss": 3.297,
+      "step": 27335
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5552850365638733,
+      "learning_rate": 0.0003501323309595321,
+      "loss": 3.193,
+      "step": 27336
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.582349419593811,
+      "learning_rate": 0.00035011715119455036,
+      "loss": 2.9565,
+      "step": 27337
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5713590383529663,
+      "learning_rate": 0.0003501019712975696,
+      "loss": 3.1794,
+      "step": 27338
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5552675127983093,
+      "learning_rate": 0.00035008679126862993,
+      "loss": 2.889,
+      "step": 27339
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.548956573009491,
+      "learning_rate": 0.00035007161110777135,
+      "loss": 3.2242,
+      "step": 27340
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5811518430709839,
+      "learning_rate": 0.0003500564308150337,
+      "loss": 3.1934,
+      "step": 27341
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5775765776634216,
+      "learning_rate": 0.0003500412503904571,
+      "loss": 3.0469,
+      "step": 27342
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5719749927520752,
+      "learning_rate": 0.00035002606983408147,
+      "loss": 2.9931,
+      "step": 27343
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5796716809272766,
+      "learning_rate": 0.00035001088914594675,
+      "loss": 3.1023,
+      "step": 27344
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5906006693840027,
+      "learning_rate": 0.000349995708326093,
+      "loss": 3.2135,
+      "step": 27345
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.58286052942276,
+      "learning_rate": 0.0003499805273745602,
+      "loss": 3.2402,
+      "step": 27346
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5934612154960632,
+      "learning_rate": 0.00034996534629138824,
+      "loss": 3.0281,
+      "step": 27347
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5473476648330688,
+      "learning_rate": 0.00034995016507661733,
+      "loss": 3.0496,
+      "step": 27348
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5469577312469482,
+      "learning_rate": 0.0003499349837302871,
+      "loss": 3.2524,
+      "step": 27349
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6002890467643738,
+      "learning_rate": 0.0003499198022524379,
+      "loss": 3.1881,
+      "step": 27350
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.564316987991333,
+      "learning_rate": 0.00034990462064310945,
+      "loss": 3.2093,
+      "step": 27351
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5971184968948364,
+      "learning_rate": 0.00034988943890234186,
+      "loss": 2.937,
+      "step": 27352
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5803214311599731,
+      "learning_rate": 0.0003498742570301751,
+      "loss": 2.9276,
+      "step": 27353
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5375666618347168,
+      "learning_rate": 0.00034985907502664917,
+      "loss": 3.1482,
+      "step": 27354
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.60368412733078,
+      "learning_rate": 0.00034984389289180397,
+      "loss": 3.1235,
+      "step": 27355
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6041290760040283,
+      "learning_rate": 0.00034982871062567956,
+      "loss": 3.1815,
+      "step": 27356
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5556178092956543,
+      "learning_rate": 0.00034981352822831594,
+      "loss": 3.0318,
+      "step": 27357
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5745914578437805,
+      "learning_rate": 0.000349798345699753,
+      "loss": 2.7708,
+      "step": 27358
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5874989628791809,
+      "learning_rate": 0.00034978316304003096,
+      "loss": 2.9932,
+      "step": 27359
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5912497043609619,
+      "learning_rate": 0.0003497679802491895,
+      "loss": 2.8349,
+      "step": 27360
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5921602845191956,
+      "learning_rate": 0.0003497527973272688,
+      "loss": 2.8376,
+      "step": 27361
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5861170887947083,
+      "learning_rate": 0.0003497376142743089,
+      "loss": 2.9823,
+      "step": 27362
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6155903339385986,
+      "learning_rate": 0.00034972243109034957,
+      "loss": 3.0171,
+      "step": 27363
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5605053901672363,
+      "learning_rate": 0.00034970724777543097,
+      "loss": 3.1852,
+      "step": 27364
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5607673525810242,
+      "learning_rate": 0.0003496920643295931,
+      "loss": 3.17,
+      "step": 27365
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5722825527191162,
+      "learning_rate": 0.00034967688075287584,
+      "loss": 3.1404,
+      "step": 27366
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5873673558235168,
+      "learning_rate": 0.0003496616970453192,
+      "loss": 3.0477,
+      "step": 27367
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6069886088371277,
+      "learning_rate": 0.00034964651320696326,
+      "loss": 2.9818,
+      "step": 27368
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5635610222816467,
+      "learning_rate": 0.00034963132923784796,
+      "loss": 2.8535,
+      "step": 27369
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5699170231819153,
+      "learning_rate": 0.0003496161451380133,
+      "loss": 3.075,
+      "step": 27370
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5741016268730164,
+      "learning_rate": 0.00034960096090749923,
+      "loss": 3.0363,
+      "step": 27371
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5453298687934875,
+      "learning_rate": 0.00034958577654634575,
+      "loss": 2.9578,
+      "step": 27372
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.577882707118988,
+      "learning_rate": 0.000349570592054593,
+      "loss": 2.9079,
+      "step": 27373
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5363011360168457,
+      "learning_rate": 0.0003495554074322807,
+      "loss": 3.054,
+      "step": 27374
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5371752381324768,
+      "learning_rate": 0.000349540222679449,
+      "loss": 3.0363,
+      "step": 27375
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5992576479911804,
+      "learning_rate": 0.000349525037796138,
+      "loss": 3.2143,
+      "step": 27376
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5904133319854736,
+      "learning_rate": 0.00034950985278238753,
+      "loss": 3.2084,
+      "step": 27377
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6310132145881653,
+      "learning_rate": 0.00034949466763823766,
+      "loss": 2.872,
+      "step": 27378
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5347175002098083,
+      "learning_rate": 0.0003494794823637283,
+      "loss": 2.8966,
+      "step": 27379
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5774542689323425,
+      "learning_rate": 0.00034946429695889954,
+      "loss": 3.0197,
+      "step": 27380
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6394713521003723,
+      "learning_rate": 0.00034944911142379136,
+      "loss": 3.0879,
+      "step": 27381
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5754245519638062,
+      "learning_rate": 0.0003494339257584437,
+      "loss": 3.056,
+      "step": 27382
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5918018817901611,
+      "learning_rate": 0.0003494187399628966,
+      "loss": 3.0525,
+      "step": 27383
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5479419231414795,
+      "learning_rate": 0.0003494035540371901,
+      "loss": 3.0767,
+      "step": 27384
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5651912093162537,
+      "learning_rate": 0.0003493883679813641,
+      "loss": 3.0674,
+      "step": 27385
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.610517680644989,
+      "learning_rate": 0.0003493731817954586,
+      "loss": 2.9939,
+      "step": 27386
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5542714595794678,
+      "learning_rate": 0.0003493579954795137,
+      "loss": 2.9815,
+      "step": 27387
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5359601974487305,
+      "learning_rate": 0.0003493428090335694,
+      "loss": 3.1892,
+      "step": 27388
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5919166207313538,
+      "learning_rate": 0.00034932762245766557,
+      "loss": 3.1598,
+      "step": 27389
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5871334075927734,
+      "learning_rate": 0.0003493124357518422,
+      "loss": 3.1153,
+      "step": 27390
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5805013179779053,
+      "learning_rate": 0.0003492972489161395,
+      "loss": 3.0343,
+      "step": 27391
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6670655012130737,
+      "learning_rate": 0.0003492820619505973,
+      "loss": 3.1204,
+      "step": 27392
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5655217170715332,
+      "learning_rate": 0.00034926687485525555,
+      "loss": 2.9918,
+      "step": 27393
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5714253187179565,
+      "learning_rate": 0.00034925168763015444,
+      "loss": 3.0133,
+      "step": 27394
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6281166076660156,
+      "learning_rate": 0.0003492365002753338,
+      "loss": 3.1404,
+      "step": 27395
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6088977456092834,
+      "learning_rate": 0.00034922131279083364,
+      "loss": 2.8393,
+      "step": 27396
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6364918351173401,
+      "learning_rate": 0.0003492061251766941,
+      "loss": 3.0798,
+      "step": 27397
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5799409747123718,
+      "learning_rate": 0.0003491909374329551,
+      "loss": 2.9797,
+      "step": 27398
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5414788126945496,
+      "learning_rate": 0.00034917574955965666,
+      "loss": 3.2742,
+      "step": 27399
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5648382902145386,
+      "learning_rate": 0.00034916056155683873,
+      "loss": 3.2028,
+      "step": 27400
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5464253425598145,
+      "learning_rate": 0.00034914537342454127,
+      "loss": 3.0764,
+      "step": 27401
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5706402063369751,
+      "learning_rate": 0.00034913018516280443,
+      "loss": 3.0302,
+      "step": 27402
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5839251279830933,
+      "learning_rate": 0.0003491149967716681,
+      "loss": 3.1927,
+      "step": 27403
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5891869068145752,
+      "learning_rate": 0.0003490998082511724,
+      "loss": 3.1018,
+      "step": 27404
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6145084500312805,
+      "learning_rate": 0.00034908461960135714,
+      "loss": 3.0066,
+      "step": 27405
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6099284887313843,
+      "learning_rate": 0.0003490694308222625,
+      "loss": 2.9118,
+      "step": 27406
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5804110765457153,
+      "learning_rate": 0.0003490542419139284,
+      "loss": 3.0253,
+      "step": 27407
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6241898536682129,
+      "learning_rate": 0.0003490390528763949,
+      "loss": 3.0083,
+      "step": 27408
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5708284974098206,
+      "learning_rate": 0.00034902386370970197,
+      "loss": 3.1177,
+      "step": 27409
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5617969036102295,
+      "learning_rate": 0.00034900867441388963,
+      "loss": 3.1507,
+      "step": 27410
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5512824058532715,
+      "learning_rate": 0.0003489934849889978,
+      "loss": 3.2521,
+      "step": 27411
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5705128908157349,
+      "learning_rate": 0.00034897829543506666,
+      "loss": 2.9359,
+      "step": 27412
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5484633445739746,
+      "learning_rate": 0.00034896310575213604,
+      "loss": 3.2642,
+      "step": 27413
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6320682764053345,
+      "learning_rate": 0.00034894791594024607,
+      "loss": 3.0843,
+      "step": 27414
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5714908242225647,
+      "learning_rate": 0.0003489327259994367,
+      "loss": 3.2425,
+      "step": 27415
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5834794640541077,
+      "learning_rate": 0.0003489175359297479,
+      "loss": 3.0362,
+      "step": 27416
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.580422580242157,
+      "learning_rate": 0.0003489023457312198,
+      "loss": 3.0221,
+      "step": 27417
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5791642665863037,
+      "learning_rate": 0.00034888715540389226,
+      "loss": 3.1245,
+      "step": 27418
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5940735936164856,
+      "learning_rate": 0.00034887196494780536,
+      "loss": 3.1225,
+      "step": 27419
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5686174035072327,
+      "learning_rate": 0.00034885677436299916,
+      "loss": 2.943,
+      "step": 27420
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5724006295204163,
+      "learning_rate": 0.00034884158364951356,
+      "loss": 3.0081,
+      "step": 27421
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.646140456199646,
+      "learning_rate": 0.0003488263928073887,
+      "loss": 3.0739,
+      "step": 27422
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6348581314086914,
+      "learning_rate": 0.0003488112018366645,
+      "loss": 3.2155,
+      "step": 27423
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5576120018959045,
+      "learning_rate": 0.000348796010737381,
+      "loss": 3.326,
+      "step": 27424
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5346035361289978,
+      "learning_rate": 0.0003487808195095782,
+      "loss": 2.9394,
+      "step": 27425
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6104670763015747,
+      "learning_rate": 0.000348765628153296,
+      "loss": 3.2695,
+      "step": 27426
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5855081081390381,
+      "learning_rate": 0.0003487504366685747,
+      "loss": 3.1842,
+      "step": 27427
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.604956865310669,
+      "learning_rate": 0.00034873524505545404,
+      "loss": 2.8405,
+      "step": 27428
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5948700904846191,
+      "learning_rate": 0.0003487200533139741,
+      "loss": 3.2639,
+      "step": 27429
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5722827911376953,
+      "learning_rate": 0.00034870486144417486,
+      "loss": 3.0473,
+      "step": 27430
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5703803300857544,
+      "learning_rate": 0.0003486896694460965,
+      "loss": 3.1484,
+      "step": 27431
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5681774616241455,
+      "learning_rate": 0.00034867447731977885,
+      "loss": 2.9344,
+      "step": 27432
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5943686962127686,
+      "learning_rate": 0.00034865928506526206,
+      "loss": 3.2241,
+      "step": 27433
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5889038443565369,
+      "learning_rate": 0.00034864409268258606,
+      "loss": 3.0752,
+      "step": 27434
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.551078736782074,
+      "learning_rate": 0.0003486289001717909,
+      "loss": 3.0876,
+      "step": 27435
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5845509767532349,
+      "learning_rate": 0.00034861370753291654,
+      "loss": 3.1042,
+      "step": 27436
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5808750987052917,
+      "learning_rate": 0.000348598514766003,
+      "loss": 3.0015,
+      "step": 27437
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5626875758171082,
+      "learning_rate": 0.00034858332187109035,
+      "loss": 3.1874,
+      "step": 27438
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5855140089988708,
+      "learning_rate": 0.0003485681288482186,
+      "loss": 3.0787,
+      "step": 27439
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5767703056335449,
+      "learning_rate": 0.00034855293569742776,
+      "loss": 3.1796,
+      "step": 27440
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5492141246795654,
+      "learning_rate": 0.0003485377424187578,
+      "loss": 2.7963,
+      "step": 27441
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5425765514373779,
+      "learning_rate": 0.0003485225490122488,
+      "loss": 3.1762,
+      "step": 27442
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5987265706062317,
+      "learning_rate": 0.00034850735547794066,
+      "loss": 3.1511,
+      "step": 27443
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5857033729553223,
+      "learning_rate": 0.00034849216181587363,
+      "loss": 3.1447,
+      "step": 27444
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6146132946014404,
+      "learning_rate": 0.0003484769680260874,
+      "loss": 3.1278,
+      "step": 27445
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5451048016548157,
+      "learning_rate": 0.00034846177410862236,
+      "loss": 3.1896,
+      "step": 27446
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6035172343254089,
+      "learning_rate": 0.00034844658006351814,
+      "loss": 3.1108,
+      "step": 27447
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5181143283843994,
+      "learning_rate": 0.0003484313858908151,
+      "loss": 3.073,
+      "step": 27448
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5882478356361389,
+      "learning_rate": 0.000348416191590553,
+      "loss": 3.1133,
+      "step": 27449
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5757107138633728,
+      "learning_rate": 0.00034840099716277207,
+      "loss": 3.1586,
+      "step": 27450
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5847183465957642,
+      "learning_rate": 0.0003483858026075122,
+      "loss": 3.0846,
+      "step": 27451
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6361482739448547,
+      "learning_rate": 0.00034837060792481333,
+      "loss": 3.0394,
+      "step": 27452
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5674738883972168,
+      "learning_rate": 0.00034835541311471573,
+      "loss": 2.8753,
+      "step": 27453
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5853537321090698,
+      "learning_rate": 0.00034834021817725923,
+      "loss": 2.982,
+      "step": 27454
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6073134541511536,
+      "learning_rate": 0.00034832502311248385,
+      "loss": 3.1412,
+      "step": 27455
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5832349061965942,
+      "learning_rate": 0.00034830982792042974,
+      "loss": 3.0898,
+      "step": 27456
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5568851828575134,
+      "learning_rate": 0.00034829463260113675,
+      "loss": 2.9267,
+      "step": 27457
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5909299850463867,
+      "learning_rate": 0.00034827943715464506,
+      "loss": 2.9748,
+      "step": 27458
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5588631629943848,
+      "learning_rate": 0.0003482642415809946,
+      "loss": 2.9251,
+      "step": 27459
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5427162051200867,
+      "learning_rate": 0.0003482490458802254,
+      "loss": 3.0714,
+      "step": 27460
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5868051052093506,
+      "learning_rate": 0.00034823385005237765,
+      "loss": 3.2121,
+      "step": 27461
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5954635143280029,
+      "learning_rate": 0.00034821865409749103,
+      "loss": 3.2533,
+      "step": 27462
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5590815544128418,
+      "learning_rate": 0.0003482034580156058,
+      "loss": 3.1354,
+      "step": 27463
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5484957695007324,
+      "learning_rate": 0.00034818826180676196,
+      "loss": 2.9512,
+      "step": 27464
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6143427491188049,
+      "learning_rate": 0.0003481730654709995,
+      "loss": 2.9583,
+      "step": 27465
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.602432131767273,
+      "learning_rate": 0.0003481578690083585,
+      "loss": 3.1544,
+      "step": 27466
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5481232404708862,
+      "learning_rate": 0.0003481426724188789,
+      "loss": 3.1557,
+      "step": 27467
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5532791018486023,
+      "learning_rate": 0.00034812747570260073,
+      "loss": 3.0697,
+      "step": 27468
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5800402164459229,
+      "learning_rate": 0.0003481122788595641,
+      "loss": 2.9307,
+      "step": 27469
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6530680656433105,
+      "learning_rate": 0.00034809708188980896,
+      "loss": 3.2462,
+      "step": 27470
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5887762308120728,
+      "learning_rate": 0.0003480818847933754,
+      "loss": 3.0021,
+      "step": 27471
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5353226661682129,
+      "learning_rate": 0.0003480666875703035,
+      "loss": 3.0513,
+      "step": 27472
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5689337253570557,
+      "learning_rate": 0.00034805149022063296,
+      "loss": 3.1072,
+      "step": 27473
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5485574007034302,
+      "learning_rate": 0.0003480362927444042,
+      "loss": 3.1626,
+      "step": 27474
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6580893993377686,
+      "learning_rate": 0.0003480210951416571,
+      "loss": 3.0229,
+      "step": 27475
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6442924737930298,
+      "learning_rate": 0.00034800589741243167,
+      "loss": 3.077,
+      "step": 27476
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5990310311317444,
+      "learning_rate": 0.00034799069955676794,
+      "loss": 3.1217,
+      "step": 27477
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5739229321479797,
+      "learning_rate": 0.0003479755015747059,
+      "loss": 3.1821,
+      "step": 27478
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6192485690116882,
+      "learning_rate": 0.0003479603034662856,
+      "loss": 3.071,
+      "step": 27479
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5898162722587585,
+      "learning_rate": 0.0003479451052315471,
+      "loss": 3.1133,
+      "step": 27480
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5484324097633362,
+      "learning_rate": 0.00034792990687053046,
+      "loss": 2.8319,
+      "step": 27481
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5545966625213623,
+      "learning_rate": 0.0003479147083832757,
+      "loss": 3.0854,
+      "step": 27482
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5734389424324036,
+      "learning_rate": 0.00034789950976982275,
+      "loss": 3.2265,
+      "step": 27483
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6044603586196899,
+      "learning_rate": 0.00034788431103021175,
+      "loss": 2.8291,
+      "step": 27484
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5482417345046997,
+      "learning_rate": 0.00034786911216448267,
+      "loss": 3.1032,
+      "step": 27485
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5807044506072998,
+      "learning_rate": 0.0003478539131726757,
+      "loss": 3.2418,
+      "step": 27486
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5764628648757935,
+      "learning_rate": 0.00034783871405483056,
+      "loss": 3.1398,
+      "step": 27487
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5609407424926758,
+      "learning_rate": 0.00034782351481098744,
+      "loss": 3.0961,
+      "step": 27488
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5483243465423584,
+      "learning_rate": 0.0003478083154411865,
+      "loss": 3.3225,
+      "step": 27489
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5866134166717529,
+      "learning_rate": 0.0003477931159454675,
+      "loss": 3.1778,
+      "step": 27490
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5812581181526184,
+      "learning_rate": 0.0003477779163238708,
+      "loss": 2.9979,
+      "step": 27491
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5664998292922974,
+      "learning_rate": 0.00034776271657643615,
+      "loss": 3.0774,
+      "step": 27492
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6252369284629822,
+      "learning_rate": 0.0003477475167032038,
+      "loss": 2.8475,
+      "step": 27493
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.647500216960907,
+      "learning_rate": 0.0003477323167042136,
+      "loss": 2.954,
+      "step": 27494
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5603214502334595,
+      "learning_rate": 0.0003477171165795057,
+      "loss": 2.9279,
+      "step": 27495
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5812236070632935,
+      "learning_rate": 0.00034770191632912,
+      "loss": 3.0352,
+      "step": 27496
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5692586898803711,
+      "learning_rate": 0.0003476867159530968,
+      "loss": 3.1431,
+      "step": 27497
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5523695945739746,
+      "learning_rate": 0.0003476715154514758,
+      "loss": 3.2377,
+      "step": 27498
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5964672565460205,
+      "learning_rate": 0.00034765631482429733,
+      "loss": 3.2893,
+      "step": 27499
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.601300835609436,
+      "learning_rate": 0.0003476411140716012,
+      "loss": 2.9938,
+      "step": 27500
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5985939502716064,
+      "learning_rate": 0.0003476259131934277,
+      "loss": 3.0204,
+      "step": 27501
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5791720151901245,
+      "learning_rate": 0.0003476107121898166,
+      "loss": 3.0052,
+      "step": 27502
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6213023662567139,
+      "learning_rate": 0.00034759551106080795,
+      "loss": 3.0617,
+      "step": 27503
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5528647899627686,
+      "learning_rate": 0.0003475803098064421,
+      "loss": 3.024,
+      "step": 27504
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5452709197998047,
+      "learning_rate": 0.0003475651084267587,
+      "loss": 3.1488,
+      "step": 27505
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5762589573860168,
+      "learning_rate": 0.000347549906921798,
+      "loss": 3.0249,
+      "step": 27506
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6029790043830872,
+      "learning_rate": 0.00034753470529160006,
+      "loss": 3.1336,
+      "step": 27507
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5660719871520996,
+      "learning_rate": 0.00034751950353620494,
+      "loss": 3.1616,
+      "step": 27508
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5759086012840271,
+      "learning_rate": 0.00034750430165565233,
+      "loss": 3.1031,
+      "step": 27509
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5439847111701965,
+      "learning_rate": 0.00034748909964998264,
+      "loss": 3.1131,
+      "step": 27510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6624140739440918,
+      "learning_rate": 0.0003474738975192359,
+      "loss": 3.0788,
+      "step": 27511
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5809428095817566,
+      "learning_rate": 0.000347458695263452,
+      "loss": 3.0249,
+      "step": 27512
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5749208927154541,
+      "learning_rate": 0.00034744349288267105,
+      "loss": 2.9156,
+      "step": 27513
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5683353543281555,
+      "learning_rate": 0.0003474282903769329,
+      "loss": 3.1445,
+      "step": 27514
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5920277833938599,
+      "learning_rate": 0.00034741308774627794,
+      "loss": 3.0885,
+      "step": 27515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.561812698841095,
+      "learning_rate": 0.000347397884990746,
+      "loss": 3.0101,
+      "step": 27516
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5506225228309631,
+      "learning_rate": 0.00034738268211037716,
+      "loss": 3.156,
+      "step": 27517
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5584924817085266,
+      "learning_rate": 0.00034736747910521135,
+      "loss": 2.9141,
+      "step": 27518
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6194260716438293,
+      "learning_rate": 0.00034735227597528884,
+      "loss": 3.0649,
+      "step": 27519
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5642139315605164,
+      "learning_rate": 0.0003473370727206495,
+      "loss": 2.6767,
+      "step": 27520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6075909733772278,
+      "learning_rate": 0.0003473218693413334,
+      "loss": 2.9534,
+      "step": 27521
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5577836632728577,
+      "learning_rate": 0.00034730666583738064,
+      "loss": 2.9497,
+      "step": 27522
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.565424382686615,
+      "learning_rate": 0.0003472914622088312,
+      "loss": 3.0355,
+      "step": 27523
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5719602704048157,
+      "learning_rate": 0.0003472762584557252,
+      "loss": 3.1099,
+      "step": 27524
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5585300922393799,
+      "learning_rate": 0.00034726105457810253,
+      "loss": 3.027,
+      "step": 27525
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5087852478027344,
+      "learning_rate": 0.0003472458505760035,
+      "loss": 3.1373,
+      "step": 27526
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5587421655654907,
+      "learning_rate": 0.00034723064644946787,
+      "loss": 3.1473,
+      "step": 27527
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5711315274238586,
+      "learning_rate": 0.0003472154421985359,
+      "loss": 3.0316,
+      "step": 27528
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5235508680343628,
+      "learning_rate": 0.0003472002378232474,
+      "loss": 3.0433,
+      "step": 27529
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5291568040847778,
+      "learning_rate": 0.00034718503332364264,
+      "loss": 3.2164,
+      "step": 27530
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5746374130249023,
+      "learning_rate": 0.00034716982869976157,
+      "loss": 3.1122,
+      "step": 27531
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5441637635231018,
+      "learning_rate": 0.0003471546239516443,
+      "loss": 2.9419,
+      "step": 27532
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5394623279571533,
+      "learning_rate": 0.00034713941907933075,
+      "loss": 3.0706,
+      "step": 27533
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5609070062637329,
+      "learning_rate": 0.00034712421408286106,
+      "loss": 2.9203,
+      "step": 27534
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5641400218009949,
+      "learning_rate": 0.0003471090089622752,
+      "loss": 3.1299,
+      "step": 27535
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5568336248397827,
+      "learning_rate": 0.0003470938037176134,
+      "loss": 3.0579,
+      "step": 27536
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5555543899536133,
+      "learning_rate": 0.00034707859834891557,
+      "loss": 2.9768,
+      "step": 27537
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5764153599739075,
+      "learning_rate": 0.0003470633928562218,
+      "loss": 2.9253,
+      "step": 27538
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5446388125419617,
+      "learning_rate": 0.00034704818723957197,
+      "loss": 3.1424,
+      "step": 27539
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5743789076805115,
+      "learning_rate": 0.0003470329814990064,
+      "loss": 3.0394,
+      "step": 27540
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5995753407478333,
+      "learning_rate": 0.00034701777563456496,
+      "loss": 3.1474,
+      "step": 27541
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5293822288513184,
+      "learning_rate": 0.00034700256964628767,
+      "loss": 3.0878,
+      "step": 27542
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5851097702980042,
+      "learning_rate": 0.00034698736353421477,
+      "loss": 3.1295,
+      "step": 27543
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5822820067405701,
+      "learning_rate": 0.00034697215729838615,
+      "loss": 3.1411,
+      "step": 27544
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5606449842453003,
+      "learning_rate": 0.00034695695093884193,
+      "loss": 3.1297,
+      "step": 27545
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5498665571212769,
+      "learning_rate": 0.00034694174445562206,
+      "loss": 3.0238,
+      "step": 27546
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5421035289764404,
+      "learning_rate": 0.0003469265378487668,
+      "loss": 3.0742,
+      "step": 27547
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5203042030334473,
+      "learning_rate": 0.000346911331118316,
+      "loss": 3.0389,
+      "step": 27548
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5701861381530762,
+      "learning_rate": 0.0003468961242643098,
+      "loss": 3.2245,
+      "step": 27549
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5597485303878784,
+      "learning_rate": 0.0003468809172867881,
+      "loss": 3.0925,
+      "step": 27550
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5598206520080566,
+      "learning_rate": 0.00034686571018579127,
+      "loss": 2.9806,
+      "step": 27551
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5926447510719299,
+      "learning_rate": 0.00034685050296135914,
+      "loss": 2.9184,
+      "step": 27552
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5669886469841003,
+      "learning_rate": 0.00034683529561353174,
+      "loss": 3.074,
+      "step": 27553
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5824142098426819,
+      "learning_rate": 0.0003468200881423493,
+      "loss": 3.116,
+      "step": 27554
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5819292068481445,
+      "learning_rate": 0.00034680488054785163,
+      "loss": 3.0986,
+      "step": 27555
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5521724820137024,
+      "learning_rate": 0.000346789672830079,
+      "loss": 2.9404,
+      "step": 27556
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5370104908943176,
+      "learning_rate": 0.0003467744649890713,
+      "loss": 3.0389,
+      "step": 27557
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5862271189689636,
+      "learning_rate": 0.0003467592570248687,
+      "loss": 3.1758,
+      "step": 27558
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5811551809310913,
+      "learning_rate": 0.0003467440489375113,
+      "loss": 2.9348,
+      "step": 27559
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.523688554763794,
+      "learning_rate": 0.00034672884072703885,
+      "loss": 2.9629,
+      "step": 27560
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5698621273040771,
+      "learning_rate": 0.0003467136323934918,
+      "loss": 3.0013,
+      "step": 27561
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5459543466567993,
+      "learning_rate": 0.00034669842393691,
+      "loss": 3.0861,
+      "step": 27562
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6095524430274963,
+      "learning_rate": 0.0003466832153573336,
+      "loss": 3.0588,
+      "step": 27563
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5736280679702759,
+      "learning_rate": 0.00034666800665480253,
+      "loss": 3.1733,
+      "step": 27564
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.579998254776001,
+      "learning_rate": 0.0003466527978293568,
+      "loss": 3.0893,
+      "step": 27565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5939657688140869,
+      "learning_rate": 0.00034663758888103677,
+      "loss": 3.0966,
+      "step": 27566
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5807148814201355,
+      "learning_rate": 0.00034662237980988226,
+      "loss": 3.03,
+      "step": 27567
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.57145756483078,
+      "learning_rate": 0.00034660717061593335,
+      "loss": 3.0573,
+      "step": 27568
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5279229283332825,
+      "learning_rate": 0.00034659196129923004,
+      "loss": 2.9326,
+      "step": 27569
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5835081338882446,
+      "learning_rate": 0.00034657675185981255,
+      "loss": 2.9735,
+      "step": 27570
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5817593932151794,
+      "learning_rate": 0.00034656154229772084,
+      "loss": 3.1158,
+      "step": 27571
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5388606786727905,
+      "learning_rate": 0.000346546332612995,
+      "loss": 2.8509,
+      "step": 27572
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.622467041015625,
+      "learning_rate": 0.0003465311228056751,
+      "loss": 3.0875,
+      "step": 27573
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5687299966812134,
+      "learning_rate": 0.00034651591287580125,
+      "loss": 2.9959,
+      "step": 27574
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5633476972579956,
+      "learning_rate": 0.00034650070282341326,
+      "loss": 2.9332,
+      "step": 27575
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5782656073570251,
+      "learning_rate": 0.00034648549264855146,
+      "loss": 3.0858,
+      "step": 27576
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.577627956867218,
+      "learning_rate": 0.0003464702823512558,
+      "loss": 3.012,
+      "step": 27577
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.559639573097229,
+      "learning_rate": 0.00034645507193156646,
+      "loss": 3.1584,
+      "step": 27578
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5748494863510132,
+      "learning_rate": 0.0003464398613895233,
+      "loss": 3.0548,
+      "step": 27579
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5871093273162842,
+      "learning_rate": 0.0003464246507251664,
+      "loss": 3.1131,
+      "step": 27580
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5574132204055786,
+      "learning_rate": 0.00034640943993853606,
+      "loss": 3.1498,
+      "step": 27581
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5610949397087097,
+      "learning_rate": 0.0003463942290296721,
+      "loss": 3.0497,
+      "step": 27582
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5792922377586365,
+      "learning_rate": 0.0003463790179986147,
+      "loss": 3.3694,
+      "step": 27583
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5634068250656128,
+      "learning_rate": 0.0003463638068454039,
+      "loss": 3.0192,
+      "step": 27584
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5394881367683411,
+      "learning_rate": 0.00034634859557007976,
+      "loss": 3.1326,
+      "step": 27585
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6030181050300598,
+      "learning_rate": 0.00034633338417268227,
+      "loss": 3.0469,
+      "step": 27586
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6113521456718445,
+      "learning_rate": 0.0003463181726532516,
+      "loss": 3.0221,
+      "step": 27587
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5368742346763611,
+      "learning_rate": 0.00034630296101182794,
+      "loss": 3.1202,
+      "step": 27588
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.588915228843689,
+      "learning_rate": 0.000346287749248451,
+      "loss": 3.0284,
+      "step": 27589
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5892957448959351,
+      "learning_rate": 0.00034627253736316103,
+      "loss": 3.1462,
+      "step": 27590
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5666577816009521,
+      "learning_rate": 0.0003462573253559982,
+      "loss": 3.0818,
+      "step": 27591
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5563762187957764,
+      "learning_rate": 0.0003462421132270024,
+      "loss": 3.2716,
+      "step": 27592
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.615742564201355,
+      "learning_rate": 0.0003462269009762138,
+      "loss": 3.0507,
+      "step": 27593
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5579602122306824,
+      "learning_rate": 0.0003462116886036725,
+      "loss": 3.0305,
+      "step": 27594
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5372714996337891,
+      "learning_rate": 0.0003461964761094184,
+      "loss": 2.8777,
+      "step": 27595
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5734666585922241,
+      "learning_rate": 0.0003461812634934917,
+      "loss": 2.9552,
+      "step": 27596
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5839183330535889,
+      "learning_rate": 0.0003461660507559325,
+      "loss": 3.0588,
+      "step": 27597
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6317874193191528,
+      "learning_rate": 0.00034615083789678075,
+      "loss": 2.8828,
+      "step": 27598
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.572235643863678,
+      "learning_rate": 0.0003461356249160767,
+      "loss": 3.0354,
+      "step": 27599
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5442335605621338,
+      "learning_rate": 0.00034612041181386014,
+      "loss": 3.099,
+      "step": 27600
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5586719512939453,
+      "learning_rate": 0.0003461051985901713,
+      "loss": 3.1294,
+      "step": 27601
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5442153811454773,
+      "learning_rate": 0.0003460899852450502,
+      "loss": 2.9258,
+      "step": 27602
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5317257642745972,
+      "learning_rate": 0.0003460747717785371,
+      "loss": 3.2071,
+      "step": 27603
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5560556054115295,
+      "learning_rate": 0.0003460595581906719,
+      "loss": 3.2065,
+      "step": 27604
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6010359525680542,
+      "learning_rate": 0.0003460443444814946,
+      "loss": 2.8649,
+      "step": 27605
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5659478306770325,
+      "learning_rate": 0.0003460291306510454,
+      "loss": 3.0582,
+      "step": 27606
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5736438035964966,
+      "learning_rate": 0.00034601391669936436,
+      "loss": 3.0973,
+      "step": 27607
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5653098821640015,
+      "learning_rate": 0.0003459987026264914,
+      "loss": 3.3044,
+      "step": 27608
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5595373511314392,
+      "learning_rate": 0.0003459834884324668,
+      "loss": 3.1815,
+      "step": 27609
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5338704586029053,
+      "learning_rate": 0.0003459682741173306,
+      "loss": 3.2657,
+      "step": 27610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.595348060131073,
+      "learning_rate": 0.0003459530596811227,
+      "loss": 3.0412,
+      "step": 27611
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5767225623130798,
+      "learning_rate": 0.0003459378451238833,
+      "loss": 3.1244,
+      "step": 27612
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5606244206428528,
+      "learning_rate": 0.00034592263044565247,
+      "loss": 2.9589,
+      "step": 27613
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.581843376159668,
+      "learning_rate": 0.0003459074156464704,
+      "loss": 3.1054,
+      "step": 27614
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5597598552703857,
+      "learning_rate": 0.0003458922007263769,
+      "loss": 3.1613,
+      "step": 27615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6114240884780884,
+      "learning_rate": 0.0003458769856854121,
+      "loss": 2.9326,
+      "step": 27616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6092174649238586,
+      "learning_rate": 0.0003458617705236163,
+      "loss": 3.1666,
+      "step": 27617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5693092346191406,
+      "learning_rate": 0.0003458465552410294,
+      "loss": 3.1785,
+      "step": 27618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5575876235961914,
+      "learning_rate": 0.00034583133983769146,
+      "loss": 3.0092,
+      "step": 27619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5795199871063232,
+      "learning_rate": 0.00034581612431364253,
+      "loss": 3.2021,
+      "step": 27620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5466296076774597,
+      "learning_rate": 0.0003458009086689228,
+      "loss": 3.271,
+      "step": 27621
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5569012761116028,
+      "learning_rate": 0.00034578569290357234,
+      "loss": 3.0389,
+      "step": 27622
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.554438591003418,
+      "learning_rate": 0.00034577047701763114,
+      "loss": 3.284,
+      "step": 27623
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.540482759475708,
+      "learning_rate": 0.00034575526101113924,
+      "loss": 3.0049,
+      "step": 27624
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5931518077850342,
+      "learning_rate": 0.0003457400448841369,
+      "loss": 3.0533,
+      "step": 27625
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5511003136634827,
+      "learning_rate": 0.0003457248286366641,
+      "loss": 2.9812,
+      "step": 27626
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5712304711341858,
+      "learning_rate": 0.00034570961226876074,
+      "loss": 3.1466,
+      "step": 27627
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5965542197227478,
+      "learning_rate": 0.0003456943957804672,
+      "loss": 3.0022,
+      "step": 27628
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5600318312644958,
+      "learning_rate": 0.0003456791791718234,
+      "loss": 3.0214,
+      "step": 27629
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5907925367355347,
+      "learning_rate": 0.00034566396244286945,
+      "loss": 3.0262,
+      "step": 27630
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6217458844184875,
+      "learning_rate": 0.00034564874559364536,
+      "loss": 3.1207,
+      "step": 27631
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5708328485488892,
+      "learning_rate": 0.00034563352862419127,
+      "loss": 3.0498,
+      "step": 27632
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5754064321517944,
+      "learning_rate": 0.00034561831153454725,
+      "loss": 3.1127,
+      "step": 27633
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5828617215156555,
+      "learning_rate": 0.00034560309432475335,
+      "loss": 3.1008,
+      "step": 27634
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.553674042224884,
+      "learning_rate": 0.0003455878769948497,
+      "loss": 3.0445,
+      "step": 27635
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5497589707374573,
+      "learning_rate": 0.0003455726595448764,
+      "loss": 3.0791,
+      "step": 27636
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5885263681411743,
+      "learning_rate": 0.00034555744197487334,
+      "loss": 2.9445,
+      "step": 27637
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5797986388206482,
+      "learning_rate": 0.0003455422242848809,
+      "loss": 3.3291,
+      "step": 27638
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5659964680671692,
+      "learning_rate": 0.000345527006474939,
+      "loss": 2.9028,
+      "step": 27639
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5410609841346741,
+      "learning_rate": 0.00034551178854508763,
+      "loss": 3.1878,
+      "step": 27640
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6087222099304199,
+      "learning_rate": 0.0003454965704953671,
+      "loss": 3.0399,
+      "step": 27641
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5863704085350037,
+      "learning_rate": 0.0003454813523258172,
+      "loss": 2.9568,
+      "step": 27642
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5730543732643127,
+      "learning_rate": 0.00034546613403647826,
+      "loss": 2.7912,
+      "step": 27643
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5473147630691528,
+      "learning_rate": 0.0003454509156273903,
+      "loss": 3.1532,
+      "step": 27644
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6057653427124023,
+      "learning_rate": 0.0003454356970985933,
+      "loss": 3.098,
+      "step": 27645
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5998139381408691,
+      "learning_rate": 0.0003454204784501275,
+      "loss": 3.0748,
+      "step": 27646
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6057741641998291,
+      "learning_rate": 0.0003454052596820328,
+      "loss": 3.0626,
+      "step": 27647
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5644229650497437,
+      "learning_rate": 0.00034539004079434945,
+      "loss": 3.1077,
+      "step": 27648
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5536103248596191,
+      "learning_rate": 0.00034537482178711743,
+      "loss": 2.9196,
+      "step": 27649
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5765447616577148,
+      "learning_rate": 0.00034535960266037695,
+      "loss": 3.158,
+      "step": 27650
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.551717221736908,
+      "learning_rate": 0.00034534438341416796,
+      "loss": 3.047,
+      "step": 27651
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.55844646692276,
+      "learning_rate": 0.0003453291640485305,
+      "loss": 3.0137,
+      "step": 27652
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5359190702438354,
+      "learning_rate": 0.00034531394456350486,
+      "loss": 3.0382,
+      "step": 27653
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5825155377388,
+      "learning_rate": 0.00034529872495913104,
+      "loss": 3.2114,
+      "step": 27654
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5790165662765503,
+      "learning_rate": 0.000345283505235449,
+      "loss": 2.943,
+      "step": 27655
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6118806004524231,
+      "learning_rate": 0.00034526828539249894,
+      "loss": 2.9642,
+      "step": 27656
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5818917751312256,
+      "learning_rate": 0.00034525306543032095,
+      "loss": 2.8687,
+      "step": 27657
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5818273425102234,
+      "learning_rate": 0.0003452378453489551,
+      "loss": 2.9843,
+      "step": 27658
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5957201719284058,
+      "learning_rate": 0.00034522262514844143,
+      "loss": 2.9917,
+      "step": 27659
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5800155997276306,
+      "learning_rate": 0.0003452074048288201,
+      "loss": 2.8889,
+      "step": 27660
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5779448747634888,
+      "learning_rate": 0.0003451921843901312,
+      "loss": 3.0047,
+      "step": 27661
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6519986391067505,
+      "learning_rate": 0.0003451769638324147,
+      "loss": 3.1293,
+      "step": 27662
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.577875554561615,
+      "learning_rate": 0.00034516174315571077,
+      "loss": 3.0927,
+      "step": 27663
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5626934766769409,
+      "learning_rate": 0.00034514652236005956,
+      "loss": 2.9672,
+      "step": 27664
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5806666612625122,
+      "learning_rate": 0.00034513130144550114,
+      "loss": 2.8056,
+      "step": 27665
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5470517873764038,
+      "learning_rate": 0.0003451160804120755,
+      "loss": 2.9702,
+      "step": 27666
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.560828685760498,
+      "learning_rate": 0.00034510085925982284,
+      "loss": 3.1832,
+      "step": 27667
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5910745859146118,
+      "learning_rate": 0.0003450856379887831,
+      "loss": 2.9672,
+      "step": 27668
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5761743187904358,
+      "learning_rate": 0.0003450704165989965,
+      "loss": 2.8815,
+      "step": 27669
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5565659403800964,
+      "learning_rate": 0.00034505519509050314,
+      "loss": 3.0549,
+      "step": 27670
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5785242319107056,
+      "learning_rate": 0.000345039973463343,
+      "loss": 2.9993,
+      "step": 27671
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5992151498794556,
+      "learning_rate": 0.00034502475171755635,
+      "loss": 2.9655,
+      "step": 27672
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5523272752761841,
+      "learning_rate": 0.00034500952985318305,
+      "loss": 3.1217,
+      "step": 27673
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.543624222278595,
+      "learning_rate": 0.0003449943078702633,
+      "loss": 3.0863,
+      "step": 27674
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5931307673454285,
+      "learning_rate": 0.0003449790857688373,
+      "loss": 3.2511,
+      "step": 27675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5931395888328552,
+      "learning_rate": 0.0003449638635489451,
+      "loss": 2.8998,
+      "step": 27676
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5682350993156433,
+      "learning_rate": 0.0003449486412106266,
+      "loss": 3.0556,
+      "step": 27677
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.590588390827179,
+      "learning_rate": 0.00034493341875392196,
+      "loss": 3.0171,
+      "step": 27678
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6109996438026428,
+      "learning_rate": 0.00034491819617887154,
+      "loss": 3.0134,
+      "step": 27679
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6124204397201538,
+      "learning_rate": 0.0003449029734855152,
+      "loss": 3.1012,
+      "step": 27680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5548977255821228,
+      "learning_rate": 0.00034488775067389294,
+      "loss": 3.282,
+      "step": 27681
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6532643437385559,
+      "learning_rate": 0.00034487252774404505,
+      "loss": 2.9508,
+      "step": 27682
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6053386330604553,
+      "learning_rate": 0.00034485730469601154,
+      "loss": 3.2382,
+      "step": 27683
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.567348837852478,
+      "learning_rate": 0.00034484208152983257,
+      "loss": 2.7893,
+      "step": 27684
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5835475325584412,
+      "learning_rate": 0.00034482685824554815,
+      "loss": 3.0216,
+      "step": 27685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5753740668296814,
+      "learning_rate": 0.00034481163484319845,
+      "loss": 3.3452,
+      "step": 27686
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6069841980934143,
+      "learning_rate": 0.0003447964113228236,
+      "loss": 3.1379,
+      "step": 27687
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5838615894317627,
+      "learning_rate": 0.0003447811876844634,
+      "loss": 2.9669,
+      "step": 27688
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.547076940536499,
+      "learning_rate": 0.00034476596392815835,
+      "loss": 3.0967,
+      "step": 27689
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6205630898475647,
+      "learning_rate": 0.00034475074005394836,
+      "loss": 3.1115,
+      "step": 27690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5432915091514587,
+      "learning_rate": 0.0003447355160618735,
+      "loss": 3.1954,
+      "step": 27691
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5629213452339172,
+      "learning_rate": 0.0003447202919519739,
+      "loss": 3.134,
+      "step": 27692
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6095461249351501,
+      "learning_rate": 0.00034470506772428966,
+      "loss": 2.9366,
+      "step": 27693
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5309693813323975,
+      "learning_rate": 0.00034468984337886085,
+      "loss": 3.2188,
+      "step": 27694
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5798507928848267,
+      "learning_rate": 0.0003446746189157276,
+      "loss": 2.9499,
+      "step": 27695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5637386441230774,
+      "learning_rate": 0.00034465939433493003,
+      "loss": 2.9753,
+      "step": 27696
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5978463888168335,
+      "learning_rate": 0.0003446441696365082,
+      "loss": 2.9246,
+      "step": 27697
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5650119781494141,
+      "learning_rate": 0.00034462894482050214,
+      "loss": 3.1587,
+      "step": 27698
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6245388984680176,
+      "learning_rate": 0.00034461371988695215,
+      "loss": 3.1073,
+      "step": 27699
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6453614234924316,
+      "learning_rate": 0.0003445984948358981,
+      "loss": 3.1702,
+      "step": 27700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5954031944274902,
+      "learning_rate": 0.0003445832696673803,
+      "loss": 3.2765,
+      "step": 27701
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5462033152580261,
+      "learning_rate": 0.0003445680443814387,
+      "loss": 2.9853,
+      "step": 27702
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5822536945343018,
+      "learning_rate": 0.00034455281897811344,
+      "loss": 3.147,
+      "step": 27703
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5529893040657043,
+      "learning_rate": 0.0003445375934574446,
+      "loss": 3.1181,
+      "step": 27704
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5892257690429688,
+      "learning_rate": 0.0003445223678194724,
+      "loss": 3.0272,
+      "step": 27705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5971611738204956,
+      "learning_rate": 0.0003445071420642368,
+      "loss": 3.0081,
+      "step": 27706
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5674407482147217,
+      "learning_rate": 0.000344491916191778,
+      "loss": 3.1847,
+      "step": 27707
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6164575219154358,
+      "learning_rate": 0.000344476690202136,
+      "loss": 3.0107,
+      "step": 27708
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5708376169204712,
+      "learning_rate": 0.0003444614640953509,
+      "loss": 3.1381,
+      "step": 27709
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5596605539321899,
+      "learning_rate": 0.00034444623787146297,
+      "loss": 3.0325,
+      "step": 27710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5818103551864624,
+      "learning_rate": 0.0003444310115305121,
+      "loss": 3.1746,
+      "step": 27711
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5505349040031433,
+      "learning_rate": 0.0003444157850725386,
+      "loss": 3.128,
+      "step": 27712
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5452769994735718,
+      "learning_rate": 0.00034440055849758246,
+      "loss": 3.2109,
+      "step": 27713
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5605431199073792,
+      "learning_rate": 0.0003443853318056837,
+      "loss": 2.9834,
+      "step": 27714
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5733485221862793,
+      "learning_rate": 0.0003443701049968826,
+      "loss": 3.1925,
+      "step": 27715
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5892324447631836,
+      "learning_rate": 0.0003443548780712192,
+      "loss": 2.9853,
+      "step": 27716
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6256640553474426,
+      "learning_rate": 0.0003443396510287335,
+      "loss": 3.0881,
+      "step": 27717
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6237658858299255,
+      "learning_rate": 0.00034432442386946575,
+      "loss": 3.0615,
+      "step": 27718
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6183185577392578,
+      "learning_rate": 0.000344309196593456,
+      "loss": 2.9544,
+      "step": 27719
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5464622974395752,
+      "learning_rate": 0.00034429396920074436,
+      "loss": 2.9632,
+      "step": 27720
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5681126117706299,
+      "learning_rate": 0.0003442787416913709,
+      "loss": 3.0697,
+      "step": 27721
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5842775106430054,
+      "learning_rate": 0.0003442635140653758,
+      "loss": 3.0525,
+      "step": 27722
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5486373901367188,
+      "learning_rate": 0.00034424828632279914,
+      "loss": 3.0561,
+      "step": 27723
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5385405421257019,
+      "learning_rate": 0.0003442330584636809,
+      "loss": 3.1888,
+      "step": 27724
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5733704566955566,
+      "learning_rate": 0.0003442178304880613,
+      "loss": 2.9091,
+      "step": 27725
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5765992999076843,
+      "learning_rate": 0.00034420260239598053,
+      "loss": 3.0318,
+      "step": 27726
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5930641889572144,
+      "learning_rate": 0.0003441873741874786,
+      "loss": 3.0233,
+      "step": 27727
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5822330117225647,
+      "learning_rate": 0.00034417214586259567,
+      "loss": 3.1026,
+      "step": 27728
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6270058751106262,
+      "learning_rate": 0.0003441569174213717,
+      "loss": 3.1726,
+      "step": 27729
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5660275220870972,
+      "learning_rate": 0.000344141688863847,
+      "loss": 2.8885,
+      "step": 27730
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6260101199150085,
+      "learning_rate": 0.00034412646019006156,
+      "loss": 3.075,
+      "step": 27731
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.577351450920105,
+      "learning_rate": 0.0003441112314000555,
+      "loss": 3.1064,
+      "step": 27732
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5536080598831177,
+      "learning_rate": 0.00034409600249386894,
+      "loss": 3.1454,
+      "step": 27733
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5863784551620483,
+      "learning_rate": 0.000344080773471542,
+      "loss": 3.0272,
+      "step": 27734
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6073148846626282,
+      "learning_rate": 0.0003440655443331148,
+      "loss": 3.2349,
+      "step": 27735
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5578989386558533,
+      "learning_rate": 0.00034405031507862746,
+      "loss": 3.078,
+      "step": 27736
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5783877372741699,
+      "learning_rate": 0.00034403508570811993,
+      "loss": 3.208,
+      "step": 27737
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5361840128898621,
+      "learning_rate": 0.0003440198562216327,
+      "loss": 3.0135,
+      "step": 27738
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5612678527832031,
+      "learning_rate": 0.0003440046266192054,
+      "loss": 2.8991,
+      "step": 27739
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.575139045715332,
+      "learning_rate": 0.0003439893969008785,
+      "loss": 2.9018,
+      "step": 27740
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5466126799583435,
+      "learning_rate": 0.000343974167066692,
+      "loss": 2.9975,
+      "step": 27741
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5556744337081909,
+      "learning_rate": 0.000343958937116686,
+      "loss": 2.9647,
+      "step": 27742
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6033949255943298,
+      "learning_rate": 0.00034394370705090067,
+      "loss": 2.8281,
+      "step": 27743
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5831649303436279,
+      "learning_rate": 0.000343928476869376,
+      "loss": 3.053,
+      "step": 27744
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5626701712608337,
+      "learning_rate": 0.0003439132465721522,
+      "loss": 3.0697,
+      "step": 27745
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5939829349517822,
+      "learning_rate": 0.0003438980161592693,
+      "loss": 3.2721,
+      "step": 27746
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.564018964767456,
+      "learning_rate": 0.0003438827856307675,
+      "loss": 2.8827,
+      "step": 27747
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5679821968078613,
+      "learning_rate": 0.0003438675549866869,
+      "loss": 3.1603,
+      "step": 27748
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.56532222032547,
+      "learning_rate": 0.0003438523242270677,
+      "loss": 3.0086,
+      "step": 27749
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5731778144836426,
+      "learning_rate": 0.00034383709335194975,
+      "loss": 3.0396,
+      "step": 27750
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5794793963432312,
+      "learning_rate": 0.00034382186236137346,
+      "loss": 2.9388,
+      "step": 27751
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5715826153755188,
+      "learning_rate": 0.0003438066312553788,
+      "loss": 3.0948,
+      "step": 27752
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6393271088600159,
+      "learning_rate": 0.0003437914000340059,
+      "loss": 2.9898,
+      "step": 27753
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5703928470611572,
+      "learning_rate": 0.00034377616869729486,
+      "loss": 3.2764,
+      "step": 27754
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5397366285324097,
+      "learning_rate": 0.0003437609372452858,
+      "loss": 3.1067,
+      "step": 27755
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5772479772567749,
+      "learning_rate": 0.0003437457056780188,
+      "loss": 3.0379,
+      "step": 27756
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5816461443901062,
+      "learning_rate": 0.0003437304739955341,
+      "loss": 3.0917,
+      "step": 27757
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5699825882911682,
+      "learning_rate": 0.00034371524219787176,
+      "loss": 2.9859,
+      "step": 27758
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6363911032676697,
+      "learning_rate": 0.0003437000102850719,
+      "loss": 2.9362,
+      "step": 27759
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6590486168861389,
+      "learning_rate": 0.00034368477825717455,
+      "loss": 3.0905,
+      "step": 27760
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5966184139251709,
+      "learning_rate": 0.00034366954611421993,
+      "loss": 2.9768,
+      "step": 27761
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6021097302436829,
+      "learning_rate": 0.00034365431385624816,
+      "loss": 3.0687,
+      "step": 27762
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5527408719062805,
+      "learning_rate": 0.0003436390814832994,
+      "loss": 3.1802,
+      "step": 27763
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5687882900238037,
+      "learning_rate": 0.00034362384899541356,
+      "loss": 3.1737,
+      "step": 27764
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5569482445716858,
+      "learning_rate": 0.00034360861639263086,
+      "loss": 3.1958,
+      "step": 27765
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5489016771316528,
+      "learning_rate": 0.0003435933836749916,
+      "loss": 2.8973,
+      "step": 27766
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5932565927505493,
+      "learning_rate": 0.00034357815084253576,
+      "loss": 2.9053,
+      "step": 27767
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5763736963272095,
+      "learning_rate": 0.00034356291789530333,
+      "loss": 3.1687,
+      "step": 27768
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6087867021560669,
+      "learning_rate": 0.00034354768483333466,
+      "loss": 2.9277,
+      "step": 27769
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5887653231620789,
+      "learning_rate": 0.0003435324516566697,
+      "loss": 2.8634,
+      "step": 27770
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5681719183921814,
+      "learning_rate": 0.0003435172183653487,
+      "loss": 2.9306,
+      "step": 27771
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6251999735832214,
+      "learning_rate": 0.0003435019849594117,
+      "loss": 2.9546,
+      "step": 27772
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5517759919166565,
+      "learning_rate": 0.00034348675143889884,
+      "loss": 3.1553,
+      "step": 27773
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5633260607719421,
+      "learning_rate": 0.00034347151780385034,
+      "loss": 3.0941,
+      "step": 27774
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5699298977851868,
+      "learning_rate": 0.000343456284054306,
+      "loss": 3.0928,
+      "step": 27775
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5600057244300842,
+      "learning_rate": 0.0003434410501903063,
+      "loss": 3.3317,
+      "step": 27776
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5598301291465759,
+      "learning_rate": 0.0003434258162118913,
+      "loss": 3.1272,
+      "step": 27777
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6225312352180481,
+      "learning_rate": 0.000343410582119101,
+      "loss": 2.9665,
+      "step": 27778
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5822789072990417,
+      "learning_rate": 0.00034339534791197557,
+      "loss": 2.927,
+      "step": 27779
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6133136749267578,
+      "learning_rate": 0.00034338011359055516,
+      "loss": 3.1081,
+      "step": 27780
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.597339391708374,
+      "learning_rate": 0.0003433648791548799,
+      "loss": 3.1425,
+      "step": 27781
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.562778115272522,
+      "learning_rate": 0.00034334964460498984,
+      "loss": 3.2396,
+      "step": 27782
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5441209673881531,
+      "learning_rate": 0.0003433344099409252,
+      "loss": 3.2741,
+      "step": 27783
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5804208517074585,
+      "learning_rate": 0.00034331917516272604,
+      "loss": 3.0572,
+      "step": 27784
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6229684352874756,
+      "learning_rate": 0.0003433039402704326,
+      "loss": 2.9544,
+      "step": 27785
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5673180818557739,
+      "learning_rate": 0.0003432887052640848,
+      "loss": 2.9424,
+      "step": 27786
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5669776797294617,
+      "learning_rate": 0.0003432734701437229,
+      "loss": 3.0883,
+      "step": 27787
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5755397081375122,
+      "learning_rate": 0.000343258234909387,
+      "loss": 3.2678,
+      "step": 27788
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8632124662399292,
+      "learning_rate": 0.00034324299956111737,
+      "loss": 3.0894,
+      "step": 27789
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5921272039413452,
+      "learning_rate": 0.00034322776409895385,
+      "loss": 2.7846,
+      "step": 27790
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5782287120819092,
+      "learning_rate": 0.0003432125285229367,
+      "loss": 2.951,
+      "step": 27791
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5763784646987915,
+      "learning_rate": 0.0003431972928331061,
+      "loss": 2.9963,
+      "step": 27792
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6159095764160156,
+      "learning_rate": 0.0003431820570295022,
+      "loss": 2.9511,
+      "step": 27793
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5671179890632629,
+      "learning_rate": 0.00034316682111216513,
+      "loss": 2.9702,
+      "step": 27794
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6086155772209167,
+      "learning_rate": 0.00034315158508113483,
+      "loss": 2.9524,
+      "step": 27795
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6278063654899597,
+      "learning_rate": 0.0003431363489364516,
+      "loss": 2.9647,
+      "step": 27796
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5405550003051758,
+      "learning_rate": 0.0003431211126781555,
+      "loss": 3.1255,
+      "step": 27797
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5947132110595703,
+      "learning_rate": 0.0003431058763062867,
+      "loss": 3.0813,
+      "step": 27798
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6197540760040283,
+      "learning_rate": 0.00034309063982088534,
+      "loss": 3.1518,
+      "step": 27799
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5950043201446533,
+      "learning_rate": 0.00034307540322199163,
+      "loss": 3.0943,
+      "step": 27800
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5821655988693237,
+      "learning_rate": 0.00034306016650964544,
+      "loss": 3.0351,
+      "step": 27801
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6075488328933716,
+      "learning_rate": 0.00034304492968388703,
+      "loss": 3.1314,
+      "step": 27802
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5653616786003113,
+      "learning_rate": 0.00034302969274475675,
+      "loss": 2.9401,
+      "step": 27803
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.556991696357727,
+      "learning_rate": 0.00034301445569229444,
+      "loss": 2.8252,
+      "step": 27804
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5913997888565063,
+      "learning_rate": 0.00034299921852654037,
+      "loss": 3.0624,
+      "step": 27805
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0726978778839111,
+      "learning_rate": 0.00034298398124753455,
+      "loss": 3.1772,
+      "step": 27806
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5975397825241089,
+      "learning_rate": 0.0003429687438553172,
+      "loss": 2.8712,
+      "step": 27807
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5815263390541077,
+      "learning_rate": 0.0003429535063499285,
+      "loss": 3.0278,
+      "step": 27808
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.563348114490509,
+      "learning_rate": 0.0003429382687314085,
+      "loss": 2.9911,
+      "step": 27809
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6096580028533936,
+      "learning_rate": 0.00034292303099979737,
+      "loss": 2.9992,
+      "step": 27810
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5842347145080566,
+      "learning_rate": 0.00034290779315513525,
+      "loss": 2.8416,
+      "step": 27811
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.558156430721283,
+      "learning_rate": 0.00034289255519746225,
+      "loss": 2.9634,
+      "step": 27812
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6108639240264893,
+      "learning_rate": 0.0003428773171268185,
+      "loss": 3.1804,
+      "step": 27813
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6039309501647949,
+      "learning_rate": 0.0003428620789432443,
+      "loss": 3.0395,
+      "step": 27814
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5984408855438232,
+      "learning_rate": 0.0003428468406467794,
+      "loss": 3.1099,
+      "step": 27815
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5448001027107239,
+      "learning_rate": 0.0003428316022374642,
+      "loss": 3.2093,
+      "step": 27816
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.577616274356842,
+      "learning_rate": 0.00034281636371533895,
+      "loss": 3.1489,
+      "step": 27817
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5480098128318787,
+      "learning_rate": 0.0003428011250804436,
+      "loss": 3.0566,
+      "step": 27818
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5774540305137634,
+      "learning_rate": 0.00034278588633281825,
+      "loss": 3.0605,
+      "step": 27819
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5654169917106628,
+      "learning_rate": 0.00034277064747250313,
+      "loss": 2.8954,
+      "step": 27820
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6514195799827576,
+      "learning_rate": 0.0003427554084995384,
+      "loss": 3.0725,
+      "step": 27821
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5787948369979858,
+      "learning_rate": 0.0003427401694139641,
+      "loss": 3.034,
+      "step": 27822
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.59248948097229,
+      "learning_rate": 0.0003427249302158205,
+      "loss": 2.8765,
+      "step": 27823
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5247685313224792,
+      "learning_rate": 0.0003427096909051475,
+      "loss": 3.0719,
+      "step": 27824
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5795184969902039,
+      "learning_rate": 0.00034269445148198553,
+      "loss": 3.2857,
+      "step": 27825
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5434640049934387,
+      "learning_rate": 0.0003426792119463746,
+      "loss": 3.0622,
+      "step": 27826
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5672294497489929,
+      "learning_rate": 0.0003426639722983547,
+      "loss": 2.9796,
+      "step": 27827
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6290318369865417,
+      "learning_rate": 0.0003426487325379662,
+      "loss": 2.8113,
+      "step": 27828
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5443248152732849,
+      "learning_rate": 0.0003426334926652492,
+      "loss": 3.0352,
+      "step": 27829
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5471267700195312,
+      "learning_rate": 0.00034261825268024374,
+      "loss": 3.1233,
+      "step": 27830
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5816265940666199,
+      "learning_rate": 0.00034260301258299,
+      "loss": 3.1235,
+      "step": 27831
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5702730417251587,
+      "learning_rate": 0.00034258777237352814,
+      "loss": 3.0953,
+      "step": 27832
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5764862895011902,
+      "learning_rate": 0.0003425725320518982,
+      "loss": 3.0093,
+      "step": 27833
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5817887187004089,
+      "learning_rate": 0.0003425572916181405,
+      "loss": 3.0378,
+      "step": 27834
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5505737066268921,
+      "learning_rate": 0.00034254205107229504,
+      "loss": 3.0721,
+      "step": 27835
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6135760545730591,
+      "learning_rate": 0.0003425268104144021,
+      "loss": 2.9754,
+      "step": 27836
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5726092457771301,
+      "learning_rate": 0.00034251156964450153,
+      "loss": 3.2546,
+      "step": 27837
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5802931785583496,
+      "learning_rate": 0.0003424963287626338,
+      "loss": 2.9391,
+      "step": 27838
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5456535816192627,
+      "learning_rate": 0.0003424810877688389,
+      "loss": 3.1814,
+      "step": 27839
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5868333578109741,
+      "learning_rate": 0.00034246584666315703,
+      "loss": 3.0369,
+      "step": 27840
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6372731924057007,
+      "learning_rate": 0.00034245060544562826,
+      "loss": 2.9961,
+      "step": 27841
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5847151279449463,
+      "learning_rate": 0.00034243536411629266,
+      "loss": 3.073,
+      "step": 27842
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5791596174240112,
+      "learning_rate": 0.0003424201226751906,
+      "loss": 3.0307,
+      "step": 27843
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5760388970375061,
+      "learning_rate": 0.0003424048811223621,
+      "loss": 2.9906,
+      "step": 27844
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5581043362617493,
+      "learning_rate": 0.0003423896394578473,
+      "loss": 3.1661,
+      "step": 27845
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6014453172683716,
+      "learning_rate": 0.0003423743976816863,
+      "loss": 3.0195,
+      "step": 27846
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6490236520767212,
+      "learning_rate": 0.0003423591557939193,
+      "loss": 2.8944,
+      "step": 27847
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5952842831611633,
+      "learning_rate": 0.00034234391379458655,
+      "loss": 3.0426,
+      "step": 27848
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5552687644958496,
+      "learning_rate": 0.00034232867168372793,
+      "loss": 2.9578,
+      "step": 27849
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5637187361717224,
+      "learning_rate": 0.0003423134294613838,
+      "loss": 3.1498,
+      "step": 27850
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5724777579307556,
+      "learning_rate": 0.00034229818712759423,
+      "loss": 3.0703,
+      "step": 27851
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6271558403968811,
+      "learning_rate": 0.0003422829446823994,
+      "loss": 3.046,
+      "step": 27852
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5826039910316467,
+      "learning_rate": 0.00034226770212583934,
+      "loss": 2.9899,
+      "step": 27853
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5479075312614441,
+      "learning_rate": 0.0003422524594579544,
+      "loss": 2.9437,
+      "step": 27854
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.572819709777832,
+      "learning_rate": 0.0003422372166787846,
+      "loss": 3.0236,
+      "step": 27855
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5536037683486938,
+      "learning_rate": 0.00034222197378837006,
+      "loss": 3.1475,
+      "step": 27856
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6179131865501404,
+      "learning_rate": 0.00034220673078675095,
+      "loss": 2.9968,
+      "step": 27857
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.562525749206543,
+      "learning_rate": 0.00034219148767396744,
+      "loss": 3.0093,
+      "step": 27858
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5559420585632324,
+      "learning_rate": 0.00034217624445005967,
+      "loss": 3.0252,
+      "step": 27859
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5643912553787231,
+      "learning_rate": 0.00034216100111506784,
+      "loss": 3.1213,
+      "step": 27860
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5565373301506042,
+      "learning_rate": 0.00034214575766903203,
+      "loss": 3.0101,
+      "step": 27861
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5774194002151489,
+      "learning_rate": 0.00034213051411199246,
+      "loss": 2.9123,
+      "step": 27862
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5667964220046997,
+      "learning_rate": 0.000342115270443989,
+      "loss": 3.2832,
+      "step": 27863
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5750254988670349,
+      "learning_rate": 0.0003421000266650622,
+      "loss": 3.1505,
+      "step": 27864
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6452105641365051,
+      "learning_rate": 0.0003420847827752521,
+      "loss": 3.0935,
+      "step": 27865
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5923091173171997,
+      "learning_rate": 0.0003420695387745987,
+      "loss": 2.9655,
+      "step": 27866
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5900555849075317,
+      "learning_rate": 0.0003420542946631422,
+      "loss": 2.9454,
+      "step": 27867
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5830941200256348,
+      "learning_rate": 0.00034203905044092274,
+      "loss": 3.0925,
+      "step": 27868
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5814031362533569,
+      "learning_rate": 0.00034202380610798066,
+      "loss": 3.0646,
+      "step": 27869
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5986719131469727,
+      "learning_rate": 0.0003420085616643558,
+      "loss": 2.8448,
+      "step": 27870
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5420025587081909,
+      "learning_rate": 0.00034199331711008864,
+      "loss": 2.9547,
+      "step": 27871
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5797125101089478,
+      "learning_rate": 0.00034197807244521904,
+      "loss": 3.1777,
+      "step": 27872
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5667112469673157,
+      "learning_rate": 0.00034196282766978727,
+      "loss": 3.1301,
+      "step": 27873
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5843361616134644,
+      "learning_rate": 0.0003419475827838335,
+      "loss": 2.8819,
+      "step": 27874
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5419958829879761,
+      "learning_rate": 0.00034193233778739797,
+      "loss": 2.8901,
+      "step": 27875
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5321311354637146,
+      "learning_rate": 0.0003419170926805207,
+      "loss": 2.8699,
+      "step": 27876
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6508222818374634,
+      "learning_rate": 0.0003419018474632418,
+      "loss": 3.2209,
+      "step": 27877
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6090697646141052,
+      "learning_rate": 0.0003418866021356015,
+      "loss": 3.0021,
+      "step": 27878
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5967793464660645,
+      "learning_rate": 0.00034187135669764,
+      "loss": 2.9383,
+      "step": 27879
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5427808165550232,
+      "learning_rate": 0.00034185611114939744,
+      "loss": 3.1093,
+      "step": 27880
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.601786732673645,
+      "learning_rate": 0.0003418408654909139,
+      "loss": 3.0333,
+      "step": 27881
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6144266724586487,
+      "learning_rate": 0.0003418256197222295,
+      "loss": 3.0911,
+      "step": 27882
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5672413110733032,
+      "learning_rate": 0.00034181037384338453,
+      "loss": 3.0872,
+      "step": 27883
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6033337712287903,
+      "learning_rate": 0.0003417951278544191,
+      "loss": 3.2807,
+      "step": 27884
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5832936763763428,
+      "learning_rate": 0.0003417798817553733,
+      "loss": 2.9746,
+      "step": 27885
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5509626269340515,
+      "learning_rate": 0.0003417646355462874,
+      "loss": 3.0896,
+      "step": 27886
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6345809698104858,
+      "learning_rate": 0.0003417493892272015,
+      "loss": 2.9941,
+      "step": 27887
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5827943682670593,
+      "learning_rate": 0.00034173414279815563,
+      "loss": 3.0285,
+      "step": 27888
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5544496774673462,
+      "learning_rate": 0.0003417188962591901,
+      "loss": 3.0064,
+      "step": 27889
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5938286185264587,
+      "learning_rate": 0.00034170364961034507,
+      "loss": 2.9717,
+      "step": 27890
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5810260772705078,
+      "learning_rate": 0.00034168840285166065,
+      "loss": 2.9629,
+      "step": 27891
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6095471978187561,
+      "learning_rate": 0.00034167315598317697,
+      "loss": 3.1627,
+      "step": 27892
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5932841897010803,
+      "learning_rate": 0.00034165790900493413,
+      "loss": 2.9353,
+      "step": 27893
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6483215093612671,
+      "learning_rate": 0.00034164266191697253,
+      "loss": 2.9442,
+      "step": 27894
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5468027591705322,
+      "learning_rate": 0.00034162741471933217,
+      "loss": 3.1029,
+      "step": 27895
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5780773758888245,
+      "learning_rate": 0.00034161216741205317,
+      "loss": 3.1393,
+      "step": 27896
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6147868037223816,
+      "learning_rate": 0.0003415969199951757,
+      "loss": 2.9525,
+      "step": 27897
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6018847227096558,
+      "learning_rate": 0.00034158167246874,
+      "loss": 3.2364,
+      "step": 27898
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6293785572052002,
+      "learning_rate": 0.0003415664248327861,
+      "loss": 3.0502,
+      "step": 27899
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5687551498413086,
+      "learning_rate": 0.0003415511770873542,
+      "loss": 2.9601,
+      "step": 27900
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5848996639251709,
+      "learning_rate": 0.00034153592923248465,
+      "loss": 2.9939,
+      "step": 27901
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5531403422355652,
+      "learning_rate": 0.00034152068126821745,
+      "loss": 2.9455,
+      "step": 27902
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6372696757316589,
+      "learning_rate": 0.0003415054331945927,
+      "loss": 3.0152,
+      "step": 27903
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6175010204315186,
+      "learning_rate": 0.00034149018501165054,
+      "loss": 3.0507,
+      "step": 27904
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5934361815452576,
+      "learning_rate": 0.00034147493671943133,
+      "loss": 3.14,
+      "step": 27905
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5523131489753723,
+      "learning_rate": 0.00034145968831797506,
+      "loss": 3.2342,
+      "step": 27906
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.595689594745636,
+      "learning_rate": 0.000341444439807322,
+      "loss": 2.9376,
+      "step": 27907
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5672516822814941,
+      "learning_rate": 0.0003414291911875122,
+      "loss": 3.2794,
+      "step": 27908
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6107340455055237,
+      "learning_rate": 0.000341413942458586,
+      "loss": 3.1038,
+      "step": 27909
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6264364719390869,
+      "learning_rate": 0.0003413986936205833,
+      "loss": 2.9496,
+      "step": 27910
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6131812334060669,
+      "learning_rate": 0.00034138344467354447,
+      "loss": 2.895,
+      "step": 27911
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5950072407722473,
+      "learning_rate": 0.00034136819561750965,
+      "loss": 3.083,
+      "step": 27912
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5509093403816223,
+      "learning_rate": 0.00034135294645251896,
+      "loss": 3.0271,
+      "step": 27913
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5661293864250183,
+      "learning_rate": 0.0003413376971786125,
+      "loss": 3.1161,
+      "step": 27914
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5739542841911316,
+      "learning_rate": 0.0003413224477958306,
+      "loss": 3.0398,
+      "step": 27915
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5414571166038513,
+      "learning_rate": 0.00034130719830421333,
+      "loss": 3.0356,
+      "step": 27916
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5896730422973633,
+      "learning_rate": 0.00034129194870380076,
+      "loss": 3.0442,
+      "step": 27917
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.606563925743103,
+      "learning_rate": 0.0003412766989946332,
+      "loss": 3.0527,
+      "step": 27918
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5917063355445862,
+      "learning_rate": 0.0003412614491767507,
+      "loss": 3.0547,
+      "step": 27919
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5469411015510559,
+      "learning_rate": 0.00034124619925019346,
+      "loss": 3.4306,
+      "step": 27920
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5653911828994751,
+      "learning_rate": 0.0003412309492150018,
+      "loss": 3.1284,
+      "step": 27921
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5671815872192383,
+      "learning_rate": 0.00034121569907121565,
+      "loss": 3.1207,
+      "step": 27922
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6383528709411621,
+      "learning_rate": 0.0003412004488188753,
+      "loss": 2.8917,
+      "step": 27923
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5371748805046082,
+      "learning_rate": 0.0003411851984580209,
+      "loss": 3.0595,
+      "step": 27924
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5818402767181396,
+      "learning_rate": 0.0003411699479886926,
+      "loss": 3.001,
+      "step": 27925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6108590960502625,
+      "learning_rate": 0.0003411546974109306,
+      "loss": 3.0168,
+      "step": 27926
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5887522101402283,
+      "learning_rate": 0.0003411394467247751,
+      "loss": 3.0405,
+      "step": 27927
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5705615878105164,
+      "learning_rate": 0.00034112419593026615,
+      "loss": 3.0193,
+      "step": 27928
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5551228523254395,
+      "learning_rate": 0.0003411089450274439,
+      "loss": 3.1167,
+      "step": 27929
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.567327082157135,
+      "learning_rate": 0.0003410936940163487,
+      "loss": 2.9874,
+      "step": 27930
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5707133412361145,
+      "learning_rate": 0.0003410784428970207,
+      "loss": 3.0144,
+      "step": 27931
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5892231464385986,
+      "learning_rate": 0.0003410631916694999,
+      "loss": 3.1033,
+      "step": 27932
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6045179963111877,
+      "learning_rate": 0.00034104794033382657,
+      "loss": 2.9812,
+      "step": 27933
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5541436076164246,
+      "learning_rate": 0.0003410326888900408,
+      "loss": 3.1712,
+      "step": 27934
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5599698424339294,
+      "learning_rate": 0.0003410174373381829,
+      "loss": 3.0917,
+      "step": 27935
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5694020986557007,
+      "learning_rate": 0.0003410021856782929,
+      "loss": 3.1889,
+      "step": 27936
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5983900427818298,
+      "learning_rate": 0.00034098693391041107,
+      "loss": 2.9114,
+      "step": 27937
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6155137419700623,
+      "learning_rate": 0.0003409716820345776,
+      "loss": 3.0607,
+      "step": 27938
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6034281849861145,
+      "learning_rate": 0.00034095643005083257,
+      "loss": 2.961,
+      "step": 27939
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5892189145088196,
+      "learning_rate": 0.0003409411779592161,
+      "loss": 3.244,
+      "step": 27940
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5782514214515686,
+      "learning_rate": 0.00034092592575976855,
+      "loss": 3.2606,
+      "step": 27941
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5579574108123779,
+      "learning_rate": 0.00034091067345253,
+      "loss": 2.9648,
+      "step": 27942
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5963599681854248,
+      "learning_rate": 0.0003408954210375405,
+      "loss": 3.0709,
+      "step": 27943
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6034666299819946,
+      "learning_rate": 0.00034088016851484043,
+      "loss": 2.8616,
+      "step": 27944
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6475614309310913,
+      "learning_rate": 0.00034086491588446985,
+      "loss": 3.0576,
+      "step": 27945
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5820233821868896,
+      "learning_rate": 0.0003408496631464689,
+      "loss": 2.9848,
+      "step": 27946
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5455297827720642,
+      "learning_rate": 0.00034083441030087784,
+      "loss": 3.0637,
+      "step": 27947
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.568981945514679,
+      "learning_rate": 0.00034081915734773675,
+      "loss": 2.9338,
+      "step": 27948
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5847499370574951,
+      "learning_rate": 0.00034080390428708595,
+      "loss": 2.8865,
+      "step": 27949
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6259151697158813,
+      "learning_rate": 0.0003407886511189654,
+      "loss": 3.0548,
+      "step": 27950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5772634744644165,
+      "learning_rate": 0.00034077339784341544,
+      "loss": 3.0336,
+      "step": 27951
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5600981712341309,
+      "learning_rate": 0.00034075814446047617,
+      "loss": 2.9458,
+      "step": 27952
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5539212226867676,
+      "learning_rate": 0.0003407428909701879,
+      "loss": 3.0916,
+      "step": 27953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.598659336566925,
+      "learning_rate": 0.0003407276373725906,
+      "loss": 3.18,
+      "step": 27954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5764479637145996,
+      "learning_rate": 0.00034071238366772456,
+      "loss": 3.0754,
+      "step": 27955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5627712607383728,
+      "learning_rate": 0.00034069712985563,
+      "loss": 2.9324,
+      "step": 27956
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6134984493255615,
+      "learning_rate": 0.00034068187593634696,
+      "loss": 2.9614,
+      "step": 27957
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.581798255443573,
+      "learning_rate": 0.0003406666219099157,
+      "loss": 2.8332,
+      "step": 27958
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5963696837425232,
+      "learning_rate": 0.0003406513677763764,
+      "loss": 2.9447,
+      "step": 27959
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5944687128067017,
+      "learning_rate": 0.0003406361135357692,
+      "loss": 2.9688,
+      "step": 27960
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6353680491447449,
+      "learning_rate": 0.00034062085918813435,
+      "loss": 2.8498,
+      "step": 27961
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5559459328651428,
+      "learning_rate": 0.00034060560473351187,
+      "loss": 2.8492,
+      "step": 27962
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5717218518257141,
+      "learning_rate": 0.00034059035017194216,
+      "loss": 2.782,
+      "step": 27963
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5911750197410583,
+      "learning_rate": 0.00034057509550346524,
+      "loss": 3.1497,
+      "step": 27964
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5874834060668945,
+      "learning_rate": 0.0003405598407281212,
+      "loss": 2.923,
+      "step": 27965
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5998968482017517,
+      "learning_rate": 0.0003405445858459505,
+      "loss": 2.8489,
+      "step": 27966
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.576008677482605,
+      "learning_rate": 0.0003405293308569932,
+      "loss": 2.9175,
+      "step": 27967
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5811209082603455,
+      "learning_rate": 0.0003405140757612893,
+      "loss": 3.0114,
+      "step": 27968
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6185311675071716,
+      "learning_rate": 0.0003404988205588792,
+      "loss": 3.1087,
+      "step": 27969
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6029673218727112,
+      "learning_rate": 0.000340483565249803,
+      "loss": 3.0329,
+      "step": 27970
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6005308628082275,
+      "learning_rate": 0.0003404683098341009,
+      "loss": 3.1651,
+      "step": 27971
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5758666396141052,
+      "learning_rate": 0.00034045305431181297,
+      "loss": 3.0992,
+      "step": 27972
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.546139657497406,
+      "learning_rate": 0.00034043779868297953,
+      "loss": 2.9213,
+      "step": 27973
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5436400771141052,
+      "learning_rate": 0.00034042254294764073,
+      "loss": 3.0611,
+      "step": 27974
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5803444385528564,
+      "learning_rate": 0.00034040728710583665,
+      "loss": 3.0152,
+      "step": 27975
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5593194961547852,
+      "learning_rate": 0.0003403920311576077,
+      "loss": 3.0609,
+      "step": 27976
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5657615661621094,
+      "learning_rate": 0.0003403767751029938,
+      "loss": 3.1276,
+      "step": 27977
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5875985622406006,
+      "learning_rate": 0.00034036151894203536,
+      "loss": 3.0625,
+      "step": 27978
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.55743807554245,
+      "learning_rate": 0.00034034626267477234,
+      "loss": 3.1621,
+      "step": 27979
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6265092492103577,
+      "learning_rate": 0.00034033100630124504,
+      "loss": 2.9605,
+      "step": 27980
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5701452493667603,
+      "learning_rate": 0.00034031574982149363,
+      "loss": 3.0356,
+      "step": 27981
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5367501974105835,
+      "learning_rate": 0.00034030049323555833,
+      "loss": 3.0456,
+      "step": 27982
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5916096568107605,
+      "learning_rate": 0.00034028523654347926,
+      "loss": 3.2683,
+      "step": 27983
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5602060556411743,
+      "learning_rate": 0.00034026997974529664,
+      "loss": 3.2333,
+      "step": 27984
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.542546272277832,
+      "learning_rate": 0.00034025472284105065,
+      "loss": 2.9991,
+      "step": 27985
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5879301428794861,
+      "learning_rate": 0.0003402394658307814,
+      "loss": 2.8912,
+      "step": 27986
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5856759548187256,
+      "learning_rate": 0.00034022420871452925,
+      "loss": 2.9014,
+      "step": 27987
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5833997130393982,
+      "learning_rate": 0.0003402089514923342,
+      "loss": 2.9192,
+      "step": 27988
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5771424770355225,
+      "learning_rate": 0.0003401936941642366,
+      "loss": 3.1536,
+      "step": 27989
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5892386436462402,
+      "learning_rate": 0.00034017843673027646,
+      "loss": 2.9945,
+      "step": 27990
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6291250586509705,
+      "learning_rate": 0.00034016317919049403,
+      "loss": 2.9909,
+      "step": 27991
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5596926212310791,
+      "learning_rate": 0.00034014792154492957,
+      "loss": 3.1103,
+      "step": 27992
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5595067739486694,
+      "learning_rate": 0.00034013266379362325,
+      "loss": 3.0245,
+      "step": 27993
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5994234085083008,
+      "learning_rate": 0.00034011740593661513,
+      "loss": 3.1417,
+      "step": 27994
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.578930675983429,
+      "learning_rate": 0.00034010214797394555,
+      "loss": 3.1635,
+      "step": 27995
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6194568872451782,
+      "learning_rate": 0.00034008688990565456,
+      "loss": 2.962,
+      "step": 27996
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5864658951759338,
+      "learning_rate": 0.00034007163173178245,
+      "loss": 3.2179,
+      "step": 27997
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5644214153289795,
+      "learning_rate": 0.0003400563734523694,
+      "loss": 2.9034,
+      "step": 27998
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5568000674247742,
+      "learning_rate": 0.00034004111506745553,
+      "loss": 3.0852,
+      "step": 27999
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5540627241134644,
+      "learning_rate": 0.0003400258565770811,
+      "loss": 3.0086,
+      "step": 28000
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5743499994277954,
+      "learning_rate": 0.0003400105979812862,
+      "loss": 2.6773,
+      "step": 28001
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5981676578521729,
+      "learning_rate": 0.0003399953392801111,
+      "loss": 3.0462,
+      "step": 28002
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5800593495368958,
+      "learning_rate": 0.00033998008047359603,
+      "loss": 3.116,
+      "step": 28003
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.590241014957428,
+      "learning_rate": 0.0003399648215617812,
+      "loss": 3.228,
+      "step": 28004
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5530853271484375,
+      "learning_rate": 0.0003399495625447066,
+      "loss": 3.315,
+      "step": 28005
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5522137880325317,
+      "learning_rate": 0.00033993430342241254,
+      "loss": 3.0778,
+      "step": 28006
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5803789496421814,
+      "learning_rate": 0.00033991904419493926,
+      "loss": 3.1161,
+      "step": 28007
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5703111886978149,
+      "learning_rate": 0.00033990378486232684,
+      "loss": 3.0318,
+      "step": 28008
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5972906351089478,
+      "learning_rate": 0.0003398885254246156,
+      "loss": 3.0836,
+      "step": 28009
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6131359338760376,
+      "learning_rate": 0.00033987326588184565,
+      "loss": 3.1311,
+      "step": 28010
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5489694476127625,
+      "learning_rate": 0.00033985800623405717,
+      "loss": 3.1116,
+      "step": 28011
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.552985668182373,
+      "learning_rate": 0.0003398427464812904,
+      "loss": 3.0374,
+      "step": 28012
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.581065833568573,
+      "learning_rate": 0.00033982748662358546,
+      "loss": 2.8298,
+      "step": 28013
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5935637354850769,
+      "learning_rate": 0.00033981222666098256,
+      "loss": 3.1071,
+      "step": 28014
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5556374788284302,
+      "learning_rate": 0.00033979696659352203,
+      "loss": 3.204,
+      "step": 28015
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5717475414276123,
+      "learning_rate": 0.00033978170642124387,
+      "loss": 2.7786,
+      "step": 28016
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5912850499153137,
+      "learning_rate": 0.0003397664461441883,
+      "loss": 3.1273,
+      "step": 28017
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5778431296348572,
+      "learning_rate": 0.00033975118576239565,
+      "loss": 2.9896,
+      "step": 28018
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5879780054092407,
+      "learning_rate": 0.000339735925275906,
+      "loss": 3.0575,
+      "step": 28019
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5398730635643005,
+      "learning_rate": 0.00033972066468475955,
+      "loss": 3.0526,
+      "step": 28020
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6177681088447571,
+      "learning_rate": 0.00033970540398899654,
+      "loss": 3.0964,
+      "step": 28021
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5534880757331848,
+      "learning_rate": 0.00033969014318865713,
+      "loss": 3.0763,
+      "step": 28022
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5374436378479004,
+      "learning_rate": 0.0003396748822837815,
+      "loss": 3.2786,
+      "step": 28023
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5656604170799255,
+      "learning_rate": 0.0003396596212744099,
+      "loss": 3.1471,
+      "step": 28024
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5550830960273743,
+      "learning_rate": 0.0003396443601605825,
+      "loss": 3.0216,
+      "step": 28025
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5797797441482544,
+      "learning_rate": 0.0003396290989423395,
+      "loss": 2.9703,
+      "step": 28026
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5749794244766235,
+      "learning_rate": 0.000339613837619721,
+      "loss": 3.1812,
+      "step": 28027
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6089553833007812,
+      "learning_rate": 0.0003395985761927673,
+      "loss": 3.0538,
+      "step": 28028
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5771617889404297,
+      "learning_rate": 0.00033958331466151865,
+      "loss": 3.1415,
+      "step": 28029
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.602817952632904,
+      "learning_rate": 0.0003395680530260151,
+      "loss": 3.1378,
+      "step": 28030
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.597654402256012,
+      "learning_rate": 0.000339552791286297,
+      "loss": 3.1961,
+      "step": 28031
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6101335287094116,
+      "learning_rate": 0.0003395375294424043,
+      "loss": 2.981,
+      "step": 28032
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5998036861419678,
+      "learning_rate": 0.0003395222674943775,
+      "loss": 3.0004,
+      "step": 28033
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9481212496757507,
+      "learning_rate": 0.0003395070054422566,
+      "loss": 3.0037,
+      "step": 28034
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5774849653244019,
+      "learning_rate": 0.00033949174328608175,
+      "loss": 3.4101,
+      "step": 28035
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.608545184135437,
+      "learning_rate": 0.0003394764810258934,
+      "loss": 3.1483,
+      "step": 28036
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5682400465011597,
+      "learning_rate": 0.00033946121866173146,
+      "loss": 3.0725,
+      "step": 28037
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.634087085723877,
+      "learning_rate": 0.0003394459561936364,
+      "loss": 3.0956,
+      "step": 28038
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5982432961463928,
+      "learning_rate": 0.00033943069362164823,
+      "loss": 3.0107,
+      "step": 28039
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5551794171333313,
+      "learning_rate": 0.0003394154309458072,
+      "loss": 2.9166,
+      "step": 28040
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.594632625579834,
+      "learning_rate": 0.00033940016816615354,
+      "loss": 2.9562,
+      "step": 28041
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5509760975837708,
+      "learning_rate": 0.0003393849052827273,
+      "loss": 3.1103,
+      "step": 28042
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5657804012298584,
+      "learning_rate": 0.0003393696422955689,
+      "loss": 3.1675,
+      "step": 28043
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5348253846168518,
+      "learning_rate": 0.0003393543792047185,
+      "loss": 3.0305,
+      "step": 28044
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.572263240814209,
+      "learning_rate": 0.0003393391160102161,
+      "loss": 3.1031,
+      "step": 28045
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.581788182258606,
+      "learning_rate": 0.00033932385271210214,
+      "loss": 2.9799,
+      "step": 28046
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5556021928787231,
+      "learning_rate": 0.0003393085893104167,
+      "loss": 3.1716,
+      "step": 28047
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5477098226547241,
+      "learning_rate": 0.00033929332580519994,
+      "loss": 3.0578,
+      "step": 28048
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6444635987281799,
+      "learning_rate": 0.0003392780621964922,
+      "loss": 3.1512,
+      "step": 28049
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5834984183311462,
+      "learning_rate": 0.0003392627984843335,
+      "loss": 3.3345,
+      "step": 28050
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5475103259086609,
+      "learning_rate": 0.00033924753466876427,
+      "loss": 3.094,
+      "step": 28051
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5777970552444458,
+      "learning_rate": 0.0003392322707498245,
+      "loss": 3.1841,
+      "step": 28052
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5672211647033691,
+      "learning_rate": 0.00033921700672755455,
+      "loss": 3.0541,
+      "step": 28053
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7378207445144653,
+      "learning_rate": 0.0003392017426019945,
+      "loss": 3.0522,
+      "step": 28054
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5990445017814636,
+      "learning_rate": 0.0003391864783731846,
+      "loss": 3.0336,
+      "step": 28055
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5967203974723816,
+      "learning_rate": 0.00033917121404116515,
+      "loss": 2.9228,
+      "step": 28056
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5567073225975037,
+      "learning_rate": 0.00033915594960597616,
+      "loss": 3.0333,
+      "step": 28057
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5595899820327759,
+      "learning_rate": 0.0003391406850676579,
+      "loss": 3.093,
+      "step": 28058
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5615085363388062,
+      "learning_rate": 0.0003391254204262507,
+      "loss": 2.9224,
+      "step": 28059
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5954691767692566,
+      "learning_rate": 0.00033911015568179466,
+      "loss": 2.9878,
+      "step": 28060
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.557086169719696,
+      "learning_rate": 0.00033909489083433,
+      "loss": 3.1065,
+      "step": 28061
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5938393473625183,
+      "learning_rate": 0.0003390796258838969,
+      "loss": 2.859,
+      "step": 28062
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5960376262664795,
+      "learning_rate": 0.00033906436083053557,
+      "loss": 3.1894,
+      "step": 28063
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5832287669181824,
+      "learning_rate": 0.0003390490956742862,
+      "loss": 2.9972,
+      "step": 28064
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8595984578132629,
+      "learning_rate": 0.0003390338304151891,
+      "loss": 3.2039,
+      "step": 28065
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5543466806411743,
+      "learning_rate": 0.0003390185650532844,
+      "loss": 3.0077,
+      "step": 28066
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6111711263656616,
+      "learning_rate": 0.00033900329958861227,
+      "loss": 2.9772,
+      "step": 28067
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5975467562675476,
+      "learning_rate": 0.0003389880340212129,
+      "loss": 3.1056,
+      "step": 28068
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5642098784446716,
+      "learning_rate": 0.00033897276835112675,
+      "loss": 2.9942,
+      "step": 28069
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5549335479736328,
+      "learning_rate": 0.0003389575025783937,
+      "loss": 3.0474,
+      "step": 28070
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6323215961456299,
+      "learning_rate": 0.00033894223670305403,
+      "loss": 2.9807,
+      "step": 28071
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5654608607292175,
+      "learning_rate": 0.000338926970725148,
+      "loss": 3.0737,
+      "step": 28072
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.574693500995636,
+      "learning_rate": 0.00033891170464471596,
+      "loss": 2.8669,
+      "step": 28073
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5713275671005249,
+      "learning_rate": 0.00033889643846179784,
+      "loss": 3.2703,
+      "step": 28074
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6364814639091492,
+      "learning_rate": 0.00033888117217643405,
+      "loss": 3.1299,
+      "step": 28075
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5909932255744934,
+      "learning_rate": 0.0003388659057886648,
+      "loss": 3.0662,
+      "step": 28076
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5821861624717712,
+      "learning_rate": 0.0003388506392985302,
+      "loss": 3.0567,
+      "step": 28077
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5835837721824646,
+      "learning_rate": 0.0003388353727060704,
+      "loss": 3.1389,
+      "step": 28078
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5665825605392456,
+      "learning_rate": 0.00033882010601132573,
+      "loss": 3.3765,
+      "step": 28079
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5594505071640015,
+      "learning_rate": 0.0003388048392143365,
+      "loss": 3.0203,
+      "step": 28080
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5456429123878479,
+      "learning_rate": 0.0003387895723151427,
+      "loss": 3.107,
+      "step": 28081
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.571052610874176,
+      "learning_rate": 0.0003387743053137846,
+      "loss": 3.1842,
+      "step": 28082
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6176466941833496,
+      "learning_rate": 0.0003387590382103025,
+      "loss": 2.9045,
+      "step": 28083
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5826286673545837,
+      "learning_rate": 0.00033874377100473653,
+      "loss": 2.9922,
+      "step": 28084
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5976090431213379,
+      "learning_rate": 0.0003387285036971268,
+      "loss": 2.9568,
+      "step": 28085
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5784058570861816,
+      "learning_rate": 0.0003387132362875138,
+      "loss": 2.9194,
+      "step": 28086
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5768011212348938,
+      "learning_rate": 0.00033869796877593756,
+      "loss": 3.0978,
+      "step": 28087
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.563029944896698,
+      "learning_rate": 0.0003386827011624383,
+      "loss": 3.0959,
+      "step": 28088
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5373599529266357,
+      "learning_rate": 0.00033866743344705626,
+      "loss": 3.0315,
+      "step": 28089
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5780729055404663,
+      "learning_rate": 0.0003386521656298316,
+      "loss": 3.013,
+      "step": 28090
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5696938037872314,
+      "learning_rate": 0.0003386368977108047,
+      "loss": 2.8036,
+      "step": 28091
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.55682373046875,
+      "learning_rate": 0.0003386216296900155,
+      "loss": 3.0036,
+      "step": 28092
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6155198812484741,
+      "learning_rate": 0.0003386063615675045,
+      "loss": 2.979,
+      "step": 28093
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5788460373878479,
+      "learning_rate": 0.00033859109334331166,
+      "loss": 3.1754,
+      "step": 28094
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6345138549804688,
+      "learning_rate": 0.0003385758250174774,
+      "loss": 3.0151,
+      "step": 28095
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5743170380592346,
+      "learning_rate": 0.0003385605565900417,
+      "loss": 3.0853,
+      "step": 28096
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6293991208076477,
+      "learning_rate": 0.00033854528806104495,
+      "loss": 2.999,
+      "step": 28097
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6683945655822754,
+      "learning_rate": 0.00033853001943052736,
+      "loss": 3.2133,
+      "step": 28098
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5892733335494995,
+      "learning_rate": 0.00033851475069852916,
+      "loss": 2.9365,
+      "step": 28099
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5764167904853821,
+      "learning_rate": 0.0003384994818650905,
+      "loss": 3.0822,
+      "step": 28100
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5932957530021667,
+      "learning_rate": 0.00033848421293025156,
+      "loss": 3.2123,
+      "step": 28101
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5980968475341797,
+      "learning_rate": 0.0003384689438940528,
+      "loss": 2.9073,
+      "step": 28102
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5964227318763733,
+      "learning_rate": 0.0003384536747565341,
+      "loss": 2.7695,
+      "step": 28103
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5735576152801514,
+      "learning_rate": 0.00033843840551773564,
+      "loss": 3.0268,
+      "step": 28104
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6148794889450073,
+      "learning_rate": 0.000338423136177698,
+      "loss": 2.8976,
+      "step": 28105
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6786152124404907,
+      "learning_rate": 0.00033840786673646134,
+      "loss": 3.1084,
+      "step": 28106
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.748644232749939,
+      "learning_rate": 0.00033839259719406554,
+      "loss": 3.0318,
+      "step": 28107
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5767213106155396,
+      "learning_rate": 0.00033837732755055114,
+      "loss": 2.9741,
+      "step": 28108
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6026296019554138,
+      "learning_rate": 0.0003383620578059582,
+      "loss": 3.0951,
+      "step": 28109
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5640915036201477,
+      "learning_rate": 0.000338346787960327,
+      "loss": 3.0876,
+      "step": 28110
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5978099703788757,
+      "learning_rate": 0.00033833151801369773,
+      "loss": 3.0631,
+      "step": 28111
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5457133054733276,
+      "learning_rate": 0.0003383162479661106,
+      "loss": 3.207,
+      "step": 28112
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5326697826385498,
+      "learning_rate": 0.00033830097781760595,
+      "loss": 3.086,
+      "step": 28113
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5941699147224426,
+      "learning_rate": 0.0003382857075682237,
+      "loss": 3.1537,
+      "step": 28114
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5777954459190369,
+      "learning_rate": 0.0003382704372180044,
+      "loss": 3.0681,
+      "step": 28115
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5638835430145264,
+      "learning_rate": 0.00033825516676698816,
+      "loss": 3.0887,
+      "step": 28116
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5932229161262512,
+      "learning_rate": 0.0003382398962152152,
+      "loss": 2.9348,
+      "step": 28117
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.566215991973877,
+      "learning_rate": 0.0003382246255627256,
+      "loss": 3.0488,
+      "step": 28118
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5923061370849609,
+      "learning_rate": 0.0003382093548095597,
+      "loss": 3.053,
+      "step": 28119
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5887033939361572,
+      "learning_rate": 0.0003381940839557578,
+      "loss": 3.1721,
+      "step": 28120
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6126934289932251,
+      "learning_rate": 0.00033817881300136,
+      "loss": 3.0277,
+      "step": 28121
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5494149923324585,
+      "learning_rate": 0.0003381635419464065,
+      "loss": 3.209,
+      "step": 28122
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5944859385490417,
+      "learning_rate": 0.00033814827079093755,
+      "loss": 3.0693,
+      "step": 28123
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5820398926734924,
+      "learning_rate": 0.0003381329995349935,
+      "loss": 3.2692,
+      "step": 28124
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.580773651599884,
+      "learning_rate": 0.00033811772817861444,
+      "loss": 3.0431,
+      "step": 28125
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5761692523956299,
+      "learning_rate": 0.00033810245672184053,
+      "loss": 3.0927,
+      "step": 28126
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5915241837501526,
+      "learning_rate": 0.00033808718516471217,
+      "loss": 3.1483,
+      "step": 28127
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5700193047523499,
+      "learning_rate": 0.00033807191350726957,
+      "loss": 3.0007,
+      "step": 28128
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.596190333366394,
+      "learning_rate": 0.00033805664174955274,
+      "loss": 2.9628,
+      "step": 28129
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5701046586036682,
+      "learning_rate": 0.0003380413698916021,
+      "loss": 3.1288,
+      "step": 28130
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5389936566352844,
+      "learning_rate": 0.00033802609793345784,
+      "loss": 3.0459,
+      "step": 28131
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5747068524360657,
+      "learning_rate": 0.0003380108258751601,
+      "loss": 2.9633,
+      "step": 28132
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.606648862361908,
+      "learning_rate": 0.00033799555371674916,
+      "loss": 2.9476,
+      "step": 28133
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6043571829795837,
+      "learning_rate": 0.00033798028145826526,
+      "loss": 3.1849,
+      "step": 28134
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5670210719108582,
+      "learning_rate": 0.0003379650090997486,
+      "loss": 2.9245,
+      "step": 28135
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5730489492416382,
+      "learning_rate": 0.00033794973664123937,
+      "loss": 3.2632,
+      "step": 28136
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6139273643493652,
+      "learning_rate": 0.000337934464082778,
+      "loss": 2.8662,
+      "step": 28137
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5544430613517761,
+      "learning_rate": 0.0003379191914244044,
+      "loss": 2.9654,
+      "step": 28138
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5451558232307434,
+      "learning_rate": 0.000337903918666159,
+      "loss": 3.077,
+      "step": 28139
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5961140990257263,
+      "learning_rate": 0.0003378886458080819,
+      "loss": 2.9309,
+      "step": 28140
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5541216135025024,
+      "learning_rate": 0.00033787337285021347,
+      "loss": 3.2097,
+      "step": 28141
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6020674705505371,
+      "learning_rate": 0.0003378580997925939,
+      "loss": 3.0919,
+      "step": 28142
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5713543891906738,
+      "learning_rate": 0.0003378428266352633,
+      "loss": 2.98,
+      "step": 28143
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5623284578323364,
+      "learning_rate": 0.00033782755337826207,
+      "loss": 3.1572,
+      "step": 28144
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.616407036781311,
+      "learning_rate": 0.00033781228002163023,
+      "loss": 3.0148,
+      "step": 28145
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6024094820022583,
+      "learning_rate": 0.0003377970065654082,
+      "loss": 3.2478,
+      "step": 28146
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5678855180740356,
+      "learning_rate": 0.0003377817330096361,
+      "loss": 3.0946,
+      "step": 28147
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5765829682350159,
+      "learning_rate": 0.00033776645935435423,
+      "loss": 3.0297,
+      "step": 28148
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5960372090339661,
+      "learning_rate": 0.0003377511855996027,
+      "loss": 3.2084,
+      "step": 28149
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5560441017150879,
+      "learning_rate": 0.00033773591174542187,
+      "loss": 3.2332,
+      "step": 28150
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5740705132484436,
+      "learning_rate": 0.0003377206377918519,
+      "loss": 3.1825,
+      "step": 28151
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5770044922828674,
+      "learning_rate": 0.00033770536373893296,
+      "loss": 2.9729,
+      "step": 28152
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5786614418029785,
+      "learning_rate": 0.0003376900895867055,
+      "loss": 2.9429,
+      "step": 28153
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5744767189025879,
+      "learning_rate": 0.00033767481533520955,
+      "loss": 3.047,
+      "step": 28154
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5745210647583008,
+      "learning_rate": 0.0003376595409844852,
+      "loss": 3.0093,
+      "step": 28155
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5670796632766724,
+      "learning_rate": 0.000337644266534573,
+      "loss": 3.3227,
+      "step": 28156
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5535241365432739,
+      "learning_rate": 0.00033762899198551313,
+      "loss": 2.8861,
+      "step": 28157
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5871549248695374,
+      "learning_rate": 0.00033761371733734573,
+      "loss": 3.0057,
+      "step": 28158
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5588001012802124,
+      "learning_rate": 0.0003375984425901109,
+      "loss": 3.0635,
+      "step": 28159
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5769440531730652,
+      "learning_rate": 0.00033758316774384905,
+      "loss": 3.0683,
+      "step": 28160
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5671083331108093,
+      "learning_rate": 0.00033756789279860045,
+      "loss": 3.2814,
+      "step": 28161
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5821115970611572,
+      "learning_rate": 0.00033755261775440516,
+      "loss": 3.0085,
+      "step": 28162
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5854321122169495,
+      "learning_rate": 0.00033753734261130354,
+      "loss": 3.0935,
+      "step": 28163
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6202100515365601,
+      "learning_rate": 0.0003375220673693359,
+      "loss": 2.9791,
+      "step": 28164
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5914233326911926,
+      "learning_rate": 0.00033750679202854215,
+      "loss": 3.1086,
+      "step": 28165
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6070291996002197,
+      "learning_rate": 0.0003374915165889628,
+      "loss": 2.9375,
+      "step": 28166
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5866225361824036,
+      "learning_rate": 0.000337476241050638,
+      "loss": 2.9277,
+      "step": 28167
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5692453980445862,
+      "learning_rate": 0.0003374609654136081,
+      "loss": 3.2167,
+      "step": 28168
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5685532689094543,
+      "learning_rate": 0.0003374456896779132,
+      "loss": 3.0796,
+      "step": 28169
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5635455846786499,
+      "learning_rate": 0.00033743041384359335,
+      "loss": 2.8701,
+      "step": 28170
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6098818182945251,
+      "learning_rate": 0.0003374151379106893,
+      "loss": 2.901,
+      "step": 28171
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.579856812953949,
+      "learning_rate": 0.00033739986187924083,
+      "loss": 3.2155,
+      "step": 28172
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.596610963344574,
+      "learning_rate": 0.0003373845857492883,
+      "loss": 2.9086,
+      "step": 28173
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5602899789810181,
+      "learning_rate": 0.000337369309520872,
+      "loss": 3.1105,
+      "step": 28174
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5983844995498657,
+      "learning_rate": 0.00033735403319403215,
+      "loss": 3.0502,
+      "step": 28175
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5782126784324646,
+      "learning_rate": 0.0003373387567688089,
+      "loss": 3.148,
+      "step": 28176
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5897262692451477,
+      "learning_rate": 0.0003373234802452425,
+      "loss": 2.9129,
+      "step": 28177
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5535157918930054,
+      "learning_rate": 0.00033730820362337335,
+      "loss": 3.263,
+      "step": 28178
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5690072178840637,
+      "learning_rate": 0.00033729292690324166,
+      "loss": 3.0606,
+      "step": 28179
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5716756582260132,
+      "learning_rate": 0.00033727765008488743,
+      "loss": 3.0934,
+      "step": 28180
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6298738718032837,
+      "learning_rate": 0.00033726237316835094,
+      "loss": 2.8377,
+      "step": 28181
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5696219205856323,
+      "learning_rate": 0.0003372470961536728,
+      "loss": 3.2661,
+      "step": 28182
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5914949178695679,
+      "learning_rate": 0.00033723181904089284,
+      "loss": 3.0636,
+      "step": 28183
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5469275116920471,
+      "learning_rate": 0.0003372165418300514,
+      "loss": 3.2017,
+      "step": 28184
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.543024480342865,
+      "learning_rate": 0.00033720126452118883,
+      "loss": 2.9891,
+      "step": 28185
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5579934120178223,
+      "learning_rate": 0.0003371859871143452,
+      "loss": 2.9713,
+      "step": 28186
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5843299627304077,
+      "learning_rate": 0.0003371707096095609,
+      "loss": 2.8568,
+      "step": 28187
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5556427240371704,
+      "learning_rate": 0.000337155432006876,
+      "loss": 3.2779,
+      "step": 28188
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5713661313056946,
+      "learning_rate": 0.000337140154306331,
+      "loss": 2.8248,
+      "step": 28189
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5614469647407532,
+      "learning_rate": 0.000337124876507966,
+      "loss": 3.1321,
+      "step": 28190
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5511723160743713,
+      "learning_rate": 0.00033710959861182107,
+      "loss": 3.0381,
+      "step": 28191
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5804082751274109,
+      "learning_rate": 0.00033709432061793663,
+      "loss": 3.0038,
+      "step": 28192
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5728508234024048,
+      "learning_rate": 0.000337079042526353,
+      "loss": 3.1805,
+      "step": 28193
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5317544341087341,
+      "learning_rate": 0.0003370637643371102,
+      "loss": 3.1494,
+      "step": 28194
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5826964378356934,
+      "learning_rate": 0.0003370484860502486,
+      "loss": 3.038,
+      "step": 28195
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5762989521026611,
+      "learning_rate": 0.0003370332076658084,
+      "loss": 3.0305,
+      "step": 28196
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.55995774269104,
+      "learning_rate": 0.00033701792918382987,
+      "loss": 3.0058,
+      "step": 28197
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6025938987731934,
+      "learning_rate": 0.0003370026506043533,
+      "loss": 3.2507,
+      "step": 28198
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5528343915939331,
+      "learning_rate": 0.0003369873719274188,
+      "loss": 3.0487,
+      "step": 28199
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5329409241676331,
+      "learning_rate": 0.0003369720931530667,
+      "loss": 2.9969,
+      "step": 28200
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5734084248542786,
+      "learning_rate": 0.00033695681428133725,
+      "loss": 3.2161,
+      "step": 28201
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.562954306602478,
+      "learning_rate": 0.00033694153531227053,
+      "loss": 3.1746,
+      "step": 28202
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5255847573280334,
+      "learning_rate": 0.0003369262562459071,
+      "loss": 2.9532,
+      "step": 28203
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5966204404830933,
+      "learning_rate": 0.000336910977082287,
+      "loss": 3.195,
+      "step": 28204
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.549161434173584,
+      "learning_rate": 0.00033689569782145045,
+      "loss": 3.1496,
+      "step": 28205
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5623190999031067,
+      "learning_rate": 0.0003368804184634376,
+      "loss": 3.1611,
+      "step": 28206
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5726973414421082,
+      "learning_rate": 0.000336865139008289,
+      "loss": 3.0249,
+      "step": 28207
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5972573161125183,
+      "learning_rate": 0.00033684985945604465,
+      "loss": 3.0751,
+      "step": 28208
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5603774785995483,
+      "learning_rate": 0.0003368345798067449,
+      "loss": 3.1158,
+      "step": 28209
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5560487508773804,
+      "learning_rate": 0.0003368193000604299,
+      "loss": 3.1079,
+      "step": 28210
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5797421336174011,
+      "learning_rate": 0.00033680402021714,
+      "loss": 3.0904,
+      "step": 28211
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5546010136604309,
+      "learning_rate": 0.0003367887402769153,
+      "loss": 3.1467,
+      "step": 28212
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6668386459350586,
+      "learning_rate": 0.0003367734602397962,
+      "loss": 3.0496,
+      "step": 28213
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5610592365264893,
+      "learning_rate": 0.0003367581801058229,
+      "loss": 3.0401,
+      "step": 28214
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5730311870574951,
+      "learning_rate": 0.00033674289987503566,
+      "loss": 3.0605,
+      "step": 28215
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5653447508811951,
+      "learning_rate": 0.0003367276195474747,
+      "loss": 3.0281,
+      "step": 28216
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5662187933921814,
+      "learning_rate": 0.0003367123391231801,
+      "loss": 2.9896,
+      "step": 28217
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5825369954109192,
+      "learning_rate": 0.00033669705860219233,
+      "loss": 3.0219,
+      "step": 28218
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5653578042984009,
+      "learning_rate": 0.0003366817779845517,
+      "loss": 3.1838,
+      "step": 28219
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5515113472938538,
+      "learning_rate": 0.00033666649727029824,
+      "loss": 2.9519,
+      "step": 28220
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5668362975120544,
+      "learning_rate": 0.0003366512164594723,
+      "loss": 3.0131,
+      "step": 28221
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5645515322685242,
+      "learning_rate": 0.0003366359355521141,
+      "loss": 3.1945,
+      "step": 28222
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5728771686553955,
+      "learning_rate": 0.0003366206545482639,
+      "loss": 3.2453,
+      "step": 28223
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5764223337173462,
+      "learning_rate": 0.00033660537344796187,
+      "loss": 3.0845,
+      "step": 28224
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5650098323822021,
+      "learning_rate": 0.00033659009225124836,
+      "loss": 2.9356,
+      "step": 28225
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5907217860221863,
+      "learning_rate": 0.0003365748109581637,
+      "loss": 3.1849,
+      "step": 28226
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5731244683265686,
+      "learning_rate": 0.0003365595295687479,
+      "loss": 3.1806,
+      "step": 28227
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5921067595481873,
+      "learning_rate": 0.0003365442480830414,
+      "loss": 3.1239,
+      "step": 28228
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5963445901870728,
+      "learning_rate": 0.00033652896650108435,
+      "loss": 2.944,
+      "step": 28229
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.565613329410553,
+      "learning_rate": 0.00033651368482291705,
+      "loss": 2.9348,
+      "step": 28230
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6107422709465027,
+      "learning_rate": 0.00033649840304857977,
+      "loss": 2.9047,
+      "step": 28231
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5712419748306274,
+      "learning_rate": 0.0003364831211781126,
+      "loss": 3.086,
+      "step": 28232
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5632501840591431,
+      "learning_rate": 0.0003364678392115561,
+      "loss": 3.0059,
+      "step": 28233
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5802181363105774,
+      "learning_rate": 0.0003364525571489502,
+      "loss": 3.0808,
+      "step": 28234
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5678035616874695,
+      "learning_rate": 0.00033643727499033533,
+      "loss": 3.0125,
+      "step": 28235
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5598823428153992,
+      "learning_rate": 0.0003364219927357516,
+      "loss": 3.0799,
+      "step": 28236
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5781440138816833,
+      "learning_rate": 0.00033640671038523943,
+      "loss": 3.0678,
+      "step": 28237
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5706907510757446,
+      "learning_rate": 0.000336391427938839,
+      "loss": 3.1238,
+      "step": 28238
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5403823852539062,
+      "learning_rate": 0.0003363761453965905,
+      "loss": 3.0808,
+      "step": 28239
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5869365334510803,
+      "learning_rate": 0.0003363608627585342,
+      "loss": 3.202,
+      "step": 28240
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.590049684047699,
+      "learning_rate": 0.00033634558002471055,
+      "loss": 2.983,
+      "step": 28241
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6182698607444763,
+      "learning_rate": 0.00033633029719515946,
+      "loss": 2.9491,
+      "step": 28242
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5630959868431091,
+      "learning_rate": 0.00033631501426992146,
+      "loss": 3.1776,
+      "step": 28243
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5790128111839294,
+      "learning_rate": 0.00033629973124903666,
+      "loss": 3.2561,
+      "step": 28244
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.612781286239624,
+      "learning_rate": 0.0003362844481325454,
+      "loss": 3.0595,
+      "step": 28245
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5653966069221497,
+      "learning_rate": 0.00033626916492048784,
+      "loss": 3.0165,
+      "step": 28246
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5676288604736328,
+      "learning_rate": 0.0003362538816129043,
+      "loss": 3.1517,
+      "step": 28247
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5745164752006531,
+      "learning_rate": 0.000336238598209835,
+      "loss": 3.2458,
+      "step": 28248
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.624270498752594,
+      "learning_rate": 0.0003362233147113202,
+      "loss": 2.9842,
+      "step": 28249
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5752471089363098,
+      "learning_rate": 0.0003362080311174002,
+      "loss": 3.08,
+      "step": 28250
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6258599758148193,
+      "learning_rate": 0.0003361927474281152,
+      "loss": 2.9069,
+      "step": 28251
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5874921083450317,
+      "learning_rate": 0.0003361774636435055,
+      "loss": 3.077,
+      "step": 28252
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5625190734863281,
+      "learning_rate": 0.0003361621797636112,
+      "loss": 2.7603,
+      "step": 28253
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5910776257514954,
+      "learning_rate": 0.00033614689578847275,
+      "loss": 2.8325,
+      "step": 28254
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5543468594551086,
+      "learning_rate": 0.0003361316117181304,
+      "loss": 3.0568,
+      "step": 28255
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6161579489707947,
+      "learning_rate": 0.0003361163275526243,
+      "loss": 3.1239,
+      "step": 28256
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5997530817985535,
+      "learning_rate": 0.0003361010432919947,
+      "loss": 3.1178,
+      "step": 28257
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6007872223854065,
+      "learning_rate": 0.0003360857589362819,
+      "loss": 2.8373,
+      "step": 28258
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5524115562438965,
+      "learning_rate": 0.00033607047448552617,
+      "loss": 3.1908,
+      "step": 28259
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5868061184883118,
+      "learning_rate": 0.00033605518993976783,
+      "loss": 3.091,
+      "step": 28260
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5432960391044617,
+      "learning_rate": 0.00033603990529904694,
+      "loss": 3.1476,
+      "step": 28261
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5888095498085022,
+      "learning_rate": 0.00033602462056340387,
+      "loss": 3.0807,
+      "step": 28262
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5865523219108582,
+      "learning_rate": 0.00033600933573287896,
+      "loss": 2.9347,
+      "step": 28263
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5949804186820984,
+      "learning_rate": 0.0003359940508075124,
+      "loss": 3.151,
+      "step": 28264
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5891909599304199,
+      "learning_rate": 0.0003359787657873444,
+      "loss": 2.8826,
+      "step": 28265
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5554693937301636,
+      "learning_rate": 0.0003359634806724153,
+      "loss": 3.0168,
+      "step": 28266
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.583011269569397,
+      "learning_rate": 0.0003359481954627653,
+      "loss": 2.8308,
+      "step": 28267
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5493490695953369,
+      "learning_rate": 0.00033593291015843456,
+      "loss": 2.6961,
+      "step": 28268
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5663343667984009,
+      "learning_rate": 0.0003359176247594635,
+      "loss": 3.1074,
+      "step": 28269
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6228228807449341,
+      "learning_rate": 0.00033590233926589246,
+      "loss": 3.2481,
+      "step": 28270
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6101515293121338,
+      "learning_rate": 0.00033588705367776145,
+      "loss": 2.8834,
+      "step": 28271
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5485780239105225,
+      "learning_rate": 0.0003358717679951109,
+      "loss": 3.0186,
+      "step": 28272
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5793207287788391,
+      "learning_rate": 0.0003358564822179809,
+      "loss": 3.0909,
+      "step": 28273
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6276209354400635,
+      "learning_rate": 0.000335841196346412,
+      "loss": 3.0662,
+      "step": 28274
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5886393785476685,
+      "learning_rate": 0.00033582591038044414,
+      "loss": 3.1548,
+      "step": 28275
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5739596486091614,
+      "learning_rate": 0.0003358106243201178,
+      "loss": 2.8666,
+      "step": 28276
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.583791196346283,
+      "learning_rate": 0.0003357953381654732,
+      "loss": 3.0231,
+      "step": 28277
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.584082305431366,
+      "learning_rate": 0.00033578005191655047,
+      "loss": 3.1848,
+      "step": 28278
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5657137632369995,
+      "learning_rate": 0.00033576476557339,
+      "loss": 2.9686,
+      "step": 28279
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5658915042877197,
+      "learning_rate": 0.00033574947913603205,
+      "loss": 3.1709,
+      "step": 28280
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5720155835151672,
+      "learning_rate": 0.00033573419260451694,
+      "loss": 3.1229,
+      "step": 28281
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5654585361480713,
+      "learning_rate": 0.00033571890597888473,
+      "loss": 2.9455,
+      "step": 28282
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5792554616928101,
+      "learning_rate": 0.00033570361925917575,
+      "loss": 3.1386,
+      "step": 28283
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5770618915557861,
+      "learning_rate": 0.0003356883324454304,
+      "loss": 3.1544,
+      "step": 28284
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5782315731048584,
+      "learning_rate": 0.00033567304553768884,
+      "loss": 3.1645,
+      "step": 28285
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.570188581943512,
+      "learning_rate": 0.0003356577585359913,
+      "loss": 3.1487,
+      "step": 28286
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5855571031570435,
+      "learning_rate": 0.00033564247144037815,
+      "loss": 2.8481,
+      "step": 28287
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5529902577400208,
+      "learning_rate": 0.0003356271842508896,
+      "loss": 2.9648,
+      "step": 28288
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5426622629165649,
+      "learning_rate": 0.00033561189696756574,
+      "loss": 3.077,
+      "step": 28289
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5838942527770996,
+      "learning_rate": 0.00033559660959044715,
+      "loss": 3.0563,
+      "step": 28290
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6251130700111389,
+      "learning_rate": 0.00033558132211957385,
+      "loss": 2.962,
+      "step": 28291
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5622515678405762,
+      "learning_rate": 0.0003355660345549863,
+      "loss": 3.0919,
+      "step": 28292
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5667559504508972,
+      "learning_rate": 0.00033555074689672464,
+      "loss": 2.9387,
+      "step": 28293
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5650799870491028,
+      "learning_rate": 0.00033553545914482907,
+      "loss": 3.1512,
+      "step": 28294
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5878683924674988,
+      "learning_rate": 0.00033552017129934,
+      "loss": 2.8555,
+      "step": 28295
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.61265629529953,
+      "learning_rate": 0.00033550488336029765,
+      "loss": 3.0915,
+      "step": 28296
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5722835659980774,
+      "learning_rate": 0.00033548959532774225,
+      "loss": 2.8942,
+      "step": 28297
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5835713744163513,
+      "learning_rate": 0.0003354743072017141,
+      "loss": 2.9738,
+      "step": 28298
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6345725655555725,
+      "learning_rate": 0.0003354590189822534,
+      "loss": 3.3849,
+      "step": 28299
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5987057685852051,
+      "learning_rate": 0.0003354437306694005,
+      "loss": 2.9575,
+      "step": 28300
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5893240571022034,
+      "learning_rate": 0.00033542844226319566,
+      "loss": 3.0071,
+      "step": 28301
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.595427930355072,
+      "learning_rate": 0.0003354131537636791,
+      "loss": 3.1305,
+      "step": 28302
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6372551321983337,
+      "learning_rate": 0.0003353978651708911,
+      "loss": 2.9446,
+      "step": 28303
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6060798168182373,
+      "learning_rate": 0.00033538257648487195,
+      "loss": 3.0027,
+      "step": 28304
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5530107021331787,
+      "learning_rate": 0.00033536728770566186,
+      "loss": 3.0878,
+      "step": 28305
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6247791051864624,
+      "learning_rate": 0.00033535199883330123,
+      "loss": 3.2343,
+      "step": 28306
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5921459794044495,
+      "learning_rate": 0.00033533670986783014,
+      "loss": 3.1694,
+      "step": 28307
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5848678946495056,
+      "learning_rate": 0.000335321420809289,
+      "loss": 3.0928,
+      "step": 28308
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6024354100227356,
+      "learning_rate": 0.00033530613165771804,
+      "loss": 2.881,
+      "step": 28309
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5735805034637451,
+      "learning_rate": 0.0003352908424131576,
+      "loss": 2.9333,
+      "step": 28310
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5353273153305054,
+      "learning_rate": 0.00033527555307564773,
+      "loss": 3.0901,
+      "step": 28311
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5976811647415161,
+      "learning_rate": 0.00033526026364522895,
+      "loss": 3.2398,
+      "step": 28312
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5682657361030579,
+      "learning_rate": 0.00033524497412194137,
+      "loss": 3.137,
+      "step": 28313
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5598923563957214,
+      "learning_rate": 0.0003352296845058253,
+      "loss": 2.9647,
+      "step": 28314
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5730248093605042,
+      "learning_rate": 0.000335214394796921,
+      "loss": 3.0873,
+      "step": 28315
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5653427839279175,
+      "learning_rate": 0.00033519910499526883,
+      "loss": 2.9421,
+      "step": 28316
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5537395477294922,
+      "learning_rate": 0.000335183815100909,
+      "loss": 2.9381,
+      "step": 28317
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5599444508552551,
+      "learning_rate": 0.00033516852511388174,
+      "loss": 2.903,
+      "step": 28318
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.552375853061676,
+      "learning_rate": 0.00033515323503422725,
+      "loss": 3.0835,
+      "step": 28319
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6051978468894958,
+      "learning_rate": 0.00033513794486198607,
+      "loss": 3.013,
+      "step": 28320
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5906317234039307,
+      "learning_rate": 0.0003351226545971983,
+      "loss": 3.0689,
+      "step": 28321
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5584150552749634,
+      "learning_rate": 0.0003351073642399041,
+      "loss": 3.0572,
+      "step": 28322
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5732271075248718,
+      "learning_rate": 0.0003350920737901439,
+      "loss": 3.0508,
+      "step": 28323
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6018269658088684,
+      "learning_rate": 0.000335076783247958,
+      "loss": 2.9892,
+      "step": 28324
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5475950837135315,
+      "learning_rate": 0.00033506149261338655,
+      "loss": 3.0967,
+      "step": 28325
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5918049812316895,
+      "learning_rate": 0.0003350462018864698,
+      "loss": 2.9518,
+      "step": 28326
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6317940950393677,
+      "learning_rate": 0.0003350309110672482,
+      "loss": 3.0599,
+      "step": 28327
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5397853255271912,
+      "learning_rate": 0.00033501562015576195,
+      "loss": 3.1353,
+      "step": 28328
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5789517760276794,
+      "learning_rate": 0.0003350003291520513,
+      "loss": 3.1184,
+      "step": 28329
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6313763856887817,
+      "learning_rate": 0.00033498503805615636,
+      "loss": 2.9678,
+      "step": 28330
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5650973916053772,
+      "learning_rate": 0.00033496974686811766,
+      "loss": 3.1104,
+      "step": 28331
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5666373372077942,
+      "learning_rate": 0.00033495445558797543,
+      "loss": 2.8425,
+      "step": 28332
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5651002526283264,
+      "learning_rate": 0.0003349391642157698,
+      "loss": 2.909,
+      "step": 28333
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5783306360244751,
+      "learning_rate": 0.0003349238727515412,
+      "loss": 3.1748,
+      "step": 28334
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5995091795921326,
+      "learning_rate": 0.00033490858119532985,
+      "loss": 2.9953,
+      "step": 28335
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5513354539871216,
+      "learning_rate": 0.00033489328954717596,
+      "loss": 2.9552,
+      "step": 28336
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.56820148229599,
+      "learning_rate": 0.0003348779978071199,
+      "loss": 3.0048,
+      "step": 28337
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5699313282966614,
+      "learning_rate": 0.0003348627059752019,
+      "loss": 3.1077,
+      "step": 28338
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5874160528182983,
+      "learning_rate": 0.00033484741405146227,
+      "loss": 3.1624,
+      "step": 28339
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5903382301330566,
+      "learning_rate": 0.00033483212203594116,
+      "loss": 3.124,
+      "step": 28340
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5418932437896729,
+      "learning_rate": 0.00033481682992867904,
+      "loss": 3.1153,
+      "step": 28341
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6493607759475708,
+      "learning_rate": 0.00033480153772971603,
+      "loss": 2.9999,
+      "step": 28342
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5959764122962952,
+      "learning_rate": 0.00033478624543909256,
+      "loss": 3.0586,
+      "step": 28343
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5970855355262756,
+      "learning_rate": 0.0003347709530568487,
+      "loss": 3.1342,
+      "step": 28344
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5649682879447937,
+      "learning_rate": 0.00033475566058302486,
+      "loss": 3.1413,
+      "step": 28345
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6004948616027832,
+      "learning_rate": 0.00033474036801766136,
+      "loss": 3.0267,
+      "step": 28346
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6019605398178101,
+      "learning_rate": 0.0003347250753607984,
+      "loss": 2.954,
+      "step": 28347
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5807358026504517,
+      "learning_rate": 0.0003347097826124762,
+      "loss": 3.1699,
+      "step": 28348
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5428721308708191,
+      "learning_rate": 0.0003346944897727352,
+      "loss": 3.1219,
+      "step": 28349
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6022670269012451,
+      "learning_rate": 0.00033467919684161554,
+      "loss": 3.1146,
+      "step": 28350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5750691294670105,
+      "learning_rate": 0.0003346639038191576,
+      "loss": 3.06,
+      "step": 28351
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5875911712646484,
+      "learning_rate": 0.0003346486107054016,
+      "loss": 3.0479,
+      "step": 28352
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.588360071182251,
+      "learning_rate": 0.0003346333175003878,
+      "loss": 3.0508,
+      "step": 28353
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5815553665161133,
+      "learning_rate": 0.00033461802420415654,
+      "loss": 3.2218,
+      "step": 28354
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5743089914321899,
+      "learning_rate": 0.00033460273081674797,
+      "loss": 3.039,
+      "step": 28355
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5957072377204895,
+      "learning_rate": 0.0003345874373382025,
+      "loss": 3.0255,
+      "step": 28356
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5829094648361206,
+      "learning_rate": 0.0003345721437685605,
+      "loss": 3.2637,
+      "step": 28357
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5787622928619385,
+      "learning_rate": 0.00033455685010786205,
+      "loss": 3.0796,
+      "step": 28358
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6022943258285522,
+      "learning_rate": 0.0003345415563561475,
+      "loss": 3.2131,
+      "step": 28359
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.59725421667099,
+      "learning_rate": 0.00033452626251345713,
+      "loss": 3.1263,
+      "step": 28360
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5658643245697021,
+      "learning_rate": 0.00033451096857983124,
+      "loss": 2.9285,
+      "step": 28361
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5705099105834961,
+      "learning_rate": 0.0003344956745553101,
+      "loss": 2.9658,
+      "step": 28362
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5385516881942749,
+      "learning_rate": 0.000334480380439934,
+      "loss": 3.1083,
+      "step": 28363
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5449607372283936,
+      "learning_rate": 0.00033446508623374317,
+      "loss": 3.1891,
+      "step": 28364
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6035528779029846,
+      "learning_rate": 0.000334449791936778,
+      "loss": 3.2337,
+      "step": 28365
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5781618356704712,
+      "learning_rate": 0.0003344344975490786,
+      "loss": 3.1241,
+      "step": 28366
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.561470091342926,
+      "learning_rate": 0.0003344192030706855,
+      "loss": 3.0613,
+      "step": 28367
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5787182450294495,
+      "learning_rate": 0.00033440390850163876,
+      "loss": 3.063,
+      "step": 28368
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5583317875862122,
+      "learning_rate": 0.0003343886138419788,
+      "loss": 3.0326,
+      "step": 28369
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5830174088478088,
+      "learning_rate": 0.0003343733190917458,
+      "loss": 3.1601,
+      "step": 28370
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5853716731071472,
+      "learning_rate": 0.00033435802425098006,
+      "loss": 2.985,
+      "step": 28371
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5587119460105896,
+      "learning_rate": 0.000334342729319722,
+      "loss": 3.1248,
+      "step": 28372
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5844123363494873,
+      "learning_rate": 0.00033432743429801176,
+      "loss": 3.1114,
+      "step": 28373
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.620872974395752,
+      "learning_rate": 0.0003343121391858896,
+      "loss": 3.1804,
+      "step": 28374
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5749686360359192,
+      "learning_rate": 0.00033429684398339596,
+      "loss": 2.8822,
+      "step": 28375
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5744134783744812,
+      "learning_rate": 0.00033428154869057097,
+      "loss": 3.0772,
+      "step": 28376
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6181778311729431,
+      "learning_rate": 0.000334266253307455,
+      "loss": 3.1124,
+      "step": 28377
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5899921655654907,
+      "learning_rate": 0.0003342509578340883,
+      "loss": 3.2172,
+      "step": 28378
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5781901478767395,
+      "learning_rate": 0.00033423566227051127,
+      "loss": 3.12,
+      "step": 28379
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5640218257904053,
+      "learning_rate": 0.000334220366616764,
+      "loss": 2.9047,
+      "step": 28380
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6122222542762756,
+      "learning_rate": 0.0003342050708728868,
+      "loss": 3.0602,
+      "step": 28381
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5799720883369446,
+      "learning_rate": 0.0003341897750389201,
+      "loss": 3.148,
+      "step": 28382
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.57460618019104,
+      "learning_rate": 0.0003341744791149042,
+      "loss": 3.3129,
+      "step": 28383
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5734221339225769,
+      "learning_rate": 0.0003341591831008792,
+      "loss": 3.0301,
+      "step": 28384
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5949209928512573,
+      "learning_rate": 0.0003341438869968855,
+      "loss": 3.1525,
+      "step": 28385
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5924066305160522,
+      "learning_rate": 0.00033412859080296335,
+      "loss": 3.2143,
+      "step": 28386
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5933117270469666,
+      "learning_rate": 0.0003341132945191531,
+      "loss": 3.2277,
+      "step": 28387
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5546301007270813,
+      "learning_rate": 0.00033409799814549496,
+      "loss": 2.9102,
+      "step": 28388
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5834781527519226,
+      "learning_rate": 0.0003340827016820293,
+      "loss": 2.8856,
+      "step": 28389
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5664399266242981,
+      "learning_rate": 0.00033406740512879635,
+      "loss": 3.0256,
+      "step": 28390
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5528917908668518,
+      "learning_rate": 0.00033405210848583636,
+      "loss": 2.9815,
+      "step": 28391
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5898980498313904,
+      "learning_rate": 0.00033403681175318974,
+      "loss": 2.9924,
+      "step": 28392
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6130119562149048,
+      "learning_rate": 0.00033402151493089663,
+      "loss": 2.812,
+      "step": 28393
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5696910619735718,
+      "learning_rate": 0.0003340062180189975,
+      "loss": 3.2702,
+      "step": 28394
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5849608778953552,
+      "learning_rate": 0.0003339909210175325,
+      "loss": 3.0146,
+      "step": 28395
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.596095621585846,
+      "learning_rate": 0.0003339756239265419,
+      "loss": 2.9181,
+      "step": 28396
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5665485262870789,
+      "learning_rate": 0.0003339603267460661,
+      "loss": 3.069,
+      "step": 28397
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5528839826583862,
+      "learning_rate": 0.0003339450294761453,
+      "loss": 3.0766,
+      "step": 28398
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5885125398635864,
+      "learning_rate": 0.00033392973211681987,
+      "loss": 2.9717,
+      "step": 28399
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6001590490341187,
+      "learning_rate": 0.00033391443466813004,
+      "loss": 2.8122,
+      "step": 28400
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5653418898582458,
+      "learning_rate": 0.0003338991371301161,
+      "loss": 3.1272,
+      "step": 28401
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.577953577041626,
+      "learning_rate": 0.0003338838395028183,
+      "loss": 3.0803,
+      "step": 28402
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5523266196250916,
+      "learning_rate": 0.00033386854178627705,
+      "loss": 3.0247,
+      "step": 28403
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5539414882659912,
+      "learning_rate": 0.0003338532439805326,
+      "loss": 3.1748,
+      "step": 28404
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5848209857940674,
+      "learning_rate": 0.0003338379460856253,
+      "loss": 2.9997,
+      "step": 28405
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.573849618434906,
+      "learning_rate": 0.00033382264810159523,
+      "loss": 3.0206,
+      "step": 28406
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.575800895690918,
+      "learning_rate": 0.00033380735002848274,
+      "loss": 3.0534,
+      "step": 28407
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5938015580177307,
+      "learning_rate": 0.0003337920518663284,
+      "loss": 3.1933,
+      "step": 28408
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5658040642738342,
+      "learning_rate": 0.0003337767536151722,
+      "loss": 2.8285,
+      "step": 28409
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5854364633560181,
+      "learning_rate": 0.0003337614552750545,
+      "loss": 2.8454,
+      "step": 28410
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5785530209541321,
+      "learning_rate": 0.00033374615684601567,
+      "loss": 3.0105,
+      "step": 28411
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5590497851371765,
+      "learning_rate": 0.000333730858328096,
+      "loss": 2.8793,
+      "step": 28412
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5437723994255066,
+      "learning_rate": 0.00033371555972133563,
+      "loss": 2.9129,
+      "step": 28413
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6012238264083862,
+      "learning_rate": 0.00033370026102577503,
+      "loss": 3.1785,
+      "step": 28414
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5660067200660706,
+      "learning_rate": 0.0003336849622414544,
+      "loss": 3.0921,
+      "step": 28415
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5316462516784668,
+      "learning_rate": 0.0003336696633684141,
+      "loss": 3.1858,
+      "step": 28416
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6181156039237976,
+      "learning_rate": 0.0003336543644066944,
+      "loss": 3.0513,
+      "step": 28417
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5790272951126099,
+      "learning_rate": 0.00033363906535633546,
+      "loss": 3.0921,
+      "step": 28418
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.553527295589447,
+      "learning_rate": 0.0003336237662173779,
+      "loss": 2.9811,
+      "step": 28419
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5918067097663879,
+      "learning_rate": 0.00033360846698986165,
+      "loss": 2.9584,
+      "step": 28420
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5768551826477051,
+      "learning_rate": 0.0003335931676738272,
+      "loss": 3.1019,
+      "step": 28421
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.573397696018219,
+      "learning_rate": 0.00033357786826931484,
+      "loss": 3.0213,
+      "step": 28422
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5723202228546143,
+      "learning_rate": 0.00033356256877636485,
+      "loss": 2.847,
+      "step": 28423
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5666738748550415,
+      "learning_rate": 0.00033354726919501747,
+      "loss": 2.8922,
+      "step": 28424
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5918300747871399,
+      "learning_rate": 0.00033353196952531304,
+      "loss": 3.0617,
+      "step": 28425
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5779464840888977,
+      "learning_rate": 0.0003335166697672919,
+      "loss": 3.0998,
+      "step": 28426
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5665059685707092,
+      "learning_rate": 0.00033350136992099416,
+      "loss": 3.207,
+      "step": 28427
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5882079601287842,
+      "learning_rate": 0.0003334860699864604,
+      "loss": 3.0049,
+      "step": 28428
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5963062047958374,
+      "learning_rate": 0.00033347076996373073,
+      "loss": 3.2008,
+      "step": 28429
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5421083569526672,
+      "learning_rate": 0.0003334554698528456,
+      "loss": 2.9144,
+      "step": 28430
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5771951675415039,
+      "learning_rate": 0.0003334401696538451,
+      "loss": 3.1604,
+      "step": 28431
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6018733978271484,
+      "learning_rate": 0.00033342486936676957,
+      "loss": 3.2034,
+      "step": 28432
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5767068862915039,
+      "learning_rate": 0.0003334095689916594,
+      "loss": 3.2174,
+      "step": 28433
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5713096857070923,
+      "learning_rate": 0.00033339426852855496,
+      "loss": 3.0759,
+      "step": 28434
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5891026258468628,
+      "learning_rate": 0.0003333789679774964,
+      "loss": 3.3295,
+      "step": 28435
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5787071585655212,
+      "learning_rate": 0.000333363667338524,
+      "loss": 3.164,
+      "step": 28436
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6013773679733276,
+      "learning_rate": 0.00033334836661167816,
+      "loss": 2.8607,
+      "step": 28437
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5678498148918152,
+      "learning_rate": 0.00033333306579699914,
+      "loss": 3.0511,
+      "step": 28438
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6372431516647339,
+      "learning_rate": 0.00033331776489452724,
+      "loss": 2.7539,
+      "step": 28439
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5957581400871277,
+      "learning_rate": 0.0003333024639043028,
+      "loss": 3.117,
+      "step": 28440
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5882934927940369,
+      "learning_rate": 0.0003332871628263661,
+      "loss": 3.0676,
+      "step": 28441
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5675431489944458,
+      "learning_rate": 0.00033327186166075723,
+      "loss": 3.2111,
+      "step": 28442
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5941580533981323,
+      "learning_rate": 0.0003332565604075168,
+      "loss": 3.0244,
+      "step": 28443
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6287396550178528,
+      "learning_rate": 0.000333241259066685,
+      "loss": 3.1406,
+      "step": 28444
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5600050091743469,
+      "learning_rate": 0.0003332259576383022,
+      "loss": 2.9896,
+      "step": 28445
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5782938599586487,
+      "learning_rate": 0.00033321065612240854,
+      "loss": 3.0307,
+      "step": 28446
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5861037969589233,
+      "learning_rate": 0.0003331953545190444,
+      "loss": 2.9698,
+      "step": 28447
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5864157676696777,
+      "learning_rate": 0.00033318005282825,
+      "loss": 2.9486,
+      "step": 28448
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5786110162734985,
+      "learning_rate": 0.0003331647510500659,
+      "loss": 2.9526,
+      "step": 28449
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5792027711868286,
+      "learning_rate": 0.0003331494491845321,
+      "loss": 3.2205,
+      "step": 28450
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.546311616897583,
+      "learning_rate": 0.00033313414723168904,
+      "loss": 2.8147,
+      "step": 28451
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5609930753707886,
+      "learning_rate": 0.0003331188451915771,
+      "loss": 3.0984,
+      "step": 28452
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6086905598640442,
+      "learning_rate": 0.0003331035430642364,
+      "loss": 3.138,
+      "step": 28453
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5497440099716187,
+      "learning_rate": 0.00033308824084970736,
+      "loss": 3.0029,
+      "step": 28454
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5903970003128052,
+      "learning_rate": 0.0003330729385480303,
+      "loss": 2.9264,
+      "step": 28455
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6043407320976257,
+      "learning_rate": 0.0003330576361592455,
+      "loss": 2.8884,
+      "step": 28456
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.559085488319397,
+      "learning_rate": 0.0003330423336833933,
+      "loss": 3.0239,
+      "step": 28457
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5664832592010498,
+      "learning_rate": 0.0003330270311205137,
+      "loss": 3.116,
+      "step": 28458
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5917896032333374,
+      "learning_rate": 0.0003330117284706475,
+      "loss": 3.139,
+      "step": 28459
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.588679850101471,
+      "learning_rate": 0.0003329964257338347,
+      "loss": 3.0135,
+      "step": 28460
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5861005187034607,
+      "learning_rate": 0.0003329811229101156,
+      "loss": 2.8976,
+      "step": 28461
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5760902166366577,
+      "learning_rate": 0.00033296581999953063,
+      "loss": 3.094,
+      "step": 28462
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5762536525726318,
+      "learning_rate": 0.00033295051700212003,
+      "loss": 3.2222,
+      "step": 28463
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5697655081748962,
+      "learning_rate": 0.0003329352139179241,
+      "loss": 3.0622,
+      "step": 28464
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6054670810699463,
+      "learning_rate": 0.0003329199107469832,
+      "loss": 2.8424,
+      "step": 28465
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5558887720108032,
+      "learning_rate": 0.00033290460748933745,
+      "loss": 3.2195,
+      "step": 28466
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6065505743026733,
+      "learning_rate": 0.0003328893041450275,
+      "loss": 3.1301,
+      "step": 28467
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6185204982757568,
+      "learning_rate": 0.00033287400071409327,
+      "loss": 3.1541,
+      "step": 28468
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5224096179008484,
+      "learning_rate": 0.0003328586971965753,
+      "loss": 2.9278,
+      "step": 28469
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5885241627693176,
+      "learning_rate": 0.0003328433935925139,
+      "loss": 3.2153,
+      "step": 28470
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5575811266899109,
+      "learning_rate": 0.00033282808990194935,
+      "loss": 3.1026,
+      "step": 28471
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6058753132820129,
+      "learning_rate": 0.00033281278612492187,
+      "loss": 2.8013,
+      "step": 28472
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.585891842842102,
+      "learning_rate": 0.0003327974822614718,
+      "loss": 3.0083,
+      "step": 28473
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5971677303314209,
+      "learning_rate": 0.0003327821783116395,
+      "loss": 3.1401,
+      "step": 28474
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5844119191169739,
+      "learning_rate": 0.0003327668742754653,
+      "loss": 2.9653,
+      "step": 28475
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6075958609580994,
+      "learning_rate": 0.0003327515701529894,
+      "loss": 3.0223,
+      "step": 28476
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5761697888374329,
+      "learning_rate": 0.00033273626594425217,
+      "loss": 3.2987,
+      "step": 28477
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7378432750701904,
+      "learning_rate": 0.0003327209616492939,
+      "loss": 3.1888,
+      "step": 28478
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5896404981613159,
+      "learning_rate": 0.0003327056572681549,
+      "loss": 3.1417,
+      "step": 28479
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.581295371055603,
+      "learning_rate": 0.00033269035280087555,
+      "loss": 2.915,
+      "step": 28480
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5805588960647583,
+      "learning_rate": 0.0003326750482474961,
+      "loss": 3.0804,
+      "step": 28481
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.631949782371521,
+      "learning_rate": 0.00033265974360805687,
+      "loss": 2.8173,
+      "step": 28482
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5704339742660522,
+      "learning_rate": 0.0003326444388825981,
+      "loss": 2.9781,
+      "step": 28483
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5693856477737427,
+      "learning_rate": 0.0003326291340711602,
+      "loss": 2.923,
+      "step": 28484
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5859050750732422,
+      "learning_rate": 0.0003326138291737835,
+      "loss": 2.9329,
+      "step": 28485
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5935141444206238,
+      "learning_rate": 0.00033259852419050814,
+      "loss": 3.1226,
+      "step": 28486
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6061211228370667,
+      "learning_rate": 0.00033258321912137465,
+      "loss": 3.1309,
+      "step": 28487
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6038804650306702,
+      "learning_rate": 0.0003325679139664231,
+      "loss": 2.8745,
+      "step": 28488
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6384372711181641,
+      "learning_rate": 0.00033255260872569405,
+      "loss": 2.9845,
+      "step": 28489
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6087737679481506,
+      "learning_rate": 0.00033253730339922766,
+      "loss": 3.0332,
+      "step": 28490
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.544606626033783,
+      "learning_rate": 0.00033252199798706426,
+      "loss": 3.0271,
+      "step": 28491
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6075654625892639,
+      "learning_rate": 0.00033250669248924416,
+      "loss": 3.1083,
+      "step": 28492
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5745288729667664,
+      "learning_rate": 0.0003324913869058077,
+      "loss": 3.118,
+      "step": 28493
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.585982620716095,
+      "learning_rate": 0.0003324760812367952,
+      "loss": 2.8791,
+      "step": 28494
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6062518358230591,
+      "learning_rate": 0.00033246077548224686,
+      "loss": 3.0527,
+      "step": 28495
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5591403841972351,
+      "learning_rate": 0.0003324454696422033,
+      "loss": 3.087,
+      "step": 28496
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6129733324050903,
+      "learning_rate": 0.00033243016371670446,
+      "loss": 3.1021,
+      "step": 28497
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5932202935218811,
+      "learning_rate": 0.00033241485770579084,
+      "loss": 2.89,
+      "step": 28498
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5794448256492615,
+      "learning_rate": 0.0003323995516095027,
+      "loss": 3.1646,
+      "step": 28499
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6294394135475159,
+      "learning_rate": 0.0003323842454278804,
+      "loss": 3.0625,
+      "step": 28500
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5836816430091858,
+      "learning_rate": 0.0003323689391609642,
+      "loss": 2.9569,
+      "step": 28501
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5736550688743591,
+      "learning_rate": 0.0003323536328087945,
+      "loss": 3.0808,
+      "step": 28502
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6020394563674927,
+      "learning_rate": 0.0003323383263714115,
+      "loss": 3.1046,
+      "step": 28503
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.566116213798523,
+      "learning_rate": 0.0003323230198488555,
+      "loss": 2.8113,
+      "step": 28504
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5596758723258972,
+      "learning_rate": 0.000332307713241167,
+      "loss": 3.1021,
+      "step": 28505
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5862348675727844,
+      "learning_rate": 0.0003322924065483862,
+      "loss": 2.9094,
+      "step": 28506
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5914575457572937,
+      "learning_rate": 0.00033227709977055344,
+      "loss": 3.1915,
+      "step": 28507
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6266211867332458,
+      "learning_rate": 0.000332261792907709,
+      "loss": 3.0326,
+      "step": 28508
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6111003756523132,
+      "learning_rate": 0.00033224648595989306,
+      "loss": 3.3007,
+      "step": 28509
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.611010730266571,
+      "learning_rate": 0.00033223117892714626,
+      "loss": 2.8517,
+      "step": 28510
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6048029065132141,
+      "learning_rate": 0.0003322158718095087,
+      "loss": 3.065,
+      "step": 28511
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6111939549446106,
+      "learning_rate": 0.0003322005646070207,
+      "loss": 2.9375,
+      "step": 28512
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5618122816085815,
+      "learning_rate": 0.0003321852573197226,
+      "loss": 3.0391,
+      "step": 28513
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6132274866104126,
+      "learning_rate": 0.00033216994994765477,
+      "loss": 3.0492,
+      "step": 28514
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6032343506813049,
+      "learning_rate": 0.0003321546424908574,
+      "loss": 2.9005,
+      "step": 28515
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6772813200950623,
+      "learning_rate": 0.00033213933494937093,
+      "loss": 3.089,
+      "step": 28516
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.544093668460846,
+      "learning_rate": 0.0003321240273232357,
+      "loss": 3.0564,
+      "step": 28517
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6309990286827087,
+      "learning_rate": 0.00033210871961249195,
+      "loss": 2.9296,
+      "step": 28518
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5908814668655396,
+      "learning_rate": 0.00033209341181717996,
+      "loss": 3.1371,
+      "step": 28519
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.55782151222229,
+      "learning_rate": 0.0003320781039373401,
+      "loss": 2.9583,
+      "step": 28520
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5843961834907532,
+      "learning_rate": 0.00033206279597301276,
+      "loss": 3.2184,
+      "step": 28521
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5808646082878113,
+      "learning_rate": 0.0003320474879242381,
+      "loss": 2.8148,
+      "step": 28522
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6347132325172424,
+      "learning_rate": 0.00033203217979105656,
+      "loss": 3.2595,
+      "step": 28523
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5245764851570129,
+      "learning_rate": 0.00033201687157350843,
+      "loss": 3.0635,
+      "step": 28524
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6027230024337769,
+      "learning_rate": 0.000332001563271634,
+      "loss": 3.0783,
+      "step": 28525
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5538972616195679,
+      "learning_rate": 0.0003319862548854736,
+      "loss": 2.8988,
+      "step": 28526
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6136893630027771,
+      "learning_rate": 0.00033197094641506763,
+      "loss": 2.9789,
+      "step": 28527
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5786117911338806,
+      "learning_rate": 0.0003319556378604563,
+      "loss": 2.8044,
+      "step": 28528
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6115828156471252,
+      "learning_rate": 0.00033194032922167994,
+      "loss": 3.114,
+      "step": 28529
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5768436193466187,
+      "learning_rate": 0.0003319250204987789,
+      "loss": 2.6763,
+      "step": 28530
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5673786401748657,
+      "learning_rate": 0.0003319097116917935,
+      "loss": 2.8227,
+      "step": 28531
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6050379276275635,
+      "learning_rate": 0.00033189440280076416,
+      "loss": 2.9611,
+      "step": 28532
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5654118061065674,
+      "learning_rate": 0.000331879093825731,
+      "loss": 3.0181,
+      "step": 28533
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5509759783744812,
+      "learning_rate": 0.00033186378476673446,
+      "loss": 3.2072,
+      "step": 28534
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5696454644203186,
+      "learning_rate": 0.0003318484756238149,
+      "loss": 3.1229,
+      "step": 28535
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.641002357006073,
+      "learning_rate": 0.0003318331663970125,
+      "loss": 3.2484,
+      "step": 28536
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6124339699745178,
+      "learning_rate": 0.00033181785708636774,
+      "loss": 3.158,
+      "step": 28537
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5635184049606323,
+      "learning_rate": 0.0003318025476919208,
+      "loss": 2.9605,
+      "step": 28538
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6408677697181702,
+      "learning_rate": 0.00033178723821371216,
+      "loss": 2.8396,
+      "step": 28539
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5962166786193848,
+      "learning_rate": 0.00033177192865178196,
+      "loss": 3.1288,
+      "step": 28540
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5760708451271057,
+      "learning_rate": 0.00033175661900617065,
+      "loss": 3.0704,
+      "step": 28541
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.667230486869812,
+      "learning_rate": 0.0003317413092769185,
+      "loss": 2.9527,
+      "step": 28542
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5681466460227966,
+      "learning_rate": 0.00033172599946406587,
+      "loss": 3.1464,
+      "step": 28543
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6880142092704773,
+      "learning_rate": 0.0003317106895676531,
+      "loss": 3.2,
+      "step": 28544
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6255538463592529,
+      "learning_rate": 0.0003316953795877204,
+      "loss": 3.1399,
+      "step": 28545
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5599697232246399,
+      "learning_rate": 0.0003316800695243082,
+      "loss": 3.1353,
+      "step": 28546
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5526019930839539,
+      "learning_rate": 0.0003316647593774568,
+      "loss": 2.8375,
+      "step": 28547
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6557570695877075,
+      "learning_rate": 0.0003316494491472065,
+      "loss": 2.9793,
+      "step": 28548
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5730993747711182,
+      "learning_rate": 0.00033163413883359764,
+      "loss": 3.0206,
+      "step": 28549
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1626458168029785,
+      "learning_rate": 0.0003316188284366706,
+      "loss": 3.0879,
+      "step": 28550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5795100927352905,
+      "learning_rate": 0.0003316035179564656,
+      "loss": 2.9078,
+      "step": 28551
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5477898120880127,
+      "learning_rate": 0.000331588207393023,
+      "loss": 2.9622,
+      "step": 28552
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5724733471870422,
+      "learning_rate": 0.0003315728967463832,
+      "loss": 3.2156,
+      "step": 28553
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5982617139816284,
+      "learning_rate": 0.0003315575860165865,
+      "loss": 2.8647,
+      "step": 28554
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6391007900238037,
+      "learning_rate": 0.0003315422752036731,
+      "loss": 3.0067,
+      "step": 28555
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6046311855316162,
+      "learning_rate": 0.00033152696430768347,
+      "loss": 2.9357,
+      "step": 28556
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5996942520141602,
+      "learning_rate": 0.0003315116533286578,
+      "loss": 3.0517,
+      "step": 28557
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5585609078407288,
+      "learning_rate": 0.00033149634226663666,
+      "loss": 3.0389,
+      "step": 28558
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6330249309539795,
+      "learning_rate": 0.0003314810311216601,
+      "loss": 3.0515,
+      "step": 28559
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.574908971786499,
+      "learning_rate": 0.0003314657198937685,
+      "loss": 3.255,
+      "step": 28560
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.597728431224823,
+      "learning_rate": 0.0003314504085830024,
+      "loss": 3.0646,
+      "step": 28561
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6056450009346008,
+      "learning_rate": 0.00033143509718940186,
+      "loss": 3.1887,
+      "step": 28562
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5824889540672302,
+      "learning_rate": 0.0003314197857130074,
+      "loss": 3.1014,
+      "step": 28563
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5638570189476013,
+      "learning_rate": 0.00033140447415385923,
+      "loss": 3.0617,
+      "step": 28564
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5699633359909058,
+      "learning_rate": 0.00033138916251199775,
+      "loss": 3.1922,
+      "step": 28565
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5394684672355652,
+      "learning_rate": 0.0003313738507874632,
+      "loss": 3.1004,
+      "step": 28566
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5617507100105286,
+      "learning_rate": 0.000331358538980296,
+      "loss": 3.1166,
+      "step": 28567
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5685690641403198,
+      "learning_rate": 0.00033134322709053643,
+      "loss": 3.0646,
+      "step": 28568
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5826153755187988,
+      "learning_rate": 0.000331327915118225,
+      "loss": 3.073,
+      "step": 28569
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5799400210380554,
+      "learning_rate": 0.0003313126030634016,
+      "loss": 3.1073,
+      "step": 28570
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6147586703300476,
+      "learning_rate": 0.0003312972909261069,
+      "loss": 3.0393,
+      "step": 28571
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5650718212127686,
+      "learning_rate": 0.0003312819787063813,
+      "loss": 2.536,
+      "step": 28572
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6030925512313843,
+      "learning_rate": 0.00033126666640426487,
+      "loss": 2.9619,
+      "step": 28573
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.573800802230835,
+      "learning_rate": 0.0003312513540197981,
+      "loss": 3.1022,
+      "step": 28574
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5658773183822632,
+      "learning_rate": 0.0003312360415530213,
+      "loss": 3.0579,
+      "step": 28575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5464009642601013,
+      "learning_rate": 0.00033122072900397466,
+      "loss": 3.1756,
+      "step": 28576
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5734522342681885,
+      "learning_rate": 0.00033120541637269874,
+      "loss": 3.1855,
+      "step": 28577
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6184408068656921,
+      "learning_rate": 0.00033119010365923374,
+      "loss": 2.8647,
+      "step": 28578
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6169735789299011,
+      "learning_rate": 0.00033117479086362,
+      "loss": 3.0088,
+      "step": 28579
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5645252466201782,
+      "learning_rate": 0.00033115947798589786,
+      "loss": 3.0772,
+      "step": 28580
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.587335467338562,
+      "learning_rate": 0.0003311441650261076,
+      "loss": 3.1952,
+      "step": 28581
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6208239197731018,
+      "learning_rate": 0.00033112885198428963,
+      "loss": 3.0525,
+      "step": 28582
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5548790097236633,
+      "learning_rate": 0.00033111353886048435,
+      "loss": 3.0921,
+      "step": 28583
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5708760619163513,
+      "learning_rate": 0.00033109822565473187,
+      "loss": 3.0512,
+      "step": 28584
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5541423559188843,
+      "learning_rate": 0.0003310829123670728,
+      "loss": 3.001,
+      "step": 28585
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6086994409561157,
+      "learning_rate": 0.00033106759899754714,
+      "loss": 2.9911,
+      "step": 28586
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5725777745246887,
+      "learning_rate": 0.0003310522855461955,
+      "loss": 3.1316,
+      "step": 28587
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5796346664428711,
+      "learning_rate": 0.0003310369720130581,
+      "loss": 3.2368,
+      "step": 28588
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5890423655509949,
+      "learning_rate": 0.00033102165839817525,
+      "loss": 3.0038,
+      "step": 28589
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5724145770072937,
+      "learning_rate": 0.0003310063447015874,
+      "loss": 3.2116,
+      "step": 28590
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.560804009437561,
+      "learning_rate": 0.0003309910309233348,
+      "loss": 2.8441,
+      "step": 28591
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5579610466957092,
+      "learning_rate": 0.0003309757170634577,
+      "loss": 3.0832,
+      "step": 28592
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5677865147590637,
+      "learning_rate": 0.0003309604031219966,
+      "loss": 3.1131,
+      "step": 28593
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5703995227813721,
+      "learning_rate": 0.00033094508909899177,
+      "loss": 3.0792,
+      "step": 28594
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5485500693321228,
+      "learning_rate": 0.00033092977499448356,
+      "loss": 3.1622,
+      "step": 28595
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5834635496139526,
+      "learning_rate": 0.0003309144608085121,
+      "loss": 3.1295,
+      "step": 28596
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5779139995574951,
+      "learning_rate": 0.000330899146541118,
+      "loss": 3.1709,
+      "step": 28597
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6009939908981323,
+      "learning_rate": 0.00033088383219234154,
+      "loss": 2.9788,
+      "step": 28598
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6160380840301514,
+      "learning_rate": 0.00033086851776222297,
+      "loss": 2.8536,
+      "step": 28599
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5951359868049622,
+      "learning_rate": 0.00033085320325080273,
+      "loss": 3.1106,
+      "step": 28600
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6209423542022705,
+      "learning_rate": 0.000330837888658121,
+      "loss": 3.0604,
+      "step": 28601
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5782553553581238,
+      "learning_rate": 0.0003308225739842182,
+      "loss": 3.2548,
+      "step": 28602
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5685989260673523,
+      "learning_rate": 0.00033080725922913476,
+      "loss": 3.0094,
+      "step": 28603
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5853109955787659,
+      "learning_rate": 0.00033079194439291086,
+      "loss": 3.1369,
+      "step": 28604
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5716216564178467,
+      "learning_rate": 0.000330776629475587,
+      "loss": 3.156,
+      "step": 28605
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6147785186767578,
+      "learning_rate": 0.00033076131447720334,
+      "loss": 2.942,
+      "step": 28606
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5691528916358948,
+      "learning_rate": 0.0003307459993978002,
+      "loss": 3.1663,
+      "step": 28607
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6025173664093018,
+      "learning_rate": 0.00033073068423741814,
+      "loss": 3.1461,
+      "step": 28608
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.59565669298172,
+      "learning_rate": 0.00033071536899609744,
+      "loss": 2.7866,
+      "step": 28609
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.543670654296875,
+      "learning_rate": 0.00033070005367387824,
+      "loss": 3.0595,
+      "step": 28610
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6227362751960754,
+      "learning_rate": 0.000330684738270801,
+      "loss": 3.0906,
+      "step": 28611
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5895864367485046,
+      "learning_rate": 0.00033066942278690614,
+      "loss": 3.08,
+      "step": 28612
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5541119575500488,
+      "learning_rate": 0.0003306541072222339,
+      "loss": 3.106,
+      "step": 28613
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5648583769798279,
+      "learning_rate": 0.0003306387915768246,
+      "loss": 3.2744,
+      "step": 28614
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6211113929748535,
+      "learning_rate": 0.00033062347585071863,
+      "loss": 2.9499,
+      "step": 28615
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6163092255592346,
+      "learning_rate": 0.00033060816004395647,
+      "loss": 3.1507,
+      "step": 28616
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6097185015678406,
+      "learning_rate": 0.0003305928441565781,
+      "loss": 3.0956,
+      "step": 28617
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6374078989028931,
+      "learning_rate": 0.0003305775281886241,
+      "loss": 3.0742,
+      "step": 28618
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6147667169570923,
+      "learning_rate": 0.0003305622121401348,
+      "loss": 3.029,
+      "step": 28619
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5697017312049866,
+      "learning_rate": 0.00033054689601115067,
+      "loss": 3.0943,
+      "step": 28620
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.563630998134613,
+      "learning_rate": 0.0003305315798017117,
+      "loss": 2.928,
+      "step": 28621
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5571799278259277,
+      "learning_rate": 0.00033051626351185843,
+      "loss": 3.0234,
+      "step": 28622
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5704235434532166,
+      "learning_rate": 0.00033050094714163136,
+      "loss": 3.1792,
+      "step": 28623
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5932474732398987,
+      "learning_rate": 0.0003304856306910705,
+      "loss": 3.0762,
+      "step": 28624
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6049175262451172,
+      "learning_rate": 0.00033047031416021637,
+      "loss": 3.0379,
+      "step": 28625
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5922101736068726,
+      "learning_rate": 0.00033045499754910933,
+      "loss": 3.1122,
+      "step": 28626
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5743208527565002,
+      "learning_rate": 0.00033043968085778977,
+      "loss": 3.0287,
+      "step": 28627
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6035799384117126,
+      "learning_rate": 0.00033042436408629785,
+      "loss": 2.9865,
+      "step": 28628
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5719898343086243,
+      "learning_rate": 0.000330409047234674,
+      "loss": 3.0642,
+      "step": 28629
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6188897490501404,
+      "learning_rate": 0.00033039373030295863,
+      "loss": 3.123,
+      "step": 28630
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5590984225273132,
+      "learning_rate": 0.00033037841329119205,
+      "loss": 3.3236,
+      "step": 28631
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5841843485832214,
+      "learning_rate": 0.0003303630961994145,
+      "loss": 3.2134,
+      "step": 28632
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5968827605247498,
+      "learning_rate": 0.00033034777902766636,
+      "loss": 2.9945,
+      "step": 28633
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5639210939407349,
+      "learning_rate": 0.00033033246177598813,
+      "loss": 3.0465,
+      "step": 28634
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5694261193275452,
+      "learning_rate": 0.00033031714444441997,
+      "loss": 3.1198,
+      "step": 28635
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5945818424224854,
+      "learning_rate": 0.0003303018270330023,
+      "loss": 3.1175,
+      "step": 28636
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6064791679382324,
+      "learning_rate": 0.0003302865095417754,
+      "loss": 3.0226,
+      "step": 28637
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5732207298278809,
+      "learning_rate": 0.0003302711919707798,
+      "loss": 3.1993,
+      "step": 28638
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.604217529296875,
+      "learning_rate": 0.00033025587432005554,
+      "loss": 2.9412,
+      "step": 28639
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.589924156665802,
+      "learning_rate": 0.0003302405565896432,
+      "loss": 3.0485,
+      "step": 28640
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5883309841156006,
+      "learning_rate": 0.00033022523877958304,
+      "loss": 2.848,
+      "step": 28641
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.641121506690979,
+      "learning_rate": 0.0003302099208899156,
+      "loss": 3.0409,
+      "step": 28642
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5524410605430603,
+      "learning_rate": 0.00033019460292068074,
+      "loss": 3.0698,
+      "step": 28643
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5567286610603333,
+      "learning_rate": 0.00033017928487191924,
+      "loss": 3.1788,
+      "step": 28644
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5724384784698486,
+      "learning_rate": 0.0003301639667436714,
+      "loss": 3.3344,
+      "step": 28645
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6331698298454285,
+      "learning_rate": 0.00033014864853597736,
+      "loss": 3.0563,
+      "step": 28646
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6174417734146118,
+      "learning_rate": 0.0003301333302488776,
+      "loss": 3.0488,
+      "step": 28647
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5746307373046875,
+      "learning_rate": 0.00033011801188241244,
+      "loss": 3.0419,
+      "step": 28648
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6337699294090271,
+      "learning_rate": 0.0003301026934366223,
+      "loss": 3.0054,
+      "step": 28649
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5989909768104553,
+      "learning_rate": 0.00033008737491154735,
+      "loss": 3.1279,
+      "step": 28650
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6146669387817383,
+      "learning_rate": 0.00033007205630722813,
+      "loss": 2.9037,
+      "step": 28651
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5555586218833923,
+      "learning_rate": 0.00033005673762370486,
+      "loss": 2.9902,
+      "step": 28652
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5725548267364502,
+      "learning_rate": 0.0003300414188610179,
+      "loss": 3.2456,
+      "step": 28653
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5860716700553894,
+      "learning_rate": 0.00033002610001920773,
+      "loss": 3.106,
+      "step": 28654
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5727003812789917,
+      "learning_rate": 0.0003300107810983145,
+      "loss": 3.0971,
+      "step": 28655
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5849184989929199,
+      "learning_rate": 0.0003299954620983788,
+      "loss": 2.8865,
+      "step": 28656
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5811796188354492,
+      "learning_rate": 0.0003299801430194407,
+      "loss": 2.9816,
+      "step": 28657
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5917805433273315,
+      "learning_rate": 0.00032996482386154054,
+      "loss": 3.1353,
+      "step": 28658
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5500307083129883,
+      "learning_rate": 0.0003299495046247189,
+      "loss": 3.0833,
+      "step": 28659
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5789609551429749,
+      "learning_rate": 0.0003299341853090162,
+      "loss": 2.9942,
+      "step": 28660
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.539592981338501,
+      "learning_rate": 0.0003299188659144725,
+      "loss": 3.0309,
+      "step": 28661
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.601234495639801,
+      "learning_rate": 0.00032990354644112823,
+      "loss": 2.9512,
+      "step": 28662
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5849595665931702,
+      "learning_rate": 0.00032988822688902376,
+      "loss": 3.0492,
+      "step": 28663
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.609544038772583,
+      "learning_rate": 0.00032987290725819955,
+      "loss": 3.0828,
+      "step": 28664
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5832148194313049,
+      "learning_rate": 0.00032985758754869575,
+      "loss": 2.9258,
+      "step": 28665
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5789518356323242,
+      "learning_rate": 0.00032984226776055283,
+      "loss": 3.1485,
+      "step": 28666
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5975507497787476,
+      "learning_rate": 0.0003298269478938112,
+      "loss": 3.1122,
+      "step": 28667
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5649939179420471,
+      "learning_rate": 0.000329811627948511,
+      "loss": 3.0364,
+      "step": 28668
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6013187170028687,
+      "learning_rate": 0.0003297963079246928,
+      "loss": 3.1851,
+      "step": 28669
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6145163774490356,
+      "learning_rate": 0.00032978098782239686,
+      "loss": 3.0284,
+      "step": 28670
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5722653865814209,
+      "learning_rate": 0.00032976566764166356,
+      "loss": 3.088,
+      "step": 28671
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.597767174243927,
+      "learning_rate": 0.0003297503473825331,
+      "loss": 3.2594,
+      "step": 28672
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5960759520530701,
+      "learning_rate": 0.000329735027045046,
+      "loss": 3.149,
+      "step": 28673
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.629086434841156,
+      "learning_rate": 0.0003297197066292426,
+      "loss": 3.1178,
+      "step": 28674
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5683742165565491,
+      "learning_rate": 0.00032970438613516323,
+      "loss": 3.1392,
+      "step": 28675
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5865838527679443,
+      "learning_rate": 0.00032968906556284815,
+      "loss": 3.1683,
+      "step": 28676
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5845195651054382,
+      "learning_rate": 0.0003296737449123378,
+      "loss": 2.9377,
+      "step": 28677
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5908820033073425,
+      "learning_rate": 0.0003296584241836726,
+      "loss": 3.0637,
+      "step": 28678
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5620695352554321,
+      "learning_rate": 0.00032964310337689273,
+      "loss": 3.1156,
+      "step": 28679
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5610857009887695,
+      "learning_rate": 0.00032962778249203865,
+      "loss": 3.1601,
+      "step": 28680
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5518953800201416,
+      "learning_rate": 0.0003296124615291507,
+      "loss": 3.0757,
+      "step": 28681
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5811406373977661,
+      "learning_rate": 0.00032959714048826923,
+      "loss": 3.0635,
+      "step": 28682
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5600650310516357,
+      "learning_rate": 0.0003295818193694346,
+      "loss": 2.9543,
+      "step": 28683
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6001834869384766,
+      "learning_rate": 0.00032956649817268707,
+      "loss": 3.0855,
+      "step": 28684
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5461838841438293,
+      "learning_rate": 0.0003295511768980672,
+      "loss": 3.0432,
+      "step": 28685
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5799773335456848,
+      "learning_rate": 0.00032953585554561514,
+      "loss": 2.9155,
+      "step": 28686
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.589601993560791,
+      "learning_rate": 0.0003295205341153713,
+      "loss": 3.1836,
+      "step": 28687
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5646551251411438,
+      "learning_rate": 0.0003295052126073761,
+      "loss": 3.2984,
+      "step": 28688
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5855870246887207,
+      "learning_rate": 0.00032948989102166983,
+      "loss": 3.1553,
+      "step": 28689
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5871852040290833,
+      "learning_rate": 0.00032947456935829284,
+      "loss": 3.1598,
+      "step": 28690
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6054069995880127,
+      "learning_rate": 0.0003294592476172856,
+      "loss": 2.9861,
+      "step": 28691
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.600292980670929,
+      "learning_rate": 0.00032944392579868826,
+      "loss": 3.0694,
+      "step": 28692
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.570586621761322,
+      "learning_rate": 0.0003294286039025414,
+      "loss": 2.9997,
+      "step": 28693
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5904337167739868,
+      "learning_rate": 0.0003294132819288851,
+      "loss": 2.9074,
+      "step": 28694
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5638096332550049,
+      "learning_rate": 0.00032939795987776,
+      "loss": 2.9564,
+      "step": 28695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5465283393859863,
+      "learning_rate": 0.0003293826377492064,
+      "loss": 3.1155,
+      "step": 28696
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5867735147476196,
+      "learning_rate": 0.00032936731554326443,
+      "loss": 3.1527,
+      "step": 28697
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5899142026901245,
+      "learning_rate": 0.0003293519932599747,
+      "loss": 3.0795,
+      "step": 28698
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5677284002304077,
+      "learning_rate": 0.00032933667089937745,
+      "loss": 2.8966,
+      "step": 28699
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5423473715782166,
+      "learning_rate": 0.000329321348461513,
+      "loss": 3.2039,
+      "step": 28700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6081904768943787,
+      "learning_rate": 0.0003293060259464218,
+      "loss": 3.1396,
+      "step": 28701
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5386385321617126,
+      "learning_rate": 0.0003292907033541441,
+      "loss": 3.0957,
+      "step": 28702
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5593324303627014,
+      "learning_rate": 0.0003292753806847205,
+      "loss": 3.0017,
+      "step": 28703
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5604330897331238,
+      "learning_rate": 0.000329260057938191,
+      "loss": 2.9814,
+      "step": 28704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6002910733222961,
+      "learning_rate": 0.0003292447351145962,
+      "loss": 2.9692,
+      "step": 28705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5873032808303833,
+      "learning_rate": 0.00032922941221397644,
+      "loss": 3.2579,
+      "step": 28706
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5412537455558777,
+      "learning_rate": 0.000329214089236372,
+      "loss": 2.7752,
+      "step": 28707
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.58359295129776,
+      "learning_rate": 0.0003291987661818233,
+      "loss": 3.2543,
+      "step": 28708
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5666883587837219,
+      "learning_rate": 0.00032918344305037057,
+      "loss": 3.2266,
+      "step": 28709
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5886451005935669,
+      "learning_rate": 0.00032916811984205427,
+      "loss": 3.0358,
+      "step": 28710
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5730712413787842,
+      "learning_rate": 0.0003291527965569149,
+      "loss": 3.3266,
+      "step": 28711
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6448611617088318,
+      "learning_rate": 0.00032913747319499265,
+      "loss": 2.9833,
+      "step": 28712
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5740808844566345,
+      "learning_rate": 0.00032912214975632783,
+      "loss": 3.1631,
+      "step": 28713
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5787137746810913,
+      "learning_rate": 0.00032910682624096087,
+      "loss": 3.0614,
+      "step": 28714
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6143020391464233,
+      "learning_rate": 0.0003290915026489322,
+      "loss": 2.9691,
+      "step": 28715
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5791074633598328,
+      "learning_rate": 0.000329076178980282,
+      "loss": 2.8866,
+      "step": 28716
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5702418088912964,
+      "learning_rate": 0.0003290608552350508,
+      "loss": 2.997,
+      "step": 28717
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5896664261817932,
+      "learning_rate": 0.000329045531413279,
+      "loss": 2.9138,
+      "step": 28718
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6688372492790222,
+      "learning_rate": 0.00032903020751500676,
+      "loss": 3.1648,
+      "step": 28719
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5435354113578796,
+      "learning_rate": 0.00032901488354027445,
+      "loss": 3.0354,
+      "step": 28720
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5935372710227966,
+      "learning_rate": 0.0003289995594891227,
+      "loss": 3.0646,
+      "step": 28721
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5761470198631287,
+      "learning_rate": 0.00032898423536159165,
+      "loss": 3.2373,
+      "step": 28722
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6226154565811157,
+      "learning_rate": 0.00032896891115772166,
+      "loss": 2.8922,
+      "step": 28723
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5801342129707336,
+      "learning_rate": 0.0003289535868775531,
+      "loss": 3.0508,
+      "step": 28724
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5750948190689087,
+      "learning_rate": 0.0003289382625211264,
+      "loss": 3.152,
+      "step": 28725
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.612395167350769,
+      "learning_rate": 0.00032892293808848183,
+      "loss": 2.9434,
+      "step": 28726
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.607111930847168,
+      "learning_rate": 0.0003289076135796599,
+      "loss": 3.1274,
+      "step": 28727
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5647713541984558,
+      "learning_rate": 0.00032889228899470085,
+      "loss": 3.1467,
+      "step": 28728
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5546459555625916,
+      "learning_rate": 0.0003288769643336451,
+      "loss": 3.2392,
+      "step": 28729
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6006819605827332,
+      "learning_rate": 0.00032886163959653286,
+      "loss": 3.1325,
+      "step": 28730
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.593948245048523,
+      "learning_rate": 0.0003288463147834047,
+      "loss": 3.1686,
+      "step": 28731
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6291419267654419,
+      "learning_rate": 0.0003288309898943009,
+      "loss": 3.0538,
+      "step": 28732
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5162724256515503,
+      "learning_rate": 0.00032881566492926187,
+      "loss": 3.0556,
+      "step": 28733
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6125902533531189,
+      "learning_rate": 0.0003288003398883279,
+      "loss": 3.1117,
+      "step": 28734
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5817069411277771,
+      "learning_rate": 0.0003287850147715393,
+      "loss": 3.0981,
+      "step": 28735
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5820333361625671,
+      "learning_rate": 0.0003287696895789366,
+      "loss": 3.18,
+      "step": 28736
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6199052929878235,
+      "learning_rate": 0.00032875436431056,
+      "loss": 2.9355,
+      "step": 28737
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6050741672515869,
+      "learning_rate": 0.00032873903896645,
+      "loss": 3.0017,
+      "step": 28738
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5911166667938232,
+      "learning_rate": 0.00032872371354664685,
+      "loss": 2.894,
+      "step": 28739
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5618557929992676,
+      "learning_rate": 0.000328708388051191,
+      "loss": 2.8401,
+      "step": 28740
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5915781259536743,
+      "learning_rate": 0.00032869306248012277,
+      "loss": 3.02,
+      "step": 28741
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5654686689376831,
+      "learning_rate": 0.00032867773683348256,
+      "loss": 2.9968,
+      "step": 28742
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.584214448928833,
+      "learning_rate": 0.0003286624111113107,
+      "loss": 3.1967,
+      "step": 28743
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5594285726547241,
+      "learning_rate": 0.0003286470853136476,
+      "loss": 3.0001,
+      "step": 28744
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5778558254241943,
+      "learning_rate": 0.0003286317594405335,
+      "loss": 3.0812,
+      "step": 28745
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6112017631530762,
+      "learning_rate": 0.0003286164334920089,
+      "loss": 2.981,
+      "step": 28746
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5561625361442566,
+      "learning_rate": 0.00032860110746811414,
+      "loss": 2.9896,
+      "step": 28747
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.57763671875,
+      "learning_rate": 0.00032858578136888957,
+      "loss": 2.7896,
+      "step": 28748
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5785109996795654,
+      "learning_rate": 0.00032857045519437555,
+      "loss": 2.9054,
+      "step": 28749
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5656970143318176,
+      "learning_rate": 0.0003285551289446124,
+      "loss": 3.0454,
+      "step": 28750
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5791580080986023,
+      "learning_rate": 0.00032853980261964054,
+      "loss": 2.8888,
+      "step": 28751
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.586801290512085,
+      "learning_rate": 0.0003285244762195004,
+      "loss": 2.8627,
+      "step": 28752
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5646578073501587,
+      "learning_rate": 0.00032850914974423225,
+      "loss": 3.1138,
+      "step": 28753
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5531555414199829,
+      "learning_rate": 0.00032849382319387647,
+      "loss": 3.193,
+      "step": 28754
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5732769966125488,
+      "learning_rate": 0.0003284784965684734,
+      "loss": 3.1193,
+      "step": 28755
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5827401876449585,
+      "learning_rate": 0.00032846316986806355,
+      "loss": 3.1374,
+      "step": 28756
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5628218650817871,
+      "learning_rate": 0.00032844784309268713,
+      "loss": 3.2275,
+      "step": 28757
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5630583167076111,
+      "learning_rate": 0.00032843251624238464,
+      "loss": 3.212,
+      "step": 28758
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5518205165863037,
+      "learning_rate": 0.00032841718931719626,
+      "loss": 2.8222,
+      "step": 28759
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.576407253742218,
+      "learning_rate": 0.00032840186231716253,
+      "loss": 3.1868,
+      "step": 28760
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5976382493972778,
+      "learning_rate": 0.00032838653524232367,
+      "loss": 3.1501,
+      "step": 28761
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5755053758621216,
+      "learning_rate": 0.0003283712080927203,
+      "loss": 2.9968,
+      "step": 28762
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5521357655525208,
+      "learning_rate": 0.00032835588086839256,
+      "loss": 3.211,
+      "step": 28763
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5395427346229553,
+      "learning_rate": 0.0003283405535693809,
+      "loss": 2.959,
+      "step": 28764
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5700335502624512,
+      "learning_rate": 0.00032832522619572564,
+      "loss": 3.0619,
+      "step": 28765
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5535926222801208,
+      "learning_rate": 0.0003283098987474672,
+      "loss": 3.1751,
+      "step": 28766
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6124323606491089,
+      "learning_rate": 0.00032829457122464595,
+      "loss": 3.1244,
+      "step": 28767
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6224178671836853,
+      "learning_rate": 0.0003282792436273022,
+      "loss": 2.879,
+      "step": 28768
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5655115842819214,
+      "learning_rate": 0.0003282639159554765,
+      "loss": 3.1099,
+      "step": 28769
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6045935153961182,
+      "learning_rate": 0.00032824858820920894,
+      "loss": 2.9087,
+      "step": 28770
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5513472557067871,
+      "learning_rate": 0.00032823326038854003,
+      "loss": 2.8912,
+      "step": 28771
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6038599610328674,
+      "learning_rate": 0.0003282179324935102,
+      "loss": 2.9113,
+      "step": 28772
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5618096590042114,
+      "learning_rate": 0.00032820260452415983,
+      "loss": 3.2247,
+      "step": 28773
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6248725652694702,
+      "learning_rate": 0.00032818727648052914,
+      "loss": 3.141,
+      "step": 28774
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5756538510322571,
+      "learning_rate": 0.0003281719483626585,
+      "loss": 3.1785,
+      "step": 28775
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5904281735420227,
+      "learning_rate": 0.0003281566201705885,
+      "loss": 2.9667,
+      "step": 28776
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5620636940002441,
+      "learning_rate": 0.00032814129190435936,
+      "loss": 3.1892,
+      "step": 28777
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5535551905632019,
+      "learning_rate": 0.0003281259635640115,
+      "loss": 3.2308,
+      "step": 28778
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.602520227432251,
+      "learning_rate": 0.0003281106351495852,
+      "loss": 2.8856,
+      "step": 28779
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5640451908111572,
+      "learning_rate": 0.000328095306661121,
+      "loss": 3.0938,
+      "step": 28780
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6214268207550049,
+      "learning_rate": 0.000328079978098659,
+      "loss": 2.7807,
+      "step": 28781
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5931678414344788,
+      "learning_rate": 0.0003280646494622399,
+      "loss": 3.0696,
+      "step": 28782
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5738914012908936,
+      "learning_rate": 0.00032804932075190384,
+      "loss": 3.1592,
+      "step": 28783
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5861164331436157,
+      "learning_rate": 0.00032803399196769135,
+      "loss": 3.0072,
+      "step": 28784
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5898017883300781,
+      "learning_rate": 0.0003280186631096426,
+      "loss": 3.1016,
+      "step": 28785
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5697299242019653,
+      "learning_rate": 0.00032800333417779806,
+      "loss": 3.2083,
+      "step": 28786
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5507596135139465,
+      "learning_rate": 0.00032798800517219826,
+      "loss": 2.8548,
+      "step": 28787
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5449158549308777,
+      "learning_rate": 0.00032797267609288344,
+      "loss": 3.2197,
+      "step": 28788
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5956053733825684,
+      "learning_rate": 0.0003279573469398939,
+      "loss": 3.011,
+      "step": 28789
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6707125902175903,
+      "learning_rate": 0.0003279420177132701,
+      "loss": 3.1171,
+      "step": 28790
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5548729300498962,
+      "learning_rate": 0.00032792668841305245,
+      "loss": 2.96,
+      "step": 28791
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5525285601615906,
+      "learning_rate": 0.00032791135903928123,
+      "loss": 3.0747,
+      "step": 28792
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5493025183677673,
+      "learning_rate": 0.0003278960295919968,
+      "loss": 3.1805,
+      "step": 28793
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5797526836395264,
+      "learning_rate": 0.00032788070007123965,
+      "loss": 2.8174,
+      "step": 28794
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5741456151008606,
+      "learning_rate": 0.00032786537047705023,
+      "loss": 3.1977,
+      "step": 28795
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5887008905410767,
+      "learning_rate": 0.0003278500408094686,
+      "loss": 2.9331,
+      "step": 28796
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6311918497085571,
+      "learning_rate": 0.00032783471106853537,
+      "loss": 2.8529,
+      "step": 28797
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5639904141426086,
+      "learning_rate": 0.00032781938125429097,
+      "loss": 2.9302,
+      "step": 28798
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5747509002685547,
+      "learning_rate": 0.0003278040513667756,
+      "loss": 3.0155,
+      "step": 28799
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5683767795562744,
+      "learning_rate": 0.00032778872140602967,
+      "loss": 3.1515,
+      "step": 28800
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5533831119537354,
+      "learning_rate": 0.00032777339137209365,
+      "loss": 2.9977,
+      "step": 28801
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5720163583755493,
+      "learning_rate": 0.0003277580612650078,
+      "loss": 3.1106,
+      "step": 28802
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.605758011341095,
+      "learning_rate": 0.0003277427310848127,
+      "loss": 3.0415,
+      "step": 28803
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5644210577011108,
+      "learning_rate": 0.00032772740083154844,
+      "loss": 2.9584,
+      "step": 28804
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.603422999382019,
+      "learning_rate": 0.0003277120705052556,
+      "loss": 2.9925,
+      "step": 28805
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5727534890174866,
+      "learning_rate": 0.00032769674010597454,
+      "loss": 3.1375,
+      "step": 28806
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5674635171890259,
+      "learning_rate": 0.0003276814096337455,
+      "loss": 3.0833,
+      "step": 28807
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5521325469017029,
+      "learning_rate": 0.00032766607908860904,
+      "loss": 2.9992,
+      "step": 28808
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5518462657928467,
+      "learning_rate": 0.0003276507484706054,
+      "loss": 3.1441,
+      "step": 28809
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5592528581619263,
+      "learning_rate": 0.0003276354177797751,
+      "loss": 3.0381,
+      "step": 28810
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5564046502113342,
+      "learning_rate": 0.00032762008701615826,
+      "loss": 3.1174,
+      "step": 28811
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5931382775306702,
+      "learning_rate": 0.0003276047561797955,
+      "loss": 3.051,
+      "step": 28812
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5513250231742859,
+      "learning_rate": 0.0003275894252707272,
+      "loss": 2.845,
+      "step": 28813
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5776057839393616,
+      "learning_rate": 0.0003275740942889936,
+      "loss": 3.111,
+      "step": 28814
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5935764908790588,
+      "learning_rate": 0.00032755876323463514,
+      "loss": 3.1766,
+      "step": 28815
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6411653757095337,
+      "learning_rate": 0.00032754343210769226,
+      "loss": 3.037,
+      "step": 28816
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5873118042945862,
+      "learning_rate": 0.0003275281009082052,
+      "loss": 3.1031,
+      "step": 28817
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.57716965675354,
+      "learning_rate": 0.0003275127696362144,
+      "loss": 3.1625,
+      "step": 28818
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5541166067123413,
+      "learning_rate": 0.00032749743829176036,
+      "loss": 2.9424,
+      "step": 28819
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5837197303771973,
+      "learning_rate": 0.0003274821068748834,
+      "loss": 3.1427,
+      "step": 28820
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6180403828620911,
+      "learning_rate": 0.00032746677538562376,
+      "loss": 3.0543,
+      "step": 28821
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5609276294708252,
+      "learning_rate": 0.00032745144382402184,
+      "loss": 3.2852,
+      "step": 28822
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5260721445083618,
+      "learning_rate": 0.00032743611219011816,
+      "loss": 3.1377,
+      "step": 28823
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.568089485168457,
+      "learning_rate": 0.00032742078048395316,
+      "loss": 3.2754,
+      "step": 28824
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5611562728881836,
+      "learning_rate": 0.000327405448705567,
+      "loss": 2.9644,
+      "step": 28825
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5755119919776917,
+      "learning_rate": 0.00032739011685500017,
+      "loss": 3.0372,
+      "step": 28826
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.579472541809082,
+      "learning_rate": 0.00032737478493229306,
+      "loss": 3.2508,
+      "step": 28827
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5705549716949463,
+      "learning_rate": 0.00032735945293748607,
+      "loss": 3.0324,
+      "step": 28828
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.582402229309082,
+      "learning_rate": 0.0003273441208706194,
+      "loss": 2.9384,
+      "step": 28829
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5975726842880249,
+      "learning_rate": 0.00032732878873173373,
+      "loss": 2.7791,
+      "step": 28830
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.546567440032959,
+      "learning_rate": 0.0003273134565208693,
+      "loss": 3.1615,
+      "step": 28831
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5662059783935547,
+      "learning_rate": 0.0003272981242380664,
+      "loss": 3.037,
+      "step": 28832
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.567099392414093,
+      "learning_rate": 0.00032728279188336544,
+      "loss": 3.0003,
+      "step": 28833
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.584934651851654,
+      "learning_rate": 0.0003272674594568069,
+      "loss": 2.908,
+      "step": 28834
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5936591029167175,
+      "learning_rate": 0.0003272521269584313,
+      "loss": 3.0655,
+      "step": 28835
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5734385251998901,
+      "learning_rate": 0.0003272367943882787,
+      "loss": 2.9705,
+      "step": 28836
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5487362742424011,
+      "learning_rate": 0.0003272214617463896,
+      "loss": 3.0344,
+      "step": 28837
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5989526510238647,
+      "learning_rate": 0.0003272061290328044,
+      "loss": 3.11,
+      "step": 28838
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5572174191474915,
+      "learning_rate": 0.00032719079624756353,
+      "loss": 2.9015,
+      "step": 28839
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5623894929885864,
+      "learning_rate": 0.00032717546339070734,
+      "loss": 2.982,
+      "step": 28840
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6148263216018677,
+      "learning_rate": 0.0003271601304622762,
+      "loss": 2.9311,
+      "step": 28841
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5677447319030762,
+      "learning_rate": 0.0003271447974623105,
+      "loss": 3.0023,
+      "step": 28842
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5840166807174683,
+      "learning_rate": 0.0003271294643908506,
+      "loss": 3.0108,
+      "step": 28843
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5517377853393555,
+      "learning_rate": 0.00032711413124793693,
+      "loss": 3.1285,
+      "step": 28844
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5746965408325195,
+      "learning_rate": 0.0003270987980336099,
+      "loss": 3.0579,
+      "step": 28845
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5635104775428772,
+      "learning_rate": 0.00032708346474790985,
+      "loss": 2.8986,
+      "step": 28846
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5739995241165161,
+      "learning_rate": 0.00032706813139087715,
+      "loss": 3.0193,
+      "step": 28847
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5986478328704834,
+      "learning_rate": 0.0003270527979625521,
+      "loss": 3.2475,
+      "step": 28848
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5273615121841431,
+      "learning_rate": 0.0003270374644629753,
+      "loss": 3.2175,
+      "step": 28849
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5814329981803894,
+      "learning_rate": 0.000327022130892187,
+      "loss": 2.9465,
+      "step": 28850
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5327739715576172,
+      "learning_rate": 0.0003270067972502276,
+      "loss": 3.0869,
+      "step": 28851
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5583226680755615,
+      "learning_rate": 0.0003269914635371375,
+      "loss": 3.0564,
+      "step": 28852
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5671470761299133,
+      "learning_rate": 0.00032697612975295706,
+      "loss": 3.3398,
+      "step": 28853
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6003071665763855,
+      "learning_rate": 0.00032696079589772674,
+      "loss": 2.9829,
+      "step": 28854
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5671707391738892,
+      "learning_rate": 0.00032694546197148676,
+      "loss": 3.1985,
+      "step": 28855
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5648956298828125,
+      "learning_rate": 0.00032693012797427773,
+      "loss": 3.1373,
+      "step": 28856
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5903165936470032,
+      "learning_rate": 0.0003269147939061399,
+      "loss": 2.9278,
+      "step": 28857
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5882886648178101,
+      "learning_rate": 0.00032689945976711366,
+      "loss": 3.102,
+      "step": 28858
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6090853810310364,
+      "learning_rate": 0.00032688412555723946,
+      "loss": 3.0155,
+      "step": 28859
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5981293320655823,
+      "learning_rate": 0.0003268687912765576,
+      "loss": 2.9891,
+      "step": 28860
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6331175565719604,
+      "learning_rate": 0.0003268534569251086,
+      "loss": 2.9693,
+      "step": 28861
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5615060925483704,
+      "learning_rate": 0.0003268381225029327,
+      "loss": 3.0143,
+      "step": 28862
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5723454356193542,
+      "learning_rate": 0.0003268227880100703,
+      "loss": 3.0968,
+      "step": 28863
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5850404500961304,
+      "learning_rate": 0.0003268074534465619,
+      "loss": 2.9922,
+      "step": 28864
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5858272910118103,
+      "learning_rate": 0.0003267921188124478,
+      "loss": 3.0211,
+      "step": 28865
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5558182597160339,
+      "learning_rate": 0.0003267767841077684,
+      "loss": 2.8762,
+      "step": 28866
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.570468544960022,
+      "learning_rate": 0.0003267614493325641,
+      "loss": 3.2,
+      "step": 28867
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5738118886947632,
+      "learning_rate": 0.0003267461144868753,
+      "loss": 2.9635,
+      "step": 28868
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8501331806182861,
+      "learning_rate": 0.00032673077957074244,
+      "loss": 3.1468,
+      "step": 28869
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.582685112953186,
+      "learning_rate": 0.00032671544458420587,
+      "loss": 2.9663,
+      "step": 28870
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5798792839050293,
+      "learning_rate": 0.0003267001095273059,
+      "loss": 3.1031,
+      "step": 28871
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9409517645835876,
+      "learning_rate": 0.000326684774400083,
+      "loss": 3.17,
+      "step": 28872
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5941800475120544,
+      "learning_rate": 0.00032666943920257754,
+      "loss": 3.1101,
+      "step": 28873
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5913896560668945,
+      "learning_rate": 0.0003266541039348298,
+      "loss": 3.1029,
+      "step": 28874
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5694225430488586,
+      "learning_rate": 0.00032663876859688045,
+      "loss": 3.1596,
+      "step": 28875
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.600014865398407,
+      "learning_rate": 0.00032662343318876964,
+      "loss": 3.0626,
+      "step": 28876
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5729812383651733,
+      "learning_rate": 0.00032660809771053784,
+      "loss": 2.9556,
+      "step": 28877
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5758359432220459,
+      "learning_rate": 0.0003265927621622254,
+      "loss": 3.1092,
+      "step": 28878
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5663775205612183,
+      "learning_rate": 0.0003265774265438727,
+      "loss": 3.0048,
+      "step": 28879
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5517311692237854,
+      "learning_rate": 0.0003265620908555203,
+      "loss": 2.9829,
+      "step": 28880
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.550212025642395,
+      "learning_rate": 0.0003265467550972084,
+      "loss": 3.1944,
+      "step": 28881
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5680566430091858,
+      "learning_rate": 0.00032653141926897747,
+      "loss": 3.0897,
+      "step": 28882
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5950158834457397,
+      "learning_rate": 0.0003265160833708679,
+      "loss": 3.0992,
+      "step": 28883
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.2692245244979858,
+      "learning_rate": 0.00032650074740291996,
+      "loss": 3.2001,
+      "step": 28884
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.560718297958374,
+      "learning_rate": 0.0003264854113651742,
+      "loss": 3.074,
+      "step": 28885
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6030946373939514,
+      "learning_rate": 0.00032647007525767106,
+      "loss": 3.1256,
+      "step": 28886
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.580146312713623,
+      "learning_rate": 0.0003264547390804508,
+      "loss": 3.1419,
+      "step": 28887
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6022080779075623,
+      "learning_rate": 0.0003264394028335538,
+      "loss": 3.111,
+      "step": 28888
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6084839105606079,
+      "learning_rate": 0.00032642406651702053,
+      "loss": 3.088,
+      "step": 28889
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.767863392829895,
+      "learning_rate": 0.00032640873013089133,
+      "loss": 3.033,
+      "step": 28890
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5766077041625977,
+      "learning_rate": 0.00032639339367520665,
+      "loss": 2.9984,
+      "step": 28891
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5633019804954529,
+      "learning_rate": 0.00032637805715000684,
+      "loss": 3.1288,
+      "step": 28892
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5685615539550781,
+      "learning_rate": 0.0003263627205553323,
+      "loss": 3.0734,
+      "step": 28893
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6013651490211487,
+      "learning_rate": 0.0003263473838912234,
+      "loss": 2.937,
+      "step": 28894
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6018322706222534,
+      "learning_rate": 0.0003263320471577205,
+      "loss": 3.0378,
+      "step": 28895
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6196945905685425,
+      "learning_rate": 0.00032631671035486423,
+      "loss": 3.0361,
+      "step": 28896
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5838954448699951,
+      "learning_rate": 0.0003263013734826948,
+      "loss": 3.1028,
+      "step": 28897
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5709912776947021,
+      "learning_rate": 0.00032628603654125246,
+      "loss": 3.1711,
+      "step": 28898
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5789138674736023,
+      "learning_rate": 0.0003262706995305778,
+      "loss": 3.2224,
+      "step": 28899
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6325254440307617,
+      "learning_rate": 0.0003262553624507113,
+      "loss": 3.0815,
+      "step": 28900
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5830736756324768,
+      "learning_rate": 0.0003262400253016931,
+      "loss": 3.0138,
+      "step": 28901
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6032917499542236,
+      "learning_rate": 0.00032622468808356377,
+      "loss": 3.0851,
+      "step": 28902
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6263208985328674,
+      "learning_rate": 0.00032620935079636363,
+      "loss": 3.0902,
+      "step": 28903
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6103036403656006,
+      "learning_rate": 0.0003261940134401331,
+      "loss": 2.7757,
+      "step": 28904
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5601415038108826,
+      "learning_rate": 0.00032617867601491266,
+      "loss": 3.0745,
+      "step": 28905
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5642362833023071,
+      "learning_rate": 0.00032616333852074255,
+      "loss": 3.1547,
+      "step": 28906
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6236977577209473,
+      "learning_rate": 0.0003261480009576632,
+      "loss": 2.9319,
+      "step": 28907
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6301894187927246,
+      "learning_rate": 0.0003261326633257152,
+      "loss": 3.04,
+      "step": 28908
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5724765658378601,
+      "learning_rate": 0.00032611732562493864,
+      "loss": 3.1385,
+      "step": 28909
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5792955756187439,
+      "learning_rate": 0.00032610198785537414,
+      "loss": 3.0763,
+      "step": 28910
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5876222848892212,
+      "learning_rate": 0.0003260866500170621,
+      "loss": 2.9057,
+      "step": 28911
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5676042437553406,
+      "learning_rate": 0.0003260713121100427,
+      "loss": 3.0592,
+      "step": 28912
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6063807010650635,
+      "learning_rate": 0.0003260559741343566,
+      "loss": 3.0202,
+      "step": 28913
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5811011791229248,
+      "learning_rate": 0.000326040636090044,
+      "loss": 3.1869,
+      "step": 28914
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5812034606933594,
+      "learning_rate": 0.0003260252979771454,
+      "loss": 2.9437,
+      "step": 28915
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.57767254114151,
+      "learning_rate": 0.00032600995979570113,
+      "loss": 3.1279,
+      "step": 28916
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5982522368431091,
+      "learning_rate": 0.00032599462154575174,
+      "loss": 3.0572,
+      "step": 28917
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6943778395652771,
+      "learning_rate": 0.0003259792832273374,
+      "loss": 3.1348,
+      "step": 28918
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5726862549781799,
+      "learning_rate": 0.0003259639448404988,
+      "loss": 3.3887,
+      "step": 28919
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5609963536262512,
+      "learning_rate": 0.000325948606385276,
+      "loss": 3.0386,
+      "step": 28920
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.573071300983429,
+      "learning_rate": 0.0003259332678617095,
+      "loss": 3.2499,
+      "step": 28921
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5977573990821838,
+      "learning_rate": 0.00032591792926983995,
+      "loss": 3.0938,
+      "step": 28922
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5285585522651672,
+      "learning_rate": 0.00032590259060970755,
+      "loss": 3.3141,
+      "step": 28923
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6297756433486938,
+      "learning_rate": 0.0003258872518813526,
+      "loss": 3.0633,
+      "step": 28924
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5830036401748657,
+      "learning_rate": 0.00032587191308481564,
+      "loss": 3.0276,
+      "step": 28925
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6116616725921631,
+      "learning_rate": 0.0003258565742201371,
+      "loss": 2.7825,
+      "step": 28926
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5653083920478821,
+      "learning_rate": 0.00032584123528735725,
+      "loss": 3.0379,
+      "step": 28927
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6250095963478088,
+      "learning_rate": 0.00032582589628651665,
+      "loss": 3.0878,
+      "step": 28928
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5555509924888611,
+      "learning_rate": 0.00032581055721765553,
+      "loss": 2.9807,
+      "step": 28929
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5599590539932251,
+      "learning_rate": 0.00032579521808081435,
+      "loss": 2.8723,
+      "step": 28930
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.589675784111023,
+      "learning_rate": 0.0003257798788760336,
+      "loss": 3.0035,
+      "step": 28931
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5757635831832886,
+      "learning_rate": 0.00032576453960335357,
+      "loss": 3.2037,
+      "step": 28932
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5922831892967224,
+      "learning_rate": 0.0003257492002628148,
+      "loss": 3.1709,
+      "step": 28933
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5696724653244019,
+      "learning_rate": 0.00032573386085445747,
+      "loss": 3.0498,
+      "step": 28934
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6184026002883911,
+      "learning_rate": 0.0003257185213783221,
+      "loss": 2.9888,
+      "step": 28935
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.552787184715271,
+      "learning_rate": 0.00032570318183444913,
+      "loss": 3.1177,
+      "step": 28936
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6209615468978882,
+      "learning_rate": 0.00032568784222287896,
+      "loss": 3.0414,
+      "step": 28937
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5979040265083313,
+      "learning_rate": 0.0003256725025436519,
+      "loss": 2.9614,
+      "step": 28938
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6105141043663025,
+      "learning_rate": 0.0003256571627968084,
+      "loss": 2.9355,
+      "step": 28939
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.564461350440979,
+      "learning_rate": 0.000325641822982389,
+      "loss": 2.9066,
+      "step": 28940
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5573233962059021,
+      "learning_rate": 0.0003256264831004338,
+      "loss": 2.9489,
+      "step": 28941
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5542727112770081,
+      "learning_rate": 0.0003256111431509835,
+      "loss": 2.9582,
+      "step": 28942
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5618546605110168,
+      "learning_rate": 0.0003255958031340783,
+      "loss": 3.084,
+      "step": 28943
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5816484093666077,
+      "learning_rate": 0.00032558046304975876,
+      "loss": 2.9806,
+      "step": 28944
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6100296974182129,
+      "learning_rate": 0.0003255651228980651,
+      "loss": 3.202,
+      "step": 28945
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5640974640846252,
+      "learning_rate": 0.00032554978267903786,
+      "loss": 3.1282,
+      "step": 28946
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5738323330879211,
+      "learning_rate": 0.00032553444239271745,
+      "loss": 3.0857,
+      "step": 28947
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6203998923301697,
+      "learning_rate": 0.0003255191020391443,
+      "loss": 3.0114,
+      "step": 28948
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.570510983467102,
+      "learning_rate": 0.00032550376161835865,
+      "loss": 3.0182,
+      "step": 28949
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5366271138191223,
+      "learning_rate": 0.0003254884211304009,
+      "loss": 2.9167,
+      "step": 28950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5703023672103882,
+      "learning_rate": 0.00032547308057531173,
+      "loss": 3.073,
+      "step": 28951
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.57989501953125,
+      "learning_rate": 0.0003254577399531313,
+      "loss": 2.995,
+      "step": 28952
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5939070582389832,
+      "learning_rate": 0.0003254423992639001,
+      "loss": 3.1934,
+      "step": 28953
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5694959759712219,
+      "learning_rate": 0.00032542705850765845,
+      "loss": 3.1665,
+      "step": 28954
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5695184469223022,
+      "learning_rate": 0.000325411717684447,
+      "loss": 2.9085,
+      "step": 28955
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6038377285003662,
+      "learning_rate": 0.00032539637679430576,
+      "loss": 2.978,
+      "step": 28956
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5637387037277222,
+      "learning_rate": 0.00032538103583727544,
+      "loss": 3.1269,
+      "step": 28957
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6748733520507812,
+      "learning_rate": 0.00032536569481339637,
+      "loss": 3.1292,
+      "step": 28958
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5864824652671814,
+      "learning_rate": 0.00032535035372270897,
+      "loss": 3.1735,
+      "step": 28959
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6003410220146179,
+      "learning_rate": 0.0003253350125652536,
+      "loss": 3.1991,
+      "step": 28960
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5306596159934998,
+      "learning_rate": 0.0003253196713410706,
+      "loss": 2.9204,
+      "step": 28961
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5750020742416382,
+      "learning_rate": 0.0003253043300502006,
+      "loss": 3.019,
+      "step": 28962
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5885328054428101,
+      "learning_rate": 0.0003252889886926838,
+      "loss": 2.971,
+      "step": 28963
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6015452146530151,
+      "learning_rate": 0.0003252736472685606,
+      "loss": 3.061,
+      "step": 28964
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5836628675460815,
+      "learning_rate": 0.00032525830577787154,
+      "loss": 3.0505,
+      "step": 28965
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5610659718513489,
+      "learning_rate": 0.00032524296422065693,
+      "loss": 3.1,
+      "step": 28966
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5679178237915039,
+      "learning_rate": 0.0003252276225969572,
+      "loss": 3.0292,
+      "step": 28967
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.541135311126709,
+      "learning_rate": 0.0003252122809068128,
+      "loss": 2.9467,
+      "step": 28968
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6564010977745056,
+      "learning_rate": 0.0003251969391502641,
+      "loss": 2.9938,
+      "step": 28969
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5690862536430359,
+      "learning_rate": 0.0003251815973273516,
+      "loss": 3.131,
+      "step": 28970
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.56672203540802,
+      "learning_rate": 0.0003251662554381154,
+      "loss": 3.0208,
+      "step": 28971
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6128435730934143,
+      "learning_rate": 0.0003251509134825963,
+      "loss": 2.9169,
+      "step": 28972
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5664635896682739,
+      "learning_rate": 0.00032513557146083454,
+      "loss": 3.0566,
+      "step": 28973
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5436082482337952,
+      "learning_rate": 0.0003251202293728705,
+      "loss": 3.0333,
+      "step": 28974
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6189156770706177,
+      "learning_rate": 0.0003251048872187446,
+      "loss": 2.9735,
+      "step": 28975
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5798361897468567,
+      "learning_rate": 0.0003250895449984972,
+      "loss": 3.1293,
+      "step": 28976
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.584924042224884,
+      "learning_rate": 0.00032507420271216884,
+      "loss": 2.9706,
+      "step": 28977
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.597545862197876,
+      "learning_rate": 0.0003250588603597998,
+      "loss": 3.2138,
+      "step": 28978
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5903199315071106,
+      "learning_rate": 0.0003250435179414306,
+      "loss": 3.1368,
+      "step": 28979
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6203091144561768,
+      "learning_rate": 0.0003250281754571016,
+      "loss": 3.1918,
+      "step": 28980
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6048563718795776,
+      "learning_rate": 0.0003250128329068531,
+      "loss": 3.1939,
+      "step": 28981
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5847277641296387,
+      "learning_rate": 0.0003249974902907257,
+      "loss": 3.2304,
+      "step": 28982
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6675156354904175,
+      "learning_rate": 0.0003249821476087597,
+      "loss": 2.9924,
+      "step": 28983
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5942085385322571,
+      "learning_rate": 0.00032496680486099566,
+      "loss": 2.9132,
+      "step": 28984
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6033164858818054,
+      "learning_rate": 0.0003249514620474737,
+      "loss": 3.0188,
+      "step": 28985
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5678269863128662,
+      "learning_rate": 0.00032493611916823437,
+      "loss": 3.0577,
+      "step": 28986
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5534139275550842,
+      "learning_rate": 0.0003249207762233181,
+      "loss": 2.9875,
+      "step": 28987
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5515173673629761,
+      "learning_rate": 0.00032490543321276547,
+      "loss": 3.0418,
+      "step": 28988
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.585387110710144,
+      "learning_rate": 0.0003248900901366166,
+      "loss": 3.2985,
+      "step": 28989
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5718480944633484,
+      "learning_rate": 0.00032487474699491206,
+      "loss": 2.9427,
+      "step": 28990
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5717833638191223,
+      "learning_rate": 0.0003248594037876922,
+      "loss": 2.9248,
+      "step": 28991
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5640171766281128,
+      "learning_rate": 0.0003248440605149974,
+      "loss": 3.0837,
+      "step": 28992
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5515601634979248,
+      "learning_rate": 0.00032482871717686817,
+      "loss": 3.1191,
+      "step": 28993
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5653744339942932,
+      "learning_rate": 0.00032481337377334487,
+      "loss": 3.1291,
+      "step": 28994
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5796131491661072,
+      "learning_rate": 0.00032479803030446803,
+      "loss": 3.0333,
+      "step": 28995
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5645726919174194,
+      "learning_rate": 0.0003247826867702778,
+      "loss": 3.1165,
+      "step": 28996
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5933886170387268,
+      "learning_rate": 0.0003247673431708148,
+      "loss": 3.0208,
+      "step": 28997
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5697206258773804,
+      "learning_rate": 0.00032475199950611933,
+      "loss": 2.9842,
+      "step": 28998
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5712945461273193,
+      "learning_rate": 0.000324736655776232,
+      "loss": 3.1261,
+      "step": 28999
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.689601480960846,
+      "learning_rate": 0.000324721311981193,
+      "loss": 3.2871,
+      "step": 29000
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5841153264045715,
+      "learning_rate": 0.00032470596812104275,
+      "loss": 3.1395,
+      "step": 29001
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6104511618614197,
+      "learning_rate": 0.0003246906241958218,
+      "loss": 3.1197,
+      "step": 29002
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.575042724609375,
+      "learning_rate": 0.0003246752802055705,
+      "loss": 2.9056,
+      "step": 29003
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6196618676185608,
+      "learning_rate": 0.0003246599361503292,
+      "loss": 2.9237,
+      "step": 29004
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.607953667640686,
+      "learning_rate": 0.00032464459203013847,
+      "loss": 2.9469,
+      "step": 29005
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.574377179145813,
+      "learning_rate": 0.00032462924784503865,
+      "loss": 2.9129,
+      "step": 29006
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5855792760848999,
+      "learning_rate": 0.00032461390359507,
+      "loss": 3.2642,
+      "step": 29007
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5718584656715393,
+      "learning_rate": 0.00032459855928027313,
+      "loss": 2.9375,
+      "step": 29008
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6099596619606018,
+      "learning_rate": 0.00032458321490068837,
+      "loss": 3.0389,
+      "step": 29009
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5883126854896545,
+      "learning_rate": 0.00032456787045635624,
+      "loss": 2.9312,
+      "step": 29010
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5781348943710327,
+      "learning_rate": 0.00032455252594731704,
+      "loss": 2.8841,
+      "step": 29011
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5610764026641846,
+      "learning_rate": 0.00032453718137361113,
+      "loss": 2.9505,
+      "step": 29012
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5973708033561707,
+      "learning_rate": 0.0003245218367352791,
+      "loss": 2.8758,
+      "step": 29013
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6375737190246582,
+      "learning_rate": 0.00032450649203236116,
+      "loss": 2.9811,
+      "step": 29014
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6129302382469177,
+      "learning_rate": 0.00032449114726489786,
+      "loss": 3.2167,
+      "step": 29015
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5988703966140747,
+      "learning_rate": 0.0003244758024329297,
+      "loss": 3.0494,
+      "step": 29016
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6101521253585815,
+      "learning_rate": 0.0003244604575364969,
+      "loss": 3.0984,
+      "step": 29017
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5767313838005066,
+      "learning_rate": 0.00032444511257564,
+      "loss": 3.0744,
+      "step": 29018
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6167458891868591,
+      "learning_rate": 0.0003244297675503993,
+      "loss": 3.1639,
+      "step": 29019
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5684388875961304,
+      "learning_rate": 0.0003244144224608154,
+      "loss": 2.9517,
+      "step": 29020
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5900117754936218,
+      "learning_rate": 0.0003243990773069286,
+      "loss": 3.0036,
+      "step": 29021
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.72846919298172,
+      "learning_rate": 0.0003243837320887792,
+      "loss": 3.1484,
+      "step": 29022
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5999704003334045,
+      "learning_rate": 0.0003243683868064078,
+      "loss": 3.0344,
+      "step": 29023
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5632615089416504,
+      "learning_rate": 0.0003243530414598549,
+      "loss": 3.179,
+      "step": 29024
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.645645022392273,
+      "learning_rate": 0.00032433769604916065,
+      "loss": 3.1495,
+      "step": 29025
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6163223385810852,
+      "learning_rate": 0.0003243223505743656,
+      "loss": 2.983,
+      "step": 29026
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5711262226104736,
+      "learning_rate": 0.0003243070050355102,
+      "loss": 3.2788,
+      "step": 29027
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5460343956947327,
+      "learning_rate": 0.00032429165943263473,
+      "loss": 2.9364,
+      "step": 29028
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5538520216941833,
+      "learning_rate": 0.00032427631376577977,
+      "loss": 2.8617,
+      "step": 29029
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5613263845443726,
+      "learning_rate": 0.00032426096803498566,
+      "loss": 3.1502,
+      "step": 29030
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5788880586624146,
+      "learning_rate": 0.00032424562224029284,
+      "loss": 3.0286,
+      "step": 29031
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6063810586929321,
+      "learning_rate": 0.00032423027638174177,
+      "loss": 2.8803,
+      "step": 29032
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5913493633270264,
+      "learning_rate": 0.00032421493045937266,
+      "loss": 2.9329,
+      "step": 29033
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6406772136688232,
+      "learning_rate": 0.0003241995844732262,
+      "loss": 3.2342,
+      "step": 29034
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5965743064880371,
+      "learning_rate": 0.00032418423842334274,
+      "loss": 2.9745,
+      "step": 29035
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5714675188064575,
+      "learning_rate": 0.00032416889230976265,
+      "loss": 2.8127,
+      "step": 29036
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5749967694282532,
+      "learning_rate": 0.0003241535461325263,
+      "loss": 3.0176,
+      "step": 29037
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5987011194229126,
+      "learning_rate": 0.000324138199891674,
+      "loss": 3.1697,
+      "step": 29038
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5846742391586304,
+      "learning_rate": 0.00032412285358724655,
+      "loss": 3.1089,
+      "step": 29039
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6523012518882751,
+      "learning_rate": 0.00032410750721928406,
+      "loss": 2.933,
+      "step": 29040
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5596588850021362,
+      "learning_rate": 0.00032409216078782706,
+      "loss": 3.1106,
+      "step": 29041
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5636228322982788,
+      "learning_rate": 0.00032407681429291594,
+      "loss": 3.2038,
+      "step": 29042
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5773536562919617,
+      "learning_rate": 0.0003240614677345911,
+      "loss": 3.2443,
+      "step": 29043
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6011802554130554,
+      "learning_rate": 0.0003240461211128929,
+      "loss": 3.0798,
+      "step": 29044
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5544044971466064,
+      "learning_rate": 0.00032403077442786195,
+      "loss": 3.0196,
+      "step": 29045
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6246320605278015,
+      "learning_rate": 0.0003240154276795387,
+      "loss": 2.8156,
+      "step": 29046
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5959727764129639,
+      "learning_rate": 0.0003240000808679633,
+      "loss": 3.1569,
+      "step": 29047
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.579071581363678,
+      "learning_rate": 0.00032398473399317617,
+      "loss": 3.1073,
+      "step": 29048
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5742684006690979,
+      "learning_rate": 0.000323969387055218,
+      "loss": 3.3636,
+      "step": 29049
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5994040966033936,
+      "learning_rate": 0.0003239540400541291,
+      "loss": 3.1828,
+      "step": 29050
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5804879069328308,
+      "learning_rate": 0.0003239386929899499,
+      "loss": 2.8866,
+      "step": 29051
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5392642021179199,
+      "learning_rate": 0.0003239233458627207,
+      "loss": 3.1973,
+      "step": 29052
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5686118602752686,
+      "learning_rate": 0.000323907998672482,
+      "loss": 2.977,
+      "step": 29053
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5497790575027466,
+      "learning_rate": 0.00032389265141927426,
+      "loss": 2.9656,
+      "step": 29054
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5894394516944885,
+      "learning_rate": 0.00032387730410313794,
+      "loss": 2.9002,
+      "step": 29055
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5542726516723633,
+      "learning_rate": 0.0003238619567241133,
+      "loss": 3.0072,
+      "step": 29056
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5575839877128601,
+      "learning_rate": 0.00032384660928224095,
+      "loss": 3.0584,
+      "step": 29057
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5948037505149841,
+      "learning_rate": 0.0003238312617775611,
+      "loss": 3.1864,
+      "step": 29058
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5639663934707642,
+      "learning_rate": 0.0003238159142101144,
+      "loss": 3.0449,
+      "step": 29059
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5659995079040527,
+      "learning_rate": 0.0003238005665799411,
+      "loss": 3.1618,
+      "step": 29060
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5502591133117676,
+      "learning_rate": 0.0003237852188870818,
+      "loss": 3.0585,
+      "step": 29061
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6027945280075073,
+      "learning_rate": 0.00032376987113157666,
+      "loss": 3.142,
+      "step": 29062
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5965318083763123,
+      "learning_rate": 0.00032375452331346634,
+      "loss": 3.0064,
+      "step": 29063
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5846608877182007,
+      "learning_rate": 0.00032373917543279115,
+      "loss": 3.1892,
+      "step": 29064
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6004025340080261,
+      "learning_rate": 0.00032372382748959165,
+      "loss": 3.1346,
+      "step": 29065
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.581133246421814,
+      "learning_rate": 0.000323708479483908,
+      "loss": 2.9457,
+      "step": 29066
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5575944781303406,
+      "learning_rate": 0.0003236931314157808,
+      "loss": 3.0139,
+      "step": 29067
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.589728832244873,
+      "learning_rate": 0.0003236777832852506,
+      "loss": 3.127,
+      "step": 29068
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5565273761749268,
+      "learning_rate": 0.00032366243509235744,
+      "loss": 3.1569,
+      "step": 29069
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5574138164520264,
+      "learning_rate": 0.00032364708683714206,
+      "loss": 3.0355,
+      "step": 29070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5988021492958069,
+      "learning_rate": 0.00032363173851964486,
+      "loss": 2.9168,
+      "step": 29071
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5684751272201538,
+      "learning_rate": 0.0003236163901399063,
+      "loss": 3.1833,
+      "step": 29072
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.619386613368988,
+      "learning_rate": 0.0003236010416979666,
+      "loss": 2.9159,
+      "step": 29073
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5681483149528503,
+      "learning_rate": 0.0003235856931938662,
+      "loss": 2.9759,
+      "step": 29074
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5597151517868042,
+      "learning_rate": 0.0003235703446276459,
+      "loss": 3.2504,
+      "step": 29075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5716343522071838,
+      "learning_rate": 0.0003235549959993456,
+      "loss": 3.0314,
+      "step": 29076
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6016743779182434,
+      "learning_rate": 0.00032353964730900606,
+      "loss": 3.0251,
+      "step": 29077
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6275364756584167,
+      "learning_rate": 0.0003235242985566676,
+      "loss": 2.8835,
+      "step": 29078
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5814191699028015,
+      "learning_rate": 0.00032350894974237065,
+      "loss": 3.0817,
+      "step": 29079
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5964180827140808,
+      "learning_rate": 0.0003234936008661557,
+      "loss": 2.8892,
+      "step": 29080
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5740237236022949,
+      "learning_rate": 0.00032347825192806314,
+      "loss": 2.9775,
+      "step": 29081
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5742438435554504,
+      "learning_rate": 0.00032346290292813325,
+      "loss": 2.7709,
+      "step": 29082
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5780929327011108,
+      "learning_rate": 0.00032344755386640677,
+      "loss": 3.0928,
+      "step": 29083
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5740398168563843,
+      "learning_rate": 0.0003234322047429238,
+      "loss": 3.1639,
+      "step": 29084
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6118278503417969,
+      "learning_rate": 0.00032341685555772496,
+      "loss": 3.0158,
+      "step": 29085
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5554642677307129,
+      "learning_rate": 0.00032340150631085074,
+      "loss": 3.0543,
+      "step": 29086
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5830543041229248,
+      "learning_rate": 0.00032338615700234135,
+      "loss": 3.0372,
+      "step": 29087
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5947121381759644,
+      "learning_rate": 0.00032337080763223736,
+      "loss": 2.7584,
+      "step": 29088
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5922091007232666,
+      "learning_rate": 0.00032335545820057904,
+      "loss": 3.1258,
+      "step": 29089
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5475114583969116,
+      "learning_rate": 0.0003233401087074071,
+      "loss": 2.8411,
+      "step": 29090
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5651057362556458,
+      "learning_rate": 0.0003233247591527617,
+      "loss": 3.0353,
+      "step": 29091
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5793414115905762,
+      "learning_rate": 0.0003233094095366834,
+      "loss": 2.8918,
+      "step": 29092
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5808545351028442,
+      "learning_rate": 0.0003232940598592126,
+      "loss": 3.0612,
+      "step": 29093
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5640668272972107,
+      "learning_rate": 0.00032327871012038977,
+      "loss": 3.0701,
+      "step": 29094
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5635700225830078,
+      "learning_rate": 0.00032326336032025525,
+      "loss": 3.1003,
+      "step": 29095
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5963321328163147,
+      "learning_rate": 0.0003232480104588495,
+      "loss": 2.8877,
+      "step": 29096
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5794896483421326,
+      "learning_rate": 0.0003232326605362131,
+      "loss": 2.9719,
+      "step": 29097
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5805581212043762,
+      "learning_rate": 0.0003232173105523862,
+      "loss": 3.0281,
+      "step": 29098
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.600281834602356,
+      "learning_rate": 0.00032320196050740935,
+      "loss": 3.3348,
+      "step": 29099
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5603992938995361,
+      "learning_rate": 0.00032318661040132307,
+      "loss": 3.0012,
+      "step": 29100
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6025208830833435,
+      "learning_rate": 0.0003231712602341678,
+      "loss": 2.9332,
+      "step": 29101
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5629807710647583,
+      "learning_rate": 0.00032315591000598375,
+      "loss": 3.1528,
+      "step": 29102
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5991537570953369,
+      "learning_rate": 0.0003231405597168116,
+      "loss": 3.1373,
+      "step": 29103
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5493240356445312,
+      "learning_rate": 0.0003231252093666916,
+      "loss": 2.9525,
+      "step": 29104
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5737966299057007,
+      "learning_rate": 0.0003231098589556642,
+      "loss": 2.9983,
+      "step": 29105
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6087252497673035,
+      "learning_rate": 0.00032309450848377,
+      "loss": 3.0698,
+      "step": 29106
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6365086436271667,
+      "learning_rate": 0.00032307915795104925,
+      "loss": 2.9886,
+      "step": 29107
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5757536292076111,
+      "learning_rate": 0.0003230638073575425,
+      "loss": 2.9974,
+      "step": 29108
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5831839442253113,
+      "learning_rate": 0.00032304845670329003,
+      "loss": 3.0675,
+      "step": 29109
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5890743136405945,
+      "learning_rate": 0.0003230331059883323,
+      "loss": 2.9943,
+      "step": 29110
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6159026622772217,
+      "learning_rate": 0.0003230177552127099,
+      "loss": 3.2641,
+      "step": 29111
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6064269542694092,
+      "learning_rate": 0.0003230024043764632,
+      "loss": 3.031,
+      "step": 29112
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6121566891670227,
+      "learning_rate": 0.0003229870534796326,
+      "loss": 3.0392,
+      "step": 29113
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6181187629699707,
+      "learning_rate": 0.0003229717025222585,
+      "loss": 3.1578,
+      "step": 29114
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6169598698616028,
+      "learning_rate": 0.00032295635150438127,
+      "loss": 3.1092,
+      "step": 29115
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5787153840065002,
+      "learning_rate": 0.0003229410004260415,
+      "loss": 3.0131,
+      "step": 29116
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5475699305534363,
+      "learning_rate": 0.00032292564928727954,
+      "loss": 3.3638,
+      "step": 29117
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5485643744468689,
+      "learning_rate": 0.0003229102980881358,
+      "loss": 2.7888,
+      "step": 29118
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5797244310379028,
+      "learning_rate": 0.0003228949468286509,
+      "loss": 3.1413,
+      "step": 29119
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5789276957511902,
+      "learning_rate": 0.00032287959550886486,
+      "loss": 2.7676,
+      "step": 29120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5958558917045593,
+      "learning_rate": 0.0003228642441288185,
+      "loss": 3.1023,
+      "step": 29121
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5825849175453186,
+      "learning_rate": 0.00032284889268855213,
+      "loss": 3.2999,
+      "step": 29122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5534700751304626,
+      "learning_rate": 0.0003228335411881063,
+      "loss": 3.1667,
+      "step": 29123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.581171989440918,
+      "learning_rate": 0.00032281818962752115,
+      "loss": 2.8703,
+      "step": 29124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5967993140220642,
+      "learning_rate": 0.00032280283800683717,
+      "loss": 3.2153,
+      "step": 29125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5790849924087524,
+      "learning_rate": 0.00032278748632609517,
+      "loss": 2.9798,
+      "step": 29126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5897535681724548,
+      "learning_rate": 0.00032277213458533515,
+      "loss": 3.0861,
+      "step": 29127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5886898636817932,
+      "learning_rate": 0.0003227567827845978,
+      "loss": 2.9372,
+      "step": 29128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5656288862228394,
+      "learning_rate": 0.00032274143092392337,
+      "loss": 3.1163,
+      "step": 29129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5811963081359863,
+      "learning_rate": 0.00032272607900335243,
+      "loss": 3.0505,
+      "step": 29130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5721598267555237,
+      "learning_rate": 0.0003227107270229254,
+      "loss": 3.1477,
+      "step": 29131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5497804880142212,
+      "learning_rate": 0.0003226953749826826,
+      "loss": 3.1137,
+      "step": 29132
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6076148748397827,
+      "learning_rate": 0.0003226800228826646,
+      "loss": 3.0502,
+      "step": 29133
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5574772357940674,
+      "learning_rate": 0.0003226646707229119,
+      "loss": 2.9401,
+      "step": 29134
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5898085236549377,
+      "learning_rate": 0.00032264931850346464,
+      "loss": 3.1227,
+      "step": 29135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5748165845870972,
+      "learning_rate": 0.0003226339662243635,
+      "loss": 3.0951,
+      "step": 29136
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6368452310562134,
+      "learning_rate": 0.0003226186138856489,
+      "loss": 3.0673,
+      "step": 29137
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5354744791984558,
+      "learning_rate": 0.0003226032614873612,
+      "loss": 3.025,
+      "step": 29138
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5715710520744324,
+      "learning_rate": 0.0003225879090295408,
+      "loss": 3.2214,
+      "step": 29139
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6316717863082886,
+      "learning_rate": 0.00032257255651222826,
+      "loss": 3.273,
+      "step": 29140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5766654014587402,
+      "learning_rate": 0.0003225572039354639,
+      "loss": 3.1074,
+      "step": 29141
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6002153754234314,
+      "learning_rate": 0.0003225418512992883,
+      "loss": 3.0406,
+      "step": 29142
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6038157343864441,
+      "learning_rate": 0.0003225264986037417,
+      "loss": 3.014,
+      "step": 29143
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5729610323905945,
+      "learning_rate": 0.00032251114584886466,
+      "loss": 2.8022,
+      "step": 29144
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5932610034942627,
+      "learning_rate": 0.00032249579303469765,
+      "loss": 3.0532,
+      "step": 29145
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5910505056381226,
+      "learning_rate": 0.000322480440161281,
+      "loss": 2.8465,
+      "step": 29146
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5964004397392273,
+      "learning_rate": 0.00032246508722865515,
+      "loss": 3.0927,
+      "step": 29147
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5830546617507935,
+      "learning_rate": 0.00032244973423686075,
+      "loss": 3.0487,
+      "step": 29148
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5955746173858643,
+      "learning_rate": 0.0003224343811859379,
+      "loss": 3.1017,
+      "step": 29149
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6120107173919678,
+      "learning_rate": 0.0003224190280759273,
+      "loss": 3.0038,
+      "step": 29150
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6444991827011108,
+      "learning_rate": 0.0003224036749068692,
+      "loss": 3.089,
+      "step": 29151
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5521337389945984,
+      "learning_rate": 0.0003223883216788042,
+      "loss": 3.0886,
+      "step": 29152
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6167043447494507,
+      "learning_rate": 0.00032237296839177266,
+      "loss": 2.9639,
+      "step": 29153
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6153653264045715,
+      "learning_rate": 0.000322357615045815,
+      "loss": 3.2767,
+      "step": 29154
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6091169118881226,
+      "learning_rate": 0.00032234226164097173,
+      "loss": 3.0682,
+      "step": 29155
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.577284574508667,
+      "learning_rate": 0.00032232690817728327,
+      "loss": 3.0631,
+      "step": 29156
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6353074908256531,
+      "learning_rate": 0.00032231155465478996,
+      "loss": 3.0545,
+      "step": 29157
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6882631778717041,
+      "learning_rate": 0.00032229620107353236,
+      "loss": 3.0973,
+      "step": 29158
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6139711141586304,
+      "learning_rate": 0.0003222808474335509,
+      "loss": 2.9405,
+      "step": 29159
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6546551585197449,
+      "learning_rate": 0.00032226549373488584,
+      "loss": 3.1871,
+      "step": 29160
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5913887619972229,
+      "learning_rate": 0.00032225013997757776,
+      "loss": 3.0151,
+      "step": 29161
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5905306935310364,
+      "learning_rate": 0.0003222347861616672,
+      "loss": 2.8942,
+      "step": 29162
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5535931587219238,
+      "learning_rate": 0.0003222194322871945,
+      "loss": 3.1094,
+      "step": 29163
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6039665341377258,
+      "learning_rate": 0.00032220407835420006,
+      "loss": 2.9505,
+      "step": 29164
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5726522207260132,
+      "learning_rate": 0.0003221887243627243,
+      "loss": 2.9469,
+      "step": 29165
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6165740489959717,
+      "learning_rate": 0.00032217337031280774,
+      "loss": 2.9994,
+      "step": 29166
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6031109690666199,
+      "learning_rate": 0.0003221580162044908,
+      "loss": 3.1434,
+      "step": 29167
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6150469779968262,
+      "learning_rate": 0.0003221426620378139,
+      "loss": 2.9249,
+      "step": 29168
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5885416865348816,
+      "learning_rate": 0.00032212730781281746,
+      "loss": 3.2877,
+      "step": 29169
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5557782649993896,
+      "learning_rate": 0.0003221119535295421,
+      "loss": 3.0221,
+      "step": 29170
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6185104846954346,
+      "learning_rate": 0.00032209659918802793,
+      "loss": 3.0152,
+      "step": 29171
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5505481958389282,
+      "learning_rate": 0.00032208124478831564,
+      "loss": 3.0942,
+      "step": 29172
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5830556154251099,
+      "learning_rate": 0.00032206589033044556,
+      "loss": 3.1301,
+      "step": 29173
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5747976303100586,
+      "learning_rate": 0.0003220505358144583,
+      "loss": 3.0054,
+      "step": 29174
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6036010980606079,
+      "learning_rate": 0.000322035181240394,
+      "loss": 2.9831,
+      "step": 29175
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6000482439994812,
+      "learning_rate": 0.0003220198266082933,
+      "loss": 2.9121,
+      "step": 29176
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5834548473358154,
+      "learning_rate": 0.00032200447191819677,
+      "loss": 3.0241,
+      "step": 29177
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5785706639289856,
+      "learning_rate": 0.00032198911717014453,
+      "loss": 3.1835,
+      "step": 29178
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5782999992370605,
+      "learning_rate": 0.0003219737623641773,
+      "loss": 3.0318,
+      "step": 29179
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5533021688461304,
+      "learning_rate": 0.00032195840750033535,
+      "loss": 3.1158,
+      "step": 29180
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5466267466545105,
+      "learning_rate": 0.00032194305257865914,
+      "loss": 3.1926,
+      "step": 29181
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.624919056892395,
+      "learning_rate": 0.0003219276975991892,
+      "loss": 2.9148,
+      "step": 29182
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5868450403213501,
+      "learning_rate": 0.00032191234256196587,
+      "loss": 2.9879,
+      "step": 29183
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5854912400245667,
+      "learning_rate": 0.00032189698746702976,
+      "loss": 2.9871,
+      "step": 29184
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5936477780342102,
+      "learning_rate": 0.0003218816323144212,
+      "loss": 3.0536,
+      "step": 29185
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5781201720237732,
+      "learning_rate": 0.00032186627710418056,
+      "loss": 3.1815,
+      "step": 29186
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5902300477027893,
+      "learning_rate": 0.0003218509218363483,
+      "loss": 3.0949,
+      "step": 29187
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5788503289222717,
+      "learning_rate": 0.00032183556651096506,
+      "loss": 2.8732,
+      "step": 29188
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5969659090042114,
+      "learning_rate": 0.0003218202111280711,
+      "loss": 2.9742,
+      "step": 29189
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.603798508644104,
+      "learning_rate": 0.0003218048556877068,
+      "loss": 3.2926,
+      "step": 29190
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5414276123046875,
+      "learning_rate": 0.00032178950018991277,
+      "loss": 3.0918,
+      "step": 29191
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5900503993034363,
+      "learning_rate": 0.0003217741446347294,
+      "loss": 3.2083,
+      "step": 29192
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6011453866958618,
+      "learning_rate": 0.00032175878902219714,
+      "loss": 3.215,
+      "step": 29193
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6004211902618408,
+      "learning_rate": 0.00032174343335235637,
+      "loss": 3.0323,
+      "step": 29194
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5935797095298767,
+      "learning_rate": 0.0003217280776252476,
+      "loss": 2.909,
+      "step": 29195
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6498414278030396,
+      "learning_rate": 0.0003217127218409113,
+      "loss": 3.0649,
+      "step": 29196
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5944969058036804,
+      "learning_rate": 0.00032169736599938777,
+      "loss": 3.234,
+      "step": 29197
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6110669374465942,
+      "learning_rate": 0.0003216820101007177,
+      "loss": 2.709,
+      "step": 29198
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5599201321601868,
+      "learning_rate": 0.0003216666541449413,
+      "loss": 3.0135,
+      "step": 29199
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5829009413719177,
+      "learning_rate": 0.0003216512981320991,
+      "loss": 2.8439,
+      "step": 29200
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6302081942558289,
+      "learning_rate": 0.0003216359420622315,
+      "loss": 2.9406,
+      "step": 29201
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6054470539093018,
+      "learning_rate": 0.000321620585935379,
+      "loss": 3.1103,
+      "step": 29202
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6377239227294922,
+      "learning_rate": 0.00032160522975158216,
+      "loss": 2.9263,
+      "step": 29203
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5627120733261108,
+      "learning_rate": 0.0003215898735108812,
+      "loss": 2.8536,
+      "step": 29204
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5820385217666626,
+      "learning_rate": 0.00032157451721331675,
+      "loss": 2.9532,
+      "step": 29205
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.593466579914093,
+      "learning_rate": 0.0003215591608589291,
+      "loss": 3.0497,
+      "step": 29206
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5768133401870728,
+      "learning_rate": 0.00032154380444775877,
+      "loss": 3.1535,
+      "step": 29207
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6087096929550171,
+      "learning_rate": 0.0003215284479798462,
+      "loss": 3.0309,
+      "step": 29208
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6114547848701477,
+      "learning_rate": 0.0003215130914552319,
+      "loss": 3.0905,
+      "step": 29209
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6147521734237671,
+      "learning_rate": 0.00032149773487395624,
+      "loss": 2.9946,
+      "step": 29210
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5623575448989868,
+      "learning_rate": 0.0003214823782360597,
+      "loss": 2.9643,
+      "step": 29211
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5678356289863586,
+      "learning_rate": 0.00032146702154158257,
+      "loss": 3.0765,
+      "step": 29212
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5572715401649475,
+      "learning_rate": 0.00032145166479056553,
+      "loss": 2.6901,
+      "step": 29213
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6046434640884399,
+      "learning_rate": 0.00032143630798304903,
+      "loss": 2.9514,
+      "step": 29214
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6713277101516724,
+      "learning_rate": 0.0003214209511190733,
+      "loss": 2.9896,
+      "step": 29215
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6149976253509521,
+      "learning_rate": 0.00032140559419867895,
+      "loss": 3.0438,
+      "step": 29216
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.603386640548706,
+      "learning_rate": 0.0003213902372219064,
+      "loss": 3.2325,
+      "step": 29217
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5698647499084473,
+      "learning_rate": 0.000321374880188796,
+      "loss": 2.9808,
+      "step": 29218
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.588376522064209,
+      "learning_rate": 0.00032135952309938836,
+      "loss": 3.0994,
+      "step": 29219
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5656581521034241,
+      "learning_rate": 0.0003213441659537238,
+      "loss": 3.2486,
+      "step": 29220
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.56963711977005,
+      "learning_rate": 0.0003213288087518429,
+      "loss": 3.1403,
+      "step": 29221
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5951929092407227,
+      "learning_rate": 0.0003213134514937859,
+      "loss": 2.8392,
+      "step": 29222
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5948726534843445,
+      "learning_rate": 0.00032129809417959346,
+      "loss": 3.0278,
+      "step": 29223
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.632287323474884,
+      "learning_rate": 0.0003212827368093059,
+      "loss": 3.0306,
+      "step": 29224
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5774150490760803,
+      "learning_rate": 0.00032126737938296376,
+      "loss": 3.0625,
+      "step": 29225
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.687145471572876,
+      "learning_rate": 0.0003212520219006074,
+      "loss": 3.0098,
+      "step": 29226
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.615565299987793,
+      "learning_rate": 0.00032123666436227735,
+      "loss": 3.0459,
+      "step": 29227
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6112501621246338,
+      "learning_rate": 0.0003212213067680139,
+      "loss": 3.0546,
+      "step": 29228
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5874766707420349,
+      "learning_rate": 0.0003212059491178577,
+      "loss": 2.7821,
+      "step": 29229
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.64829021692276,
+      "learning_rate": 0.0003211905914118491,
+      "loss": 3.0968,
+      "step": 29230
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6121979355812073,
+      "learning_rate": 0.0003211752336500285,
+      "loss": 3.0179,
+      "step": 29231
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6101468801498413,
+      "learning_rate": 0.00032115987583243653,
+      "loss": 2.9941,
+      "step": 29232
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5940099358558655,
+      "learning_rate": 0.0003211445179591134,
+      "loss": 3.0411,
+      "step": 29233
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5766475796699524,
+      "learning_rate": 0.00032112916003009965,
+      "loss": 2.939,
+      "step": 29234
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5749085545539856,
+      "learning_rate": 0.00032111380204543586,
+      "loss": 3.0331,
+      "step": 29235
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5809697508811951,
+      "learning_rate": 0.0003210984440051624,
+      "loss": 3.1916,
+      "step": 29236
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5791935920715332,
+      "learning_rate": 0.00032108308590931965,
+      "loss": 2.8306,
+      "step": 29237
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6096273064613342,
+      "learning_rate": 0.00032106772775794806,
+      "loss": 3.0382,
+      "step": 29238
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5717841386795044,
+      "learning_rate": 0.00032105236955108825,
+      "loss": 2.9114,
+      "step": 29239
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.589238166809082,
+      "learning_rate": 0.0003210370112887805,
+      "loss": 3.0616,
+      "step": 29240
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6342265009880066,
+      "learning_rate": 0.0003210216529710653,
+      "loss": 3.1,
+      "step": 29241
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.609419047832489,
+      "learning_rate": 0.00032100629459798303,
+      "loss": 3.0961,
+      "step": 29242
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5756333470344543,
+      "learning_rate": 0.0003209909361695743,
+      "loss": 2.8789,
+      "step": 29243
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5944945812225342,
+      "learning_rate": 0.0003209755776858795,
+      "loss": 2.9332,
+      "step": 29244
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.612972617149353,
+      "learning_rate": 0.00032096021914693906,
+      "loss": 3.2249,
+      "step": 29245
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6202994585037231,
+      "learning_rate": 0.0003209448605527934,
+      "loss": 2.9364,
+      "step": 29246
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6110982298851013,
+      "learning_rate": 0.0003209295019034831,
+      "loss": 3.1885,
+      "step": 29247
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5498843193054199,
+      "learning_rate": 0.0003209141431990484,
+      "loss": 3.1529,
+      "step": 29248
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5546060800552368,
+      "learning_rate": 0.00032089878443952984,
+      "loss": 3.0771,
+      "step": 29249
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5500269532203674,
+      "learning_rate": 0.0003208834256249681,
+      "loss": 3.0521,
+      "step": 29250
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5937256217002869,
+      "learning_rate": 0.0003208680667554033,
+      "loss": 3.1337,
+      "step": 29251
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5819010734558105,
+      "learning_rate": 0.00032085270783087605,
+      "loss": 3.4204,
+      "step": 29252
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5727535486221313,
+      "learning_rate": 0.00032083734885142673,
+      "loss": 2.879,
+      "step": 29253
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6039817333221436,
+      "learning_rate": 0.0003208219898170959,
+      "loss": 2.9596,
+      "step": 29254
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5760222673416138,
+      "learning_rate": 0.000320806630727924,
+      "loss": 3.2594,
+      "step": 29255
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5537810921669006,
+      "learning_rate": 0.00032079127158395136,
+      "loss": 3.0045,
+      "step": 29256
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5716868042945862,
+      "learning_rate": 0.00032077591238521857,
+      "loss": 2.921,
+      "step": 29257
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5600032210350037,
+      "learning_rate": 0.00032076055313176593,
+      "loss": 3.0156,
+      "step": 29258
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5719262361526489,
+      "learning_rate": 0.00032074519382363406,
+      "loss": 3.1698,
+      "step": 29259
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5441270470619202,
+      "learning_rate": 0.0003207298344608634,
+      "loss": 3.0605,
+      "step": 29260
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5567929744720459,
+      "learning_rate": 0.00032071447504349433,
+      "loss": 3.0992,
+      "step": 29261
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5503496527671814,
+      "learning_rate": 0.0003206991155715672,
+      "loss": 2.9937,
+      "step": 29262
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.576248049736023,
+      "learning_rate": 0.0003206837560451227,
+      "loss": 2.9549,
+      "step": 29263
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5835015773773193,
+      "learning_rate": 0.000320668396464201,
+      "loss": 3.1543,
+      "step": 29264
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5682410597801208,
+      "learning_rate": 0.00032065303682884293,
+      "loss": 3.2322,
+      "step": 29265
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5933088660240173,
+      "learning_rate": 0.0003206376771390887,
+      "loss": 3.2144,
+      "step": 29266
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5696515440940857,
+      "learning_rate": 0.0003206223173949787,
+      "loss": 2.8856,
+      "step": 29267
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6060404181480408,
+      "learning_rate": 0.0003206069575965535,
+      "loss": 3.1012,
+      "step": 29268
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5765473246574402,
+      "learning_rate": 0.0003205915977438536,
+      "loss": 3.0763,
+      "step": 29269
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5708712339401245,
+      "learning_rate": 0.00032057623783691936,
+      "loss": 3.1127,
+      "step": 29270
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5930168032646179,
+      "learning_rate": 0.0003205608778757912,
+      "loss": 2.9539,
+      "step": 29271
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5647690892219543,
+      "learning_rate": 0.0003205455178605099,
+      "loss": 2.9147,
+      "step": 29272
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6091307401657104,
+      "learning_rate": 0.00032053015779111543,
+      "loss": 3.0897,
+      "step": 29273
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5425414443016052,
+      "learning_rate": 0.00032051479766764845,
+      "loss": 2.9467,
+      "step": 29274
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.589598536491394,
+      "learning_rate": 0.00032049943749014954,
+      "loss": 3.1705,
+      "step": 29275
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6278284192085266,
+      "learning_rate": 0.00032048407725865905,
+      "loss": 2.9084,
+      "step": 29276
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5885589718818665,
+      "learning_rate": 0.0003204687169732174,
+      "loss": 3.0185,
+      "step": 29277
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.589991569519043,
+      "learning_rate": 0.0003204533566338652,
+      "loss": 3.0392,
+      "step": 29278
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5731289982795715,
+      "learning_rate": 0.0003204379962406427,
+      "loss": 2.9205,
+      "step": 29279
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5898691415786743,
+      "learning_rate": 0.00032042263579359047,
+      "loss": 3.116,
+      "step": 29280
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5861981511116028,
+      "learning_rate": 0.0003204072752927489,
+      "loss": 3.2275,
+      "step": 29281
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5974080562591553,
+      "learning_rate": 0.00032039191473815854,
+      "loss": 2.8365,
+      "step": 29282
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.569999635219574,
+      "learning_rate": 0.0003203765541298598,
+      "loss": 2.9994,
+      "step": 29283
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5541989207267761,
+      "learning_rate": 0.0003203611934678931,
+      "loss": 2.6546,
+      "step": 29284
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5642061233520508,
+      "learning_rate": 0.00032034583275229896,
+      "loss": 3.0176,
+      "step": 29285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5687254667282104,
+      "learning_rate": 0.00032033047198311786,
+      "loss": 3.0303,
+      "step": 29286
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5605220198631287,
+      "learning_rate": 0.0003203151111603902,
+      "loss": 3.0205,
+      "step": 29287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5649875998497009,
+      "learning_rate": 0.0003202997502841564,
+      "loss": 2.9837,
+      "step": 29288
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5587196350097656,
+      "learning_rate": 0.00032028438935445693,
+      "loss": 3.0814,
+      "step": 29289
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5571485161781311,
+      "learning_rate": 0.0003202690283713324,
+      "loss": 3.0345,
+      "step": 29290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6157510876655579,
+      "learning_rate": 0.0003202536673348231,
+      "loss": 3.1435,
+      "step": 29291
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5682488083839417,
+      "learning_rate": 0.0003202383062449695,
+      "loss": 3.1119,
+      "step": 29292
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5798583626747131,
+      "learning_rate": 0.0003202229451018121,
+      "loss": 2.833,
+      "step": 29293
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5703584551811218,
+      "learning_rate": 0.00032020758390539136,
+      "loss": 3.1149,
+      "step": 29294
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5614786148071289,
+      "learning_rate": 0.00032019222265574775,
+      "loss": 3.0893,
+      "step": 29295
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.541699230670929,
+      "learning_rate": 0.0003201768613529217,
+      "loss": 3.0179,
+      "step": 29296
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.615832507610321,
+      "learning_rate": 0.00032016149999695374,
+      "loss": 3.193,
+      "step": 29297
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5494009852409363,
+      "learning_rate": 0.0003201461385878842,
+      "loss": 3.1504,
+      "step": 29298
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5631309747695923,
+      "learning_rate": 0.00032013077712575357,
+      "loss": 3.1214,
+      "step": 29299
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5645419359207153,
+      "learning_rate": 0.00032011541561060246,
+      "loss": 3.1237,
+      "step": 29300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.577457070350647,
+      "learning_rate": 0.0003201000540424712,
+      "loss": 2.8725,
+      "step": 29301
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6075976490974426,
+      "learning_rate": 0.0003200846924214002,
+      "loss": 3.0414,
+      "step": 29302
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5680819749832153,
+      "learning_rate": 0.00032006933074743,
+      "loss": 3.1522,
+      "step": 29303
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.573302149772644,
+      "learning_rate": 0.00032005396902060106,
+      "loss": 3.2003,
+      "step": 29304
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.572464108467102,
+      "learning_rate": 0.0003200386072409538,
+      "loss": 2.9017,
+      "step": 29305
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.562468945980072,
+      "learning_rate": 0.00032002324540852874,
+      "loss": 3.0256,
+      "step": 29306
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.615037202835083,
+      "learning_rate": 0.00032000788352336626,
+      "loss": 2.9762,
+      "step": 29307
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5618579387664795,
+      "learning_rate": 0.0003199925215855069,
+      "loss": 3.0673,
+      "step": 29308
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5961542725563049,
+      "learning_rate": 0.0003199771595949912,
+      "loss": 3.2632,
+      "step": 29309
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6055513620376587,
+      "learning_rate": 0.00031996179755185925,
+      "loss": 2.8112,
+      "step": 29310
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5619701147079468,
+      "learning_rate": 0.000319946435456152,
+      "loss": 2.9445,
+      "step": 29311
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5580838322639465,
+      "learning_rate": 0.00031993107330790957,
+      "loss": 2.8617,
+      "step": 29312
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5237293839454651,
+      "learning_rate": 0.00031991571110717254,
+      "loss": 2.8157,
+      "step": 29313
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.545191764831543,
+      "learning_rate": 0.00031990034885398137,
+      "loss": 3.2033,
+      "step": 29314
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5724726319313049,
+      "learning_rate": 0.00031988498654837645,
+      "loss": 3.0922,
+      "step": 29315
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6132506728172302,
+      "learning_rate": 0.0003198696241903984,
+      "loss": 2.8161,
+      "step": 29316
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5981416702270508,
+      "learning_rate": 0.0003198542617800876,
+      "loss": 3.0212,
+      "step": 29317
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6016727685928345,
+      "learning_rate": 0.0003198388993174844,
+      "loss": 3.171,
+      "step": 29318
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5826240181922913,
+      "learning_rate": 0.00031982353680262937,
+      "loss": 2.9993,
+      "step": 29319
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5482321381568909,
+      "learning_rate": 0.000319808174235563,
+      "loss": 3.193,
+      "step": 29320
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5854382514953613,
+      "learning_rate": 0.0003197928116163257,
+      "loss": 3.4126,
+      "step": 29321
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5981870293617249,
+      "learning_rate": 0.0003197774489449579,
+      "loss": 3.1969,
+      "step": 29322
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6282458305358887,
+      "learning_rate": 0.00031976208622150023,
+      "loss": 2.8896,
+      "step": 29323
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5802420377731323,
+      "learning_rate": 0.0003197467234459929,
+      "loss": 3.0195,
+      "step": 29324
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6020224690437317,
+      "learning_rate": 0.0003197313606184765,
+      "loss": 3.0872,
+      "step": 29325
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.542820394039154,
+      "learning_rate": 0.0003197159977389915,
+      "loss": 3.0749,
+      "step": 29326
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5725154280662537,
+      "learning_rate": 0.00031970063480757843,
+      "loss": 3.0649,
+      "step": 29327
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.602067232131958,
+      "learning_rate": 0.00031968527182427766,
+      "loss": 2.877,
+      "step": 29328
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6186937689781189,
+      "learning_rate": 0.0003196699087891297,
+      "loss": 3.1819,
+      "step": 29329
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6243546009063721,
+      "learning_rate": 0.00031965454570217485,
+      "loss": 3.1825,
+      "step": 29330
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.586065948009491,
+      "learning_rate": 0.0003196391825634538,
+      "loss": 2.9946,
+      "step": 29331
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6126989722251892,
+      "learning_rate": 0.00031962381937300694,
+      "loss": 3.0819,
+      "step": 29332
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5854355692863464,
+      "learning_rate": 0.00031960845613087464,
+      "loss": 3.2089,
+      "step": 29333
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5818975567817688,
+      "learning_rate": 0.0003195930928370976,
+      "loss": 3.1475,
+      "step": 29334
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5570355653762817,
+      "learning_rate": 0.000319577729491716,
+      "loss": 3.0968,
+      "step": 29335
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5599480867385864,
+      "learning_rate": 0.0003195623660947704,
+      "loss": 2.9759,
+      "step": 29336
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5848993062973022,
+      "learning_rate": 0.00031954700264630136,
+      "loss": 3.2071,
+      "step": 29337
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5863507390022278,
+      "learning_rate": 0.00031953163914634936,
+      "loss": 3.3054,
+      "step": 29338
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5722285509109497,
+      "learning_rate": 0.0003195162755949547,
+      "loss": 3.1209,
+      "step": 29339
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5752810835838318,
+      "learning_rate": 0.00031950091199215775,
+      "loss": 3.1011,
+      "step": 29340
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6314206123352051,
+      "learning_rate": 0.0003194855483379994,
+      "loss": 3.1192,
+      "step": 29341
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5731692910194397,
+      "learning_rate": 0.00031947018463251975,
+      "loss": 2.9817,
+      "step": 29342
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5550987124443054,
+      "learning_rate": 0.0003194548208757594,
+      "loss": 3.079,
+      "step": 29343
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5599222183227539,
+      "learning_rate": 0.0003194394570677588,
+      "loss": 3.0314,
+      "step": 29344
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5602944493293762,
+      "learning_rate": 0.00031942409320855845,
+      "loss": 3.015,
+      "step": 29345
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5581718683242798,
+      "learning_rate": 0.00031940872929819874,
+      "loss": 3.0226,
+      "step": 29346
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6010013818740845,
+      "learning_rate": 0.00031939336533672014,
+      "loss": 3.1017,
+      "step": 29347
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5561316013336182,
+      "learning_rate": 0.00031937800132416315,
+      "loss": 3.0053,
+      "step": 29348
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.620624840259552,
+      "learning_rate": 0.0003193626372605683,
+      "loss": 3.0979,
+      "step": 29349
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5798450708389282,
+      "learning_rate": 0.000319347273145976,
+      "loss": 3.1279,
+      "step": 29350
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6187215447425842,
+      "learning_rate": 0.0003193319089804266,
+      "loss": 3.1367,
+      "step": 29351
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6134929656982422,
+      "learning_rate": 0.00031931654476396077,
+      "loss": 3.0664,
+      "step": 29352
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5657615661621094,
+      "learning_rate": 0.0003193011804966188,
+      "loss": 3.0826,
+      "step": 29353
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5925348401069641,
+      "learning_rate": 0.00031928581617844134,
+      "loss": 2.892,
+      "step": 29354
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5548837780952454,
+      "learning_rate": 0.0003192704518094686,
+      "loss": 2.9854,
+      "step": 29355
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5412815809249878,
+      "learning_rate": 0.00031925508738974134,
+      "loss": 2.9617,
+      "step": 29356
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5807236433029175,
+      "learning_rate": 0.0003192397229192998,
+      "loss": 3.1411,
+      "step": 29357
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5956936478614807,
+      "learning_rate": 0.00031922435839818456,
+      "loss": 3.1444,
+      "step": 29358
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5684614181518555,
+      "learning_rate": 0.00031920899382643605,
+      "loss": 2.9935,
+      "step": 29359
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6149534583091736,
+      "learning_rate": 0.0003191936292040948,
+      "loss": 2.8515,
+      "step": 29360
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5479702353477478,
+      "learning_rate": 0.00031917826453120113,
+      "loss": 3.1102,
+      "step": 29361
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5614345073699951,
+      "learning_rate": 0.00031916289980779566,
+      "loss": 3.0526,
+      "step": 29362
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6308231353759766,
+      "learning_rate": 0.0003191475350339189,
+      "loss": 3.0099,
+      "step": 29363
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5764047503471375,
+      "learning_rate": 0.0003191321702096111,
+      "loss": 3.1026,
+      "step": 29364
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5685030221939087,
+      "learning_rate": 0.00031911680533491284,
+      "loss": 2.9923,
+      "step": 29365
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5890762805938721,
+      "learning_rate": 0.0003191014404098645,
+      "loss": 2.9788,
+      "step": 29366
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5421032309532166,
+      "learning_rate": 0.00031908607543450684,
+      "loss": 3.1369,
+      "step": 29367
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6088888049125671,
+      "learning_rate": 0.00031907071040888007,
+      "loss": 3.1285,
+      "step": 29368
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.585292637348175,
+      "learning_rate": 0.00031905534533302465,
+      "loss": 3.0663,
+      "step": 29369
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5948856472969055,
+      "learning_rate": 0.0003190399802069812,
+      "loss": 3.0343,
+      "step": 29370
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6100438833236694,
+      "learning_rate": 0.0003190246150307901,
+      "loss": 2.9583,
+      "step": 29371
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5696083307266235,
+      "learning_rate": 0.00031900924980449174,
+      "loss": 3.1466,
+      "step": 29372
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5895338654518127,
+      "learning_rate": 0.0003189938845281267,
+      "loss": 3.001,
+      "step": 29373
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5648195147514343,
+      "learning_rate": 0.0003189785192017355,
+      "loss": 3.0352,
+      "step": 29374
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5830442905426025,
+      "learning_rate": 0.0003189631538253585,
+      "loss": 3.0942,
+      "step": 29375
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5957970023155212,
+      "learning_rate": 0.00031894778839903614,
+      "loss": 2.8905,
+      "step": 29376
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6936584115028381,
+      "learning_rate": 0.000318932422922809,
+      "loss": 3.0502,
+      "step": 29377
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6041005849838257,
+      "learning_rate": 0.0003189170573967176,
+      "loss": 3.1924,
+      "step": 29378
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5614645481109619,
+      "learning_rate": 0.00031890169182080216,
+      "loss": 3.0978,
+      "step": 29379
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7858589291572571,
+      "learning_rate": 0.0003188863261951033,
+      "loss": 3.061,
+      "step": 29380
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6027712225914001,
+      "learning_rate": 0.00031887096051966154,
+      "loss": 3.1522,
+      "step": 29381
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6058451533317566,
+      "learning_rate": 0.00031885559479451735,
+      "loss": 3.0498,
+      "step": 29382
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6025016903877258,
+      "learning_rate": 0.0003188402290197111,
+      "loss": 3.1248,
+      "step": 29383
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6525845527648926,
+      "learning_rate": 0.0003188248631952833,
+      "loss": 2.9143,
+      "step": 29384
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5875680446624756,
+      "learning_rate": 0.0003188094973212745,
+      "loss": 2.9911,
+      "step": 29385
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5825726985931396,
+      "learning_rate": 0.00031879413139772507,
+      "loss": 3.0715,
+      "step": 29386
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5770814418792725,
+      "learning_rate": 0.00031877876542467544,
+      "loss": 2.9507,
+      "step": 29387
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5879018306732178,
+      "learning_rate": 0.00031876339940216624,
+      "loss": 2.9064,
+      "step": 29388
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5837126970291138,
+      "learning_rate": 0.0003187480333302379,
+      "loss": 2.9365,
+      "step": 29389
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5690444111824036,
+      "learning_rate": 0.00031873266720893075,
+      "loss": 3.1901,
+      "step": 29390
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5976502299308777,
+      "learning_rate": 0.00031871730103828544,
+      "loss": 3.0219,
+      "step": 29391
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6792353987693787,
+      "learning_rate": 0.0003187019348183423,
+      "loss": 2.9063,
+      "step": 29392
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5715599656105042,
+      "learning_rate": 0.00031868656854914184,
+      "loss": 3.1014,
+      "step": 29393
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5927602648735046,
+      "learning_rate": 0.0003186712022307246,
+      "loss": 3.0256,
+      "step": 29394
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.583453893661499,
+      "learning_rate": 0.00031865583586313103,
+      "loss": 3.0197,
+      "step": 29395
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5917752385139465,
+      "learning_rate": 0.0003186404694464016,
+      "loss": 3.064,
+      "step": 29396
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5617307424545288,
+      "learning_rate": 0.0003186251029805767,
+      "loss": 3.1082,
+      "step": 29397
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5545293688774109,
+      "learning_rate": 0.00031860973646569683,
+      "loss": 2.9116,
+      "step": 29398
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5984163880348206,
+      "learning_rate": 0.00031859436990180255,
+      "loss": 3.0933,
+      "step": 29399
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6275175213813782,
+      "learning_rate": 0.0003185790032889343,
+      "loss": 3.1102,
+      "step": 29400
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5929582715034485,
+      "learning_rate": 0.0003185636366271325,
+      "loss": 3.0094,
+      "step": 29401
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5918399095535278,
+      "learning_rate": 0.0003185482699164376,
+      "loss": 2.9954,
+      "step": 29402
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6211867928504944,
+      "learning_rate": 0.0003185329031568902,
+      "loss": 3.0263,
+      "step": 29403
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.615907609462738,
+      "learning_rate": 0.0003185175363485307,
+      "loss": 3.3152,
+      "step": 29404
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.594133734703064,
+      "learning_rate": 0.0003185021694913995,
+      "loss": 3.0418,
+      "step": 29405
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5710883140563965,
+      "learning_rate": 0.0003184868025855373,
+      "loss": 3.1399,
+      "step": 29406
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.544840395450592,
+      "learning_rate": 0.00031847143563098427,
+      "loss": 2.8117,
+      "step": 29407
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6463403701782227,
+      "learning_rate": 0.0003184560686277811,
+      "loss": 3.0027,
+      "step": 29408
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6022986173629761,
+      "learning_rate": 0.00031844070157596813,
+      "loss": 2.9204,
+      "step": 29409
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.568325400352478,
+      "learning_rate": 0.00031842533447558596,
+      "loss": 3.0302,
+      "step": 29410
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.573235809803009,
+      "learning_rate": 0.00031840996732667505,
+      "loss": 3.1247,
+      "step": 29411
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5684030055999756,
+      "learning_rate": 0.0003183946001292757,
+      "loss": 3.1886,
+      "step": 29412
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5798808336257935,
+      "learning_rate": 0.0003183792328834286,
+      "loss": 3.2167,
+      "step": 29413
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5618143677711487,
+      "learning_rate": 0.0003183638655891742,
+      "loss": 3.0885,
+      "step": 29414
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5687463879585266,
+      "learning_rate": 0.00031834849824655283,
+      "loss": 2.8543,
+      "step": 29415
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5497147440910339,
+      "learning_rate": 0.00031833313085560507,
+      "loss": 3.104,
+      "step": 29416
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5957204103469849,
+      "learning_rate": 0.0003183177634163713,
+      "loss": 3.0252,
+      "step": 29417
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6223195195198059,
+      "learning_rate": 0.0003183023959288922,
+      "loss": 2.9171,
+      "step": 29418
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5928622484207153,
+      "learning_rate": 0.000318287028393208,
+      "loss": 3.1098,
+      "step": 29419
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.567246675491333,
+      "learning_rate": 0.0003182716608093593,
+      "loss": 3.0079,
+      "step": 29420
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6125072836875916,
+      "learning_rate": 0.00031825629317738655,
+      "loss": 2.9814,
+      "step": 29421
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5972563624382019,
+      "learning_rate": 0.0003182409254973303,
+      "loss": 3.1398,
+      "step": 29422
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5955824255943298,
+      "learning_rate": 0.0003182255577692309,
+      "loss": 2.8948,
+      "step": 29423
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5571625828742981,
+      "learning_rate": 0.00031821018999312895,
+      "loss": 3.1142,
+      "step": 29424
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6010100841522217,
+      "learning_rate": 0.00031819482216906487,
+      "loss": 3.1236,
+      "step": 29425
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5668064951896667,
+      "learning_rate": 0.0003181794542970791,
+      "loss": 2.9231,
+      "step": 29426
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5944365859031677,
+      "learning_rate": 0.0003181640863772121,
+      "loss": 3.2458,
+      "step": 29427
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5827503800392151,
+      "learning_rate": 0.0003181487184095044,
+      "loss": 3.2371,
+      "step": 29428
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5670848488807678,
+      "learning_rate": 0.00031813335039399654,
+      "loss": 3.0366,
+      "step": 29429
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5577699542045593,
+      "learning_rate": 0.0003181179823307289,
+      "loss": 3.1032,
+      "step": 29430
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5776100754737854,
+      "learning_rate": 0.000318102614219742,
+      "loss": 3.0566,
+      "step": 29431
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6043997406959534,
+      "learning_rate": 0.0003180872460610762,
+      "loss": 2.892,
+      "step": 29432
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.584821343421936,
+      "learning_rate": 0.0003180718778547722,
+      "loss": 3.0156,
+      "step": 29433
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.591511070728302,
+      "learning_rate": 0.0003180565096008703,
+      "loss": 3.0376,
+      "step": 29434
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5555319786071777,
+      "learning_rate": 0.00031804114129941095,
+      "loss": 3.0716,
+      "step": 29435
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6005762815475464,
+      "learning_rate": 0.0003180257729504349,
+      "loss": 3.0269,
+      "step": 29436
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5544703602790833,
+      "learning_rate": 0.0003180104045539823,
+      "loss": 3.006,
+      "step": 29437
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5543263554573059,
+      "learning_rate": 0.00031799503611009366,
+      "loss": 2.9448,
+      "step": 29438
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5978908538818359,
+      "learning_rate": 0.0003179796676188097,
+      "loss": 3.1896,
+      "step": 29439
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5939019322395325,
+      "learning_rate": 0.0003179642990801708,
+      "loss": 3.0236,
+      "step": 29440
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5759717226028442,
+      "learning_rate": 0.00031794893049421724,
+      "loss": 3.0154,
+      "step": 29441
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5859683156013489,
+      "learning_rate": 0.00031793356186098976,
+      "loss": 3.1299,
+      "step": 29442
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6028755307197571,
+      "learning_rate": 0.0003179181931805287,
+      "loss": 3.0498,
+      "step": 29443
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5671459436416626,
+      "learning_rate": 0.0003179028244528746,
+      "loss": 3.1373,
+      "step": 29444
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5760989785194397,
+      "learning_rate": 0.00031788745567806786,
+      "loss": 3.0959,
+      "step": 29445
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5798001885414124,
+      "learning_rate": 0.00031787208685614903,
+      "loss": 3.223,
+      "step": 29446
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5691527724266052,
+      "learning_rate": 0.0003178567179871586,
+      "loss": 2.9604,
+      "step": 29447
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5820776224136353,
+      "learning_rate": 0.00031784134907113687,
+      "loss": 2.9364,
+      "step": 29448
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.576815664768219,
+      "learning_rate": 0.0003178259801081246,
+      "loss": 2.951,
+      "step": 29449
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5452558994293213,
+      "learning_rate": 0.000317810611098162,
+      "loss": 3.0218,
+      "step": 29450
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6160311102867126,
+      "learning_rate": 0.00031779524204128985,
+      "loss": 3.0444,
+      "step": 29451
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6155081391334534,
+      "learning_rate": 0.0003177798729375484,
+      "loss": 3.1733,
+      "step": 29452
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5370228886604309,
+      "learning_rate": 0.00031776450378697807,
+      "loss": 2.9489,
+      "step": 29453
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5774576663970947,
+      "learning_rate": 0.0003177491345896196,
+      "loss": 3.1492,
+      "step": 29454
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.580348789691925,
+      "learning_rate": 0.0003177337653455132,
+      "loss": 3.0479,
+      "step": 29455
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.554613471031189,
+      "learning_rate": 0.0003177183960546996,
+      "loss": 2.9147,
+      "step": 29456
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5800740122795105,
+      "learning_rate": 0.00031770302671721906,
+      "loss": 2.9172,
+      "step": 29457
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5789035558700562,
+      "learning_rate": 0.0003176876573331123,
+      "loss": 2.9797,
+      "step": 29458
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6241493821144104,
+      "learning_rate": 0.0003176722879024194,
+      "loss": 2.9661,
+      "step": 29459
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6004448533058167,
+      "learning_rate": 0.00031765691842518127,
+      "loss": 3.0091,
+      "step": 29460
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6146655678749084,
+      "learning_rate": 0.00031764154890143815,
+      "loss": 2.7731,
+      "step": 29461
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5921909213066101,
+      "learning_rate": 0.0003176261793312307,
+      "loss": 3.0267,
+      "step": 29462
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6653693914413452,
+      "learning_rate": 0.0003176108097145992,
+      "loss": 2.9743,
+      "step": 29463
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.575783371925354,
+      "learning_rate": 0.00031759544005158415,
+      "loss": 3.0894,
+      "step": 29464
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5989192128181458,
+      "learning_rate": 0.0003175800703422262,
+      "loss": 2.8907,
+      "step": 29465
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5665627121925354,
+      "learning_rate": 0.0003175647005865657,
+      "loss": 3.0642,
+      "step": 29466
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.595417320728302,
+      "learning_rate": 0.0003175493307846432,
+      "loss": 3.1442,
+      "step": 29467
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5861067771911621,
+      "learning_rate": 0.00031753396093649904,
+      "loss": 2.803,
+      "step": 29468
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5747168064117432,
+      "learning_rate": 0.0003175185910421739,
+      "loss": 2.9859,
+      "step": 29469
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5734466314315796,
+      "learning_rate": 0.0003175032211017081,
+      "loss": 3.217,
+      "step": 29470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5525097846984863,
+      "learning_rate": 0.0003174878511151422,
+      "loss": 3.0217,
+      "step": 29471
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5869901776313782,
+      "learning_rate": 0.0003174724810825166,
+      "loss": 2.9827,
+      "step": 29472
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5891143679618835,
+      "learning_rate": 0.00031745711100387205,
+      "loss": 3.0641,
+      "step": 29473
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5963634252548218,
+      "learning_rate": 0.0003174417408792486,
+      "loss": 2.9668,
+      "step": 29474
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5769605040550232,
+      "learning_rate": 0.00031742637070868705,
+      "loss": 3.2163,
+      "step": 29475
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5539005398750305,
+      "learning_rate": 0.00031741100049222787,
+      "loss": 3.0295,
+      "step": 29476
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5648486018180847,
+      "learning_rate": 0.00031739563022991144,
+      "loss": 3.2033,
+      "step": 29477
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6375812292098999,
+      "learning_rate": 0.00031738025992177814,
+      "loss": 3.0786,
+      "step": 29478
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5986570119857788,
+      "learning_rate": 0.00031736488956786863,
+      "loss": 3.2335,
+      "step": 29479
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5965058207511902,
+      "learning_rate": 0.00031734951916822347,
+      "loss": 3.0896,
+      "step": 29480
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5741937756538391,
+      "learning_rate": 0.0003173341487228829,
+      "loss": 3.1851,
+      "step": 29481
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5639917254447937,
+      "learning_rate": 0.00031731877823188754,
+      "loss": 3.0339,
+      "step": 29482
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5802062153816223,
+      "learning_rate": 0.00031730340769527783,
+      "loss": 3.1775,
+      "step": 29483
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5636395812034607,
+      "learning_rate": 0.0003172880371130943,
+      "loss": 3.0593,
+      "step": 29484
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.57658851146698,
+      "learning_rate": 0.00031727266648537737,
+      "loss": 3.1828,
+      "step": 29485
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5845613479614258,
+      "learning_rate": 0.0003172572958121676,
+      "loss": 2.8095,
+      "step": 29486
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5840170979499817,
+      "learning_rate": 0.0003172419250935055,
+      "loss": 2.8715,
+      "step": 29487
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5798373222351074,
+      "learning_rate": 0.0003172265543294314,
+      "loss": 2.9696,
+      "step": 29488
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5842916965484619,
+      "learning_rate": 0.0003172111835199858,
+      "loss": 3.2531,
+      "step": 29489
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.57811439037323,
+      "learning_rate": 0.00031719581266520937,
+      "loss": 3.1352,
+      "step": 29490
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6172042489051819,
+      "learning_rate": 0.0003171804417651425,
+      "loss": 3.3323,
+      "step": 29491
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6265311241149902,
+      "learning_rate": 0.0003171650708198256,
+      "loss": 2.9154,
+      "step": 29492
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5730522274971008,
+      "learning_rate": 0.0003171496998292992,
+      "loss": 3.0178,
+      "step": 29493
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5934553146362305,
+      "learning_rate": 0.0003171343287936038,
+      "loss": 3.076,
+      "step": 29494
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6215176582336426,
+      "learning_rate": 0.00031711895771277987,
+      "loss": 3.1317,
+      "step": 29495
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5459911227226257,
+      "learning_rate": 0.0003171035865868679,
+      "loss": 3.2385,
+      "step": 29496
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5924990177154541,
+      "learning_rate": 0.0003170882154159084,
+      "loss": 2.8411,
+      "step": 29497
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5878446102142334,
+      "learning_rate": 0.00031707284419994183,
+      "loss": 2.9754,
+      "step": 29498
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6080270409584045,
+      "learning_rate": 0.0003170574729390087,
+      "loss": 3.1571,
+      "step": 29499
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5498298406600952,
+      "learning_rate": 0.00031704210163314934,
+      "loss": 3.1642,
+      "step": 29500
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5822784900665283,
+      "learning_rate": 0.0003170267302824044,
+      "loss": 3.0001,
+      "step": 29501
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5680847764015198,
+      "learning_rate": 0.0003170113588868145,
+      "loss": 2.9522,
+      "step": 29502
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.589328944683075,
+      "learning_rate": 0.0003169959874464198,
+      "loss": 3.1955,
+      "step": 29503
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6632125973701477,
+      "learning_rate": 0.00031698061596126096,
+      "loss": 3.0567,
+      "step": 29504
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5721989274024963,
+      "learning_rate": 0.00031696524443137846,
+      "loss": 2.9241,
+      "step": 29505
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6211953163146973,
+      "learning_rate": 0.0003169498728568128,
+      "loss": 3.0061,
+      "step": 29506
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5938788056373596,
+      "learning_rate": 0.00031693450123760437,
+      "loss": 3.0157,
+      "step": 29507
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5666404366493225,
+      "learning_rate": 0.0003169191295737938,
+      "loss": 3.0017,
+      "step": 29508
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5512176752090454,
+      "learning_rate": 0.00031690375786542146,
+      "loss": 3.1269,
+      "step": 29509
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5751276612281799,
+      "learning_rate": 0.00031688838611252783,
+      "loss": 2.9614,
+      "step": 29510
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6172812581062317,
+      "learning_rate": 0.00031687301431515346,
+      "loss": 2.8939,
+      "step": 29511
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5911383628845215,
+      "learning_rate": 0.0003168576424733389,
+      "loss": 3.2453,
+      "step": 29512
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5967884659767151,
+      "learning_rate": 0.0003168422705871245,
+      "loss": 3.016,
+      "step": 29513
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5799013376235962,
+      "learning_rate": 0.0003168268986565508,
+      "loss": 3.1217,
+      "step": 29514
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6159305572509766,
+      "learning_rate": 0.0003168115266816582,
+      "loss": 2.9896,
+      "step": 29515
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5831592679023743,
+      "learning_rate": 0.00031679615466248746,
+      "loss": 3.1185,
+      "step": 29516
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5765199065208435,
+      "learning_rate": 0.0003167807825990787,
+      "loss": 3.116,
+      "step": 29517
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6035337448120117,
+      "learning_rate": 0.00031676541049147273,
+      "loss": 3.001,
+      "step": 29518
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6357006430625916,
+      "learning_rate": 0.0003167500383397098,
+      "loss": 2.9424,
+      "step": 29519
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.595038115978241,
+      "learning_rate": 0.0003167346661438305,
+      "loss": 3.0421,
+      "step": 29520
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6188342571258545,
+      "learning_rate": 0.0003167192939038754,
+      "loss": 3.2593,
+      "step": 29521
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6148853302001953,
+      "learning_rate": 0.00031670392161988484,
+      "loss": 2.9146,
+      "step": 29522
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6067038774490356,
+      "learning_rate": 0.00031668854929189927,
+      "loss": 2.9484,
+      "step": 29523
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5747334361076355,
+      "learning_rate": 0.00031667317691995943,
+      "loss": 3.1063,
+      "step": 29524
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6633566617965698,
+      "learning_rate": 0.0003166578045041055,
+      "loss": 3.0578,
+      "step": 29525
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6091784238815308,
+      "learning_rate": 0.0003166424320443782,
+      "loss": 2.8992,
+      "step": 29526
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5521370768547058,
+      "learning_rate": 0.00031662705954081807,
+      "loss": 3.3263,
+      "step": 29527
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5992710590362549,
+      "learning_rate": 0.0003166116869934653,
+      "loss": 2.9598,
+      "step": 29528
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5992154479026794,
+      "learning_rate": 0.0003165963144023606,
+      "loss": 3.0288,
+      "step": 29529
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.563355028629303,
+      "learning_rate": 0.0003165809417675443,
+      "loss": 3.1072,
+      "step": 29530
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.575053334236145,
+      "learning_rate": 0.0003165655690890571,
+      "loss": 3.067,
+      "step": 29531
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5506818890571594,
+      "learning_rate": 0.00031655019636693933,
+      "loss": 3.2032,
+      "step": 29532
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5816770195960999,
+      "learning_rate": 0.0003165348236012316,
+      "loss": 3.2077,
+      "step": 29533
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5859194993972778,
+      "learning_rate": 0.00031651945079197425,
+      "loss": 3.084,
+      "step": 29534
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6002169847488403,
+      "learning_rate": 0.0003165040779392079,
+      "loss": 3.0688,
+      "step": 29535
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5905258655548096,
+      "learning_rate": 0.00031648870504297287,
+      "loss": 2.9794,
+      "step": 29536
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5949984788894653,
+      "learning_rate": 0.0003164733321033099,
+      "loss": 2.9201,
+      "step": 29537
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6448434591293335,
+      "learning_rate": 0.00031645795912025937,
+      "loss": 2.964,
+      "step": 29538
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5883697867393494,
+      "learning_rate": 0.0003164425860938616,
+      "loss": 3.0997,
+      "step": 29539
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5910248160362244,
+      "learning_rate": 0.0003164272130241573,
+      "loss": 3.3129,
+      "step": 29540
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5907135009765625,
+      "learning_rate": 0.0003164118399111868,
+      "loss": 2.9919,
+      "step": 29541
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5622631907463074,
+      "learning_rate": 0.0003163964667549908,
+      "loss": 2.835,
+      "step": 29542
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.609990656375885,
+      "learning_rate": 0.00031638109355560957,
+      "loss": 3.2596,
+      "step": 29543
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5881011486053467,
+      "learning_rate": 0.00031636572031308374,
+      "loss": 3.0279,
+      "step": 29544
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6357592940330505,
+      "learning_rate": 0.0003163503470274537,
+      "loss": 3.2961,
+      "step": 29545
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5936046242713928,
+      "learning_rate": 0.00031633497369876006,
+      "loss": 2.9672,
+      "step": 29546
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6165637969970703,
+      "learning_rate": 0.00031631960032704325,
+      "loss": 2.9565,
+      "step": 29547
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6325369477272034,
+      "learning_rate": 0.0003163042269123436,
+      "loss": 2.9206,
+      "step": 29548
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6465588212013245,
+      "learning_rate": 0.000316288853454702,
+      "loss": 2.9257,
+      "step": 29549
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.585541307926178,
+      "learning_rate": 0.00031627347995415856,
+      "loss": 2.8962,
+      "step": 29550
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5976788997650146,
+      "learning_rate": 0.00031625810641075383,
+      "loss": 2.9734,
+      "step": 29551
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5807003378868103,
+      "learning_rate": 0.00031624273282452844,
+      "loss": 3.1595,
+      "step": 29552
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.586675763130188,
+      "learning_rate": 0.00031622735919552297,
+      "loss": 2.9426,
+      "step": 29553
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.583739161491394,
+      "learning_rate": 0.0003162119855237776,
+      "loss": 2.9561,
+      "step": 29554
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5614110231399536,
+      "learning_rate": 0.00031619661180933294,
+      "loss": 2.9349,
+      "step": 29555
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5753343105316162,
+      "learning_rate": 0.00031618123805222964,
+      "loss": 3.2066,
+      "step": 29556
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6021926403045654,
+      "learning_rate": 0.00031616586425250797,
+      "loss": 3.0151,
+      "step": 29557
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6058591604232788,
+      "learning_rate": 0.00031615049041020855,
+      "loss": 3.2783,
+      "step": 29558
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.608304500579834,
+      "learning_rate": 0.0003161351165253719,
+      "loss": 2.7891,
+      "step": 29559
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5610315203666687,
+      "learning_rate": 0.0003161197425980385,
+      "loss": 3.0305,
+      "step": 29560
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5982028245925903,
+      "learning_rate": 0.0003161043686282487,
+      "loss": 2.8197,
+      "step": 29561
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5895063877105713,
+      "learning_rate": 0.00031608899461604313,
+      "loss": 3.0352,
+      "step": 29562
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6164365410804749,
+      "learning_rate": 0.0003160736205614622,
+      "loss": 3.0017,
+      "step": 29563
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5825421214103699,
+      "learning_rate": 0.00031605824646454655,
+      "loss": 2.9548,
+      "step": 29564
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.58476722240448,
+      "learning_rate": 0.00031604287232533647,
+      "loss": 3.157,
+      "step": 29565
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5678737163543701,
+      "learning_rate": 0.00031602749814387253,
+      "loss": 2.9115,
+      "step": 29566
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5935869812965393,
+      "learning_rate": 0.00031601212392019535,
+      "loss": 2.9762,
+      "step": 29567
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5967699885368347,
+      "learning_rate": 0.00031599674965434525,
+      "loss": 3.1632,
+      "step": 29568
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6144185066223145,
+      "learning_rate": 0.00031598137534636283,
+      "loss": 2.9605,
+      "step": 29569
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5704265236854553,
+      "learning_rate": 0.0003159660009962885,
+      "loss": 3.0765,
+      "step": 29570
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5627557635307312,
+      "learning_rate": 0.0003159506266041628,
+      "loss": 3.1848,
+      "step": 29571
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5492954254150391,
+      "learning_rate": 0.0003159352521700262,
+      "loss": 2.9258,
+      "step": 29572
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5822631120681763,
+      "learning_rate": 0.0003159198776939193,
+      "loss": 2.9043,
+      "step": 29573
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6288532614707947,
+      "learning_rate": 0.0003159045031758824,
+      "loss": 2.9592,
+      "step": 29574
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5812193751335144,
+      "learning_rate": 0.00031588912861595623,
+      "loss": 3.0854,
+      "step": 29575
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6251932382583618,
+      "learning_rate": 0.00031587375401418105,
+      "loss": 2.8665,
+      "step": 29576
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5608442425727844,
+      "learning_rate": 0.00031585837937059746,
+      "loss": 3.2142,
+      "step": 29577
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6160196661949158,
+      "learning_rate": 0.000315843004685246,
+      "loss": 3.0009,
+      "step": 29578
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5881204009056091,
+      "learning_rate": 0.0003158276299581671,
+      "loss": 2.8357,
+      "step": 29579
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5906282067298889,
+      "learning_rate": 0.00031581225518940124,
+      "loss": 3.1925,
+      "step": 29580
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5734764337539673,
+      "learning_rate": 0.00031579688037898897,
+      "loss": 3.1714,
+      "step": 29581
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5693907141685486,
+      "learning_rate": 0.00031578150552697077,
+      "loss": 3.1624,
+      "step": 29582
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6162329316139221,
+      "learning_rate": 0.00031576613063338707,
+      "loss": 2.9136,
+      "step": 29583
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6050886511802673,
+      "learning_rate": 0.0003157507556982785,
+      "loss": 2.9093,
+      "step": 29584
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5759333968162537,
+      "learning_rate": 0.0003157353807216854,
+      "loss": 3.1633,
+      "step": 29585
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6068954467773438,
+      "learning_rate": 0.0003157200057036484,
+      "loss": 3.0777,
+      "step": 29586
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5611181259155273,
+      "learning_rate": 0.0003157046306442078,
+      "loss": 3.1995,
+      "step": 29587
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5561584830284119,
+      "learning_rate": 0.0003156892555434043,
+      "loss": 3.1371,
+      "step": 29588
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5815036296844482,
+      "learning_rate": 0.0003156738804012784,
+      "loss": 3.0672,
+      "step": 29589
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5883771777153015,
+      "learning_rate": 0.00031565850521787043,
+      "loss": 3.0313,
+      "step": 29590
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5724140405654907,
+      "learning_rate": 0.000315643129993221,
+      "loss": 3.099,
+      "step": 29591
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5555368065834045,
+      "learning_rate": 0.0003156277547273704,
+      "loss": 3.1058,
+      "step": 29592
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5770111083984375,
+      "learning_rate": 0.0003156123794203596,
+      "loss": 3.0728,
+      "step": 29593
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6504952311515808,
+      "learning_rate": 0.00031559700407222865,
+      "loss": 3.0687,
+      "step": 29594
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6245805025100708,
+      "learning_rate": 0.0003155816286830182,
+      "loss": 3.0529,
+      "step": 29595
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6823046207427979,
+      "learning_rate": 0.0003155662532527687,
+      "loss": 3.0419,
+      "step": 29596
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5609052181243896,
+      "learning_rate": 0.00031555087778152075,
+      "loss": 2.8952,
+      "step": 29597
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6359115839004517,
+      "learning_rate": 0.00031553550226931477,
+      "loss": 2.9609,
+      "step": 29598
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5841073989868164,
+      "learning_rate": 0.0003155201267161912,
+      "loss": 2.9918,
+      "step": 29599
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5604556202888489,
+      "learning_rate": 0.00031550475112219074,
+      "loss": 2.8732,
+      "step": 29600
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6146168112754822,
+      "learning_rate": 0.0003154893754873537,
+      "loss": 2.8485,
+      "step": 29601
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5645619034767151,
+      "learning_rate": 0.0003154739998117205,
+      "loss": 3.0853,
+      "step": 29602
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6000223159790039,
+      "learning_rate": 0.0003154586240953318,
+      "loss": 3.0441,
+      "step": 29603
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5900564193725586,
+      "learning_rate": 0.00031544324833822823,
+      "loss": 3.1783,
+      "step": 29604
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6112537384033203,
+      "learning_rate": 0.00031542787254044995,
+      "loss": 3.0663,
+      "step": 29605
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6015599966049194,
+      "learning_rate": 0.00031541249670203773,
+      "loss": 2.9529,
+      "step": 29606
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.562324583530426,
+      "learning_rate": 0.00031539712082303184,
+      "loss": 3.2351,
+      "step": 29607
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6208454370498657,
+      "learning_rate": 0.000315381744903473,
+      "loss": 3.4744,
+      "step": 29608
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5760998725891113,
+      "learning_rate": 0.0003153663689434016,
+      "loss": 3.0427,
+      "step": 29609
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5720778107643127,
+      "learning_rate": 0.00031535099294285803,
+      "loss": 2.9921,
+      "step": 29610
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5598829388618469,
+      "learning_rate": 0.000315335616901883,
+      "loss": 3.1476,
+      "step": 29611
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5947191119194031,
+      "learning_rate": 0.00031532024082051686,
+      "loss": 3.2268,
+      "step": 29612
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5800713300704956,
+      "learning_rate": 0.00031530486469880017,
+      "loss": 2.9083,
+      "step": 29613
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5774427652359009,
+      "learning_rate": 0.0003152894885367734,
+      "loss": 2.8321,
+      "step": 29614
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5743990540504456,
+      "learning_rate": 0.0003152741123344771,
+      "loss": 3.0104,
+      "step": 29615
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.557272732257843,
+      "learning_rate": 0.0003152587360919517,
+      "loss": 3.0485,
+      "step": 29616
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5970534682273865,
+      "learning_rate": 0.0003152433598092378,
+      "loss": 3.011,
+      "step": 29617
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5622311234474182,
+      "learning_rate": 0.0003152279834863757,
+      "loss": 3.1436,
+      "step": 29618
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5816734433174133,
+      "learning_rate": 0.00031521260712340606,
+      "loss": 3.0806,
+      "step": 29619
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5890375375747681,
+      "learning_rate": 0.0003151972307203694,
+      "loss": 3.2086,
+      "step": 29620
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6203916072845459,
+      "learning_rate": 0.00031518185427730606,
+      "loss": 3.126,
+      "step": 29621
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5634053945541382,
+      "learning_rate": 0.00031516647779425677,
+      "loss": 3.3819,
+      "step": 29622
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5932179689407349,
+      "learning_rate": 0.0003151511012712617,
+      "loss": 3.0579,
+      "step": 29623
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5462847948074341,
+      "learning_rate": 0.0003151357247083617,
+      "loss": 3.0788,
+      "step": 29624
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6403987407684326,
+      "learning_rate": 0.00031512034810559703,
+      "loss": 3.0505,
+      "step": 29625
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6161810755729675,
+      "learning_rate": 0.0003151049714630084,
+      "loss": 3.0068,
+      "step": 29626
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5708884596824646,
+      "learning_rate": 0.0003150895947806361,
+      "loss": 2.9304,
+      "step": 29627
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5830209851264954,
+      "learning_rate": 0.0003150742180585207,
+      "loss": 3.0144,
+      "step": 29628
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6132103204727173,
+      "learning_rate": 0.0003150588412967028,
+      "loss": 3.1479,
+      "step": 29629
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5505216717720032,
+      "learning_rate": 0.0003150434644952227,
+      "loss": 3.3665,
+      "step": 29630
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5873746275901794,
+      "learning_rate": 0.00031502808765412107,
+      "loss": 3.2011,
+      "step": 29631
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6037224531173706,
+      "learning_rate": 0.0003150127107734383,
+      "loss": 3.2015,
+      "step": 29632
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6088622212409973,
+      "learning_rate": 0.00031499733385321495,
+      "loss": 3.0153,
+      "step": 29633
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6723757386207581,
+      "learning_rate": 0.0003149819568934915,
+      "loss": 3.1598,
+      "step": 29634
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5801018476486206,
+      "learning_rate": 0.0003149665798943085,
+      "loss": 2.9169,
+      "step": 29635
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5815666913986206,
+      "learning_rate": 0.00031495120285570635,
+      "loss": 3.0191,
+      "step": 29636
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5963479280471802,
+      "learning_rate": 0.00031493582577772573,
+      "loss": 3.0861,
+      "step": 29637
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5587741732597351,
+      "learning_rate": 0.0003149204486604069,
+      "loss": 3.1379,
+      "step": 29638
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.574137806892395,
+      "learning_rate": 0.0003149050715037905,
+      "loss": 2.9059,
+      "step": 29639
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6238990426063538,
+      "learning_rate": 0.0003148896943079171,
+      "loss": 3.1322,
+      "step": 29640
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6018304824829102,
+      "learning_rate": 0.000314874317072827,
+      "loss": 3.0981,
+      "step": 29641
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5812610983848572,
+      "learning_rate": 0.0003148589397985609,
+      "loss": 2.9599,
+      "step": 29642
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6216371655464172,
+      "learning_rate": 0.00031484356248515907,
+      "loss": 3.1933,
+      "step": 29643
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.621842086315155,
+      "learning_rate": 0.0003148281851326623,
+      "loss": 2.7123,
+      "step": 29644
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5874238014221191,
+      "learning_rate": 0.0003148128077411109,
+      "loss": 3.2495,
+      "step": 29645
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.63470059633255,
+      "learning_rate": 0.0003147974303105454,
+      "loss": 2.9006,
+      "step": 29646
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5815867781639099,
+      "learning_rate": 0.0003147820528410063,
+      "loss": 3.078,
+      "step": 29647
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5960510969161987,
+      "learning_rate": 0.00031476667533253416,
+      "loss": 3.0143,
+      "step": 29648
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5853951573371887,
+      "learning_rate": 0.0003147512977851694,
+      "loss": 3.179,
+      "step": 29649
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5538525581359863,
+      "learning_rate": 0.00031473592019895253,
+      "loss": 2.9866,
+      "step": 29650
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6282745599746704,
+      "learning_rate": 0.0003147205425739242,
+      "loss": 3.2866,
+      "step": 29651
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6330800652503967,
+      "learning_rate": 0.00031470516491012476,
+      "loss": 2.8979,
+      "step": 29652
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6331319808959961,
+      "learning_rate": 0.00031468978720759473,
+      "loss": 2.9076,
+      "step": 29653
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.587941288948059,
+      "learning_rate": 0.0003146744094663745,
+      "loss": 3.2152,
+      "step": 29654
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5891446471214294,
+      "learning_rate": 0.0003146590316865049,
+      "loss": 3.0108,
+      "step": 29655
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5783140063285828,
+      "learning_rate": 0.0003146436538680261,
+      "loss": 2.781,
+      "step": 29656
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5963090658187866,
+      "learning_rate": 0.00031462827601097873,
+      "loss": 3.212,
+      "step": 29657
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6018965840339661,
+      "learning_rate": 0.0003146128981154034,
+      "loss": 3.201,
+      "step": 29658
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5825430750846863,
+      "learning_rate": 0.0003145975201813404,
+      "loss": 3.1323,
+      "step": 29659
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.58023601770401,
+      "learning_rate": 0.0003145821422088304,
+      "loss": 2.8852,
+      "step": 29660
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6018133759498596,
+      "learning_rate": 0.00031456676419791384,
+      "loss": 3.2422,
+      "step": 29661
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6052143573760986,
+      "learning_rate": 0.0003145513861486313,
+      "loss": 3.0118,
+      "step": 29662
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5762540102005005,
+      "learning_rate": 0.0003145360080610231,
+      "loss": 3.1484,
+      "step": 29663
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5601885914802551,
+      "learning_rate": 0.0003145206299351298,
+      "loss": 3.2171,
+      "step": 29664
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6563633680343628,
+      "learning_rate": 0.00031450525177099205,
+      "loss": 2.9215,
+      "step": 29665
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5905854105949402,
+      "learning_rate": 0.0003144898735686503,
+      "loss": 3.0631,
+      "step": 29666
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5945887565612793,
+      "learning_rate": 0.00031447449532814495,
+      "loss": 2.9145,
+      "step": 29667
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6269522309303284,
+      "learning_rate": 0.00031445911704951657,
+      "loss": 3.1563,
+      "step": 29668
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.577925980091095,
+      "learning_rate": 0.00031444373873280564,
+      "loss": 2.9717,
+      "step": 29669
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.572482705116272,
+      "learning_rate": 0.00031442836037805276,
+      "loss": 3.1004,
+      "step": 29670
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6007272005081177,
+      "learning_rate": 0.0003144129819852983,
+      "loss": 3.157,
+      "step": 29671
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5722142457962036,
+      "learning_rate": 0.00031439760355458284,
+      "loss": 2.966,
+      "step": 29672
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5537738800048828,
+      "learning_rate": 0.00031438222508594683,
+      "loss": 3.095,
+      "step": 29673
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5694095492362976,
+      "learning_rate": 0.0003143668465794308,
+      "loss": 3.0124,
+      "step": 29674
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6016697287559509,
+      "learning_rate": 0.0003143514680350752,
+      "loss": 3.0182,
+      "step": 29675
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5762494206428528,
+      "learning_rate": 0.0003143360894529207,
+      "loss": 2.9215,
+      "step": 29676
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5735913515090942,
+      "learning_rate": 0.00031432071083300773,
+      "loss": 3.148,
+      "step": 29677
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5936062932014465,
+      "learning_rate": 0.0003143053321753767,
+      "loss": 2.9584,
+      "step": 29678
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5691812634468079,
+      "learning_rate": 0.00031428995348006816,
+      "loss": 3.23,
+      "step": 29679
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5691925883293152,
+      "learning_rate": 0.00031427457474712274,
+      "loss": 3.1767,
+      "step": 29680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6052525043487549,
+      "learning_rate": 0.00031425919597658067,
+      "loss": 2.9384,
+      "step": 29681
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6288476586341858,
+      "learning_rate": 0.00031424381716848275,
+      "loss": 2.9858,
+      "step": 29682
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5947161316871643,
+      "learning_rate": 0.0003142284383228693,
+      "loss": 3.2128,
+      "step": 29683
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7295497059822083,
+      "learning_rate": 0.00031421305943978093,
+      "loss": 3.237,
+      "step": 29684
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5625701546669006,
+      "learning_rate": 0.000314197680519258,
+      "loss": 2.8254,
+      "step": 29685
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6221656203269958,
+      "learning_rate": 0.0003141823015613412,
+      "loss": 2.8866,
+      "step": 29686
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6753302812576294,
+      "learning_rate": 0.0003141669225660709,
+      "loss": 3.0037,
+      "step": 29687
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.579978883266449,
+      "learning_rate": 0.0003141515435334878,
+      "loss": 2.7757,
+      "step": 29688
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5823621153831482,
+      "learning_rate": 0.000314136164463632,
+      "loss": 3.2779,
+      "step": 29689
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6142686605453491,
+      "learning_rate": 0.0003141207853565444,
+      "loss": 2.8307,
+      "step": 29690
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5641099810600281,
+      "learning_rate": 0.0003141054062122654,
+      "loss": 2.9068,
+      "step": 29691
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6020845770835876,
+      "learning_rate": 0.00031409002703083547,
+      "loss": 3.0636,
+      "step": 29692
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5882956981658936,
+      "learning_rate": 0.00031407464781229503,
+      "loss": 3.0913,
+      "step": 29693
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.593789279460907,
+      "learning_rate": 0.00031405926855668475,
+      "loss": 3.1935,
+      "step": 29694
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5638978481292725,
+      "learning_rate": 0.00031404388926404507,
+      "loss": 3.1643,
+      "step": 29695
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.595602810382843,
+      "learning_rate": 0.00031402850993441647,
+      "loss": 3.1038,
+      "step": 29696
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6187957525253296,
+      "learning_rate": 0.00031401313056783945,
+      "loss": 2.9744,
+      "step": 29697
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5919390916824341,
+      "learning_rate": 0.0003139977511643546,
+      "loss": 3.2703,
+      "step": 29698
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.566708505153656,
+      "learning_rate": 0.00031398237172400236,
+      "loss": 3.0879,
+      "step": 29699
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5841004848480225,
+      "learning_rate": 0.0003139669922468231,
+      "loss": 3.036,
+      "step": 29700
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5895936489105225,
+      "learning_rate": 0.00031395161273285765,
+      "loss": 2.9205,
+      "step": 29701
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5551169514656067,
+      "learning_rate": 0.00031393623318214635,
+      "loss": 2.9754,
+      "step": 29702
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5898941159248352,
+      "learning_rate": 0.00031392085359472954,
+      "loss": 2.9792,
+      "step": 29703
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6206042170524597,
+      "learning_rate": 0.000313905473970648,
+      "loss": 3.1062,
+      "step": 29704
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6914759278297424,
+      "learning_rate": 0.00031389009430994195,
+      "loss": 3.0206,
+      "step": 29705
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.619304358959198,
+      "learning_rate": 0.00031387471461265227,
+      "loss": 2.9323,
+      "step": 29706
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.573853075504303,
+      "learning_rate": 0.00031385933487881916,
+      "loss": 3.3458,
+      "step": 29707
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5841895341873169,
+      "learning_rate": 0.0003138439551084833,
+      "loss": 2.78,
+      "step": 29708
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5773389339447021,
+      "learning_rate": 0.0003138285753016851,
+      "loss": 3.123,
+      "step": 29709
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5712064504623413,
+      "learning_rate": 0.00031381319545846503,
+      "loss": 3.2139,
+      "step": 29710
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6082553267478943,
+      "learning_rate": 0.00031379781557886366,
+      "loss": 3.19,
+      "step": 29711
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6170501112937927,
+      "learning_rate": 0.00031378243566292156,
+      "loss": 3.0305,
+      "step": 29712
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5564154982566833,
+      "learning_rate": 0.00031376705571067924,
+      "loss": 3.0066,
+      "step": 29713
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6043330430984497,
+      "learning_rate": 0.000313751675722177,
+      "loss": 3.0141,
+      "step": 29714
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.633105456829071,
+      "learning_rate": 0.00031373629569745546,
+      "loss": 2.8554,
+      "step": 29715
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5764904022216797,
+      "learning_rate": 0.0003137209156365553,
+      "loss": 2.9793,
+      "step": 29716
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6007531881332397,
+      "learning_rate": 0.00031370553553951687,
+      "loss": 3.0194,
+      "step": 29717
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.597282350063324,
+      "learning_rate": 0.0003136901554063807,
+      "loss": 3.086,
+      "step": 29718
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6131802797317505,
+      "learning_rate": 0.0003136747752371873,
+      "loss": 3.1779,
+      "step": 29719
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5542619824409485,
+      "learning_rate": 0.0003136593950319771,
+      "loss": 3.0502,
+      "step": 29720
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.662255048751831,
+      "learning_rate": 0.00031364401479079074,
+      "loss": 3.0517,
+      "step": 29721
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5674471259117126,
+      "learning_rate": 0.00031362863451366864,
+      "loss": 2.9217,
+      "step": 29722
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5355050563812256,
+      "learning_rate": 0.0003136132542006514,
+      "loss": 3.1893,
+      "step": 29723
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5745908617973328,
+      "learning_rate": 0.00031359787385177944,
+      "loss": 3.2364,
+      "step": 29724
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5746610760688782,
+      "learning_rate": 0.0003135824934670932,
+      "loss": 3.1486,
+      "step": 29725
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5605559945106506,
+      "learning_rate": 0.00031356711304663336,
+      "loss": 3.0998,
+      "step": 29726
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5893903970718384,
+      "learning_rate": 0.0003135517325904403,
+      "loss": 3.0817,
+      "step": 29727
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.583816647529602,
+      "learning_rate": 0.00031353635209855474,
+      "loss": 3.0743,
+      "step": 29728
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.616067111492157,
+      "learning_rate": 0.00031352097157101696,
+      "loss": 3.1714,
+      "step": 29729
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6139724850654602,
+      "learning_rate": 0.00031350559100786743,
+      "loss": 3.0468,
+      "step": 29730
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5720040202140808,
+      "learning_rate": 0.0003134902104091469,
+      "loss": 3.2753,
+      "step": 29731
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6097339391708374,
+      "learning_rate": 0.00031347482977489577,
+      "loss": 3.0575,
+      "step": 29732
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6279599666595459,
+      "learning_rate": 0.00031345944910515447,
+      "loss": 3.0261,
+      "step": 29733
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5827269554138184,
+      "learning_rate": 0.00031344406839996355,
+      "loss": 3.2295,
+      "step": 29734
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5751627683639526,
+      "learning_rate": 0.0003134286876593637,
+      "loss": 2.9174,
+      "step": 29735
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5884239077568054,
+      "learning_rate": 0.000313413306883395,
+      "loss": 3.0503,
+      "step": 29736
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5882185697555542,
+      "learning_rate": 0.00031339792607209844,
+      "loss": 2.9154,
+      "step": 29737
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5342128276824951,
+      "learning_rate": 0.00031338254522551417,
+      "loss": 3.0905,
+      "step": 29738
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.566861093044281,
+      "learning_rate": 0.00031336716434368303,
+      "loss": 3.0281,
+      "step": 29739
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5697751045227051,
+      "learning_rate": 0.00031335178342664524,
+      "loss": 2.994,
+      "step": 29740
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6492906212806702,
+      "learning_rate": 0.00031333640247444136,
+      "loss": 3.1775,
+      "step": 29741
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6085817813873291,
+      "learning_rate": 0.00031332102148711214,
+      "loss": 3.1993,
+      "step": 29742
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5619418621063232,
+      "learning_rate": 0.0003133056404646978,
+      "loss": 2.982,
+      "step": 29743
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5778747200965881,
+      "learning_rate": 0.000313290259407239,
+      "loss": 2.8579,
+      "step": 29744
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.597019374370575,
+      "learning_rate": 0.00031327487831477616,
+      "loss": 2.9197,
+      "step": 29745
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5481294989585876,
+      "learning_rate": 0.00031325949718734986,
+      "loss": 3.0998,
+      "step": 29746
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5638428926467896,
+      "learning_rate": 0.0003132441160250006,
+      "loss": 3.1415,
+      "step": 29747
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5789369940757751,
+      "learning_rate": 0.0003132287348277689,
+      "loss": 3.1912,
+      "step": 29748
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5999630689620972,
+      "learning_rate": 0.00031321335359569527,
+      "loss": 3.0397,
+      "step": 29749
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5871257185935974,
+      "learning_rate": 0.0003131979723288203,
+      "loss": 2.9909,
+      "step": 29750
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6134441494941711,
+      "learning_rate": 0.0003131825910271842,
+      "loss": 3.1867,
+      "step": 29751
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.622126579284668,
+      "learning_rate": 0.0003131672096908278,
+      "loss": 3.1849,
+      "step": 29752
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5536166429519653,
+      "learning_rate": 0.00031315182831979157,
+      "loss": 3.1054,
+      "step": 29753
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5597190856933594,
+      "learning_rate": 0.0003131364469141159,
+      "loss": 3.1826,
+      "step": 29754
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5869278311729431,
+      "learning_rate": 0.00031312106547384133,
+      "loss": 3.029,
+      "step": 29755
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6084804534912109,
+      "learning_rate": 0.0003131056839990084,
+      "loss": 2.6037,
+      "step": 29756
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5875492691993713,
+      "learning_rate": 0.00031309030248965764,
+      "loss": 3.1481,
+      "step": 29757
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6422194242477417,
+      "learning_rate": 0.00031307492094582957,
+      "loss": 3.0253,
+      "step": 29758
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5797473788261414,
+      "learning_rate": 0.0003130595393675646,
+      "loss": 3.0883,
+      "step": 29759
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6029950976371765,
+      "learning_rate": 0.0003130441577549034,
+      "loss": 2.8537,
+      "step": 29760
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5672838687896729,
+      "learning_rate": 0.00031302877610788637,
+      "loss": 2.8068,
+      "step": 29761
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5959682464599609,
+      "learning_rate": 0.00031301339442655403,
+      "loss": 3.1368,
+      "step": 29762
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6413449645042419,
+      "learning_rate": 0.00031299801271094687,
+      "loss": 2.7994,
+      "step": 29763
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5850305557250977,
+      "learning_rate": 0.0003129826309611056,
+      "loss": 3.1345,
+      "step": 29764
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6164899468421936,
+      "learning_rate": 0.0003129672491770705,
+      "loss": 3.048,
+      "step": 29765
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6072206497192383,
+      "learning_rate": 0.00031295186735888204,
+      "loss": 3.0531,
+      "step": 29766
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5849358439445496,
+      "learning_rate": 0.000312936485506581,
+      "loss": 3.0533,
+      "step": 29767
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5805505514144897,
+      "learning_rate": 0.0003129211036202077,
+      "loss": 2.9414,
+      "step": 29768
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5903421640396118,
+      "learning_rate": 0.0003129057216998027,
+      "loss": 3.0021,
+      "step": 29769
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5872484445571899,
+      "learning_rate": 0.0003128903397454066,
+      "loss": 2.9317,
+      "step": 29770
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6200918555259705,
+      "learning_rate": 0.00031287495775705963,
+      "loss": 3.0879,
+      "step": 29771
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5617815256118774,
+      "learning_rate": 0.00031285957573480263,
+      "loss": 3.0599,
+      "step": 29772
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5852945446968079,
+      "learning_rate": 0.00031284419367867596,
+      "loss": 2.9581,
+      "step": 29773
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6024762392044067,
+      "learning_rate": 0.0003128288115887202,
+      "loss": 2.8992,
+      "step": 29774
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6069262623786926,
+      "learning_rate": 0.00031281342946497584,
+      "loss": 3.1949,
+      "step": 29775
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6076410412788391,
+      "learning_rate": 0.00031279804730748327,
+      "loss": 2.93,
+      "step": 29776
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.599779486656189,
+      "learning_rate": 0.00031278266511628307,
+      "loss": 3.0456,
+      "step": 29777
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6234508156776428,
+      "learning_rate": 0.0003127672828914159,
+      "loss": 3.0092,
+      "step": 29778
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5925348997116089,
+      "learning_rate": 0.00031275190063292214,
+      "loss": 3.2223,
+      "step": 29779
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6192617416381836,
+      "learning_rate": 0.00031273651834084234,
+      "loss": 2.9969,
+      "step": 29780
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6267848014831543,
+      "learning_rate": 0.000312721136015217,
+      "loss": 2.8548,
+      "step": 29781
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5980367064476013,
+      "learning_rate": 0.00031270575365608663,
+      "loss": 2.9697,
+      "step": 29782
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5981336236000061,
+      "learning_rate": 0.0003126903712634917,
+      "loss": 3.051,
+      "step": 29783
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6255171298980713,
+      "learning_rate": 0.00031267498883747277,
+      "loss": 3.1955,
+      "step": 29784
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5721108913421631,
+      "learning_rate": 0.00031265960637807044,
+      "loss": 3.077,
+      "step": 29785
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5838683843612671,
+      "learning_rate": 0.00031264422388532515,
+      "loss": 3.2894,
+      "step": 29786
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5979468822479248,
+      "learning_rate": 0.0003126288413592773,
+      "loss": 2.9987,
+      "step": 29787
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.601974368095398,
+      "learning_rate": 0.0003126134587999676,
+      "loss": 2.9889,
+      "step": 29788
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5563658475875854,
+      "learning_rate": 0.0003125980762074364,
+      "loss": 2.9098,
+      "step": 29789
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.628157913684845,
+      "learning_rate": 0.0003125826935817244,
+      "loss": 3.2493,
+      "step": 29790
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5461869239807129,
+      "learning_rate": 0.00031256731092287195,
+      "loss": 3.1618,
+      "step": 29791
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6051981449127197,
+      "learning_rate": 0.0003125519282309196,
+      "loss": 3.0727,
+      "step": 29792
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5399933457374573,
+      "learning_rate": 0.00031253654550590795,
+      "loss": 2.8477,
+      "step": 29793
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5897714495658875,
+      "learning_rate": 0.00031252116274787736,
+      "loss": 3.0971,
+      "step": 29794
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5866346955299377,
+      "learning_rate": 0.00031250577995686854,
+      "loss": 3.0104,
+      "step": 29795
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6055521368980408,
+      "learning_rate": 0.00031249039713292183,
+      "loss": 3.2093,
+      "step": 29796
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.581646203994751,
+      "learning_rate": 0.0003124750142760778,
+      "loss": 3.2742,
+      "step": 29797
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5622509717941284,
+      "learning_rate": 0.00031245963138637706,
+      "loss": 3.1615,
+      "step": 29798
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5802165269851685,
+      "learning_rate": 0.00031244424846386,
+      "loss": 2.7669,
+      "step": 29799
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.578447699546814,
+      "learning_rate": 0.0003124288655085671,
+      "loss": 3.0275,
+      "step": 29800
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6063286066055298,
+      "learning_rate": 0.00031241348252053914,
+      "loss": 3.052,
+      "step": 29801
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5866938829421997,
+      "learning_rate": 0.0003123980994998163,
+      "loss": 2.9231,
+      "step": 29802
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5843689441680908,
+      "learning_rate": 0.00031238271644643933,
+      "loss": 3.0812,
+      "step": 29803
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6158820986747742,
+      "learning_rate": 0.00031236733336044876,
+      "loss": 2.9661,
+      "step": 29804
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6195018887519836,
+      "learning_rate": 0.00031235195024188484,
+      "loss": 2.955,
+      "step": 29805
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5677433013916016,
+      "learning_rate": 0.00031233656709078836,
+      "loss": 3.0007,
+      "step": 29806
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5664675831794739,
+      "learning_rate": 0.0003123211839071997,
+      "loss": 2.8784,
+      "step": 29807
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5745792984962463,
+      "learning_rate": 0.00031230580069115945,
+      "loss": 2.9201,
+      "step": 29808
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5681601762771606,
+      "learning_rate": 0.000312290417442708,
+      "loss": 2.9938,
+      "step": 29809
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.580299437046051,
+      "learning_rate": 0.00031227503416188604,
+      "loss": 2.8414,
+      "step": 29810
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5805426239967346,
+      "learning_rate": 0.00031225965084873396,
+      "loss": 3.0765,
+      "step": 29811
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5639891624450684,
+      "learning_rate": 0.00031224426750329245,
+      "loss": 3.0892,
+      "step": 29812
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5976626873016357,
+      "learning_rate": 0.0003122288841256017,
+      "loss": 3.1597,
+      "step": 29813
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6398038864135742,
+      "learning_rate": 0.0003122135007157025,
+      "loss": 2.881,
+      "step": 29814
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5923072695732117,
+      "learning_rate": 0.0003121981172736354,
+      "loss": 3.1807,
+      "step": 29815
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5970144271850586,
+      "learning_rate": 0.00031218273379944063,
+      "loss": 3.1799,
+      "step": 29816
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6289509534835815,
+      "learning_rate": 0.00031216735029315903,
+      "loss": 2.7998,
+      "step": 29817
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.609671413898468,
+      "learning_rate": 0.00031215196675483074,
+      "loss": 2.9472,
+      "step": 29818
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5885813236236572,
+      "learning_rate": 0.00031213658318449676,
+      "loss": 3.249,
+      "step": 29819
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5870927572250366,
+      "learning_rate": 0.0003121211995821973,
+      "loss": 2.9601,
+      "step": 29820
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5752508044242859,
+      "learning_rate": 0.00031210581594797284,
+      "loss": 3.1265,
+      "step": 29821
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5618848204612732,
+      "learning_rate": 0.00031209043228186406,
+      "loss": 2.8466,
+      "step": 29822
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5932725667953491,
+      "learning_rate": 0.00031207504858391145,
+      "loss": 3.1074,
+      "step": 29823
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.567245602607727,
+      "learning_rate": 0.00031205966485415535,
+      "loss": 2.9608,
+      "step": 29824
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5531851649284363,
+      "learning_rate": 0.0003120442810926365,
+      "loss": 3.1427,
+      "step": 29825
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5592467188835144,
+      "learning_rate": 0.00031202889729939543,
+      "loss": 3.1028,
+      "step": 29826
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6010973453521729,
+      "learning_rate": 0.0003120135134744724,
+      "loss": 3.1718,
+      "step": 29827
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.587733805179596,
+      "learning_rate": 0.0003119981296179081,
+      "loss": 2.8077,
+      "step": 29828
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5497106313705444,
+      "learning_rate": 0.0003119827457297431,
+      "loss": 3.0282,
+      "step": 29829
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5502404570579529,
+      "learning_rate": 0.00031196736181001784,
+      "loss": 2.9867,
+      "step": 29830
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5896203517913818,
+      "learning_rate": 0.0003119519778587729,
+      "loss": 2.9452,
+      "step": 29831
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5907154679298401,
+      "learning_rate": 0.0003119365938760486,
+      "loss": 2.919,
+      "step": 29832
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5724136829376221,
+      "learning_rate": 0.00031192120986188573,
+      "loss": 3.052,
+      "step": 29833
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5927634835243225,
+      "learning_rate": 0.0003119058258163247,
+      "loss": 3.0744,
+      "step": 29834
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5511701107025146,
+      "learning_rate": 0.00031189044173940596,
+      "loss": 2.9463,
+      "step": 29835
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6173757314682007,
+      "learning_rate": 0.00031187505763117016,
+      "loss": 2.9852,
+      "step": 29836
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6480523347854614,
+      "learning_rate": 0.0003118596734916577,
+      "loss": 3.0302,
+      "step": 29837
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6319320201873779,
+      "learning_rate": 0.00031184428932090906,
+      "loss": 3.1108,
+      "step": 29838
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5854665040969849,
+      "learning_rate": 0.0003118289051189649,
+      "loss": 3.1069,
+      "step": 29839
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5656838417053223,
+      "learning_rate": 0.00031181352088586567,
+      "loss": 3.0943,
+      "step": 29840
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5562570691108704,
+      "learning_rate": 0.000311798136621652,
+      "loss": 2.9549,
+      "step": 29841
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.635686457157135,
+      "learning_rate": 0.0003117827523263642,
+      "loss": 3.1737,
+      "step": 29842
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5831470489501953,
+      "learning_rate": 0.0003117673680000429,
+      "loss": 2.9611,
+      "step": 29843
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5976853966712952,
+      "learning_rate": 0.00031175198364272866,
+      "loss": 2.9907,
+      "step": 29844
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6116719841957092,
+      "learning_rate": 0.000311736599254462,
+      "loss": 2.9301,
+      "step": 29845
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6009196639060974,
+      "learning_rate": 0.0003117212148352833,
+      "loss": 3.0093,
+      "step": 29846
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5840451717376709,
+      "learning_rate": 0.00031170583038523324,
+      "loss": 3.1293,
+      "step": 29847
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5879883766174316,
+      "learning_rate": 0.0003116904459043523,
+      "loss": 2.8642,
+      "step": 29848
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5799393057823181,
+      "learning_rate": 0.00031167506139268084,
+      "loss": 3.0245,
+      "step": 29849
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5922977328300476,
+      "learning_rate": 0.0003116596768502596,
+      "loss": 3.0836,
+      "step": 29850
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5858464241027832,
+      "learning_rate": 0.000311644292277129,
+      "loss": 3.0924,
+      "step": 29851
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5689262747764587,
+      "learning_rate": 0.00031162890767332966,
+      "loss": 2.9704,
+      "step": 29852
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6013664603233337,
+      "learning_rate": 0.0003116135230389019,
+      "loss": 2.994,
+      "step": 29853
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5736306309700012,
+      "learning_rate": 0.00031159813837388636,
+      "loss": 2.8971,
+      "step": 29854
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6058791279792786,
+      "learning_rate": 0.00031158275367832366,
+      "loss": 3.0407,
+      "step": 29855
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.563359797000885,
+      "learning_rate": 0.00031156736895225417,
+      "loss": 3.3032,
+      "step": 29856
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6421247720718384,
+      "learning_rate": 0.00031155198419571844,
+      "loss": 3.0148,
+      "step": 29857
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5899890065193176,
+      "learning_rate": 0.00031153659940875696,
+      "loss": 2.8782,
+      "step": 29858
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5498698353767395,
+      "learning_rate": 0.00031152121459141033,
+      "loss": 3.1737,
+      "step": 29859
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.564635694026947,
+      "learning_rate": 0.00031150582974371905,
+      "loss": 3.1709,
+      "step": 29860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5672764778137207,
+      "learning_rate": 0.0003114904448657236,
+      "loss": 3.0535,
+      "step": 29861
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6894760727882385,
+      "learning_rate": 0.0003114750599574646,
+      "loss": 3.0999,
+      "step": 29862
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6003525257110596,
+      "learning_rate": 0.0003114596750189825,
+      "loss": 3.137,
+      "step": 29863
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5983224511146545,
+      "learning_rate": 0.0003114442900503177,
+      "loss": 3.0206,
+      "step": 29864
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5564965009689331,
+      "learning_rate": 0.000311428905051511,
+      "loss": 2.9906,
+      "step": 29865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5878624320030212,
+      "learning_rate": 0.0003114135200226027,
+      "loss": 2.8516,
+      "step": 29866
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5519507527351379,
+      "learning_rate": 0.00031139813496363337,
+      "loss": 3.2566,
+      "step": 29867
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5717609524726868,
+      "learning_rate": 0.0003113827498746435,
+      "loss": 3.0263,
+      "step": 29868
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5674951076507568,
+      "learning_rate": 0.00031136736475567364,
+      "loss": 3.1298,
+      "step": 29869
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6507361531257629,
+      "learning_rate": 0.0003113519796067645,
+      "loss": 3.1109,
+      "step": 29870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6233397722244263,
+      "learning_rate": 0.0003113365944279563,
+      "loss": 2.8556,
+      "step": 29871
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5794210433959961,
+      "learning_rate": 0.0003113212092192897,
+      "loss": 2.9724,
+      "step": 29872
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5545445084571838,
+      "learning_rate": 0.0003113058239808052,
+      "loss": 2.9606,
+      "step": 29873
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5938491821289062,
+      "learning_rate": 0.00031129043871254344,
+      "loss": 3.1588,
+      "step": 29874
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5351107716560364,
+      "learning_rate": 0.0003112750534145448,
+      "loss": 3.0698,
+      "step": 29875
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5751978754997253,
+      "learning_rate": 0.0003112596680868497,
+      "loss": 3.298,
+      "step": 29876
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6225106120109558,
+      "learning_rate": 0.00031124428272949896,
+      "loss": 3.1264,
+      "step": 29877
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5818681716918945,
+      "learning_rate": 0.00031122889734253294,
+      "loss": 3.1766,
+      "step": 29878
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5759578943252563,
+      "learning_rate": 0.000311213511925992,
+      "loss": 3.0989,
+      "step": 29879
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5782442688941956,
+      "learning_rate": 0.0003111981264799169,
+      "loss": 3.1062,
+      "step": 29880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.606756865978241,
+      "learning_rate": 0.00031118274100434823,
+      "loss": 2.9649,
+      "step": 29881
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5722266435623169,
+      "learning_rate": 0.0003111673554993263,
+      "loss": 2.9258,
+      "step": 29882
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5965888500213623,
+      "learning_rate": 0.0003111519699648916,
+      "loss": 2.7842,
+      "step": 29883
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5483811497688293,
+      "learning_rate": 0.0003111365844010849,
+      "loss": 2.9596,
+      "step": 29884
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5529528856277466,
+      "learning_rate": 0.00031112119880794654,
+      "loss": 3.1932,
+      "step": 29885
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.624657392501831,
+      "learning_rate": 0.000311105813185517,
+      "loss": 2.9514,
+      "step": 29886
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5750808119773865,
+      "learning_rate": 0.0003110904275338369,
+      "loss": 3.1854,
+      "step": 29887
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5993642807006836,
+      "learning_rate": 0.0003110750418529469,
+      "loss": 2.9088,
+      "step": 29888
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6091515421867371,
+      "learning_rate": 0.00031105965614288723,
+      "loss": 3.1339,
+      "step": 29889
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6110482811927795,
+      "learning_rate": 0.0003110442704036985,
+      "loss": 3.286,
+      "step": 29890
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6325213313102722,
+      "learning_rate": 0.00031102888463542137,
+      "loss": 3.0859,
+      "step": 29891
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.586706280708313,
+      "learning_rate": 0.0003110134988380964,
+      "loss": 2.8581,
+      "step": 29892
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5791521668434143,
+      "learning_rate": 0.0003109981130117639,
+      "loss": 3.0277,
+      "step": 29893
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6084620952606201,
+      "learning_rate": 0.0003109827271564644,
+      "loss": 3.1767,
+      "step": 29894
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6085389852523804,
+      "learning_rate": 0.0003109673412722386,
+      "loss": 3.0384,
+      "step": 29895
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6390563249588013,
+      "learning_rate": 0.0003109519553591269,
+      "loss": 2.8968,
+      "step": 29896
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5879514813423157,
+      "learning_rate": 0.0003109365694171699,
+      "loss": 3.0246,
+      "step": 29897
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6184220314025879,
+      "learning_rate": 0.000310921183446408,
+      "loss": 2.9875,
+      "step": 29898
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.629254162311554,
+      "learning_rate": 0.00031090579744688196,
+      "loss": 3.1362,
+      "step": 29899
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6382981538772583,
+      "learning_rate": 0.00031089041141863194,
+      "loss": 3.0567,
+      "step": 29900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5889188647270203,
+      "learning_rate": 0.00031087502536169875,
+      "loss": 3.0674,
+      "step": 29901
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6274064183235168,
+      "learning_rate": 0.0003108596392761229,
+      "loss": 3.0911,
+      "step": 29902
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6702287197113037,
+      "learning_rate": 0.0003108442531619448,
+      "loss": 2.8349,
+      "step": 29903
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5988737940788269,
+      "learning_rate": 0.00031082886701920505,
+      "loss": 3.0651,
+      "step": 29904
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6140710115432739,
+      "learning_rate": 0.00031081348084794404,
+      "loss": 2.8459,
+      "step": 29905
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5804564356803894,
+      "learning_rate": 0.00031079809464820257,
+      "loss": 3.0062,
+      "step": 29906
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5837451219558716,
+      "learning_rate": 0.0003107827084200209,
+      "loss": 3.1248,
+      "step": 29907
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6358888149261475,
+      "learning_rate": 0.0003107673221634397,
+      "loss": 3.0773,
+      "step": 29908
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6011119484901428,
+      "learning_rate": 0.00031075193587849943,
+      "loss": 3.0292,
+      "step": 29909
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6402463316917419,
+      "learning_rate": 0.00031073654956524063,
+      "loss": 2.7842,
+      "step": 29910
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5692670345306396,
+      "learning_rate": 0.0003107211632237037,
+      "loss": 3.114,
+      "step": 29911
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5846226215362549,
+      "learning_rate": 0.0003107057768539295,
+      "loss": 3.1666,
+      "step": 29912
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5675181150436401,
+      "learning_rate": 0.0003106903904559582,
+      "loss": 2.7647,
+      "step": 29913
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6056368350982666,
+      "learning_rate": 0.00031067500402983056,
+      "loss": 3.0578,
+      "step": 29914
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5673414468765259,
+      "learning_rate": 0.00031065961757558694,
+      "loss": 2.9745,
+      "step": 29915
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6073873043060303,
+      "learning_rate": 0.00031064423109326795,
+      "loss": 3.1665,
+      "step": 29916
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5638627409934998,
+      "learning_rate": 0.0003106288445829142,
+      "loss": 3.1064,
+      "step": 29917
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5839649438858032,
+      "learning_rate": 0.00031061345804456604,
+      "loss": 2.7305,
+      "step": 29918
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5768024325370789,
+      "learning_rate": 0.00031059807147826405,
+      "loss": 2.9832,
+      "step": 29919
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6130768656730652,
+      "learning_rate": 0.0003105826848840488,
+      "loss": 2.9809,
+      "step": 29920
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5641641616821289,
+      "learning_rate": 0.0003105672982619608,
+      "loss": 2.9487,
+      "step": 29921
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5932754278182983,
+      "learning_rate": 0.0003105519116120405,
+      "loss": 3.1855,
+      "step": 29922
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6482731699943542,
+      "learning_rate": 0.00031053652493432857,
+      "loss": 3.0039,
+      "step": 29923
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.586216926574707,
+      "learning_rate": 0.0003105211382288655,
+      "loss": 3.0655,
+      "step": 29924
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5854665637016296,
+      "learning_rate": 0.0003105057514956917,
+      "loss": 3.1271,
+      "step": 29925
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.592591404914856,
+      "learning_rate": 0.0003104903647348478,
+      "loss": 2.8566,
+      "step": 29926
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5938833951950073,
+      "learning_rate": 0.0003104749779463743,
+      "loss": 2.9205,
+      "step": 29927
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5724154114723206,
+      "learning_rate": 0.00031045959113031187,
+      "loss": 2.8484,
+      "step": 29928
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5788401961326599,
+      "learning_rate": 0.00031044420428670067,
+      "loss": 3.0497,
+      "step": 29929
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5472137928009033,
+      "learning_rate": 0.0003104288174155816,
+      "loss": 3.1811,
+      "step": 29930
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6054922938346863,
+      "learning_rate": 0.0003104134305169948,
+      "loss": 3.059,
+      "step": 29931
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6112335920333862,
+      "learning_rate": 0.0003103980435909812,
+      "loss": 2.9871,
+      "step": 29932
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5911417007446289,
+      "learning_rate": 0.0003103826566375812,
+      "loss": 3.0777,
+      "step": 29933
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.724625825881958,
+      "learning_rate": 0.0003103672696568352,
+      "loss": 3.1604,
+      "step": 29934
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5843529105186462,
+      "learning_rate": 0.0003103518826487838,
+      "loss": 2.8306,
+      "step": 29935
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5531800389289856,
+      "learning_rate": 0.0003103364956134676,
+      "loss": 2.9553,
+      "step": 29936
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6143542528152466,
+      "learning_rate": 0.00031032110855092705,
+      "loss": 2.8877,
+      "step": 29937
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5682981014251709,
+      "learning_rate": 0.0003103057214612026,
+      "loss": 3.1894,
+      "step": 29938
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6502469182014465,
+      "learning_rate": 0.000310290334344335,
+      "loss": 3.2924,
+      "step": 29939
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5962908267974854,
+      "learning_rate": 0.00031027494720036456,
+      "loss": 2.8696,
+      "step": 29940
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6846643090248108,
+      "learning_rate": 0.0003102595600293318,
+      "loss": 2.8676,
+      "step": 29941
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5837406516075134,
+      "learning_rate": 0.00031024417283127744,
+      "loss": 3.0501,
+      "step": 29942
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6407919526100159,
+      "learning_rate": 0.0003102287856062419,
+      "loss": 3.2333,
+      "step": 29943
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.615726888179779,
+      "learning_rate": 0.0003102133983542657,
+      "loss": 3.1496,
+      "step": 29944
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5835103392601013,
+      "learning_rate": 0.00031019801107538933,
+      "loss": 3.0538,
+      "step": 29945
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5882119536399841,
+      "learning_rate": 0.0003101826237696534,
+      "loss": 2.8205,
+      "step": 29946
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5969010591506958,
+      "learning_rate": 0.0003101672364370984,
+      "loss": 2.8979,
+      "step": 29947
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.561356782913208,
+      "learning_rate": 0.00031015184907776485,
+      "loss": 2.9959,
+      "step": 29948
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5852020978927612,
+      "learning_rate": 0.00031013646169169326,
+      "loss": 3.0055,
+      "step": 29949
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.613553524017334,
+      "learning_rate": 0.0003101210742789242,
+      "loss": 2.9737,
+      "step": 29950
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5886765122413635,
+      "learning_rate": 0.00031010568683949817,
+      "loss": 2.8426,
+      "step": 29951
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6295921206474304,
+      "learning_rate": 0.00031009029937345567,
+      "loss": 3.2008,
+      "step": 29952
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5718304514884949,
+      "learning_rate": 0.0003100749118808373,
+      "loss": 3.2872,
+      "step": 29953
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5818148851394653,
+      "learning_rate": 0.0003100595243616836,
+      "loss": 3.315,
+      "step": 29954
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5743296146392822,
+      "learning_rate": 0.00031004413681603505,
+      "loss": 3.1359,
+      "step": 29955
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6252809166908264,
+      "learning_rate": 0.00031002874924393204,
+      "loss": 2.96,
+      "step": 29956
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5608223676681519,
+      "learning_rate": 0.00031001336164541537,
+      "loss": 3.1832,
+      "step": 29957
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5842495560646057,
+      "learning_rate": 0.00030999797402052543,
+      "loss": 3.2502,
+      "step": 29958
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.569668710231781,
+      "learning_rate": 0.00030998258636930266,
+      "loss": 3.2578,
+      "step": 29959
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5936859250068665,
+      "learning_rate": 0.00030996719869178776,
+      "loss": 3.2201,
+      "step": 29960
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5749362111091614,
+      "learning_rate": 0.00030995181098802114,
+      "loss": 3.1107,
+      "step": 29961
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6007623076438904,
+      "learning_rate": 0.00030993642325804333,
+      "loss": 3.0569,
+      "step": 29962
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6313218474388123,
+      "learning_rate": 0.000309921035501895,
+      "loss": 3.0859,
+      "step": 29963
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6418139934539795,
+      "learning_rate": 0.00030990564771961646,
+      "loss": 3.1744,
+      "step": 29964
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5875469446182251,
+      "learning_rate": 0.00030989025991124845,
+      "loss": 2.8325,
+      "step": 29965
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5943783521652222,
+      "learning_rate": 0.00030987487207683134,
+      "loss": 3.1794,
+      "step": 29966
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5599361658096313,
+      "learning_rate": 0.00030985948421640564,
+      "loss": 3.1811,
+      "step": 29967
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.605446994304657,
+      "learning_rate": 0.00030984409633001215,
+      "loss": 3.1107,
+      "step": 29968
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.609178900718689,
+      "learning_rate": 0.00030982870841769105,
+      "loss": 2.9717,
+      "step": 29969
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5545887351036072,
+      "learning_rate": 0.00030981332047948305,
+      "loss": 2.9905,
+      "step": 29970
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5839653611183167,
+      "learning_rate": 0.0003097979325154287,
+      "loss": 2.9023,
+      "step": 29971
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5799376964569092,
+      "learning_rate": 0.0003097825445255684,
+      "loss": 3.0182,
+      "step": 29972
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.560981273651123,
+      "learning_rate": 0.00030976715650994283,
+      "loss": 3.1762,
+      "step": 29973
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6042661666870117,
+      "learning_rate": 0.00030975176846859245,
+      "loss": 3.1572,
+      "step": 29974
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6044111847877502,
+      "learning_rate": 0.0003097363804015577,
+      "loss": 3.0491,
+      "step": 29975
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5796766877174377,
+      "learning_rate": 0.00030972099230887936,
+      "loss": 3.0295,
+      "step": 29976
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7335373163223267,
+      "learning_rate": 0.0003097056041905977,
+      "loss": 2.8727,
+      "step": 29977
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6482856273651123,
+      "learning_rate": 0.00030969021604675335,
+      "loss": 3.0558,
+      "step": 29978
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5912559628486633,
+      "learning_rate": 0.00030967482787738683,
+      "loss": 2.9217,
+      "step": 29979
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.577346920967102,
+      "learning_rate": 0.0003096594396825387,
+      "loss": 3.1919,
+      "step": 29980
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.586104154586792,
+      "learning_rate": 0.0003096440514622495,
+      "loss": 3.0131,
+      "step": 29981
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5916348695755005,
+      "learning_rate": 0.0003096286632165596,
+      "loss": 3.0654,
+      "step": 29982
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6343048810958862,
+      "learning_rate": 0.00030961327494550975,
+      "loss": 2.8386,
+      "step": 29983
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.60175621509552,
+      "learning_rate": 0.00030959788664914034,
+      "loss": 2.8354,
+      "step": 29984
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5782569050788879,
+      "learning_rate": 0.00030958249832749193,
+      "loss": 3.1649,
+      "step": 29985
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5811314582824707,
+      "learning_rate": 0.00030956710998060515,
+      "loss": 2.8642,
+      "step": 29986
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6273580193519592,
+      "learning_rate": 0.0003095517216085203,
+      "loss": 3.082,
+      "step": 29987
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.595939576625824,
+      "learning_rate": 0.0003095363332112782,
+      "loss": 3.0947,
+      "step": 29988
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6219014525413513,
+      "learning_rate": 0.00030952094478891917,
+      "loss": 2.9882,
+      "step": 29989
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6124245524406433,
+      "learning_rate": 0.0003095055563414839,
+      "loss": 2.9679,
+      "step": 29990
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5397445559501648,
+      "learning_rate": 0.0003094901678690127,
+      "loss": 3.1045,
+      "step": 29991
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6168830990791321,
+      "learning_rate": 0.0003094747793715462,
+      "loss": 3.0952,
+      "step": 29992
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6066709756851196,
+      "learning_rate": 0.000309459390849125,
+      "loss": 2.8169,
+      "step": 29993
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6061710119247437,
+      "learning_rate": 0.00030944400230178965,
+      "loss": 3.1075,
+      "step": 29994
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5859954953193665,
+      "learning_rate": 0.00030942861372958055,
+      "loss": 3.0391,
+      "step": 29995
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6048535704612732,
+      "learning_rate": 0.0003094132251325383,
+      "loss": 3.171,
+      "step": 29996
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6408707499504089,
+      "learning_rate": 0.00030939783651070347,
+      "loss": 3.0741,
+      "step": 29997
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5734691023826599,
+      "learning_rate": 0.0003093824478641165,
+      "loss": 2.9984,
+      "step": 29998
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5938532948493958,
+      "learning_rate": 0.00030936705919281797,
+      "loss": 2.9143,
+      "step": 29999
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5862475037574768,
+      "learning_rate": 0.00030935167049684845,
+      "loss": 2.9718,
+      "step": 30000
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5853474140167236,
+      "learning_rate": 0.00030933628177624845,
+      "loss": 2.987,
+      "step": 30001
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5849444270133972,
+      "learning_rate": 0.00030932089303105836,
+      "loss": 3.0514,
+      "step": 30002
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5930509567260742,
+      "learning_rate": 0.0003093055042613189,
+      "loss": 3.1043,
+      "step": 30003
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5608189702033997,
+      "learning_rate": 0.0003092901154670705,
+      "loss": 3.0665,
+      "step": 30004
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5968251824378967,
+      "learning_rate": 0.00030927472664835385,
+      "loss": 2.9385,
+      "step": 30005
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6607335805892944,
+      "learning_rate": 0.0003092593378052093,
+      "loss": 2.8233,
+      "step": 30006
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5770511031150818,
+      "learning_rate": 0.00030924394893767736,
+      "loss": 3.1056,
+      "step": 30007
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.671971321105957,
+      "learning_rate": 0.0003092285600457987,
+      "loss": 2.833,
+      "step": 30008
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5936998724937439,
+      "learning_rate": 0.0003092131711296137,
+      "loss": 3.1278,
+      "step": 30009
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6061517596244812,
+      "learning_rate": 0.00030919778218916304,
+      "loss": 3.2403,
+      "step": 30010
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6386778354644775,
+      "learning_rate": 0.00030918239322448723,
+      "loss": 3.0888,
+      "step": 30011
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5995616316795349,
+      "learning_rate": 0.00030916700423562677,
+      "loss": 2.953,
+      "step": 30012
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6046136021614075,
+      "learning_rate": 0.0003091516152226221,
+      "loss": 3.1343,
+      "step": 30013
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6061043739318848,
+      "learning_rate": 0.00030913622618551384,
+      "loss": 3.0659,
+      "step": 30014
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6259770393371582,
+      "learning_rate": 0.00030912083712434263,
+      "loss": 3.0636,
+      "step": 30015
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5781980156898499,
+      "learning_rate": 0.00030910544803914886,
+      "loss": 3.0077,
+      "step": 30016
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6106329560279846,
+      "learning_rate": 0.0003090900589299731,
+      "loss": 2.8722,
+      "step": 30017
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6144434213638306,
+      "learning_rate": 0.00030907466979685573,
+      "loss": 2.9317,
+      "step": 30018
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5561577081680298,
+      "learning_rate": 0.0003090592806398376,
+      "loss": 3.058,
+      "step": 30019
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5754068493843079,
+      "learning_rate": 0.00030904389145895895,
+      "loss": 2.7952,
+      "step": 30020
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.566835343837738,
+      "learning_rate": 0.00030902850225426046,
+      "loss": 3.3087,
+      "step": 30021
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6028580069541931,
+      "learning_rate": 0.00030901311302578265,
+      "loss": 2.8337,
+      "step": 30022
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6136353611946106,
+      "learning_rate": 0.000308997723773566,
+      "loss": 2.9233,
+      "step": 30023
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5877104997634888,
+      "learning_rate": 0.00030898233449765117,
+      "loss": 3.2272,
+      "step": 30024
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5864350199699402,
+      "learning_rate": 0.0003089669451980785,
+      "loss": 3.0262,
+      "step": 30025
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5969464182853699,
+      "learning_rate": 0.00030895155587488867,
+      "loss": 3.0745,
+      "step": 30026
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6056321859359741,
+      "learning_rate": 0.00030893616652812224,
+      "loss": 3.0233,
+      "step": 30027
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5771879553794861,
+      "learning_rate": 0.0003089207771578195,
+      "loss": 3.1142,
+      "step": 30028
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5768468976020813,
+      "learning_rate": 0.00030890538776402125,
+      "loss": 3.2566,
+      "step": 30029
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6096814870834351,
+      "learning_rate": 0.0003088899983467679,
+      "loss": 3.0353,
+      "step": 30030
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5729770660400391,
+      "learning_rate": 0.0003088746089061,
+      "loss": 3.0693,
+      "step": 30031
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6186831593513489,
+      "learning_rate": 0.0003088592194420581,
+      "loss": 3.2713,
+      "step": 30032
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5830119848251343,
+      "learning_rate": 0.00030884382995468267,
+      "loss": 3.2221,
+      "step": 30033
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5844249129295349,
+      "learning_rate": 0.00030882844044401446,
+      "loss": 2.9112,
+      "step": 30034
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6245279312133789,
+      "learning_rate": 0.0003088130509100937,
+      "loss": 3.1312,
+      "step": 30035
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6162157654762268,
+      "learning_rate": 0.00030879766135296103,
+      "loss": 3.0903,
+      "step": 30036
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6323968172073364,
+      "learning_rate": 0.00030878227177265707,
+      "loss": 3.0378,
+      "step": 30037
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5921278595924377,
+      "learning_rate": 0.0003087668821692223,
+      "loss": 3.1487,
+      "step": 30038
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5776242613792419,
+      "learning_rate": 0.00030875149254269723,
+      "loss": 2.8284,
+      "step": 30039
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5871045589447021,
+      "learning_rate": 0.0003087361028931224,
+      "loss": 3.0864,
+      "step": 30040
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6193739175796509,
+      "learning_rate": 0.0003087207132205384,
+      "loss": 3.0912,
+      "step": 30041
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5853241086006165,
+      "learning_rate": 0.00030870532352498577,
+      "loss": 3.095,
+      "step": 30042
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5713692903518677,
+      "learning_rate": 0.00030868993380650495,
+      "loss": 2.9933,
+      "step": 30043
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.589024543762207,
+      "learning_rate": 0.00030867454406513636,
+      "loss": 3.1075,
+      "step": 30044
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.638248085975647,
+      "learning_rate": 0.0003086591543009209,
+      "loss": 3.1879,
+      "step": 30045
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5836499333381653,
+      "learning_rate": 0.00030864376451389885,
+      "loss": 3.2182,
+      "step": 30046
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6031222343444824,
+      "learning_rate": 0.0003086283747041107,
+      "loss": 3.1655,
+      "step": 30047
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5775892734527588,
+      "learning_rate": 0.00030861298487159714,
+      "loss": 3.1287,
+      "step": 30048
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5872842669487,
+      "learning_rate": 0.00030859759501639865,
+      "loss": 2.8683,
+      "step": 30049
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6183463931083679,
+      "learning_rate": 0.00030858220513855567,
+      "loss": 3.0174,
+      "step": 30050
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.56663578748703,
+      "learning_rate": 0.0003085668152381088,
+      "loss": 3.0062,
+      "step": 30051
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.610784649848938,
+      "learning_rate": 0.0003085514253150988,
+      "loss": 3.0837,
+      "step": 30052
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5991683602333069,
+      "learning_rate": 0.0003085360353695658,
+      "loss": 3.1117,
+      "step": 30053
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5896287560462952,
+      "learning_rate": 0.0003085206454015505,
+      "loss": 3.2123,
+      "step": 30054
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5315856337547302,
+      "learning_rate": 0.00030850525541109347,
+      "loss": 3.1274,
+      "step": 30055
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.600836992263794,
+      "learning_rate": 0.0003084898653982354,
+      "loss": 3.0838,
+      "step": 30056
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6084719300270081,
+      "learning_rate": 0.00030847447536301657,
+      "loss": 2.7858,
+      "step": 30057
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.597867488861084,
+      "learning_rate": 0.0003084590853054775,
+      "loss": 3.0085,
+      "step": 30058
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6317042112350464,
+      "learning_rate": 0.000308443695225659,
+      "loss": 2.9331,
+      "step": 30059
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6137262582778931,
+      "learning_rate": 0.00030842830512360127,
+      "loss": 3.237,
+      "step": 30060
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6111647486686707,
+      "learning_rate": 0.0003084129149993451,
+      "loss": 3.0709,
+      "step": 30061
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6254605650901794,
+      "learning_rate": 0.0003083975248529309,
+      "loss": 3.1905,
+      "step": 30062
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5825523138046265,
+      "learning_rate": 0.00030838213468439927,
+      "loss": 3.2801,
+      "step": 30063
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6181331276893616,
+      "learning_rate": 0.00030836674449379064,
+      "loss": 2.8734,
+      "step": 30064
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6186519861221313,
+      "learning_rate": 0.0003083513542811456,
+      "loss": 3.1002,
+      "step": 30065
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5745614767074585,
+      "learning_rate": 0.00030833596404650476,
+      "loss": 3.0784,
+      "step": 30066
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5756679773330688,
+      "learning_rate": 0.0003083205737899087,
+      "loss": 2.988,
+      "step": 30067
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6421597003936768,
+      "learning_rate": 0.0003083051835113977,
+      "loss": 3.2194,
+      "step": 30068
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.610306978225708,
+      "learning_rate": 0.00030828979321101245,
+      "loss": 3.0658,
+      "step": 30069
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5943095088005066,
+      "learning_rate": 0.0003082744028887936,
+      "loss": 3.0469,
+      "step": 30070
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5513453483581543,
+      "learning_rate": 0.0003082590125447815,
+      "loss": 3.1213,
+      "step": 30071
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.577889084815979,
+      "learning_rate": 0.0003082436221790167,
+      "loss": 2.8904,
+      "step": 30072
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5584579706192017,
+      "learning_rate": 0.0003082282317915398,
+      "loss": 2.9921,
+      "step": 30073
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5914309620857239,
+      "learning_rate": 0.00030821284138239137,
+      "loss": 3.0308,
+      "step": 30074
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5481913089752197,
+      "learning_rate": 0.00030819745095161184,
+      "loss": 3.0113,
+      "step": 30075
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6211130619049072,
+      "learning_rate": 0.00030818206049924184,
+      "loss": 3.1022,
+      "step": 30076
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5855414271354675,
+      "learning_rate": 0.0003081666700253218,
+      "loss": 2.867,
+      "step": 30077
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.626336395740509,
+      "learning_rate": 0.00030815127952989245,
+      "loss": 3.2556,
+      "step": 30078
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5970550179481506,
+      "learning_rate": 0.00030813588901299404,
+      "loss": 3.2407,
+      "step": 30079
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6191651821136475,
+      "learning_rate": 0.00030812049847466736,
+      "loss": 3.1078,
+      "step": 30080
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6230210661888123,
+      "learning_rate": 0.00030810510791495294,
+      "loss": 3.3527,
+      "step": 30081
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6455768942832947,
+      "learning_rate": 0.00030808971733389107,
+      "loss": 2.8692,
+      "step": 30082
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6259360909461975,
+      "learning_rate": 0.00030807432673152257,
+      "loss": 3.0794,
+      "step": 30083
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5726274251937866,
+      "learning_rate": 0.00030805893610788776,
+      "loss": 3.0452,
+      "step": 30084
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5579006671905518,
+      "learning_rate": 0.00030804354546302726,
+      "loss": 2.9137,
+      "step": 30085
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.582209050655365,
+      "learning_rate": 0.0003080281547969816,
+      "loss": 2.8087,
+      "step": 30086
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5763527750968933,
+      "learning_rate": 0.00030801276410979137,
+      "loss": 3.0132,
+      "step": 30087
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.606497585773468,
+      "learning_rate": 0.00030799737340149707,
+      "loss": 2.8572,
+      "step": 30088
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6698089241981506,
+      "learning_rate": 0.0003079819826721392,
+      "loss": 2.9752,
+      "step": 30089
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6636598110198975,
+      "learning_rate": 0.00030796659192175826,
+      "loss": 3.3192,
+      "step": 30090
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5712068676948547,
+      "learning_rate": 0.0003079512011503949,
+      "loss": 2.9899,
+      "step": 30091
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5981823205947876,
+      "learning_rate": 0.00030793581035808967,
+      "loss": 3.0397,
+      "step": 30092
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5917591452598572,
+      "learning_rate": 0.0003079204195448831,
+      "loss": 3.2434,
+      "step": 30093
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.616977870464325,
+      "learning_rate": 0.00030790502871081555,
+      "loss": 3.1531,
+      "step": 30094
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5743931531906128,
+      "learning_rate": 0.0003078896378559276,
+      "loss": 2.8923,
+      "step": 30095
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6095641851425171,
+      "learning_rate": 0.00030787424698026,
+      "loss": 3.196,
+      "step": 30096
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5719312429428101,
+      "learning_rate": 0.0003078588560838531,
+      "loss": 2.9232,
+      "step": 30097
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5918893218040466,
+      "learning_rate": 0.00030784346516674743,
+      "loss": 3.1492,
+      "step": 30098
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6037421822547913,
+      "learning_rate": 0.00030782807422898364,
+      "loss": 2.899,
+      "step": 30099
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5701082944869995,
+      "learning_rate": 0.00030781268327060223,
+      "loss": 2.8979,
+      "step": 30100
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5683816075325012,
+      "learning_rate": 0.0003077972922916437,
+      "loss": 2.8652,
+      "step": 30101
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5692213177680969,
+      "learning_rate": 0.0003077819012921486,
+      "loss": 3.1262,
+      "step": 30102
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5966072082519531,
+      "learning_rate": 0.00030776651027215754,
+      "loss": 3.1198,
+      "step": 30103
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.59067302942276,
+      "learning_rate": 0.0003077511192317109,
+      "loss": 3.0482,
+      "step": 30104
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6178147196769714,
+      "learning_rate": 0.00030773572817084925,
+      "loss": 3.0714,
+      "step": 30105
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5838093161582947,
+      "learning_rate": 0.0003077203370896132,
+      "loss": 3.0316,
+      "step": 30106
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5537781715393066,
+      "learning_rate": 0.00030770494598804344,
+      "loss": 3.1259,
+      "step": 30107
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5612012147903442,
+      "learning_rate": 0.00030768955486618017,
+      "loss": 3.0059,
+      "step": 30108
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5880160927772522,
+      "learning_rate": 0.0003076741637240641,
+      "loss": 2.9501,
+      "step": 30109
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5850368142127991,
+      "learning_rate": 0.0003076587725617358,
+      "loss": 2.9588,
+      "step": 30110
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5933541655540466,
+      "learning_rate": 0.00030764338137923583,
+      "loss": 3.0509,
+      "step": 30111
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6294256448745728,
+      "learning_rate": 0.00030762799017660454,
+      "loss": 3.0801,
+      "step": 30112
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5729942917823792,
+      "learning_rate": 0.00030761259895388266,
+      "loss": 3.0384,
+      "step": 30113
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.575340986251831,
+      "learning_rate": 0.0003075972077111107,
+      "loss": 3.067,
+      "step": 30114
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5830142498016357,
+      "learning_rate": 0.000307581816448329,
+      "loss": 3.1674,
+      "step": 30115
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5919670462608337,
+      "learning_rate": 0.00030756642516557836,
+      "loss": 3.2883,
+      "step": 30116
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5591514110565186,
+      "learning_rate": 0.00030755103386289923,
+      "loss": 2.8813,
+      "step": 30117
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6111984252929688,
+      "learning_rate": 0.0003075356425403322,
+      "loss": 3.2448,
+      "step": 30118
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5747086405754089,
+      "learning_rate": 0.0003075202511979176,
+      "loss": 3.2162,
+      "step": 30119
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6217588782310486,
+      "learning_rate": 0.0003075048598356961,
+      "loss": 3.0186,
+      "step": 30120
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5869501829147339,
+      "learning_rate": 0.00030748946845370837,
+      "loss": 3.1229,
+      "step": 30121
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6082846522331238,
+      "learning_rate": 0.00030747407705199474,
+      "loss": 3.0859,
+      "step": 30122
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6041231751441956,
+      "learning_rate": 0.00030745868563059585,
+      "loss": 3.1107,
+      "step": 30123
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5768772959709167,
+      "learning_rate": 0.0003074432941895522,
+      "loss": 3.1982,
+      "step": 30124
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5537108182907104,
+      "learning_rate": 0.00030742790272890436,
+      "loss": 2.9797,
+      "step": 30125
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.582661509513855,
+      "learning_rate": 0.0003074125112486928,
+      "loss": 3.0316,
+      "step": 30126
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5950396656990051,
+      "learning_rate": 0.00030739711974895814,
+      "loss": 2.7202,
+      "step": 30127
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5830130577087402,
+      "learning_rate": 0.00030738172822974096,
+      "loss": 2.9846,
+      "step": 30128
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5635205507278442,
+      "learning_rate": 0.0003073663366910817,
+      "loss": 2.9789,
+      "step": 30129
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.577469527721405,
+      "learning_rate": 0.00030735094513302085,
+      "loss": 2.9118,
+      "step": 30130
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.603352963924408,
+      "learning_rate": 0.00030733555355559903,
+      "loss": 2.9981,
+      "step": 30131
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5949218273162842,
+      "learning_rate": 0.0003073201619588569,
+      "loss": 2.8709,
+      "step": 30132
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5829793214797974,
+      "learning_rate": 0.00030730477034283477,
+      "loss": 2.947,
+      "step": 30133
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5699021220207214,
+      "learning_rate": 0.0003072893787075733,
+      "loss": 3.1005,
+      "step": 30134
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5696962475776672,
+      "learning_rate": 0.000307273987053113,
+      "loss": 3.1134,
+      "step": 30135
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5930047631263733,
+      "learning_rate": 0.0003072585953794945,
+      "loss": 2.9946,
+      "step": 30136
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6077128052711487,
+      "learning_rate": 0.0003072432036867581,
+      "loss": 3.1896,
+      "step": 30137
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6007574796676636,
+      "learning_rate": 0.0003072278119749446,
+      "loss": 3.0898,
+      "step": 30138
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.613781750202179,
+      "learning_rate": 0.00030721242024409447,
+      "loss": 2.9471,
+      "step": 30139
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5998162031173706,
+      "learning_rate": 0.0003071970284942482,
+      "loss": 3.1317,
+      "step": 30140
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5879024863243103,
+      "learning_rate": 0.0003071816367254462,
+      "loss": 3.2085,
+      "step": 30141
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5877510905265808,
+      "learning_rate": 0.00030716624493772927,
+      "loss": 3.1028,
+      "step": 30142
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6133305430412292,
+      "learning_rate": 0.00030715085313113785,
+      "loss": 2.9439,
+      "step": 30143
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5921503901481628,
+      "learning_rate": 0.0003071354613057124,
+      "loss": 2.9244,
+      "step": 30144
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5787282586097717,
+      "learning_rate": 0.0003071200694614936,
+      "loss": 2.927,
+      "step": 30145
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5823509097099304,
+      "learning_rate": 0.00030710467759852177,
+      "loss": 3.1512,
+      "step": 30146
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6131330132484436,
+      "learning_rate": 0.0003070892857168377,
+      "loss": 3.0768,
+      "step": 30147
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6605939865112305,
+      "learning_rate": 0.00030707389381648176,
+      "loss": 3.1089,
+      "step": 30148
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5737425088882446,
+      "learning_rate": 0.0003070585018974946,
+      "loss": 3.2414,
+      "step": 30149
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5738173127174377,
+      "learning_rate": 0.00030704310995991667,
+      "loss": 3.024,
+      "step": 30150
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6123209595680237,
+      "learning_rate": 0.00030702771800378856,
+      "loss": 3.0543,
+      "step": 30151
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5468517541885376,
+      "learning_rate": 0.00030701232602915073,
+      "loss": 3.029,
+      "step": 30152
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5879467725753784,
+      "learning_rate": 0.0003069969340360438,
+      "loss": 3.1501,
+      "step": 30153
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.609234631061554,
+      "learning_rate": 0.0003069815420245084,
+      "loss": 3.1284,
+      "step": 30154
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5802608132362366,
+      "learning_rate": 0.00030696614999458494,
+      "loss": 3.3857,
+      "step": 30155
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5865260362625122,
+      "learning_rate": 0.00030695075794631385,
+      "loss": 2.9436,
+      "step": 30156
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5951496362686157,
+      "learning_rate": 0.00030693536587973584,
+      "loss": 3.2445,
+      "step": 30157
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.59162837266922,
+      "learning_rate": 0.0003069199737948915,
+      "loss": 3.0387,
+      "step": 30158
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.574717104434967,
+      "learning_rate": 0.0003069045816918212,
+      "loss": 2.8136,
+      "step": 30159
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6385117769241333,
+      "learning_rate": 0.00030688918957056565,
+      "loss": 2.971,
+      "step": 30160
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6385549902915955,
+      "learning_rate": 0.0003068737974311652,
+      "loss": 3.1969,
+      "step": 30161
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5969220399856567,
+      "learning_rate": 0.0003068584052736606,
+      "loss": 3.2271,
+      "step": 30162
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5814893841743469,
+      "learning_rate": 0.0003068430130980921,
+      "loss": 2.9764,
+      "step": 30163
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5654752850532532,
+      "learning_rate": 0.0003068276209045005,
+      "loss": 3.1884,
+      "step": 30164
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5874380469322205,
+      "learning_rate": 0.0003068122286929264,
+      "loss": 3.2886,
+      "step": 30165
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5981268286705017,
+      "learning_rate": 0.0003067968364634101,
+      "loss": 2.7295,
+      "step": 30166
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6147415041923523,
+      "learning_rate": 0.00030678144421599217,
+      "loss": 3.2141,
+      "step": 30167
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5960661768913269,
+      "learning_rate": 0.00030676605195071325,
+      "loss": 3.0085,
+      "step": 30168
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5469895601272583,
+      "learning_rate": 0.00030675065966761394,
+      "loss": 3.0473,
+      "step": 30169
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6334496736526489,
+      "learning_rate": 0.0003067352673667347,
+      "loss": 2.9559,
+      "step": 30170
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6375976800918579,
+      "learning_rate": 0.000306719875048116,
+      "loss": 3.0748,
+      "step": 30171
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5987905859947205,
+      "learning_rate": 0.0003067044827117984,
+      "loss": 2.7641,
+      "step": 30172
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6431529521942139,
+      "learning_rate": 0.00030668909035782253,
+      "loss": 2.9902,
+      "step": 30173
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5719122290611267,
+      "learning_rate": 0.00030667369798622886,
+      "loss": 3.1214,
+      "step": 30174
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5882480144500732,
+      "learning_rate": 0.00030665830559705804,
+      "loss": 3.3428,
+      "step": 30175
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5951686501502991,
+      "learning_rate": 0.00030664291319035047,
+      "loss": 3.058,
+      "step": 30176
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.594463050365448,
+      "learning_rate": 0.00030662752076614664,
+      "loss": 2.9867,
+      "step": 30177
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6419702768325806,
+      "learning_rate": 0.0003066121283244873,
+      "loss": 3.0913,
+      "step": 30178
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6004655361175537,
+      "learning_rate": 0.0003065967358654129,
+      "loss": 3.0112,
+      "step": 30179
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6011263728141785,
+      "learning_rate": 0.00030658134338896396,
+      "loss": 2.9962,
+      "step": 30180
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5436926484107971,
+      "learning_rate": 0.000306565950895181,
+      "loss": 3.1637,
+      "step": 30181
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6261427998542786,
+      "learning_rate": 0.0003065505583841045,
+      "loss": 2.9735,
+      "step": 30182
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6169971227645874,
+      "learning_rate": 0.00030653516585577527,
+      "loss": 2.9467,
+      "step": 30183
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6584175825119019,
+      "learning_rate": 0.0003065197733102336,
+      "loss": 3.1038,
+      "step": 30184
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5873148441314697,
+      "learning_rate": 0.00030650438074752013,
+      "loss": 2.959,
+      "step": 30185
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5871749520301819,
+      "learning_rate": 0.0003064889881676753,
+      "loss": 2.9619,
+      "step": 30186
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6506713628768921,
+      "learning_rate": 0.00030647359557073973,
+      "loss": 3.1184,
+      "step": 30187
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5763834118843079,
+      "learning_rate": 0.000306458202956754,
+      "loss": 3.2246,
+      "step": 30188
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5880481600761414,
+      "learning_rate": 0.00030644281032575857,
+      "loss": 2.8984,
+      "step": 30189
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5734893679618835,
+      "learning_rate": 0.00030642741767779406,
+      "loss": 3.1565,
+      "step": 30190
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6054985523223877,
+      "learning_rate": 0.000306412025012901,
+      "loss": 2.9848,
+      "step": 30191
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5837637186050415,
+      "learning_rate": 0.00030639663233111977,
+      "loss": 3.0006,
+      "step": 30192
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5740242004394531,
+      "learning_rate": 0.00030638123963249114,
+      "loss": 2.8531,
+      "step": 30193
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5803605318069458,
+      "learning_rate": 0.0003063658469170556,
+      "loss": 2.9359,
+      "step": 30194
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6074859499931335,
+      "learning_rate": 0.0003063504541848535,
+      "loss": 3.085,
+      "step": 30195
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5865538120269775,
+      "learning_rate": 0.00030633506143592566,
+      "loss": 3.0822,
+      "step": 30196
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.589298665523529,
+      "learning_rate": 0.0003063196686703123,
+      "loss": 3.2545,
+      "step": 30197
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5684748291969299,
+      "learning_rate": 0.0003063042758880544,
+      "loss": 2.933,
+      "step": 30198
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5615243911743164,
+      "learning_rate": 0.0003062888830891921,
+      "loss": 3.0079,
+      "step": 30199
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6056434512138367,
+      "learning_rate": 0.0003062734902737661,
+      "loss": 2.8702,
+      "step": 30200
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.61896151304245,
+      "learning_rate": 0.00030625809744181694,
+      "loss": 3.1829,
+      "step": 30201
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5565640330314636,
+      "learning_rate": 0.0003062427045933853,
+      "loss": 3.0956,
+      "step": 30202
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.634736180305481,
+      "learning_rate": 0.00030622731172851134,
+      "loss": 3.0891,
+      "step": 30203
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5769646763801575,
+      "learning_rate": 0.0003062119188472359,
+      "loss": 3.0662,
+      "step": 30204
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5608484148979187,
+      "learning_rate": 0.0003061965259495996,
+      "loss": 3.0664,
+      "step": 30205
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5996660590171814,
+      "learning_rate": 0.0003061811330356427,
+      "loss": 3.0626,
+      "step": 30206
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.606673538684845,
+      "learning_rate": 0.00030616574010540595,
+      "loss": 3.0254,
+      "step": 30207
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6172423958778381,
+      "learning_rate": 0.0003061503471589298,
+      "loss": 2.9156,
+      "step": 30208
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5741183757781982,
+      "learning_rate": 0.00030613495419625484,
+      "loss": 3.0138,
+      "step": 30209
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5955791473388672,
+      "learning_rate": 0.00030611956121742164,
+      "loss": 3.0918,
+      "step": 30210
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5800381898880005,
+      "learning_rate": 0.0003061041682224706,
+      "loss": 3.1426,
+      "step": 30211
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6998795866966248,
+      "learning_rate": 0.0003060887752114424,
+      "loss": 3.0431,
+      "step": 30212
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5997923016548157,
+      "learning_rate": 0.00030607338218437754,
+      "loss": 2.8662,
+      "step": 30213
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5872188210487366,
+      "learning_rate": 0.00030605798914131653,
+      "loss": 2.9839,
+      "step": 30214
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6395425200462341,
+      "learning_rate": 0.00030604259608229997,
+      "loss": 2.9624,
+      "step": 30215
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5830799341201782,
+      "learning_rate": 0.0003060272030073684,
+      "loss": 3.0724,
+      "step": 30216
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5782269835472107,
+      "learning_rate": 0.00030601180991656233,
+      "loss": 3.0847,
+      "step": 30217
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5734887719154358,
+      "learning_rate": 0.0003059964168099222,
+      "loss": 3.0501,
+      "step": 30218
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6264581680297852,
+      "learning_rate": 0.00030598102368748875,
+      "loss": 3.0401,
+      "step": 30219
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6159687042236328,
+      "learning_rate": 0.00030596563054930244,
+      "loss": 3.3187,
+      "step": 30220
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.601514458656311,
+      "learning_rate": 0.00030595023739540383,
+      "loss": 2.9992,
+      "step": 30221
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6012780070304871,
+      "learning_rate": 0.0003059348442258334,
+      "loss": 2.8195,
+      "step": 30222
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6135539412498474,
+      "learning_rate": 0.00030591945104063175,
+      "loss": 3.0922,
+      "step": 30223
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6214767098426819,
+      "learning_rate": 0.0003059040578398394,
+      "loss": 3.2023,
+      "step": 30224
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5887531042098999,
+      "learning_rate": 0.0003058886646234969,
+      "loss": 2.8366,
+      "step": 30225
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5788924694061279,
+      "learning_rate": 0.00030587327139164474,
+      "loss": 3.0441,
+      "step": 30226
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6139992475509644,
+      "learning_rate": 0.0003058578781443236,
+      "loss": 3.1177,
+      "step": 30227
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5603842735290527,
+      "learning_rate": 0.00030584248488157393,
+      "loss": 2.965,
+      "step": 30228
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5821728110313416,
+      "learning_rate": 0.00030582709160343615,
+      "loss": 2.9993,
+      "step": 30229
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5948326587677002,
+      "learning_rate": 0.00030581169830995106,
+      "loss": 2.8071,
+      "step": 30230
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6164664030075073,
+      "learning_rate": 0.0003057963050011591,
+      "loss": 3.2478,
+      "step": 30231
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6196146607398987,
+      "learning_rate": 0.00030578091167710075,
+      "loss": 2.7503,
+      "step": 30232
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6732343435287476,
+      "learning_rate": 0.00030576551833781646,
+      "loss": 3.0339,
+      "step": 30233
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5696006417274475,
+      "learning_rate": 0.00030575012498334705,
+      "loss": 2.8955,
+      "step": 30234
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5879353284835815,
+      "learning_rate": 0.0003057347316137329,
+      "loss": 3.0795,
+      "step": 30235
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5767134428024292,
+      "learning_rate": 0.0003057193382290146,
+      "loss": 2.9579,
+      "step": 30236
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5511177182197571,
+      "learning_rate": 0.00030570394482923254,
+      "loss": 3.0081,
+      "step": 30237
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5799663662910461,
+      "learning_rate": 0.0003056885514144276,
+      "loss": 3.1636,
+      "step": 30238
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.582383394241333,
+      "learning_rate": 0.00030567315798463987,
+      "loss": 2.8857,
+      "step": 30239
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.554225742816925,
+      "learning_rate": 0.00030565776453991025,
+      "loss": 2.9747,
+      "step": 30240
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6220203638076782,
+      "learning_rate": 0.00030564237108027915,
+      "loss": 2.9467,
+      "step": 30241
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5915405750274658,
+      "learning_rate": 0.0003056269776057872,
+      "loss": 2.7895,
+      "step": 30242
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5988735556602478,
+      "learning_rate": 0.0003056115841164748,
+      "loss": 2.9361,
+      "step": 30243
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6032905578613281,
+      "learning_rate": 0.0003055961906123825,
+      "loss": 3.1007,
+      "step": 30244
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6156014204025269,
+      "learning_rate": 0.00030558079709355106,
+      "loss": 2.9049,
+      "step": 30245
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6102397441864014,
+      "learning_rate": 0.0003055654035600208,
+      "loss": 2.9683,
+      "step": 30246
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6204525232315063,
+      "learning_rate": 0.0003055500100118323,
+      "loss": 3.0618,
+      "step": 30247
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6160491108894348,
+      "learning_rate": 0.0003055346164490262,
+      "loss": 3.1209,
+      "step": 30248
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5589972734451294,
+      "learning_rate": 0.000305519222871643,
+      "loss": 3.1358,
+      "step": 30249
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5976723432540894,
+      "learning_rate": 0.00030550382927972317,
+      "loss": 2.9208,
+      "step": 30250
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5512217283248901,
+      "learning_rate": 0.0003054884356733073,
+      "loss": 3.0062,
+      "step": 30251
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5562359690666199,
+      "learning_rate": 0.000305473042052436,
+      "loss": 3.0616,
+      "step": 30252
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6103971600532532,
+      "learning_rate": 0.0003054576484171498,
+      "loss": 2.8764,
+      "step": 30253
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5834851861000061,
+      "learning_rate": 0.00030544225476748907,
+      "loss": 3.1122,
+      "step": 30254
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5874277949333191,
+      "learning_rate": 0.0003054268611034946,
+      "loss": 3.0748,
+      "step": 30255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5750007629394531,
+      "learning_rate": 0.00030541146742520684,
+      "loss": 3.0289,
+      "step": 30256
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5605846047401428,
+      "learning_rate": 0.0003053960737326663,
+      "loss": 2.9117,
+      "step": 30257
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.577288031578064,
+      "learning_rate": 0.0003053806800259135,
+      "loss": 2.9049,
+      "step": 30258
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5720159411430359,
+      "learning_rate": 0.00030536528630498893,
+      "loss": 3.2772,
+      "step": 30259
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5890551805496216,
+      "learning_rate": 0.0003053498925699334,
+      "loss": 2.8408,
+      "step": 30260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6078507900238037,
+      "learning_rate": 0.00030533449882078723,
+      "loss": 3.1933,
+      "step": 30261
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5361073017120361,
+      "learning_rate": 0.000305319105057591,
+      "loss": 3.0317,
+      "step": 30262
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6214322447776794,
+      "learning_rate": 0.00030530371128038527,
+      "loss": 2.9402,
+      "step": 30263
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5904223322868347,
+      "learning_rate": 0.00030528831748921056,
+      "loss": 3.1021,
+      "step": 30264
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6872094869613647,
+      "learning_rate": 0.0003052729236841075,
+      "loss": 3.0022,
+      "step": 30265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6223731637001038,
+      "learning_rate": 0.0003052575298651165,
+      "loss": 2.7194,
+      "step": 30266
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.592032253742218,
+      "learning_rate": 0.00030524213603227827,
+      "loss": 3.336,
+      "step": 30267
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5875898599624634,
+      "learning_rate": 0.0003052267421856332,
+      "loss": 3.031,
+      "step": 30268
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6270338892936707,
+      "learning_rate": 0.00030521134832522186,
+      "loss": 2.9393,
+      "step": 30269
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5827121734619141,
+      "learning_rate": 0.00030519595445108487,
+      "loss": 3.0373,
+      "step": 30270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5908574461936951,
+      "learning_rate": 0.0003051805605632628,
+      "loss": 2.9094,
+      "step": 30271
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6578699946403503,
+      "learning_rate": 0.00030516516666179605,
+      "loss": 3.124,
+      "step": 30272
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6305190324783325,
+      "learning_rate": 0.0003051497727467253,
+      "loss": 3.0748,
+      "step": 30273
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5935174226760864,
+      "learning_rate": 0.000305134378818091,
+      "loss": 2.9313,
+      "step": 30274
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5942251682281494,
+      "learning_rate": 0.00030511898487593374,
+      "loss": 2.8511,
+      "step": 30275
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5874195694923401,
+      "learning_rate": 0.00030510359092029407,
+      "loss": 3.1529,
+      "step": 30276
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5620760917663574,
+      "learning_rate": 0.00030508819695121246,
+      "loss": 3.1332,
+      "step": 30277
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.573728084564209,
+      "learning_rate": 0.0003050728029687297,
+      "loss": 3.0645,
+      "step": 30278
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5932513475418091,
+      "learning_rate": 0.000305057408972886,
+      "loss": 3.1181,
+      "step": 30279
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6048024296760559,
+      "learning_rate": 0.00030504201496372205,
+      "loss": 3.0854,
+      "step": 30280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5609991550445557,
+      "learning_rate": 0.0003050266209412784,
+      "loss": 3.0611,
+      "step": 30281
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.598209023475647,
+      "learning_rate": 0.00030501122690559576,
+      "loss": 3.0283,
+      "step": 30282
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5940065979957581,
+      "learning_rate": 0.00030499583285671433,
+      "loss": 3.0705,
+      "step": 30283
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6010187268257141,
+      "learning_rate": 0.0003049804387946749,
+      "loss": 3.1678,
+      "step": 30284
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6073846220970154,
+      "learning_rate": 0.00030496504471951794,
+      "loss": 3.0441,
+      "step": 30285
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6591005325317383,
+      "learning_rate": 0.000304949650631284,
+      "loss": 2.9177,
+      "step": 30286
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6189901828765869,
+      "learning_rate": 0.0003049342565300137,
+      "loss": 3.0777,
+      "step": 30287
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6542283892631531,
+      "learning_rate": 0.00030491886241574743,
+      "loss": 2.9349,
+      "step": 30288
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5866391062736511,
+      "learning_rate": 0.00030490346828852594,
+      "loss": 3.0771,
+      "step": 30289
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6572050452232361,
+      "learning_rate": 0.00030488807414838953,
+      "loss": 3.0787,
+      "step": 30290
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6037701368331909,
+      "learning_rate": 0.00030487267999537894,
+      "loss": 3.1791,
+      "step": 30291
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6054988503456116,
+      "learning_rate": 0.0003048572858295346,
+      "loss": 2.9688,
+      "step": 30292
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6137593984603882,
+      "learning_rate": 0.00030484189165089714,
+      "loss": 3.1096,
+      "step": 30293
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5879693031311035,
+      "learning_rate": 0.00030482649745950714,
+      "loss": 3.1275,
+      "step": 30294
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.588402271270752,
+      "learning_rate": 0.00030481110325540487,
+      "loss": 3.0677,
+      "step": 30295
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5765928030014038,
+      "learning_rate": 0.00030479570903863126,
+      "loss": 3.2167,
+      "step": 30296
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6110025644302368,
+      "learning_rate": 0.0003047803148092267,
+      "loss": 3.1337,
+      "step": 30297
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6048384308815002,
+      "learning_rate": 0.00030476492056723156,
+      "loss": 2.8711,
+      "step": 30298
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6859794855117798,
+      "learning_rate": 0.0003047495263126867,
+      "loss": 3.0686,
+      "step": 30299
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6263118982315063,
+      "learning_rate": 0.0003047341320456324,
+      "loss": 3.052,
+      "step": 30300
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5739185214042664,
+      "learning_rate": 0.00030471873776610926,
+      "loss": 3.1724,
+      "step": 30301
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6050714254379272,
+      "learning_rate": 0.00030470334347415794,
+      "loss": 2.7828,
+      "step": 30302
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5985298156738281,
+      "learning_rate": 0.0003046879491698189,
+      "loss": 3.2557,
+      "step": 30303
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6600600481033325,
+      "learning_rate": 0.00030467255485313276,
+      "loss": 3.1668,
+      "step": 30304
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6146441698074341,
+      "learning_rate": 0.00030465716052413994,
+      "loss": 2.9902,
+      "step": 30305
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5583391785621643,
+      "learning_rate": 0.00030464176618288106,
+      "loss": 3.1554,
+      "step": 30306
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5961558222770691,
+      "learning_rate": 0.00030462637182939676,
+      "loss": 3.1391,
+      "step": 30307
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5696814656257629,
+      "learning_rate": 0.00030461097746372744,
+      "loss": 2.9831,
+      "step": 30308
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5873986482620239,
+      "learning_rate": 0.00030459558308591365,
+      "loss": 3.1334,
+      "step": 30309
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5988091230392456,
+      "learning_rate": 0.0003045801886959959,
+      "loss": 2.9959,
+      "step": 30310
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5809959173202515,
+      "learning_rate": 0.000304564794294015,
+      "loss": 3.0092,
+      "step": 30311
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.647441565990448,
+      "learning_rate": 0.00030454939988001114,
+      "loss": 3.0226,
+      "step": 30312
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5705923438072205,
+      "learning_rate": 0.0003045340054540251,
+      "loss": 3.0666,
+      "step": 30313
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5672298669815063,
+      "learning_rate": 0.00030451861101609746,
+      "loss": 2.9993,
+      "step": 30314
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6036275029182434,
+      "learning_rate": 0.0003045032165662686,
+      "loss": 3.2952,
+      "step": 30315
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5786696672439575,
+      "learning_rate": 0.00030448782210457906,
+      "loss": 3.0459,
+      "step": 30316
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5882992148399353,
+      "learning_rate": 0.0003044724276310695,
+      "loss": 3.1442,
+      "step": 30317
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5856705904006958,
+      "learning_rate": 0.0003044570331457805,
+      "loss": 3.0808,
+      "step": 30318
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5906365513801575,
+      "learning_rate": 0.0003044416386487525,
+      "loss": 3.0573,
+      "step": 30319
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.587424635887146,
+      "learning_rate": 0.00030442624414002607,
+      "loss": 2.9874,
+      "step": 30320
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5522072315216064,
+      "learning_rate": 0.00030441084961964164,
+      "loss": 3.0988,
+      "step": 30321
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.617243230342865,
+      "learning_rate": 0.0003043954550876401,
+      "loss": 2.9937,
+      "step": 30322
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5649192929267883,
+      "learning_rate": 0.00030438006054406166,
+      "loss": 3.0672,
+      "step": 30323
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5680512189865112,
+      "learning_rate": 0.000304364665988947,
+      "loss": 2.9709,
+      "step": 30324
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.560331404209137,
+      "learning_rate": 0.00030434927142233663,
+      "loss": 2.9678,
+      "step": 30325
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.652197539806366,
+      "learning_rate": 0.00030433387684427113,
+      "loss": 2.8437,
+      "step": 30326
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5792146921157837,
+      "learning_rate": 0.000304318482254791,
+      "loss": 3.1141,
+      "step": 30327
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.589543342590332,
+      "learning_rate": 0.0003043030876539368,
+      "loss": 3.0839,
+      "step": 30328
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.613621175289154,
+      "learning_rate": 0.0003042876930417493,
+      "loss": 3.0055,
+      "step": 30329
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5933493971824646,
+      "learning_rate": 0.00030427229841826863,
+      "loss": 2.9089,
+      "step": 30330
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5833031535148621,
+      "learning_rate": 0.00030425690378353557,
+      "loss": 3.0106,
+      "step": 30331
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6349642872810364,
+      "learning_rate": 0.00030424150913759067,
+      "loss": 3.0396,
+      "step": 30332
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6486304998397827,
+      "learning_rate": 0.0003042261144804745,
+      "loss": 3.0428,
+      "step": 30333
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6241731643676758,
+      "learning_rate": 0.0003042107198122275,
+      "loss": 3.0911,
+      "step": 30334
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6061176657676697,
+      "learning_rate": 0.0003041953251328903,
+      "loss": 2.8672,
+      "step": 30335
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5616305470466614,
+      "learning_rate": 0.00030417993044250344,
+      "loss": 3.0808,
+      "step": 30336
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5880496501922607,
+      "learning_rate": 0.00030416453574110734,
+      "loss": 3.1441,
+      "step": 30337
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5873252749443054,
+      "learning_rate": 0.0003041491410287428,
+      "loss": 3.2033,
+      "step": 30338
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5877861380577087,
+      "learning_rate": 0.0003041337463054501,
+      "loss": 3.1364,
+      "step": 30339
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6224821209907532,
+      "learning_rate": 0.00030411835157127007,
+      "loss": 3.01,
+      "step": 30340
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5856956839561462,
+      "learning_rate": 0.0003041029568262429,
+      "loss": 3.1922,
+      "step": 30341
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6419745683670044,
+      "learning_rate": 0.0003040875620704094,
+      "loss": 3.0602,
+      "step": 30342
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7789932489395142,
+      "learning_rate": 0.0003040721673038101,
+      "loss": 2.9055,
+      "step": 30343
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5722334384918213,
+      "learning_rate": 0.00030405677252648547,
+      "loss": 3.3164,
+      "step": 30344
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6217427849769592,
+      "learning_rate": 0.0003040413777384761,
+      "loss": 3.1159,
+      "step": 30345
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6086733937263489,
+      "learning_rate": 0.00030402598293982245,
+      "loss": 2.9625,
+      "step": 30346
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5802361965179443,
+      "learning_rate": 0.00030401058813056526,
+      "loss": 3.132,
+      "step": 30347
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.617330014705658,
+      "learning_rate": 0.00030399519331074485,
+      "loss": 2.8512,
+      "step": 30348
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5500723123550415,
+      "learning_rate": 0.00030397979848040187,
+      "loss": 3.1054,
+      "step": 30349
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5912415981292725,
+      "learning_rate": 0.00030396440363957683,
+      "loss": 2.6677,
+      "step": 30350
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5975785255432129,
+      "learning_rate": 0.0003039490087883104,
+      "loss": 3.1374,
+      "step": 30351
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6210824251174927,
+      "learning_rate": 0.000303933613926643,
+      "loss": 3.0865,
+      "step": 30352
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5919585824012756,
+      "learning_rate": 0.00030391821905461524,
+      "loss": 2.9867,
+      "step": 30353
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6664869785308838,
+      "learning_rate": 0.00030390282417226764,
+      "loss": 3.0154,
+      "step": 30354
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6210564970970154,
+      "learning_rate": 0.00030388742927964083,
+      "loss": 3.0155,
+      "step": 30355
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6251997947692871,
+      "learning_rate": 0.0003038720343767752,
+      "loss": 3.2516,
+      "step": 30356
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6180994510650635,
+      "learning_rate": 0.0003038566394637113,
+      "loss": 2.9035,
+      "step": 30357
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.590447187423706,
+      "learning_rate": 0.00030384124454048986,
+      "loss": 3.1161,
+      "step": 30358
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5581066608428955,
+      "learning_rate": 0.00030382584960715137,
+      "loss": 2.945,
+      "step": 30359
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6092578768730164,
+      "learning_rate": 0.0003038104546637362,
+      "loss": 3.1729,
+      "step": 30360
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5907177329063416,
+      "learning_rate": 0.00030379505971028504,
+      "loss": 2.8266,
+      "step": 30361
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5585633516311646,
+      "learning_rate": 0.0003037796647468385,
+      "loss": 3.0163,
+      "step": 30362
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5564131736755371,
+      "learning_rate": 0.000303764269773437,
+      "loss": 3.1012,
+      "step": 30363
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.578472375869751,
+      "learning_rate": 0.00030374887479012115,
+      "loss": 2.98,
+      "step": 30364
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5936715602874756,
+      "learning_rate": 0.00030373347979693145,
+      "loss": 3.014,
+      "step": 30365
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6830310821533203,
+      "learning_rate": 0.00030371808479390857,
+      "loss": 3.0181,
+      "step": 30366
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5550583600997925,
+      "learning_rate": 0.0003037026897810928,
+      "loss": 3.1111,
+      "step": 30367
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5920956134796143,
+      "learning_rate": 0.000303687294758525,
+      "loss": 3.0784,
+      "step": 30368
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6229501366615295,
+      "learning_rate": 0.00030367189972624564,
+      "loss": 3.1064,
+      "step": 30369
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6036462783813477,
+      "learning_rate": 0.0003036565046842951,
+      "loss": 2.8816,
+      "step": 30370
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5995955467224121,
+      "learning_rate": 0.00030364110963271404,
+      "loss": 3.0448,
+      "step": 30371
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5678886771202087,
+      "learning_rate": 0.00030362571457154296,
+      "loss": 3.1554,
+      "step": 30372
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5802719593048096,
+      "learning_rate": 0.0003036103195008225,
+      "loss": 3.1493,
+      "step": 30373
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5747050642967224,
+      "learning_rate": 0.00030359492442059315,
+      "loss": 3.1413,
+      "step": 30374
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6112601161003113,
+      "learning_rate": 0.0003035795293308954,
+      "loss": 3.0674,
+      "step": 30375
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5865578055381775,
+      "learning_rate": 0.0003035641342317699,
+      "loss": 2.955,
+      "step": 30376
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7048220038414001,
+      "learning_rate": 0.0003035487391232572,
+      "loss": 3.0602,
+      "step": 30377
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.585225522518158,
+      "learning_rate": 0.00030353334400539777,
+      "loss": 2.9674,
+      "step": 30378
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6342913508415222,
+      "learning_rate": 0.0003035179488782322,
+      "loss": 3.0666,
+      "step": 30379
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5397252440452576,
+      "learning_rate": 0.0003035025537418011,
+      "loss": 3.041,
+      "step": 30380
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5967416167259216,
+      "learning_rate": 0.0003034871585961448,
+      "loss": 3.2197,
+      "step": 30381
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5769542455673218,
+      "learning_rate": 0.000303471763441304,
+      "loss": 3.0792,
+      "step": 30382
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5737739205360413,
+      "learning_rate": 0.00030345636827731936,
+      "loss": 3.0399,
+      "step": 30383
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5969375967979431,
+      "learning_rate": 0.00030344097310423124,
+      "loss": 3.2229,
+      "step": 30384
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5473321676254272,
+      "learning_rate": 0.00030342557792208027,
+      "loss": 3.1632,
+      "step": 30385
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5815901756286621,
+      "learning_rate": 0.000303410182730907,
+      "loss": 3.0973,
+      "step": 30386
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5692906975746155,
+      "learning_rate": 0.00030339478753075194,
+      "loss": 2.9992,
+      "step": 30387
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5809762477874756,
+      "learning_rate": 0.0003033793923216557,
+      "loss": 3.1701,
+      "step": 30388
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5709482431411743,
+      "learning_rate": 0.00030336399710365877,
+      "loss": 3.102,
+      "step": 30389
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.578090250492096,
+      "learning_rate": 0.0003033486018768017,
+      "loss": 3.0336,
+      "step": 30390
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6145226359367371,
+      "learning_rate": 0.0003033332066411252,
+      "loss": 2.9575,
+      "step": 30391
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5879982113838196,
+      "learning_rate": 0.00030331781139666947,
+      "loss": 2.979,
+      "step": 30392
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5679128766059875,
+      "learning_rate": 0.0003033024161434753,
+      "loss": 3.4305,
+      "step": 30393
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6054519414901733,
+      "learning_rate": 0.00030328702088158326,
+      "loss": 3.1972,
+      "step": 30394
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5590806603431702,
+      "learning_rate": 0.0003032716256110339,
+      "loss": 3.0287,
+      "step": 30395
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.660683274269104,
+      "learning_rate": 0.00030325623033186763,
+      "loss": 3.0268,
+      "step": 30396
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5682603716850281,
+      "learning_rate": 0.0003032408350441251,
+      "loss": 3.1007,
+      "step": 30397
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5996838808059692,
+      "learning_rate": 0.00030322543974784676,
+      "loss": 2.8001,
+      "step": 30398
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5741966962814331,
+      "learning_rate": 0.0003032100444430733,
+      "loss": 3.041,
+      "step": 30399
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.604943573474884,
+      "learning_rate": 0.0003031946491298452,
+      "loss": 3.1721,
+      "step": 30400
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5835703015327454,
+      "learning_rate": 0.000303179253808203,
+      "loss": 2.8629,
+      "step": 30401
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5605635046958923,
+      "learning_rate": 0.0003031638584781873,
+      "loss": 2.9483,
+      "step": 30402
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5778949856758118,
+      "learning_rate": 0.0003031484631398385,
+      "loss": 2.8903,
+      "step": 30403
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5777247548103333,
+      "learning_rate": 0.0003031330677931973,
+      "loss": 3.0132,
+      "step": 30404
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6003771424293518,
+      "learning_rate": 0.0003031176724383042,
+      "loss": 2.9525,
+      "step": 30405
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6186116337776184,
+      "learning_rate": 0.00030310227707519983,
+      "loss": 2.8372,
+      "step": 30406
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7651210427284241,
+      "learning_rate": 0.0003030868817039246,
+      "loss": 3.118,
+      "step": 30407
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5799486637115479,
+      "learning_rate": 0.00030307148632451904,
+      "loss": 2.9309,
+      "step": 30408
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5795913934707642,
+      "learning_rate": 0.0003030560909370239,
+      "loss": 2.9225,
+      "step": 30409
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6140474081039429,
+      "learning_rate": 0.0003030406955414796,
+      "loss": 2.9041,
+      "step": 30410
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5724575519561768,
+      "learning_rate": 0.00030302530013792656,
+      "loss": 3.0642,
+      "step": 30411
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5742631554603577,
+      "learning_rate": 0.0003030099047264056,
+      "loss": 2.8497,
+      "step": 30412
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6531141996383667,
+      "learning_rate": 0.0003029945093069571,
+      "loss": 3.0945,
+      "step": 30413
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5915018916130066,
+      "learning_rate": 0.0003029791138796216,
+      "loss": 3.1048,
+      "step": 30414
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5942657589912415,
+      "learning_rate": 0.0003029637184444397,
+      "loss": 2.9497,
+      "step": 30415
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5846244096755981,
+      "learning_rate": 0.00030294832300145196,
+      "loss": 2.9391,
+      "step": 30416
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5968058109283447,
+      "learning_rate": 0.0003029329275506989,
+      "loss": 2.9962,
+      "step": 30417
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5921707153320312,
+      "learning_rate": 0.00030291753209222097,
+      "loss": 3.0783,
+      "step": 30418
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6028823852539062,
+      "learning_rate": 0.00030290213662605896,
+      "loss": 3.063,
+      "step": 30419
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.577496349811554,
+      "learning_rate": 0.00030288674115225327,
+      "loss": 3.1577,
+      "step": 30420
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5886926054954529,
+      "learning_rate": 0.0003028713456708444,
+      "loss": 3.0918,
+      "step": 30421
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6270083785057068,
+      "learning_rate": 0.000302855950181873,
+      "loss": 3.2039,
+      "step": 30422
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5588559508323669,
+      "learning_rate": 0.00030284055468537946,
+      "loss": 2.9458,
+      "step": 30423
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5802051424980164,
+      "learning_rate": 0.0003028251591814046,
+      "loss": 3.1478,
+      "step": 30424
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5683912038803101,
+      "learning_rate": 0.00030280976366998876,
+      "loss": 2.8673,
+      "step": 30425
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5958796143531799,
+      "learning_rate": 0.0003027943681511726,
+      "loss": 3.0611,
+      "step": 30426
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5795177817344666,
+      "learning_rate": 0.0003027789726249965,
+      "loss": 2.9482,
+      "step": 30427
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6204610466957092,
+      "learning_rate": 0.0003027635770915012,
+      "loss": 3.1761,
+      "step": 30428
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5817151069641113,
+      "learning_rate": 0.0003027481815507271,
+      "loss": 3.0828,
+      "step": 30429
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5747313499450684,
+      "learning_rate": 0.00030273278600271485,
+      "loss": 2.9711,
+      "step": 30430
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6377949714660645,
+      "learning_rate": 0.00030271739044750504,
+      "loss": 2.9626,
+      "step": 30431
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.599992036819458,
+      "learning_rate": 0.0003027019948851381,
+      "loss": 3.3038,
+      "step": 30432
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5743573307991028,
+      "learning_rate": 0.00030268659931565465,
+      "loss": 3.0086,
+      "step": 30433
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5959254503250122,
+      "learning_rate": 0.00030267120373909506,
+      "loss": 3.1205,
+      "step": 30434
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5806088447570801,
+      "learning_rate": 0.00030265580815550024,
+      "loss": 2.9793,
+      "step": 30435
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6778531074523926,
+      "learning_rate": 0.0003026404125649105,
+      "loss": 2.9428,
+      "step": 30436
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5825092792510986,
+      "learning_rate": 0.0003026250169673663,
+      "loss": 2.9743,
+      "step": 30437
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5714017748832703,
+      "learning_rate": 0.00030260962136290835,
+      "loss": 2.9662,
+      "step": 30438
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5609602928161621,
+      "learning_rate": 0.0003025942257515772,
+      "loss": 2.913,
+      "step": 30439
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5940423011779785,
+      "learning_rate": 0.00030257883013341336,
+      "loss": 3.082,
+      "step": 30440
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5933805704116821,
+      "learning_rate": 0.0003025634345084573,
+      "loss": 3.3069,
+      "step": 30441
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6016649603843689,
+      "learning_rate": 0.0003025480388767498,
+      "loss": 3.1425,
+      "step": 30442
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5964870452880859,
+      "learning_rate": 0.0003025326432383312,
+      "loss": 3.0939,
+      "step": 30443
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6471694111824036,
+      "learning_rate": 0.0003025172475932419,
+      "loss": 3.2347,
+      "step": 30444
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5884523987770081,
+      "learning_rate": 0.00030250185194152286,
+      "loss": 2.8428,
+      "step": 30445
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6636972427368164,
+      "learning_rate": 0.0003024864562832144,
+      "loss": 3.0002,
+      "step": 30446
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6032358407974243,
+      "learning_rate": 0.00030247106061835713,
+      "loss": 3.0426,
+      "step": 30447
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6321635842323303,
+      "learning_rate": 0.0003024556649469915,
+      "loss": 2.8345,
+      "step": 30448
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5757449865341187,
+      "learning_rate": 0.00030244026926915814,
+      "loss": 2.9545,
+      "step": 30449
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5811001658439636,
+      "learning_rate": 0.00030242487358489754,
+      "loss": 3.0403,
+      "step": 30450
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.582600474357605,
+      "learning_rate": 0.00030240947789425033,
+      "loss": 2.9595,
+      "step": 30451
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6360583305358887,
+      "learning_rate": 0.000302394082197257,
+      "loss": 2.8013,
+      "step": 30452
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6164677739143372,
+      "learning_rate": 0.00030237868649395815,
+      "loss": 3.1337,
+      "step": 30453
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5845510363578796,
+      "learning_rate": 0.0003023632907843942,
+      "loss": 3.1204,
+      "step": 30454
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5880382657051086,
+      "learning_rate": 0.00030234789506860594,
+      "loss": 3.1311,
+      "step": 30455
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6087186932563782,
+      "learning_rate": 0.0003023324993466337,
+      "loss": 3.1319,
+      "step": 30456
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6022242307662964,
+      "learning_rate": 0.00030231710361851814,
+      "loss": 3.0978,
+      "step": 30457
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5847160816192627,
+      "learning_rate": 0.0003023017078842998,
+      "loss": 2.8099,
+      "step": 30458
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5767192840576172,
+      "learning_rate": 0.00030228631214401905,
+      "loss": 3.2496,
+      "step": 30459
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5725535154342651,
+      "learning_rate": 0.00030227091639771676,
+      "loss": 3.0633,
+      "step": 30460
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6081965565681458,
+      "learning_rate": 0.00030225552064543326,
+      "loss": 3.1437,
+      "step": 30461
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6389308571815491,
+      "learning_rate": 0.0003022401248872091,
+      "loss": 3.1606,
+      "step": 30462
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5722255110740662,
+      "learning_rate": 0.0003022247291230849,
+      "loss": 3.1019,
+      "step": 30463
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6214210987091064,
+      "learning_rate": 0.00030220933335310126,
+      "loss": 3.0384,
+      "step": 30464
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5633954405784607,
+      "learning_rate": 0.0003021939375772986,
+      "loss": 2.9655,
+      "step": 30465
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.583755373954773,
+      "learning_rate": 0.00030217854179571756,
+      "loss": 3.0302,
+      "step": 30466
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5567439794540405,
+      "learning_rate": 0.00030216314600839866,
+      "loss": 3.1762,
+      "step": 30467
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6298270225524902,
+      "learning_rate": 0.00030214775021538244,
+      "loss": 2.9571,
+      "step": 30468
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5982662439346313,
+      "learning_rate": 0.00030213235441670945,
+      "loss": 2.9339,
+      "step": 30469
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5737664103507996,
+      "learning_rate": 0.00030211695861242024,
+      "loss": 3.0476,
+      "step": 30470
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5396885275840759,
+      "learning_rate": 0.00030210156280255547,
+      "loss": 3.04,
+      "step": 30471
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6123176217079163,
+      "learning_rate": 0.00030208616698715546,
+      "loss": 2.9829,
+      "step": 30472
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6010289788246155,
+      "learning_rate": 0.00030207077116626093,
+      "loss": 3.0169,
+      "step": 30473
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5837356448173523,
+      "learning_rate": 0.0003020553753399124,
+      "loss": 3.0827,
+      "step": 30474
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6117029190063477,
+      "learning_rate": 0.0003020399795081504,
+      "loss": 2.9524,
+      "step": 30475
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.595624566078186,
+      "learning_rate": 0.0003020245836710155,
+      "loss": 3.2267,
+      "step": 30476
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5812867879867554,
+      "learning_rate": 0.0003020091878285482,
+      "loss": 3.0827,
+      "step": 30477
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6493427157402039,
+      "learning_rate": 0.0003019937919807891,
+      "loss": 3.1577,
+      "step": 30478
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5776815414428711,
+      "learning_rate": 0.0003019783961277788,
+      "loss": 2.8263,
+      "step": 30479
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5874110460281372,
+      "learning_rate": 0.00030196300026955763,
+      "loss": 3.2444,
+      "step": 30480
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5695233345031738,
+      "learning_rate": 0.0003019476044061664,
+      "loss": 2.9263,
+      "step": 30481
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5816549062728882,
+      "learning_rate": 0.0003019322085376456,
+      "loss": 3.0367,
+      "step": 30482
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5536554455757141,
+      "learning_rate": 0.0003019168126640357,
+      "loss": 2.8394,
+      "step": 30483
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5806564688682556,
+      "learning_rate": 0.00030190141678537724,
+      "loss": 3.1847,
+      "step": 30484
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5865193605422974,
+      "learning_rate": 0.00030188602090171077,
+      "loss": 3.1498,
+      "step": 30485
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5693684220314026,
+      "learning_rate": 0.000301870625013077,
+      "loss": 3.0875,
+      "step": 30486
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6434977650642395,
+      "learning_rate": 0.00030185522911951634,
+      "loss": 2.9599,
+      "step": 30487
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6276722550392151,
+      "learning_rate": 0.0003018398332210693,
+      "loss": 3.3054,
+      "step": 30488
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6278494000434875,
+      "learning_rate": 0.0003018244373177765,
+      "loss": 3.1279,
+      "step": 30489
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.639779806137085,
+      "learning_rate": 0.0003018090414096786,
+      "loss": 2.9098,
+      "step": 30490
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.58705735206604,
+      "learning_rate": 0.00030179364549681595,
+      "loss": 3.1877,
+      "step": 30491
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6035036444664001,
+      "learning_rate": 0.00030177824957922913,
+      "loss": 3.1528,
+      "step": 30492
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6042265295982361,
+      "learning_rate": 0.00030176285365695883,
+      "loss": 2.8595,
+      "step": 30493
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5860368013381958,
+      "learning_rate": 0.0003017474577300455,
+      "loss": 3.2413,
+      "step": 30494
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.580884575843811,
+      "learning_rate": 0.00030173206179852965,
+      "loss": 2.9413,
+      "step": 30495
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6143849492073059,
+      "learning_rate": 0.00030171666586245187,
+      "loss": 2.9255,
+      "step": 30496
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5543144941329956,
+      "learning_rate": 0.00030170126992185286,
+      "loss": 3.1471,
+      "step": 30497
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5888075232505798,
+      "learning_rate": 0.0003016858739767729,
+      "loss": 2.8926,
+      "step": 30498
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6028613448143005,
+      "learning_rate": 0.0003016704780272527,
+      "loss": 3.1541,
+      "step": 30499
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5923327207565308,
+      "learning_rate": 0.00030165508207333284,
+      "loss": 2.9867,
+      "step": 30500
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5775189399719238,
+      "learning_rate": 0.00030163968611505375,
+      "loss": 3.1233,
+      "step": 30501
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6112174391746521,
+      "learning_rate": 0.000301624290152456,
+      "loss": 3.0625,
+      "step": 30502
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.562967836856842,
+      "learning_rate": 0.0003016088941855803,
+      "loss": 2.9785,
+      "step": 30503
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6123626828193665,
+      "learning_rate": 0.00030159349821446703,
+      "loss": 2.9996,
+      "step": 30504
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.578737199306488,
+      "learning_rate": 0.00030157810223915675,
+      "loss": 2.9256,
+      "step": 30505
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6503991484642029,
+      "learning_rate": 0.00030156270625969006,
+      "loss": 3.0334,
+      "step": 30506
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5757296085357666,
+      "learning_rate": 0.00030154731027610753,
+      "loss": 2.9777,
+      "step": 30507
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5990521907806396,
+      "learning_rate": 0.00030153191428844976,
+      "loss": 3.0282,
+      "step": 30508
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5953813791275024,
+      "learning_rate": 0.0003015165182967572,
+      "loss": 3.1037,
+      "step": 30509
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5794540047645569,
+      "learning_rate": 0.00030150112230107023,
+      "loss": 3.0526,
+      "step": 30510
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.564243495464325,
+      "learning_rate": 0.0003014857263014298,
+      "loss": 2.8519,
+      "step": 30511
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5954510569572449,
+      "learning_rate": 0.0003014703302978762,
+      "loss": 3.1894,
+      "step": 30512
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6306339502334595,
+      "learning_rate": 0.00030145493429044995,
+      "loss": 3.0282,
+      "step": 30513
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5806563496589661,
+      "learning_rate": 0.00030143953827919175,
+      "loss": 2.8584,
+      "step": 30514
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6290732026100159,
+      "learning_rate": 0.00030142414226414215,
+      "loss": 3.2384,
+      "step": 30515
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5588606595993042,
+      "learning_rate": 0.00030140874624534153,
+      "loss": 2.962,
+      "step": 30516
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5796079635620117,
+      "learning_rate": 0.00030139335022283056,
+      "loss": 2.8573,
+      "step": 30517
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5561217069625854,
+      "learning_rate": 0.0003013779541966498,
+      "loss": 2.8819,
+      "step": 30518
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5866872072219849,
+      "learning_rate": 0.00030136255816683985,
+      "loss": 3.0356,
+      "step": 30519
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5967084765434265,
+      "learning_rate": 0.0003013471621334411,
+      "loss": 2.724,
+      "step": 30520
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5654155611991882,
+      "learning_rate": 0.0003013317660964941,
+      "loss": 3.045,
+      "step": 30521
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5480029582977295,
+      "learning_rate": 0.0003013163700560396,
+      "loss": 3.0384,
+      "step": 30522
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5529598593711853,
+      "learning_rate": 0.000301300974012118,
+      "loss": 2.8905,
+      "step": 30523
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6232104301452637,
+      "learning_rate": 0.0003012855779647699,
+      "loss": 3.1526,
+      "step": 30524
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5751103758811951,
+      "learning_rate": 0.0003012701819140358,
+      "loss": 3.1856,
+      "step": 30525
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5563510060310364,
+      "learning_rate": 0.0003012547858599563,
+      "loss": 2.8839,
+      "step": 30526
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.597868800163269,
+      "learning_rate": 0.00030123938980257197,
+      "loss": 2.8558,
+      "step": 30527
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5711178779602051,
+      "learning_rate": 0.00030122399374192326,
+      "loss": 3.1168,
+      "step": 30528
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5784346461296082,
+      "learning_rate": 0.0003012085976780509,
+      "loss": 3.1221,
+      "step": 30529
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5460003614425659,
+      "learning_rate": 0.0003011932016109953,
+      "loss": 2.9339,
+      "step": 30530
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5536607503890991,
+      "learning_rate": 0.0003011778055407969,
+      "loss": 2.7398,
+      "step": 30531
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5944807529449463,
+      "learning_rate": 0.0003011624094674964,
+      "loss": 2.9107,
+      "step": 30532
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5808088183403015,
+      "learning_rate": 0.00030114701339113454,
+      "loss": 3.1568,
+      "step": 30533
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6210165619850159,
+      "learning_rate": 0.0003011316173117515,
+      "loss": 3.0295,
+      "step": 30534
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5955290794372559,
+      "learning_rate": 0.00030111622122938803,
+      "loss": 3.1209,
+      "step": 30535
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.608069121837616,
+      "learning_rate": 0.0003011008251440846,
+      "loss": 3.093,
+      "step": 30536
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5967391133308411,
+      "learning_rate": 0.00030108542905588195,
+      "loss": 3.195,
+      "step": 30537
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5866439938545227,
+      "learning_rate": 0.0003010700329648204,
+      "loss": 3.0273,
+      "step": 30538
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6006977558135986,
+      "learning_rate": 0.0003010546368709406,
+      "loss": 3.0791,
+      "step": 30539
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5997205972671509,
+      "learning_rate": 0.000301039240774283,
+      "loss": 2.8495,
+      "step": 30540
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5799188613891602,
+      "learning_rate": 0.0003010238446748884,
+      "loss": 3.0779,
+      "step": 30541
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6030257940292358,
+      "learning_rate": 0.0003010084485727971,
+      "loss": 2.7769,
+      "step": 30542
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6359684467315674,
+      "learning_rate": 0.0003009930524680497,
+      "loss": 3.0085,
+      "step": 30543
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5771874189376831,
+      "learning_rate": 0.0003009776563606869,
+      "loss": 2.8918,
+      "step": 30544
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5872456431388855,
+      "learning_rate": 0.0003009622602507491,
+      "loss": 3.0065,
+      "step": 30545
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6064155697822571,
+      "learning_rate": 0.00030094686413827685,
+      "loss": 3.1301,
+      "step": 30546
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6292130351066589,
+      "learning_rate": 0.0003009314680233107,
+      "loss": 3.0003,
+      "step": 30547
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6809849739074707,
+      "learning_rate": 0.00030091607190589143,
+      "loss": 3.2596,
+      "step": 30548
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6155319213867188,
+      "learning_rate": 0.00030090067578605927,
+      "loss": 2.9799,
+      "step": 30549
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5968411564826965,
+      "learning_rate": 0.00030088527966385493,
+      "loss": 3.0207,
+      "step": 30550
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5945718288421631,
+      "learning_rate": 0.0003008698835393189,
+      "loss": 3.0553,
+      "step": 30551
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.654512882232666,
+      "learning_rate": 0.0003008544874124918,
+      "loss": 3.2307,
+      "step": 30552
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5625406503677368,
+      "learning_rate": 0.0003008390912834141,
+      "loss": 3.1419,
+      "step": 30553
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6100373864173889,
+      "learning_rate": 0.00030082369515212643,
+      "loss": 3.0421,
+      "step": 30554
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6178447008132935,
+      "learning_rate": 0.0003008082990186694,
+      "loss": 3.1594,
+      "step": 30555
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5651780962944031,
+      "learning_rate": 0.0003007929028830833,
+      "loss": 3.0438,
+      "step": 30556
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5665957927703857,
+      "learning_rate": 0.00030077750674540883,
+      "loss": 2.9222,
+      "step": 30557
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5908525586128235,
+      "learning_rate": 0.00030076211060568666,
+      "loss": 2.9867,
+      "step": 30558
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5936393141746521,
+      "learning_rate": 0.00030074671446395725,
+      "loss": 2.962,
+      "step": 30559
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6061809659004211,
+      "learning_rate": 0.0003007313183202611,
+      "loss": 3.1466,
+      "step": 30560
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5717028975486755,
+      "learning_rate": 0.00030071592217463885,
+      "loss": 2.985,
+      "step": 30561
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.596230149269104,
+      "learning_rate": 0.00030070052602713097,
+      "loss": 3.0271,
+      "step": 30562
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6109982132911682,
+      "learning_rate": 0.000300685129877778,
+      "loss": 3.0328,
+      "step": 30563
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5894909501075745,
+      "learning_rate": 0.0003006697337266206,
+      "loss": 2.791,
+      "step": 30564
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5837761163711548,
+      "learning_rate": 0.0003006543375736991,
+      "loss": 2.8944,
+      "step": 30565
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5935973525047302,
+      "learning_rate": 0.0003006389414190544,
+      "loss": 2.9046,
+      "step": 30566
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6078287959098816,
+      "learning_rate": 0.0003006235452627267,
+      "loss": 3.2071,
+      "step": 30567
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6492429971694946,
+      "learning_rate": 0.0003006081491047567,
+      "loss": 3.0934,
+      "step": 30568
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5882315635681152,
+      "learning_rate": 0.00030059275294518507,
+      "loss": 2.9978,
+      "step": 30569
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6093411445617676,
+      "learning_rate": 0.0003005773567840522,
+      "loss": 3.0642,
+      "step": 30570
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5847349762916565,
+      "learning_rate": 0.0003005619606213987,
+      "loss": 2.9372,
+      "step": 30571
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5897664427757263,
+      "learning_rate": 0.00030054656445726495,
+      "loss": 3.059,
+      "step": 30572
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5819639563560486,
+      "learning_rate": 0.0003005311682916918,
+      "loss": 2.9334,
+      "step": 30573
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5720083713531494,
+      "learning_rate": 0.00030051577212471964,
+      "loss": 2.8956,
+      "step": 30574
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6046839356422424,
+      "learning_rate": 0.00030050037595638905,
+      "loss": 3.0251,
+      "step": 30575
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.618550717830658,
+      "learning_rate": 0.00030048497978674057,
+      "loss": 2.8597,
+      "step": 30576
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5971773266792297,
+      "learning_rate": 0.0003004695836158147,
+      "loss": 2.9743,
+      "step": 30577
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6202574968338013,
+      "learning_rate": 0.00030045418744365204,
+      "loss": 3.063,
+      "step": 30578
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6273716688156128,
+      "learning_rate": 0.00030043879127029314,
+      "loss": 3.1348,
+      "step": 30579
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6152206063270569,
+      "learning_rate": 0.0003004233950957785,
+      "loss": 3.1494,
+      "step": 30580
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5668372511863708,
+      "learning_rate": 0.0003004079989201489,
+      "loss": 3.1972,
+      "step": 30581
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5945584177970886,
+      "learning_rate": 0.0003003926027434445,
+      "loss": 2.9756,
+      "step": 30582
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.68867427110672,
+      "learning_rate": 0.0003003772065657061,
+      "loss": 2.9829,
+      "step": 30583
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6505774259567261,
+      "learning_rate": 0.00030036181038697434,
+      "loss": 3.1015,
+      "step": 30584
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5589194297790527,
+      "learning_rate": 0.00030034641420728956,
+      "loss": 2.9296,
+      "step": 30585
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5537344217300415,
+      "learning_rate": 0.0003003310180266924,
+      "loss": 2.9285,
+      "step": 30586
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6337488889694214,
+      "learning_rate": 0.00030031562184522326,
+      "loss": 3.0206,
+      "step": 30587
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6181750297546387,
+      "learning_rate": 0.00030030022566292306,
+      "loss": 2.8851,
+      "step": 30588
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5827994346618652,
+      "learning_rate": 0.00030028482947983204,
+      "loss": 3.1003,
+      "step": 30589
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6145060658454895,
+      "learning_rate": 0.0003002694332959908,
+      "loss": 3.0419,
+      "step": 30590
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6019697189331055,
+      "learning_rate": 0.00030025403711143997,
+      "loss": 2.9749,
+      "step": 30591
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.625700056552887,
+      "learning_rate": 0.0003002386409262201,
+      "loss": 3.0995,
+      "step": 30592
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5388084650039673,
+      "learning_rate": 0.00030022324474037154,
+      "loss": 2.9934,
+      "step": 30593
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.626770555973053,
+      "learning_rate": 0.0003002078485539351,
+      "loss": 3.071,
+      "step": 30594
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5812084078788757,
+      "learning_rate": 0.00030019245236695126,
+      "loss": 2.9377,
+      "step": 30595
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6054795384407043,
+      "learning_rate": 0.0003001770561794605,
+      "loss": 2.7962,
+      "step": 30596
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5243318676948547,
+      "learning_rate": 0.00030016165999150336,
+      "loss": 3.0612,
+      "step": 30597
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.562089741230011,
+      "learning_rate": 0.0003001462638031204,
+      "loss": 3.0272,
+      "step": 30598
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5881873369216919,
+      "learning_rate": 0.0003001308676143524,
+      "loss": 3.2251,
+      "step": 30599
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5781797170639038,
+      "learning_rate": 0.00030011547142523955,
+      "loss": 2.8792,
+      "step": 30600
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6735518574714661,
+      "learning_rate": 0.0003001000752358226,
+      "loss": 3.1419,
+      "step": 30601
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5663983821868896,
+      "learning_rate": 0.0003000846790461421,
+      "loss": 2.9512,
+      "step": 30602
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6189948320388794,
+      "learning_rate": 0.00030006928285623857,
+      "loss": 2.9013,
+      "step": 30603
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6442484855651855,
+      "learning_rate": 0.00030005388666615255,
+      "loss": 3.1847,
+      "step": 30604
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.629422128200531,
+      "learning_rate": 0.0003000384904759246,
+      "loss": 3.0761,
+      "step": 30605
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5894017219543457,
+      "learning_rate": 0.0003000230942855954,
+      "loss": 3.1667,
+      "step": 30606
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5657310485839844,
+      "learning_rate": 0.00030000769809520526,
+      "loss": 3.1474,
+      "step": 30607
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.664206862449646,
+      "learning_rate": 0.0002999923019047948,
+      "loss": 2.9338,
+      "step": 30608
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5959284901618958,
+      "learning_rate": 0.0002999769057144046,
+      "loss": 3.0428,
+      "step": 30609
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5982975959777832,
+      "learning_rate": 0.00029996150952407534,
+      "loss": 2.9871,
+      "step": 30610
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6166161894798279,
+      "learning_rate": 0.00029994611333384745,
+      "loss": 3.0232,
+      "step": 30611
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5803634524345398,
+      "learning_rate": 0.0002999307171437613,
+      "loss": 3.2022,
+      "step": 30612
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5776461958885193,
+      "learning_rate": 0.0002999153209538579,
+      "loss": 2.9408,
+      "step": 30613
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5784315466880798,
+      "learning_rate": 0.00029989992476417735,
+      "loss": 3.2184,
+      "step": 30614
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.635517418384552,
+      "learning_rate": 0.00029988452857476045,
+      "loss": 3.1963,
+      "step": 30615
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6005640625953674,
+      "learning_rate": 0.0002998691323856476,
+      "loss": 3.0256,
+      "step": 30616
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5672207474708557,
+      "learning_rate": 0.0002998537361968795,
+      "loss": 3.0099,
+      "step": 30617
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6029596328735352,
+      "learning_rate": 0.00029983834000849664,
+      "loss": 3.125,
+      "step": 30618
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5798458456993103,
+      "learning_rate": 0.0002998229438205395,
+      "loss": 3.0502,
+      "step": 30619
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5979757308959961,
+      "learning_rate": 0.0002998075476330488,
+      "loss": 3.0172,
+      "step": 30620
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6162039041519165,
+      "learning_rate": 0.0002997921514460649,
+      "loss": 3.0594,
+      "step": 30621
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5752620100975037,
+      "learning_rate": 0.00029977675525962835,
+      "loss": 3.0725,
+      "step": 30622
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6150903701782227,
+      "learning_rate": 0.00029976135907377994,
+      "loss": 2.9637,
+      "step": 30623
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.607782781124115,
+      "learning_rate": 0.00029974596288856003,
+      "loss": 2.8398,
+      "step": 30624
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5732407569885254,
+      "learning_rate": 0.00029973056670400925,
+      "loss": 2.9802,
+      "step": 30625
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5856612324714661,
+      "learning_rate": 0.0002997151705201679,
+      "loss": 2.9968,
+      "step": 30626
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5715699791908264,
+      "learning_rate": 0.00029969977433707683,
+      "loss": 3.0805,
+      "step": 30627
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6128286123275757,
+      "learning_rate": 0.0002996843781547767,
+      "loss": 2.8269,
+      "step": 30628
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5702658891677856,
+      "learning_rate": 0.0002996689819733076,
+      "loss": 2.8871,
+      "step": 30629
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5865205526351929,
+      "learning_rate": 0.0002996535857927105,
+      "loss": 2.8866,
+      "step": 30630
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5985194444656372,
+      "learning_rate": 0.0002996381896130256,
+      "loss": 3.1501,
+      "step": 30631
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6110131144523621,
+      "learning_rate": 0.0002996227934342938,
+      "loss": 2.9409,
+      "step": 30632
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5706114768981934,
+      "learning_rate": 0.00029960739725655547,
+      "loss": 3.0904,
+      "step": 30633
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6214529275894165,
+      "learning_rate": 0.0002995920010798511,
+      "loss": 3.0926,
+      "step": 30634
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6152591109275818,
+      "learning_rate": 0.00029957660490422146,
+      "loss": 3.1173,
+      "step": 30635
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6281127333641052,
+      "learning_rate": 0.00029956120872970686,
+      "loss": 3.0511,
+      "step": 30636
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6203930377960205,
+      "learning_rate": 0.0002995458125563479,
+      "loss": 3.079,
+      "step": 30637
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.603507399559021,
+      "learning_rate": 0.0002995304163841853,
+      "loss": 3.2659,
+      "step": 30638
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5846010446548462,
+      "learning_rate": 0.0002995150202132595,
+      "loss": 3.1515,
+      "step": 30639
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6017332673072815,
+      "learning_rate": 0.000299499624043611,
+      "loss": 2.8583,
+      "step": 30640
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6173145771026611,
+      "learning_rate": 0.0002994842278752803,
+      "loss": 2.9607,
+      "step": 30641
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6321989297866821,
+      "learning_rate": 0.0002994688317083081,
+      "loss": 3.1546,
+      "step": 30642
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6010352373123169,
+      "learning_rate": 0.00029945343554273505,
+      "loss": 3.1106,
+      "step": 30643
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6526492238044739,
+      "learning_rate": 0.00029943803937860136,
+      "loss": 3.2372,
+      "step": 30644
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5786515474319458,
+      "learning_rate": 0.0002994226432159479,
+      "loss": 2.9052,
+      "step": 30645
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5871065258979797,
+      "learning_rate": 0.00029940724705481493,
+      "loss": 3.2204,
+      "step": 30646
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5986695289611816,
+      "learning_rate": 0.0002993918508952432,
+      "loss": 3.0846,
+      "step": 30647
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5756219625473022,
+      "learning_rate": 0.0002993764547372733,
+      "loss": 2.9358,
+      "step": 30648
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5979821681976318,
+      "learning_rate": 0.0002993610585809456,
+      "loss": 3.1195,
+      "step": 30649
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6420942544937134,
+      "learning_rate": 0.00029934566242630084,
+      "loss": 3.1625,
+      "step": 30650
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7682424783706665,
+      "learning_rate": 0.0002993302662733795,
+      "loss": 3.1441,
+      "step": 30651
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5857537984848022,
+      "learning_rate": 0.0002993148701222219,
+      "loss": 2.9752,
+      "step": 30652
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6025992035865784,
+      "learning_rate": 0.00029929947397286903,
+      "loss": 3.0435,
+      "step": 30653
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6120580434799194,
+      "learning_rate": 0.00029928407782536115,
+      "loss": 3.14,
+      "step": 30654
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5636025667190552,
+      "learning_rate": 0.0002992686816797389,
+      "loss": 3.1197,
+      "step": 30655
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5800879001617432,
+      "learning_rate": 0.0002992532855360427,
+      "loss": 2.947,
+      "step": 30656
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6197709441184998,
+      "learning_rate": 0.00029923788939431324,
+      "loss": 3.0448,
+      "step": 30657
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6334044337272644,
+      "learning_rate": 0.0002992224932545911,
+      "loss": 2.8652,
+      "step": 30658
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6068458557128906,
+      "learning_rate": 0.0002992070971169166,
+      "loss": 2.781,
+      "step": 30659
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6140934824943542,
+      "learning_rate": 0.0002991917009813307,
+      "loss": 2.8921,
+      "step": 30660
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.593880295753479,
+      "learning_rate": 0.00029917630484787357,
+      "loss": 3.1028,
+      "step": 30661
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5681495666503906,
+      "learning_rate": 0.0002991609087165858,
+      "loss": 3.0147,
+      "step": 30662
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6396278738975525,
+      "learning_rate": 0.0002991455125875082,
+      "loss": 3.0758,
+      "step": 30663
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6014585494995117,
+      "learning_rate": 0.00029913011646068104,
+      "loss": 3.1103,
+      "step": 30664
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6441971659660339,
+      "learning_rate": 0.0002991147203361451,
+      "loss": 2.9616,
+      "step": 30665
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5746193528175354,
+      "learning_rate": 0.00029909932421394073,
+      "loss": 2.9842,
+      "step": 30666
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6001632213592529,
+      "learning_rate": 0.0002990839280941086,
+      "loss": 3.1733,
+      "step": 30667
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5855227708816528,
+      "learning_rate": 0.00029906853197668925,
+      "loss": 3.0555,
+      "step": 30668
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6073086857795715,
+      "learning_rate": 0.0002990531358617231,
+      "loss": 2.9114,
+      "step": 30669
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6142853498458862,
+      "learning_rate": 0.00029903773974925094,
+      "loss": 3.2542,
+      "step": 30670
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6134935617446899,
+      "learning_rate": 0.00029902234363931305,
+      "loss": 3.0456,
+      "step": 30671
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6078312993049622,
+      "learning_rate": 0.0002990069475319502,
+      "loss": 3.1121,
+      "step": 30672
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5812007784843445,
+      "learning_rate": 0.0002989915514272029,
+      "loss": 2.9169,
+      "step": 30673
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5966639518737793,
+      "learning_rate": 0.00029897615532511156,
+      "loss": 3.0635,
+      "step": 30674
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6440772414207458,
+      "learning_rate": 0.00029896075922571693,
+      "loss": 3.085,
+      "step": 30675
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6082709431648254,
+      "learning_rate": 0.00029894536312905943,
+      "loss": 3.0807,
+      "step": 30676
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.585754930973053,
+      "learning_rate": 0.0002989299670351795,
+      "loss": 3.2426,
+      "step": 30677
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5902261734008789,
+      "learning_rate": 0.00029891457094411805,
+      "loss": 2.7597,
+      "step": 30678
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6010871529579163,
+      "learning_rate": 0.0002988991748559153,
+      "loss": 2.9683,
+      "step": 30679
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6037760972976685,
+      "learning_rate": 0.00029888377877061197,
+      "loss": 3.1568,
+      "step": 30680
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6149284839630127,
+      "learning_rate": 0.0002988683826882484,
+      "loss": 3.2423,
+      "step": 30681
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.897918164730072,
+      "learning_rate": 0.00029885298660886546,
+      "loss": 2.9964,
+      "step": 30682
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5681318044662476,
+      "learning_rate": 0.0002988375905325035,
+      "loss": 3.1145,
+      "step": 30683
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6103178262710571,
+      "learning_rate": 0.00029882219445920305,
+      "loss": 2.9291,
+      "step": 30684
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5969374179840088,
+      "learning_rate": 0.0002988067983890048,
+      "loss": 3.1072,
+      "step": 30685
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5936666131019592,
+      "learning_rate": 0.00029879140232194916,
+      "loss": 3.129,
+      "step": 30686
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6052805185317993,
+      "learning_rate": 0.0002987760062580768,
+      "loss": 2.9546,
+      "step": 30687
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5850530862808228,
+      "learning_rate": 0.00029876061019742803,
+      "loss": 2.9499,
+      "step": 30688
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5777554512023926,
+      "learning_rate": 0.0002987452141400436,
+      "loss": 2.9389,
+      "step": 30689
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5716623067855835,
+      "learning_rate": 0.0002987298180859642,
+      "loss": 2.9837,
+      "step": 30690
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5977976322174072,
+      "learning_rate": 0.0002987144220352301,
+      "loss": 3.116,
+      "step": 30691
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6488404273986816,
+      "learning_rate": 0.000298699025987882,
+      "loss": 3.0038,
+      "step": 30692
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6428252458572388,
+      "learning_rate": 0.0002986836299439604,
+      "loss": 3.0782,
+      "step": 30693
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6638103127479553,
+      "learning_rate": 0.0002986682339035058,
+      "loss": 3.1157,
+      "step": 30694
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5968785285949707,
+      "learning_rate": 0.00029865283786655896,
+      "loss": 3.2548,
+      "step": 30695
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6197184324264526,
+      "learning_rate": 0.00029863744183316015,
+      "loss": 3.1877,
+      "step": 30696
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6367472410202026,
+      "learning_rate": 0.00029862204580335017,
+      "loss": 3.0486,
+      "step": 30697
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.644905149936676,
+      "learning_rate": 0.00029860664977716944,
+      "loss": 2.8002,
+      "step": 30698
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6608884334564209,
+      "learning_rate": 0.0002985912537546584,
+      "loss": 3.0254,
+      "step": 30699
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6148144006729126,
+      "learning_rate": 0.00029857585773585785,
+      "loss": 3.0313,
+      "step": 30700
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6347461342811584,
+      "learning_rate": 0.0002985604617208082,
+      "loss": 2.7207,
+      "step": 30701
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6112677454948425,
+      "learning_rate": 0.00029854506570955005,
+      "loss": 3.0828,
+      "step": 30702
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5936033129692078,
+      "learning_rate": 0.0002985296697021238,
+      "loss": 3.1319,
+      "step": 30703
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6078819632530212,
+      "learning_rate": 0.0002985142736985702,
+      "loss": 3.0869,
+      "step": 30704
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6138762831687927,
+      "learning_rate": 0.00029849887769892977,
+      "loss": 2.7921,
+      "step": 30705
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6070495247840881,
+      "learning_rate": 0.0002984834817032429,
+      "loss": 3.1173,
+      "step": 30706
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6134839653968811,
+      "learning_rate": 0.0002984680857115503,
+      "loss": 2.9879,
+      "step": 30707
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5913065671920776,
+      "learning_rate": 0.00029845268972389247,
+      "loss": 3.2981,
+      "step": 30708
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5801486372947693,
+      "learning_rate": 0.00029843729374030983,
+      "loss": 3.3145,
+      "step": 30709
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6102262139320374,
+      "learning_rate": 0.00029842189776084325,
+      "loss": 3.0449,
+      "step": 30710
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5686670541763306,
+      "learning_rate": 0.0002984065017855329,
+      "loss": 2.991,
+      "step": 30711
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6398444175720215,
+      "learning_rate": 0.00029839110581441974,
+      "loss": 3.1699,
+      "step": 30712
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5702043175697327,
+      "learning_rate": 0.00029837570984754395,
+      "loss": 2.7375,
+      "step": 30713
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5870870351791382,
+      "learning_rate": 0.0002983603138849462,
+      "loss": 3.0903,
+      "step": 30714
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5614882707595825,
+      "learning_rate": 0.00029834491792666716,
+      "loss": 3.1141,
+      "step": 30715
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.56441730260849,
+      "learning_rate": 0.00029832952197274726,
+      "loss": 3.1426,
+      "step": 30716
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5936001539230347,
+      "learning_rate": 0.00029831412602322716,
+      "loss": 3.0108,
+      "step": 30717
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5969215035438538,
+      "learning_rate": 0.00029829873007814714,
+      "loss": 3.0834,
+      "step": 30718
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6016257405281067,
+      "learning_rate": 0.000298283334137548,
+      "loss": 2.9115,
+      "step": 30719
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6029444336891174,
+      "learning_rate": 0.00029826793820147035,
+      "loss": 3.0821,
+      "step": 30720
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5698739886283875,
+      "learning_rate": 0.00029825254226995446,
+      "loss": 3.0591,
+      "step": 30721
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.595145046710968,
+      "learning_rate": 0.00029823714634304117,
+      "loss": 3.0994,
+      "step": 30722
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6198235750198364,
+      "learning_rate": 0.0002982217504207708,
+      "loss": 2.9106,
+      "step": 30723
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5928401350975037,
+      "learning_rate": 0.000298206354503184,
+      "loss": 2.9303,
+      "step": 30724
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5822376608848572,
+      "learning_rate": 0.0002981909585903214,
+      "loss": 3.0786,
+      "step": 30725
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5875365138053894,
+      "learning_rate": 0.00029817556268222343,
+      "loss": 2.81,
+      "step": 30726
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6183528900146484,
+      "learning_rate": 0.00029816016677893073,
+      "loss": 3.1806,
+      "step": 30727
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5537392497062683,
+      "learning_rate": 0.00029814477088048366,
+      "loss": 3.1093,
+      "step": 30728
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5720838904380798,
+      "learning_rate": 0.000298129374986923,
+      "loss": 3.1148,
+      "step": 30729
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6122817397117615,
+      "learning_rate": 0.00029811397909828923,
+      "loss": 3.0031,
+      "step": 30730
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5819237232208252,
+      "learning_rate": 0.00029809858321462276,
+      "loss": 3.1225,
+      "step": 30731
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6177088618278503,
+      "learning_rate": 0.0002980831873359644,
+      "loss": 2.9787,
+      "step": 30732
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6123387813568115,
+      "learning_rate": 0.00029806779146235437,
+      "loss": 3.1079,
+      "step": 30733
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6111491918563843,
+      "learning_rate": 0.0002980523955938335,
+      "loss": 2.9765,
+      "step": 30734
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.566396176815033,
+      "learning_rate": 0.0002980369997304423,
+      "loss": 2.815,
+      "step": 30735
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.588200032711029,
+      "learning_rate": 0.0002980216038722212,
+      "loss": 3.112,
+      "step": 30736
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5717114806175232,
+      "learning_rate": 0.00029800620801921087,
+      "loss": 3.0413,
+      "step": 30737
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6023069620132446,
+      "learning_rate": 0.00029799081217145177,
+      "loss": 3.1242,
+      "step": 30738
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5799679756164551,
+      "learning_rate": 0.00029797541632898444,
+      "loss": 2.8477,
+      "step": 30739
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5831611752510071,
+      "learning_rate": 0.0002979600204918496,
+      "loss": 3.0047,
+      "step": 30740
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6182281374931335,
+      "learning_rate": 0.00029794462466008756,
+      "loss": 3.0421,
+      "step": 30741
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5878432989120483,
+      "learning_rate": 0.00029792922883373907,
+      "loss": 2.879,
+      "step": 30742
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5995323061943054,
+      "learning_rate": 0.0002979138330128445,
+      "loss": 2.9612,
+      "step": 30743
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5848367810249329,
+      "learning_rate": 0.00029789843719744453,
+      "loss": 3.1774,
+      "step": 30744
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5988871455192566,
+      "learning_rate": 0.00029788304138757976,
+      "loss": 3.0458,
+      "step": 30745
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6023797988891602,
+      "learning_rate": 0.0002978676455832905,
+      "loss": 3.1648,
+      "step": 30746
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6229612827301025,
+      "learning_rate": 0.00029785224978461756,
+      "loss": 3.0046,
+      "step": 30747
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6516324877738953,
+      "learning_rate": 0.00029783685399160134,
+      "loss": 3.0324,
+      "step": 30748
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6502434611320496,
+      "learning_rate": 0.0002978214582042824,
+      "loss": 2.9143,
+      "step": 30749
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5838092565536499,
+      "learning_rate": 0.00029780606242270135,
+      "loss": 2.9889,
+      "step": 30750
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5811021327972412,
+      "learning_rate": 0.0002977906666468987,
+      "loss": 3.0722,
+      "step": 30751
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5942781567573547,
+      "learning_rate": 0.0002977752708769151,
+      "loss": 2.9056,
+      "step": 30752
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5923861861228943,
+      "learning_rate": 0.00029775987511279086,
+      "loss": 3.1602,
+      "step": 30753
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6492545008659363,
+      "learning_rate": 0.00029774447935456663,
+      "loss": 3.1609,
+      "step": 30754
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6185020804405212,
+      "learning_rate": 0.00029772908360228324,
+      "loss": 2.8854,
+      "step": 30755
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6205618381500244,
+      "learning_rate": 0.00029771368785598085,
+      "loss": 2.8574,
+      "step": 30756
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.590815544128418,
+      "learning_rate": 0.0002976982921157003,
+      "loss": 2.9544,
+      "step": 30757
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6202199459075928,
+      "learning_rate": 0.0002976828963814818,
+      "loss": 3.0796,
+      "step": 30758
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6064924597740173,
+      "learning_rate": 0.00029766750065336625,
+      "loss": 2.9535,
+      "step": 30759
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6110441088676453,
+      "learning_rate": 0.00029765210493139406,
+      "loss": 3.1612,
+      "step": 30760
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5866239666938782,
+      "learning_rate": 0.00029763670921560567,
+      "loss": 2.9289,
+      "step": 30761
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5617367625236511,
+      "learning_rate": 0.00029762131350604185,
+      "loss": 3.1119,
+      "step": 30762
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.551780104637146,
+      "learning_rate": 0.000297605917802743,
+      "loss": 3.0988,
+      "step": 30763
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5915526747703552,
+      "learning_rate": 0.0002975905221057497,
+      "loss": 3.0443,
+      "step": 30764
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5596851110458374,
+      "learning_rate": 0.0002975751264151024,
+      "loss": 2.8931,
+      "step": 30765
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6421846151351929,
+      "learning_rate": 0.00029755973073084186,
+      "loss": 2.9407,
+      "step": 30766
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6003358364105225,
+      "learning_rate": 0.00029754433505300857,
+      "loss": 3.0634,
+      "step": 30767
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5622581839561462,
+      "learning_rate": 0.00029752893938164287,
+      "loss": 3.1398,
+      "step": 30768
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6357739567756653,
+      "learning_rate": 0.0002975135437167856,
+      "loss": 3.0894,
+      "step": 30769
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6039325594902039,
+      "learning_rate": 0.0002974981480584771,
+      "loss": 3.1304,
+      "step": 30770
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5737800002098083,
+      "learning_rate": 0.00029748275240675795,
+      "loss": 2.8875,
+      "step": 30771
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6089869141578674,
+      "learning_rate": 0.0002974673567616689,
+      "loss": 2.9146,
+      "step": 30772
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6089136600494385,
+      "learning_rate": 0.0002974519611232502,
+      "loss": 3.2106,
+      "step": 30773
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5950151681900024,
+      "learning_rate": 0.00029743656549154263,
+      "loss": 2.8026,
+      "step": 30774
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6248009204864502,
+      "learning_rate": 0.00029742116986658664,
+      "loss": 3.1683,
+      "step": 30775
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6114397048950195,
+      "learning_rate": 0.00029740577424842274,
+      "loss": 2.9792,
+      "step": 30776
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6554459929466248,
+      "learning_rate": 0.00029739037863709165,
+      "loss": 3.2027,
+      "step": 30777
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6299651265144348,
+      "learning_rate": 0.00029737498303263374,
+      "loss": 3.015,
+      "step": 30778
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6014512777328491,
+      "learning_rate": 0.00029735958743508964,
+      "loss": 3.058,
+      "step": 30779
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6071848273277283,
+      "learning_rate": 0.00029734419184449976,
+      "loss": 3.0521,
+      "step": 30780
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6110485792160034,
+      "learning_rate": 0.00029732879626090484,
+      "loss": 2.9316,
+      "step": 30781
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5651352405548096,
+      "learning_rate": 0.0002973134006843454,
+      "loss": 3.0355,
+      "step": 30782
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6445112824440002,
+      "learning_rate": 0.00029729800511486186,
+      "loss": 3.1326,
+      "step": 30783
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.582943320274353,
+      "learning_rate": 0.00029728260955249496,
+      "loss": 3.223,
+      "step": 30784
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.611379861831665,
+      "learning_rate": 0.0002972672139972851,
+      "loss": 2.8036,
+      "step": 30785
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6126112937927246,
+      "learning_rate": 0.0002972518184492728,
+      "loss": 2.797,
+      "step": 30786
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5877750515937805,
+      "learning_rate": 0.0002972364229084988,
+      "loss": 2.9347,
+      "step": 30787
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6406726837158203,
+      "learning_rate": 0.0002972210273750035,
+      "loss": 3.1474,
+      "step": 30788
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6040596961975098,
+      "learning_rate": 0.0002972056318488275,
+      "loss": 3.1666,
+      "step": 30789
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6338227987289429,
+      "learning_rate": 0.0002971902363300112,
+      "loss": 3.1337,
+      "step": 30790
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6211509704589844,
+      "learning_rate": 0.00029717484081859534,
+      "loss": 2.8265,
+      "step": 30791
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5889843106269836,
+      "learning_rate": 0.0002971594453146205,
+      "loss": 3.1715,
+      "step": 30792
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5887772440910339,
+      "learning_rate": 0.00029714404981812695,
+      "loss": 2.9016,
+      "step": 30793
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5816041231155396,
+      "learning_rate": 0.00029712865432915566,
+      "loss": 2.9766,
+      "step": 30794
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5636284947395325,
+      "learning_rate": 0.00029711325884774673,
+      "loss": 3.1463,
+      "step": 30795
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5685577392578125,
+      "learning_rate": 0.000297097863373941,
+      "loss": 2.8435,
+      "step": 30796
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6058424115180969,
+      "learning_rate": 0.000297082467907779,
+      "loss": 3.2246,
+      "step": 30797
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6104329824447632,
+      "learning_rate": 0.0002970670724493011,
+      "loss": 3.0745,
+      "step": 30798
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5837059020996094,
+      "learning_rate": 0.00029705167699854804,
+      "loss": 2.9483,
+      "step": 30799
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6254569888114929,
+      "learning_rate": 0.0002970362815555603,
+      "loss": 2.8121,
+      "step": 30800
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5854609608650208,
+      "learning_rate": 0.00029702088612037835,
+      "loss": 3.0961,
+      "step": 30801
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5687035322189331,
+      "learning_rate": 0.0002970054906930429,
+      "loss": 2.8367,
+      "step": 30802
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5669450759887695,
+      "learning_rate": 0.00029699009527359437,
+      "loss": 3.0268,
+      "step": 30803
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5802821516990662,
+      "learning_rate": 0.00029697469986207344,
+      "loss": 3.1936,
+      "step": 30804
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6279524564743042,
+      "learning_rate": 0.0002969593044585204,
+      "loss": 3.0683,
+      "step": 30805
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5692258477210999,
+      "learning_rate": 0.0002969439090629761,
+      "loss": 3.2225,
+      "step": 30806
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5868411064147949,
+      "learning_rate": 0.00029692851367548096,
+      "loss": 2.9523,
+      "step": 30807
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5972650647163391,
+      "learning_rate": 0.0002969131182960754,
+      "loss": 2.8724,
+      "step": 30808
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6030793786048889,
+      "learning_rate": 0.0002968977229248002,
+      "loss": 3.2454,
+      "step": 30809
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6051518321037292,
+      "learning_rate": 0.00029688232756169575,
+      "loss": 3.1048,
+      "step": 30810
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5919243097305298,
+      "learning_rate": 0.0002968669322068026,
+      "loss": 3.0596,
+      "step": 30811
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6317167282104492,
+      "learning_rate": 0.00029685153686016146,
+      "loss": 2.9332,
+      "step": 30812
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6132222414016724,
+      "learning_rate": 0.0002968361415218127,
+      "loss": 3.0974,
+      "step": 30813
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.621911346912384,
+      "learning_rate": 0.000296820746191797,
+      "loss": 3.1756,
+      "step": 30814
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6402205228805542,
+      "learning_rate": 0.0002968053508701548,
+      "loss": 3.0177,
+      "step": 30815
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5846377015113831,
+      "learning_rate": 0.0002967899555569266,
+      "loss": 3.0623,
+      "step": 30816
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6023883819580078,
+      "learning_rate": 0.00029677456025215324,
+      "loss": 2.9456,
+      "step": 30817
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6518478393554688,
+      "learning_rate": 0.0002967591649558749,
+      "loss": 3.2577,
+      "step": 30818
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5860528945922852,
+      "learning_rate": 0.0002967437696681324,
+      "loss": 2.9268,
+      "step": 30819
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5669897198677063,
+      "learning_rate": 0.0002967283743889661,
+      "loss": 2.9486,
+      "step": 30820
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5880569815635681,
+      "learning_rate": 0.00029671297911841663,
+      "loss": 3.3732,
+      "step": 30821
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.588849663734436,
+      "learning_rate": 0.00029669758385652466,
+      "loss": 3.0296,
+      "step": 30822
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5880824327468872,
+      "learning_rate": 0.0002966821886033305,
+      "loss": 3.1179,
+      "step": 30823
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5957514047622681,
+      "learning_rate": 0.00029666679335887486,
+      "loss": 3.0971,
+      "step": 30824
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.646440327167511,
+      "learning_rate": 0.00029665139812319824,
+      "loss": 3.0703,
+      "step": 30825
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6188963055610657,
+      "learning_rate": 0.0002966360028963412,
+      "loss": 2.8634,
+      "step": 30826
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5998916625976562,
+      "learning_rate": 0.00029662060767834427,
+      "loss": 3.0819,
+      "step": 30827
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5960819721221924,
+      "learning_rate": 0.000296605212469248,
+      "loss": 3.1007,
+      "step": 30828
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6161120533943176,
+      "learning_rate": 0.00029658981726909305,
+      "loss": 2.9573,
+      "step": 30829
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6249703168869019,
+      "learning_rate": 0.0002965744220779197,
+      "loss": 3.0573,
+      "step": 30830
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6068581342697144,
+      "learning_rate": 0.0002965590268957687,
+      "loss": 3.1307,
+      "step": 30831
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5972431302070618,
+      "learning_rate": 0.0002965436317226807,
+      "loss": 3.2302,
+      "step": 30832
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5628926753997803,
+      "learning_rate": 0.0002965282365586959,
+      "loss": 3.204,
+      "step": 30833
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6214185357093811,
+      "learning_rate": 0.00029651284140385525,
+      "loss": 3.0294,
+      "step": 30834
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6001722812652588,
+      "learning_rate": 0.0002964974462581989,
+      "loss": 2.8884,
+      "step": 30835
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6169463396072388,
+      "learning_rate": 0.0002964820511217678,
+      "loss": 2.9654,
+      "step": 30836
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5828376412391663,
+      "learning_rate": 0.00029646665599460223,
+      "loss": 3.0695,
+      "step": 30837
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6448942422866821,
+      "learning_rate": 0.0002964512608767427,
+      "loss": 3.1534,
+      "step": 30838
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5935970544815063,
+      "learning_rate": 0.00029643586576823006,
+      "loss": 2.8783,
+      "step": 30839
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6004209518432617,
+      "learning_rate": 0.0002964204706691046,
+      "loss": 2.9735,
+      "step": 30840
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5771257877349854,
+      "learning_rate": 0.0002964050755794068,
+      "loss": 2.7251,
+      "step": 30841
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5918455123901367,
+      "learning_rate": 0.0002963896804991775,
+      "loss": 2.9045,
+      "step": 30842
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5754095911979675,
+      "learning_rate": 0.000296374285428457,
+      "loss": 3.2429,
+      "step": 30843
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5826681852340698,
+      "learning_rate": 0.000296358890367286,
+      "loss": 3.0659,
+      "step": 30844
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.610785722732544,
+      "learning_rate": 0.00029634349531570487,
+      "loss": 3.0146,
+      "step": 30845
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5650796890258789,
+      "learning_rate": 0.0002963281002737544,
+      "loss": 2.9735,
+      "step": 30846
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6261953711509705,
+      "learning_rate": 0.000296312705241475,
+      "loss": 2.9293,
+      "step": 30847
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5803135633468628,
+      "learning_rate": 0.00029629731021890704,
+      "loss": 3.1329,
+      "step": 30848
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.611264169216156,
+      "learning_rate": 0.0002962819152060915,
+      "loss": 3.0638,
+      "step": 30849
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.587598979473114,
+      "learning_rate": 0.00029626652020306855,
+      "loss": 3.1472,
+      "step": 30850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6119747161865234,
+      "learning_rate": 0.0002962511252098789,
+      "loss": 3.0915,
+      "step": 30851
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6030398011207581,
+      "learning_rate": 0.000296235730226563,
+      "loss": 2.7529,
+      "step": 30852
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6091843843460083,
+      "learning_rate": 0.0002962203352531615,
+      "loss": 3.1295,
+      "step": 30853
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5616978406906128,
+      "learning_rate": 0.00029620494028971496,
+      "loss": 3.268,
+      "step": 30854
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6114065647125244,
+      "learning_rate": 0.0002961895453362638,
+      "loss": 3.12,
+      "step": 30855
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6025994420051575,
+      "learning_rate": 0.00029617415039284874,
+      "loss": 2.9599,
+      "step": 30856
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5990439057350159,
+      "learning_rate": 0.0002961587554595101,
+      "loss": 3.1287,
+      "step": 30857
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5876929759979248,
+      "learning_rate": 0.0002961433605362886,
+      "loss": 3.0938,
+      "step": 30858
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6120889186859131,
+      "learning_rate": 0.00029612796562322486,
+      "loss": 3.1546,
+      "step": 30859
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5933310985565186,
+      "learning_rate": 0.00029611257072035917,
+      "loss": 3.0761,
+      "step": 30860
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.646134614944458,
+      "learning_rate": 0.0002960971758277323,
+      "loss": 3.1372,
+      "step": 30861
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5914332866668701,
+      "learning_rate": 0.0002960817809453847,
+      "loss": 2.8682,
+      "step": 30862
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6064649820327759,
+      "learning_rate": 0.00029606638607335687,
+      "loss": 3.0682,
+      "step": 30863
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5667405128479004,
+      "learning_rate": 0.00029605099121168956,
+      "loss": 3.1389,
+      "step": 30864
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5688351988792419,
+      "learning_rate": 0.0002960355963604231,
+      "loss": 3.0559,
+      "step": 30865
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5366978645324707,
+      "learning_rate": 0.0002960202015195982,
+      "loss": 3.0342,
+      "step": 30866
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6322748064994812,
+      "learning_rate": 0.00029600480668925516,
+      "loss": 3.4253,
+      "step": 30867
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6188519597053528,
+      "learning_rate": 0.00029598941186943475,
+      "loss": 3.2374,
+      "step": 30868
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.567204475402832,
+      "learning_rate": 0.00029597401706017755,
+      "loss": 3.1257,
+      "step": 30869
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6143216490745544,
+      "learning_rate": 0.00029595862226152386,
+      "loss": 2.9782,
+      "step": 30870
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6275702118873596,
+      "learning_rate": 0.00029594322747351453,
+      "loss": 3.0836,
+      "step": 30871
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7560949325561523,
+      "learning_rate": 0.0002959278326961899,
+      "loss": 3.0159,
+      "step": 30872
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6060622334480286,
+      "learning_rate": 0.0002959124379295905,
+      "loss": 3.2309,
+      "step": 30873
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5544068813323975,
+      "learning_rate": 0.0002958970431737571,
+      "loss": 2.9914,
+      "step": 30874
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5971145629882812,
+      "learning_rate": 0.00029588164842872994,
+      "loss": 3.0419,
+      "step": 30875
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6468105912208557,
+      "learning_rate": 0.0002958662536945499,
+      "loss": 2.7897,
+      "step": 30876
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6096239686012268,
+      "learning_rate": 0.0002958508589712572,
+      "loss": 3.1458,
+      "step": 30877
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6860781908035278,
+      "learning_rate": 0.00029583546425889255,
+      "loss": 2.8722,
+      "step": 30878
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5959381461143494,
+      "learning_rate": 0.0002958200695574966,
+      "loss": 3.0762,
+      "step": 30879
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.587128221988678,
+      "learning_rate": 0.00029580467486710967,
+      "loss": 3.1168,
+      "step": 30880
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6013617515563965,
+      "learning_rate": 0.00029578928018777253,
+      "loss": 3.0795,
+      "step": 30881
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6354458928108215,
+      "learning_rate": 0.0002957738855195255,
+      "loss": 3.2363,
+      "step": 30882
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6061522960662842,
+      "learning_rate": 0.0002957584908624093,
+      "loss": 3.1117,
+      "step": 30883
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6070138216018677,
+      "learning_rate": 0.00029574309621646443,
+      "loss": 2.9919,
+      "step": 30884
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6107861399650574,
+      "learning_rate": 0.0002957277015817314,
+      "loss": 3.265,
+      "step": 30885
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5738034844398499,
+      "learning_rate": 0.0002957123069582508,
+      "loss": 3.1544,
+      "step": 30886
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6216245889663696,
+      "learning_rate": 0.00029569691234606314,
+      "loss": 2.9596,
+      "step": 30887
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6213082671165466,
+      "learning_rate": 0.0002956815177452089,
+      "loss": 2.9898,
+      "step": 30888
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6034766435623169,
+      "learning_rate": 0.0002956661231557289,
+      "loss": 3.1961,
+      "step": 30889
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5714133977890015,
+      "learning_rate": 0.0002956507285776634,
+      "loss": 3.2041,
+      "step": 30890
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5667616724967957,
+      "learning_rate": 0.0002956353340110531,
+      "loss": 3.1601,
+      "step": 30891
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5944485068321228,
+      "learning_rate": 0.00029561993945593834,
+      "loss": 3.0033,
+      "step": 30892
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5806108117103577,
+      "learning_rate": 0.0002956045449123599,
+      "loss": 2.8649,
+      "step": 30893
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.621354877948761,
+      "learning_rate": 0.0002955891503803583,
+      "loss": 3.0581,
+      "step": 30894
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6162626147270203,
+      "learning_rate": 0.00029557375585997393,
+      "loss": 2.9909,
+      "step": 30895
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5693962574005127,
+      "learning_rate": 0.00029555836135124755,
+      "loss": 3.117,
+      "step": 30896
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7049458622932434,
+      "learning_rate": 0.00029554296685421945,
+      "loss": 3.0019,
+      "step": 30897
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5936644077301025,
+      "learning_rate": 0.0002955275723689304,
+      "loss": 3.3479,
+      "step": 30898
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5873288512229919,
+      "learning_rate": 0.00029551217789542094,
+      "loss": 2.8661,
+      "step": 30899
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5874931812286377,
+      "learning_rate": 0.00029549678343373137,
+      "loss": 3.2489,
+      "step": 30900
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6320956945419312,
+      "learning_rate": 0.00029548138898390254,
+      "loss": 2.8875,
+      "step": 30901
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5752156972885132,
+      "learning_rate": 0.00029546599454597484,
+      "loss": 2.9791,
+      "step": 30902
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5847448110580444,
+      "learning_rate": 0.00029545060011998875,
+      "loss": 2.9069,
+      "step": 30903
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5820603966712952,
+      "learning_rate": 0.00029543520570598505,
+      "loss": 3.1561,
+      "step": 30904
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6131413578987122,
+      "learning_rate": 0.000295419811304004,
+      "loss": 3.1054,
+      "step": 30905
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6036339402198792,
+      "learning_rate": 0.0002954044169140864,
+      "loss": 2.9931,
+      "step": 30906
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6479763984680176,
+      "learning_rate": 0.0002953890225362726,
+      "loss": 3.1477,
+      "step": 30907
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5721249580383301,
+      "learning_rate": 0.00029537362817060324,
+      "loss": 3.0113,
+      "step": 30908
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6328310370445251,
+      "learning_rate": 0.0002953582338171189,
+      "loss": 2.9072,
+      "step": 30909
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6006953716278076,
+      "learning_rate": 0.00029534283947586,
+      "loss": 2.95,
+      "step": 30910
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6052888631820679,
+      "learning_rate": 0.00029532744514686724,
+      "loss": 2.8012,
+      "step": 30911
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6413729190826416,
+      "learning_rate": 0.00029531205083018104,
+      "loss": 3.1825,
+      "step": 30912
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5984901785850525,
+      "learning_rate": 0.00029529665652584195,
+      "loss": 2.9248,
+      "step": 30913
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5955290198326111,
+      "learning_rate": 0.0002952812622338907,
+      "loss": 2.9168,
+      "step": 30914
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6276412606239319,
+      "learning_rate": 0.00029526586795436757,
+      "loss": 3.1651,
+      "step": 30915
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5960054397583008,
+      "learning_rate": 0.0002952504736873134,
+      "loss": 3.1858,
+      "step": 30916
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5850929021835327,
+      "learning_rate": 0.0002952350794327684,
+      "loss": 3.048,
+      "step": 30917
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.616568386554718,
+      "learning_rate": 0.00029521968519077325,
+      "loss": 2.8989,
+      "step": 30918
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.623950183391571,
+      "learning_rate": 0.0002952042909613687,
+      "loss": 3.1168,
+      "step": 30919
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5852279663085938,
+      "learning_rate": 0.000295188896744595,
+      "loss": 3.0668,
+      "step": 30920
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5998043417930603,
+      "learning_rate": 0.00029517350254049297,
+      "loss": 3.0312,
+      "step": 30921
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6115825772285461,
+      "learning_rate": 0.0002951581083491028,
+      "loss": 2.8682,
+      "step": 30922
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6442579627037048,
+      "learning_rate": 0.0002951427141704654,
+      "loss": 3.1506,
+      "step": 30923
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6034064292907715,
+      "learning_rate": 0.0002951273200046211,
+      "loss": 3.1786,
+      "step": 30924
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5963617563247681,
+      "learning_rate": 0.0002951119258516104,
+      "loss": 2.975,
+      "step": 30925
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6269310712814331,
+      "learning_rate": 0.0002950965317114741,
+      "loss": 3.1549,
+      "step": 30926
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5554681420326233,
+      "learning_rate": 0.0002950811375842526,
+      "loss": 3.099,
+      "step": 30927
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6449235677719116,
+      "learning_rate": 0.0002950657434699864,
+      "loss": 3.0467,
+      "step": 30928
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.633747398853302,
+      "learning_rate": 0.00029505034936871597,
+      "loss": 2.9924,
+      "step": 30929
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5843096971511841,
+      "learning_rate": 0.00029503495528048206,
+      "loss": 3.2848,
+      "step": 30930
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6081466674804688,
+      "learning_rate": 0.00029501956120532515,
+      "loss": 3.054,
+      "step": 30931
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5923753976821899,
+      "learning_rate": 0.0002950041671432856,
+      "loss": 2.893,
+      "step": 30932
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.638497531414032,
+      "learning_rate": 0.00029498877309440435,
+      "loss": 3.1434,
+      "step": 30933
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5726613998413086,
+      "learning_rate": 0.00029497337905872155,
+      "loss": 3.0945,
+      "step": 30934
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6029412746429443,
+      "learning_rate": 0.0002949579850362779,
+      "loss": 3.0213,
+      "step": 30935
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6141619086265564,
+      "learning_rate": 0.00029494259102711405,
+      "loss": 2.8401,
+      "step": 30936
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5874308347702026,
+      "learning_rate": 0.0002949271970312703,
+      "loss": 3.1897,
+      "step": 30937
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5788091421127319,
+      "learning_rate": 0.0002949118030487875,
+      "loss": 3.0942,
+      "step": 30938
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6286618709564209,
+      "learning_rate": 0.00029489640907970593,
+      "loss": 2.9367,
+      "step": 30939
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.588944137096405,
+      "learning_rate": 0.00029488101512406615,
+      "loss": 3.0126,
+      "step": 30940
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5936257243156433,
+      "learning_rate": 0.000294865621181909,
+      "loss": 2.9992,
+      "step": 30941
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5838150382041931,
+      "learning_rate": 0.0002948502272532747,
+      "loss": 3.0763,
+      "step": 30942
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5851708650588989,
+      "learning_rate": 0.000294834833338204,
+      "loss": 3.0234,
+      "step": 30943
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6120582818984985,
+      "learning_rate": 0.00029481943943673715,
+      "loss": 2.7278,
+      "step": 30944
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6074814796447754,
+      "learning_rate": 0.00029480404554891503,
+      "loss": 3.1052,
+      "step": 30945
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5912714004516602,
+      "learning_rate": 0.00029478865167477814,
+      "loss": 2.8922,
+      "step": 30946
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5891293287277222,
+      "learning_rate": 0.00029477325781436677,
+      "loss": 3.0284,
+      "step": 30947
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.714016854763031,
+      "learning_rate": 0.0002947578639677218,
+      "loss": 2.9723,
+      "step": 30948
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5967224836349487,
+      "learning_rate": 0.0002947424701348835,
+      "loss": 3.0212,
+      "step": 30949
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5975646376609802,
+      "learning_rate": 0.00029472707631589246,
+      "loss": 3.0859,
+      "step": 30950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6335617899894714,
+      "learning_rate": 0.00029471168251078944,
+      "loss": 2.9693,
+      "step": 30951
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6031022071838379,
+      "learning_rate": 0.00029469628871961473,
+      "loss": 2.8384,
+      "step": 30952
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5797258019447327,
+      "learning_rate": 0.0002946808949424091,
+      "loss": 2.9484,
+      "step": 30953
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5635722875595093,
+      "learning_rate": 0.00029466550117921277,
+      "loss": 3.0065,
+      "step": 30954
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6008172631263733,
+      "learning_rate": 0.00029465010743006656,
+      "loss": 3.1646,
+      "step": 30955
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5772174596786499,
+      "learning_rate": 0.0002946347136950111,
+      "loss": 2.8605,
+      "step": 30956
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6074007749557495,
+      "learning_rate": 0.00029461931997408654,
+      "loss": 3.1638,
+      "step": 30957
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5578022003173828,
+      "learning_rate": 0.0002946039262673338,
+      "loss": 2.9493,
+      "step": 30958
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6266377568244934,
+      "learning_rate": 0.0002945885325747931,
+      "loss": 2.9573,
+      "step": 30959
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6051025986671448,
+      "learning_rate": 0.0002945731388965053,
+      "loss": 2.8036,
+      "step": 30960
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6597091555595398,
+      "learning_rate": 0.0002945577452325109,
+      "loss": 3.0133,
+      "step": 30961
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5791193842887878,
+      "learning_rate": 0.00029454235158285014,
+      "loss": 3.1355,
+      "step": 30962
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6521401405334473,
+      "learning_rate": 0.00029452695794756397,
+      "loss": 3.2595,
+      "step": 30963
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.614771842956543,
+      "learning_rate": 0.00029451156432669265,
+      "loss": 3.0447,
+      "step": 30964
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6045995354652405,
+      "learning_rate": 0.0002944961707202768,
+      "loss": 3.0624,
+      "step": 30965
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5819627642631531,
+      "learning_rate": 0.000294480777128357,
+      "loss": 3.2695,
+      "step": 30966
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5987125039100647,
+      "learning_rate": 0.0002944653835509738,
+      "loss": 3.1373,
+      "step": 30967
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5703858137130737,
+      "learning_rate": 0.00029444998998816774,
+      "loss": 3.2609,
+      "step": 30968
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6514813303947449,
+      "learning_rate": 0.0002944345964399792,
+      "loss": 3.0562,
+      "step": 30969
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6097198724746704,
+      "learning_rate": 0.00029441920290644894,
+      "loss": 3.1765,
+      "step": 30970
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5842579007148743,
+      "learning_rate": 0.00029440380938761747,
+      "loss": 3.0327,
+      "step": 30971
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5636705756187439,
+      "learning_rate": 0.0002943884158835252,
+      "loss": 2.9826,
+      "step": 30972
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5914285778999329,
+      "learning_rate": 0.00029437302239421285,
+      "loss": 2.9421,
+      "step": 30973
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6153814792633057,
+      "learning_rate": 0.00029435762891972085,
+      "loss": 3.1329,
+      "step": 30974
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5827009081840515,
+      "learning_rate": 0.0002943422354600897,
+      "loss": 3.1252,
+      "step": 30975
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5662691593170166,
+      "learning_rate": 0.00029432684201536013,
+      "loss": 3.2731,
+      "step": 30976
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6179322600364685,
+      "learning_rate": 0.00029431144858557245,
+      "loss": 3.048,
+      "step": 30977
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6341006755828857,
+      "learning_rate": 0.0002942960551707674,
+      "loss": 3.1041,
+      "step": 30978
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.626679539680481,
+      "learning_rate": 0.0002942806617709854,
+      "loss": 3.1037,
+      "step": 30979
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5652845501899719,
+      "learning_rate": 0.000294265268386267,
+      "loss": 2.7815,
+      "step": 30980
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6046680212020874,
+      "learning_rate": 0.0002942498750166529,
+      "loss": 2.9942,
+      "step": 30981
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5820021629333496,
+      "learning_rate": 0.00029423448166218343,
+      "loss": 3.126,
+      "step": 30982
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6065924167633057,
+      "learning_rate": 0.0002942190883228993,
+      "loss": 2.8712,
+      "step": 30983
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6059271693229675,
+      "learning_rate": 0.00029420369499884086,
+      "loss": 3.0145,
+      "step": 30984
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6043378710746765,
+      "learning_rate": 0.0002941883016900489,
+      "loss": 3.1985,
+      "step": 30985
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5847151875495911,
+      "learning_rate": 0.0002941729083965638,
+      "loss": 3.1297,
+      "step": 30986
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6242455244064331,
+      "learning_rate": 0.000294157515118426,
+      "loss": 3.034,
+      "step": 30987
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6258992552757263,
+      "learning_rate": 0.00029414212185567643,
+      "loss": 2.9636,
+      "step": 30988
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6174604892730713,
+      "learning_rate": 0.0002941267286083552,
+      "loss": 3.1873,
+      "step": 30989
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6053414940834045,
+      "learning_rate": 0.00029411133537650305,
+      "loss": 3.0621,
+      "step": 30990
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6238487362861633,
+      "learning_rate": 0.00029409594216016057,
+      "loss": 3.3516,
+      "step": 30991
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6055889129638672,
+      "learning_rate": 0.0002940805489593682,
+      "loss": 2.9447,
+      "step": 30992
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6507366895675659,
+      "learning_rate": 0.0002940651557741666,
+      "loss": 3.007,
+      "step": 30993
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5801891684532166,
+      "learning_rate": 0.0002940497626045962,
+      "loss": 3.0388,
+      "step": 30994
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5716410875320435,
+      "learning_rate": 0.0002940343694506975,
+      "loss": 2.8096,
+      "step": 30995
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5872518420219421,
+      "learning_rate": 0.00029401897631251125,
+      "loss": 3.0487,
+      "step": 30996
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6373841762542725,
+      "learning_rate": 0.0002940035831900777,
+      "loss": 3.104,
+      "step": 30997
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6203256845474243,
+      "learning_rate": 0.0002939881900834377,
+      "loss": 2.8221,
+      "step": 30998
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5925654768943787,
+      "learning_rate": 0.0002939727969926316,
+      "loss": 3.0561,
+      "step": 30999
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.58648282289505,
+      "learning_rate": 0.00029395740391770003,
+      "loss": 3.1059,
+      "step": 31000
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5882959365844727,
+      "learning_rate": 0.00029394201085868347,
+      "loss": 3.2673,
+      "step": 31001
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6323480606079102,
+      "learning_rate": 0.0002939266178156224,
+      "loss": 3.2371,
+      "step": 31002
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5646398067474365,
+      "learning_rate": 0.0002939112247885576,
+      "loss": 3.3075,
+      "step": 31003
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9151919484138489,
+      "learning_rate": 0.0002938958317775294,
+      "loss": 3.2684,
+      "step": 31004
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5926310420036316,
+      "learning_rate": 0.00029388043878257847,
+      "loss": 3.0728,
+      "step": 31005
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.580113410949707,
+      "learning_rate": 0.0002938650458037451,
+      "loss": 3.203,
+      "step": 31006
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5675933361053467,
+      "learning_rate": 0.00029384965284107016,
+      "loss": 3.0887,
+      "step": 31007
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6494491696357727,
+      "learning_rate": 0.00029383425989459405,
+      "loss": 3.0097,
+      "step": 31008
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5970069169998169,
+      "learning_rate": 0.00029381886696435724,
+      "loss": 2.7491,
+      "step": 31009
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5872333645820618,
+      "learning_rate": 0.0002938034740504004,
+      "loss": 3.293,
+      "step": 31010
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5623251795768738,
+      "learning_rate": 0.00029378808115276405,
+      "loss": 3.0746,
+      "step": 31011
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5938485860824585,
+      "learning_rate": 0.00029377268827148855,
+      "loss": 3.1148,
+      "step": 31012
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5834391713142395,
+      "learning_rate": 0.0002937572954066148,
+      "loss": 2.8762,
+      "step": 31013
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5958360433578491,
+      "learning_rate": 0.000293741902558183,
+      "loss": 2.9429,
+      "step": 31014
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5586824417114258,
+      "learning_rate": 0.00029372650972623395,
+      "loss": 3.0833,
+      "step": 31015
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6077871322631836,
+      "learning_rate": 0.00029371111691080785,
+      "loss": 2.9211,
+      "step": 31016
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.566399097442627,
+      "learning_rate": 0.00029369572411194555,
+      "loss": 3.0479,
+      "step": 31017
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5618554949760437,
+      "learning_rate": 0.00029368033132968764,
+      "loss": 2.9492,
+      "step": 31018
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6119033694267273,
+      "learning_rate": 0.00029366493856407435,
+      "loss": 3.0998,
+      "step": 31019
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5961689352989197,
+      "learning_rate": 0.0002936495458151465,
+      "loss": 2.8835,
+      "step": 31020
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.563315212726593,
+      "learning_rate": 0.0002936341530829444,
+      "loss": 3.0595,
+      "step": 31021
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5843584537506104,
+      "learning_rate": 0.0002936187603675088,
+      "loss": 3.117,
+      "step": 31022
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6035363674163818,
+      "learning_rate": 0.00029360336766888023,
+      "loss": 2.9555,
+      "step": 31023
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6223370432853699,
+      "learning_rate": 0.00029358797498709897,
+      "loss": 3.3564,
+      "step": 31024
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5917815566062927,
+      "learning_rate": 0.00029357258232220595,
+      "loss": 3.1168,
+      "step": 31025
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.567468523979187,
+      "learning_rate": 0.00029355718967424144,
+      "loss": 3.0815,
+      "step": 31026
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6053411364555359,
+      "learning_rate": 0.00029354179704324594,
+      "loss": 2.9032,
+      "step": 31027
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6136942505836487,
+      "learning_rate": 0.0002935264044292602,
+      "loss": 2.9913,
+      "step": 31028
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5973330736160278,
+      "learning_rate": 0.00029351101183232466,
+      "loss": 3.091,
+      "step": 31029
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6054880619049072,
+      "learning_rate": 0.00029349561925248,
+      "loss": 2.8339,
+      "step": 31030
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5846839547157288,
+      "learning_rate": 0.0002934802266897664,
+      "loss": 2.973,
+      "step": 31031
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5589364767074585,
+      "learning_rate": 0.0002934648341442247,
+      "loss": 3.0979,
+      "step": 31032
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5708386898040771,
+      "learning_rate": 0.0002934494416158954,
+      "loss": 3.19,
+      "step": 31033
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5712014436721802,
+      "learning_rate": 0.000293434049104819,
+      "loss": 3.1109,
+      "step": 31034
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6233195662498474,
+      "learning_rate": 0.00029341865661103604,
+      "loss": 3.1072,
+      "step": 31035
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5828282833099365,
+      "learning_rate": 0.0002934032641345871,
+      "loss": 3.1385,
+      "step": 31036
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6146054863929749,
+      "learning_rate": 0.00029338787167551263,
+      "loss": 2.9546,
+      "step": 31037
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5709480047225952,
+      "learning_rate": 0.0002933724792338533,
+      "loss": 2.9423,
+      "step": 31038
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5619285702705383,
+      "learning_rate": 0.00029335708680964953,
+      "loss": 2.9497,
+      "step": 31039
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6475139856338501,
+      "learning_rate": 0.000293341694402942,
+      "loss": 2.9842,
+      "step": 31040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5895599722862244,
+      "learning_rate": 0.0002933263020137711,
+      "loss": 2.8115,
+      "step": 31041
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5933480858802795,
+      "learning_rate": 0.0002933109096421774,
+      "loss": 3.0456,
+      "step": 31042
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6113942861557007,
+      "learning_rate": 0.00029329551728820153,
+      "loss": 3.0089,
+      "step": 31043
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5870497226715088,
+      "learning_rate": 0.000293280124951884,
+      "loss": 3.2366,
+      "step": 31044
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6341589689254761,
+      "learning_rate": 0.00029326473263326536,
+      "loss": 3.2206,
+      "step": 31045
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6039684414863586,
+      "learning_rate": 0.000293249340332386,
+      "loss": 2.9294,
+      "step": 31046
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.609302282333374,
+      "learning_rate": 0.0002932339480492866,
+      "loss": 3.1502,
+      "step": 31047
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5908909440040588,
+      "learning_rate": 0.0002932185557840078,
+      "loss": 3.0183,
+      "step": 31048
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6392471790313721,
+      "learning_rate": 0.0002932031635365899,
+      "loss": 2.9455,
+      "step": 31049
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6417034864425659,
+      "learning_rate": 0.00029318777130707364,
+      "loss": 2.9563,
+      "step": 31050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.637570858001709,
+      "learning_rate": 0.0002931723790954994,
+      "loss": 2.9481,
+      "step": 31051
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5919541716575623,
+      "learning_rate": 0.0002931569869019078,
+      "loss": 3.021,
+      "step": 31052
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6471995115280151,
+      "learning_rate": 0.00029314159472633943,
+      "loss": 2.9314,
+      "step": 31053
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5962557196617126,
+      "learning_rate": 0.00029312620256883475,
+      "loss": 2.9913,
+      "step": 31054
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.57488614320755,
+      "learning_rate": 0.0002931108104294344,
+      "loss": 3.0118,
+      "step": 31055
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6746479272842407,
+      "learning_rate": 0.00029309541830817873,
+      "loss": 2.932,
+      "step": 31056
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6285524964332581,
+      "learning_rate": 0.0002930800262051085,
+      "loss": 3.079,
+      "step": 31057
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6768153309822083,
+      "learning_rate": 0.0002930646341202641,
+      "loss": 3.0896,
+      "step": 31058
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6034467220306396,
+      "learning_rate": 0.0002930492420536861,
+      "loss": 3.1119,
+      "step": 31059
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5936095714569092,
+      "learning_rate": 0.00029303385000541517,
+      "loss": 3.1263,
+      "step": 31060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6120097041130066,
+      "learning_rate": 0.00029301845797549155,
+      "loss": 2.9978,
+      "step": 31061
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6264371871948242,
+      "learning_rate": 0.00029300306596395604,
+      "loss": 3.0816,
+      "step": 31062
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6259018182754517,
+      "learning_rate": 0.00029298767397084927,
+      "loss": 2.8979,
+      "step": 31063
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6216327548027039,
+      "learning_rate": 0.0002929722819962114,
+      "loss": 3.02,
+      "step": 31064
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6106691956520081,
+      "learning_rate": 0.00029295689004008333,
+      "loss": 3.1225,
+      "step": 31065
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6300420761108398,
+      "learning_rate": 0.0002929414981025054,
+      "loss": 3.0543,
+      "step": 31066
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6276101469993591,
+      "learning_rate": 0.00029292610618351814,
+      "loss": 3.0879,
+      "step": 31067
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5739606618881226,
+      "learning_rate": 0.00029291071428316225,
+      "loss": 2.9895,
+      "step": 31068
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6423936486244202,
+      "learning_rate": 0.0002928953224014782,
+      "loss": 2.9886,
+      "step": 31069
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.563739001750946,
+      "learning_rate": 0.00029287993053850645,
+      "loss": 2.9262,
+      "step": 31070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6203590631484985,
+      "learning_rate": 0.0002928645386942875,
+      "loss": 3.1504,
+      "step": 31071
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6195458173751831,
+      "learning_rate": 0.0002928491468688621,
+      "loss": 3.0775,
+      "step": 31072
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6083816289901733,
+      "learning_rate": 0.00029283375506227073,
+      "loss": 3.0681,
+      "step": 31073
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.567173182964325,
+      "learning_rate": 0.0002928183632745537,
+      "loss": 2.9375,
+      "step": 31074
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5811203122138977,
+      "learning_rate": 0.00029280297150575187,
+      "loss": 2.9456,
+      "step": 31075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5663540363311768,
+      "learning_rate": 0.0002927875797559056,
+      "loss": 2.8892,
+      "step": 31076
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5827478766441345,
+      "learning_rate": 0.00029277218802505544,
+      "loss": 2.8407,
+      "step": 31077
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6202985048294067,
+      "learning_rate": 0.0002927567963132419,
+      "loss": 3.1736,
+      "step": 31078
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.601956307888031,
+      "learning_rate": 0.00029274140462050546,
+      "loss": 3.0033,
+      "step": 31079
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6013195514678955,
+      "learning_rate": 0.000292726012946887,
+      "loss": 2.8182,
+      "step": 31080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5847362279891968,
+      "learning_rate": 0.00029271062129242664,
+      "loss": 3.0162,
+      "step": 31081
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5564871430397034,
+      "learning_rate": 0.0002926952296571653,
+      "loss": 3.1941,
+      "step": 31082
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5841847062110901,
+      "learning_rate": 0.0002926798380411431,
+      "loss": 3.0432,
+      "step": 31083
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6471392512321472,
+      "learning_rate": 0.00029266444644440086,
+      "loss": 2.9476,
+      "step": 31084
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5753255486488342,
+      "learning_rate": 0.00029264905486697915,
+      "loss": 2.876,
+      "step": 31085
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6308822631835938,
+      "learning_rate": 0.0002926336633089183,
+      "loss": 2.907,
+      "step": 31086
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6044859886169434,
+      "learning_rate": 0.0002926182717702591,
+      "loss": 3.1926,
+      "step": 31087
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.610822319984436,
+      "learning_rate": 0.0002926028802510418,
+      "loss": 2.9054,
+      "step": 31088
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6061519384384155,
+      "learning_rate": 0.00029258748875130714,
+      "loss": 3.0566,
+      "step": 31089
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6459476351737976,
+      "learning_rate": 0.00029257209727109564,
+      "loss": 2.9014,
+      "step": 31090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6128569841384888,
+      "learning_rate": 0.0002925567058104478,
+      "loss": 2.6935,
+      "step": 31091
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6115437746047974,
+      "learning_rate": 0.0002925413143694042,
+      "loss": 3.3244,
+      "step": 31092
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6120768785476685,
+      "learning_rate": 0.00029252592294800526,
+      "loss": 2.8478,
+      "step": 31093
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5760208368301392,
+      "learning_rate": 0.00029251053154629163,
+      "loss": 2.9368,
+      "step": 31094
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5910771489143372,
+      "learning_rate": 0.0002924951401643039,
+      "loss": 2.9577,
+      "step": 31095
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6013192534446716,
+      "learning_rate": 0.00029247974880208235,
+      "loss": 3.07,
+      "step": 31096
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6419551372528076,
+      "learning_rate": 0.00029246435745966786,
+      "loss": 2.997,
+      "step": 31097
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6099100708961487,
+      "learning_rate": 0.00029244896613710077,
+      "loss": 3.0675,
+      "step": 31098
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5890765190124512,
+      "learning_rate": 0.00029243357483442153,
+      "loss": 2.9062,
+      "step": 31099
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6350886821746826,
+      "learning_rate": 0.00029241818355167095,
+      "loss": 3.1908,
+      "step": 31100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6403512954711914,
+      "learning_rate": 0.00029240279228888926,
+      "loss": 3.0824,
+      "step": 31101
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5865958333015442,
+      "learning_rate": 0.00029238740104611734,
+      "loss": 3.1828,
+      "step": 31102
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6323754191398621,
+      "learning_rate": 0.0002923720098233954,
+      "loss": 3.1062,
+      "step": 31103
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6078318953514099,
+      "learning_rate": 0.0002923566186207641,
+      "loss": 3.2757,
+      "step": 31104
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5803014636039734,
+      "learning_rate": 0.00029234122743826413,
+      "loss": 3.2241,
+      "step": 31105
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5942589044570923,
+      "learning_rate": 0.00029232583627593583,
+      "loss": 3.378,
+      "step": 31106
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6468483209609985,
+      "learning_rate": 0.0002923104451338199,
+      "loss": 3.0655,
+      "step": 31107
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6610134243965149,
+      "learning_rate": 0.00029229505401195656,
+      "loss": 2.9371,
+      "step": 31108
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5979439616203308,
+      "learning_rate": 0.0002922796629103867,
+      "loss": 3.087,
+      "step": 31109
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5915126204490662,
+      "learning_rate": 0.00029226427182915075,
+      "loss": 3.043,
+      "step": 31110
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5519029498100281,
+      "learning_rate": 0.0002922488807682891,
+      "loss": 3.1751,
+      "step": 31111
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5895189642906189,
+      "learning_rate": 0.0002922334897278425,
+      "loss": 3.1257,
+      "step": 31112
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.627914547920227,
+      "learning_rate": 0.00029221809870785137,
+      "loss": 3.2162,
+      "step": 31113
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5837568044662476,
+      "learning_rate": 0.0002922027077083562,
+      "loss": 3.0215,
+      "step": 31114
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5984426736831665,
+      "learning_rate": 0.0002921873167293977,
+      "loss": 2.9769,
+      "step": 31115
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.600309431552887,
+      "learning_rate": 0.0002921719257710163,
+      "loss": 3.3203,
+      "step": 31116
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5865994095802307,
+      "learning_rate": 0.00029215653483325257,
+      "loss": 3.0811,
+      "step": 31117
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5973957777023315,
+      "learning_rate": 0.0002921411439161469,
+      "loss": 3.1921,
+      "step": 31118
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6090832352638245,
+      "learning_rate": 0.00029212575301974,
+      "loss": 3.0701,
+      "step": 31119
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5993704795837402,
+      "learning_rate": 0.00029211036214407243,
+      "loss": 2.9587,
+      "step": 31120
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5656693577766418,
+      "learning_rate": 0.0002920949712891845,
+      "loss": 2.8517,
+      "step": 31121
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6290675401687622,
+      "learning_rate": 0.000292079580455117,
+      "loss": 2.9327,
+      "step": 31122
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6300106048583984,
+      "learning_rate": 0.0002920641896419103,
+      "loss": 3.0199,
+      "step": 31123
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5924240946769714,
+      "learning_rate": 0.00029204879884960495,
+      "loss": 3.042,
+      "step": 31124
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6742143630981445,
+      "learning_rate": 0.0002920334080782417,
+      "loss": 3.1415,
+      "step": 31125
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6075363159179688,
+      "learning_rate": 0.0002920180173278608,
+      "loss": 3.1435,
+      "step": 31126
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6138061881065369,
+      "learning_rate": 0.00029200262659850293,
+      "loss": 2.8322,
+      "step": 31127
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6019613742828369,
+      "learning_rate": 0.0002919872358902086,
+      "loss": 2.9297,
+      "step": 31128
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.584140956401825,
+      "learning_rate": 0.00029197184520301833,
+      "loss": 3.0311,
+      "step": 31129
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6052551865577698,
+      "learning_rate": 0.00029195645453697274,
+      "loss": 3.276,
+      "step": 31130
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5705941915512085,
+      "learning_rate": 0.00029194106389211224,
+      "loss": 3.1659,
+      "step": 31131
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6407089233398438,
+      "learning_rate": 0.0002919256732684775,
+      "loss": 2.7949,
+      "step": 31132
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6144587397575378,
+      "learning_rate": 0.0002919102826661089,
+      "loss": 3.1167,
+      "step": 31133
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5976653099060059,
+      "learning_rate": 0.00029189489208504706,
+      "loss": 2.9931,
+      "step": 31134
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6328117251396179,
+      "learning_rate": 0.0002918795015253326,
+      "loss": 3.169,
+      "step": 31135
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5923229455947876,
+      "learning_rate": 0.00029186411098700585,
+      "loss": 3.1009,
+      "step": 31136
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5932513475418091,
+      "learning_rate": 0.00029184872047010755,
+      "loss": 3.17,
+      "step": 31137
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6198069453239441,
+      "learning_rate": 0.00029183332997467814,
+      "loss": 2.9827,
+      "step": 31138
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6290854215621948,
+      "learning_rate": 0.0002918179395007581,
+      "loss": 2.9822,
+      "step": 31139
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.58577960729599,
+      "learning_rate": 0.00029180254904838816,
+      "loss": 3.062,
+      "step": 31140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5880772471427917,
+      "learning_rate": 0.0002917871586176086,
+      "loss": 3.0266,
+      "step": 31141
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6108989715576172,
+      "learning_rate": 0.0002917717682084602,
+      "loss": 3.0569,
+      "step": 31142
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5859941840171814,
+      "learning_rate": 0.00029175637782098327,
+      "loss": 3.1125,
+      "step": 31143
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5857529044151306,
+      "learning_rate": 0.0002917409874552184,
+      "loss": 3.0528,
+      "step": 31144
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6427134871482849,
+      "learning_rate": 0.0002917255971112064,
+      "loss": 3.147,
+      "step": 31145
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5859655737876892,
+      "learning_rate": 0.00029171020678898745,
+      "loss": 3.154,
+      "step": 31146
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5824823379516602,
+      "learning_rate": 0.0002916948164886023,
+      "loss": 3.174,
+      "step": 31147
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6048821210861206,
+      "learning_rate": 0.0002916794262100913,
+      "loss": 3.0062,
+      "step": 31148
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5758355855941772,
+      "learning_rate": 0.00029166403595349514,
+      "loss": 2.8761,
+      "step": 31149
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5995056629180908,
+      "learning_rate": 0.00029164864571885435,
+      "loss": 2.9244,
+      "step": 31150
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6515876650810242,
+      "learning_rate": 0.00029163325550620925,
+      "loss": 2.7845,
+      "step": 31151
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5933794975280762,
+      "learning_rate": 0.00029161786531560073,
+      "loss": 3.0369,
+      "step": 31152
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6130641102790833,
+      "learning_rate": 0.0002916024751470691,
+      "loss": 2.8529,
+      "step": 31153
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6391645669937134,
+      "learning_rate": 0.00029158708500065495,
+      "loss": 3.2121,
+      "step": 31154
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6190594434738159,
+      "learning_rate": 0.0002915716948763987,
+      "loss": 2.8578,
+      "step": 31155
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6114823818206787,
+      "learning_rate": 0.000291556304774341,
+      "loss": 3.2051,
+      "step": 31156
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5992326736450195,
+      "learning_rate": 0.0002915409146945225,
+      "loss": 2.9415,
+      "step": 31157
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6097759008407593,
+      "learning_rate": 0.0002915255246369835,
+      "loss": 2.8587,
+      "step": 31158
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5706651210784912,
+      "learning_rate": 0.00029151013460176466,
+      "loss": 3.1365,
+      "step": 31159
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5817975997924805,
+      "learning_rate": 0.0002914947445889065,
+      "loss": 2.8617,
+      "step": 31160
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5653377771377563,
+      "learning_rate": 0.00029147935459844943,
+      "loss": 2.9423,
+      "step": 31161
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6266032457351685,
+      "learning_rate": 0.0002914639646304343,
+      "loss": 2.8249,
+      "step": 31162
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6154088973999023,
+      "learning_rate": 0.00029144857468490126,
+      "loss": 3.3106,
+      "step": 31163
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5928178429603577,
+      "learning_rate": 0.00029143318476189113,
+      "loss": 3.1093,
+      "step": 31164
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6203148365020752,
+      "learning_rate": 0.00029141779486144433,
+      "loss": 3.0661,
+      "step": 31165
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5982689261436462,
+      "learning_rate": 0.0002914024049836013,
+      "loss": 2.9455,
+      "step": 31166
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5757526755332947,
+      "learning_rate": 0.00029138701512840286,
+      "loss": 3.1113,
+      "step": 31167
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5914579033851624,
+      "learning_rate": 0.00029137162529588924,
+      "loss": 2.9128,
+      "step": 31168
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6134317517280579,
+      "learning_rate": 0.00029135623548610126,
+      "loss": 2.9946,
+      "step": 31169
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5648379921913147,
+      "learning_rate": 0.0002913408456990791,
+      "loss": 3.4179,
+      "step": 31170
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6595600247383118,
+      "learning_rate": 0.00029132545593486353,
+      "loss": 3.0651,
+      "step": 31171
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5977067351341248,
+      "learning_rate": 0.00029131006619349516,
+      "loss": 2.9647,
+      "step": 31172
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5756570100784302,
+      "learning_rate": 0.00029129467647501424,
+      "loss": 3.0136,
+      "step": 31173
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5817214250564575,
+      "learning_rate": 0.0002912792867794616,
+      "loss": 3.013,
+      "step": 31174
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.628970742225647,
+      "learning_rate": 0.00029126389710687755,
+      "loss": 2.8144,
+      "step": 31175
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5823667049407959,
+      "learning_rate": 0.00029124850745730266,
+      "loss": 3.2412,
+      "step": 31176
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6125685572624207,
+      "learning_rate": 0.0002912331178307777,
+      "loss": 3.0372,
+      "step": 31177
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6078442335128784,
+      "learning_rate": 0.0002912177282273429,
+      "loss": 3.1155,
+      "step": 31178
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5867834091186523,
+      "learning_rate": 0.000291202338647039,
+      "loss": 2.7659,
+      "step": 31179
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.605891227722168,
+      "learning_rate": 0.0002911869490899063,
+      "loss": 3.1373,
+      "step": 31180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6134265065193176,
+      "learning_rate": 0.0002911715595559855,
+      "loss": 3.0259,
+      "step": 31181
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6155909299850464,
+      "learning_rate": 0.0002911561700453173,
+      "loss": 3.0505,
+      "step": 31182
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6217151284217834,
+      "learning_rate": 0.00029114078055794186,
+      "loss": 3.1017,
+      "step": 31183
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6171595454216003,
+      "learning_rate": 0.00029112539109390004,
+      "loss": 3.1822,
+      "step": 31184
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5772479176521301,
+      "learning_rate": 0.00029111000165323205,
+      "loss": 2.9001,
+      "step": 31185
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5680612921714783,
+      "learning_rate": 0.0002910946122359787,
+      "loss": 3.0623,
+      "step": 31186
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5654654502868652,
+      "learning_rate": 0.0002910792228421805,
+      "loss": 3.1616,
+      "step": 31187
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5805103182792664,
+      "learning_rate": 0.00029106383347187777,
+      "loss": 3.1398,
+      "step": 31188
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5948848724365234,
+      "learning_rate": 0.0002910484441251113,
+      "loss": 2.8847,
+      "step": 31189
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6186020970344543,
+      "learning_rate": 0.00029103305480192147,
+      "loss": 3.1687,
+      "step": 31190
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5708625316619873,
+      "learning_rate": 0.0002910176655023488,
+      "loss": 3.048,
+      "step": 31191
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5748517513275146,
+      "learning_rate": 0.00029100227622643393,
+      "loss": 3.2699,
+      "step": 31192
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6108618378639221,
+      "learning_rate": 0.0002909868869742173,
+      "loss": 3.1266,
+      "step": 31193
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5868430733680725,
+      "learning_rate": 0.0002909714977457396,
+      "loss": 2.9538,
+      "step": 31194
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5879116058349609,
+      "learning_rate": 0.00029095610854104105,
+      "loss": 3.023,
+      "step": 31195
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6066574454307556,
+      "learning_rate": 0.0002909407193601624,
+      "loss": 3.202,
+      "step": 31196
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5830428004264832,
+      "learning_rate": 0.00029092533020314427,
+      "loss": 3.1226,
+      "step": 31197
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6408023238182068,
+      "learning_rate": 0.0002909099410700269,
+      "loss": 3.0443,
+      "step": 31198
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6012759804725647,
+      "learning_rate": 0.0002908945519608512,
+      "loss": 3.0593,
+      "step": 31199
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6043107509613037,
+      "learning_rate": 0.00029087916287565737,
+      "loss": 3.059,
+      "step": 31200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.590762197971344,
+      "learning_rate": 0.00029086377381448605,
+      "loss": 3.1015,
+      "step": 31201
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5998045206069946,
+      "learning_rate": 0.00029084838477737784,
+      "loss": 2.8935,
+      "step": 31202
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6675084233283997,
+      "learning_rate": 0.0002908329957643732,
+      "loss": 3.0359,
+      "step": 31203
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5763409733772278,
+      "learning_rate": 0.0002908176067755127,
+      "loss": 2.9924,
+      "step": 31204
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5802990794181824,
+      "learning_rate": 0.0002908022178108369,
+      "loss": 3.0601,
+      "step": 31205
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.592325747013092,
+      "learning_rate": 0.0002907868288703862,
+      "loss": 3.0388,
+      "step": 31206
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5971692800521851,
+      "learning_rate": 0.0002907714399542013,
+      "loss": 2.8794,
+      "step": 31207
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6179129481315613,
+      "learning_rate": 0.00029075605106232264,
+      "loss": 2.9758,
+      "step": 31208
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.648285984992981,
+      "learning_rate": 0.0002907406621947908,
+      "loss": 2.772,
+      "step": 31209
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6502874493598938,
+      "learning_rate": 0.00029072527335164615,
+      "loss": 2.8967,
+      "step": 31210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6211252212524414,
+      "learning_rate": 0.0002907098845329294,
+      "loss": 2.8207,
+      "step": 31211
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6525001525878906,
+      "learning_rate": 0.0002906944957386811,
+      "loss": 2.9625,
+      "step": 31212
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6617982983589172,
+      "learning_rate": 0.0002906791069689416,
+      "loss": 3.0828,
+      "step": 31213
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6012643575668335,
+      "learning_rate": 0.0002906637182237516,
+      "loss": 2.9603,
+      "step": 31214
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5715975165367126,
+      "learning_rate": 0.00029064832950315156,
+      "loss": 3.043,
+      "step": 31215
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6091911196708679,
+      "learning_rate": 0.0002906329408071819,
+      "loss": 2.9976,
+      "step": 31216
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6170886158943176,
+      "learning_rate": 0.00029061755213588347,
+      "loss": 2.9807,
+      "step": 31217
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5898032188415527,
+      "learning_rate": 0.00029060216348929653,
+      "loss": 3.1002,
+      "step": 31218
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6269931197166443,
+      "learning_rate": 0.0002905867748674617,
+      "loss": 3.2008,
+      "step": 31219
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6010162830352783,
+      "learning_rate": 0.0002905713862704194,
+      "loss": 2.7986,
+      "step": 31220
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.664634644985199,
+      "learning_rate": 0.00029055599769821036,
+      "loss": 3.1664,
+      "step": 31221
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5801214575767517,
+      "learning_rate": 0.00029054060915087497,
+      "loss": 2.9423,
+      "step": 31222
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5644407272338867,
+      "learning_rate": 0.00029052522062845374,
+      "loss": 3.0543,
+      "step": 31223
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5935441255569458,
+      "learning_rate": 0.0002905098321309873,
+      "loss": 3.0877,
+      "step": 31224
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6010169982910156,
+      "learning_rate": 0.0002904944436585161,
+      "loss": 3.1078,
+      "step": 31225
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6892803311347961,
+      "learning_rate": 0.0002904790552110808,
+      "loss": 3.0326,
+      "step": 31226
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5705881118774414,
+      "learning_rate": 0.00029046366678872176,
+      "loss": 2.9728,
+      "step": 31227
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6575391292572021,
+      "learning_rate": 0.00029044827839147953,
+      "loss": 3.0386,
+      "step": 31228
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5752348303794861,
+      "learning_rate": 0.00029043289001939485,
+      "loss": 3.0842,
+      "step": 31229
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5966557860374451,
+      "learning_rate": 0.000290417501672508,
+      "loss": 3.0958,
+      "step": 31230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6041060090065002,
+      "learning_rate": 0.00029040211335085956,
+      "loss": 2.974,
+      "step": 31231
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.590144693851471,
+      "learning_rate": 0.00029038672505449025,
+      "loss": 3.0159,
+      "step": 31232
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6653158068656921,
+      "learning_rate": 0.00029037133678344037,
+      "loss": 2.6909,
+      "step": 31233
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6356053948402405,
+      "learning_rate": 0.00029035594853775063,
+      "loss": 2.886,
+      "step": 31234
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.616893470287323,
+      "learning_rate": 0.00029034056031746125,
+      "loss": 3.085,
+      "step": 31235
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5892876386642456,
+      "learning_rate": 0.0002903251721226132,
+      "loss": 3.0894,
+      "step": 31236
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5598612427711487,
+      "learning_rate": 0.0002903097839532467,
+      "loss": 3.0063,
+      "step": 31237
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6560896039009094,
+      "learning_rate": 0.00029029439580940226,
+      "loss": 3.0115,
+      "step": 31238
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5958468317985535,
+      "learning_rate": 0.00029027900769112064,
+      "loss": 3.0396,
+      "step": 31239
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5762074589729309,
+      "learning_rate": 0.00029026361959844224,
+      "loss": 3.3083,
+      "step": 31240
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5689447522163391,
+      "learning_rate": 0.0002902482315314076,
+      "loss": 3.2403,
+      "step": 31241
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5754348039627075,
+      "learning_rate": 0.0002902328434900571,
+      "loss": 3.07,
+      "step": 31242
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6313903331756592,
+      "learning_rate": 0.00029021745547443145,
+      "loss": 3.122,
+      "step": 31243
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6057738065719604,
+      "learning_rate": 0.00029020206748457135,
+      "loss": 3.025,
+      "step": 31244
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6703782677650452,
+      "learning_rate": 0.0002901866795205169,
+      "loss": 3.1737,
+      "step": 31245
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5860815048217773,
+      "learning_rate": 0.000290171291582309,
+      "loss": 3.0211,
+      "step": 31246
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5635547041893005,
+      "learning_rate": 0.00029015590366998785,
+      "loss": 3.2277,
+      "step": 31247
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.598173201084137,
+      "learning_rate": 0.00029014051578359426,
+      "loss": 2.9847,
+      "step": 31248
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6276450157165527,
+      "learning_rate": 0.0002901251279231687,
+      "loss": 3.1291,
+      "step": 31249
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5984366536140442,
+      "learning_rate": 0.00029010974008875155,
+      "loss": 3.23,
+      "step": 31250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5806760191917419,
+      "learning_rate": 0.00029009435228038354,
+      "loss": 2.9694,
+      "step": 31251
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6110997796058655,
+      "learning_rate": 0.00029007896449810506,
+      "loss": 3.054,
+      "step": 31252
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5997822880744934,
+      "learning_rate": 0.00029006357674195657,
+      "loss": 3.1742,
+      "step": 31253
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6100072860717773,
+      "learning_rate": 0.00029004818901197886,
+      "loss": 2.9635,
+      "step": 31254
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5825597047805786,
+      "learning_rate": 0.00029003280130821224,
+      "loss": 3.119,
+      "step": 31255
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.636887788772583,
+      "learning_rate": 0.00029001741363069735,
+      "loss": 3.1825,
+      "step": 31256
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5912345051765442,
+      "learning_rate": 0.00029000202597947457,
+      "loss": 3.078,
+      "step": 31257
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5856207609176636,
+      "learning_rate": 0.0002899866383545846,
+      "loss": 3.2083,
+      "step": 31258
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6235297322273254,
+      "learning_rate": 0.0002899712507560679,
+      "loss": 3.106,
+      "step": 31259
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6140345335006714,
+      "learning_rate": 0.00028995586318396495,
+      "loss": 3.2502,
+      "step": 31260
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5801708698272705,
+      "learning_rate": 0.0002899404756383164,
+      "loss": 3.1254,
+      "step": 31261
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5877318382263184,
+      "learning_rate": 0.0002899250881191627,
+      "loss": 2.9534,
+      "step": 31262
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5814765095710754,
+      "learning_rate": 0.0002899097006265442,
+      "loss": 2.9297,
+      "step": 31263
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6114090085029602,
+      "learning_rate": 0.00028989431316050183,
+      "loss": 2.9979,
+      "step": 31264
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.601153552532196,
+      "learning_rate": 0.00028987892572107573,
+      "loss": 2.9141,
+      "step": 31265
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6009148955345154,
+      "learning_rate": 0.00028986353830830674,
+      "loss": 2.9535,
+      "step": 31266
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5960407257080078,
+      "learning_rate": 0.0002898481509222352,
+      "loss": 3.208,
+      "step": 31267
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6321334838867188,
+      "learning_rate": 0.00028983276356290156,
+      "loss": 2.9113,
+      "step": 31268
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6519079804420471,
+      "learning_rate": 0.00028981737623034657,
+      "loss": 2.9775,
+      "step": 31269
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.652167797088623,
+      "learning_rate": 0.0002898019889246106,
+      "loss": 3.239,
+      "step": 31270
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5852517485618591,
+      "learning_rate": 0.00028978660164573437,
+      "loss": 2.943,
+      "step": 31271
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5924550294876099,
+      "learning_rate": 0.0002897712143937581,
+      "loss": 3.0424,
+      "step": 31272
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5765801072120667,
+      "learning_rate": 0.0002897558271687225,
+      "loss": 3.1681,
+      "step": 31273
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6032983660697937,
+      "learning_rate": 0.0002897404399706682,
+      "loss": 3.0438,
+      "step": 31274
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6318213939666748,
+      "learning_rate": 0.0002897250527996355,
+      "loss": 3.053,
+      "step": 31275
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6157848238945007,
+      "learning_rate": 0.0002897096656556651,
+      "loss": 2.8284,
+      "step": 31276
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6649987697601318,
+      "learning_rate": 0.0002896942785387974,
+      "loss": 3.1921,
+      "step": 31277
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5770165920257568,
+      "learning_rate": 0.0002896788914490729,
+      "loss": 3.0402,
+      "step": 31278
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5739706158638,
+      "learning_rate": 0.0002896635043865324,
+      "loss": 3.0012,
+      "step": 31279
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6319332122802734,
+      "learning_rate": 0.00028964811735121614,
+      "loss": 3.2183,
+      "step": 31280
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6746277809143066,
+      "learning_rate": 0.00028963273034316486,
+      "loss": 3.0346,
+      "step": 31281
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5536617636680603,
+      "learning_rate": 0.0002896173433624188,
+      "loss": 3.1045,
+      "step": 31282
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6022813320159912,
+      "learning_rate": 0.0002896019564090187,
+      "loss": 2.8612,
+      "step": 31283
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6644525527954102,
+      "learning_rate": 0.0002895865694830052,
+      "loss": 2.758,
+      "step": 31284
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6138014197349548,
+      "learning_rate": 0.0002895711825844185,
+      "loss": 3.1694,
+      "step": 31285
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5913397669792175,
+      "learning_rate": 0.0002895557957132994,
+      "loss": 2.8191,
+      "step": 31286
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5768195390701294,
+      "learning_rate": 0.0002895404088696882,
+      "loss": 2.8873,
+      "step": 31287
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6131641268730164,
+      "learning_rate": 0.0002895250220536256,
+      "loss": 3.1928,
+      "step": 31288
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5838510394096375,
+      "learning_rate": 0.00028950963526515214,
+      "loss": 3.0279,
+      "step": 31289
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6200519800186157,
+      "learning_rate": 0.00028949424850430817,
+      "loss": 3.0139,
+      "step": 31290
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.60508131980896,
+      "learning_rate": 0.00028947886177113446,
+      "loss": 3.1849,
+      "step": 31291
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6125930547714233,
+      "learning_rate": 0.0002894634750656714,
+      "loss": 3.0983,
+      "step": 31292
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5791587829589844,
+      "learning_rate": 0.0002894480883879593,
+      "loss": 3.1379,
+      "step": 31293
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6140015721321106,
+      "learning_rate": 0.0002894327017380392,
+      "loss": 3.2646,
+      "step": 31294
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7361003756523132,
+      "learning_rate": 0.00028941731511595117,
+      "loss": 3.1597,
+      "step": 31295
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5974676012992859,
+      "learning_rate": 0.00028940192852173596,
+      "loss": 3.1636,
+      "step": 31296
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6096514463424683,
+      "learning_rate": 0.00028938654195543396,
+      "loss": 2.9766,
+      "step": 31297
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6190111637115479,
+      "learning_rate": 0.0002893711554170858,
+      "loss": 2.8579,
+      "step": 31298
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6046838164329529,
+      "learning_rate": 0.00028935576890673205,
+      "loss": 3.1151,
+      "step": 31299
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6145317554473877,
+      "learning_rate": 0.000289340382424413,
+      "loss": 3.1992,
+      "step": 31300
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6103838086128235,
+      "learning_rate": 0.00028932499597016944,
+      "loss": 3.0945,
+      "step": 31301
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.600480854511261,
+      "learning_rate": 0.0002893096095440418,
+      "loss": 2.9785,
+      "step": 31302
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5839743614196777,
+      "learning_rate": 0.0002892942231460704,
+      "loss": 3.1328,
+      "step": 31303
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6166554689407349,
+      "learning_rate": 0.0002892788367762962,
+      "loss": 3.0722,
+      "step": 31304
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6772105693817139,
+      "learning_rate": 0.0002892634504347593,
+      "loss": 2.8744,
+      "step": 31305
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6089929938316345,
+      "learning_rate": 0.00028924806412150057,
+      "loss": 3.1004,
+      "step": 31306
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5962395668029785,
+      "learning_rate": 0.00028923267783656026,
+      "loss": 2.9843,
+      "step": 31307
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6941888928413391,
+      "learning_rate": 0.00028921729157997894,
+      "loss": 3.0849,
+      "step": 31308
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.607742965221405,
+      "learning_rate": 0.0002892019053517974,
+      "loss": 3.2729,
+      "step": 31309
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5975240468978882,
+      "learning_rate": 0.00028918651915205585,
+      "loss": 3.1408,
+      "step": 31310
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5710057616233826,
+      "learning_rate": 0.000289171132980795,
+      "loss": 3.3382,
+      "step": 31311
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6463126540184021,
+      "learning_rate": 0.00028915574683805514,
+      "loss": 2.967,
+      "step": 31312
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7494412064552307,
+      "learning_rate": 0.00028914036072387716,
+      "loss": 2.9294,
+      "step": 31313
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5747896432876587,
+      "learning_rate": 0.00028912497463830125,
+      "loss": 3.2201,
+      "step": 31314
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5977765917778015,
+      "learning_rate": 0.00028910958858136796,
+      "loss": 3.0647,
+      "step": 31315
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6861127614974976,
+      "learning_rate": 0.00028909420255311815,
+      "loss": 3.0568,
+      "step": 31316
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6028975248336792,
+      "learning_rate": 0.00028907881655359195,
+      "loss": 3.0533,
+      "step": 31317
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5990371108055115,
+      "learning_rate": 0.0002890634305828302,
+      "loss": 3.0781,
+      "step": 31318
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6500790119171143,
+      "learning_rate": 0.0002890480446408731,
+      "loss": 2.9856,
+      "step": 31319
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6449432969093323,
+      "learning_rate": 0.0002890326587277614,
+      "loss": 2.9427,
+      "step": 31320
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6211156845092773,
+      "learning_rate": 0.0002890172728435356,
+      "loss": 2.9741,
+      "step": 31321
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6242359280586243,
+      "learning_rate": 0.0002890018869882361,
+      "loss": 3.1796,
+      "step": 31322
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6206004619598389,
+      "learning_rate": 0.00028898650116190367,
+      "loss": 3.1543,
+      "step": 31323
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5952155590057373,
+      "learning_rate": 0.0002889711153645786,
+      "loss": 2.88,
+      "step": 31324
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6566175222396851,
+      "learning_rate": 0.00028895572959630137,
+      "loss": 3.3069,
+      "step": 31325
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5943531394004822,
+      "learning_rate": 0.00028894034385711277,
+      "loss": 3.0009,
+      "step": 31326
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6611933708190918,
+      "learning_rate": 0.0002889249581470531,
+      "loss": 3.1812,
+      "step": 31327
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6006651520729065,
+      "learning_rate": 0.000288909572466163,
+      "loss": 3.2205,
+      "step": 31328
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6034921407699585,
+      "learning_rate": 0.000288894186814483,
+      "loss": 3.079,
+      "step": 31329
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5508182644844055,
+      "learning_rate": 0.0002888788011920534,
+      "loss": 2.9034,
+      "step": 31330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.596584141254425,
+      "learning_rate": 0.0002888634155989151,
+      "loss": 3.1913,
+      "step": 31331
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5692295432090759,
+      "learning_rate": 0.0002888480300351083,
+      "loss": 3.1871,
+      "step": 31332
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6114552617073059,
+      "learning_rate": 0.0002888326445006738,
+      "loss": 2.9601,
+      "step": 31333
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6093324422836304,
+      "learning_rate": 0.00028881725899565177,
+      "loss": 3.0075,
+      "step": 31334
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.608741044998169,
+      "learning_rate": 0.00028880187352008297,
+      "loss": 3.0645,
+      "step": 31335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.598630428314209,
+      "learning_rate": 0.00028878648807400794,
+      "loss": 3.1638,
+      "step": 31336
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6114171743392944,
+      "learning_rate": 0.00028877110265746706,
+      "loss": 2.9811,
+      "step": 31337
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6134542226791382,
+      "learning_rate": 0.0002887557172705011,
+      "loss": 3.0926,
+      "step": 31338
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.61723792552948,
+      "learning_rate": 0.00028874033191315023,
+      "loss": 3.0028,
+      "step": 31339
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5689911842346191,
+      "learning_rate": 0.00028872494658545517,
+      "loss": 3.1265,
+      "step": 31340
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6355405449867249,
+      "learning_rate": 0.00028870956128745657,
+      "loss": 3.0033,
+      "step": 31341
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6212851405143738,
+      "learning_rate": 0.00028869417601919475,
+      "loss": 3.1319,
+      "step": 31342
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.600463330745697,
+      "learning_rate": 0.00028867879078071034,
+      "loss": 3.0882,
+      "step": 31343
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6733449101448059,
+      "learning_rate": 0.0002886634055720437,
+      "loss": 2.8782,
+      "step": 31344
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6149190068244934,
+      "learning_rate": 0.00028864802039323546,
+      "loss": 3.1114,
+      "step": 31345
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6464491486549377,
+      "learning_rate": 0.0002886326352443263,
+      "loss": 2.9551,
+      "step": 31346
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5949641466140747,
+      "learning_rate": 0.0002886172501253564,
+      "loss": 3.2199,
+      "step": 31347
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6497898697853088,
+      "learning_rate": 0.0002886018650363667,
+      "loss": 2.9258,
+      "step": 31348
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6072144508361816,
+      "learning_rate": 0.0002885864799773973,
+      "loss": 2.9873,
+      "step": 31349
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5734956860542297,
+      "learning_rate": 0.00028857109494848896,
+      "loss": 2.9834,
+      "step": 31350
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6253546476364136,
+      "learning_rate": 0.00028855570994968223,
+      "loss": 2.9745,
+      "step": 31351
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5898392796516418,
+      "learning_rate": 0.00028854032498101746,
+      "loss": 2.8968,
+      "step": 31352
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6153894662857056,
+      "learning_rate": 0.00028852494004253537,
+      "loss": 2.9332,
+      "step": 31353
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6599181890487671,
+      "learning_rate": 0.00028850955513427633,
+      "loss": 3.1384,
+      "step": 31354
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5734416842460632,
+      "learning_rate": 0.00028849417025628084,
+      "loss": 3.0424,
+      "step": 31355
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5493830442428589,
+      "learning_rate": 0.0002884787854085896,
+      "loss": 3.0597,
+      "step": 31356
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6523565649986267,
+      "learning_rate": 0.000288463400591243,
+      "loss": 3.3818,
+      "step": 31357
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6114045977592468,
+      "learning_rate": 0.00028844801580428167,
+      "loss": 2.8229,
+      "step": 31358
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5856912136077881,
+      "learning_rate": 0.00028843263104774583,
+      "loss": 3.0191,
+      "step": 31359
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6086015105247498,
+      "learning_rate": 0.00028841724632167634,
+      "loss": 3.0111,
+      "step": 31360
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5889554619789124,
+      "learning_rate": 0.00028840186162611364,
+      "loss": 3.1478,
+      "step": 31361
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6440091133117676,
+      "learning_rate": 0.00028838647696109807,
+      "loss": 3.1964,
+      "step": 31362
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6442698240280151,
+      "learning_rate": 0.0002883710923266704,
+      "loss": 3.0283,
+      "step": 31363
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6014328002929688,
+      "learning_rate": 0.000288355707722871,
+      "loss": 3.1843,
+      "step": 31364
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5934261679649353,
+      "learning_rate": 0.0002883403231497403,
+      "loss": 2.9702,
+      "step": 31365
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6464361548423767,
+      "learning_rate": 0.00028832493860731916,
+      "loss": 3.0837,
+      "step": 31366
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5907967686653137,
+      "learning_rate": 0.0002883095540956477,
+      "loss": 2.9704,
+      "step": 31367
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5807782411575317,
+      "learning_rate": 0.0002882941696147668,
+      "loss": 3.0911,
+      "step": 31368
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5743398666381836,
+      "learning_rate": 0.0002882787851647167,
+      "loss": 2.9713,
+      "step": 31369
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5994200110435486,
+      "learning_rate": 0.00028826340074553797,
+      "loss": 2.9487,
+      "step": 31370
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6094851493835449,
+      "learning_rate": 0.0002882480163572713,
+      "loss": 2.9808,
+      "step": 31371
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6090477108955383,
+      "learning_rate": 0.00028823263199995706,
+      "loss": 3.0441,
+      "step": 31372
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5785759091377258,
+      "learning_rate": 0.00028821724767363584,
+      "loss": 3.0751,
+      "step": 31373
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5603683590888977,
+      "learning_rate": 0.000288201863378348,
+      "loss": 3.0937,
+      "step": 31374
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5782989263534546,
+      "learning_rate": 0.0002881864791141342,
+      "loss": 2.9813,
+      "step": 31375
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6115820407867432,
+      "learning_rate": 0.0002881710948810351,
+      "loss": 2.9611,
+      "step": 31376
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6622043251991272,
+      "learning_rate": 0.00028815571067909083,
+      "loss": 3.142,
+      "step": 31377
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6727275252342224,
+      "learning_rate": 0.0002881403265083423,
+      "loss": 3.1263,
+      "step": 31378
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6339889764785767,
+      "learning_rate": 0.0002881249423688299,
+      "loss": 3.0685,
+      "step": 31379
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6273970007896423,
+      "learning_rate": 0.000288109558260594,
+      "loss": 3.2247,
+      "step": 31380
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6286791563034058,
+      "learning_rate": 0.0002880941741836753,
+      "loss": 3.0735,
+      "step": 31381
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5971516370773315,
+      "learning_rate": 0.0002880787901381142,
+      "loss": 3.1196,
+      "step": 31382
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5588302612304688,
+      "learning_rate": 0.0002880634061239514,
+      "loss": 2.7368,
+      "step": 31383
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6372582912445068,
+      "learning_rate": 0.0002880480221412271,
+      "loss": 3.0035,
+      "step": 31384
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6087130904197693,
+      "learning_rate": 0.0002880326381899821,
+      "loss": 3.1687,
+      "step": 31385
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6057910919189453,
+      "learning_rate": 0.00028801725427025687,
+      "loss": 2.9205,
+      "step": 31386
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5742743015289307,
+      "learning_rate": 0.0002880018703820918,
+      "loss": 3.0471,
+      "step": 31387
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5914288759231567,
+      "learning_rate": 0.0002879864865255276,
+      "loss": 3.14,
+      "step": 31388
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5800380706787109,
+      "learning_rate": 0.00028797110270060457,
+      "loss": 2.8114,
+      "step": 31389
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5977968573570251,
+      "learning_rate": 0.0002879557189073635,
+      "loss": 3.0108,
+      "step": 31390
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6388533711433411,
+      "learning_rate": 0.0002879403351458446,
+      "loss": 2.8719,
+      "step": 31391
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5507797002792358,
+      "learning_rate": 0.0002879249514160885,
+      "loss": 3.0939,
+      "step": 31392
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5805448293685913,
+      "learning_rate": 0.0002879095677181359,
+      "loss": 2.9328,
+      "step": 31393
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5536087155342102,
+      "learning_rate": 0.0002878941840520271,
+      "loss": 3.1754,
+      "step": 31394
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5905557870864868,
+      "learning_rate": 0.00028787880041780277,
+      "loss": 3.0814,
+      "step": 31395
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6131752729415894,
+      "learning_rate": 0.0002878634168155032,
+      "loss": 2.9661,
+      "step": 31396
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6142122149467468,
+      "learning_rate": 0.0002878480332451691,
+      "loss": 3.0028,
+      "step": 31397
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6468976140022278,
+      "learning_rate": 0.0002878326497068411,
+      "loss": 2.9483,
+      "step": 31398
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5881075859069824,
+      "learning_rate": 0.0002878172662005593,
+      "loss": 2.9752,
+      "step": 31399
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5991531610488892,
+      "learning_rate": 0.0002878018827263647,
+      "loss": 2.8782,
+      "step": 31400
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.589621365070343,
+      "learning_rate": 0.00028778649928429744,
+      "loss": 3.1235,
+      "step": 31401
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5604541301727295,
+      "learning_rate": 0.0002877711158743982,
+      "loss": 3.1901,
+      "step": 31402
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6096484661102295,
+      "learning_rate": 0.0002877557324967076,
+      "loss": 3.0457,
+      "step": 31403
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.645102858543396,
+      "learning_rate": 0.000287740349151266,
+      "loss": 2.9014,
+      "step": 31404
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6197732090950012,
+      "learning_rate": 0.00028772496583811397,
+      "loss": 2.9379,
+      "step": 31405
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5692452788352966,
+      "learning_rate": 0.00028770958255729194,
+      "loss": 3.0261,
+      "step": 31406
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6242355704307556,
+      "learning_rate": 0.00028769419930884044,
+      "loss": 2.9257,
+      "step": 31407
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.578525722026825,
+      "learning_rate": 0.0002876788160928003,
+      "loss": 3.1085,
+      "step": 31408
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6010854840278625,
+      "learning_rate": 0.0002876634329092116,
+      "loss": 3.0628,
+      "step": 31409
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5785937309265137,
+      "learning_rate": 0.00028764804975811516,
+      "loss": 2.8512,
+      "step": 31410
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6024236083030701,
+      "learning_rate": 0.00028763266663955125,
+      "loss": 2.9083,
+      "step": 31411
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5748933553695679,
+      "learning_rate": 0.00028761728355356056,
+      "loss": 2.9888,
+      "step": 31412
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5996086597442627,
+      "learning_rate": 0.00028760190050018366,
+      "loss": 3.0,
+      "step": 31413
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6583248376846313,
+      "learning_rate": 0.0002875865174794608,
+      "loss": 2.8246,
+      "step": 31414
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.64971524477005,
+      "learning_rate": 0.00028757113449143283,
+      "loss": 2.9814,
+      "step": 31415
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6212257146835327,
+      "learning_rate": 0.00028755575153614,
+      "loss": 2.9499,
+      "step": 31416
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6192092895507812,
+      "learning_rate": 0.0002875403686136229,
+      "loss": 3.067,
+      "step": 31417
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.588834285736084,
+      "learning_rate": 0.0002875249857239222,
+      "loss": 3.134,
+      "step": 31418
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5852153897285461,
+      "learning_rate": 0.0002875096028670782,
+      "loss": 3.1103,
+      "step": 31419
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6158280372619629,
+      "learning_rate": 0.00028749422004313157,
+      "loss": 2.9628,
+      "step": 31420
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6489527225494385,
+      "learning_rate": 0.0002874788372521226,
+      "loss": 3.1299,
+      "step": 31421
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5668579339981079,
+      "learning_rate": 0.00028746345449409206,
+      "loss": 2.8412,
+      "step": 31422
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6310492753982544,
+      "learning_rate": 0.0002874480717690804,
+      "loss": 3.1859,
+      "step": 31423
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5989445447921753,
+      "learning_rate": 0.00028743268907712805,
+      "loss": 3.0767,
+      "step": 31424
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5920668244361877,
+      "learning_rate": 0.0002874173064182756,
+      "loss": 2.8711,
+      "step": 31425
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5631740093231201,
+      "learning_rate": 0.00028740192379256356,
+      "loss": 3.1315,
+      "step": 31426
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.613795816898346,
+      "learning_rate": 0.00028738654120003237,
+      "loss": 2.8111,
+      "step": 31427
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6500447988510132,
+      "learning_rate": 0.0002873711586407227,
+      "loss": 3.0035,
+      "step": 31428
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.595923662185669,
+      "learning_rate": 0.00028735577611467485,
+      "loss": 3.0004,
+      "step": 31429
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5796763896942139,
+      "learning_rate": 0.00028734039362192956,
+      "loss": 3.0754,
+      "step": 31430
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6116151213645935,
+      "learning_rate": 0.0002873250111625272,
+      "loss": 3.1244,
+      "step": 31431
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6706094741821289,
+      "learning_rate": 0.00028730962873650825,
+      "loss": 3.3772,
+      "step": 31432
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5975269079208374,
+      "learning_rate": 0.0002872942463439134,
+      "loss": 3.1309,
+      "step": 31433
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5869545340538025,
+      "learning_rate": 0.000287278863984783,
+      "loss": 3.2301,
+      "step": 31434
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5768601298332214,
+      "learning_rate": 0.0002872634816591577,
+      "loss": 2.9476,
+      "step": 31435
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.614723265171051,
+      "learning_rate": 0.0002872480993670778,
+      "loss": 3.1834,
+      "step": 31436
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5859989523887634,
+      "learning_rate": 0.00028723271710858405,
+      "loss": 3.0171,
+      "step": 31437
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6272194981575012,
+      "learning_rate": 0.0002872173348837169,
+      "loss": 3.0156,
+      "step": 31438
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6279314756393433,
+      "learning_rate": 0.0002872019526925167,
+      "loss": 2.7363,
+      "step": 31439
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5947081446647644,
+      "learning_rate": 0.00028718657053502427,
+      "loss": 3.1848,
+      "step": 31440
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.619102954864502,
+      "learning_rate": 0.0002871711884112798,
+      "loss": 2.929,
+      "step": 31441
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6086522936820984,
+      "learning_rate": 0.00028715580632132393,
+      "loss": 3.2284,
+      "step": 31442
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6687970757484436,
+      "learning_rate": 0.0002871404242651973,
+      "loss": 2.9145,
+      "step": 31443
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5713397860527039,
+      "learning_rate": 0.0002871250422429403,
+      "loss": 2.8666,
+      "step": 31444
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5800654292106628,
+      "learning_rate": 0.00028710966025459353,
+      "loss": 3.0257,
+      "step": 31445
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6090438365936279,
+      "learning_rate": 0.00028709427830019723,
+      "loss": 2.9861,
+      "step": 31446
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5880951285362244,
+      "learning_rate": 0.00028707889637979224,
+      "loss": 3.0089,
+      "step": 31447
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6175966858863831,
+      "learning_rate": 0.000287063514493419,
+      "loss": 3.0327,
+      "step": 31448
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.602347195148468,
+      "learning_rate": 0.0002870481326411179,
+      "loss": 3.0266,
+      "step": 31449
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6162763833999634,
+      "learning_rate": 0.0002870327508229296,
+      "loss": 3.1335,
+      "step": 31450
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6215873956680298,
+      "learning_rate": 0.0002870173690388944,
+      "loss": 3.0356,
+      "step": 31451
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6079941391944885,
+      "learning_rate": 0.000287001987289053,
+      "loss": 3.0005,
+      "step": 31452
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5799838304519653,
+      "learning_rate": 0.00028698660557344597,
+      "loss": 3.1835,
+      "step": 31453
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5891263484954834,
+      "learning_rate": 0.0002869712238921136,
+      "loss": 2.8755,
+      "step": 31454
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5852018594741821,
+      "learning_rate": 0.00028695584224509656,
+      "loss": 3.0743,
+      "step": 31455
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6360331177711487,
+      "learning_rate": 0.00028694046063243537,
+      "loss": 3.1777,
+      "step": 31456
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5894647240638733,
+      "learning_rate": 0.0002869250790541704,
+      "loss": 3.2482,
+      "step": 31457
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5866122841835022,
+      "learning_rate": 0.0002869096975103423,
+      "loss": 2.9541,
+      "step": 31458
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5750760436058044,
+      "learning_rate": 0.00028689431600099154,
+      "loss": 3.1921,
+      "step": 31459
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6584492325782776,
+      "learning_rate": 0.0002868789345261587,
+      "loss": 2.9085,
+      "step": 31460
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6034347414970398,
+      "learning_rate": 0.00028686355308588405,
+      "loss": 3.332,
+      "step": 31461
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.581174910068512,
+      "learning_rate": 0.0002868481716802084,
+      "loss": 2.9919,
+      "step": 31462
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.616607129573822,
+      "learning_rate": 0.00028683279030917216,
+      "loss": 2.9166,
+      "step": 31463
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6416230201721191,
+      "learning_rate": 0.0002868174089728157,
+      "loss": 2.944,
+      "step": 31464
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.616322934627533,
+      "learning_rate": 0.0002868020276711798,
+      "loss": 2.9355,
+      "step": 31465
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5935969948768616,
+      "learning_rate": 0.0002867866464043047,
+      "loss": 3.1532,
+      "step": 31466
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6162576079368591,
+      "learning_rate": 0.00028677126517223113,
+      "loss": 2.867,
+      "step": 31467
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6141955852508545,
+      "learning_rate": 0.00028675588397499935,
+      "loss": 2.8846,
+      "step": 31468
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5820909142494202,
+      "learning_rate": 0.00028674050281265003,
+      "loss": 3.1472,
+      "step": 31469
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6148683428764343,
+      "learning_rate": 0.00028672512168522384,
+      "loss": 2.9993,
+      "step": 31470
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5928391218185425,
+      "learning_rate": 0.000286709740592761,
+      "loss": 2.9203,
+      "step": 31471
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5849189758300781,
+      "learning_rate": 0.0002866943595353023,
+      "loss": 2.995,
+      "step": 31472
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6054073572158813,
+      "learning_rate": 0.00028667897851288786,
+      "loss": 3.1151,
+      "step": 31473
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5830089449882507,
+      "learning_rate": 0.00028666359752555854,
+      "loss": 3.001,
+      "step": 31474
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.565097451210022,
+      "learning_rate": 0.00028664821657335476,
+      "loss": 3.1232,
+      "step": 31475
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7242405414581299,
+      "learning_rate": 0.0002866328356563169,
+      "loss": 2.8098,
+      "step": 31476
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5993044376373291,
+      "learning_rate": 0.0002866174547744858,
+      "loss": 3.0741,
+      "step": 31477
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5751321911811829,
+      "learning_rate": 0.0002866020739279016,
+      "loss": 2.8981,
+      "step": 31478
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6080617308616638,
+      "learning_rate": 0.0002865866931166049,
+      "loss": 3.0646,
+      "step": 31479
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6069605350494385,
+      "learning_rate": 0.0002865713123406364,
+      "loss": 2.967,
+      "step": 31480
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6173204183578491,
+      "learning_rate": 0.0002865559316000364,
+      "loss": 2.9911,
+      "step": 31481
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5786520838737488,
+      "learning_rate": 0.0002865405508948456,
+      "loss": 2.7597,
+      "step": 31482
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6026302576065063,
+      "learning_rate": 0.0002865251702251042,
+      "loss": 3.1053,
+      "step": 31483
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5935983061790466,
+      "learning_rate": 0.00028650978959085305,
+      "loss": 3.092,
+      "step": 31484
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6370457410812378,
+      "learning_rate": 0.0002864944089921325,
+      "loss": 3.0275,
+      "step": 31485
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6145215630531311,
+      "learning_rate": 0.00028647902842898304,
+      "loss": 3.1223,
+      "step": 31486
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6142358779907227,
+      "learning_rate": 0.00028646364790144526,
+      "loss": 3.0543,
+      "step": 31487
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5921164155006409,
+      "learning_rate": 0.00028644826740955963,
+      "loss": 2.9785,
+      "step": 31488
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6127771139144897,
+      "learning_rate": 0.0002864328869533666,
+      "loss": 3.1838,
+      "step": 31489
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6095576882362366,
+      "learning_rate": 0.0002864175065329068,
+      "loss": 2.9457,
+      "step": 31490
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5726305842399597,
+      "learning_rate": 0.00028640212614822056,
+      "loss": 2.9453,
+      "step": 31491
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5958991050720215,
+      "learning_rate": 0.0002863867457993486,
+      "loss": 2.8155,
+      "step": 31492
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5724688768386841,
+      "learning_rate": 0.00028637136548633136,
+      "loss": 2.9964,
+      "step": 31493
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6108876466751099,
+      "learning_rate": 0.0002863559852092092,
+      "loss": 3.0757,
+      "step": 31494
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5869536399841309,
+      "learning_rate": 0.00028634060496802283,
+      "loss": 2.9439,
+      "step": 31495
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6103752255439758,
+      "learning_rate": 0.0002863252247628127,
+      "loss": 3.0208,
+      "step": 31496
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6240681409835815,
+      "learning_rate": 0.0002863098445936194,
+      "loss": 2.8288,
+      "step": 31497
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5823042988777161,
+      "learning_rate": 0.0002862944644604831,
+      "loss": 2.9939,
+      "step": 31498
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5619244575500488,
+      "learning_rate": 0.00028627908436344463,
+      "loss": 3.0009,
+      "step": 31499
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.652056097984314,
+      "learning_rate": 0.0002862637043025445,
+      "loss": 3.3113,
+      "step": 31500
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5911815166473389,
+      "learning_rate": 0.0002862483242778229,
+      "loss": 3.1164,
+      "step": 31501
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6071313619613647,
+      "learning_rate": 0.0002862329442893208,
+      "loss": 3.1199,
+      "step": 31502
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6353355646133423,
+      "learning_rate": 0.00028621756433707844,
+      "loss": 2.8222,
+      "step": 31503
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6127616167068481,
+      "learning_rate": 0.00028620218442113623,
+      "loss": 3.1173,
+      "step": 31504
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6205554008483887,
+      "learning_rate": 0.00028618680454153497,
+      "loss": 3.0187,
+      "step": 31505
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5830715298652649,
+      "learning_rate": 0.000286171424698315,
+      "loss": 2.8732,
+      "step": 31506
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6005604267120361,
+      "learning_rate": 0.0002861560448915168,
+      "loss": 3.1406,
+      "step": 31507
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5887455344200134,
+      "learning_rate": 0.0002861406651211808,
+      "loss": 3.1733,
+      "step": 31508
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5852056741714478,
+      "learning_rate": 0.0002861252853873476,
+      "loss": 3.0398,
+      "step": 31509
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6041200160980225,
+      "learning_rate": 0.000286109905690058,
+      "loss": 2.9279,
+      "step": 31510
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.590787410736084,
+      "learning_rate": 0.000286094526029352,
+      "loss": 2.9335,
+      "step": 31511
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5768125653266907,
+      "learning_rate": 0.00028607914640527046,
+      "loss": 3.1014,
+      "step": 31512
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6092056035995483,
+      "learning_rate": 0.00028606376681785366,
+      "loss": 3.0408,
+      "step": 31513
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6362078189849854,
+      "learning_rate": 0.0002860483872671423,
+      "loss": 2.8638,
+      "step": 31514
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5419045686721802,
+      "learning_rate": 0.00028603300775317683,
+      "loss": 3.0463,
+      "step": 31515
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5694605112075806,
+      "learning_rate": 0.00028601762827599764,
+      "loss": 2.8361,
+      "step": 31516
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6117928624153137,
+      "learning_rate": 0.00028600224883564544,
+      "loss": 3.0969,
+      "step": 31517
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5907307863235474,
+      "learning_rate": 0.00028598686943216055,
+      "loss": 3.1432,
+      "step": 31518
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5980222821235657,
+      "learning_rate": 0.00028597149006558343,
+      "loss": 2.9284,
+      "step": 31519
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6280714869499207,
+      "learning_rate": 0.0002859561107359549,
+      "loss": 2.811,
+      "step": 31520
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5964347720146179,
+      "learning_rate": 0.0002859407314433152,
+      "loss": 2.8092,
+      "step": 31521
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5739207863807678,
+      "learning_rate": 0.00028592535218770497,
+      "loss": 3.0099,
+      "step": 31522
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5974461436271667,
+      "learning_rate": 0.00028590997296916454,
+      "loss": 2.9818,
+      "step": 31523
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6324328184127808,
+      "learning_rate": 0.00028589459378773456,
+      "loss": 3.3268,
+      "step": 31524
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5549293756484985,
+      "learning_rate": 0.0002858792146434556,
+      "loss": 3.1289,
+      "step": 31525
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5910172462463379,
+      "learning_rate": 0.00028586383553636793,
+      "loss": 3.0805,
+      "step": 31526
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5912405848503113,
+      "learning_rate": 0.0002858484564665123,
+      "loss": 3.2192,
+      "step": 31527
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6898330450057983,
+      "learning_rate": 0.0002858330774339291,
+      "loss": 2.9639,
+      "step": 31528
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6082789897918701,
+      "learning_rate": 0.00028581769843865875,
+      "loss": 2.9877,
+      "step": 31529
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6206566691398621,
+      "learning_rate": 0.000285802319480742,
+      "loss": 3.1703,
+      "step": 31530
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5829499363899231,
+      "learning_rate": 0.000285786940560219,
+      "loss": 3.1095,
+      "step": 31531
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5791610479354858,
+      "learning_rate": 0.0002857715616771307,
+      "loss": 2.9349,
+      "step": 31532
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6252768635749817,
+      "learning_rate": 0.00028575618283151725,
+      "loss": 3.0773,
+      "step": 31533
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5908172130584717,
+      "learning_rate": 0.0002857408040234192,
+      "loss": 2.9155,
+      "step": 31534
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5993767976760864,
+      "learning_rate": 0.0002857254252528773,
+      "loss": 3.1915,
+      "step": 31535
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.615296483039856,
+      "learning_rate": 0.0002857100465199318,
+      "loss": 2.9795,
+      "step": 31536
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.57916659116745,
+      "learning_rate": 0.00028569466782462337,
+      "loss": 3.1722,
+      "step": 31537
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.597743570804596,
+      "learning_rate": 0.0002856792891669922,
+      "loss": 2.8156,
+      "step": 31538
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5725197792053223,
+      "learning_rate": 0.0002856639105470792,
+      "loss": 2.9045,
+      "step": 31539
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5700836777687073,
+      "learning_rate": 0.00028564853196492475,
+      "loss": 3.203,
+      "step": 31540
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6204779148101807,
+      "learning_rate": 0.00028563315342056914,
+      "loss": 3.0341,
+      "step": 31541
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6638866066932678,
+      "learning_rate": 0.00028561777491405317,
+      "loss": 2.8039,
+      "step": 31542
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6386014819145203,
+      "learning_rate": 0.0002856023964454172,
+      "loss": 3.016,
+      "step": 31543
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5822014212608337,
+      "learning_rate": 0.0002855870180147018,
+      "loss": 2.9548,
+      "step": 31544
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.586982786655426,
+      "learning_rate": 0.00028557163962194724,
+      "loss": 2.9768,
+      "step": 31545
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5915098786354065,
+      "learning_rate": 0.0002855562612671943,
+      "loss": 3.1178,
+      "step": 31546
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5712721943855286,
+      "learning_rate": 0.0002855408829504835,
+      "loss": 2.9357,
+      "step": 31547
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.643669605255127,
+      "learning_rate": 0.000285525504671855,
+      "loss": 2.9061,
+      "step": 31548
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5977221727371216,
+      "learning_rate": 0.0002855101264313497,
+      "loss": 2.989,
+      "step": 31549
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6182084679603577,
+      "learning_rate": 0.0002854947482290079,
+      "loss": 3.1102,
+      "step": 31550
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6026540994644165,
+      "learning_rate": 0.0002854793700648701,
+      "loss": 3.0948,
+      "step": 31551
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6409699320793152,
+      "learning_rate": 0.00028546399193897694,
+      "loss": 2.8929,
+      "step": 31552
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5884620547294617,
+      "learning_rate": 0.0002854486138513687,
+      "loss": 3.3798,
+      "step": 31553
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5860359072685242,
+      "learning_rate": 0.0002854332358020861,
+      "loss": 3.1201,
+      "step": 31554
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6279225945472717,
+      "learning_rate": 0.0002854178577911696,
+      "loss": 3.0676,
+      "step": 31555
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6590830087661743,
+      "learning_rate": 0.00028540247981865944,
+      "loss": 3.0733,
+      "step": 31556
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6351860165596008,
+      "learning_rate": 0.0002853871018845966,
+      "loss": 2.8937,
+      "step": 31557
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6308092474937439,
+      "learning_rate": 0.0002853717239890212,
+      "loss": 2.96,
+      "step": 31558
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6065652370452881,
+      "learning_rate": 0.00028535634613197395,
+      "loss": 3.285,
+      "step": 31559
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5910200476646423,
+      "learning_rate": 0.0002853409683134951,
+      "loss": 2.7821,
+      "step": 31560
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.624097466468811,
+      "learning_rate": 0.0002853255905336254,
+      "loss": 3.0966,
+      "step": 31561
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6241322159767151,
+      "learning_rate": 0.0002853102127924054,
+      "loss": 3.0899,
+      "step": 31562
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6398972272872925,
+      "learning_rate": 0.00028529483508987524,
+      "loss": 3.0438,
+      "step": 31563
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.601036548614502,
+      "learning_rate": 0.0002852794574260758,
+      "loss": 3.2424,
+      "step": 31564
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.610298216342926,
+      "learning_rate": 0.0002852640798010474,
+      "loss": 3.0497,
+      "step": 31565
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5706086754798889,
+      "learning_rate": 0.00028524870221483054,
+      "loss": 2.9957,
+      "step": 31566
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5881780385971069,
+      "learning_rate": 0.00028523332466746584,
+      "loss": 3.1404,
+      "step": 31567
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.558418333530426,
+      "learning_rate": 0.0002852179471589937,
+      "loss": 2.9538,
+      "step": 31568
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5936317443847656,
+      "learning_rate": 0.00028520256968945465,
+      "loss": 2.7039,
+      "step": 31569
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6009182929992676,
+      "learning_rate": 0.0002851871922588891,
+      "loss": 3.1115,
+      "step": 31570
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6169144511222839,
+      "learning_rate": 0.00028517181486733764,
+      "loss": 3.127,
+      "step": 31571
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6200720071792603,
+      "learning_rate": 0.00028515643751484093,
+      "loss": 3.0668,
+      "step": 31572
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6260064244270325,
+      "learning_rate": 0.00028514106020143914,
+      "loss": 3.2649,
+      "step": 31573
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5613835453987122,
+      "learning_rate": 0.00028512568292717304,
+      "loss": 3.1029,
+      "step": 31574
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6062586903572083,
+      "learning_rate": 0.0002851103056920829,
+      "loss": 2.9356,
+      "step": 31575
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6101064085960388,
+      "learning_rate": 0.00028509492849620945,
+      "loss": 2.9583,
+      "step": 31576
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5925574898719788,
+      "learning_rate": 0.0002850795513395931,
+      "loss": 3.092,
+      "step": 31577
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5718997120857239,
+      "learning_rate": 0.00028506417422227427,
+      "loss": 3.0322,
+      "step": 31578
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6322290897369385,
+      "learning_rate": 0.0002850487971442936,
+      "loss": 2.9443,
+      "step": 31579
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.585167646408081,
+      "learning_rate": 0.00028503342010569147,
+      "loss": 2.9556,
+      "step": 31580
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5464421510696411,
+      "learning_rate": 0.00028501804310650844,
+      "loss": 2.9251,
+      "step": 31581
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5760757327079773,
+      "learning_rate": 0.00028500266614678505,
+      "loss": 3.1459,
+      "step": 31582
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5659793615341187,
+      "learning_rate": 0.0002849872892265617,
+      "loss": 3.1723,
+      "step": 31583
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.64796382188797,
+      "learning_rate": 0.00028497191234587904,
+      "loss": 2.8646,
+      "step": 31584
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7371762990951538,
+      "learning_rate": 0.0002849565355047773,
+      "loss": 3.0312,
+      "step": 31585
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5944815278053284,
+      "learning_rate": 0.00028494115870329723,
+      "loss": 2.9846,
+      "step": 31586
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5763345956802368,
+      "learning_rate": 0.0002849257819414793,
+      "loss": 2.9265,
+      "step": 31587
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5970070958137512,
+      "learning_rate": 0.00028491040521936385,
+      "loss": 3.0161,
+      "step": 31588
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.594948410987854,
+      "learning_rate": 0.00028489502853699163,
+      "loss": 3.1838,
+      "step": 31589
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6008180975914001,
+      "learning_rate": 0.0002848796518944029,
+      "loss": 3.0902,
+      "step": 31590
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5951430201530457,
+      "learning_rate": 0.00028486427529163824,
+      "loss": 3.0587,
+      "step": 31591
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.607059121131897,
+      "learning_rate": 0.0002848488987287382,
+      "loss": 3.0581,
+      "step": 31592
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.600946307182312,
+      "learning_rate": 0.00028483352220574324,
+      "loss": 3.1581,
+      "step": 31593
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5530339479446411,
+      "learning_rate": 0.0002848181457226939,
+      "loss": 3.1023,
+      "step": 31594
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6456699371337891,
+      "learning_rate": 0.0002848027692796306,
+      "loss": 3.1343,
+      "step": 31595
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5949581265449524,
+      "learning_rate": 0.00028478739287659383,
+      "loss": 3.0981,
+      "step": 31596
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6152155995368958,
+      "learning_rate": 0.00028477201651362427,
+      "loss": 2.957,
+      "step": 31597
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6344307065010071,
+      "learning_rate": 0.0002847566401907622,
+      "loss": 2.9609,
+      "step": 31598
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5822964906692505,
+      "learning_rate": 0.00028474126390804833,
+      "loss": 3.0524,
+      "step": 31599
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5847934484481812,
+      "learning_rate": 0.00028472588766552284,
+      "loss": 2.9667,
+      "step": 31600
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5926794409751892,
+      "learning_rate": 0.0002847105114632265,
+      "loss": 3.0349,
+      "step": 31601
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5681343078613281,
+      "learning_rate": 0.00028469513530119983,
+      "loss": 3.2188,
+      "step": 31602
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5752440690994263,
+      "learning_rate": 0.0002846797591794831,
+      "loss": 3.0376,
+      "step": 31603
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6142379641532898,
+      "learning_rate": 0.00028466438309811704,
+      "loss": 3.1508,
+      "step": 31604
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6223300099372864,
+      "learning_rate": 0.00028464900705714197,
+      "loss": 2.9159,
+      "step": 31605
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6309249401092529,
+      "learning_rate": 0.00028463363105659837,
+      "loss": 2.8823,
+      "step": 31606
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5628587603569031,
+      "learning_rate": 0.000284618255096527,
+      "loss": 3.1089,
+      "step": 31607
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6215999126434326,
+      "learning_rate": 0.0002846028791769681,
+      "loss": 3.1475,
+      "step": 31608
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5789870619773865,
+      "learning_rate": 0.0002845875032979624,
+      "loss": 3.0857,
+      "step": 31609
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5938981771469116,
+      "learning_rate": 0.00028457212745955,
+      "loss": 2.958,
+      "step": 31610
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5996585488319397,
+      "learning_rate": 0.00028455675166177177,
+      "loss": 2.9466,
+      "step": 31611
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5623268485069275,
+      "learning_rate": 0.00028454137590466815,
+      "loss": 2.8726,
+      "step": 31612
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5604333281517029,
+      "learning_rate": 0.00028452600018827947,
+      "loss": 3.3238,
+      "step": 31613
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6165540218353271,
+      "learning_rate": 0.0002845106245126464,
+      "loss": 2.9483,
+      "step": 31614
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5510979890823364,
+      "learning_rate": 0.00028449524887780926,
+      "loss": 2.9379,
+      "step": 31615
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5766331553459167,
+      "learning_rate": 0.0002844798732838087,
+      "loss": 3.0955,
+      "step": 31616
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6156991124153137,
+      "learning_rate": 0.00028446449773068523,
+      "loss": 2.9406,
+      "step": 31617
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6004645824432373,
+      "learning_rate": 0.0002844491222184792,
+      "loss": 3.0694,
+      "step": 31618
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6063745021820068,
+      "learning_rate": 0.0002844337467472313,
+      "loss": 2.8684,
+      "step": 31619
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5919575691223145,
+      "learning_rate": 0.0002844183713169818,
+      "loss": 3.1631,
+      "step": 31620
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5818456411361694,
+      "learning_rate": 0.0002844029959277713,
+      "loss": 2.9513,
+      "step": 31621
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5850279927253723,
+      "learning_rate": 0.0002843876205796404,
+      "loss": 3.0365,
+      "step": 31622
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.590082049369812,
+      "learning_rate": 0.00028437224527262943,
+      "loss": 2.9559,
+      "step": 31623
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8791807889938354,
+      "learning_rate": 0.0002843568700067791,
+      "loss": 2.9216,
+      "step": 31624
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.593088686466217,
+      "learning_rate": 0.00028434149478212957,
+      "loss": 3.1052,
+      "step": 31625
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6907350420951843,
+      "learning_rate": 0.00028432611959872166,
+      "loss": 3.1177,
+      "step": 31626
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5803176164627075,
+      "learning_rate": 0.00028431074445659573,
+      "loss": 3.0874,
+      "step": 31627
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.627086877822876,
+      "learning_rate": 0.0002842953693557921,
+      "loss": 3.0477,
+      "step": 31628
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6301358938217163,
+      "learning_rate": 0.0002842799942963517,
+      "loss": 3.1221,
+      "step": 31629
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5863131284713745,
+      "learning_rate": 0.0002842646192783146,
+      "loss": 3.176,
+      "step": 31630
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6010187864303589,
+      "learning_rate": 0.0002842492443017216,
+      "loss": 3.2,
+      "step": 31631
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5959247946739197,
+      "learning_rate": 0.0002842338693666129,
+      "loss": 2.8586,
+      "step": 31632
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5971164107322693,
+      "learning_rate": 0.0002842184944730292,
+      "loss": 2.8364,
+      "step": 31633
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6169106960296631,
+      "learning_rate": 0.000284203119621011,
+      "loss": 2.9412,
+      "step": 31634
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6330366730690002,
+      "learning_rate": 0.0002841877448105987,
+      "loss": 3.2115,
+      "step": 31635
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6117235422134399,
+      "learning_rate": 0.00028417237004183297,
+      "loss": 2.8728,
+      "step": 31636
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5777655839920044,
+      "learning_rate": 0.00028415699531475394,
+      "loss": 2.9479,
+      "step": 31637
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5683594346046448,
+      "learning_rate": 0.0002841416206294025,
+      "loss": 3.1955,
+      "step": 31638
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6049903631210327,
+      "learning_rate": 0.00028412624598581895,
+      "loss": 3.1881,
+      "step": 31639
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5954413414001465,
+      "learning_rate": 0.0002841108713840437,
+      "loss": 3.0535,
+      "step": 31640
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6370765566825867,
+      "learning_rate": 0.00028409549682411754,
+      "loss": 3.0142,
+      "step": 31641
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5675783753395081,
+      "learning_rate": 0.0002840801223060807,
+      "loss": 3.1618,
+      "step": 31642
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6547769904136658,
+      "learning_rate": 0.00028406474782997367,
+      "loss": 3.2298,
+      "step": 31643
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6375356316566467,
+      "learning_rate": 0.0002840493733958372,
+      "loss": 3.0107,
+      "step": 31644
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6045966744422913,
+      "learning_rate": 0.00028403399900371146,
+      "loss": 3.0618,
+      "step": 31645
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5673590302467346,
+      "learning_rate": 0.0002840186246536372,
+      "loss": 2.9589,
+      "step": 31646
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6037384867668152,
+      "learning_rate": 0.0002840032503456547,
+      "loss": 3.1438,
+      "step": 31647
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6107524633407593,
+      "learning_rate": 0.0002839878760798046,
+      "loss": 3.062,
+      "step": 31648
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5911772847175598,
+      "learning_rate": 0.00028397250185612747,
+      "loss": 2.9119,
+      "step": 31649
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6252596974372864,
+      "learning_rate": 0.0002839571276746635,
+      "loss": 2.9791,
+      "step": 31650
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5904194116592407,
+      "learning_rate": 0.0002839417535354535,
+      "loss": 2.805,
+      "step": 31651
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5900318026542664,
+      "learning_rate": 0.0002839263794385378,
+      "loss": 3.1029,
+      "step": 31652
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6207619309425354,
+      "learning_rate": 0.0002839110053839568,
+      "loss": 2.9533,
+      "step": 31653
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7129519581794739,
+      "learning_rate": 0.0002838956313717513,
+      "loss": 2.9724,
+      "step": 31654
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5762568712234497,
+      "learning_rate": 0.0002838802574019615,
+      "loss": 3.0598,
+      "step": 31655
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6000192761421204,
+      "learning_rate": 0.0002838648834746281,
+      "loss": 3.0808,
+      "step": 31656
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6309391260147095,
+      "learning_rate": 0.0002838495095897914,
+      "loss": 3.1001,
+      "step": 31657
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5648094415664673,
+      "learning_rate": 0.0002838341357474919,
+      "loss": 2.8382,
+      "step": 31658
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.601072371006012,
+      "learning_rate": 0.00028381876194777037,
+      "loss": 2.9228,
+      "step": 31659
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6006704568862915,
+      "learning_rate": 0.000283803388190667,
+      "loss": 3.1873,
+      "step": 31660
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5812244415283203,
+      "learning_rate": 0.00028378801447622245,
+      "loss": 2.9829,
+      "step": 31661
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5904305577278137,
+      "learning_rate": 0.00028377264080447703,
+      "loss": 2.9529,
+      "step": 31662
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5870997905731201,
+      "learning_rate": 0.0002837572671754714,
+      "loss": 2.9428,
+      "step": 31663
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6143147945404053,
+      "learning_rate": 0.0002837418935892461,
+      "loss": 3.1222,
+      "step": 31664
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8274781107902527,
+      "learning_rate": 0.0002837265200458414,
+      "loss": 3.1928,
+      "step": 31665
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5731198787689209,
+      "learning_rate": 0.00028371114654529805,
+      "loss": 3.1566,
+      "step": 31666
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.582653820514679,
+      "learning_rate": 0.0002836957730876563,
+      "loss": 3.0699,
+      "step": 31667
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6127033829689026,
+      "learning_rate": 0.0002836803996729567,
+      "loss": 3.074,
+      "step": 31668
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.578920304775238,
+      "learning_rate": 0.00028366502630123994,
+      "loss": 3.1217,
+      "step": 31669
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.574611246585846,
+      "learning_rate": 0.0002836496529725462,
+      "loss": 3.0978,
+      "step": 31670
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6003169417381287,
+      "learning_rate": 0.0002836342796869163,
+      "loss": 2.872,
+      "step": 31671
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6231915950775146,
+      "learning_rate": 0.0002836189064443904,
+      "loss": 2.9799,
+      "step": 31672
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6026855111122131,
+      "learning_rate": 0.00028360353324500913,
+      "loss": 2.9663,
+      "step": 31673
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.639599621295929,
+      "learning_rate": 0.0002835881600888132,
+      "loss": 3.0197,
+      "step": 31674
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6096246242523193,
+      "learning_rate": 0.0002835727869758427,
+      "loss": 3.1729,
+      "step": 31675
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.585129976272583,
+      "learning_rate": 0.00028355741390613846,
+      "loss": 2.8763,
+      "step": 31676
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5901762843132019,
+      "learning_rate": 0.00028354204087974064,
+      "loss": 3.2103,
+      "step": 31677
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6148892641067505,
+      "learning_rate": 0.00028352666789669003,
+      "loss": 2.9013,
+      "step": 31678
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5568849444389343,
+      "learning_rate": 0.0002835112949570271,
+      "loss": 2.9905,
+      "step": 31679
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5767827033996582,
+      "learning_rate": 0.00028349592206079207,
+      "loss": 3.0622,
+      "step": 31680
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6218153834342957,
+      "learning_rate": 0.00028348054920802575,
+      "loss": 2.9914,
+      "step": 31681
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6101388335227966,
+      "learning_rate": 0.0002834651763987684,
+      "loss": 3.0335,
+      "step": 31682
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5884225368499756,
+      "learning_rate": 0.00028344980363306056,
+      "loss": 2.9399,
+      "step": 31683
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6422539353370667,
+      "learning_rate": 0.0002834344309109429,
+      "loss": 2.9614,
+      "step": 31684
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5843330025672913,
+      "learning_rate": 0.0002834190582324557,
+      "loss": 3.2347,
+      "step": 31685
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5898882746696472,
+      "learning_rate": 0.0002834036855976395,
+      "loss": 3.0284,
+      "step": 31686
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6142812967300415,
+      "learning_rate": 0.0002833883130065347,
+      "loss": 3.0773,
+      "step": 31687
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.591521143913269,
+      "learning_rate": 0.00028337294045918194,
+      "loss": 3.2177,
+      "step": 31688
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6057526469230652,
+      "learning_rate": 0.00028335756795562175,
+      "loss": 2.7478,
+      "step": 31689
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5793746113777161,
+      "learning_rate": 0.0002833421954958944,
+      "loss": 3.3766,
+      "step": 31690
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6439757943153381,
+      "learning_rate": 0.00028332682308004057,
+      "loss": 3.1779,
+      "step": 31691
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6085169911384583,
+      "learning_rate": 0.0002833114507081007,
+      "loss": 3.1525,
+      "step": 31692
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6145389080047607,
+      "learning_rate": 0.0002832960783801151,
+      "loss": 3.0357,
+      "step": 31693
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5741784572601318,
+      "learning_rate": 0.0002832807060961246,
+      "loss": 3.3173,
+      "step": 31694
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6480720639228821,
+      "learning_rate": 0.0002832653338561694,
+      "loss": 2.9654,
+      "step": 31695
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.600227415561676,
+      "learning_rate": 0.00028324996166029015,
+      "loss": 2.8275,
+      "step": 31696
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6081393957138062,
+      "learning_rate": 0.00028323458950852727,
+      "loss": 3.164,
+      "step": 31697
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5734052658081055,
+      "learning_rate": 0.00028321921740092113,
+      "loss": 3.0086,
+      "step": 31698
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6059851050376892,
+      "learning_rate": 0.00028320384533751255,
+      "loss": 2.9567,
+      "step": 31699
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5766007304191589,
+      "learning_rate": 0.0002831884733183417,
+      "loss": 3.1765,
+      "step": 31700
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5918262600898743,
+      "learning_rate": 0.00028317310134344927,
+      "loss": 3.1488,
+      "step": 31701
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5856273770332336,
+      "learning_rate": 0.0002831577294128755,
+      "loss": 3.2332,
+      "step": 31702
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6248118877410889,
+      "learning_rate": 0.0002831423575266611,
+      "loss": 2.9965,
+      "step": 31703
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6196810603141785,
+      "learning_rate": 0.0002831269856848465,
+      "loss": 3.1695,
+      "step": 31704
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.693333089351654,
+      "learning_rate": 0.0002831116138874721,
+      "loss": 3.2062,
+      "step": 31705
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6046402454376221,
+      "learning_rate": 0.00028309624213457854,
+      "loss": 3.0393,
+      "step": 31706
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6249449253082275,
+      "learning_rate": 0.0002830808704262062,
+      "loss": 2.9533,
+      "step": 31707
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5609068870544434,
+      "learning_rate": 0.0002830654987623957,
+      "loss": 2.9595,
+      "step": 31708
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6460155844688416,
+      "learning_rate": 0.00028305012714318717,
+      "loss": 2.9995,
+      "step": 31709
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6091744303703308,
+      "learning_rate": 0.0002830347555686215,
+      "loss": 3.2087,
+      "step": 31710
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5952534079551697,
+      "learning_rate": 0.00028301938403873904,
+      "loss": 2.9537,
+      "step": 31711
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5830929279327393,
+      "learning_rate": 0.0002830040125535802,
+      "loss": 2.989,
+      "step": 31712
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.598037838935852,
+      "learning_rate": 0.00028298864111318557,
+      "loss": 3.0506,
+      "step": 31713
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5895772576332092,
+      "learning_rate": 0.00028297326971759554,
+      "loss": 3.0416,
+      "step": 31714
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6120174527168274,
+      "learning_rate": 0.00028295789836685055,
+      "loss": 3.1145,
+      "step": 31715
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5766130089759827,
+      "learning_rate": 0.00028294252706099137,
+      "loss": 3.0686,
+      "step": 31716
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6399678587913513,
+      "learning_rate": 0.0002829271558000581,
+      "loss": 2.9891,
+      "step": 31717
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6096529364585876,
+      "learning_rate": 0.0002829117845840916,
+      "loss": 2.9512,
+      "step": 31718
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.632032036781311,
+      "learning_rate": 0.00028289641341313205,
+      "loss": 3.1313,
+      "step": 31719
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6017841696739197,
+      "learning_rate": 0.00028288104228722,
+      "loss": 3.0703,
+      "step": 31720
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6441850066184998,
+      "learning_rate": 0.00028286567120639616,
+      "loss": 2.9992,
+      "step": 31721
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5980612635612488,
+      "learning_rate": 0.00028285030017070076,
+      "loss": 3.1623,
+      "step": 31722
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5885056257247925,
+      "learning_rate": 0.0002828349291801744,
+      "loss": 2.9903,
+      "step": 31723
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5749340057373047,
+      "learning_rate": 0.00028281955823485745,
+      "loss": 2.9712,
+      "step": 31724
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5880852341651917,
+      "learning_rate": 0.00028280418733479053,
+      "loss": 3.0399,
+      "step": 31725
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6071673631668091,
+      "learning_rate": 0.00028278881648001415,
+      "loss": 2.9693,
+      "step": 31726
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6033803820610046,
+      "learning_rate": 0.0002827734456705686,
+      "loss": 3.0939,
+      "step": 31727
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5925212502479553,
+      "learning_rate": 0.00028275807490649456,
+      "loss": 3.0581,
+      "step": 31728
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6496951580047607,
+      "learning_rate": 0.0002827427041878324,
+      "loss": 2.8309,
+      "step": 31729
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.568051815032959,
+      "learning_rate": 0.00028272733351462253,
+      "loss": 3.1127,
+      "step": 31730
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5681890249252319,
+      "learning_rate": 0.0002827119628869057,
+      "loss": 2.8544,
+      "step": 31731
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5864232182502747,
+      "learning_rate": 0.0002826965923047222,
+      "loss": 2.8054,
+      "step": 31732
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6019498705863953,
+      "learning_rate": 0.0002826812217681125,
+      "loss": 2.9317,
+      "step": 31733
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.637272298336029,
+      "learning_rate": 0.00028266585127711713,
+      "loss": 3.1232,
+      "step": 31734
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6020846366882324,
+      "learning_rate": 0.0002826504808317765,
+      "loss": 3.0031,
+      "step": 31735
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6123052835464478,
+      "learning_rate": 0.00028263511043213137,
+      "loss": 3.0354,
+      "step": 31736
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5840319395065308,
+      "learning_rate": 0.0002826197400782218,
+      "loss": 2.7779,
+      "step": 31737
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6461364030838013,
+      "learning_rate": 0.0002826043697700887,
+      "loss": 3.3123,
+      "step": 31738
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6259379386901855,
+      "learning_rate": 0.00028258899950777213,
+      "loss": 3.0375,
+      "step": 31739
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5921148657798767,
+      "learning_rate": 0.0002825736292913129,
+      "loss": 2.9987,
+      "step": 31740
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5672594904899597,
+      "learning_rate": 0.00028255825912075136,
+      "loss": 2.9653,
+      "step": 31741
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5888809561729431,
+      "learning_rate": 0.00028254288899612795,
+      "loss": 3.078,
+      "step": 31742
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5958986282348633,
+      "learning_rate": 0.00028252751891748334,
+      "loss": 3.1048,
+      "step": 31743
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6289757490158081,
+      "learning_rate": 0.0002825121488848578,
+      "loss": 3.0964,
+      "step": 31744
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5901963710784912,
+      "learning_rate": 0.00028249677889829187,
+      "loss": 2.9512,
+      "step": 31745
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6648480296134949,
+      "learning_rate": 0.0002824814089578261,
+      "loss": 3.0944,
+      "step": 31746
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6112911105155945,
+      "learning_rate": 0.0002824660390635009,
+      "loss": 2.8058,
+      "step": 31747
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6173609495162964,
+      "learning_rate": 0.00028245066921535686,
+      "loss": 3.1509,
+      "step": 31748
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6812626123428345,
+      "learning_rate": 0.00028243529941343425,
+      "loss": 2.8451,
+      "step": 31749
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5810101628303528,
+      "learning_rate": 0.0002824199296577738,
+      "loss": 2.7511,
+      "step": 31750
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5572994947433472,
+      "learning_rate": 0.00028240455994841586,
+      "loss": 3.047,
+      "step": 31751
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6092203259468079,
+      "learning_rate": 0.0002823891902854008,
+      "loss": 3.0467,
+      "step": 31752
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6097322106361389,
+      "learning_rate": 0.00028237382066876937,
+      "loss": 3.1722,
+      "step": 31753
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5815500020980835,
+      "learning_rate": 0.0002823584510985618,
+      "loss": 3.1401,
+      "step": 31754
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6111709475517273,
+      "learning_rate": 0.0002823430815748187,
+      "loss": 3.1151,
+      "step": 31755
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6202290058135986,
+      "learning_rate": 0.0002823277120975806,
+      "loss": 2.981,
+      "step": 31756
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6156234741210938,
+      "learning_rate": 0.00028231234266688776,
+      "loss": 3.0407,
+      "step": 31757
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5927169322967529,
+      "learning_rate": 0.00028229697328278094,
+      "loss": 2.9775,
+      "step": 31758
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5997295379638672,
+      "learning_rate": 0.0002822816039453004,
+      "loss": 3.031,
+      "step": 31759
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5959268808364868,
+      "learning_rate": 0.0002822662346544867,
+      "loss": 3.1206,
+      "step": 31760
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6445311307907104,
+      "learning_rate": 0.0002822508654103804,
+      "loss": 2.9471,
+      "step": 31761
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6009385585784912,
+      "learning_rate": 0.0002822354962130219,
+      "loss": 2.9307,
+      "step": 31762
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5991648435592651,
+      "learning_rate": 0.00028222012706245167,
+      "loss": 3.112,
+      "step": 31763
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5902231335639954,
+      "learning_rate": 0.0002822047579587101,
+      "loss": 2.8237,
+      "step": 31764
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8094449639320374,
+      "learning_rate": 0.0002821893889018379,
+      "loss": 3.1565,
+      "step": 31765
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6150186657905579,
+      "learning_rate": 0.00028217401989187547,
+      "loss": 2.9342,
+      "step": 31766
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5921657681465149,
+      "learning_rate": 0.000282158650928863,
+      "loss": 2.9887,
+      "step": 31767
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5835986137390137,
+      "learning_rate": 0.00028214328201284146,
+      "loss": 3.1866,
+      "step": 31768
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5715051889419556,
+      "learning_rate": 0.00028212791314385097,
+      "loss": 2.8587,
+      "step": 31769
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6232491135597229,
+      "learning_rate": 0.0002821125443219321,
+      "loss": 3.0507,
+      "step": 31770
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5849910974502563,
+      "learning_rate": 0.0002820971755471254,
+      "loss": 2.9724,
+      "step": 31771
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6116089820861816,
+      "learning_rate": 0.0002820818068194713,
+      "loss": 3.2148,
+      "step": 31772
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6128947138786316,
+      "learning_rate": 0.00028206643813901024,
+      "loss": 3.0349,
+      "step": 31773
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.591354250907898,
+      "learning_rate": 0.0002820510695057827,
+      "loss": 3.119,
+      "step": 31774
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6188297867774963,
+      "learning_rate": 0.0002820357009198292,
+      "loss": 3.2393,
+      "step": 31775
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6261090636253357,
+      "learning_rate": 0.0002820203323811903,
+      "loss": 2.9624,
+      "step": 31776
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.56405109167099,
+      "learning_rate": 0.00028200496388990624,
+      "loss": 3.1006,
+      "step": 31777
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6306495666503906,
+      "learning_rate": 0.00028198959544601776,
+      "loss": 3.1529,
+      "step": 31778
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6113765239715576,
+      "learning_rate": 0.0002819742270495651,
+      "loss": 2.888,
+      "step": 31779
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5723241567611694,
+      "learning_rate": 0.000281958858700589,
+      "loss": 2.8904,
+      "step": 31780
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5704621076583862,
+      "learning_rate": 0.0002819434903991297,
+      "loss": 2.879,
+      "step": 31781
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5865147113800049,
+      "learning_rate": 0.00028192812214522777,
+      "loss": 3.0125,
+      "step": 31782
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6170030832290649,
+      "learning_rate": 0.00028191275393892376,
+      "loss": 2.9818,
+      "step": 31783
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6150692701339722,
+      "learning_rate": 0.00028189738578025803,
+      "loss": 3.019,
+      "step": 31784
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6316766142845154,
+      "learning_rate": 0.0002818820176692711,
+      "loss": 2.8793,
+      "step": 31785
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5430260300636292,
+      "learning_rate": 0.00028186664960600346,
+      "loss": 2.9567,
+      "step": 31786
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5936475992202759,
+      "learning_rate": 0.0002818512815904955,
+      "loss": 3.1525,
+      "step": 31787
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6307163834571838,
+      "learning_rate": 0.0002818359136227879,
+      "loss": 3.0512,
+      "step": 31788
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6203303933143616,
+      "learning_rate": 0.0002818205457029209,
+      "loss": 3.0299,
+      "step": 31789
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5848477482795715,
+      "learning_rate": 0.0002818051778309352,
+      "loss": 3.0923,
+      "step": 31790
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.611884355545044,
+      "learning_rate": 0.00028178981000687105,
+      "loss": 3.0312,
+      "step": 31791
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.585080623626709,
+      "learning_rate": 0.00028177444223076903,
+      "loss": 3.14,
+      "step": 31792
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5883607268333435,
+      "learning_rate": 0.00028175907450266974,
+      "loss": 3.2011,
+      "step": 31793
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6267507672309875,
+      "learning_rate": 0.0002817437068226134,
+      "loss": 3.17,
+      "step": 31794
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6080348491668701,
+      "learning_rate": 0.00028172833919064077,
+      "loss": 2.996,
+      "step": 31795
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5978765487670898,
+      "learning_rate": 0.000281712971606792,
+      "loss": 2.9237,
+      "step": 31796
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6047960519790649,
+      "learning_rate": 0.0002816976040711078,
+      "loss": 2.8924,
+      "step": 31797
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6580313444137573,
+      "learning_rate": 0.0002816822365836287,
+      "loss": 2.9914,
+      "step": 31798
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.620141863822937,
+      "learning_rate": 0.00028166686914439493,
+      "loss": 3.1461,
+      "step": 31799
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.604900062084198,
+      "learning_rate": 0.0002816515017534472,
+      "loss": 3.0193,
+      "step": 31800
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6572511196136475,
+      "learning_rate": 0.0002816361344108258,
+      "loss": 2.9982,
+      "step": 31801
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6069661974906921,
+      "learning_rate": 0.00028162076711657133,
+      "loss": 3.0559,
+      "step": 31802
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6206227540969849,
+      "learning_rate": 0.00028160539987072425,
+      "loss": 3.0348,
+      "step": 31803
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5691961050033569,
+      "learning_rate": 0.0002815900326733249,
+      "loss": 2.9718,
+      "step": 31804
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5769830942153931,
+      "learning_rate": 0.000281574665524414,
+      "loss": 2.8864,
+      "step": 31805
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6015021800994873,
+      "learning_rate": 0.0002815592984240318,
+      "loss": 3.0994,
+      "step": 31806
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6178338527679443,
+      "learning_rate": 0.00028154393137221886,
+      "loss": 3.1458,
+      "step": 31807
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5842515826225281,
+      "learning_rate": 0.00028152856436901573,
+      "loss": 3.1004,
+      "step": 31808
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5896745920181274,
+      "learning_rate": 0.00028151319741446276,
+      "loss": 3.0969,
+      "step": 31809
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6141567230224609,
+      "learning_rate": 0.0002814978305086005,
+      "loss": 2.9886,
+      "step": 31810
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6013665795326233,
+      "learning_rate": 0.00028148246365146927,
+      "loss": 3.0342,
+      "step": 31811
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6272392868995667,
+      "learning_rate": 0.00028146709684310974,
+      "loss": 3.2857,
+      "step": 31812
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5829760432243347,
+      "learning_rate": 0.0002814517300835624,
+      "loss": 3.1099,
+      "step": 31813
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.629869282245636,
+      "learning_rate": 0.0002814363633728675,
+      "loss": 2.8432,
+      "step": 31814
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6304134726524353,
+      "learning_rate": 0.00028142099671106573,
+      "loss": 3.305,
+      "step": 31815
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5929440855979919,
+      "learning_rate": 0.00028140563009819745,
+      "loss": 3.1774,
+      "step": 31816
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6079155206680298,
+      "learning_rate": 0.0002813902635343031,
+      "loss": 3.167,
+      "step": 31817
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6250888705253601,
+      "learning_rate": 0.00028137489701942335,
+      "loss": 2.9714,
+      "step": 31818
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6257724165916443,
+      "learning_rate": 0.00028135953055359836,
+      "loss": 2.7605,
+      "step": 31819
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6094701290130615,
+      "learning_rate": 0.00028134416413686897,
+      "loss": 3.1054,
+      "step": 31820
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6528363823890686,
+      "learning_rate": 0.00028132879776927535,
+      "loss": 2.792,
+      "step": 31821
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.647331953048706,
+      "learning_rate": 0.00028131343145085805,
+      "loss": 2.9551,
+      "step": 31822
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5757825374603271,
+      "learning_rate": 0.0002812980651816577,
+      "loss": 3.011,
+      "step": 31823
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5802479982376099,
+      "learning_rate": 0.00028128269896171457,
+      "loss": 2.9968,
+      "step": 31824
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.583878755569458,
+      "learning_rate": 0.00028126733279106925,
+      "loss": 3.0743,
+      "step": 31825
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5796141624450684,
+      "learning_rate": 0.0002812519666697621,
+      "loss": 3.1808,
+      "step": 31826
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6166408658027649,
+      "learning_rate": 0.0002812366005978337,
+      "loss": 3.1588,
+      "step": 31827
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5788041353225708,
+      "learning_rate": 0.0002812212345753245,
+      "loss": 2.872,
+      "step": 31828
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5892295837402344,
+      "learning_rate": 0.0002812058686022749,
+      "loss": 3.0678,
+      "step": 31829
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6605625748634338,
+      "learning_rate": 0.0002811905026787255,
+      "loss": 2.7785,
+      "step": 31830
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6186345219612122,
+      "learning_rate": 0.00028117513680471663,
+      "loss": 3.0413,
+      "step": 31831
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.561852216720581,
+      "learning_rate": 0.0002811597709802888,
+      "loss": 2.7898,
+      "step": 31832
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5842719674110413,
+      "learning_rate": 0.0002811444052054826,
+      "loss": 3.0802,
+      "step": 31833
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6352782249450684,
+      "learning_rate": 0.0002811290394803384,
+      "loss": 2.9058,
+      "step": 31834
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6168009638786316,
+      "learning_rate": 0.0002811136738048967,
+      "loss": 2.9086,
+      "step": 31835
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6582854986190796,
+      "learning_rate": 0.0002810983081791978,
+      "loss": 3.0012,
+      "step": 31836
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5972253084182739,
+      "learning_rate": 0.0002810829426032824,
+      "loss": 3.2943,
+      "step": 31837
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6005479693412781,
+      "learning_rate": 0.000281067577077191,
+      "loss": 2.9417,
+      "step": 31838
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5901066660881042,
+      "learning_rate": 0.0002810522116009638,
+      "loss": 2.9913,
+      "step": 31839
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6512740850448608,
+      "learning_rate": 0.0002810368461746415,
+      "loss": 3.1729,
+      "step": 31840
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6124058961868286,
+      "learning_rate": 0.00028102148079826443,
+      "loss": 2.9708,
+      "step": 31841
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5982481241226196,
+      "learning_rate": 0.0002810061154718732,
+      "loss": 3.1866,
+      "step": 31842
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5993276238441467,
+      "learning_rate": 0.00028099075019550826,
+      "loss": 2.8918,
+      "step": 31843
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5819422602653503,
+      "learning_rate": 0.00028097538496920987,
+      "loss": 3.1729,
+      "step": 31844
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6041077375411987,
+      "learning_rate": 0.0002809600197930188,
+      "loss": 2.9457,
+      "step": 31845
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6543545126914978,
+      "learning_rate": 0.0002809446546669753,
+      "loss": 3.07,
+      "step": 31846
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5843205451965332,
+      "learning_rate": 0.0002809292895911199,
+      "loss": 3.1067,
+      "step": 31847
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5955379605293274,
+      "learning_rate": 0.0002809139245654931,
+      "loss": 3.0711,
+      "step": 31848
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6512821912765503,
+      "learning_rate": 0.00028089855959013537,
+      "loss": 2.8107,
+      "step": 31849
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6019611954689026,
+      "learning_rate": 0.0002808831946650872,
+      "loss": 3.0046,
+      "step": 31850
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5904426574707031,
+      "learning_rate": 0.0002808678297903889,
+      "loss": 2.968,
+      "step": 31851
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6119365096092224,
+      "learning_rate": 0.00028085246496608115,
+      "loss": 3.2099,
+      "step": 31852
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6144260168075562,
+      "learning_rate": 0.00028083710019220434,
+      "loss": 3.1469,
+      "step": 31853
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5753419399261475,
+      "learning_rate": 0.00028082173546879876,
+      "loss": 3.1069,
+      "step": 31854
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5890007019042969,
+      "learning_rate": 0.0002808063707959052,
+      "loss": 2.9953,
+      "step": 31855
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6050313711166382,
+      "learning_rate": 0.0002807910061735639,
+      "loss": 3.1714,
+      "step": 31856
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6816890835762024,
+      "learning_rate": 0.0002807756416018155,
+      "loss": 3.093,
+      "step": 31857
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6396083831787109,
+      "learning_rate": 0.0002807602770807002,
+      "loss": 3.1671,
+      "step": 31858
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5927896499633789,
+      "learning_rate": 0.0002807449126102586,
+      "loss": 2.9865,
+      "step": 31859
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6155912280082703,
+      "learning_rate": 0.00028072954819053134,
+      "loss": 2.914,
+      "step": 31860
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5896044373512268,
+      "learning_rate": 0.0002807141838215587,
+      "loss": 3.0633,
+      "step": 31861
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6200414896011353,
+      "learning_rate": 0.00028069881950338124,
+      "loss": 2.9665,
+      "step": 31862
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5940514206886292,
+      "learning_rate": 0.00028068345523603924,
+      "loss": 3.2252,
+      "step": 31863
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6118853092193604,
+      "learning_rate": 0.00028066809101957337,
+      "loss": 2.9935,
+      "step": 31864
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.628420889377594,
+      "learning_rate": 0.00028065272685402407,
+      "loss": 2.6308,
+      "step": 31865
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6051673293113708,
+      "learning_rate": 0.0002806373627394316,
+      "loss": 3.186,
+      "step": 31866
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6414685845375061,
+      "learning_rate": 0.0002806219986758368,
+      "loss": 2.9328,
+      "step": 31867
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6532812118530273,
+      "learning_rate": 0.00028060663466327986,
+      "loss": 3.215,
+      "step": 31868
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6058521270751953,
+      "learning_rate": 0.0002805912707018012,
+      "loss": 3.1998,
+      "step": 31869
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5776923298835754,
+      "learning_rate": 0.0002805759067914416,
+      "loss": 2.9845,
+      "step": 31870
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6667451858520508,
+      "learning_rate": 0.00028056054293224117,
+      "loss": 3.1901,
+      "step": 31871
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6361920833587646,
+      "learning_rate": 0.00028054517912424066,
+      "loss": 3.0641,
+      "step": 31872
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.587245762348175,
+      "learning_rate": 0.0002805298153674802,
+      "loss": 3.1569,
+      "step": 31873
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6225107908248901,
+      "learning_rate": 0.0002805144516620006,
+      "loss": 3.0355,
+      "step": 31874
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6272902488708496,
+      "learning_rate": 0.0002804990880078422,
+      "loss": 3.0836,
+      "step": 31875
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6180784106254578,
+      "learning_rate": 0.0002804837244050454,
+      "loss": 3.0903,
+      "step": 31876
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6579588651657104,
+      "learning_rate": 0.00028046836085365075,
+      "loss": 3.1705,
+      "step": 31877
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6163791418075562,
+      "learning_rate": 0.00028045299735369864,
+      "loss": 3.1334,
+      "step": 31878
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5973280668258667,
+      "learning_rate": 0.0002804376339052295,
+      "loss": 2.9392,
+      "step": 31879
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5973444581031799,
+      "learning_rate": 0.000280422270508284,
+      "loss": 3.068,
+      "step": 31880
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5980919599533081,
+      "learning_rate": 0.0002804069071629024,
+      "loss": 3.0488,
+      "step": 31881
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5866699814796448,
+      "learning_rate": 0.0002803915438691253,
+      "loss": 2.9993,
+      "step": 31882
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6229212880134583,
+      "learning_rate": 0.00028037618062699306,
+      "loss": 3.168,
+      "step": 31883
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.629422664642334,
+      "learning_rate": 0.0002803608174365461,
+      "loss": 2.9593,
+      "step": 31884
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5796070694923401,
+      "learning_rate": 0.0002803454542978251,
+      "loss": 3.0037,
+      "step": 31885
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6071597933769226,
+      "learning_rate": 0.0002803300912108703,
+      "loss": 2.9927,
+      "step": 31886
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6303873658180237,
+      "learning_rate": 0.0002803147281757224,
+      "loss": 3.144,
+      "step": 31887
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5881187915802002,
+      "learning_rate": 0.0002802993651924215,
+      "loss": 3.1171,
+      "step": 31888
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5947704315185547,
+      "learning_rate": 0.00028028400226100844,
+      "loss": 3.233,
+      "step": 31889
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6180113554000854,
+      "learning_rate": 0.00028026863938152355,
+      "loss": 3.0989,
+      "step": 31890
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6293803453445435,
+      "learning_rate": 0.00028025327655400707,
+      "loss": 3.1379,
+      "step": 31891
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6309683322906494,
+      "learning_rate": 0.0002802379137784999,
+      "loss": 3.0034,
+      "step": 31892
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5890076160430908,
+      "learning_rate": 0.0002802225510550421,
+      "loss": 3.1913,
+      "step": 31893
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5986996293067932,
+      "learning_rate": 0.00028020718838367426,
+      "loss": 3.1027,
+      "step": 31894
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6149293184280396,
+      "learning_rate": 0.000280191825764437,
+      "loss": 2.8922,
+      "step": 31895
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5837023258209229,
+      "learning_rate": 0.00028017646319737063,
+      "loss": 2.8853,
+      "step": 31896
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5797613859176636,
+      "learning_rate": 0.00028016110068251567,
+      "loss": 3.1853,
+      "step": 31897
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6410925984382629,
+      "learning_rate": 0.00028014573821991243,
+      "loss": 3.1645,
+      "step": 31898
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5977853536605835,
+      "learning_rate": 0.0002801303758096015,
+      "loss": 3.1972,
+      "step": 31899
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6081350445747375,
+      "learning_rate": 0.00028011501345162355,
+      "loss": 3.0046,
+      "step": 31900
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.598984956741333,
+      "learning_rate": 0.0002800996511460186,
+      "loss": 2.9773,
+      "step": 31901
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5773735046386719,
+      "learning_rate": 0.0002800842888928275,
+      "loss": 2.9064,
+      "step": 31902
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5690639615058899,
+      "learning_rate": 0.0002800689266920904,
+      "loss": 2.9076,
+      "step": 31903
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6134076118469238,
+      "learning_rate": 0.00028005356454384797,
+      "loss": 2.9833,
+      "step": 31904
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5963901281356812,
+      "learning_rate": 0.0002800382024481407,
+      "loss": 2.8316,
+      "step": 31905
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5689248442649841,
+      "learning_rate": 0.0002800228404050088,
+      "loss": 3.0716,
+      "step": 31906
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6227253079414368,
+      "learning_rate": 0.00028000747841449306,
+      "loss": 2.9693,
+      "step": 31907
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6223208904266357,
+      "learning_rate": 0.0002799921164766337,
+      "loss": 3.0232,
+      "step": 31908
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5878385305404663,
+      "learning_rate": 0.00027997675459147115,
+      "loss": 2.9983,
+      "step": 31909
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6404999494552612,
+      "learning_rate": 0.00027996139275904616,
+      "loss": 3.1131,
+      "step": 31910
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6232870817184448,
+      "learning_rate": 0.0002799460309793989,
+      "loss": 3.0633,
+      "step": 31911
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6137735843658447,
+      "learning_rate": 0.00027993066925257004,
+      "loss": 2.8579,
+      "step": 31912
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5834466218948364,
+      "learning_rate": 0.0002799153075785998,
+      "loss": 3.0556,
+      "step": 31913
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6220276951789856,
+      "learning_rate": 0.0002798999459575288,
+      "loss": 3.0881,
+      "step": 31914
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5749549269676208,
+      "learning_rate": 0.00027988458438939754,
+      "loss": 3.0783,
+      "step": 31915
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5834282636642456,
+      "learning_rate": 0.0002798692228742463,
+      "loss": 3.078,
+      "step": 31916
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.59247225522995,
+      "learning_rate": 0.00027985386141211584,
+      "loss": 2.9249,
+      "step": 31917
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5814173221588135,
+      "learning_rate": 0.00027983850000304626,
+      "loss": 2.897,
+      "step": 31918
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6004323959350586,
+      "learning_rate": 0.0002798231386470782,
+      "loss": 3.0616,
+      "step": 31919
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5897133350372314,
+      "learning_rate": 0.0002798077773442522,
+      "loss": 3.1485,
+      "step": 31920
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6207877397537231,
+      "learning_rate": 0.00027979241609460854,
+      "loss": 2.8907,
+      "step": 31921
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5839530229568481,
+      "learning_rate": 0.00027977705489818787,
+      "loss": 3.0955,
+      "step": 31922
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6286572813987732,
+      "learning_rate": 0.0002797616937550305,
+      "loss": 2.9629,
+      "step": 31923
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5850244760513306,
+      "learning_rate": 0.00027974633266517686,
+      "loss": 3.3621,
+      "step": 31924
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5959276556968689,
+      "learning_rate": 0.0002797309716286676,
+      "loss": 3.0628,
+      "step": 31925
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6070095896720886,
+      "learning_rate": 0.00027971561064554296,
+      "loss": 2.8452,
+      "step": 31926
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6149983406066895,
+      "learning_rate": 0.0002797002497158436,
+      "loss": 2.8937,
+      "step": 31927
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.594487726688385,
+      "learning_rate": 0.0002796848888396098,
+      "loss": 2.9891,
+      "step": 31928
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.623718798160553,
+      "learning_rate": 0.0002796695280168821,
+      "loss": 3.0839,
+      "step": 31929
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6588249206542969,
+      "learning_rate": 0.000279654167247701,
+      "loss": 2.8436,
+      "step": 31930
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6797537803649902,
+      "learning_rate": 0.0002796388065321068,
+      "loss": 3.0175,
+      "step": 31931
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6062451601028442,
+      "learning_rate": 0.00027962344587014024,
+      "loss": 3.0375,
+      "step": 31932
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5665577054023743,
+      "learning_rate": 0.00027960808526184146,
+      "loss": 3.332,
+      "step": 31933
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5775216817855835,
+      "learning_rate": 0.00027959272470725113,
+      "loss": 3.1504,
+      "step": 31934
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6084179282188416,
+      "learning_rate": 0.0002795773642064096,
+      "loss": 3.3357,
+      "step": 31935
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6588486433029175,
+      "learning_rate": 0.0002795620037593573,
+      "loss": 2.8731,
+      "step": 31936
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6479736566543579,
+      "learning_rate": 0.0002795466433661349,
+      "loss": 3.0273,
+      "step": 31937
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5970213413238525,
+      "learning_rate": 0.00027953128302678253,
+      "loss": 2.9507,
+      "step": 31938
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5876474380493164,
+      "learning_rate": 0.00027951592274134095,
+      "loss": 2.9572,
+      "step": 31939
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5981870293617249,
+      "learning_rate": 0.00027950056250985046,
+      "loss": 3.128,
+      "step": 31940
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6277065277099609,
+      "learning_rate": 0.0002794852023323515,
+      "loss": 2.9634,
+      "step": 31941
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6169352531433105,
+      "learning_rate": 0.0002794698422088846,
+      "loss": 2.9605,
+      "step": 31942
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6255897283554077,
+      "learning_rate": 0.0002794544821394901,
+      "loss": 3.017,
+      "step": 31943
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.584186851978302,
+      "learning_rate": 0.00027943912212420873,
+      "loss": 3.0064,
+      "step": 31944
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5788871049880981,
+      "learning_rate": 0.00027942376216308064,
+      "loss": 2.9854,
+      "step": 31945
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.601641833782196,
+      "learning_rate": 0.00027940840225614634,
+      "loss": 2.8917,
+      "step": 31946
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5883685946464539,
+      "learning_rate": 0.00027939304240344644,
+      "loss": 3.2907,
+      "step": 31947
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5894083380699158,
+      "learning_rate": 0.0002793776826050213,
+      "loss": 3.0337,
+      "step": 31948
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5871928930282593,
+      "learning_rate": 0.0002793623228609114,
+      "loss": 3.1671,
+      "step": 31949
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.599830150604248,
+      "learning_rate": 0.00027934696317115707,
+      "loss": 3.0923,
+      "step": 31950
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5850623846054077,
+      "learning_rate": 0.0002793316035357989,
+      "loss": 2.9835,
+      "step": 31951
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6004115343093872,
+      "learning_rate": 0.00027931624395487734,
+      "loss": 3.0782,
+      "step": 31952
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6150354743003845,
+      "learning_rate": 0.00027930088442843274,
+      "loss": 3.1377,
+      "step": 31953
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5694727897644043,
+      "learning_rate": 0.0002792855249565058,
+      "loss": 2.9826,
+      "step": 31954
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6149030327796936,
+      "learning_rate": 0.0002792701655391366,
+      "loss": 3.1122,
+      "step": 31955
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5870358347892761,
+      "learning_rate": 0.00027925480617636584,
+      "loss": 2.9392,
+      "step": 31956
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.576116681098938,
+      "learning_rate": 0.000279239446868234,
+      "loss": 2.9264,
+      "step": 31957
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.586956262588501,
+      "learning_rate": 0.00027922408761478143,
+      "loss": 2.9759,
+      "step": 31958
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5969283580780029,
+      "learning_rate": 0.0002792087284160487,
+      "loss": 3.1103,
+      "step": 31959
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5975672602653503,
+      "learning_rate": 0.000279193369272076,
+      "loss": 2.7399,
+      "step": 31960
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5579419732093811,
+      "learning_rate": 0.000279178010182904,
+      "loss": 3.1493,
+      "step": 31961
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5934404134750366,
+      "learning_rate": 0.0002791626511485733,
+      "loss": 3.0314,
+      "step": 31962
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6547814607620239,
+      "learning_rate": 0.00027914729216912395,
+      "loss": 3.1367,
+      "step": 31963
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6256775856018066,
+      "learning_rate": 0.00027913193324459674,
+      "loss": 3.0097,
+      "step": 31964
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6281748414039612,
+      "learning_rate": 0.0002791165743750319,
+      "loss": 3.119,
+      "step": 31965
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.584899365901947,
+      "learning_rate": 0.00027910121556047005,
+      "loss": 3.0945,
+      "step": 31966
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5488184690475464,
+      "learning_rate": 0.00027908585680095163,
+      "loss": 2.9363,
+      "step": 31967
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6100363731384277,
+      "learning_rate": 0.0002790704980965169,
+      "loss": 3.102,
+      "step": 31968
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5827188491821289,
+      "learning_rate": 0.0002790551394472066,
+      "loss": 2.9794,
+      "step": 31969
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.592393696308136,
+      "learning_rate": 0.00027903978085306095,
+      "loss": 3.0996,
+      "step": 31970
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5915204882621765,
+      "learning_rate": 0.0002790244223141204,
+      "loss": 3.125,
+      "step": 31971
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6166325211524963,
+      "learning_rate": 0.0002790090638304257,
+      "loss": 3.1128,
+      "step": 31972
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6143838167190552,
+      "learning_rate": 0.00027899370540201697,
+      "loss": 2.8731,
+      "step": 31973
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5882230997085571,
+      "learning_rate": 0.0002789783470289348,
+      "loss": 3.0723,
+      "step": 31974
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5999899506568909,
+      "learning_rate": 0.0002789629887112195,
+      "loss": 3.0853,
+      "step": 31975
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6208217740058899,
+      "learning_rate": 0.00027894763044891176,
+      "loss": 2.8768,
+      "step": 31976
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6412845253944397,
+      "learning_rate": 0.0002789322722420519,
+      "loss": 3.1244,
+      "step": 31977
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5920975208282471,
+      "learning_rate": 0.0002789169140906803,
+      "loss": 3.2523,
+      "step": 31978
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.576759934425354,
+      "learning_rate": 0.0002789015559948376,
+      "loss": 3.0485,
+      "step": 31979
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6126590967178345,
+      "learning_rate": 0.0002788861979545641,
+      "loss": 2.9728,
+      "step": 31980
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6673457026481628,
+      "learning_rate": 0.0002788708399699002,
+      "loss": 2.889,
+      "step": 31981
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6051563620567322,
+      "learning_rate": 0.0002788554820408866,
+      "loss": 3.1627,
+      "step": 31982
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5865967273712158,
+      "learning_rate": 0.00027884012416756347,
+      "loss": 3.1364,
+      "step": 31983
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6102021336555481,
+      "learning_rate": 0.0002788247663499715,
+      "loss": 3.0828,
+      "step": 31984
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5698949694633484,
+      "learning_rate": 0.0002788094085881509,
+      "loss": 3.0331,
+      "step": 31985
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5987421274185181,
+      "learning_rate": 0.00027879405088214226,
+      "loss": 3.156,
+      "step": 31986
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6104761362075806,
+      "learning_rate": 0.00027877869323198606,
+      "loss": 3.0174,
+      "step": 31987
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5552363395690918,
+      "learning_rate": 0.0002787633356377227,
+      "loss": 3.0127,
+      "step": 31988
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5822331309318542,
+      "learning_rate": 0.00027874797809939264,
+      "loss": 3.1789,
+      "step": 31989
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5854426622390747,
+      "learning_rate": 0.0002787326206170362,
+      "loss": 3.1524,
+      "step": 31990
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6006827354431152,
+      "learning_rate": 0.000278717263190694,
+      "loss": 2.9734,
+      "step": 31991
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5953511595726013,
+      "learning_rate": 0.0002787019058204065,
+      "loss": 2.9979,
+      "step": 31992
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5749074220657349,
+      "learning_rate": 0.00027868654850621397,
+      "loss": 2.9802,
+      "step": 31993
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.638322114944458,
+      "learning_rate": 0.0002786711912481571,
+      "loss": 3.0781,
+      "step": 31994
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5966150760650635,
+      "learning_rate": 0.00027865583404627616,
+      "loss": 2.9011,
+      "step": 31995
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5970824956893921,
+      "learning_rate": 0.00027864047690061153,
+      "loss": 3.1345,
+      "step": 31996
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6031424403190613,
+      "learning_rate": 0.00027862511981120394,
+      "loss": 3.2617,
+      "step": 31997
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5995854735374451,
+      "learning_rate": 0.0002786097627780936,
+      "loss": 2.9957,
+      "step": 31998
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6159862279891968,
+      "learning_rate": 0.00027859440580132105,
+      "loss": 2.9234,
+      "step": 31999
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6073964238166809,
+      "learning_rate": 0.0002785790488809267,
+      "loss": 2.867,
+      "step": 32000
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2414467334747314,
+      "learning_rate": 0.00027856369201695097,
+      "loss": 3.1922,
+      "step": 32001
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.616241455078125,
+      "learning_rate": 0.0002785483352094344,
+      "loss": 2.9149,
+      "step": 32002
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6222730875015259,
+      "learning_rate": 0.0002785329784584173,
+      "loss": 3.1566,
+      "step": 32003
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5811864137649536,
+      "learning_rate": 0.0002785176217639404,
+      "loss": 3.0071,
+      "step": 32004
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7227095365524292,
+      "learning_rate": 0.00027850226512604376,
+      "loss": 2.9084,
+      "step": 32005
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5832785964012146,
+      "learning_rate": 0.00027848690854476805,
+      "loss": 3.0637,
+      "step": 32006
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6182896494865417,
+      "learning_rate": 0.0002784715520201538,
+      "loss": 3.1451,
+      "step": 32007
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6133847832679749,
+      "learning_rate": 0.0002784561955522412,
+      "loss": 2.8646,
+      "step": 32008
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6060160994529724,
+      "learning_rate": 0.0002784408391410709,
+      "loss": 3.0987,
+      "step": 32009
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6133733987808228,
+      "learning_rate": 0.0002784254827866833,
+      "loss": 2.9392,
+      "step": 32010
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5962675213813782,
+      "learning_rate": 0.0002784101264891187,
+      "loss": 3.1465,
+      "step": 32011
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5825144648551941,
+      "learning_rate": 0.00027839477024841784,
+      "loss": 3.1008,
+      "step": 32012
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5747712254524231,
+      "learning_rate": 0.0002783794140646209,
+      "loss": 3.088,
+      "step": 32013
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.624232292175293,
+      "learning_rate": 0.0002783640579377685,
+      "loss": 3.1401,
+      "step": 32014
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6441041231155396,
+      "learning_rate": 0.00027834870186790086,
+      "loss": 2.9997,
+      "step": 32015
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6096225380897522,
+      "learning_rate": 0.00027833334585505875,
+      "loss": 2.6779,
+      "step": 32016
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6041115522384644,
+      "learning_rate": 0.0002783179898992824,
+      "loss": 3.0651,
+      "step": 32017
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6201046705245972,
+      "learning_rate": 0.0002783026340006121,
+      "loss": 2.9325,
+      "step": 32018
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5794186592102051,
+      "learning_rate": 0.0002782872781590887,
+      "loss": 3.0974,
+      "step": 32019
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5904180407524109,
+      "learning_rate": 0.0002782719223747524,
+      "loss": 3.1501,
+      "step": 32020
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5724608898162842,
+      "learning_rate": 0.0002782565666476437,
+      "loss": 3.0647,
+      "step": 32021
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6130523681640625,
+      "learning_rate": 0.00027824121097780286,
+      "loss": 2.9517,
+      "step": 32022
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5855075716972351,
+      "learning_rate": 0.00027822585536527053,
+      "loss": 2.8407,
+      "step": 32023
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5709105730056763,
+      "learning_rate": 0.0002782104998100872,
+      "loss": 3.1079,
+      "step": 32024
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6038665175437927,
+      "learning_rate": 0.00027819514431229314,
+      "loss": 2.9033,
+      "step": 32025
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6230751872062683,
+      "learning_rate": 0.00027817978887192897,
+      "loss": 2.9427,
+      "step": 32026
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5995590090751648,
+      "learning_rate": 0.00027816443348903494,
+      "loss": 3.0794,
+      "step": 32027
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5960817337036133,
+      "learning_rate": 0.0002781490781636516,
+      "loss": 3.079,
+      "step": 32028
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5772078633308411,
+      "learning_rate": 0.0002781337228958195,
+      "loss": 2.9819,
+      "step": 32029
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5906760692596436,
+      "learning_rate": 0.0002781183676855788,
+      "loss": 3.1876,
+      "step": 32030
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6208398938179016,
+      "learning_rate": 0.00027810301253297024,
+      "loss": 2.9969,
+      "step": 32031
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6172988414764404,
+      "learning_rate": 0.0002780876574380341,
+      "loss": 3.1743,
+      "step": 32032
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6380127668380737,
+      "learning_rate": 0.0002780723024008107,
+      "loss": 3.0405,
+      "step": 32033
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.639022946357727,
+      "learning_rate": 0.0002780569474213408,
+      "loss": 2.8838,
+      "step": 32034
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6695070862770081,
+      "learning_rate": 0.00027804159249966465,
+      "loss": 2.9628,
+      "step": 32035
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6148467063903809,
+      "learning_rate": 0.0002780262376358228,
+      "loss": 2.9927,
+      "step": 32036
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6534361839294434,
+      "learning_rate": 0.0002780108828298554,
+      "loss": 3.1393,
+      "step": 32037
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5975500345230103,
+      "learning_rate": 0.00027799552808180323,
+      "loss": 3.2256,
+      "step": 32038
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6091464161872864,
+      "learning_rate": 0.00027798017339170664,
+      "loss": 2.9941,
+      "step": 32039
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5968720316886902,
+      "learning_rate": 0.00027796481875960597,
+      "loss": 3.15,
+      "step": 32040
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5805733799934387,
+      "learning_rate": 0.00027794946418554183,
+      "loss": 3.0558,
+      "step": 32041
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.638577401638031,
+      "learning_rate": 0.00027793410966955444,
+      "loss": 3.0992,
+      "step": 32042
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5825785398483276,
+      "learning_rate": 0.0002779187552116843,
+      "loss": 3.1721,
+      "step": 32043
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6435587406158447,
+      "learning_rate": 0.00027790340081197207,
+      "loss": 3.0042,
+      "step": 32044
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5820013880729675,
+      "learning_rate": 0.0002778880464704579,
+      "loss": 2.998,
+      "step": 32045
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5931047797203064,
+      "learning_rate": 0.0002778726921871825,
+      "loss": 3.1897,
+      "step": 32046
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6294622421264648,
+      "learning_rate": 0.0002778573379621861,
+      "loss": 2.9959,
+      "step": 32047
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5972573161125183,
+      "learning_rate": 0.0002778419837955091,
+      "loss": 3.0585,
+      "step": 32048
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6380653381347656,
+      "learning_rate": 0.0002778266296871922,
+      "loss": 3.0712,
+      "step": 32049
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6002389192581177,
+      "learning_rate": 0.0002778112756372757,
+      "loss": 3.1067,
+      "step": 32050
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6016014218330383,
+      "learning_rate": 0.0002777959216458,
+      "loss": 3.0874,
+      "step": 32051
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5852669477462769,
+      "learning_rate": 0.00027778056771280546,
+      "loss": 3.1815,
+      "step": 32052
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6085726022720337,
+      "learning_rate": 0.0002777652138383327,
+      "loss": 3.1229,
+      "step": 32053
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6045544147491455,
+      "learning_rate": 0.0002777498600224222,
+      "loss": 2.7204,
+      "step": 32054
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5995080471038818,
+      "learning_rate": 0.0002777345062651141,
+      "loss": 3.1933,
+      "step": 32055
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6304450035095215,
+      "learning_rate": 0.00027771915256644917,
+      "loss": 2.8745,
+      "step": 32056
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5954857468605042,
+      "learning_rate": 0.00027770379892646765,
+      "loss": 2.9642,
+      "step": 32057
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6180627346038818,
+      "learning_rate": 0.00027768844534520993,
+      "loss": 2.9956,
+      "step": 32058
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5869137644767761,
+      "learning_rate": 0.00027767309182271673,
+      "loss": 2.8287,
+      "step": 32059
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5612866282463074,
+      "learning_rate": 0.0002776577383590282,
+      "loss": 2.8453,
+      "step": 32060
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5565618872642517,
+      "learning_rate": 0.000277642384954185,
+      "loss": 3.1548,
+      "step": 32061
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6152032017707825,
+      "learning_rate": 0.0002776270316082273,
+      "loss": 2.8864,
+      "step": 32062
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5819271206855774,
+      "learning_rate": 0.0002776116783211957,
+      "loss": 3.0542,
+      "step": 32063
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5854024887084961,
+      "learning_rate": 0.0002775963250931308,
+      "loss": 2.9896,
+      "step": 32064
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5745155215263367,
+      "learning_rate": 0.0002775809719240727,
+      "loss": 2.971,
+      "step": 32065
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5686827898025513,
+      "learning_rate": 0.00027756561881406213,
+      "loss": 3.1338,
+      "step": 32066
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5938876867294312,
+      "learning_rate": 0.0002775502657631393,
+      "loss": 3.1614,
+      "step": 32067
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5606462359428406,
+      "learning_rate": 0.00027753491277134474,
+      "loss": 3.1713,
+      "step": 32068
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6324983239173889,
+      "learning_rate": 0.00027751955983871905,
+      "loss": 3.2225,
+      "step": 32069
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5873457193374634,
+      "learning_rate": 0.0002775042069653023,
+      "loss": 2.8439,
+      "step": 32070
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5858466625213623,
+      "learning_rate": 0.0002774888541511353,
+      "loss": 3.1174,
+      "step": 32071
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6003888249397278,
+      "learning_rate": 0.0002774735013962583,
+      "loss": 3.1971,
+      "step": 32072
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.574512243270874,
+      "learning_rate": 0.00027745814870071164,
+      "loss": 2.901,
+      "step": 32073
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5840027928352356,
+      "learning_rate": 0.00027744279606453605,
+      "loss": 2.9237,
+      "step": 32074
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5972057580947876,
+      "learning_rate": 0.0002774274434877717,
+      "loss": 3.0789,
+      "step": 32075
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6341174840927124,
+      "learning_rate": 0.0002774120909704592,
+      "loss": 3.0405,
+      "step": 32076
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6080915331840515,
+      "learning_rate": 0.0002773967385126388,
+      "loss": 3.1676,
+      "step": 32077
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5747969150543213,
+      "learning_rate": 0.00027738138611435104,
+      "loss": 3.1673,
+      "step": 32078
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6224098205566406,
+      "learning_rate": 0.00027736603377563645,
+      "loss": 3.1983,
+      "step": 32079
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6277689337730408,
+      "learning_rate": 0.0002773506814965353,
+      "loss": 2.974,
+      "step": 32080
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6136027574539185,
+      "learning_rate": 0.00027733532927708815,
+      "loss": 3.1981,
+      "step": 32081
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6523263454437256,
+      "learning_rate": 0.00027731997711733533,
+      "loss": 3.1751,
+      "step": 32082
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6148524880409241,
+      "learning_rate": 0.0002773046250173173,
+      "loss": 2.9798,
+      "step": 32083
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5856233835220337,
+      "learning_rate": 0.0002772892729770746,
+      "loss": 2.8006,
+      "step": 32084
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5549863576889038,
+      "learning_rate": 0.00027727392099664746,
+      "loss": 3.0683,
+      "step": 32085
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6235625743865967,
+      "learning_rate": 0.0002772585690760766,
+      "loss": 3.0228,
+      "step": 32086
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6109411716461182,
+      "learning_rate": 0.00027724321721540223,
+      "loss": 3.031,
+      "step": 32087
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.641261637210846,
+      "learning_rate": 0.00027722786541466474,
+      "loss": 3.1146,
+      "step": 32088
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6297993063926697,
+      "learning_rate": 0.00027721251367390483,
+      "loss": 2.9968,
+      "step": 32089
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6300150156021118,
+      "learning_rate": 0.0002771971619931627,
+      "loss": 3.0939,
+      "step": 32090
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6364719271659851,
+      "learning_rate": 0.00027718181037247896,
+      "loss": 3.026,
+      "step": 32091
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.613513171672821,
+      "learning_rate": 0.0002771664588118937,
+      "loss": 2.8987,
+      "step": 32092
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6249509453773499,
+      "learning_rate": 0.0002771511073114478,
+      "loss": 2.9277,
+      "step": 32093
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.605344295501709,
+      "learning_rate": 0.0002771357558711815,
+      "loss": 2.9627,
+      "step": 32094
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7609037160873413,
+      "learning_rate": 0.00027712040449113503,
+      "loss": 2.957,
+      "step": 32095
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5966310501098633,
+      "learning_rate": 0.0002771050531713492,
+      "loss": 3.1181,
+      "step": 32096
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6008804440498352,
+      "learning_rate": 0.00027708970191186414,
+      "loss": 2.9581,
+      "step": 32097
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5897909998893738,
+      "learning_rate": 0.0002770743507127205,
+      "loss": 3.3047,
+      "step": 32098
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.590558648109436,
+      "learning_rate": 0.0002770589995739585,
+      "loss": 3.0585,
+      "step": 32099
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5807520747184753,
+      "learning_rate": 0.0002770436484956187,
+      "loss": 2.9507,
+      "step": 32100
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6075061559677124,
+      "learning_rate": 0.0002770282974777416,
+      "loss": 2.8376,
+      "step": 32101
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6187633275985718,
+      "learning_rate": 0.0002770129465203674,
+      "loss": 3.172,
+      "step": 32102
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5896881222724915,
+      "learning_rate": 0.00027699759562353684,
+      "loss": 3.0375,
+      "step": 32103
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5933699011802673,
+      "learning_rate": 0.00027698224478729005,
+      "loss": 3.0564,
+      "step": 32104
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5940541625022888,
+      "learning_rate": 0.0002769668940116676,
+      "loss": 2.9197,
+      "step": 32105
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6250037550926208,
+      "learning_rate": 0.00027695154329671003,
+      "loss": 3.0014,
+      "step": 32106
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6533704400062561,
+      "learning_rate": 0.0002769361926424575,
+      "loss": 2.8502,
+      "step": 32107
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5797675848007202,
+      "learning_rate": 0.00027692084204895075,
+      "loss": 3.0491,
+      "step": 32108
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5860679745674133,
+      "learning_rate": 0.00027690549151623,
+      "loss": 3.176,
+      "step": 32109
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6044549345970154,
+      "learning_rate": 0.0002768901410443357,
+      "loss": 2.8871,
+      "step": 32110
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6011185646057129,
+      "learning_rate": 0.0002768747906333084,
+      "loss": 3.0566,
+      "step": 32111
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5532136559486389,
+      "learning_rate": 0.00027685944028318845,
+      "loss": 2.9295,
+      "step": 32112
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5941985249519348,
+      "learning_rate": 0.00027684408999401625,
+      "loss": 2.9165,
+      "step": 32113
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.599427342414856,
+      "learning_rate": 0.0002768287397658322,
+      "loss": 3.1695,
+      "step": 32114
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6106255650520325,
+      "learning_rate": 0.00027681338959867683,
+      "loss": 3.0471,
+      "step": 32115
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6107386946678162,
+      "learning_rate": 0.0002767980394925906,
+      "loss": 3.2058,
+      "step": 32116
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6064965128898621,
+      "learning_rate": 0.00027678268944761374,
+      "loss": 2.885,
+      "step": 32117
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6419435143470764,
+      "learning_rate": 0.00027676733946378697,
+      "loss": 3.086,
+      "step": 32118
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6065117716789246,
+      "learning_rate": 0.00027675198954115046,
+      "loss": 3.0474,
+      "step": 32119
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7930900454521179,
+      "learning_rate": 0.0002767366396797447,
+      "loss": 2.9981,
+      "step": 32120
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6305885910987854,
+      "learning_rate": 0.00027672128987961023,
+      "loss": 3.0314,
+      "step": 32121
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.599500834941864,
+      "learning_rate": 0.00027670594014078735,
+      "loss": 3.3042,
+      "step": 32122
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6495477557182312,
+      "learning_rate": 0.00027669059046331665,
+      "loss": 3.0577,
+      "step": 32123
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5801239013671875,
+      "learning_rate": 0.00027667524084723825,
+      "loss": 2.8206,
+      "step": 32124
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6565015316009521,
+      "learning_rate": 0.00027665989129259285,
+      "loss": 2.8886,
+      "step": 32125
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.612785279750824,
+      "learning_rate": 0.00027664454179942096,
+      "loss": 3.0586,
+      "step": 32126
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6405092477798462,
+      "learning_rate": 0.00027662919236776264,
+      "loss": 3.2097,
+      "step": 32127
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6418353319168091,
+      "learning_rate": 0.0002766138429976587,
+      "loss": 2.8991,
+      "step": 32128
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.598755419254303,
+      "learning_rate": 0.00027659849368914926,
+      "loss": 3.0289,
+      "step": 32129
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6165720820426941,
+      "learning_rate": 0.00027658314444227493,
+      "loss": 3.2701,
+      "step": 32130
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6199347972869873,
+      "learning_rate": 0.00027656779525707614,
+      "loss": 2.834,
+      "step": 32131
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6760289669036865,
+      "learning_rate": 0.0002765524461335932,
+      "loss": 3.128,
+      "step": 32132
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6090681552886963,
+      "learning_rate": 0.0002765370970718667,
+      "loss": 2.799,
+      "step": 32133
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6263375282287598,
+      "learning_rate": 0.0002765217480719369,
+      "loss": 2.939,
+      "step": 32134
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5948789715766907,
+      "learning_rate": 0.00027650639913384423,
+      "loss": 2.971,
+      "step": 32135
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6212787628173828,
+      "learning_rate": 0.0002764910502576293,
+      "loss": 3.0933,
+      "step": 32136
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6157195568084717,
+      "learning_rate": 0.0002764757014433324,
+      "loss": 2.8863,
+      "step": 32137
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6237986087799072,
+      "learning_rate": 0.000276460352690994,
+      "loss": 3.1061,
+      "step": 32138
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6217524409294128,
+      "learning_rate": 0.00027644500400065435,
+      "loss": 3.0781,
+      "step": 32139
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6218914985656738,
+      "learning_rate": 0.00027642965537235417,
+      "loss": 2.946,
+      "step": 32140
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6007027626037598,
+      "learning_rate": 0.0002764143068061337,
+      "loss": 3.0428,
+      "step": 32141
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5837012529373169,
+      "learning_rate": 0.0002763989583020334,
+      "loss": 2.8888,
+      "step": 32142
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5839048624038696,
+      "learning_rate": 0.00027638360986009374,
+      "loss": 3.089,
+      "step": 32143
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6786932945251465,
+      "learning_rate": 0.0002763682614803551,
+      "loss": 2.9294,
+      "step": 32144
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6156800985336304,
+      "learning_rate": 0.0002763529131628578,
+      "loss": 3.2307,
+      "step": 32145
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6580122709274292,
+      "learning_rate": 0.0002763375649076425,
+      "loss": 2.9384,
+      "step": 32146
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6277601718902588,
+      "learning_rate": 0.0002763222167147494,
+      "loss": 2.9617,
+      "step": 32147
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6300899386405945,
+      "learning_rate": 0.0002763068685842191,
+      "loss": 3.0255,
+      "step": 32148
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5806252956390381,
+      "learning_rate": 0.000276291520516092,
+      "loss": 2.7774,
+      "step": 32149
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5649163126945496,
+      "learning_rate": 0.0002762761725104083,
+      "loss": 3.0464,
+      "step": 32150
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6552037000656128,
+      "learning_rate": 0.0002762608245672088,
+      "loss": 2.938,
+      "step": 32151
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5685431957244873,
+      "learning_rate": 0.0002762454766865336,
+      "loss": 3.0988,
+      "step": 32152
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5887003540992737,
+      "learning_rate": 0.00027623012886842334,
+      "loss": 2.9508,
+      "step": 32153
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6270775198936462,
+      "learning_rate": 0.0002762147811129182,
+      "loss": 3.3599,
+      "step": 32154
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5925376415252686,
+      "learning_rate": 0.0002761994334200588,
+      "loss": 2.8112,
+      "step": 32155
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6008676290512085,
+      "learning_rate": 0.00027618408578988557,
+      "loss": 3.0296,
+      "step": 32156
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6249087452888489,
+      "learning_rate": 0.0002761687382224388,
+      "loss": 3.0094,
+      "step": 32157
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6997944116592407,
+      "learning_rate": 0.00027615339071775905,
+      "loss": 2.9996,
+      "step": 32158
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6397016644477844,
+      "learning_rate": 0.00027613804327588666,
+      "loss": 3.184,
+      "step": 32159
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6241958141326904,
+      "learning_rate": 0.000276122695896862,
+      "loss": 3.1826,
+      "step": 32160
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5984312295913696,
+      "learning_rate": 0.0002761073485807257,
+      "loss": 2.841,
+      "step": 32161
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6074608564376831,
+      "learning_rate": 0.00027609200132751793,
+      "loss": 2.9895,
+      "step": 32162
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6068947315216064,
+      "learning_rate": 0.0002760766541372794,
+      "loss": 2.7853,
+      "step": 32163
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6099889278411865,
+      "learning_rate": 0.0002760613070100501,
+      "loss": 3.0638,
+      "step": 32164
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.640886127948761,
+      "learning_rate": 0.0002760459599458709,
+      "loss": 3.0396,
+      "step": 32165
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6190171241760254,
+      "learning_rate": 0.00027603061294478203,
+      "loss": 3.0839,
+      "step": 32166
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5623420476913452,
+      "learning_rate": 0.0002760152660068238,
+      "loss": 2.9209,
+      "step": 32167
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.588102400302887,
+      "learning_rate": 0.0002759999191320368,
+      "loss": 3.0652,
+      "step": 32168
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5879389047622681,
+      "learning_rate": 0.0002759845723204613,
+      "loss": 2.8914,
+      "step": 32169
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6012025475502014,
+      "learning_rate": 0.000275969225572138,
+      "loss": 3.1915,
+      "step": 32170
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6043398380279541,
+      "learning_rate": 0.00027595387888710703,
+      "loss": 3.0019,
+      "step": 32171
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6456342935562134,
+      "learning_rate": 0.00027593853226540887,
+      "loss": 2.8778,
+      "step": 32172
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5889996886253357,
+      "learning_rate": 0.0002759231857070841,
+      "loss": 2.9956,
+      "step": 32173
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.697205126285553,
+      "learning_rate": 0.00027590783921217294,
+      "loss": 3.0563,
+      "step": 32174
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5691419243812561,
+      "learning_rate": 0.000275892492780716,
+      "loss": 3.1269,
+      "step": 32175
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6309320330619812,
+      "learning_rate": 0.00027587714641275345,
+      "loss": 2.7833,
+      "step": 32176
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5982531905174255,
+      "learning_rate": 0.0002758618001083259,
+      "loss": 2.9703,
+      "step": 32177
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5961381793022156,
+      "learning_rate": 0.0002758464538674738,
+      "loss": 3.017,
+      "step": 32178
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6061530113220215,
+      "learning_rate": 0.00027583110769023735,
+      "loss": 3.2634,
+      "step": 32179
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5718122124671936,
+      "learning_rate": 0.00027581576157665726,
+      "loss": 3.1522,
+      "step": 32180
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.632178544998169,
+      "learning_rate": 0.0002758004155267738,
+      "loss": 2.8639,
+      "step": 32181
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6122092008590698,
+      "learning_rate": 0.0002757850695406272,
+      "loss": 2.8233,
+      "step": 32182
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6195982694625854,
+      "learning_rate": 0.00027576972361825823,
+      "loss": 3.092,
+      "step": 32183
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6117836833000183,
+      "learning_rate": 0.0002757543777597071,
+      "loss": 3.2046,
+      "step": 32184
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6399086117744446,
+      "learning_rate": 0.0002757390319650144,
+      "loss": 3.0408,
+      "step": 32185
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6001808047294617,
+      "learning_rate": 0.0002757236862342202,
+      "loss": 2.9358,
+      "step": 32186
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5729873776435852,
+      "learning_rate": 0.00027570834056736516,
+      "loss": 3.1652,
+      "step": 32187
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5765482187271118,
+      "learning_rate": 0.00027569299496448983,
+      "loss": 3.0379,
+      "step": 32188
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5675467848777771,
+      "learning_rate": 0.00027567764942563436,
+      "loss": 3.0612,
+      "step": 32189
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5885112285614014,
+      "learning_rate": 0.0002756623039508394,
+      "loss": 2.9927,
+      "step": 32190
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6263838410377502,
+      "learning_rate": 0.0002756469585401451,
+      "loss": 3.1661,
+      "step": 32191
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.580174446105957,
+      "learning_rate": 0.00027563161319359207,
+      "loss": 3.057,
+      "step": 32192
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5984945297241211,
+      "learning_rate": 0.0002756162679112208,
+      "loss": 3.1215,
+      "step": 32193
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.588066041469574,
+      "learning_rate": 0.0002756009226930714,
+      "loss": 3.1108,
+      "step": 32194
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6735916137695312,
+      "learning_rate": 0.0002755855775391846,
+      "loss": 2.7924,
+      "step": 32195
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6410754919052124,
+      "learning_rate": 0.0002755702324496007,
+      "loss": 2.906,
+      "step": 32196
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5920937061309814,
+      "learning_rate": 0.00027555488742436,
+      "loss": 3.1199,
+      "step": 32197
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5957187414169312,
+      "learning_rate": 0.0002755395424635031,
+      "loss": 2.8848,
+      "step": 32198
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5822932124137878,
+      "learning_rate": 0.0002755241975670703,
+      "loss": 2.7998,
+      "step": 32199
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6525944471359253,
+      "learning_rate": 0.00027550885273510214,
+      "loss": 3.3601,
+      "step": 32200
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6136645674705505,
+      "learning_rate": 0.0002754935079676388,
+      "loss": 2.8207,
+      "step": 32201
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5625642538070679,
+      "learning_rate": 0.0002754781632647209,
+      "loss": 2.9618,
+      "step": 32202
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6007310748100281,
+      "learning_rate": 0.0002754628186263889,
+      "loss": 3.0686,
+      "step": 32203
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5860627889633179,
+      "learning_rate": 0.00027544747405268297,
+      "loss": 2.9696,
+      "step": 32204
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6014703512191772,
+      "learning_rate": 0.00027543212954364376,
+      "loss": 3.1983,
+      "step": 32205
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6181617975234985,
+      "learning_rate": 0.0002754167850993116,
+      "loss": 3.234,
+      "step": 32206
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6260382533073425,
+      "learning_rate": 0.00027540144071972676,
+      "loss": 2.9967,
+      "step": 32207
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5667868256568909,
+      "learning_rate": 0.00027538609640492996,
+      "loss": 2.902,
+      "step": 32208
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5965689420700073,
+      "learning_rate": 0.0002753707521549613,
+      "loss": 3.1618,
+      "step": 32209
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6163270473480225,
+      "learning_rate": 0.0002753554079698615,
+      "loss": 2.993,
+      "step": 32210
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6091551780700684,
+      "learning_rate": 0.0002753400638496707,
+      "loss": 3.1026,
+      "step": 32211
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.615360677242279,
+      "learning_rate": 0.00027532471979442943,
+      "loss": 3.3256,
+      "step": 32212
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.58626389503479,
+      "learning_rate": 0.00027530937580417814,
+      "loss": 3.1256,
+      "step": 32213
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6162660717964172,
+      "learning_rate": 0.0002752940318789572,
+      "loss": 2.9119,
+      "step": 32214
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.630847692489624,
+      "learning_rate": 0.0002752786880188071,
+      "loss": 3.0364,
+      "step": 32215
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5829058289527893,
+      "learning_rate": 0.000275263344223768,
+      "loss": 3.141,
+      "step": 32216
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6056093573570251,
+      "learning_rate": 0.00027524800049388056,
+      "loss": 3.0728,
+      "step": 32217
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6000507473945618,
+      "learning_rate": 0.0002752326568291852,
+      "loss": 2.8014,
+      "step": 32218
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5783975124359131,
+      "learning_rate": 0.0002752173132297221,
+      "loss": 2.9118,
+      "step": 32219
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5872189402580261,
+      "learning_rate": 0.0002752019696955321,
+      "loss": 2.8498,
+      "step": 32220
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6652278900146484,
+      "learning_rate": 0.0002751866262266551,
+      "loss": 3.0035,
+      "step": 32221
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5831838846206665,
+      "learning_rate": 0.0002751712828231317,
+      "loss": 3.2491,
+      "step": 32222
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6078551411628723,
+      "learning_rate": 0.00027515593948500256,
+      "loss": 3.0388,
+      "step": 32223
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5786187648773193,
+      "learning_rate": 0.0002751405962123078,
+      "loss": 2.9939,
+      "step": 32224
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5971993207931519,
+      "learning_rate": 0.00027512525300508805,
+      "loss": 3.2239,
+      "step": 32225
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5976489782333374,
+      "learning_rate": 0.0002751099098633834,
+      "loss": 3.0328,
+      "step": 32226
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5616191029548645,
+      "learning_rate": 0.0002750945667872345,
+      "loss": 2.9874,
+      "step": 32227
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6311202049255371,
+      "learning_rate": 0.0002750792237766819,
+      "loss": 2.9954,
+      "step": 32228
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5888649821281433,
+      "learning_rate": 0.0002750638808317656,
+      "loss": 3.1133,
+      "step": 32229
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6184988617897034,
+      "learning_rate": 0.0002750485379525264,
+      "loss": 3.1047,
+      "step": 32230
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6103339791297913,
+      "learning_rate": 0.0002750331951390044,
+      "loss": 3.2664,
+      "step": 32231
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5738555788993835,
+      "learning_rate": 0.0002750178523912402,
+      "loss": 3.1224,
+      "step": 32232
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6099984049797058,
+      "learning_rate": 0.00027500250970927424,
+      "loss": 3.1296,
+      "step": 32233
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6913991570472717,
+      "learning_rate": 0.0002749871670931468,
+      "loss": 2.9424,
+      "step": 32234
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5921345353126526,
+      "learning_rate": 0.0002749718245428984,
+      "loss": 3.1274,
+      "step": 32235
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5641476511955261,
+      "learning_rate": 0.00027495648205856936,
+      "loss": 3.1257,
+      "step": 32236
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6268711090087891,
+      "learning_rate": 0.0002749411396402001,
+      "loss": 3.0799,
+      "step": 32237
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6031218767166138,
+      "learning_rate": 0.0002749257972878311,
+      "loss": 2.9939,
+      "step": 32238
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5958530902862549,
+      "learning_rate": 0.00027491045500150274,
+      "loss": 3.1693,
+      "step": 32239
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5875588059425354,
+      "learning_rate": 0.00027489511278125546,
+      "loss": 3.2846,
+      "step": 32240
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5789769887924194,
+      "learning_rate": 0.0002748797706271295,
+      "loss": 2.9361,
+      "step": 32241
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5652769207954407,
+      "learning_rate": 0.0002748644285391654,
+      "loss": 3.0612,
+      "step": 32242
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.57442307472229,
+      "learning_rate": 0.00027484908651740367,
+      "loss": 3.177,
+      "step": 32243
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6496897339820862,
+      "learning_rate": 0.00027483374456188447,
+      "loss": 2.985,
+      "step": 32244
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5900055170059204,
+      "learning_rate": 0.00027481840267264845,
+      "loss": 2.9613,
+      "step": 32245
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6131547093391418,
+      "learning_rate": 0.0002748030608497359,
+      "loss": 3.0876,
+      "step": 32246
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6449847221374512,
+      "learning_rate": 0.00027478771909318725,
+      "loss": 2.7354,
+      "step": 32247
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5974999666213989,
+      "learning_rate": 0.0002747723774030428,
+      "loss": 3.2546,
+      "step": 32248
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6547728776931763,
+      "learning_rate": 0.000274757035779343,
+      "loss": 3.1246,
+      "step": 32249
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6210682988166809,
+      "learning_rate": 0.00027474169422212846,
+      "loss": 3.1677,
+      "step": 32250
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.924889326095581,
+      "learning_rate": 0.00027472635273143937,
+      "loss": 3.046,
+      "step": 32251
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5950158834457397,
+      "learning_rate": 0.0002747110113073163,
+      "loss": 3.2527,
+      "step": 32252
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6571494340896606,
+      "learning_rate": 0.00027469566994979945,
+      "loss": 3.0866,
+      "step": 32253
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6341148614883423,
+      "learning_rate": 0.00027468032865892936,
+      "loss": 2.8808,
+      "step": 32254
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5878071188926697,
+      "learning_rate": 0.0002746649874347465,
+      "loss": 2.9232,
+      "step": 32255
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6338520646095276,
+      "learning_rate": 0.00027464964627729103,
+      "loss": 3.1164,
+      "step": 32256
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5606010556221008,
+      "learning_rate": 0.00027463430518660364,
+      "loss": 3.1759,
+      "step": 32257
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6088926196098328,
+      "learning_rate": 0.00027461896416272456,
+      "loss": 3.0883,
+      "step": 32258
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.580753743648529,
+      "learning_rate": 0.00027460362320569414,
+      "loss": 2.9957,
+      "step": 32259
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5840525031089783,
+      "learning_rate": 0.0002745882823155531,
+      "loss": 3.1201,
+      "step": 32260
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5990431904792786,
+      "learning_rate": 0.0002745729414923415,
+      "loss": 3.1314,
+      "step": 32261
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6257085204124451,
+      "learning_rate": 0.00027455760073609995,
+      "loss": 3.0842,
+      "step": 32262
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5750504732131958,
+      "learning_rate": 0.00027454226004686866,
+      "loss": 3.148,
+      "step": 32263
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6597176790237427,
+      "learning_rate": 0.0002745269194246882,
+      "loss": 3.263,
+      "step": 32264
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5871362686157227,
+      "learning_rate": 0.000274511578869599,
+      "loss": 3.0359,
+      "step": 32265
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5987416505813599,
+      "learning_rate": 0.00027449623838164135,
+      "loss": 2.9311,
+      "step": 32266
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6555812358856201,
+      "learning_rate": 0.00027448089796085575,
+      "loss": 3.0914,
+      "step": 32267
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5952374339103699,
+      "learning_rate": 0.00027446555760728255,
+      "loss": 2.9107,
+      "step": 32268
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5956671237945557,
+      "learning_rate": 0.00027445021732096204,
+      "loss": 3.183,
+      "step": 32269
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5921088457107544,
+      "learning_rate": 0.00027443487710193486,
+      "loss": 3.1328,
+      "step": 32270
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5908499360084534,
+      "learning_rate": 0.0002744195369502412,
+      "loss": 3.1227,
+      "step": 32271
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6202715635299683,
+      "learning_rate": 0.0002744041968659217,
+      "loss": 2.9904,
+      "step": 32272
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7672480344772339,
+      "learning_rate": 0.00027438885684901647,
+      "loss": 2.8866,
+      "step": 32273
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6302787065505981,
+      "learning_rate": 0.00027437351689956613,
+      "loss": 3.0427,
+      "step": 32274
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5923793911933899,
+      "learning_rate": 0.000274358177017611,
+      "loss": 2.9583,
+      "step": 32275
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.584662139415741,
+      "learning_rate": 0.00027434283720319155,
+      "loss": 2.9949,
+      "step": 32276
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5973477959632874,
+      "learning_rate": 0.00027432749745634813,
+      "loss": 3.0155,
+      "step": 32277
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5955763459205627,
+      "learning_rate": 0.00027431215777712104,
+      "loss": 3.1771,
+      "step": 32278
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5979658365249634,
+      "learning_rate": 0.0002742968181655508,
+      "loss": 3.0354,
+      "step": 32279
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6078369617462158,
+      "learning_rate": 0.0002742814786216779,
+      "loss": 3.127,
+      "step": 32280
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6542069315910339,
+      "learning_rate": 0.00027426613914554253,
+      "loss": 2.921,
+      "step": 32281
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6545595526695251,
+      "learning_rate": 0.0002742507997371853,
+      "loss": 3.0068,
+      "step": 32282
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5855773091316223,
+      "learning_rate": 0.00027423546039664643,
+      "loss": 3.1546,
+      "step": 32283
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6614902019500732,
+      "learning_rate": 0.0002742201211239663,
+      "loss": 3.1364,
+      "step": 32284
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6412147879600525,
+      "learning_rate": 0.0002742047819191856,
+      "loss": 2.7693,
+      "step": 32285
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6186434030532837,
+      "learning_rate": 0.00027418944278234447,
+      "loss": 2.9929,
+      "step": 32286
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6143002510070801,
+      "learning_rate": 0.0002741741037134834,
+      "loss": 2.8651,
+      "step": 32287
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6143453121185303,
+      "learning_rate": 0.0002741587647126427,
+      "loss": 3.2461,
+      "step": 32288
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6308348774909973,
+      "learning_rate": 0.00027414342577986283,
+      "loss": 2.9028,
+      "step": 32289
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.588534414768219,
+      "learning_rate": 0.00027412808691518436,
+      "loss": 3.1194,
+      "step": 32290
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6186642646789551,
+      "learning_rate": 0.00027411274811864735,
+      "loss": 3.0384,
+      "step": 32291
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5939307808876038,
+      "learning_rate": 0.0002740974093902925,
+      "loss": 3.1793,
+      "step": 32292
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6168650984764099,
+      "learning_rate": 0.00027408207073016,
+      "loss": 3.1562,
+      "step": 32293
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6111552119255066,
+      "learning_rate": 0.00027406673213829037,
+      "loss": 3.2366,
+      "step": 32294
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.612614631652832,
+      "learning_rate": 0.000274051393614724,
+      "loss": 2.9828,
+      "step": 32295
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5782878398895264,
+      "learning_rate": 0.0002740360551595012,
+      "loss": 3.1741,
+      "step": 32296
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6033691167831421,
+      "learning_rate": 0.0002740207167726626,
+      "loss": 2.99,
+      "step": 32297
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5962322950363159,
+      "learning_rate": 0.00027400537845424826,
+      "loss": 3.293,
+      "step": 32298
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6039497256278992,
+      "learning_rate": 0.0002739900402042987,
+      "loss": 3.2375,
+      "step": 32299
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6100921630859375,
+      "learning_rate": 0.0002739747020228546,
+      "loss": 3.096,
+      "step": 32300
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5819284319877625,
+      "learning_rate": 0.000273959363909956,
+      "loss": 3.0614,
+      "step": 32301
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6027130484580994,
+      "learning_rate": 0.0002739440258656435,
+      "loss": 2.9864,
+      "step": 32302
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5925067663192749,
+      "learning_rate": 0.00027392868788995723,
+      "loss": 3.2436,
+      "step": 32303
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6477892398834229,
+      "learning_rate": 0.0002739133499829379,
+      "loss": 3.0252,
+      "step": 32304
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6437166929244995,
+      "learning_rate": 0.00027389801214462587,
+      "loss": 3.0984,
+      "step": 32305
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.60708087682724,
+      "learning_rate": 0.00027388267437506125,
+      "loss": 2.945,
+      "step": 32306
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6095942258834839,
+      "learning_rate": 0.0002738673366742848,
+      "loss": 3.1028,
+      "step": 32307
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5710780024528503,
+      "learning_rate": 0.00027385199904233674,
+      "loss": 2.9135,
+      "step": 32308
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5993046164512634,
+      "learning_rate": 0.0002738366614792574,
+      "loss": 2.9611,
+      "step": 32309
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6444818377494812,
+      "learning_rate": 0.00027382132398508734,
+      "loss": 3.0658,
+      "step": 32310
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6341570615768433,
+      "learning_rate": 0.0002738059865598668,
+      "loss": 3.1248,
+      "step": 32311
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6394447684288025,
+      "learning_rate": 0.00027379064920363637,
+      "loss": 2.9208,
+      "step": 32312
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.613353431224823,
+      "learning_rate": 0.00027377531191643623,
+      "loss": 3.16,
+      "step": 32313
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5999470949172974,
+      "learning_rate": 0.0002737599746983068,
+      "loss": 2.9717,
+      "step": 32314
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6254415512084961,
+      "learning_rate": 0.00027374463754928867,
+      "loss": 3.171,
+      "step": 32315
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6086492538452148,
+      "learning_rate": 0.0002737293004694221,
+      "loss": 3.0757,
+      "step": 32316
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6170065402984619,
+      "learning_rate": 0.00027371396345874754,
+      "loss": 2.8206,
+      "step": 32317
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6064412593841553,
+      "learning_rate": 0.0002736986265173052,
+      "loss": 3.0026,
+      "step": 32318
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6228976249694824,
+      "learning_rate": 0.0002736832896451357,
+      "loss": 3.1017,
+      "step": 32319
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6091843247413635,
+      "learning_rate": 0.0002736679528422794,
+      "loss": 2.9123,
+      "step": 32320
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5790774822235107,
+      "learning_rate": 0.0002736526161087765,
+      "loss": 2.8931,
+      "step": 32321
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6052638292312622,
+      "learning_rate": 0.00027363727944466773,
+      "loss": 3.2133,
+      "step": 32322
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6797242164611816,
+      "learning_rate": 0.00027362194284999316,
+      "loss": 3.1776,
+      "step": 32323
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5846360921859741,
+      "learning_rate": 0.00027360660632479346,
+      "loss": 2.9818,
+      "step": 32324
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5824291706085205,
+      "learning_rate": 0.00027359126986910867,
+      "loss": 2.9715,
+      "step": 32325
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5969865322113037,
+      "learning_rate": 0.00027357593348297947,
+      "loss": 3.0331,
+      "step": 32326
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6006813049316406,
+      "learning_rate": 0.00027356059716644623,
+      "loss": 3.1214,
+      "step": 32327
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6340086460113525,
+      "learning_rate": 0.0002735452609195492,
+      "loss": 3.1599,
+      "step": 32328
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.610417902469635,
+      "learning_rate": 0.00027352992474232894,
+      "loss": 3.0844,
+      "step": 32329
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6211175322532654,
+      "learning_rate": 0.00027351458863482576,
+      "loss": 2.9915,
+      "step": 32330
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6344559192657471,
+      "learning_rate": 0.00027349925259708,
+      "loss": 2.8198,
+      "step": 32331
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5910095572471619,
+      "learning_rate": 0.00027348391662913217,
+      "loss": 3.0135,
+      "step": 32332
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6326538920402527,
+      "learning_rate": 0.0002734685807310225,
+      "loss": 3.0308,
+      "step": 32333
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6006748676300049,
+      "learning_rate": 0.0002734532449027916,
+      "loss": 2.9874,
+      "step": 32334
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6194648742675781,
+      "learning_rate": 0.0002734379091444797,
+      "loss": 3.0995,
+      "step": 32335
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6255232095718384,
+      "learning_rate": 0.00027342257345612715,
+      "loss": 2.9889,
+      "step": 32336
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6315654516220093,
+      "learning_rate": 0.0002734072378377746,
+      "loss": 3.1515,
+      "step": 32337
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6415723562240601,
+      "learning_rate": 0.0002733919022894622,
+      "loss": 2.9541,
+      "step": 32338
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6148661375045776,
+      "learning_rate": 0.0002733765668112304,
+      "loss": 3.0006,
+      "step": 32339
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7170533537864685,
+      "learning_rate": 0.00027336123140311955,
+      "loss": 2.9005,
+      "step": 32340
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5991891622543335,
+      "learning_rate": 0.0002733458960651701,
+      "loss": 2.9749,
+      "step": 32341
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5969253182411194,
+      "learning_rate": 0.00027333056079742257,
+      "loss": 2.8458,
+      "step": 32342
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6092698574066162,
+      "learning_rate": 0.000273315225599917,
+      "loss": 3.0701,
+      "step": 32343
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6403960585594177,
+      "learning_rate": 0.00027329989047269413,
+      "loss": 2.9416,
+      "step": 32344
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5866044759750366,
+      "learning_rate": 0.0002732845554157942,
+      "loss": 3.0471,
+      "step": 32345
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6154597997665405,
+      "learning_rate": 0.00027326922042925745,
+      "loss": 3.141,
+      "step": 32346
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6291300058364868,
+      "learning_rate": 0.0002732538855131246,
+      "loss": 2.9708,
+      "step": 32347
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.615016520023346,
+      "learning_rate": 0.0002732385506674359,
+      "loss": 2.903,
+      "step": 32348
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6233264803886414,
+      "learning_rate": 0.0002732232158922316,
+      "loss": 3.0574,
+      "step": 32349
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5890025496482849,
+      "learning_rate": 0.0002732078811875522,
+      "loss": 3.0299,
+      "step": 32350
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5658774971961975,
+      "learning_rate": 0.00027319254655343804,
+      "loss": 2.972,
+      "step": 32351
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6112178564071655,
+      "learning_rate": 0.00027317721198992967,
+      "loss": 2.9405,
+      "step": 32352
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6809776425361633,
+      "learning_rate": 0.00027316187749706737,
+      "loss": 3.1299,
+      "step": 32353
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6142784953117371,
+      "learning_rate": 0.0002731465430748915,
+      "loss": 3.2225,
+      "step": 32354
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5717936158180237,
+      "learning_rate": 0.00027313120872344236,
+      "loss": 3.1395,
+      "step": 32355
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6221408843994141,
+      "learning_rate": 0.0002731158744427605,
+      "loss": 3.1424,
+      "step": 32356
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6313767433166504,
+      "learning_rate": 0.00027310054023288634,
+      "loss": 3.0057,
+      "step": 32357
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.577363908290863,
+      "learning_rate": 0.00027308520609386003,
+      "loss": 3.0129,
+      "step": 32358
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6358572840690613,
+      "learning_rate": 0.00027306987202572227,
+      "loss": 3.0684,
+      "step": 32359
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6030645370483398,
+      "learning_rate": 0.0002730545380285132,
+      "loss": 3.105,
+      "step": 32360
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.574252188205719,
+      "learning_rate": 0.0002730392041022732,
+      "loss": 3.2618,
+      "step": 32361
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5889049172401428,
+      "learning_rate": 0.0002730238702470429,
+      "loss": 3.1371,
+      "step": 32362
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5928773283958435,
+      "learning_rate": 0.00027300853646286245,
+      "loss": 2.8395,
+      "step": 32363
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6077459454536438,
+      "learning_rate": 0.00027299320274977245,
+      "loss": 2.9993,
+      "step": 32364
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6264618635177612,
+      "learning_rate": 0.000272977869107813,
+      "loss": 3.1145,
+      "step": 32365
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6416333913803101,
+      "learning_rate": 0.0002729625355370247,
+      "loss": 3.0723,
+      "step": 32366
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6093899607658386,
+      "learning_rate": 0.0002729472020374479,
+      "loss": 2.8618,
+      "step": 32367
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6002346277236938,
+      "learning_rate": 0.00027293186860912285,
+      "loss": 3.0874,
+      "step": 32368
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5929070115089417,
+      "learning_rate": 0.0002729165352520902,
+      "loss": 3.2543,
+      "step": 32369
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6260755062103271,
+      "learning_rate": 0.0002729012019663901,
+      "loss": 3.0492,
+      "step": 32370
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6383898258209229,
+      "learning_rate": 0.00027288586875206296,
+      "loss": 2.941,
+      "step": 32371
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6028842926025391,
+      "learning_rate": 0.0002728705356091494,
+      "loss": 3.1908,
+      "step": 32372
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6282774209976196,
+      "learning_rate": 0.00027285520253768944,
+      "loss": 3.0125,
+      "step": 32373
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5710699558258057,
+      "learning_rate": 0.00027283986953772376,
+      "loss": 3.0275,
+      "step": 32374
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6519010066986084,
+      "learning_rate": 0.00027282453660929266,
+      "loss": 2.6907,
+      "step": 32375
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5964446067810059,
+      "learning_rate": 0.0002728092037524364,
+      "loss": 3.0842,
+      "step": 32376
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.564836323261261,
+      "learning_rate": 0.00027279387096719557,
+      "loss": 3.0647,
+      "step": 32377
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5880719423294067,
+      "learning_rate": 0.00027277853825361035,
+      "loss": 3.0947,
+      "step": 32378
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6163837909698486,
+      "learning_rate": 0.00027276320561172136,
+      "loss": 3.095,
+      "step": 32379
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6298701763153076,
+      "learning_rate": 0.0002727478730415687,
+      "loss": 2.9519,
+      "step": 32380
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6387132406234741,
+      "learning_rate": 0.000272732540543193,
+      "loss": 3.0831,
+      "step": 32381
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5862566232681274,
+      "learning_rate": 0.0002727172081166345,
+      "loss": 3.0494,
+      "step": 32382
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5777621269226074,
+      "learning_rate": 0.00027270187576193354,
+      "loss": 3.0085,
+      "step": 32383
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5978364944458008,
+      "learning_rate": 0.00027268654347913074,
+      "loss": 2.8669,
+      "step": 32384
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6363843679428101,
+      "learning_rate": 0.0002726712112682662,
+      "loss": 2.9042,
+      "step": 32385
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5844437479972839,
+      "learning_rate": 0.00027265587912938047,
+      "loss": 3.1127,
+      "step": 32386
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5706456899642944,
+      "learning_rate": 0.00027264054706251393,
+      "loss": 2.8488,
+      "step": 32387
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5758098363876343,
+      "learning_rate": 0.0002726252150677069,
+      "loss": 3.4247,
+      "step": 32388
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.647951602935791,
+      "learning_rate": 0.00027260988314499983,
+      "loss": 3.069,
+      "step": 32389
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6163773536682129,
+      "learning_rate": 0.000272594551294433,
+      "loss": 3.0409,
+      "step": 32390
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5810932517051697,
+      "learning_rate": 0.00027257921951604673,
+      "loss": 3.0328,
+      "step": 32391
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6113194823265076,
+      "learning_rate": 0.0002725638878098818,
+      "loss": 3.0184,
+      "step": 32392
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6727303266525269,
+      "learning_rate": 0.00027254855617597806,
+      "loss": 3.1082,
+      "step": 32393
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5940728783607483,
+      "learning_rate": 0.00027253322461437635,
+      "loss": 3.158,
+      "step": 32394
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6131221652030945,
+      "learning_rate": 0.0002725178931251166,
+      "loss": 3.0268,
+      "step": 32395
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.591697633266449,
+      "learning_rate": 0.0002725025617082396,
+      "loss": 3.1224,
+      "step": 32396
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5704348683357239,
+      "learning_rate": 0.0002724872303637855,
+      "loss": 2.9679,
+      "step": 32397
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6032789349555969,
+      "learning_rate": 0.0002724718990917947,
+      "loss": 2.959,
+      "step": 32398
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5963385701179504,
+      "learning_rate": 0.00027245656789230775,
+      "loss": 3.1237,
+      "step": 32399
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.625329315662384,
+      "learning_rate": 0.00027244123676536486,
+      "loss": 2.9193,
+      "step": 32400
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5919100046157837,
+      "learning_rate": 0.0002724259057110063,
+      "loss": 3.141,
+      "step": 32401
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6092190146446228,
+      "learning_rate": 0.0002724105747292728,
+      "loss": 2.9224,
+      "step": 32402
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5872280597686768,
+      "learning_rate": 0.0002723952438202044,
+      "loss": 3.3013,
+      "step": 32403
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5837467908859253,
+      "learning_rate": 0.00027237991298384174,
+      "loss": 2.8526,
+      "step": 32404
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6181141138076782,
+      "learning_rate": 0.0002723645822202249,
+      "loss": 2.9058,
+      "step": 32405
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6013671159744263,
+      "learning_rate": 0.0002723492515293946,
+      "loss": 3.0388,
+      "step": 32406
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5945082306861877,
+      "learning_rate": 0.00027233392091139096,
+      "loss": 2.9561,
+      "step": 32407
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.565525233745575,
+      "learning_rate": 0.00027231859036625444,
+      "loss": 3.0511,
+      "step": 32408
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6196669936180115,
+      "learning_rate": 0.0002723032598940255,
+      "loss": 3.0336,
+      "step": 32409
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6390196681022644,
+      "learning_rate": 0.0002722879294947444,
+      "loss": 2.7013,
+      "step": 32410
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6164066791534424,
+      "learning_rate": 0.0002722725991684516,
+      "loss": 3.1329,
+      "step": 32411
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.570892870426178,
+      "learning_rate": 0.00027225726891518735,
+      "loss": 3.1695,
+      "step": 32412
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6367611289024353,
+      "learning_rate": 0.000272241938734992,
+      "loss": 3.1406,
+      "step": 32413
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6313506960868835,
+      "learning_rate": 0.0002722266086279063,
+      "loss": 3.0672,
+      "step": 32414
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5836746096611023,
+      "learning_rate": 0.0002722112785939703,
+      "loss": 3.2579,
+      "step": 32415
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5823087692260742,
+      "learning_rate": 0.00027219594863322447,
+      "loss": 2.9865,
+      "step": 32416
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6018955111503601,
+      "learning_rate": 0.00027218061874570903,
+      "loss": 2.9428,
+      "step": 32417
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6570754647254944,
+      "learning_rate": 0.0002721652889314645,
+      "loss": 3.081,
+      "step": 32418
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6093233227729797,
+      "learning_rate": 0.00027214995919053143,
+      "loss": 2.591,
+      "step": 32419
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5643780827522278,
+      "learning_rate": 0.00027213462952294977,
+      "loss": 3.0753,
+      "step": 32420
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5750892162322998,
+      "learning_rate": 0.0002721192999287603,
+      "loss": 2.9966,
+      "step": 32421
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6011072993278503,
+      "learning_rate": 0.00027210397040800313,
+      "loss": 3.0969,
+      "step": 32422
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5999905467033386,
+      "learning_rate": 0.0002720886409607187,
+      "loss": 3.0539,
+      "step": 32423
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6333760023117065,
+      "learning_rate": 0.0002720733115869476,
+      "loss": 3.1335,
+      "step": 32424
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6337881684303284,
+      "learning_rate": 0.0002720579822867299,
+      "loss": 2.828,
+      "step": 32425
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5925475358963013,
+      "learning_rate": 0.00027204265306010616,
+      "loss": 2.9094,
+      "step": 32426
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6454971432685852,
+      "learning_rate": 0.00027202732390711656,
+      "loss": 3.0657,
+      "step": 32427
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6801785826683044,
+      "learning_rate": 0.0002720119948278017,
+      "loss": 2.9696,
+      "step": 32428
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6122978329658508,
+      "learning_rate": 0.0002719966658222019,
+      "loss": 2.901,
+      "step": 32429
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6025977730751038,
+      "learning_rate": 0.0002719813368903573,
+      "loss": 3.0501,
+      "step": 32430
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5820662975311279,
+      "learning_rate": 0.0002719660080323087,
+      "loss": 3.1117,
+      "step": 32431
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6449814438819885,
+      "learning_rate": 0.00027195067924809616,
+      "loss": 3.0745,
+      "step": 32432
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6674616932868958,
+      "learning_rate": 0.00027193535053776006,
+      "loss": 2.9594,
+      "step": 32433
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6257567405700684,
+      "learning_rate": 0.00027192002190134097,
+      "loss": 3.1974,
+      "step": 32434
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6641944050788879,
+      "learning_rate": 0.000271904693338879,
+      "loss": 3.2397,
+      "step": 32435
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6264591813087463,
+      "learning_rate": 0.0002718893648504148,
+      "loss": 3.1784,
+      "step": 32436
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6466692686080933,
+      "learning_rate": 0.00027187403643598847,
+      "loss": 2.7668,
+      "step": 32437
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6122263073921204,
+      "learning_rate": 0.00027185870809564054,
+      "loss": 2.9252,
+      "step": 32438
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5788053274154663,
+      "learning_rate": 0.00027184337982941143,
+      "loss": 2.9313,
+      "step": 32439
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6403984427452087,
+      "learning_rate": 0.00027182805163734143,
+      "loss": 3.004,
+      "step": 32440
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6190902590751648,
+      "learning_rate": 0.00027181272351947097,
+      "loss": 3.2076,
+      "step": 32441
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6131787300109863,
+      "learning_rate": 0.0002717973954758402,
+      "loss": 3.0188,
+      "step": 32442
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.574184000492096,
+      "learning_rate": 0.00027178206750648975,
+      "loss": 3.0804,
+      "step": 32443
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6264970302581787,
+      "learning_rate": 0.00027176673961145997,
+      "loss": 2.8913,
+      "step": 32444
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.613540768623352,
+      "learning_rate": 0.000271751411790791,
+      "loss": 3.0355,
+      "step": 32445
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5924250483512878,
+      "learning_rate": 0.0002717360840445236,
+      "loss": 2.7862,
+      "step": 32446
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5855326652526855,
+      "learning_rate": 0.00027172075637269774,
+      "loss": 2.8646,
+      "step": 32447
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6014474034309387,
+      "learning_rate": 0.000271705428775354,
+      "loss": 3.0086,
+      "step": 32448
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6462697386741638,
+      "learning_rate": 0.0002716901012525328,
+      "loss": 3.0801,
+      "step": 32449
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.582455039024353,
+      "learning_rate": 0.00027167477380427436,
+      "loss": 3.0355,
+      "step": 32450
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5739898085594177,
+      "learning_rate": 0.00027165944643061914,
+      "loss": 3.1304,
+      "step": 32451
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6203562617301941,
+      "learning_rate": 0.0002716441191316074,
+      "loss": 3.2051,
+      "step": 32452
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6219992637634277,
+      "learning_rate": 0.0002716287919072796,
+      "loss": 2.9347,
+      "step": 32453
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6303226351737976,
+      "learning_rate": 0.0002716134647576763,
+      "loss": 3.0828,
+      "step": 32454
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6271916627883911,
+      "learning_rate": 0.0002715981376828374,
+      "loss": 3.0765,
+      "step": 32455
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6190890073776245,
+      "learning_rate": 0.00027158281068280374,
+      "loss": 2.999,
+      "step": 32456
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6210778951644897,
+      "learning_rate": 0.00027156748375761536,
+      "loss": 3.0662,
+      "step": 32457
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5740706324577332,
+      "learning_rate": 0.0002715521569073128,
+      "loss": 2.9903,
+      "step": 32458
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5983421802520752,
+      "learning_rate": 0.00027153683013193645,
+      "loss": 2.9476,
+      "step": 32459
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5819885730743408,
+      "learning_rate": 0.0002715215034315265,
+      "loss": 3.1873,
+      "step": 32460
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6093148589134216,
+      "learning_rate": 0.00027150617680612353,
+      "loss": 2.9995,
+      "step": 32461
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.607557475566864,
+      "learning_rate": 0.00027149085025576775,
+      "loss": 3.0563,
+      "step": 32462
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6053733825683594,
+      "learning_rate": 0.0002714755237804995,
+      "loss": 3.2042,
+      "step": 32463
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6200517416000366,
+      "learning_rate": 0.0002714601973803594,
+      "loss": 3.1293,
+      "step": 32464
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6223552823066711,
+      "learning_rate": 0.00027144487105538757,
+      "loss": 2.9668,
+      "step": 32465
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6040474772453308,
+      "learning_rate": 0.00027142954480562456,
+      "loss": 2.9209,
+      "step": 32466
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5464975237846375,
+      "learning_rate": 0.00027141421863111044,
+      "loss": 3.1782,
+      "step": 32467
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6203432679176331,
+      "learning_rate": 0.00027139889253188586,
+      "loss": 2.926,
+      "step": 32468
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5865224599838257,
+      "learning_rate": 0.0002713835665079911,
+      "loss": 3.2632,
+      "step": 32469
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.577903687953949,
+      "learning_rate": 0.00027136824055946647,
+      "loss": 3.0893,
+      "step": 32470
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6584243178367615,
+      "learning_rate": 0.00027135291468635247,
+      "loss": 2.8712,
+      "step": 32471
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.597705364227295,
+      "learning_rate": 0.00027133758888868935,
+      "loss": 3.0972,
+      "step": 32472
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5774487853050232,
+      "learning_rate": 0.0002713222631665174,
+      "loss": 2.9983,
+      "step": 32473
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6193967461585999,
+      "learning_rate": 0.00027130693751987724,
+      "loss": 3.0678,
+      "step": 32474
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.692612886428833,
+      "learning_rate": 0.0002712916119488089,
+      "loss": 3.0259,
+      "step": 32475
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.611100971698761,
+      "learning_rate": 0.0002712762864533531,
+      "loss": 2.943,
+      "step": 32476
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6383076310157776,
+      "learning_rate": 0.00027126096103355,
+      "loss": 3.1925,
+      "step": 32477
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5976969003677368,
+      "learning_rate": 0.0002712456356894399,
+      "loss": 2.983,
+      "step": 32478
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6111800670623779,
+      "learning_rate": 0.0002712303104210634,
+      "loss": 3.1147,
+      "step": 32479
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5945995450019836,
+      "learning_rate": 0.00027121498522846065,
+      "loss": 3.1164,
+      "step": 32480
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5979011654853821,
+      "learning_rate": 0.00027119966011167215,
+      "loss": 3.0233,
+      "step": 32481
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6658178567886353,
+      "learning_rate": 0.0002711843350707381,
+      "loss": 3.0965,
+      "step": 32482
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5945828557014465,
+      "learning_rate": 0.0002711690101056991,
+      "loss": 3.0912,
+      "step": 32483
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6201061606407166,
+      "learning_rate": 0.00027115368521659527,
+      "loss": 2.9176,
+      "step": 32484
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6284053325653076,
+      "learning_rate": 0.00027113836040346703,
+      "loss": 2.9548,
+      "step": 32485
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5955159068107605,
+      "learning_rate": 0.000271123035666355,
+      "loss": 3.2301,
+      "step": 32486
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6049878597259521,
+      "learning_rate": 0.00027110771100529916,
+      "loss": 2.9669,
+      "step": 32487
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5916286706924438,
+      "learning_rate": 0.0002710923864203402,
+      "loss": 3.0258,
+      "step": 32488
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5768219232559204,
+      "learning_rate": 0.0002710770619115181,
+      "loss": 3.0384,
+      "step": 32489
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6287925839424133,
+      "learning_rate": 0.0002710617374788736,
+      "loss": 3.0563,
+      "step": 32490
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.659156858921051,
+      "learning_rate": 0.00027104641312244694,
+      "loss": 3.0919,
+      "step": 32491
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5746557712554932,
+      "learning_rate": 0.0002710310888422784,
+      "loss": 3.063,
+      "step": 32492
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.613263726234436,
+      "learning_rate": 0.00027101576463840846,
+      "loss": 2.9219,
+      "step": 32493
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.610640287399292,
+      "learning_rate": 0.00027100044051087737,
+      "loss": 3.182,
+      "step": 32494
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6279901266098022,
+      "learning_rate": 0.00027098511645972544,
+      "loss": 2.7512,
+      "step": 32495
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6687607169151306,
+      "learning_rate": 0.0002709697924849933,
+      "loss": 3.1327,
+      "step": 32496
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6255040764808655,
+      "learning_rate": 0.000270954468586721,
+      "loss": 2.8758,
+      "step": 32497
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5970877408981323,
+      "learning_rate": 0.00027093914476494915,
+      "loss": 3.0503,
+      "step": 32498
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5644693970680237,
+      "learning_rate": 0.00027092382101971793,
+      "loss": 2.8474,
+      "step": 32499
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.591464638710022,
+      "learning_rate": 0.00027090849735106776,
+      "loss": 2.9457,
+      "step": 32500
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6773257255554199,
+      "learning_rate": 0.0002708931737590391,
+      "loss": 3.0178,
+      "step": 32501
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.59250807762146,
+      "learning_rate": 0.00027087785024367217,
+      "loss": 3.1202,
+      "step": 32502
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6139160990715027,
+      "learning_rate": 0.0002708625268050074,
+      "loss": 2.8928,
+      "step": 32503
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6274259686470032,
+      "learning_rate": 0.00027084720344308503,
+      "loss": 3.168,
+      "step": 32504
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5896390080451965,
+      "learning_rate": 0.0002708318801579456,
+      "loss": 2.9981,
+      "step": 32505
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6155339479446411,
+      "learning_rate": 0.00027081655694962943,
+      "loss": 2.9269,
+      "step": 32506
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5823910236358643,
+      "learning_rate": 0.00027080123381817666,
+      "loss": 2.9131,
+      "step": 32507
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6796182990074158,
+      "learning_rate": 0.00027078591076362804,
+      "loss": 3.0235,
+      "step": 32508
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5854375958442688,
+      "learning_rate": 0.00027077058778602356,
+      "loss": 2.883,
+      "step": 32509
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5948371291160583,
+      "learning_rate": 0.0002707552648854037,
+      "loss": 2.9561,
+      "step": 32510
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6129918694496155,
+      "learning_rate": 0.00027073994206180897,
+      "loss": 3.2477,
+      "step": 32511
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6146805882453918,
+      "learning_rate": 0.0002707246193152796,
+      "loss": 3.0606,
+      "step": 32512
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5845103859901428,
+      "learning_rate": 0.00027070929664585594,
+      "loss": 3.1949,
+      "step": 32513
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.606810450553894,
+      "learning_rate": 0.0002706939740535782,
+      "loss": 2.9231,
+      "step": 32514
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5757008194923401,
+      "learning_rate": 0.0002706786515384869,
+      "loss": 2.9504,
+      "step": 32515
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.620858371257782,
+      "learning_rate": 0.0002706633291006226,
+      "loss": 2.9856,
+      "step": 32516
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6243740320205688,
+      "learning_rate": 0.00027064800674002533,
+      "loss": 3.1278,
+      "step": 32517
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5853642225265503,
+      "learning_rate": 0.00027063268445673557,
+      "loss": 3.1047,
+      "step": 32518
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6242586374282837,
+      "learning_rate": 0.0002706173622507936,
+      "loss": 3.009,
+      "step": 32519
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.636432409286499,
+      "learning_rate": 0.0002706020401222399,
+      "loss": 2.9736,
+      "step": 32520
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6194183826446533,
+      "learning_rate": 0.00027058671807111484,
+      "loss": 3.1121,
+      "step": 32521
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6781464219093323,
+      "learning_rate": 0.00027057139609745854,
+      "loss": 2.7746,
+      "step": 32522
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5913337469100952,
+      "learning_rate": 0.0002705560742013117,
+      "loss": 3.0744,
+      "step": 32523
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5661170482635498,
+      "learning_rate": 0.0002705407523827144,
+      "loss": 3.1307,
+      "step": 32524
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6512724161148071,
+      "learning_rate": 0.000270525430641707,
+      "loss": 3.1466,
+      "step": 32525
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.631696879863739,
+      "learning_rate": 0.0002705101089783301,
+      "loss": 3.0165,
+      "step": 32526
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6301171183586121,
+      "learning_rate": 0.00027049478739262384,
+      "loss": 2.9184,
+      "step": 32527
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6100221276283264,
+      "learning_rate": 0.00027047946588462874,
+      "loss": 3.1662,
+      "step": 32528
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6037270426750183,
+      "learning_rate": 0.00027046414445438486,
+      "loss": 2.9379,
+      "step": 32529
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6437107920646667,
+      "learning_rate": 0.0002704488231019328,
+      "loss": 3.0219,
+      "step": 32530
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5775290131568909,
+      "learning_rate": 0.00027043350182731293,
+      "loss": 3.1123,
+      "step": 32531
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5918175578117371,
+      "learning_rate": 0.0002704181806305654,
+      "loss": 3.0319,
+      "step": 32532
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.567966103553772,
+      "learning_rate": 0.0002704028595117308,
+      "loss": 3.1738,
+      "step": 32533
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6273672580718994,
+      "learning_rate": 0.0002703875384708493,
+      "loss": 3.2175,
+      "step": 32534
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6246945858001709,
+      "learning_rate": 0.0002703722175079613,
+      "loss": 2.9896,
+      "step": 32535
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5851449370384216,
+      "learning_rate": 0.0002703568966231073,
+      "loss": 3.0777,
+      "step": 32536
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6178532838821411,
+      "learning_rate": 0.00027034157581632735,
+      "loss": 3.3369,
+      "step": 32537
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.626719057559967,
+      "learning_rate": 0.0002703262550876622,
+      "loss": 3.2271,
+      "step": 32538
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5996720194816589,
+      "learning_rate": 0.00027031093443715185,
+      "loss": 3.2784,
+      "step": 32539
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5899909138679504,
+      "learning_rate": 0.0002702956138648367,
+      "loss": 3.2182,
+      "step": 32540
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5896705985069275,
+      "learning_rate": 0.0002702802933707574,
+      "loss": 2.862,
+      "step": 32541
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6266672015190125,
+      "learning_rate": 0.00027026497295495395,
+      "loss": 3.039,
+      "step": 32542
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.615408718585968,
+      "learning_rate": 0.0002702496526174669,
+      "loss": 3.0288,
+      "step": 32543
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6363191604614258,
+      "learning_rate": 0.0002702343323583365,
+      "loss": 3.1345,
+      "step": 32544
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6553589701652527,
+      "learning_rate": 0.0002702190121776031,
+      "loss": 3.1664,
+      "step": 32545
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6160081624984741,
+      "learning_rate": 0.0002702036920753072,
+      "loss": 2.9866,
+      "step": 32546
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6299121975898743,
+      "learning_rate": 0.0002701883720514889,
+      "loss": 3.0313,
+      "step": 32547
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5947949290275574,
+      "learning_rate": 0.00027017305210618885,
+      "loss": 3.2526,
+      "step": 32548
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5934600830078125,
+      "learning_rate": 0.00027015773223944717,
+      "loss": 2.9821,
+      "step": 32549
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6083028316497803,
+      "learning_rate": 0.00027014241245130414,
+      "loss": 2.9853,
+      "step": 32550
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5787101984024048,
+      "learning_rate": 0.0002701270927418005,
+      "loss": 2.9115,
+      "step": 32551
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.580090343952179,
+      "learning_rate": 0.0002701117731109762,
+      "loss": 3.02,
+      "step": 32552
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6392739415168762,
+      "learning_rate": 0.0002700964535588718,
+      "loss": 3.2771,
+      "step": 32553
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.607865035533905,
+      "learning_rate": 0.0002700811340855275,
+      "loss": 2.8881,
+      "step": 32554
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6404851078987122,
+      "learning_rate": 0.0002700658146909837,
+      "loss": 3.1971,
+      "step": 32555
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6085619330406189,
+      "learning_rate": 0.00027005049537528103,
+      "loss": 3.0046,
+      "step": 32556
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5817078351974487,
+      "learning_rate": 0.00027003517613845935,
+      "loss": 2.956,
+      "step": 32557
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6575085520744324,
+      "learning_rate": 0.0002700198569805594,
+      "loss": 3.2202,
+      "step": 32558
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6530008912086487,
+      "learning_rate": 0.0002700045379016212,
+      "loss": 2.9829,
+      "step": 32559
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5922717452049255,
+      "learning_rate": 0.00026998921890168545,
+      "loss": 3.005,
+      "step": 32560
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5945848822593689,
+      "learning_rate": 0.00026997389998079227,
+      "loss": 2.9596,
+      "step": 32561
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5837536454200745,
+      "learning_rate": 0.0002699585811389819,
+      "loss": 2.9112,
+      "step": 32562
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6117374897003174,
+      "learning_rate": 0.0002699432623762951,
+      "loss": 2.9097,
+      "step": 32563
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5952165126800537,
+      "learning_rate": 0.00026992794369277187,
+      "loss": 3.0588,
+      "step": 32564
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6353278756141663,
+      "learning_rate": 0.00026991262508845265,
+      "loss": 3.1253,
+      "step": 32565
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6072468161582947,
+      "learning_rate": 0.0002698973065633777,
+      "loss": 2.8889,
+      "step": 32566
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6333851218223572,
+      "learning_rate": 0.0002698819881175875,
+      "loss": 3.0873,
+      "step": 32567
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5800299644470215,
+      "learning_rate": 0.0002698666697511224,
+      "loss": 3.1187,
+      "step": 32568
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6245841979980469,
+      "learning_rate": 0.00026985135146402264,
+      "loss": 3.1307,
+      "step": 32569
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6200968623161316,
+      "learning_rate": 0.0002698360332563287,
+      "loss": 3.0176,
+      "step": 32570
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5942978858947754,
+      "learning_rate": 0.00026982071512808076,
+      "loss": 3.2657,
+      "step": 32571
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6148001551628113,
+      "learning_rate": 0.0002698053970793192,
+      "loss": 2.9059,
+      "step": 32572
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6103355288505554,
+      "learning_rate": 0.00026979007911008454,
+      "loss": 3.1997,
+      "step": 32573
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5524942278862,
+      "learning_rate": 0.0002697747612204169,
+      "loss": 3.0229,
+      "step": 32574
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.590096652507782,
+      "learning_rate": 0.00026975944341035683,
+      "loss": 3.2664,
+      "step": 32575
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6268036961555481,
+      "learning_rate": 0.0002697441256799444,
+      "loss": 2.976,
+      "step": 32576
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6254855394363403,
+      "learning_rate": 0.00026972880802922014,
+      "loss": 3.2555,
+      "step": 32577
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6256861090660095,
+      "learning_rate": 0.00026971349045822453,
+      "loss": 3.0236,
+      "step": 32578
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6017988324165344,
+      "learning_rate": 0.0002696981729669977,
+      "loss": 3.025,
+      "step": 32579
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6067514419555664,
+      "learning_rate": 0.00026968285555558003,
+      "loss": 3.0145,
+      "step": 32580
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6188127398490906,
+      "learning_rate": 0.0002696675382240118,
+      "loss": 3.2419,
+      "step": 32581
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7204647064208984,
+      "learning_rate": 0.00026965222097233353,
+      "loss": 3.1838,
+      "step": 32582
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6136758923530579,
+      "learning_rate": 0.00026963690380058555,
+      "loss": 2.9887,
+      "step": 32583
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6154049634933472,
+      "learning_rate": 0.00026962158670880795,
+      "loss": 3.1223,
+      "step": 32584
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6027714014053345,
+      "learning_rate": 0.00026960626969704137,
+      "loss": 3.0375,
+      "step": 32585
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5715473890304565,
+      "learning_rate": 0.00026959095276532594,
+      "loss": 2.9966,
+      "step": 32586
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6251134872436523,
+      "learning_rate": 0.0002695756359137021,
+      "loss": 2.9965,
+      "step": 32587
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6239544153213501,
+      "learning_rate": 0.00026956031914221023,
+      "loss": 3.2285,
+      "step": 32588
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6197227239608765,
+      "learning_rate": 0.0002695450024508906,
+      "loss": 2.9335,
+      "step": 32589
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6238695979118347,
+      "learning_rate": 0.00026952968583978363,
+      "loss": 3.1362,
+      "step": 32590
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6636531949043274,
+      "learning_rate": 0.00026951436930892946,
+      "loss": 2.9115,
+      "step": 32591
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5675901770591736,
+      "learning_rate": 0.0002694990528583687,
+      "loss": 2.8736,
+      "step": 32592
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6764063239097595,
+      "learning_rate": 0.00026948373648814157,
+      "loss": 3.0807,
+      "step": 32593
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6123232841491699,
+      "learning_rate": 0.00026946842019828825,
+      "loss": 2.9812,
+      "step": 32594
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7344564199447632,
+      "learning_rate": 0.00026945310398884944,
+      "loss": 2.9918,
+      "step": 32595
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6294760704040527,
+      "learning_rate": 0.00026943778785986516,
+      "loss": 3.0817,
+      "step": 32596
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6110149621963501,
+      "learning_rate": 0.0002694224718113758,
+      "loss": 3.1679,
+      "step": 32597
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6438841819763184,
+      "learning_rate": 0.00026940715584342187,
+      "loss": 3.1391,
+      "step": 32598
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6176292896270752,
+      "learning_rate": 0.00026939183995604353,
+      "loss": 2.913,
+      "step": 32599
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5883298516273499,
+      "learning_rate": 0.0002693765241492813,
+      "loss": 3.054,
+      "step": 32600
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6791853308677673,
+      "learning_rate": 0.0002693612084231754,
+      "loss": 3.0085,
+      "step": 32601
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6615046262741089,
+      "learning_rate": 0.00026934589277776604,
+      "loss": 3.0181,
+      "step": 32602
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6171573996543884,
+      "learning_rate": 0.00026933057721309386,
+      "loss": 2.9615,
+      "step": 32603
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6239962577819824,
+      "learning_rate": 0.0002693152617291989,
+      "loss": 2.9545,
+      "step": 32604
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.61008620262146,
+      "learning_rate": 0.0002692999463261218,
+      "loss": 3.1183,
+      "step": 32605
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6085340976715088,
+      "learning_rate": 0.00026928463100390256,
+      "loss": 3.1372,
+      "step": 32606
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6348270177841187,
+      "learning_rate": 0.00026926931576258176,
+      "loss": 2.9547,
+      "step": 32607
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5983766913414001,
+      "learning_rate": 0.00026925400060219973,
+      "loss": 3.3116,
+      "step": 32608
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6149789690971375,
+      "learning_rate": 0.00026923868552279666,
+      "loss": 3.0122,
+      "step": 32609
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5611794590950012,
+      "learning_rate": 0.0002692233705244131,
+      "loss": 3.1514,
+      "step": 32610
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6104076504707336,
+      "learning_rate": 0.00026920805560708914,
+      "loss": 3.0857,
+      "step": 32611
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6021207571029663,
+      "learning_rate": 0.0002691927407708652,
+      "loss": 2.9064,
+      "step": 32612
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.58101487159729,
+      "learning_rate": 0.00026917742601578173,
+      "loss": 3.0131,
+      "step": 32613
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5953050851821899,
+      "learning_rate": 0.00026916211134187895,
+      "loss": 2.9169,
+      "step": 32614
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6245440244674683,
+      "learning_rate": 0.0002691467967491973,
+      "loss": 3.0768,
+      "step": 32615
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5973414182662964,
+      "learning_rate": 0.000269131482237777,
+      "loss": 2.9519,
+      "step": 32616
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.583087146282196,
+      "learning_rate": 0.0002691161678076583,
+      "loss": 3.0333,
+      "step": 32617
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6332690715789795,
+      "learning_rate": 0.000269100853458882,
+      "loss": 3.0362,
+      "step": 32618
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6058811545372009,
+      "learning_rate": 0.0002690855391914878,
+      "loss": 2.9249,
+      "step": 32619
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6081041693687439,
+      "learning_rate": 0.00026907022500551655,
+      "loss": 3.0316,
+      "step": 32620
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6076980829238892,
+      "learning_rate": 0.0002690549109010082,
+      "loss": 2.9194,
+      "step": 32621
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6190997362136841,
+      "learning_rate": 0.0002690395968780033,
+      "loss": 3.1308,
+      "step": 32622
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6132203936576843,
+      "learning_rate": 0.0002690242829365423,
+      "loss": 3.0724,
+      "step": 32623
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5711997151374817,
+      "learning_rate": 0.00026900896907666516,
+      "loss": 2.984,
+      "step": 32624
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6215646862983704,
+      "learning_rate": 0.00026899365529841257,
+      "loss": 3.1405,
+      "step": 32625
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5729935765266418,
+      "learning_rate": 0.0002689783416018247,
+      "loss": 3.1276,
+      "step": 32626
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5910037755966187,
+      "learning_rate": 0.0002689630279869418,
+      "loss": 2.9777,
+      "step": 32627
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6306336522102356,
+      "learning_rate": 0.00026894771445380445,
+      "loss": 2.9572,
+      "step": 32628
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6902565360069275,
+      "learning_rate": 0.0002689324010024528,
+      "loss": 2.8442,
+      "step": 32629
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5936403870582581,
+      "learning_rate": 0.0002689170876329273,
+      "loss": 3.0951,
+      "step": 32630
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6323724389076233,
+      "learning_rate": 0.0002689017743452681,
+      "loss": 2.9881,
+      "step": 32631
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5704877972602844,
+      "learning_rate": 0.0002688864611395156,
+      "loss": 2.9077,
+      "step": 32632
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.643515944480896,
+      "learning_rate": 0.0002688711480157103,
+      "loss": 3.0715,
+      "step": 32633
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6297488212585449,
+      "learning_rate": 0.00026885583497389235,
+      "loss": 3.0259,
+      "step": 32634
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5987085103988647,
+      "learning_rate": 0.00026884052201410215,
+      "loss": 3.0065,
+      "step": 32635
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5853610634803772,
+      "learning_rate": 0.00026882520913638,
+      "loss": 2.9778,
+      "step": 32636
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5946998000144958,
+      "learning_rate": 0.0002688098963407664,
+      "loss": 2.9844,
+      "step": 32637
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6260942220687866,
+      "learning_rate": 0.00026879458362730126,
+      "loss": 2.9159,
+      "step": 32638
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7542493343353271,
+      "learning_rate": 0.00026877927099602523,
+      "loss": 3.1824,
+      "step": 32639
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5645418763160706,
+      "learning_rate": 0.0002687639584469787,
+      "loss": 2.9704,
+      "step": 32640
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6517764329910278,
+      "learning_rate": 0.0002687486459802019,
+      "loss": 3.099,
+      "step": 32641
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6776633262634277,
+      "learning_rate": 0.0002687333335957352,
+      "loss": 3.0951,
+      "step": 32642
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6206831336021423,
+      "learning_rate": 0.00026871802129361865,
+      "loss": 3.1086,
+      "step": 32643
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.633886992931366,
+      "learning_rate": 0.00026870270907389303,
+      "loss": 3.0899,
+      "step": 32644
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6100537776947021,
+      "learning_rate": 0.0002686873969365984,
+      "loss": 2.9998,
+      "step": 32645
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.588862419128418,
+      "learning_rate": 0.0002686720848817751,
+      "loss": 3.1437,
+      "step": 32646
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6336972713470459,
+      "learning_rate": 0.0002686567729094635,
+      "loss": 3.0714,
+      "step": 32647
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6299551129341125,
+      "learning_rate": 0.000268641461019704,
+      "loss": 3.0617,
+      "step": 32648
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5796019434928894,
+      "learning_rate": 0.0002686261492125367,
+      "loss": 2.8452,
+      "step": 32649
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6125264763832092,
+      "learning_rate": 0.00026861083748800225,
+      "loss": 3.037,
+      "step": 32650
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6127404570579529,
+      "learning_rate": 0.00026859552584614077,
+      "loss": 3.182,
+      "step": 32651
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6474936604499817,
+      "learning_rate": 0.00026858021428699264,
+      "loss": 3.2452,
+      "step": 32652
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5996273159980774,
+      "learning_rate": 0.0002685649028105981,
+      "loss": 3.1117,
+      "step": 32653
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6106273531913757,
+      "learning_rate": 0.0002685495914169976,
+      "loss": 3.2262,
+      "step": 32654
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6102750897407532,
+      "learning_rate": 0.00026853428010623146,
+      "loss": 3.1025,
+      "step": 32655
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6115225553512573,
+      "learning_rate": 0.0002685189688783399,
+      "loss": 3.0833,
+      "step": 32656
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6346925497055054,
+      "learning_rate": 0.0002685036577333634,
+      "loss": 3.0,
+      "step": 32657
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6187523603439331,
+      "learning_rate": 0.0002684883466713422,
+      "loss": 3.0215,
+      "step": 32658
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7155755758285522,
+      "learning_rate": 0.0002684730356923165,
+      "loss": 2.8853,
+      "step": 32659
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6200131177902222,
+      "learning_rate": 0.0002684577247963269,
+      "loss": 3.0541,
+      "step": 32660
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6308820843696594,
+      "learning_rate": 0.0002684424139834135,
+      "loss": 3.0811,
+      "step": 32661
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5879965424537659,
+      "learning_rate": 0.00026842710325361677,
+      "loss": 2.9076,
+      "step": 32662
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.629178524017334,
+      "learning_rate": 0.00026841179260697693,
+      "loss": 2.9654,
+      "step": 32663
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.621136486530304,
+      "learning_rate": 0.0002683964820435343,
+      "loss": 3.0303,
+      "step": 32664
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6378344297409058,
+      "learning_rate": 0.0002683811715633294,
+      "loss": 2.8025,
+      "step": 32665
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.642932116985321,
+      "learning_rate": 0.0002683658611664023,
+      "loss": 2.7823,
+      "step": 32666
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6393907070159912,
+      "learning_rate": 0.0002683505508527935,
+      "loss": 3.1127,
+      "step": 32667
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6391983032226562,
+      "learning_rate": 0.0002683352406225432,
+      "loss": 2.9901,
+      "step": 32668
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6028316020965576,
+      "learning_rate": 0.00026831993047569177,
+      "loss": 3.0726,
+      "step": 32669
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.655157208442688,
+      "learning_rate": 0.0002683046204122796,
+      "loss": 2.786,
+      "step": 32670
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.61063551902771,
+      "learning_rate": 0.0002682893104323469,
+      "loss": 3.0121,
+      "step": 32671
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6077100038528442,
+      "learning_rate": 0.00026827400053593413,
+      "loss": 3.0462,
+      "step": 32672
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6022846698760986,
+      "learning_rate": 0.0002682586907230815,
+      "loss": 3.0828,
+      "step": 32673
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6264395117759705,
+      "learning_rate": 0.0002682433809938293,
+      "loss": 3.1968,
+      "step": 32674
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6148096323013306,
+      "learning_rate": 0.00026822807134821804,
+      "loss": 3.2081,
+      "step": 32675
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6186606287956238,
+      "learning_rate": 0.0002682127617862879,
+      "loss": 2.9582,
+      "step": 32676
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6134831309318542,
+      "learning_rate": 0.00026819745230807925,
+      "loss": 3.0577,
+      "step": 32677
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5858891010284424,
+      "learning_rate": 0.00026818214291363226,
+      "loss": 2.9476,
+      "step": 32678
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6068850755691528,
+      "learning_rate": 0.0002681668336029874,
+      "loss": 3.0758,
+      "step": 32679
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5995755195617676,
+      "learning_rate": 0.0002681515243761851,
+      "loss": 2.9556,
+      "step": 32680
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5753273367881775,
+      "learning_rate": 0.0002681362152332655,
+      "loss": 2.7898,
+      "step": 32681
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6415500044822693,
+      "learning_rate": 0.00026812090617426903,
+      "loss": 3.2757,
+      "step": 32682
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5945836305618286,
+      "learning_rate": 0.0002681055971992358,
+      "loss": 3.1363,
+      "step": 32683
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6280573010444641,
+      "learning_rate": 0.0002680902883082064,
+      "loss": 2.8068,
+      "step": 32684
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6142799258232117,
+      "learning_rate": 0.00026807497950122105,
+      "loss": 3.0977,
+      "step": 32685
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6052511930465698,
+      "learning_rate": 0.00026805967077832,
+      "loss": 2.9415,
+      "step": 32686
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5932520031929016,
+      "learning_rate": 0.0002680443621395437,
+      "loss": 2.9637,
+      "step": 32687
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5896964073181152,
+      "learning_rate": 0.00026802905358493237,
+      "loss": 2.875,
+      "step": 32688
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6037048697471619,
+      "learning_rate": 0.0002680137451145263,
+      "loss": 3.1561,
+      "step": 32689
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6128067374229431,
+      "learning_rate": 0.00026799843672836596,
+      "loss": 2.8895,
+      "step": 32690
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5946832299232483,
+      "learning_rate": 0.00026798312842649157,
+      "loss": 2.9309,
+      "step": 32691
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6187651753425598,
+      "learning_rate": 0.0002679678202089435,
+      "loss": 2.8649,
+      "step": 32692
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5798802971839905,
+      "learning_rate": 0.00026795251207576185,
+      "loss": 3.0354,
+      "step": 32693
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5939930081367493,
+      "learning_rate": 0.00026793720402698725,
+      "loss": 3.1686,
+      "step": 32694
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5916917324066162,
+      "learning_rate": 0.0002679218960626599,
+      "loss": 3.1373,
+      "step": 32695
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6304068565368652,
+      "learning_rate": 0.00026790658818282,
+      "loss": 3.0483,
+      "step": 32696
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5936771035194397,
+      "learning_rate": 0.0002678912803875081,
+      "loss": 2.9578,
+      "step": 32697
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5950177907943726,
+      "learning_rate": 0.0002678759726767643,
+      "loss": 3.0496,
+      "step": 32698
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6186109781265259,
+      "learning_rate": 0.00026786066505062896,
+      "loss": 3.0393,
+      "step": 32699
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6025801301002502,
+      "learning_rate": 0.0002678453575091426,
+      "loss": 3.2257,
+      "step": 32700
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5880355834960938,
+      "learning_rate": 0.0002678300500523452,
+      "loss": 2.7253,
+      "step": 32701
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6294435858726501,
+      "learning_rate": 0.0002678147426802774,
+      "loss": 3.0772,
+      "step": 32702
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6276559233665466,
+      "learning_rate": 0.0002677994353929793,
+      "loss": 2.8216,
+      "step": 32703
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6946702003479004,
+      "learning_rate": 0.00026778412819049123,
+      "loss": 3.0334,
+      "step": 32704
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5705452561378479,
+      "learning_rate": 0.00026776882107285374,
+      "loss": 2.9975,
+      "step": 32705
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5945636034011841,
+      "learning_rate": 0.00026775351404010684,
+      "loss": 2.9102,
+      "step": 32706
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5809672474861145,
+      "learning_rate": 0.0002677382070922911,
+      "loss": 2.9264,
+      "step": 32707
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5987460613250732,
+      "learning_rate": 0.0002677229002294465,
+      "loss": 3.0214,
+      "step": 32708
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6070874929428101,
+      "learning_rate": 0.00026770759345161375,
+      "loss": 3.1131,
+      "step": 32709
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.632290780544281,
+      "learning_rate": 0.00026769228675883297,
+      "loss": 2.8364,
+      "step": 32710
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6187470555305481,
+      "learning_rate": 0.0002676769801511444,
+      "loss": 2.9534,
+      "step": 32711
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5995469689369202,
+      "learning_rate": 0.00026766167362858855,
+      "loss": 2.9997,
+      "step": 32712
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6334401965141296,
+      "learning_rate": 0.0002676463671912056,
+      "loss": 3.0416,
+      "step": 32713
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6241161227226257,
+      "learning_rate": 0.0002676310608390359,
+      "loss": 2.9658,
+      "step": 32714
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6149295568466187,
+      "learning_rate": 0.00026761575457211965,
+      "loss": 3.071,
+      "step": 32715
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6074439287185669,
+      "learning_rate": 0.0002676004483904973,
+      "loss": 3.0599,
+      "step": 32716
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5981096625328064,
+      "learning_rate": 0.0002675851422942093,
+      "loss": 3.1352,
+      "step": 32717
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5963551998138428,
+      "learning_rate": 0.00026756983628329554,
+      "loss": 2.8888,
+      "step": 32718
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6065682768821716,
+      "learning_rate": 0.0002675545303577968,
+      "loss": 3.0682,
+      "step": 32719
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6360631585121155,
+      "learning_rate": 0.0002675392245177531,
+      "loss": 3.0089,
+      "step": 32720
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5898069143295288,
+      "learning_rate": 0.00026752391876320476,
+      "loss": 2.876,
+      "step": 32721
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6404922604560852,
+      "learning_rate": 0.0002675086130941923,
+      "loss": 3.1414,
+      "step": 32722
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6144474744796753,
+      "learning_rate": 0.0002674933075107558,
+      "loss": 2.9401,
+      "step": 32723
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6159098744392395,
+      "learning_rate": 0.00026747800201293575,
+      "loss": 3.0381,
+      "step": 32724
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6114851236343384,
+      "learning_rate": 0.0002674626966007724,
+      "loss": 3.0802,
+      "step": 32725
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6240290999412537,
+      "learning_rate": 0.0002674473912743059,
+      "loss": 3.0702,
+      "step": 32726
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.598355770111084,
+      "learning_rate": 0.00026743208603357687,
+      "loss": 2.8799,
+      "step": 32727
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6011687517166138,
+      "learning_rate": 0.00026741678087862535,
+      "loss": 3.171,
+      "step": 32728
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6154776811599731,
+      "learning_rate": 0.00026740147580949186,
+      "loss": 3.1596,
+      "step": 32729
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5840276479721069,
+      "learning_rate": 0.0002673861708262165,
+      "loss": 3.1403,
+      "step": 32730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6225008368492126,
+      "learning_rate": 0.00026737086592883976,
+      "loss": 2.9208,
+      "step": 32731
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5753346681594849,
+      "learning_rate": 0.0002673555611174019,
+      "loss": 3.0474,
+      "step": 32732
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6378356218338013,
+      "learning_rate": 0.0002673402563919431,
+      "loss": 3.0054,
+      "step": 32733
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5706433653831482,
+      "learning_rate": 0.0002673249517525039,
+      "loss": 3.058,
+      "step": 32734
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6072514057159424,
+      "learning_rate": 0.00026730964719912445,
+      "loss": 2.9545,
+      "step": 32735
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6374166011810303,
+      "learning_rate": 0.00026729434273184497,
+      "loss": 3.1019,
+      "step": 32736
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6123475432395935,
+      "learning_rate": 0.00026727903835070606,
+      "loss": 3.0819,
+      "step": 32737
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6655570268630981,
+      "learning_rate": 0.00026726373405574783,
+      "loss": 3.1967,
+      "step": 32738
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5814079642295837,
+      "learning_rate": 0.0002672484298470107,
+      "loss": 3.0927,
+      "step": 32739
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7113948464393616,
+      "learning_rate": 0.0002672331257245347,
+      "loss": 3.0828,
+      "step": 32740
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6144413352012634,
+      "learning_rate": 0.0002672178216883604,
+      "loss": 2.8819,
+      "step": 32741
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5857863426208496,
+      "learning_rate": 0.0002672025177385282,
+      "loss": 3.2243,
+      "step": 32742
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5955747365951538,
+      "learning_rate": 0.00026718721387507813,
+      "loss": 3.1915,
+      "step": 32743
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6116932034492493,
+      "learning_rate": 0.0002671719100980507,
+      "loss": 2.8832,
+      "step": 32744
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5967369079589844,
+      "learning_rate": 0.000267156606407486,
+      "loss": 2.9476,
+      "step": 32745
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6008356213569641,
+      "learning_rate": 0.0002671413028034246,
+      "loss": 3.111,
+      "step": 32746
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5891445279121399,
+      "learning_rate": 0.00026712599928590674,
+      "loss": 3.1365,
+      "step": 32747
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5794476270675659,
+      "learning_rate": 0.0002671106958549725,
+      "loss": 3.1771,
+      "step": 32748
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6033952236175537,
+      "learning_rate": 0.0002670953925106625,
+      "loss": 3.1056,
+      "step": 32749
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6501690745353699,
+      "learning_rate": 0.00026708008925301687,
+      "loss": 2.7864,
+      "step": 32750
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6086935997009277,
+      "learning_rate": 0.00026706478608207586,
+      "loss": 3.0754,
+      "step": 32751
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6224468350410461,
+      "learning_rate": 0.00026704948299787997,
+      "loss": 3.0004,
+      "step": 32752
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5442426800727844,
+      "learning_rate": 0.0002670341800004693,
+      "loss": 3.0873,
+      "step": 32753
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5749726891517639,
+      "learning_rate": 0.00026701887708988444,
+      "loss": 2.8662,
+      "step": 32754
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6050938367843628,
+      "learning_rate": 0.0002670035742661653,
+      "loss": 2.9914,
+      "step": 32755
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5932197570800781,
+      "learning_rate": 0.00026698827152935245,
+      "loss": 3.0756,
+      "step": 32756
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.592960000038147,
+      "learning_rate": 0.0002669729688794863,
+      "loss": 3.055,
+      "step": 32757
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5769248008728027,
+      "learning_rate": 0.0002669576663166068,
+      "loss": 3.3109,
+      "step": 32758
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6258298754692078,
+      "learning_rate": 0.0002669423638407545,
+      "loss": 3.139,
+      "step": 32759
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5944938659667969,
+      "learning_rate": 0.0002669270614519697,
+      "loss": 3.1027,
+      "step": 32760
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6634683012962341,
+      "learning_rate": 0.00026691175915029253,
+      "loss": 3.0298,
+      "step": 32761
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6014466881752014,
+      "learning_rate": 0.0002668964569357636,
+      "loss": 2.9555,
+      "step": 32762
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.620914876461029,
+      "learning_rate": 0.00026688115480842287,
+      "loss": 3.0418,
+      "step": 32763
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5882852673530579,
+      "learning_rate": 0.0002668658527683109,
+      "loss": 2.9286,
+      "step": 32764
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.610764741897583,
+      "learning_rate": 0.0002668505508154679,
+      "loss": 3.0676,
+      "step": 32765
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6545993685722351,
+      "learning_rate": 0.00026683524894993406,
+      "loss": 2.8785,
+      "step": 32766
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6135849356651306,
+      "learning_rate": 0.00026681994717174995,
+      "loss": 3.0813,
+      "step": 32767
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.596288800239563,
+      "learning_rate": 0.0002668046454809556,
+      "loss": 3.1225,
+      "step": 32768
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6006448864936829,
+      "learning_rate": 0.0002667893438775915,
+      "loss": 2.8525,
+      "step": 32769
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5774670243263245,
+      "learning_rate": 0.0002667740423616978,
+      "loss": 3.2201,
+      "step": 32770
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6070205569267273,
+      "learning_rate": 0.00026675874093331494,
+      "loss": 2.946,
+      "step": 32771
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5863068103790283,
+      "learning_rate": 0.0002667434395924832,
+      "loss": 3.2808,
+      "step": 32772
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6191710233688354,
+      "learning_rate": 0.0002667281383392427,
+      "loss": 2.894,
+      "step": 32773
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.621767520904541,
+      "learning_rate": 0.000266712837173634,
+      "loss": 3.1097,
+      "step": 32774
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5857933163642883,
+      "learning_rate": 0.0002666975360956972,
+      "loss": 2.9612,
+      "step": 32775
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6357319355010986,
+      "learning_rate": 0.0002666822351054727,
+      "loss": 3.0118,
+      "step": 32776
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5903388857841492,
+      "learning_rate": 0.00026666693420300086,
+      "loss": 2.7496,
+      "step": 32777
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6077666282653809,
+      "learning_rate": 0.0002666516333883218,
+      "loss": 3.0988,
+      "step": 32778
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6337267756462097,
+      "learning_rate": 0.000266636332661476,
+      "loss": 3.1458,
+      "step": 32779
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.567110538482666,
+      "learning_rate": 0.00026662103202250355,
+      "loss": 2.8474,
+      "step": 32780
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6637780666351318,
+      "learning_rate": 0.00026660573147144494,
+      "loss": 3.2734,
+      "step": 32781
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6085167527198792,
+      "learning_rate": 0.00026659043100834054,
+      "loss": 3.1563,
+      "step": 32782
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6335734128952026,
+      "learning_rate": 0.0002665751306332304,
+      "loss": 3.1045,
+      "step": 32783
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6378978490829468,
+      "learning_rate": 0.00026655983034615493,
+      "loss": 3.1083,
+      "step": 32784
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5712326765060425,
+      "learning_rate": 0.0002665445301471544,
+      "loss": 3.1253,
+      "step": 32785
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5776566863059998,
+      "learning_rate": 0.00026652923003626916,
+      "loss": 3.0857,
+      "step": 32786
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5935328602790833,
+      "learning_rate": 0.00026651393001353956,
+      "loss": 2.9172,
+      "step": 32787
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5658502578735352,
+      "learning_rate": 0.0002664986300790057,
+      "loss": 2.8991,
+      "step": 32788
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6034933924674988,
+      "learning_rate": 0.0002664833302327081,
+      "loss": 3.1756,
+      "step": 32789
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5854072570800781,
+      "learning_rate": 0.00026646803047468697,
+      "loss": 2.8792,
+      "step": 32790
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5877964496612549,
+      "learning_rate": 0.00026645273080498264,
+      "loss": 2.9641,
+      "step": 32791
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6241223216056824,
+      "learning_rate": 0.00026643743122363515,
+      "loss": 2.9318,
+      "step": 32792
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5755746364593506,
+      "learning_rate": 0.00026642213173068516,
+      "loss": 2.9953,
+      "step": 32793
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6470869779586792,
+      "learning_rate": 0.00026640683232617284,
+      "loss": 2.8269,
+      "step": 32794
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6277267932891846,
+      "learning_rate": 0.0002663915330101383,
+      "loss": 3.0639,
+      "step": 32795
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6223757266998291,
+      "learning_rate": 0.0002663762337826222,
+      "loss": 2.9395,
+      "step": 32796
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5803819894790649,
+      "learning_rate": 0.0002663609346436645,
+      "loss": 2.8632,
+      "step": 32797
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.659708559513092,
+      "learning_rate": 0.00026634563559330555,
+      "loss": 2.9963,
+      "step": 32798
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6480075716972351,
+      "learning_rate": 0.0002663303366315859,
+      "loss": 3.0471,
+      "step": 32799
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5975663661956787,
+      "learning_rate": 0.00026631503775854557,
+      "loss": 2.7081,
+      "step": 32800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5907651782035828,
+      "learning_rate": 0.000266299738974225,
+      "loss": 3.1879,
+      "step": 32801
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6050690412521362,
+      "learning_rate": 0.00026628444027866437,
+      "loss": 3.0749,
+      "step": 32802
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.638195276260376,
+      "learning_rate": 0.00026626914167190394,
+      "loss": 3.0082,
+      "step": 32803
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5860411524772644,
+      "learning_rate": 0.0002662538431539843,
+      "loss": 2.9251,
+      "step": 32804
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5866068005561829,
+      "learning_rate": 0.00026623854472494546,
+      "loss": 2.9912,
+      "step": 32805
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6038371324539185,
+      "learning_rate": 0.00026622324638482785,
+      "loss": 2.9756,
+      "step": 32806
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5844926834106445,
+      "learning_rate": 0.0002662079481336716,
+      "loss": 3.011,
+      "step": 32807
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5914148092269897,
+      "learning_rate": 0.00026619264997151715,
+      "loss": 3.2044,
+      "step": 32808
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5837469100952148,
+      "learning_rate": 0.0002661773518984048,
+      "loss": 2.9881,
+      "step": 32809
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.60676109790802,
+      "learning_rate": 0.0002661620539143747,
+      "loss": 3.1325,
+      "step": 32810
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6246742010116577,
+      "learning_rate": 0.0002661467560194674,
+      "loss": 2.9884,
+      "step": 32811
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5762494206428528,
+      "learning_rate": 0.0002661314582137229,
+      "loss": 3.31,
+      "step": 32812
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5754257440567017,
+      "learning_rate": 0.00026611616049718157,
+      "loss": 3.0474,
+      "step": 32813
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6173507571220398,
+      "learning_rate": 0.0002661008628698839,
+      "loss": 3.0469,
+      "step": 32814
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5958468914031982,
+      "learning_rate": 0.00026608556533186997,
+      "loss": 2.8676,
+      "step": 32815
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6056309342384338,
+      "learning_rate": 0.0002660702678831802,
+      "loss": 3.2593,
+      "step": 32816
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6290296316146851,
+      "learning_rate": 0.00026605497052385464,
+      "loss": 3.0761,
+      "step": 32817
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6019764542579651,
+      "learning_rate": 0.00026603967325393386,
+      "loss": 2.9699,
+      "step": 32818
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6389729976654053,
+      "learning_rate": 0.00026602437607345814,
+      "loss": 3.1099,
+      "step": 32819
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6280241012573242,
+      "learning_rate": 0.0002660090789824675,
+      "loss": 2.9096,
+      "step": 32820
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6099143028259277,
+      "learning_rate": 0.00026599378198100253,
+      "loss": 3.0422,
+      "step": 32821
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6680817604064941,
+      "learning_rate": 0.0002659784850691033,
+      "loss": 3.2611,
+      "step": 32822
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7214831113815308,
+      "learning_rate": 0.0002659631882468102,
+      "loss": 3.0891,
+      "step": 32823
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6249724626541138,
+      "learning_rate": 0.0002659478915141636,
+      "loss": 2.9744,
+      "step": 32824
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6301910877227783,
+      "learning_rate": 0.0002659325948712036,
+      "loss": 2.9372,
+      "step": 32825
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6053838133811951,
+      "learning_rate": 0.0002659172983179707,
+      "loss": 3.2256,
+      "step": 32826
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6509674191474915,
+      "learning_rate": 0.00026590200185450504,
+      "loss": 3.0197,
+      "step": 32827
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6163961887359619,
+      "learning_rate": 0.0002658867054808468,
+      "loss": 3.0188,
+      "step": 32828
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5931441783905029,
+      "learning_rate": 0.00026587140919703665,
+      "loss": 3.0105,
+      "step": 32829
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6105912923812866,
+      "learning_rate": 0.0002658561130031145,
+      "loss": 2.8814,
+      "step": 32830
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6719006896018982,
+      "learning_rate": 0.0002658408168991209,
+      "loss": 2.9579,
+      "step": 32831
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5960902571678162,
+      "learning_rate": 0.0002658255208850958,
+      "loss": 3.1361,
+      "step": 32832
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5977107882499695,
+      "learning_rate": 0.0002658102249610798,
+      "loss": 3.2039,
+      "step": 32833
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6679555177688599,
+      "learning_rate": 0.00026579492912711316,
+      "loss": 3.3237,
+      "step": 32834
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6147051453590393,
+      "learning_rate": 0.000265779633383236,
+      "loss": 3.0583,
+      "step": 32835
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6416993141174316,
+      "learning_rate": 0.0002657643377294888,
+      "loss": 3.0023,
+      "step": 32836
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5568658113479614,
+      "learning_rate": 0.0002657490421659117,
+      "loss": 2.8273,
+      "step": 32837
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6182296872138977,
+      "learning_rate": 0.0002657337466925449,
+      "loss": 3.1808,
+      "step": 32838
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.61872398853302,
+      "learning_rate": 0.00026571845130942903,
+      "loss": 2.9039,
+      "step": 32839
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5836578011512756,
+      "learning_rate": 0.000265703156016604,
+      "loss": 2.8411,
+      "step": 32840
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6261762976646423,
+      "learning_rate": 0.0002656878608141104,
+      "loss": 3.0058,
+      "step": 32841
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7778844237327576,
+      "learning_rate": 0.00026567256570198824,
+      "loss": 2.8564,
+      "step": 32842
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6407623887062073,
+      "learning_rate": 0.0002656572706802779,
+      "loss": 2.9384,
+      "step": 32843
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5836843252182007,
+      "learning_rate": 0.00026564197574901994,
+      "loss": 3.0409,
+      "step": 32844
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6171208024024963,
+      "learning_rate": 0.00026562668090825413,
+      "loss": 2.9327,
+      "step": 32845
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6322222352027893,
+      "learning_rate": 0.0002656113861580212,
+      "loss": 3.12,
+      "step": 32846
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0868051052093506,
+      "learning_rate": 0.0002655960914983612,
+      "loss": 3.0806,
+      "step": 32847
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6040574908256531,
+      "learning_rate": 0.00026558079692931444,
+      "loss": 3.019,
+      "step": 32848
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.693903386592865,
+      "learning_rate": 0.00026556550245092136,
+      "loss": 2.8479,
+      "step": 32849
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6436097025871277,
+      "learning_rate": 0.00026555020806322196,
+      "loss": 3.0285,
+      "step": 32850
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6326521039009094,
+      "learning_rate": 0.0002655349137662568,
+      "loss": 3.076,
+      "step": 32851
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6029640436172485,
+      "learning_rate": 0.00026551961956006604,
+      "loss": 2.957,
+      "step": 32852
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5923823118209839,
+      "learning_rate": 0.0002655043254446898,
+      "loss": 2.9685,
+      "step": 32853
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6047025322914124,
+      "learning_rate": 0.00026548903142016876,
+      "loss": 3.0945,
+      "step": 32854
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6205102205276489,
+      "learning_rate": 0.0002654737374865428,
+      "loss": 2.7298,
+      "step": 32855
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.636955976486206,
+      "learning_rate": 0.00026545844364385255,
+      "loss": 3.0996,
+      "step": 32856
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6210900545120239,
+      "learning_rate": 0.00026544314989213795,
+      "loss": 2.9344,
+      "step": 32857
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5878751873970032,
+      "learning_rate": 0.0002654278562314395,
+      "loss": 2.929,
+      "step": 32858
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6176540851593018,
+      "learning_rate": 0.00026541256266179744,
+      "loss": 3.2169,
+      "step": 32859
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7335445284843445,
+      "learning_rate": 0.0002653972691832519,
+      "loss": 3.117,
+      "step": 32860
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6168895363807678,
+      "learning_rate": 0.0002653819757958435,
+      "loss": 3.0956,
+      "step": 32861
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6440188884735107,
+      "learning_rate": 0.0002653666824996122,
+      "loss": 3.0447,
+      "step": 32862
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5899813771247864,
+      "learning_rate": 0.00026535138929459834,
+      "loss": 2.9709,
+      "step": 32863
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6249847412109375,
+      "learning_rate": 0.0002653360961808424,
+      "loss": 3.1298,
+      "step": 32864
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5660920143127441,
+      "learning_rate": 0.00026532080315838435,
+      "loss": 3.0273,
+      "step": 32865
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6059000492095947,
+      "learning_rate": 0.00026530551022726476,
+      "loss": 3.014,
+      "step": 32866
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6058610081672668,
+      "learning_rate": 0.00026529021738752373,
+      "loss": 3.0652,
+      "step": 32867
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6035959124565125,
+      "learning_rate": 0.00026527492463920154,
+      "loss": 3.1049,
+      "step": 32868
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5923382639884949,
+      "learning_rate": 0.00026525963198233864,
+      "loss": 3.1006,
+      "step": 32869
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6309588551521301,
+      "learning_rate": 0.0002652443394169751,
+      "loss": 3.1369,
+      "step": 32870
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5777470469474792,
+      "learning_rate": 0.0002652290469431513,
+      "loss": 2.9681,
+      "step": 32871
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.604954183101654,
+      "learning_rate": 0.00026521375456090745,
+      "loss": 3.0512,
+      "step": 32872
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.618948757648468,
+      "learning_rate": 0.000265198462270284,
+      "loss": 3.152,
+      "step": 32873
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5669094920158386,
+      "learning_rate": 0.000265183170071321,
+      "loss": 3.0005,
+      "step": 32874
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6252792477607727,
+      "learning_rate": 0.0002651678779640588,
+      "loss": 3.0027,
+      "step": 32875
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6201028823852539,
+      "learning_rate": 0.0002651525859485378,
+      "loss": 2.932,
+      "step": 32876
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5937361717224121,
+      "learning_rate": 0.0002651372940247981,
+      "loss": 3.0645,
+      "step": 32877
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6492998003959656,
+      "learning_rate": 0.00026512200219288017,
+      "loss": 3.1435,
+      "step": 32878
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.625295877456665,
+      "learning_rate": 0.000265106710452824,
+      "loss": 2.965,
+      "step": 32879
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5870344042778015,
+      "learning_rate": 0.00026509141880467016,
+      "loss": 2.9514,
+      "step": 32880
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5917270183563232,
+      "learning_rate": 0.00026507612724845883,
+      "loss": 2.9134,
+      "step": 32881
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6095778346061707,
+      "learning_rate": 0.00026506083578423014,
+      "loss": 3.1039,
+      "step": 32882
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.66627436876297,
+      "learning_rate": 0.00026504554441202457,
+      "loss": 2.996,
+      "step": 32883
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6477625370025635,
+      "learning_rate": 0.00026503025313188235,
+      "loss": 2.9929,
+      "step": 32884
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5901625156402588,
+      "learning_rate": 0.0002650149619438436,
+      "loss": 3.2471,
+      "step": 32885
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6187769174575806,
+      "learning_rate": 0.0002649996708479488,
+      "loss": 3.0292,
+      "step": 32886
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6475508809089661,
+      "learning_rate": 0.000264984379844238,
+      "loss": 3.0723,
+      "step": 32887
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5832803845405579,
+      "learning_rate": 0.0002649690889327518,
+      "loss": 2.8318,
+      "step": 32888
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6353322863578796,
+      "learning_rate": 0.00026495379811353014,
+      "loss": 3.0975,
+      "step": 32889
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5817992091178894,
+      "learning_rate": 0.0002649385073866134,
+      "loss": 3.0773,
+      "step": 32890
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.596846342086792,
+      "learning_rate": 0.000264923216752042,
+      "loss": 3.1376,
+      "step": 32891
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6275048851966858,
+      "learning_rate": 0.00026490792620985603,
+      "loss": 3.2158,
+      "step": 32892
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5930907130241394,
+      "learning_rate": 0.0002648926357600959,
+      "loss": 3.2069,
+      "step": 32893
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.597496747970581,
+      "learning_rate": 0.00026487734540280174,
+      "loss": 2.8155,
+      "step": 32894
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6274101734161377,
+      "learning_rate": 0.0002648620551380139,
+      "loss": 3.1146,
+      "step": 32895
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6199111342430115,
+      "learning_rate": 0.0002648467649657727,
+      "loss": 3.0811,
+      "step": 32896
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6213255524635315,
+      "learning_rate": 0.0002648314748861182,
+      "loss": 3.0629,
+      "step": 32897
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6072750091552734,
+      "learning_rate": 0.000264816184899091,
+      "loss": 3.0835,
+      "step": 32898
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7357712388038635,
+      "learning_rate": 0.0002648008950047312,
+      "loss": 2.8921,
+      "step": 32899
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.637231707572937,
+      "learning_rate": 0.00026478560520307886,
+      "loss": 3.1188,
+      "step": 32900
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5956443548202515,
+      "learning_rate": 0.00026477031549417473,
+      "loss": 3.0897,
+      "step": 32901
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5935696959495544,
+      "learning_rate": 0.0002647550258780587,
+      "loss": 2.942,
+      "step": 32902
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.659156322479248,
+      "learning_rate": 0.00026473973635477116,
+      "loss": 3.1397,
+      "step": 32903
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6243013739585876,
+      "learning_rate": 0.00026472444692435227,
+      "loss": 2.8611,
+      "step": 32904
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5944880843162537,
+      "learning_rate": 0.00026470915758684235,
+      "loss": 2.9702,
+      "step": 32905
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6213318705558777,
+      "learning_rate": 0.0002646938683422819,
+      "loss": 3.0618,
+      "step": 32906
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6070886254310608,
+      "learning_rate": 0.0002646785791907109,
+      "loss": 3.1214,
+      "step": 32907
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6019971966743469,
+      "learning_rate": 0.00026466329013216986,
+      "loss": 3.0693,
+      "step": 32908
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5632584691047668,
+      "learning_rate": 0.00026464800116669877,
+      "loss": 2.8446,
+      "step": 32909
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6107988953590393,
+      "learning_rate": 0.0002646327122943381,
+      "loss": 2.9715,
+      "step": 32910
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6390516757965088,
+      "learning_rate": 0.0002646174235151281,
+      "loss": 3.0538,
+      "step": 32911
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5974604487419128,
+      "learning_rate": 0.00026460213482910883,
+      "loss": 2.9292,
+      "step": 32912
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.603056788444519,
+      "learning_rate": 0.0002645868462363209,
+      "loss": 3.1131,
+      "step": 32913
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6022782921791077,
+      "learning_rate": 0.00026457155773680434,
+      "loss": 3.1211,
+      "step": 32914
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6246730089187622,
+      "learning_rate": 0.00026455626933059943,
+      "loss": 3.1058,
+      "step": 32915
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6355071067810059,
+      "learning_rate": 0.0002645409810177466,
+      "loss": 3.1005,
+      "step": 32916
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5840960741043091,
+      "learning_rate": 0.0002645256927982859,
+      "loss": 3.2979,
+      "step": 32917
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.614286482334137,
+      "learning_rate": 0.0002645104046722578,
+      "loss": 2.919,
+      "step": 32918
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6425104141235352,
+      "learning_rate": 0.00026449511663970235,
+      "loss": 3.1202,
+      "step": 32919
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.617282509803772,
+      "learning_rate": 0.00026447982870065993,
+      "loss": 3.072,
+      "step": 32920
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5912017822265625,
+      "learning_rate": 0.00026446454085517093,
+      "loss": 2.8702,
+      "step": 32921
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6111149787902832,
+      "learning_rate": 0.00026444925310327536,
+      "loss": 2.9778,
+      "step": 32922
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6221730709075928,
+      "learning_rate": 0.0002644339654450137,
+      "loss": 3.289,
+      "step": 32923
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6156650185585022,
+      "learning_rate": 0.0002644186778804261,
+      "loss": 3.1326,
+      "step": 32924
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6163793802261353,
+      "learning_rate": 0.0002644033904095528,
+      "loss": 3.0494,
+      "step": 32925
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6287680268287659,
+      "learning_rate": 0.00026438810303243426,
+      "loss": 3.0893,
+      "step": 32926
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6106224060058594,
+      "learning_rate": 0.00026437281574911043,
+      "loss": 2.9948,
+      "step": 32927
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6386653184890747,
+      "learning_rate": 0.0002643575285596219,
+      "loss": 3.023,
+      "step": 32928
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6105353236198425,
+      "learning_rate": 0.00026434224146400866,
+      "loss": 3.0687,
+      "step": 32929
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5927025079727173,
+      "learning_rate": 0.0002643269544623111,
+      "loss": 2.9087,
+      "step": 32930
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6284357309341431,
+      "learning_rate": 0.00026431166755456955,
+      "loss": 2.934,
+      "step": 32931
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6147943139076233,
+      "learning_rate": 0.0002642963807408242,
+      "loss": 2.9622,
+      "step": 32932
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6074808239936829,
+      "learning_rate": 0.00026428109402111533,
+      "loss": 2.9748,
+      "step": 32933
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6237158179283142,
+      "learning_rate": 0.00026426580739548306,
+      "loss": 3.109,
+      "step": 32934
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.594880223274231,
+      "learning_rate": 0.0002642505208639679,
+      "loss": 3.1491,
+      "step": 32935
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5945537686347961,
+      "learning_rate": 0.00026423523442661,
+      "loss": 2.9643,
+      "step": 32936
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6391527652740479,
+      "learning_rate": 0.0002642199480834495,
+      "loss": 2.9662,
+      "step": 32937
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6021575927734375,
+      "learning_rate": 0.00026420466183452683,
+      "loss": 3.1065,
+      "step": 32938
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6353442072868347,
+      "learning_rate": 0.0002641893756798822,
+      "loss": 2.9833,
+      "step": 32939
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6768826246261597,
+      "learning_rate": 0.00026417408961955575,
+      "loss": 2.8308,
+      "step": 32940
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5737194418907166,
+      "learning_rate": 0.00026415880365358803,
+      "loss": 3.0519,
+      "step": 32941
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5978490710258484,
+      "learning_rate": 0.000264143517782019,
+      "loss": 2.8875,
+      "step": 32942
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6324152946472168,
+      "learning_rate": 0.0002641282320048892,
+      "loss": 3.2518,
+      "step": 32943
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6049727201461792,
+      "learning_rate": 0.0002641129463222385,
+      "loss": 2.9631,
+      "step": 32944
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.584157407283783,
+      "learning_rate": 0.0002640976607341075,
+      "loss": 3.0125,
+      "step": 32945
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.603410005569458,
+      "learning_rate": 0.00026408237524053646,
+      "loss": 2.7752,
+      "step": 32946
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6187100410461426,
+      "learning_rate": 0.0002640670898415654,
+      "loss": 3.0289,
+      "step": 32947
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5807082653045654,
+      "learning_rate": 0.0002640518045372348,
+      "loss": 3.1831,
+      "step": 32948
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6443330645561218,
+      "learning_rate": 0.00026403651932758467,
+      "loss": 3.1691,
+      "step": 32949
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6212769746780396,
+      "learning_rate": 0.0002640212342126556,
+      "loss": 2.9774,
+      "step": 32950
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5869172811508179,
+      "learning_rate": 0.00026400594919248757,
+      "loss": 2.8276,
+      "step": 32951
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6382247805595398,
+      "learning_rate": 0.00026399066426712094,
+      "loss": 3.0276,
+      "step": 32952
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6578730344772339,
+      "learning_rate": 0.0002639753794365961,
+      "loss": 3.0336,
+      "step": 32953
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6229750514030457,
+      "learning_rate": 0.000263960094700953,
+      "loss": 3.0683,
+      "step": 32954
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6478546261787415,
+      "learning_rate": 0.0002639448100602323,
+      "loss": 2.9494,
+      "step": 32955
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.607498288154602,
+      "learning_rate": 0.0002639295255144738,
+      "loss": 3.0408,
+      "step": 32956
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6057003736495972,
+      "learning_rate": 0.00026391424106371805,
+      "loss": 3.2572,
+      "step": 32957
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5550916194915771,
+      "learning_rate": 0.00026389895670800534,
+      "loss": 2.9359,
+      "step": 32958
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6296790242195129,
+      "learning_rate": 0.0002638836724473757,
+      "loss": 3.0151,
+      "step": 32959
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5951831936836243,
+      "learning_rate": 0.00026386838828186966,
+      "loss": 2.8441,
+      "step": 32960
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6196625828742981,
+      "learning_rate": 0.00026385310421152725,
+      "loss": 2.8942,
+      "step": 32961
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6182433366775513,
+      "learning_rate": 0.00026383782023638874,
+      "loss": 2.9938,
+      "step": 32962
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6165900230407715,
+      "learning_rate": 0.0002638225363564946,
+      "loss": 2.8868,
+      "step": 32963
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6178959012031555,
+      "learning_rate": 0.0002638072525718848,
+      "loss": 2.8964,
+      "step": 32964
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6362462639808655,
+      "learning_rate": 0.00026379196888259984,
+      "loss": 2.8124,
+      "step": 32965
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6361078023910522,
+      "learning_rate": 0.00026377668528867977,
+      "loss": 3.0088,
+      "step": 32966
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7130830883979797,
+      "learning_rate": 0.00026376140179016494,
+      "loss": 3.1082,
+      "step": 32967
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6127350926399231,
+      "learning_rate": 0.0002637461183870957,
+      "loss": 3.1208,
+      "step": 32968
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5895925164222717,
+      "learning_rate": 0.0002637308350795121,
+      "loss": 2.8528,
+      "step": 32969
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6005771160125732,
+      "learning_rate": 0.00026371555186745465,
+      "loss": 3.0934,
+      "step": 32970
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6108828783035278,
+      "learning_rate": 0.0002637002687509633,
+      "loss": 2.9517,
+      "step": 32971
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5996699333190918,
+      "learning_rate": 0.0002636849857300785,
+      "loss": 2.9753,
+      "step": 32972
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5930362939834595,
+      "learning_rate": 0.0002636697028048405,
+      "loss": 2.9364,
+      "step": 32973
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6767230033874512,
+      "learning_rate": 0.0002636544199752894,
+      "loss": 3.1138,
+      "step": 32974
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6063745021820068,
+      "learning_rate": 0.0002636391372414657,
+      "loss": 3.1365,
+      "step": 32975
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6611247658729553,
+      "learning_rate": 0.00026362385460340946,
+      "loss": 3.0819,
+      "step": 32976
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6121769547462463,
+      "learning_rate": 0.00026360857206116095,
+      "loss": 2.8333,
+      "step": 32977
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6047820448875427,
+      "learning_rate": 0.00026359328961476057,
+      "loss": 3.1222,
+      "step": 32978
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6275504231452942,
+      "learning_rate": 0.00026357800726424833,
+      "loss": 3.3318,
+      "step": 32979
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6080473065376282,
+      "learning_rate": 0.0002635627250096647,
+      "loss": 2.9658,
+      "step": 32980
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6527457237243652,
+      "learning_rate": 0.00026354744285104977,
+      "loss": 3.0866,
+      "step": 32981
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6376957297325134,
+      "learning_rate": 0.0002635321607884439,
+      "loss": 3.1455,
+      "step": 32982
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6095510125160217,
+      "learning_rate": 0.0002635168788218874,
+      "loss": 2.9285,
+      "step": 32983
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6063658595085144,
+      "learning_rate": 0.0002635015969514202,
+      "loss": 3.2578,
+      "step": 32984
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5850301384925842,
+      "learning_rate": 0.00026348631517708295,
+      "loss": 3.193,
+      "step": 32985
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6624311804771423,
+      "learning_rate": 0.00026347103349891565,
+      "loss": 3.1238,
+      "step": 32986
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5807081460952759,
+      "learning_rate": 0.00026345575191695855,
+      "loss": 3.0628,
+      "step": 32987
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5957835912704468,
+      "learning_rate": 0.0002634404704312521,
+      "loss": 2.947,
+      "step": 32988
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6118351817131042,
+      "learning_rate": 0.00026342518904183624,
+      "loss": 3.0652,
+      "step": 32989
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5782414674758911,
+      "learning_rate": 0.0002634099077487516,
+      "loss": 2.9368,
+      "step": 32990
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6376646161079407,
+      "learning_rate": 0.0002633946265520381,
+      "loss": 2.9095,
+      "step": 32991
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7035831809043884,
+      "learning_rate": 0.00026337934545173606,
+      "loss": 2.9032,
+      "step": 32992
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6374669075012207,
+      "learning_rate": 0.0002633640644478859,
+      "loss": 3.1269,
+      "step": 32993
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6482071280479431,
+      "learning_rate": 0.0002633487835405277,
+      "loss": 2.9167,
+      "step": 32994
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5882799029350281,
+      "learning_rate": 0.0002633335027297018,
+      "loss": 2.8802,
+      "step": 32995
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6047265529632568,
+      "learning_rate": 0.00026331822201544825,
+      "loss": 3.0355,
+      "step": 32996
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5921814441680908,
+      "learning_rate": 0.00026330294139780756,
+      "loss": 2.9725,
+      "step": 32997
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6499700546264648,
+      "learning_rate": 0.00026328766087681986,
+      "loss": 2.9638,
+      "step": 32998
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6229745149612427,
+      "learning_rate": 0.00026327238045252533,
+      "loss": 3.066,
+      "step": 32999
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6632835268974304,
+      "learning_rate": 0.00026325710012496434,
+      "loss": 2.9409,
+      "step": 33000
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5706971287727356,
+      "learning_rate": 0.00026324181989417707,
+      "loss": 2.9932,
+      "step": 33001
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6048433184623718,
+      "learning_rate": 0.0002632265397602037,
+      "loss": 2.9595,
+      "step": 33002
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6280288100242615,
+      "learning_rate": 0.0002632112597230847,
+      "loss": 3.1984,
+      "step": 33003
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6192353963851929,
+      "learning_rate": 0.00026319597978286,
+      "loss": 2.9718,
+      "step": 33004
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6082794070243835,
+      "learning_rate": 0.0002631806999395702,
+      "loss": 3.0911,
+      "step": 33005
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6182049512863159,
+      "learning_rate": 0.00026316542019325513,
+      "loss": 3.064,
+      "step": 33006
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6134260296821594,
+      "learning_rate": 0.0002631501405439553,
+      "loss": 3.0449,
+      "step": 33007
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6304604411125183,
+      "learning_rate": 0.0002631348609917111,
+      "loss": 2.9606,
+      "step": 33008
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5884554386138916,
+      "learning_rate": 0.00026311958153656237,
+      "loss": 3.0864,
+      "step": 33009
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5836405158042908,
+      "learning_rate": 0.00026310430217854966,
+      "loss": 3.4003,
+      "step": 33010
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6103641986846924,
+      "learning_rate": 0.000263089022917713,
+      "loss": 3.1367,
+      "step": 33011
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6257080435752869,
+      "learning_rate": 0.0002630737437540928,
+      "loss": 3.2796,
+      "step": 33012
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6084847450256348,
+      "learning_rate": 0.0002630584646877294,
+      "loss": 3.099,
+      "step": 33013
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6533360481262207,
+      "learning_rate": 0.0002630431857186627,
+      "loss": 3.201,
+      "step": 33014
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.65522700548172,
+      "learning_rate": 0.0002630279068469333,
+      "loss": 2.9477,
+      "step": 33015
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6127893328666687,
+      "learning_rate": 0.0002630126280725812,
+      "loss": 2.9227,
+      "step": 33016
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6066474318504333,
+      "learning_rate": 0.00026299734939564664,
+      "loss": 2.9635,
+      "step": 33017
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6110500693321228,
+      "learning_rate": 0.0002629820708161701,
+      "loss": 3.1665,
+      "step": 33018
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6653189063072205,
+      "learning_rate": 0.00026296679233419153,
+      "loss": 3.1245,
+      "step": 33019
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5800127983093262,
+      "learning_rate": 0.00026295151394975145,
+      "loss": 2.8873,
+      "step": 33020
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6234315037727356,
+      "learning_rate": 0.0002629362356628898,
+      "loss": 2.9345,
+      "step": 33021
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6198316216468811,
+      "learning_rate": 0.000262920957473647,
+      "loss": 2.8847,
+      "step": 33022
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6274871230125427,
+      "learning_rate": 0.0002629056793820634,
+      "loss": 3.0915,
+      "step": 33023
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6605402231216431,
+      "learning_rate": 0.0002628904013881789,
+      "loss": 2.9215,
+      "step": 33024
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.634557843208313,
+      "learning_rate": 0.00026287512349203407,
+      "loss": 2.955,
+      "step": 33025
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6039037108421326,
+      "learning_rate": 0.000262859845693669,
+      "loss": 3.1472,
+      "step": 33026
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6004440784454346,
+      "learning_rate": 0.000262844567993124,
+      "loss": 3.1694,
+      "step": 33027
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5967251658439636,
+      "learning_rate": 0.0002628292903904391,
+      "loss": 2.922,
+      "step": 33028
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5802043676376343,
+      "learning_rate": 0.0002628140128856547,
+      "loss": 3.0168,
+      "step": 33029
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5677957534790039,
+      "learning_rate": 0.00026279873547881117,
+      "loss": 2.8979,
+      "step": 33030
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6296228170394897,
+      "learning_rate": 0.0002627834581699486,
+      "loss": 2.9727,
+      "step": 33031
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6011860966682434,
+      "learning_rate": 0.0002627681809591072,
+      "loss": 3.1614,
+      "step": 33032
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6390479803085327,
+      "learning_rate": 0.0002627529038463272,
+      "loss": 3.0619,
+      "step": 33033
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7037544846534729,
+      "learning_rate": 0.0002627376268316489,
+      "loss": 3.0301,
+      "step": 33034
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6357463598251343,
+      "learning_rate": 0.00026272234991511263,
+      "loss": 3.0546,
+      "step": 33035
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5906470417976379,
+      "learning_rate": 0.0002627070730967584,
+      "loss": 3.1207,
+      "step": 33036
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6273382306098938,
+      "learning_rate": 0.0002626917963766266,
+      "loss": 3.0675,
+      "step": 33037
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5891216993331909,
+      "learning_rate": 0.00026267651975475745,
+      "loss": 3.158,
+      "step": 33038
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6417419910430908,
+      "learning_rate": 0.00026266124323119107,
+      "loss": 3.1118,
+      "step": 33039
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5798816680908203,
+      "learning_rate": 0.0002626459668059679,
+      "loss": 3.1448,
+      "step": 33040
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6085364818572998,
+      "learning_rate": 0.000262630690479128,
+      "loss": 3.1527,
+      "step": 33041
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6986152529716492,
+      "learning_rate": 0.00026261541425071176,
+      "loss": 2.9744,
+      "step": 33042
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5539711713790894,
+      "learning_rate": 0.0002626001381207592,
+      "loss": 3.1058,
+      "step": 33043
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6884754300117493,
+      "learning_rate": 0.00026258486208931075,
+      "loss": 3.0357,
+      "step": 33044
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6211069226264954,
+      "learning_rate": 0.0002625695861564066,
+      "loss": 3.2542,
+      "step": 33045
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6055322885513306,
+      "learning_rate": 0.00026255431032208684,
+      "loss": 3.275,
+      "step": 33046
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.581346333026886,
+      "learning_rate": 0.00026253903458639197,
+      "loss": 2.9475,
+      "step": 33047
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5834121704101562,
+      "learning_rate": 0.00026252375894936193,
+      "loss": 3.3297,
+      "step": 33048
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.610347330570221,
+      "learning_rate": 0.0002625084834110371,
+      "loss": 2.9988,
+      "step": 33049
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6127068400382996,
+      "learning_rate": 0.00026249320797145785,
+      "loss": 2.843,
+      "step": 33050
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5858662128448486,
+      "learning_rate": 0.0002624779326306641,
+      "loss": 3.0937,
+      "step": 33051
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5876694917678833,
+      "learning_rate": 0.00026246265738869647,
+      "loss": 2.9783,
+      "step": 33052
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.622711718082428,
+      "learning_rate": 0.0002624473822455948,
+      "loss": 2.9802,
+      "step": 33053
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5976312160491943,
+      "learning_rate": 0.0002624321072013995,
+      "loss": 3.0595,
+      "step": 33054
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5904166102409363,
+      "learning_rate": 0.0002624168322561509,
+      "loss": 2.9977,
+      "step": 33055
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7579708099365234,
+      "learning_rate": 0.0002624015574098891,
+      "loss": 2.9153,
+      "step": 33056
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6306504011154175,
+      "learning_rate": 0.0002623862826626544,
+      "loss": 3.1642,
+      "step": 33057
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6161608695983887,
+      "learning_rate": 0.0002623710080144868,
+      "loss": 3.1045,
+      "step": 33058
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6324542760848999,
+      "learning_rate": 0.0002623557334654269,
+      "loss": 3.1015,
+      "step": 33059
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6091515421867371,
+      "learning_rate": 0.00026234045901551474,
+      "loss": 3.0746,
+      "step": 33060
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6115736961364746,
+      "learning_rate": 0.0002623251846647905,
+      "loss": 3.0863,
+      "step": 33061
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6223326325416565,
+      "learning_rate": 0.00026230991041329457,
+      "loss": 2.998,
+      "step": 33062
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5899375677108765,
+      "learning_rate": 0.000262294636261067,
+      "loss": 3.0513,
+      "step": 33063
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.577978789806366,
+      "learning_rate": 0.00026227936220814805,
+      "loss": 3.0171,
+      "step": 33064
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5958923697471619,
+      "learning_rate": 0.00026226408825457813,
+      "loss": 2.9265,
+      "step": 33065
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6276746988296509,
+      "learning_rate": 0.0002622488144003973,
+      "loss": 2.8662,
+      "step": 33066
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5911863446235657,
+      "learning_rate": 0.0002622335406456459,
+      "loss": 3.0632,
+      "step": 33067
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6043195724487305,
+      "learning_rate": 0.0002622182669903639,
+      "loss": 2.8693,
+      "step": 33068
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6209897994995117,
+      "learning_rate": 0.0002622029934345917,
+      "loss": 2.8144,
+      "step": 33069
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5737490653991699,
+      "learning_rate": 0.0002621877199783697,
+      "loss": 3.1074,
+      "step": 33070
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5962395668029785,
+      "learning_rate": 0.00026217244662173793,
+      "loss": 2.9208,
+      "step": 33071
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6508395671844482,
+      "learning_rate": 0.00026215717336473674,
+      "loss": 2.7527,
+      "step": 33072
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6254932284355164,
+      "learning_rate": 0.0002621419002074061,
+      "loss": 2.7828,
+      "step": 33073
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6272798180580139,
+      "learning_rate": 0.0002621266271497865,
+      "loss": 3.0308,
+      "step": 33074
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5974328517913818,
+      "learning_rate": 0.0002621113541919181,
+      "loss": 2.8549,
+      "step": 33075
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6224279403686523,
+      "learning_rate": 0.000262096081333841,
+      "loss": 2.9043,
+      "step": 33076
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6068066358566284,
+      "learning_rate": 0.0002620808085755956,
+      "loss": 3.1066,
+      "step": 33077
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6328063011169434,
+      "learning_rate": 0.0002620655359172221,
+      "loss": 3.0899,
+      "step": 33078
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6107156276702881,
+      "learning_rate": 0.00026205026335876047,
+      "loss": 3.0569,
+      "step": 33079
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6129263639450073,
+      "learning_rate": 0.00026203499090025136,
+      "loss": 3.0112,
+      "step": 33080
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6082353591918945,
+      "learning_rate": 0.00026201971854173474,
+      "loss": 3.0659,
+      "step": 33081
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6048126816749573,
+      "learning_rate": 0.0002620044462832509,
+      "loss": 2.9859,
+      "step": 33082
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5699526071548462,
+      "learning_rate": 0.0002619891741248399,
+      "loss": 3.1299,
+      "step": 33083
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5666524767875671,
+      "learning_rate": 0.0002619739020665422,
+      "loss": 3.215,
+      "step": 33084
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5848353505134583,
+      "learning_rate": 0.00026195863010839793,
+      "loss": 3.0975,
+      "step": 33085
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5911352634429932,
+      "learning_rate": 0.0002619433582504472,
+      "loss": 3.1146,
+      "step": 33086
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6226699352264404,
+      "learning_rate": 0.0002619280864927305,
+      "loss": 2.8967,
+      "step": 33087
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5983354449272156,
+      "learning_rate": 0.0002619128148352878,
+      "loss": 3.0173,
+      "step": 33088
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.643765389919281,
+      "learning_rate": 0.00026189754327815936,
+      "loss": 3.1155,
+      "step": 33089
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5957856774330139,
+      "learning_rate": 0.0002618822718213856,
+      "loss": 2.9854,
+      "step": 33090
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.565482497215271,
+      "learning_rate": 0.00026186700046500644,
+      "loss": 2.9022,
+      "step": 33091
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5913927555084229,
+      "learning_rate": 0.0002618517292090624,
+      "loss": 3.1603,
+      "step": 33092
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6006325483322144,
+      "learning_rate": 0.00026183645805359353,
+      "loss": 3.2807,
+      "step": 33093
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6765154004096985,
+      "learning_rate": 0.00026182118699863996,
+      "loss": 3.0531,
+      "step": 33094
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6251957416534424,
+      "learning_rate": 0.0002618059160442422,
+      "loss": 2.9957,
+      "step": 33095
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6182376742362976,
+      "learning_rate": 0.00026179064519044027,
+      "loss": 3.0766,
+      "step": 33096
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5697817206382751,
+      "learning_rate": 0.00026177537443727443,
+      "loss": 2.8416,
+      "step": 33097
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6249179244041443,
+      "learning_rate": 0.00026176010378478484,
+      "loss": 3.007,
+      "step": 33098
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6290808320045471,
+      "learning_rate": 0.0002617448332330118,
+      "loss": 3.235,
+      "step": 33099
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6982139348983765,
+      "learning_rate": 0.0002617295627819956,
+      "loss": 3.0333,
+      "step": 33100
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6556278467178345,
+      "learning_rate": 0.00026171429243177614,
+      "loss": 3.0093,
+      "step": 33101
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6192794442176819,
+      "learning_rate": 0.0002616990221823941,
+      "loss": 3.0252,
+      "step": 33102
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6167252659797668,
+      "learning_rate": 0.0002616837520338894,
+      "loss": 2.9509,
+      "step": 33103
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.651763379573822,
+      "learning_rate": 0.0002616684819863023,
+      "loss": 3.0517,
+      "step": 33104
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6661128997802734,
+      "learning_rate": 0.00026165321203967297,
+      "loss": 2.9474,
+      "step": 33105
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6066205501556396,
+      "learning_rate": 0.00026163794219404176,
+      "loss": 3.1624,
+      "step": 33106
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6290202736854553,
+      "learning_rate": 0.0002616226724494489,
+      "loss": 2.8767,
+      "step": 33107
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6024856567382812,
+      "learning_rate": 0.0002616074028059344,
+      "loss": 2.9091,
+      "step": 33108
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5915390253067017,
+      "learning_rate": 0.00026159213326353877,
+      "loss": 2.9278,
+      "step": 33109
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6625373363494873,
+      "learning_rate": 0.00026157686382230195,
+      "loss": 3.3122,
+      "step": 33110
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7008400559425354,
+      "learning_rate": 0.00026156159448226425,
+      "loss": 3.002,
+      "step": 33111
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7132464051246643,
+      "learning_rate": 0.00026154632524346604,
+      "loss": 3.1318,
+      "step": 33112
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6390926837921143,
+      "learning_rate": 0.00026153105610594727,
+      "loss": 2.9645,
+      "step": 33113
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6167672872543335,
+      "learning_rate": 0.0002615157870697484,
+      "loss": 2.9288,
+      "step": 33114
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5881397724151611,
+      "learning_rate": 0.0002615005181349095,
+      "loss": 2.8726,
+      "step": 33115
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6101930737495422,
+      "learning_rate": 0.00026148524930147074,
+      "loss": 3.1134,
+      "step": 33116
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.602695643901825,
+      "learning_rate": 0.0002614699805694726,
+      "loss": 2.9318,
+      "step": 33117
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6428423523902893,
+      "learning_rate": 0.000261454711938955,
+      "loss": 3.018,
+      "step": 33118
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.62602299451828,
+      "learning_rate": 0.00026143944340995836,
+      "loss": 2.8108,
+      "step": 33119
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6125110387802124,
+      "learning_rate": 0.0002614241749825227,
+      "loss": 3.1053,
+      "step": 33120
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5930248498916626,
+      "learning_rate": 0.00026140890665668834,
+      "loss": 2.7055,
+      "step": 33121
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6099435091018677,
+      "learning_rate": 0.0002613936384324956,
+      "loss": 3.0313,
+      "step": 33122
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.599867045879364,
+      "learning_rate": 0.00026137837030998444,
+      "loss": 3.0244,
+      "step": 33123
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6972360014915466,
+      "learning_rate": 0.00026136310228919537,
+      "loss": 3.1866,
+      "step": 33124
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6721414923667908,
+      "learning_rate": 0.0002613478343701684,
+      "loss": 3.0925,
+      "step": 33125
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6666560769081116,
+      "learning_rate": 0.0002613325665529437,
+      "loss": 3.2204,
+      "step": 33126
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.577688992023468,
+      "learning_rate": 0.0002613172988375617,
+      "loss": 3.1363,
+      "step": 33127
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.60993492603302,
+      "learning_rate": 0.00026130203122406244,
+      "loss": 3.2141,
+      "step": 33128
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.601895809173584,
+      "learning_rate": 0.0002612867637124862,
+      "loss": 2.8998,
+      "step": 33129
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5851340889930725,
+      "learning_rate": 0.00026127149630287313,
+      "loss": 3.0916,
+      "step": 33130
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5805398225784302,
+      "learning_rate": 0.0002612562289952634,
+      "loss": 2.9505,
+      "step": 33131
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6383261680603027,
+      "learning_rate": 0.0002612409617896975,
+      "loss": 2.8642,
+      "step": 33132
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6343017816543579,
+      "learning_rate": 0.00026122569468621537,
+      "loss": 3.29,
+      "step": 33133
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6100292801856995,
+      "learning_rate": 0.00026121042768485737,
+      "loss": 2.9963,
+      "step": 33134
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5947556495666504,
+      "learning_rate": 0.0002611951607856635,
+      "loss": 2.851,
+      "step": 33135
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6661785244941711,
+      "learning_rate": 0.00026117989398867416,
+      "loss": 3.2415,
+      "step": 33136
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6562435030937195,
+      "learning_rate": 0.00026116462729392956,
+      "loss": 3.1879,
+      "step": 33137
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6270297765731812,
+      "learning_rate": 0.0002611493607014698,
+      "loss": 3.1056,
+      "step": 33138
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6291563510894775,
+      "learning_rate": 0.00026113409421133524,
+      "loss": 3.0792,
+      "step": 33139
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6328475475311279,
+      "learning_rate": 0.0002611188278235659,
+      "loss": 2.9278,
+      "step": 33140
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5976963639259338,
+      "learning_rate": 0.000261103561538202,
+      "loss": 3.1471,
+      "step": 33141
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5877504348754883,
+      "learning_rate": 0.000261088295355284,
+      "loss": 2.8909,
+      "step": 33142
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6378092169761658,
+      "learning_rate": 0.0002610730292748519,
+      "loss": 2.7738,
+      "step": 33143
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.686267614364624,
+      "learning_rate": 0.00026105776329694597,
+      "loss": 3.0721,
+      "step": 33144
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6485110521316528,
+      "learning_rate": 0.0002610424974216063,
+      "loss": 3.1614,
+      "step": 33145
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6028688549995422,
+      "learning_rate": 0.00026102723164887325,
+      "loss": 3.1251,
+      "step": 33146
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5872344374656677,
+      "learning_rate": 0.00026101196597878704,
+      "loss": 3.2345,
+      "step": 33147
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.661635160446167,
+      "learning_rate": 0.0002609967004113877,
+      "loss": 3.0462,
+      "step": 33148
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6713125705718994,
+      "learning_rate": 0.0002609814349467156,
+      "loss": 2.9479,
+      "step": 33149
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5891671776771545,
+      "learning_rate": 0.00026096616958481094,
+      "loss": 3.0252,
+      "step": 33150
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6562174558639526,
+      "learning_rate": 0.0002609509043257137,
+      "loss": 3.0676,
+      "step": 33151
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6034083366394043,
+      "learning_rate": 0.00026093563916946443,
+      "loss": 3.2758,
+      "step": 33152
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5874406099319458,
+      "learning_rate": 0.00026092037411610305,
+      "loss": 2.7679,
+      "step": 33153
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.583278238773346,
+      "learning_rate": 0.00026090510916567,
+      "loss": 3.2504,
+      "step": 33154
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6118584871292114,
+      "learning_rate": 0.00026088984431820535,
+      "loss": 3.032,
+      "step": 33155
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6118886470794678,
+      "learning_rate": 0.0002608745795737492,
+      "loss": 2.9379,
+      "step": 33156
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5924785137176514,
+      "learning_rate": 0.00026085931493234204,
+      "loss": 3.0218,
+      "step": 33157
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6517102718353271,
+      "learning_rate": 0.00026084405039402385,
+      "loss": 2.8742,
+      "step": 33158
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6144333481788635,
+      "learning_rate": 0.0002608287859588349,
+      "loss": 2.952,
+      "step": 33159
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5906076431274414,
+      "learning_rate": 0.00026081352162681535,
+      "loss": 3.0309,
+      "step": 33160
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6224218606948853,
+      "learning_rate": 0.00026079825739800544,
+      "loss": 3.0958,
+      "step": 33161
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6227795481681824,
+      "learning_rate": 0.00026078299327244545,
+      "loss": 2.8246,
+      "step": 33162
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5719353556632996,
+      "learning_rate": 0.0002607677292501754,
+      "loss": 2.9469,
+      "step": 33163
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6005598902702332,
+      "learning_rate": 0.00026075246533123573,
+      "loss": 3.0891,
+      "step": 33164
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6221149563789368,
+      "learning_rate": 0.00026073720151566644,
+      "loss": 2.8739,
+      "step": 33165
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6196457147598267,
+      "learning_rate": 0.00026072193780350776,
+      "loss": 2.9828,
+      "step": 33166
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6265254616737366,
+      "learning_rate": 0.0002607066741948,
+      "loss": 2.8732,
+      "step": 33167
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.653234601020813,
+      "learning_rate": 0.0002606914106895833,
+      "loss": 2.9117,
+      "step": 33168
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6232619285583496,
+      "learning_rate": 0.0002606761472878979,
+      "loss": 3.2086,
+      "step": 33169
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5829308032989502,
+      "learning_rate": 0.00026066088398978384,
+      "loss": 3.2108,
+      "step": 33170
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5806586742401123,
+      "learning_rate": 0.00026064562079528144,
+      "loss": 3.133,
+      "step": 33171
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5950442552566528,
+      "learning_rate": 0.0002606303577044311,
+      "loss": 3.0696,
+      "step": 33172
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6098839640617371,
+      "learning_rate": 0.0002606150947172726,
+      "loss": 3.2242,
+      "step": 33173
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6669129729270935,
+      "learning_rate": 0.0002605998318338465,
+      "loss": 2.9275,
+      "step": 33174
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5990313291549683,
+      "learning_rate": 0.0002605845690541928,
+      "loss": 3.1519,
+      "step": 33175
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.614076554775238,
+      "learning_rate": 0.0002605693063783517,
+      "loss": 2.8938,
+      "step": 33176
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6242024302482605,
+      "learning_rate": 0.0002605540438063636,
+      "loss": 3.0608,
+      "step": 33177
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5861707329750061,
+      "learning_rate": 0.00026053878133826844,
+      "loss": 3.0707,
+      "step": 33178
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6213533282279968,
+      "learning_rate": 0.00026052351897410666,
+      "loss": 2.9031,
+      "step": 33179
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.560062825679779,
+      "learning_rate": 0.0002605082567139182,
+      "loss": 3.173,
+      "step": 33180
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6374415159225464,
+      "learning_rate": 0.00026049299455774353,
+      "loss": 2.9135,
+      "step": 33181
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6181164979934692,
+      "learning_rate": 0.0002604777325056225,
+      "loss": 3.0568,
+      "step": 33182
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6000651121139526,
+      "learning_rate": 0.0002604624705575957,
+      "loss": 3.0938,
+      "step": 33183
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6053292155265808,
+      "learning_rate": 0.0002604472087137031,
+      "loss": 3.0198,
+      "step": 33184
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6192270517349243,
+      "learning_rate": 0.0002604319469739849,
+      "loss": 2.9611,
+      "step": 33185
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.599314272403717,
+      "learning_rate": 0.00026041668533848135,
+      "loss": 3.0997,
+      "step": 33186
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.613339364528656,
+      "learning_rate": 0.00026040142380723264,
+      "loss": 3.1241,
+      "step": 33187
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6284672021865845,
+      "learning_rate": 0.0002603861623802789,
+      "loss": 2.9876,
+      "step": 33188
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6124971508979797,
+      "learning_rate": 0.0002603709010576605,
+      "loss": 2.9233,
+      "step": 33189
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6334354281425476,
+      "learning_rate": 0.0002603556398394175,
+      "loss": 3.187,
+      "step": 33190
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5786687135696411,
+      "learning_rate": 0.0002603403787255901,
+      "loss": 2.8804,
+      "step": 33191
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6184731721878052,
+      "learning_rate": 0.00026032511771621844,
+      "loss": 2.9722,
+      "step": 33192
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.591331422328949,
+      "learning_rate": 0.00026030985681134276,
+      "loss": 2.9679,
+      "step": 33193
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6321749091148376,
+      "learning_rate": 0.0002602945960110034,
+      "loss": 3.0021,
+      "step": 33194
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6103703379631042,
+      "learning_rate": 0.0002602793353152404,
+      "loss": 3.1078,
+      "step": 33195
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5984880924224854,
+      "learning_rate": 0.000260264074724094,
+      "loss": 3.0928,
+      "step": 33196
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5951915979385376,
+      "learning_rate": 0.0002602488142376043,
+      "loss": 2.8636,
+      "step": 33197
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.656234622001648,
+      "learning_rate": 0.00026023355385581165,
+      "loss": 3.2596,
+      "step": 33198
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6170872449874878,
+      "learning_rate": 0.0002602182935787562,
+      "loss": 3.0438,
+      "step": 33199
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6340166926383972,
+      "learning_rate": 0.00026020303340647797,
+      "loss": 3.1691,
+      "step": 33200
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6261439323425293,
+      "learning_rate": 0.00026018777333901744,
+      "loss": 2.905,
+      "step": 33201
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5833534598350525,
+      "learning_rate": 0.00026017251337641454,
+      "loss": 3.0735,
+      "step": 33202
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6300815343856812,
+      "learning_rate": 0.00026015725351870956,
+      "loss": 3.054,
+      "step": 33203
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6257278919219971,
+      "learning_rate": 0.00026014199376594283,
+      "loss": 3.032,
+      "step": 33204
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.607824444770813,
+      "learning_rate": 0.00026012673411815435,
+      "loss": 2.8231,
+      "step": 33205
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6163040399551392,
+      "learning_rate": 0.00026011147457538446,
+      "loss": 3.0407,
+      "step": 33206
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6370983719825745,
+      "learning_rate": 0.0002600962151376731,
+      "loss": 2.9238,
+      "step": 33207
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5754092335700989,
+      "learning_rate": 0.0002600809558050607,
+      "loss": 2.9017,
+      "step": 33208
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6198747754096985,
+      "learning_rate": 0.00026006569657758747,
+      "loss": 3.0832,
+      "step": 33209
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6094009876251221,
+      "learning_rate": 0.00026005043745529335,
+      "loss": 3.0943,
+      "step": 33210
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6221636533737183,
+      "learning_rate": 0.00026003517843821883,
+      "loss": 2.969,
+      "step": 33211
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6165490746498108,
+      "learning_rate": 0.0002600199195264039,
+      "loss": 3.0619,
+      "step": 33212
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6335442066192627,
+      "learning_rate": 0.0002600046607198888,
+      "loss": 3.0032,
+      "step": 33213
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.608069658279419,
+      "learning_rate": 0.00025998940201871375,
+      "loss": 2.9305,
+      "step": 33214
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5837422609329224,
+      "learning_rate": 0.00025997414342291883,
+      "loss": 3.1584,
+      "step": 33215
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6601833701133728,
+      "learning_rate": 0.00025995888493254447,
+      "loss": 3.0164,
+      "step": 33216
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6254077553749084,
+      "learning_rate": 0.0002599436265476306,
+      "loss": 2.9673,
+      "step": 33217
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6094150543212891,
+      "learning_rate": 0.0002599283682682175,
+      "loss": 3.217,
+      "step": 33218
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6002028584480286,
+      "learning_rate": 0.00025991311009434544,
+      "loss": 2.9381,
+      "step": 33219
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6100465655326843,
+      "learning_rate": 0.0002598978520260545,
+      "loss": 3.05,
+      "step": 33220
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6161542534828186,
+      "learning_rate": 0.0002598825940633849,
+      "loss": 3.0822,
+      "step": 33221
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6322788000106812,
+      "learning_rate": 0.0002598673362063768,
+      "loss": 3.0368,
+      "step": 33222
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.617686927318573,
+      "learning_rate": 0.0002598520784550704,
+      "loss": 3.1602,
+      "step": 33223
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6013792753219604,
+      "learning_rate": 0.00025983682080950597,
+      "loss": 3.0759,
+      "step": 33224
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5862300992012024,
+      "learning_rate": 0.0002598215632697235,
+      "loss": 3.2734,
+      "step": 33225
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5786439776420593,
+      "learning_rate": 0.00025980630583576347,
+      "loss": 2.8391,
+      "step": 33226
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6323323249816895,
+      "learning_rate": 0.0002597910485076658,
+      "loss": 3.1474,
+      "step": 33227
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5810719728469849,
+      "learning_rate": 0.0002597757912854707,
+      "loss": 3.1847,
+      "step": 33228
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6033740639686584,
+      "learning_rate": 0.00025976053416921856,
+      "loss": 3.0101,
+      "step": 33229
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6337068676948547,
+      "learning_rate": 0.00025974527715894936,
+      "loss": 2.9007,
+      "step": 33230
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6159489750862122,
+      "learning_rate": 0.0002597300202547034,
+      "loss": 3.0461,
+      "step": 33231
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5909284949302673,
+      "learning_rate": 0.0002597147634565207,
+      "loss": 3.0588,
+      "step": 33232
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5779708623886108,
+      "learning_rate": 0.0002596995067644416,
+      "loss": 3.0699,
+      "step": 33233
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.618440568447113,
+      "learning_rate": 0.0002596842501785063,
+      "loss": 3.1235,
+      "step": 33234
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6726293563842773,
+      "learning_rate": 0.0002596689936987549,
+      "loss": 2.8195,
+      "step": 33235
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6066452264785767,
+      "learning_rate": 0.0002596537373252277,
+      "loss": 3.019,
+      "step": 33236
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6022939085960388,
+      "learning_rate": 0.00025963848105796465,
+      "loss": 2.9119,
+      "step": 33237
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.590732991695404,
+      "learning_rate": 0.00025962322489700613,
+      "loss": 2.8292,
+      "step": 33238
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5825474262237549,
+      "learning_rate": 0.0002596079688423923,
+      "loss": 2.9209,
+      "step": 33239
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6330576539039612,
+      "learning_rate": 0.0002595927128941632,
+      "loss": 2.9113,
+      "step": 33240
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5879871845245361,
+      "learning_rate": 0.00025957745705235927,
+      "loss": 2.9726,
+      "step": 33241
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6738741993904114,
+      "learning_rate": 0.0002595622013170204,
+      "loss": 2.8329,
+      "step": 33242
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6530206203460693,
+      "learning_rate": 0.0002595469456881869,
+      "loss": 3.0572,
+      "step": 33243
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6073220372200012,
+      "learning_rate": 0.00025953169016589917,
+      "loss": 2.9772,
+      "step": 33244
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6424898505210876,
+      "learning_rate": 0.000259516434750197,
+      "loss": 3.0354,
+      "step": 33245
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6124472618103027,
+      "learning_rate": 0.00025950117944112085,
+      "loss": 3.0896,
+      "step": 33246
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6103495955467224,
+      "learning_rate": 0.00025948592423871063,
+      "loss": 3.0801,
+      "step": 33247
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6181695461273193,
+      "learning_rate": 0.0002594706691430068,
+      "loss": 3.0102,
+      "step": 33248
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6263618469238281,
+      "learning_rate": 0.0002594554141540495,
+      "loss": 2.9957,
+      "step": 33249
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6513804793357849,
+      "learning_rate": 0.00025944015927187867,
+      "loss": 3.1224,
+      "step": 33250
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6265279054641724,
+      "learning_rate": 0.0002594249044965348,
+      "loss": 2.7665,
+      "step": 33251
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.651552140712738,
+      "learning_rate": 0.0002594096498280579,
+      "loss": 2.906,
+      "step": 33252
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6316579580307007,
+      "learning_rate": 0.00025939439526648803,
+      "loss": 2.9716,
+      "step": 33253
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5931977033615112,
+      "learning_rate": 0.0002593791408118657,
+      "loss": 3.1162,
+      "step": 33254
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6142486333847046,
+      "learning_rate": 0.0002593638864642307,
+      "loss": 3.0778,
+      "step": 33255
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5705798268318176,
+      "learning_rate": 0.0002593486322236236,
+      "loss": 3.136,
+      "step": 33256
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6078622341156006,
+      "learning_rate": 0.0002593333780900843,
+      "loss": 3.0166,
+      "step": 33257
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5794697999954224,
+      "learning_rate": 0.00025931812406365293,
+      "loss": 2.999,
+      "step": 33258
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6560611724853516,
+      "learning_rate": 0.00025930287014437,
+      "loss": 2.8884,
+      "step": 33259
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6472424268722534,
+      "learning_rate": 0.00025928761633227534,
+      "loss": 3.0264,
+      "step": 33260
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6192428469657898,
+      "learning_rate": 0.0002592723626274094,
+      "loss": 2.9969,
+      "step": 33261
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6398415565490723,
+      "learning_rate": 0.0002592571090298121,
+      "loss": 3.0716,
+      "step": 33262
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5866214036941528,
+      "learning_rate": 0.0002592418555395238,
+      "loss": 3.1818,
+      "step": 33263
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6058687567710876,
+      "learning_rate": 0.00025922660215658457,
+      "loss": 3.0925,
+      "step": 33264
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5742605924606323,
+      "learning_rate": 0.00025921134888103454,
+      "loss": 2.9712,
+      "step": 33265
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6119689345359802,
+      "learning_rate": 0.0002591960957129141,
+      "loss": 2.8314,
+      "step": 33266
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.645915150642395,
+      "learning_rate": 0.00025918084265226326,
+      "loss": 2.8669,
+      "step": 33267
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.630989134311676,
+      "learning_rate": 0.0002591655896991222,
+      "loss": 3.1193,
+      "step": 33268
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5954846739768982,
+      "learning_rate": 0.00025915033685353106,
+      "loss": 2.9961,
+      "step": 33269
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5985564589500427,
+      "learning_rate": 0.00025913508411553015,
+      "loss": 3.0826,
+      "step": 33270
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.623395562171936,
+      "learning_rate": 0.0002591198314851596,
+      "loss": 3.1322,
+      "step": 33271
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5947928428649902,
+      "learning_rate": 0.00025910457896245943,
+      "loss": 3.0461,
+      "step": 33272
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.594084620475769,
+      "learning_rate": 0.00025908932654747005,
+      "loss": 2.9222,
+      "step": 33273
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6043190956115723,
+      "learning_rate": 0.00025907407424023146,
+      "loss": 3.0105,
+      "step": 33274
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6263400912284851,
+      "learning_rate": 0.0002590588220407838,
+      "loss": 3.1322,
+      "step": 33275
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5728640556335449,
+      "learning_rate": 0.0002590435699491675,
+      "loss": 2.9339,
+      "step": 33276
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6283183097839355,
+      "learning_rate": 0.00025902831796542234,
+      "loss": 3.0203,
+      "step": 33277
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6112415790557861,
+      "learning_rate": 0.0002590130660895889,
+      "loss": 2.7254,
+      "step": 33278
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5928776860237122,
+      "learning_rate": 0.0002589978143217071,
+      "loss": 3.0446,
+      "step": 33279
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6818798780441284,
+      "learning_rate": 0.00025898256266181704,
+      "loss": 2.9687,
+      "step": 33280
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5973649621009827,
+      "learning_rate": 0.00025896731110995917,
+      "loss": 3.2809,
+      "step": 33281
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6226446628570557,
+      "learning_rate": 0.00025895205966617343,
+      "loss": 3.058,
+      "step": 33282
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6349427103996277,
+      "learning_rate": 0.00025893680833050017,
+      "loss": 3.167,
+      "step": 33283
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6562497019767761,
+      "learning_rate": 0.0002589215571029793,
+      "loss": 3.0447,
+      "step": 33284
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5976384878158569,
+      "learning_rate": 0.0002589063059836512,
+      "loss": 3.0494,
+      "step": 33285
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.62012779712677,
+      "learning_rate": 0.00025889105497255605,
+      "loss": 3.164,
+      "step": 33286
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5701225996017456,
+      "learning_rate": 0.00025887580406973385,
+      "loss": 3.0588,
+      "step": 33287
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6170254945755005,
+      "learning_rate": 0.0002588605532752249,
+      "loss": 2.9552,
+      "step": 33288
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6369534730911255,
+      "learning_rate": 0.0002588453025890694,
+      "loss": 3.1661,
+      "step": 33289
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6616371273994446,
+      "learning_rate": 0.00025883005201130736,
+      "loss": 3.0857,
+      "step": 33290
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6203917264938354,
+      "learning_rate": 0.0002588148015419791,
+      "loss": 3.0843,
+      "step": 33291
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6321138739585876,
+      "learning_rate": 0.0002587995511811247,
+      "loss": 2.8944,
+      "step": 33292
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6112622618675232,
+      "learning_rate": 0.00025878430092878446,
+      "loss": 3.1176,
+      "step": 33293
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6042934060096741,
+      "learning_rate": 0.0002587690507849983,
+      "loss": 2.9723,
+      "step": 33294
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6640942692756653,
+      "learning_rate": 0.00025875380074980644,
+      "loss": 2.8202,
+      "step": 33295
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5991891622543335,
+      "learning_rate": 0.0002587385508232493,
+      "loss": 2.9682,
+      "step": 33296
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6276793479919434,
+      "learning_rate": 0.00025872330100536683,
+      "loss": 2.91,
+      "step": 33297
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5799841284751892,
+      "learning_rate": 0.0002587080512961993,
+      "loss": 2.9584,
+      "step": 33298
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6178253293037415,
+      "learning_rate": 0.0002586928016957867,
+      "loss": 3.0259,
+      "step": 33299
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6273776292800903,
+      "learning_rate": 0.00025867755220416934,
+      "loss": 3.0089,
+      "step": 33300
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6046527624130249,
+      "learning_rate": 0.0002586623028213875,
+      "loss": 3.0268,
+      "step": 33301
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6330516934394836,
+      "learning_rate": 0.000258647053547481,
+      "loss": 3.0706,
+      "step": 33302
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6148566603660583,
+      "learning_rate": 0.0002586318043824903,
+      "loss": 3.0584,
+      "step": 33303
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6204240918159485,
+      "learning_rate": 0.0002586165553264555,
+      "loss": 3.0304,
+      "step": 33304
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6241191029548645,
+      "learning_rate": 0.0002586013063794166,
+      "loss": 2.7981,
+      "step": 33305
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6470986604690552,
+      "learning_rate": 0.000258586057541414,
+      "loss": 3.0272,
+      "step": 33306
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6138694286346436,
+      "learning_rate": 0.00025857080881248776,
+      "loss": 2.8922,
+      "step": 33307
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5973373055458069,
+      "learning_rate": 0.00025855556019267805,
+      "loss": 2.907,
+      "step": 33308
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.599695086479187,
+      "learning_rate": 0.0002585403116820249,
+      "loss": 3.1989,
+      "step": 33309
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6345245838165283,
+      "learning_rate": 0.0002585250632805686,
+      "loss": 3.1087,
+      "step": 33310
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6359992623329163,
+      "learning_rate": 0.00025850981498834946,
+      "loss": 2.968,
+      "step": 33311
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5912989377975464,
+      "learning_rate": 0.0002584945668054073,
+      "loss": 2.9463,
+      "step": 33312
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6323924660682678,
+      "learning_rate": 0.00025847931873178266,
+      "loss": 3.3634,
+      "step": 33313
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6180119514465332,
+      "learning_rate": 0.00025846407076751535,
+      "loss": 3.0597,
+      "step": 33314
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5933579802513123,
+      "learning_rate": 0.0002584488229126457,
+      "loss": 3.1696,
+      "step": 33315
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6219573020935059,
+      "learning_rate": 0.0002584335751672139,
+      "loss": 2.9026,
+      "step": 33316
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6562425494194031,
+      "learning_rate": 0.00025841832753126,
+      "loss": 3.0916,
+      "step": 33317
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6309749484062195,
+      "learning_rate": 0.00025840308000482426,
+      "loss": 3.0988,
+      "step": 33318
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6028389930725098,
+      "learning_rate": 0.00025838783258794683,
+      "loss": 2.8867,
+      "step": 33319
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5963541269302368,
+      "learning_rate": 0.0002583725852806678,
+      "loss": 2.9588,
+      "step": 33320
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6047212481498718,
+      "learning_rate": 0.0002583573380830274,
+      "loss": 2.9557,
+      "step": 33321
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5715580582618713,
+      "learning_rate": 0.00025834209099506576,
+      "loss": 3.0006,
+      "step": 33322
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6066077351570129,
+      "learning_rate": 0.00025832684401682303,
+      "loss": 3.1405,
+      "step": 33323
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6018175482749939,
+      "learning_rate": 0.00025831159714833935,
+      "loss": 2.902,
+      "step": 33324
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6172803044319153,
+      "learning_rate": 0.0002582963503896549,
+      "loss": 2.888,
+      "step": 33325
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6167096495628357,
+      "learning_rate": 0.0002582811037408099,
+      "loss": 2.9556,
+      "step": 33326
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6224731802940369,
+      "learning_rate": 0.0002582658572018443,
+      "loss": 2.9202,
+      "step": 33327
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6047587990760803,
+      "learning_rate": 0.00025825061077279856,
+      "loss": 2.9544,
+      "step": 33328
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6331458687782288,
+      "learning_rate": 0.00025823536445371265,
+      "loss": 2.9418,
+      "step": 33329
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6286932826042175,
+      "learning_rate": 0.00025822011824462663,
+      "loss": 3.2196,
+      "step": 33330
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6198918223381042,
+      "learning_rate": 0.0002582048721455809,
+      "loss": 2.9849,
+      "step": 33331
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6311217546463013,
+      "learning_rate": 0.0002581896261566154,
+      "loss": 2.9469,
+      "step": 33332
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6329818964004517,
+      "learning_rate": 0.0002581743802777705,
+      "loss": 2.7676,
+      "step": 33333
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6387625932693481,
+      "learning_rate": 0.0002581591345090861,
+      "loss": 2.994,
+      "step": 33334
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6207517385482788,
+      "learning_rate": 0.0002581438888506025,
+      "loss": 3.1672,
+      "step": 33335
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5961693525314331,
+      "learning_rate": 0.00025812864330236005,
+      "loss": 2.8354,
+      "step": 33336
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6982114911079407,
+      "learning_rate": 0.00025811339786439846,
+      "loss": 3.005,
+      "step": 33337
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6450947523117065,
+      "learning_rate": 0.0002580981525367582,
+      "loss": 2.9924,
+      "step": 33338
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6073309779167175,
+      "learning_rate": 0.0002580829073194793,
+      "loss": 2.9336,
+      "step": 33339
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6175057291984558,
+      "learning_rate": 0.00025806766221260204,
+      "loss": 2.8445,
+      "step": 33340
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6069908738136292,
+      "learning_rate": 0.00025805241721616644,
+      "loss": 3.0606,
+      "step": 33341
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6620157957077026,
+      "learning_rate": 0.0002580371723302126,
+      "loss": 3.2177,
+      "step": 33342
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6427522897720337,
+      "learning_rate": 0.00025802192755478096,
+      "loss": 3.0998,
+      "step": 33343
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6400549411773682,
+      "learning_rate": 0.0002580066828899114,
+      "loss": 2.9967,
+      "step": 33344
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5848243832588196,
+      "learning_rate": 0.0002579914383356442,
+      "loss": 3.0598,
+      "step": 33345
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5872244238853455,
+      "learning_rate": 0.00025797619389201934,
+      "loss": 3.002,
+      "step": 33346
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5892220735549927,
+      "learning_rate": 0.0002579609495590772,
+      "loss": 3.0065,
+      "step": 33347
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5997381210327148,
+      "learning_rate": 0.0002579457053368578,
+      "loss": 2.9662,
+      "step": 33348
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.645177960395813,
+      "learning_rate": 0.00025793046122540126,
+      "loss": 3.0608,
+      "step": 33349
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6021146178245544,
+      "learning_rate": 0.00025791521722474795,
+      "loss": 2.8478,
+      "step": 33350
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6015944480895996,
+      "learning_rate": 0.0002578999733349378,
+      "loss": 3.0736,
+      "step": 33351
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6046991944313049,
+      "learning_rate": 0.00025788472955601087,
+      "loss": 3.1848,
+      "step": 33352
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6653998494148254,
+      "learning_rate": 0.00025786948588800765,
+      "loss": 3.1648,
+      "step": 33353
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.599899172782898,
+      "learning_rate": 0.00025785424233096797,
+      "loss": 3.2521,
+      "step": 33354
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6322702169418335,
+      "learning_rate": 0.0002578389988849322,
+      "loss": 2.8027,
+      "step": 33355
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5862153768539429,
+      "learning_rate": 0.0002578237555499403,
+      "loss": 2.9012,
+      "step": 33356
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6028646230697632,
+      "learning_rate": 0.0002578085123260325,
+      "loss": 3.1227,
+      "step": 33357
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6330054402351379,
+      "learning_rate": 0.00025779326921324905,
+      "loss": 2.8262,
+      "step": 33358
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.609877347946167,
+      "learning_rate": 0.00025777802621162994,
+      "loss": 2.9177,
+      "step": 33359
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5922818183898926,
+      "learning_rate": 0.0002577627833212155,
+      "loss": 2.9613,
+      "step": 33360
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5810437798500061,
+      "learning_rate": 0.0002577475405420456,
+      "loss": 3.0444,
+      "step": 33361
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6439691781997681,
+      "learning_rate": 0.0002577322978741606,
+      "loss": 2.9434,
+      "step": 33362
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6181412935256958,
+      "learning_rate": 0.0002577170553176007,
+      "loss": 3.0512,
+      "step": 33363
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5877893567085266,
+      "learning_rate": 0.0002577018128724057,
+      "loss": 3.1475,
+      "step": 33364
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6160455942153931,
+      "learning_rate": 0.0002576865705386162,
+      "loss": 2.9531,
+      "step": 33365
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5825440883636475,
+      "learning_rate": 0.00025767132831627207,
+      "loss": 3.0221,
+      "step": 33366
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5844451189041138,
+      "learning_rate": 0.0002576560862054134,
+      "loss": 3.0393,
+      "step": 33367
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6060708165168762,
+      "learning_rate": 0.00025764084420608064,
+      "loss": 3.1585,
+      "step": 33368
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.641826868057251,
+      "learning_rate": 0.00025762560231831365,
+      "loss": 3.0958,
+      "step": 33369
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6380789279937744,
+      "learning_rate": 0.00025761036054215274,
+      "loss": 2.9916,
+      "step": 33370
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.624642014503479,
+      "learning_rate": 0.0002575951188776378,
+      "loss": 2.8842,
+      "step": 33371
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5973908305168152,
+      "learning_rate": 0.00025757987732480934,
+      "loss": 2.6799,
+      "step": 33372
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6174383163452148,
+      "learning_rate": 0.0002575646358837073,
+      "loss": 3.2208,
+      "step": 33373
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5949251055717468,
+      "learning_rate": 0.00025754939455437174,
+      "loss": 2.959,
+      "step": 33374
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6241964101791382,
+      "learning_rate": 0.000257534153336843,
+      "loss": 3.0481,
+      "step": 33375
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6426224112510681,
+      "learning_rate": 0.0002575189122311611,
+      "loss": 2.8685,
+      "step": 33376
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6326531171798706,
+      "learning_rate": 0.0002575036712373661,
+      "loss": 2.9726,
+      "step": 33377
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6101643443107605,
+      "learning_rate": 0.0002574884303554984,
+      "loss": 2.9489,
+      "step": 33378
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6922885179519653,
+      "learning_rate": 0.0002574731895855979,
+      "loss": 2.8549,
+      "step": 33379
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5953141450881958,
+      "learning_rate": 0.0002574579489277049,
+      "loss": 3.1971,
+      "step": 33380
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6089001893997192,
+      "learning_rate": 0.0002574427083818595,
+      "loss": 3.1966,
+      "step": 33381
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6315808296203613,
+      "learning_rate": 0.00025742746794810167,
+      "loss": 2.9393,
+      "step": 33382
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6140810251235962,
+      "learning_rate": 0.00025741222762647187,
+      "loss": 3.0597,
+      "step": 33383
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6030791997909546,
+      "learning_rate": 0.00025739698741701,
+      "loss": 2.9677,
+      "step": 33384
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6273488998413086,
+      "learning_rate": 0.0002573817473197563,
+      "loss": 3.0209,
+      "step": 33385
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5986915230751038,
+      "learning_rate": 0.00025736650733475074,
+      "loss": 3.0809,
+      "step": 33386
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6623584032058716,
+      "learning_rate": 0.0002573512674620337,
+      "loss": 3.1532,
+      "step": 33387
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5963658690452576,
+      "learning_rate": 0.0002573360277016453,
+      "loss": 3.0506,
+      "step": 33388
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6195214986801147,
+      "learning_rate": 0.0002573207880536254,
+      "loss": 3.316,
+      "step": 33389
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6641665697097778,
+      "learning_rate": 0.0002573055485180145,
+      "loss": 3.044,
+      "step": 33390
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6132287383079529,
+      "learning_rate": 0.00025729030909485244,
+      "loss": 2.805,
+      "step": 33391
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6133543252944946,
+      "learning_rate": 0.00025727506978417947,
+      "loss": 2.8478,
+      "step": 33392
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.652361273765564,
+      "learning_rate": 0.0002572598305860359,
+      "loss": 3.0461,
+      "step": 33393
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5797291398048401,
+      "learning_rate": 0.00025724459150046157,
+      "loss": 3.3286,
+      "step": 33394
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6221191883087158,
+      "learning_rate": 0.0002572293525274969,
+      "loss": 2.8018,
+      "step": 33395
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6194936633110046,
+      "learning_rate": 0.0002572141136671817,
+      "loss": 2.847,
+      "step": 33396
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6495451331138611,
+      "learning_rate": 0.00025719887491955633,
+      "loss": 3.1063,
+      "step": 33397
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6331799030303955,
+      "learning_rate": 0.000257183636284661,
+      "loss": 3.1819,
+      "step": 33398
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5986855626106262,
+      "learning_rate": 0.0002571683977625357,
+      "loss": 3.0035,
+      "step": 33399
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6081706881523132,
+      "learning_rate": 0.0002571531593532206,
+      "loss": 3.0621,
+      "step": 33400
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.638114333152771,
+      "learning_rate": 0.00025713792105675577,
+      "loss": 2.9936,
+      "step": 33401
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6606442928314209,
+      "learning_rate": 0.00025712268287318143,
+      "loss": 3.0516,
+      "step": 33402
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.642713189125061,
+      "learning_rate": 0.00025710744480253776,
+      "loss": 2.943,
+      "step": 33403
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6082857847213745,
+      "learning_rate": 0.0002570922068448647,
+      "loss": 3.0018,
+      "step": 33404
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5871419310569763,
+      "learning_rate": 0.00025707696900020263,
+      "loss": 3.0449,
+      "step": 33405
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6455428600311279,
+      "learning_rate": 0.0002570617312685915,
+      "loss": 2.959,
+      "step": 33406
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5815387964248657,
+      "learning_rate": 0.00025704649365007143,
+      "loss": 2.7549,
+      "step": 33407
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6647642254829407,
+      "learning_rate": 0.0002570312561446828,
+      "loss": 3.2245,
+      "step": 33408
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6292085647583008,
+      "learning_rate": 0.00025701601875246545,
+      "loss": 2.8913,
+      "step": 33409
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6244023442268372,
+      "learning_rate": 0.00025700078147345974,
+      "loss": 2.9362,
+      "step": 33410
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.620573878288269,
+      "learning_rate": 0.0002569855443077055,
+      "loss": 2.8535,
+      "step": 33411
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.612984299659729,
+      "learning_rate": 0.00025697030725524325,
+      "loss": 2.9519,
+      "step": 33412
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6438848376274109,
+      "learning_rate": 0.0002569550703161129,
+      "loss": 2.8849,
+      "step": 33413
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5936183929443359,
+      "learning_rate": 0.0002569398334903545,
+      "loss": 2.7897,
+      "step": 33414
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.614172101020813,
+      "learning_rate": 0.0002569245967780084,
+      "loss": 3.0311,
+      "step": 33415
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6163328289985657,
+      "learning_rate": 0.0002569093601791146,
+      "loss": 3.0904,
+      "step": 33416
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6245632767677307,
+      "learning_rate": 0.0002568941236937133,
+      "loss": 2.9677,
+      "step": 33417
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6340919733047485,
+      "learning_rate": 0.00025687888732184446,
+      "loss": 3.2047,
+      "step": 33418
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.58950275182724,
+      "learning_rate": 0.0002568636510635483,
+      "loss": 2.9382,
+      "step": 33419
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5915112495422363,
+      "learning_rate": 0.00025684841491886517,
+      "loss": 2.7972,
+      "step": 33420
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6251772046089172,
+      "learning_rate": 0.0002568331788878349,
+      "loss": 2.9719,
+      "step": 33421
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5975916385650635,
+      "learning_rate": 0.0002568179429704978,
+      "loss": 2.9929,
+      "step": 33422
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6500588655471802,
+      "learning_rate": 0.0002568027071668938,
+      "loss": 2.9747,
+      "step": 33423
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6431196331977844,
+      "learning_rate": 0.00025678747147706325,
+      "loss": 2.9508,
+      "step": 33424
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6361487507820129,
+      "learning_rate": 0.0002567722359010462,
+      "loss": 2.9058,
+      "step": 33425
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6039143204689026,
+      "learning_rate": 0.00025675700043888264,
+      "loss": 2.9206,
+      "step": 33426
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6220351457595825,
+      "learning_rate": 0.000256741765090613,
+      "loss": 3.2211,
+      "step": 33427
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6292129158973694,
+      "learning_rate": 0.0002567265298562771,
+      "loss": 3.0292,
+      "step": 33428
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6125423312187195,
+      "learning_rate": 0.00025671129473591516,
+      "loss": 2.9991,
+      "step": 33429
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6337186098098755,
+      "learning_rate": 0.0002566960597295674,
+      "loss": 3.1525,
+      "step": 33430
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6597618460655212,
+      "learning_rate": 0.0002566808248372739,
+      "loss": 2.9801,
+      "step": 33431
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6206303238868713,
+      "learning_rate": 0.00025666559005907484,
+      "loss": 3.1365,
+      "step": 33432
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5949422717094421,
+      "learning_rate": 0.0002566503553950101,
+      "loss": 2.8709,
+      "step": 33433
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5526446104049683,
+      "learning_rate": 0.0002566351208451201,
+      "loss": 2.9796,
+      "step": 33434
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6217344999313354,
+      "learning_rate": 0.00025661988640944484,
+      "loss": 2.7793,
+      "step": 33435
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5836462378501892,
+      "learning_rate": 0.0002566046520880244,
+      "loss": 2.9644,
+      "step": 33436
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6020750999450684,
+      "learning_rate": 0.000256589417880899,
+      "loss": 3.1975,
+      "step": 33437
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6435813903808594,
+      "learning_rate": 0.00025657418378810873,
+      "loss": 2.9718,
+      "step": 33438
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6334481835365295,
+      "learning_rate": 0.00025655894980969356,
+      "loss": 2.8116,
+      "step": 33439
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6242516040802002,
+      "learning_rate": 0.0002565437159456939,
+      "loss": 2.8935,
+      "step": 33440
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6350675821304321,
+      "learning_rate": 0.0002565284821961497,
+      "loss": 2.9426,
+      "step": 33441
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6530657410621643,
+      "learning_rate": 0.00025651324856110116,
+      "loss": 3.0214,
+      "step": 33442
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6366276144981384,
+      "learning_rate": 0.00025649801504058826,
+      "loss": 2.9178,
+      "step": 33443
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.629357635974884,
+      "learning_rate": 0.0002564827816346512,
+      "loss": 3.052,
+      "step": 33444
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5792384743690491,
+      "learning_rate": 0.00025646754834333024,
+      "loss": 2.9482,
+      "step": 33445
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.615355908870697,
+      "learning_rate": 0.00025645231516666534,
+      "loss": 3.1283,
+      "step": 33446
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6263376474380493,
+      "learning_rate": 0.00025643708210469667,
+      "loss": 3.1832,
+      "step": 33447
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6375722289085388,
+      "learning_rate": 0.0002564218491574643,
+      "loss": 2.9832,
+      "step": 33448
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6010499596595764,
+      "learning_rate": 0.0002564066163250084,
+      "loss": 3.1208,
+      "step": 33449
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6063414216041565,
+      "learning_rate": 0.0002563913836073691,
+      "loss": 3.1275,
+      "step": 33450
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6149500012397766,
+      "learning_rate": 0.0002563761510045864,
+      "loss": 3.0182,
+      "step": 33451
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6116110682487488,
+      "learning_rate": 0.0002563609185167007,
+      "loss": 3.1028,
+      "step": 33452
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6712843775749207,
+      "learning_rate": 0.00025634568614375184,
+      "loss": 3.0418,
+      "step": 33453
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.597224235534668,
+      "learning_rate": 0.00025633045388577996,
+      "loss": 3.0967,
+      "step": 33454
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.609315037727356,
+      "learning_rate": 0.00025631522174282545,
+      "loss": 3.0672,
+      "step": 33455
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6235337853431702,
+      "learning_rate": 0.00025629998971492816,
+      "loss": 2.9888,
+      "step": 33456
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5854429602622986,
+      "learning_rate": 0.00025628475780212825,
+      "loss": 3.0874,
+      "step": 33457
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6504743695259094,
+      "learning_rate": 0.0002562695260044659,
+      "loss": 3.1094,
+      "step": 33458
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6945144534111023,
+      "learning_rate": 0.0002562542943219811,
+      "loss": 2.8553,
+      "step": 33459
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5794476866722107,
+      "learning_rate": 0.0002562390627547142,
+      "loss": 2.9821,
+      "step": 33460
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5813378691673279,
+      "learning_rate": 0.00025622383130270514,
+      "loss": 2.9139,
+      "step": 33461
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6076094508171082,
+      "learning_rate": 0.00025620859996599416,
+      "loss": 2.9726,
+      "step": 33462
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6358339190483093,
+      "learning_rate": 0.0002561933687446212,
+      "loss": 2.9508,
+      "step": 33463
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6296307444572449,
+      "learning_rate": 0.0002561781376386265,
+      "loss": 3.04,
+      "step": 33464
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6401169896125793,
+      "learning_rate": 0.00025616290664805025,
+      "loss": 3.2167,
+      "step": 33465
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6119227409362793,
+      "learning_rate": 0.00025614767577293227,
+      "loss": 3.2228,
+      "step": 33466
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6023973226547241,
+      "learning_rate": 0.00025613244501331304,
+      "loss": 3.0254,
+      "step": 33467
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5672397613525391,
+      "learning_rate": 0.0002561172143692325,
+      "loss": 3.1091,
+      "step": 33468
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6069930195808411,
+      "learning_rate": 0.0002561019838407306,
+      "loss": 2.9391,
+      "step": 33469
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6205309629440308,
+      "learning_rate": 0.0002560867534278478,
+      "loss": 2.9794,
+      "step": 33470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6706638336181641,
+      "learning_rate": 0.000256071523130624,
+      "loss": 2.9495,
+      "step": 33471
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6044138073921204,
+      "learning_rate": 0.00025605629294909944,
+      "loss": 3.0051,
+      "step": 33472
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6162787079811096,
+      "learning_rate": 0.00025604106288331394,
+      "loss": 2.7831,
+      "step": 33473
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6222827434539795,
+      "learning_rate": 0.00025602583293330794,
+      "loss": 2.9045,
+      "step": 33474
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6912936568260193,
+      "learning_rate": 0.0002560106030991215,
+      "loss": 3.111,
+      "step": 33475
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5932526588439941,
+      "learning_rate": 0.0002559953733807945,
+      "loss": 2.973,
+      "step": 33476
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6283918619155884,
+      "learning_rate": 0.00025598014377836736,
+      "loss": 2.8764,
+      "step": 33477
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6525346636772156,
+      "learning_rate": 0.00025596491429188,
+      "loss": 2.9557,
+      "step": 33478
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.595056414604187,
+      "learning_rate": 0.0002559496849213725,
+      "loss": 3.0048,
+      "step": 33479
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6286141276359558,
+      "learning_rate": 0.0002559344556668852,
+      "loss": 2.9229,
+      "step": 33480
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6390392184257507,
+      "learning_rate": 0.0002559192265284579,
+      "loss": 3.024,
+      "step": 33481
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6058946251869202,
+      "learning_rate": 0.00025590399750613106,
+      "loss": 2.8562,
+      "step": 33482
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7172413468360901,
+      "learning_rate": 0.0002558887685999445,
+      "loss": 2.993,
+      "step": 33483
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6925371289253235,
+      "learning_rate": 0.0002558735398099384,
+      "loss": 2.83,
+      "step": 33484
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6435196995735168,
+      "learning_rate": 0.000255858311136153,
+      "loss": 3.1041,
+      "step": 33485
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6237314939498901,
+      "learning_rate": 0.00025584308257862824,
+      "loss": 2.8744,
+      "step": 33486
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5538667440414429,
+      "learning_rate": 0.0002558278541374044,
+      "loss": 2.9134,
+      "step": 33487
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6051622629165649,
+      "learning_rate": 0.00025581262581252133,
+      "loss": 2.8513,
+      "step": 33488
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.61176997423172,
+      "learning_rate": 0.0002557973976040194,
+      "loss": 3.0499,
+      "step": 33489
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6022197604179382,
+      "learning_rate": 0.00025578216951193864,
+      "loss": 3.1049,
+      "step": 33490
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6228072047233582,
+      "learning_rate": 0.00025576694153631906,
+      "loss": 3.0617,
+      "step": 33491
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6194964051246643,
+      "learning_rate": 0.0002557517136772009,
+      "loss": 2.8614,
+      "step": 33492
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6054251790046692,
+      "learning_rate": 0.00025573648593462425,
+      "loss": 3.0247,
+      "step": 33493
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6623701453208923,
+      "learning_rate": 0.00025572125830862914,
+      "loss": 3.0102,
+      "step": 33494
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6244812607765198,
+      "learning_rate": 0.00025570603079925564,
+      "loss": 2.8992,
+      "step": 33495
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6170223951339722,
+      "learning_rate": 0.00025569080340654397,
+      "loss": 3.1205,
+      "step": 33496
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5932803750038147,
+      "learning_rate": 0.00025567557613053425,
+      "loss": 3.1165,
+      "step": 33497
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6121468544006348,
+      "learning_rate": 0.00025566034897126644,
+      "loss": 3.0744,
+      "step": 33498
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6298878788948059,
+      "learning_rate": 0.00025564512192878086,
+      "loss": 2.8936,
+      "step": 33499
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6140074729919434,
+      "learning_rate": 0.0002556298950031174,
+      "loss": 2.9988,
+      "step": 33500
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5965002179145813,
+      "learning_rate": 0.00025561466819431627,
+      "loss": 3.168,
+      "step": 33501
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.637391209602356,
+      "learning_rate": 0.0002555994415024176,
+      "loss": 3.2728,
+      "step": 33502
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6739944219589233,
+      "learning_rate": 0.00025558421492746135,
+      "loss": 3.0144,
+      "step": 33503
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6153771877288818,
+      "learning_rate": 0.00025556898846948786,
+      "loss": 3.0694,
+      "step": 33504
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5968106985092163,
+      "learning_rate": 0.00025555376212853704,
+      "loss": 2.9543,
+      "step": 33505
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.610713541507721,
+      "learning_rate": 0.000255538535904649,
+      "loss": 3.0342,
+      "step": 33506
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6176581382751465,
+      "learning_rate": 0.000255523309797864,
+      "loss": 3.1373,
+      "step": 33507
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6341189742088318,
+      "learning_rate": 0.000255508083808222,
+      "loss": 2.9842,
+      "step": 33508
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5881686806678772,
+      "learning_rate": 0.00025549285793576324,
+      "loss": 3.0403,
+      "step": 33509
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6181926727294922,
+      "learning_rate": 0.0002554776321805276,
+      "loss": 3.124,
+      "step": 33510
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6077942848205566,
+      "learning_rate": 0.00025546240654255535,
+      "loss": 2.9548,
+      "step": 33511
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6073379516601562,
+      "learning_rate": 0.00025544718102188656,
+      "loss": 2.8528,
+      "step": 33512
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6286836266517639,
+      "learning_rate": 0.00025543195561856125,
+      "loss": 2.9042,
+      "step": 33513
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6575766801834106,
+      "learning_rate": 0.0002554167303326197,
+      "loss": 3.1573,
+      "step": 33514
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5965009331703186,
+      "learning_rate": 0.0002554015051641019,
+      "loss": 3.0997,
+      "step": 33515
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6113033890724182,
+      "learning_rate": 0.0002553862801130478,
+      "loss": 2.9108,
+      "step": 33516
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6287041306495667,
+      "learning_rate": 0.00025537105517949786,
+      "loss": 2.875,
+      "step": 33517
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6469335556030273,
+      "learning_rate": 0.00025535583036349186,
+      "loss": 3.1125,
+      "step": 33518
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6197258830070496,
+      "learning_rate": 0.0002553406056650701,
+      "loss": 3.0977,
+      "step": 33519
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6109700202941895,
+      "learning_rate": 0.0002553253810842724,
+      "loss": 2.9279,
+      "step": 33520
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6416023373603821,
+      "learning_rate": 0.0002553101566211391,
+      "loss": 2.9842,
+      "step": 33521
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5842431783676147,
+      "learning_rate": 0.00025529493227571034,
+      "loss": 3.105,
+      "step": 33522
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6129705905914307,
+      "learning_rate": 0.0002552797080480261,
+      "loss": 2.9826,
+      "step": 33523
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5900617837905884,
+      "learning_rate": 0.00025526448393812655,
+      "loss": 3.2003,
+      "step": 33524
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5936287641525269,
+      "learning_rate": 0.00025524925994605164,
+      "loss": 3.2126,
+      "step": 33525
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5882431864738464,
+      "learning_rate": 0.0002552340360718416,
+      "loss": 2.9639,
+      "step": 33526
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5971298217773438,
+      "learning_rate": 0.00025521881231553654,
+      "loss": 3.096,
+      "step": 33527
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6381330490112305,
+      "learning_rate": 0.0002552035886771764,
+      "loss": 2.9462,
+      "step": 33528
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5903002023696899,
+      "learning_rate": 0.00025518836515680155,
+      "loss": 2.8188,
+      "step": 33529
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5994799733161926,
+      "learning_rate": 0.0002551731417544518,
+      "loss": 2.9601,
+      "step": 33530
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5971114039421082,
+      "learning_rate": 0.0002551579184701673,
+      "loss": 2.9885,
+      "step": 33531
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6384589076042175,
+      "learning_rate": 0.0002551426953039884,
+      "loss": 3.1067,
+      "step": 33532
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.591693639755249,
+      "learning_rate": 0.0002551274722559549,
+      "loss": 3.0607,
+      "step": 33533
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6068736910820007,
+      "learning_rate": 0.0002551122493261071,
+      "loss": 3.1471,
+      "step": 33534
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6241335868835449,
+      "learning_rate": 0.00025509702651448486,
+      "loss": 3.0201,
+      "step": 33535
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6447053551673889,
+      "learning_rate": 0.00025508180382112847,
+      "loss": 2.9881,
+      "step": 33536
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.600275456905365,
+      "learning_rate": 0.000255066581246078,
+      "loss": 3.0984,
+      "step": 33537
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5895094275474548,
+      "learning_rate": 0.00025505135878937344,
+      "loss": 3.1595,
+      "step": 33538
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6206028461456299,
+      "learning_rate": 0.000255036136451055,
+      "loss": 3.1758,
+      "step": 33539
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6393868923187256,
+      "learning_rate": 0.00025502091423116273,
+      "loss": 2.9527,
+      "step": 33540
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.672612190246582,
+      "learning_rate": 0.0002550056921297366,
+      "loss": 2.9975,
+      "step": 33541
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6368157863616943,
+      "learning_rate": 0.00025499047014681695,
+      "loss": 3.0863,
+      "step": 33542
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.621232271194458,
+      "learning_rate": 0.00025497524828244365,
+      "loss": 3.052,
+      "step": 33543
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6463088989257812,
+      "learning_rate": 0.00025496002653665697,
+      "loss": 2.6657,
+      "step": 33544
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6010912656784058,
+      "learning_rate": 0.00025494480490949686,
+      "loss": 2.9976,
+      "step": 33545
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6122451424598694,
+      "learning_rate": 0.00025492958340100344,
+      "loss": 2.9951,
+      "step": 33546
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6520957350730896,
+      "learning_rate": 0.00025491436201121687,
+      "loss": 2.9447,
+      "step": 33547
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6786126494407654,
+      "learning_rate": 0.00025489914074017716,
+      "loss": 3.0369,
+      "step": 33548
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5904258489608765,
+      "learning_rate": 0.0002548839195879245,
+      "loss": 2.9875,
+      "step": 33549
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6132463216781616,
+      "learning_rate": 0.0002548686985544988,
+      "loss": 2.8802,
+      "step": 33550
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6229579448699951,
+      "learning_rate": 0.00025485347763994033,
+      "loss": 3.1264,
+      "step": 33551
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5817505717277527,
+      "learning_rate": 0.0002548382568442892,
+      "loss": 2.837,
+      "step": 33552
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6309376955032349,
+      "learning_rate": 0.00025482303616758525,
+      "loss": 2.9213,
+      "step": 33553
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6208295822143555,
+      "learning_rate": 0.00025480781560986883,
+      "loss": 2.9553,
+      "step": 33554
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6129673719406128,
+      "learning_rate": 0.0002547925951711799,
+      "loss": 2.7496,
+      "step": 33555
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6032195091247559,
+      "learning_rate": 0.0002547773748515585,
+      "loss": 2.9992,
+      "step": 33556
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6073378920555115,
+      "learning_rate": 0.0002547621546510449,
+      "loss": 2.8163,
+      "step": 33557
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6146911382675171,
+      "learning_rate": 0.00025474693456967905,
+      "loss": 2.9754,
+      "step": 33558
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6286951899528503,
+      "learning_rate": 0.0002547317146075011,
+      "loss": 2.9883,
+      "step": 33559
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6064934134483337,
+      "learning_rate": 0.000254716494764551,
+      "loss": 3.0692,
+      "step": 33560
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6015488505363464,
+      "learning_rate": 0.0002547012750408689,
+      "loss": 3.0608,
+      "step": 33561
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6412321925163269,
+      "learning_rate": 0.0002546860554364951,
+      "loss": 3.3523,
+      "step": 33562
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6053003072738647,
+      "learning_rate": 0.0002546708359514694,
+      "loss": 2.8756,
+      "step": 33563
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6287083029747009,
+      "learning_rate": 0.00025465561658583204,
+      "loss": 3.047,
+      "step": 33564
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6067085266113281,
+      "learning_rate": 0.000254640397339623,
+      "loss": 3.1418,
+      "step": 33565
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6111128330230713,
+      "learning_rate": 0.00025462517821288246,
+      "loss": 3.0786,
+      "step": 33566
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6282280087471008,
+      "learning_rate": 0.00025460995920565055,
+      "loss": 3.0406,
+      "step": 33567
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6765866875648499,
+      "learning_rate": 0.0002545947403179671,
+      "loss": 3.1285,
+      "step": 33568
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6518777012825012,
+      "learning_rate": 0.0002545795215498725,
+      "loss": 3.0133,
+      "step": 33569
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6171415448188782,
+      "learning_rate": 0.0002545643029014067,
+      "loss": 3.1624,
+      "step": 33570
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6353912949562073,
+      "learning_rate": 0.0002545490843726098,
+      "loss": 2.9545,
+      "step": 33571
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6414846181869507,
+      "learning_rate": 0.0002545338659635217,
+      "loss": 3.1739,
+      "step": 33572
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6102538704872131,
+      "learning_rate": 0.00025451864767418275,
+      "loss": 3.1158,
+      "step": 33573
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6371604204177856,
+      "learning_rate": 0.00025450342950463303,
+      "loss": 3.0714,
+      "step": 33574
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5931726694107056,
+      "learning_rate": 0.0002544882114549123,
+      "loss": 3.1613,
+      "step": 33575
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.614202618598938,
+      "learning_rate": 0.00025447299352506105,
+      "loss": 2.8824,
+      "step": 33576
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6412155628204346,
+      "learning_rate": 0.0002544577757151191,
+      "loss": 2.6015,
+      "step": 33577
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6213472485542297,
+      "learning_rate": 0.0002544425580251265,
+      "loss": 3.0039,
+      "step": 33578
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6479688286781311,
+      "learning_rate": 0.0002544273404551236,
+      "loss": 2.9294,
+      "step": 33579
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6218825578689575,
+      "learning_rate": 0.00025441212300515026,
+      "loss": 3.1985,
+      "step": 33580
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6288421750068665,
+      "learning_rate": 0.0002543969056752467,
+      "loss": 2.974,
+      "step": 33581
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6092920303344727,
+      "learning_rate": 0.00025438168846545275,
+      "loss": 3.1613,
+      "step": 33582
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6967260241508484,
+      "learning_rate": 0.0002543664713758087,
+      "loss": 3.1719,
+      "step": 33583
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6147881150245667,
+      "learning_rate": 0.00025435125440635464,
+      "loss": 2.9597,
+      "step": 33584
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6140986680984497,
+      "learning_rate": 0.00025433603755713055,
+      "loss": 2.8277,
+      "step": 33585
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6095908880233765,
+      "learning_rate": 0.00025432082082817663,
+      "loss": 2.9332,
+      "step": 33586
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6159641146659851,
+      "learning_rate": 0.0002543056042195327,
+      "loss": 2.96,
+      "step": 33587
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5917350053787231,
+      "learning_rate": 0.00025429038773123916,
+      "loss": 2.9545,
+      "step": 33588
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6371214985847473,
+      "learning_rate": 0.000254275171363336,
+      "loss": 2.9997,
+      "step": 33589
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.609025239944458,
+      "learning_rate": 0.00025425995511586303,
+      "loss": 3.2217,
+      "step": 33590
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6314291954040527,
+      "learning_rate": 0.0002542447389888607,
+      "loss": 3.135,
+      "step": 33591
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6252784729003906,
+      "learning_rate": 0.00025422952298236886,
+      "loss": 3.072,
+      "step": 33592
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5745939612388611,
+      "learning_rate": 0.0002542143070964276,
+      "loss": 2.9007,
+      "step": 33593
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6394028663635254,
+      "learning_rate": 0.0002541990913310772,
+      "loss": 3.1977,
+      "step": 33594
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5794762969017029,
+      "learning_rate": 0.00025418387568635747,
+      "loss": 3.063,
+      "step": 33595
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6263272762298584,
+      "learning_rate": 0.00025416866016230865,
+      "loss": 2.9177,
+      "step": 33596
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6222317218780518,
+      "learning_rate": 0.0002541534447589706,
+      "loss": 2.9814,
+      "step": 33597
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6260561347007751,
+      "learning_rate": 0.0002541382294763837,
+      "loss": 3.0318,
+      "step": 33598
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6080615520477295,
+      "learning_rate": 0.00025412301431458785,
+      "loss": 3.0894,
+      "step": 33599
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.617087721824646,
+      "learning_rate": 0.0002541077992736231,
+      "loss": 2.9801,
+      "step": 33600
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6065132021903992,
+      "learning_rate": 0.0002540925843535297,
+      "loss": 3.0377,
+      "step": 33601
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5966345071792603,
+      "learning_rate": 0.0002540773695543475,
+      "loss": 2.8676,
+      "step": 33602
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5771703720092773,
+      "learning_rate": 0.0002540621548761166,
+      "loss": 2.9713,
+      "step": 33603
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6102816462516785,
+      "learning_rate": 0.0002540469403188773,
+      "loss": 3.0297,
+      "step": 33604
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5968355536460876,
+      "learning_rate": 0.00025403172588266934,
+      "loss": 2.9866,
+      "step": 33605
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6273283958435059,
+      "learning_rate": 0.00025401651156753316,
+      "loss": 3.0427,
+      "step": 33606
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6076398491859436,
+      "learning_rate": 0.00025400129737350853,
+      "loss": 2.7761,
+      "step": 33607
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6098259091377258,
+      "learning_rate": 0.0002539860833006356,
+      "loss": 3.0196,
+      "step": 33608
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5994309186935425,
+      "learning_rate": 0.0002539708693489546,
+      "loss": 3.0615,
+      "step": 33609
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6185238361358643,
+      "learning_rate": 0.00025395565551850535,
+      "loss": 2.9607,
+      "step": 33610
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6246052384376526,
+      "learning_rate": 0.0002539404418093282,
+      "loss": 2.9347,
+      "step": 33611
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.616183876991272,
+      "learning_rate": 0.00025392522822146284,
+      "loss": 2.8803,
+      "step": 33612
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5878469944000244,
+      "learning_rate": 0.00025391001475494964,
+      "loss": 3.1297,
+      "step": 33613
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6638484597206116,
+      "learning_rate": 0.0002538948014098287,
+      "loss": 2.9255,
+      "step": 33614
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5929738879203796,
+      "learning_rate": 0.00025387958818613986,
+      "loss": 3.0306,
+      "step": 33615
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6322368383407593,
+      "learning_rate": 0.00025386437508392336,
+      "loss": 3.0476,
+      "step": 33616
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6140685081481934,
+      "learning_rate": 0.00025384916210321925,
+      "loss": 3.4163,
+      "step": 33617
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.685403048992157,
+      "learning_rate": 0.00025383394924406743,
+      "loss": 2.907,
+      "step": 33618
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6040143966674805,
+      "learning_rate": 0.00025381873650650823,
+      "loss": 3.2015,
+      "step": 33619
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6561526656150818,
+      "learning_rate": 0.00025380352389058156,
+      "loss": 2.9761,
+      "step": 33620
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5854026675224304,
+      "learning_rate": 0.0002537883113963276,
+      "loss": 3.077,
+      "step": 33621
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5917167067527771,
+      "learning_rate": 0.0002537730990237862,
+      "loss": 2.972,
+      "step": 33622
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6419681906700134,
+      "learning_rate": 0.0002537578867729975,
+      "loss": 3.0175,
+      "step": 33623
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6021273732185364,
+      "learning_rate": 0.0002537426746440018,
+      "loss": 3.1386,
+      "step": 33624
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6152304410934448,
+      "learning_rate": 0.0002537274626368389,
+      "loss": 2.7386,
+      "step": 33625
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6640774011611938,
+      "learning_rate": 0.00025371225075154904,
+      "loss": 2.9014,
+      "step": 33626
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6067616939544678,
+      "learning_rate": 0.00025369703898817206,
+      "loss": 2.9653,
+      "step": 33627
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5870032906532288,
+      "learning_rate": 0.00025368182734674825,
+      "loss": 3.0218,
+      "step": 33628
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5886638164520264,
+      "learning_rate": 0.0002536666158273177,
+      "loss": 3.0415,
+      "step": 33629
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.640203595161438,
+      "learning_rate": 0.0002536514044299202,
+      "loss": 3.0191,
+      "step": 33630
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6461527943611145,
+      "learning_rate": 0.00025363619315459606,
+      "loss": 2.9528,
+      "step": 33631
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6292715668678284,
+      "learning_rate": 0.00025362098200138524,
+      "loss": 2.9719,
+      "step": 33632
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6264772415161133,
+      "learning_rate": 0.0002536057709703278,
+      "loss": 3.0541,
+      "step": 33633
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6099303960800171,
+      "learning_rate": 0.00025359056006146394,
+      "loss": 2.9152,
+      "step": 33634
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6216869950294495,
+      "learning_rate": 0.0002535753492748335,
+      "loss": 3.1676,
+      "step": 33635
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6083691716194153,
+      "learning_rate": 0.0002535601386104768,
+      "loss": 3.0394,
+      "step": 33636
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5741828083992004,
+      "learning_rate": 0.00025354492806843354,
+      "loss": 3.0786,
+      "step": 33637
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.634376049041748,
+      "learning_rate": 0.00025352971764874415,
+      "loss": 3.148,
+      "step": 33638
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6192229390144348,
+      "learning_rate": 0.00025351450735144854,
+      "loss": 3.0148,
+      "step": 33639
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6180740594863892,
+      "learning_rate": 0.0002534992971765867,
+      "loss": 3.1087,
+      "step": 33640
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5931870937347412,
+      "learning_rate": 0.0002534840871241988,
+      "loss": 3.1503,
+      "step": 33641
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5776154398918152,
+      "learning_rate": 0.00025346887719432485,
+      "loss": 3.0434,
+      "step": 33642
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.61803138256073,
+      "learning_rate": 0.0002534536673870049,
+      "loss": 3.0197,
+      "step": 33643
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6241032481193542,
+      "learning_rate": 0.0002534384577022791,
+      "loss": 3.0501,
+      "step": 33644
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6268068552017212,
+      "learning_rate": 0.00025342324814018734,
+      "loss": 2.8237,
+      "step": 33645
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5747461318969727,
+      "learning_rate": 0.0002534080387007699,
+      "loss": 3.3158,
+      "step": 33646
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6452773213386536,
+      "learning_rate": 0.0002533928293840667,
+      "loss": 3.1601,
+      "step": 33647
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6670741438865662,
+      "learning_rate": 0.0002533776201901177,
+      "loss": 3.0166,
+      "step": 33648
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.585372269153595,
+      "learning_rate": 0.0002533624111189632,
+      "loss": 3.0088,
+      "step": 33649
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5669026970863342,
+      "learning_rate": 0.0002533472021706431,
+      "loss": 2.8939,
+      "step": 33650
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6074293851852417,
+      "learning_rate": 0.0002533319933451975,
+      "loss": 3.2963,
+      "step": 33651
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5969696640968323,
+      "learning_rate": 0.0002533167846426664,
+      "loss": 2.9618,
+      "step": 33652
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6585623621940613,
+      "learning_rate": 0.00025330157606309,
+      "loss": 2.9734,
+      "step": 33653
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5954491496086121,
+      "learning_rate": 0.0002532863676065082,
+      "loss": 2.9167,
+      "step": 33654
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6330291628837585,
+      "learning_rate": 0.00025327115927296104,
+      "loss": 3.005,
+      "step": 33655
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6088626980781555,
+      "learning_rate": 0.0002532559510624888,
+      "loss": 3.2036,
+      "step": 33656
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6002334356307983,
+      "learning_rate": 0.0002532407429751313,
+      "loss": 3.0143,
+      "step": 33657
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6352154612541199,
+      "learning_rate": 0.0002532255350109287,
+      "loss": 3.0687,
+      "step": 33658
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5860254764556885,
+      "learning_rate": 0.000253210327169921,
+      "loss": 3.0877,
+      "step": 33659
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6036262512207031,
+      "learning_rate": 0.0002531951194521483,
+      "loss": 2.8906,
+      "step": 33660
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5720845460891724,
+      "learning_rate": 0.00025317991185765077,
+      "loss": 2.8707,
+      "step": 33661
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5837738513946533,
+      "learning_rate": 0.0002531647043864682,
+      "loss": 3.0003,
+      "step": 33662
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5987973213195801,
+      "learning_rate": 0.00025314949703864086,
+      "loss": 3.1969,
+      "step": 33663
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6149620413780212,
+      "learning_rate": 0.00025313428981420873,
+      "loss": 2.9412,
+      "step": 33664
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6203813552856445,
+      "learning_rate": 0.00025311908271321177,
+      "loss": 3.0063,
+      "step": 33665
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6189368367195129,
+      "learning_rate": 0.00025310387573569026,
+      "loss": 3.0654,
+      "step": 33666
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6118205189704895,
+      "learning_rate": 0.000253088668881684,
+      "loss": 2.9288,
+      "step": 33667
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6222481727600098,
+      "learning_rate": 0.0002530734621512332,
+      "loss": 3.0051,
+      "step": 33668
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.631308913230896,
+      "learning_rate": 0.0002530582555443779,
+      "loss": 3.1259,
+      "step": 33669
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6014788150787354,
+      "learning_rate": 0.000253043049061158,
+      "loss": 2.9009,
+      "step": 33670
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5971441268920898,
+      "learning_rate": 0.0002530278427016138,
+      "loss": 3.2945,
+      "step": 33671
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.606205940246582,
+      "learning_rate": 0.00025301263646578524,
+      "loss": 2.8732,
+      "step": 33672
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5801162123680115,
+      "learning_rate": 0.00025299743035371233,
+      "loss": 2.9203,
+      "step": 33673
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6619035005569458,
+      "learning_rate": 0.0002529822243654351,
+      "loss": 2.8384,
+      "step": 33674
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.614162027835846,
+      "learning_rate": 0.0002529670185009936,
+      "loss": 2.9881,
+      "step": 33675
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6145156621932983,
+      "learning_rate": 0.0002529518127604281,
+      "loss": 3.2938,
+      "step": 33676
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6185023784637451,
+      "learning_rate": 0.0002529366071437782,
+      "loss": 2.9915,
+      "step": 33677
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.640781581401825,
+      "learning_rate": 0.0002529214016510845,
+      "loss": 2.879,
+      "step": 33678
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.616585910320282,
+      "learning_rate": 0.00025290619628238656,
+      "loss": 2.9472,
+      "step": 33679
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6268201470375061,
+      "learning_rate": 0.0002528909910377247,
+      "loss": 3.0141,
+      "step": 33680
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6437908411026001,
+      "learning_rate": 0.00025287578591713894,
+      "loss": 2.9349,
+      "step": 33681
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6435546875,
+      "learning_rate": 0.00025286058092066925,
+      "loss": 2.9744,
+      "step": 33682
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6060557961463928,
+      "learning_rate": 0.0002528453760483558,
+      "loss": 3.0914,
+      "step": 33683
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5863337516784668,
+      "learning_rate": 0.00025283017130023843,
+      "loss": 3.1313,
+      "step": 33684
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6209391951560974,
+      "learning_rate": 0.0002528149666763573,
+      "loss": 3.1179,
+      "step": 33685
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6263371706008911,
+      "learning_rate": 0.0002527997621767526,
+      "loss": 3.0119,
+      "step": 33686
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6627563238143921,
+      "learning_rate": 0.0002527845578014642,
+      "loss": 2.9974,
+      "step": 33687
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7270352244377136,
+      "learning_rate": 0.0002527693535505322,
+      "loss": 3.0584,
+      "step": 33688
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5933760404586792,
+      "learning_rate": 0.0002527541494239965,
+      "loss": 2.9056,
+      "step": 33689
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5804953575134277,
+      "learning_rate": 0.00025273894542189736,
+      "loss": 3.1185,
+      "step": 33690
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6337752938270569,
+      "learning_rate": 0.0002527237415442748,
+      "loss": 2.8029,
+      "step": 33691
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5905575156211853,
+      "learning_rate": 0.00025270853779116873,
+      "loss": 2.9804,
+      "step": 33692
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8519971370697021,
+      "learning_rate": 0.00025269333416261936,
+      "loss": 3.0931,
+      "step": 33693
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5886369347572327,
+      "learning_rate": 0.00025267813065866655,
+      "loss": 3.0487,
+      "step": 33694
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5819816589355469,
+      "learning_rate": 0.00025266292727935043,
+      "loss": 2.9745,
+      "step": 33695
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5780503749847412,
+      "learning_rate": 0.00025264772402471117,
+      "loss": 3.2257,
+      "step": 33696
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6502557396888733,
+      "learning_rate": 0.0002526325208947886,
+      "loss": 3.067,
+      "step": 33697
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6687465310096741,
+      "learning_rate": 0.00025261731788962295,
+      "loss": 3.1206,
+      "step": 33698
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6016365885734558,
+      "learning_rate": 0.000252602115009254,
+      "loss": 3.0287,
+      "step": 33699
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6240675449371338,
+      "learning_rate": 0.000252586912253722,
+      "loss": 2.9064,
+      "step": 33700
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.582499086856842,
+      "learning_rate": 0.00025257170962306706,
+      "loss": 3.2111,
+      "step": 33701
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6195961236953735,
+      "learning_rate": 0.00025255650711732895,
+      "loss": 3.008,
+      "step": 33702
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5937759280204773,
+      "learning_rate": 0.00025254130473654807,
+      "loss": 2.8966,
+      "step": 33703
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6716198921203613,
+      "learning_rate": 0.00025252610248076415,
+      "loss": 2.8721,
+      "step": 33704
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6009177565574646,
+      "learning_rate": 0.0002525109003500172,
+      "loss": 3.0408,
+      "step": 33705
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6476081013679504,
+      "learning_rate": 0.0002524956983443476,
+      "loss": 2.9311,
+      "step": 33706
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6230241060256958,
+      "learning_rate": 0.00025248049646379506,
+      "loss": 2.6999,
+      "step": 33707
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6477628946304321,
+      "learning_rate": 0.0002524652947083999,
+      "loss": 3.2673,
+      "step": 33708
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6434798240661621,
+      "learning_rate": 0.0002524500930782019,
+      "loss": 2.8493,
+      "step": 33709
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6150158047676086,
+      "learning_rate": 0.00025243489157324113,
+      "loss": 2.9382,
+      "step": 33710
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6013590693473816,
+      "learning_rate": 0.0002524196901935579,
+      "loss": 2.9773,
+      "step": 33711
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6192983388900757,
+      "learning_rate": 0.00025240448893919194,
+      "loss": 3.2487,
+      "step": 33712
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6279726624488831,
+      "learning_rate": 0.00025238928781018345,
+      "loss": 2.9022,
+      "step": 33713
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6031069755554199,
+      "learning_rate": 0.00025237408680657227,
+      "loss": 3.0851,
+      "step": 33714
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6207053065299988,
+      "learning_rate": 0.0002523588859283987,
+      "loss": 3.0118,
+      "step": 33715
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5942187309265137,
+      "learning_rate": 0.00025234368517570267,
+      "loss": 3.1582,
+      "step": 33716
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6146659851074219,
+      "learning_rate": 0.0002523284845485241,
+      "loss": 3.0741,
+      "step": 33717
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6310451030731201,
+      "learning_rate": 0.00025231328404690323,
+      "loss": 3.027,
+      "step": 33718
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6505426168441772,
+      "learning_rate": 0.00025229808367087993,
+      "loss": 2.864,
+      "step": 33719
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6088349223136902,
+      "learning_rate": 0.00025228288342049425,
+      "loss": 2.9772,
+      "step": 33720
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6298375725746155,
+      "learning_rate": 0.0002522676832957864,
+      "loss": 3.1535,
+      "step": 33721
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6122135519981384,
+      "learning_rate": 0.00025225248329679626,
+      "loss": 3.1195,
+      "step": 33722
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5897640585899353,
+      "learning_rate": 0.00025223728342356385,
+      "loss": 3.0419,
+      "step": 33723
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6169995069503784,
+      "learning_rate": 0.0002522220836761292,
+      "loss": 3.0293,
+      "step": 33724
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6491032838821411,
+      "learning_rate": 0.00025220688405453237,
+      "loss": 2.9347,
+      "step": 33725
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5842862725257874,
+      "learning_rate": 0.0002521916845588135,
+      "loss": 3.1327,
+      "step": 33726
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6430948376655579,
+      "learning_rate": 0.0002521764851890125,
+      "loss": 2.9445,
+      "step": 33727
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6189959645271301,
+      "learning_rate": 0.00025216128594516955,
+      "loss": 3.0916,
+      "step": 33728
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.608040452003479,
+      "learning_rate": 0.0002521460868273244,
+      "loss": 3.0271,
+      "step": 33729
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6393671035766602,
+      "learning_rate": 0.00025213088783551733,
+      "loss": 2.8933,
+      "step": 33730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6282805800437927,
+      "learning_rate": 0.00025211568896978825,
+      "loss": 2.9764,
+      "step": 33731
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5948448181152344,
+      "learning_rate": 0.0002521004902301772,
+      "loss": 2.9167,
+      "step": 33732
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6147938370704651,
+      "learning_rate": 0.0002520852916167243,
+      "loss": 3.0297,
+      "step": 33733
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6409828662872314,
+      "learning_rate": 0.0002520700931294695,
+      "loss": 2.8486,
+      "step": 33734
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.718755841255188,
+      "learning_rate": 0.0002520548947684529,
+      "loss": 2.9996,
+      "step": 33735
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6078330874443054,
+      "learning_rate": 0.00025203969653371435,
+      "loss": 2.9222,
+      "step": 33736
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6095114350318909,
+      "learning_rate": 0.0002520244984252941,
+      "loss": 2.8218,
+      "step": 33737
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.670408308506012,
+      "learning_rate": 0.0002520093004432322,
+      "loss": 2.9352,
+      "step": 33738
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6077500581741333,
+      "learning_rate": 0.0002519941025875683,
+      "loss": 3.225,
+      "step": 33739
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6481585502624512,
+      "learning_rate": 0.0002519789048583429,
+      "loss": 2.9603,
+      "step": 33740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5873180627822876,
+      "learning_rate": 0.0002519637072555958,
+      "loss": 3.0317,
+      "step": 33741
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6104382872581482,
+      "learning_rate": 0.00025194850977936693,
+      "loss": 3.0169,
+      "step": 33742
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5977993607521057,
+      "learning_rate": 0.00025193331242969656,
+      "loss": 3.1041,
+      "step": 33743
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6217960119247437,
+      "learning_rate": 0.0002519181152066246,
+      "loss": 2.8854,
+      "step": 33744
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6322939991950989,
+      "learning_rate": 0.00025190291811019104,
+      "loss": 2.9709,
+      "step": 33745
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5825021862983704,
+      "learning_rate": 0.00025188772114043584,
+      "loss": 3.1036,
+      "step": 33746
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6149255633354187,
+      "learning_rate": 0.00025187252429739916,
+      "loss": 3.0629,
+      "step": 33747
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5964076519012451,
+      "learning_rate": 0.0002518573275811211,
+      "loss": 2.949,
+      "step": 33748
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5930870175361633,
+      "learning_rate": 0.0002518421309916415,
+      "loss": 3.2019,
+      "step": 33749
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6285166144371033,
+      "learning_rate": 0.0002518269345290005,
+      "loss": 3.1833,
+      "step": 33750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5890181064605713,
+      "learning_rate": 0.000251811738193238,
+      "loss": 3.284,
+      "step": 33751
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5849533677101135,
+      "learning_rate": 0.00025179654198439415,
+      "loss": 3.0216,
+      "step": 33752
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6618516445159912,
+      "learning_rate": 0.000251781345902509,
+      "loss": 3.1006,
+      "step": 33753
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6029493808746338,
+      "learning_rate": 0.0002517661499476224,
+      "loss": 2.8438,
+      "step": 33754
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.628272294998169,
+      "learning_rate": 0.00025175095411977453,
+      "loss": 3.0567,
+      "step": 33755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6236722469329834,
+      "learning_rate": 0.0002517357584190054,
+      "loss": 3.056,
+      "step": 33756
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5944138169288635,
+      "learning_rate": 0.00025172056284535483,
+      "loss": 3.1059,
+      "step": 33757
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5836665034294128,
+      "learning_rate": 0.0002517053673988632,
+      "loss": 3.1147,
+      "step": 33758
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6636527180671692,
+      "learning_rate": 0.00025169017207957026,
+      "loss": 2.7778,
+      "step": 33759
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6323641538619995,
+      "learning_rate": 0.00025167497688751615,
+      "loss": 3.0488,
+      "step": 33760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6302624940872192,
+      "learning_rate": 0.00025165978182274077,
+      "loss": 3.1198,
+      "step": 33761
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6282402873039246,
+      "learning_rate": 0.0002516445868852842,
+      "loss": 3.0372,
+      "step": 33762
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5997104048728943,
+      "learning_rate": 0.0002516293920751867,
+      "loss": 2.7741,
+      "step": 33763
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6254650950431824,
+      "learning_rate": 0.0002516141973924878,
+      "loss": 3.0331,
+      "step": 33764
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6326666474342346,
+      "learning_rate": 0.000251599002837228,
+      "loss": 2.9852,
+      "step": 33765
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5990265011787415,
+      "learning_rate": 0.00025158380840944696,
+      "loss": 2.9702,
+      "step": 33766
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6204058527946472,
+      "learning_rate": 0.00025156861410918487,
+      "loss": 3.0526,
+      "step": 33767
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.747573971748352,
+      "learning_rate": 0.0002515534199364818,
+      "loss": 2.9046,
+      "step": 33768
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6106170415878296,
+      "learning_rate": 0.00025153822589137764,
+      "loss": 3.1094,
+      "step": 33769
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6426050066947937,
+      "learning_rate": 0.0002515230319739126,
+      "loss": 2.9108,
+      "step": 33770
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6021739840507507,
+      "learning_rate": 0.0002515078381841264,
+      "loss": 3.1094,
+      "step": 33771
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5746902823448181,
+      "learning_rate": 0.00025149264452205923,
+      "loss": 3.0454,
+      "step": 33772
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5948019027709961,
+      "learning_rate": 0.00025147745098775116,
+      "loss": 2.9203,
+      "step": 33773
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6490957736968994,
+      "learning_rate": 0.0002514622575812422,
+      "loss": 2.8963,
+      "step": 33774
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6122204065322876,
+      "learning_rate": 0.0002514470643025723,
+      "loss": 2.8563,
+      "step": 33775
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5842893123626709,
+      "learning_rate": 0.00025143187115178134,
+      "loss": 2.8726,
+      "step": 33776
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.62315434217453,
+      "learning_rate": 0.0002514166781289096,
+      "loss": 2.9784,
+      "step": 33777
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6309679746627808,
+      "learning_rate": 0.00025140148523399697,
+      "loss": 3.1414,
+      "step": 33778
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6170636415481567,
+      "learning_rate": 0.0002513862924670834,
+      "loss": 3.0576,
+      "step": 33779
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6519901156425476,
+      "learning_rate": 0.00025137109982820914,
+      "loss": 3.0581,
+      "step": 33780
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6126569509506226,
+      "learning_rate": 0.00025135590731741394,
+      "loss": 3.0446,
+      "step": 33781
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5987186431884766,
+      "learning_rate": 0.00025134071493473784,
+      "loss": 2.9068,
+      "step": 33782
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6122428178787231,
+      "learning_rate": 0.0002513255226802211,
+      "loss": 2.9485,
+      "step": 33783
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6258891224861145,
+      "learning_rate": 0.00025131033055390345,
+      "loss": 3.0306,
+      "step": 33784
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6125542521476746,
+      "learning_rate": 0.00025129513855582514,
+      "loss": 3.1461,
+      "step": 33785
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.610899806022644,
+      "learning_rate": 0.00025127994668602593,
+      "loss": 2.9816,
+      "step": 33786
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6190662384033203,
+      "learning_rate": 0.0002512647549445459,
+      "loss": 2.9831,
+      "step": 33787
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6101117134094238,
+      "learning_rate": 0.00025124956333142534,
+      "loss": 3.1565,
+      "step": 33788
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6318680047988892,
+      "learning_rate": 0.0002512343718467039,
+      "loss": 3.1456,
+      "step": 33789
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6346719264984131,
+      "learning_rate": 0.0002512191804904219,
+      "loss": 3.0291,
+      "step": 33790
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6252193450927734,
+      "learning_rate": 0.00025120398926261895,
+      "loss": 2.9251,
+      "step": 33791
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6464138627052307,
+      "learning_rate": 0.00025118879816333545,
+      "loss": 3.0129,
+      "step": 33792
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6083313226699829,
+      "learning_rate": 0.0002511736071926113,
+      "loss": 2.9965,
+      "step": 33793
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5951755046844482,
+      "learning_rate": 0.0002511584163504863,
+      "loss": 2.9157,
+      "step": 33794
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.604433000087738,
+      "learning_rate": 0.00025114322563700084,
+      "loss": 3.1928,
+      "step": 33795
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6072720289230347,
+      "learning_rate": 0.0002511280350521946,
+      "loss": 2.8213,
+      "step": 33796
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.611731231212616,
+      "learning_rate": 0.0002511128445961077,
+      "loss": 3.159,
+      "step": 33797
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5745161175727844,
+      "learning_rate": 0.00025109765426878025,
+      "loss": 2.951,
+      "step": 33798
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5982853770256042,
+      "learning_rate": 0.0002510824640702521,
+      "loss": 3.1052,
+      "step": 33799
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6100951433181763,
+      "learning_rate": 0.00025106727400056334,
+      "loss": 3.0418,
+      "step": 33800
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6522631645202637,
+      "learning_rate": 0.00025105208405975393,
+      "loss": 2.8211,
+      "step": 33801
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6171745657920837,
+      "learning_rate": 0.0002510368942478639,
+      "loss": 2.8845,
+      "step": 33802
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6560838222503662,
+      "learning_rate": 0.0002510217045649334,
+      "loss": 3.0927,
+      "step": 33803
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6423225998878479,
+      "learning_rate": 0.0002510065150110021,
+      "loss": 2.9911,
+      "step": 33804
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6459085941314697,
+      "learning_rate": 0.00025099132558611037,
+      "loss": 2.8735,
+      "step": 33805
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6427082419395447,
+      "learning_rate": 0.000250976136290298,
+      "loss": 3.0119,
+      "step": 33806
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6143639087677002,
+      "learning_rate": 0.0002509609471236051,
+      "loss": 3.0391,
+      "step": 33807
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6215494871139526,
+      "learning_rate": 0.00025094575808607154,
+      "loss": 2.9713,
+      "step": 33808
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7413198351860046,
+      "learning_rate": 0.0002509305691777374,
+      "loss": 2.8757,
+      "step": 33809
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5840345621109009,
+      "learning_rate": 0.0002509153803986428,
+      "loss": 2.9908,
+      "step": 33810
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5827749967575073,
+      "learning_rate": 0.0002509001917488276,
+      "loss": 3.0006,
+      "step": 33811
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6079306602478027,
+      "learning_rate": 0.0002508850032283319,
+      "loss": 3.0942,
+      "step": 33812
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5841607451438904,
+      "learning_rate": 0.0002508698148371955,
+      "loss": 2.917,
+      "step": 33813
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6027677059173584,
+      "learning_rate": 0.0002508546265754587,
+      "loss": 3.0351,
+      "step": 33814
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5869815349578857,
+      "learning_rate": 0.0002508394384431613,
+      "loss": 2.9288,
+      "step": 33815
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6005563139915466,
+      "learning_rate": 0.00025082425044034335,
+      "loss": 3.2591,
+      "step": 33816
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6811476945877075,
+      "learning_rate": 0.00025080906256704486,
+      "loss": 2.9065,
+      "step": 33817
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5969555377960205,
+      "learning_rate": 0.0002507938748233059,
+      "loss": 3.1142,
+      "step": 33818
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6238395571708679,
+      "learning_rate": 0.0002507786872091662,
+      "loss": 2.9812,
+      "step": 33819
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5949825644493103,
+      "learning_rate": 0.0002507634997246662,
+      "loss": 2.9406,
+      "step": 33820
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5956006646156311,
+      "learning_rate": 0.00025074831236984556,
+      "loss": 3.0653,
+      "step": 33821
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6946335434913635,
+      "learning_rate": 0.00025073312514474445,
+      "loss": 2.8979,
+      "step": 33822
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6324100494384766,
+      "learning_rate": 0.0002507179380494027,
+      "loss": 3.1288,
+      "step": 33823
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6476609706878662,
+      "learning_rate": 0.00025070275108386047,
+      "loss": 3.1207,
+      "step": 33824
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6044153571128845,
+      "learning_rate": 0.0002506875642481578,
+      "loss": 3.0552,
+      "step": 33825
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6101298332214355,
+      "learning_rate": 0.00025067237754233443,
+      "loss": 3.069,
+      "step": 33826
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5707805752754211,
+      "learning_rate": 0.00025065719096643065,
+      "loss": 3.0449,
+      "step": 33827
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5993731617927551,
+      "learning_rate": 0.0002506420045204863,
+      "loss": 2.9649,
+      "step": 33828
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6095276474952698,
+      "learning_rate": 0.00025062681820454133,
+      "loss": 3.2008,
+      "step": 33829
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6098636984825134,
+      "learning_rate": 0.0002506116320186359,
+      "loss": 3.1267,
+      "step": 33830
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5973874926567078,
+      "learning_rate": 0.0002505964459628098,
+      "loss": 2.8534,
+      "step": 33831
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.657274067401886,
+      "learning_rate": 0.0002505812600371034,
+      "loss": 2.9407,
+      "step": 33832
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6024130582809448,
+      "learning_rate": 0.0002505660742415563,
+      "loss": 3.1447,
+      "step": 33833
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6005028486251831,
+      "learning_rate": 0.0002505508885762086,
+      "loss": 3.1241,
+      "step": 33834
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6737073063850403,
+      "learning_rate": 0.0002505357030411004,
+      "loss": 3.0029,
+      "step": 33835
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.603809118270874,
+      "learning_rate": 0.0002505205176362717,
+      "loss": 3.0558,
+      "step": 33836
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.593603789806366,
+      "learning_rate": 0.0002505053323617624,
+      "loss": 3.203,
+      "step": 33837
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.639643132686615,
+      "learning_rate": 0.0002504901472176124,
+      "loss": 3.0018,
+      "step": 33838
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6236515045166016,
+      "learning_rate": 0.00025047496220386193,
+      "loss": 3.0459,
+      "step": 33839
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6432270407676697,
+      "learning_rate": 0.00025045977732055094,
+      "loss": 2.9573,
+      "step": 33840
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6288147568702698,
+      "learning_rate": 0.00025044459256771923,
+      "loss": 2.9896,
+      "step": 33841
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6194936037063599,
+      "learning_rate": 0.0002504294079454071,
+      "loss": 3.1635,
+      "step": 33842
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6322032809257507,
+      "learning_rate": 0.00025041422345365425,
+      "loss": 3.0478,
+      "step": 33843
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6737121939659119,
+      "learning_rate": 0.0002503990390925007,
+      "loss": 3.3406,
+      "step": 33844
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6102167963981628,
+      "learning_rate": 0.0002503838548619867,
+      "loss": 3.1013,
+      "step": 33845
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6064083576202393,
+      "learning_rate": 0.00025036867076215204,
+      "loss": 3.0313,
+      "step": 33846
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6235889196395874,
+      "learning_rate": 0.0002503534867930368,
+      "loss": 3.0489,
+      "step": 33847
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6064871549606323,
+      "learning_rate": 0.0002503383029546808,
+      "loss": 2.9426,
+      "step": 33848
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6244485378265381,
+      "learning_rate": 0.0002503231192471241,
+      "loss": 2.994,
+      "step": 33849
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6445832252502441,
+      "learning_rate": 0.0002503079356704069,
+      "loss": 3.1693,
+      "step": 33850
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6180044412612915,
+      "learning_rate": 0.000250292752224569,
+      "loss": 3.1618,
+      "step": 33851
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6148596405982971,
+      "learning_rate": 0.00025027756890965043,
+      "loss": 2.8679,
+      "step": 33852
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6558221578598022,
+      "learning_rate": 0.00025026238572569104,
+      "loss": 3.1456,
+      "step": 33853
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5984779596328735,
+      "learning_rate": 0.00025024720267273113,
+      "loss": 3.0931,
+      "step": 33854
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5910894274711609,
+      "learning_rate": 0.00025023201975081044,
+      "loss": 2.8309,
+      "step": 33855
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6671299338340759,
+      "learning_rate": 0.000250216836959969,
+      "loss": 2.9913,
+      "step": 33856
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.617927610874176,
+      "learning_rate": 0.00025020165430024693,
+      "loss": 3.0802,
+      "step": 33857
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6495857238769531,
+      "learning_rate": 0.000250186471771684,
+      "loss": 2.8715,
+      "step": 33858
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6269500851631165,
+      "learning_rate": 0.00025017128937432034,
+      "loss": 2.7541,
+      "step": 33859
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6183517575263977,
+      "learning_rate": 0.00025015610710819603,
+      "loss": 3.0808,
+      "step": 33860
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6176064014434814,
+      "learning_rate": 0.00025014092497335083,
+      "loss": 3.178,
+      "step": 33861
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.609695315361023,
+      "learning_rate": 0.00025012574296982495,
+      "loss": 3.0251,
+      "step": 33862
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6329081654548645,
+      "learning_rate": 0.0002501105610976581,
+      "loss": 3.1642,
+      "step": 33863
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6028696894645691,
+      "learning_rate": 0.0002500953793568905,
+      "loss": 2.8275,
+      "step": 33864
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6294543147087097,
+      "learning_rate": 0.00025008019774756216,
+      "loss": 3.264,
+      "step": 33865
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5953629016876221,
+      "learning_rate": 0.0002500650162697128,
+      "loss": 3.0756,
+      "step": 33866
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6343550682067871,
+      "learning_rate": 0.00025004983492338273,
+      "loss": 3.1477,
+      "step": 33867
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.581670343875885,
+      "learning_rate": 0.0002500346537086117,
+      "loss": 3.1717,
+      "step": 33868
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6213530898094177,
+      "learning_rate": 0.00025001947262543974,
+      "loss": 2.9365,
+      "step": 33869
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6029473543167114,
+      "learning_rate": 0.000250004291673907,
+      "loss": 3.0754,
+      "step": 33870
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6134632229804993,
+      "learning_rate": 0.00024998911085405315,
+      "loss": 2.8977,
+      "step": 33871
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6177032589912415,
+      "learning_rate": 0.00024997393016591853,
+      "loss": 3.1839,
+      "step": 33872
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6571273803710938,
+      "learning_rate": 0.0002499587496095429,
+      "loss": 3.0477,
+      "step": 33873
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6338697075843811,
+      "learning_rate": 0.0002499435691849662,
+      "loss": 2.9806,
+      "step": 33874
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6008161902427673,
+      "learning_rate": 0.00024992838889222865,
+      "loss": 2.9013,
+      "step": 33875
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6442448496818542,
+      "learning_rate": 0.00024991320873137,
+      "loss": 2.9811,
+      "step": 33876
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6157088875770569,
+      "learning_rate": 0.0002498980287024304,
+      "loss": 3.0265,
+      "step": 33877
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5938071012496948,
+      "learning_rate": 0.0002498828488054496,
+      "loss": 2.9337,
+      "step": 33878
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6022875905036926,
+      "learning_rate": 0.00024986766904046785,
+      "loss": 3.1691,
+      "step": 33879
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6283571720123291,
+      "learning_rate": 0.00024985248940752504,
+      "loss": 3.0208,
+      "step": 33880
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6032127141952515,
+      "learning_rate": 0.000249837309906661,
+      "loss": 2.9129,
+      "step": 33881
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6097046136856079,
+      "learning_rate": 0.00024982213053791595,
+      "loss": 3.0705,
+      "step": 33882
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5926503539085388,
+      "learning_rate": 0.00024980695130132963,
+      "loss": 3.0113,
+      "step": 33883
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6158053874969482,
+      "learning_rate": 0.0002497917721969423,
+      "loss": 3.1628,
+      "step": 33884
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6430631279945374,
+      "learning_rate": 0.0002497765932247936,
+      "loss": 3.0387,
+      "step": 33885
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5978126525878906,
+      "learning_rate": 0.0002497614143849238,
+      "loss": 2.8056,
+      "step": 33886
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5920930504798889,
+      "learning_rate": 0.0002497462356773728,
+      "loss": 2.8876,
+      "step": 33887
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5528014302253723,
+      "learning_rate": 0.0002497310571021804,
+      "loss": 2.97,
+      "step": 33888
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5907645225524902,
+      "learning_rate": 0.00024971587865938686,
+      "loss": 3.0249,
+      "step": 33889
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5735505223274231,
+      "learning_rate": 0.00024970070034903197,
+      "loss": 2.9847,
+      "step": 33890
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6211549639701843,
+      "learning_rate": 0.0002496855221711557,
+      "loss": 3.1012,
+      "step": 33891
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6353756785392761,
+      "learning_rate": 0.0002496703441257982,
+      "loss": 3.062,
+      "step": 33892
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5956559181213379,
+      "learning_rate": 0.00024965516621299914,
+      "loss": 2.9391,
+      "step": 33893
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.598713755607605,
+      "learning_rate": 0.00024963998843279884,
+      "loss": 2.964,
+      "step": 33894
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6306696534156799,
+      "learning_rate": 0.0002496248107852371,
+      "loss": 2.8183,
+      "step": 33895
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6647626161575317,
+      "learning_rate": 0.00024960963327035383,
+      "loss": 2.9461,
+      "step": 33896
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6233382821083069,
+      "learning_rate": 0.00024959445588818924,
+      "loss": 3.1564,
+      "step": 33897
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6287595629692078,
+      "learning_rate": 0.00024957927863878304,
+      "loss": 2.9366,
+      "step": 33898
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6304962635040283,
+      "learning_rate": 0.0002495641015221754,
+      "loss": 3.1073,
+      "step": 33899
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6396434307098389,
+      "learning_rate": 0.00024954892453840606,
+      "loss": 2.9138,
+      "step": 33900
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5777924656867981,
+      "learning_rate": 0.0002495337476875152,
+      "loss": 2.9275,
+      "step": 33901
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6460705995559692,
+      "learning_rate": 0.00024951857096954286,
+      "loss": 2.9709,
+      "step": 33902
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.607254683971405,
+      "learning_rate": 0.0002495033943845287,
+      "loss": 3.1602,
+      "step": 33903
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5625812411308289,
+      "learning_rate": 0.0002494882179325131,
+      "loss": 3.0044,
+      "step": 33904
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.597536563873291,
+      "learning_rate": 0.00024947304161353564,
+      "loss": 3.281,
+      "step": 33905
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7313137650489807,
+      "learning_rate": 0.0002494578654276365,
+      "loss": 3.086,
+      "step": 33906
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6183013319969177,
+      "learning_rate": 0.00024944268937485565,
+      "loss": 2.8716,
+      "step": 33907
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6373917460441589,
+      "learning_rate": 0.000249427513455233,
+      "loss": 3.0631,
+      "step": 33908
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6329418420791626,
+      "learning_rate": 0.0002494123376688086,
+      "loss": 3.0995,
+      "step": 33909
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6428967118263245,
+      "learning_rate": 0.00024939716201562227,
+      "loss": 3.1429,
+      "step": 33910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5830432772636414,
+      "learning_rate": 0.00024938198649571407,
+      "loss": 2.8414,
+      "step": 33911
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6218639612197876,
+      "learning_rate": 0.0002493668111091241,
+      "loss": 3.0054,
+      "step": 33912
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6176791787147522,
+      "learning_rate": 0.0002493516358558921,
+      "loss": 2.9076,
+      "step": 33913
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6287698745727539,
+      "learning_rate": 0.00024933646073605826,
+      "loss": 3.1468,
+      "step": 33914
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6077466011047363,
+      "learning_rate": 0.00024932128574966226,
+      "loss": 3.0103,
+      "step": 33915
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6049577593803406,
+      "learning_rate": 0.00024930611089674434,
+      "loss": 2.6801,
+      "step": 33916
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5947195887565613,
+      "learning_rate": 0.0002492909361773444,
+      "loss": 3.1092,
+      "step": 33917
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6179590225219727,
+      "learning_rate": 0.0002492757615915023,
+      "loss": 3.0234,
+      "step": 33918
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6137421727180481,
+      "learning_rate": 0.0002492605871392582,
+      "loss": 3.0543,
+      "step": 33919
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6110509037971497,
+      "learning_rate": 0.00024924541282065184,
+      "loss": 3.0235,
+      "step": 33920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.604406476020813,
+      "learning_rate": 0.00024923023863572324,
+      "loss": 2.8201,
+      "step": 33921
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5563648343086243,
+      "learning_rate": 0.00024921506458451255,
+      "loss": 3.0609,
+      "step": 33922
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6176638007164001,
+      "learning_rate": 0.00024919989066705955,
+      "loss": 2.9723,
+      "step": 33923
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5700177550315857,
+      "learning_rate": 0.00024918471688340437,
+      "loss": 3.0213,
+      "step": 33924
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6631049513816833,
+      "learning_rate": 0.0002491695432335867,
+      "loss": 2.908,
+      "step": 33925
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6572071313858032,
+      "learning_rate": 0.0002491543697176467,
+      "loss": 3.0808,
+      "step": 33926
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5589262843132019,
+      "learning_rate": 0.0002491391963356244,
+      "loss": 2.9899,
+      "step": 33927
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6265024542808533,
+      "learning_rate": 0.00024912402308755957,
+      "loss": 2.9103,
+      "step": 33928
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5696996450424194,
+      "learning_rate": 0.00024910884997349236,
+      "loss": 3.0077,
+      "step": 33929
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6142954230308533,
+      "learning_rate": 0.0002490936769934626,
+      "loss": 3.0136,
+      "step": 33930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5850833654403687,
+      "learning_rate": 0.00024907850414751016,
+      "loss": 3.0001,
+      "step": 33931
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6385449171066284,
+      "learning_rate": 0.0002490633314356754,
+      "loss": 3.005,
+      "step": 33932
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6238567233085632,
+      "learning_rate": 0.0002490481588579978,
+      "loss": 3.1686,
+      "step": 33933
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6315261721611023,
+      "learning_rate": 0.0002490329864145177,
+      "loss": 3.2475,
+      "step": 33934
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6120766997337341,
+      "learning_rate": 0.0002490178141052748,
+      "loss": 2.8656,
+      "step": 33935
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6094756126403809,
+      "learning_rate": 0.0002490026419303092,
+      "loss": 2.949,
+      "step": 33936
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6236438751220703,
+      "learning_rate": 0.0002489874698896609,
+      "loss": 3.0454,
+      "step": 33937
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7023098468780518,
+      "learning_rate": 0.00024897229798336967,
+      "loss": 2.8311,
+      "step": 33938
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.609567403793335,
+      "learning_rate": 0.00024895712621147574,
+      "loss": 3.0063,
+      "step": 33939
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6077160835266113,
+      "learning_rate": 0.00024894195457401876,
+      "loss": 3.0588,
+      "step": 33940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6205511093139648,
+      "learning_rate": 0.00024892678307103885,
+      "loss": 3.0441,
+      "step": 33941
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.626202404499054,
+      "learning_rate": 0.00024891161170257606,
+      "loss": 3.137,
+      "step": 33942
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5801575779914856,
+      "learning_rate": 0.0002488964404686702,
+      "loss": 2.9521,
+      "step": 33943
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6224533319473267,
+      "learning_rate": 0.00024888126936936135,
+      "loss": 3.0051,
+      "step": 33944
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6122703552246094,
+      "learning_rate": 0.00024886609840468933,
+      "loss": 2.7792,
+      "step": 33945
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6122666001319885,
+      "learning_rate": 0.0002488509275746941,
+      "loss": 2.8985,
+      "step": 33946
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5584424734115601,
+      "learning_rate": 0.0002488357568794158,
+      "loss": 2.9583,
+      "step": 33947
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.601530909538269,
+      "learning_rate": 0.0002488205863188942,
+      "loss": 2.8557,
+      "step": 33948
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6443337798118591,
+      "learning_rate": 0.0002488054158931694,
+      "loss": 2.9457,
+      "step": 33949
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5796291828155518,
+      "learning_rate": 0.0002487902456022812,
+      "loss": 2.9709,
+      "step": 33950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5982805490493774,
+      "learning_rate": 0.0002487750754462696,
+      "loss": 3.193,
+      "step": 33951
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5929266810417175,
+      "learning_rate": 0.0002487599054251747,
+      "loss": 3.1062,
+      "step": 33952
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6260059475898743,
+      "learning_rate": 0.0002487447355390363,
+      "loss": 2.9633,
+      "step": 33953
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6307040452957153,
+      "learning_rate": 0.0002487295657878945,
+      "loss": 3.0025,
+      "step": 33954
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6263796091079712,
+      "learning_rate": 0.000248714396171789,
+      "loss": 2.8704,
+      "step": 33955
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5983632206916809,
+      "learning_rate": 0.00024869922669076,
+      "loss": 3.1199,
+      "step": 33956
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6320955157279968,
+      "learning_rate": 0.0002486840573448474,
+      "loss": 3.0689,
+      "step": 33957
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6292237639427185,
+      "learning_rate": 0.00024866888813409097,
+      "loss": 2.9389,
+      "step": 33958
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6104917526245117,
+      "learning_rate": 0.00024865371905853097,
+      "loss": 3.0677,
+      "step": 33959
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6262634992599487,
+      "learning_rate": 0.0002486385501182071,
+      "loss": 2.9518,
+      "step": 33960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6107706427574158,
+      "learning_rate": 0.0002486233813131595,
+      "loss": 2.9798,
+      "step": 33961
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6529039144515991,
+      "learning_rate": 0.0002486082126434279,
+      "loss": 3.1314,
+      "step": 33962
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6255234479904175,
+      "learning_rate": 0.0002485930441090524,
+      "loss": 3.112,
+      "step": 33963
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5849068760871887,
+      "learning_rate": 0.0002485778757100731,
+      "loss": 3.0832,
+      "step": 33964
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6300506591796875,
+      "learning_rate": 0.0002485627074465295,
+      "loss": 2.8947,
+      "step": 33965
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6072397828102112,
+      "learning_rate": 0.00024854753931846207,
+      "loss": 2.9836,
+      "step": 33966
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6096535325050354,
+      "learning_rate": 0.0002485323713259104,
+      "loss": 2.876,
+      "step": 33967
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6128242611885071,
+      "learning_rate": 0.0002485172034689145,
+      "loss": 3.1772,
+      "step": 33968
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6279151439666748,
+      "learning_rate": 0.00024850203574751454,
+      "loss": 2.9902,
+      "step": 33969
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7040567398071289,
+      "learning_rate": 0.0002484868681617502,
+      "loss": 3.0256,
+      "step": 33970
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6096146702766418,
+      "learning_rate": 0.0002484717007116616,
+      "loss": 2.8562,
+      "step": 33971
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5939738750457764,
+      "learning_rate": 0.0002484565333972885,
+      "loss": 3.1068,
+      "step": 33972
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6173325181007385,
+      "learning_rate": 0.00024844136621867096,
+      "loss": 3.0958,
+      "step": 33973
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6269416213035583,
+      "learning_rate": 0.0002484261991758491,
+      "loss": 3.0309,
+      "step": 33974
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6126830577850342,
+      "learning_rate": 0.00024841103226886263,
+      "loss": 3.0065,
+      "step": 33975
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6426466107368469,
+      "learning_rate": 0.00024839586549775164,
+      "loss": 3.0531,
+      "step": 33976
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6027688384056091,
+      "learning_rate": 0.0002483806988625559,
+      "loss": 3.0285,
+      "step": 33977
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6397801637649536,
+      "learning_rate": 0.0002483655323633155,
+      "loss": 3.0949,
+      "step": 33978
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.591986894607544,
+      "learning_rate": 0.0002483503660000704,
+      "loss": 3.1487,
+      "step": 33979
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6391310095787048,
+      "learning_rate": 0.00024833519977286044,
+      "loss": 3.0401,
+      "step": 33980
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6004000306129456,
+      "learning_rate": 0.00024832003368172566,
+      "loss": 3.0843,
+      "step": 33981
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5886471271514893,
+      "learning_rate": 0.00024830486772670595,
+      "loss": 3.0734,
+      "step": 33982
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.616348385810852,
+      "learning_rate": 0.0002482897019078412,
+      "loss": 2.848,
+      "step": 33983
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6533381342887878,
+      "learning_rate": 0.0002482745362251715,
+      "loss": 3.0357,
+      "step": 33984
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6193398237228394,
+      "learning_rate": 0.00024825937067873675,
+      "loss": 3.0845,
+      "step": 33985
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.604232668876648,
+      "learning_rate": 0.0002482442052685769,
+      "loss": 3.1351,
+      "step": 33986
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6174710392951965,
+      "learning_rate": 0.0002482290399947317,
+      "loss": 3.1375,
+      "step": 33987
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5886451005935669,
+      "learning_rate": 0.0002482138748572413,
+      "loss": 3.0255,
+      "step": 33988
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6079821586608887,
+      "learning_rate": 0.00024819870985614564,
+      "loss": 2.9344,
+      "step": 33989
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.669161856174469,
+      "learning_rate": 0.00024818354499148447,
+      "loss": 3.0288,
+      "step": 33990
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6441835761070251,
+      "learning_rate": 0.00024816838026329806,
+      "loss": 2.9344,
+      "step": 33991
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6017826795578003,
+      "learning_rate": 0.00024815321567162605,
+      "loss": 2.8813,
+      "step": 33992
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6216018199920654,
+      "learning_rate": 0.0002481380512165084,
+      "loss": 2.8692,
+      "step": 33993
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6077483892440796,
+      "learning_rate": 0.00024812288689798537,
+      "loss": 3.0153,
+      "step": 33994
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8035667538642883,
+      "learning_rate": 0.00024810772271609646,
+      "loss": 3.3342,
+      "step": 33995
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6364055275917053,
+      "learning_rate": 0.000248092558670882,
+      "loss": 2.9934,
+      "step": 33996
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5845009684562683,
+      "learning_rate": 0.00024807739476238167,
+      "loss": 3.0476,
+      "step": 33997
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6098141670227051,
+      "learning_rate": 0.0002480622309906354,
+      "loss": 2.9357,
+      "step": 33998
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6033996343612671,
+      "learning_rate": 0.00024804706735568336,
+      "loss": 3.0799,
+      "step": 33999
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6313714385032654,
+      "learning_rate": 0.0002480319038575653,
+      "loss": 3.0056,
+      "step": 34000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6305382251739502,
+      "learning_rate": 0.00024801674049632126,
+      "loss": 2.9569,
+      "step": 34001
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6073910593986511,
+      "learning_rate": 0.000248001577271991,
+      "loss": 2.9318,
+      "step": 34002
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.594529926776886,
+      "learning_rate": 0.0002479864141846146,
+      "loss": 2.9888,
+      "step": 34003
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6212973594665527,
+      "learning_rate": 0.0002479712512342321,
+      "loss": 2.9173,
+      "step": 34004
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6293397545814514,
+      "learning_rate": 0.0002479560884208831,
+      "loss": 2.8407,
+      "step": 34005
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6491313576698303,
+      "learning_rate": 0.00024794092574460797,
+      "loss": 2.8996,
+      "step": 34006
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.583442747592926,
+      "learning_rate": 0.0002479257632054463,
+      "loss": 2.928,
+      "step": 34007
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6074749827384949,
+      "learning_rate": 0.0002479106008034381,
+      "loss": 2.8657,
+      "step": 34008
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5930288434028625,
+      "learning_rate": 0.0002478954385386235,
+      "loss": 3.0198,
+      "step": 34009
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6107326745986938,
+      "learning_rate": 0.0002478802764110422,
+      "loss": 2.9917,
+      "step": 34010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5948980450630188,
+      "learning_rate": 0.0002478651144207342,
+      "loss": 3.0944,
+      "step": 34011
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6396520137786865,
+      "learning_rate": 0.0002478499525677395,
+      "loss": 2.9329,
+      "step": 34012
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6055231094360352,
+      "learning_rate": 0.0002478347908520979,
+      "loss": 3.2016,
+      "step": 34013
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6053661108016968,
+      "learning_rate": 0.00024781962927384953,
+      "loss": 2.8238,
+      "step": 34014
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6582406163215637,
+      "learning_rate": 0.00024780446783303415,
+      "loss": 2.8245,
+      "step": 34015
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6152645349502563,
+      "learning_rate": 0.00024778930652969187,
+      "loss": 3.0896,
+      "step": 34016
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6121913194656372,
+      "learning_rate": 0.00024777414536386236,
+      "loss": 3.1472,
+      "step": 34017
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6189271807670593,
+      "learning_rate": 0.00024775898433558575,
+      "loss": 3.2943,
+      "step": 34018
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6674231886863708,
+      "learning_rate": 0.000247743823444902,
+      "loss": 2.8963,
+      "step": 34019
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6085709929466248,
+      "learning_rate": 0.00024772866269185084,
+      "loss": 2.9968,
+      "step": 34020
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6044761538505554,
+      "learning_rate": 0.0002477135020764724,
+      "loss": 2.9718,
+      "step": 34021
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6364777088165283,
+      "learning_rate": 0.00024769834159880656,
+      "loss": 3.092,
+      "step": 34022
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6292357444763184,
+      "learning_rate": 0.00024768318125889305,
+      "loss": 2.865,
+      "step": 34023
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6410455703735352,
+      "learning_rate": 0.0002476680210567722,
+      "loss": 2.8619,
+      "step": 34024
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6323256492614746,
+      "learning_rate": 0.0002476528609924836,
+      "loss": 2.9672,
+      "step": 34025
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6153978705406189,
+      "learning_rate": 0.00024763770106606733,
+      "loss": 3.0129,
+      "step": 34026
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6172310709953308,
+      "learning_rate": 0.0002476225412775632,
+      "loss": 3.1617,
+      "step": 34027
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6111814379692078,
+      "learning_rate": 0.00024760738162701125,
+      "loss": 2.8775,
+      "step": 34028
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6040524840354919,
+      "learning_rate": 0.00024759222211445144,
+      "loss": 3.0851,
+      "step": 34029
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6083324551582336,
+      "learning_rate": 0.0002475770627399235,
+      "loss": 3.2039,
+      "step": 34030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6631869673728943,
+      "learning_rate": 0.0002475619035034676,
+      "loss": 3.0856,
+      "step": 34031
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.605511486530304,
+      "learning_rate": 0.00024754674440512355,
+      "loss": 3.0132,
+      "step": 34032
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6278396248817444,
+      "learning_rate": 0.0002475315854449312,
+      "loss": 2.935,
+      "step": 34033
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6231174468994141,
+      "learning_rate": 0.00024751642662293065,
+      "loss": 2.9998,
+      "step": 34034
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6460698246955872,
+      "learning_rate": 0.0002475012679391616,
+      "loss": 3.1249,
+      "step": 34035
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6108884811401367,
+      "learning_rate": 0.00024748610939366426,
+      "loss": 3.1124,
+      "step": 34036
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.67753005027771,
+      "learning_rate": 0.00024747095098647833,
+      "loss": 3.0618,
+      "step": 34037
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6046687364578247,
+      "learning_rate": 0.00024745579271764375,
+      "loss": 3.1446,
+      "step": 34038
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.701672375202179,
+      "learning_rate": 0.0002474406345872006,
+      "loss": 2.9236,
+      "step": 34039
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.60172438621521,
+      "learning_rate": 0.00024742547659518867,
+      "loss": 3.0169,
+      "step": 34040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5958285927772522,
+      "learning_rate": 0.000247410318741648,
+      "loss": 3.0876,
+      "step": 34041
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6142698526382446,
+      "learning_rate": 0.0002473951610266182,
+      "loss": 3.2667,
+      "step": 34042
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5789704322814941,
+      "learning_rate": 0.00024738000345013965,
+      "loss": 2.9274,
+      "step": 34043
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6565642356872559,
+      "learning_rate": 0.0002473648460122519,
+      "loss": 3.2293,
+      "step": 34044
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6168193221092224,
+      "learning_rate": 0.00024734968871299506,
+      "loss": 3.166,
+      "step": 34045
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6158468127250671,
+      "learning_rate": 0.00024733453155240906,
+      "loss": 3.1144,
+      "step": 34046
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6314502954483032,
+      "learning_rate": 0.0002473193745305337,
+      "loss": 2.9168,
+      "step": 34047
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6529750823974609,
+      "learning_rate": 0.00024730421764740905,
+      "loss": 3.116,
+      "step": 34048
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6014294028282166,
+      "learning_rate": 0.00024728906090307483,
+      "loss": 3.053,
+      "step": 34049
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6091155409812927,
+      "learning_rate": 0.0002472739042975712,
+      "loss": 2.8216,
+      "step": 34050
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.641144871711731,
+      "learning_rate": 0.00024725874783093795,
+      "loss": 3.0557,
+      "step": 34051
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6270335912704468,
+      "learning_rate": 0.00024724359150321486,
+      "loss": 3.1368,
+      "step": 34052
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5929427146911621,
+      "learning_rate": 0.00024722843531444215,
+      "loss": 3.031,
+      "step": 34053
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6024312376976013,
+      "learning_rate": 0.00024721327926465955,
+      "loss": 2.8996,
+      "step": 34054
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6244577765464783,
+      "learning_rate": 0.0002471981233539069,
+      "loss": 2.9659,
+      "step": 34055
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6653500199317932,
+      "learning_rate": 0.0002471829675822244,
+      "loss": 2.9293,
+      "step": 34056
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.583678662776947,
+      "learning_rate": 0.0002471678119496516,
+      "loss": 3.0174,
+      "step": 34057
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6141014099121094,
+      "learning_rate": 0.00024715265645622883,
+      "loss": 3.1254,
+      "step": 34058
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6307711601257324,
+      "learning_rate": 0.0002471375011019957,
+      "loss": 2.8053,
+      "step": 34059
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5954844951629639,
+      "learning_rate": 0.0002471223458869922,
+      "loss": 2.8284,
+      "step": 34060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6285943388938904,
+      "learning_rate": 0.00024710719081125833,
+      "loss": 2.9632,
+      "step": 34061
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6757904887199402,
+      "learning_rate": 0.00024709203587483384,
+      "loss": 2.9003,
+      "step": 34062
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5749251842498779,
+      "learning_rate": 0.0002470768810777589,
+      "loss": 3.0031,
+      "step": 34063
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6063864231109619,
+      "learning_rate": 0.0002470617264200731,
+      "loss": 2.9874,
+      "step": 34064
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6300905346870422,
+      "learning_rate": 0.0002470465719018166,
+      "loss": 2.8773,
+      "step": 34065
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6623172760009766,
+      "learning_rate": 0.00024703141752302927,
+      "loss": 2.9534,
+      "step": 34066
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6001306772232056,
+      "learning_rate": 0.0002470162632837509,
+      "loss": 2.909,
+      "step": 34067
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6355345249176025,
+      "learning_rate": 0.0002470011091840216,
+      "loss": 3.036,
+      "step": 34068
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6132904887199402,
+      "learning_rate": 0.00024698595522388114,
+      "loss": 3.0769,
+      "step": 34069
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6177525520324707,
+      "learning_rate": 0.0002469708014033694,
+      "loss": 2.792,
+      "step": 34070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6004658937454224,
+      "learning_rate": 0.0002469556477225265,
+      "loss": 3.1412,
+      "step": 34071
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6170461177825928,
+      "learning_rate": 0.0002469404941813921,
+      "loss": 3.1152,
+      "step": 34072
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5980595350265503,
+      "learning_rate": 0.0002469253407800063,
+      "loss": 3.123,
+      "step": 34073
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6601492762565613,
+      "learning_rate": 0.0002469101875184089,
+      "loss": 2.977,
+      "step": 34074
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6366380453109741,
+      "learning_rate": 0.00024689503439663974,
+      "loss": 2.9476,
+      "step": 34075
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.656537652015686,
+      "learning_rate": 0.000246879881414739,
+      "loss": 3.1397,
+      "step": 34076
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.629987359046936,
+      "learning_rate": 0.0002468647285727464,
+      "loss": 2.9363,
+      "step": 34077
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6459150314331055,
+      "learning_rate": 0.0002468495758707019,
+      "loss": 3.0619,
+      "step": 34078
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6273544430732727,
+      "learning_rate": 0.0002468344233086453,
+      "loss": 2.9486,
+      "step": 34079
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6106494665145874,
+      "learning_rate": 0.00024681927088661664,
+      "loss": 3.2629,
+      "step": 34080
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6219441890716553,
+      "learning_rate": 0.0002468041186046558,
+      "loss": 3.2484,
+      "step": 34081
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6537150144577026,
+      "learning_rate": 0.0002467889664628026,
+      "loss": 2.9886,
+      "step": 34082
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6156889200210571,
+      "learning_rate": 0.0002467738144610972,
+      "loss": 3.1849,
+      "step": 34083
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6050704717636108,
+      "learning_rate": 0.0002467586625995792,
+      "loss": 2.9982,
+      "step": 34084
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6336327195167542,
+      "learning_rate": 0.00024674351087828857,
+      "loss": 3.1285,
+      "step": 34085
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6327105164527893,
+      "learning_rate": 0.00024672835929726543,
+      "loss": 2.9639,
+      "step": 34086
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6417514681816101,
+      "learning_rate": 0.0002467132078565495,
+      "loss": 3.111,
+      "step": 34087
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6045467257499695,
+      "learning_rate": 0.0002466980565561807,
+      "loss": 3.2154,
+      "step": 34088
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9114331007003784,
+      "learning_rate": 0.0002466829053961989,
+      "loss": 3.1789,
+      "step": 34089
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5843703746795654,
+      "learning_rate": 0.0002466677543766441,
+      "loss": 3.1093,
+      "step": 34090
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6175855398178101,
+      "learning_rate": 0.0002466526034975563,
+      "loss": 3.1485,
+      "step": 34091
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6193850636482239,
+      "learning_rate": 0.0002466374527589751,
+      "loss": 3.0645,
+      "step": 34092
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6025353670120239,
+      "learning_rate": 0.0002466223021609407,
+      "loss": 2.9797,
+      "step": 34093
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6327268481254578,
+      "learning_rate": 0.00024660715170349285,
+      "loss": 3.2005,
+      "step": 34094
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6247721910476685,
+      "learning_rate": 0.0002465920013866714,
+      "loss": 3.0138,
+      "step": 34095
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6387418508529663,
+      "learning_rate": 0.0002465768512105165,
+      "loss": 3.1151,
+      "step": 34096
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6263685822486877,
+      "learning_rate": 0.00024656170117506773,
+      "loss": 2.9747,
+      "step": 34097
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6204245686531067,
+      "learning_rate": 0.00024654655128036526,
+      "loss": 2.9732,
+      "step": 34098
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5845150947570801,
+      "learning_rate": 0.0002465314015264489,
+      "loss": 3.0868,
+      "step": 34099
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6110867857933044,
+      "learning_rate": 0.0002465162519133584,
+      "loss": 2.8015,
+      "step": 34100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6433117389678955,
+      "learning_rate": 0.0002465011024411339,
+      "loss": 2.9421,
+      "step": 34101
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6174062490463257,
+      "learning_rate": 0.0002464859531098152,
+      "loss": 2.854,
+      "step": 34102
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5987910032272339,
+      "learning_rate": 0.0002464708039194423,
+      "loss": 3.1167,
+      "step": 34103
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6083313226699829,
+      "learning_rate": 0.0002464556548700548,
+      "loss": 3.0495,
+      "step": 34104
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.633455753326416,
+      "learning_rate": 0.0002464405059616929,
+      "loss": 2.8616,
+      "step": 34105
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6219099760055542,
+      "learning_rate": 0.00024642535719439644,
+      "loss": 3.1337,
+      "step": 34106
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6289411783218384,
+      "learning_rate": 0.0002464102085682052,
+      "loss": 3.1779,
+      "step": 34107
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6306152939796448,
+      "learning_rate": 0.00024639506008315926,
+      "loss": 3.1277,
+      "step": 34108
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6876551508903503,
+      "learning_rate": 0.0002463799117392983,
+      "loss": 3.0214,
+      "step": 34109
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.598209023475647,
+      "learning_rate": 0.00024636476353666234,
+      "loss": 3.0505,
+      "step": 34110
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.645162045955658,
+      "learning_rate": 0.0002463496154752914,
+      "loss": 3.0546,
+      "step": 34111
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6210871934890747,
+      "learning_rate": 0.0002463344675552251,
+      "loss": 3.1667,
+      "step": 34112
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5721740126609802,
+      "learning_rate": 0.0002463193197765036,
+      "loss": 2.9053,
+      "step": 34113
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5755153298377991,
+      "learning_rate": 0.00024630417213916656,
+      "loss": 2.8611,
+      "step": 34114
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.615121066570282,
+      "learning_rate": 0.000246289024643254,
+      "loss": 2.9583,
+      "step": 34115
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.641639232635498,
+      "learning_rate": 0.0002462738772888059,
+      "loss": 3.1712,
+      "step": 34116
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5973453521728516,
+      "learning_rate": 0.00024625873007586203,
+      "loss": 3.1632,
+      "step": 34117
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6583412289619446,
+      "learning_rate": 0.0002462435830044624,
+      "loss": 2.9733,
+      "step": 34118
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6712580323219299,
+      "learning_rate": 0.0002462284360746467,
+      "loss": 3.1107,
+      "step": 34119
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6605498194694519,
+      "learning_rate": 0.00024621328928645504,
+      "loss": 3.0804,
+      "step": 34120
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6001577377319336,
+      "learning_rate": 0.0002461981426399272,
+      "loss": 2.8523,
+      "step": 34121
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6218976974487305,
+      "learning_rate": 0.000246182996135103,
+      "loss": 3.1601,
+      "step": 34122
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6491854190826416,
+      "learning_rate": 0.0002461678497720226,
+      "loss": 2.9465,
+      "step": 34123
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6104328036308289,
+      "learning_rate": 0.0002461527035507256,
+      "loss": 3.0921,
+      "step": 34124
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.622700035572052,
+      "learning_rate": 0.00024613755747125216,
+      "loss": 2.9134,
+      "step": 34125
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6061636805534363,
+      "learning_rate": 0.00024612241153364184,
+      "loss": 3.047,
+      "step": 34126
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6331983804702759,
+      "learning_rate": 0.0002461072657379348,
+      "loss": 2.7637,
+      "step": 34127
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6341940760612488,
+      "learning_rate": 0.00024609212008417096,
+      "loss": 3.0107,
+      "step": 34128
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6093514561653137,
+      "learning_rate": 0.00024607697457238993,
+      "loss": 2.9378,
+      "step": 34129
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6078907251358032,
+      "learning_rate": 0.0002460618292026319,
+      "loss": 2.928,
+      "step": 34130
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6402571201324463,
+      "learning_rate": 0.00024604668397493657,
+      "loss": 3.0684,
+      "step": 34131
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6203433871269226,
+      "learning_rate": 0.0002460315388893438,
+      "loss": 2.8733,
+      "step": 34132
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6072234511375427,
+      "learning_rate": 0.0002460163939458937,
+      "loss": 3.0711,
+      "step": 34133
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6317029595375061,
+      "learning_rate": 0.000246001249144626,
+      "loss": 2.9225,
+      "step": 34134
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6035718321800232,
+      "learning_rate": 0.0002459861044855807,
+      "loss": 3.0206,
+      "step": 34135
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6024402976036072,
+      "learning_rate": 0.0002459709599687974,
+      "loss": 2.8322,
+      "step": 34136
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7386209964752197,
+      "learning_rate": 0.00024595581559431627,
+      "loss": 3.1102,
+      "step": 34137
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.646673321723938,
+      "learning_rate": 0.00024594067136217716,
+      "loss": 3.1028,
+      "step": 34138
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6153397560119629,
+      "learning_rate": 0.00024592552727242,
+      "loss": 3.1711,
+      "step": 34139
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6744325160980225,
+      "learning_rate": 0.0002459103833250845,
+      "loss": 2.8508,
+      "step": 34140
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5932391881942749,
+      "learning_rate": 0.0002458952395202106,
+      "loss": 2.9682,
+      "step": 34141
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.570640504360199,
+      "learning_rate": 0.0002458800958578383,
+      "loss": 3.0059,
+      "step": 34142
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6285528540611267,
+      "learning_rate": 0.00024586495233800743,
+      "loss": 3.0278,
+      "step": 34143
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6127492189407349,
+      "learning_rate": 0.0002458498089607577,
+      "loss": 3.117,
+      "step": 34144
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6140125393867493,
+      "learning_rate": 0.00024583466572612934,
+      "loss": 3.2226,
+      "step": 34145
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6389678716659546,
+      "learning_rate": 0.000245819522634162,
+      "loss": 3.1822,
+      "step": 34146
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6139385104179382,
+      "learning_rate": 0.0002458043796848955,
+      "loss": 2.8201,
+      "step": 34147
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6733065843582153,
+      "learning_rate": 0.00024578923687837,
+      "loss": 3.0222,
+      "step": 34148
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6101844906806946,
+      "learning_rate": 0.0002457740942146251,
+      "loss": 2.8743,
+      "step": 34149
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5869734883308411,
+      "learning_rate": 0.0002457589516937009,
+      "loss": 3.0927,
+      "step": 34150
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6247618198394775,
+      "learning_rate": 0.00024574380931563706,
+      "loss": 3.2443,
+      "step": 34151
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6441221833229065,
+      "learning_rate": 0.00024572866708047367,
+      "loss": 2.9533,
+      "step": 34152
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6270751357078552,
+      "learning_rate": 0.0002457135249882505,
+      "loss": 3.0383,
+      "step": 34153
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6131313443183899,
+      "learning_rate": 0.0002456983830390074,
+      "loss": 3.2256,
+      "step": 34154
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6170026063919067,
+      "learning_rate": 0.0002456832412327844,
+      "loss": 2.9657,
+      "step": 34155
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6400575637817383,
+      "learning_rate": 0.00024566809956962126,
+      "loss": 3.0925,
+      "step": 34156
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.618373453617096,
+      "learning_rate": 0.00024565295804955783,
+      "loss": 2.9742,
+      "step": 34157
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.67595374584198,
+      "learning_rate": 0.0002456378166726342,
+      "loss": 3.2046,
+      "step": 34158
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6886033415794373,
+      "learning_rate": 0.0002456226754388899,
+      "loss": 2.941,
+      "step": 34159
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.62932288646698,
+      "learning_rate": 0.00024560753434836515,
+      "loss": 2.9249,
+      "step": 34160
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6120386123657227,
+      "learning_rate": 0.00024559239340109966,
+      "loss": 3.2161,
+      "step": 34161
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6127865314483643,
+      "learning_rate": 0.00024557725259713323,
+      "loss": 2.8264,
+      "step": 34162
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6054245829582214,
+      "learning_rate": 0.00024556211193650597,
+      "loss": 3.1627,
+      "step": 34163
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6147023439407349,
+      "learning_rate": 0.0002455469714192576,
+      "loss": 3.0502,
+      "step": 34164
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6629253029823303,
+      "learning_rate": 0.0002455318310454281,
+      "loss": 2.9995,
+      "step": 34165
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6122319102287292,
+      "learning_rate": 0.0002455166908150571,
+      "loss": 3.0473,
+      "step": 34166
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6370877027511597,
+      "learning_rate": 0.00024550155072818476,
+      "loss": 2.8365,
+      "step": 34167
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6131877899169922,
+      "learning_rate": 0.0002454864107848509,
+      "loss": 2.9348,
+      "step": 34168
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6444746255874634,
+      "learning_rate": 0.0002454712709850952,
+      "loss": 3.127,
+      "step": 34169
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6460678577423096,
+      "learning_rate": 0.0002454561313289578,
+      "loss": 2.8952,
+      "step": 34170
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6373785734176636,
+      "learning_rate": 0.00024544099181647843,
+      "loss": 3.0078,
+      "step": 34171
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6953406929969788,
+      "learning_rate": 0.00024542585244769687,
+      "loss": 3.0514,
+      "step": 34172
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6235325336456299,
+      "learning_rate": 0.00024541071322265324,
+      "loss": 3.0022,
+      "step": 34173
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6241675019264221,
+      "learning_rate": 0.00024539557414138724,
+      "loss": 3.1864,
+      "step": 34174
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6402691602706909,
+      "learning_rate": 0.00024538043520393887,
+      "loss": 2.88,
+      "step": 34175
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6314650774002075,
+      "learning_rate": 0.0002453652964103478,
+      "loss": 3.0372,
+      "step": 34176
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6837327480316162,
+      "learning_rate": 0.000245350157760654,
+      "loss": 2.8986,
+      "step": 34177
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6558158993721008,
+      "learning_rate": 0.0002453350192548975,
+      "loss": 3.008,
+      "step": 34178
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6023430228233337,
+      "learning_rate": 0.00024531988089311797,
+      "loss": 2.9874,
+      "step": 34179
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6416671276092529,
+      "learning_rate": 0.00024530474267535547,
+      "loss": 2.7857,
+      "step": 34180
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6319637298583984,
+      "learning_rate": 0.00024528960460164955,
+      "loss": 2.914,
+      "step": 34181
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6269749999046326,
+      "learning_rate": 0.0002452744666720404,
+      "loss": 3.0919,
+      "step": 34182
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6353631615638733,
+      "learning_rate": 0.0002452593288865678,
+      "loss": 2.9123,
+      "step": 34183
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6360670924186707,
+      "learning_rate": 0.00024524419124527145,
+      "loss": 2.8882,
+      "step": 34184
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6149998307228088,
+      "learning_rate": 0.0002452290537481915,
+      "loss": 3.2487,
+      "step": 34185
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5942299365997314,
+      "learning_rate": 0.0002452139163953676,
+      "loss": 3.0568,
+      "step": 34186
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5962042808532715,
+      "learning_rate": 0.00024519877918683974,
+      "loss": 3.0002,
+      "step": 34187
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6600490212440491,
+      "learning_rate": 0.00024518364212264775,
+      "loss": 3.2226,
+      "step": 34188
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6743106842041016,
+      "learning_rate": 0.00024516850520283154,
+      "loss": 3.1296,
+      "step": 34189
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6211302876472473,
+      "learning_rate": 0.0002451533684274309,
+      "loss": 3.1217,
+      "step": 34190
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7367480397224426,
+      "learning_rate": 0.00024513823179648565,
+      "loss": 2.9123,
+      "step": 34191
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6232231855392456,
+      "learning_rate": 0.00024512309531003583,
+      "loss": 3.0276,
+      "step": 34192
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6443117260932922,
+      "learning_rate": 0.0002451079589681212,
+      "loss": 3.2087,
+      "step": 34193
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6081936955451965,
+      "learning_rate": 0.0002450928227707816,
+      "loss": 2.993,
+      "step": 34194
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.600371241569519,
+      "learning_rate": 0.00024507768671805705,
+      "loss": 3.051,
+      "step": 34195
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6255161762237549,
+      "learning_rate": 0.0002450625508099872,
+      "loss": 3.0649,
+      "step": 34196
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.632082462310791,
+      "learning_rate": 0.00024504741504661207,
+      "loss": 2.9378,
+      "step": 34197
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5944221615791321,
+      "learning_rate": 0.00024503227942797144,
+      "loss": 3.1465,
+      "step": 34198
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5801829695701599,
+      "learning_rate": 0.00024501714395410514,
+      "loss": 2.8567,
+      "step": 34199
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5647958517074585,
+      "learning_rate": 0.00024500200862505317,
+      "loss": 3.0342,
+      "step": 34200
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6428757309913635,
+      "learning_rate": 0.0002449868734408554,
+      "loss": 3.0151,
+      "step": 34201
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5983733534812927,
+      "learning_rate": 0.00024497173840155157,
+      "loss": 2.9189,
+      "step": 34202
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6872225999832153,
+      "learning_rate": 0.0002449566035071815,
+      "loss": 2.9386,
+      "step": 34203
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6514347195625305,
+      "learning_rate": 0.00024494146875778523,
+      "loss": 2.9179,
+      "step": 34204
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6127699613571167,
+      "learning_rate": 0.00024492633415340256,
+      "loss": 2.8151,
+      "step": 34205
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6513962149620056,
+      "learning_rate": 0.0002449111996940732,
+      "loss": 2.9692,
+      "step": 34206
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6389971971511841,
+      "learning_rate": 0.00024489606537983727,
+      "loss": 2.9922,
+      "step": 34207
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6056563854217529,
+      "learning_rate": 0.00024488093121073445,
+      "loss": 2.9475,
+      "step": 34208
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6097168326377869,
+      "learning_rate": 0.0002448657971868046,
+      "loss": 2.9191,
+      "step": 34209
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.640527069568634,
+      "learning_rate": 0.0002448506633080877,
+      "loss": 3.0529,
+      "step": 34210
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6365799307823181,
+      "learning_rate": 0.00024483552957462357,
+      "loss": 2.9739,
+      "step": 34211
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6156507730484009,
+      "learning_rate": 0.00024482039598645205,
+      "loss": 2.9809,
+      "step": 34212
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6057137846946716,
+      "learning_rate": 0.00024480526254361284,
+      "loss": 3.0669,
+      "step": 34213
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6164252758026123,
+      "learning_rate": 0.000244790129246146,
+      "loss": 3.0025,
+      "step": 34214
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6119879484176636,
+      "learning_rate": 0.00024477499609409144,
+      "loss": 3.0289,
+      "step": 34215
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6054874062538147,
+      "learning_rate": 0.00024475986308748873,
+      "loss": 3.1761,
+      "step": 34216
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6626514196395874,
+      "learning_rate": 0.0002447447302263781,
+      "loss": 3.0249,
+      "step": 34217
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6329224109649658,
+      "learning_rate": 0.0002447295975107991,
+      "loss": 2.9464,
+      "step": 34218
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6414870619773865,
+      "learning_rate": 0.0002447144649407917,
+      "loss": 3.1747,
+      "step": 34219
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6235871315002441,
+      "learning_rate": 0.00024469933251639584,
+      "loss": 3.0888,
+      "step": 34220
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6273425221443176,
+      "learning_rate": 0.00024468420023765117,
+      "loss": 3.0888,
+      "step": 34221
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6260068416595459,
+      "learning_rate": 0.00024466906810459785,
+      "loss": 3.174,
+      "step": 34222
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.721221923828125,
+      "learning_rate": 0.00024465393611727544,
+      "loss": 2.8587,
+      "step": 34223
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5994982719421387,
+      "learning_rate": 0.0002446388042757238,
+      "loss": 2.8307,
+      "step": 34224
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5987682342529297,
+      "learning_rate": 0.0002446236725799831,
+      "loss": 3.0374,
+      "step": 34225
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6004258990287781,
+      "learning_rate": 0.0002446085410300929,
+      "loss": 2.8079,
+      "step": 34226
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5990030169487,
+      "learning_rate": 0.0002445934096260932,
+      "loss": 3.0825,
+      "step": 34227
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7027409076690674,
+      "learning_rate": 0.0002445782783680237,
+      "loss": 2.9353,
+      "step": 34228
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6076624989509583,
+      "learning_rate": 0.0002445631472559244,
+      "loss": 3.0427,
+      "step": 34229
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6400550603866577,
+      "learning_rate": 0.0002445480162898352,
+      "loss": 3.0594,
+      "step": 34230
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5895517468452454,
+      "learning_rate": 0.0002445328854697957,
+      "loss": 3.1673,
+      "step": 34231
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6176828742027283,
+      "learning_rate": 0.000244517754795846,
+      "loss": 3.1844,
+      "step": 34232
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6322025060653687,
+      "learning_rate": 0.0002445026242680258,
+      "loss": 3.1415,
+      "step": 34233
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6357572078704834,
+      "learning_rate": 0.00024448749388637493,
+      "loss": 2.9114,
+      "step": 34234
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6500570178031921,
+      "learning_rate": 0.0002444723636509335,
+      "loss": 2.9372,
+      "step": 34235
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6248914003372192,
+      "learning_rate": 0.0002444572335617411,
+      "loss": 2.9285,
+      "step": 34236
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5908624529838562,
+      "learning_rate": 0.0002444421036188377,
+      "loss": 3.1703,
+      "step": 34237
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6131687760353088,
+      "learning_rate": 0.00024442697382226296,
+      "loss": 2.8722,
+      "step": 34238
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6797118186950684,
+      "learning_rate": 0.00024441184417205684,
+      "loss": 3.0294,
+      "step": 34239
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6466017365455627,
+      "learning_rate": 0.00024439671466825944,
+      "loss": 2.9447,
+      "step": 34240
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6415280699729919,
+      "learning_rate": 0.0002443815853109103,
+      "loss": 3.1305,
+      "step": 34241
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.613725483417511,
+      "learning_rate": 0.0002443664561000494,
+      "loss": 2.773,
+      "step": 34242
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6166124939918518,
+      "learning_rate": 0.00024435132703571645,
+      "loss": 2.8056,
+      "step": 34243
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6828124523162842,
+      "learning_rate": 0.00024433619811795146,
+      "loss": 2.9771,
+      "step": 34244
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6250672936439514,
+      "learning_rate": 0.00024432106934679424,
+      "loss": 3.0102,
+      "step": 34245
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6510005593299866,
+      "learning_rate": 0.0002443059407222845,
+      "loss": 2.931,
+      "step": 34246
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6244876980781555,
+      "learning_rate": 0.0002442908122444623,
+      "loss": 2.9292,
+      "step": 34247
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.631175696849823,
+      "learning_rate": 0.0002442756839133674,
+      "loss": 3.1205,
+      "step": 34248
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6444908380508423,
+      "learning_rate": 0.0002442605557290395,
+      "loss": 3.013,
+      "step": 34249
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6660930514335632,
+      "learning_rate": 0.0002442454276915187,
+      "loss": 3.1253,
+      "step": 34250
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6830661296844482,
+      "learning_rate": 0.00024423029980084464,
+      "loss": 3.1829,
+      "step": 34251
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6448426246643066,
+      "learning_rate": 0.00024421517205705734,
+      "loss": 3.0845,
+      "step": 34252
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6062400341033936,
+      "learning_rate": 0.00024420004446019635,
+      "loss": 3.12,
+      "step": 34253
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5658940672874451,
+      "learning_rate": 0.00024418491701030183,
+      "loss": 2.9393,
+      "step": 34254
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6128112077713013,
+      "learning_rate": 0.00024416978970741353,
+      "loss": 2.9671,
+      "step": 34255
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5911213159561157,
+      "learning_rate": 0.0002441546625515711,
+      "loss": 3.0101,
+      "step": 34256
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6012565493583679,
+      "learning_rate": 0.00024413953554281473,
+      "loss": 2.9874,
+      "step": 34257
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6286384463310242,
+      "learning_rate": 0.00024412440868118397,
+      "loss": 2.9793,
+      "step": 34258
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6148403286933899,
+      "learning_rate": 0.0002441092819667187,
+      "loss": 2.9738,
+      "step": 34259
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.616051435470581,
+      "learning_rate": 0.00024409415539945896,
+      "loss": 3.0531,
+      "step": 34260
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5852324962615967,
+      "learning_rate": 0.00024407902897944435,
+      "loss": 3.0542,
+      "step": 34261
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6138484477996826,
+      "learning_rate": 0.00024406390270671492,
+      "loss": 3.1578,
+      "step": 34262
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6208502054214478,
+      "learning_rate": 0.00024404877658131032,
+      "loss": 3.067,
+      "step": 34263
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6173650026321411,
+      "learning_rate": 0.00024403365060327045,
+      "loss": 2.8049,
+      "step": 34264
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6221855878829956,
+      "learning_rate": 0.0002440185247726353,
+      "loss": 3.171,
+      "step": 34265
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6358213424682617,
+      "learning_rate": 0.0002440033990894445,
+      "loss": 3.2996,
+      "step": 34266
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5805011987686157,
+      "learning_rate": 0.00024398827355373803,
+      "loss": 3.1994,
+      "step": 34267
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6460946798324585,
+      "learning_rate": 0.00024397314816555557,
+      "loss": 2.9848,
+      "step": 34268
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6628650426864624,
+      "learning_rate": 0.00024395802292493708,
+      "loss": 2.9467,
+      "step": 34269
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6098015308380127,
+      "learning_rate": 0.00024394289783192245,
+      "loss": 3.0061,
+      "step": 34270
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6058465242385864,
+      "learning_rate": 0.00024392777288655132,
+      "loss": 2.85,
+      "step": 34271
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6222949624061584,
+      "learning_rate": 0.0002439126480888638,
+      "loss": 3.2935,
+      "step": 34272
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6029030084609985,
+      "learning_rate": 0.00024389752343889943,
+      "loss": 3.1337,
+      "step": 34273
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5769012570381165,
+      "learning_rate": 0.00024388239893669833,
+      "loss": 2.9523,
+      "step": 34274
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6386813521385193,
+      "learning_rate": 0.0002438672745823,
+      "loss": 3.0349,
+      "step": 34275
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6211322546005249,
+      "learning_rate": 0.00024385215037574457,
+      "loss": 3.046,
+      "step": 34276
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6359053254127502,
+      "learning_rate": 0.00024383702631707182,
+      "loss": 2.9339,
+      "step": 34277
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6204125285148621,
+      "learning_rate": 0.0002438219024063214,
+      "loss": 3.1282,
+      "step": 34278
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6863940358161926,
+      "learning_rate": 0.00024380677864353344,
+      "loss": 3.1766,
+      "step": 34279
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6086069345474243,
+      "learning_rate": 0.00024379165502874743,
+      "loss": 2.8825,
+      "step": 34280
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6797551512718201,
+      "learning_rate": 0.00024377653156200347,
+      "loss": 2.948,
+      "step": 34281
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6163539290428162,
+      "learning_rate": 0.0002437614082433414,
+      "loss": 3.045,
+      "step": 34282
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6408080458641052,
+      "learning_rate": 0.0002437462850728008,
+      "loss": 3.0271,
+      "step": 34283
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6323572993278503,
+      "learning_rate": 0.00024373116205042176,
+      "loss": 2.9313,
+      "step": 34284
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.611010730266571,
+      "learning_rate": 0.00024371603917624398,
+      "loss": 3.0656,
+      "step": 34285
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.615911602973938,
+      "learning_rate": 0.00024370091645030726,
+      "loss": 3.2463,
+      "step": 34286
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6529790163040161,
+      "learning_rate": 0.0002436857938726516,
+      "loss": 2.9683,
+      "step": 34287
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.64628005027771,
+      "learning_rate": 0.00024367067144331667,
+      "loss": 2.9319,
+      "step": 34288
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5937123894691467,
+      "learning_rate": 0.00024365554916234243,
+      "loss": 3.0921,
+      "step": 34289
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6053999066352844,
+      "learning_rate": 0.00024364042702976847,
+      "loss": 2.9979,
+      "step": 34290
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.635191023349762,
+      "learning_rate": 0.00024362530504563486,
+      "loss": 2.9769,
+      "step": 34291
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6321050524711609,
+      "learning_rate": 0.00024361018320998137,
+      "loss": 2.9325,
+      "step": 34292
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6223572492599487,
+      "learning_rate": 0.00024359506152284776,
+      "loss": 3.0409,
+      "step": 34293
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6281905770301819,
+      "learning_rate": 0.00024357993998427396,
+      "loss": 3.1466,
+      "step": 34294
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5953301191329956,
+      "learning_rate": 0.0002435648185942997,
+      "loss": 3.0765,
+      "step": 34295
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6165432333946228,
+      "learning_rate": 0.0002435496973529648,
+      "loss": 3.0896,
+      "step": 34296
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5897827744483948,
+      "learning_rate": 0.00024353457626030922,
+      "loss": 2.9357,
+      "step": 34297
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6467866897583008,
+      "learning_rate": 0.00024351945531637267,
+      "loss": 2.8904,
+      "step": 34298
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6166879534721375,
+      "learning_rate": 0.00024350433452119505,
+      "loss": 3.0529,
+      "step": 34299
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6443753242492676,
+      "learning_rate": 0.00024348921387481602,
+      "loss": 2.7317,
+      "step": 34300
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.755827009677887,
+      "learning_rate": 0.00024347409337727554,
+      "loss": 3.1424,
+      "step": 34301
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5986933708190918,
+      "learning_rate": 0.00024345897302861355,
+      "loss": 3.1745,
+      "step": 34302
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6184709668159485,
+      "learning_rate": 0.00024344385282886962,
+      "loss": 3.0542,
+      "step": 34303
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9942894577980042,
+      "learning_rate": 0.0002434287327780838,
+      "loss": 3.0583,
+      "step": 34304
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6022435426712036,
+      "learning_rate": 0.00024341361287629572,
+      "loss": 3.154,
+      "step": 34305
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6629198789596558,
+      "learning_rate": 0.0002433984931235453,
+      "loss": 3.0552,
+      "step": 34306
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5723194479942322,
+      "learning_rate": 0.00024338337351987246,
+      "loss": 3.1065,
+      "step": 34307
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6211422085762024,
+      "learning_rate": 0.00024336825406531676,
+      "loss": 3.1316,
+      "step": 34308
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6164812445640564,
+      "learning_rate": 0.00024335313475991834,
+      "loss": 2.9869,
+      "step": 34309
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0711008310317993,
+      "learning_rate": 0.00024333801560371676,
+      "loss": 2.9982,
+      "step": 34310
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6362112760543823,
+      "learning_rate": 0.00024332289659675192,
+      "loss": 3.0657,
+      "step": 34311
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5985422134399414,
+      "learning_rate": 0.00024330777773906375,
+      "loss": 3.1279,
+      "step": 34312
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6208683848381042,
+      "learning_rate": 0.00024329265903069193,
+      "loss": 2.8673,
+      "step": 34313
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6138306856155396,
+      "learning_rate": 0.00024327754047167641,
+      "loss": 3.024,
+      "step": 34314
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5941352844238281,
+      "learning_rate": 0.00024326242206205678,
+      "loss": 2.8803,
+      "step": 34315
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6221591830253601,
+      "learning_rate": 0.0002432473038018731,
+      "loss": 3.1186,
+      "step": 34316
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6871387958526611,
+      "learning_rate": 0.00024323218569116515,
+      "loss": 3.0352,
+      "step": 34317
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5821581482887268,
+      "learning_rate": 0.0002432170677299726,
+      "loss": 3.2781,
+      "step": 34318
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6137875914573669,
+      "learning_rate": 0.00024320194991833544,
+      "loss": 2.8125,
+      "step": 34319
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6038451194763184,
+      "learning_rate": 0.0002431868322562934,
+      "loss": 3.1545,
+      "step": 34320
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6101359128952026,
+      "learning_rate": 0.00024317171474388618,
+      "loss": 3.026,
+      "step": 34321
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7233001589775085,
+      "learning_rate": 0.00024315659738115392,
+      "loss": 2.9277,
+      "step": 34322
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6210648417472839,
+      "learning_rate": 0.0002431414801681361,
+      "loss": 3.0142,
+      "step": 34323
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.598692774772644,
+      "learning_rate": 0.00024312636310487278,
+      "loss": 2.9345,
+      "step": 34324
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5950759053230286,
+      "learning_rate": 0.00024311124619140365,
+      "loss": 3.062,
+      "step": 34325
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5895894765853882,
+      "learning_rate": 0.00024309612942776846,
+      "loss": 3.2376,
+      "step": 34326
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.638976514339447,
+      "learning_rate": 0.00024308101281400726,
+      "loss": 2.917,
+      "step": 34327
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6100997924804688,
+      "learning_rate": 0.00024306589635015961,
+      "loss": 3.0652,
+      "step": 34328
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6079884767532349,
+      "learning_rate": 0.00024305078003626553,
+      "loss": 3.0837,
+      "step": 34329
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5833554267883301,
+      "learning_rate": 0.0002430356638723646,
+      "loss": 3.0955,
+      "step": 34330
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5608810782432556,
+      "learning_rate": 0.00024302054785849684,
+      "loss": 3.0168,
+      "step": 34331
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6720162630081177,
+      "learning_rate": 0.00024300543199470208,
+      "loss": 3.0969,
+      "step": 34332
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6119675636291504,
+      "learning_rate": 0.00024299031628101988,
+      "loss": 3.1501,
+      "step": 34333
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6166000366210938,
+      "learning_rate": 0.00024297520071749033,
+      "loss": 3.1214,
+      "step": 34334
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7879331707954407,
+      "learning_rate": 0.00024296008530415308,
+      "loss": 3.0078,
+      "step": 34335
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.630114734172821,
+      "learning_rate": 0.0002429449700410479,
+      "loss": 3.1951,
+      "step": 34336
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6173840761184692,
+      "learning_rate": 0.00024292985492821488,
+      "loss": 3.1601,
+      "step": 34337
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6316830515861511,
+      "learning_rate": 0.0002429147399656935,
+      "loss": 3.086,
+      "step": 34338
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6645943522453308,
+      "learning_rate": 0.00024289962515352384,
+      "loss": 2.9076,
+      "step": 34339
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6618261933326721,
+      "learning_rate": 0.0002428845104917454,
+      "loss": 3.1904,
+      "step": 34340
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6228432655334473,
+      "learning_rate": 0.0002428693959803982,
+      "loss": 3.0376,
+      "step": 34341
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6474012732505798,
+      "learning_rate": 0.00024285428161952214,
+      "loss": 3.1362,
+      "step": 34342
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5851452350616455,
+      "learning_rate": 0.0002428391674091568,
+      "loss": 3.0867,
+      "step": 34343
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6338858604431152,
+      "learning_rate": 0.00024282405334934221,
+      "loss": 3.013,
+      "step": 34344
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6032165288925171,
+      "learning_rate": 0.0002428089394401179,
+      "loss": 3.1117,
+      "step": 34345
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6296061873435974,
+      "learning_rate": 0.00024279382568152393,
+      "loss": 2.954,
+      "step": 34346
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5732358694076538,
+      "learning_rate": 0.00024277871207360003,
+      "loss": 2.9257,
+      "step": 34347
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6399105787277222,
+      "learning_rate": 0.0002427635986163859,
+      "loss": 2.9714,
+      "step": 34348
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6738459467887878,
+      "learning_rate": 0.00024274848530992155,
+      "loss": 2.9373,
+      "step": 34349
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6233891844749451,
+      "learning_rate": 0.0002427333721542466,
+      "loss": 3.1315,
+      "step": 34350
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6309840083122253,
+      "learning_rate": 0.00024271825914940102,
+      "loss": 3.0287,
+      "step": 34351
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6268077492713928,
+      "learning_rate": 0.00024270314629542435,
+      "loss": 3.0762,
+      "step": 34352
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5958818793296814,
+      "learning_rate": 0.00024268803359235667,
+      "loss": 2.9662,
+      "step": 34353
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6950878500938416,
+      "learning_rate": 0.0002426729210402377,
+      "loss": 2.8479,
+      "step": 34354
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6276915073394775,
+      "learning_rate": 0.00024265780863910716,
+      "loss": 3.1364,
+      "step": 34355
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.647860050201416,
+      "learning_rate": 0.00024264269638900497,
+      "loss": 2.5724,
+      "step": 34356
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.599606454372406,
+      "learning_rate": 0.00024262758428997088,
+      "loss": 2.9408,
+      "step": 34357
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6251691579818726,
+      "learning_rate": 0.0002426124723420446,
+      "loss": 3.0815,
+      "step": 34358
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.597493052482605,
+      "learning_rate": 0.00024259736054526614,
+      "loss": 2.9994,
+      "step": 34359
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6401041150093079,
+      "learning_rate": 0.00024258224889967512,
+      "loss": 3.0447,
+      "step": 34360
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6328083276748657,
+      "learning_rate": 0.00024256713740531148,
+      "loss": 2.9255,
+      "step": 34361
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6329466700553894,
+      "learning_rate": 0.00024255202606221485,
+      "loss": 2.923,
+      "step": 34362
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6171144247055054,
+      "learning_rate": 0.00024253691487042511,
+      "loss": 2.8842,
+      "step": 34363
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6171637773513794,
+      "learning_rate": 0.00024252180382998217,
+      "loss": 3.1929,
+      "step": 34364
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6588241457939148,
+      "learning_rate": 0.0002425066929409257,
+      "loss": 2.9894,
+      "step": 34365
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6128795742988586,
+      "learning_rate": 0.0002424915822032956,
+      "loss": 2.8934,
+      "step": 34366
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.584743082523346,
+      "learning_rate": 0.00024247647161713148,
+      "loss": 3.0518,
+      "step": 34367
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6354223489761353,
+      "learning_rate": 0.00024246136118247333,
+      "loss": 2.8951,
+      "step": 34368
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6299620866775513,
+      "learning_rate": 0.00024244625089936091,
+      "loss": 2.9908,
+      "step": 34369
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6883636713027954,
+      "learning_rate": 0.0002424311407678339,
+      "loss": 2.9329,
+      "step": 34370
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5922103524208069,
+      "learning_rate": 0.0002424160307879323,
+      "loss": 3.091,
+      "step": 34371
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6945319175720215,
+      "learning_rate": 0.0002424009209596957,
+      "loss": 3.1215,
+      "step": 34372
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6298499703407288,
+      "learning_rate": 0.00024238581128316396,
+      "loss": 3.0817,
+      "step": 34373
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.632078230381012,
+      "learning_rate": 0.00024237070175837705,
+      "loss": 3.1141,
+      "step": 34374
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.633488118648529,
+      "learning_rate": 0.0002423555923853745,
+      "loss": 2.97,
+      "step": 34375
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6549461483955383,
+      "learning_rate": 0.00024234048316419633,
+      "loss": 3.0173,
+      "step": 34376
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6033638715744019,
+      "learning_rate": 0.0002423253740948821,
+      "loss": 2.9334,
+      "step": 34377
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6423242688179016,
+      "learning_rate": 0.00024231026517747178,
+      "loss": 3.1711,
+      "step": 34378
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.586613118648529,
+      "learning_rate": 0.0002422951564120052,
+      "loss": 2.9276,
+      "step": 34379
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6340028643608093,
+      "learning_rate": 0.00024228004779852194,
+      "loss": 3.0458,
+      "step": 34380
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6134786009788513,
+      "learning_rate": 0.00024226493933706208,
+      "loss": 3.0783,
+      "step": 34381
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.641250491142273,
+      "learning_rate": 0.00024224983102766516,
+      "loss": 3.0496,
+      "step": 34382
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6067681908607483,
+      "learning_rate": 0.000242234722870371,
+      "loss": 2.8943,
+      "step": 34383
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5972951650619507,
+      "learning_rate": 0.00024221961486521962,
+      "loss": 3.0228,
+      "step": 34384
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6391837000846863,
+      "learning_rate": 0.0002422045070122505,
+      "loss": 2.9397,
+      "step": 34385
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5825820565223694,
+      "learning_rate": 0.00024218939931150375,
+      "loss": 3.1374,
+      "step": 34386
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6349732875823975,
+      "learning_rate": 0.00024217429176301889,
+      "loss": 2.9957,
+      "step": 34387
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6446109414100647,
+      "learning_rate": 0.00024215918436683577,
+      "loss": 3.1967,
+      "step": 34388
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.636913001537323,
+      "learning_rate": 0.00024214407712299437,
+      "loss": 3.1058,
+      "step": 34389
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5909577012062073,
+      "learning_rate": 0.00024212897003153425,
+      "loss": 3.0518,
+      "step": 34390
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6335484981536865,
+      "learning_rate": 0.00024211386309249539,
+      "loss": 3.0988,
+      "step": 34391
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6210086941719055,
+      "learning_rate": 0.0002420987563059173,
+      "loss": 2.9179,
+      "step": 34392
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6080520153045654,
+      "learning_rate": 0.00024208364967184,
+      "loss": 3.227,
+      "step": 34393
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6113629341125488,
+      "learning_rate": 0.00024206854319030333,
+      "loss": 2.9451,
+      "step": 34394
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6378176808357239,
+      "learning_rate": 0.00024205343686134682,
+      "loss": 3.1103,
+      "step": 34395
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6006256341934204,
+      "learning_rate": 0.00024203833068501052,
+      "loss": 2.9884,
+      "step": 34396
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6118775010108948,
+      "learning_rate": 0.00024202322466133406,
+      "loss": 2.7407,
+      "step": 34397
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6467645168304443,
+      "learning_rate": 0.00024200811879035718,
+      "loss": 2.7859,
+      "step": 34398
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6246896386146545,
+      "learning_rate": 0.00024199301307211988,
+      "loss": 3.0706,
+      "step": 34399
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6289390325546265,
+      "learning_rate": 0.00024197790750666177,
+      "loss": 3.0411,
+      "step": 34400
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6309362649917603,
+      "learning_rate": 0.00024196280209402276,
+      "loss": 2.9929,
+      "step": 34401
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6463305950164795,
+      "learning_rate": 0.0002419476968342424,
+      "loss": 3.0134,
+      "step": 34402
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6127621531486511,
+      "learning_rate": 0.00024193259172736067,
+      "loss": 3.0831,
+      "step": 34403
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6621184945106506,
+      "learning_rate": 0.00024191748677341744,
+      "loss": 2.9493,
+      "step": 34404
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6097039580345154,
+      "learning_rate": 0.0002419023819724523,
+      "loss": 3.0993,
+      "step": 34405
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.660682201385498,
+      "learning_rate": 0.00024188727732450517,
+      "loss": 3.2129,
+      "step": 34406
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6187965869903564,
+      "learning_rate": 0.00024187217282961563,
+      "loss": 2.8815,
+      "step": 34407
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5884891152381897,
+      "learning_rate": 0.00024185706848782365,
+      "loss": 3.0827,
+      "step": 34408
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5681439638137817,
+      "learning_rate": 0.00024184196429916904,
+      "loss": 3.1114,
+      "step": 34409
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6478064060211182,
+      "learning_rate": 0.00024182686026369142,
+      "loss": 3.0638,
+      "step": 34410
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6609353423118591,
+      "learning_rate": 0.00024181175638143072,
+      "loss": 3.0618,
+      "step": 34411
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6224457025527954,
+      "learning_rate": 0.0002417966526524266,
+      "loss": 2.9672,
+      "step": 34412
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6122366189956665,
+      "learning_rate": 0.00024178154907671887,
+      "loss": 2.8498,
+      "step": 34413
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6315528154373169,
+      "learning_rate": 0.00024176644565434744,
+      "loss": 2.8217,
+      "step": 34414
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.714515209197998,
+      "learning_rate": 0.0002417513423853519,
+      "loss": 2.774,
+      "step": 34415
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6689561009407043,
+      "learning_rate": 0.00024173623926977223,
+      "loss": 2.9304,
+      "step": 34416
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6164539456367493,
+      "learning_rate": 0.00024172113630764792,
+      "loss": 3.1242,
+      "step": 34417
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6706748604774475,
+      "learning_rate": 0.000241706033499019,
+      "loss": 3.3955,
+      "step": 34418
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5987676382064819,
+      "learning_rate": 0.00024169093084392524,
+      "loss": 3.03,
+      "step": 34419
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6364895105361938,
+      "learning_rate": 0.0002416758283424062,
+      "loss": 3.2174,
+      "step": 34420
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.755660891532898,
+      "learning_rate": 0.00024166072599450192,
+      "loss": 2.9115,
+      "step": 34421
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6371123194694519,
+      "learning_rate": 0.00024164562380025203,
+      "loss": 3.0867,
+      "step": 34422
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6260061264038086,
+      "learning_rate": 0.00024163052175969627,
+      "loss": 2.9373,
+      "step": 34423
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.601723849773407,
+      "learning_rate": 0.00024161541987287457,
+      "loss": 3.0002,
+      "step": 34424
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6575174331665039,
+      "learning_rate": 0.00024160031813982657,
+      "loss": 3.0804,
+      "step": 34425
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6266413927078247,
+      "learning_rate": 0.00024158521656059213,
+      "loss": 3.0032,
+      "step": 34426
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6671347618103027,
+      "learning_rate": 0.000241570115135211,
+      "loss": 2.9415,
+      "step": 34427
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6078870296478271,
+      "learning_rate": 0.00024155501386372282,
+      "loss": 2.9793,
+      "step": 34428
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6798886060714722,
+      "learning_rate": 0.00024153991274616763,
+      "loss": 2.7276,
+      "step": 34429
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6024205088615417,
+      "learning_rate": 0.00024152481178258501,
+      "loss": 3.1548,
+      "step": 34430
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6181690692901611,
+      "learning_rate": 0.00024150971097301486,
+      "loss": 2.8255,
+      "step": 34431
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6173415780067444,
+      "learning_rate": 0.00024149461031749668,
+      "loss": 3.0516,
+      "step": 34432
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6175852417945862,
+      "learning_rate": 0.00024147950981607064,
+      "loss": 2.9044,
+      "step": 34433
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6466128826141357,
+      "learning_rate": 0.0002414644094687762,
+      "loss": 2.893,
+      "step": 34434
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.626565158367157,
+      "learning_rate": 0.00024144930927565318,
+      "loss": 3.1384,
+      "step": 34435
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6032750606536865,
+      "learning_rate": 0.00024143420923674157,
+      "loss": 2.773,
+      "step": 34436
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6452335715293884,
+      "learning_rate": 0.0002414191093520809,
+      "loss": 2.8955,
+      "step": 34437
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6146382093429565,
+      "learning_rate": 0.00024140400962171106,
+      "loss": 2.935,
+      "step": 34438
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5858275294303894,
+      "learning_rate": 0.00024138891004567167,
+      "loss": 3.0345,
+      "step": 34439
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6086834669113159,
+      "learning_rate": 0.00024137381062400268,
+      "loss": 2.9822,
+      "step": 34440
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6375246047973633,
+      "learning_rate": 0.00024135871135674388,
+      "loss": 3.1297,
+      "step": 34441
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6939441561698914,
+      "learning_rate": 0.00024134361224393479,
+      "loss": 3.0535,
+      "step": 34442
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6073477864265442,
+      "learning_rate": 0.00024132851328561544,
+      "loss": 2.944,
+      "step": 34443
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6158416271209717,
+      "learning_rate": 0.0002413134144818254,
+      "loss": 3.151,
+      "step": 34444
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5689690709114075,
+      "learning_rate": 0.0002412983158326046,
+      "loss": 2.9117,
+      "step": 34445
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6501169800758362,
+      "learning_rate": 0.0002412832173379928,
+      "loss": 3.1277,
+      "step": 34446
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6253321766853333,
+      "learning_rate": 0.00024126811899802958,
+      "loss": 3.1249,
+      "step": 34447
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6375982761383057,
+      "learning_rate": 0.00024125302081275492,
+      "loss": 3.1011,
+      "step": 34448
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6295775175094604,
+      "learning_rate": 0.00024123792278220847,
+      "loss": 3.1153,
+      "step": 34449
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6375122666358948,
+      "learning_rate": 0.00024122282490642998,
+      "loss": 2.9863,
+      "step": 34450
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6656721234321594,
+      "learning_rate": 0.00024120772718545935,
+      "loss": 3.0008,
+      "step": 34451
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.584492564201355,
+      "learning_rate": 0.00024119262961933618,
+      "loss": 2.9974,
+      "step": 34452
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5878311395645142,
+      "learning_rate": 0.0002411775322081004,
+      "loss": 3.0679,
+      "step": 34453
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6173829436302185,
+      "learning_rate": 0.00024116243495179152,
+      "loss": 2.82,
+      "step": 34454
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6128963828086853,
+      "learning_rate": 0.00024114733785044953,
+      "loss": 2.8603,
+      "step": 34455
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6082571744918823,
+      "learning_rate": 0.0002411322409041142,
+      "loss": 2.9488,
+      "step": 34456
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6463648080825806,
+      "learning_rate": 0.0002411171441128251,
+      "loss": 3.0709,
+      "step": 34457
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6087504625320435,
+      "learning_rate": 0.00024110204747662225,
+      "loss": 3.1574,
+      "step": 34458
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6248615980148315,
+      "learning_rate": 0.0002410869509955452,
+      "loss": 3.0775,
+      "step": 34459
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6408868432044983,
+      "learning_rate": 0.00024107185466963369,
+      "loss": 3.1373,
+      "step": 34460
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6176990866661072,
+      "learning_rate": 0.00024105675849892772,
+      "loss": 3.0042,
+      "step": 34461
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6030988693237305,
+      "learning_rate": 0.00024104166248346685,
+      "loss": 2.9704,
+      "step": 34462
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6126129031181335,
+      "learning_rate": 0.00024102656662329094,
+      "loss": 2.9439,
+      "step": 34463
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.620477020740509,
+      "learning_rate": 0.0002410114709184396,
+      "loss": 3.2136,
+      "step": 34464
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6435030102729797,
+      "learning_rate": 0.00024099637536895263,
+      "loss": 2.8834,
+      "step": 34465
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6259873509407043,
+      "learning_rate": 0.00024098127997487006,
+      "loss": 3.2122,
+      "step": 34466
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.630929708480835,
+      "learning_rate": 0.00024096618473623134,
+      "loss": 3.1022,
+      "step": 34467
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6018049120903015,
+      "learning_rate": 0.0002409510896530764,
+      "loss": 2.6932,
+      "step": 34468
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6094369292259216,
+      "learning_rate": 0.0002409359947254448,
+      "loss": 2.9726,
+      "step": 34469
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6517297625541687,
+      "learning_rate": 0.0002409208999533765,
+      "loss": 3.0395,
+      "step": 34470
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6332142949104309,
+      "learning_rate": 0.0002409058053369112,
+      "loss": 2.9317,
+      "step": 34471
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6489332914352417,
+      "learning_rate": 0.00024089071087608852,
+      "loss": 2.8955,
+      "step": 34472
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6442323327064514,
+      "learning_rate": 0.0002408756165709485,
+      "loss": 3.0071,
+      "step": 34473
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6622292995452881,
+      "learning_rate": 0.0002408605224215306,
+      "loss": 2.5943,
+      "step": 34474
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6429809927940369,
+      "learning_rate": 0.00024084542842787467,
+      "loss": 2.9143,
+      "step": 34475
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.658168375492096,
+      "learning_rate": 0.00024083033459002066,
+      "loss": 3.0773,
+      "step": 34476
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6099553108215332,
+      "learning_rate": 0.00024081524090800805,
+      "loss": 3.1176,
+      "step": 34477
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6159991025924683,
+      "learning_rate": 0.00024080014738187677,
+      "loss": 2.9891,
+      "step": 34478
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.589513897895813,
+      "learning_rate": 0.0002407850540116664,
+      "loss": 2.9638,
+      "step": 34479
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6264899969100952,
+      "learning_rate": 0.00024076996079741686,
+      "loss": 2.898,
+      "step": 34480
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7650187611579895,
+      "learning_rate": 0.00024075486773916793,
+      "loss": 3.0772,
+      "step": 34481
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6351287364959717,
+      "learning_rate": 0.0002407397748369591,
+      "loss": 3.1892,
+      "step": 34482
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6214819550514221,
+      "learning_rate": 0.00024072468209083045,
+      "loss": 2.9347,
+      "step": 34483
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6406108736991882,
+      "learning_rate": 0.00024070958950082153,
+      "loss": 2.9502,
+      "step": 34484
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6591122150421143,
+      "learning_rate": 0.00024069449706697207,
+      "loss": 3.3276,
+      "step": 34485
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6212307214736938,
+      "learning_rate": 0.000240679404789322,
+      "loss": 3.2392,
+      "step": 34486
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7033385634422302,
+      "learning_rate": 0.00024066431266791085,
+      "loss": 2.9312,
+      "step": 34487
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5959001779556274,
+      "learning_rate": 0.0002406492207027786,
+      "loss": 2.7623,
+      "step": 34488
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5990002155303955,
+      "learning_rate": 0.00024063412889396485,
+      "loss": 2.971,
+      "step": 34489
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6115400195121765,
+      "learning_rate": 0.00024061903724150927,
+      "loss": 2.8386,
+      "step": 34490
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6301840543746948,
+      "learning_rate": 0.00024060394574545186,
+      "loss": 2.8966,
+      "step": 34491
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6668326258659363,
+      "learning_rate": 0.00024058885440583216,
+      "loss": 2.8776,
+      "step": 34492
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6199162602424622,
+      "learning_rate": 0.00024057376322269007,
+      "loss": 3.1312,
+      "step": 34493
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6100919842720032,
+      "learning_rate": 0.0002405586721960651,
+      "loss": 3.1269,
+      "step": 34494
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6069178581237793,
+      "learning_rate": 0.00024054358132599723,
+      "loss": 2.9928,
+      "step": 34495
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6124541163444519,
+      "learning_rate": 0.00024052849061252616,
+      "loss": 3.018,
+      "step": 34496
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5998200178146362,
+      "learning_rate": 0.0002405134000556915,
+      "loss": 2.8799,
+      "step": 34497
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6153830885887146,
+      "learning_rate": 0.00024049830965553322,
+      "loss": 3.1137,
+      "step": 34498
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6186378002166748,
+      "learning_rate": 0.00024048321941209087,
+      "loss": 3.0032,
+      "step": 34499
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.625507652759552,
+      "learning_rate": 0.00024046812932540417,
+      "loss": 3.2128,
+      "step": 34500
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.628273069858551,
+      "learning_rate": 0.00024045303939551315,
+      "loss": 3.0745,
+      "step": 34501
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6160310506820679,
+      "learning_rate": 0.00024043794962245725,
+      "loss": 3.0487,
+      "step": 34502
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6147609353065491,
+      "learning_rate": 0.0002404228600062764,
+      "loss": 3.0751,
+      "step": 34503
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6561556458473206,
+      "learning_rate": 0.0002404077705470102,
+      "loss": 3.1311,
+      "step": 34504
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6268799901008606,
+      "learning_rate": 0.00024039268124469837,
+      "loss": 2.981,
+      "step": 34505
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6238470077514648,
+      "learning_rate": 0.00024037759209938092,
+      "loss": 2.9657,
+      "step": 34506
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5943543314933777,
+      "learning_rate": 0.0002403625031110974,
+      "loss": 2.8509,
+      "step": 34507
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6060153841972351,
+      "learning_rate": 0.00024034741427988757,
+      "loss": 3.0004,
+      "step": 34508
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6190601587295532,
+      "learning_rate": 0.00024033232560579106,
+      "loss": 3.0331,
+      "step": 34509
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6295119524002075,
+      "learning_rate": 0.00024031723708884786,
+      "loss": 2.895,
+      "step": 34510
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6272006034851074,
+      "learning_rate": 0.00024030214872909747,
+      "loss": 3.0444,
+      "step": 34511
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6380195021629333,
+      "learning_rate": 0.00024028706052657973,
+      "loss": 3.0712,
+      "step": 34512
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6345843076705933,
+      "learning_rate": 0.00024027197248133446,
+      "loss": 3.0436,
+      "step": 34513
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.586628258228302,
+      "learning_rate": 0.00024025688459340127,
+      "loss": 3.2935,
+      "step": 34514
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6063928008079529,
+      "learning_rate": 0.00024024179686282,
+      "loss": 2.8827,
+      "step": 34515
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5976722836494446,
+      "learning_rate": 0.00024022670928963024,
+      "loss": 3.0413,
+      "step": 34516
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6519937515258789,
+      "learning_rate": 0.00024021162187387184,
+      "loss": 2.9673,
+      "step": 34517
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6437779068946838,
+      "learning_rate": 0.0002401965346155846,
+      "loss": 3.0316,
+      "step": 34518
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5971271395683289,
+      "learning_rate": 0.0002401814475148081,
+      "loss": 3.1481,
+      "step": 34519
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6342564225196838,
+      "learning_rate": 0.00024016636057158224,
+      "loss": 3.0502,
+      "step": 34520
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5975090265274048,
+      "learning_rate": 0.00024015127378594663,
+      "loss": 2.9394,
+      "step": 34521
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6094161868095398,
+      "learning_rate": 0.00024013618715794095,
+      "loss": 2.9791,
+      "step": 34522
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6090801954269409,
+      "learning_rate": 0.00024012110068760519,
+      "loss": 3.0233,
+      "step": 34523
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6072431802749634,
+      "learning_rate": 0.00024010601437497882,
+      "loss": 2.9527,
+      "step": 34524
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7141648530960083,
+      "learning_rate": 0.0002400909282201018,
+      "loss": 2.8589,
+      "step": 34525
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6420031785964966,
+      "learning_rate": 0.00024007584222301357,
+      "loss": 3.0201,
+      "step": 34526
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6151692867279053,
+      "learning_rate": 0.00024006075638375408,
+      "loss": 3.039,
+      "step": 34527
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6226093173027039,
+      "learning_rate": 0.00024004567070236312,
+      "loss": 3.0654,
+      "step": 34528
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6651445031166077,
+      "learning_rate": 0.00024003058517888028,
+      "loss": 2.8437,
+      "step": 34529
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5924610495567322,
+      "learning_rate": 0.00024001549981334541,
+      "loss": 2.7907,
+      "step": 34530
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6144448518753052,
+      "learning_rate": 0.00024000041460579802,
+      "loss": 3.0333,
+      "step": 34531
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6835182905197144,
+      "learning_rate": 0.00023998532955627803,
+      "loss": 2.8505,
+      "step": 34532
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6623280048370361,
+      "learning_rate": 0.00023997024466482523,
+      "loss": 3.0159,
+      "step": 34533
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6474083065986633,
+      "learning_rate": 0.00023995515993147911,
+      "loss": 2.9979,
+      "step": 34534
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6150049567222595,
+      "learning_rate": 0.00023994007535627968,
+      "loss": 3.1431,
+      "step": 34535
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6174440383911133,
+      "learning_rate": 0.00023992499093926646,
+      "loss": 2.996,
+      "step": 34536
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6341604590415955,
+      "learning_rate": 0.00023990990668047921,
+      "loss": 3.0105,
+      "step": 34537
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.624151349067688,
+      "learning_rate": 0.00023989482257995783,
+      "loss": 3.1166,
+      "step": 34538
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5957610011100769,
+      "learning_rate": 0.00023987973863774184,
+      "loss": 2.9692,
+      "step": 34539
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.655441403388977,
+      "learning_rate": 0.00023986465485387114,
+      "loss": 2.9397,
+      "step": 34540
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6130520701408386,
+      "learning_rate": 0.0002398495712283852,
+      "loss": 3.0796,
+      "step": 34541
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6145035624504089,
+      "learning_rate": 0.00023983448776132397,
+      "loss": 3.0219,
+      "step": 34542
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6132924556732178,
+      "learning_rate": 0.00023981940445272723,
+      "loss": 2.9801,
+      "step": 34543
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6345343589782715,
+      "learning_rate": 0.00023980432130263444,
+      "loss": 2.9226,
+      "step": 34544
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6521178483963013,
+      "learning_rate": 0.00023978923831108562,
+      "loss": 3.0685,
+      "step": 34545
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6504637598991394,
+      "learning_rate": 0.00023977415547812029,
+      "loss": 2.8594,
+      "step": 34546
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6152077913284302,
+      "learning_rate": 0.00023975907280377815,
+      "loss": 3.1033,
+      "step": 34547
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6426700949668884,
+      "learning_rate": 0.00023974399028809916,
+      "loss": 2.9754,
+      "step": 34548
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6243044137954712,
+      "learning_rate": 0.00023972890793112281,
+      "loss": 2.958,
+      "step": 34549
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6287975311279297,
+      "learning_rate": 0.00023971382573288903,
+      "loss": 3.0594,
+      "step": 34550
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6270124912261963,
+      "learning_rate": 0.00023969874369343734,
+      "loss": 2.976,
+      "step": 34551
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6438772678375244,
+      "learning_rate": 0.00023968366181280755,
+      "loss": 3.0046,
+      "step": 34552
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6129069328308105,
+      "learning_rate": 0.00023966858009103945,
+      "loss": 2.9007,
+      "step": 34553
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6193562150001526,
+      "learning_rate": 0.00023965349852817267,
+      "loss": 2.9481,
+      "step": 34554
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6159753203392029,
+      "learning_rate": 0.00023963841712424703,
+      "loss": 2.9394,
+      "step": 34555
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5913207530975342,
+      "learning_rate": 0.00023962333587930203,
+      "loss": 2.9432,
+      "step": 34556
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6216997504234314,
+      "learning_rate": 0.00023960825479337763,
+      "loss": 3.0619,
+      "step": 34557
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6246969699859619,
+      "learning_rate": 0.00023959317386651352,
+      "loss": 2.8386,
+      "step": 34558
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6128319501876831,
+      "learning_rate": 0.0002395780930987492,
+      "loss": 3.0849,
+      "step": 34559
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6397287249565125,
+      "learning_rate": 0.00023956301249012472,
+      "loss": 3.0055,
+      "step": 34560
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5970715880393982,
+      "learning_rate": 0.00023954793204067956,
+      "loss": 2.909,
+      "step": 34561
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6784390807151794,
+      "learning_rate": 0.00023953285175045346,
+      "loss": 3.0605,
+      "step": 34562
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.611601710319519,
+      "learning_rate": 0.00023951777161948634,
+      "loss": 3.0419,
+      "step": 34563
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6383347511291504,
+      "learning_rate": 0.00023950269164781766,
+      "loss": 2.9873,
+      "step": 34564
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5747080445289612,
+      "learning_rate": 0.00023948761183548734,
+      "loss": 2.8731,
+      "step": 34565
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.611871600151062,
+      "learning_rate": 0.00023947253218253484,
+      "loss": 3.0842,
+      "step": 34566
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6260141730308533,
+      "learning_rate": 0.00023945745268900008,
+      "loss": 3.1116,
+      "step": 34567
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5983566641807556,
+      "learning_rate": 0.00023944237335492285,
+      "loss": 3.0337,
+      "step": 34568
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.599165678024292,
+      "learning_rate": 0.0002394272941803427,
+      "loss": 2.9508,
+      "step": 34569
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6618895530700684,
+      "learning_rate": 0.00023941221516529944,
+      "loss": 3.1141,
+      "step": 34570
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5907045006752014,
+      "learning_rate": 0.00023939713630983267,
+      "loss": 2.8549,
+      "step": 34571
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6474612951278687,
+      "learning_rate": 0.0002393820576139822,
+      "loss": 2.9113,
+      "step": 34572
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5951142311096191,
+      "learning_rate": 0.00023936697907778778,
+      "loss": 2.9788,
+      "step": 34573
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6441212296485901,
+      "learning_rate": 0.00023935190070128895,
+      "loss": 2.8805,
+      "step": 34574
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6195458173751831,
+      "learning_rate": 0.0002393368224845257,
+      "loss": 3.1319,
+      "step": 34575
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.602984607219696,
+      "learning_rate": 0.00023932174442753748,
+      "loss": 3.1744,
+      "step": 34576
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.616973340511322,
+      "learning_rate": 0.00023930666653036408,
+      "loss": 3.1541,
+      "step": 34577
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6357001066207886,
+      "learning_rate": 0.00023929158879304534,
+      "loss": 3.112,
+      "step": 34578
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6247024536132812,
+      "learning_rate": 0.00023927651121562082,
+      "loss": 3.2426,
+      "step": 34579
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6100210547447205,
+      "learning_rate": 0.00023926143379813034,
+      "loss": 3.0513,
+      "step": 34580
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5985271334648132,
+      "learning_rate": 0.00023924635654061342,
+      "loss": 2.9466,
+      "step": 34581
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6380882859230042,
+      "learning_rate": 0.00023923127944310997,
+      "loss": 3.0692,
+      "step": 34582
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6195142269134521,
+      "learning_rate": 0.0002392162025056597,
+      "loss": 3.1293,
+      "step": 34583
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6390209197998047,
+      "learning_rate": 0.00023920112572830212,
+      "loss": 2.8975,
+      "step": 34584
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6426082849502563,
+      "learning_rate": 0.00023918604911107722,
+      "loss": 2.9478,
+      "step": 34585
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6332377195358276,
+      "learning_rate": 0.00023917097265402452,
+      "loss": 2.9731,
+      "step": 34586
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6898009181022644,
+      "learning_rate": 0.0002391558963571838,
+      "loss": 3.0346,
+      "step": 34587
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6284445524215698,
+      "learning_rate": 0.0002391408202205946,
+      "loss": 3.1084,
+      "step": 34588
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5925804376602173,
+      "learning_rate": 0.00023912574424429678,
+      "loss": 3.0499,
+      "step": 34589
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6358410716056824,
+      "learning_rate": 0.0002391106684283302,
+      "loss": 2.8512,
+      "step": 34590
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6737104654312134,
+      "learning_rate": 0.0002390955927727343,
+      "loss": 2.946,
+      "step": 34591
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.604070782661438,
+      "learning_rate": 0.00023908051727754897,
+      "loss": 3.0541,
+      "step": 34592
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5812761187553406,
+      "learning_rate": 0.00023906544194281367,
+      "loss": 3.1825,
+      "step": 34593
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5896308422088623,
+      "learning_rate": 0.00023905036676856835,
+      "loss": 3.1379,
+      "step": 34594
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6765037775039673,
+      "learning_rate": 0.00023903529175485272,
+      "loss": 2.9756,
+      "step": 34595
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6317283511161804,
+      "learning_rate": 0.00023902021690170623,
+      "loss": 3.07,
+      "step": 34596
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6281628012657166,
+      "learning_rate": 0.0002390051422091689,
+      "loss": 3.049,
+      "step": 34597
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6032910943031311,
+      "learning_rate": 0.00023899006767728024,
+      "loss": 2.9015,
+      "step": 34598
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6200457811355591,
+      "learning_rate": 0.00023897499330607992,
+      "loss": 3.0544,
+      "step": 34599
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5972832441329956,
+      "learning_rate": 0.00023895991909560785,
+      "loss": 3.0304,
+      "step": 34600
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6621426343917847,
+      "learning_rate": 0.00023894484504590355,
+      "loss": 3.1561,
+      "step": 34601
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6243399977684021,
+      "learning_rate": 0.00023892977115700685,
+      "loss": 2.9101,
+      "step": 34602
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6167004108428955,
+      "learning_rate": 0.00023891469742895723,
+      "loss": 2.9986,
+      "step": 34603
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7089727520942688,
+      "learning_rate": 0.0002388996238617946,
+      "loss": 3.242,
+      "step": 34604
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5827861428260803,
+      "learning_rate": 0.00023888455045555868,
+      "loss": 2.9402,
+      "step": 34605
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6445075273513794,
+      "learning_rate": 0.000238869477210289,
+      "loss": 3.0979,
+      "step": 34606
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6177965402603149,
+      "learning_rate": 0.0002388544041260254,
+      "loss": 2.9209,
+      "step": 34607
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6588315963745117,
+      "learning_rate": 0.00023883933120280747,
+      "loss": 2.9099,
+      "step": 34608
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6694788336753845,
+      "learning_rate": 0.000238824258440675,
+      "loss": 2.9249,
+      "step": 34609
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6350728869438171,
+      "learning_rate": 0.00023880918583966777,
+      "loss": 3.1211,
+      "step": 34610
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6018306016921997,
+      "learning_rate": 0.0002387941133998252,
+      "loss": 3.068,
+      "step": 34611
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6151270866394043,
+      "learning_rate": 0.00023877904112118726,
+      "loss": 2.7378,
+      "step": 34612
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6024587154388428,
+      "learning_rate": 0.0002387639690037935,
+      "loss": 3.111,
+      "step": 34613
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6354588866233826,
+      "learning_rate": 0.00023874889704768366,
+      "loss": 3.0932,
+      "step": 34614
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6052812337875366,
+      "learning_rate": 0.0002387338252528975,
+      "loss": 3.0522,
+      "step": 34615
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6550926566123962,
+      "learning_rate": 0.00023871875361947462,
+      "loss": 3.0193,
+      "step": 34616
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6020631790161133,
+      "learning_rate": 0.00023870368214745478,
+      "loss": 2.9907,
+      "step": 34617
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6404764652252197,
+      "learning_rate": 0.00023868861083687753,
+      "loss": 3.0878,
+      "step": 34618
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6281757950782776,
+      "learning_rate": 0.00023867353968778273,
+      "loss": 2.9839,
+      "step": 34619
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6446701884269714,
+      "learning_rate": 0.0002386584687002101,
+      "loss": 3.0934,
+      "step": 34620
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6888076663017273,
+      "learning_rate": 0.00023864339787419916,
+      "loss": 3.2198,
+      "step": 34621
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7113660573959351,
+      "learning_rate": 0.0002386283272097898,
+      "loss": 3.125,
+      "step": 34622
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6033316254615784,
+      "learning_rate": 0.00023861325670702154,
+      "loss": 2.9347,
+      "step": 34623
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6461472511291504,
+      "learning_rate": 0.00023859818636593407,
+      "loss": 2.9211,
+      "step": 34624
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6217671632766724,
+      "learning_rate": 0.00023858311618656734,
+      "loss": 2.9407,
+      "step": 34625
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7843379378318787,
+      "learning_rate": 0.00023856804616896075,
+      "loss": 3.0633,
+      "step": 34626
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6176000237464905,
+      "learning_rate": 0.0002385529763131542,
+      "loss": 3.0751,
+      "step": 34627
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6316340565681458,
+      "learning_rate": 0.00023853790661918715,
+      "loss": 2.9783,
+      "step": 34628
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.63956218957901,
+      "learning_rate": 0.00023852283708709944,
+      "loss": 3.0346,
+      "step": 34629
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6014463901519775,
+      "learning_rate": 0.00023850776771693082,
+      "loss": 3.088,
+      "step": 34630
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6172289252281189,
+      "learning_rate": 0.0002384926985087209,
+      "loss": 2.9347,
+      "step": 34631
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6280187964439392,
+      "learning_rate": 0.00023847762946250942,
+      "loss": 2.9345,
+      "step": 34632
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5740066170692444,
+      "learning_rate": 0.0002384625605783359,
+      "loss": 2.9651,
+      "step": 34633
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6673811078071594,
+      "learning_rate": 0.0002384474918562402,
+      "loss": 3.2161,
+      "step": 34634
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6165456175804138,
+      "learning_rate": 0.000238432423296262,
+      "loss": 3.0175,
+      "step": 34635
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6148802042007446,
+      "learning_rate": 0.0002384173548984409,
+      "loss": 2.9729,
+      "step": 34636
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6252344846725464,
+      "learning_rate": 0.0002384022866628167,
+      "loss": 3.0979,
+      "step": 34637
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6333229541778564,
+      "learning_rate": 0.00023838721858942898,
+      "loss": 2.8437,
+      "step": 34638
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6194198131561279,
+      "learning_rate": 0.0002383721506783174,
+      "loss": 2.8774,
+      "step": 34639
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6095027327537537,
+      "learning_rate": 0.00023835708292952183,
+      "loss": 2.9816,
+      "step": 34640
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.613484799861908,
+      "learning_rate": 0.0002383420153430818,
+      "loss": 3.0787,
+      "step": 34641
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5842860341072083,
+      "learning_rate": 0.00023832694791903707,
+      "loss": 3.0523,
+      "step": 34642
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6341204047203064,
+      "learning_rate": 0.0002383118806574272,
+      "loss": 3.1984,
+      "step": 34643
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6389890313148499,
+      "learning_rate": 0.00023829681355829203,
+      "loss": 2.8034,
+      "step": 34644
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5947014689445496,
+      "learning_rate": 0.00023828174662167122,
+      "loss": 3.1272,
+      "step": 34645
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6640393733978271,
+      "learning_rate": 0.00023826667984760426,
+      "loss": 2.9868,
+      "step": 34646
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6023179292678833,
+      "learning_rate": 0.00023825161323613113,
+      "loss": 3.1548,
+      "step": 34647
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6089648604393005,
+      "learning_rate": 0.0002382365467872913,
+      "loss": 2.9102,
+      "step": 34648
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6076211333274841,
+      "learning_rate": 0.00023822148050112448,
+      "loss": 3.0588,
+      "step": 34649
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7229902744293213,
+      "learning_rate": 0.0002382064143776705,
+      "loss": 2.927,
+      "step": 34650
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5934354662895203,
+      "learning_rate": 0.00023819134841696882,
+      "loss": 2.9372,
+      "step": 34651
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6203776597976685,
+      "learning_rate": 0.00023817628261905933,
+      "loss": 3.0984,
+      "step": 34652
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6217381358146667,
+      "learning_rate": 0.0002381612169839816,
+      "loss": 3.1926,
+      "step": 34653
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5950883030891418,
+      "learning_rate": 0.0002381461515117752,
+      "loss": 2.8172,
+      "step": 34654
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.622291624546051,
+      "learning_rate": 0.0002381310862024801,
+      "loss": 2.9426,
+      "step": 34655
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6245651841163635,
+      "learning_rate": 0.00023811602105613575,
+      "loss": 3.0114,
+      "step": 34656
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6118585467338562,
+      "learning_rate": 0.0002381009560727819,
+      "loss": 3.003,
+      "step": 34657
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7004932761192322,
+      "learning_rate": 0.00023808589125245814,
+      "loss": 2.9431,
+      "step": 34658
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6721332669258118,
+      "learning_rate": 0.0002380708265952043,
+      "loss": 2.9977,
+      "step": 34659
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6560796499252319,
+      "learning_rate": 0.00023805576210106003,
+      "loss": 2.9513,
+      "step": 34660
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5827532410621643,
+      "learning_rate": 0.0002380406977700648,
+      "loss": 2.8129,
+      "step": 34661
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6369756460189819,
+      "learning_rate": 0.00023802563360225863,
+      "loss": 3.1372,
+      "step": 34662
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6369935870170593,
+      "learning_rate": 0.00023801056959768093,
+      "loss": 2.876,
+      "step": 34663
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6207792162895203,
+      "learning_rate": 0.00023799550575637153,
+      "loss": 3.0598,
+      "step": 34664
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5954633355140686,
+      "learning_rate": 0.0002379804420783699,
+      "loss": 2.9627,
+      "step": 34665
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6295991539955139,
+      "learning_rate": 0.00023796537856371589,
+      "loss": 3.0739,
+      "step": 34666
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6417800188064575,
+      "learning_rate": 0.00023795031521244922,
+      "loss": 3.0968,
+      "step": 34667
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6085421442985535,
+      "learning_rate": 0.00023793525202460937,
+      "loss": 2.824,
+      "step": 34668
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6392890214920044,
+      "learning_rate": 0.00023792018900023622,
+      "loss": 2.9634,
+      "step": 34669
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6218932271003723,
+      "learning_rate": 0.00023790512613936925,
+      "loss": 3.0132,
+      "step": 34670
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.642927885055542,
+      "learning_rate": 0.00023789006344204823,
+      "loss": 2.981,
+      "step": 34671
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6467803120613098,
+      "learning_rate": 0.00023787500090831294,
+      "loss": 3.0479,
+      "step": 34672
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6575911045074463,
+      "learning_rate": 0.00023785993853820282,
+      "loss": 2.999,
+      "step": 34673
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6105745434761047,
+      "learning_rate": 0.00023784487633175778,
+      "loss": 3.017,
+      "step": 34674
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6278857588768005,
+      "learning_rate": 0.0002378298142890173,
+      "loss": 3.0142,
+      "step": 34675
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7186394929885864,
+      "learning_rate": 0.00023781475241002107,
+      "loss": 3.0599,
+      "step": 34676
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6153616309165955,
+      "learning_rate": 0.00023779969069480896,
+      "loss": 3.1681,
+      "step": 34677
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6550021171569824,
+      "learning_rate": 0.00023778462914342038,
+      "loss": 3.1913,
+      "step": 34678
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6431772112846375,
+      "learning_rate": 0.0002377695677558952,
+      "loss": 3.0687,
+      "step": 34679
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5983546376228333,
+      "learning_rate": 0.0002377545065322729,
+      "loss": 2.9564,
+      "step": 34680
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6056777238845825,
+      "learning_rate": 0.00023773944547259327,
+      "loss": 3.1077,
+      "step": 34681
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6222465634346008,
+      "learning_rate": 0.00023772438457689607,
+      "loss": 3.2957,
+      "step": 34682
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5904668569564819,
+      "learning_rate": 0.00023770932384522072,
+      "loss": 3.1094,
+      "step": 34683
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6390523910522461,
+      "learning_rate": 0.00023769426327760712,
+      "loss": 3.1585,
+      "step": 34684
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6436138153076172,
+      "learning_rate": 0.0002376792028740948,
+      "loss": 2.9761,
+      "step": 34685
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6362774968147278,
+      "learning_rate": 0.00023766414263472338,
+      "loss": 3.1662,
+      "step": 34686
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6165968775749207,
+      "learning_rate": 0.00023764908255953277,
+      "loss": 2.9115,
+      "step": 34687
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6462124586105347,
+      "learning_rate": 0.00023763402264856238,
+      "loss": 3.1004,
+      "step": 34688
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6038845777511597,
+      "learning_rate": 0.00023761896290185207,
+      "loss": 3.167,
+      "step": 34689
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6276283860206604,
+      "learning_rate": 0.00023760390331944127,
+      "loss": 3.0901,
+      "step": 34690
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6404023766517639,
+      "learning_rate": 0.00023758884390136974,
+      "loss": 3.031,
+      "step": 34691
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5949774384498596,
+      "learning_rate": 0.00023757378464767735,
+      "loss": 3.1607,
+      "step": 34692
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5838576555252075,
+      "learning_rate": 0.00023755872555840355,
+      "loss": 2.8934,
+      "step": 34693
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6016432046890259,
+      "learning_rate": 0.00023754366663358804,
+      "loss": 2.9678,
+      "step": 34694
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6535770297050476,
+      "learning_rate": 0.00023752860787327044,
+      "loss": 3.088,
+      "step": 34695
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5867074131965637,
+      "learning_rate": 0.0002375135492774905,
+      "loss": 3.1053,
+      "step": 34696
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6075217127799988,
+      "learning_rate": 0.0002374984908462879,
+      "loss": 2.8846,
+      "step": 34697
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5963104367256165,
+      "learning_rate": 0.00023748343257970215,
+      "loss": 3.1219,
+      "step": 34698
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5905624032020569,
+      "learning_rate": 0.00023746837447777313,
+      "loss": 3.1523,
+      "step": 34699
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6578564643859863,
+      "learning_rate": 0.00023745331654054027,
+      "loss": 2.8213,
+      "step": 34700
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.609445333480835,
+      "learning_rate": 0.0002374382587680433,
+      "loss": 3.1463,
+      "step": 34701
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6245207190513611,
+      "learning_rate": 0.00023742320116032202,
+      "loss": 3.0789,
+      "step": 34702
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6319349408149719,
+      "learning_rate": 0.00023740814371741597,
+      "loss": 2.9953,
+      "step": 34703
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6266147494316101,
+      "learning_rate": 0.00023739308643936486,
+      "loss": 2.9907,
+      "step": 34704
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6235140562057495,
+      "learning_rate": 0.00023737802932620819,
+      "loss": 3.0847,
+      "step": 34705
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6102908253669739,
+      "learning_rate": 0.0002373629723779858,
+      "loss": 3.0529,
+      "step": 34706
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5892670750617981,
+      "learning_rate": 0.00023734791559473732,
+      "loss": 2.8094,
+      "step": 34707
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.658932089805603,
+      "learning_rate": 0.0002373328589765023,
+      "loss": 3.0447,
+      "step": 34708
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6005896925926208,
+      "learning_rate": 0.00023731780252332058,
+      "loss": 3.0984,
+      "step": 34709
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6177623867988586,
+      "learning_rate": 0.00023730274623523162,
+      "loss": 3.033,
+      "step": 34710
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6345233917236328,
+      "learning_rate": 0.00023728769011227509,
+      "loss": 3.0754,
+      "step": 34711
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7004926800727844,
+      "learning_rate": 0.00023727263415449087,
+      "loss": 3.1004,
+      "step": 34712
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6280829906463623,
+      "learning_rate": 0.0002372575783619183,
+      "loss": 2.9972,
+      "step": 34713
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.661109983921051,
+      "learning_rate": 0.00023724252273459736,
+      "loss": 2.9467,
+      "step": 34714
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6094497442245483,
+      "learning_rate": 0.00023722746727256744,
+      "loss": 2.869,
+      "step": 34715
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6951985359191895,
+      "learning_rate": 0.00023721241197586826,
+      "loss": 3.0284,
+      "step": 34716
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6294052004814148,
+      "learning_rate": 0.0002371973568445396,
+      "loss": 2.898,
+      "step": 34717
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6243933439254761,
+      "learning_rate": 0.00023718230187862098,
+      "loss": 2.8577,
+      "step": 34718
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6630285978317261,
+      "learning_rate": 0.00023716724707815217,
+      "loss": 2.9685,
+      "step": 34719
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6620771288871765,
+      "learning_rate": 0.00023715219244317256,
+      "loss": 3.0159,
+      "step": 34720
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6197815537452698,
+      "learning_rate": 0.00023713713797372202,
+      "loss": 2.9542,
+      "step": 34721
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6559625864028931,
+      "learning_rate": 0.0002371220836698403,
+      "loss": 2.9549,
+      "step": 34722
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6281477212905884,
+      "learning_rate": 0.00023710702953156675,
+      "loss": 2.9301,
+      "step": 34723
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6305189728736877,
+      "learning_rate": 0.00023709197555894128,
+      "loss": 2.9736,
+      "step": 34724
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6746921539306641,
+      "learning_rate": 0.0002370769217520034,
+      "loss": 2.9676,
+      "step": 34725
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6075299382209778,
+      "learning_rate": 0.00023706186811079278,
+      "loss": 2.7907,
+      "step": 34726
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.61752849817276,
+      "learning_rate": 0.00023704681463534915,
+      "loss": 3.1077,
+      "step": 34727
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5963026285171509,
+      "learning_rate": 0.00023703176132571207,
+      "loss": 2.9804,
+      "step": 34728
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6269551515579224,
+      "learning_rate": 0.00023701670818192122,
+      "loss": 2.8909,
+      "step": 34729
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6345133781433105,
+      "learning_rate": 0.00023700165520401618,
+      "loss": 3.0245,
+      "step": 34730
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6552383899688721,
+      "learning_rate": 0.00023698660239203663,
+      "loss": 2.9909,
+      "step": 34731
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6076620817184448,
+      "learning_rate": 0.00023697154974602235,
+      "loss": 3.0518,
+      "step": 34732
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6243478655815125,
+      "learning_rate": 0.00023695649726601284,
+      "loss": 2.9917,
+      "step": 34733
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6108534336090088,
+      "learning_rate": 0.00023694144495204782,
+      "loss": 2.8576,
+      "step": 34734
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.658390998840332,
+      "learning_rate": 0.00023692639280416679,
+      "loss": 3.1793,
+      "step": 34735
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.649669349193573,
+      "learning_rate": 0.00023691134082240957,
+      "loss": 3.0141,
+      "step": 34736
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6026470065116882,
+      "learning_rate": 0.0002368962890068158,
+      "loss": 2.8894,
+      "step": 34737
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5986788868904114,
+      "learning_rate": 0.00023688123735742493,
+      "loss": 2.9405,
+      "step": 34738
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6287503838539124,
+      "learning_rate": 0.00023686618587427685,
+      "loss": 2.9873,
+      "step": 34739
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.617247998714447,
+      "learning_rate": 0.00023685113455741103,
+      "loss": 2.9822,
+      "step": 34740
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6538761258125305,
+      "learning_rate": 0.00023683608340686722,
+      "loss": 3.2603,
+      "step": 34741
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5831221342086792,
+      "learning_rate": 0.00023682103242268489,
+      "loss": 3.0979,
+      "step": 34742
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6267368197441101,
+      "learning_rate": 0.00023680598160490387,
+      "loss": 2.7693,
+      "step": 34743
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6173529624938965,
+      "learning_rate": 0.0002367909309535638,
+      "loss": 2.883,
+      "step": 34744
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6151677370071411,
+      "learning_rate": 0.00023677588046870413,
+      "loss": 3.2401,
+      "step": 34745
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6226587891578674,
+      "learning_rate": 0.00023676083015036474,
+      "loss": 2.8953,
+      "step": 34746
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5849825739860535,
+      "learning_rate": 0.00023674577999858507,
+      "loss": 2.8515,
+      "step": 34747
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5755380988121033,
+      "learning_rate": 0.0002367307300134048,
+      "loss": 3.012,
+      "step": 34748
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6167582273483276,
+      "learning_rate": 0.00023671568019486376,
+      "loss": 2.9798,
+      "step": 34749
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5916376709938049,
+      "learning_rate": 0.00023670063054300134,
+      "loss": 3.0067,
+      "step": 34750
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.639107882976532,
+      "learning_rate": 0.00023668558105785736,
+      "loss": 3.067,
+      "step": 34751
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.610746443271637,
+      "learning_rate": 0.00023667053173947122,
+      "loss": 2.9683,
+      "step": 34752
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6915435194969177,
+      "learning_rate": 0.00023665548258788274,
+      "loss": 3.0443,
+      "step": 34753
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6022907495498657,
+      "learning_rate": 0.00023664043360313166,
+      "loss": 3.0091,
+      "step": 34754
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6308808326721191,
+      "learning_rate": 0.0002366253847852574,
+      "loss": 2.9871,
+      "step": 34755
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6034700870513916,
+      "learning_rate": 0.00023661033613429976,
+      "loss": 3.1442,
+      "step": 34756
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5891356468200684,
+      "learning_rate": 0.00023659528765029815,
+      "loss": 3.0425,
+      "step": 34757
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6369116902351379,
+      "learning_rate": 0.00023658023933329243,
+      "loss": 3.1113,
+      "step": 34758
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.605215311050415,
+      "learning_rate": 0.0002365651911833222,
+      "loss": 3.1997,
+      "step": 34759
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6010726094245911,
+      "learning_rate": 0.00023655014320042692,
+      "loss": 2.907,
+      "step": 34760
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.615598201751709,
+      "learning_rate": 0.00023653509538464648,
+      "loss": 2.9991,
+      "step": 34761
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6162781119346619,
+      "learning_rate": 0.00023652004773602034,
+      "loss": 3.1683,
+      "step": 34762
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6095041036605835,
+      "learning_rate": 0.0002365050002545881,
+      "loss": 2.9159,
+      "step": 34763
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5748782753944397,
+      "learning_rate": 0.0002364899529403896,
+      "loss": 2.7686,
+      "step": 34764
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6566309332847595,
+      "learning_rate": 0.00023647490579346432,
+      "loss": 2.8777,
+      "step": 34765
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5985787510871887,
+      "learning_rate": 0.0002364598588138519,
+      "loss": 2.9995,
+      "step": 34766
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6187407374382019,
+      "learning_rate": 0.0002364448120015919,
+      "loss": 2.8916,
+      "step": 34767
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6135745644569397,
+      "learning_rate": 0.00023642976535672407,
+      "loss": 3.0603,
+      "step": 34768
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6574584245681763,
+      "learning_rate": 0.0002364147188792881,
+      "loss": 2.9804,
+      "step": 34769
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6308838725090027,
+      "learning_rate": 0.00023639967256932338,
+      "loss": 2.9685,
+      "step": 34770
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6589849591255188,
+      "learning_rate": 0.00023638462642686984,
+      "loss": 3.1744,
+      "step": 34771
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6437431573867798,
+      "learning_rate": 0.00023636958045196676,
+      "loss": 2.9364,
+      "step": 34772
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6591246128082275,
+      "learning_rate": 0.00023635453464465408,
+      "loss": 3.0042,
+      "step": 34773
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6104686260223389,
+      "learning_rate": 0.00023633948900497135,
+      "loss": 2.9775,
+      "step": 34774
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6379290819168091,
+      "learning_rate": 0.00023632444353295802,
+      "loss": 2.9928,
+      "step": 34775
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6709439158439636,
+      "learning_rate": 0.00023630939822865397,
+      "loss": 2.8491,
+      "step": 34776
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6231077909469604,
+      "learning_rate": 0.00023629435309209865,
+      "loss": 2.9117,
+      "step": 34777
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6265411376953125,
+      "learning_rate": 0.00023627930812333172,
+      "loss": 3.1632,
+      "step": 34778
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6154355406761169,
+      "learning_rate": 0.00023626426332239292,
+      "loss": 3.1184,
+      "step": 34779
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6437824964523315,
+      "learning_rate": 0.0002362492186893217,
+      "loss": 2.844,
+      "step": 34780
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6491339206695557,
+      "learning_rate": 0.00023623417422415788,
+      "loss": 2.8048,
+      "step": 34781
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6449874639511108,
+      "learning_rate": 0.0002362191299269408,
+      "loss": 3.0313,
+      "step": 34782
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6093539595603943,
+      "learning_rate": 0.00023620408579771038,
+      "loss": 3.1829,
+      "step": 34783
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6419656872749329,
+      "learning_rate": 0.00023618904183650614,
+      "loss": 3.3079,
+      "step": 34784
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6378693580627441,
+      "learning_rate": 0.00023617399804336755,
+      "loss": 2.9831,
+      "step": 34785
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6107136607170105,
+      "learning_rate": 0.0002361589544183345,
+      "loss": 2.9476,
+      "step": 34786
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6500281691551208,
+      "learning_rate": 0.0002361439109614464,
+      "loss": 3.0999,
+      "step": 34787
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5922091007232666,
+      "learning_rate": 0.00023612886767274293,
+      "loss": 3.0714,
+      "step": 34788
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6356111764907837,
+      "learning_rate": 0.00023611382455226385,
+      "loss": 2.8917,
+      "step": 34789
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6688171029090881,
+      "learning_rate": 0.00023609878160004856,
+      "loss": 3.0792,
+      "step": 34790
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6286024451255798,
+      "learning_rate": 0.00023608373881613685,
+      "loss": 3.1178,
+      "step": 34791
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6304290294647217,
+      "learning_rate": 0.00023606869620056817,
+      "loss": 2.9926,
+      "step": 34792
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6616683006286621,
+      "learning_rate": 0.0002360536537533822,
+      "loss": 2.761,
+      "step": 34793
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5931280851364136,
+      "learning_rate": 0.0002360386114746188,
+      "loss": 2.9022,
+      "step": 34794
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6022833585739136,
+      "learning_rate": 0.00023602356936431728,
+      "loss": 3.0123,
+      "step": 34795
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6116443276405334,
+      "learning_rate": 0.00023600852742251742,
+      "loss": 2.9503,
+      "step": 34796
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.651864230632782,
+      "learning_rate": 0.00023599348564925867,
+      "loss": 2.966,
+      "step": 34797
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6232309341430664,
+      "learning_rate": 0.0002359784440445808,
+      "loss": 2.9551,
+      "step": 34798
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5941501259803772,
+      "learning_rate": 0.00023596340260852348,
+      "loss": 3.1523,
+      "step": 34799
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6085199117660522,
+      "learning_rate": 0.0002359483613411261,
+      "loss": 2.8691,
+      "step": 34800
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6391234397888184,
+      "learning_rate": 0.00023593332024242854,
+      "loss": 2.9873,
+      "step": 34801
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6489654183387756,
+      "learning_rate": 0.0002359182793124702,
+      "loss": 2.9056,
+      "step": 34802
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6384288668632507,
+      "learning_rate": 0.0002359032385512907,
+      "loss": 3.1661,
+      "step": 34803
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.660001814365387,
+      "learning_rate": 0.00023588819795892994,
+      "loss": 3.021,
+      "step": 34804
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6211850047111511,
+      "learning_rate": 0.0002358731575354272,
+      "loss": 3.0002,
+      "step": 34805
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5897373557090759,
+      "learning_rate": 0.0002358581172808223,
+      "loss": 3.1594,
+      "step": 34806
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5892007350921631,
+      "learning_rate": 0.00023584307719515469,
+      "loss": 2.7936,
+      "step": 34807
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5857629776000977,
+      "learning_rate": 0.00023582803727846405,
+      "loss": 2.9798,
+      "step": 34808
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6502203941345215,
+      "learning_rate": 0.0002358129975307901,
+      "loss": 3.3007,
+      "step": 34809
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6255640983581543,
+      "learning_rate": 0.00023579795795217225,
+      "loss": 2.9249,
+      "step": 34810
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6112305521965027,
+      "learning_rate": 0.00023578291854265033,
+      "loss": 3.198,
+      "step": 34811
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6151379346847534,
+      "learning_rate": 0.00023576787930226385,
+      "loss": 2.7124,
+      "step": 34812
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6541122794151306,
+      "learning_rate": 0.00023575284023105229,
+      "loss": 2.8483,
+      "step": 34813
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6102315187454224,
+      "learning_rate": 0.0002357378013290555,
+      "loss": 2.942,
+      "step": 34814
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6251721382141113,
+      "learning_rate": 0.00023572276259631287,
+      "loss": 3.1909,
+      "step": 34815
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5804009437561035,
+      "learning_rate": 0.00023570772403286425,
+      "loss": 3.049,
+      "step": 34816
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6197915077209473,
+      "learning_rate": 0.00023569268563874903,
+      "loss": 3.0851,
+      "step": 34817
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.605315089225769,
+      "learning_rate": 0.00023567764741400682,
+      "loss": 3.1499,
+      "step": 34818
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6538598537445068,
+      "learning_rate": 0.00023566260935867746,
+      "loss": 3.0767,
+      "step": 34819
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6209130883216858,
+      "learning_rate": 0.00023564757147280033,
+      "loss": 3.1515,
+      "step": 34820
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6308000087738037,
+      "learning_rate": 0.00023563253375641517,
+      "loss": 3.0422,
+      "step": 34821
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5922633409500122,
+      "learning_rate": 0.00023561749620956143,
+      "loss": 3.0058,
+      "step": 34822
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6279627084732056,
+      "learning_rate": 0.00023560245883227893,
+      "loss": 2.9634,
+      "step": 34823
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.611582338809967,
+      "learning_rate": 0.00023558742162460707,
+      "loss": 2.9571,
+      "step": 34824
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5956346392631531,
+      "learning_rate": 0.00023557238458658553,
+      "loss": 3.0411,
+      "step": 34825
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6095432639122009,
+      "learning_rate": 0.00023555734771825402,
+      "loss": 3.1169,
+      "step": 34826
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5974093079566956,
+      "learning_rate": 0.000235542311019652,
+      "loss": 2.9938,
+      "step": 34827
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.696919322013855,
+      "learning_rate": 0.0002355272744908192,
+      "loss": 3.144,
+      "step": 34828
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6090503931045532,
+      "learning_rate": 0.00023551223813179502,
+      "loss": 3.1853,
+      "step": 34829
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6434307098388672,
+      "learning_rate": 0.00023549720194261924,
+      "loss": 3.0533,
+      "step": 34830
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6265668272972107,
+      "learning_rate": 0.00023548216592333152,
+      "loss": 2.8724,
+      "step": 34831
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6203542947769165,
+      "learning_rate": 0.00023546713007397115,
+      "loss": 3.1396,
+      "step": 34832
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6344203948974609,
+      "learning_rate": 0.00023545209439457813,
+      "loss": 3.0088,
+      "step": 34833
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6461747288703918,
+      "learning_rate": 0.00023543705888519174,
+      "loss": 2.848,
+      "step": 34834
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6251636743545532,
+      "learning_rate": 0.0002354220235458518,
+      "loss": 2.8499,
+      "step": 34835
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6171122789382935,
+      "learning_rate": 0.00023540698837659784,
+      "loss": 2.9701,
+      "step": 34836
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6390722990036011,
+      "learning_rate": 0.0002353919533774693,
+      "loss": 3.0639,
+      "step": 34837
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6113898158073425,
+      "learning_rate": 0.0002353769185485061,
+      "loss": 3.0786,
+      "step": 34838
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.640236496925354,
+      "learning_rate": 0.00023536188388974757,
+      "loss": 2.9315,
+      "step": 34839
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6200087666511536,
+      "learning_rate": 0.00023534684940123332,
+      "loss": 3.1699,
+      "step": 34840
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6289457678794861,
+      "learning_rate": 0.00023533181508300318,
+      "loss": 3.1179,
+      "step": 34841
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6203035116195679,
+      "learning_rate": 0.00023531678093509655,
+      "loss": 3.1601,
+      "step": 34842
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6237162351608276,
+      "learning_rate": 0.00023530174695755306,
+      "loss": 3.0658,
+      "step": 34843
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6613100171089172,
+      "learning_rate": 0.00023528671315041226,
+      "loss": 2.8084,
+      "step": 34844
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6259462833404541,
+      "learning_rate": 0.00023527167951371383,
+      "loss": 3.0736,
+      "step": 34845
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6277626156806946,
+      "learning_rate": 0.0002352566460474974,
+      "loss": 3.0055,
+      "step": 34846
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6393147110939026,
+      "learning_rate": 0.0002352416127518024,
+      "loss": 3.144,
+      "step": 34847
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6422019600868225,
+      "learning_rate": 0.00023522657962666866,
+      "loss": 3.0748,
+      "step": 34848
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6111181378364563,
+      "learning_rate": 0.00023521154667213555,
+      "loss": 3.024,
+      "step": 34849
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6628186702728271,
+      "learning_rate": 0.0002351965138882427,
+      "loss": 3.0101,
+      "step": 34850
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6172040700912476,
+      "learning_rate": 0.0002351814812750299,
+      "loss": 3.1652,
+      "step": 34851
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6279569268226624,
+      "learning_rate": 0.00023516644883253654,
+      "loss": 3.0395,
+      "step": 34852
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6078111529350281,
+      "learning_rate": 0.00023515141656080236,
+      "loss": 2.9895,
+      "step": 34853
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6228066682815552,
+      "learning_rate": 0.0002351363844598667,
+      "loss": 2.9287,
+      "step": 34854
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7036370038986206,
+      "learning_rate": 0.00023512135252976935,
+      "loss": 3.1124,
+      "step": 34855
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5945392847061157,
+      "learning_rate": 0.00023510632077054997,
+      "loss": 2.9471,
+      "step": 34856
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6402318477630615,
+      "learning_rate": 0.00023509128918224802,
+      "loss": 3.1473,
+      "step": 34857
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6327133774757385,
+      "learning_rate": 0.00023507625776490317,
+      "loss": 2.855,
+      "step": 34858
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6053986549377441,
+      "learning_rate": 0.0002350612265185548,
+      "loss": 3.0087,
+      "step": 34859
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5792701244354248,
+      "learning_rate": 0.00023504619544324277,
+      "loss": 2.8177,
+      "step": 34860
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6614009141921997,
+      "learning_rate": 0.00023503116453900663,
+      "loss": 3.218,
+      "step": 34861
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6756806969642639,
+      "learning_rate": 0.00023501613380588575,
+      "loss": 3.1788,
+      "step": 34862
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6036829352378845,
+      "learning_rate": 0.00023500110324392004,
+      "loss": 3.0215,
+      "step": 34863
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6031563878059387,
+      "learning_rate": 0.0002349860728531488,
+      "loss": 3.146,
+      "step": 34864
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6657361388206482,
+      "learning_rate": 0.00023497104263361167,
+      "loss": 3.0158,
+      "step": 34865
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6230410933494568,
+      "learning_rate": 0.00023495601258534843,
+      "loss": 2.758,
+      "step": 34866
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.598953127861023,
+      "learning_rate": 0.00023494098270839849,
+      "loss": 3.1044,
+      "step": 34867
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6193811893463135,
+      "learning_rate": 0.00023492595300280152,
+      "loss": 2.843,
+      "step": 34868
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6279089450836182,
+      "learning_rate": 0.00023491092346859693,
+      "loss": 3.1108,
+      "step": 34869
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6103789806365967,
+      "learning_rate": 0.00023489589410582455,
+      "loss": 3.1877,
+      "step": 34870
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6290521621704102,
+      "learning_rate": 0.0002348808649145239,
+      "loss": 3.0293,
+      "step": 34871
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5811523795127869,
+      "learning_rate": 0.00023486583589473438,
+      "loss": 3.0623,
+      "step": 34872
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6363834738731384,
+      "learning_rate": 0.00023485080704649586,
+      "loss": 2.9667,
+      "step": 34873
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6560509204864502,
+      "learning_rate": 0.0002348357783698477,
+      "loss": 2.975,
+      "step": 34874
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6376746892929077,
+      "learning_rate": 0.0002348207498648295,
+      "loss": 3.156,
+      "step": 34875
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6634820699691772,
+      "learning_rate": 0.00023480572153148103,
+      "loss": 2.9336,
+      "step": 34876
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6263517141342163,
+      "learning_rate": 0.00023479069336984157,
+      "loss": 3.0352,
+      "step": 34877
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6027346253395081,
+      "learning_rate": 0.00023477566537995106,
+      "loss": 2.884,
+      "step": 34878
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6649355292320251,
+      "learning_rate": 0.0002347606375618488,
+      "loss": 2.9949,
+      "step": 34879
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6449239253997803,
+      "learning_rate": 0.00023474560991557442,
+      "loss": 3.0944,
+      "step": 34880
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6204206943511963,
+      "learning_rate": 0.00023473058244116767,
+      "loss": 2.8473,
+      "step": 34881
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6074206829071045,
+      "learning_rate": 0.00023471555513866795,
+      "loss": 2.8826,
+      "step": 34882
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.607448935508728,
+      "learning_rate": 0.00023470052800811495,
+      "loss": 2.9217,
+      "step": 34883
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6031710505485535,
+      "learning_rate": 0.00023468550104954804,
+      "loss": 2.9217,
+      "step": 34884
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5931732654571533,
+      "learning_rate": 0.00023467047426300704,
+      "loss": 2.8402,
+      "step": 34885
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6108601689338684,
+      "learning_rate": 0.00023465544764853148,
+      "loss": 3.0791,
+      "step": 34886
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6241343021392822,
+      "learning_rate": 0.0002346404212061608,
+      "loss": 3.0006,
+      "step": 34887
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6183168292045593,
+      "learning_rate": 0.00023462539493593478,
+      "loss": 3.1661,
+      "step": 34888
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6017194986343384,
+      "learning_rate": 0.00023461036883789282,
+      "loss": 3.0034,
+      "step": 34889
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.659395694732666,
+      "learning_rate": 0.00023459534291207448,
+      "loss": 3.0346,
+      "step": 34890
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6340118646621704,
+      "learning_rate": 0.00023458031715851957,
+      "loss": 3.1279,
+      "step": 34891
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6464155912399292,
+      "learning_rate": 0.00023456529157726746,
+      "loss": 2.861,
+      "step": 34892
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6099604368209839,
+      "learning_rate": 0.00023455026616835784,
+      "loss": 3.066,
+      "step": 34893
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6244305968284607,
+      "learning_rate": 0.00023453524093183009,
+      "loss": 2.9749,
+      "step": 34894
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5913376212120056,
+      "learning_rate": 0.00023452021586772392,
+      "loss": 3.099,
+      "step": 34895
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6622759699821472,
+      "learning_rate": 0.000234505190976079,
+      "loss": 3.0457,
+      "step": 34896
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6322427988052368,
+      "learning_rate": 0.00023449016625693477,
+      "loss": 3.0076,
+      "step": 34897
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6200683116912842,
+      "learning_rate": 0.0002344751417103309,
+      "loss": 2.9269,
+      "step": 34898
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6351677775382996,
+      "learning_rate": 0.00023446011733630674,
+      "loss": 2.9341,
+      "step": 34899
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6051552891731262,
+      "learning_rate": 0.0002344450931349022,
+      "loss": 3.1266,
+      "step": 34900
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.657870352268219,
+      "learning_rate": 0.00023443006910615652,
+      "loss": 3.2018,
+      "step": 34901
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6787803173065186,
+      "learning_rate": 0.00023441504525010943,
+      "loss": 3.0304,
+      "step": 34902
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6219592690467834,
+      "learning_rate": 0.0002344000215668006,
+      "loss": 3.2,
+      "step": 34903
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6439167857170105,
+      "learning_rate": 0.0002343849980562694,
+      "loss": 3.0774,
+      "step": 34904
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6104590892791748,
+      "learning_rate": 0.00023436997471855558,
+      "loss": 3.0582,
+      "step": 34905
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6476345062255859,
+      "learning_rate": 0.00023435495155369848,
+      "loss": 3.014,
+      "step": 34906
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6276230812072754,
+      "learning_rate": 0.00023433992856173786,
+      "loss": 3.0314,
+      "step": 34907
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6397703886032104,
+      "learning_rate": 0.00023432490574271328,
+      "loss": 2.9595,
+      "step": 34908
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6316772699356079,
+      "learning_rate": 0.00023430988309666414,
+      "loss": 2.9005,
+      "step": 34909
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6077426671981812,
+      "learning_rate": 0.00023429486062363027,
+      "loss": 2.9963,
+      "step": 34910
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6244196891784668,
+      "learning_rate": 0.00023427983832365106,
+      "loss": 3.0356,
+      "step": 34911
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6547126770019531,
+      "learning_rate": 0.00023426481619676597,
+      "loss": 2.9158,
+      "step": 34912
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5975217223167419,
+      "learning_rate": 0.00023424979424301488,
+      "loss": 3.005,
+      "step": 34913
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6610994935035706,
+      "learning_rate": 0.00023423477246243708,
+      "loss": 2.8519,
+      "step": 34914
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6371542811393738,
+      "learning_rate": 0.00023421975085507232,
+      "loss": 2.9664,
+      "step": 34915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6088597178459167,
+      "learning_rate": 0.00023420472942095995,
+      "loss": 2.9404,
+      "step": 34916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.654114305973053,
+      "learning_rate": 0.00023418970816013965,
+      "loss": 2.9306,
+      "step": 34917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.637015700340271,
+      "learning_rate": 0.0002341746870726511,
+      "loss": 2.8428,
+      "step": 34918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6335122585296631,
+      "learning_rate": 0.00023415966615853372,
+      "loss": 2.9795,
+      "step": 34919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6221902966499329,
+      "learning_rate": 0.0002341446454178272,
+      "loss": 3.0404,
+      "step": 34920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6287366151809692,
+      "learning_rate": 0.00023412962485057086,
+      "loss": 3.1862,
+      "step": 34921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6433752775192261,
+      "learning_rate": 0.00023411460445680444,
+      "loss": 3.119,
+      "step": 34922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6408416032791138,
+      "learning_rate": 0.00023409958423656755,
+      "loss": 3.0431,
+      "step": 34923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6352258920669556,
+      "learning_rate": 0.00023408456418989957,
+      "loss": 2.9922,
+      "step": 34924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6453670263290405,
+      "learning_rate": 0.00023406954431684027,
+      "loss": 3.1621,
+      "step": 34925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5970926284790039,
+      "learning_rate": 0.00023405452461742904,
+      "loss": 2.9972,
+      "step": 34926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6512788534164429,
+      "learning_rate": 0.00023403950509170543,
+      "loss": 3.1177,
+      "step": 34927
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6225276589393616,
+      "learning_rate": 0.00023402448573970925,
+      "loss": 2.9748,
+      "step": 34928
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6059454083442688,
+      "learning_rate": 0.00023400946656147977,
+      "loss": 3.0446,
+      "step": 34929
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6157578825950623,
+      "learning_rate": 0.0002339944475570567,
+      "loss": 3.1555,
+      "step": 34930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6617079973220825,
+      "learning_rate": 0.00023397942872647944,
+      "loss": 2.9416,
+      "step": 34931
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6348146200180054,
+      "learning_rate": 0.00023396441006978775,
+      "loss": 3.1802,
+      "step": 34932
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6032415628433228,
+      "learning_rate": 0.00023394939158702112,
+      "loss": 2.9522,
+      "step": 34933
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6263419985771179,
+      "learning_rate": 0.00023393437327821895,
+      "loss": 2.8259,
+      "step": 34934
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6232619285583496,
+      "learning_rate": 0.0002339193551434211,
+      "loss": 2.8977,
+      "step": 34935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6191720962524414,
+      "learning_rate": 0.00023390433718266685,
+      "loss": 2.9033,
+      "step": 34936
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7007677555084229,
+      "learning_rate": 0.0002338893193959958,
+      "loss": 3.1077,
+      "step": 34937
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6107507944107056,
+      "learning_rate": 0.0002338743017834477,
+      "loss": 3.1321,
+      "step": 34938
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6792271137237549,
+      "learning_rate": 0.00023385928434506182,
+      "loss": 2.9655,
+      "step": 34939
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.645459771156311,
+      "learning_rate": 0.00023384426708087798,
+      "loss": 3.0854,
+      "step": 34940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6554004549980164,
+      "learning_rate": 0.00023382924999093555,
+      "loss": 2.7691,
+      "step": 34941
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6682091355323792,
+      "learning_rate": 0.00023381423307527408,
+      "loss": 2.9499,
+      "step": 34942
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.653188169002533,
+      "learning_rate": 0.0002337992163339333,
+      "loss": 3.0563,
+      "step": 34943
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6153813600540161,
+      "learning_rate": 0.0002337841997669526,
+      "loss": 2.972,
+      "step": 34944
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6357757449150085,
+      "learning_rate": 0.00023376918337437164,
+      "loss": 2.9628,
+      "step": 34945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6298278570175171,
+      "learning_rate": 0.00023375416715622977,
+      "loss": 2.7598,
+      "step": 34946
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.581403911113739,
+      "learning_rate": 0.00023373915111256675,
+      "loss": 2.8313,
+      "step": 34947
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6330929398536682,
+      "learning_rate": 0.00023372413524342208,
+      "loss": 2.9305,
+      "step": 34948
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6617899537086487,
+      "learning_rate": 0.00023370911954883518,
+      "loss": 2.8998,
+      "step": 34949
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6223428249359131,
+      "learning_rate": 0.00023369410402884582,
+      "loss": 3.0054,
+      "step": 34950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6279831528663635,
+      "learning_rate": 0.00023367908868349337,
+      "loss": 3.0671,
+      "step": 34951
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5640294551849365,
+      "learning_rate": 0.00023366407351281735,
+      "loss": 3.0893,
+      "step": 34952
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5992966294288635,
+      "learning_rate": 0.0002336490585168576,
+      "loss": 3.0811,
+      "step": 34953
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6405746340751648,
+      "learning_rate": 0.00023363404369565328,
+      "loss": 3.0204,
+      "step": 34954
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.651674747467041,
+      "learning_rate": 0.00023361902904924428,
+      "loss": 2.7941,
+      "step": 34955
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6315301656723022,
+      "learning_rate": 0.0002336040145776698,
+      "loss": 2.9018,
+      "step": 34956
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6295512914657593,
+      "learning_rate": 0.00023358900028096959,
+      "loss": 3.013,
+      "step": 34957
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5874469876289368,
+      "learning_rate": 0.0002335739861591833,
+      "loss": 3.203,
+      "step": 34958
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6032001376152039,
+      "learning_rate": 0.00023355897221235026,
+      "loss": 3.0261,
+      "step": 34959
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6361725330352783,
+      "learning_rate": 0.00023354395844051018,
+      "loss": 3.0568,
+      "step": 34960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6287243962287903,
+      "learning_rate": 0.0002335289448437024,
+      "loss": 3.1561,
+      "step": 34961
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6232472062110901,
+      "learning_rate": 0.00023351393142196665,
+      "loss": 3.0336,
+      "step": 34962
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6193660497665405,
+      "learning_rate": 0.00023349891817534246,
+      "loss": 2.7832,
+      "step": 34963
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6325574517250061,
+      "learning_rate": 0.00023348390510386917,
+      "loss": 2.9592,
+      "step": 34964
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6106416583061218,
+      "learning_rate": 0.00023346889220758667,
+      "loss": 3.0528,
+      "step": 34965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6092174053192139,
+      "learning_rate": 0.0002334538794865342,
+      "loss": 2.9849,
+      "step": 34966
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6205219030380249,
+      "learning_rate": 0.00023343886694075135,
+      "loss": 2.9281,
+      "step": 34967
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6163828372955322,
+      "learning_rate": 0.00023342385457027781,
+      "loss": 2.7753,
+      "step": 34968
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6253390312194824,
+      "learning_rate": 0.000233408842375153,
+      "loss": 2.8628,
+      "step": 34969
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.650109589099884,
+      "learning_rate": 0.00023339383035541654,
+      "loss": 2.9486,
+      "step": 34970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6107782125473022,
+      "learning_rate": 0.00023337881851110776,
+      "loss": 2.8765,
+      "step": 34971
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6146960854530334,
+      "learning_rate": 0.00023336380684226644,
+      "loss": 3.0767,
+      "step": 34972
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6434600949287415,
+      "learning_rate": 0.0002333487953489321,
+      "loss": 3.0933,
+      "step": 34973
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5741561055183411,
+      "learning_rate": 0.00023333378403114405,
+      "loss": 3.2363,
+      "step": 34974
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6086795330047607,
+      "learning_rate": 0.00023331877288894213,
+      "loss": 2.7989,
+      "step": 34975
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5960971713066101,
+      "learning_rate": 0.00023330376192236564,
+      "loss": 3.1249,
+      "step": 34976
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6542201638221741,
+      "learning_rate": 0.00023328875113145432,
+      "loss": 2.8669,
+      "step": 34977
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6497222185134888,
+      "learning_rate": 0.0002332737405162474,
+      "loss": 3.1174,
+      "step": 34978
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6587827205657959,
+      "learning_rate": 0.00023325873007678464,
+      "loss": 2.9742,
+      "step": 34979
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6213930249214172,
+      "learning_rate": 0.0002332437198131057,
+      "loss": 3.1319,
+      "step": 34980
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6171470880508423,
+      "learning_rate": 0.00023322870972524983,
+      "loss": 2.9525,
+      "step": 34981
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6292840242385864,
+      "learning_rate": 0.00023321369981325676,
+      "loss": 3.0539,
+      "step": 34982
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6087939143180847,
+      "learning_rate": 0.00023319869007716587,
+      "loss": 3.0538,
+      "step": 34983
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6223199367523193,
+      "learning_rate": 0.0002331836805170168,
+      "loss": 2.8537,
+      "step": 34984
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6605082154273987,
+      "learning_rate": 0.00023316867113284913,
+      "loss": 3.0738,
+      "step": 34985
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6185541749000549,
+      "learning_rate": 0.0002331536619247022,
+      "loss": 2.9984,
+      "step": 34986
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.631881594657898,
+      "learning_rate": 0.00023313865289261576,
+      "loss": 2.9922,
+      "step": 34987
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.618061363697052,
+      "learning_rate": 0.0002331236440366292,
+      "loss": 2.996,
+      "step": 34988
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.609032928943634,
+      "learning_rate": 0.00023310863535678202,
+      "loss": 2.6868,
+      "step": 34989
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6071302890777588,
+      "learning_rate": 0.00023309362685311396,
+      "loss": 3.0175,
+      "step": 34990
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6098959445953369,
+      "learning_rate": 0.00023307861852566434,
+      "loss": 2.9696,
+      "step": 34991
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.638105034828186,
+      "learning_rate": 0.00023306361037447282,
+      "loss": 2.9288,
+      "step": 34992
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6155096292495728,
+      "learning_rate": 0.00023304860239957873,
+      "loss": 3.1519,
+      "step": 34993
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.628525972366333,
+      "learning_rate": 0.0002330335946010218,
+      "loss": 2.9343,
+      "step": 34994
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7091739177703857,
+      "learning_rate": 0.00023301858697884157,
+      "loss": 3.1016,
+      "step": 34995
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.62142014503479,
+      "learning_rate": 0.0002330035795330774,
+      "loss": 2.8768,
+      "step": 34996
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6097868084907532,
+      "learning_rate": 0.00023298857226376895,
+      "loss": 3.0771,
+      "step": 34997
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6286457777023315,
+      "learning_rate": 0.00023297356517095562,
+      "loss": 3.0466,
+      "step": 34998
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6783491373062134,
+      "learning_rate": 0.00023295855825467712,
+      "loss": 3.2109,
+      "step": 34999
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7100011110305786,
+      "learning_rate": 0.00023294355151497287,
+      "loss": 2.953,
+      "step": 35000
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5850964188575745,
+      "learning_rate": 0.00023292854495188232,
+      "loss": 2.9883,
+      "step": 35001
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6173315048217773,
+      "learning_rate": 0.0002329135385654452,
+      "loss": 3.0854,
+      "step": 35002
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6593319773674011,
+      "learning_rate": 0.00023289853235570083,
+      "loss": 2.7224,
+      "step": 35003
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6051262617111206,
+      "learning_rate": 0.00023288352632268877,
+      "loss": 3.0075,
+      "step": 35004
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6009809970855713,
+      "learning_rate": 0.00023286852046644867,
+      "loss": 2.7894,
+      "step": 35005
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5898076295852661,
+      "learning_rate": 0.00023285351478701994,
+      "loss": 3.0432,
+      "step": 35006
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6174237728118896,
+      "learning_rate": 0.0002328385092844422,
+      "loss": 2.9945,
+      "step": 35007
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6520997881889343,
+      "learning_rate": 0.00023282350395875472,
+      "loss": 2.9187,
+      "step": 35008
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6250781416893005,
+      "learning_rate": 0.0002328084988099973,
+      "loss": 3.1153,
+      "step": 35009
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6104256510734558,
+      "learning_rate": 0.00023279349383820944,
+      "loss": 3.1432,
+      "step": 35010
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6256521344184875,
+      "learning_rate": 0.00023277848904343045,
+      "loss": 2.8728,
+      "step": 35011
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5990914702415466,
+      "learning_rate": 0.0002327634844257001,
+      "loss": 3.0632,
+      "step": 35012
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6210713982582092,
+      "learning_rate": 0.00023274847998505774,
+      "loss": 2.9762,
+      "step": 35013
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6181674003601074,
+      "learning_rate": 0.00023273347572154288,
+      "loss": 3.0125,
+      "step": 35014
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8895490169525146,
+      "learning_rate": 0.0002327184716351952,
+      "loss": 2.9884,
+      "step": 35015
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6275051236152649,
+      "learning_rate": 0.00023270346772605408,
+      "loss": 2.9187,
+      "step": 35016
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5979534387588501,
+      "learning_rate": 0.0002326884639941591,
+      "loss": 2.9628,
+      "step": 35017
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6114262938499451,
+      "learning_rate": 0.0002326734604395497,
+      "loss": 2.9154,
+      "step": 35018
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6187745332717896,
+      "learning_rate": 0.00023265845706226535,
+      "loss": 2.9844,
+      "step": 35019
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5852953791618347,
+      "learning_rate": 0.00023264345386234586,
+      "loss": 3.0157,
+      "step": 35020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5987752676010132,
+      "learning_rate": 0.00023262845083983047,
+      "loss": 3.0215,
+      "step": 35021
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6052468419075012,
+      "learning_rate": 0.00023261344799475886,
+      "loss": 2.8387,
+      "step": 35022
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.653300940990448,
+      "learning_rate": 0.00023259844532717028,
+      "loss": 3.0387,
+      "step": 35023
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.645679235458374,
+      "learning_rate": 0.00023258344283710452,
+      "loss": 2.861,
+      "step": 35024
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6186267137527466,
+      "learning_rate": 0.00023256844052460106,
+      "loss": 2.9967,
+      "step": 35025
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6053884029388428,
+      "learning_rate": 0.0002325534383896992,
+      "loss": 3.0525,
+      "step": 35026
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6017190217971802,
+      "learning_rate": 0.00023253843643243876,
+      "loss": 2.8209,
+      "step": 35027
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6449417471885681,
+      "learning_rate": 0.000232523434652859,
+      "loss": 3.0937,
+      "step": 35028
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6535047888755798,
+      "learning_rate": 0.0002325084330509995,
+      "loss": 3.1532,
+      "step": 35029
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6600343585014343,
+      "learning_rate": 0.0002324934316268999,
+      "loss": 3.0011,
+      "step": 35030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6052358150482178,
+      "learning_rate": 0.00023247843038059956,
+      "loss": 3.2281,
+      "step": 35031
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6027965545654297,
+      "learning_rate": 0.00023246342931213809,
+      "loss": 2.9464,
+      "step": 35032
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6339533925056458,
+      "learning_rate": 0.00023244842842155483,
+      "loss": 3.0856,
+      "step": 35033
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6081569194793701,
+      "learning_rate": 0.00023243342770888948,
+      "loss": 2.8734,
+      "step": 35034
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6404123902320862,
+      "learning_rate": 0.00023241842717418152,
+      "loss": 3.113,
+      "step": 35035
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6517124176025391,
+      "learning_rate": 0.00023240342681747031,
+      "loss": 3.0599,
+      "step": 35036
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6136491894721985,
+      "learning_rate": 0.00023238842663879557,
+      "loss": 2.9569,
+      "step": 35037
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.585811972618103,
+      "learning_rate": 0.00023237342663819666,
+      "loss": 2.9896,
+      "step": 35038
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6353388428688049,
+      "learning_rate": 0.00023235842681571306,
+      "loss": 2.8659,
+      "step": 35039
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.625332236289978,
+      "learning_rate": 0.00023234342717138448,
+      "loss": 3.2114,
+      "step": 35040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6419073939323425,
+      "learning_rate": 0.00023232842770525014,
+      "loss": 2.8107,
+      "step": 35041
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6179407835006714,
+      "learning_rate": 0.00023231342841734985,
+      "loss": 2.9328,
+      "step": 35042
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6118654608726501,
+      "learning_rate": 0.0002322984293077229,
+      "loss": 2.9489,
+      "step": 35043
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6000199913978577,
+      "learning_rate": 0.00023228343037640879,
+      "loss": 3.0466,
+      "step": 35044
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6850727796554565,
+      "learning_rate": 0.0002322684316234472,
+      "loss": 3.1185,
+      "step": 35045
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6070858240127563,
+      "learning_rate": 0.0002322534330488775,
+      "loss": 3.0039,
+      "step": 35046
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7080966830253601,
+      "learning_rate": 0.00023223843465273924,
+      "loss": 3.0643,
+      "step": 35047
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6055739521980286,
+      "learning_rate": 0.00023222343643507177,
+      "loss": 2.9921,
+      "step": 35048
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6497558355331421,
+      "learning_rate": 0.00023220843839591484,
+      "loss": 2.9583,
+      "step": 35049
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6094196438789368,
+      "learning_rate": 0.00023219344053530786,
+      "loss": 2.7776,
+      "step": 35050
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6469390988349915,
+      "learning_rate": 0.0002321784428532902,
+      "loss": 3.0497,
+      "step": 35051
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6989441514015198,
+      "learning_rate": 0.0002321634453499016,
+      "loss": 3.1574,
+      "step": 35052
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6314229965209961,
+      "learning_rate": 0.00023214844802518135,
+      "loss": 2.9443,
+      "step": 35053
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6381677389144897,
+      "learning_rate": 0.0002321334508791691,
+      "loss": 2.9694,
+      "step": 35054
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6836425065994263,
+      "learning_rate": 0.00023211845391190415,
+      "loss": 2.8559,
+      "step": 35055
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6240857243537903,
+      "learning_rate": 0.00023210345712342619,
+      "loss": 2.7767,
+      "step": 35056
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6270095705986023,
+      "learning_rate": 0.00023208846051377474,
+      "loss": 2.999,
+      "step": 35057
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6575677990913391,
+      "learning_rate": 0.00023207346408298906,
+      "loss": 2.943,
+      "step": 35058
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6204233765602112,
+      "learning_rate": 0.00023205846783110894,
+      "loss": 3.0342,
+      "step": 35059
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6046788692474365,
+      "learning_rate": 0.00023204347175817362,
+      "loss": 2.9394,
+      "step": 35060
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6025890707969666,
+      "learning_rate": 0.0002320284758642228,
+      "loss": 2.9541,
+      "step": 35061
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5898043513298035,
+      "learning_rate": 0.0002320134801492959,
+      "loss": 3.1216,
+      "step": 35062
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6248551607131958,
+      "learning_rate": 0.0002319984846134323,
+      "loss": 2.9211,
+      "step": 35063
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.677790641784668,
+      "learning_rate": 0.00023198348925667177,
+      "loss": 2.8475,
+      "step": 35064
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6180427074432373,
+      "learning_rate": 0.00023196849407905353,
+      "loss": 3.1241,
+      "step": 35065
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6461980938911438,
+      "learning_rate": 0.00023195349908061716,
+      "loss": 3.1005,
+      "step": 35066
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6379979848861694,
+      "learning_rate": 0.0002319385042614023,
+      "loss": 2.9287,
+      "step": 35067
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.631192684173584,
+      "learning_rate": 0.0002319235096214482,
+      "loss": 3.0211,
+      "step": 35068
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6260473728179932,
+      "learning_rate": 0.00023190851516079463,
+      "loss": 2.9484,
+      "step": 35069
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6406779885292053,
+      "learning_rate": 0.0002318935208794807,
+      "loss": 2.9191,
+      "step": 35070
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6329416036605835,
+      "learning_rate": 0.00023187852677754626,
+      "loss": 2.8363,
+      "step": 35071
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6531370878219604,
+      "learning_rate": 0.00023186353285503076,
+      "loss": 2.9056,
+      "step": 35072
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6264121532440186,
+      "learning_rate": 0.0002318485391119734,
+      "loss": 3.0358,
+      "step": 35073
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6566066145896912,
+      "learning_rate": 0.00023183354554841403,
+      "loss": 2.8215,
+      "step": 35074
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.652664840221405,
+      "learning_rate": 0.00023181855216439194,
+      "loss": 3.1829,
+      "step": 35075
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6445363163948059,
+      "learning_rate": 0.00023180355895994654,
+      "loss": 3.0111,
+      "step": 35076
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5922958254814148,
+      "learning_rate": 0.00023178856593511761,
+      "loss": 3.0851,
+      "step": 35077
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6356951594352722,
+      "learning_rate": 0.00023177357308994443,
+      "loss": 3.0073,
+      "step": 35078
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6576932072639465,
+      "learning_rate": 0.00023175858042446653,
+      "loss": 3.1066,
+      "step": 35079
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6770666837692261,
+      "learning_rate": 0.0002317435879387233,
+      "loss": 2.8589,
+      "step": 35080
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6689457297325134,
+      "learning_rate": 0.00023172859563275436,
+      "loss": 2.9456,
+      "step": 35081
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6123393774032593,
+      "learning_rate": 0.00023171360350659922,
+      "loss": 3.3388,
+      "step": 35082
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6080477833747864,
+      "learning_rate": 0.0002316986115602973,
+      "loss": 2.8955,
+      "step": 35083
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5717896819114685,
+      "learning_rate": 0.0002316836197938881,
+      "loss": 2.9882,
+      "step": 35084
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5806421637535095,
+      "learning_rate": 0.000231668628207411,
+      "loss": 3.0781,
+      "step": 35085
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6210070252418518,
+      "learning_rate": 0.00023165363680090567,
+      "loss": 3.1186,
+      "step": 35086
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6110129356384277,
+      "learning_rate": 0.00023163864557441152,
+      "loss": 3.156,
+      "step": 35087
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6130040884017944,
+      "learning_rate": 0.00023162365452796793,
+      "loss": 2.936,
+      "step": 35088
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6246054768562317,
+      "learning_rate": 0.0002316086636616146,
+      "loss": 2.9497,
+      "step": 35089
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6145926117897034,
+      "learning_rate": 0.00023159367297539082,
+      "loss": 3.0549,
+      "step": 35090
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5754942893981934,
+      "learning_rate": 0.0002315786824693361,
+      "loss": 3.034,
+      "step": 35091
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.639498770236969,
+      "learning_rate": 0.00023156369214349004,
+      "loss": 3.0611,
+      "step": 35092
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6060043573379517,
+      "learning_rate": 0.00023154870199789199,
+      "loss": 3.1878,
+      "step": 35093
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6273714900016785,
+      "learning_rate": 0.00023153371203258157,
+      "loss": 3.0544,
+      "step": 35094
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6444366574287415,
+      "learning_rate": 0.00023151872224759805,
+      "loss": 3.2422,
+      "step": 35095
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6600435376167297,
+      "learning_rate": 0.0002315037326429811,
+      "loss": 3.1418,
+      "step": 35096
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.605183482170105,
+      "learning_rate": 0.00023148874321877018,
+      "loss": 2.9719,
+      "step": 35097
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6069551110267639,
+      "learning_rate": 0.0002314737539750046,
+      "loss": 2.6923,
+      "step": 35098
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5764786005020142,
+      "learning_rate": 0.0002314587649117241,
+      "loss": 2.9678,
+      "step": 35099
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6361105442047119,
+      "learning_rate": 0.000231443776028968,
+      "loss": 3.2081,
+      "step": 35100
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5919163227081299,
+      "learning_rate": 0.00023142878732677566,
+      "loss": 3.0231,
+      "step": 35101
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6047539114952087,
+      "learning_rate": 0.00023141379880518685,
+      "loss": 3.0358,
+      "step": 35102
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6241350173950195,
+      "learning_rate": 0.00023139881046424079,
+      "loss": 2.9605,
+      "step": 35103
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6277053356170654,
+      "learning_rate": 0.00023138382230397718,
+      "loss": 3.0339,
+      "step": 35104
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6247433423995972,
+      "learning_rate": 0.0002313688343244353,
+      "loss": 2.979,
+      "step": 35105
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5854073762893677,
+      "learning_rate": 0.00023135384652565465,
+      "loss": 2.9685,
+      "step": 35106
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6047874093055725,
+      "learning_rate": 0.0002313388589076749,
+      "loss": 3.2763,
+      "step": 35107
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5953376293182373,
+      "learning_rate": 0.00023132387147053527,
+      "loss": 2.8974,
+      "step": 35108
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5756540894508362,
+      "learning_rate": 0.00023130888421427545,
+      "loss": 3.0288,
+      "step": 35109
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6189519166946411,
+      "learning_rate": 0.00023129389713893466,
+      "loss": 2.7638,
+      "step": 35110
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5780520439147949,
+      "learning_rate": 0.0002312789102445526,
+      "loss": 3.0979,
+      "step": 35111
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6746885180473328,
+      "learning_rate": 0.00023126392353116872,
+      "loss": 3.0076,
+      "step": 35112
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6331716179847717,
+      "learning_rate": 0.00023124893699882231,
+      "loss": 3.2743,
+      "step": 35113
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6570529937744141,
+      "learning_rate": 0.00023123395064755312,
+      "loss": 3.0482,
+      "step": 35114
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6318064332008362,
+      "learning_rate": 0.00023121896447740036,
+      "loss": 3.0843,
+      "step": 35115
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6018270254135132,
+      "learning_rate": 0.00023120397848840359,
+      "loss": 2.9797,
+      "step": 35116
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5887584090232849,
+      "learning_rate": 0.00023118899268060241,
+      "loss": 3.1461,
+      "step": 35117
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6415103077888489,
+      "learning_rate": 0.00023117400705403613,
+      "loss": 3.0801,
+      "step": 35118
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6064682006835938,
+      "learning_rate": 0.0002311590216087443,
+      "loss": 2.779,
+      "step": 35119
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6046452522277832,
+      "learning_rate": 0.00023114403634476625,
+      "loss": 2.99,
+      "step": 35120
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6241592764854431,
+      "learning_rate": 0.00023112905126214156,
+      "loss": 2.903,
+      "step": 35121
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6440034508705139,
+      "learning_rate": 0.00023111406636090983,
+      "loss": 2.928,
+      "step": 35122
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6371393799781799,
+      "learning_rate": 0.0002310990816411103,
+      "loss": 3.198,
+      "step": 35123
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.635123610496521,
+      "learning_rate": 0.00023108409710278263,
+      "loss": 2.8926,
+      "step": 35124
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6079972386360168,
+      "learning_rate": 0.000231069112745966,
+      "loss": 2.9344,
+      "step": 35125
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5995076298713684,
+      "learning_rate": 0.0002310541285707002,
+      "loss": 2.9177,
+      "step": 35126
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6506484746932983,
+      "learning_rate": 0.00023103914457702455,
+      "loss": 3.03,
+      "step": 35127
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6203939914703369,
+      "learning_rate": 0.00023102416076497843,
+      "loss": 3.0106,
+      "step": 35128
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6446170210838318,
+      "learning_rate": 0.00023100917713460154,
+      "loss": 3.1954,
+      "step": 35129
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.676665723323822,
+      "learning_rate": 0.00023099419368593311,
+      "loss": 2.9198,
+      "step": 35130
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6401715278625488,
+      "learning_rate": 0.00023097921041901275,
+      "loss": 3.1917,
+      "step": 35131
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6062941551208496,
+      "learning_rate": 0.00023096422733387978,
+      "loss": 2.7867,
+      "step": 35132
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.574828565120697,
+      "learning_rate": 0.00023094924443057378,
+      "loss": 2.9809,
+      "step": 35133
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6718885898590088,
+      "learning_rate": 0.00023093426170913426,
+      "loss": 3.095,
+      "step": 35134
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.595146656036377,
+      "learning_rate": 0.00023091927916960042,
+      "loss": 2.8781,
+      "step": 35135
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6414282321929932,
+      "learning_rate": 0.0002309042968120121,
+      "loss": 2.8138,
+      "step": 35136
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6145898699760437,
+      "learning_rate": 0.00023088931463640847,
+      "loss": 2.856,
+      "step": 35137
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6160130500793457,
+      "learning_rate": 0.00023087433264282905,
+      "loss": 2.7365,
+      "step": 35138
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6244746446609497,
+      "learning_rate": 0.00023085935083131342,
+      "loss": 2.838,
+      "step": 35139
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6224129796028137,
+      "learning_rate": 0.00023084436920190092,
+      "loss": 2.8034,
+      "step": 35140
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6752472519874573,
+      "learning_rate": 0.0002308293877546311,
+      "loss": 2.9116,
+      "step": 35141
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6341972947120667,
+      "learning_rate": 0.00023081440648954323,
+      "loss": 2.8963,
+      "step": 35142
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6429397463798523,
+      "learning_rate": 0.00023079942540667688,
+      "loss": 3.1747,
+      "step": 35143
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6000608205795288,
+      "learning_rate": 0.00023078444450607164,
+      "loss": 2.8118,
+      "step": 35144
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6079022884368896,
+      "learning_rate": 0.00023076946378776687,
+      "loss": 2.9218,
+      "step": 35145
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6264728307723999,
+      "learning_rate": 0.00023075448325180203,
+      "loss": 2.8652,
+      "step": 35146
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6049911379814148,
+      "learning_rate": 0.00023073950289821638,
+      "loss": 3.0077,
+      "step": 35147
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6240792870521545,
+      "learning_rate": 0.00023072452272704966,
+      "loss": 3.0171,
+      "step": 35148
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6277632117271423,
+      "learning_rate": 0.00023070954273834128,
+      "loss": 2.9082,
+      "step": 35149
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5884895324707031,
+      "learning_rate": 0.0002306945629321305,
+      "loss": 3.139,
+      "step": 35150
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6733899712562561,
+      "learning_rate": 0.00023067958330845703,
+      "loss": 3.0002,
+      "step": 35151
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6554712057113647,
+      "learning_rate": 0.0002306646038673601,
+      "loss": 3.1105,
+      "step": 35152
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6073846817016602,
+      "learning_rate": 0.00023064962460887925,
+      "loss": 2.9152,
+      "step": 35153
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5951540470123291,
+      "learning_rate": 0.000230634645533054,
+      "loss": 3.0983,
+      "step": 35154
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6558448672294617,
+      "learning_rate": 0.00023061966663992376,
+      "loss": 2.9839,
+      "step": 35155
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.632959246635437,
+      "learning_rate": 0.00023060468792952796,
+      "loss": 3.2383,
+      "step": 35156
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6814397573471069,
+      "learning_rate": 0.00023058970940190596,
+      "loss": 2.7762,
+      "step": 35157
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.62791907787323,
+      "learning_rate": 0.0002305747310570974,
+      "loss": 3.0783,
+      "step": 35158
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6006518602371216,
+      "learning_rate": 0.00023055975289514166,
+      "loss": 3.2219,
+      "step": 35159
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6083630919456482,
+      "learning_rate": 0.00023054477491607804,
+      "loss": 3.1753,
+      "step": 35160
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5891106128692627,
+      "learning_rate": 0.00023052979711994624,
+      "loss": 2.9886,
+      "step": 35161
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6096766591072083,
+      "learning_rate": 0.00023051481950678545,
+      "loss": 3.0412,
+      "step": 35162
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.64244544506073,
+      "learning_rate": 0.00023049984207663537,
+      "loss": 2.8953,
+      "step": 35163
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6248708963394165,
+      "learning_rate": 0.00023048486482953536,
+      "loss": 3.0307,
+      "step": 35164
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6246823072433472,
+      "learning_rate": 0.00023046988776552469,
+      "loss": 3.14,
+      "step": 35165
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6114916205406189,
+      "learning_rate": 0.00023045491088464313,
+      "loss": 3.003,
+      "step": 35166
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6664708256721497,
+      "learning_rate": 0.00023043993418692986,
+      "loss": 2.9394,
+      "step": 35167
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6526961326599121,
+      "learning_rate": 0.00023042495767242434,
+      "loss": 3.0523,
+      "step": 35168
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.637554943561554,
+      "learning_rate": 0.00023040998134116623,
+      "loss": 3.157,
+      "step": 35169
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6036433577537537,
+      "learning_rate": 0.00023039500519319477,
+      "loss": 3.2725,
+      "step": 35170
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6289345622062683,
+      "learning_rate": 0.00023038002922854957,
+      "loss": 3.088,
+      "step": 35171
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6497759222984314,
+      "learning_rate": 0.0002303650534472698,
+      "loss": 3.1527,
+      "step": 35172
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6652274131774902,
+      "learning_rate": 0.00023035007784939516,
+      "loss": 2.8838,
+      "step": 35173
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6045557260513306,
+      "learning_rate": 0.00023033510243496507,
+      "loss": 2.9615,
+      "step": 35174
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6381337642669678,
+      "learning_rate": 0.00023032012720401878,
+      "loss": 3.0682,
+      "step": 35175
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6169887781143188,
+      "learning_rate": 0.000230305152156596,
+      "loss": 3.2183,
+      "step": 35176
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6736177206039429,
+      "learning_rate": 0.00023029017729273596,
+      "loss": 2.9374,
+      "step": 35177
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6414783596992493,
+      "learning_rate": 0.00023027520261247812,
+      "loss": 3.0381,
+      "step": 35178
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6434562802314758,
+      "learning_rate": 0.00023026022811586215,
+      "loss": 2.8393,
+      "step": 35179
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6093327403068542,
+      "learning_rate": 0.00023024525380292718,
+      "loss": 3.0741,
+      "step": 35180
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5809537172317505,
+      "learning_rate": 0.0002302302796737129,
+      "loss": 3.0564,
+      "step": 35181
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.635099470615387,
+      "learning_rate": 0.00023021530572825847,
+      "loss": 3.1763,
+      "step": 35182
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6042500734329224,
+      "learning_rate": 0.00023020033196660355,
+      "loss": 3.0027,
+      "step": 35183
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6080394387245178,
+      "learning_rate": 0.00023018535838878758,
+      "loss": 3.0439,
+      "step": 35184
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.666522204875946,
+      "learning_rate": 0.00023017038499484994,
+      "loss": 3.0731,
+      "step": 35185
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6572619080543518,
+      "learning_rate": 0.0002301554117848301,
+      "loss": 3.0514,
+      "step": 35186
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6002278327941895,
+      "learning_rate": 0.00023014043875876734,
+      "loss": 2.9164,
+      "step": 35187
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6185051202774048,
+      "learning_rate": 0.00023012546591670127,
+      "loss": 2.9961,
+      "step": 35188
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.61963951587677,
+      "learning_rate": 0.00023011049325867135,
+      "loss": 2.9879,
+      "step": 35189
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5980280041694641,
+      "learning_rate": 0.0002300955207847168,
+      "loss": 3.2725,
+      "step": 35190
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5959213972091675,
+      "learning_rate": 0.00023008054849487733,
+      "loss": 3.1077,
+      "step": 35191
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5841280221939087,
+      "learning_rate": 0.00023006557638919223,
+      "loss": 3.0087,
+      "step": 35192
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.618198573589325,
+      "learning_rate": 0.0002300506044677008,
+      "loss": 2.9676,
+      "step": 35193
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6227092742919922,
+      "learning_rate": 0.00023003563273044276,
+      "loss": 3.1052,
+      "step": 35194
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6252568364143372,
+      "learning_rate": 0.00023002066117745737,
+      "loss": 3.0911,
+      "step": 35195
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.623429536819458,
+      "learning_rate": 0.00023000568980878415,
+      "loss": 2.9064,
+      "step": 35196
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6618231534957886,
+      "learning_rate": 0.0002299907186244623,
+      "loss": 3.1137,
+      "step": 35197
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6143259406089783,
+      "learning_rate": 0.00022997574762453156,
+      "loss": 3.0053,
+      "step": 35198
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6106548309326172,
+      "learning_rate": 0.0002299607768090312,
+      "loss": 3.0564,
+      "step": 35199
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.595967710018158,
+      "learning_rate": 0.0002299458061780006,
+      "loss": 2.9528,
+      "step": 35200
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6516256928443909,
+      "learning_rate": 0.0002299308357314794,
+      "loss": 3.0523,
+      "step": 35201
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6277573704719543,
+      "learning_rate": 0.0002299158654695068,
+      "loss": 3.1288,
+      "step": 35202
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6570903062820435,
+      "learning_rate": 0.0002299008953921223,
+      "loss": 2.9566,
+      "step": 35203
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.638201892375946,
+      "learning_rate": 0.00022988592549936542,
+      "loss": 2.756,
+      "step": 35204
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6191340088844299,
+      "learning_rate": 0.00022987095579127544,
+      "loss": 3.0523,
+      "step": 35205
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6047921776771545,
+      "learning_rate": 0.00022985598626789196,
+      "loss": 3.0704,
+      "step": 35206
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6013028025627136,
+      "learning_rate": 0.00022984101692925425,
+      "loss": 3.1367,
+      "step": 35207
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.681874692440033,
+      "learning_rate": 0.00022982604777540175,
+      "loss": 3.125,
+      "step": 35208
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6835533976554871,
+      "learning_rate": 0.0002298110788063741,
+      "loss": 2.946,
+      "step": 35209
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6232330799102783,
+      "learning_rate": 0.00022979611002221044,
+      "loss": 2.9882,
+      "step": 35210
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6557362675666809,
+      "learning_rate": 0.00022978114142295036,
+      "loss": 2.8999,
+      "step": 35211
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6331151127815247,
+      "learning_rate": 0.00022976617300863317,
+      "loss": 3.0894,
+      "step": 35212
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6033350825309753,
+      "learning_rate": 0.00022975120477929846,
+      "loss": 2.892,
+      "step": 35213
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.64558345079422,
+      "learning_rate": 0.0002297362367349855,
+      "loss": 2.8062,
+      "step": 35214
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6405138373374939,
+      "learning_rate": 0.00022972126887573372,
+      "loss": 2.9401,
+      "step": 35215
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6379008293151855,
+      "learning_rate": 0.0002297063012015827,
+      "loss": 3.1785,
+      "step": 35216
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6051523089408875,
+      "learning_rate": 0.0002296913337125717,
+      "loss": 3.099,
+      "step": 35217
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6115164160728455,
+      "learning_rate": 0.00022967636640874026,
+      "loss": 2.9392,
+      "step": 35218
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6115735769271851,
+      "learning_rate": 0.0002296613992901276,
+      "loss": 3.0127,
+      "step": 35219
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6742574572563171,
+      "learning_rate": 0.00022964643235677337,
+      "loss": 3.2632,
+      "step": 35220
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6841800212860107,
+      "learning_rate": 0.00022963146560871692,
+      "loss": 2.976,
+      "step": 35221
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6270367503166199,
+      "learning_rate": 0.00022961649904599754,
+      "loss": 3.1518,
+      "step": 35222
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6001424193382263,
+      "learning_rate": 0.00022960153266865488,
+      "loss": 3.0636,
+      "step": 35223
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5948289632797241,
+      "learning_rate": 0.0002295865664767281,
+      "loss": 3.0079,
+      "step": 35224
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6431679725646973,
+      "learning_rate": 0.00022957160047025683,
+      "loss": 2.9789,
+      "step": 35225
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6434341669082642,
+      "learning_rate": 0.00022955663464928048,
+      "loss": 3.2493,
+      "step": 35226
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6969561576843262,
+      "learning_rate": 0.00022954166901383827,
+      "loss": 2.903,
+      "step": 35227
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7470771074295044,
+      "learning_rate": 0.00022952670356396986,
+      "loss": 3.0102,
+      "step": 35228
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6371145844459534,
+      "learning_rate": 0.00022951173829971447,
+      "loss": 2.9863,
+      "step": 35229
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6292941570281982,
+      "learning_rate": 0.00022949677322111155,
+      "loss": 3.1331,
+      "step": 35230
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6096308827400208,
+      "learning_rate": 0.0002294818083282007,
+      "loss": 3.084,
+      "step": 35231
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6575873494148254,
+      "learning_rate": 0.0002294668436210211,
+      "loss": 3.0223,
+      "step": 35232
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6215528845787048,
+      "learning_rate": 0.00022945187909961237,
+      "loss": 2.9783,
+      "step": 35233
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6451046466827393,
+      "learning_rate": 0.00022943691476401363,
+      "loss": 2.8736,
+      "step": 35234
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5992505550384521,
+      "learning_rate": 0.0002294219506142646,
+      "loss": 3.1373,
+      "step": 35235
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6010803580284119,
+      "learning_rate": 0.00022940698665040459,
+      "loss": 2.8851,
+      "step": 35236
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6390721201896667,
+      "learning_rate": 0.00022939202287247285,
+      "loss": 3.0608,
+      "step": 35237
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6065561771392822,
+      "learning_rate": 0.00022937705928050911,
+      "loss": 2.9655,
+      "step": 35238
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6406224370002747,
+      "learning_rate": 0.0002293620958745525,
+      "loss": 3.0668,
+      "step": 35239
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6230671405792236,
+      "learning_rate": 0.00022934713265464248,
+      "loss": 2.97,
+      "step": 35240
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6472780704498291,
+      "learning_rate": 0.00022933216962081867,
+      "loss": 3.097,
+      "step": 35241
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6151770949363708,
+      "learning_rate": 0.00022931720677312023,
+      "loss": 2.9176,
+      "step": 35242
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6394203901290894,
+      "learning_rate": 0.00022930224411158673,
+      "loss": 2.904,
+      "step": 35243
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5917284488677979,
+      "learning_rate": 0.0002292872816362574,
+      "loss": 3.172,
+      "step": 35244
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6539353728294373,
+      "learning_rate": 0.00022927231934717176,
+      "loss": 3.0997,
+      "step": 35245
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.592643678188324,
+      "learning_rate": 0.00022925735724436933,
+      "loss": 3.1776,
+      "step": 35246
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6221964955329895,
+      "learning_rate": 0.00022924239532788935,
+      "loss": 3.0839,
+      "step": 35247
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6389697194099426,
+      "learning_rate": 0.00022922743359777138,
+      "loss": 3.1729,
+      "step": 35248
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6145962476730347,
+      "learning_rate": 0.00022921247205405458,
+      "loss": 3.2908,
+      "step": 35249
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6351203322410583,
+      "learning_rate": 0.00022919751069677856,
+      "loss": 3.214,
+      "step": 35250
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6602262258529663,
+      "learning_rate": 0.00022918254952598272,
+      "loss": 2.904,
+      "step": 35251
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6813586354255676,
+      "learning_rate": 0.0002291675885417063,
+      "loss": 2.9873,
+      "step": 35252
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6100876331329346,
+      "learning_rate": 0.00022915262774398897,
+      "loss": 3.0853,
+      "step": 35253
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.614741861820221,
+      "learning_rate": 0.0002291376671328699,
+      "loss": 3.0462,
+      "step": 35254
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.629533588886261,
+      "learning_rate": 0.0002291227067083885,
+      "loss": 2.9814,
+      "step": 35255
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6625435948371887,
+      "learning_rate": 0.0002291077464705844,
+      "loss": 3.1346,
+      "step": 35256
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6223627924919128,
+      "learning_rate": 0.0002290927864194968,
+      "loss": 2.9611,
+      "step": 35257
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5756930112838745,
+      "learning_rate": 0.0002290778265551652,
+      "loss": 2.9672,
+      "step": 35258
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6262645125389099,
+      "learning_rate": 0.0002290628668776288,
+      "loss": 2.8283,
+      "step": 35259
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6052729487419128,
+      "learning_rate": 0.00022904790738692725,
+      "loss": 2.8735,
+      "step": 35260
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6487675309181213,
+      "learning_rate": 0.0002290329480830999,
+      "loss": 2.8281,
+      "step": 35261
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6117686629295349,
+      "learning_rate": 0.00022901798896618602,
+      "loss": 3.0395,
+      "step": 35262
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.602601170539856,
+      "learning_rate": 0.0002290030300362252,
+      "loss": 2.974,
+      "step": 35263
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6140736937522888,
+      "learning_rate": 0.00022898807129325665,
+      "loss": 2.8676,
+      "step": 35264
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6174287796020508,
+      "learning_rate": 0.00022897311273731985,
+      "loss": 3.0077,
+      "step": 35265
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6516817212104797,
+      "learning_rate": 0.00022895815436845427,
+      "loss": 3.1008,
+      "step": 35266
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6091328859329224,
+      "learning_rate": 0.0002289431961866991,
+      "loss": 2.8788,
+      "step": 35267
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6192858815193176,
+      "learning_rate": 0.00022892823819209407,
+      "loss": 2.9493,
+      "step": 35268
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.641791820526123,
+      "learning_rate": 0.0002289132803846783,
+      "loss": 3.0263,
+      "step": 35269
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6411734223365784,
+      "learning_rate": 0.00022889832276449116,
+      "loss": 3.1047,
+      "step": 35270
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6926196217536926,
+      "learning_rate": 0.00022888336533157232,
+      "loss": 3.1334,
+      "step": 35271
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.622738778591156,
+      "learning_rate": 0.00022886840808596096,
+      "loss": 3.1372,
+      "step": 35272
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5840880870819092,
+      "learning_rate": 0.00022885345102769655,
+      "loss": 3.0984,
+      "step": 35273
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.60886549949646,
+      "learning_rate": 0.00022883849415681834,
+      "loss": 2.9715,
+      "step": 35274
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6631279587745667,
+      "learning_rate": 0.00022882353747336593,
+      "loss": 2.959,
+      "step": 35275
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6146132946014404,
+      "learning_rate": 0.00022880858097737864,
+      "loss": 2.9739,
+      "step": 35276
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6303625702857971,
+      "learning_rate": 0.00022879362466889577,
+      "loss": 3.0501,
+      "step": 35277
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6738470792770386,
+      "learning_rate": 0.0002287786685479569,
+      "loss": 2.9648,
+      "step": 35278
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6016108989715576,
+      "learning_rate": 0.00022876371261460122,
+      "loss": 3.1268,
+      "step": 35279
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6048367023468018,
+      "learning_rate": 0.00022874875686886817,
+      "loss": 3.1259,
+      "step": 35280
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6452575922012329,
+      "learning_rate": 0.0002287338013107973,
+      "loss": 3.1736,
+      "step": 35281
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6293025612831116,
+      "learning_rate": 0.00022871884594042783,
+      "loss": 2.9434,
+      "step": 35282
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6301412582397461,
+      "learning_rate": 0.00022870389075779925,
+      "loss": 2.8099,
+      "step": 35283
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6559537053108215,
+      "learning_rate": 0.0002286889357629508,
+      "loss": 2.8507,
+      "step": 35284
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6444092392921448,
+      "learning_rate": 0.00022867398095592194,
+      "loss": 2.8865,
+      "step": 35285
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5924097895622253,
+      "learning_rate": 0.0002286590263367522,
+      "loss": 3.3732,
+      "step": 35286
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6571818590164185,
+      "learning_rate": 0.00022864407190548085,
+      "loss": 3.0194,
+      "step": 35287
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6414764523506165,
+      "learning_rate": 0.00022862911766214733,
+      "loss": 2.895,
+      "step": 35288
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5858622789382935,
+      "learning_rate": 0.00022861416360679082,
+      "loss": 2.973,
+      "step": 35289
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6514439582824707,
+      "learning_rate": 0.00022859920973945097,
+      "loss": 2.8787,
+      "step": 35290
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6287581324577332,
+      "learning_rate": 0.0002285842560601671,
+      "loss": 3.1014,
+      "step": 35291
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6182932257652283,
+      "learning_rate": 0.0002285693025689784,
+      "loss": 3.1868,
+      "step": 35292
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.739643394947052,
+      "learning_rate": 0.00022855434926592454,
+      "loss": 2.7995,
+      "step": 35293
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6869654655456543,
+      "learning_rate": 0.00022853939615104476,
+      "loss": 2.9529,
+      "step": 35294
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6125555038452148,
+      "learning_rate": 0.00022852444322437847,
+      "loss": 2.8912,
+      "step": 35295
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5996557474136353,
+      "learning_rate": 0.00022850949048596492,
+      "loss": 3.0872,
+      "step": 35296
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6288072466850281,
+      "learning_rate": 0.00022849453793584372,
+      "loss": 2.9651,
+      "step": 35297
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6631354093551636,
+      "learning_rate": 0.00022847958557405416,
+      "loss": 2.9076,
+      "step": 35298
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6144710779190063,
+      "learning_rate": 0.0002284646334006355,
+      "loss": 2.9343,
+      "step": 35299
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6347219347953796,
+      "learning_rate": 0.00022844968141562738,
+      "loss": 3.1248,
+      "step": 35300
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6205817461013794,
+      "learning_rate": 0.00022843472961906893,
+      "loss": 2.8253,
+      "step": 35301
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6230130791664124,
+      "learning_rate": 0.0002284197780109996,
+      "loss": 2.9309,
+      "step": 35302
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.673992931842804,
+      "learning_rate": 0.0002284048265914589,
+      "loss": 3.2319,
+      "step": 35303
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7464253902435303,
+      "learning_rate": 0.00022838987536048604,
+      "loss": 3.1167,
+      "step": 35304
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6007848381996155,
+      "learning_rate": 0.00022837492431812054,
+      "loss": 2.8541,
+      "step": 35305
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6575406789779663,
+      "learning_rate": 0.0002283599734644016,
+      "loss": 3.1411,
+      "step": 35306
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5982678532600403,
+      "learning_rate": 0.00022834502279936867,
+      "loss": 2.9991,
+      "step": 35307
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6275074481964111,
+      "learning_rate": 0.00022833007232306128,
+      "loss": 2.88,
+      "step": 35308
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6013560891151428,
+      "learning_rate": 0.00022831512203551864,
+      "loss": 2.8328,
+      "step": 35309
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6640160083770752,
+      "learning_rate": 0.00022830017193678024,
+      "loss": 2.9587,
+      "step": 35310
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6699552536010742,
+      "learning_rate": 0.00022828522202688527,
+      "loss": 3.1212,
+      "step": 35311
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6032391786575317,
+      "learning_rate": 0.00022827027230587328,
+      "loss": 2.9259,
+      "step": 35312
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6313595175743103,
+      "learning_rate": 0.00022825532277378363,
+      "loss": 3.02,
+      "step": 35313
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6690878868103027,
+      "learning_rate": 0.00022824037343065556,
+      "loss": 3.0762,
+      "step": 35314
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5943594574928284,
+      "learning_rate": 0.00022822542427652863,
+      "loss": 2.8547,
+      "step": 35315
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6819391846656799,
+      "learning_rate": 0.00022821047531144207,
+      "loss": 3.2038,
+      "step": 35316
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6711174845695496,
+      "learning_rate": 0.00022819552653543525,
+      "loss": 2.8025,
+      "step": 35317
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.667477011680603,
+      "learning_rate": 0.0002281805779485477,
+      "loss": 3.1893,
+      "step": 35318
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6368096470832825,
+      "learning_rate": 0.0002281656295508186,
+      "loss": 3.1114,
+      "step": 35319
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6909788846969604,
+      "learning_rate": 0.0002281506813422875,
+      "loss": 2.9798,
+      "step": 35320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6262080073356628,
+      "learning_rate": 0.0002281357333229936,
+      "loss": 3.0459,
+      "step": 35321
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6632434725761414,
+      "learning_rate": 0.00022812078549297633,
+      "loss": 3.0869,
+      "step": 35322
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6851413249969482,
+      "learning_rate": 0.0002281058378522752,
+      "loss": 2.9552,
+      "step": 35323
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6063692569732666,
+      "learning_rate": 0.0002280908904009293,
+      "loss": 2.8897,
+      "step": 35324
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6380870342254639,
+      "learning_rate": 0.00022807594313897827,
+      "loss": 3.108,
+      "step": 35325
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6056703925132751,
+      "learning_rate": 0.00022806099606646125,
+      "loss": 3.0071,
+      "step": 35326
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6386041045188904,
+      "learning_rate": 0.0002280460491834178,
+      "loss": 3.0436,
+      "step": 35327
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6316136717796326,
+      "learning_rate": 0.00022803110248988725,
+      "loss": 2.9004,
+      "step": 35328
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.624875545501709,
+      "learning_rate": 0.0002280161559859088,
+      "loss": 3.0124,
+      "step": 35329
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6380939483642578,
+      "learning_rate": 0.0002280012096715221,
+      "loss": 3.0695,
+      "step": 35330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6007676720619202,
+      "learning_rate": 0.00022798626354676626,
+      "loss": 3.1734,
+      "step": 35331
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6163882613182068,
+      "learning_rate": 0.00022797131761168068,
+      "loss": 2.8966,
+      "step": 35332
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6753647327423096,
+      "learning_rate": 0.00022795637186630494,
+      "loss": 2.9465,
+      "step": 35333
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5936952233314514,
+      "learning_rate": 0.00022794142631067814,
+      "loss": 2.8909,
+      "step": 35334
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6279860734939575,
+      "learning_rate": 0.00022792648094483987,
+      "loss": 3.0901,
+      "step": 35335
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6184460520744324,
+      "learning_rate": 0.00022791153576882922,
+      "loss": 2.9562,
+      "step": 35336
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6539486646652222,
+      "learning_rate": 0.0002278965907826858,
+      "loss": 2.9344,
+      "step": 35337
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6066672205924988,
+      "learning_rate": 0.00022788164598644894,
+      "loss": 3.0856,
+      "step": 35338
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.613047182559967,
+      "learning_rate": 0.00022786670138015783,
+      "loss": 3.1087,
+      "step": 35339
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6032065749168396,
+      "learning_rate": 0.00022785175696385203,
+      "loss": 3.0789,
+      "step": 35340
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6134180426597595,
+      "learning_rate": 0.00022783681273757077,
+      "loss": 3.1098,
+      "step": 35341
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6650903820991516,
+      "learning_rate": 0.00022782186870135343,
+      "loss": 2.9518,
+      "step": 35342
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6557191610336304,
+      "learning_rate": 0.0002278069248552395,
+      "loss": 2.8869,
+      "step": 35343
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6179054975509644,
+      "learning_rate": 0.00022779198119926819,
+      "loss": 3.0837,
+      "step": 35344
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6232442855834961,
+      "learning_rate": 0.00022777703773347894,
+      "loss": 3.1992,
+      "step": 35345
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6513820290565491,
+      "learning_rate": 0.00022776209445791093,
+      "loss": 3.0261,
+      "step": 35346
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6319857239723206,
+      "learning_rate": 0.00022774715137260367,
+      "loss": 2.9248,
+      "step": 35347
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6437803506851196,
+      "learning_rate": 0.00022773220847759667,
+      "loss": 3.0184,
+      "step": 35348
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.645805299282074,
+      "learning_rate": 0.00022771726577292903,
+      "loss": 2.9881,
+      "step": 35349
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6178550124168396,
+      "learning_rate": 0.0002277023232586403,
+      "loss": 3.097,
+      "step": 35350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6673634052276611,
+      "learning_rate": 0.00022768738093476956,
+      "loss": 3.2377,
+      "step": 35351
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.601243257522583,
+      "learning_rate": 0.00022767243880135641,
+      "loss": 2.6505,
+      "step": 35352
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6067811846733093,
+      "learning_rate": 0.00022765749685844025,
+      "loss": 2.8767,
+      "step": 35353
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5918362736701965,
+      "learning_rate": 0.00022764255510606014,
+      "loss": 2.9484,
+      "step": 35354
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6136022806167603,
+      "learning_rate": 0.00022762761354425576,
+      "loss": 3.0941,
+      "step": 35355
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6263452768325806,
+      "learning_rate": 0.0002276126721730663,
+      "loss": 2.9776,
+      "step": 35356
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6397753953933716,
+      "learning_rate": 0.000227597730992531,
+      "loss": 2.8435,
+      "step": 35357
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6426587700843811,
+      "learning_rate": 0.0002275827900026895,
+      "loss": 2.9557,
+      "step": 35358
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6584495902061462,
+      "learning_rate": 0.00022756784920358093,
+      "loss": 2.8092,
+      "step": 35359
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6029062271118164,
+      "learning_rate": 0.00022755290859524476,
+      "loss": 3.1722,
+      "step": 35360
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5749930739402771,
+      "learning_rate": 0.00022753796817772014,
+      "loss": 2.9569,
+      "step": 35361
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7005466222763062,
+      "learning_rate": 0.00022752302795104665,
+      "loss": 3.0455,
+      "step": 35362
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6490772366523743,
+      "learning_rate": 0.00022750808791526363,
+      "loss": 2.8489,
+      "step": 35363
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6193538904190063,
+      "learning_rate": 0.00022749314807041023,
+      "loss": 2.97,
+      "step": 35364
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6467298865318298,
+      "learning_rate": 0.00022747820841652607,
+      "loss": 3.0202,
+      "step": 35365
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5905973315238953,
+      "learning_rate": 0.00022746326895365025,
+      "loss": 3.1219,
+      "step": 35366
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6227057576179504,
+      "learning_rate": 0.00022744832968182228,
+      "loss": 2.9582,
+      "step": 35367
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6363484263420105,
+      "learning_rate": 0.00022743339060108135,
+      "loss": 3.2402,
+      "step": 35368
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6553448438644409,
+      "learning_rate": 0.00022741845171146685,
+      "loss": 3.0608,
+      "step": 35369
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6466527581214905,
+      "learning_rate": 0.00022740351301301837,
+      "loss": 3.0484,
+      "step": 35370
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5950348377227783,
+      "learning_rate": 0.00022738857450577496,
+      "loss": 3.2043,
+      "step": 35371
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6238282918930054,
+      "learning_rate": 0.00022737363618977613,
+      "loss": 2.9925,
+      "step": 35372
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6248958110809326,
+      "learning_rate": 0.00022735869806506106,
+      "loss": 3.1858,
+      "step": 35373
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7205170392990112,
+      "learning_rate": 0.00022734376013166927,
+      "loss": 2.9612,
+      "step": 35374
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6162315607070923,
+      "learning_rate": 0.0002273288223896401,
+      "loss": 2.8959,
+      "step": 35375
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6196463704109192,
+      "learning_rate": 0.00022731388483901267,
+      "loss": 3.0026,
+      "step": 35376
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6567784547805786,
+      "learning_rate": 0.00022729894747982663,
+      "loss": 2.9925,
+      "step": 35377
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6151658892631531,
+      "learning_rate": 0.00022728401031212107,
+      "loss": 3.1042,
+      "step": 35378
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5919417142868042,
+      "learning_rate": 0.0002272690733359354,
+      "loss": 3.0079,
+      "step": 35379
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6030619740486145,
+      "learning_rate": 0.00022725413655130916,
+      "loss": 2.9471,
+      "step": 35380
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6577292084693909,
+      "learning_rate": 0.0002272391999582814,
+      "loss": 2.8635,
+      "step": 35381
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6286402940750122,
+      "learning_rate": 0.00022722426355689166,
+      "loss": 2.8672,
+      "step": 35382
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6577459573745728,
+      "learning_rate": 0.0002272093273471791,
+      "loss": 3.0737,
+      "step": 35383
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.692613959312439,
+      "learning_rate": 0.00022719439132918318,
+      "loss": 3.2409,
+      "step": 35384
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.628135085105896,
+      "learning_rate": 0.0002271794555029433,
+      "loss": 2.9021,
+      "step": 35385
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.629897952079773,
+      "learning_rate": 0.00022716451986849861,
+      "loss": 2.9762,
+      "step": 35386
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6259167790412903,
+      "learning_rate": 0.00022714958442588868,
+      "loss": 2.949,
+      "step": 35387
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6222914457321167,
+      "learning_rate": 0.0002271346491751526,
+      "loss": 2.9188,
+      "step": 35388
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6155588030815125,
+      "learning_rate": 0.00022711971411632992,
+      "loss": 3.0117,
+      "step": 35389
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6510941982269287,
+      "learning_rate": 0.0002271047792494599,
+      "loss": 3.1024,
+      "step": 35390
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6198724508285522,
+      "learning_rate": 0.00022708984457458175,
+      "loss": 2.7966,
+      "step": 35391
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6015543341636658,
+      "learning_rate": 0.00022707491009173506,
+      "loss": 3.0348,
+      "step": 35392
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6300435662269592,
+      "learning_rate": 0.00022705997580095894,
+      "loss": 3.0702,
+      "step": 35393
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6055490970611572,
+      "learning_rate": 0.00022704504170229276,
+      "loss": 3.1463,
+      "step": 35394
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5714443922042847,
+      "learning_rate": 0.00022703010779577602,
+      "loss": 3.1352,
+      "step": 35395
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6332013607025146,
+      "learning_rate": 0.00022701517408144789,
+      "loss": 3.0562,
+      "step": 35396
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5947373509407043,
+      "learning_rate": 0.00022700024055934777,
+      "loss": 3.167,
+      "step": 35397
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6268903613090515,
+      "learning_rate": 0.00022698530722951486,
+      "loss": 2.9227,
+      "step": 35398
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6999599933624268,
+      "learning_rate": 0.00022697037409198865,
+      "loss": 2.8985,
+      "step": 35399
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6626995205879211,
+      "learning_rate": 0.00022695544114680852,
+      "loss": 3.0245,
+      "step": 35400
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6301445960998535,
+      "learning_rate": 0.0002269405083940135,
+      "loss": 3.0983,
+      "step": 35401
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6395618915557861,
+      "learning_rate": 0.00022692557583364333,
+      "loss": 2.9343,
+      "step": 35402
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5879999995231628,
+      "learning_rate": 0.00022691064346573703,
+      "loss": 2.7305,
+      "step": 35403
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6151583194732666,
+      "learning_rate": 0.00022689571129033397,
+      "loss": 3.0668,
+      "step": 35404
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6331624984741211,
+      "learning_rate": 0.00022688077930747367,
+      "loss": 3.0581,
+      "step": 35405
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6111859083175659,
+      "learning_rate": 0.00022686584751719524,
+      "loss": 3.1264,
+      "step": 35406
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6400036811828613,
+      "learning_rate": 0.00022685091591953817,
+      "loss": 3.1299,
+      "step": 35407
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6397266983985901,
+      "learning_rate": 0.00022683598451454158,
+      "loss": 3.2165,
+      "step": 35408
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6920884847640991,
+      "learning_rate": 0.0002268210533022449,
+      "loss": 3.1075,
+      "step": 35409
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6338069438934326,
+      "learning_rate": 0.00022680612228268766,
+      "loss": 3.0068,
+      "step": 35410
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6007820963859558,
+      "learning_rate": 0.0002267911914559089,
+      "loss": 3.0907,
+      "step": 35411
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6300089955329895,
+      "learning_rate": 0.00022677626082194812,
+      "loss": 3.1833,
+      "step": 35412
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6192349791526794,
+      "learning_rate": 0.00022676133038084447,
+      "loss": 3.0577,
+      "step": 35413
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6377847790718079,
+      "learning_rate": 0.00022674640013263745,
+      "loss": 3.0741,
+      "step": 35414
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6192604303359985,
+      "learning_rate": 0.00022673147007736636,
+      "loss": 3.0948,
+      "step": 35415
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.605172336101532,
+      "learning_rate": 0.00022671654021507037,
+      "loss": 3.0666,
+      "step": 35416
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6593872904777527,
+      "learning_rate": 0.00022670161054578903,
+      "loss": 3.0207,
+      "step": 35417
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6469155550003052,
+      "learning_rate": 0.00022668668106956148,
+      "loss": 2.9885,
+      "step": 35418
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6012063026428223,
+      "learning_rate": 0.00022667175178642701,
+      "loss": 3.0121,
+      "step": 35419
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6011958718299866,
+      "learning_rate": 0.0002266568226964252,
+      "loss": 2.9255,
+      "step": 35420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6600900888442993,
+      "learning_rate": 0.0002266418937995951,
+      "loss": 3.0529,
+      "step": 35421
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6183411478996277,
+      "learning_rate": 0.00022662696509597623,
+      "loss": 2.9691,
+      "step": 35422
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6118205189704895,
+      "learning_rate": 0.0002266120365856077,
+      "loss": 3.101,
+      "step": 35423
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6012549996376038,
+      "learning_rate": 0.00022659710826852895,
+      "loss": 3.0369,
+      "step": 35424
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7324210405349731,
+      "learning_rate": 0.00022658218014477938,
+      "loss": 3.0671,
+      "step": 35425
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.616727888584137,
+      "learning_rate": 0.00022656725221439808,
+      "loss": 2.9368,
+      "step": 35426
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5840345621109009,
+      "learning_rate": 0.00022655232447742467,
+      "loss": 3.1202,
+      "step": 35427
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6287633180618286,
+      "learning_rate": 0.0002265373969338982,
+      "loss": 2.9494,
+      "step": 35428
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6012067198753357,
+      "learning_rate": 0.00022652246958385807,
+      "loss": 3.0535,
+      "step": 35429
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6259651184082031,
+      "learning_rate": 0.00022650754242734371,
+      "loss": 3.2078,
+      "step": 35430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6013410687446594,
+      "learning_rate": 0.00022649261546439422,
+      "loss": 3.0495,
+      "step": 35431
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6384536623954773,
+      "learning_rate": 0.00022647768869504912,
+      "loss": 2.9705,
+      "step": 35432
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6175307631492615,
+      "learning_rate": 0.00022646276211934764,
+      "loss": 3.1426,
+      "step": 35433
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.630964457988739,
+      "learning_rate": 0.000226447835737329,
+      "loss": 3.0491,
+      "step": 35434
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6299197673797607,
+      "learning_rate": 0.00022643290954903273,
+      "loss": 2.9989,
+      "step": 35435
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6441041231155396,
+      "learning_rate": 0.00022641798355449796,
+      "loss": 3.2179,
+      "step": 35436
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6124367117881775,
+      "learning_rate": 0.00022640305775376409,
+      "loss": 2.9684,
+      "step": 35437
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5946165323257446,
+      "learning_rate": 0.00022638813214687033,
+      "loss": 2.781,
+      "step": 35438
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6153435707092285,
+      "learning_rate": 0.0002263732067338561,
+      "loss": 3.2042,
+      "step": 35439
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.614129364490509,
+      "learning_rate": 0.0002263582815147607,
+      "loss": 3.0863,
+      "step": 35440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6298364400863647,
+      "learning_rate": 0.00022634335648962333,
+      "loss": 2.8166,
+      "step": 35441
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6218454241752625,
+      "learning_rate": 0.0002263284316584835,
+      "loss": 3.0104,
+      "step": 35442
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.63499915599823,
+      "learning_rate": 0.00022631350702138038,
+      "loss": 3.0578,
+      "step": 35443
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5916718244552612,
+      "learning_rate": 0.0002262985825783533,
+      "loss": 3.2383,
+      "step": 35444
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6168555617332458,
+      "learning_rate": 0.0002262836583294415,
+      "loss": 3.2438,
+      "step": 35445
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6397154927253723,
+      "learning_rate": 0.00022626873427468439,
+      "loss": 2.8624,
+      "step": 35446
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.594692587852478,
+      "learning_rate": 0.00022625381041412127,
+      "loss": 3.0968,
+      "step": 35447
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6055973172187805,
+      "learning_rate": 0.00022623888674779135,
+      "loss": 2.9747,
+      "step": 35448
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6591005921363831,
+      "learning_rate": 0.0002262239632757341,
+      "loss": 2.9765,
+      "step": 35449
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6846657991409302,
+      "learning_rate": 0.00022620903999798862,
+      "loss": 2.827,
+      "step": 35450
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6533744931221008,
+      "learning_rate": 0.00022619411691459442,
+      "loss": 3.1813,
+      "step": 35451
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6433231830596924,
+      "learning_rate": 0.00022617919402559075,
+      "loss": 2.9903,
+      "step": 35452
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6263206005096436,
+      "learning_rate": 0.00022616427133101673,
+      "loss": 2.9961,
+      "step": 35453
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6297306418418884,
+      "learning_rate": 0.00022614934883091197,
+      "loss": 3.0246,
+      "step": 35454
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6420456171035767,
+      "learning_rate": 0.00022613442652531551,
+      "loss": 3.0221,
+      "step": 35455
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6900204420089722,
+      "learning_rate": 0.00022611950441426675,
+      "loss": 3.232,
+      "step": 35456
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6068527698516846,
+      "learning_rate": 0.00022610458249780507,
+      "loss": 2.8889,
+      "step": 35457
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6444588303565979,
+      "learning_rate": 0.00022608966077596965,
+      "loss": 2.9091,
+      "step": 35458
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6167253255844116,
+      "learning_rate": 0.0002260747392487999,
+      "loss": 2.8771,
+      "step": 35459
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6610934138298035,
+      "learning_rate": 0.00022605981791633492,
+      "loss": 2.8356,
+      "step": 35460
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6797847151756287,
+      "learning_rate": 0.00022604489677861426,
+      "loss": 3.1294,
+      "step": 35461
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6337746381759644,
+      "learning_rate": 0.00022602997583567712,
+      "loss": 3.0702,
+      "step": 35462
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.619875967502594,
+      "learning_rate": 0.00022601505508756267,
+      "loss": 2.879,
+      "step": 35463
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6586485505104065,
+      "learning_rate": 0.0002260001345343105,
+      "loss": 3.1996,
+      "step": 35464
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6685471534729004,
+      "learning_rate": 0.0002259852141759596,
+      "loss": 3.0143,
+      "step": 35465
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6211610436439514,
+      "learning_rate": 0.00022597029401254937,
+      "loss": 3.2222,
+      "step": 35466
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6223245859146118,
+      "learning_rate": 0.0002259553740441193,
+      "loss": 2.8696,
+      "step": 35467
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.603645384311676,
+      "learning_rate": 0.00022594045427070842,
+      "loss": 2.9405,
+      "step": 35468
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6659462451934814,
+      "learning_rate": 0.0002259255346923562,
+      "loss": 3.0429,
+      "step": 35469
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6123594045639038,
+      "learning_rate": 0.00022591061530910175,
+      "loss": 2.9071,
+      "step": 35470
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6161990165710449,
+      "learning_rate": 0.00022589569612098447,
+      "loss": 3.1323,
+      "step": 35471
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6867458820343018,
+      "learning_rate": 0.0002258807771280438,
+      "loss": 2.9695,
+      "step": 35472
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5981655716896057,
+      "learning_rate": 0.00022586585833031883,
+      "loss": 2.9782,
+      "step": 35473
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6146122217178345,
+      "learning_rate": 0.000225850939727849,
+      "loss": 3.0785,
+      "step": 35474
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6437488794326782,
+      "learning_rate": 0.00022583602132067335,
+      "loss": 2.8745,
+      "step": 35475
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6147376894950867,
+      "learning_rate": 0.00022582110310883147,
+      "loss": 2.9838,
+      "step": 35476
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5768126845359802,
+      "learning_rate": 0.00022580618509236258,
+      "loss": 3.0706,
+      "step": 35477
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6257051825523376,
+      "learning_rate": 0.00022579126727130576,
+      "loss": 2.857,
+      "step": 35478
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6463055610656738,
+      "learning_rate": 0.0002257763496457006,
+      "loss": 3.0198,
+      "step": 35479
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6310402154922485,
+      "learning_rate": 0.00022576143221558618,
+      "loss": 2.7418,
+      "step": 35480
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6200835704803467,
+      "learning_rate": 0.00022574651498100186,
+      "loss": 3.0547,
+      "step": 35481
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6173551082611084,
+      "learning_rate": 0.00022573159794198695,
+      "loss": 2.8957,
+      "step": 35482
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6188896894454956,
+      "learning_rate": 0.00022571668109858074,
+      "loss": 2.9646,
+      "step": 35483
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6149911880493164,
+      "learning_rate": 0.00022570176445082252,
+      "loss": 2.963,
+      "step": 35484
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.568781316280365,
+      "learning_rate": 0.00022568684799875144,
+      "loss": 3.0014,
+      "step": 35485
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7318620681762695,
+      "learning_rate": 0.00022567193174240693,
+      "loss": 3.0284,
+      "step": 35486
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6288902759552002,
+      "learning_rate": 0.00022565701568182835,
+      "loss": 2.8994,
+      "step": 35487
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6177136301994324,
+      "learning_rate": 0.00022564209981705473,
+      "loss": 3.065,
+      "step": 35488
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6629291772842407,
+      "learning_rate": 0.00022562718414812568,
+      "loss": 3.1049,
+      "step": 35489
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6288484334945679,
+      "learning_rate": 0.00022561226867508018,
+      "loss": 2.9863,
+      "step": 35490
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.611689031124115,
+      "learning_rate": 0.00022559735339795767,
+      "loss": 2.8179,
+      "step": 35491
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6365360617637634,
+      "learning_rate": 0.00022558243831679749,
+      "loss": 3.0323,
+      "step": 35492
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.586676299571991,
+      "learning_rate": 0.00022556752343163872,
+      "loss": 2.9651,
+      "step": 35493
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6502875685691833,
+      "learning_rate": 0.00022555260874252092,
+      "loss": 3.2247,
+      "step": 35494
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6236476302146912,
+      "learning_rate": 0.00022553769424948314,
+      "loss": 3.2573,
+      "step": 35495
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.622161865234375,
+      "learning_rate": 0.00022552277995256466,
+      "loss": 3.0476,
+      "step": 35496
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5968731045722961,
+      "learning_rate": 0.000225507865851805,
+      "loss": 3.1644,
+      "step": 35497
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5831536054611206,
+      "learning_rate": 0.00022549295194724322,
+      "loss": 2.9254,
+      "step": 35498
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6409755945205688,
+      "learning_rate": 0.0002254780382389187,
+      "loss": 3.0804,
+      "step": 35499
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6048734188079834,
+      "learning_rate": 0.0002254631247268706,
+      "loss": 2.9089,
+      "step": 35500
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6509040594100952,
+      "learning_rate": 0.00022544821141113832,
+      "loss": 2.6318,
+      "step": 35501
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.637199342250824,
+      "learning_rate": 0.00022543329829176115,
+      "loss": 3.145,
+      "step": 35502
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6597538590431213,
+      "learning_rate": 0.00022541838536877822,
+      "loss": 2.7563,
+      "step": 35503
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6191924214363098,
+      "learning_rate": 0.00022540347264222904,
+      "loss": 3.2077,
+      "step": 35504
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6519606113433838,
+      "learning_rate": 0.00022538856011215265,
+      "loss": 2.8337,
+      "step": 35505
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6172239184379578,
+      "learning_rate": 0.00022537364777858843,
+      "loss": 3.2083,
+      "step": 35506
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6710629463195801,
+      "learning_rate": 0.00022535873564157575,
+      "loss": 2.7566,
+      "step": 35507
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6321614980697632,
+      "learning_rate": 0.00022534382370115375,
+      "loss": 2.8906,
+      "step": 35508
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6277992129325867,
+      "learning_rate": 0.00022532891195736182,
+      "loss": 3.0721,
+      "step": 35509
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6334667205810547,
+      "learning_rate": 0.000225314000410239,
+      "loss": 2.9217,
+      "step": 35510
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6273242831230164,
+      "learning_rate": 0.00022529908905982477,
+      "loss": 2.844,
+      "step": 35511
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6677801012992859,
+      "learning_rate": 0.0002252841779061585,
+      "loss": 3.0954,
+      "step": 35512
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6611440181732178,
+      "learning_rate": 0.00022526926694927922,
+      "loss": 3.0319,
+      "step": 35513
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6573300957679749,
+      "learning_rate": 0.00022525435618922637,
+      "loss": 3.0643,
+      "step": 35514
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6609066724777222,
+      "learning_rate": 0.00022523944562603904,
+      "loss": 3.2395,
+      "step": 35515
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6033406853675842,
+      "learning_rate": 0.0002252245352597567,
+      "loss": 3.2297,
+      "step": 35516
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6859046220779419,
+      "learning_rate": 0.00022520962509041864,
+      "loss": 2.9935,
+      "step": 35517
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6412832140922546,
+      "learning_rate": 0.00022519471511806384,
+      "loss": 2.8779,
+      "step": 35518
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6439999341964722,
+      "learning_rate": 0.00022517980534273192,
+      "loss": 2.958,
+      "step": 35519
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6545829772949219,
+      "learning_rate": 0.00022516489576446196,
+      "loss": 3.0343,
+      "step": 35520
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6581568121910095,
+      "learning_rate": 0.0002251499863832933,
+      "loss": 3.0944,
+      "step": 35521
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5977330207824707,
+      "learning_rate": 0.00022513507719926505,
+      "loss": 3.0737,
+      "step": 35522
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.611193835735321,
+      "learning_rate": 0.00022512016821241664,
+      "loss": 3.0715,
+      "step": 35523
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6156790256500244,
+      "learning_rate": 0.0002251052594227874,
+      "loss": 2.9858,
+      "step": 35524
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6435781717300415,
+      "learning_rate": 0.00022509035083041631,
+      "loss": 3.1172,
+      "step": 35525
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6316915154457092,
+      "learning_rate": 0.00022507544243534296,
+      "loss": 3.0121,
+      "step": 35526
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6427628397941589,
+      "learning_rate": 0.00022506053423760645,
+      "loss": 3.1036,
+      "step": 35527
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7132937908172607,
+      "learning_rate": 0.000225045626237246,
+      "loss": 3.1381,
+      "step": 35528
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6076539754867554,
+      "learning_rate": 0.00022503071843430102,
+      "loss": 2.9185,
+      "step": 35529
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.59321129322052,
+      "learning_rate": 0.0002250158108288107,
+      "loss": 3.2645,
+      "step": 35530
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6133169531822205,
+      "learning_rate": 0.00022500090342081433,
+      "loss": 2.9951,
+      "step": 35531
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5968568325042725,
+      "learning_rate": 0.00022498599621035103,
+      "loss": 2.9241,
+      "step": 35532
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6295343637466431,
+      "learning_rate": 0.00022497108919746018,
+      "loss": 2.9182,
+      "step": 35533
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6084388494491577,
+      "learning_rate": 0.0002249561823821812,
+      "loss": 3.0067,
+      "step": 35534
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6127825379371643,
+      "learning_rate": 0.00022494127576455305,
+      "loss": 2.998,
+      "step": 35535
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6642194986343384,
+      "learning_rate": 0.00022492636934461524,
+      "loss": 2.9918,
+      "step": 35536
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5972631573677063,
+      "learning_rate": 0.00022491146312240683,
+      "loss": 2.8956,
+      "step": 35537
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6442238092422485,
+      "learning_rate": 0.0002248965570979672,
+      "loss": 2.8439,
+      "step": 35538
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6403670907020569,
+      "learning_rate": 0.00022488165127133565,
+      "loss": 2.867,
+      "step": 35539
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6831574440002441,
+      "learning_rate": 0.00022486674564255128,
+      "loss": 3.0751,
+      "step": 35540
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6002147793769836,
+      "learning_rate": 0.0002248518402116535,
+      "loss": 3.0445,
+      "step": 35541
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6270040273666382,
+      "learning_rate": 0.0002248369349786815,
+      "loss": 3.0538,
+      "step": 35542
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6381674408912659,
+      "learning_rate": 0.00022482202994367448,
+      "loss": 3.0805,
+      "step": 35543
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6307418942451477,
+      "learning_rate": 0.00022480712510667195,
+      "loss": 3.0251,
+      "step": 35544
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6338752508163452,
+      "learning_rate": 0.00022479222046771282,
+      "loss": 3.2437,
+      "step": 35545
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5932478904724121,
+      "learning_rate": 0.00022477731602683662,
+      "loss": 2.7564,
+      "step": 35546
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6139539480209351,
+      "learning_rate": 0.00022476241178408232,
+      "loss": 3.0832,
+      "step": 35547
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6189956068992615,
+      "learning_rate": 0.00022474750773948944,
+      "loss": 3.0366,
+      "step": 35548
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6055632829666138,
+      "learning_rate": 0.0002247326038930972,
+      "loss": 3.0145,
+      "step": 35549
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6140079498291016,
+      "learning_rate": 0.0002247177002449447,
+      "loss": 3.0551,
+      "step": 35550
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.656378984451294,
+      "learning_rate": 0.00022470279679507143,
+      "loss": 3.0311,
+      "step": 35551
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.668050229549408,
+      "learning_rate": 0.0002246878935435163,
+      "loss": 3.0796,
+      "step": 35552
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6154969930648804,
+      "learning_rate": 0.00022467299049031887,
+      "loss": 3.0686,
+      "step": 35553
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6256099343299866,
+      "learning_rate": 0.0002246580876355184,
+      "loss": 2.9475,
+      "step": 35554
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6148830056190491,
+      "learning_rate": 0.0002246431849791538,
+      "loss": 2.8056,
+      "step": 35555
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6625372171401978,
+      "learning_rate": 0.00022462828252126473,
+      "loss": 2.8291,
+      "step": 35556
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6332937479019165,
+      "learning_rate": 0.0002246133802618902,
+      "loss": 3.0003,
+      "step": 35557
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5910854339599609,
+      "learning_rate": 0.0002245984782010694,
+      "loss": 3.0787,
+      "step": 35558
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6351834535598755,
+      "learning_rate": 0.0002245835763388419,
+      "loss": 3.2896,
+      "step": 35559
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6503569483757019,
+      "learning_rate": 0.00022456867467524663,
+      "loss": 3.0756,
+      "step": 35560
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.686381995677948,
+      "learning_rate": 0.00022455377321032306,
+      "loss": 2.7848,
+      "step": 35561
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6213240027427673,
+      "learning_rate": 0.00022453887194411016,
+      "loss": 3.0433,
+      "step": 35562
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6782155632972717,
+      "learning_rate": 0.00022452397087664743,
+      "loss": 3.0367,
+      "step": 35563
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8070650100708008,
+      "learning_rate": 0.00022450907000797412,
+      "loss": 3.0191,
+      "step": 35564
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6163588166236877,
+      "learning_rate": 0.00022449416933812924,
+      "loss": 2.9667,
+      "step": 35565
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6353696584701538,
+      "learning_rate": 0.00022447926886715231,
+      "loss": 3.0485,
+      "step": 35566
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6798037886619568,
+      "learning_rate": 0.0002244643685950824,
+      "loss": 2.9272,
+      "step": 35567
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5956858992576599,
+      "learning_rate": 0.00022444946852195874,
+      "loss": 3.0098,
+      "step": 35568
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6191090941429138,
+      "learning_rate": 0.00022443456864782078,
+      "loss": 2.8499,
+      "step": 35569
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6584221720695496,
+      "learning_rate": 0.00022441966897270753,
+      "loss": 3.2436,
+      "step": 35570
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6226842403411865,
+      "learning_rate": 0.0002244047694966584,
+      "loss": 3.0886,
+      "step": 35571
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6515241265296936,
+      "learning_rate": 0.00022438987021971242,
+      "loss": 3.0962,
+      "step": 35572
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6351691484451294,
+      "learning_rate": 0.000224374971141909,
+      "loss": 2.9709,
+      "step": 35573
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6209082007408142,
+      "learning_rate": 0.00022436007226328748,
+      "loss": 2.9444,
+      "step": 35574
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5917410850524902,
+      "learning_rate": 0.00022434517358388688,
+      "loss": 3.1332,
+      "step": 35575
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6047651171684265,
+      "learning_rate": 0.0002243302751037466,
+      "loss": 2.9769,
+      "step": 35576
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5979304909706116,
+      "learning_rate": 0.0002243153768229057,
+      "loss": 3.0874,
+      "step": 35577
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6439438462257385,
+      "learning_rate": 0.00022430047874140357,
+      "loss": 2.9494,
+      "step": 35578
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.590167760848999,
+      "learning_rate": 0.0002242855808592795,
+      "loss": 2.9652,
+      "step": 35579
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6090808510780334,
+      "learning_rate": 0.00022427068317657252,
+      "loss": 2.9947,
+      "step": 35580
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.607502281665802,
+      "learning_rate": 0.0002242557856933221,
+      "loss": 2.9917,
+      "step": 35581
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6465458273887634,
+      "learning_rate": 0.0002242408884095673,
+      "loss": 2.9978,
+      "step": 35582
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6103089451789856,
+      "learning_rate": 0.00022422599132534737,
+      "loss": 2.8718,
+      "step": 35583
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6132879853248596,
+      "learning_rate": 0.0002242110944407017,
+      "loss": 2.7641,
+      "step": 35584
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6607589721679688,
+      "learning_rate": 0.00022419619775566934,
+      "loss": 3.0388,
+      "step": 35585
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6323291063308716,
+      "learning_rate": 0.00022418130127028972,
+      "loss": 3.0716,
+      "step": 35586
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6502161622047424,
+      "learning_rate": 0.0002241664049846018,
+      "loss": 3.0507,
+      "step": 35587
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6421080231666565,
+      "learning_rate": 0.00022415150889864503,
+      "loss": 3.0403,
+      "step": 35588
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6806257367134094,
+      "learning_rate": 0.00022413661301245872,
+      "loss": 2.9999,
+      "step": 35589
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6115500330924988,
+      "learning_rate": 0.00022412171732608176,
+      "loss": 2.9546,
+      "step": 35590
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6313839554786682,
+      "learning_rate": 0.00022410682183955375,
+      "loss": 3.0004,
+      "step": 35591
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6472813487052917,
+      "learning_rate": 0.0002240919265529137,
+      "loss": 3.1098,
+      "step": 35592
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6413887143135071,
+      "learning_rate": 0.00022407703146620086,
+      "loss": 3.1327,
+      "step": 35593
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5937476754188538,
+      "learning_rate": 0.00022406213657945464,
+      "loss": 2.833,
+      "step": 35594
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6520037055015564,
+      "learning_rate": 0.00022404724189271398,
+      "loss": 3.0171,
+      "step": 35595
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6196516156196594,
+      "learning_rate": 0.0002240323474060184,
+      "loss": 2.9223,
+      "step": 35596
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6525871753692627,
+      "learning_rate": 0.00022401745311940695,
+      "loss": 3.1616,
+      "step": 35597
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.617184042930603,
+      "learning_rate": 0.00022400255903291886,
+      "loss": 3.071,
+      "step": 35598
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6290184259414673,
+      "learning_rate": 0.00022398766514659347,
+      "loss": 3.1684,
+      "step": 35599
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6642170548439026,
+      "learning_rate": 0.00022397277146046996,
+      "loss": 2.8508,
+      "step": 35600
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6329320073127747,
+      "learning_rate": 0.0002239578779745875,
+      "loss": 2.8697,
+      "step": 35601
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6439346671104431,
+      "learning_rate": 0.00022394298468898533,
+      "loss": 3.0158,
+      "step": 35602
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6102849245071411,
+      "learning_rate": 0.0002239280916037028,
+      "loss": 3.0351,
+      "step": 35603
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6272664070129395,
+      "learning_rate": 0.00022391319871877888,
+      "loss": 2.9033,
+      "step": 35604
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6411396265029907,
+      "learning_rate": 0.000223898306034253,
+      "loss": 2.9436,
+      "step": 35605
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6268124580383301,
+      "learning_rate": 0.00022388341355016443,
+      "loss": 2.9714,
+      "step": 35606
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.608563244342804,
+      "learning_rate": 0.00022386852126655223,
+      "loss": 3.1994,
+      "step": 35607
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6222233176231384,
+      "learning_rate": 0.00022385362918345573,
+      "loss": 2.9368,
+      "step": 35608
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6342154145240784,
+      "learning_rate": 0.00022383873730091403,
+      "loss": 2.9475,
+      "step": 35609
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6370996832847595,
+      "learning_rate": 0.00022382384561896648,
+      "loss": 2.7946,
+      "step": 35610
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6122251749038696,
+      "learning_rate": 0.0002238089541376523,
+      "loss": 2.809,
+      "step": 35611
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6048957109451294,
+      "learning_rate": 0.0002237940628570106,
+      "loss": 2.9704,
+      "step": 35612
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.619551420211792,
+      "learning_rate": 0.0002237791717770808,
+      "loss": 2.8918,
+      "step": 35613
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6651395559310913,
+      "learning_rate": 0.0002237642808979018,
+      "loss": 3.0585,
+      "step": 35614
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6424996852874756,
+      "learning_rate": 0.00022374939021951314,
+      "loss": 2.9588,
+      "step": 35615
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.678436815738678,
+      "learning_rate": 0.00022373449974195396,
+      "loss": 2.9305,
+      "step": 35616
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6074570417404175,
+      "learning_rate": 0.0002237196094652633,
+      "loss": 2.7319,
+      "step": 35617
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6088811755180359,
+      "learning_rate": 0.0002237047193894806,
+      "loss": 2.9016,
+      "step": 35618
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6111059188842773,
+      "learning_rate": 0.00022368982951464497,
+      "loss": 3.071,
+      "step": 35619
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.594634473323822,
+      "learning_rate": 0.00022367493984079553,
+      "loss": 3.0388,
+      "step": 35620
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5961778163909912,
+      "learning_rate": 0.0002236600503679718,
+      "loss": 3.0913,
+      "step": 35621
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6522762179374695,
+      "learning_rate": 0.0002236451610962127,
+      "loss": 2.9231,
+      "step": 35622
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6687374711036682,
+      "learning_rate": 0.00022363027202555762,
+      "loss": 3.0651,
+      "step": 35623
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6435642838478088,
+      "learning_rate": 0.00022361538315604558,
+      "loss": 2.9751,
+      "step": 35624
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.667826235294342,
+      "learning_rate": 0.00022360049448771598,
+      "loss": 3.1817,
+      "step": 35625
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6378101706504822,
+      "learning_rate": 0.00022358560602060807,
+      "loss": 2.9707,
+      "step": 35626
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.620212197303772,
+      "learning_rate": 0.0002235707177547608,
+      "loss": 3.1727,
+      "step": 35627
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6127507090568542,
+      "learning_rate": 0.0002235558296902137,
+      "loss": 3.027,
+      "step": 35628
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6060869693756104,
+      "learning_rate": 0.00022354094182700575,
+      "loss": 3.2523,
+      "step": 35629
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6330663561820984,
+      "learning_rate": 0.00022352605416517618,
+      "loss": 2.9645,
+      "step": 35630
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6215522289276123,
+      "learning_rate": 0.0002235111667047644,
+      "loss": 2.9589,
+      "step": 35631
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6048828363418579,
+      "learning_rate": 0.00022349627944580946,
+      "loss": 2.9766,
+      "step": 35632
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.607565701007843,
+      "learning_rate": 0.00022348139238835064,
+      "loss": 3.177,
+      "step": 35633
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6367427110671997,
+      "learning_rate": 0.00022346650553242695,
+      "loss": 3.0012,
+      "step": 35634
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6365235447883606,
+      "learning_rate": 0.00022345161887807774,
+      "loss": 3.0022,
+      "step": 35635
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6165895462036133,
+      "learning_rate": 0.00022343673242534244,
+      "loss": 2.8973,
+      "step": 35636
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6315513849258423,
+      "learning_rate": 0.00022342184617425993,
+      "loss": 2.887,
+      "step": 35637
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6446321606636047,
+      "learning_rate": 0.0002234069601248696,
+      "loss": 3.0301,
+      "step": 35638
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6023744940757751,
+      "learning_rate": 0.00022339207427721049,
+      "loss": 2.8683,
+      "step": 35639
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6446340680122375,
+      "learning_rate": 0.00022337718863132196,
+      "loss": 3.1663,
+      "step": 35640
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6004554629325867,
+      "learning_rate": 0.00022336230318724325,
+      "loss": 3.0679,
+      "step": 35641
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6370178461074829,
+      "learning_rate": 0.00022334741794501333,
+      "loss": 2.9213,
+      "step": 35642
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6428967118263245,
+      "learning_rate": 0.0002233325329046717,
+      "loss": 2.9016,
+      "step": 35643
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.685459554195404,
+      "learning_rate": 0.0002233176480662574,
+      "loss": 2.7608,
+      "step": 35644
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6399477124214172,
+      "learning_rate": 0.00022330276342980956,
+      "loss": 2.9918,
+      "step": 35645
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.64689040184021,
+      "learning_rate": 0.0002232878789953676,
+      "loss": 3.0101,
+      "step": 35646
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6764118671417236,
+      "learning_rate": 0.00022327299476297054,
+      "loss": 3.169,
+      "step": 35647
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6111961603164673,
+      "learning_rate": 0.00022325811073265774,
+      "loss": 2.777,
+      "step": 35648
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6330588459968567,
+      "learning_rate": 0.00022324322690446815,
+      "loss": 2.8875,
+      "step": 35649
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6471860408782959,
+      "learning_rate": 0.00022322834327844117,
+      "loss": 3.055,
+      "step": 35650
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6193934679031372,
+      "learning_rate": 0.00022321345985461605,
+      "loss": 2.9988,
+      "step": 35651
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.627863883972168,
+      "learning_rate": 0.00022319857663303177,
+      "loss": 3.1429,
+      "step": 35652
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6523374915122986,
+      "learning_rate": 0.0002231836936137278,
+      "loss": 3.1287,
+      "step": 35653
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6731512546539307,
+      "learning_rate": 0.00022316881079674305,
+      "loss": 2.9187,
+      "step": 35654
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6700525879859924,
+      "learning_rate": 0.00022315392818211694,
+      "loss": 2.8942,
+      "step": 35655
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6407337784767151,
+      "learning_rate": 0.00022313904576988866,
+      "loss": 3.2872,
+      "step": 35656
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6617950797080994,
+      "learning_rate": 0.00022312416356009722,
+      "loss": 3.0562,
+      "step": 35657
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6472514867782593,
+      "learning_rate": 0.000223109281552782,
+      "loss": 3.138,
+      "step": 35658
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.669858992099762,
+      "learning_rate": 0.00022309439974798213,
+      "loss": 2.8543,
+      "step": 35659
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5877886414527893,
+      "learning_rate": 0.00022307951814573676,
+      "loss": 3.2045,
+      "step": 35660
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5913372039794922,
+      "learning_rate": 0.00022306463674608525,
+      "loss": 3.2414,
+      "step": 35661
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6737897992134094,
+      "learning_rate": 0.0002230497555490666,
+      "loss": 3.0833,
+      "step": 35662
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6321114897727966,
+      "learning_rate": 0.00022303487455472017,
+      "loss": 2.9484,
+      "step": 35663
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6263748407363892,
+      "learning_rate": 0.00022301999376308492,
+      "loss": 3.1179,
+      "step": 35664
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7027498483657837,
+      "learning_rate": 0.00022300511317420023,
+      "loss": 2.8271,
+      "step": 35665
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6153057813644409,
+      "learning_rate": 0.0002229902327881054,
+      "loss": 3.2247,
+      "step": 35666
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6329478621482849,
+      "learning_rate": 0.0002229753526048393,
+      "loss": 2.8801,
+      "step": 35667
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6193848848342896,
+      "learning_rate": 0.0002229604726244414,
+      "loss": 2.9691,
+      "step": 35668
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6393882036209106,
+      "learning_rate": 0.00022294559284695076,
+      "loss": 2.9422,
+      "step": 35669
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6423365473747253,
+      "learning_rate": 0.00022293071327240656,
+      "loss": 3.0597,
+      "step": 35670
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6023625135421753,
+      "learning_rate": 0.00022291583390084808,
+      "loss": 2.9063,
+      "step": 35671
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6385714411735535,
+      "learning_rate": 0.00022290095473231444,
+      "loss": 2.9283,
+      "step": 35672
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6258054971694946,
+      "learning_rate": 0.0002228860757668449,
+      "loss": 3.0131,
+      "step": 35673
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6036873459815979,
+      "learning_rate": 0.0002228711970044785,
+      "loss": 3.2195,
+      "step": 35674
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6359200477600098,
+      "learning_rate": 0.0002228563184452545,
+      "loss": 3.214,
+      "step": 35675
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6024333834648132,
+      "learning_rate": 0.00022284144008921224,
+      "loss": 2.9977,
+      "step": 35676
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6681496500968933,
+      "learning_rate": 0.00022282656193639067,
+      "loss": 3.1888,
+      "step": 35677
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6251303553581238,
+      "learning_rate": 0.0002228116839868292,
+      "loss": 2.8912,
+      "step": 35678
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6392276287078857,
+      "learning_rate": 0.00022279680624056678,
+      "loss": 3.1644,
+      "step": 35679
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6303515434265137,
+      "learning_rate": 0.00022278192869764282,
+      "loss": 3.0278,
+      "step": 35680
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6725468635559082,
+      "learning_rate": 0.00022276705135809631,
+      "loss": 3.1027,
+      "step": 35681
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6221615672111511,
+      "learning_rate": 0.0002227521742219665,
+      "loss": 3.0331,
+      "step": 35682
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6534598469734192,
+      "learning_rate": 0.00022273729728929272,
+      "loss": 3.0542,
+      "step": 35683
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6483638286590576,
+      "learning_rate": 0.00022272242056011396,
+      "loss": 2.9716,
+      "step": 35684
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.636474072933197,
+      "learning_rate": 0.0002227075440344695,
+      "loss": 3.0775,
+      "step": 35685
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6173744201660156,
+      "learning_rate": 0.00022269266771239842,
+      "loss": 2.879,
+      "step": 35686
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6293776631355286,
+      "learning_rate": 0.00022267779159394,
+      "loss": 3.0981,
+      "step": 35687
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.624398946762085,
+      "learning_rate": 0.0002226629156791335,
+      "loss": 2.7682,
+      "step": 35688
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.647941529750824,
+      "learning_rate": 0.00022264803996801787,
+      "loss": 3.103,
+      "step": 35689
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6048708558082581,
+      "learning_rate": 0.00022263316446063247,
+      "loss": 2.9964,
+      "step": 35690
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6444445848464966,
+      "learning_rate": 0.00022261828915701642,
+      "loss": 3.1067,
+      "step": 35691
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6575992703437805,
+      "learning_rate": 0.00022260341405720886,
+      "loss": 3.024,
+      "step": 35692
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6296408772468567,
+      "learning_rate": 0.00022258853916124908,
+      "loss": 3.1385,
+      "step": 35693
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6155877709388733,
+      "learning_rate": 0.00022257366446917616,
+      "loss": 3.0299,
+      "step": 35694
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6629244089126587,
+      "learning_rate": 0.00022255878998102936,
+      "loss": 2.886,
+      "step": 35695
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6616221070289612,
+      "learning_rate": 0.00022254391569684768,
+      "loss": 3.066,
+      "step": 35696
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.630438506603241,
+      "learning_rate": 0.00022252904161667043,
+      "loss": 3.0545,
+      "step": 35697
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6666058301925659,
+      "learning_rate": 0.00022251416774053686,
+      "loss": 2.9689,
+      "step": 35698
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6253411173820496,
+      "learning_rate": 0.00022249929406848609,
+      "loss": 2.8768,
+      "step": 35699
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6282232403755188,
+      "learning_rate": 0.00022248442060055724,
+      "loss": 2.9593,
+      "step": 35700
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5989751219749451,
+      "learning_rate": 0.0002224695473367894,
+      "loss": 2.8903,
+      "step": 35701
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6522663831710815,
+      "learning_rate": 0.00022245467427722194,
+      "loss": 2.9982,
+      "step": 35702
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6247538328170776,
+      "learning_rate": 0.00022243980142189397,
+      "loss": 2.8845,
+      "step": 35703
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6098034977912903,
+      "learning_rate": 0.00022242492877084453,
+      "loss": 2.941,
+      "step": 35704
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6762757301330566,
+      "learning_rate": 0.00022241005632411305,
+      "loss": 3.0948,
+      "step": 35705
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6396958827972412,
+      "learning_rate": 0.00022239518408173842,
+      "loss": 3.0555,
+      "step": 35706
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6262756586074829,
+      "learning_rate": 0.00022238031204375994,
+      "loss": 3.0633,
+      "step": 35707
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6355347037315369,
+      "learning_rate": 0.0002223654402102169,
+      "loss": 3.2732,
+      "step": 35708
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6595379114151001,
+      "learning_rate": 0.00022235056858114824,
+      "loss": 2.9131,
+      "step": 35709
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5768058896064758,
+      "learning_rate": 0.00022233569715659335,
+      "loss": 3.0761,
+      "step": 35710
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6388674974441528,
+      "learning_rate": 0.00022232082593659112,
+      "loss": 3.2021,
+      "step": 35711
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6107366681098938,
+      "learning_rate": 0.00022230595492118098,
+      "loss": 2.9706,
+      "step": 35712
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6297752857208252,
+      "learning_rate": 0.00022229108411040205,
+      "loss": 3.1894,
+      "step": 35713
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6354250311851501,
+      "learning_rate": 0.00022227621350429333,
+      "loss": 3.2159,
+      "step": 35714
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6811100244522095,
+      "learning_rate": 0.00022226134310289426,
+      "loss": 3.0646,
+      "step": 35715
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6282296776771545,
+      "learning_rate": 0.00022224647290624365,
+      "loss": 3.1562,
+      "step": 35716
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6105203628540039,
+      "learning_rate": 0.00022223160291438097,
+      "loss": 3.07,
+      "step": 35717
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5900524258613586,
+      "learning_rate": 0.00022221673312734536,
+      "loss": 3.1058,
+      "step": 35718
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6298518180847168,
+      "learning_rate": 0.00022220186354517576,
+      "loss": 3.0704,
+      "step": 35719
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6291689872741699,
+      "learning_rate": 0.00022218699416791163,
+      "loss": 2.9188,
+      "step": 35720
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5981619358062744,
+      "learning_rate": 0.00022217212499559192,
+      "loss": 3.0848,
+      "step": 35721
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6223424673080444,
+      "learning_rate": 0.00022215725602825573,
+      "loss": 2.9289,
+      "step": 35722
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6437315344810486,
+      "learning_rate": 0.00022214238726594253,
+      "loss": 3.2998,
+      "step": 35723
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6619381904602051,
+      "learning_rate": 0.0002221275187086912,
+      "loss": 2.8725,
+      "step": 35724
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.624908983707428,
+      "learning_rate": 0.0002221126503565411,
+      "loss": 3.1948,
+      "step": 35725
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6466329097747803,
+      "learning_rate": 0.00022209778220953115,
+      "loss": 2.9206,
+      "step": 35726
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6283143162727356,
+      "learning_rate": 0.00022208291426770068,
+      "loss": 3.0594,
+      "step": 35727
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6625270843505859,
+      "learning_rate": 0.00022206804653108895,
+      "loss": 2.9291,
+      "step": 35728
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6456392407417297,
+      "learning_rate": 0.0002220531789997348,
+      "loss": 2.9168,
+      "step": 35729
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.799959659576416,
+      "learning_rate": 0.00022203831167367773,
+      "loss": 3.0285,
+      "step": 35730
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6306584477424622,
+      "learning_rate": 0.00022202344455295663,
+      "loss": 2.7562,
+      "step": 35731
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7043753266334534,
+      "learning_rate": 0.00022200857763761078,
+      "loss": 2.8602,
+      "step": 35732
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.656856119632721,
+      "learning_rate": 0.00022199371092767944,
+      "loss": 2.9566,
+      "step": 35733
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.638849139213562,
+      "learning_rate": 0.00022197884442320154,
+      "loss": 2.9435,
+      "step": 35734
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6164271831512451,
+      "learning_rate": 0.00022196397812421646,
+      "loss": 2.8935,
+      "step": 35735
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6180526614189148,
+      "learning_rate": 0.0002219491120307631,
+      "loss": 3.05,
+      "step": 35736
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6781109571456909,
+      "learning_rate": 0.00022193424614288076,
+      "loss": 3.0915,
+      "step": 35737
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6297280788421631,
+      "learning_rate": 0.0002219193804606087,
+      "loss": 2.8888,
+      "step": 35738
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6124048829078674,
+      "learning_rate": 0.00022190451498398594,
+      "loss": 3.0255,
+      "step": 35739
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6476908922195435,
+      "learning_rate": 0.00022188964971305169,
+      "loss": 3.111,
+      "step": 35740
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.602753221988678,
+      "learning_rate": 0.0002218747846478449,
+      "loss": 3.2071,
+      "step": 35741
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6379673480987549,
+      "learning_rate": 0.00022185991978840503,
+      "loss": 2.9443,
+      "step": 35742
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.675123393535614,
+      "learning_rate": 0.00022184505513477115,
+      "loss": 2.9927,
+      "step": 35743
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.615736186504364,
+      "learning_rate": 0.0002218301906869822,
+      "loss": 2.928,
+      "step": 35744
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6254475116729736,
+      "learning_rate": 0.0002218153264450776,
+      "loss": 3.0137,
+      "step": 35745
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5977519750595093,
+      "learning_rate": 0.00022180046240909634,
+      "loss": 2.7181,
+      "step": 35746
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.659858763217926,
+      "learning_rate": 0.00022178559857907759,
+      "loss": 3.1556,
+      "step": 35747
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.634911060333252,
+      "learning_rate": 0.00022177073495506058,
+      "loss": 3.1182,
+      "step": 35748
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6552413105964661,
+      "learning_rate": 0.00022175587153708438,
+      "loss": 3.0402,
+      "step": 35749
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5962386131286621,
+      "learning_rate": 0.00022174100832518819,
+      "loss": 3.092,
+      "step": 35750
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6224254965782166,
+      "learning_rate": 0.00022172614531941102,
+      "loss": 2.9382,
+      "step": 35751
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6392555236816406,
+      "learning_rate": 0.00022171128251979218,
+      "loss": 3.0492,
+      "step": 35752
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6384632587432861,
+      "learning_rate": 0.0002216964199263708,
+      "loss": 3.043,
+      "step": 35753
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6350359320640564,
+      "learning_rate": 0.00022168155753918585,
+      "loss": 3.0483,
+      "step": 35754
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6259121298789978,
+      "learning_rate": 0.00022166669535827674,
+      "loss": 3.1861,
+      "step": 35755
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6385367512702942,
+      "learning_rate": 0.0002216518333836824,
+      "loss": 2.9332,
+      "step": 35756
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7472400069236755,
+      "learning_rate": 0.00022163697161544217,
+      "loss": 2.9411,
+      "step": 35757
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6142936944961548,
+      "learning_rate": 0.00022162211005359486,
+      "loss": 3.0595,
+      "step": 35758
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6409468054771423,
+      "learning_rate": 0.0002216072486981799,
+      "loss": 3.0586,
+      "step": 35759
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6137709617614746,
+      "learning_rate": 0.00022159238754923647,
+      "loss": 2.9876,
+      "step": 35760
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6863341331481934,
+      "learning_rate": 0.00022157752660680353,
+      "loss": 3.1946,
+      "step": 35761
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.615785539150238,
+      "learning_rate": 0.00022156266587092038,
+      "loss": 2.8858,
+      "step": 35762
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6471198201179504,
+      "learning_rate": 0.00022154780534162593,
+      "loss": 3.07,
+      "step": 35763
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6607215404510498,
+      "learning_rate": 0.0002215329450189595,
+      "loss": 3.1074,
+      "step": 35764
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6380542516708374,
+      "learning_rate": 0.0002215180849029603,
+      "loss": 3.0175,
+      "step": 35765
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5926660895347595,
+      "learning_rate": 0.00022150322499366718,
+      "loss": 3.1769,
+      "step": 35766
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6166107058525085,
+      "learning_rate": 0.00022148836529111964,
+      "loss": 2.812,
+      "step": 35767
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6104620695114136,
+      "learning_rate": 0.00022147350579535652,
+      "loss": 3.1454,
+      "step": 35768
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6072012186050415,
+      "learning_rate": 0.00022145864650641704,
+      "loss": 3.0505,
+      "step": 35769
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6567019820213318,
+      "learning_rate": 0.00022144378742434054,
+      "loss": 2.9977,
+      "step": 35770
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6445237398147583,
+      "learning_rate": 0.00022142892854916586,
+      "loss": 2.9995,
+      "step": 35771
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6039271354675293,
+      "learning_rate": 0.0002214140698809323,
+      "loss": 3.1029,
+      "step": 35772
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6350820660591125,
+      "learning_rate": 0.00022139921141967885,
+      "loss": 2.942,
+      "step": 35773
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6840111017227173,
+      "learning_rate": 0.00022138435316544487,
+      "loss": 2.9071,
+      "step": 35774
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7613235712051392,
+      "learning_rate": 0.00022136949511826942,
+      "loss": 2.7632,
+      "step": 35775
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.651785671710968,
+      "learning_rate": 0.00022135463727819142,
+      "loss": 2.9729,
+      "step": 35776
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.62009596824646,
+      "learning_rate": 0.00022133977964525033,
+      "loss": 3.1663,
+      "step": 35777
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6254829168319702,
+      "learning_rate": 0.00022132492221948498,
+      "loss": 2.9613,
+      "step": 35778
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6415801048278809,
+      "learning_rate": 0.00022131006500093468,
+      "loss": 3.0607,
+      "step": 35779
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6973337531089783,
+      "learning_rate": 0.00022129520798963863,
+      "loss": 2.6692,
+      "step": 35780
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.614383339881897,
+      "learning_rate": 0.0002212803511856357,
+      "loss": 2.8457,
+      "step": 35781
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6090778708457947,
+      "learning_rate": 0.00022126549458896535,
+      "loss": 3.0376,
+      "step": 35782
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6137256622314453,
+      "learning_rate": 0.00022125063819966644,
+      "loss": 2.8509,
+      "step": 35783
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6132293939590454,
+      "learning_rate": 0.0002212357820177781,
+      "loss": 3.2059,
+      "step": 35784
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6477642059326172,
+      "learning_rate": 0.00022122092604333975,
+      "loss": 3.0761,
+      "step": 35785
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6460610032081604,
+      "learning_rate": 0.00022120607027639017,
+      "loss": 3.0912,
+      "step": 35786
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.628158450126648,
+      "learning_rate": 0.00022119121471696874,
+      "loss": 3.2221,
+      "step": 35787
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6395934820175171,
+      "learning_rate": 0.00022117635936511437,
+      "loss": 2.687,
+      "step": 35788
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6341500282287598,
+      "learning_rate": 0.00022116150422086633,
+      "loss": 2.9476,
+      "step": 35789
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5902708172798157,
+      "learning_rate": 0.0002211466492842638,
+      "loss": 3.0437,
+      "step": 35790
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6335721611976624,
+      "learning_rate": 0.00022113179455534573,
+      "loss": 2.9241,
+      "step": 35791
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6263195872306824,
+      "learning_rate": 0.00022111694003415145,
+      "loss": 2.9259,
+      "step": 35792
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6031926870346069,
+      "learning_rate": 0.00022110208572071984,
+      "loss": 2.9516,
+      "step": 35793
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6146538257598877,
+      "learning_rate": 0.00022108723161509014,
+      "loss": 2.9278,
+      "step": 35794
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6239001154899597,
+      "learning_rate": 0.0002210723777173016,
+      "loss": 2.8726,
+      "step": 35795
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6321032047271729,
+      "learning_rate": 0.00022105752402739313,
+      "loss": 3.0842,
+      "step": 35796
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6525657773017883,
+      "learning_rate": 0.0002210426705454041,
+      "loss": 2.8471,
+      "step": 35797
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6774393320083618,
+      "learning_rate": 0.00022102781727137327,
+      "loss": 3.2696,
+      "step": 35798
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7013652324676514,
+      "learning_rate": 0.00022101296420534,
+      "loss": 2.9296,
+      "step": 35799
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6332024931907654,
+      "learning_rate": 0.0002209981113473435,
+      "loss": 2.9423,
+      "step": 35800
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.64900803565979,
+      "learning_rate": 0.00022098325869742273,
+      "loss": 2.9368,
+      "step": 35801
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6335112452507019,
+      "learning_rate": 0.0002209684062556169,
+      "loss": 3.258,
+      "step": 35802
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5994589328765869,
+      "learning_rate": 0.00022095355402196494,
+      "loss": 3.1889,
+      "step": 35803
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6108216643333435,
+      "learning_rate": 0.0002209387019965062,
+      "loss": 3.0253,
+      "step": 35804
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6158538460731506,
+      "learning_rate": 0.00022092385017927977,
+      "loss": 2.9247,
+      "step": 35805
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6153451204299927,
+      "learning_rate": 0.0002209089985703245,
+      "loss": 3.0619,
+      "step": 35806
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6863804459571838,
+      "learning_rate": 0.00022089414716967992,
+      "loss": 2.8898,
+      "step": 35807
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6681492924690247,
+      "learning_rate": 0.00022087929597738482,
+      "loss": 2.9785,
+      "step": 35808
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6478404998779297,
+      "learning_rate": 0.00022086444499347835,
+      "loss": 3.0399,
+      "step": 35809
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6117250919342041,
+      "learning_rate": 0.00022084959421799988,
+      "loss": 3.1075,
+      "step": 35810
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6180821657180786,
+      "learning_rate": 0.00022083474365098825,
+      "loss": 3.2999,
+      "step": 35811
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6459303498268127,
+      "learning_rate": 0.00022081989329248276,
+      "loss": 2.9466,
+      "step": 35812
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6436235904693604,
+      "learning_rate": 0.00022080504314252227,
+      "loss": 3.0516,
+      "step": 35813
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.627779483795166,
+      "learning_rate": 0.0002207901932011461,
+      "loss": 3.0361,
+      "step": 35814
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5869961977005005,
+      "learning_rate": 0.00022077534346839344,
+      "loss": 2.9429,
+      "step": 35815
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6305521726608276,
+      "learning_rate": 0.0002207604939443031,
+      "loss": 3.2122,
+      "step": 35816
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.621493399143219,
+      "learning_rate": 0.00022074564462891454,
+      "loss": 2.9271,
+      "step": 35817
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6501767039299011,
+      "learning_rate": 0.0002207307955222666,
+      "loss": 2.8088,
+      "step": 35818
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6253821849822998,
+      "learning_rate": 0.00022071594662439846,
+      "loss": 3.0041,
+      "step": 35819
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6286084651947021,
+      "learning_rate": 0.00022070109793534935,
+      "loss": 2.8641,
+      "step": 35820
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8458594083786011,
+      "learning_rate": 0.00022068624945515818,
+      "loss": 3.1558,
+      "step": 35821
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6667223572731018,
+      "learning_rate": 0.00022067140118386432,
+      "loss": 2.9657,
+      "step": 35822
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6399785280227661,
+      "learning_rate": 0.00022065655312150664,
+      "loss": 2.9304,
+      "step": 35823
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6804106831550598,
+      "learning_rate": 0.00022064170526812425,
+      "loss": 3.1718,
+      "step": 35824
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6263512969017029,
+      "learning_rate": 0.00022062685762375646,
+      "loss": 2.7909,
+      "step": 35825
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6424779891967773,
+      "learning_rate": 0.00022061201018844218,
+      "loss": 3.1985,
+      "step": 35826
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7617523074150085,
+      "learning_rate": 0.00022059716296222065,
+      "loss": 2.9046,
+      "step": 35827
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6302613019943237,
+      "learning_rate": 0.0002205823159451308,
+      "loss": 3.043,
+      "step": 35828
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6046407222747803,
+      "learning_rate": 0.00022056746913721192,
+      "loss": 3.1451,
+      "step": 35829
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.634103536605835,
+      "learning_rate": 0.0002205526225385031,
+      "loss": 3.1131,
+      "step": 35830
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6643496155738831,
+      "learning_rate": 0.0002205377761490432,
+      "loss": 3.097,
+      "step": 35831
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6304755806922913,
+      "learning_rate": 0.00022052292996887173,
+      "loss": 3.1338,
+      "step": 35832
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6238082051277161,
+      "learning_rate": 0.00022050808399802743,
+      "loss": 2.871,
+      "step": 35833
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6616991758346558,
+      "learning_rate": 0.00022049323823654962,
+      "loss": 3.0218,
+      "step": 35834
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6222161650657654,
+      "learning_rate": 0.0002204783926844772,
+      "loss": 3.1102,
+      "step": 35835
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6417911648750305,
+      "learning_rate": 0.00022046354734184945,
+      "loss": 2.897,
+      "step": 35836
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5990554094314575,
+      "learning_rate": 0.00022044870220870545,
+      "loss": 2.9511,
+      "step": 35837
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5991533994674683,
+      "learning_rate": 0.00022043385728508417,
+      "loss": 2.89,
+      "step": 35838
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6272563934326172,
+      "learning_rate": 0.0002204190125710249,
+      "loss": 3.0124,
+      "step": 35839
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6233096122741699,
+      "learning_rate": 0.00022040416806656652,
+      "loss": 3.1823,
+      "step": 35840
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6353172659873962,
+      "learning_rate": 0.0002203893237717483,
+      "loss": 2.8572,
+      "step": 35841
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6314014792442322,
+      "learning_rate": 0.00022037447968660935,
+      "loss": 2.7837,
+      "step": 35842
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6511485576629639,
+      "learning_rate": 0.00022035963581118855,
+      "loss": 3.0148,
+      "step": 35843
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6228102445602417,
+      "learning_rate": 0.00022034479214552528,
+      "loss": 2.9736,
+      "step": 35844
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5977052450180054,
+      "learning_rate": 0.00022032994868965844,
+      "loss": 3.0123,
+      "step": 35845
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6566368341445923,
+      "learning_rate": 0.0002203151054436271,
+      "loss": 3.0192,
+      "step": 35846
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.620186984539032,
+      "learning_rate": 0.00022030026240747058,
+      "loss": 3.0604,
+      "step": 35847
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6430531144142151,
+      "learning_rate": 0.00022028541958122775,
+      "loss": 3.0086,
+      "step": 35848
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6357125639915466,
+      "learning_rate": 0.00022027057696493785,
+      "loss": 2.8469,
+      "step": 35849
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6521148085594177,
+      "learning_rate": 0.0002202557345586398,
+      "loss": 2.7497,
+      "step": 35850
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6499366164207458,
+      "learning_rate": 0.00022024089236237287,
+      "loss": 3.2067,
+      "step": 35851
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6402422785758972,
+      "learning_rate": 0.00022022605037617612,
+      "loss": 3.1193,
+      "step": 35852
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6213492155075073,
+      "learning_rate": 0.00022021120860008848,
+      "loss": 3.0745,
+      "step": 35853
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6546719074249268,
+      "learning_rate": 0.00022019636703414928,
+      "loss": 3.1135,
+      "step": 35854
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.627903163433075,
+      "learning_rate": 0.00022018152567839747,
+      "loss": 3.1131,
+      "step": 35855
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.628490686416626,
+      "learning_rate": 0.00022016668453287203,
+      "loss": 2.8469,
+      "step": 35856
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6247190833091736,
+      "learning_rate": 0.00022015184359761237,
+      "loss": 3.0752,
+      "step": 35857
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6584323048591614,
+      "learning_rate": 0.00022013700287265727,
+      "loss": 2.7668,
+      "step": 35858
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6163456439971924,
+      "learning_rate": 0.000220122162358046,
+      "loss": 3.1157,
+      "step": 35859
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6366605162620544,
+      "learning_rate": 0.0002201073220538175,
+      "loss": 2.9302,
+      "step": 35860
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7113758325576782,
+      "learning_rate": 0.0002200924819600109,
+      "loss": 3.1147,
+      "step": 35861
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5954180955886841,
+      "learning_rate": 0.00022007764207666546,
+      "loss": 2.9149,
+      "step": 35862
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6021776795387268,
+      "learning_rate": 0.00022006280240382007,
+      "loss": 3.0365,
+      "step": 35863
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.617344856262207,
+      "learning_rate": 0.00022004796294151392,
+      "loss": 3.2413,
+      "step": 35864
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6328892707824707,
+      "learning_rate": 0.00022003312368978597,
+      "loss": 3.1775,
+      "step": 35865
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6340292096138,
+      "learning_rate": 0.0002200182846486754,
+      "loss": 2.7447,
+      "step": 35866
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6463967561721802,
+      "learning_rate": 0.00022000344581822137,
+      "loss": 2.9128,
+      "step": 35867
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6408669352531433,
+      "learning_rate": 0.0002199886071984627,
+      "loss": 3.0877,
+      "step": 35868
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6117002367973328,
+      "learning_rate": 0.0002199737687894388,
+      "loss": 3.2649,
+      "step": 35869
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6672056913375854,
+      "learning_rate": 0.0002199589305911885,
+      "loss": 2.74,
+      "step": 35870
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6419618129730225,
+      "learning_rate": 0.0002199440926037509,
+      "loss": 2.9445,
+      "step": 35871
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.165690302848816,
+      "learning_rate": 0.00021992925482716534,
+      "loss": 3.0519,
+      "step": 35872
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6113609671592712,
+      "learning_rate": 0.0002199144172614706,
+      "loss": 3.0843,
+      "step": 35873
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6271687746047974,
+      "learning_rate": 0.00021989957990670595,
+      "loss": 2.9394,
+      "step": 35874
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6276607513427734,
+      "learning_rate": 0.00021988474276291028,
+      "loss": 3.0057,
+      "step": 35875
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6973236203193665,
+      "learning_rate": 0.00021986990583012283,
+      "loss": 2.699,
+      "step": 35876
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6413176655769348,
+      "learning_rate": 0.00021985506910838267,
+      "loss": 3.0356,
+      "step": 35877
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6662982106208801,
+      "learning_rate": 0.00021984023259772874,
+      "loss": 2.9587,
+      "step": 35878
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6634508967399597,
+      "learning_rate": 0.0002198253962982003,
+      "loss": 2.69,
+      "step": 35879
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6144524812698364,
+      "learning_rate": 0.00021981056020983623,
+      "loss": 2.9196,
+      "step": 35880
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6477610468864441,
+      "learning_rate": 0.0002197957243326758,
+      "loss": 3.0727,
+      "step": 35881
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6386900544166565,
+      "learning_rate": 0.00021978088866675803,
+      "loss": 2.9197,
+      "step": 35882
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6349610686302185,
+      "learning_rate": 0.00021976605321212184,
+      "loss": 3.2623,
+      "step": 35883
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6554539799690247,
+      "learning_rate": 0.00021975121796880657,
+      "loss": 3.2208,
+      "step": 35884
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6648715138435364,
+      "learning_rate": 0.00021973638293685104,
+      "loss": 2.9701,
+      "step": 35885
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6611798405647278,
+      "learning_rate": 0.00021972154811629438,
+      "loss": 3.0644,
+      "step": 35886
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6232627034187317,
+      "learning_rate": 0.00021970671350717588,
+      "loss": 3.0797,
+      "step": 35887
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6150153875350952,
+      "learning_rate": 0.00021969187910953435,
+      "loss": 2.9063,
+      "step": 35888
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6366429328918457,
+      "learning_rate": 0.000219677044923409,
+      "loss": 2.9047,
+      "step": 35889
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6754364967346191,
+      "learning_rate": 0.00021966221094883877,
+      "loss": 3.1375,
+      "step": 35890
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5876303315162659,
+      "learning_rate": 0.00021964737718586287,
+      "loss": 2.973,
+      "step": 35891
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7163461446762085,
+      "learning_rate": 0.00021963254363452036,
+      "loss": 2.9597,
+      "step": 35892
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6201485991477966,
+      "learning_rate": 0.0002196177102948501,
+      "loss": 2.992,
+      "step": 35893
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6001695990562439,
+      "learning_rate": 0.0002196028771668915,
+      "loss": 3.0862,
+      "step": 35894
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6142009496688843,
+      "learning_rate": 0.0002195880442506834,
+      "loss": 2.9972,
+      "step": 35895
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6313881278038025,
+      "learning_rate": 0.00021957321154626483,
+      "loss": 3.204,
+      "step": 35896
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6144121885299683,
+      "learning_rate": 0.00021955837905367507,
+      "loss": 2.8235,
+      "step": 35897
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6347367763519287,
+      "learning_rate": 0.00021954354677295297,
+      "loss": 2.9528,
+      "step": 35898
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.655368447303772,
+      "learning_rate": 0.00021952871470413778,
+      "loss": 2.9214,
+      "step": 35899
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6603178977966309,
+      "learning_rate": 0.00021951388284726834,
+      "loss": 2.9844,
+      "step": 35900
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6305208802223206,
+      "learning_rate": 0.00021949905120238384,
+      "loss": 2.933,
+      "step": 35901
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6413267850875854,
+      "learning_rate": 0.00021948421976952345,
+      "loss": 2.8795,
+      "step": 35902
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6816225647926331,
+      "learning_rate": 0.00021946938854872607,
+      "loss": 3.0087,
+      "step": 35903
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6837611794471741,
+      "learning_rate": 0.00021945455754003088,
+      "loss": 2.9229,
+      "step": 35904
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7396782040596008,
+      "learning_rate": 0.0002194397267434768,
+      "loss": 3.0543,
+      "step": 35905
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6411307454109192,
+      "learning_rate": 0.00021942489615910303,
+      "loss": 2.8873,
+      "step": 35906
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6512781381607056,
+      "learning_rate": 0.00021941006578694863,
+      "loss": 3.1282,
+      "step": 35907
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6359968185424805,
+      "learning_rate": 0.00021939523562705242,
+      "loss": 3.0821,
+      "step": 35908
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6434915661811829,
+      "learning_rate": 0.00021938040567945387,
+      "loss": 3.2044,
+      "step": 35909
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6505936980247498,
+      "learning_rate": 0.0002193655759441917,
+      "loss": 2.8933,
+      "step": 35910
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6163533926010132,
+      "learning_rate": 0.00021935074642130515,
+      "loss": 2.9099,
+      "step": 35911
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6336808800697327,
+      "learning_rate": 0.0002193359171108331,
+      "loss": 2.9613,
+      "step": 35912
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6180527806282043,
+      "learning_rate": 0.00021932108801281478,
+      "loss": 3.2311,
+      "step": 35913
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6565324068069458,
+      "learning_rate": 0.00021930625912728926,
+      "loss": 3.0751,
+      "step": 35914
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6686374545097351,
+      "learning_rate": 0.00021929143045429535,
+      "loss": 3.0744,
+      "step": 35915
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6567208170890808,
+      "learning_rate": 0.00021927660199387247,
+      "loss": 3.1457,
+      "step": 35916
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6428233981132507,
+      "learning_rate": 0.00021926177374605938,
+      "loss": 3.1215,
+      "step": 35917
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6695310473442078,
+      "learning_rate": 0.00021924694571089518,
+      "loss": 3.1795,
+      "step": 35918
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6748586297035217,
+      "learning_rate": 0.0002192321178884191,
+      "loss": 3.2222,
+      "step": 35919
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6294280290603638,
+      "learning_rate": 0.0002192172902786701,
+      "loss": 2.8973,
+      "step": 35920
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6525007486343384,
+      "learning_rate": 0.00021920246288168718,
+      "loss": 2.9294,
+      "step": 35921
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6621073484420776,
+      "learning_rate": 0.00021918763569750934,
+      "loss": 3.0736,
+      "step": 35922
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.620073139667511,
+      "learning_rate": 0.00021917280872617566,
+      "loss": 3.2611,
+      "step": 35923
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.641943633556366,
+      "learning_rate": 0.00021915798196772544,
+      "loss": 2.9386,
+      "step": 35924
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6125879287719727,
+      "learning_rate": 0.00021914315542219742,
+      "loss": 2.8696,
+      "step": 35925
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6328999996185303,
+      "learning_rate": 0.0002191283290896309,
+      "loss": 2.9842,
+      "step": 35926
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.582988977432251,
+      "learning_rate": 0.00021911350297006458,
+      "loss": 2.8296,
+      "step": 35927
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6015204787254333,
+      "learning_rate": 0.00021909867706353787,
+      "loss": 3.0296,
+      "step": 35928
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6688986420631409,
+      "learning_rate": 0.00021908385137008968,
+      "loss": 2.8411,
+      "step": 35929
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6483864188194275,
+      "learning_rate": 0.00021906902588975896,
+      "loss": 3.0242,
+      "step": 35930
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6931516528129578,
+      "learning_rate": 0.00021905420062258497,
+      "loss": 2.9332,
+      "step": 35931
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6430986523628235,
+      "learning_rate": 0.00021903937556860657,
+      "loss": 2.9444,
+      "step": 35932
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6340141296386719,
+      "learning_rate": 0.00021902455072786282,
+      "loss": 2.7925,
+      "step": 35933
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.608032763004303,
+      "learning_rate": 0.0002190097261003929,
+      "loss": 3.0971,
+      "step": 35934
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6480880379676819,
+      "learning_rate": 0.0002189949016862358,
+      "loss": 2.8061,
+      "step": 35935
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6553376317024231,
+      "learning_rate": 0.00021898007748543053,
+      "loss": 3.0343,
+      "step": 35936
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6663204431533813,
+      "learning_rate": 0.00021896525349801606,
+      "loss": 3.0759,
+      "step": 35937
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6248182654380798,
+      "learning_rate": 0.00021895042972403154,
+      "loss": 2.7778,
+      "step": 35938
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6360815167427063,
+      "learning_rate": 0.00021893560616351607,
+      "loss": 2.9929,
+      "step": 35939
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6211484670639038,
+      "learning_rate": 0.00021892078281650847,
+      "loss": 3.0523,
+      "step": 35940
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6059033274650574,
+      "learning_rate": 0.00021890595968304807,
+      "loss": 2.8255,
+      "step": 35941
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6170439124107361,
+      "learning_rate": 0.00021889113676317365,
+      "loss": 3.0655,
+      "step": 35942
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6702554821968079,
+      "learning_rate": 0.00021887631405692444,
+      "loss": 3.0742,
+      "step": 35943
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6257311105728149,
+      "learning_rate": 0.00021886149156433944,
+      "loss": 2.9236,
+      "step": 35944
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6383312344551086,
+      "learning_rate": 0.0002188466692854575,
+      "loss": 3.0114,
+      "step": 35945
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6520895957946777,
+      "learning_rate": 0.000218831847220318,
+      "loss": 3.0147,
+      "step": 35946
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6120529174804688,
+      "learning_rate": 0.0002188170253689597,
+      "loss": 2.9981,
+      "step": 35947
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6751158237457275,
+      "learning_rate": 0.0002188022037314217,
+      "loss": 3.0975,
+      "step": 35948
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6635515689849854,
+      "learning_rate": 0.00021878738230774315,
+      "loss": 2.8741,
+      "step": 35949
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6444944143295288,
+      "learning_rate": 0.00021877256109796299,
+      "loss": 2.788,
+      "step": 35950
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6126725673675537,
+      "learning_rate": 0.00021875774010212027,
+      "loss": 2.8899,
+      "step": 35951
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6406420469284058,
+      "learning_rate": 0.00021874291932025394,
+      "loss": 3.0294,
+      "step": 35952
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6252112984657288,
+      "learning_rate": 0.00021872809875240318,
+      "loss": 3.0093,
+      "step": 35953
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6004651188850403,
+      "learning_rate": 0.00021871327839860703,
+      "loss": 3.0745,
+      "step": 35954
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6479066014289856,
+      "learning_rate": 0.00021869845825890435,
+      "loss": 3.2259,
+      "step": 35955
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6312503218650818,
+      "learning_rate": 0.00021868363833333439,
+      "loss": 3.0985,
+      "step": 35956
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6492436528205872,
+      "learning_rate": 0.00021866881862193604,
+      "loss": 3.0955,
+      "step": 35957
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6168732643127441,
+      "learning_rate": 0.00021865399912474824,
+      "loss": 2.8809,
+      "step": 35958
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7132405638694763,
+      "learning_rate": 0.00021863917984181035,
+      "loss": 3.2223,
+      "step": 35959
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6588032245635986,
+      "learning_rate": 0.00021862436077316113,
+      "loss": 3.2283,
+      "step": 35960
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8370882272720337,
+      "learning_rate": 0.00021860954191883973,
+      "loss": 2.8941,
+      "step": 35961
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.634719729423523,
+      "learning_rate": 0.00021859472327888496,
+      "loss": 3.1808,
+      "step": 35962
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6163418292999268,
+      "learning_rate": 0.00021857990485333608,
+      "loss": 2.9578,
+      "step": 35963
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6694480180740356,
+      "learning_rate": 0.00021856508664223217,
+      "loss": 2.8498,
+      "step": 35964
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6347652673721313,
+      "learning_rate": 0.0002185502686456121,
+      "loss": 2.8425,
+      "step": 35965
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6317447423934937,
+      "learning_rate": 0.000218535450863515,
+      "loss": 3.0694,
+      "step": 35966
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6160470843315125,
+      "learning_rate": 0.00021852063329597972,
+      "loss": 2.9192,
+      "step": 35967
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5880303978919983,
+      "learning_rate": 0.00021850581594304547,
+      "loss": 2.6224,
+      "step": 35968
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7462405562400818,
+      "learning_rate": 0.00021849099880475126,
+      "loss": 3.0498,
+      "step": 35969
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6336656212806702,
+      "learning_rate": 0.000218476181881136,
+      "loss": 3.2386,
+      "step": 35970
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.638992965221405,
+      "learning_rate": 0.0002184613651722389,
+      "loss": 2.9897,
+      "step": 35971
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6258759498596191,
+      "learning_rate": 0.00021844654867809877,
+      "loss": 3.0165,
+      "step": 35972
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6453997492790222,
+      "learning_rate": 0.00021843173239875468,
+      "loss": 2.7978,
+      "step": 35973
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6440539360046387,
+      "learning_rate": 0.00021841691633424587,
+      "loss": 2.8896,
+      "step": 35974
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6205512285232544,
+      "learning_rate": 0.0002184021004846111,
+      "loss": 3.1415,
+      "step": 35975
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6318713426589966,
+      "learning_rate": 0.00021838728484988956,
+      "loss": 2.9567,
+      "step": 35976
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6605975031852722,
+      "learning_rate": 0.0002183724694301201,
+      "loss": 2.9058,
+      "step": 35977
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6602401733398438,
+      "learning_rate": 0.00021835765422534192,
+      "loss": 2.8578,
+      "step": 35978
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6470891237258911,
+      "learning_rate": 0.000218342839235594,
+      "loss": 2.8326,
+      "step": 35979
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6551663279533386,
+      "learning_rate": 0.00021832802446091523,
+      "loss": 3.1398,
+      "step": 35980
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6770197153091431,
+      "learning_rate": 0.00021831320990134485,
+      "loss": 2.9125,
+      "step": 35981
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6041737794876099,
+      "learning_rate": 0.00021829839555692165,
+      "loss": 3.1245,
+      "step": 35982
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6864397525787354,
+      "learning_rate": 0.0002182835814276847,
+      "loss": 2.9672,
+      "step": 35983
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.628705620765686,
+      "learning_rate": 0.00021826876751367323,
+      "loss": 2.9742,
+      "step": 35984
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6601821184158325,
+      "learning_rate": 0.00021825395381492593,
+      "loss": 2.941,
+      "step": 35985
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.679912269115448,
+      "learning_rate": 0.00021823914033148215,
+      "loss": 3.0776,
+      "step": 35986
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7171043753623962,
+      "learning_rate": 0.00021822432706338065,
+      "loss": 3.0341,
+      "step": 35987
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6408360600471497,
+      "learning_rate": 0.0002182095140106605,
+      "loss": 2.934,
+      "step": 35988
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6534038186073303,
+      "learning_rate": 0.00021819470117336083,
+      "loss": 2.9033,
+      "step": 35989
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6343661546707153,
+      "learning_rate": 0.00021817988855152051,
+      "loss": 3.1754,
+      "step": 35990
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6617305278778076,
+      "learning_rate": 0.00021816507614517874,
+      "loss": 3.0125,
+      "step": 35991
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.636042594909668,
+      "learning_rate": 0.00021815026395437422,
+      "loss": 3.2112,
+      "step": 35992
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6248968243598938,
+      "learning_rate": 0.0002181354519791463,
+      "loss": 2.7923,
+      "step": 35993
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6503058671951294,
+      "learning_rate": 0.0002181206402195338,
+      "loss": 2.963,
+      "step": 35994
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6492918133735657,
+      "learning_rate": 0.0002181058286755757,
+      "loss": 2.9537,
+      "step": 35995
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6048479676246643,
+      "learning_rate": 0.0002180910173473112,
+      "loss": 2.8097,
+      "step": 35996
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6944062113761902,
+      "learning_rate": 0.00021807620623477916,
+      "loss": 2.8929,
+      "step": 35997
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6303960084915161,
+      "learning_rate": 0.00021806139533801867,
+      "loss": 3.0977,
+      "step": 35998
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6644660234451294,
+      "learning_rate": 0.00021804658465706853,
+      "loss": 2.9469,
+      "step": 35999
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5891596078872681,
+      "learning_rate": 0.00021803177419196806,
+      "loss": 3.255,
+      "step": 36000
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6277598142623901,
+      "learning_rate": 0.00021801696394275615,
+      "loss": 3.1176,
+      "step": 36001
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6298516392707825,
+      "learning_rate": 0.00021800215390947164,
+      "loss": 2.9568,
+      "step": 36002
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6497096419334412,
+      "learning_rate": 0.0002179873440921538,
+      "loss": 2.9517,
+      "step": 36003
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6442708969116211,
+      "learning_rate": 0.00021797253449084143,
+      "loss": 3.0019,
+      "step": 36004
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6389861702919006,
+      "learning_rate": 0.00021795772510557363,
+      "loss": 2.8269,
+      "step": 36005
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6288984417915344,
+      "learning_rate": 0.00021794291593638947,
+      "loss": 2.7268,
+      "step": 36006
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6967528462409973,
+      "learning_rate": 0.00021792810698332773,
+      "loss": 2.8301,
+      "step": 36007
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7129323482513428,
+      "learning_rate": 0.00021791329824642774,
+      "loss": 3.0183,
+      "step": 36008
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0993932485580444,
+      "learning_rate": 0.0002178984897257282,
+      "loss": 2.9254,
+      "step": 36009
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6599279642105103,
+      "learning_rate": 0.00021788368142126822,
+      "loss": 2.9239,
+      "step": 36010
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6514578461647034,
+      "learning_rate": 0.00021786887333308694,
+      "loss": 2.9384,
+      "step": 36011
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6423161029815674,
+      "learning_rate": 0.00021785406546122316,
+      "loss": 3.0236,
+      "step": 36012
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6230955719947815,
+      "learning_rate": 0.00021783925780571604,
+      "loss": 3.0301,
+      "step": 36013
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6230641007423401,
+      "learning_rate": 0.00021782445036660437,
+      "loss": 2.9346,
+      "step": 36014
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6333297491073608,
+      "learning_rate": 0.00021780964314392734,
+      "loss": 2.7931,
+      "step": 36015
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6389443874359131,
+      "learning_rate": 0.000217794836137724,
+      "loss": 2.7033,
+      "step": 36016
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6404136419296265,
+      "learning_rate": 0.00021778002934803306,
+      "loss": 2.8099,
+      "step": 36017
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6511685848236084,
+      "learning_rate": 0.00021776522277489386,
+      "loss": 3.1501,
+      "step": 36018
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6114490628242493,
+      "learning_rate": 0.00021775041641834513,
+      "loss": 2.9718,
+      "step": 36019
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6275057792663574,
+      "learning_rate": 0.00021773561027842595,
+      "loss": 2.9344,
+      "step": 36020
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6325558423995972,
+      "learning_rate": 0.00021772080435517545,
+      "loss": 3.1454,
+      "step": 36021
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6032198667526245,
+      "learning_rate": 0.00021770599864863246,
+      "loss": 2.9158,
+      "step": 36022
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5971514582633972,
+      "learning_rate": 0.0002176911931588361,
+      "loss": 2.8928,
+      "step": 36023
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6245431900024414,
+      "learning_rate": 0.00021767638788582517,
+      "loss": 3.1731,
+      "step": 36024
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6920629143714905,
+      "learning_rate": 0.00021766158282963876,
+      "loss": 3.0704,
+      "step": 36025
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6170417666435242,
+      "learning_rate": 0.00021764677799031607,
+      "loss": 2.942,
+      "step": 36026
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6285954713821411,
+      "learning_rate": 0.0002176319733678958,
+      "loss": 2.9836,
+      "step": 36027
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6193482875823975,
+      "learning_rate": 0.00021761716896241716,
+      "loss": 2.8479,
+      "step": 36028
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6073896884918213,
+      "learning_rate": 0.00021760236477391887,
+      "loss": 3.2373,
+      "step": 36029
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6277605295181274,
+      "learning_rate": 0.00021758756080244023,
+      "loss": 3.0198,
+      "step": 36030
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6205739974975586,
+      "learning_rate": 0.0002175727570480201,
+      "loss": 3.0834,
+      "step": 36031
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9096717238426208,
+      "learning_rate": 0.00021755795351069738,
+      "loss": 3.0458,
+      "step": 36032
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6602234840393066,
+      "learning_rate": 0.00021754315019051126,
+      "loss": 3.2214,
+      "step": 36033
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.650767982006073,
+      "learning_rate": 0.00021752834708750054,
+      "loss": 3.0056,
+      "step": 36034
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5995840430259705,
+      "learning_rate": 0.0002175135442017042,
+      "loss": 3.1619,
+      "step": 36035
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6354784369468689,
+      "learning_rate": 0.00021749874153316148,
+      "loss": 3.2268,
+      "step": 36036
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6659834980964661,
+      "learning_rate": 0.0002174839390819111,
+      "loss": 3.1833,
+      "step": 36037
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.649095892906189,
+      "learning_rate": 0.00021746913684799222,
+      "loss": 2.8661,
+      "step": 36038
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7001377940177917,
+      "learning_rate": 0.0002174543348314436,
+      "loss": 2.85,
+      "step": 36039
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6997196674346924,
+      "learning_rate": 0.00021743953303230448,
+      "loss": 3.1971,
+      "step": 36040
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6453948020935059,
+      "learning_rate": 0.00021742473145061378,
+      "loss": 3.01,
+      "step": 36041
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.639747679233551,
+      "learning_rate": 0.0002174099300864103,
+      "loss": 2.9441,
+      "step": 36042
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.619999885559082,
+      "learning_rate": 0.00021739512893973335,
+      "loss": 2.8838,
+      "step": 36043
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6154755353927612,
+      "learning_rate": 0.00021738032801062155,
+      "loss": 3.1061,
+      "step": 36044
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6470183730125427,
+      "learning_rate": 0.00021736552729911417,
+      "loss": 3.0322,
+      "step": 36045
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6488195657730103,
+      "learning_rate": 0.00021735072680525015,
+      "loss": 2.9479,
+      "step": 36046
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6353014707565308,
+      "learning_rate": 0.00021733592652906826,
+      "loss": 3.135,
+      "step": 36047
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6322101354598999,
+      "learning_rate": 0.00021732112647060775,
+      "loss": 3.1161,
+      "step": 36048
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5940348505973816,
+      "learning_rate": 0.00021730632662990743,
+      "loss": 2.9735,
+      "step": 36049
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5832706689834595,
+      "learning_rate": 0.00021729152700700628,
+      "loss": 2.9688,
+      "step": 36050
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6051971316337585,
+      "learning_rate": 0.00021727672760194343,
+      "loss": 2.9654,
+      "step": 36051
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6165816783905029,
+      "learning_rate": 0.0002172619284147577,
+      "loss": 2.8744,
+      "step": 36052
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6479892134666443,
+      "learning_rate": 0.0002172471294454882,
+      "loss": 2.8019,
+      "step": 36053
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6285097002983093,
+      "learning_rate": 0.00021723233069417374,
+      "loss": 2.8948,
+      "step": 36054
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6039531826972961,
+      "learning_rate": 0.0002172175321608534,
+      "loss": 2.8057,
+      "step": 36055
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6159113049507141,
+      "learning_rate": 0.00021720273384556624,
+      "loss": 2.9787,
+      "step": 36056
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.642290472984314,
+      "learning_rate": 0.00021718793574835102,
+      "loss": 3.0758,
+      "step": 36057
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6488417387008667,
+      "learning_rate": 0.00021717313786924695,
+      "loss": 2.9597,
+      "step": 36058
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6642580628395081,
+      "learning_rate": 0.00021715834020829284,
+      "loss": 3.1399,
+      "step": 36059
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.686511218547821,
+      "learning_rate": 0.00021714354276552766,
+      "loss": 2.9219,
+      "step": 36060
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7143248319625854,
+      "learning_rate": 0.00021712874554099057,
+      "loss": 2.8625,
+      "step": 36061
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6069398522377014,
+      "learning_rate": 0.00021711394853472037,
+      "loss": 2.6463,
+      "step": 36062
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5863265991210938,
+      "learning_rate": 0.00021709915174675614,
+      "loss": 3.0262,
+      "step": 36063
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6850087642669678,
+      "learning_rate": 0.00021708435517713663,
+      "loss": 3.182,
+      "step": 36064
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6968920230865479,
+      "learning_rate": 0.00021706955882590096,
+      "loss": 2.9166,
+      "step": 36065
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.662936270236969,
+      "learning_rate": 0.0002170547626930883,
+      "loss": 2.9049,
+      "step": 36066
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6453676223754883,
+      "learning_rate": 0.0002170399667787373,
+      "loss": 3.0578,
+      "step": 36067
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6160590052604675,
+      "learning_rate": 0.00021702517108288715,
+      "loss": 3.0075,
+      "step": 36068
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6253478527069092,
+      "learning_rate": 0.00021701037560557663,
+      "loss": 3.0597,
+      "step": 36069
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6356991529464722,
+      "learning_rate": 0.0002169955803468449,
+      "loss": 2.8705,
+      "step": 36070
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6288435459136963,
+      "learning_rate": 0.0002169807853067308,
+      "loss": 3.2712,
+      "step": 36071
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6195838451385498,
+      "learning_rate": 0.00021696599048527326,
+      "loss": 2.9045,
+      "step": 36072
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6254141330718994,
+      "learning_rate": 0.00021695119588251144,
+      "loss": 3.2025,
+      "step": 36073
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6363393664360046,
+      "learning_rate": 0.00021693640149848416,
+      "loss": 3.0731,
+      "step": 36074
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5991094708442688,
+      "learning_rate": 0.00021692160733323044,
+      "loss": 3.021,
+      "step": 36075
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6361525654792786,
+      "learning_rate": 0.0002169068133867891,
+      "loss": 2.9463,
+      "step": 36076
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.653708279132843,
+      "learning_rate": 0.00021689201965919929,
+      "loss": 2.9209,
+      "step": 36077
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6295130848884583,
+      "learning_rate": 0.00021687722615049994,
+      "loss": 2.9804,
+      "step": 36078
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6194791197776794,
+      "learning_rate": 0.00021686243286072988,
+      "loss": 2.979,
+      "step": 36079
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6836393475532532,
+      "learning_rate": 0.0002168476397899283,
+      "loss": 2.8899,
+      "step": 36080
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5709426403045654,
+      "learning_rate": 0.00021683284693813396,
+      "loss": 3.0018,
+      "step": 36081
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.637849748134613,
+      "learning_rate": 0.00021681805430538587,
+      "loss": 3.072,
+      "step": 36082
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5979206562042236,
+      "learning_rate": 0.00021680326189172307,
+      "loss": 2.9527,
+      "step": 36083
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6736457347869873,
+      "learning_rate": 0.00021678846969718445,
+      "loss": 3.1289,
+      "step": 36084
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6516305208206177,
+      "learning_rate": 0.00021677367772180906,
+      "loss": 3.0119,
+      "step": 36085
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6282474994659424,
+      "learning_rate": 0.0002167588859656356,
+      "loss": 2.9219,
+      "step": 36086
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6169061660766602,
+      "learning_rate": 0.00021674409442870327,
+      "loss": 3.0883,
+      "step": 36087
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7025720477104187,
+      "learning_rate": 0.00021672930311105107,
+      "loss": 3.1129,
+      "step": 36088
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6823790073394775,
+      "learning_rate": 0.0002167145120127178,
+      "loss": 2.9719,
+      "step": 36089
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6688551902770996,
+      "learning_rate": 0.00021669972113374258,
+      "loss": 2.8693,
+      "step": 36090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6762685775756836,
+      "learning_rate": 0.0002166849304741641,
+      "loss": 2.9417,
+      "step": 36091
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6573099493980408,
+      "learning_rate": 0.00021667014003402154,
+      "loss": 2.9705,
+      "step": 36092
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6102715730667114,
+      "learning_rate": 0.00021665534981335388,
+      "loss": 3.0654,
+      "step": 36093
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6486304998397827,
+      "learning_rate": 0.00021664055981219984,
+      "loss": 2.9239,
+      "step": 36094
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5909685492515564,
+      "learning_rate": 0.00021662577003059869,
+      "loss": 3.0793,
+      "step": 36095
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6453856229782104,
+      "learning_rate": 0.00021661098046858908,
+      "loss": 2.9635,
+      "step": 36096
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6692531108856201,
+      "learning_rate": 0.0002165961911262101,
+      "loss": 3.0343,
+      "step": 36097
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6578263640403748,
+      "learning_rate": 0.0002165814020035008,
+      "loss": 3.1648,
+      "step": 36098
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6265367865562439,
+      "learning_rate": 0.0002165666131005,
+      "loss": 3.0647,
+      "step": 36099
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.615085780620575,
+      "learning_rate": 0.00021655182441724673,
+      "loss": 3.1935,
+      "step": 36100
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6076129078865051,
+      "learning_rate": 0.0002165370359537797,
+      "loss": 3.0154,
+      "step": 36101
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6174434423446655,
+      "learning_rate": 0.0002165222477101382,
+      "loss": 3.0799,
+      "step": 36102
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6437174081802368,
+      "learning_rate": 0.00021650745968636108,
+      "loss": 2.9149,
+      "step": 36103
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.614182710647583,
+      "learning_rate": 0.0002164926718824871,
+      "loss": 3.0644,
+      "step": 36104
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6699419617652893,
+      "learning_rate": 0.0002164778842985555,
+      "loss": 2.8719,
+      "step": 36105
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.636222243309021,
+      "learning_rate": 0.00021646309693460495,
+      "loss": 3.0052,
+      "step": 36106
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6297125816345215,
+      "learning_rate": 0.00021644830979067456,
+      "loss": 3.184,
+      "step": 36107
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6243493556976318,
+      "learning_rate": 0.00021643352286680333,
+      "loss": 2.8433,
+      "step": 36108
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6312358975410461,
+      "learning_rate": 0.00021641873616303,
+      "loss": 3.1941,
+      "step": 36109
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6219147443771362,
+      "learning_rate": 0.00021640394967939374,
+      "loss": 3.0337,
+      "step": 36110
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6636743545532227,
+      "learning_rate": 0.0002163891634159333,
+      "loss": 3.2584,
+      "step": 36111
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6117985844612122,
+      "learning_rate": 0.00021637437737268773,
+      "loss": 3.284,
+      "step": 36112
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6436432003974915,
+      "learning_rate": 0.00021635959154969603,
+      "loss": 2.99,
+      "step": 36113
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6268394589424133,
+      "learning_rate": 0.00021634480594699698,
+      "loss": 3.0048,
+      "step": 36114
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6493527889251709,
+      "learning_rate": 0.0002163300205646297,
+      "loss": 3.037,
+      "step": 36115
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6208270192146301,
+      "learning_rate": 0.00021631523540263292,
+      "loss": 2.973,
+      "step": 36116
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6634708642959595,
+      "learning_rate": 0.00021630045046104577,
+      "loss": 2.9894,
+      "step": 36117
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6609952449798584,
+      "learning_rate": 0.00021628566573990717,
+      "loss": 3.1561,
+      "step": 36118
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6642773747444153,
+      "learning_rate": 0.0002162708812392559,
+      "loss": 2.9295,
+      "step": 36119
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6038638353347778,
+      "learning_rate": 0.00021625609695913116,
+      "loss": 2.869,
+      "step": 36120
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6279966831207275,
+      "learning_rate": 0.00021624131289957167,
+      "loss": 2.9704,
+      "step": 36121
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6348757147789001,
+      "learning_rate": 0.00021622652906061636,
+      "loss": 3.0251,
+      "step": 36122
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6414353847503662,
+      "learning_rate": 0.0002162117454423044,
+      "loss": 2.9168,
+      "step": 36123
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6199612617492676,
+      "learning_rate": 0.00021619696204467452,
+      "loss": 2.956,
+      "step": 36124
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6189448833465576,
+      "learning_rate": 0.00021618217886776577,
+      "loss": 3.0051,
+      "step": 36125
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6277577877044678,
+      "learning_rate": 0.0002161673959116169,
+      "loss": 3.0413,
+      "step": 36126
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6269301772117615,
+      "learning_rate": 0.00021615261317626692,
+      "loss": 3.1164,
+      "step": 36127
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6273691058158875,
+      "learning_rate": 0.00021613783066175505,
+      "loss": 2.7998,
+      "step": 36128
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6344377398490906,
+      "learning_rate": 0.00021612304836811984,
+      "loss": 3.1015,
+      "step": 36129
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6333027482032776,
+      "learning_rate": 0.00021610826629540048,
+      "loss": 2.893,
+      "step": 36130
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6379823088645935,
+      "learning_rate": 0.0002160934844436357,
+      "loss": 3.0585,
+      "step": 36131
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6152437925338745,
+      "learning_rate": 0.00021607870281286455,
+      "loss": 3.0948,
+      "step": 36132
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8633241057395935,
+      "learning_rate": 0.00021606392140312606,
+      "loss": 2.9737,
+      "step": 36133
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6464073061943054,
+      "learning_rate": 0.00021604914021445886,
+      "loss": 2.9933,
+      "step": 36134
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6235373616218567,
+      "learning_rate": 0.00021603435924690223,
+      "loss": 3.0425,
+      "step": 36135
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7127751708030701,
+      "learning_rate": 0.00021601957850049487,
+      "loss": 2.9248,
+      "step": 36136
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6603729724884033,
+      "learning_rate": 0.00021600479797527573,
+      "loss": 3.1996,
+      "step": 36137
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6313992738723755,
+      "learning_rate": 0.00021599001767128387,
+      "loss": 3.0105,
+      "step": 36138
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6753054857254028,
+      "learning_rate": 0.00021597523758855813,
+      "loss": 2.942,
+      "step": 36139
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6346279978752136,
+      "learning_rate": 0.0002159604577271375,
+      "loss": 2.9444,
+      "step": 36140
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6592559218406677,
+      "learning_rate": 0.00021594567808706067,
+      "loss": 3.264,
+      "step": 36141
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6272072196006775,
+      "learning_rate": 0.00021593089866836683,
+      "loss": 3.0916,
+      "step": 36142
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6413544416427612,
+      "learning_rate": 0.0002159161194710949,
+      "loss": 3.0556,
+      "step": 36143
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6144356727600098,
+      "learning_rate": 0.00021590134049528364,
+      "loss": 3.0897,
+      "step": 36144
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6145825982093811,
+      "learning_rate": 0.00021588656174097214,
+      "loss": 3.1299,
+      "step": 36145
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6415427327156067,
+      "learning_rate": 0.0002158717832081992,
+      "loss": 3.0902,
+      "step": 36146
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6182569265365601,
+      "learning_rate": 0.00021585700489700383,
+      "loss": 2.9496,
+      "step": 36147
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6118202209472656,
+      "learning_rate": 0.00021584222680742484,
+      "loss": 2.9401,
+      "step": 36148
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.630115807056427,
+      "learning_rate": 0.00021582744893950122,
+      "loss": 2.9302,
+      "step": 36149
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6311905384063721,
+      "learning_rate": 0.000215812671293272,
+      "loss": 3.1874,
+      "step": 36150
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.632701575756073,
+      "learning_rate": 0.0002157978938687759,
+      "loss": 2.9891,
+      "step": 36151
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6322634220123291,
+      "learning_rate": 0.00021578311666605208,
+      "loss": 2.8718,
+      "step": 36152
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6122119426727295,
+      "learning_rate": 0.00021576833968513917,
+      "loss": 3.0455,
+      "step": 36153
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6663072109222412,
+      "learning_rate": 0.0002157535629260763,
+      "loss": 2.8699,
+      "step": 36154
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6152451038360596,
+      "learning_rate": 0.0002157387863889024,
+      "loss": 3.0347,
+      "step": 36155
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.614471435546875,
+      "learning_rate": 0.0002157240100736562,
+      "loss": 3.1776,
+      "step": 36156
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1464407444000244,
+      "learning_rate": 0.00021570923398037684,
+      "loss": 3.0467,
+      "step": 36157
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.626950204372406,
+      "learning_rate": 0.0002156944581091031,
+      "loss": 2.7787,
+      "step": 36158
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6673448085784912,
+      "learning_rate": 0.00021567968245987386,
+      "loss": 3.0078,
+      "step": 36159
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6380613446235657,
+      "learning_rate": 0.00021566490703272824,
+      "loss": 2.8045,
+      "step": 36160
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6106299161911011,
+      "learning_rate": 0.00021565013182770493,
+      "loss": 3.2147,
+      "step": 36161
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6322900652885437,
+      "learning_rate": 0.00021563535684484305,
+      "loss": 3.0431,
+      "step": 36162
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.603399932384491,
+      "learning_rate": 0.00021562058208418123,
+      "loss": 2.9781,
+      "step": 36163
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6214710474014282,
+      "learning_rate": 0.00021560580754575865,
+      "loss": 3.0257,
+      "step": 36164
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6370988488197327,
+      "learning_rate": 0.00021559103322961419,
+      "loss": 3.0746,
+      "step": 36165
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6562209129333496,
+      "learning_rate": 0.00021557625913578654,
+      "loss": 2.9225,
+      "step": 36166
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6525112986564636,
+      "learning_rate": 0.00021556148526431497,
+      "loss": 2.9325,
+      "step": 36167
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6724076867103577,
+      "learning_rate": 0.00021554671161523806,
+      "loss": 2.9713,
+      "step": 36168
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6039425134658813,
+      "learning_rate": 0.00021553193818859488,
+      "loss": 3.0612,
+      "step": 36169
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6516081690788269,
+      "learning_rate": 0.0002155171649844244,
+      "loss": 2.8402,
+      "step": 36170
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6352439522743225,
+      "learning_rate": 0.00021550239200276536,
+      "loss": 3.0824,
+      "step": 36171
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6601967215538025,
+      "learning_rate": 0.00021548761924365683,
+      "loss": 2.9832,
+      "step": 36172
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6137501001358032,
+      "learning_rate": 0.0002154728467071376,
+      "loss": 3.0011,
+      "step": 36173
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6737114787101746,
+      "learning_rate": 0.00021545807439324655,
+      "loss": 2.8073,
+      "step": 36174
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6427306532859802,
+      "learning_rate": 0.0002154433023020228,
+      "loss": 2.913,
+      "step": 36175
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6364057660102844,
+      "learning_rate": 0.00021542853043350507,
+      "loss": 3.0798,
+      "step": 36176
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6504160165786743,
+      "learning_rate": 0.00021541375878773237,
+      "loss": 2.9817,
+      "step": 36177
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.669624924659729,
+      "learning_rate": 0.00021539898736474338,
+      "loss": 2.8296,
+      "step": 36178
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6107624769210815,
+      "learning_rate": 0.0002153842161645773,
+      "loss": 3.046,
+      "step": 36179
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6231289505958557,
+      "learning_rate": 0.00021536944518727296,
+      "loss": 3.1231,
+      "step": 36180
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6516662836074829,
+      "learning_rate": 0.0002153546744328691,
+      "loss": 3.0611,
+      "step": 36181
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6371469497680664,
+      "learning_rate": 0.00021533990390140484,
+      "loss": 3.0647,
+      "step": 36182
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6329872608184814,
+      "learning_rate": 0.00021532513359291886,
+      "loss": 2.9429,
+      "step": 36183
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6683485507965088,
+      "learning_rate": 0.00021531036350745018,
+      "loss": 2.8126,
+      "step": 36184
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6202999949455261,
+      "learning_rate": 0.00021529559364503784,
+      "loss": 3.0846,
+      "step": 36185
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6209010481834412,
+      "learning_rate": 0.00021528082400572055,
+      "loss": 2.9513,
+      "step": 36186
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6557936072349548,
+      "learning_rate": 0.00021526605458953733,
+      "loss": 3.028,
+      "step": 36187
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6131925582885742,
+      "learning_rate": 0.0002152512853965268,
+      "loss": 2.867,
+      "step": 36188
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6493362188339233,
+      "learning_rate": 0.00021523651642672815,
+      "loss": 2.935,
+      "step": 36189
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8060845136642456,
+      "learning_rate": 0.00021522174768018033,
+      "loss": 3.1432,
+      "step": 36190
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5938668251037598,
+      "learning_rate": 0.00021520697915692203,
+      "loss": 2.9079,
+      "step": 36191
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6498206257820129,
+      "learning_rate": 0.0002151922108569923,
+      "loss": 2.9356,
+      "step": 36192
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6662101745605469,
+      "learning_rate": 0.00021517744278042982,
+      "loss": 2.9902,
+      "step": 36193
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.686980128288269,
+      "learning_rate": 0.00021516267492727373,
+      "loss": 2.9666,
+      "step": 36194
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6905176639556885,
+      "learning_rate": 0.00021514790729756285,
+      "loss": 2.9056,
+      "step": 36195
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6709184050559998,
+      "learning_rate": 0.00021513313989133596,
+      "loss": 3.0483,
+      "step": 36196
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6174628734588623,
+      "learning_rate": 0.00021511837270863214,
+      "loss": 3.0596,
+      "step": 36197
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6258738040924072,
+      "learning_rate": 0.0002151036057494902,
+      "loss": 3.0001,
+      "step": 36198
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6265406012535095,
+      "learning_rate": 0.00021508883901394886,
+      "loss": 3.172,
+      "step": 36199
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6480788588523865,
+      "learning_rate": 0.00021507407250204736,
+      "loss": 2.9221,
+      "step": 36200
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6287563443183899,
+      "learning_rate": 0.00021505930621382437,
+      "loss": 3.0638,
+      "step": 36201
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6304842829704285,
+      "learning_rate": 0.00021504454014931884,
+      "loss": 2.9461,
+      "step": 36202
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6325621604919434,
+      "learning_rate": 0.0002150297743085695,
+      "loss": 2.892,
+      "step": 36203
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6825348734855652,
+      "learning_rate": 0.00021501500869161548,
+      "loss": 2.8529,
+      "step": 36204
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6398288607597351,
+      "learning_rate": 0.00021500024329849563,
+      "loss": 2.9031,
+      "step": 36205
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5954134464263916,
+      "learning_rate": 0.00021498547812924865,
+      "loss": 2.8321,
+      "step": 36206
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6618627309799194,
+      "learning_rate": 0.00021497071318391368,
+      "loss": 3.1237,
+      "step": 36207
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6097262501716614,
+      "learning_rate": 0.00021495594846252937,
+      "loss": 2.97,
+      "step": 36208
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6133797764778137,
+      "learning_rate": 0.00021494118396513475,
+      "loss": 3.1634,
+      "step": 36209
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7019491791725159,
+      "learning_rate": 0.0002149264196917688,
+      "loss": 2.9923,
+      "step": 36210
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6512349843978882,
+      "learning_rate": 0.00021491165564247011,
+      "loss": 3.1578,
+      "step": 36211
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6159108281135559,
+      "learning_rate": 0.0002148968918172779,
+      "loss": 2.9716,
+      "step": 36212
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6668187975883484,
+      "learning_rate": 0.00021488212821623085,
+      "loss": 2.9477,
+      "step": 36213
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6413983106613159,
+      "learning_rate": 0.0002148673648393678,
+      "loss": 3.215,
+      "step": 36214
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6101641058921814,
+      "learning_rate": 0.0002148526016867279,
+      "loss": 2.7737,
+      "step": 36215
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.622040867805481,
+      "learning_rate": 0.00021483783875834972,
+      "loss": 2.9313,
+      "step": 36216
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6527264714241028,
+      "learning_rate": 0.0002148230760542724,
+      "loss": 3.2522,
+      "step": 36217
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6784375905990601,
+      "learning_rate": 0.00021480831357453456,
+      "loss": 2.8937,
+      "step": 36218
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6425294876098633,
+      "learning_rate": 0.0002147935513191753,
+      "loss": 2.905,
+      "step": 36219
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6469739675521851,
+      "learning_rate": 0.0002147787892882335,
+      "loss": 2.9366,
+      "step": 36220
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6778955459594727,
+      "learning_rate": 0.00021476402748174783,
+      "loss": 3.0863,
+      "step": 36221
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5806019306182861,
+      "learning_rate": 0.00021474926589975745,
+      "loss": 2.7963,
+      "step": 36222
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6367674469947815,
+      "learning_rate": 0.000214734504542301,
+      "loss": 2.9159,
+      "step": 36223
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6361095905303955,
+      "learning_rate": 0.00021471974340941754,
+      "loss": 3.1667,
+      "step": 36224
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6673516631126404,
+      "learning_rate": 0.0002147049825011457,
+      "loss": 3.1121,
+      "step": 36225
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6467750072479248,
+      "learning_rate": 0.00021469022181752465,
+      "loss": 3.1073,
+      "step": 36226
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6565433740615845,
+      "learning_rate": 0.00021467546135859315,
+      "loss": 2.8489,
+      "step": 36227
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6047545075416565,
+      "learning_rate": 0.00021466070112438997,
+      "loss": 3.0128,
+      "step": 36228
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6131613254547119,
+      "learning_rate": 0.00021464594111495419,
+      "loss": 2.8185,
+      "step": 36229
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6076128482818604,
+      "learning_rate": 0.0002146311813303244,
+      "loss": 3.0758,
+      "step": 36230
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6186553835868835,
+      "learning_rate": 0.0002146164217705398,
+      "loss": 2.9951,
+      "step": 36231
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6514021754264832,
+      "learning_rate": 0.00021460166243563914,
+      "loss": 3.0689,
+      "step": 36232
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6270773410797119,
+      "learning_rate": 0.00021458690332566113,
+      "loss": 3.036,
+      "step": 36233
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6818325519561768,
+      "learning_rate": 0.00021457214444064492,
+      "loss": 2.9342,
+      "step": 36234
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6430490016937256,
+      "learning_rate": 0.00021455738578062916,
+      "loss": 2.9211,
+      "step": 36235
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6284935474395752,
+      "learning_rate": 0.00021454262734565277,
+      "loss": 3.1729,
+      "step": 36236
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.646930456161499,
+      "learning_rate": 0.00021452786913575474,
+      "loss": 2.983,
+      "step": 36237
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6334282159805298,
+      "learning_rate": 0.00021451311115097378,
+      "loss": 3.1183,
+      "step": 36238
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6594672203063965,
+      "learning_rate": 0.00021449835339134893,
+      "loss": 3.0003,
+      "step": 36239
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6455363035202026,
+      "learning_rate": 0.00021448359585691883,
+      "loss": 3.0601,
+      "step": 36240
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6710865497589111,
+      "learning_rate": 0.00021446883854772255,
+      "loss": 2.9493,
+      "step": 36241
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6526823043823242,
+      "learning_rate": 0.00021445408146379894,
+      "loss": 3.1155,
+      "step": 36242
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6468623876571655,
+      "learning_rate": 0.0002144393246051867,
+      "loss": 2.7825,
+      "step": 36243
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6826371550559998,
+      "learning_rate": 0.00021442456797192492,
+      "loss": 3.0516,
+      "step": 36244
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6300487518310547,
+      "learning_rate": 0.0002144098115640523,
+      "loss": 3.0863,
+      "step": 36245
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6709060072898865,
+      "learning_rate": 0.00021439505538160776,
+      "loss": 3.0067,
+      "step": 36246
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6281489729881287,
+      "learning_rate": 0.0002143802994246302,
+      "loss": 3.1286,
+      "step": 36247
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6108605861663818,
+      "learning_rate": 0.00021436554369315845,
+      "loss": 3.0594,
+      "step": 36248
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.692886233329773,
+      "learning_rate": 0.00021435078818723144,
+      "loss": 2.9815,
+      "step": 36249
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6155291199684143,
+      "learning_rate": 0.0002143360329068878,
+      "loss": 2.9099,
+      "step": 36250
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5946487784385681,
+      "learning_rate": 0.0002143212778521666,
+      "loss": 3.0502,
+      "step": 36251
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6302567720413208,
+      "learning_rate": 0.0002143065230231068,
+      "loss": 3.019,
+      "step": 36252
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.610625684261322,
+      "learning_rate": 0.00021429176841974702,
+      "loss": 2.9311,
+      "step": 36253
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6491797566413879,
+      "learning_rate": 0.00021427701404212634,
+      "loss": 2.978,
+      "step": 36254
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6238184571266174,
+      "learning_rate": 0.00021426225989028334,
+      "loss": 3.1842,
+      "step": 36255
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6471647024154663,
+      "learning_rate": 0.00021424750596425713,
+      "loss": 3.1372,
+      "step": 36256
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6920592188835144,
+      "learning_rate": 0.00021423275226408654,
+      "loss": 3.1878,
+      "step": 36257
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6681246757507324,
+      "learning_rate": 0.00021421799878981027,
+      "loss": 2.9845,
+      "step": 36258
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6273609399795532,
+      "learning_rate": 0.0002142032455414674,
+      "loss": 3.0139,
+      "step": 36259
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6397222876548767,
+      "learning_rate": 0.0002141884925190966,
+      "loss": 2.9536,
+      "step": 36260
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6393402814865112,
+      "learning_rate": 0.00021417373972273672,
+      "loss": 2.983,
+      "step": 36261
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.619844913482666,
+      "learning_rate": 0.00021415898715242682,
+      "loss": 3.053,
+      "step": 36262
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6226319074630737,
+      "learning_rate": 0.00021414423480820558,
+      "loss": 3.3036,
+      "step": 36263
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7460436820983887,
+      "learning_rate": 0.000214129482690112,
+      "loss": 2.7747,
+      "step": 36264
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6607783436775208,
+      "learning_rate": 0.00021411473079818464,
+      "loss": 3.0174,
+      "step": 36265
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6538329124450684,
+      "learning_rate": 0.00021409997913246262,
+      "loss": 3.0033,
+      "step": 36266
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.704361081123352,
+      "learning_rate": 0.0002140852276929848,
+      "loss": 3.0254,
+      "step": 36267
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7130153775215149,
+      "learning_rate": 0.00021407047647978988,
+      "loss": 3.0591,
+      "step": 36268
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.598651111125946,
+      "learning_rate": 0.00021405572549291685,
+      "loss": 3.1185,
+      "step": 36269
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6102203726768494,
+      "learning_rate": 0.0002140409747324044,
+      "loss": 3.2775,
+      "step": 36270
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6122444868087769,
+      "learning_rate": 0.00021402622419829155,
+      "loss": 3.1641,
+      "step": 36271
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6745235919952393,
+      "learning_rate": 0.00021401147389061716,
+      "loss": 2.8793,
+      "step": 36272
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6349541544914246,
+      "learning_rate": 0.0002139967238094198,
+      "loss": 3.0279,
+      "step": 36273
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.631254255771637,
+      "learning_rate": 0.0002139819739547387,
+      "loss": 2.9795,
+      "step": 36274
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6068877577781677,
+      "learning_rate": 0.00021396722432661245,
+      "loss": 2.9905,
+      "step": 36275
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.677623987197876,
+      "learning_rate": 0.0002139524749250799,
+      "loss": 2.8977,
+      "step": 36276
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6311962604522705,
+      "learning_rate": 0.00021393772575018012,
+      "loss": 3.075,
+      "step": 36277
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6224701404571533,
+      "learning_rate": 0.00021392297680195173,
+      "loss": 3.0223,
+      "step": 36278
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6306285858154297,
+      "learning_rate": 0.00021390822808043373,
+      "loss": 3.163,
+      "step": 36279
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6172642707824707,
+      "learning_rate": 0.00021389347958566474,
+      "loss": 3.0831,
+      "step": 36280
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6490658521652222,
+      "learning_rate": 0.0002138787313176838,
+      "loss": 2.973,
+      "step": 36281
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6317246556282043,
+      "learning_rate": 0.00021386398327652981,
+      "loss": 3.3257,
+      "step": 36282
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6703366041183472,
+      "learning_rate": 0.00021384923546224137,
+      "loss": 3.067,
+      "step": 36283
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6874723434448242,
+      "learning_rate": 0.0002138344878748576,
+      "loss": 2.9391,
+      "step": 36284
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6609402894973755,
+      "learning_rate": 0.0002138197405144171,
+      "loss": 2.9994,
+      "step": 36285
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6028761267662048,
+      "learning_rate": 0.0002138049933809588,
+      "loss": 3.0421,
+      "step": 36286
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6413368582725525,
+      "learning_rate": 0.00021379024647452167,
+      "loss": 3.0925,
+      "step": 36287
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.660110354423523,
+      "learning_rate": 0.00021377549979514438,
+      "loss": 2.8916,
+      "step": 36288
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6747713088989258,
+      "learning_rate": 0.0002137607533428659,
+      "loss": 2.8905,
+      "step": 36289
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.600744366645813,
+      "learning_rate": 0.00021374600711772488,
+      "loss": 3.1411,
+      "step": 36290
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6018193960189819,
+      "learning_rate": 0.00021373126111976027,
+      "loss": 3.0956,
+      "step": 36291
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.611671507358551,
+      "learning_rate": 0.00021371651534901102,
+      "loss": 3.064,
+      "step": 36292
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6379413604736328,
+      "learning_rate": 0.00021370176980551583,
+      "loss": 2.9962,
+      "step": 36293
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.644549548625946,
+      "learning_rate": 0.00021368702448931363,
+      "loss": 2.9289,
+      "step": 36294
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6469461917877197,
+      "learning_rate": 0.00021367227940044303,
+      "loss": 3.3018,
+      "step": 36295
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6457450985908508,
+      "learning_rate": 0.00021365753453894313,
+      "loss": 3.0339,
+      "step": 36296
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6415975689888,
+      "learning_rate": 0.00021364278990485273,
+      "loss": 3.1975,
+      "step": 36297
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6394268870353699,
+      "learning_rate": 0.0002136280454982105,
+      "loss": 3.05,
+      "step": 36298
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6522897481918335,
+      "learning_rate": 0.00021361330131905543,
+      "loss": 3.0616,
+      "step": 36299
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6101564764976501,
+      "learning_rate": 0.0002135985573674263,
+      "loss": 3.218,
+      "step": 36300
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6102758049964905,
+      "learning_rate": 0.000213583813643362,
+      "loss": 2.7773,
+      "step": 36301
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5993099212646484,
+      "learning_rate": 0.00021356907014690113,
+      "loss": 2.6897,
+      "step": 36302
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.629689633846283,
+      "learning_rate": 0.0002135543268780828,
+      "loss": 3.0496,
+      "step": 36303
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5861761569976807,
+      "learning_rate": 0.00021353958383694575,
+      "loss": 3.0468,
+      "step": 36304
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6414809226989746,
+      "learning_rate": 0.00021352484102352872,
+      "loss": 2.7947,
+      "step": 36305
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6364729404449463,
+      "learning_rate": 0.00021351009843787074,
+      "loss": 3.0792,
+      "step": 36306
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6140216588973999,
+      "learning_rate": 0.00021349535608001044,
+      "loss": 2.8048,
+      "step": 36307
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6230908036231995,
+      "learning_rate": 0.00021348061394998663,
+      "loss": 2.9799,
+      "step": 36308
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6611496210098267,
+      "learning_rate": 0.00021346587204783843,
+      "loss": 2.7688,
+      "step": 36309
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.638316810131073,
+      "learning_rate": 0.00021345113037360438,
+      "loss": 3.2603,
+      "step": 36310
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6499404311180115,
+      "learning_rate": 0.00021343638892732343,
+      "loss": 2.8078,
+      "step": 36311
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6417861580848694,
+      "learning_rate": 0.00021342164770903426,
+      "loss": 2.8695,
+      "step": 36312
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6680542230606079,
+      "learning_rate": 0.0002134069067187758,
+      "loss": 3.202,
+      "step": 36313
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6242054104804993,
+      "learning_rate": 0.000213392165956587,
+      "loss": 2.8746,
+      "step": 36314
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6265760064125061,
+      "learning_rate": 0.00021337742542250652,
+      "loss": 3.0401,
+      "step": 36315
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6692229509353638,
+      "learning_rate": 0.00021336268511657332,
+      "loss": 2.6561,
+      "step": 36316
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7591632008552551,
+      "learning_rate": 0.00021334794503882596,
+      "loss": 3.0317,
+      "step": 36317
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6061434745788574,
+      "learning_rate": 0.00021333320518930355,
+      "loss": 3.0425,
+      "step": 36318
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6159639954566956,
+      "learning_rate": 0.00021331846556804483,
+      "loss": 2.9656,
+      "step": 36319
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5951846241950989,
+      "learning_rate": 0.00021330372617508846,
+      "loss": 2.867,
+      "step": 36320
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5990233421325684,
+      "learning_rate": 0.00021328898701047352,
+      "loss": 2.9664,
+      "step": 36321
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6403229236602783,
+      "learning_rate": 0.00021327424807423865,
+      "loss": 3.045,
+      "step": 36322
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6548666954040527,
+      "learning_rate": 0.00021325950936642264,
+      "loss": 3.1726,
+      "step": 36323
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6505340933799744,
+      "learning_rate": 0.00021324477088706452,
+      "loss": 3.0317,
+      "step": 36324
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6289445757865906,
+      "learning_rate": 0.0002132300326362029,
+      "loss": 3.0337,
+      "step": 36325
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6409799456596375,
+      "learning_rate": 0.00021321529461387675,
+      "loss": 3.0774,
+      "step": 36326
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6213870048522949,
+      "learning_rate": 0.00021320055682012468,
+      "loss": 3.1011,
+      "step": 36327
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5979524254798889,
+      "learning_rate": 0.00021318581925498572,
+      "loss": 2.9269,
+      "step": 36328
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6520103812217712,
+      "learning_rate": 0.0002131710819184987,
+      "loss": 2.9932,
+      "step": 36329
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6027911305427551,
+      "learning_rate": 0.00021315634481070214,
+      "loss": 3.1689,
+      "step": 36330
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6382830142974854,
+      "learning_rate": 0.0002131416079316352,
+      "loss": 3.0135,
+      "step": 36331
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6174009442329407,
+      "learning_rate": 0.00021312687128133645,
+      "loss": 2.9106,
+      "step": 36332
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.640430212020874,
+      "learning_rate": 0.00021311213485984486,
+      "loss": 3.1547,
+      "step": 36333
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6170399785041809,
+      "learning_rate": 0.00021309739866719922,
+      "loss": 3.18,
+      "step": 36334
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6654266715049744,
+      "learning_rate": 0.0002130826627034382,
+      "loss": 3.0434,
+      "step": 36335
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6619197130203247,
+      "learning_rate": 0.00021306792696860086,
+      "loss": 2.9076,
+      "step": 36336
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6682602167129517,
+      "learning_rate": 0.0002130531914627258,
+      "loss": 2.9377,
+      "step": 36337
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8006364107131958,
+      "learning_rate": 0.00021303845618585182,
+      "loss": 3.014,
+      "step": 36338
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6219844818115234,
+      "learning_rate": 0.00021302372113801796,
+      "loss": 2.8757,
+      "step": 36339
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6059569716453552,
+      "learning_rate": 0.00021300898631926275,
+      "loss": 2.9424,
+      "step": 36340
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6212844848632812,
+      "learning_rate": 0.00021299425172962528,
+      "loss": 2.9297,
+      "step": 36341
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6634481549263,
+      "learning_rate": 0.00021297951736914402,
+      "loss": 3.2482,
+      "step": 36342
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6362781524658203,
+      "learning_rate": 0.00021296478323785804,
+      "loss": 3.0874,
+      "step": 36343
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.721335768699646,
+      "learning_rate": 0.00021295004933580614,
+      "loss": 2.8786,
+      "step": 36344
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6178750395774841,
+      "learning_rate": 0.00021293531566302692,
+      "loss": 2.8094,
+      "step": 36345
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6878019571304321,
+      "learning_rate": 0.00021292058221955947,
+      "loss": 2.9951,
+      "step": 36346
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6131126284599304,
+      "learning_rate": 0.0002129058490054424,
+      "loss": 2.993,
+      "step": 36347
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6379331946372986,
+      "learning_rate": 0.00021289111602071448,
+      "loss": 2.8931,
+      "step": 36348
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6440445780754089,
+      "learning_rate": 0.00021287638326541468,
+      "loss": 3.142,
+      "step": 36349
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6755915880203247,
+      "learning_rate": 0.00021286165073958175,
+      "loss": 3.0775,
+      "step": 36350
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6246413588523865,
+      "learning_rate": 0.0002128469184432545,
+      "loss": 2.8214,
+      "step": 36351
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6133143305778503,
+      "learning_rate": 0.0002128321863764715,
+      "loss": 2.7959,
+      "step": 36352
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6247037649154663,
+      "learning_rate": 0.0002128174545392718,
+      "loss": 2.9561,
+      "step": 36353
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6576348543167114,
+      "learning_rate": 0.00021280272293169424,
+      "loss": 3.043,
+      "step": 36354
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.633882462978363,
+      "learning_rate": 0.00021278799155377746,
+      "loss": 3.1632,
+      "step": 36355
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6401931643486023,
+      "learning_rate": 0.0002127732604055604,
+      "loss": 3.1631,
+      "step": 36356
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6456989645957947,
+      "learning_rate": 0.0002127585294870817,
+      "loss": 2.7861,
+      "step": 36357
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5906917452812195,
+      "learning_rate": 0.00021274379879838027,
+      "loss": 3.2061,
+      "step": 36358
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6478468179702759,
+      "learning_rate": 0.0002127290683394949,
+      "loss": 3.1233,
+      "step": 36359
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.643356204032898,
+      "learning_rate": 0.00021271433811046432,
+      "loss": 2.7415,
+      "step": 36360
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.651016116142273,
+      "learning_rate": 0.00021269960811132747,
+      "loss": 2.8069,
+      "step": 36361
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6640231609344482,
+      "learning_rate": 0.000212684878342123,
+      "loss": 2.8503,
+      "step": 36362
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6326684951782227,
+      "learning_rate": 0.0002126701488028897,
+      "loss": 2.9992,
+      "step": 36363
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6268022656440735,
+      "learning_rate": 0.00021265541949366652,
+      "loss": 3.0261,
+      "step": 36364
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6220657229423523,
+      "learning_rate": 0.00021264069041449207,
+      "loss": 2.7213,
+      "step": 36365
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6234197616577148,
+      "learning_rate": 0.00021262596156540535,
+      "loss": 3.0385,
+      "step": 36366
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6607543230056763,
+      "learning_rate": 0.00021261123294644485,
+      "loss": 3.0352,
+      "step": 36367
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6178925037384033,
+      "learning_rate": 0.00021259650455764966,
+      "loss": 3.0404,
+      "step": 36368
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6788184642791748,
+      "learning_rate": 0.00021258177639905848,
+      "loss": 2.9027,
+      "step": 36369
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6631069779396057,
+      "learning_rate": 0.00021256704847071,
+      "loss": 3.1795,
+      "step": 36370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6293043494224548,
+      "learning_rate": 0.00021255232077264314,
+      "loss": 2.953,
+      "step": 36371
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6410183906555176,
+      "learning_rate": 0.00021253759330489656,
+      "loss": 3.0083,
+      "step": 36372
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6360601782798767,
+      "learning_rate": 0.00021252286606750914,
+      "loss": 3.0385,
+      "step": 36373
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6407144069671631,
+      "learning_rate": 0.00021250813906051978,
+      "loss": 3.0849,
+      "step": 36374
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5937209129333496,
+      "learning_rate": 0.00021249341228396702,
+      "loss": 2.8081,
+      "step": 36375
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.612413763999939,
+      "learning_rate": 0.00021247868573788986,
+      "loss": 3.133,
+      "step": 36376
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6570436358451843,
+      "learning_rate": 0.00021246395942232693,
+      "loss": 3.0547,
+      "step": 36377
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6165720820426941,
+      "learning_rate": 0.00021244923333731702,
+      "loss": 3.0068,
+      "step": 36378
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6688787341117859,
+      "learning_rate": 0.00021243450748289916,
+      "loss": 2.9723,
+      "step": 36379
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6865296363830566,
+      "learning_rate": 0.00021241978185911184,
+      "loss": 2.8357,
+      "step": 36380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6525271534919739,
+      "learning_rate": 0.00021240505646599405,
+      "loss": 3.1706,
+      "step": 36381
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6239628195762634,
+      "learning_rate": 0.00021239033130358433,
+      "loss": 3.0773,
+      "step": 36382
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6260185241699219,
+      "learning_rate": 0.00021237560637192178,
+      "loss": 2.9808,
+      "step": 36383
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6808270812034607,
+      "learning_rate": 0.00021236088167104494,
+      "loss": 2.7452,
+      "step": 36384
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6365841031074524,
+      "learning_rate": 0.00021234615720099257,
+      "loss": 3.1358,
+      "step": 36385
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6161235570907593,
+      "learning_rate": 0.00021233143296180374,
+      "loss": 3.1963,
+      "step": 36386
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6361461281776428,
+      "learning_rate": 0.00021231670895351695,
+      "loss": 3.1814,
+      "step": 36387
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6474431157112122,
+      "learning_rate": 0.00021230198517617114,
+      "loss": 2.9853,
+      "step": 36388
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.679755687713623,
+      "learning_rate": 0.0002122872616298049,
+      "loss": 3.062,
+      "step": 36389
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6086769700050354,
+      "learning_rate": 0.00021227253831445717,
+      "loss": 3.1973,
+      "step": 36390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6710407733917236,
+      "learning_rate": 0.0002122578152301668,
+      "loss": 2.8719,
+      "step": 36391
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.664715051651001,
+      "learning_rate": 0.0002122430923769723,
+      "loss": 2.9214,
+      "step": 36392
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.625869631767273,
+      "learning_rate": 0.00021222836975491274,
+      "loss": 3.0165,
+      "step": 36393
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6978575587272644,
+      "learning_rate": 0.00021221364736402664,
+      "loss": 3.0272,
+      "step": 36394
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6914448142051697,
+      "learning_rate": 0.00021219892520435294,
+      "loss": 2.8926,
+      "step": 36395
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6650540828704834,
+      "learning_rate": 0.00021218420327593043,
+      "loss": 3.0207,
+      "step": 36396
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6832113862037659,
+      "learning_rate": 0.00021216948157879776,
+      "loss": 2.9073,
+      "step": 36397
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6525067687034607,
+      "learning_rate": 0.00021215476011299383,
+      "loss": 2.9059,
+      "step": 36398
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6345611214637756,
+      "learning_rate": 0.00021214003887855737,
+      "loss": 2.8414,
+      "step": 36399
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6637008786201477,
+      "learning_rate": 0.000212125317875527,
+      "loss": 2.9859,
+      "step": 36400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6102834343910217,
+      "learning_rate": 0.00021211059710394178,
+      "loss": 3.1545,
+      "step": 36401
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6296905875205994,
+      "learning_rate": 0.0002120958765638403,
+      "loss": 2.6042,
+      "step": 36402
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6705476641654968,
+      "learning_rate": 0.00021208115625526142,
+      "loss": 2.9978,
+      "step": 36403
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6274638772010803,
+      "learning_rate": 0.0002120664361782437,
+      "loss": 3.0183,
+      "step": 36404
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6929411888122559,
+      "learning_rate": 0.00021205171633282614,
+      "loss": 3.3224,
+      "step": 36405
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6245270371437073,
+      "learning_rate": 0.0002120369967190475,
+      "loss": 3.1516,
+      "step": 36406
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6495602130889893,
+      "learning_rate": 0.00021202227733694635,
+      "loss": 2.9227,
+      "step": 36407
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6302369832992554,
+      "learning_rate": 0.0002120075581865617,
+      "loss": 3.09,
+      "step": 36408
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6170496344566345,
+      "learning_rate": 0.00021199283926793218,
+      "loss": 2.8749,
+      "step": 36409
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6352107524871826,
+      "learning_rate": 0.00021197812058109652,
+      "loss": 2.9232,
+      "step": 36410
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6431768536567688,
+      "learning_rate": 0.00021196340212609362,
+      "loss": 3.116,
+      "step": 36411
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6413148045539856,
+      "learning_rate": 0.00021194868390296217,
+      "loss": 3.0991,
+      "step": 36412
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.607494592666626,
+      "learning_rate": 0.00021193396591174098,
+      "loss": 3.0078,
+      "step": 36413
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6016876101493835,
+      "learning_rate": 0.00021191924815246865,
+      "loss": 2.9614,
+      "step": 36414
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6435602903366089,
+      "learning_rate": 0.00021190453062518407,
+      "loss": 2.8346,
+      "step": 36415
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6488229036331177,
+      "learning_rate": 0.00021188981332992613,
+      "loss": 2.989,
+      "step": 36416
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.623367428779602,
+      "learning_rate": 0.00021187509626673342,
+      "loss": 3.1408,
+      "step": 36417
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6525915265083313,
+      "learning_rate": 0.0002118603794356448,
+      "loss": 3.0817,
+      "step": 36418
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6700499653816223,
+      "learning_rate": 0.00021184566283669883,
+      "loss": 3.0911,
+      "step": 36419
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6590023636817932,
+      "learning_rate": 0.0002118309464699345,
+      "loss": 2.8538,
+      "step": 36420
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6719303131103516,
+      "learning_rate": 0.00021181623033539052,
+      "loss": 3.0935,
+      "step": 36421
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6459015607833862,
+      "learning_rate": 0.00021180151443310556,
+      "loss": 2.809,
+      "step": 36422
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6407613158226013,
+      "learning_rate": 0.0002117867987631185,
+      "loss": 2.7572,
+      "step": 36423
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6750385165214539,
+      "learning_rate": 0.000211772083325468,
+      "loss": 2.9058,
+      "step": 36424
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6313875317573547,
+      "learning_rate": 0.00021175736812019275,
+      "loss": 3.0272,
+      "step": 36425
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6446053385734558,
+      "learning_rate": 0.0002117426531473318,
+      "loss": 2.9416,
+      "step": 36426
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6613894104957581,
+      "learning_rate": 0.00021172793840692358,
+      "loss": 2.9213,
+      "step": 36427
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6796953082084656,
+      "learning_rate": 0.00021171322389900715,
+      "loss": 2.9999,
+      "step": 36428
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6764907836914062,
+      "learning_rate": 0.00021169850962362088,
+      "loss": 2.9982,
+      "step": 36429
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6619737148284912,
+      "learning_rate": 0.00021168379558080385,
+      "loss": 2.8365,
+      "step": 36430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6781612038612366,
+      "learning_rate": 0.00021166908177059474,
+      "loss": 2.7274,
+      "step": 36431
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6520367860794067,
+      "learning_rate": 0.00021165436819303214,
+      "loss": 3.0871,
+      "step": 36432
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6807015538215637,
+      "learning_rate": 0.00021163965484815507,
+      "loss": 3.1435,
+      "step": 36433
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.688068687915802,
+      "learning_rate": 0.00021162494173600202,
+      "loss": 3.1147,
+      "step": 36434
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6110426187515259,
+      "learning_rate": 0.00021161022885661195,
+      "loss": 2.9672,
+      "step": 36435
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6746833920478821,
+      "learning_rate": 0.00021159551621002356,
+      "loss": 2.9923,
+      "step": 36436
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6675704717636108,
+      "learning_rate": 0.00021158080379627547,
+      "loss": 3.146,
+      "step": 36437
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6473252773284912,
+      "learning_rate": 0.00021156609161540664,
+      "loss": 2.9734,
+      "step": 36438
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6055066585540771,
+      "learning_rate": 0.0002115513796674556,
+      "loss": 2.9343,
+      "step": 36439
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6825117468833923,
+      "learning_rate": 0.00021153666795246118,
+      "loss": 2.9982,
+      "step": 36440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7194039821624756,
+      "learning_rate": 0.00021152195647046226,
+      "loss": 3.1512,
+      "step": 36441
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6608765125274658,
+      "learning_rate": 0.00021150724522149742,
+      "loss": 2.9463,
+      "step": 36442
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6718124747276306,
+      "learning_rate": 0.00021149253420560556,
+      "loss": 2.7968,
+      "step": 36443
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6314850449562073,
+      "learning_rate": 0.00021147782342282519,
+      "loss": 2.8741,
+      "step": 36444
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6046271324157715,
+      "learning_rate": 0.00021146311287319522,
+      "loss": 3.0566,
+      "step": 36445
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6407644152641296,
+      "learning_rate": 0.00021144840255675446,
+      "loss": 2.9809,
+      "step": 36446
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.632928192615509,
+      "learning_rate": 0.00021143369247354147,
+      "loss": 3.0315,
+      "step": 36447
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6315308213233948,
+      "learning_rate": 0.00021141898262359515,
+      "loss": 3.1262,
+      "step": 36448
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6301126480102539,
+      "learning_rate": 0.00021140427300695417,
+      "loss": 2.8378,
+      "step": 36449
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6344707012176514,
+      "learning_rate": 0.00021138956362365722,
+      "loss": 3.2058,
+      "step": 36450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6632879972457886,
+      "learning_rate": 0.00021137485447374324,
+      "loss": 3.1878,
+      "step": 36451
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5994957089424133,
+      "learning_rate": 0.00021136014555725075,
+      "loss": 3.1941,
+      "step": 36452
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5996524691581726,
+      "learning_rate": 0.00021134543687421865,
+      "loss": 3.1972,
+      "step": 36453
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6582194566726685,
+      "learning_rate": 0.00021133072842468545,
+      "loss": 2.8146,
+      "step": 36454
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6702811121940613,
+      "learning_rate": 0.00021131602020869013,
+      "loss": 3.0313,
+      "step": 36455
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6455747485160828,
+      "learning_rate": 0.0002113013122262714,
+      "loss": 3.1189,
+      "step": 36456
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6200703382492065,
+      "learning_rate": 0.00021128660447746787,
+      "loss": 2.7576,
+      "step": 36457
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6644160747528076,
+      "learning_rate": 0.00021127189696231846,
+      "loss": 3.0674,
+      "step": 36458
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6624540090560913,
+      "learning_rate": 0.00021125718968086163,
+      "loss": 2.9791,
+      "step": 36459
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6029409766197205,
+      "learning_rate": 0.00021124248263313644,
+      "loss": 3.0949,
+      "step": 36460
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6144557595252991,
+      "learning_rate": 0.0002112277758191814,
+      "loss": 3.1555,
+      "step": 36461
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6617509722709656,
+      "learning_rate": 0.00021121306923903524,
+      "loss": 2.9906,
+      "step": 36462
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6040384769439697,
+      "learning_rate": 0.0002111983628927369,
+      "loss": 2.941,
+      "step": 36463
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6424322724342346,
+      "learning_rate": 0.0002111836567803249,
+      "loss": 3.1002,
+      "step": 36464
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6235278844833374,
+      "learning_rate": 0.00021116895090183817,
+      "loss": 3.0474,
+      "step": 36465
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6009097099304199,
+      "learning_rate": 0.00021115424525731515,
+      "loss": 3.0715,
+      "step": 36466
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6081644892692566,
+      "learning_rate": 0.00021113953984679484,
+      "loss": 3.1374,
+      "step": 36467
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6132400631904602,
+      "learning_rate": 0.00021112483467031594,
+      "loss": 2.8668,
+      "step": 36468
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6370624303817749,
+      "learning_rate": 0.000211110129727917,
+      "loss": 2.8138,
+      "step": 36469
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6113362908363342,
+      "learning_rate": 0.00021109542501963695,
+      "loss": 3.1049,
+      "step": 36470
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6134870052337646,
+      "learning_rate": 0.00021108072054551444,
+      "loss": 3.051,
+      "step": 36471
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6292585730552673,
+      "learning_rate": 0.0002110660163055881,
+      "loss": 2.9592,
+      "step": 36472
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6173241138458252,
+      "learning_rate": 0.0002110513122998969,
+      "loss": 3.2329,
+      "step": 36473
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6382222771644592,
+      "learning_rate": 0.00021103660852847937,
+      "loss": 2.9698,
+      "step": 36474
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6229817867279053,
+      "learning_rate": 0.00021102190499137433,
+      "loss": 2.9552,
+      "step": 36475
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6382546424865723,
+      "learning_rate": 0.00021100720168862038,
+      "loss": 3.0138,
+      "step": 36476
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.63431316614151,
+      "learning_rate": 0.0002109924986202563,
+      "loss": 2.7953,
+      "step": 36477
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.629733681678772,
+      "learning_rate": 0.00021097779578632095,
+      "loss": 3.2056,
+      "step": 36478
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6950914859771729,
+      "learning_rate": 0.00021096309318685292,
+      "loss": 3.1653,
+      "step": 36479
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6350083351135254,
+      "learning_rate": 0.00021094839082189104,
+      "loss": 2.9533,
+      "step": 36480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6433498859405518,
+      "learning_rate": 0.00021093368869147384,
+      "loss": 2.9834,
+      "step": 36481
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6216697692871094,
+      "learning_rate": 0.0002109189867956402,
+      "loss": 3.0459,
+      "step": 36482
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6351829767227173,
+      "learning_rate": 0.00021090428513442888,
+      "loss": 3.0169,
+      "step": 36483
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6634454727172852,
+      "learning_rate": 0.00021088958370787838,
+      "loss": 2.9469,
+      "step": 36484
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6411239504814148,
+      "learning_rate": 0.00021087488251602776,
+      "loss": 3.1319,
+      "step": 36485
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6567073464393616,
+      "learning_rate": 0.00021086018155891541,
+      "loss": 2.8596,
+      "step": 36486
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6086127758026123,
+      "learning_rate": 0.00021084548083658017,
+      "loss": 3.0581,
+      "step": 36487
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6434993743896484,
+      "learning_rate": 0.00021083078034906085,
+      "loss": 3.1005,
+      "step": 36488
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.685982882976532,
+      "learning_rate": 0.0002108160800963961,
+      "loss": 3.0861,
+      "step": 36489
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6597147583961487,
+      "learning_rate": 0.00021080138007862465,
+      "loss": 3.0276,
+      "step": 36490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6205618977546692,
+      "learning_rate": 0.00021078668029578506,
+      "loss": 3.1633,
+      "step": 36491
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6635233759880066,
+      "learning_rate": 0.00021077198074791626,
+      "loss": 2.885,
+      "step": 36492
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6489311456680298,
+      "learning_rate": 0.00021075728143505698,
+      "loss": 2.8726,
+      "step": 36493
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6681040525436401,
+      "learning_rate": 0.0002107425823572457,
+      "loss": 2.8963,
+      "step": 36494
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6183825731277466,
+      "learning_rate": 0.0002107278835145214,
+      "loss": 3.2596,
+      "step": 36495
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9229050278663635,
+      "learning_rate": 0.00021071318490692255,
+      "loss": 3.1778,
+      "step": 36496
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6755209565162659,
+      "learning_rate": 0.00021069848653448804,
+      "loss": 2.8989,
+      "step": 36497
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6122171878814697,
+      "learning_rate": 0.00021068378839725663,
+      "loss": 3.1738,
+      "step": 36498
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6368758082389832,
+      "learning_rate": 0.00021066909049526677,
+      "loss": 2.9932,
+      "step": 36499
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6233503222465515,
+      "learning_rate": 0.00021065439282855748,
+      "loss": 3.0565,
+      "step": 36500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6276006698608398,
+      "learning_rate": 0.00021063969539716727,
+      "loss": 2.7239,
+      "step": 36501
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6331776976585388,
+      "learning_rate": 0.0002106249982011348,
+      "loss": 2.9001,
+      "step": 36502
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6143416166305542,
+      "learning_rate": 0.00021061030124049903,
+      "loss": 3.2166,
+      "step": 36503
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6579669713973999,
+      "learning_rate": 0.0002105956045152985,
+      "loss": 2.8755,
+      "step": 36504
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6055128574371338,
+      "learning_rate": 0.00021058090802557198,
+      "loss": 2.7474,
+      "step": 36505
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6531729102134705,
+      "learning_rate": 0.000210566211771358,
+      "loss": 3.0321,
+      "step": 36506
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6186026334762573,
+      "learning_rate": 0.00021055151575269547,
+      "loss": 3.0292,
+      "step": 36507
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6538664102554321,
+      "learning_rate": 0.0002105368199696231,
+      "loss": 3.0134,
+      "step": 36508
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5929544568061829,
+      "learning_rate": 0.0002105221244221794,
+      "loss": 3.0028,
+      "step": 36509
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6697933673858643,
+      "learning_rate": 0.0002105074291104033,
+      "loss": 2.9312,
+      "step": 36510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6528447270393372,
+      "learning_rate": 0.00021049273403433337,
+      "loss": 2.8436,
+      "step": 36511
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6155452728271484,
+      "learning_rate": 0.0002104780391940083,
+      "loss": 3.1464,
+      "step": 36512
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6547124981880188,
+      "learning_rate": 0.00021046334458946697,
+      "loss": 2.8919,
+      "step": 36513
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6116988658905029,
+      "learning_rate": 0.0002104486502207479,
+      "loss": 3.0086,
+      "step": 36514
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6030555367469788,
+      "learning_rate": 0.00021043395608788993,
+      "loss": 3.0789,
+      "step": 36515
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6490576267242432,
+      "learning_rate": 0.0002104192621909315,
+      "loss": 2.8825,
+      "step": 36516
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.643524169921875,
+      "learning_rate": 0.00021040456852991157,
+      "loss": 2.931,
+      "step": 36517
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6385647654533386,
+      "learning_rate": 0.0002103898751048688,
+      "loss": 3.002,
+      "step": 36518
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6039292812347412,
+      "learning_rate": 0.00021037518191584188,
+      "loss": 3.1068,
+      "step": 36519
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6315106153488159,
+      "learning_rate": 0.0002103604889628695,
+      "loss": 3.1007,
+      "step": 36520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5782997012138367,
+      "learning_rate": 0.0002103457962459902,
+      "loss": 3.0973,
+      "step": 36521
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6795260310173035,
+      "learning_rate": 0.00021033110376524291,
+      "loss": 3.1494,
+      "step": 36522
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6649677157402039,
+      "learning_rate": 0.00021031641152066628,
+      "loss": 3.0697,
+      "step": 36523
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6292036771774292,
+      "learning_rate": 0.0002103017195122988,
+      "loss": 3.1645,
+      "step": 36524
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6512131690979004,
+      "learning_rate": 0.00021028702774017954,
+      "loss": 2.9064,
+      "step": 36525
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6254218816757202,
+      "learning_rate": 0.00021027233620434687,
+      "loss": 3.1291,
+      "step": 36526
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6537070274353027,
+      "learning_rate": 0.00021025764490483956,
+      "loss": 3.0194,
+      "step": 36527
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6867677569389343,
+      "learning_rate": 0.00021024295384169646,
+      "loss": 2.7768,
+      "step": 36528
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6541476249694824,
+      "learning_rate": 0.00021022826301495605,
+      "loss": 2.8653,
+      "step": 36529
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6634353995323181,
+      "learning_rate": 0.0002102135724246572,
+      "loss": 3.1185,
+      "step": 36530
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6761667728424072,
+      "learning_rate": 0.0002101988820708384,
+      "loss": 2.9844,
+      "step": 36531
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6391075849533081,
+      "learning_rate": 0.00021018419195353852,
+      "loss": 2.9843,
+      "step": 36532
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6631091833114624,
+      "learning_rate": 0.00021016950207279627,
+      "loss": 2.9588,
+      "step": 36533
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6815654039382935,
+      "learning_rate": 0.00021015481242865012,
+      "loss": 2.9905,
+      "step": 36534
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6456697583198547,
+      "learning_rate": 0.00021014012302113903,
+      "loss": 3.0181,
+      "step": 36535
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6634407639503479,
+      "learning_rate": 0.00021012543385030147,
+      "loss": 3.1435,
+      "step": 36536
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.678929328918457,
+      "learning_rate": 0.00021011074491617636,
+      "loss": 2.8749,
+      "step": 36537
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6539340019226074,
+      "learning_rate": 0.00021009605621880208,
+      "loss": 2.8567,
+      "step": 36538
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6301677227020264,
+      "learning_rate": 0.00021008136775821752,
+      "loss": 3.1227,
+      "step": 36539
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6353392601013184,
+      "learning_rate": 0.0002100666795344614,
+      "loss": 3.1078,
+      "step": 36540
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6301056742668152,
+      "learning_rate": 0.0002100519915475723,
+      "loss": 3.1015,
+      "step": 36541
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6307017803192139,
+      "learning_rate": 0.000210037303797589,
+      "loss": 2.948,
+      "step": 36542
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5895671248435974,
+      "learning_rate": 0.00021002261628455,
+      "loss": 3.1825,
+      "step": 36543
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6554360389709473,
+      "learning_rate": 0.00021000792900849422,
+      "loss": 3.0251,
+      "step": 36544
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6623850464820862,
+      "learning_rate": 0.00020999324196946026,
+      "loss": 2.9006,
+      "step": 36545
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6320748329162598,
+      "learning_rate": 0.00020997855516748665,
+      "loss": 2.932,
+      "step": 36546
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.655175507068634,
+      "learning_rate": 0.0002099638686026123,
+      "loss": 2.8594,
+      "step": 36547
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.719323456287384,
+      "learning_rate": 0.00020994918227487577,
+      "loss": 3.0602,
+      "step": 36548
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6324592232704163,
+      "learning_rate": 0.0002099344961843157,
+      "loss": 2.763,
+      "step": 36549
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5989832282066345,
+      "learning_rate": 0.00020991981033097095,
+      "loss": 2.6408,
+      "step": 36550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6369208693504333,
+      "learning_rate": 0.00020990512471488003,
+      "loss": 3.09,
+      "step": 36551
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6124125123023987,
+      "learning_rate": 0.00020989043933608169,
+      "loss": 3.0194,
+      "step": 36552
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6564167141914368,
+      "learning_rate": 0.00020987575419461452,
+      "loss": 3.2123,
+      "step": 36553
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6500516533851624,
+      "learning_rate": 0.00020986106929051732,
+      "loss": 3.0145,
+      "step": 36554
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6239467859268188,
+      "learning_rate": 0.00020984638462382877,
+      "loss": 3.0401,
+      "step": 36555
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.640312910079956,
+      "learning_rate": 0.00020983170019458737,
+      "loss": 3.0706,
+      "step": 36556
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6156535744667053,
+      "learning_rate": 0.00020981701600283204,
+      "loss": 2.9437,
+      "step": 36557
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6147194504737854,
+      "learning_rate": 0.00020980233204860122,
+      "loss": 3.0339,
+      "step": 36558
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.64543616771698,
+      "learning_rate": 0.00020978764833193376,
+      "loss": 2.9397,
+      "step": 36559
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.627521276473999,
+      "learning_rate": 0.0002097729648528683,
+      "loss": 2.9861,
+      "step": 36560
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6274319887161255,
+      "learning_rate": 0.0002097582816114434,
+      "loss": 2.9149,
+      "step": 36561
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6580536961555481,
+      "learning_rate": 0.00020974359860769792,
+      "loss": 2.9476,
+      "step": 36562
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6078628301620483,
+      "learning_rate": 0.0002097289158416704,
+      "loss": 2.911,
+      "step": 36563
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6547374725341797,
+      "learning_rate": 0.00020971423331339945,
+      "loss": 3.0069,
+      "step": 36564
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6162861585617065,
+      "learning_rate": 0.00020969955102292394,
+      "loss": 2.9177,
+      "step": 36565
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6419432163238525,
+      "learning_rate": 0.00020968486897028242,
+      "loss": 3.0997,
+      "step": 36566
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6692230105400085,
+      "learning_rate": 0.00020967018715551362,
+      "loss": 3.0527,
+      "step": 36567
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6301555633544922,
+      "learning_rate": 0.00020965550557865606,
+      "loss": 3.1243,
+      "step": 36568
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6723204851150513,
+      "learning_rate": 0.00020964082423974853,
+      "loss": 3.1511,
+      "step": 36569
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6410068273544312,
+      "learning_rate": 0.00020962614313882978,
+      "loss": 3.0315,
+      "step": 36570
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6543024182319641,
+      "learning_rate": 0.00020961146227593821,
+      "loss": 2.8805,
+      "step": 36571
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6214228868484497,
+      "learning_rate": 0.00020959678165111283,
+      "loss": 2.9974,
+      "step": 36572
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5998808145523071,
+      "learning_rate": 0.00020958210126439204,
+      "loss": 3.0079,
+      "step": 36573
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6029256582260132,
+      "learning_rate": 0.00020956742111581447,
+      "loss": 2.9974,
+      "step": 36574
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6017154455184937,
+      "learning_rate": 0.00020955274120541915,
+      "loss": 3.0842,
+      "step": 36575
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6353750228881836,
+      "learning_rate": 0.00020953806153324437,
+      "loss": 2.8448,
+      "step": 36576
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6240317821502686,
+      "learning_rate": 0.000209523382099329,
+      "loss": 3.0912,
+      "step": 36577
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6466373205184937,
+      "learning_rate": 0.0002095087029037115,
+      "loss": 2.9799,
+      "step": 36578
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7123924493789673,
+      "learning_rate": 0.00020949402394643062,
+      "loss": 3.0003,
+      "step": 36579
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6275160312652588,
+      "learning_rate": 0.00020947934522752524,
+      "loss": 2.9326,
+      "step": 36580
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5951224565505981,
+      "learning_rate": 0.00020946466674703378,
+      "loss": 2.9995,
+      "step": 36581
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6290155053138733,
+      "learning_rate": 0.00020944998850499501,
+      "loss": 2.9477,
+      "step": 36582
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.645408570766449,
+      "learning_rate": 0.00020943531050144745,
+      "loss": 3.0292,
+      "step": 36583
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6126695871353149,
+      "learning_rate": 0.00020942063273642988,
+      "loss": 2.7644,
+      "step": 36584
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6352578401565552,
+      "learning_rate": 0.000209405955209981,
+      "loss": 2.9679,
+      "step": 36585
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6484764218330383,
+      "learning_rate": 0.00020939127792213928,
+      "loss": 3.058,
+      "step": 36586
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6158818602561951,
+      "learning_rate": 0.00020937660087294366,
+      "loss": 3.0141,
+      "step": 36587
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6046254634857178,
+      "learning_rate": 0.0002093619240624325,
+      "loss": 3.0229,
+      "step": 36588
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6500994563102722,
+      "learning_rate": 0.00020934724749064457,
+      "loss": 2.7405,
+      "step": 36589
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6411953568458557,
+      "learning_rate": 0.00020933257115761866,
+      "loss": 2.9773,
+      "step": 36590
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6058011651039124,
+      "learning_rate": 0.00020931789506339324,
+      "loss": 3.0519,
+      "step": 36591
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6400535702705383,
+      "learning_rate": 0.0002093032192080071,
+      "loss": 2.8411,
+      "step": 36592
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6149111986160278,
+      "learning_rate": 0.0002092885435914987,
+      "loss": 3.0371,
+      "step": 36593
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.64310222864151,
+      "learning_rate": 0.00020927386821390688,
+      "loss": 2.8115,
+      "step": 36594
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.620502769947052,
+      "learning_rate": 0.0002092591930752703,
+      "loss": 2.8578,
+      "step": 36595
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6260932683944702,
+      "learning_rate": 0.00020924451817562738,
+      "loss": 3.0976,
+      "step": 36596
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6364988088607788,
+      "learning_rate": 0.00020922984351501707,
+      "loss": 3.0012,
+      "step": 36597
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6061363220214844,
+      "learning_rate": 0.00020921516909347778,
+      "loss": 3.0253,
+      "step": 36598
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6426119804382324,
+      "learning_rate": 0.00020920049491104833,
+      "loss": 3.0864,
+      "step": 36599
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6786094903945923,
+      "learning_rate": 0.00020918582096776737,
+      "loss": 2.8078,
+      "step": 36600
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.60618656873703,
+      "learning_rate": 0.0002091711472636733,
+      "loss": 2.9329,
+      "step": 36601
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6546763777732849,
+      "learning_rate": 0.00020915647379880515,
+      "loss": 3.0702,
+      "step": 36602
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6468024849891663,
+      "learning_rate": 0.00020914180057320127,
+      "loss": 3.0346,
+      "step": 36603
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6212300062179565,
+      "learning_rate": 0.00020912712758690033,
+      "loss": 3.0142,
+      "step": 36604
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6541531682014465,
+      "learning_rate": 0.00020911245483994117,
+      "loss": 2.9084,
+      "step": 36605
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.680400550365448,
+      "learning_rate": 0.00020909778233236224,
+      "loss": 2.9649,
+      "step": 36606
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6672208309173584,
+      "learning_rate": 0.00020908311006420234,
+      "loss": 2.812,
+      "step": 36607
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6206700205802917,
+      "learning_rate": 0.0002090684380354999,
+      "loss": 3.1004,
+      "step": 36608
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6560788154602051,
+      "learning_rate": 0.00020905376624629377,
+      "loss": 2.8389,
+      "step": 36609
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.61861652135849,
+      "learning_rate": 0.00020903909469662253,
+      "loss": 3.2776,
+      "step": 36610
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6285805702209473,
+      "learning_rate": 0.00020902442338652474,
+      "loss": 3.0981,
+      "step": 36611
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6283038258552551,
+      "learning_rate": 0.0002090097523160392,
+      "loss": 2.9732,
+      "step": 36612
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6750781536102295,
+      "learning_rate": 0.00020899508148520442,
+      "loss": 2.8903,
+      "step": 36613
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6676583886146545,
+      "learning_rate": 0.00020898041089405912,
+      "loss": 3.013,
+      "step": 36614
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6667542457580566,
+      "learning_rate": 0.00020896574054264183,
+      "loss": 2.9005,
+      "step": 36615
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6553710699081421,
+      "learning_rate": 0.00020895107043099125,
+      "loss": 2.7584,
+      "step": 36616
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6519506573677063,
+      "learning_rate": 0.00020893640055914613,
+      "loss": 3.0171,
+      "step": 36617
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6207144260406494,
+      "learning_rate": 0.00020892173092714491,
+      "loss": 2.9434,
+      "step": 36618
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6540098190307617,
+      "learning_rate": 0.0002089070615350264,
+      "loss": 3.1374,
+      "step": 36619
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6426064372062683,
+      "learning_rate": 0.00020889239238282902,
+      "loss": 2.8674,
+      "step": 36620
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6309428811073303,
+      "learning_rate": 0.00020887772347059165,
+      "loss": 2.9389,
+      "step": 36621
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6225348711013794,
+      "learning_rate": 0.00020886305479835286,
+      "loss": 2.9193,
+      "step": 36622
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.626919686794281,
+      "learning_rate": 0.00020884838636615113,
+      "loss": 3.2561,
+      "step": 36623
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.637225329875946,
+      "learning_rate": 0.00020883371817402533,
+      "loss": 2.9154,
+      "step": 36624
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7885748744010925,
+      "learning_rate": 0.00020881905022201392,
+      "loss": 2.9749,
+      "step": 36625
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6135236620903015,
+      "learning_rate": 0.00020880438251015548,
+      "loss": 3.3823,
+      "step": 36626
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.634709358215332,
+      "learning_rate": 0.00020878971503848892,
+      "loss": 2.9316,
+      "step": 36627
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6408893465995789,
+      "learning_rate": 0.00020877504780705258,
+      "loss": 3.0149,
+      "step": 36628
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8755736351013184,
+      "learning_rate": 0.00020876038081588534,
+      "loss": 2.8576,
+      "step": 36629
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6238410472869873,
+      "learning_rate": 0.00020874571406502551,
+      "loss": 2.9549,
+      "step": 36630
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.633625864982605,
+      "learning_rate": 0.00020873104755451198,
+      "loss": 2.7371,
+      "step": 36631
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6014534831047058,
+      "learning_rate": 0.00020871638128438342,
+      "loss": 3.0362,
+      "step": 36632
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6362054944038391,
+      "learning_rate": 0.00020870171525467812,
+      "loss": 2.9729,
+      "step": 36633
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6586917638778687,
+      "learning_rate": 0.00020868704946543518,
+      "loss": 3.0641,
+      "step": 36634
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6921262145042419,
+      "learning_rate": 0.00020867238391669284,
+      "loss": 3.0557,
+      "step": 36635
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6216939687728882,
+      "learning_rate": 0.0002086577186084898,
+      "loss": 3.0349,
+      "step": 36636
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6409044861793518,
+      "learning_rate": 0.00020864305354086489,
+      "loss": 3.1367,
+      "step": 36637
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6375213265419006,
+      "learning_rate": 0.00020862838871385655,
+      "loss": 3.0367,
+      "step": 36638
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6405977010726929,
+      "learning_rate": 0.00020861372412750347,
+      "loss": 3.1688,
+      "step": 36639
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6716753244400024,
+      "learning_rate": 0.00020859905978184415,
+      "loss": 3.0811,
+      "step": 36640
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7119506597518921,
+      "learning_rate": 0.0002085843956769173,
+      "loss": 3.0607,
+      "step": 36641
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6559655070304871,
+      "learning_rate": 0.0002085697318127617,
+      "loss": 3.1062,
+      "step": 36642
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6267669796943665,
+      "learning_rate": 0.0002085550681894158,
+      "loss": 3.0712,
+      "step": 36643
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6666412949562073,
+      "learning_rate": 0.00020854040480691824,
+      "loss": 3.0376,
+      "step": 36644
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6373438835144043,
+      "learning_rate": 0.0002085257416653076,
+      "loss": 3.1018,
+      "step": 36645
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6477558612823486,
+      "learning_rate": 0.00020851107876462257,
+      "loss": 3.0902,
+      "step": 36646
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6476315259933472,
+      "learning_rate": 0.00020849641610490185,
+      "loss": 3.0337,
+      "step": 36647
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6021722555160522,
+      "learning_rate": 0.0002084817536861838,
+      "loss": 3.0909,
+      "step": 36648
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6318761110305786,
+      "learning_rate": 0.00020846709150850736,
+      "loss": 2.9111,
+      "step": 36649
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6342579126358032,
+      "learning_rate": 0.00020845242957191087,
+      "loss": 3.0186,
+      "step": 36650
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6395136713981628,
+      "learning_rate": 0.00020843776787643302,
+      "loss": 3.0277,
+      "step": 36651
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.613551914691925,
+      "learning_rate": 0.0002084231064221126,
+      "loss": 3.0372,
+      "step": 36652
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6066928505897522,
+      "learning_rate": 0.00020840844520898804,
+      "loss": 3.0808,
+      "step": 36653
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6370232701301575,
+      "learning_rate": 0.00020839378423709805,
+      "loss": 3.1052,
+      "step": 36654
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6237879991531372,
+      "learning_rate": 0.0002083791235064811,
+      "loss": 2.753,
+      "step": 36655
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6108696460723877,
+      "learning_rate": 0.00020836446301717598,
+      "loss": 2.9691,
+      "step": 36656
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6503059267997742,
+      "learning_rate": 0.00020834980276922125,
+      "loss": 3.0589,
+      "step": 36657
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6078692078590393,
+      "learning_rate": 0.0002083351427626554,
+      "loss": 3.2604,
+      "step": 36658
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6196121573448181,
+      "learning_rate": 0.00020832048299751726,
+      "loss": 3.1257,
+      "step": 36659
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.651357889175415,
+      "learning_rate": 0.0002083058234738452,
+      "loss": 2.9815,
+      "step": 36660
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5992416143417358,
+      "learning_rate": 0.000208291164191678,
+      "loss": 2.7149,
+      "step": 36661
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6600878238677979,
+      "learning_rate": 0.0002082765051510543,
+      "loss": 3.1287,
+      "step": 36662
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6135623455047607,
+      "learning_rate": 0.00020826184635201254,
+      "loss": 2.8863,
+      "step": 36663
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.614033579826355,
+      "learning_rate": 0.0002082471877945915,
+      "loss": 3.1524,
+      "step": 36664
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6445769667625427,
+      "learning_rate": 0.00020823252947882969,
+      "loss": 2.993,
+      "step": 36665
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6738548874855042,
+      "learning_rate": 0.00020821787140476564,
+      "loss": 3.0463,
+      "step": 36666
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.678357720375061,
+      "learning_rate": 0.00020820321357243818,
+      "loss": 2.7473,
+      "step": 36667
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6646535396575928,
+      "learning_rate": 0.00020818855598188574,
+      "loss": 2.9655,
+      "step": 36668
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6294272541999817,
+      "learning_rate": 0.000208173898633147,
+      "loss": 3.0103,
+      "step": 36669
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6239528059959412,
+      "learning_rate": 0.0002081592415262605,
+      "loss": 3.0741,
+      "step": 36670
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.630536675453186,
+      "learning_rate": 0.00020814458466126485,
+      "loss": 3.0073,
+      "step": 36671
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6464529633522034,
+      "learning_rate": 0.0002081299280381988,
+      "loss": 3.1235,
+      "step": 36672
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6047073602676392,
+      "learning_rate": 0.00020811527165710069,
+      "loss": 2.9439,
+      "step": 36673
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6120465397834778,
+      "learning_rate": 0.0002081006155180094,
+      "loss": 2.8782,
+      "step": 36674
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.631659984588623,
+      "learning_rate": 0.00020808595962096336,
+      "loss": 3.1335,
+      "step": 36675
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6841307282447815,
+      "learning_rate": 0.00020807130396600115,
+      "loss": 2.9434,
+      "step": 36676
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.626285970211029,
+      "learning_rate": 0.00020805664855316153,
+      "loss": 2.8426,
+      "step": 36677
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6159030795097351,
+      "learning_rate": 0.00020804199338248292,
+      "loss": 3.0015,
+      "step": 36678
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.593059778213501,
+      "learning_rate": 0.00020802733845400411,
+      "loss": 2.8796,
+      "step": 36679
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7196869254112244,
+      "learning_rate": 0.00020801268376776348,
+      "loss": 2.7227,
+      "step": 36680
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6337694525718689,
+      "learning_rate": 0.0002079980293237997,
+      "loss": 3.0181,
+      "step": 36681
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.612427830696106,
+      "learning_rate": 0.00020798337512215154,
+      "loss": 2.9788,
+      "step": 36682
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.629952073097229,
+      "learning_rate": 0.0002079687211628574,
+      "loss": 2.9669,
+      "step": 36683
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.657768964767456,
+      "learning_rate": 0.00020795406744595604,
+      "loss": 2.9131,
+      "step": 36684
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6409008502960205,
+      "learning_rate": 0.00020793941397148576,
+      "loss": 3.0031,
+      "step": 36685
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6390206217765808,
+      "learning_rate": 0.00020792476073948544,
+      "loss": 3.0613,
+      "step": 36686
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.663670539855957,
+      "learning_rate": 0.00020791010774999367,
+      "loss": 3.0369,
+      "step": 36687
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6246560215950012,
+      "learning_rate": 0.00020789545500304873,
+      "loss": 3.0468,
+      "step": 36688
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6665436625480652,
+      "learning_rate": 0.00020788080249868965,
+      "loss": 2.7913,
+      "step": 36689
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6187042593955994,
+      "learning_rate": 0.00020786615023695473,
+      "loss": 2.8444,
+      "step": 36690
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.659570038318634,
+      "learning_rate": 0.0002078514982178827,
+      "loss": 2.9988,
+      "step": 36691
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6804064512252808,
+      "learning_rate": 0.00020783684644151194,
+      "loss": 3.032,
+      "step": 36692
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6458125710487366,
+      "learning_rate": 0.00020782219490788126,
+      "loss": 2.9141,
+      "step": 36693
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6113386154174805,
+      "learning_rate": 0.00020780754361702925,
+      "loss": 2.9778,
+      "step": 36694
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6785792112350464,
+      "learning_rate": 0.00020779289256899435,
+      "loss": 3.1989,
+      "step": 36695
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5986670851707458,
+      "learning_rate": 0.0002077782417638153,
+      "loss": 2.8518,
+      "step": 36696
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6187350749969482,
+      "learning_rate": 0.00020776359120153054,
+      "loss": 2.9814,
+      "step": 36697
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6327255368232727,
+      "learning_rate": 0.00020774894088217868,
+      "loss": 3.1217,
+      "step": 36698
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6451022028923035,
+      "learning_rate": 0.0002077342908057985,
+      "loss": 3.1749,
+      "step": 36699
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6500363945960999,
+      "learning_rate": 0.00020771964097242827,
+      "loss": 2.8666,
+      "step": 36700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.659353494644165,
+      "learning_rate": 0.00020770499138210696,
+      "loss": 3.2217,
+      "step": 36701
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6432385444641113,
+      "learning_rate": 0.00020769034203487274,
+      "loss": 2.7555,
+      "step": 36702
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7030990123748779,
+      "learning_rate": 0.00020767569293076444,
+      "loss": 2.9348,
+      "step": 36703
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6431054472923279,
+      "learning_rate": 0.00020766104406982067,
+      "loss": 2.866,
+      "step": 36704
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6525076627731323,
+      "learning_rate": 0.00020764639545207988,
+      "loss": 3.0982,
+      "step": 36705
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6153354644775391,
+      "learning_rate": 0.0002076317470775808,
+      "loss": 2.7865,
+      "step": 36706
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6521280407905579,
+      "learning_rate": 0.0002076170989463618,
+      "loss": 2.9171,
+      "step": 36707
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.677487313747406,
+      "learning_rate": 0.00020760245105846164,
+      "loss": 3.1123,
+      "step": 36708
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6668803095817566,
+      "learning_rate": 0.00020758780341391888,
+      "loss": 2.9094,
+      "step": 36709
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6921082139015198,
+      "learning_rate": 0.00020757315601277197,
+      "loss": 3.0438,
+      "step": 36710
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6118077039718628,
+      "learning_rate": 0.00020755850885505965,
+      "loss": 3.111,
+      "step": 36711
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6800283789634705,
+      "learning_rate": 0.00020754386194082037,
+      "loss": 2.9849,
+      "step": 36712
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6486184597015381,
+      "learning_rate": 0.00020752921527009277,
+      "loss": 2.8923,
+      "step": 36713
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.631157636642456,
+      "learning_rate": 0.0002075145688429155,
+      "loss": 3.034,
+      "step": 36714
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.610759973526001,
+      "learning_rate": 0.00020749992265932697,
+      "loss": 3.0504,
+      "step": 36715
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6135387420654297,
+      "learning_rate": 0.00020748527671936596,
+      "loss": 2.8504,
+      "step": 36716
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6542940139770508,
+      "learning_rate": 0.00020747063102307078,
+      "loss": 3.0165,
+      "step": 36717
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7032894492149353,
+      "learning_rate": 0.00020745598557048018,
+      "loss": 2.9044,
+      "step": 36718
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6352038383483887,
+      "learning_rate": 0.0002074413403616328,
+      "loss": 2.8658,
+      "step": 36719
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6255574822425842,
+      "learning_rate": 0.000207426695396567,
+      "loss": 3.1381,
+      "step": 36720
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6670256853103638,
+      "learning_rate": 0.00020741205067532158,
+      "loss": 2.993,
+      "step": 36721
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6089436411857605,
+      "learning_rate": 0.00020739740619793486,
+      "loss": 3.0545,
+      "step": 36722
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6588670015335083,
+      "learning_rate": 0.0002073827619644456,
+      "loss": 3.0358,
+      "step": 36723
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6333926916122437,
+      "learning_rate": 0.00020736811797489243,
+      "loss": 2.8797,
+      "step": 36724
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6069063544273376,
+      "learning_rate": 0.00020735347422931366,
+      "loss": 3.113,
+      "step": 36725
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6227714419364929,
+      "learning_rate": 0.00020733883072774815,
+      "loss": 3.0379,
+      "step": 36726
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.635350227355957,
+      "learning_rate": 0.00020732418747023424,
+      "loss": 3.0705,
+      "step": 36727
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6420906186103821,
+      "learning_rate": 0.00020730954445681053,
+      "loss": 2.8377,
+      "step": 36728
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6361296772956848,
+      "learning_rate": 0.0002072949016875158,
+      "loss": 3.1277,
+      "step": 36729
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6772409677505493,
+      "learning_rate": 0.00020728025916238836,
+      "loss": 2.8932,
+      "step": 36730
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6195856928825378,
+      "learning_rate": 0.00020726561688146696,
+      "loss": 3.0179,
+      "step": 36731
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6762132048606873,
+      "learning_rate": 0.0002072509748447899,
+      "loss": 3.0329,
+      "step": 36732
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6098534464836121,
+      "learning_rate": 0.000207236333052396,
+      "loss": 2.9847,
+      "step": 36733
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6518588066101074,
+      "learning_rate": 0.00020722169150432383,
+      "loss": 3.0305,
+      "step": 36734
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6830816268920898,
+      "learning_rate": 0.00020720705020061173,
+      "loss": 3.1915,
+      "step": 36735
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6383629441261292,
+      "learning_rate": 0.00020719240914129854,
+      "loss": 3.1503,
+      "step": 36736
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5932306051254272,
+      "learning_rate": 0.0002071777683264226,
+      "loss": 3.0398,
+      "step": 36737
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6360545754432678,
+      "learning_rate": 0.00020716312775602254,
+      "loss": 3.0491,
+      "step": 36738
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5986540913581848,
+      "learning_rate": 0.000207148487430137,
+      "loss": 2.8491,
+      "step": 36739
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6128685474395752,
+      "learning_rate": 0.00020713384734880445,
+      "loss": 2.9966,
+      "step": 36740
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5811918377876282,
+      "learning_rate": 0.00020711920751206354,
+      "loss": 2.828,
+      "step": 36741
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6315313577651978,
+      "learning_rate": 0.00020710456791995266,
+      "loss": 3.0263,
+      "step": 36742
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5910419821739197,
+      "learning_rate": 0.0002070899285725104,
+      "loss": 2.8827,
+      "step": 36743
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6899007558822632,
+      "learning_rate": 0.00020707528946977555,
+      "loss": 2.9803,
+      "step": 36744
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6172998547554016,
+      "learning_rate": 0.00020706065061178644,
+      "loss": 3.0134,
+      "step": 36745
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6163078546524048,
+      "learning_rate": 0.00020704601199858172,
+      "loss": 3.1796,
+      "step": 36746
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6540799140930176,
+      "learning_rate": 0.00020703137363019985,
+      "loss": 2.8401,
+      "step": 36747
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6538397073745728,
+      "learning_rate": 0.0002070167355066795,
+      "loss": 2.8745,
+      "step": 36748
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7262605428695679,
+      "learning_rate": 0.00020700209762805922,
+      "loss": 3.2279,
+      "step": 36749
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6116203665733337,
+      "learning_rate": 0.00020698745999437736,
+      "loss": 3.0517,
+      "step": 36750
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6457718014717102,
+      "learning_rate": 0.0002069728226056728,
+      "loss": 2.8412,
+      "step": 36751
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6367230415344238,
+      "learning_rate": 0.00020695818546198383,
+      "loss": 3.0288,
+      "step": 36752
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6102105379104614,
+      "learning_rate": 0.00020694354856334908,
+      "loss": 2.8687,
+      "step": 36753
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.679972767829895,
+      "learning_rate": 0.00020692891190980716,
+      "loss": 2.9442,
+      "step": 36754
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6194684505462646,
+      "learning_rate": 0.00020691427550139656,
+      "loss": 2.9587,
+      "step": 36755
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7451155781745911,
+      "learning_rate": 0.00020689963933815593,
+      "loss": 3.0858,
+      "step": 36756
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.615552544593811,
+      "learning_rate": 0.00020688500342012355,
+      "loss": 2.8748,
+      "step": 36757
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6059805750846863,
+      "learning_rate": 0.00020687036774733823,
+      "loss": 2.7704,
+      "step": 36758
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6106909513473511,
+      "learning_rate": 0.00020685573231983852,
+      "loss": 2.8743,
+      "step": 36759
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6602317094802856,
+      "learning_rate": 0.00020684109713766278,
+      "loss": 3.0426,
+      "step": 36760
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6297293305397034,
+      "learning_rate": 0.00020682646220084974,
+      "loss": 3.0277,
+      "step": 36761
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6300654411315918,
+      "learning_rate": 0.00020681182750943776,
+      "loss": 3.0978,
+      "step": 36762
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6634638905525208,
+      "learning_rate": 0.00020679719306346557,
+      "loss": 3.0136,
+      "step": 36763
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6517794728279114,
+      "learning_rate": 0.00020678255886297171,
+      "loss": 2.971,
+      "step": 36764
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6523165106773376,
+      "learning_rate": 0.00020676792490799447,
+      "loss": 3.0286,
+      "step": 36765
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6734626889228821,
+      "learning_rate": 0.00020675329119857275,
+      "loss": 3.1295,
+      "step": 36766
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.632370114326477,
+      "learning_rate": 0.00020673865773474477,
+      "loss": 3.0338,
+      "step": 36767
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6716340780258179,
+      "learning_rate": 0.0002067240245165494,
+      "loss": 3.0109,
+      "step": 36768
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.641351044178009,
+      "learning_rate": 0.00020670939154402478,
+      "loss": 2.9696,
+      "step": 36769
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.610369086265564,
+      "learning_rate": 0.00020669475881720975,
+      "loss": 3.0894,
+      "step": 36770
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6596361398696899,
+      "learning_rate": 0.00020668012633614287,
+      "loss": 3.0013,
+      "step": 36771
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6531994938850403,
+      "learning_rate": 0.00020666549410086244,
+      "loss": 3.2047,
+      "step": 36772
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6469523906707764,
+      "learning_rate": 0.00020665086211140723,
+      "loss": 2.853,
+      "step": 36773
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6518253684043884,
+      "learning_rate": 0.00020663623036781565,
+      "loss": 2.9245,
+      "step": 36774
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.617925763130188,
+      "learning_rate": 0.00020662159887012616,
+      "loss": 2.7722,
+      "step": 36775
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6405405402183533,
+      "learning_rate": 0.00020660696761837761,
+      "loss": 3.2794,
+      "step": 36776
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6911517977714539,
+      "learning_rate": 0.00020659233661260822,
+      "loss": 2.9798,
+      "step": 36777
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.627225935459137,
+      "learning_rate": 0.0002065777058528567,
+      "loss": 3.0352,
+      "step": 36778
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6320593953132629,
+      "learning_rate": 0.00020656307533916142,
+      "loss": 2.9623,
+      "step": 36779
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6159602403640747,
+      "learning_rate": 0.00020654844507156107,
+      "loss": 3.0543,
+      "step": 36780
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6653395891189575,
+      "learning_rate": 0.00020653381505009418,
+      "loss": 2.9946,
+      "step": 36781
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6371135711669922,
+      "learning_rate": 0.00020651918527479914,
+      "loss": 3.1782,
+      "step": 36782
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6792336106300354,
+      "learning_rate": 0.0002065045557457147,
+      "loss": 3.0116,
+      "step": 36783
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6385068297386169,
+      "learning_rate": 0.0002064899264628791,
+      "loss": 2.9848,
+      "step": 36784
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6543515920639038,
+      "learning_rate": 0.00020647529742633117,
+      "loss": 2.9478,
+      "step": 36785
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6753814816474915,
+      "learning_rate": 0.00020646066863610931,
+      "loss": 3.0857,
+      "step": 36786
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6674555540084839,
+      "learning_rate": 0.00020644604009225191,
+      "loss": 3.0606,
+      "step": 36787
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6608149409294128,
+      "learning_rate": 0.00020643141179479775,
+      "loss": 3.1004,
+      "step": 36788
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6545112729072571,
+      "learning_rate": 0.00020641678374378527,
+      "loss": 2.8895,
+      "step": 36789
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6464564800262451,
+      "learning_rate": 0.0002064021559392528,
+      "loss": 3.1448,
+      "step": 36790
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6789308190345764,
+      "learning_rate": 0.00020638752838123925,
+      "loss": 3.0127,
+      "step": 36791
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6076009273529053,
+      "learning_rate": 0.00020637290106978283,
+      "loss": 3.0271,
+      "step": 36792
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6446407437324524,
+      "learning_rate": 0.00020635827400492223,
+      "loss": 2.9744,
+      "step": 36793
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6316608786582947,
+      "learning_rate": 0.00020634364718669584,
+      "loss": 3.008,
+      "step": 36794
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6050397157669067,
+      "learning_rate": 0.0002063290206151423,
+      "loss": 2.8838,
+      "step": 36795
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6316719651222229,
+      "learning_rate": 0.00020631439429030012,
+      "loss": 2.9403,
+      "step": 36796
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6683484315872192,
+      "learning_rate": 0.00020629976821220767,
+      "loss": 3.0497,
+      "step": 36797
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.663461446762085,
+      "learning_rate": 0.00020628514238090373,
+      "loss": 3.0129,
+      "step": 36798
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6077750325202942,
+      "learning_rate": 0.00020627051679642666,
+      "loss": 2.9524,
+      "step": 36799
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.632233738899231,
+      "learning_rate": 0.0002062558914588149,
+      "loss": 2.8329,
+      "step": 36800
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6921725273132324,
+      "learning_rate": 0.00020624126636810725,
+      "loss": 3.0573,
+      "step": 36801
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6208415031433105,
+      "learning_rate": 0.000206226641524342,
+      "loss": 2.8597,
+      "step": 36802
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6545069217681885,
+      "learning_rate": 0.00020621201692755777,
+      "loss": 3.0219,
+      "step": 36803
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.649098813533783,
+      "learning_rate": 0.0002061973925777929,
+      "loss": 3.1196,
+      "step": 36804
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6079701781272888,
+      "learning_rate": 0.00020618276847508604,
+      "loss": 2.788,
+      "step": 36805
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6139565110206604,
+      "learning_rate": 0.00020616814461947588,
+      "loss": 2.9477,
+      "step": 36806
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6319758296012878,
+      "learning_rate": 0.0002061535210110007,
+      "loss": 3.0009,
+      "step": 36807
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6935994029045105,
+      "learning_rate": 0.0002061388976496991,
+      "loss": 2.9628,
+      "step": 36808
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6240067481994629,
+      "learning_rate": 0.00020612427453560946,
+      "loss": 2.795,
+      "step": 36809
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6220824122428894,
+      "learning_rate": 0.00020610965166877054,
+      "loss": 3.0931,
+      "step": 36810
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6552855968475342,
+      "learning_rate": 0.00020609502904922074,
+      "loss": 2.8704,
+      "step": 36811
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6498665809631348,
+      "learning_rate": 0.00020608040667699845,
+      "loss": 3.0281,
+      "step": 36812
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6534984707832336,
+      "learning_rate": 0.00020606578455214242,
+      "loss": 3.0395,
+      "step": 36813
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7032010555267334,
+      "learning_rate": 0.00020605116267469097,
+      "loss": 3.0231,
+      "step": 36814
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6376596093177795,
+      "learning_rate": 0.0002060365410446826,
+      "loss": 3.1822,
+      "step": 36815
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6410520672798157,
+      "learning_rate": 0.00020602191966215605,
+      "loss": 3.0589,
+      "step": 36816
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6917757391929626,
+      "learning_rate": 0.00020600729852714958,
+      "loss": 2.9896,
+      "step": 36817
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6153084635734558,
+      "learning_rate": 0.0002059926776397019,
+      "loss": 3.0361,
+      "step": 36818
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.720618724822998,
+      "learning_rate": 0.00020597805699985124,
+      "loss": 3.0778,
+      "step": 36819
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6662018895149231,
+      "learning_rate": 0.00020596343660763642,
+      "loss": 2.7993,
+      "step": 36820
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6655521392822266,
+      "learning_rate": 0.0002059488164630958,
+      "loss": 2.942,
+      "step": 36821
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6253843307495117,
+      "learning_rate": 0.00020593419656626778,
+      "loss": 3.0623,
+      "step": 36822
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6495913863182068,
+      "learning_rate": 0.00020591957691719112,
+      "loss": 2.8426,
+      "step": 36823
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6686586737632751,
+      "learning_rate": 0.00020590495751590408,
+      "loss": 2.9081,
+      "step": 36824
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6453542113304138,
+      "learning_rate": 0.00020589033836244534,
+      "loss": 3.1822,
+      "step": 36825
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6410621404647827,
+      "learning_rate": 0.0002058757194568534,
+      "loss": 2.809,
+      "step": 36826
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6789267659187317,
+      "learning_rate": 0.00020586110079916654,
+      "loss": 2.9868,
+      "step": 36827
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6374423503875732,
+      "learning_rate": 0.00020584648238942358,
+      "loss": 3.154,
+      "step": 36828
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6617890000343323,
+      "learning_rate": 0.0002058318642276628,
+      "loss": 3.0522,
+      "step": 36829
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6505021452903748,
+      "learning_rate": 0.00020581724631392267,
+      "loss": 3.2386,
+      "step": 36830
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6423779726028442,
+      "learning_rate": 0.00020580262864824193,
+      "loss": 3.1164,
+      "step": 36831
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6466403603553772,
+      "learning_rate": 0.0002057880112306589,
+      "loss": 3.0535,
+      "step": 36832
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6527153849601746,
+      "learning_rate": 0.00020577339406121215,
+      "loss": 2.8668,
+      "step": 36833
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6217302083969116,
+      "learning_rate": 0.00020575877713994002,
+      "loss": 3.0263,
+      "step": 36834
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6141082048416138,
+      "learning_rate": 0.00020574416046688124,
+      "loss": 3.1425,
+      "step": 36835
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6378345489501953,
+      "learning_rate": 0.00020572954404207423,
+      "loss": 3.1281,
+      "step": 36836
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6423177719116211,
+      "learning_rate": 0.00020571492786555733,
+      "loss": 3.0076,
+      "step": 36837
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6553202867507935,
+      "learning_rate": 0.0002057003119373693,
+      "loss": 2.9455,
+      "step": 36838
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.627715528011322,
+      "learning_rate": 0.00020568569625754844,
+      "loss": 2.9104,
+      "step": 36839
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6472757458686829,
+      "learning_rate": 0.00020567108082613324,
+      "loss": 2.9738,
+      "step": 36840
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6219276785850525,
+      "learning_rate": 0.00020565646564316237,
+      "loss": 3.0335,
+      "step": 36841
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6370683312416077,
+      "learning_rate": 0.00020564185070867416,
+      "loss": 3.1745,
+      "step": 36842
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6401722431182861,
+      "learning_rate": 0.0002056272360227072,
+      "loss": 3.118,
+      "step": 36843
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7073693871498108,
+      "learning_rate": 0.00020561262158529984,
+      "loss": 2.9147,
+      "step": 36844
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.637370228767395,
+      "learning_rate": 0.00020559800739649065,
+      "loss": 3.0401,
+      "step": 36845
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.658949077129364,
+      "learning_rate": 0.00020558339345631827,
+      "loss": 2.925,
+      "step": 36846
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6802607774734497,
+      "learning_rate": 0.000205568779764821,
+      "loss": 2.9548,
+      "step": 36847
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6683867573738098,
+      "learning_rate": 0.00020555416632203749,
+      "loss": 3.0257,
+      "step": 36848
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6157516241073608,
+      "learning_rate": 0.00020553955312800596,
+      "loss": 3.0218,
+      "step": 36849
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6470605134963989,
+      "learning_rate": 0.0002055249401827652,
+      "loss": 2.8591,
+      "step": 36850
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7139477729797363,
+      "learning_rate": 0.0002055103274863535,
+      "loss": 2.8813,
+      "step": 36851
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5850152373313904,
+      "learning_rate": 0.00020549571503880937,
+      "loss": 2.9898,
+      "step": 36852
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6298661828041077,
+      "learning_rate": 0.00020548110284017144,
+      "loss": 3.1898,
+      "step": 36853
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6115210056304932,
+      "learning_rate": 0.00020546649089047805,
+      "loss": 2.9295,
+      "step": 36854
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6343759894371033,
+      "learning_rate": 0.0002054518791897678,
+      "loss": 3.2084,
+      "step": 36855
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6652531623840332,
+      "learning_rate": 0.00020543726773807896,
+      "loss": 2.8345,
+      "step": 36856
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6212843060493469,
+      "learning_rate": 0.00020542265653545022,
+      "loss": 2.9031,
+      "step": 36857
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6636337637901306,
+      "learning_rate": 0.00020540804558192007,
+      "loss": 3.254,
+      "step": 36858
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6373041868209839,
+      "learning_rate": 0.0002053934348775268,
+      "loss": 2.9458,
+      "step": 36859
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6539564728736877,
+      "learning_rate": 0.00020537882442230916,
+      "loss": 2.9724,
+      "step": 36860
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6322240829467773,
+      "learning_rate": 0.00020536421421630537,
+      "loss": 2.9376,
+      "step": 36861
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6140871047973633,
+      "learning_rate": 0.000205349604259554,
+      "loss": 3.0766,
+      "step": 36862
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6811372637748718,
+      "learning_rate": 0.00020533499455209364,
+      "loss": 2.9396,
+      "step": 36863
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6444950699806213,
+      "learning_rate": 0.0002053203850939626,
+      "loss": 2.9081,
+      "step": 36864
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6304511427879333,
+      "learning_rate": 0.00020530577588519968,
+      "loss": 3.3028,
+      "step": 36865
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6575809121131897,
+      "learning_rate": 0.00020529116692584286,
+      "loss": 2.9909,
+      "step": 36866
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6433678865432739,
+      "learning_rate": 0.00020527655821593092,
+      "loss": 3.0047,
+      "step": 36867
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6921300888061523,
+      "learning_rate": 0.00020526194975550242,
+      "loss": 3.0075,
+      "step": 36868
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6458017230033875,
+      "learning_rate": 0.00020524734154459562,
+      "loss": 3.1427,
+      "step": 36869
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6515001058578491,
+      "learning_rate": 0.00020523273358324912,
+      "loss": 2.9976,
+      "step": 36870
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6600772142410278,
+      "learning_rate": 0.0002052181258715013,
+      "loss": 2.9814,
+      "step": 36871
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6418211460113525,
+      "learning_rate": 0.00020520351840939074,
+      "loss": 3.1198,
+      "step": 36872
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6537681221961975,
+      "learning_rate": 0.00020518891119695592,
+      "loss": 3.0484,
+      "step": 36873
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.669597327709198,
+      "learning_rate": 0.00020517430423423513,
+      "loss": 2.9601,
+      "step": 36874
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6508209109306335,
+      "learning_rate": 0.0002051596975212671,
+      "loss": 2.887,
+      "step": 36875
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.580568790435791,
+      "learning_rate": 0.00020514509105809013,
+      "loss": 3.0521,
+      "step": 36876
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6386202573776245,
+      "learning_rate": 0.00020513048484474265,
+      "loss": 2.9794,
+      "step": 36877
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6095972657203674,
+      "learning_rate": 0.00020511587888126338,
+      "loss": 3.0316,
+      "step": 36878
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6701319813728333,
+      "learning_rate": 0.00020510127316769055,
+      "loss": 2.9153,
+      "step": 36879
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6238023042678833,
+      "learning_rate": 0.00020508666770406277,
+      "loss": 3.0534,
+      "step": 36880
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6750576496124268,
+      "learning_rate": 0.0002050720624904183,
+      "loss": 2.9337,
+      "step": 36881
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.67641282081604,
+      "learning_rate": 0.0002050574575267958,
+      "loss": 3.0228,
+      "step": 36882
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6155452132225037,
+      "learning_rate": 0.00020504285281323375,
+      "loss": 3.0172,
+      "step": 36883
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6545352339744568,
+      "learning_rate": 0.0002050282483497704,
+      "loss": 2.9602,
+      "step": 36884
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.630028247833252,
+      "learning_rate": 0.00020501364413644455,
+      "loss": 3.0454,
+      "step": 36885
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6380795836448669,
+      "learning_rate": 0.00020499904017329433,
+      "loss": 3.0451,
+      "step": 36886
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6383315324783325,
+      "learning_rate": 0.0002049844364603584,
+      "loss": 3.048,
+      "step": 36887
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6407351493835449,
+      "learning_rate": 0.00020496983299767528,
+      "loss": 3.145,
+      "step": 36888
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6724693775177002,
+      "learning_rate": 0.00020495522978528316,
+      "loss": 3.084,
+      "step": 36889
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.625685453414917,
+      "learning_rate": 0.00020494062682322085,
+      "loss": 3.1366,
+      "step": 36890
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6715691089630127,
+      "learning_rate": 0.00020492602411152655,
+      "loss": 3.0932,
+      "step": 36891
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.635994017124176,
+      "learning_rate": 0.00020491142165023874,
+      "loss": 2.9926,
+      "step": 36892
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6208715438842773,
+      "learning_rate": 0.00020489681943939604,
+      "loss": 3.1724,
+      "step": 36893
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6574182510375977,
+      "learning_rate": 0.0002048822174790368,
+      "loss": 3.1069,
+      "step": 36894
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6716117858886719,
+      "learning_rate": 0.0002048676157691995,
+      "loss": 2.7032,
+      "step": 36895
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6087467670440674,
+      "learning_rate": 0.00020485301430992254,
+      "loss": 3.0921,
+      "step": 36896
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6368101835250854,
+      "learning_rate": 0.00020483841310124445,
+      "loss": 3.028,
+      "step": 36897
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.634164035320282,
+      "learning_rate": 0.0002048238121432037,
+      "loss": 3.1461,
+      "step": 36898
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6230823993682861,
+      "learning_rate": 0.00020480921143583864,
+      "loss": 2.879,
+      "step": 36899
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.656328022480011,
+      "learning_rate": 0.00020479461097918787,
+      "loss": 3.0447,
+      "step": 36900
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6206152439117432,
+      "learning_rate": 0.00020478001077328973,
+      "loss": 2.7903,
+      "step": 36901
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6653019785881042,
+      "learning_rate": 0.00020476541081818268,
+      "loss": 3.1942,
+      "step": 36902
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6336932182312012,
+      "learning_rate": 0.00020475081111390533,
+      "loss": 2.9204,
+      "step": 36903
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.66854327917099,
+      "learning_rate": 0.00020473621166049592,
+      "loss": 2.9795,
+      "step": 36904
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6467189788818359,
+      "learning_rate": 0.0002047216124579931,
+      "loss": 2.8518,
+      "step": 36905
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6561312079429626,
+      "learning_rate": 0.00020470701350643503,
+      "loss": 2.9816,
+      "step": 36906
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6752562522888184,
+      "learning_rate": 0.0002046924148058604,
+      "loss": 3.0072,
+      "step": 36907
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6663380861282349,
+      "learning_rate": 0.00020467781635630773,
+      "loss": 3.0714,
+      "step": 36908
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6871305108070374,
+      "learning_rate": 0.00020466321815781525,
+      "loss": 2.7063,
+      "step": 36909
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6497218608856201,
+      "learning_rate": 0.0002046486202104216,
+      "loss": 2.948,
+      "step": 36910
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.633631706237793,
+      "learning_rate": 0.000204634022514165,
+      "loss": 2.9519,
+      "step": 36911
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6532074809074402,
+      "learning_rate": 0.00020461942506908408,
+      "loss": 2.8991,
+      "step": 36912
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6703870892524719,
+      "learning_rate": 0.0002046048278752173,
+      "loss": 3.0202,
+      "step": 36913
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6608051657676697,
+      "learning_rate": 0.00020459023093260295,
+      "loss": 3.1402,
+      "step": 36914
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6233830451965332,
+      "learning_rate": 0.00020457563424127967,
+      "loss": 3.1028,
+      "step": 36915
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6100077629089355,
+      "learning_rate": 0.00020456103780128575,
+      "loss": 2.8412,
+      "step": 36916
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.60553377866745,
+      "learning_rate": 0.00020454644161265958,
+      "loss": 2.9459,
+      "step": 36917
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6492863893508911,
+      "learning_rate": 0.00020453184567543992,
+      "loss": 3.0801,
+      "step": 36918
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6354946494102478,
+      "learning_rate": 0.00020451724998966488,
+      "loss": 3.0188,
+      "step": 36919
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.617892861366272,
+      "learning_rate": 0.00020450265455537308,
+      "loss": 2.9104,
+      "step": 36920
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6572944521903992,
+      "learning_rate": 0.00020448805937260287,
+      "loss": 2.9129,
+      "step": 36921
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7081995010375977,
+      "learning_rate": 0.00020447346444139275,
+      "loss": 2.9972,
+      "step": 36922
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6335873007774353,
+      "learning_rate": 0.00020445886976178115,
+      "loss": 3.0526,
+      "step": 36923
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6316371560096741,
+      "learning_rate": 0.00020444427533380642,
+      "loss": 2.906,
+      "step": 36924
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6577939391136169,
+      "learning_rate": 0.00020442968115750722,
+      "loss": 2.9412,
+      "step": 36925
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6616895794868469,
+      "learning_rate": 0.0002044150872329217,
+      "loss": 2.8354,
+      "step": 36926
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6275897026062012,
+      "learning_rate": 0.0002044004935600886,
+      "loss": 3.1783,
+      "step": 36927
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6467520594596863,
+      "learning_rate": 0.00020438590013904605,
+      "loss": 3.1131,
+      "step": 36928
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6437161564826965,
+      "learning_rate": 0.00020437130696983265,
+      "loss": 3.0071,
+      "step": 36929
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6312966346740723,
+      "learning_rate": 0.00020435671405248694,
+      "loss": 3.2288,
+      "step": 36930
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6609817743301392,
+      "learning_rate": 0.00020434212138704716,
+      "loss": 3.136,
+      "step": 36931
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6926374435424805,
+      "learning_rate": 0.0002043275289735519,
+      "loss": 2.8855,
+      "step": 36932
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6676579713821411,
+      "learning_rate": 0.00020431293681203938,
+      "loss": 3.1017,
+      "step": 36933
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.676917314529419,
+      "learning_rate": 0.00020429834490254823,
+      "loss": 3.0384,
+      "step": 36934
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6208324432373047,
+      "learning_rate": 0.0002042837532451169,
+      "loss": 3.0039,
+      "step": 36935
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6595355868339539,
+      "learning_rate": 0.00020426916183978363,
+      "loss": 2.996,
+      "step": 36936
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6129118800163269,
+      "learning_rate": 0.00020425457068658706,
+      "loss": 3.043,
+      "step": 36937
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5993137955665588,
+      "learning_rate": 0.00020423997978556547,
+      "loss": 3.0566,
+      "step": 36938
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.620607316493988,
+      "learning_rate": 0.00020422538913675726,
+      "loss": 2.8563,
+      "step": 36939
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6551635265350342,
+      "learning_rate": 0.00020421079874020112,
+      "loss": 3.1258,
+      "step": 36940
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6164003610610962,
+      "learning_rate": 0.00020419620859593517,
+      "loss": 3.1104,
+      "step": 36941
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6258066296577454,
+      "learning_rate": 0.00020418161870399808,
+      "loss": 3.1211,
+      "step": 36942
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6458246111869812,
+      "learning_rate": 0.00020416702906442805,
+      "loss": 3.0477,
+      "step": 36943
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6024858951568604,
+      "learning_rate": 0.00020415243967726362,
+      "loss": 2.9224,
+      "step": 36944
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6260032653808594,
+      "learning_rate": 0.00020413785054254337,
+      "loss": 2.8545,
+      "step": 36945
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6995924711227417,
+      "learning_rate": 0.00020412326166030539,
+      "loss": 2.8598,
+      "step": 36946
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.634792149066925,
+      "learning_rate": 0.0002041086730305884,
+      "loss": 2.72,
+      "step": 36947
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.657988965511322,
+      "learning_rate": 0.00020409408465343066,
+      "loss": 2.8743,
+      "step": 36948
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6405299305915833,
+      "learning_rate": 0.00020407949652887064,
+      "loss": 2.9527,
+      "step": 36949
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6475312113761902,
+      "learning_rate": 0.0002040649086569468,
+      "loss": 3.0618,
+      "step": 36950
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6201655864715576,
+      "learning_rate": 0.00020405032103769746,
+      "loss": 3.1179,
+      "step": 36951
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6073014140129089,
+      "learning_rate": 0.0002040357336711612,
+      "loss": 2.9965,
+      "step": 36952
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6534898281097412,
+      "learning_rate": 0.00020402114655737635,
+      "loss": 2.9554,
+      "step": 36953
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.630856990814209,
+      "learning_rate": 0.0002040065596963812,
+      "loss": 2.8513,
+      "step": 36954
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6377066969871521,
+      "learning_rate": 0.00020399197308821444,
+      "loss": 3.083,
+      "step": 36955
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6434876322746277,
+      "learning_rate": 0.00020397738673291432,
+      "loss": 2.7419,
+      "step": 36956
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6340557932853699,
+      "learning_rate": 0.0002039628006305193,
+      "loss": 2.9313,
+      "step": 36957
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6318641304969788,
+      "learning_rate": 0.00020394821478106768,
+      "loss": 2.8285,
+      "step": 36958
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6800630688667297,
+      "learning_rate": 0.000203933629184598,
+      "loss": 3.031,
+      "step": 36959
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6454338431358337,
+      "learning_rate": 0.00020391904384114875,
+      "loss": 2.8901,
+      "step": 36960
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7657660245895386,
+      "learning_rate": 0.00020390445875075812,
+      "loss": 3.1257,
+      "step": 36961
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6502024531364441,
+      "learning_rate": 0.00020388987391346477,
+      "loss": 3.2133,
+      "step": 36962
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6643862128257751,
+      "learning_rate": 0.00020387528932930691,
+      "loss": 3.0175,
+      "step": 36963
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6402032375335693,
+      "learning_rate": 0.00020386070499832303,
+      "loss": 2.9454,
+      "step": 36964
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6753803491592407,
+      "learning_rate": 0.00020384612092055166,
+      "loss": 3.1392,
+      "step": 36965
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6361395716667175,
+      "learning_rate": 0.000203831537096031,
+      "loss": 2.8411,
+      "step": 36966
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6042212843894958,
+      "learning_rate": 0.00020381695352479967,
+      "loss": 2.9097,
+      "step": 36967
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.724984347820282,
+      "learning_rate": 0.00020380237020689582,
+      "loss": 3.0332,
+      "step": 36968
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6241597533226013,
+      "learning_rate": 0.00020378778714235803,
+      "loss": 2.9793,
+      "step": 36969
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6064009070396423,
+      "learning_rate": 0.00020377320433122482,
+      "loss": 3.2852,
+      "step": 36970
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6465131044387817,
+      "learning_rate": 0.00020375862177353436,
+      "loss": 3.1043,
+      "step": 36971
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6775936484336853,
+      "learning_rate": 0.0002037440394693253,
+      "loss": 3.1991,
+      "step": 36972
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6569826602935791,
+      "learning_rate": 0.00020372945741863575,
+      "loss": 2.8564,
+      "step": 36973
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6317790746688843,
+      "learning_rate": 0.00020371487562150436,
+      "loss": 3.0964,
+      "step": 36974
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.642630398273468,
+      "learning_rate": 0.00020370029407796951,
+      "loss": 3.1208,
+      "step": 36975
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6569013595581055,
+      "learning_rate": 0.00020368571278806943,
+      "loss": 3.1136,
+      "step": 36976
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6564520597457886,
+      "learning_rate": 0.00020367113175184278,
+      "loss": 2.9458,
+      "step": 36977
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6364352703094482,
+      "learning_rate": 0.00020365655096932773,
+      "loss": 3.0549,
+      "step": 36978
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6397393345832825,
+      "learning_rate": 0.00020364197044056277,
+      "loss": 2.7911,
+      "step": 36979
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6152938008308411,
+      "learning_rate": 0.00020362739016558642,
+      "loss": 2.8501,
+      "step": 36980
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6240938305854797,
+      "learning_rate": 0.00020361281014443693,
+      "loss": 2.9638,
+      "step": 36981
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6312567591667175,
+      "learning_rate": 0.0002035982303771528,
+      "loss": 2.9071,
+      "step": 36982
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6281710267066956,
+      "learning_rate": 0.00020358365086377224,
+      "loss": 2.9096,
+      "step": 36983
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6307974457740784,
+      "learning_rate": 0.00020356907160433386,
+      "loss": 2.7694,
+      "step": 36984
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6692641377449036,
+      "learning_rate": 0.00020355449259887607,
+      "loss": 3.0278,
+      "step": 36985
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6274914145469666,
+      "learning_rate": 0.000203539913847437,
+      "loss": 3.0309,
+      "step": 36986
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6962708234786987,
+      "learning_rate": 0.00020352533535005545,
+      "loss": 2.827,
+      "step": 36987
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6521985530853271,
+      "learning_rate": 0.0002035107571067694,
+      "loss": 3.0369,
+      "step": 36988
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6316908597946167,
+      "learning_rate": 0.00020349617911761753,
+      "loss": 2.8481,
+      "step": 36989
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6510263085365295,
+      "learning_rate": 0.00020348160138263817,
+      "loss": 3.0658,
+      "step": 36990
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6218271255493164,
+      "learning_rate": 0.00020346702390186964,
+      "loss": 2.7787,
+      "step": 36991
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6902745366096497,
+      "learning_rate": 0.0002034524466753505,
+      "loss": 2.9119,
+      "step": 36992
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6275454163551331,
+      "learning_rate": 0.00020343786970311895,
+      "loss": 2.893,
+      "step": 36993
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6849902868270874,
+      "learning_rate": 0.00020342329298521345,
+      "loss": 2.9944,
+      "step": 36994
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7208778262138367,
+      "learning_rate": 0.00020340871652167248,
+      "loss": 2.9227,
+      "step": 36995
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6108225584030151,
+      "learning_rate": 0.00020339414031253432,
+      "loss": 3.1913,
+      "step": 36996
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6657624840736389,
+      "learning_rate": 0.00020337956435783749,
+      "loss": 2.6087,
+      "step": 36997
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6316897869110107,
+      "learning_rate": 0.00020336498865762014,
+      "loss": 3.0783,
+      "step": 36998
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6183792352676392,
+      "learning_rate": 0.00020335041321192088,
+      "loss": 2.9349,
+      "step": 36999
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6724928617477417,
+      "learning_rate": 0.00020333583802077811,
+      "loss": 2.9243,
+      "step": 37000
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5862394571304321,
+      "learning_rate": 0.00020332126308422998,
+      "loss": 3.1631,
+      "step": 37001
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6738343834877014,
+      "learning_rate": 0.00020330668840231515,
+      "loss": 2.9031,
+      "step": 37002
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.637536346912384,
+      "learning_rate": 0.0002032921139750719,
+      "loss": 2.9616,
+      "step": 37003
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6663991212844849,
+      "learning_rate": 0.00020327753980253863,
+      "loss": 3.0612,
+      "step": 37004
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6609890460968018,
+      "learning_rate": 0.00020326296588475356,
+      "loss": 3.0707,
+      "step": 37005
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.645164966583252,
+      "learning_rate": 0.0002032483922217553,
+      "loss": 2.9354,
+      "step": 37006
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.675213098526001,
+      "learning_rate": 0.00020323381881358224,
+      "loss": 3.0725,
+      "step": 37007
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6274236440658569,
+      "learning_rate": 0.00020321924566027254,
+      "loss": 3.0605,
+      "step": 37008
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6301320791244507,
+      "learning_rate": 0.00020320467276186484,
+      "loss": 3.1105,
+      "step": 37009
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6352813839912415,
+      "learning_rate": 0.00020319010011839728,
+      "loss": 3.2368,
+      "step": 37010
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6481574773788452,
+      "learning_rate": 0.00020317552772990843,
+      "loss": 2.9269,
+      "step": 37011
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6353967189788818,
+      "learning_rate": 0.00020316095559643667,
+      "loss": 3.0593,
+      "step": 37012
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6281538009643555,
+      "learning_rate": 0.0002031463837180202,
+      "loss": 2.9999,
+      "step": 37013
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6500021815299988,
+      "learning_rate": 0.00020313181209469762,
+      "loss": 3.1799,
+      "step": 37014
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.611717164516449,
+      "learning_rate": 0.00020311724072650717,
+      "loss": 2.9209,
+      "step": 37015
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.657289981842041,
+      "learning_rate": 0.00020310266961348714,
+      "loss": 2.8855,
+      "step": 37016
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6263495683670044,
+      "learning_rate": 0.00020308809875567622,
+      "loss": 2.8046,
+      "step": 37017
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6198089122772217,
+      "learning_rate": 0.00020307352815311248,
+      "loss": 2.8548,
+      "step": 37018
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6605949401855469,
+      "learning_rate": 0.00020305895780583453,
+      "loss": 3.0922,
+      "step": 37019
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6008433103561401,
+      "learning_rate": 0.0002030443877138805,
+      "loss": 2.9261,
+      "step": 37020
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6339818835258484,
+      "learning_rate": 0.0002030298178772889,
+      "loss": 2.9345,
+      "step": 37021
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6539232730865479,
+      "learning_rate": 0.0002030152482960982,
+      "loss": 2.9125,
+      "step": 37022
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5978472232818604,
+      "learning_rate": 0.00020300067897034655,
+      "loss": 3.0355,
+      "step": 37023
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6346756219863892,
+      "learning_rate": 0.00020298610990007255,
+      "loss": 2.9446,
+      "step": 37024
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6300790905952454,
+      "learning_rate": 0.00020297154108531442,
+      "loss": 2.8722,
+      "step": 37025
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6479358077049255,
+      "learning_rate": 0.0002029569725261105,
+      "loss": 3.0601,
+      "step": 37026
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6204195618629456,
+      "learning_rate": 0.00020294240422249937,
+      "loss": 2.9397,
+      "step": 37027
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7228114008903503,
+      "learning_rate": 0.0002029278361745192,
+      "loss": 2.7625,
+      "step": 37028
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6030262112617493,
+      "learning_rate": 0.0002029132683822085,
+      "loss": 2.9429,
+      "step": 37029
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.623123049736023,
+      "learning_rate": 0.00020289870084560542,
+      "loss": 2.9507,
+      "step": 37030
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6438615322113037,
+      "learning_rate": 0.0002028841335647485,
+      "loss": 3.0401,
+      "step": 37031
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7645979523658752,
+      "learning_rate": 0.0002028695665396762,
+      "loss": 3.062,
+      "step": 37032
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6688880920410156,
+      "learning_rate": 0.0002028549997704267,
+      "loss": 2.9529,
+      "step": 37033
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6559062004089355,
+      "learning_rate": 0.0002028404332570385,
+      "loss": 3.097,
+      "step": 37034
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6532920002937317,
+      "learning_rate": 0.00020282586699954977,
+      "loss": 2.7324,
+      "step": 37035
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6466086506843567,
+      "learning_rate": 0.0002028113009979991,
+      "loss": 3.0552,
+      "step": 37036
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6443438529968262,
+      "learning_rate": 0.00020279673525242477,
+      "loss": 2.9941,
+      "step": 37037
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6671386957168579,
+      "learning_rate": 0.00020278216976286504,
+      "loss": 2.95,
+      "step": 37038
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6276841163635254,
+      "learning_rate": 0.00020276760452935848,
+      "loss": 3.0235,
+      "step": 37039
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.650061845779419,
+      "learning_rate": 0.0002027530395519433,
+      "loss": 3.0023,
+      "step": 37040
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6341570019721985,
+      "learning_rate": 0.0002027384748306578,
+      "loss": 2.9555,
+      "step": 37041
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6999353766441345,
+      "learning_rate": 0.00020272391036554058,
+      "loss": 3.009,
+      "step": 37042
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.630683183670044,
+      "learning_rate": 0.00020270934615662978,
+      "loss": 3.1733,
+      "step": 37043
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6502330303192139,
+      "learning_rate": 0.00020269478220396393,
+      "loss": 2.8276,
+      "step": 37044
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.599058210849762,
+      "learning_rate": 0.00020268021850758115,
+      "loss": 2.8763,
+      "step": 37045
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5828586220741272,
+      "learning_rate": 0.00020266565506752002,
+      "loss": 2.9274,
+      "step": 37046
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6535758376121521,
+      "learning_rate": 0.00020265109188381887,
+      "loss": 3.153,
+      "step": 37047
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.624528169631958,
+      "learning_rate": 0.00020263652895651587,
+      "loss": 2.9203,
+      "step": 37048
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6232485175132751,
+      "learning_rate": 0.0002026219662856497,
+      "loss": 2.9709,
+      "step": 37049
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6133432388305664,
+      "learning_rate": 0.00020260740387125834,
+      "loss": 2.9712,
+      "step": 37050
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6627100706100464,
+      "learning_rate": 0.00020259284171338042,
+      "loss": 2.9813,
+      "step": 37051
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6731305122375488,
+      "learning_rate": 0.0002025782798120543,
+      "loss": 3.0706,
+      "step": 37052
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6375182867050171,
+      "learning_rate": 0.0002025637181673181,
+      "loss": 3.0896,
+      "step": 37053
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.684897243976593,
+      "learning_rate": 0.0002025491567792104,
+      "loss": 2.8677,
+      "step": 37054
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6456209421157837,
+      "learning_rate": 0.00020253459564776945,
+      "loss": 3.1135,
+      "step": 37055
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6744033098220825,
+      "learning_rate": 0.00020252003477303353,
+      "loss": 3.0566,
+      "step": 37056
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6282669305801392,
+      "learning_rate": 0.0002025054741550412,
+      "loss": 3.0302,
+      "step": 37057
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6827451586723328,
+      "learning_rate": 0.00020249091379383066,
+      "loss": 2.8749,
+      "step": 37058
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.631802499294281,
+      "learning_rate": 0.0002024763536894403,
+      "loss": 3.0074,
+      "step": 37059
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6587291359901428,
+      "learning_rate": 0.00020246179384190836,
+      "loss": 2.7468,
+      "step": 37060
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.655880868434906,
+      "learning_rate": 0.00020244723425127335,
+      "loss": 3.1387,
+      "step": 37061
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6162218451499939,
+      "learning_rate": 0.0002024326749175736,
+      "loss": 3.1554,
+      "step": 37062
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6140300631523132,
+      "learning_rate": 0.0002024181158408473,
+      "loss": 3.0222,
+      "step": 37063
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6147719025611877,
+      "learning_rate": 0.000202403557021133,
+      "loss": 2.9423,
+      "step": 37064
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6162436604499817,
+      "learning_rate": 0.00020238899845846889,
+      "loss": 2.9565,
+      "step": 37065
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6531295776367188,
+      "learning_rate": 0.00020237444015289336,
+      "loss": 3.1544,
+      "step": 37066
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6456297039985657,
+      "learning_rate": 0.00020235988210444482,
+      "loss": 2.9994,
+      "step": 37067
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6418159008026123,
+      "learning_rate": 0.00020234532431316157,
+      "loss": 2.8051,
+      "step": 37068
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6710839867591858,
+      "learning_rate": 0.00020233076677908196,
+      "loss": 2.9623,
+      "step": 37069
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6115990877151489,
+      "learning_rate": 0.00020231620950224419,
+      "loss": 2.956,
+      "step": 37070
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.601249098777771,
+      "learning_rate": 0.0002023016524826867,
+      "loss": 3.0491,
+      "step": 37071
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6476225256919861,
+      "learning_rate": 0.00020228709572044803,
+      "loss": 3.0574,
+      "step": 37072
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6286438703536987,
+      "learning_rate": 0.00020227253921556624,
+      "loss": 2.8102,
+      "step": 37073
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6054357886314392,
+      "learning_rate": 0.00020225798296807986,
+      "loss": 3.1466,
+      "step": 37074
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.681904673576355,
+      "learning_rate": 0.00020224342697802705,
+      "loss": 3.0506,
+      "step": 37075
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6787092089653015,
+      "learning_rate": 0.00020222887124544625,
+      "loss": 3.0671,
+      "step": 37076
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6227284669876099,
+      "learning_rate": 0.00020221431577037584,
+      "loss": 3.0675,
+      "step": 37077
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6418759226799011,
+      "learning_rate": 0.000202199760552854,
+      "loss": 3.0101,
+      "step": 37078
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6091601252555847,
+      "learning_rate": 0.00020218520559291928,
+      "loss": 2.8531,
+      "step": 37079
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6450102925300598,
+      "learning_rate": 0.0002021706508906099,
+      "loss": 3.0782,
+      "step": 37080
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6219157576560974,
+      "learning_rate": 0.00020215609644596422,
+      "loss": 3.0429,
+      "step": 37081
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6502217650413513,
+      "learning_rate": 0.0002021415422590204,
+      "loss": 2.9689,
+      "step": 37082
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6541635990142822,
+      "learning_rate": 0.000202126988329817,
+      "loss": 2.9687,
+      "step": 37083
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6724274754524231,
+      "learning_rate": 0.00020211243465839237,
+      "loss": 2.9509,
+      "step": 37084
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6167418360710144,
+      "learning_rate": 0.0002020978812447846,
+      "loss": 3.0759,
+      "step": 37085
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6464261412620544,
+      "learning_rate": 0.0002020833280890323,
+      "loss": 2.7714,
+      "step": 37086
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6340137720108032,
+      "learning_rate": 0.0002020687751911736,
+      "loss": 3.0417,
+      "step": 37087
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.631371259689331,
+      "learning_rate": 0.00020205422255124684,
+      "loss": 2.8924,
+      "step": 37088
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6182030439376831,
+      "learning_rate": 0.00020203967016929053,
+      "loss": 3.1116,
+      "step": 37089
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6417081356048584,
+      "learning_rate": 0.00020202511804534278,
+      "loss": 3.0076,
+      "step": 37090
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6194267868995667,
+      "learning_rate": 0.00020201056617944217,
+      "loss": 3.0782,
+      "step": 37091
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6341774463653564,
+      "learning_rate": 0.00020199601457162667,
+      "loss": 2.9582,
+      "step": 37092
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6601331233978271,
+      "learning_rate": 0.00020198146322193483,
+      "loss": 3.0365,
+      "step": 37093
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6085912585258484,
+      "learning_rate": 0.00020196691213040507,
+      "loss": 3.1794,
+      "step": 37094
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.621552586555481,
+      "learning_rate": 0.00020195236129707552,
+      "loss": 3.1996,
+      "step": 37095
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6433894634246826,
+      "learning_rate": 0.00020193781072198468,
+      "loss": 2.7632,
+      "step": 37096
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.633199155330658,
+      "learning_rate": 0.00020192326040517064,
+      "loss": 3.1733,
+      "step": 37097
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6254742741584778,
+      "learning_rate": 0.0002019087103466719,
+      "loss": 3.0559,
+      "step": 37098
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6049128770828247,
+      "learning_rate": 0.0002018941605465268,
+      "loss": 3.0833,
+      "step": 37099
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6894818544387817,
+      "learning_rate": 0.0002018796110047735,
+      "loss": 2.7616,
+      "step": 37100
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.596031665802002,
+      "learning_rate": 0.0002018650617214505,
+      "loss": 2.9947,
+      "step": 37101
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6057289838790894,
+      "learning_rate": 0.000201850512696596,
+      "loss": 3.1525,
+      "step": 37102
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6352101564407349,
+      "learning_rate": 0.00020183596393024832,
+      "loss": 3.0793,
+      "step": 37103
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6023186445236206,
+      "learning_rate": 0.00020182141542244593,
+      "loss": 3.0214,
+      "step": 37104
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6341410875320435,
+      "learning_rate": 0.000201806867173227,
+      "loss": 2.868,
+      "step": 37105
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6453239917755127,
+      "learning_rate": 0.00020179231918262988,
+      "loss": 3.0226,
+      "step": 37106
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6963949203491211,
+      "learning_rate": 0.00020177777145069284,
+      "loss": 3.0891,
+      "step": 37107
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6552883386611938,
+      "learning_rate": 0.00020176322397745427,
+      "loss": 3.1864,
+      "step": 37108
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6849330067634583,
+      "learning_rate": 0.00020174867676295256,
+      "loss": 3.0045,
+      "step": 37109
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7277083992958069,
+      "learning_rate": 0.00020173412980722575,
+      "loss": 3.0083,
+      "step": 37110
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6667925119400024,
+      "learning_rate": 0.00020171958311031245,
+      "loss": 2.9899,
+      "step": 37111
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6309471726417542,
+      "learning_rate": 0.00020170503667225076,
+      "loss": 3.0762,
+      "step": 37112
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6317311525344849,
+      "learning_rate": 0.00020169049049307918,
+      "loss": 2.9453,
+      "step": 37113
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6759694814682007,
+      "learning_rate": 0.00020167594457283595,
+      "loss": 3.2062,
+      "step": 37114
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6160515546798706,
+      "learning_rate": 0.0002016613989115592,
+      "loss": 3.099,
+      "step": 37115
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6787965297698975,
+      "learning_rate": 0.00020164685350928757,
+      "loss": 2.8703,
+      "step": 37116
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6101091504096985,
+      "learning_rate": 0.00020163230836605912,
+      "loss": 3.2368,
+      "step": 37117
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6226139068603516,
+      "learning_rate": 0.00020161776348191223,
+      "loss": 3.125,
+      "step": 37118
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7263720035552979,
+      "learning_rate": 0.00020160321885688527,
+      "loss": 3.0666,
+      "step": 37119
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6440698504447937,
+      "learning_rate": 0.00020158867449101645,
+      "loss": 2.9552,
+      "step": 37120
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6304377317428589,
+      "learning_rate": 0.0002015741303843442,
+      "loss": 3.2824,
+      "step": 37121
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6487731337547302,
+      "learning_rate": 0.00020155958653690664,
+      "loss": 2.8647,
+      "step": 37122
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6729735732078552,
+      "learning_rate": 0.00020154504294874224,
+      "loss": 2.879,
+      "step": 37123
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6333503127098083,
+      "learning_rate": 0.0002015304996198893,
+      "loss": 3.0657,
+      "step": 37124
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6147133708000183,
+      "learning_rate": 0.0002015159565503859,
+      "loss": 3.0688,
+      "step": 37125
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6360808610916138,
+      "learning_rate": 0.0002015014137402707,
+      "loss": 3.21,
+      "step": 37126
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6403216123580933,
+      "learning_rate": 0.00020148687118958175,
+      "loss": 3.0954,
+      "step": 37127
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6334720849990845,
+      "learning_rate": 0.00020147232889835735,
+      "loss": 3.0747,
+      "step": 37128
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.688264787197113,
+      "learning_rate": 0.000201457786866636,
+      "loss": 2.8259,
+      "step": 37129
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6472107172012329,
+      "learning_rate": 0.00020144324509445584,
+      "loss": 3.025,
+      "step": 37130
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6436012983322144,
+      "learning_rate": 0.0002014287035818552,
+      "loss": 2.9706,
+      "step": 37131
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6599947810173035,
+      "learning_rate": 0.00020141416232887232,
+      "loss": 2.8894,
+      "step": 37132
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.65586918592453,
+      "learning_rate": 0.00020139962133554552,
+      "loss": 3.006,
+      "step": 37133
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7479538321495056,
+      "learning_rate": 0.0002013850806019133,
+      "loss": 2.8773,
+      "step": 37134
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6416367888450623,
+      "learning_rate": 0.00020137054012801377,
+      "loss": 3.0427,
+      "step": 37135
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5910057425498962,
+      "learning_rate": 0.00020135599991388524,
+      "loss": 2.9003,
+      "step": 37136
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.604770839214325,
+      "learning_rate": 0.00020134145995956597,
+      "loss": 2.9453,
+      "step": 37137
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6588855385780334,
+      "learning_rate": 0.0002013269202650943,
+      "loss": 2.9322,
+      "step": 37138
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6265450716018677,
+      "learning_rate": 0.00020131238083050863,
+      "loss": 2.8692,
+      "step": 37139
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.607767641544342,
+      "learning_rate": 0.00020129784165584703,
+      "loss": 3.0773,
+      "step": 37140
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.660254716873169,
+      "learning_rate": 0.00020128330274114803,
+      "loss": 2.9246,
+      "step": 37141
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6739830374717712,
+      "learning_rate": 0.00020126876408644976,
+      "loss": 2.7952,
+      "step": 37142
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6582867503166199,
+      "learning_rate": 0.00020125422569179053,
+      "loss": 2.9389,
+      "step": 37143
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6650797724723816,
+      "learning_rate": 0.00020123968755720876,
+      "loss": 2.9109,
+      "step": 37144
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7328686118125916,
+      "learning_rate": 0.00020122514968274257,
+      "loss": 3.0024,
+      "step": 37145
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6152870059013367,
+      "learning_rate": 0.00020121061206843042,
+      "loss": 2.9541,
+      "step": 37146
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.603813648223877,
+      "learning_rate": 0.00020119607471431034,
+      "loss": 2.8763,
+      "step": 37147
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.646477222442627,
+      "learning_rate": 0.00020118153762042088,
+      "loss": 3.0344,
+      "step": 37148
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7055385112762451,
+      "learning_rate": 0.0002011670007868003,
+      "loss": 3.0673,
+      "step": 37149
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6229759454727173,
+      "learning_rate": 0.00020115246421348668,
+      "loss": 3.1857,
+      "step": 37150
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6314061284065247,
+      "learning_rate": 0.00020113792790051854,
+      "loss": 3.0096,
+      "step": 37151
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6126551032066345,
+      "learning_rate": 0.000201123391847934,
+      "loss": 2.9149,
+      "step": 37152
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6453806161880493,
+      "learning_rate": 0.00020110885605577146,
+      "loss": 3.0979,
+      "step": 37153
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6280014514923096,
+      "learning_rate": 0.0002010943205240692,
+      "loss": 2.8085,
+      "step": 37154
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6255245208740234,
+      "learning_rate": 0.00020107978525286537,
+      "loss": 3.2936,
+      "step": 37155
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6780776977539062,
+      "learning_rate": 0.0002010652502421985,
+      "loss": 2.9829,
+      "step": 37156
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7052608132362366,
+      "learning_rate": 0.0002010507154921066,
+      "loss": 2.8949,
+      "step": 37157
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6485824584960938,
+      "learning_rate": 0.00020103618100262815,
+      "loss": 2.9753,
+      "step": 37158
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6526051163673401,
+      "learning_rate": 0.00020102164677380122,
+      "loss": 3.1261,
+      "step": 37159
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6495037078857422,
+      "learning_rate": 0.00020100711280566431,
+      "loss": 2.9435,
+      "step": 37160
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5939731597900391,
+      "learning_rate": 0.00020099257909825567,
+      "loss": 3.1176,
+      "step": 37161
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6784581542015076,
+      "learning_rate": 0.00020097804565161342,
+      "loss": 3.1203,
+      "step": 37162
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6253423690795898,
+      "learning_rate": 0.000200963512465776,
+      "loss": 3.0622,
+      "step": 37163
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6293935179710388,
+      "learning_rate": 0.00020094897954078163,
+      "loss": 3.1353,
+      "step": 37164
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7159479856491089,
+      "learning_rate": 0.0002009344468766685,
+      "loss": 3.2139,
+      "step": 37165
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6202086806297302,
+      "learning_rate": 0.0002009199144734751,
+      "loss": 3.0405,
+      "step": 37166
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6366106867790222,
+      "learning_rate": 0.00020090538233123948,
+      "loss": 3.0614,
+      "step": 37167
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6750251650810242,
+      "learning_rate": 0.00020089085045000012,
+      "loss": 3.0452,
+      "step": 37168
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6211447715759277,
+      "learning_rate": 0.00020087631882979504,
+      "loss": 3.0842,
+      "step": 37169
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6459939479827881,
+      "learning_rate": 0.00020086178747066272,
+      "loss": 2.9701,
+      "step": 37170
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6417645812034607,
+      "learning_rate": 0.00020084725637264147,
+      "loss": 2.9645,
+      "step": 37171
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.610141396522522,
+      "learning_rate": 0.00020083272553576929,
+      "loss": 2.9528,
+      "step": 37172
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6073371171951294,
+      "learning_rate": 0.00020081819496008477,
+      "loss": 3.0393,
+      "step": 37173
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6643429398536682,
+      "learning_rate": 0.00020080366464562594,
+      "loss": 2.9228,
+      "step": 37174
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6577807664871216,
+      "learning_rate": 0.00020078913459243119,
+      "loss": 3.1136,
+      "step": 37175
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5988003015518188,
+      "learning_rate": 0.00020077460480053883,
+      "loss": 2.9344,
+      "step": 37176
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6243207454681396,
+      "learning_rate": 0.00020076007526998696,
+      "loss": 2.9445,
+      "step": 37177
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6623547077178955,
+      "learning_rate": 0.00020074554600081406,
+      "loss": 2.9545,
+      "step": 37178
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6492032408714294,
+      "learning_rate": 0.00020073101699305822,
+      "loss": 2.8503,
+      "step": 37179
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6891651153564453,
+      "learning_rate": 0.00020071648824675773,
+      "loss": 3.1972,
+      "step": 37180
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6541674733161926,
+      "learning_rate": 0.00020070195976195105,
+      "loss": 2.9552,
+      "step": 37181
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6345112323760986,
+      "learning_rate": 0.0002006874315386762,
+      "loss": 2.9296,
+      "step": 37182
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6238939762115479,
+      "learning_rate": 0.00020067290357697162,
+      "loss": 2.9192,
+      "step": 37183
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6258445978164673,
+      "learning_rate": 0.00020065837587687535,
+      "loss": 2.798,
+      "step": 37184
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6790409088134766,
+      "learning_rate": 0.00020064384843842588,
+      "loss": 3.1312,
+      "step": 37185
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6279799342155457,
+      "learning_rate": 0.00020062932126166147,
+      "loss": 2.8466,
+      "step": 37186
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6571778059005737,
+      "learning_rate": 0.00020061479434662017,
+      "loss": 3.1089,
+      "step": 37187
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6268858909606934,
+      "learning_rate": 0.0002006002676933405,
+      "loss": 2.922,
+      "step": 37188
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6086148023605347,
+      "learning_rate": 0.0002005857413018605,
+      "loss": 2.881,
+      "step": 37189
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6448004245758057,
+      "learning_rate": 0.0002005712151722185,
+      "loss": 2.9411,
+      "step": 37190
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6258628964424133,
+      "learning_rate": 0.0002005566893044529,
+      "loss": 3.1739,
+      "step": 37191
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6352370381355286,
+      "learning_rate": 0.00020054216369860177,
+      "loss": 2.9686,
+      "step": 37192
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6704267859458923,
+      "learning_rate": 0.0002005276383547035,
+      "loss": 3.0863,
+      "step": 37193
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6442691087722778,
+      "learning_rate": 0.00020051311327279618,
+      "loss": 2.9182,
+      "step": 37194
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6933001279830933,
+      "learning_rate": 0.00020049858845291814,
+      "loss": 3.0166,
+      "step": 37195
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6700044870376587,
+      "learning_rate": 0.00020048406389510786,
+      "loss": 2.9624,
+      "step": 37196
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6275744438171387,
+      "learning_rate": 0.00020046953959940324,
+      "loss": 2.9598,
+      "step": 37197
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6295133829116821,
+      "learning_rate": 0.00020045501556584285,
+      "loss": 2.9946,
+      "step": 37198
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6550989747047424,
+      "learning_rate": 0.00020044049179446463,
+      "loss": 2.8582,
+      "step": 37199
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6450958251953125,
+      "learning_rate": 0.00020042596828530706,
+      "loss": 2.9827,
+      "step": 37200
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6483104228973389,
+      "learning_rate": 0.00020041144503840836,
+      "loss": 2.8909,
+      "step": 37201
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6364044547080994,
+      "learning_rate": 0.00020039692205380666,
+      "loss": 2.9006,
+      "step": 37202
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6128686666488647,
+      "learning_rate": 0.00020038239933154042,
+      "loss": 2.91,
+      "step": 37203
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6625557541847229,
+      "learning_rate": 0.0002003678768716477,
+      "loss": 2.8524,
+      "step": 37204
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6292389631271362,
+      "learning_rate": 0.00020035335467416676,
+      "loss": 3.1534,
+      "step": 37205
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6541522741317749,
+      "learning_rate": 0.000200338832739136,
+      "loss": 2.8124,
+      "step": 37206
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6432307958602905,
+      "learning_rate": 0.00020032431106659353,
+      "loss": 2.9206,
+      "step": 37207
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6428229212760925,
+      "learning_rate": 0.00020030978965657775,
+      "loss": 2.7471,
+      "step": 37208
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6233975887298584,
+      "learning_rate": 0.00020029526850912663,
+      "loss": 2.965,
+      "step": 37209
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6788715720176697,
+      "learning_rate": 0.00020028074762427865,
+      "loss": 3.0143,
+      "step": 37210
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6513542532920837,
+      "learning_rate": 0.00020026622700207206,
+      "loss": 3.0382,
+      "step": 37211
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6843482255935669,
+      "learning_rate": 0.00020025170664254488,
+      "loss": 2.7828,
+      "step": 37212
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5919236540794373,
+      "learning_rate": 0.00020023718654573567,
+      "loss": 2.951,
+      "step": 37213
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6735742092132568,
+      "learning_rate": 0.0002002226667116824,
+      "loss": 3.0408,
+      "step": 37214
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6410362720489502,
+      "learning_rate": 0.0002002081471404234,
+      "loss": 3.0412,
+      "step": 37215
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.623496949672699,
+      "learning_rate": 0.0002001936278319971,
+      "loss": 3.1536,
+      "step": 37216
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5960062742233276,
+      "learning_rate": 0.0002001791087864414,
+      "loss": 3.2316,
+      "step": 37217
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6214715242385864,
+      "learning_rate": 0.00020016459000379483,
+      "loss": 2.8934,
+      "step": 37218
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6391525268554688,
+      "learning_rate": 0.00020015007148409548,
+      "loss": 2.8583,
+      "step": 37219
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.616982102394104,
+      "learning_rate": 0.00020013555322738157,
+      "loss": 3.1318,
+      "step": 37220
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.665523111820221,
+      "learning_rate": 0.0002001210352336915,
+      "loss": 2.9528,
+      "step": 37221
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6505708694458008,
+      "learning_rate": 0.00020010651750306332,
+      "loss": 2.9226,
+      "step": 37222
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6481497883796692,
+      "learning_rate": 0.00020009200003553545,
+      "loss": 3.2122,
+      "step": 37223
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6302831768989563,
+      "learning_rate": 0.00020007748283114586,
+      "loss": 3.0263,
+      "step": 37224
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6752156019210815,
+      "learning_rate": 0.00020006296588993304,
+      "loss": 3.0486,
+      "step": 37225
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6285132169723511,
+      "learning_rate": 0.0002000484492119352,
+      "loss": 2.983,
+      "step": 37226
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.651867151260376,
+      "learning_rate": 0.0002000339327971904,
+      "loss": 3.0873,
+      "step": 37227
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6278545260429382,
+      "learning_rate": 0.00020001941664573705,
+      "loss": 3.0538,
+      "step": 37228
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6236552000045776,
+      "learning_rate": 0.0002000049007576133,
+      "loss": 2.9181,
+      "step": 37229
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.647704005241394,
+      "learning_rate": 0.00019999038513285732,
+      "loss": 2.9183,
+      "step": 37230
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6166619658470154,
+      "learning_rate": 0.00019997586977150752,
+      "loss": 3.0254,
+      "step": 37231
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.636547863483429,
+      "learning_rate": 0.000199961354673602,
+      "loss": 3.1082,
+      "step": 37232
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6511338353157043,
+      "learning_rate": 0.00019994683983917905,
+      "loss": 2.9792,
+      "step": 37233
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6390637755393982,
+      "learning_rate": 0.0001999323252682768,
+      "loss": 3.0991,
+      "step": 37234
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6388706564903259,
+      "learning_rate": 0.0001999178109609335,
+      "loss": 3.0653,
+      "step": 37235
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.629075825214386,
+      "learning_rate": 0.00019990329691718753,
+      "loss": 2.9239,
+      "step": 37236
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6319878101348877,
+      "learning_rate": 0.00019988878313707698,
+      "loss": 3.2043,
+      "step": 37237
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6719925403594971,
+      "learning_rate": 0.0001998742696206402,
+      "loss": 3.0834,
+      "step": 37238
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6301445960998535,
+      "learning_rate": 0.00019985975636791517,
+      "loss": 3.2402,
+      "step": 37239
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6451971530914307,
+      "learning_rate": 0.00019984524337894038,
+      "loss": 3.2114,
+      "step": 37240
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6485806107521057,
+      "learning_rate": 0.00019983073065375393,
+      "loss": 2.8606,
+      "step": 37241
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6495173573493958,
+      "learning_rate": 0.00019981621819239396,
+      "loss": 2.9877,
+      "step": 37242
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6307535171508789,
+      "learning_rate": 0.0001998017059948989,
+      "loss": 2.8788,
+      "step": 37243
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6974294781684875,
+      "learning_rate": 0.0001997871940613068,
+      "loss": 3.1262,
+      "step": 37244
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6444449424743652,
+      "learning_rate": 0.00019977268239165604,
+      "loss": 3.0831,
+      "step": 37245
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6444473266601562,
+      "learning_rate": 0.0001997581709859846,
+      "loss": 2.886,
+      "step": 37246
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6505341529846191,
+      "learning_rate": 0.0001997436598443309,
+      "loss": 2.8883,
+      "step": 37247
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6732316017150879,
+      "learning_rate": 0.00019972914896673318,
+      "loss": 3.0509,
+      "step": 37248
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6295874714851379,
+      "learning_rate": 0.00019971463835322947,
+      "loss": 3.2458,
+      "step": 37249
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6318504214286804,
+      "learning_rate": 0.00019970012800385822,
+      "loss": 2.9675,
+      "step": 37250
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6528605818748474,
+      "learning_rate": 0.00019968561791865744,
+      "loss": 2.8613,
+      "step": 37251
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.649989902973175,
+      "learning_rate": 0.0001996711080976654,
+      "loss": 3.1373,
+      "step": 37252
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6412109136581421,
+      "learning_rate": 0.00019965659854092048,
+      "loss": 2.7814,
+      "step": 37253
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6689494848251343,
+      "learning_rate": 0.0001996420892484606,
+      "loss": 2.877,
+      "step": 37254
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6729632616043091,
+      "learning_rate": 0.00019962758022032441,
+      "loss": 3.1106,
+      "step": 37255
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6817084550857544,
+      "learning_rate": 0.00019961307145654962,
+      "loss": 3.0034,
+      "step": 37256
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6519462466239929,
+      "learning_rate": 0.0001995985629571747,
+      "loss": 3.2566,
+      "step": 37257
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6344906687736511,
+      "learning_rate": 0.00019958405472223796,
+      "loss": 3.128,
+      "step": 37258
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6581130027770996,
+      "learning_rate": 0.0001995695467517774,
+      "loss": 3.1058,
+      "step": 37259
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6368814706802368,
+      "learning_rate": 0.00019955503904583142,
+      "loss": 2.973,
+      "step": 37260
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6296173334121704,
+      "learning_rate": 0.000199540531604438,
+      "loss": 2.9619,
+      "step": 37261
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6157219409942627,
+      "learning_rate": 0.00019952602442763553,
+      "loss": 3.0496,
+      "step": 37262
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.645746111869812,
+      "learning_rate": 0.0001995115175154623,
+      "loss": 2.9254,
+      "step": 37263
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6452582478523254,
+      "learning_rate": 0.00019949701086795625,
+      "loss": 2.8589,
+      "step": 37264
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6056995987892151,
+      "learning_rate": 0.00019948250448515578,
+      "loss": 2.9054,
+      "step": 37265
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.649617612361908,
+      "learning_rate": 0.00019946799836709908,
+      "loss": 2.7798,
+      "step": 37266
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6778903007507324,
+      "learning_rate": 0.0001994534925138242,
+      "loss": 2.7532,
+      "step": 37267
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6305443644523621,
+      "learning_rate": 0.0001994389869253696,
+      "loss": 3.0986,
+      "step": 37268
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6616302132606506,
+      "learning_rate": 0.00019942448160177335,
+      "loss": 3.2042,
+      "step": 37269
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6154059767723083,
+      "learning_rate": 0.00019940997654307367,
+      "loss": 2.8201,
+      "step": 37270
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6408655643463135,
+      "learning_rate": 0.00019939547174930867,
+      "loss": 3.0539,
+      "step": 37271
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6339004635810852,
+      "learning_rate": 0.00019938096722051666,
+      "loss": 2.887,
+      "step": 37272
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6301832795143127,
+      "learning_rate": 0.00019936646295673586,
+      "loss": 3.0111,
+      "step": 37273
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6578487157821655,
+      "learning_rate": 0.00019935195895800435,
+      "loss": 3.0007,
+      "step": 37274
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6201621294021606,
+      "learning_rate": 0.00019933745522436052,
+      "loss": 3.0404,
+      "step": 37275
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.63294517993927,
+      "learning_rate": 0.00019932295175584234,
+      "loss": 2.8971,
+      "step": 37276
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6506541967391968,
+      "learning_rate": 0.00019930844855248816,
+      "loss": 3.0214,
+      "step": 37277
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6443052887916565,
+      "learning_rate": 0.00019929394561433628,
+      "loss": 2.9546,
+      "step": 37278
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.641716480255127,
+      "learning_rate": 0.00019927944294142455,
+      "loss": 3.1024,
+      "step": 37279
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6462839245796204,
+      "learning_rate": 0.00019926494053379155,
+      "loss": 3.033,
+      "step": 37280
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6586825251579285,
+      "learning_rate": 0.00019925043839147527,
+      "loss": 2.9807,
+      "step": 37281
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6649317145347595,
+      "learning_rate": 0.00019923593651451386,
+      "loss": 3.1034,
+      "step": 37282
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6174647212028503,
+      "learning_rate": 0.00019922143490294576,
+      "loss": 2.9981,
+      "step": 37283
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7386031746864319,
+      "learning_rate": 0.00019920693355680893,
+      "loss": 2.9414,
+      "step": 37284
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7577504515647888,
+      "learning_rate": 0.00019919243247614168,
+      "loss": 3.0452,
+      "step": 37285
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.624578058719635,
+      "learning_rate": 0.00019917793166098204,
+      "loss": 3.1585,
+      "step": 37286
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6715158224105835,
+      "learning_rate": 0.0001991634311113684,
+      "loss": 3.2085,
+      "step": 37287
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.672528862953186,
+      "learning_rate": 0.00019914893082733888,
+      "loss": 3.1,
+      "step": 37288
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6003857254981995,
+      "learning_rate": 0.0001991344308089316,
+      "loss": 3.0433,
+      "step": 37289
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6297593116760254,
+      "learning_rate": 0.00019911993105618494,
+      "loss": 2.9949,
+      "step": 37290
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6269044280052185,
+      "learning_rate": 0.00019910543156913688,
+      "loss": 3.0166,
+      "step": 37291
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6183241009712219,
+      "learning_rate": 0.00019909093234782565,
+      "loss": 3.0963,
+      "step": 37292
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6317001581192017,
+      "learning_rate": 0.0001990764333922896,
+      "loss": 3.1374,
+      "step": 37293
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6698792576789856,
+      "learning_rate": 0.00019906193470256678,
+      "loss": 2.7781,
+      "step": 37294
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6696431040763855,
+      "learning_rate": 0.00019904743627869543,
+      "loss": 3.0622,
+      "step": 37295
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6266303062438965,
+      "learning_rate": 0.00019903293812071357,
+      "loss": 3.1569,
+      "step": 37296
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6246880292892456,
+      "learning_rate": 0.0001990184402286595,
+      "loss": 3.3158,
+      "step": 37297
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7014569044113159,
+      "learning_rate": 0.0001990039426025716,
+      "loss": 2.9063,
+      "step": 37298
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5825400352478027,
+      "learning_rate": 0.0001989894452424878,
+      "loss": 2.9569,
+      "step": 37299
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6470881104469299,
+      "learning_rate": 0.00019897494814844643,
+      "loss": 2.9374,
+      "step": 37300
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6545881628990173,
+      "learning_rate": 0.00019896045132048544,
+      "loss": 3.0919,
+      "step": 37301
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.701209306716919,
+      "learning_rate": 0.00019894595475864325,
+      "loss": 3.0953,
+      "step": 37302
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7519078254699707,
+      "learning_rate": 0.00019893145846295806,
+      "loss": 2.9928,
+      "step": 37303
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6551284193992615,
+      "learning_rate": 0.00019891696243346782,
+      "loss": 3.2031,
+      "step": 37304
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6280120611190796,
+      "learning_rate": 0.00019890246667021095,
+      "loss": 3.0596,
+      "step": 37305
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.62892746925354,
+      "learning_rate": 0.0001988879711732255,
+      "loss": 2.8681,
+      "step": 37306
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6775756478309631,
+      "learning_rate": 0.00019887347594254958,
+      "loss": 2.9004,
+      "step": 37307
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6382337808609009,
+      "learning_rate": 0.0001988589809782216,
+      "loss": 2.8997,
+      "step": 37308
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.639171838760376,
+      "learning_rate": 0.00019884448628027955,
+      "loss": 3.1746,
+      "step": 37309
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.643237829208374,
+      "learning_rate": 0.0001988299918487617,
+      "loss": 3.0516,
+      "step": 37310
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6277839541435242,
+      "learning_rate": 0.00019881549768370604,
+      "loss": 2.9276,
+      "step": 37311
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6903034448623657,
+      "learning_rate": 0.000198801003785151,
+      "loss": 2.7113,
+      "step": 37312
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6506561636924744,
+      "learning_rate": 0.00019878651015313464,
+      "loss": 3.142,
+      "step": 37313
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6318259835243225,
+      "learning_rate": 0.00019877201678769504,
+      "loss": 3.018,
+      "step": 37314
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.660740852355957,
+      "learning_rate": 0.0001987575236888706,
+      "loss": 3.2023,
+      "step": 37315
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6094775199890137,
+      "learning_rate": 0.00019874303085669923,
+      "loss": 2.8146,
+      "step": 37316
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.64765465259552,
+      "learning_rate": 0.0001987285382912194,
+      "loss": 3.1413,
+      "step": 37317
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6887152791023254,
+      "learning_rate": 0.00019871404599246893,
+      "loss": 3.0623,
+      "step": 37318
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6461672782897949,
+      "learning_rate": 0.0001986995539604862,
+      "loss": 3.0541,
+      "step": 37319
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6719304323196411,
+      "learning_rate": 0.0001986850621953094,
+      "loss": 3.1318,
+      "step": 37320
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6449258327484131,
+      "learning_rate": 0.00019867057069697665,
+      "loss": 2.7289,
+      "step": 37321
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6388571858406067,
+      "learning_rate": 0.00019865607946552616,
+      "loss": 2.9219,
+      "step": 37322
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6894384622573853,
+      "learning_rate": 0.00019864158850099595,
+      "loss": 3.1242,
+      "step": 37323
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6467133164405823,
+      "learning_rate": 0.00019862709780342432,
+      "loss": 3.1348,
+      "step": 37324
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6446911096572876,
+      "learning_rate": 0.0001986126073728495,
+      "loss": 2.8621,
+      "step": 37325
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6543609499931335,
+      "learning_rate": 0.0001985981172093094,
+      "loss": 2.8432,
+      "step": 37326
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6412947773933411,
+      "learning_rate": 0.00019858362731284247,
+      "loss": 2.9529,
+      "step": 37327
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6755062341690063,
+      "learning_rate": 0.0001985691376834867,
+      "loss": 3.0239,
+      "step": 37328
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6204738616943359,
+      "learning_rate": 0.00019855464832128024,
+      "loss": 3.0975,
+      "step": 37329
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6035327315330505,
+      "learning_rate": 0.00019854015922626141,
+      "loss": 3.0631,
+      "step": 37330
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6611145734786987,
+      "learning_rate": 0.00019852567039846825,
+      "loss": 2.9024,
+      "step": 37331
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6760570406913757,
+      "learning_rate": 0.00019851118183793898,
+      "loss": 2.8843,
+      "step": 37332
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6354165077209473,
+      "learning_rate": 0.00019849669354471162,
+      "loss": 3.0704,
+      "step": 37333
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6146211624145508,
+      "learning_rate": 0.00019848220551882447,
+      "loss": 3.0125,
+      "step": 37334
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.726223349571228,
+      "learning_rate": 0.00019846771776031572,
+      "loss": 2.838,
+      "step": 37335
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6218348145484924,
+      "learning_rate": 0.00019845323026922339,
+      "loss": 3.0697,
+      "step": 37336
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6220173835754395,
+      "learning_rate": 0.00019843874304558578,
+      "loss": 3.224,
+      "step": 37337
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6685629487037659,
+      "learning_rate": 0.00019842425608944085,
+      "loss": 3.0376,
+      "step": 37338
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6236324906349182,
+      "learning_rate": 0.00019840976940082696,
+      "loss": 3.0332,
+      "step": 37339
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6823529601097107,
+      "learning_rate": 0.00019839528297978227,
+      "loss": 2.8418,
+      "step": 37340
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6390722393989563,
+      "learning_rate": 0.0001983807968263447,
+      "loss": 3.0181,
+      "step": 37341
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6559518575668335,
+      "learning_rate": 0.00019836631094055266,
+      "loss": 2.774,
+      "step": 37342
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6333982348442078,
+      "learning_rate": 0.0001983518253224442,
+      "loss": 2.8799,
+      "step": 37343
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6427645683288574,
+      "learning_rate": 0.00019833733997205735,
+      "loss": 2.924,
+      "step": 37344
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6432594656944275,
+      "learning_rate": 0.0001983228548894305,
+      "loss": 2.9778,
+      "step": 37345
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6509485244750977,
+      "learning_rate": 0.00019830837007460162,
+      "loss": 3.065,
+      "step": 37346
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6645312905311584,
+      "learning_rate": 0.000198293885527609,
+      "loss": 3.0997,
+      "step": 37347
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6213983297348022,
+      "learning_rate": 0.00019827940124849061,
+      "loss": 3.0474,
+      "step": 37348
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7174972295761108,
+      "learning_rate": 0.00019826491723728474,
+      "loss": 3.0375,
+      "step": 37349
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6308889389038086,
+      "learning_rate": 0.00019825043349402954,
+      "loss": 2.937,
+      "step": 37350
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6400142312049866,
+      "learning_rate": 0.00019823595001876301,
+      "loss": 3.1328,
+      "step": 37351
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6188294887542725,
+      "learning_rate": 0.00019822146681152354,
+      "loss": 2.9726,
+      "step": 37352
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6347734928131104,
+      "learning_rate": 0.0001982069838723491,
+      "loss": 2.8248,
+      "step": 37353
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6330137252807617,
+      "learning_rate": 0.00019819250120127778,
+      "loss": 3.0705,
+      "step": 37354
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6539512872695923,
+      "learning_rate": 0.00019817801879834794,
+      "loss": 2.9606,
+      "step": 37355
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6931697726249695,
+      "learning_rate": 0.00019816353666359752,
+      "loss": 2.8545,
+      "step": 37356
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6446145176887512,
+      "learning_rate": 0.0001981490547970648,
+      "loss": 3.0497,
+      "step": 37357
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6381685137748718,
+      "learning_rate": 0.00019813457319878782,
+      "loss": 3.0407,
+      "step": 37358
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7098675966262817,
+      "learning_rate": 0.0001981200918688047,
+      "loss": 3.0888,
+      "step": 37359
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.623608410358429,
+      "learning_rate": 0.00019810561080715384,
+      "loss": 2.9005,
+      "step": 37360
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6890565156936646,
+      "learning_rate": 0.00019809113001387305,
+      "loss": 2.7902,
+      "step": 37361
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6344024538993835,
+      "learning_rate": 0.00019807664948900071,
+      "loss": 2.9944,
+      "step": 37362
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6248427033424377,
+      "learning_rate": 0.00019806216923257475,
+      "loss": 3.179,
+      "step": 37363
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6155762672424316,
+      "learning_rate": 0.00019804768924463348,
+      "loss": 2.9478,
+      "step": 37364
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6568313837051392,
+      "learning_rate": 0.00019803320952521501,
+      "loss": 3.0381,
+      "step": 37365
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6843270063400269,
+      "learning_rate": 0.0001980187300743573,
+      "loss": 2.9444,
+      "step": 37366
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6298990845680237,
+      "learning_rate": 0.00019800425089209881,
+      "loss": 3.0637,
+      "step": 37367
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6255677938461304,
+      "learning_rate": 0.00019798977197847743,
+      "loss": 3.1191,
+      "step": 37368
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6226163506507874,
+      "learning_rate": 0.00019797529333353127,
+      "loss": 3.1296,
+      "step": 37369
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6100264191627502,
+      "learning_rate": 0.00019796081495729868,
+      "loss": 3.1796,
+      "step": 37370
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.644378662109375,
+      "learning_rate": 0.00019794633684981758,
+      "loss": 2.7442,
+      "step": 37371
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6284385919570923,
+      "learning_rate": 0.00019793185901112627,
+      "loss": 3.1884,
+      "step": 37372
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.659661054611206,
+      "learning_rate": 0.0001979173814412627,
+      "loss": 3.0421,
+      "step": 37373
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6167201399803162,
+      "learning_rate": 0.00019790290414026514,
+      "loss": 3.1204,
+      "step": 37374
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6492601037025452,
+      "learning_rate": 0.00019788842710817173,
+      "loss": 3.0217,
+      "step": 37375
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8590417504310608,
+      "learning_rate": 0.00019787395034502047,
+      "loss": 2.93,
+      "step": 37376
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6162007451057434,
+      "learning_rate": 0.00019785947385084966,
+      "loss": 2.8552,
+      "step": 37377
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6711397767066956,
+      "learning_rate": 0.0001978449976256972,
+      "loss": 3.0362,
+      "step": 37378
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6198904514312744,
+      "learning_rate": 0.00019783052166960147,
+      "loss": 2.8681,
+      "step": 37379
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6615649461746216,
+      "learning_rate": 0.00019781604598260048,
+      "loss": 3.008,
+      "step": 37380
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6402595639228821,
+      "learning_rate": 0.0001978015705647323,
+      "loss": 2.8823,
+      "step": 37381
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6121137738227844,
+      "learning_rate": 0.0001977870954160352,
+      "loss": 3.1466,
+      "step": 37382
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.63599693775177,
+      "learning_rate": 0.00019777262053654714,
+      "loss": 2.923,
+      "step": 37383
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6528685092926025,
+      "learning_rate": 0.0001977581459263063,
+      "loss": 2.9125,
+      "step": 37384
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6184086799621582,
+      "learning_rate": 0.00019774367158535095,
+      "loss": 3.0302,
+      "step": 37385
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6583097577095032,
+      "learning_rate": 0.00019772919751371903,
+      "loss": 3.0069,
+      "step": 37386
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6171321868896484,
+      "learning_rate": 0.00019771472371144874,
+      "loss": 2.911,
+      "step": 37387
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6588572263717651,
+      "learning_rate": 0.0001977002501785781,
+      "loss": 3.247,
+      "step": 37388
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6455833911895752,
+      "learning_rate": 0.00019768577691514538,
+      "loss": 2.9604,
+      "step": 37389
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6067509651184082,
+      "learning_rate": 0.00019767130392118868,
+      "loss": 2.9472,
+      "step": 37390
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6183087229728699,
+      "learning_rate": 0.00019765683119674595,
+      "loss": 3.1244,
+      "step": 37391
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7128816843032837,
+      "learning_rate": 0.0001976423587418555,
+      "loss": 3.1313,
+      "step": 37392
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6646550893783569,
+      "learning_rate": 0.0001976278865565554,
+      "loss": 2.901,
+      "step": 37393
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6627399325370789,
+      "learning_rate": 0.0001976134146408838,
+      "loss": 3.0871,
+      "step": 37394
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6553120017051697,
+      "learning_rate": 0.00019759894299487858,
+      "loss": 3.2268,
+      "step": 37395
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6263793110847473,
+      "learning_rate": 0.00019758447161857812,
+      "loss": 3.1037,
+      "step": 37396
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.647152304649353,
+      "learning_rate": 0.0001975700005120205,
+      "loss": 2.8602,
+      "step": 37397
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6689240336418152,
+      "learning_rate": 0.00019755552967524372,
+      "loss": 3.1095,
+      "step": 37398
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6836102604866028,
+      "learning_rate": 0.00019754105910828603,
+      "loss": 2.922,
+      "step": 37399
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6442393064498901,
+      "learning_rate": 0.00019752658881118536,
+      "loss": 2.9583,
+      "step": 37400
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6044512391090393,
+      "learning_rate": 0.00019751211878398002,
+      "loss": 3.21,
+      "step": 37401
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.627376139163971,
+      "learning_rate": 0.00019749764902670804,
+      "loss": 3.017,
+      "step": 37402
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6755691766738892,
+      "learning_rate": 0.00019748317953940744,
+      "loss": 3.0251,
+      "step": 37403
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.621547520160675,
+      "learning_rate": 0.00019746871032211657,
+      "loss": 3.0488,
+      "step": 37404
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6141914129257202,
+      "learning_rate": 0.00019745424137487328,
+      "loss": 3.1355,
+      "step": 37405
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6534335017204285,
+      "learning_rate": 0.00019743977269771573,
+      "loss": 3.0441,
+      "step": 37406
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6744629144668579,
+      "learning_rate": 0.0001974253042906822,
+      "loss": 3.0055,
+      "step": 37407
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6473421454429626,
+      "learning_rate": 0.0001974108361538106,
+      "loss": 2.8622,
+      "step": 37408
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6605812311172485,
+      "learning_rate": 0.00019739636828713922,
+      "loss": 3.0587,
+      "step": 37409
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6951032876968384,
+      "learning_rate": 0.0001973819006907059,
+      "loss": 3.0455,
+      "step": 37410
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6221358180046082,
+      "learning_rate": 0.000197367433364549,
+      "loss": 3.1964,
+      "step": 37411
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6235369443893433,
+      "learning_rate": 0.00019735296630870656,
+      "loss": 2.9421,
+      "step": 37412
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6678821444511414,
+      "learning_rate": 0.00019733849952321655,
+      "loss": 2.8644,
+      "step": 37413
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6573663949966431,
+      "learning_rate": 0.00019732403300811729,
+      "loss": 3.0983,
+      "step": 37414
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6279765367507935,
+      "learning_rate": 0.00019730956676344667,
+      "loss": 2.7258,
+      "step": 37415
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6428644061088562,
+      "learning_rate": 0.0001972951007892429,
+      "loss": 2.9142,
+      "step": 37416
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6456839442253113,
+      "learning_rate": 0.0001972806350855441,
+      "loss": 2.9727,
+      "step": 37417
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6260696649551392,
+      "learning_rate": 0.00019726616965238828,
+      "loss": 3.1471,
+      "step": 37418
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6509988307952881,
+      "learning_rate": 0.00019725170448981377,
+      "loss": 2.923,
+      "step": 37419
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6517656445503235,
+      "learning_rate": 0.0001972372395978583,
+      "loss": 2.9567,
+      "step": 37420
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6638441681861877,
+      "learning_rate": 0.00019722277497656018,
+      "loss": 3.1524,
+      "step": 37421
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6186351776123047,
+      "learning_rate": 0.00019720831062595759,
+      "loss": 3.002,
+      "step": 37422
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6337801814079285,
+      "learning_rate": 0.00019719384654608849,
+      "loss": 2.9353,
+      "step": 37423
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6668829917907715,
+      "learning_rate": 0.00019717938273699107,
+      "loss": 3.0473,
+      "step": 37424
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6589478850364685,
+      "learning_rate": 0.00019716491919870324,
+      "loss": 3.1595,
+      "step": 37425
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6218236684799194,
+      "learning_rate": 0.00019715045593126325,
+      "loss": 2.9327,
+      "step": 37426
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6272194981575012,
+      "learning_rate": 0.00019713599293470928,
+      "loss": 2.9345,
+      "step": 37427
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.655690610408783,
+      "learning_rate": 0.00019712153020907917,
+      "loss": 3.021,
+      "step": 37428
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6610934734344482,
+      "learning_rate": 0.0001971070677544113,
+      "loss": 3.0346,
+      "step": 37429
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6667243242263794,
+      "learning_rate": 0.00019709260557074348,
+      "loss": 3.0287,
+      "step": 37430
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6456306576728821,
+      "learning_rate": 0.00019707814365811391,
+      "loss": 2.8273,
+      "step": 37431
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6371907591819763,
+      "learning_rate": 0.00019706368201656086,
+      "loss": 3.0458,
+      "step": 37432
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6370196342468262,
+      "learning_rate": 0.00019704922064612217,
+      "loss": 3.024,
+      "step": 37433
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6374519467353821,
+      "learning_rate": 0.00019703475954683606,
+      "loss": 2.9035,
+      "step": 37434
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7033529877662659,
+      "learning_rate": 0.00019702029871874048,
+      "loss": 3.0132,
+      "step": 37435
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6438736915588379,
+      "learning_rate": 0.00019700583816187363,
+      "loss": 3.004,
+      "step": 37436
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8195931911468506,
+      "learning_rate": 0.00019699137787627369,
+      "loss": 2.8433,
+      "step": 37437
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6257565021514893,
+      "learning_rate": 0.0001969769178619785,
+      "loss": 3.02,
+      "step": 37438
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6688423752784729,
+      "learning_rate": 0.0001969624581190264,
+      "loss": 3.1861,
+      "step": 37439
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6222291588783264,
+      "learning_rate": 0.00019694799864745525,
+      "loss": 2.9031,
+      "step": 37440
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6126012206077576,
+      "learning_rate": 0.00019693353944730325,
+      "loss": 2.9896,
+      "step": 37441
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6390916705131531,
+      "learning_rate": 0.0001969190805186086,
+      "loss": 2.7175,
+      "step": 37442
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6593005061149597,
+      "learning_rate": 0.00019690462186140907,
+      "loss": 2.9528,
+      "step": 37443
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6446021199226379,
+      "learning_rate": 0.00019689016347574308,
+      "loss": 3.0649,
+      "step": 37444
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6597580909729004,
+      "learning_rate": 0.0001968757053616485,
+      "loss": 2.949,
+      "step": 37445
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6188734173774719,
+      "learning_rate": 0.0001968612475191634,
+      "loss": 3.0953,
+      "step": 37446
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6276201009750366,
+      "learning_rate": 0.00019684678994832605,
+      "loss": 2.9462,
+      "step": 37447
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6137672662734985,
+      "learning_rate": 0.00019683233264917435,
+      "loss": 2.7369,
+      "step": 37448
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6281610727310181,
+      "learning_rate": 0.00019681787562174645,
+      "loss": 2.8975,
+      "step": 37449
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.675428032875061,
+      "learning_rate": 0.00019680341886608034,
+      "loss": 3.0545,
+      "step": 37450
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6224192976951599,
+      "learning_rate": 0.0001967889623822142,
+      "loss": 2.9262,
+      "step": 37451
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6284218430519104,
+      "learning_rate": 0.00019677450617018613,
+      "loss": 2.9402,
+      "step": 37452
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6586654186248779,
+      "learning_rate": 0.000196760050230034,
+      "loss": 2.9344,
+      "step": 37453
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6120323538780212,
+      "learning_rate": 0.0001967455945617962,
+      "loss": 2.8945,
+      "step": 37454
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6563277840614319,
+      "learning_rate": 0.00019673113916551053,
+      "loss": 3.0017,
+      "step": 37455
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6687843799591064,
+      "learning_rate": 0.0001967166840412151,
+      "loss": 3.0225,
+      "step": 37456
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6557430028915405,
+      "learning_rate": 0.0001967022291889482,
+      "loss": 3.1305,
+      "step": 37457
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6802811026573181,
+      "learning_rate": 0.00019668777460874766,
+      "loss": 3.053,
+      "step": 37458
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6781749725341797,
+      "learning_rate": 0.00019667332030065172,
+      "loss": 2.8287,
+      "step": 37459
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6726826429367065,
+      "learning_rate": 0.00019665886626469824,
+      "loss": 3.1735,
+      "step": 37460
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6283700466156006,
+      "learning_rate": 0.00019664441250092542,
+      "loss": 2.9045,
+      "step": 37461
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6707321405410767,
+      "learning_rate": 0.00019662995900937143,
+      "loss": 3.1,
+      "step": 37462
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6527551412582397,
+      "learning_rate": 0.0001966155057900742,
+      "loss": 2.8893,
+      "step": 37463
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6310652494430542,
+      "learning_rate": 0.0001966010528430719,
+      "loss": 3.1475,
+      "step": 37464
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.626528799533844,
+      "learning_rate": 0.00019658660016840237,
+      "loss": 3.2232,
+      "step": 37465
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6604952812194824,
+      "learning_rate": 0.00019657214776610388,
+      "loss": 3.2123,
+      "step": 37466
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6406633257865906,
+      "learning_rate": 0.00019655769563621457,
+      "loss": 2.8845,
+      "step": 37467
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6594672799110413,
+      "learning_rate": 0.00019654324377877224,
+      "loss": 3.0813,
+      "step": 37468
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6008974313735962,
+      "learning_rate": 0.0001965287921938152,
+      "loss": 3.1406,
+      "step": 37469
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6440598964691162,
+      "learning_rate": 0.00019651434088138135,
+      "loss": 3.0571,
+      "step": 37470
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6732915043830872,
+      "learning_rate": 0.00019649988984150887,
+      "loss": 2.9316,
+      "step": 37471
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6244654059410095,
+      "learning_rate": 0.00019648543907423567,
+      "loss": 3.0407,
+      "step": 37472
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6534292697906494,
+      "learning_rate": 0.0001964709885795999,
+      "loss": 2.9746,
+      "step": 37473
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.626899778842926,
+      "learning_rate": 0.0001964565383576397,
+      "loss": 2.8462,
+      "step": 37474
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6582539677619934,
+      "learning_rate": 0.00019644208840839295,
+      "loss": 3.0684,
+      "step": 37475
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6599977016448975,
+      "learning_rate": 0.00019642763873189796,
+      "loss": 2.8524,
+      "step": 37476
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6857508420944214,
+      "learning_rate": 0.00019641318932819255,
+      "loss": 2.9823,
+      "step": 37477
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6616058349609375,
+      "learning_rate": 0.00019639874019731476,
+      "loss": 3.2488,
+      "step": 37478
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6988261938095093,
+      "learning_rate": 0.0001963842913393029,
+      "loss": 2.8922,
+      "step": 37479
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6614097952842712,
+      "learning_rate": 0.0001963698427541948,
+      "loss": 2.8559,
+      "step": 37480
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6063705086708069,
+      "learning_rate": 0.00019635539444202874,
+      "loss": 2.972,
+      "step": 37481
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6515814065933228,
+      "learning_rate": 0.00019634094640284244,
+      "loss": 2.9072,
+      "step": 37482
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6901261210441589,
+      "learning_rate": 0.00019632649863667414,
+      "loss": 3.0161,
+      "step": 37483
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.628663957118988,
+      "learning_rate": 0.000196312051143562,
+      "loss": 3.0743,
+      "step": 37484
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6244681477546692,
+      "learning_rate": 0.0001962976039235439,
+      "loss": 3.1726,
+      "step": 37485
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6625844836235046,
+      "learning_rate": 0.000196283156976658,
+      "loss": 2.9569,
+      "step": 37486
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6282349228858948,
+      "learning_rate": 0.0001962687103029422,
+      "loss": 3.1225,
+      "step": 37487
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6397706866264343,
+      "learning_rate": 0.00019625426390243472,
+      "loss": 3.0557,
+      "step": 37488
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6789543032646179,
+      "learning_rate": 0.0001962398177751736,
+      "loss": 2.9712,
+      "step": 37489
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6492639780044556,
+      "learning_rate": 0.0001962253719211967,
+      "loss": 2.9644,
+      "step": 37490
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6266045570373535,
+      "learning_rate": 0.00019621092634054235,
+      "loss": 3.053,
+      "step": 37491
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.633960485458374,
+      "learning_rate": 0.00019619648103324835,
+      "loss": 3.073,
+      "step": 37492
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.680493175983429,
+      "learning_rate": 0.00019618203599935282,
+      "loss": 3.0203,
+      "step": 37493
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6308938264846802,
+      "learning_rate": 0.0001961675912388939,
+      "loss": 3.03,
+      "step": 37494
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6352134943008423,
+      "learning_rate": 0.00019615314675190955,
+      "loss": 3.1219,
+      "step": 37495
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.608165442943573,
+      "learning_rate": 0.00019613870253843785,
+      "loss": 3.0078,
+      "step": 37496
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6532719731330872,
+      "learning_rate": 0.00019612425859851673,
+      "loss": 2.8869,
+      "step": 37497
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6677577495574951,
+      "learning_rate": 0.00019610981493218437,
+      "loss": 2.8604,
+      "step": 37498
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5935999751091003,
+      "learning_rate": 0.0001960953715394788,
+      "loss": 3.175,
+      "step": 37499
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6511732935905457,
+      "learning_rate": 0.0001960809284204379,
+      "loss": 2.9567,
+      "step": 37500
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6400048136711121,
+      "learning_rate": 0.00019606648557509996,
+      "loss": 2.9456,
+      "step": 37501
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6501739025115967,
+      "learning_rate": 0.0001960520430035028,
+      "loss": 2.9489,
+      "step": 37502
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6374745965003967,
+      "learning_rate": 0.0001960376007056846,
+      "loss": 3.0406,
+      "step": 37503
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6723983883857727,
+      "learning_rate": 0.0001960231586816834,
+      "loss": 2.8661,
+      "step": 37504
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6227818727493286,
+      "learning_rate": 0.0001960087169315371,
+      "loss": 2.7228,
+      "step": 37505
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6718969345092773,
+      "learning_rate": 0.00019599427545528392,
+      "loss": 3.0161,
+      "step": 37506
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6838817596435547,
+      "learning_rate": 0.00019597983425296176,
+      "loss": 2.8973,
+      "step": 37507
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.710553765296936,
+      "learning_rate": 0.00019596539332460863,
+      "loss": 2.8391,
+      "step": 37508
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6641603708267212,
+      "learning_rate": 0.00019595095267026277,
+      "loss": 3.1281,
+      "step": 37509
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6101526618003845,
+      "learning_rate": 0.00019593651228996201,
+      "loss": 3.0391,
+      "step": 37510
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6275225281715393,
+      "learning_rate": 0.00019592207218374448,
+      "loss": 2.9472,
+      "step": 37511
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6456922888755798,
+      "learning_rate": 0.0001959076323516481,
+      "loss": 3.2232,
+      "step": 37512
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6447487473487854,
+      "learning_rate": 0.000195893192793711,
+      "loss": 2.7937,
+      "step": 37513
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6500653624534607,
+      "learning_rate": 0.00019587875350997127,
+      "loss": 3.1145,
+      "step": 37514
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5911208391189575,
+      "learning_rate": 0.00019586431450046676,
+      "loss": 3.0935,
+      "step": 37515
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6784209609031677,
+      "learning_rate": 0.00019584987576523572,
+      "loss": 2.9249,
+      "step": 37516
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6476365923881531,
+      "learning_rate": 0.000195835437304316,
+      "loss": 2.9984,
+      "step": 37517
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6495740413665771,
+      "learning_rate": 0.0001958209991177456,
+      "loss": 2.7983,
+      "step": 37518
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6220631003379822,
+      "learning_rate": 0.0001958065612055628,
+      "loss": 3.1553,
+      "step": 37519
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6869008541107178,
+      "learning_rate": 0.0001957921235678054,
+      "loss": 2.9559,
+      "step": 37520
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6284224390983582,
+      "learning_rate": 0.00019577768620451154,
+      "loss": 2.9683,
+      "step": 37521
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6778497695922852,
+      "learning_rate": 0.00019576324911571908,
+      "loss": 3.1064,
+      "step": 37522
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6434932947158813,
+      "learning_rate": 0.00019574881230146615,
+      "loss": 3.1636,
+      "step": 37523
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7298178672790527,
+      "learning_rate": 0.0001957343757617909,
+      "loss": 3.0143,
+      "step": 37524
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6708294749259949,
+      "learning_rate": 0.0001957199394967312,
+      "loss": 2.9184,
+      "step": 37525
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6386827826499939,
+      "learning_rate": 0.0001957055035063252,
+      "loss": 3.0899,
+      "step": 37526
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7549542784690857,
+      "learning_rate": 0.00019569106779061067,
+      "loss": 2.9044,
+      "step": 37527
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7159115076065063,
+      "learning_rate": 0.00019567663234962586,
+      "loss": 3.0351,
+      "step": 37528
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.643603503704071,
+      "learning_rate": 0.00019566219718340877,
+      "loss": 3.097,
+      "step": 37529
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6065301895141602,
+      "learning_rate": 0.00019564776229199726,
+      "loss": 2.8429,
+      "step": 37530
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6380607485771179,
+      "learning_rate": 0.00019563332767542964,
+      "loss": 3.0454,
+      "step": 37531
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6211923956871033,
+      "learning_rate": 0.0001956188933337436,
+      "loss": 2.7386,
+      "step": 37532
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6757135987281799,
+      "learning_rate": 0.0001956044592669773,
+      "loss": 2.9283,
+      "step": 37533
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5980146527290344,
+      "learning_rate": 0.00019559002547516887,
+      "loss": 3.097,
+      "step": 37534
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6270644664764404,
+      "learning_rate": 0.00019557559195835614,
+      "loss": 2.8905,
+      "step": 37535
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6243385076522827,
+      "learning_rate": 0.00019556115871657728,
+      "loss": 3.0983,
+      "step": 37536
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6248341798782349,
+      "learning_rate": 0.00019554672574987011,
+      "loss": 3.1228,
+      "step": 37537
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6501051187515259,
+      "learning_rate": 0.00019553229305827282,
+      "loss": 3.0654,
+      "step": 37538
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7006635069847107,
+      "learning_rate": 0.0001955178606418234,
+      "loss": 2.8608,
+      "step": 37539
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6526589393615723,
+      "learning_rate": 0.00019550342850055974,
+      "loss": 3.0982,
+      "step": 37540
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6131229996681213,
+      "learning_rate": 0.00019548899663452007,
+      "loss": 2.8902,
+      "step": 37541
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6289817094802856,
+      "learning_rate": 0.00019547456504374213,
+      "loss": 3.0206,
+      "step": 37542
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6378176212310791,
+      "learning_rate": 0.00019546013372826411,
+      "loss": 3.0923,
+      "step": 37543
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6519317030906677,
+      "learning_rate": 0.00019544570268812406,
+      "loss": 3.0326,
+      "step": 37544
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7056100368499756,
+      "learning_rate": 0.00019543127192335979,
+      "loss": 2.8943,
+      "step": 37545
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6589441895484924,
+      "learning_rate": 0.00019541684143400953,
+      "loss": 3.1757,
+      "step": 37546
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6761717200279236,
+      "learning_rate": 0.00019540241122011111,
+      "loss": 2.8114,
+      "step": 37547
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6543384790420532,
+      "learning_rate": 0.00019538798128170272,
+      "loss": 3.2523,
+      "step": 37548
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.633036196231842,
+      "learning_rate": 0.0001953735516188221,
+      "loss": 2.858,
+      "step": 37549
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6491231322288513,
+      "learning_rate": 0.00019535912223150745,
+      "loss": 3.1472,
+      "step": 37550
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6392214894294739,
+      "learning_rate": 0.00019534469311979685,
+      "loss": 2.9719,
+      "step": 37551
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6699644923210144,
+      "learning_rate": 0.00019533026428372803,
+      "loss": 2.9833,
+      "step": 37552
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.695570707321167,
+      "learning_rate": 0.00019531583572333924,
+      "loss": 2.985,
+      "step": 37553
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7695941925048828,
+      "learning_rate": 0.0001953014074386684,
+      "loss": 2.8777,
+      "step": 37554
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6346907615661621,
+      "learning_rate": 0.0001952869794297534,
+      "loss": 3.1383,
+      "step": 37555
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6936262249946594,
+      "learning_rate": 0.00019527255169663244,
+      "loss": 3.0131,
+      "step": 37556
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6635444760322571,
+      "learning_rate": 0.0001952581242393434,
+      "loss": 3.0441,
+      "step": 37557
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6767592430114746,
+      "learning_rate": 0.0001952436970579244,
+      "loss": 2.9566,
+      "step": 37558
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6856628656387329,
+      "learning_rate": 0.00019522927015241316,
+      "loss": 3.2645,
+      "step": 37559
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6861822009086609,
+      "learning_rate": 0.00019521484352284793,
+      "loss": 3.02,
+      "step": 37560
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6693269610404968,
+      "learning_rate": 0.00019520041716926674,
+      "loss": 3.0023,
+      "step": 37561
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6560589075088501,
+      "learning_rate": 0.00019518599109170727,
+      "loss": 2.8327,
+      "step": 37562
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6303350925445557,
+      "learning_rate": 0.00019517156529020793,
+      "loss": 3.0216,
+      "step": 37563
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6482394337654114,
+      "learning_rate": 0.0001951571397648064,
+      "loss": 3.0575,
+      "step": 37564
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6522819399833679,
+      "learning_rate": 0.00019514271451554076,
+      "loss": 3.136,
+      "step": 37565
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6398262977600098,
+      "learning_rate": 0.00019512828954244913,
+      "loss": 2.9499,
+      "step": 37566
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6437414288520813,
+      "learning_rate": 0.00019511386484556933,
+      "loss": 2.9961,
+      "step": 37567
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6147285103797913,
+      "learning_rate": 0.00019509944042493953,
+      "loss": 3.1381,
+      "step": 37568
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6700540781021118,
+      "learning_rate": 0.0001950850162805975,
+      "loss": 2.8628,
+      "step": 37569
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6323624849319458,
+      "learning_rate": 0.00019507059241258133,
+      "loss": 2.9542,
+      "step": 37570
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6200016140937805,
+      "learning_rate": 0.00019505616882092917,
+      "loss": 3.0798,
+      "step": 37571
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6565930247306824,
+      "learning_rate": 0.00019504174550567875,
+      "loss": 2.9682,
+      "step": 37572
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6288124918937683,
+      "learning_rate": 0.00019502732246686825,
+      "loss": 3.1476,
+      "step": 37573
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7063016891479492,
+      "learning_rate": 0.00019501289970453545,
+      "loss": 3.0224,
+      "step": 37574
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.626835286617279,
+      "learning_rate": 0.00019499847721871855,
+      "loss": 2.9865,
+      "step": 37575
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6863059401512146,
+      "learning_rate": 0.0001949840550094555,
+      "loss": 3.0436,
+      "step": 37576
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6676422357559204,
+      "learning_rate": 0.0001949696330767841,
+      "loss": 3.0392,
+      "step": 37577
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.649277925491333,
+      "learning_rate": 0.00019495521142074264,
+      "loss": 3.0748,
+      "step": 37578
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6728671789169312,
+      "learning_rate": 0.00019494079004136886,
+      "loss": 3.0207,
+      "step": 37579
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6115674376487732,
+      "learning_rate": 0.00019492636893870074,
+      "loss": 3.0674,
+      "step": 37580
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6326895356178284,
+      "learning_rate": 0.00019491194811277647,
+      "loss": 3.042,
+      "step": 37581
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6360620260238647,
+      "learning_rate": 0.00019489752756363378,
+      "loss": 2.882,
+      "step": 37582
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6419123411178589,
+      "learning_rate": 0.00019488310729131095,
+      "loss": 3.1109,
+      "step": 37583
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6531715989112854,
+      "learning_rate": 0.00019486868729584565,
+      "loss": 2.713,
+      "step": 37584
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6210694909095764,
+      "learning_rate": 0.00019485426757727593,
+      "loss": 3.0199,
+      "step": 37585
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6679146885871887,
+      "learning_rate": 0.00019483984813564002,
+      "loss": 2.7999,
+      "step": 37586
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6969517469406128,
+      "learning_rate": 0.0001948254289709756,
+      "loss": 2.928,
+      "step": 37587
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6617041826248169,
+      "learning_rate": 0.00019481101008332082,
+      "loss": 3.1227,
+      "step": 37588
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6362742185592651,
+      "learning_rate": 0.00019479659147271354,
+      "loss": 2.732,
+      "step": 37589
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6810065507888794,
+      "learning_rate": 0.00019478217313919176,
+      "loss": 3.0502,
+      "step": 37590
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.643323540687561,
+      "learning_rate": 0.00019476775508279361,
+      "loss": 2.8579,
+      "step": 37591
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6467273235321045,
+      "learning_rate": 0.00019475333730355681,
+      "loss": 3.1401,
+      "step": 37592
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6345007419586182,
+      "learning_rate": 0.00019473891980151958,
+      "loss": 2.7846,
+      "step": 37593
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6979330778121948,
+      "learning_rate": 0.00019472450257671973,
+      "loss": 2.8787,
+      "step": 37594
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8724361062049866,
+      "learning_rate": 0.00019471008562919518,
+      "loss": 3.1034,
+      "step": 37595
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6298291087150574,
+      "learning_rate": 0.00019469566895898412,
+      "loss": 3.1297,
+      "step": 37596
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6355240941047668,
+      "learning_rate": 0.0001946812525661244,
+      "loss": 3.1283,
+      "step": 37597
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6391889452934265,
+      "learning_rate": 0.00019466683645065402,
+      "loss": 3.0097,
+      "step": 37598
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6222915649414062,
+      "learning_rate": 0.0001946524206126108,
+      "loss": 3.0043,
+      "step": 37599
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6683493256568909,
+      "learning_rate": 0.00019463800505203284,
+      "loss": 3.053,
+      "step": 37600
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6785693168640137,
+      "learning_rate": 0.00019462358976895822,
+      "loss": 3.0694,
+      "step": 37601
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6351543068885803,
+      "learning_rate": 0.00019460917476342464,
+      "loss": 2.9983,
+      "step": 37602
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6268134713172913,
+      "learning_rate": 0.00019459476003547035,
+      "loss": 2.9023,
+      "step": 37603
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6076909303665161,
+      "learning_rate": 0.00019458034558513304,
+      "loss": 3.0121,
+      "step": 37604
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6232816576957703,
+      "learning_rate": 0.00019456593141245086,
+      "loss": 2.9033,
+      "step": 37605
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6748607754707336,
+      "learning_rate": 0.00019455151751746183,
+      "loss": 3.0505,
+      "step": 37606
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6353899836540222,
+      "learning_rate": 0.00019453710390020365,
+      "loss": 3.1189,
+      "step": 37607
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6453273892402649,
+      "learning_rate": 0.00019452269056071457,
+      "loss": 3.3284,
+      "step": 37608
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6326226592063904,
+      "learning_rate": 0.00019450827749903238,
+      "loss": 2.9296,
+      "step": 37609
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6521055102348328,
+      "learning_rate": 0.00019449386471519502,
+      "loss": 3.0117,
+      "step": 37610
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.663453221321106,
+      "learning_rate": 0.00019447945220924063,
+      "loss": 2.8768,
+      "step": 37611
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6682723760604858,
+      "learning_rate": 0.000194465039981207,
+      "loss": 2.9598,
+      "step": 37612
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6204665899276733,
+      "learning_rate": 0.00019445062803113218,
+      "loss": 3.1624,
+      "step": 37613
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6165514588356018,
+      "learning_rate": 0.000194436216359054,
+      "loss": 3.3258,
+      "step": 37614
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6465550065040588,
+      "learning_rate": 0.00019442180496501054,
+      "loss": 2.9518,
+      "step": 37615
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6707178354263306,
+      "learning_rate": 0.00019440739384903982,
+      "loss": 2.9864,
+      "step": 37616
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6141831278800964,
+      "learning_rate": 0.00019439298301117955,
+      "loss": 2.9937,
+      "step": 37617
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6654483675956726,
+      "learning_rate": 0.00019437857245146797,
+      "loss": 2.8547,
+      "step": 37618
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6087260842323303,
+      "learning_rate": 0.00019436416216994285,
+      "loss": 2.8802,
+      "step": 37619
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7105410099029541,
+      "learning_rate": 0.00019434975216664217,
+      "loss": 3.0672,
+      "step": 37620
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6254940629005432,
+      "learning_rate": 0.000194335342441604,
+      "loss": 2.9234,
+      "step": 37621
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6136469841003418,
+      "learning_rate": 0.00019432093299486612,
+      "loss": 3.0558,
+      "step": 37622
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6719083786010742,
+      "learning_rate": 0.00019430652382646665,
+      "loss": 2.9483,
+      "step": 37623
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6281088590621948,
+      "learning_rate": 0.00019429211493644335,
+      "loss": 2.958,
+      "step": 37624
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.642126739025116,
+      "learning_rate": 0.00019427770632483423,
+      "loss": 3.0231,
+      "step": 37625
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6834142208099365,
+      "learning_rate": 0.00019426329799167742,
+      "loss": 2.9678,
+      "step": 37626
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6461591124534607,
+      "learning_rate": 0.00019424888993701072,
+      "loss": 2.944,
+      "step": 37627
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6434966325759888,
+      "learning_rate": 0.0001942344821608721,
+      "loss": 2.8844,
+      "step": 37628
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.62734055519104,
+      "learning_rate": 0.00019422007466329938,
+      "loss": 3.0886,
+      "step": 37629
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6289473176002502,
+      "learning_rate": 0.00019420566744433074,
+      "loss": 2.8589,
+      "step": 37630
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6288855671882629,
+      "learning_rate": 0.000194191260504004,
+      "loss": 2.986,
+      "step": 37631
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6580570936203003,
+      "learning_rate": 0.000194176853842357,
+      "loss": 3.0736,
+      "step": 37632
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6353893280029297,
+      "learning_rate": 0.00019416244745942797,
+      "loss": 3.0207,
+      "step": 37633
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6546821594238281,
+      "learning_rate": 0.0001941480413552546,
+      "loss": 2.9067,
+      "step": 37634
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6692838668823242,
+      "learning_rate": 0.00019413363552987496,
+      "loss": 3.0202,
+      "step": 37635
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6453374028205872,
+      "learning_rate": 0.00019411922998332682,
+      "loss": 3.0015,
+      "step": 37636
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6587013006210327,
+      "learning_rate": 0.00019410482471564838,
+      "loss": 2.9682,
+      "step": 37637
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6185652613639832,
+      "learning_rate": 0.00019409041972687745,
+      "loss": 2.9989,
+      "step": 37638
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6610234379768372,
+      "learning_rate": 0.0001940760150170519,
+      "loss": 3.0296,
+      "step": 37639
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.610658586025238,
+      "learning_rate": 0.00019406161058620983,
+      "loss": 3.0412,
+      "step": 37640
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6481900215148926,
+      "learning_rate": 0.000194047206434389,
+      "loss": 2.9011,
+      "step": 37641
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6394829750061035,
+      "learning_rate": 0.00019403280256162744,
+      "loss": 3.0083,
+      "step": 37642
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6193023920059204,
+      "learning_rate": 0.0001940183989679632,
+      "loss": 3.0251,
+      "step": 37643
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6388663053512573,
+      "learning_rate": 0.00019400399565343393,
+      "loss": 3.0921,
+      "step": 37644
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6341210603713989,
+      "learning_rate": 0.00019398959261807794,
+      "loss": 3.1424,
+      "step": 37645
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.662173330783844,
+      "learning_rate": 0.00019397518986193277,
+      "loss": 2.9784,
+      "step": 37646
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6525291800498962,
+      "learning_rate": 0.00019396078738503659,
+      "loss": 2.8043,
+      "step": 37647
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6806787252426147,
+      "learning_rate": 0.00019394638518742733,
+      "loss": 2.8409,
+      "step": 37648
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6746661067008972,
+      "learning_rate": 0.0001939319832691429,
+      "loss": 3.0212,
+      "step": 37649
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6305448412895203,
+      "learning_rate": 0.00019391758163022125,
+      "loss": 3.1187,
+      "step": 37650
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6361280679702759,
+      "learning_rate": 0.00019390318027070013,
+      "loss": 3.0798,
+      "step": 37651
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6559592485427856,
+      "learning_rate": 0.0001938887791906177,
+      "loss": 3.1859,
+      "step": 37652
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6498610973358154,
+      "learning_rate": 0.00019387437839001184,
+      "loss": 3.1285,
+      "step": 37653
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6713485717773438,
+      "learning_rate": 0.0001938599778689203,
+      "loss": 2.8758,
+      "step": 37654
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6473253965377808,
+      "learning_rate": 0.00019384557762738134,
+      "loss": 2.882,
+      "step": 37655
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6119524240493774,
+      "learning_rate": 0.0001938311776654326,
+      "loss": 2.7755,
+      "step": 37656
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6408953070640564,
+      "learning_rate": 0.00019381677798311207,
+      "loss": 3.2346,
+      "step": 37657
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6806536912918091,
+      "learning_rate": 0.0001938023785804578,
+      "loss": 3.0154,
+      "step": 37658
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6582728028297424,
+      "learning_rate": 0.00019378797945750756,
+      "loss": 2.8902,
+      "step": 37659
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6802835464477539,
+      "learning_rate": 0.00019377358061429943,
+      "loss": 3.0137,
+      "step": 37660
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6308269500732422,
+      "learning_rate": 0.00019375918205087112,
+      "loss": 3.1318,
+      "step": 37661
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7181369662284851,
+      "learning_rate": 0.00019374478376726074,
+      "loss": 3.1762,
+      "step": 37662
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6746264100074768,
+      "learning_rate": 0.0001937303857635062,
+      "loss": 2.9109,
+      "step": 37663
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.653678834438324,
+      "learning_rate": 0.00019371598803964523,
+      "loss": 2.9796,
+      "step": 37664
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6139053106307983,
+      "learning_rate": 0.00019370159059571608,
+      "loss": 2.995,
+      "step": 37665
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6195376515388489,
+      "learning_rate": 0.0001936871934317563,
+      "loss": 2.9449,
+      "step": 37666
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6081912517547607,
+      "learning_rate": 0.0001936727965478041,
+      "loss": 3.0824,
+      "step": 37667
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7261514067649841,
+      "learning_rate": 0.0001936583999438973,
+      "loss": 3.1328,
+      "step": 37668
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6595963835716248,
+      "learning_rate": 0.0001936440036200737,
+      "loss": 3.1274,
+      "step": 37669
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6450942158699036,
+      "learning_rate": 0.00019362960757637147,
+      "loss": 3.0672,
+      "step": 37670
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6324230432510376,
+      "learning_rate": 0.00019361521181282833,
+      "loss": 3.0313,
+      "step": 37671
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6502272486686707,
+      "learning_rate": 0.00019360081632948216,
+      "loss": 3.1039,
+      "step": 37672
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6629947423934937,
+      "learning_rate": 0.0001935864211263711,
+      "loss": 2.9381,
+      "step": 37673
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6200689077377319,
+      "learning_rate": 0.00019357202620353284,
+      "loss": 3.0525,
+      "step": 37674
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6390634775161743,
+      "learning_rate": 0.00019355763156100544,
+      "loss": 3.0149,
+      "step": 37675
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6882568597793579,
+      "learning_rate": 0.00019354323719882667,
+      "loss": 2.9673,
+      "step": 37676
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6751754283905029,
+      "learning_rate": 0.00019352884311703457,
+      "loss": 3.0688,
+      "step": 37677
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6495875716209412,
+      "learning_rate": 0.00019351444931566707,
+      "loss": 3.0199,
+      "step": 37678
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6434128880500793,
+      "learning_rate": 0.00019350005579476186,
+      "loss": 2.9602,
+      "step": 37679
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6609219312667847,
+      "learning_rate": 0.0001934856625543572,
+      "loss": 3.1063,
+      "step": 37680
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6295854449272156,
+      "learning_rate": 0.00019347126959449068,
+      "loss": 2.9994,
+      "step": 37681
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6681686639785767,
+      "learning_rate": 0.00019345687691520028,
+      "loss": 3.0875,
+      "step": 37682
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7291771173477173,
+      "learning_rate": 0.00019344248451652412,
+      "loss": 3.2816,
+      "step": 37683
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6248905062675476,
+      "learning_rate": 0.00019342809239849987,
+      "loss": 2.9861,
+      "step": 37684
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6470739841461182,
+      "learning_rate": 0.00019341370056116556,
+      "loss": 2.7754,
+      "step": 37685
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6730750203132629,
+      "learning_rate": 0.00019339930900455894,
+      "loss": 2.8785,
+      "step": 37686
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6697092652320862,
+      "learning_rate": 0.00019338491772871806,
+      "loss": 3.0509,
+      "step": 37687
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6595104336738586,
+      "learning_rate": 0.00019337052673368087,
+      "loss": 3.1409,
+      "step": 37688
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6694124937057495,
+      "learning_rate": 0.00019335613601948511,
+      "loss": 3.0137,
+      "step": 37689
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.649567186832428,
+      "learning_rate": 0.00019334174558616887,
+      "loss": 3.1952,
+      "step": 37690
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6771290302276611,
+      "learning_rate": 0.0001933273554337698,
+      "loss": 2.8044,
+      "step": 37691
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6197347044944763,
+      "learning_rate": 0.00019331296556232598,
+      "loss": 3.1106,
+      "step": 37692
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6491581797599792,
+      "learning_rate": 0.0001932985759718754,
+      "loss": 3.0107,
+      "step": 37693
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6176472902297974,
+      "learning_rate": 0.00019328418666245568,
+      "loss": 2.936,
+      "step": 37694
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6549054980278015,
+      "learning_rate": 0.00019326979763410501,
+      "loss": 2.8333,
+      "step": 37695
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6261773109436035,
+      "learning_rate": 0.0001932554088868611,
+      "loss": 2.8714,
+      "step": 37696
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6560527086257935,
+      "learning_rate": 0.00019324102042076176,
+      "loss": 2.9073,
+      "step": 37697
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6137068867683411,
+      "learning_rate": 0.00019322663223584523,
+      "loss": 2.9635,
+      "step": 37698
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.71515953540802,
+      "learning_rate": 0.0001932122443321491,
+      "loss": 2.8384,
+      "step": 37699
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6288661360740662,
+      "learning_rate": 0.00019319785670971143,
+      "loss": 2.9895,
+      "step": 37700
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6185125112533569,
+      "learning_rate": 0.00019318346936856998,
+      "loss": 3.028,
+      "step": 37701
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6613661646842957,
+      "learning_rate": 0.00019316908230876276,
+      "loss": 3.1521,
+      "step": 37702
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6700349450111389,
+      "learning_rate": 0.00019315469553032766,
+      "loss": 2.9103,
+      "step": 37703
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6402429938316345,
+      "learning_rate": 0.00019314030903330238,
+      "loss": 3.1274,
+      "step": 37704
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6753808856010437,
+      "learning_rate": 0.00019312592281772509,
+      "loss": 2.8903,
+      "step": 37705
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6379236578941345,
+      "learning_rate": 0.00019311153688363346,
+      "loss": 3.0307,
+      "step": 37706
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6957564353942871,
+      "learning_rate": 0.00019309715123106558,
+      "loss": 2.9885,
+      "step": 37707
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.694107711315155,
+      "learning_rate": 0.00019308276586005915,
+      "loss": 3.096,
+      "step": 37708
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7162430882453918,
+      "learning_rate": 0.00019306838077065208,
+      "loss": 2.9719,
+      "step": 37709
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6701442003250122,
+      "learning_rate": 0.0001930539959628825,
+      "loss": 3.0133,
+      "step": 37710
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6491703391075134,
+      "learning_rate": 0.00019303961143678797,
+      "loss": 2.9792,
+      "step": 37711
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6918108463287354,
+      "learning_rate": 0.00019302522719240664,
+      "loss": 2.9754,
+      "step": 37712
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6854971647262573,
+      "learning_rate": 0.00019301084322977612,
+      "loss": 3.0784,
+      "step": 37713
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6159098148345947,
+      "learning_rate": 0.00019299645954893452,
+      "loss": 2.9788,
+      "step": 37714
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6346889138221741,
+      "learning_rate": 0.0001929820761499197,
+      "loss": 2.8104,
+      "step": 37715
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6619777679443359,
+      "learning_rate": 0.0001929676930327694,
+      "loss": 3.0457,
+      "step": 37716
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6497694253921509,
+      "learning_rate": 0.00019295331019752166,
+      "loss": 3.2556,
+      "step": 37717
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6476145386695862,
+      "learning_rate": 0.0001929389276442143,
+      "loss": 2.798,
+      "step": 37718
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6143190264701843,
+      "learning_rate": 0.00019292454537288514,
+      "loss": 3.0311,
+      "step": 37719
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.635507345199585,
+      "learning_rate": 0.00019291016338357218,
+      "loss": 2.9349,
+      "step": 37720
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6005894541740417,
+      "learning_rate": 0.00019289578167631322,
+      "loss": 3.1017,
+      "step": 37721
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6846247315406799,
+      "learning_rate": 0.00019288140025114622,
+      "loss": 3.0524,
+      "step": 37722
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6498640179634094,
+      "learning_rate": 0.0001928670191081089,
+      "loss": 3.1925,
+      "step": 37723
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6696533560752869,
+      "learning_rate": 0.00019285263824723927,
+      "loss": 2.9829,
+      "step": 37724
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6601403951644897,
+      "learning_rate": 0.00019283825766857522,
+      "loss": 3.0455,
+      "step": 37725
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.650351345539093,
+      "learning_rate": 0.00019282387737215448,
+      "loss": 3.0277,
+      "step": 37726
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5891628265380859,
+      "learning_rate": 0.0001928094973580151,
+      "loss": 2.8122,
+      "step": 37727
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7312580943107605,
+      "learning_rate": 0.00019279511762619478,
+      "loss": 2.9223,
+      "step": 37728
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6708784103393555,
+      "learning_rate": 0.00019278073817673155,
+      "loss": 3.0713,
+      "step": 37729
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6447456479072571,
+      "learning_rate": 0.0001927663590096633,
+      "loss": 2.892,
+      "step": 37730
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6298274993896484,
+      "learning_rate": 0.00019275198012502767,
+      "loss": 3.187,
+      "step": 37731
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.683275043964386,
+      "learning_rate": 0.0001927376015228628,
+      "loss": 3.1043,
+      "step": 37732
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6236386299133301,
+      "learning_rate": 0.0001927232232032064,
+      "loss": 3.0443,
+      "step": 37733
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6350182294845581,
+      "learning_rate": 0.00019270884516609633,
+      "loss": 2.9401,
+      "step": 37734
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6639793515205383,
+      "learning_rate": 0.0001926944674115706,
+      "loss": 3.0851,
+      "step": 37735
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.611555814743042,
+      "learning_rate": 0.000192680089939667,
+      "loss": 3.0449,
+      "step": 37736
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7056550979614258,
+      "learning_rate": 0.0001926657127504234,
+      "loss": 3.1318,
+      "step": 37737
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6342179775238037,
+      "learning_rate": 0.00019265133584387754,
+      "loss": 2.7571,
+      "step": 37738
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6653878688812256,
+      "learning_rate": 0.00019263695922006746,
+      "loss": 3.1323,
+      "step": 37739
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6867647171020508,
+      "learning_rate": 0.00019262258287903105,
+      "loss": 3.0003,
+      "step": 37740
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6258794069290161,
+      "learning_rate": 0.00019260820682080596,
+      "loss": 3.0193,
+      "step": 37741
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6566473841667175,
+      "learning_rate": 0.00019259383104543027,
+      "loss": 3.1575,
+      "step": 37742
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6476341485977173,
+      "learning_rate": 0.0001925794555529417,
+      "loss": 3.0576,
+      "step": 37743
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6693728566169739,
+      "learning_rate": 0.00019256508034337817,
+      "loss": 2.9551,
+      "step": 37744
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.623833954334259,
+      "learning_rate": 0.0001925507054167776,
+      "loss": 3.0357,
+      "step": 37745
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6552188992500305,
+      "learning_rate": 0.00019253633077317772,
+      "loss": 2.9078,
+      "step": 37746
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6924782991409302,
+      "learning_rate": 0.00019252195641261663,
+      "loss": 2.8484,
+      "step": 37747
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6190445423126221,
+      "learning_rate": 0.00019250758233513185,
+      "loss": 3.0351,
+      "step": 37748
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7065290212631226,
+      "learning_rate": 0.0001924932085407614,
+      "loss": 3.1009,
+      "step": 37749
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6103439331054688,
+      "learning_rate": 0.00019247883502954326,
+      "loss": 3.1459,
+      "step": 37750
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6249575018882751,
+      "learning_rate": 0.00019246446180151515,
+      "loss": 3.0093,
+      "step": 37751
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6286122798919678,
+      "learning_rate": 0.00019245008885671494,
+      "loss": 3.0115,
+      "step": 37752
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6574020385742188,
+      "learning_rate": 0.00019243571619518044,
+      "loss": 2.8001,
+      "step": 37753
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6263217329978943,
+      "learning_rate": 0.00019242134381694962,
+      "loss": 3.0588,
+      "step": 37754
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6395589709281921,
+      "learning_rate": 0.0001924069717220603,
+      "loss": 2.9058,
+      "step": 37755
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6206561326980591,
+      "learning_rate": 0.00019239259991055022,
+      "loss": 3.1611,
+      "step": 37756
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6132239699363708,
+      "learning_rate": 0.00019237822838245744,
+      "loss": 2.9598,
+      "step": 37757
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6488093137741089,
+      "learning_rate": 0.00019236385713781965,
+      "loss": 3.0767,
+      "step": 37758
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6497656106948853,
+      "learning_rate": 0.00019234948617667463,
+      "loss": 3.0172,
+      "step": 37759
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6418945789337158,
+      "learning_rate": 0.00019233511549906053,
+      "loss": 2.9671,
+      "step": 37760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6646747589111328,
+      "learning_rate": 0.0001923207451050149,
+      "loss": 3.0554,
+      "step": 37761
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5927530527114868,
+      "learning_rate": 0.0001923063749945758,
+      "loss": 2.8786,
+      "step": 37762
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6495875716209412,
+      "learning_rate": 0.00019229200516778083,
+      "loss": 3.0241,
+      "step": 37763
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6332212686538696,
+      "learning_rate": 0.00019227763562466808,
+      "loss": 3.0432,
+      "step": 37764
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6797601580619812,
+      "learning_rate": 0.00019226326636527538,
+      "loss": 2.9573,
+      "step": 37765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6078984141349792,
+      "learning_rate": 0.00019224889738964033,
+      "loss": 2.9746,
+      "step": 37766
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6482206583023071,
+      "learning_rate": 0.0001922345286978011,
+      "loss": 3.1526,
+      "step": 37767
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6135627627372742,
+      "learning_rate": 0.00019222016028979527,
+      "loss": 2.9846,
+      "step": 37768
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6265912055969238,
+      "learning_rate": 0.00019220579216566082,
+      "loss": 3.0213,
+      "step": 37769
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.648912787437439,
+      "learning_rate": 0.00019219142432543566,
+      "loss": 3.17,
+      "step": 37770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6402798295021057,
+      "learning_rate": 0.00019217705676915744,
+      "loss": 3.302,
+      "step": 37771
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6315393447875977,
+      "learning_rate": 0.00019216268949686415,
+      "loss": 3.0794,
+      "step": 37772
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6344591975212097,
+      "learning_rate": 0.00019214832250859357,
+      "loss": 3.0817,
+      "step": 37773
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6381920576095581,
+      "learning_rate": 0.0001921339558043835,
+      "loss": 3.1171,
+      "step": 37774
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6676338315010071,
+      "learning_rate": 0.00019211958938427192,
+      "loss": 3.1187,
+      "step": 37775
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6421663761138916,
+      "learning_rate": 0.00019210522324829656,
+      "loss": 3.0546,
+      "step": 37776
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6194716691970825,
+      "learning_rate": 0.00019209085739649529,
+      "loss": 3.1053,
+      "step": 37777
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6466617584228516,
+      "learning_rate": 0.00019207649182890584,
+      "loss": 3.0838,
+      "step": 37778
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6185702085494995,
+      "learning_rate": 0.00019206212654556619,
+      "loss": 3.0664,
+      "step": 37779
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6287671327590942,
+      "learning_rate": 0.0001920477615465142,
+      "loss": 3.0368,
+      "step": 37780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6562327146530151,
+      "learning_rate": 0.00019203339683178755,
+      "loss": 2.8597,
+      "step": 37781
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.609699010848999,
+      "learning_rate": 0.0001920190324014242,
+      "loss": 3.0641,
+      "step": 37782
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.631375789642334,
+      "learning_rate": 0.00019200466825546189,
+      "loss": 2.9766,
+      "step": 37783
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6531733870506287,
+      "learning_rate": 0.0001919903043939386,
+      "loss": 2.9142,
+      "step": 37784
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6809682846069336,
+      "learning_rate": 0.0001919759408168919,
+      "loss": 2.7791,
+      "step": 37785
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6292451024055481,
+      "learning_rate": 0.00019196157752435986,
+      "loss": 3.1373,
+      "step": 37786
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6322467923164368,
+      "learning_rate": 0.0001919472145163803,
+      "loss": 2.9073,
+      "step": 37787
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6325113773345947,
+      "learning_rate": 0.00019193285179299084,
+      "loss": 3.0091,
+      "step": 37788
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6427119374275208,
+      "learning_rate": 0.0001919184893542296,
+      "loss": 2.873,
+      "step": 37789
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6506245732307434,
+      "learning_rate": 0.0001919041272001341,
+      "loss": 2.9152,
+      "step": 37790
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7006027698516846,
+      "learning_rate": 0.0001918897653307424,
+      "loss": 3.2097,
+      "step": 37791
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6487759947776794,
+      "learning_rate": 0.00019187540374609235,
+      "loss": 3.1275,
+      "step": 37792
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6356369256973267,
+      "learning_rate": 0.00019186104244622152,
+      "loss": 2.8266,
+      "step": 37793
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6854275465011597,
+      "learning_rate": 0.000191846681431168,
+      "loss": 2.9221,
+      "step": 37794
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6778642535209656,
+      "learning_rate": 0.0001918323207009695,
+      "loss": 3.0824,
+      "step": 37795
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6329032182693481,
+      "learning_rate": 0.00019181796025566375,
+      "loss": 2.8184,
+      "step": 37796
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6632543206214905,
+      "learning_rate": 0.0001918036000952888,
+      "loss": 2.7941,
+      "step": 37797
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6179139614105225,
+      "learning_rate": 0.00019178924021988225,
+      "loss": 2.9767,
+      "step": 37798
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6899917721748352,
+      "learning_rate": 0.00019177488062948215,
+      "loss": 3.0879,
+      "step": 37799
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6436970829963684,
+      "learning_rate": 0.00019176052132412598,
+      "loss": 3.2205,
+      "step": 37800
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6844620108604431,
+      "learning_rate": 0.00019174616230385188,
+      "loss": 3.1832,
+      "step": 37801
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6446673274040222,
+      "learning_rate": 0.0001917318035686976,
+      "loss": 3.1017,
+      "step": 37802
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.676476001739502,
+      "learning_rate": 0.0001917174451187008,
+      "loss": 3.0669,
+      "step": 37803
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6377735733985901,
+      "learning_rate": 0.00019170308695389956,
+      "loss": 3.1079,
+      "step": 37804
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6439332962036133,
+      "learning_rate": 0.00019168872907433142,
+      "loss": 2.9107,
+      "step": 37805
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6758326888084412,
+      "learning_rate": 0.00019167437148003428,
+      "loss": 2.9954,
+      "step": 37806
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6786984801292419,
+      "learning_rate": 0.00019166001417104616,
+      "loss": 3.1813,
+      "step": 37807
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6583569645881653,
+      "learning_rate": 0.00019164565714740458,
+      "loss": 3.01,
+      "step": 37808
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6564452648162842,
+      "learning_rate": 0.0001916313004091476,
+      "loss": 2.9767,
+      "step": 37809
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6123052835464478,
+      "learning_rate": 0.00019161694395631283,
+      "loss": 2.8546,
+      "step": 37810
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6195050477981567,
+      "learning_rate": 0.00019160258778893813,
+      "loss": 2.8823,
+      "step": 37811
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6337482333183289,
+      "learning_rate": 0.00019158823190706148,
+      "loss": 2.9034,
+      "step": 37812
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6698657274246216,
+      "learning_rate": 0.0001915738763107205,
+      "loss": 3.1051,
+      "step": 37813
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6584785580635071,
+      "learning_rate": 0.00019155952099995311,
+      "loss": 3.0516,
+      "step": 37814
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.715290367603302,
+      "learning_rate": 0.00019154516597479697,
+      "loss": 3.1058,
+      "step": 37815
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6434201598167419,
+      "learning_rate": 0.00019153081123529005,
+      "loss": 2.9279,
+      "step": 37816
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6424633264541626,
+      "learning_rate": 0.00019151645678147015,
+      "loss": 3.0143,
+      "step": 37817
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6202142238616943,
+      "learning_rate": 0.00019150210261337488,
+      "loss": 3.1368,
+      "step": 37818
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6413096189498901,
+      "learning_rate": 0.00019148774873104237,
+      "loss": 3.1654,
+      "step": 37819
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6295573711395264,
+      "learning_rate": 0.00019147339513451016,
+      "loss": 3.0959,
+      "step": 37820
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6364253163337708,
+      "learning_rate": 0.00019145904182381606,
+      "loss": 3.0315,
+      "step": 37821
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6383337378501892,
+      "learning_rate": 0.00019144468879899807,
+      "loss": 2.9825,
+      "step": 37822
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6247068643569946,
+      "learning_rate": 0.00019143033606009384,
+      "loss": 2.9444,
+      "step": 37823
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6323005557060242,
+      "learning_rate": 0.00019141598360714128,
+      "loss": 3.0731,
+      "step": 37824
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6761791706085205,
+      "learning_rate": 0.00019140163144017798,
+      "loss": 2.9698,
+      "step": 37825
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.710123598575592,
+      "learning_rate": 0.00019138727955924197,
+      "loss": 2.9033,
+      "step": 37826
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6326896548271179,
+      "learning_rate": 0.000191372927964371,
+      "loss": 3.2679,
+      "step": 37827
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6370804309844971,
+      "learning_rate": 0.00019135857665560272,
+      "loss": 3.0771,
+      "step": 37828
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8834513425827026,
+      "learning_rate": 0.00019134422563297514,
+      "loss": 2.8565,
+      "step": 37829
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6568150520324707,
+      "learning_rate": 0.00019132987489652584,
+      "loss": 3.1293,
+      "step": 37830
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6432253122329712,
+      "learning_rate": 0.00019131552444629282,
+      "loss": 2.9941,
+      "step": 37831
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6338484883308411,
+      "learning_rate": 0.00019130117428231384,
+      "loss": 2.9888,
+      "step": 37832
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6891887187957764,
+      "learning_rate": 0.00019128682440462653,
+      "loss": 3.0169,
+      "step": 37833
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6474312543869019,
+      "learning_rate": 0.00019127247481326895,
+      "loss": 2.8492,
+      "step": 37834
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6678059101104736,
+      "learning_rate": 0.00019125812550827866,
+      "loss": 3.1827,
+      "step": 37835
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.663590133190155,
+      "learning_rate": 0.00019124377648969346,
+      "loss": 3.037,
+      "step": 37836
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7006127238273621,
+      "learning_rate": 0.00019122942775755134,
+      "loss": 2.9233,
+      "step": 37837
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6066510677337646,
+      "learning_rate": 0.00019121507931188993,
+      "loss": 3.0044,
+      "step": 37838
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6588680744171143,
+      "learning_rate": 0.00019120073115274716,
+      "loss": 2.8884,
+      "step": 37839
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6584458351135254,
+      "learning_rate": 0.00019118638328016055,
+      "loss": 3.0469,
+      "step": 37840
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6320998668670654,
+      "learning_rate": 0.00019117203569416814,
+      "loss": 3.0939,
+      "step": 37841
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.699286162853241,
+      "learning_rate": 0.00019115768839480772,
+      "loss": 2.8838,
+      "step": 37842
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6417528986930847,
+      "learning_rate": 0.00019114334138211692,
+      "loss": 2.9171,
+      "step": 37843
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.614177942276001,
+      "learning_rate": 0.00019112899465613365,
+      "loss": 2.8298,
+      "step": 37844
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6882364153862,
+      "learning_rate": 0.00019111464821689565,
+      "loss": 2.9626,
+      "step": 37845
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6353859305381775,
+      "learning_rate": 0.0001911003020644406,
+      "loss": 2.9776,
+      "step": 37846
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6176462769508362,
+      "learning_rate": 0.00019108595619880657,
+      "loss": 3.1251,
+      "step": 37847
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6521666049957275,
+      "learning_rate": 0.00019107161062003106,
+      "loss": 2.8021,
+      "step": 37848
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6447446346282959,
+      "learning_rate": 0.0001910572653281521,
+      "loss": 3.2144,
+      "step": 37849
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6715746521949768,
+      "learning_rate": 0.00019104292032320712,
+      "loss": 3.0234,
+      "step": 37850
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6217811107635498,
+      "learning_rate": 0.0001910285756052342,
+      "loss": 2.7758,
+      "step": 37851
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6977801322937012,
+      "learning_rate": 0.0001910142311742711,
+      "loss": 2.9053,
+      "step": 37852
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6272152066230774,
+      "learning_rate": 0.0001909998870303555,
+      "loss": 3.087,
+      "step": 37853
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6529848575592041,
+      "learning_rate": 0.0001909855431735253,
+      "loss": 3.0135,
+      "step": 37854
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6610116362571716,
+      "learning_rate": 0.00019097119960381805,
+      "loss": 2.8011,
+      "step": 37855
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6474286913871765,
+      "learning_rate": 0.00019095685632127172,
+      "loss": 3.1778,
+      "step": 37856
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7384703755378723,
+      "learning_rate": 0.00019094251332592418,
+      "loss": 3.0711,
+      "step": 37857
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.668545126914978,
+      "learning_rate": 0.0001909281706178129,
+      "loss": 2.834,
+      "step": 37858
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.640242874622345,
+      "learning_rate": 0.00019091382819697597,
+      "loss": 3.1228,
+      "step": 37859
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6718035936355591,
+      "learning_rate": 0.00019089948606345095,
+      "loss": 2.866,
+      "step": 37860
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6487429141998291,
+      "learning_rate": 0.00019088514421727572,
+      "loss": 2.8989,
+      "step": 37861
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6288765668869019,
+      "learning_rate": 0.00019087080265848795,
+      "loss": 2.8211,
+      "step": 37862
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.644413411617279,
+      "learning_rate": 0.0001908564613871255,
+      "loss": 3.0838,
+      "step": 37863
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6302248239517212,
+      "learning_rate": 0.00019084212040322622,
+      "loss": 2.9669,
+      "step": 37864
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6651326417922974,
+      "learning_rate": 0.00019082777970682763,
+      "loss": 2.8824,
+      "step": 37865
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.631876528263092,
+      "learning_rate": 0.0001908134392979678,
+      "loss": 3.0457,
+      "step": 37866
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7018389105796814,
+      "learning_rate": 0.00019079909917668432,
+      "loss": 3.0692,
+      "step": 37867
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6403344869613647,
+      "learning_rate": 0.00019078475934301492,
+      "loss": 2.8391,
+      "step": 37868
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6221745014190674,
+      "learning_rate": 0.00019077041979699753,
+      "loss": 3.002,
+      "step": 37869
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6578125357627869,
+      "learning_rate": 0.00019075608053866974,
+      "loss": 3.0516,
+      "step": 37870
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6671814918518066,
+      "learning_rate": 0.00019074174156806956,
+      "loss": 2.9886,
+      "step": 37871
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6524838805198669,
+      "learning_rate": 0.00019072740288523452,
+      "loss": 2.9684,
+      "step": 37872
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6703388690948486,
+      "learning_rate": 0.00019071306449020236,
+      "loss": 2.9369,
+      "step": 37873
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6536246538162231,
+      "learning_rate": 0.00019069872638301116,
+      "loss": 3.1842,
+      "step": 37874
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6407559514045715,
+      "learning_rate": 0.00019068438856369838,
+      "loss": 3.1175,
+      "step": 37875
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6341151595115662,
+      "learning_rate": 0.00019067005103230196,
+      "loss": 3.2322,
+      "step": 37876
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6281828880310059,
+      "learning_rate": 0.00019065571378885942,
+      "loss": 2.9174,
+      "step": 37877
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6292392611503601,
+      "learning_rate": 0.0001906413768334088,
+      "loss": 2.999,
+      "step": 37878
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7081571221351624,
+      "learning_rate": 0.00019062704016598777,
+      "loss": 3.2075,
+      "step": 37879
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6541411876678467,
+      "learning_rate": 0.00019061270378663395,
+      "loss": 3.3458,
+      "step": 37880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6368046402931213,
+      "learning_rate": 0.00019059836769538534,
+      "loss": 2.9417,
+      "step": 37881
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6417937874794006,
+      "learning_rate": 0.00019058403189227952,
+      "loss": 3.0076,
+      "step": 37882
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6951587200164795,
+      "learning_rate": 0.00019056969637735422,
+      "loss": 2.9381,
+      "step": 37883
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6338977813720703,
+      "learning_rate": 0.00019055536115064747,
+      "loss": 3.0194,
+      "step": 37884
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6625250577926636,
+      "learning_rate": 0.0001905410262121967,
+      "loss": 3.0054,
+      "step": 37885
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6405186653137207,
+      "learning_rate": 0.0001905266915620399,
+      "loss": 3.0492,
+      "step": 37886
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.644585132598877,
+      "learning_rate": 0.0001905123572002146,
+      "loss": 2.828,
+      "step": 37887
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6336441040039062,
+      "learning_rate": 0.00019049802312675872,
+      "loss": 2.9381,
+      "step": 37888
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6756220459938049,
+      "learning_rate": 0.00019048368934171004,
+      "loss": 2.8513,
+      "step": 37889
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6149240136146545,
+      "learning_rate": 0.00019046935584510614,
+      "loss": 2.9183,
+      "step": 37890
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6757703423500061,
+      "learning_rate": 0.00019045502263698497,
+      "loss": 2.9603,
+      "step": 37891
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6618746519088745,
+      "learning_rate": 0.0001904406897173841,
+      "loss": 2.9956,
+      "step": 37892
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6471590995788574,
+      "learning_rate": 0.0001904263570863414,
+      "loss": 2.9915,
+      "step": 37893
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6565477252006531,
+      "learning_rate": 0.00019041202474389467,
+      "loss": 2.9342,
+      "step": 37894
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6662431955337524,
+      "learning_rate": 0.00019039769269008148,
+      "loss": 2.9803,
+      "step": 37895
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6561885476112366,
+      "learning_rate": 0.00019038336092493975,
+      "loss": 3.0549,
+      "step": 37896
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6919810771942139,
+      "learning_rate": 0.0001903690294485071,
+      "loss": 3.1325,
+      "step": 37897
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6764747500419617,
+      "learning_rate": 0.00019035469826082126,
+      "loss": 2.9849,
+      "step": 37898
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6264052987098694,
+      "learning_rate": 0.00019034036736192018,
+      "loss": 2.9931,
+      "step": 37899
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.652153491973877,
+      "learning_rate": 0.00019032603675184142,
+      "loss": 2.9467,
+      "step": 37900
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6279745101928711,
+      "learning_rate": 0.00019031170643062282,
+      "loss": 3.1106,
+      "step": 37901
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6377081871032715,
+      "learning_rate": 0.00019029737639830193,
+      "loss": 2.8446,
+      "step": 37902
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6817088723182678,
+      "learning_rate": 0.0001902830466549167,
+      "loss": 2.9539,
+      "step": 37903
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6729699969291687,
+      "learning_rate": 0.0001902687172005049,
+      "loss": 3.0809,
+      "step": 37904
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.680040180683136,
+      "learning_rate": 0.00019025438803510405,
+      "loss": 2.9381,
+      "step": 37905
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6663704514503479,
+      "learning_rate": 0.00019024005915875214,
+      "loss": 3.1179,
+      "step": 37906
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6278336048126221,
+      "learning_rate": 0.00019022573057148675,
+      "loss": 2.857,
+      "step": 37907
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6568112373352051,
+      "learning_rate": 0.0001902114022733456,
+      "loss": 3.1337,
+      "step": 37908
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6545429229736328,
+      "learning_rate": 0.00019019707426436654,
+      "loss": 3.0168,
+      "step": 37909
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6433941125869751,
+      "learning_rate": 0.0001901827465445872,
+      "loss": 3.0279,
+      "step": 37910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6414997577667236,
+      "learning_rate": 0.00019016841911404555,
+      "loss": 3.0132,
+      "step": 37911
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6626715660095215,
+      "learning_rate": 0.00019015409197277892,
+      "loss": 2.9186,
+      "step": 37912
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6758489608764648,
+      "learning_rate": 0.00019013976512082532,
+      "loss": 2.8539,
+      "step": 37913
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.661674439907074,
+      "learning_rate": 0.00019012543855822253,
+      "loss": 3.1409,
+      "step": 37914
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6630668640136719,
+      "learning_rate": 0.00019011111228500814,
+      "loss": 3.066,
+      "step": 37915
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6342833638191223,
+      "learning_rate": 0.00019009678630121997,
+      "loss": 2.8859,
+      "step": 37916
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6333215832710266,
+      "learning_rate": 0.00019008246060689562,
+      "loss": 3.1166,
+      "step": 37917
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6316998600959778,
+      "learning_rate": 0.00019006813520207295,
+      "loss": 2.9157,
+      "step": 37918
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6606517434120178,
+      "learning_rate": 0.00019005381008678973,
+      "loss": 2.9766,
+      "step": 37919
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6711786985397339,
+      "learning_rate": 0.0001900394852610835,
+      "loss": 3.1096,
+      "step": 37920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7660440802574158,
+      "learning_rate": 0.00019002516072499223,
+      "loss": 2.9022,
+      "step": 37921
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7094885110855103,
+      "learning_rate": 0.00019001083647855347,
+      "loss": 2.9963,
+      "step": 37922
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6385478377342224,
+      "learning_rate": 0.00018999651252180486,
+      "loss": 3.0875,
+      "step": 37923
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.632442831993103,
+      "learning_rate": 0.00018998218885478446,
+      "loss": 3.0835,
+      "step": 37924
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6296939849853516,
+      "learning_rate": 0.00018996786547752974,
+      "loss": 3.0832,
+      "step": 37925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6628495454788208,
+      "learning_rate": 0.00018995354239007852,
+      "loss": 3.208,
+      "step": 37926
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6310528516769409,
+      "learning_rate": 0.0001899392195924684,
+      "loss": 3.0514,
+      "step": 37927
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6705329418182373,
+      "learning_rate": 0.00018992489708473723,
+      "loss": 2.9223,
+      "step": 37928
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6532049179077148,
+      "learning_rate": 0.00018991057486692276,
+      "loss": 2.8727,
+      "step": 37929
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6473034024238586,
+      "learning_rate": 0.00018989625293906252,
+      "loss": 2.9542,
+      "step": 37930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7011779546737671,
+      "learning_rate": 0.00018988193130119448,
+      "loss": 3.082,
+      "step": 37931
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6515766382217407,
+      "learning_rate": 0.00018986760995335614,
+      "loss": 2.863,
+      "step": 37932
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6576693654060364,
+      "learning_rate": 0.0001898532888955854,
+      "loss": 2.8862,
+      "step": 37933
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6526912450790405,
+      "learning_rate": 0.0001898389681279199,
+      "loss": 2.8443,
+      "step": 37934
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6956529021263123,
+      "learning_rate": 0.00018982464765039726,
+      "loss": 3.0819,
+      "step": 37935
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6562744975090027,
+      "learning_rate": 0.00018981032746305542,
+      "loss": 2.8812,
+      "step": 37936
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6449558138847351,
+      "learning_rate": 0.00018979600756593193,
+      "loss": 2.9919,
+      "step": 37937
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6720443367958069,
+      "learning_rate": 0.00018978168795906458,
+      "loss": 3.1166,
+      "step": 37938
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6303071975708008,
+      "learning_rate": 0.00018976736864249094,
+      "loss": 3.0804,
+      "step": 37939
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6402961015701294,
+      "learning_rate": 0.00018975304961624893,
+      "loss": 2.8876,
+      "step": 37940
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.654398500919342,
+      "learning_rate": 0.00018973873088037616,
+      "loss": 2.8619,
+      "step": 37941
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6452508568763733,
+      "learning_rate": 0.00018972441243491028,
+      "loss": 2.9692,
+      "step": 37942
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6764039397239685,
+      "learning_rate": 0.00018971009427988918,
+      "loss": 3.1798,
+      "step": 37943
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6741904616355896,
+      "learning_rate": 0.00018969577641535042,
+      "loss": 3.0523,
+      "step": 37944
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6469334959983826,
+      "learning_rate": 0.00018968145884133169,
+      "loss": 3.0406,
+      "step": 37945
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.678054690361023,
+      "learning_rate": 0.0001896671415578709,
+      "loss": 3.0538,
+      "step": 37946
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.678111732006073,
+      "learning_rate": 0.00018965282456500553,
+      "loss": 3.0945,
+      "step": 37947
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6882033348083496,
+      "learning_rate": 0.00018963850786277346,
+      "loss": 3.1515,
+      "step": 37948
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6521106958389282,
+      "learning_rate": 0.0001896241914512122,
+      "loss": 2.9909,
+      "step": 37949
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6649858355522156,
+      "learning_rate": 0.00018960987533035963,
+      "loss": 2.954,
+      "step": 37950
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6313856244087219,
+      "learning_rate": 0.0001895955595002535,
+      "loss": 3.0676,
+      "step": 37951
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6400474309921265,
+      "learning_rate": 0.0001895812439609313,
+      "loss": 3.0938,
+      "step": 37952
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6551691293716431,
+      "learning_rate": 0.0001895669287124309,
+      "loss": 2.9153,
+      "step": 37953
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6542170643806458,
+      "learning_rate": 0.0001895526137547899,
+      "loss": 3.0044,
+      "step": 37954
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6334107518196106,
+      "learning_rate": 0.00018953829908804608,
+      "loss": 2.9579,
+      "step": 37955
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6488938927650452,
+      "learning_rate": 0.0001895239847122372,
+      "loss": 3.2192,
+      "step": 37956
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7389078736305237,
+      "learning_rate": 0.00018950967062740076,
+      "loss": 2.7784,
+      "step": 37957
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6441724300384521,
+      "learning_rate": 0.0001894953568335747,
+      "loss": 3.2414,
+      "step": 37958
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6733617186546326,
+      "learning_rate": 0.00018948104333079658,
+      "loss": 3.0559,
+      "step": 37959
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6783096790313721,
+      "learning_rate": 0.00018946673011910404,
+      "loss": 2.9956,
+      "step": 37960
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6289350390434265,
+      "learning_rate": 0.00018945241719853497,
+      "loss": 3.144,
+      "step": 37961
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7053776383399963,
+      "learning_rate": 0.0001894381045691269,
+      "loss": 3.1361,
+      "step": 37962
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6618072986602783,
+      "learning_rate": 0.00018942379223091765,
+      "loss": 2.9112,
+      "step": 37963
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6476094722747803,
+      "learning_rate": 0.00018940948018394472,
+      "loss": 3.1336,
+      "step": 37964
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6161331534385681,
+      "learning_rate": 0.00018939516842824599,
+      "loss": 3.2245,
+      "step": 37965
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6259023547172546,
+      "learning_rate": 0.00018938085696385917,
+      "loss": 2.9133,
+      "step": 37966
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6382219195365906,
+      "learning_rate": 0.00018936654579082175,
+      "loss": 2.9625,
+      "step": 37967
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6519684791564941,
+      "learning_rate": 0.0001893522349091717,
+      "loss": 3.0695,
+      "step": 37968
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7257705926895142,
+      "learning_rate": 0.0001893379243189465,
+      "loss": 3.1289,
+      "step": 37969
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7212842702865601,
+      "learning_rate": 0.00018932361402018388,
+      "loss": 3.1242,
+      "step": 37970
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6503854990005493,
+      "learning_rate": 0.00018930930401292163,
+      "loss": 2.8469,
+      "step": 37971
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6265800595283508,
+      "learning_rate": 0.00018929499429719726,
+      "loss": 2.973,
+      "step": 37972
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6678257584571838,
+      "learning_rate": 0.0001892806848730488,
+      "loss": 3.1497,
+      "step": 37973
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6247530579566956,
+      "learning_rate": 0.00018926637574051346,
+      "loss": 2.8594,
+      "step": 37974
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7319062352180481,
+      "learning_rate": 0.0001892520668996292,
+      "loss": 3.0092,
+      "step": 37975
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6562508940696716,
+      "learning_rate": 0.00018923775835043383,
+      "loss": 2.9743,
+      "step": 37976
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6403142213821411,
+      "learning_rate": 0.00018922345009296477,
+      "loss": 2.8766,
+      "step": 37977
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6423630714416504,
+      "learning_rate": 0.00018920914212725994,
+      "loss": 2.9736,
+      "step": 37978
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6349278688430786,
+      "learning_rate": 0.00018919483445335676,
+      "loss": 3.0937,
+      "step": 37979
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6362356543540955,
+      "learning_rate": 0.00018918052707129313,
+      "loss": 2.9766,
+      "step": 37980
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.642997145652771,
+      "learning_rate": 0.0001891662199811067,
+      "loss": 2.864,
+      "step": 37981
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6604081988334656,
+      "learning_rate": 0.00018915191318283502,
+      "loss": 2.9531,
+      "step": 37982
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6715200543403625,
+      "learning_rate": 0.00018913760667651602,
+      "loss": 3.0857,
+      "step": 37983
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6149404644966125,
+      "learning_rate": 0.0001891233004621871,
+      "loss": 3.145,
+      "step": 37984
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6347184181213379,
+      "learning_rate": 0.00018910899453988603,
+      "loss": 2.9672,
+      "step": 37985
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6498271822929382,
+      "learning_rate": 0.00018909468890965068,
+      "loss": 3.1004,
+      "step": 37986
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6263914108276367,
+      "learning_rate": 0.00018908038357151849,
+      "loss": 2.935,
+      "step": 37987
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6811741590499878,
+      "learning_rate": 0.00018906607852552728,
+      "loss": 2.8293,
+      "step": 37988
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6744228601455688,
+      "learning_rate": 0.00018905177377171457,
+      "loss": 3.1593,
+      "step": 37989
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6605095267295837,
+      "learning_rate": 0.00018903746931011816,
+      "loss": 3.1381,
+      "step": 37990
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.625415563583374,
+      "learning_rate": 0.00018902316514077578,
+      "loss": 3.067,
+      "step": 37991
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6394377946853638,
+      "learning_rate": 0.00018900886126372492,
+      "loss": 3.0368,
+      "step": 37992
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6430702209472656,
+      "learning_rate": 0.00018899455767900345,
+      "loss": 3.0768,
+      "step": 37993
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7407238483428955,
+      "learning_rate": 0.00018898025438664884,
+      "loss": 2.9264,
+      "step": 37994
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6624389290809631,
+      "learning_rate": 0.00018896595138669895,
+      "loss": 2.9439,
+      "step": 37995
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.649860143661499,
+      "learning_rate": 0.00018895164867919142,
+      "loss": 3.0471,
+      "step": 37996
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.652580201625824,
+      "learning_rate": 0.00018893734626416374,
+      "loss": 3.1846,
+      "step": 37997
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6489927172660828,
+      "learning_rate": 0.00018892304414165387,
+      "loss": 3.0071,
+      "step": 37998
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6793820858001709,
+      "learning_rate": 0.00018890874231169925,
+      "loss": 2.9431,
+      "step": 37999
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6238475441932678,
+      "learning_rate": 0.00018889444077433757,
+      "loss": 3.0595,
+      "step": 38000
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6435956954956055,
+      "learning_rate": 0.00018888013952960667,
+      "loss": 2.8716,
+      "step": 38001
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6251559257507324,
+      "learning_rate": 0.00018886583857754405,
+      "loss": 3.0129,
+      "step": 38002
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6116903424263,
+      "learning_rate": 0.00018885153791818742,
+      "loss": 2.9931,
+      "step": 38003
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6142878532409668,
+      "learning_rate": 0.0001888372375515744,
+      "loss": 2.9865,
+      "step": 38004
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6752248406410217,
+      "learning_rate": 0.00018882293747774276,
+      "loss": 2.8574,
+      "step": 38005
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6765214204788208,
+      "learning_rate": 0.0001888086376967301,
+      "loss": 3.0034,
+      "step": 38006
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7171432375907898,
+      "learning_rate": 0.00018879433820857402,
+      "loss": 2.8449,
+      "step": 38007
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6618970632553101,
+      "learning_rate": 0.00018878003901331235,
+      "loss": 3.0472,
+      "step": 38008
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6702781915664673,
+      "learning_rate": 0.00018876574011098262,
+      "loss": 3.0298,
+      "step": 38009
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6527828574180603,
+      "learning_rate": 0.00018875144150162245,
+      "loss": 2.8084,
+      "step": 38010
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6515904068946838,
+      "learning_rate": 0.00018873714318526968,
+      "loss": 3.1088,
+      "step": 38011
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6282628774642944,
+      "learning_rate": 0.00018872284516196182,
+      "loss": 3.0204,
+      "step": 38012
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6256953477859497,
+      "learning_rate": 0.0001887085474317366,
+      "loss": 3.097,
+      "step": 38013
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7299720048904419,
+      "learning_rate": 0.00018869424999463154,
+      "loss": 2.9651,
+      "step": 38014
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6656356453895569,
+      "learning_rate": 0.00018867995285068442,
+      "loss": 3.061,
+      "step": 38015
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7072047591209412,
+      "learning_rate": 0.000188665655999933,
+      "loss": 2.9438,
+      "step": 38016
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6499008536338806,
+      "learning_rate": 0.00018865135944241476,
+      "loss": 3.1369,
+      "step": 38017
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6523368954658508,
+      "learning_rate": 0.00018863706317816747,
+      "loss": 3.0557,
+      "step": 38018
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6852509379386902,
+      "learning_rate": 0.0001886227672072286,
+      "loss": 2.9922,
+      "step": 38019
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.695910632610321,
+      "learning_rate": 0.00018860847152963602,
+      "loss": 2.8368,
+      "step": 38020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.687065839767456,
+      "learning_rate": 0.00018859417614542725,
+      "loss": 2.9775,
+      "step": 38021
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6046010851860046,
+      "learning_rate": 0.00018857988105463991,
+      "loss": 2.9377,
+      "step": 38022
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6309628486633301,
+      "learning_rate": 0.00018856558625731187,
+      "loss": 3.0841,
+      "step": 38023
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6558736562728882,
+      "learning_rate": 0.00018855129175348052,
+      "loss": 2.9364,
+      "step": 38024
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6964730620384216,
+      "learning_rate": 0.0001885369975431837,
+      "loss": 3.0563,
+      "step": 38025
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.722326934337616,
+      "learning_rate": 0.00018852270362645885,
+      "loss": 3.048,
+      "step": 38026
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.626826286315918,
+      "learning_rate": 0.00018850841000334382,
+      "loss": 2.9608,
+      "step": 38027
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.656711995601654,
+      "learning_rate": 0.00018849411667387625,
+      "loss": 3.0312,
+      "step": 38028
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6362355351448059,
+      "learning_rate": 0.0001884798236380936,
+      "loss": 2.9385,
+      "step": 38029
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7519611716270447,
+      "learning_rate": 0.00018846553089603372,
+      "loss": 2.8248,
+      "step": 38030
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6189745664596558,
+      "learning_rate": 0.00018845123844773412,
+      "loss": 2.846,
+      "step": 38031
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6334044337272644,
+      "learning_rate": 0.00018843694629323244,
+      "loss": 2.9264,
+      "step": 38032
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6451401114463806,
+      "learning_rate": 0.0001884226544325665,
+      "loss": 2.9789,
+      "step": 38033
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6401827931404114,
+      "learning_rate": 0.00018840836286577366,
+      "loss": 3.155,
+      "step": 38034
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6569207906723022,
+      "learning_rate": 0.00018839407159289192,
+      "loss": 3.1407,
+      "step": 38035
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6262386441230774,
+      "learning_rate": 0.00018837978061395856,
+      "loss": 2.9666,
+      "step": 38036
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6198394894599915,
+      "learning_rate": 0.0001883654899290113,
+      "loss": 3.0311,
+      "step": 38037
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6523306369781494,
+      "learning_rate": 0.00018835119953808805,
+      "loss": 2.9531,
+      "step": 38038
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6170300245285034,
+      "learning_rate": 0.00018833690944122615,
+      "loss": 3.0921,
+      "step": 38039
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6404187679290771,
+      "learning_rate": 0.00018832261963846344,
+      "loss": 3.2716,
+      "step": 38040
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6466928124427795,
+      "learning_rate": 0.00018830833012983727,
+      "loss": 2.9561,
+      "step": 38041
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6404629349708557,
+      "learning_rate": 0.0001882940409153856,
+      "loss": 3.045,
+      "step": 38042
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6768361926078796,
+      "learning_rate": 0.00018827975199514597,
+      "loss": 3.0527,
+      "step": 38043
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6463939547538757,
+      "learning_rate": 0.0001882654633691558,
+      "loss": 3.1942,
+      "step": 38044
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6092776656150818,
+      "learning_rate": 0.00018825117503745308,
+      "loss": 3.0076,
+      "step": 38045
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6425269246101379,
+      "learning_rate": 0.00018823688700007516,
+      "loss": 2.8472,
+      "step": 38046
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6583645939826965,
+      "learning_rate": 0.00018822259925705973,
+      "loss": 2.9726,
+      "step": 38047
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6546190977096558,
+      "learning_rate": 0.00018820831180844455,
+      "loss": 3.0767,
+      "step": 38048
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6298822164535522,
+      "learning_rate": 0.00018819402465426716,
+      "loss": 2.9643,
+      "step": 38049
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6522963047027588,
+      "learning_rate": 0.0001881797377945652,
+      "loss": 3.0874,
+      "step": 38050
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6804884672164917,
+      "learning_rate": 0.0001881654512293762,
+      "loss": 3.01,
+      "step": 38051
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6772292852401733,
+      "learning_rate": 0.00018815116495873792,
+      "loss": 3.0163,
+      "step": 38052
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6438421607017517,
+      "learning_rate": 0.00018813687898268798,
+      "loss": 3.2455,
+      "step": 38053
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6451230645179749,
+      "learning_rate": 0.00018812259330126392,
+      "loss": 2.9777,
+      "step": 38054
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6543946862220764,
+      "learning_rate": 0.0001881083079145035,
+      "loss": 2.9877,
+      "step": 38055
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6809148192405701,
+      "learning_rate": 0.00018809402282244414,
+      "loss": 3.0758,
+      "step": 38056
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.644823431968689,
+      "learning_rate": 0.00018807973802512367,
+      "loss": 2.9312,
+      "step": 38057
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6681671142578125,
+      "learning_rate": 0.00018806545352257963,
+      "loss": 3.0616,
+      "step": 38058
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6262026429176331,
+      "learning_rate": 0.0001880511693148496,
+      "loss": 2.9225,
+      "step": 38059
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6308271884918213,
+      "learning_rate": 0.00018803688540197132,
+      "loss": 3.0833,
+      "step": 38060
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6894946098327637,
+      "learning_rate": 0.00018802260178398228,
+      "loss": 2.9456,
+      "step": 38061
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6232718825340271,
+      "learning_rate": 0.0001880083184609201,
+      "loss": 2.8134,
+      "step": 38062
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6857800483703613,
+      "learning_rate": 0.00018799403543282259,
+      "loss": 2.9333,
+      "step": 38063
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6674224734306335,
+      "learning_rate": 0.00018797975269972718,
+      "loss": 3.0611,
+      "step": 38064
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6705251336097717,
+      "learning_rate": 0.00018796547026167157,
+      "loss": 3.1249,
+      "step": 38065
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6390047073364258,
+      "learning_rate": 0.00018795118811869323,
+      "loss": 3.0813,
+      "step": 38066
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6156495809555054,
+      "learning_rate": 0.00018793690627082999,
+      "loss": 2.861,
+      "step": 38067
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6561599373817444,
+      "learning_rate": 0.00018792262471811942,
+      "loss": 2.9476,
+      "step": 38068
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6761384606361389,
+      "learning_rate": 0.00018790834346059894,
+      "loss": 3.1729,
+      "step": 38069
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6904240846633911,
+      "learning_rate": 0.00018789406249830647,
+      "loss": 2.988,
+      "step": 38070
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6512880325317383,
+      "learning_rate": 0.00018787978183127939,
+      "loss": 3.1239,
+      "step": 38071
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6401209831237793,
+      "learning_rate": 0.00018786550145955534,
+      "loss": 3.0033,
+      "step": 38072
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6807664632797241,
+      "learning_rate": 0.00018785122138317208,
+      "loss": 2.7812,
+      "step": 38073
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6754370927810669,
+      "learning_rate": 0.00018783694160216701,
+      "loss": 2.8129,
+      "step": 38074
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6359378695487976,
+      "learning_rate": 0.000187822662116578,
+      "loss": 3.2968,
+      "step": 38075
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6342616081237793,
+      "learning_rate": 0.00018780838292644235,
+      "loss": 2.9636,
+      "step": 38076
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6342082023620605,
+      "learning_rate": 0.00018779410403179782,
+      "loss": 3.1626,
+      "step": 38077
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6407181024551392,
+      "learning_rate": 0.00018777982543268215,
+      "loss": 3.0483,
+      "step": 38078
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6541846394538879,
+      "learning_rate": 0.0001877655471291328,
+      "loss": 3.0294,
+      "step": 38079
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6525475978851318,
+      "learning_rate": 0.0001877512691211874,
+      "loss": 2.8933,
+      "step": 38080
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6654950976371765,
+      "learning_rate": 0.00018773699140888346,
+      "loss": 3.043,
+      "step": 38081
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.684727668762207,
+      "learning_rate": 0.00018772271399225877,
+      "loss": 3.1511,
+      "step": 38082
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6460987329483032,
+      "learning_rate": 0.00018770843687135083,
+      "loss": 3.2699,
+      "step": 38083
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6055113673210144,
+      "learning_rate": 0.00018769416004619718,
+      "loss": 2.9939,
+      "step": 38084
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6769329309463501,
+      "learning_rate": 0.00018767988351683562,
+      "loss": 2.9545,
+      "step": 38085
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6439892053604126,
+      "learning_rate": 0.00018766560728330359,
+      "loss": 3.0184,
+      "step": 38086
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6827326416969299,
+      "learning_rate": 0.00018765133134563864,
+      "loss": 3.0605,
+      "step": 38087
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6329249739646912,
+      "learning_rate": 0.0001876370557038786,
+      "loss": 3.1776,
+      "step": 38088
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6486325263977051,
+      "learning_rate": 0.00018762278035806083,
+      "loss": 3.0524,
+      "step": 38089
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6577807664871216,
+      "learning_rate": 0.00018760850530822313,
+      "loss": 2.9419,
+      "step": 38090
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6746330857276917,
+      "learning_rate": 0.00018759423055440285,
+      "loss": 2.9862,
+      "step": 38091
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6533883213996887,
+      "learning_rate": 0.00018757995609663782,
+      "loss": 3.0289,
+      "step": 38092
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6736960411071777,
+      "learning_rate": 0.00018756568193496558,
+      "loss": 3.0924,
+      "step": 38093
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6582047343254089,
+      "learning_rate": 0.0001875514080694236,
+      "loss": 2.9104,
+      "step": 38094
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6109486818313599,
+      "learning_rate": 0.00018753713450004968,
+      "loss": 2.941,
+      "step": 38095
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6732144355773926,
+      "learning_rate": 0.0001875228612268812,
+      "loss": 3.1589,
+      "step": 38096
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7052947282791138,
+      "learning_rate": 0.00018750858824995598,
+      "loss": 2.7537,
+      "step": 38097
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6800371408462524,
+      "learning_rate": 0.00018749431556931142,
+      "loss": 3.0416,
+      "step": 38098
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.675338864326477,
+      "learning_rate": 0.00018748004318498512,
+      "loss": 2.8279,
+      "step": 38099
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6601358652114868,
+      "learning_rate": 0.00018746577109701482,
+      "loss": 3.0966,
+      "step": 38100
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6854960322380066,
+      "learning_rate": 0.000187451499305438,
+      "loss": 3.2466,
+      "step": 38101
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6765010356903076,
+      "learning_rate": 0.00018743722781029235,
+      "loss": 3.161,
+      "step": 38102
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6960597634315491,
+      "learning_rate": 0.0001874229566116152,
+      "loss": 3.2019,
+      "step": 38103
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6276587247848511,
+      "learning_rate": 0.0001874086857094444,
+      "loss": 3.1574,
+      "step": 38104
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6926313042640686,
+      "learning_rate": 0.00018739441510381752,
+      "loss": 2.9851,
+      "step": 38105
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6597981452941895,
+      "learning_rate": 0.00018738014479477197,
+      "loss": 2.8751,
+      "step": 38106
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6823752522468567,
+      "learning_rate": 0.00018736587478234554,
+      "loss": 3.0197,
+      "step": 38107
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6578922271728516,
+      "learning_rate": 0.00018735160506657566,
+      "loss": 2.9142,
+      "step": 38108
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7423891425132751,
+      "learning_rate": 0.0001873373356474999,
+      "loss": 2.882,
+      "step": 38109
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6776301264762878,
+      "learning_rate": 0.00018732306652515608,
+      "loss": 2.981,
+      "step": 38110
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6616842150688171,
+      "learning_rate": 0.0001873087976995815,
+      "loss": 2.7471,
+      "step": 38111
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.706602156162262,
+      "learning_rate": 0.000187294529170814,
+      "loss": 3.0262,
+      "step": 38112
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6833293437957764,
+      "learning_rate": 0.00018728026093889084,
+      "loss": 3.0769,
+      "step": 38113
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6360612511634827,
+      "learning_rate": 0.00018726599300384985,
+      "loss": 2.8711,
+      "step": 38114
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6546128392219543,
+      "learning_rate": 0.0001872517253657286,
+      "loss": 3.115,
+      "step": 38115
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.614295482635498,
+      "learning_rate": 0.00018723745802456447,
+      "loss": 2.9583,
+      "step": 38116
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6379879713058472,
+      "learning_rate": 0.00018722319098039532,
+      "loss": 3.0077,
+      "step": 38117
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6464399099349976,
+      "learning_rate": 0.00018720892423325846,
+      "loss": 2.9822,
+      "step": 38118
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6336324214935303,
+      "learning_rate": 0.00018719465778319161,
+      "loss": 3.0851,
+      "step": 38119
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6779120564460754,
+      "learning_rate": 0.00018718039163023246,
+      "loss": 2.8303,
+      "step": 38120
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6924359202384949,
+      "learning_rate": 0.00018716612577441828,
+      "loss": 3.1107,
+      "step": 38121
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.643219530582428,
+      "learning_rate": 0.0001871518602157869,
+      "loss": 3.1419,
+      "step": 38122
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6407486200332642,
+      "learning_rate": 0.00018713759495437576,
+      "loss": 2.9626,
+      "step": 38123
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6377759575843811,
+      "learning_rate": 0.00018712332999022246,
+      "loss": 3.0036,
+      "step": 38124
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6537527441978455,
+      "learning_rate": 0.00018710906532336466,
+      "loss": 3.0181,
+      "step": 38125
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6885370016098022,
+      "learning_rate": 0.00018709480095383977,
+      "loss": 2.8599,
+      "step": 38126
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6301379203796387,
+      "learning_rate": 0.00018708053688168554,
+      "loss": 2.8084,
+      "step": 38127
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6435237526893616,
+      "learning_rate": 0.00018706627310693933,
+      "loss": 3.1439,
+      "step": 38128
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6685776114463806,
+      "learning_rate": 0.00018705200962963885,
+      "loss": 2.9715,
+      "step": 38129
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6370143294334412,
+      "learning_rate": 0.00018703774644982177,
+      "loss": 3.0221,
+      "step": 38130
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6487584710121155,
+      "learning_rate": 0.00018702348356752535,
+      "loss": 2.9997,
+      "step": 38131
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6534705758094788,
+      "learning_rate": 0.00018700922098278746,
+      "loss": 2.8816,
+      "step": 38132
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6294593214988708,
+      "learning_rate": 0.00018699495869564547,
+      "loss": 2.9037,
+      "step": 38133
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8882706761360168,
+      "learning_rate": 0.00018698069670613696,
+      "loss": 2.9112,
+      "step": 38134
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6950933933258057,
+      "learning_rate": 0.00018696643501429967,
+      "loss": 2.9052,
+      "step": 38135
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6648854613304138,
+      "learning_rate": 0.0001869521736201709,
+      "loss": 2.9956,
+      "step": 38136
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.712222695350647,
+      "learning_rate": 0.00018693791252378855,
+      "loss": 3.2136,
+      "step": 38137
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6462759375572205,
+      "learning_rate": 0.0001869236517251898,
+      "loss": 3.0022,
+      "step": 38138
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6655752658843994,
+      "learning_rate": 0.00018690939122441237,
+      "loss": 2.9743,
+      "step": 38139
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6417601108551025,
+      "learning_rate": 0.00018689513102149394,
+      "loss": 3.051,
+      "step": 38140
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6326993107795715,
+      "learning_rate": 0.00018688087111647195,
+      "loss": 3.0341,
+      "step": 38141
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.651273787021637,
+      "learning_rate": 0.00018686661150938403,
+      "loss": 3.1457,
+      "step": 38142
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.668491005897522,
+      "learning_rate": 0.00018685235220026756,
+      "loss": 2.8735,
+      "step": 38143
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6334226727485657,
+      "learning_rate": 0.00018683809318916023,
+      "loss": 2.9159,
+      "step": 38144
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6658082008361816,
+      "learning_rate": 0.0001868238344760997,
+      "loss": 2.978,
+      "step": 38145
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6330041289329529,
+      "learning_rate": 0.00018680957606112325,
+      "loss": 2.9432,
+      "step": 38146
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6996323466300964,
+      "learning_rate": 0.00018679531794426875,
+      "loss": 3.1075,
+      "step": 38147
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6674297451972961,
+      "learning_rate": 0.00018678106012557353,
+      "loss": 3.0515,
+      "step": 38148
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.632754921913147,
+      "learning_rate": 0.00018676680260507511,
+      "loss": 2.9696,
+      "step": 38149
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6195089221000671,
+      "learning_rate": 0.0001867525453828113,
+      "loss": 3.0214,
+      "step": 38150
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6678367853164673,
+      "learning_rate": 0.00018673828845881943,
+      "loss": 2.8939,
+      "step": 38151
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6400881409645081,
+      "learning_rate": 0.00018672403183313716,
+      "loss": 3.1865,
+      "step": 38152
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6747406125068665,
+      "learning_rate": 0.00018670977550580185,
+      "loss": 3.1998,
+      "step": 38153
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6130285263061523,
+      "learning_rate": 0.00018669551947685128,
+      "loss": 3.1601,
+      "step": 38154
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6921128034591675,
+      "learning_rate": 0.00018668126374632294,
+      "loss": 2.9525,
+      "step": 38155
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6503788232803345,
+      "learning_rate": 0.00018666700831425424,
+      "loss": 3.0248,
+      "step": 38156
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.624090850353241,
+      "learning_rate": 0.00018665275318068296,
+      "loss": 3.0273,
+      "step": 38157
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6640616059303284,
+      "learning_rate": 0.00018663849834564634,
+      "loss": 3.284,
+      "step": 38158
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7166267037391663,
+      "learning_rate": 0.00018662424380918221,
+      "loss": 2.9603,
+      "step": 38159
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6211531758308411,
+      "learning_rate": 0.00018660998957132805,
+      "loss": 3.0816,
+      "step": 38160
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6357093453407288,
+      "learning_rate": 0.00018659573563212123,
+      "loss": 3.1217,
+      "step": 38161
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6868173480033875,
+      "learning_rate": 0.00018658148199159953,
+      "loss": 2.9255,
+      "step": 38162
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6988732218742371,
+      "learning_rate": 0.00018656722864980035,
+      "loss": 2.914,
+      "step": 38163
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.620403528213501,
+      "learning_rate": 0.0001865529756067612,
+      "loss": 3.0708,
+      "step": 38164
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7112835049629211,
+      "learning_rate": 0.00018653872286251974,
+      "loss": 3.1756,
+      "step": 38165
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6396851539611816,
+      "learning_rate": 0.00018652447041711344,
+      "loss": 2.86,
+      "step": 38166
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6468586325645447,
+      "learning_rate": 0.00018651021827057992,
+      "loss": 2.9201,
+      "step": 38167
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6509426832199097,
+      "learning_rate": 0.00018649596642295645,
+      "loss": 2.9801,
+      "step": 38168
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.684259295463562,
+      "learning_rate": 0.00018648171487428087,
+      "loss": 3.1006,
+      "step": 38169
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6615915298461914,
+      "learning_rate": 0.00018646746362459065,
+      "loss": 3.0024,
+      "step": 38170
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6847988963127136,
+      "learning_rate": 0.0001864532126739232,
+      "loss": 2.9606,
+      "step": 38171
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6112357974052429,
+      "learning_rate": 0.00018643896202231623,
+      "loss": 3.182,
+      "step": 38172
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6447160243988037,
+      "learning_rate": 0.00018642471166980707,
+      "loss": 3.0568,
+      "step": 38173
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6257251501083374,
+      "learning_rate": 0.00018641046161643353,
+      "loss": 3.1314,
+      "step": 38174
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6292673349380493,
+      "learning_rate": 0.00018639621186223278,
+      "loss": 2.9662,
+      "step": 38175
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6028404235839844,
+      "learning_rate": 0.0001863819624072426,
+      "loss": 3.1069,
+      "step": 38176
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6941775679588318,
+      "learning_rate": 0.00018636771325150056,
+      "loss": 2.98,
+      "step": 38177
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6567956805229187,
+      "learning_rate": 0.00018635346439504395,
+      "loss": 2.9761,
+      "step": 38178
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6520336866378784,
+      "learning_rate": 0.00018633921583791055,
+      "loss": 3.1373,
+      "step": 38179
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6364136338233948,
+      "learning_rate": 0.00018632496758013768,
+      "loss": 3.017,
+      "step": 38180
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6324988603591919,
+      "learning_rate": 0.00018631071962176303,
+      "loss": 2.9674,
+      "step": 38181
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6260587573051453,
+      "learning_rate": 0.00018629647196282412,
+      "loss": 2.8975,
+      "step": 38182
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6402279138565063,
+      "learning_rate": 0.0001862822246033583,
+      "loss": 3.0804,
+      "step": 38183
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6180815100669861,
+      "learning_rate": 0.00018626797754340332,
+      "loss": 2.917,
+      "step": 38184
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6797022819519043,
+      "learning_rate": 0.00018625373078299658,
+      "loss": 2.9086,
+      "step": 38185
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6699610352516174,
+      "learning_rate": 0.00018623948432217554,
+      "loss": 3.0638,
+      "step": 38186
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6808142066001892,
+      "learning_rate": 0.0001862252381609779,
+      "loss": 2.9938,
+      "step": 38187
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6589908599853516,
+      "learning_rate": 0.00018621099229944107,
+      "loss": 3.1829,
+      "step": 38188
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6513395309448242,
+      "learning_rate": 0.00018619674673760262,
+      "loss": 2.9098,
+      "step": 38189
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6903216242790222,
+      "learning_rate": 0.00018618250147549995,
+      "loss": 2.9158,
+      "step": 38190
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.71368807554245,
+      "learning_rate": 0.0001861682565131707,
+      "loss": 3.026,
+      "step": 38191
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.659199595451355,
+      "learning_rate": 0.00018615401185065238,
+      "loss": 2.8882,
+      "step": 38192
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6380458474159241,
+      "learning_rate": 0.0001861397674879824,
+      "loss": 3.1365,
+      "step": 38193
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6615623235702515,
+      "learning_rate": 0.00018612552342519849,
+      "loss": 2.7819,
+      "step": 38194
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6508620977401733,
+      "learning_rate": 0.00018611127966233797,
+      "loss": 2.9649,
+      "step": 38195
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6376355290412903,
+      "learning_rate": 0.00018609703619943833,
+      "loss": 2.9156,
+      "step": 38196
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6251353621482849,
+      "learning_rate": 0.00018608279303653733,
+      "loss": 3.0193,
+      "step": 38197
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6305531859397888,
+      "learning_rate": 0.00018606855017367223,
+      "loss": 2.8546,
+      "step": 38198
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6258016228675842,
+      "learning_rate": 0.00018605430761088073,
+      "loss": 2.9225,
+      "step": 38199
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6111253499984741,
+      "learning_rate": 0.00018604006534820016,
+      "loss": 2.9206,
+      "step": 38200
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6775639057159424,
+      "learning_rate": 0.0001860258233856681,
+      "loss": 2.9885,
+      "step": 38201
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6494293212890625,
+      "learning_rate": 0.0001860115817233222,
+      "loss": 3.0052,
+      "step": 38202
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6443254947662354,
+      "learning_rate": 0.0001859973403611998,
+      "loss": 3.0249,
+      "step": 38203
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5837622880935669,
+      "learning_rate": 0.00018598309929933853,
+      "loss": 2.9753,
+      "step": 38204
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6752540469169617,
+      "learning_rate": 0.00018596885853777574,
+      "loss": 2.9707,
+      "step": 38205
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6625667810440063,
+      "learning_rate": 0.00018595461807654906,
+      "loss": 3.0735,
+      "step": 38206
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6508365273475647,
+      "learning_rate": 0.00018594037791569602,
+      "loss": 3.2523,
+      "step": 38207
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6375356316566467,
+      "learning_rate": 0.00018592613805525396,
+      "loss": 3.043,
+      "step": 38208
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6329519152641296,
+      "learning_rate": 0.00018591189849526062,
+      "loss": 2.8927,
+      "step": 38209
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6348059177398682,
+      "learning_rate": 0.00018589765923575333,
+      "loss": 3.1013,
+      "step": 38210
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6728472709655762,
+      "learning_rate": 0.00018588342027676957,
+      "loss": 3.0327,
+      "step": 38211
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6233043074607849,
+      "learning_rate": 0.00018586918161834706,
+      "loss": 2.726,
+      "step": 38212
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7845079302787781,
+      "learning_rate": 0.00018585494326052307,
+      "loss": 2.8162,
+      "step": 38213
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.662473738193512,
+      "learning_rate": 0.00018584070520333528,
+      "loss": 2.9308,
+      "step": 38214
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6366502642631531,
+      "learning_rate": 0.000185826467446821,
+      "loss": 2.8987,
+      "step": 38215
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6556762456893921,
+      "learning_rate": 0.00018581222999101784,
+      "loss": 2.9589,
+      "step": 38216
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.626169741153717,
+      "learning_rate": 0.0001857979928359634,
+      "loss": 3.0476,
+      "step": 38217
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6835764050483704,
+      "learning_rate": 0.0001857837559816949,
+      "loss": 3.0164,
+      "step": 38218
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6800299286842346,
+      "learning_rate": 0.00018576951942825018,
+      "loss": 2.9386,
+      "step": 38219
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6377986073493958,
+      "learning_rate": 0.0001857552831756664,
+      "loss": 2.98,
+      "step": 38220
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6307637095451355,
+      "learning_rate": 0.0001857410472239813,
+      "loss": 2.9888,
+      "step": 38221
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6545106768608093,
+      "learning_rate": 0.00018572681157323233,
+      "loss": 3.1622,
+      "step": 38222
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6275971531867981,
+      "learning_rate": 0.00018571257622345685,
+      "loss": 3.1065,
+      "step": 38223
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6176348328590393,
+      "learning_rate": 0.00018569834117469254,
+      "loss": 2.9758,
+      "step": 38224
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6735263466835022,
+      "learning_rate": 0.00018568410642697678,
+      "loss": 2.9344,
+      "step": 38225
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6405964493751526,
+      "learning_rate": 0.00018566987198034696,
+      "loss": 3.0465,
+      "step": 38226
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6843787431716919,
+      "learning_rate": 0.0001856556378348409,
+      "loss": 2.8972,
+      "step": 38227
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.687495231628418,
+      "learning_rate": 0.00018564140399049574,
+      "loss": 3.133,
+      "step": 38228
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6218973994255066,
+      "learning_rate": 0.00018562717044734924,
+      "loss": 3.1439,
+      "step": 38229
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6834261417388916,
+      "learning_rate": 0.0001856129372054386,
+      "loss": 2.9368,
+      "step": 38230
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6899203658103943,
+      "learning_rate": 0.0001855987042648015,
+      "loss": 3.1648,
+      "step": 38231
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6193641424179077,
+      "learning_rate": 0.00018558447162547557,
+      "loss": 3.1351,
+      "step": 38232
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6698924899101257,
+      "learning_rate": 0.0001855702392874979,
+      "loss": 3.0177,
+      "step": 38233
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6577365398406982,
+      "learning_rate": 0.0001855560072509064,
+      "loss": 2.8135,
+      "step": 38234
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.705444872379303,
+      "learning_rate": 0.0001855417755157382,
+      "loss": 2.8679,
+      "step": 38235
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6559796929359436,
+      "learning_rate": 0.00018552754408203092,
+      "loss": 3.0659,
+      "step": 38236
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6978949904441833,
+      "learning_rate": 0.00018551331294982214,
+      "loss": 2.8357,
+      "step": 38237
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6330851316452026,
+      "learning_rate": 0.00018549908211914928,
+      "loss": 2.945,
+      "step": 38238
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6387531161308289,
+      "learning_rate": 0.00018548485159004978,
+      "loss": 3.0267,
+      "step": 38239
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6540210843086243,
+      "learning_rate": 0.00018547062136256107,
+      "loss": 3.0548,
+      "step": 38240
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6874194741249084,
+      "learning_rate": 0.00018545639143672068,
+      "loss": 2.8508,
+      "step": 38241
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6516175270080566,
+      "learning_rate": 0.00018544216181256624,
+      "loss": 3.0576,
+      "step": 38242
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6130465865135193,
+      "learning_rate": 0.00018542793249013505,
+      "loss": 2.9599,
+      "step": 38243
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6488755941390991,
+      "learning_rate": 0.00018541370346946467,
+      "loss": 3.0966,
+      "step": 38244
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6888542771339417,
+      "learning_rate": 0.00018539947475059245,
+      "loss": 2.9866,
+      "step": 38245
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7215191721916199,
+      "learning_rate": 0.000185385246333556,
+      "loss": 3.2059,
+      "step": 38246
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6577127575874329,
+      "learning_rate": 0.00018537101821839282,
+      "loss": 2.8346,
+      "step": 38247
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6458978056907654,
+      "learning_rate": 0.0001853567904051402,
+      "loss": 3.181,
+      "step": 38248
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6692994236946106,
+      "learning_rate": 0.00018534256289383583,
+      "loss": 2.9622,
+      "step": 38249
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6456918716430664,
+      "learning_rate": 0.00018532833568451706,
+      "loss": 3.1947,
+      "step": 38250
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6313451528549194,
+      "learning_rate": 0.00018531410877722142,
+      "loss": 3.005,
+      "step": 38251
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0270297527313232,
+      "learning_rate": 0.00018529988217198622,
+      "loss": 2.9805,
+      "step": 38252
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6870240569114685,
+      "learning_rate": 0.00018528565586884915,
+      "loss": 3.0719,
+      "step": 38253
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.631271481513977,
+      "learning_rate": 0.00018527142986784763,
+      "loss": 2.9909,
+      "step": 38254
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6817666292190552,
+      "learning_rate": 0.00018525720416901894,
+      "loss": 3.0439,
+      "step": 38255
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6389595866203308,
+      "learning_rate": 0.00018524297877240084,
+      "loss": 3.2042,
+      "step": 38256
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6673776507377625,
+      "learning_rate": 0.00018522875367803053,
+      "loss": 2.914,
+      "step": 38257
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6235078573226929,
+      "learning_rate": 0.0001852145288859456,
+      "loss": 2.8799,
+      "step": 38258
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6437073945999146,
+      "learning_rate": 0.00018520030439618364,
+      "loss": 3.0012,
+      "step": 38259
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.634657621383667,
+      "learning_rate": 0.00018518608020878184,
+      "loss": 2.9875,
+      "step": 38260
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6310805678367615,
+      "learning_rate": 0.00018517185632377787,
+      "loss": 3.0513,
+      "step": 38261
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6541125178337097,
+      "learning_rate": 0.00018515763274120912,
+      "loss": 3.0226,
+      "step": 38262
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6451280117034912,
+      "learning_rate": 0.00018514340946111302,
+      "loss": 2.9589,
+      "step": 38263
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6415655016899109,
+      "learning_rate": 0.00018512918648352718,
+      "loss": 3.0097,
+      "step": 38264
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6348758339881897,
+      "learning_rate": 0.0001851149638084889,
+      "loss": 2.8844,
+      "step": 38265
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6602434515953064,
+      "learning_rate": 0.00018510074143603575,
+      "loss": 2.8191,
+      "step": 38266
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6239683032035828,
+      "learning_rate": 0.00018508651936620504,
+      "loss": 2.9728,
+      "step": 38267
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7023791074752808,
+      "learning_rate": 0.00018507229759903436,
+      "loss": 2.9814,
+      "step": 38268
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7200928330421448,
+      "learning_rate": 0.00018505807613456116,
+      "loss": 2.8517,
+      "step": 38269
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6196789741516113,
+      "learning_rate": 0.00018504385497282281,
+      "loss": 3.0157,
+      "step": 38270
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6774440407752991,
+      "learning_rate": 0.00018502963411385688,
+      "loss": 2.9213,
+      "step": 38271
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6709611415863037,
+      "learning_rate": 0.00018501541355770074,
+      "loss": 3.0191,
+      "step": 38272
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6436893343925476,
+      "learning_rate": 0.00018500119330439182,
+      "loss": 3.0043,
+      "step": 38273
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6333373785018921,
+      "learning_rate": 0.00018498697335396773,
+      "loss": 2.9307,
+      "step": 38274
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6157318949699402,
+      "learning_rate": 0.00018497275370646575,
+      "loss": 3.2422,
+      "step": 38275
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6362408995628357,
+      "learning_rate": 0.00018495853436192348,
+      "loss": 2.9209,
+      "step": 38276
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6697286367416382,
+      "learning_rate": 0.00018494431532037816,
+      "loss": 2.9718,
+      "step": 38277
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6749752759933472,
+      "learning_rate": 0.00018493009658186743,
+      "loss": 2.9866,
+      "step": 38278
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6495433449745178,
+      "learning_rate": 0.00018491587814642874,
+      "loss": 3.0554,
+      "step": 38279
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6848012208938599,
+      "learning_rate": 0.00018490166001409935,
+      "loss": 3.0164,
+      "step": 38280
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7173035740852356,
+      "learning_rate": 0.00018488744218491694,
+      "loss": 3.077,
+      "step": 38281
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6794202327728271,
+      "learning_rate": 0.00018487322465891875,
+      "loss": 3.0307,
+      "step": 38282
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.640831470489502,
+      "learning_rate": 0.00018485900743614244,
+      "loss": 2.995,
+      "step": 38283
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6711544394493103,
+      "learning_rate": 0.00018484479051662534,
+      "loss": 2.9166,
+      "step": 38284
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.654302179813385,
+      "learning_rate": 0.00018483057390040482,
+      "loss": 3.2323,
+      "step": 38285
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6835788488388062,
+      "learning_rate": 0.00018481635758751852,
+      "loss": 3.0623,
+      "step": 38286
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6237308382987976,
+      "learning_rate": 0.00018480214157800368,
+      "loss": 3.0551,
+      "step": 38287
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.652237594127655,
+      "learning_rate": 0.00018478792587189774,
+      "loss": 2.9883,
+      "step": 38288
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6164937019348145,
+      "learning_rate": 0.0001847737104692384,
+      "loss": 3.1875,
+      "step": 38289
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6360146403312683,
+      "learning_rate": 0.0001847594953700629,
+      "loss": 3.3418,
+      "step": 38290
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.639178991317749,
+      "learning_rate": 0.0001847452805744087,
+      "loss": 2.9557,
+      "step": 38291
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6623160243034363,
+      "learning_rate": 0.0001847310660823132,
+      "loss": 2.8778,
+      "step": 38292
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7035412788391113,
+      "learning_rate": 0.0001847168518938139,
+      "loss": 3.1368,
+      "step": 38293
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6533422470092773,
+      "learning_rate": 0.00018470263800894827,
+      "loss": 3.0838,
+      "step": 38294
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7027701139450073,
+      "learning_rate": 0.00018468842442775362,
+      "loss": 2.8786,
+      "step": 38295
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6815171241760254,
+      "learning_rate": 0.0001846742111502676,
+      "loss": 2.8839,
+      "step": 38296
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6483194828033447,
+      "learning_rate": 0.0001846599981765274,
+      "loss": 2.8239,
+      "step": 38297
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6512202024459839,
+      "learning_rate": 0.00018464578550657053,
+      "loss": 2.9617,
+      "step": 38298
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6246964335441589,
+      "learning_rate": 0.00018463157314043462,
+      "loss": 2.7436,
+      "step": 38299
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6432050466537476,
+      "learning_rate": 0.00018461736107815675,
+      "loss": 3.1272,
+      "step": 38300
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7117769718170166,
+      "learning_rate": 0.00018460314931977481,
+      "loss": 2.9599,
+      "step": 38301
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6452391743659973,
+      "learning_rate": 0.0001845889378653257,
+      "loss": 3.0305,
+      "step": 38302
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6579294800758362,
+      "learning_rate": 0.00018457472671484717,
+      "loss": 3.053,
+      "step": 38303
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5954685211181641,
+      "learning_rate": 0.00018456051586837672,
+      "loss": 2.9531,
+      "step": 38304
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.673682451248169,
+      "learning_rate": 0.00018454630532595154,
+      "loss": 2.987,
+      "step": 38305
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6650246381759644,
+      "learning_rate": 0.0001845320950876093,
+      "loss": 2.9635,
+      "step": 38306
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6644670367240906,
+      "learning_rate": 0.00018451788515338715,
+      "loss": 3.0036,
+      "step": 38307
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6150769591331482,
+      "learning_rate": 0.00018450367552332272,
+      "loss": 3.0552,
+      "step": 38308
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6673305630683899,
+      "learning_rate": 0.00018448946619745342,
+      "loss": 2.8861,
+      "step": 38309
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6792820692062378,
+      "learning_rate": 0.00018447525717581656,
+      "loss": 3.0457,
+      "step": 38310
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6672223210334778,
+      "learning_rate": 0.00018446104845844976,
+      "loss": 2.9729,
+      "step": 38311
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6425515413284302,
+      "learning_rate": 0.0001844468400453902,
+      "loss": 2.9787,
+      "step": 38312
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6276784539222717,
+      "learning_rate": 0.00018443263193667543,
+      "loss": 2.833,
+      "step": 38313
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6771143674850464,
+      "learning_rate": 0.00018441842413234296,
+      "loss": 2.79,
+      "step": 38314
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6400710940361023,
+      "learning_rate": 0.00018440421663243009,
+      "loss": 3.1405,
+      "step": 38315
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6542527079582214,
+      "learning_rate": 0.00018439000943697427,
+      "loss": 2.8716,
+      "step": 38316
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.648661196231842,
+      "learning_rate": 0.0001843758025460128,
+      "loss": 2.9464,
+      "step": 38317
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6391183733940125,
+      "learning_rate": 0.00018436159595958332,
+      "loss": 3.1304,
+      "step": 38318
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6920697093009949,
+      "learning_rate": 0.0001843473896777232,
+      "loss": 2.9166,
+      "step": 38319
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6402799487113953,
+      "learning_rate": 0.00018433318370046964,
+      "loss": 2.9813,
+      "step": 38320
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6674555540084839,
+      "learning_rate": 0.00018431897802786035,
+      "loss": 3.056,
+      "step": 38321
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6335968971252441,
+      "learning_rate": 0.00018430477265993247,
+      "loss": 3.0924,
+      "step": 38322
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6576682329177856,
+      "learning_rate": 0.00018429056759672362,
+      "loss": 2.7713,
+      "step": 38323
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6932410001754761,
+      "learning_rate": 0.00018427636283827124,
+      "loss": 3.0169,
+      "step": 38324
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6399856209754944,
+      "learning_rate": 0.0001842621583846125,
+      "loss": 3.008,
+      "step": 38325
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.65628582239151,
+      "learning_rate": 0.0001842479542357851,
+      "loss": 3.279,
+      "step": 38326
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6356684565544128,
+      "learning_rate": 0.00018423375039182624,
+      "loss": 2.8128,
+      "step": 38327
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6769824028015137,
+      "learning_rate": 0.00018421954685277347,
+      "loss": 2.8816,
+      "step": 38328
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6468422412872314,
+      "learning_rate": 0.00018420534361866396,
+      "loss": 2.7391,
+      "step": 38329
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.653346598148346,
+      "learning_rate": 0.0001841911406895354,
+      "loss": 3.1066,
+      "step": 38330
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.671597957611084,
+      "learning_rate": 0.00018417693806542512,
+      "loss": 2.9741,
+      "step": 38331
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6430724859237671,
+      "learning_rate": 0.00018416273574637043,
+      "loss": 3.0607,
+      "step": 38332
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6749132871627808,
+      "learning_rate": 0.00018414853373240888,
+      "loss": 3.0446,
+      "step": 38333
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6767664551734924,
+      "learning_rate": 0.00018413433202357774,
+      "loss": 2.8677,
+      "step": 38334
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.633886992931366,
+      "learning_rate": 0.0001841201306199144,
+      "loss": 2.7959,
+      "step": 38335
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6434829831123352,
+      "learning_rate": 0.00018410592952145647,
+      "loss": 2.8927,
+      "step": 38336
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6419342160224915,
+      "learning_rate": 0.00018409172872824116,
+      "loss": 3.1004,
+      "step": 38337
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6574919819831848,
+      "learning_rate": 0.00018407752824030598,
+      "loss": 3.0997,
+      "step": 38338
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6425691246986389,
+      "learning_rate": 0.00018406332805768813,
+      "loss": 2.7971,
+      "step": 38339
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.677888810634613,
+      "learning_rate": 0.0001840491281804253,
+      "loss": 2.7808,
+      "step": 38340
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6419331431388855,
+      "learning_rate": 0.00018403492860855474,
+      "loss": 2.9517,
+      "step": 38341
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6205721497535706,
+      "learning_rate": 0.00018402072934211377,
+      "loss": 3.0624,
+      "step": 38342
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6749025583267212,
+      "learning_rate": 0.00018400653038114,
+      "loss": 2.9666,
+      "step": 38343
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6735125184059143,
+      "learning_rate": 0.00018399233172567057,
+      "loss": 2.8711,
+      "step": 38344
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.640764594078064,
+      "learning_rate": 0.00018397813337574307,
+      "loss": 3.1523,
+      "step": 38345
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6429493427276611,
+      "learning_rate": 0.00018396393533139492,
+      "loss": 2.983,
+      "step": 38346
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6589473485946655,
+      "learning_rate": 0.0001839497375926633,
+      "loss": 2.9291,
+      "step": 38347
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6674602031707764,
+      "learning_rate": 0.00018393554015958588,
+      "loss": 3.0551,
+      "step": 38348
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6474887132644653,
+      "learning_rate": 0.00018392134303219984,
+      "loss": 2.9849,
+      "step": 38349
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6512155532836914,
+      "learning_rate": 0.00018390714621054253,
+      "loss": 3.0877,
+      "step": 38350
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6206179261207581,
+      "learning_rate": 0.00018389294969465165,
+      "loss": 3.1422,
+      "step": 38351
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6520265340805054,
+      "learning_rate": 0.00018387875348456427,
+      "loss": 2.8446,
+      "step": 38352
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6797086596488953,
+      "learning_rate": 0.000183864557580318,
+      "loss": 3.1531,
+      "step": 38353
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6597195863723755,
+      "learning_rate": 0.00018385036198195,
+      "loss": 2.9861,
+      "step": 38354
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6018645167350769,
+      "learning_rate": 0.0001838361666894979,
+      "loss": 3.0342,
+      "step": 38355
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6335331797599792,
+      "learning_rate": 0.000183821971702999,
+      "loss": 2.9257,
+      "step": 38356
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6828946471214294,
+      "learning_rate": 0.0001838077770224905,
+      "loss": 2.9478,
+      "step": 38357
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6436010599136353,
+      "learning_rate": 0.00018379358264801018,
+      "loss": 3.0097,
+      "step": 38358
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6501964926719666,
+      "learning_rate": 0.00018377938857959509,
+      "loss": 2.8804,
+      "step": 38359
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6490870118141174,
+      "learning_rate": 0.00018376519481728263,
+      "loss": 2.8949,
+      "step": 38360
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6395012140274048,
+      "learning_rate": 0.00018375100136111043,
+      "loss": 3.0246,
+      "step": 38361
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6163665056228638,
+      "learning_rate": 0.00018373680821111562,
+      "loss": 3.0793,
+      "step": 38362
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.635441243648529,
+      "learning_rate": 0.00018372261536733582,
+      "loss": 3.1323,
+      "step": 38363
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6642354726791382,
+      "learning_rate": 0.0001837084228298081,
+      "loss": 2.957,
+      "step": 38364
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.666587769985199,
+      "learning_rate": 0.00018369423059857,
+      "loss": 3.0932,
+      "step": 38365
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6363824605941772,
+      "learning_rate": 0.00018368003867365909,
+      "loss": 2.7929,
+      "step": 38366
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6824597716331482,
+      "learning_rate": 0.0001836658470551125,
+      "loss": 3.0304,
+      "step": 38367
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6351178288459778,
+      "learning_rate": 0.00018365165574296766,
+      "loss": 3.0697,
+      "step": 38368
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6279469728469849,
+      "learning_rate": 0.00018363746473726194,
+      "loss": 2.9615,
+      "step": 38369
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6463478803634644,
+      "learning_rate": 0.00018362327403803277,
+      "loss": 3.1272,
+      "step": 38370
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6736152768135071,
+      "learning_rate": 0.00018360908364531756,
+      "loss": 3.0822,
+      "step": 38371
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.627798318862915,
+      "learning_rate": 0.0001835948935591535,
+      "loss": 3.2404,
+      "step": 38372
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6646186709403992,
+      "learning_rate": 0.00018358070377957824,
+      "loss": 2.8242,
+      "step": 38373
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6356680393218994,
+      "learning_rate": 0.00018356651430662888,
+      "loss": 2.7888,
+      "step": 38374
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6446818709373474,
+      "learning_rate": 0.0001835523251403429,
+      "loss": 2.9223,
+      "step": 38375
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7403597235679626,
+      "learning_rate": 0.00018353813628075775,
+      "loss": 3.102,
+      "step": 38376
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6128467321395874,
+      "learning_rate": 0.00018352394772791071,
+      "loss": 2.79,
+      "step": 38377
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6641044020652771,
+      "learning_rate": 0.00018350975948183926,
+      "loss": 2.8962,
+      "step": 38378
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6116800308227539,
+      "learning_rate": 0.00018349557154258056,
+      "loss": 3.0184,
+      "step": 38379
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6737449765205383,
+      "learning_rate": 0.00018348138391017214,
+      "loss": 3.0201,
+      "step": 38380
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6416964530944824,
+      "learning_rate": 0.0001834671965846514,
+      "loss": 3.0682,
+      "step": 38381
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6830700635910034,
+      "learning_rate": 0.00018345300956605553,
+      "loss": 2.9621,
+      "step": 38382
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6257073283195496,
+      "learning_rate": 0.0001834388228544221,
+      "loss": 3.1164,
+      "step": 38383
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6321331858634949,
+      "learning_rate": 0.00018342463644978825,
+      "loss": 3.0862,
+      "step": 38384
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6541038155555725,
+      "learning_rate": 0.00018341045035219153,
+      "loss": 3.0973,
+      "step": 38385
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.691582202911377,
+      "learning_rate": 0.00018339626456166934,
+      "loss": 3.0136,
+      "step": 38386
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6166252493858337,
+      "learning_rate": 0.00018338207907825878,
+      "loss": 3.0924,
+      "step": 38387
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6885400414466858,
+      "learning_rate": 0.00018336789390199757,
+      "loss": 3.0155,
+      "step": 38388
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6976376175880432,
+      "learning_rate": 0.00018335370903292277,
+      "loss": 3.1868,
+      "step": 38389
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6312760710716248,
+      "learning_rate": 0.0001833395244710718,
+      "loss": 3.015,
+      "step": 38390
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6591302752494812,
+      "learning_rate": 0.0001833253402164822,
+      "loss": 2.9672,
+      "step": 38391
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.647793710231781,
+      "learning_rate": 0.00018331115626919113,
+      "loss": 2.8549,
+      "step": 38392
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6505529284477234,
+      "learning_rate": 0.0001832969726292361,
+      "loss": 2.9445,
+      "step": 38393
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7024715542793274,
+      "learning_rate": 0.00018328278929665424,
+      "loss": 3.0583,
+      "step": 38394
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6506431698799133,
+      "learning_rate": 0.00018326860627148312,
+      "loss": 2.8853,
+      "step": 38395
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6874439120292664,
+      "learning_rate": 0.0001832544235537601,
+      "loss": 3.0919,
+      "step": 38396
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6847971081733704,
+      "learning_rate": 0.00018324024114352227,
+      "loss": 3.0551,
+      "step": 38397
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6514188647270203,
+      "learning_rate": 0.00018322605904080738,
+      "loss": 3.0029,
+      "step": 38398
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6254274249076843,
+      "learning_rate": 0.00018321187724565244,
+      "loss": 2.9289,
+      "step": 38399
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7166518568992615,
+      "learning_rate": 0.00018319769575809498,
+      "loss": 3.158,
+      "step": 38400
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6519727110862732,
+      "learning_rate": 0.00018318351457817235,
+      "loss": 3.1565,
+      "step": 38401
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6201913356781006,
+      "learning_rate": 0.00018316933370592182,
+      "loss": 2.9268,
+      "step": 38402
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6431904435157776,
+      "learning_rate": 0.00018315515314138085,
+      "loss": 2.8744,
+      "step": 38403
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6314598917961121,
+      "learning_rate": 0.0001831409728845866,
+      "loss": 3.0051,
+      "step": 38404
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6558628082275391,
+      "learning_rate": 0.0001831267929355765,
+      "loss": 2.7542,
+      "step": 38405
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6240569353103638,
+      "learning_rate": 0.0001831126132943881,
+      "loss": 3.0997,
+      "step": 38406
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6257813572883606,
+      "learning_rate": 0.00018309843396105852,
+      "loss": 2.8727,
+      "step": 38407
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6663781404495239,
+      "learning_rate": 0.0001830842549356252,
+      "loss": 2.9911,
+      "step": 38408
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.634472131729126,
+      "learning_rate": 0.00018307007621812534,
+      "loss": 3.0498,
+      "step": 38409
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6562722325325012,
+      "learning_rate": 0.0001830558978085965,
+      "loss": 2.6464,
+      "step": 38410
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6011884212493896,
+      "learning_rate": 0.0001830417197070759,
+      "loss": 2.8752,
+      "step": 38411
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6544702053070068,
+      "learning_rate": 0.00018302754191360082,
+      "loss": 2.94,
+      "step": 38412
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6202430129051208,
+      "learning_rate": 0.00018301336442820882,
+      "loss": 3.1568,
+      "step": 38413
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6908934712409973,
+      "learning_rate": 0.00018299918725093702,
+      "loss": 3.1212,
+      "step": 38414
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6435255408287048,
+      "learning_rate": 0.0001829850103818229,
+      "loss": 2.9659,
+      "step": 38415
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6239068508148193,
+      "learning_rate": 0.00018297083382090364,
+      "loss": 2.8303,
+      "step": 38416
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6315820813179016,
+      "learning_rate": 0.00018295665756821674,
+      "loss": 3.0459,
+      "step": 38417
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6654682755470276,
+      "learning_rate": 0.00018294248162379953,
+      "loss": 3.1875,
+      "step": 38418
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6455845832824707,
+      "learning_rate": 0.00018292830598768916,
+      "loss": 3.2678,
+      "step": 38419
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6427332758903503,
+      "learning_rate": 0.00018291413065992326,
+      "loss": 2.9188,
+      "step": 38420
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6541956067085266,
+      "learning_rate": 0.0001828999556405389,
+      "loss": 3.0159,
+      "step": 38421
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6762649416923523,
+      "learning_rate": 0.0001828857809295735,
+      "loss": 3.1659,
+      "step": 38422
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6119546890258789,
+      "learning_rate": 0.00018287160652706455,
+      "loss": 2.9543,
+      "step": 38423
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6221927404403687,
+      "learning_rate": 0.00018285743243304903,
+      "loss": 2.8137,
+      "step": 38424
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6315422654151917,
+      "learning_rate": 0.00018284325864756468,
+      "loss": 3.1488,
+      "step": 38425
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6791592240333557,
+      "learning_rate": 0.0001828290851706486,
+      "loss": 3.1033,
+      "step": 38426
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6015139222145081,
+      "learning_rate": 0.00018281491200233807,
+      "loss": 2.8296,
+      "step": 38427
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6524919867515564,
+      "learning_rate": 0.00018280073914267065,
+      "loss": 3.0343,
+      "step": 38428
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.663172721862793,
+      "learning_rate": 0.00018278656659168345,
+      "loss": 3.046,
+      "step": 38429
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.634985625743866,
+      "learning_rate": 0.0001827723943494139,
+      "loss": 3.1256,
+      "step": 38430
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6688196063041687,
+      "learning_rate": 0.00018275822241589927,
+      "loss": 2.8842,
+      "step": 38431
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6805686354637146,
+      "learning_rate": 0.0001827440507911769,
+      "loss": 2.8134,
+      "step": 38432
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6817598342895508,
+      "learning_rate": 0.00018272987947528423,
+      "loss": 3.2789,
+      "step": 38433
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6709921956062317,
+      "learning_rate": 0.00018271570846825833,
+      "loss": 3.0719,
+      "step": 38434
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6306078433990479,
+      "learning_rate": 0.00018270153777013684,
+      "loss": 2.9929,
+      "step": 38435
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6629069447517395,
+      "learning_rate": 0.00018268736738095688,
+      "loss": 3.0485,
+      "step": 38436
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6367194652557373,
+      "learning_rate": 0.00018267319730075573,
+      "loss": 2.9225,
+      "step": 38437
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6755772829055786,
+      "learning_rate": 0.0001826590275295709,
+      "loss": 2.9001,
+      "step": 38438
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6794211268424988,
+      "learning_rate": 0.0001826448580674396,
+      "loss": 3.0339,
+      "step": 38439
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7521235942840576,
+      "learning_rate": 0.0001826306889143992,
+      "loss": 2.8449,
+      "step": 38440
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6859071254730225,
+      "learning_rate": 0.0001826165200704868,
+      "loss": 3.0413,
+      "step": 38441
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6678243279457092,
+      "learning_rate": 0.00018260235153574,
+      "loss": 2.9641,
+      "step": 38442
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6535446643829346,
+      "learning_rate": 0.00018258818331019606,
+      "loss": 3.1897,
+      "step": 38443
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6140435338020325,
+      "learning_rate": 0.00018257401539389218,
+      "loss": 2.9456,
+      "step": 38444
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6568311452865601,
+      "learning_rate": 0.00018255984778686583,
+      "loss": 3.0298,
+      "step": 38445
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6603125333786011,
+      "learning_rate": 0.0001825456804891541,
+      "loss": 2.9342,
+      "step": 38446
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.623427152633667,
+      "learning_rate": 0.0001825315135007945,
+      "loss": 3.0201,
+      "step": 38447
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6774618029594421,
+      "learning_rate": 0.00018251734682182437,
+      "loss": 3.215,
+      "step": 38448
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6762425899505615,
+      "learning_rate": 0.00018250318045228081,
+      "loss": 2.904,
+      "step": 38449
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6597325205802917,
+      "learning_rate": 0.0001824890143922014,
+      "loss": 3.0122,
+      "step": 38450
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6491790413856506,
+      "learning_rate": 0.00018247484864162322,
+      "loss": 3.1313,
+      "step": 38451
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6365981101989746,
+      "learning_rate": 0.00018246068320058366,
+      "loss": 3.0639,
+      "step": 38452
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6332027912139893,
+      "learning_rate": 0.0001824465180691201,
+      "loss": 2.9355,
+      "step": 38453
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6683153510093689,
+      "learning_rate": 0.00018243235324726974,
+      "loss": 3.1984,
+      "step": 38454
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6562347412109375,
+      "learning_rate": 0.00018241818873507003,
+      "loss": 2.9301,
+      "step": 38455
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6307790279388428,
+      "learning_rate": 0.00018240402453255803,
+      "loss": 2.8323,
+      "step": 38456
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.626919686794281,
+      "learning_rate": 0.00018238986063977126,
+      "loss": 3.0347,
+      "step": 38457
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6405884027481079,
+      "learning_rate": 0.00018237569705674704,
+      "loss": 3.0111,
+      "step": 38458
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6130412817001343,
+      "learning_rate": 0.00018236153378352248,
+      "loss": 3.0065,
+      "step": 38459
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6970798969268799,
+      "learning_rate": 0.00018234737082013506,
+      "loss": 2.88,
+      "step": 38460
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6523925065994263,
+      "learning_rate": 0.00018233320816662206,
+      "loss": 2.9462,
+      "step": 38461
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6778101921081543,
+      "learning_rate": 0.0001823190458230206,
+      "loss": 3.0931,
+      "step": 38462
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6423189640045166,
+      "learning_rate": 0.00018230488378936826,
+      "loss": 3.1616,
+      "step": 38463
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6352372765541077,
+      "learning_rate": 0.00018229072206570212,
+      "loss": 2.9447,
+      "step": 38464
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.627549946308136,
+      "learning_rate": 0.0001822765606520597,
+      "loss": 3.098,
+      "step": 38465
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6376739740371704,
+      "learning_rate": 0.00018226239954847805,
+      "loss": 3.143,
+      "step": 38466
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7025358080863953,
+      "learning_rate": 0.00018224823875499451,
+      "loss": 3.028,
+      "step": 38467
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6670701503753662,
+      "learning_rate": 0.00018223407827164663,
+      "loss": 2.9701,
+      "step": 38468
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6625725626945496,
+      "learning_rate": 0.00018221991809847142,
+      "loss": 2.9864,
+      "step": 38469
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6725914478302002,
+      "learning_rate": 0.00018220575823550635,
+      "loss": 3.1172,
+      "step": 38470
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6515083909034729,
+      "learning_rate": 0.00018219159868278852,
+      "loss": 3.0334,
+      "step": 38471
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6303300261497498,
+      "learning_rate": 0.0001821774394403554,
+      "loss": 2.7711,
+      "step": 38472
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6697096228599548,
+      "learning_rate": 0.0001821632805082443,
+      "loss": 2.9665,
+      "step": 38473
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.657671332359314,
+      "learning_rate": 0.00018214912188649232,
+      "loss": 2.8763,
+      "step": 38474
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6618888974189758,
+      "learning_rate": 0.00018213496357513703,
+      "loss": 2.7808,
+      "step": 38475
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.652420163154602,
+      "learning_rate": 0.00018212080557421543,
+      "loss": 3.0118,
+      "step": 38476
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6736677289009094,
+      "learning_rate": 0.00018210664788376491,
+      "loss": 3.188,
+      "step": 38477
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6545586585998535,
+      "learning_rate": 0.0001820924905038229,
+      "loss": 3.0641,
+      "step": 38478
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6806058883666992,
+      "learning_rate": 0.00018207833343442657,
+      "loss": 2.9763,
+      "step": 38479
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6442785263061523,
+      "learning_rate": 0.00018206417667561322,
+      "loss": 3.1018,
+      "step": 38480
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6341136693954468,
+      "learning_rate": 0.00018205002022742006,
+      "loss": 2.9789,
+      "step": 38481
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6478147506713867,
+      "learning_rate": 0.00018203586408988446,
+      "loss": 3.0614,
+      "step": 38482
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6580110788345337,
+      "learning_rate": 0.00018202170826304374,
+      "loss": 2.8771,
+      "step": 38483
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6745438575744629,
+      "learning_rate": 0.00018200755274693506,
+      "loss": 3.0367,
+      "step": 38484
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6208056211471558,
+      "learning_rate": 0.0001819933975415959,
+      "loss": 3.0012,
+      "step": 38485
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6185520887374878,
+      "learning_rate": 0.00018197924264706325,
+      "loss": 2.8297,
+      "step": 38486
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6431528925895691,
+      "learning_rate": 0.00018196508806337476,
+      "loss": 3.266,
+      "step": 38487
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.660499095916748,
+      "learning_rate": 0.0001819509337905674,
+      "loss": 2.9249,
+      "step": 38488
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.636012077331543,
+      "learning_rate": 0.00018193677982867847,
+      "loss": 3.2491,
+      "step": 38489
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6246479153633118,
+      "learning_rate": 0.00018192262617774553,
+      "loss": 2.9983,
+      "step": 38490
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6834371089935303,
+      "learning_rate": 0.00018190847283780555,
+      "loss": 2.9373,
+      "step": 38491
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6637904047966003,
+      "learning_rate": 0.000181894319808896,
+      "loss": 2.8806,
+      "step": 38492
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6326305270195007,
+      "learning_rate": 0.000181880167091054,
+      "loss": 3.0709,
+      "step": 38493
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6314154267311096,
+      "learning_rate": 0.00018186601468431695,
+      "loss": 2.9915,
+      "step": 38494
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6132782697677612,
+      "learning_rate": 0.00018185186258872212,
+      "loss": 3.0361,
+      "step": 38495
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6412111520767212,
+      "learning_rate": 0.0001818377108043066,
+      "loss": 2.7906,
+      "step": 38496
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6674846410751343,
+      "learning_rate": 0.000181823559331108,
+      "loss": 2.857,
+      "step": 38497
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8719302415847778,
+      "learning_rate": 0.00018180940816916327,
+      "loss": 2.9331,
+      "step": 38498
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6447135210037231,
+      "learning_rate": 0.00018179525731850982,
+      "loss": 3.0817,
+      "step": 38499
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6724530458450317,
+      "learning_rate": 0.000181781106779185,
+      "loss": 2.9817,
+      "step": 38500
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6975722908973694,
+      "learning_rate": 0.00018176695655122596,
+      "loss": 3.1141,
+      "step": 38501
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6536849141120911,
+      "learning_rate": 0.00018175280663467003,
+      "loss": 2.867,
+      "step": 38502
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6566682457923889,
+      "learning_rate": 0.0001817386570295543,
+      "loss": 2.8168,
+      "step": 38503
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7100996971130371,
+      "learning_rate": 0.0001817245077359163,
+      "loss": 2.9156,
+      "step": 38504
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6142645478248596,
+      "learning_rate": 0.00018171035875379324,
+      "loss": 3.0425,
+      "step": 38505
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6454697251319885,
+      "learning_rate": 0.0001816962100832222,
+      "loss": 3.0476,
+      "step": 38506
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6202601790428162,
+      "learning_rate": 0.0001816820617242407,
+      "loss": 2.9302,
+      "step": 38507
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6279861927032471,
+      "learning_rate": 0.00018166791367688574,
+      "loss": 2.9165,
+      "step": 38508
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6671816110610962,
+      "learning_rate": 0.0001816537659411948,
+      "loss": 2.832,
+      "step": 38509
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6781206130981445,
+      "learning_rate": 0.00018163961851720513,
+      "loss": 2.9789,
+      "step": 38510
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6522396802902222,
+      "learning_rate": 0.0001816254714049538,
+      "loss": 2.8528,
+      "step": 38511
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.643043577671051,
+      "learning_rate": 0.0001816113246044783,
+      "loss": 2.8771,
+      "step": 38512
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6434921026229858,
+      "learning_rate": 0.00018159717811581575,
+      "loss": 2.9918,
+      "step": 38513
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6609666347503662,
+      "learning_rate": 0.00018158303193900336,
+      "loss": 3.013,
+      "step": 38514
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6703172326087952,
+      "learning_rate": 0.0001815688860740786,
+      "loss": 3.0498,
+      "step": 38515
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6868824362754822,
+      "learning_rate": 0.00018155474052107857,
+      "loss": 3.0622,
+      "step": 38516
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.632976770401001,
+      "learning_rate": 0.00018154059528004057,
+      "loss": 3.2264,
+      "step": 38517
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7155187129974365,
+      "learning_rate": 0.00018152645035100177,
+      "loss": 2.9952,
+      "step": 38518
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6054431796073914,
+      "learning_rate": 0.00018151230573399955,
+      "loss": 2.9829,
+      "step": 38519
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6408286690711975,
+      "learning_rate": 0.00018149816142907114,
+      "loss": 2.8755,
+      "step": 38520
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6466242074966431,
+      "learning_rate": 0.00018148401743625368,
+      "loss": 3.2028,
+      "step": 38521
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6467182636260986,
+      "learning_rate": 0.00018146987375558462,
+      "loss": 2.9694,
+      "step": 38522
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6525686979293823,
+      "learning_rate": 0.00018145573038710105,
+      "loss": 2.9897,
+      "step": 38523
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6323734521865845,
+      "learning_rate": 0.00018144158733084023,
+      "loss": 2.8444,
+      "step": 38524
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6349821090698242,
+      "learning_rate": 0.00018142744458683952,
+      "loss": 3.0805,
+      "step": 38525
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6364793181419373,
+      "learning_rate": 0.000181413302155136,
+      "loss": 2.8423,
+      "step": 38526
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6155493259429932,
+      "learning_rate": 0.00018139916003576722,
+      "loss": 3.0647,
+      "step": 38527
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.62147456407547,
+      "learning_rate": 0.00018138501822877003,
+      "loss": 2.977,
+      "step": 38528
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6351068615913391,
+      "learning_rate": 0.00018137087673418184,
+      "loss": 3.0379,
+      "step": 38529
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6419429779052734,
+      "learning_rate": 0.00018135673555204012,
+      "loss": 2.8855,
+      "step": 38530
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6373555660247803,
+      "learning_rate": 0.00018134259468238182,
+      "loss": 2.8341,
+      "step": 38531
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.687985360622406,
+      "learning_rate": 0.0001813284541252444,
+      "loss": 2.7618,
+      "step": 38532
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6934620141983032,
+      "learning_rate": 0.0001813143138806648,
+      "loss": 2.9882,
+      "step": 38533
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6637023687362671,
+      "learning_rate": 0.00018130017394868056,
+      "loss": 3.0607,
+      "step": 38534
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6849845051765442,
+      "learning_rate": 0.0001812860343293289,
+      "loss": 2.9292,
+      "step": 38535
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6725578308105469,
+      "learning_rate": 0.0001812718950226468,
+      "loss": 3.3626,
+      "step": 38536
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6605555415153503,
+      "learning_rate": 0.0001812577560286718,
+      "loss": 2.9055,
+      "step": 38537
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6455115079879761,
+      "learning_rate": 0.000181243617347441,
+      "loss": 2.9905,
+      "step": 38538
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6488058567047119,
+      "learning_rate": 0.0001812294789789916,
+      "loss": 2.8396,
+      "step": 38539
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6056329011917114,
+      "learning_rate": 0.00018121534092336097,
+      "loss": 2.8001,
+      "step": 38540
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6151305437088013,
+      "learning_rate": 0.00018120120318058624,
+      "loss": 3.2172,
+      "step": 38541
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6283738613128662,
+      "learning_rate": 0.00018118706575070474,
+      "loss": 2.85,
+      "step": 38542
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6253724098205566,
+      "learning_rate": 0.00018117292863375353,
+      "loss": 2.6845,
+      "step": 38543
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6234411597251892,
+      "learning_rate": 0.00018115879182977,
+      "loss": 3.0064,
+      "step": 38544
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6679108142852783,
+      "learning_rate": 0.00018114465533879142,
+      "loss": 2.8951,
+      "step": 38545
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6431871056556702,
+      "learning_rate": 0.0001811305191608548,
+      "loss": 2.9186,
+      "step": 38546
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6422107219696045,
+      "learning_rate": 0.00018111638329599763,
+      "loss": 2.9075,
+      "step": 38547
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6774196028709412,
+      "learning_rate": 0.00018110224774425696,
+      "loss": 2.915,
+      "step": 38548
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6358468532562256,
+      "learning_rate": 0.00018108811250567005,
+      "loss": 2.9972,
+      "step": 38549
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6383891701698303,
+      "learning_rate": 0.00018107397758027432,
+      "loss": 2.936,
+      "step": 38550
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6550706028938293,
+      "learning_rate": 0.00018105984296810668,
+      "loss": 3.0692,
+      "step": 38551
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6626957654953003,
+      "learning_rate": 0.00018104570866920466,
+      "loss": 3.0563,
+      "step": 38552
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6499350666999817,
+      "learning_rate": 0.00018103157468360526,
+      "loss": 3.0258,
+      "step": 38553
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6834374666213989,
+      "learning_rate": 0.00018101744101134577,
+      "loss": 2.8288,
+      "step": 38554
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5994073152542114,
+      "learning_rate": 0.00018100330765246356,
+      "loss": 3.1779,
+      "step": 38555
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.675519585609436,
+      "learning_rate": 0.0001809891746069957,
+      "loss": 2.8748,
+      "step": 38556
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.624893069267273,
+      "learning_rate": 0.00018097504187497944,
+      "loss": 3.2098,
+      "step": 38557
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.637237548828125,
+      "learning_rate": 0.0001809609094564519,
+      "loss": 2.9335,
+      "step": 38558
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6514628529548645,
+      "learning_rate": 0.00018094677735145057,
+      "loss": 3.2457,
+      "step": 38559
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6226351857185364,
+      "learning_rate": 0.0001809326455600125,
+      "loss": 2.851,
+      "step": 38560
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6673992276191711,
+      "learning_rate": 0.0001809185140821748,
+      "loss": 3.0208,
+      "step": 38561
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6154362559318542,
+      "learning_rate": 0.00018090438291797496,
+      "loss": 2.9796,
+      "step": 38562
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6860130429267883,
+      "learning_rate": 0.00018089025206744999,
+      "loss": 2.9236,
+      "step": 38563
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7032354474067688,
+      "learning_rate": 0.00018087612153063723,
+      "loss": 2.9412,
+      "step": 38564
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6549293994903564,
+      "learning_rate": 0.00018086199130757375,
+      "loss": 3.201,
+      "step": 38565
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6441744565963745,
+      "learning_rate": 0.00018084786139829686,
+      "loss": 2.9833,
+      "step": 38566
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6373230218887329,
+      "learning_rate": 0.00018083373180284386,
+      "loss": 2.9114,
+      "step": 38567
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7155922651290894,
+      "learning_rate": 0.00018081960252125178,
+      "loss": 2.8217,
+      "step": 38568
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6718721985816956,
+      "learning_rate": 0.00018080547355355802,
+      "loss": 2.8134,
+      "step": 38569
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7212534546852112,
+      "learning_rate": 0.00018079134489979958,
+      "loss": 2.8709,
+      "step": 38570
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6554144024848938,
+      "learning_rate": 0.0001807772165600139,
+      "loss": 2.8055,
+      "step": 38571
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6695471405982971,
+      "learning_rate": 0.0001807630885342381,
+      "loss": 2.8279,
+      "step": 38572
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.696575403213501,
+      "learning_rate": 0.00018074896082250928,
+      "loss": 3.1577,
+      "step": 38573
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6690512895584106,
+      "learning_rate": 0.00018073483342486481,
+      "loss": 3.0392,
+      "step": 38574
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6713618040084839,
+      "learning_rate": 0.00018072070634134185,
+      "loss": 3.108,
+      "step": 38575
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6604920625686646,
+      "learning_rate": 0.00018070657957197753,
+      "loss": 2.8892,
+      "step": 38576
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6354912519454956,
+      "learning_rate": 0.0001806924531168092,
+      "loss": 3.2366,
+      "step": 38577
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6330511569976807,
+      "learning_rate": 0.00018067832697587393,
+      "loss": 2.9807,
+      "step": 38578
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6298176646232605,
+      "learning_rate": 0.00018066420114920906,
+      "loss": 3.1927,
+      "step": 38579
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6028285622596741,
+      "learning_rate": 0.0001806500756368516,
+      "loss": 3.075,
+      "step": 38580
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6283663511276245,
+      "learning_rate": 0.00018063595043883894,
+      "loss": 2.7543,
+      "step": 38581
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6311411261558533,
+      "learning_rate": 0.00018062182555520824,
+      "loss": 3.0354,
+      "step": 38582
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.708346426486969,
+      "learning_rate": 0.00018060770098599657,
+      "loss": 3.1341,
+      "step": 38583
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6508063673973083,
+      "learning_rate": 0.0001805935767312414,
+      "loss": 2.8863,
+      "step": 38584
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6476802825927734,
+      "learning_rate": 0.00018057945279097964,
+      "loss": 2.9496,
+      "step": 38585
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6408866047859192,
+      "learning_rate": 0.0001805653291652486,
+      "loss": 3.0421,
+      "step": 38586
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6279966831207275,
+      "learning_rate": 0.0001805512058540856,
+      "loss": 3.1745,
+      "step": 38587
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6622072458267212,
+      "learning_rate": 0.00018053708285752763,
+      "loss": 2.9178,
+      "step": 38588
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6587921977043152,
+      "learning_rate": 0.00018052296017561213,
+      "loss": 3.1572,
+      "step": 38589
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6754596829414368,
+      "learning_rate": 0.00018050883780837607,
+      "loss": 2.918,
+      "step": 38590
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6614739298820496,
+      "learning_rate": 0.0001804947157558567,
+      "loss": 2.8633,
+      "step": 38591
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6809832453727722,
+      "learning_rate": 0.00018048059401809134,
+      "loss": 2.962,
+      "step": 38592
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6957449316978455,
+      "learning_rate": 0.00018046647259511704,
+      "loss": 2.9203,
+      "step": 38593
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6387645602226257,
+      "learning_rate": 0.0001804523514869711,
+      "loss": 3.0615,
+      "step": 38594
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6960155963897705,
+      "learning_rate": 0.00018043823069369057,
+      "loss": 2.8501,
+      "step": 38595
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6919668316841125,
+      "learning_rate": 0.00018042411021531273,
+      "loss": 3.0659,
+      "step": 38596
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6586015820503235,
+      "learning_rate": 0.0001804099900518749,
+      "loss": 3.0833,
+      "step": 38597
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6569781303405762,
+      "learning_rate": 0.000180395870203414,
+      "loss": 3.1045,
+      "step": 38598
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6461341977119446,
+      "learning_rate": 0.00018038175066996745,
+      "loss": 3.1601,
+      "step": 38599
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6257455945014954,
+      "learning_rate": 0.0001803676314515723,
+      "loss": 2.9019,
+      "step": 38600
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6397156119346619,
+      "learning_rate": 0.00018035351254826575,
+      "loss": 2.9765,
+      "step": 38601
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6896104216575623,
+      "learning_rate": 0.00018033939396008513,
+      "loss": 3.2906,
+      "step": 38602
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6564487814903259,
+      "learning_rate": 0.00018032527568706745,
+      "loss": 2.951,
+      "step": 38603
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6752212047576904,
+      "learning_rate": 0.00018031115772925005,
+      "loss": 2.8648,
+      "step": 38604
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6492764353752136,
+      "learning_rate": 0.00018029704008666988,
+      "loss": 3.0707,
+      "step": 38605
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.668976366519928,
+      "learning_rate": 0.00018028292275936431,
+      "loss": 3.0693,
+      "step": 38606
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6379303932189941,
+      "learning_rate": 0.00018026880574737055,
+      "loss": 3.0406,
+      "step": 38607
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8051410913467407,
+      "learning_rate": 0.00018025468905072563,
+      "loss": 3.0841,
+      "step": 38608
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6822032928466797,
+      "learning_rate": 0.0001802405726694669,
+      "loss": 2.7943,
+      "step": 38609
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6568723917007446,
+      "learning_rate": 0.00018022645660363133,
+      "loss": 2.8804,
+      "step": 38610
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6965839266777039,
+      "learning_rate": 0.00018021234085325628,
+      "loss": 3.0882,
+      "step": 38611
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6454933881759644,
+      "learning_rate": 0.00018019822541837896,
+      "loss": 3.1348,
+      "step": 38612
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6876438856124878,
+      "learning_rate": 0.00018018411029903632,
+      "loss": 2.9058,
+      "step": 38613
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6394537687301636,
+      "learning_rate": 0.0001801699954952658,
+      "loss": 2.9671,
+      "step": 38614
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6840302348136902,
+      "learning_rate": 0.00018015588100710439,
+      "loss": 2.8984,
+      "step": 38615
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6187319159507751,
+      "learning_rate": 0.00018014176683458927,
+      "loss": 3.1231,
+      "step": 38616
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.605553150177002,
+      "learning_rate": 0.00018012765297775778,
+      "loss": 2.8378,
+      "step": 38617
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6925824880599976,
+      "learning_rate": 0.0001801135394366469,
+      "loss": 2.8799,
+      "step": 38618
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.651581883430481,
+      "learning_rate": 0.00018009942621129398,
+      "loss": 3.0663,
+      "step": 38619
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6642777323722839,
+      "learning_rate": 0.00018008531330173598,
+      "loss": 2.7444,
+      "step": 38620
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6268136501312256,
+      "learning_rate": 0.00018007120070801023,
+      "loss": 3.073,
+      "step": 38621
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6606512069702148,
+      "learning_rate": 0.00018005708843015392,
+      "loss": 3.14,
+      "step": 38622
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7048143744468689,
+      "learning_rate": 0.000180042976468204,
+      "loss": 2.6478,
+      "step": 38623
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6552999019622803,
+      "learning_rate": 0.00018002886482219798,
+      "loss": 2.8626,
+      "step": 38624
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6531506180763245,
+      "learning_rate": 0.00018001475349217275,
+      "loss": 2.9233,
+      "step": 38625
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.682345986366272,
+      "learning_rate": 0.0001800006424781655,
+      "loss": 2.9008,
+      "step": 38626
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6473051905632019,
+      "learning_rate": 0.0001799865317802136,
+      "loss": 3.1767,
+      "step": 38627
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6639640927314758,
+      "learning_rate": 0.0001799724213983539,
+      "loss": 2.9966,
+      "step": 38628
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6246669888496399,
+      "learning_rate": 0.00017995831133262402,
+      "loss": 3.0385,
+      "step": 38629
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6045641899108887,
+      "learning_rate": 0.0001799442015830606,
+      "loss": 3.017,
+      "step": 38630
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7342897057533264,
+      "learning_rate": 0.00017993009214970106,
+      "loss": 3.1561,
+      "step": 38631
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7383832335472107,
+      "learning_rate": 0.00017991598303258266,
+      "loss": 2.8414,
+      "step": 38632
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7174133062362671,
+      "learning_rate": 0.00017990187423174242,
+      "loss": 2.9566,
+      "step": 38633
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6822835206985474,
+      "learning_rate": 0.00017988776574721758,
+      "loss": 3.0601,
+      "step": 38634
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6779569387435913,
+      "learning_rate": 0.0001798736575790451,
+      "loss": 2.7663,
+      "step": 38635
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6900989413261414,
+      "learning_rate": 0.00017985954972726237,
+      "loss": 2.9155,
+      "step": 38636
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6946017146110535,
+      "learning_rate": 0.00017984544219190653,
+      "loss": 3.2055,
+      "step": 38637
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6267918944358826,
+      "learning_rate": 0.00017983133497301453,
+      "loss": 2.9432,
+      "step": 38638
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7059924006462097,
+      "learning_rate": 0.0001798172280706238,
+      "loss": 3.08,
+      "step": 38639
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6676067113876343,
+      "learning_rate": 0.0001798031214847713,
+      "loss": 2.9552,
+      "step": 38640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6539041996002197,
+      "learning_rate": 0.00017978901521549428,
+      "loss": 3.0205,
+      "step": 38641
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6838967204093933,
+      "learning_rate": 0.0001797749092628298,
+      "loss": 2.9309,
+      "step": 38642
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6737627983093262,
+      "learning_rate": 0.0001797608036268151,
+      "loss": 3.1568,
+      "step": 38643
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6425014734268188,
+      "learning_rate": 0.00017974669830748735,
+      "loss": 3.0889,
+      "step": 38644
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6968850493431091,
+      "learning_rate": 0.00017973259330488354,
+      "loss": 2.8293,
+      "step": 38645
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6469689011573792,
+      "learning_rate": 0.00017971848861904105,
+      "loss": 2.9373,
+      "step": 38646
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6483293175697327,
+      "learning_rate": 0.00017970438424999688,
+      "loss": 3.098,
+      "step": 38647
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6595166325569153,
+      "learning_rate": 0.0001796902801977881,
+      "loss": 3.055,
+      "step": 38648
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6388629078865051,
+      "learning_rate": 0.00017967617646245213,
+      "loss": 2.9905,
+      "step": 38649
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6459506750106812,
+      "learning_rate": 0.00017966207304402583,
+      "loss": 2.9324,
+      "step": 38650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7083884477615356,
+      "learning_rate": 0.00017964796994254656,
+      "loss": 2.7932,
+      "step": 38651
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5959693193435669,
+      "learning_rate": 0.00017963386715805137,
+      "loss": 2.8373,
+      "step": 38652
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6729698181152344,
+      "learning_rate": 0.0001796197646905773,
+      "loss": 3.1081,
+      "step": 38653
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6342985033988953,
+      "learning_rate": 0.00017960566254016176,
+      "loss": 3.0374,
+      "step": 38654
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7330729365348816,
+      "learning_rate": 0.00017959156070684163,
+      "loss": 3.2231,
+      "step": 38655
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6425516605377197,
+      "learning_rate": 0.00017957745919065427,
+      "loss": 2.9848,
+      "step": 38656
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6559476256370544,
+      "learning_rate": 0.00017956335799163654,
+      "loss": 3.0129,
+      "step": 38657
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7110540866851807,
+      "learning_rate": 0.00017954925710982582,
+      "loss": 3.008,
+      "step": 38658
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6435421109199524,
+      "learning_rate": 0.00017953515654525924,
+      "loss": 2.9261,
+      "step": 38659
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6702364683151245,
+      "learning_rate": 0.00017952105629797375,
+      "loss": 2.9763,
+      "step": 38660
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.645945131778717,
+      "learning_rate": 0.00017950695636800672,
+      "loss": 3.0355,
+      "step": 38661
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6295376420021057,
+      "learning_rate": 0.00017949285675539515,
+      "loss": 3.0032,
+      "step": 38662
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.640484631061554,
+      "learning_rate": 0.0001794787574601761,
+      "loss": 3.0606,
+      "step": 38663
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6546970009803772,
+      "learning_rate": 0.00017946465848238697,
+      "loss": 2.9448,
+      "step": 38664
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6797558665275574,
+      "learning_rate": 0.00017945055982206465,
+      "loss": 2.8821,
+      "step": 38665
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6290567517280579,
+      "learning_rate": 0.00017943646147924639,
+      "loss": 2.9998,
+      "step": 38666
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6903430819511414,
+      "learning_rate": 0.0001794223634539692,
+      "loss": 3.0059,
+      "step": 38667
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7165091037750244,
+      "learning_rate": 0.00017940826574627038,
+      "loss": 2.9054,
+      "step": 38668
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6272432208061218,
+      "learning_rate": 0.00017939416835618702,
+      "loss": 3.0671,
+      "step": 38669
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6418277621269226,
+      "learning_rate": 0.00017938007128375609,
+      "loss": 2.8885,
+      "step": 38670
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.671662449836731,
+      "learning_rate": 0.00017936597452901493,
+      "loss": 2.9865,
+      "step": 38671
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6581867337226868,
+      "learning_rate": 0.00017935187809200046,
+      "loss": 2.9492,
+      "step": 38672
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6353850364685059,
+      "learning_rate": 0.00017933778197275002,
+      "loss": 2.9451,
+      "step": 38673
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6606138348579407,
+      "learning_rate": 0.0001793236861713007,
+      "loss": 2.9165,
+      "step": 38674
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7428664565086365,
+      "learning_rate": 0.00017930959068768944,
+      "loss": 2.9639,
+      "step": 38675
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6777299046516418,
+      "learning_rate": 0.0001792954955219536,
+      "loss": 3.3009,
+      "step": 38676
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.639334499835968,
+      "learning_rate": 0.0001792814006741301,
+      "loss": 2.9518,
+      "step": 38677
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6822625994682312,
+      "learning_rate": 0.00017926730614425615,
+      "loss": 3.0377,
+      "step": 38678
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6702042818069458,
+      "learning_rate": 0.00017925321193236897,
+      "loss": 2.9242,
+      "step": 38679
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6273990273475647,
+      "learning_rate": 0.0001792391180385056,
+      "loss": 2.8685,
+      "step": 38680
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6678882837295532,
+      "learning_rate": 0.00017922502446270314,
+      "loss": 2.9388,
+      "step": 38681
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6841011047363281,
+      "learning_rate": 0.0001792109312049986,
+      "loss": 3.025,
+      "step": 38682
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6396910548210144,
+      "learning_rate": 0.0001791968382654293,
+      "loss": 3.0088,
+      "step": 38683
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.63197261095047,
+      "learning_rate": 0.00017918274564403236,
+      "loss": 3.0014,
+      "step": 38684
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6431912779808044,
+      "learning_rate": 0.00017916865334084464,
+      "loss": 3.0681,
+      "step": 38685
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6578310132026672,
+      "learning_rate": 0.00017915456135590363,
+      "loss": 2.9865,
+      "step": 38686
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6344066262245178,
+      "learning_rate": 0.00017914046968924612,
+      "loss": 2.8682,
+      "step": 38687
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6525806784629822,
+      "learning_rate": 0.00017912637834090932,
+      "loss": 2.86,
+      "step": 38688
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6194393038749695,
+      "learning_rate": 0.00017911228731093052,
+      "loss": 2.8638,
+      "step": 38689
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6648958325386047,
+      "learning_rate": 0.0001790981965993465,
+      "loss": 2.9185,
+      "step": 38690
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6630010008811951,
+      "learning_rate": 0.0001790841062061948,
+      "loss": 2.8877,
+      "step": 38691
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.643262505531311,
+      "learning_rate": 0.00017907001613151212,
+      "loss": 2.8769,
+      "step": 38692
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6801116466522217,
+      "learning_rate": 0.00017905592637533572,
+      "loss": 2.9138,
+      "step": 38693
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6177447438240051,
+      "learning_rate": 0.00017904183693770285,
+      "loss": 3.2762,
+      "step": 38694
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6487329602241516,
+      "learning_rate": 0.00017902774781865043,
+      "loss": 2.8668,
+      "step": 38695
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6444410085678101,
+      "learning_rate": 0.00017901365901821573,
+      "loss": 3.0458,
+      "step": 38696
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6466326713562012,
+      "learning_rate": 0.00017899957053643567,
+      "loss": 3.0129,
+      "step": 38697
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6659789085388184,
+      "learning_rate": 0.00017898548237334748,
+      "loss": 3.0764,
+      "step": 38698
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6836056709289551,
+      "learning_rate": 0.0001789713945289883,
+      "loss": 2.9231,
+      "step": 38699
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6317378282546997,
+      "learning_rate": 0.00017895730700339508,
+      "loss": 2.9773,
+      "step": 38700
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7079015374183655,
+      "learning_rate": 0.0001789432197966051,
+      "loss": 3.0086,
+      "step": 38701
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6322833895683289,
+      "learning_rate": 0.00017892913290865533,
+      "loss": 2.9109,
+      "step": 38702
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.620879590511322,
+      "learning_rate": 0.00017891504633958286,
+      "loss": 3.1562,
+      "step": 38703
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6418386697769165,
+      "learning_rate": 0.00017890096008942502,
+      "loss": 3.0976,
+      "step": 38704
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6648834943771362,
+      "learning_rate": 0.00017888687415821864,
+      "loss": 3.162,
+      "step": 38705
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6503156423568726,
+      "learning_rate": 0.000178872788546001,
+      "loss": 2.9803,
+      "step": 38706
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6655746698379517,
+      "learning_rate": 0.000178858703252809,
+      "loss": 2.8608,
+      "step": 38707
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7488763928413391,
+      "learning_rate": 0.0001788446182786799,
+      "loss": 3.0323,
+      "step": 38708
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6233779788017273,
+      "learning_rate": 0.00017883053362365088,
+      "loss": 3.0267,
+      "step": 38709
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6482000946998596,
+      "learning_rate": 0.0001788164492877587,
+      "loss": 2.9888,
+      "step": 38710
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6482628583908081,
+      "learning_rate": 0.00017880236527104086,
+      "loss": 2.9677,
+      "step": 38711
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6418460011482239,
+      "learning_rate": 0.00017878828157353417,
+      "loss": 2.7318,
+      "step": 38712
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6458458304405212,
+      "learning_rate": 0.00017877419819527582,
+      "loss": 3.0026,
+      "step": 38713
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6432719826698303,
+      "learning_rate": 0.00017876011513630297,
+      "loss": 3.0038,
+      "step": 38714
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6705681085586548,
+      "learning_rate": 0.00017874603239665255,
+      "loss": 2.8765,
+      "step": 38715
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6631069779396057,
+      "learning_rate": 0.00017873194997636182,
+      "loss": 3.0594,
+      "step": 38716
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6457616090774536,
+      "learning_rate": 0.00017871786787546776,
+      "loss": 2.9703,
+      "step": 38717
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6328929662704468,
+      "learning_rate": 0.0001787037860940076,
+      "loss": 2.7767,
+      "step": 38718
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6492588520050049,
+      "learning_rate": 0.00017868970463201815,
+      "loss": 3.0519,
+      "step": 38719
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6573117971420288,
+      "learning_rate": 0.0001786756234895367,
+      "loss": 3.0632,
+      "step": 38720
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6802584528923035,
+      "learning_rate": 0.00017866154266660045,
+      "loss": 3.218,
+      "step": 38721
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6282143592834473,
+      "learning_rate": 0.0001786474621632462,
+      "loss": 2.9557,
+      "step": 38722
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6655977964401245,
+      "learning_rate": 0.00017863338197951126,
+      "loss": 2.8931,
+      "step": 38723
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6891822218894958,
+      "learning_rate": 0.00017861930211543264,
+      "loss": 2.9934,
+      "step": 38724
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6471545100212097,
+      "learning_rate": 0.0001786052225710473,
+      "loss": 3.1537,
+      "step": 38725
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6515899896621704,
+      "learning_rate": 0.0001785911433463926,
+      "loss": 3.0171,
+      "step": 38726
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6359424591064453,
+      "learning_rate": 0.00017857706444150535,
+      "loss": 3.0864,
+      "step": 38727
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6311108469963074,
+      "learning_rate": 0.00017856298585642284,
+      "loss": 3.1186,
+      "step": 38728
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6485701203346252,
+      "learning_rate": 0.00017854890759118192,
+      "loss": 3.1168,
+      "step": 38729
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6353026628494263,
+      "learning_rate": 0.00017853482964581992,
+      "loss": 3.0755,
+      "step": 38730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6661299467086792,
+      "learning_rate": 0.00017852075202037378,
+      "loss": 2.908,
+      "step": 38731
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6596245765686035,
+      "learning_rate": 0.00017850667471488053,
+      "loss": 2.9477,
+      "step": 38732
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.67014080286026,
+      "learning_rate": 0.00017849259772937743,
+      "loss": 3.2171,
+      "step": 38733
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6512858867645264,
+      "learning_rate": 0.00017847852106390132,
+      "loss": 2.8545,
+      "step": 38734
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6994531154632568,
+      "learning_rate": 0.00017846444471848944,
+      "loss": 2.9238,
+      "step": 38735
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6024457216262817,
+      "learning_rate": 0.0001784503686931789,
+      "loss": 2.9689,
+      "step": 38736
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6673498153686523,
+      "learning_rate": 0.00017843629298800655,
+      "loss": 2.8969,
+      "step": 38737
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6520382761955261,
+      "learning_rate": 0.00017842221760300977,
+      "loss": 3.0239,
+      "step": 38738
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6899427175521851,
+      "learning_rate": 0.0001784081425382254,
+      "loss": 3.0428,
+      "step": 38739
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6339791417121887,
+      "learning_rate": 0.00017839406779369052,
+      "loss": 2.8405,
+      "step": 38740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6099035739898682,
+      "learning_rate": 0.00017837999336944236,
+      "loss": 2.9079,
+      "step": 38741
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6363784074783325,
+      "learning_rate": 0.00017836591926551786,
+      "loss": 3.0636,
+      "step": 38742
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6817944049835205,
+      "learning_rate": 0.00017835184548195418,
+      "loss": 2.865,
+      "step": 38743
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6493276953697205,
+      "learning_rate": 0.0001783377720187882,
+      "loss": 2.8713,
+      "step": 38744
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6394343376159668,
+      "learning_rate": 0.00017832369887605718,
+      "loss": 3.0309,
+      "step": 38745
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6780338883399963,
+      "learning_rate": 0.00017830962605379813,
+      "loss": 3.0202,
+      "step": 38746
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6320281028747559,
+      "learning_rate": 0.00017829555355204803,
+      "loss": 2.8572,
+      "step": 38747
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6657953858375549,
+      "learning_rate": 0.00017828148137084411,
+      "loss": 3.0652,
+      "step": 38748
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6262355446815491,
+      "learning_rate": 0.00017826740951022336,
+      "loss": 3.1262,
+      "step": 38749
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7236253023147583,
+      "learning_rate": 0.0001782533379702227,
+      "loss": 2.9621,
+      "step": 38750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6585520505905151,
+      "learning_rate": 0.00017823926675087945,
+      "loss": 3.0541,
+      "step": 38751
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6381440162658691,
+      "learning_rate": 0.0001782251958522304,
+      "loss": 2.9744,
+      "step": 38752
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6845026612281799,
+      "learning_rate": 0.00017821112527431287,
+      "loss": 3.0564,
+      "step": 38753
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6539150476455688,
+      "learning_rate": 0.0001781970550171638,
+      "loss": 2.9587,
+      "step": 38754
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7004772424697876,
+      "learning_rate": 0.00017818298508082015,
+      "loss": 3.0137,
+      "step": 38755
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6276587247848511,
+      "learning_rate": 0.00017816891546531914,
+      "loss": 2.9447,
+      "step": 38756
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6287436485290527,
+      "learning_rate": 0.00017815484617069775,
+      "loss": 2.9367,
+      "step": 38757
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7259926795959473,
+      "learning_rate": 0.00017814077719699312,
+      "loss": 3.1074,
+      "step": 38758
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6566614508628845,
+      "learning_rate": 0.00017812670854424207,
+      "loss": 3.2498,
+      "step": 38759
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6689494848251343,
+      "learning_rate": 0.0001781126402124819,
+      "loss": 3.0541,
+      "step": 38760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6920581459999084,
+      "learning_rate": 0.00017809857220174962,
+      "loss": 2.9865,
+      "step": 38761
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.627387285232544,
+      "learning_rate": 0.00017808450451208213,
+      "loss": 3.109,
+      "step": 38762
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6338323354721069,
+      "learning_rate": 0.0001780704371435167,
+      "loss": 2.9502,
+      "step": 38763
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6786047220230103,
+      "learning_rate": 0.00017805637009609025,
+      "loss": 3.134,
+      "step": 38764
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6783745884895325,
+      "learning_rate": 0.00017804230336983975,
+      "loss": 3.161,
+      "step": 38765
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6686196327209473,
+      "learning_rate": 0.0001780282369648025,
+      "loss": 3.1341,
+      "step": 38766
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6732924580574036,
+      "learning_rate": 0.00017801417088101533,
+      "loss": 2.8498,
+      "step": 38767
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6651960015296936,
+      "learning_rate": 0.00017800010511851542,
+      "loss": 3.1484,
+      "step": 38768
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6445776224136353,
+      "learning_rate": 0.00017798603967733963,
+      "loss": 3.0089,
+      "step": 38769
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7008956074714661,
+      "learning_rate": 0.0001779719745575252,
+      "loss": 3.0585,
+      "step": 38770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7105851173400879,
+      "learning_rate": 0.00017795790975910913,
+      "loss": 2.9633,
+      "step": 38771
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6775414347648621,
+      "learning_rate": 0.00017794384528212835,
+      "loss": 2.9435,
+      "step": 38772
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6525750160217285,
+      "learning_rate": 0.00017792978112662013,
+      "loss": 2.8013,
+      "step": 38773
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6795325875282288,
+      "learning_rate": 0.0001779157172926212,
+      "loss": 3.1657,
+      "step": 38774
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6364821791648865,
+      "learning_rate": 0.00017790165378016884,
+      "loss": 2.9929,
+      "step": 38775
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6568708419799805,
+      "learning_rate": 0.00017788759058930015,
+      "loss": 3.088,
+      "step": 38776
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6670739650726318,
+      "learning_rate": 0.00017787352772005184,
+      "loss": 2.9832,
+      "step": 38777
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6435451507568359,
+      "learning_rate": 0.00017785946517246136,
+      "loss": 2.9281,
+      "step": 38778
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6362979412078857,
+      "learning_rate": 0.00017784540294656543,
+      "loss": 2.9789,
+      "step": 38779
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6384310722351074,
+      "learning_rate": 0.00017783134104240117,
+      "loss": 2.8507,
+      "step": 38780
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6559550166130066,
+      "learning_rate": 0.00017781727946000574,
+      "loss": 2.9265,
+      "step": 38781
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.650518000125885,
+      "learning_rate": 0.000177803218199416,
+      "loss": 2.9284,
+      "step": 38782
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6557312607765198,
+      "learning_rate": 0.00017778915726066917,
+      "loss": 3.059,
+      "step": 38783
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.613975465297699,
+      "learning_rate": 0.00017777509664380206,
+      "loss": 2.9277,
+      "step": 38784
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6250554323196411,
+      "learning_rate": 0.00017776103634885185,
+      "loss": 2.9664,
+      "step": 38785
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6322828531265259,
+      "learning_rate": 0.00017774697637585563,
+      "loss": 2.8287,
+      "step": 38786
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6861053109169006,
+      "learning_rate": 0.00017773291672485022,
+      "loss": 3.1791,
+      "step": 38787
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.630342960357666,
+      "learning_rate": 0.00017771885739587288,
+      "loss": 2.999,
+      "step": 38788
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6999707818031311,
+      "learning_rate": 0.0001777047983889605,
+      "loss": 3.0539,
+      "step": 38789
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6790862679481506,
+      "learning_rate": 0.00017769073970415005,
+      "loss": 3.0786,
+      "step": 38790
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6557303667068481,
+      "learning_rate": 0.0001776766813414788,
+      "loss": 2.7987,
+      "step": 38791
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6694257855415344,
+      "learning_rate": 0.0001776626233009835,
+      "loss": 2.9747,
+      "step": 38792
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7797983288764954,
+      "learning_rate": 0.00017764856558270153,
+      "loss": 2.932,
+      "step": 38793
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6885984539985657,
+      "learning_rate": 0.00017763450818666945,
+      "loss": 2.8699,
+      "step": 38794
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6026552319526672,
+      "learning_rate": 0.00017762045111292455,
+      "loss": 2.9986,
+      "step": 38795
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7198888063430786,
+      "learning_rate": 0.00017760639436150398,
+      "loss": 2.9756,
+      "step": 38796
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.646704375743866,
+      "learning_rate": 0.00017759233793244452,
+      "loss": 2.9899,
+      "step": 38797
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.00017757828182578332,
+      "loss": 2.8782,
+      "step": 38798
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.629848062992096,
+      "learning_rate": 0.00017756422604155728,
+      "loss": 2.9032,
+      "step": 38799
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6715112924575806,
+      "learning_rate": 0.00017755017057980362,
+      "loss": 2.9572,
+      "step": 38800
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6943894028663635,
+      "learning_rate": 0.00017753611544055916,
+      "loss": 3.0577,
+      "step": 38801
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6915984153747559,
+      "learning_rate": 0.00017752206062386096,
+      "loss": 2.9434,
+      "step": 38802
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6393081545829773,
+      "learning_rate": 0.00017750800612974626,
+      "loss": 2.9935,
+      "step": 38803
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6286363005638123,
+      "learning_rate": 0.00017749395195825173,
+      "loss": 2.7904,
+      "step": 38804
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6637555956840515,
+      "learning_rate": 0.0001774798981094147,
+      "loss": 2.8944,
+      "step": 38805
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6585133075714111,
+      "learning_rate": 0.00017746584458327189,
+      "loss": 2.8192,
+      "step": 38806
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6345171332359314,
+      "learning_rate": 0.0001774517913798605,
+      "loss": 3.0814,
+      "step": 38807
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6817945241928101,
+      "learning_rate": 0.00017743773849921758,
+      "loss": 3.1959,
+      "step": 38808
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.673495888710022,
+      "learning_rate": 0.00017742368594137993,
+      "loss": 2.8187,
+      "step": 38809
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6089171171188354,
+      "learning_rate": 0.00017740963370638485,
+      "loss": 3.0861,
+      "step": 38810
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6373415589332581,
+      "learning_rate": 0.00017739558179426915,
+      "loss": 2.9421,
+      "step": 38811
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7025952935218811,
+      "learning_rate": 0.00017738153020506982,
+      "loss": 3.0454,
+      "step": 38812
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.66726154088974,
+      "learning_rate": 0.0001773674789388241,
+      "loss": 3.1199,
+      "step": 38813
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.630937397480011,
+      "learning_rate": 0.00017735342799556864,
+      "loss": 2.9137,
+      "step": 38814
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6461975574493408,
+      "learning_rate": 0.00017733937737534082,
+      "loss": 2.9248,
+      "step": 38815
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6686995029449463,
+      "learning_rate": 0.0001773253270781774,
+      "loss": 2.8179,
+      "step": 38816
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6307097673416138,
+      "learning_rate": 0.0001773112771041154,
+      "loss": 2.9402,
+      "step": 38817
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6663997769355774,
+      "learning_rate": 0.000177297227453192,
+      "loss": 2.7335,
+      "step": 38818
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.660294771194458,
+      "learning_rate": 0.00017728317812544406,
+      "loss": 2.9602,
+      "step": 38819
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7105454206466675,
+      "learning_rate": 0.0001772691291209086,
+      "loss": 2.9683,
+      "step": 38820
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6138967275619507,
+      "learning_rate": 0.00017725508043962257,
+      "loss": 2.9629,
+      "step": 38821
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7000978589057922,
+      "learning_rate": 0.00017724103208162308,
+      "loss": 3.1553,
+      "step": 38822
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6332494020462036,
+      "learning_rate": 0.00017722698404694715,
+      "loss": 3.03,
+      "step": 38823
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5949512720108032,
+      "learning_rate": 0.00017721293633563164,
+      "loss": 3.0115,
+      "step": 38824
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6465287208557129,
+      "learning_rate": 0.0001771988889477137,
+      "loss": 2.9231,
+      "step": 38825
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6742639541625977,
+      "learning_rate": 0.00017718484188323014,
+      "loss": 3.1172,
+      "step": 38826
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6116287708282471,
+      "learning_rate": 0.00017717079514221806,
+      "loss": 3.1104,
+      "step": 38827
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7612192034721375,
+      "learning_rate": 0.0001771567487247146,
+      "loss": 2.9702,
+      "step": 38828
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6515193581581116,
+      "learning_rate": 0.0001771427026307566,
+      "loss": 3.0939,
+      "step": 38829
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6422175765037537,
+      "learning_rate": 0.00017712865686038106,
+      "loss": 2.9786,
+      "step": 38830
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6645070314407349,
+      "learning_rate": 0.0001771146114136249,
+      "loss": 2.7252,
+      "step": 38831
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6393051743507385,
+      "learning_rate": 0.00017710056629052526,
+      "loss": 2.7584,
+      "step": 38832
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6298916935920715,
+      "learning_rate": 0.00017708652149111913,
+      "loss": 3.3263,
+      "step": 38833
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6236022710800171,
+      "learning_rate": 0.00017707247701544335,
+      "loss": 2.9878,
+      "step": 38834
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6486828923225403,
+      "learning_rate": 0.0001770584328635351,
+      "loss": 3.0399,
+      "step": 38835
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6714288592338562,
+      "learning_rate": 0.00017704438903543116,
+      "loss": 2.9953,
+      "step": 38836
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6709398627281189,
+      "learning_rate": 0.0001770303455311687,
+      "loss": 2.9666,
+      "step": 38837
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6724570393562317,
+      "learning_rate": 0.0001770163023507847,
+      "loss": 3.0689,
+      "step": 38838
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6929112076759338,
+      "learning_rate": 0.00017700225949431598,
+      "loss": 3.0246,
+      "step": 38839
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6533306837081909,
+      "learning_rate": 0.00017698821696179978,
+      "loss": 2.8858,
+      "step": 38840
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6718860268592834,
+      "learning_rate": 0.00017697417475327288,
+      "loss": 2.9085,
+      "step": 38841
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.650233805179596,
+      "learning_rate": 0.00017696013286877226,
+      "loss": 3.1286,
+      "step": 38842
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7158471941947937,
+      "learning_rate": 0.0001769460913083351,
+      "loss": 2.9035,
+      "step": 38843
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6609078049659729,
+      "learning_rate": 0.00017693205007199812,
+      "loss": 3.0562,
+      "step": 38844
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.672481894493103,
+      "learning_rate": 0.00017691800915979855,
+      "loss": 2.8383,
+      "step": 38845
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6177345514297485,
+      "learning_rate": 0.00017690396857177317,
+      "loss": 2.8635,
+      "step": 38846
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6650989651679993,
+      "learning_rate": 0.00017688992830795905,
+      "loss": 3.1194,
+      "step": 38847
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6523521542549133,
+      "learning_rate": 0.00017687588836839323,
+      "loss": 2.92,
+      "step": 38848
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6387304067611694,
+      "learning_rate": 0.00017686184875311255,
+      "loss": 3.0017,
+      "step": 38849
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.650626540184021,
+      "learning_rate": 0.00017684780946215412,
+      "loss": 3.224,
+      "step": 38850
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7072687745094299,
+      "learning_rate": 0.00017683377049555483,
+      "loss": 3.068,
+      "step": 38851
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6670072674751282,
+      "learning_rate": 0.00017681973185335165,
+      "loss": 2.7412,
+      "step": 38852
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6184037327766418,
+      "learning_rate": 0.0001768056935355817,
+      "loss": 3.0716,
+      "step": 38853
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6479713916778564,
+      "learning_rate": 0.0001767916555422817,
+      "loss": 3.0369,
+      "step": 38854
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6216630339622498,
+      "learning_rate": 0.00017677761787348892,
+      "loss": 3.1023,
+      "step": 38855
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.667361319065094,
+      "learning_rate": 0.00017676358052924008,
+      "loss": 2.9674,
+      "step": 38856
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6166158318519592,
+      "learning_rate": 0.00017674954350957217,
+      "loss": 2.8018,
+      "step": 38857
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6484586596488953,
+      "learning_rate": 0.00017673550681452242,
+      "loss": 2.9074,
+      "step": 38858
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6432170271873474,
+      "learning_rate": 0.00017672147044412757,
+      "loss": 2.7897,
+      "step": 38859
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.645042359828949,
+      "learning_rate": 0.00017670743439842466,
+      "loss": 3.1734,
+      "step": 38860
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6494126915931702,
+      "learning_rate": 0.00017669339867745052,
+      "loss": 3.0803,
+      "step": 38861
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6544557809829712,
+      "learning_rate": 0.00017667936328124233,
+      "loss": 2.9205,
+      "step": 38862
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.666323184967041,
+      "learning_rate": 0.000176665328209837,
+      "loss": 2.8539,
+      "step": 38863
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6175612211227417,
+      "learning_rate": 0.00017665129346327134,
+      "loss": 3.1707,
+      "step": 38864
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7000352740287781,
+      "learning_rate": 0.00017663725904158253,
+      "loss": 3.1623,
+      "step": 38865
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7238625884056091,
+      "learning_rate": 0.00017662322494480743,
+      "loss": 2.7223,
+      "step": 38866
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6424652338027954,
+      "learning_rate": 0.00017660919117298292,
+      "loss": 2.8492,
+      "step": 38867
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6737831234931946,
+      "learning_rate": 0.00017659515772614618,
+      "loss": 3.1558,
+      "step": 38868
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6455912590026855,
+      "learning_rate": 0.00017658112460433397,
+      "loss": 2.8725,
+      "step": 38869
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6594234704971313,
+      "learning_rate": 0.00017656709180758343,
+      "loss": 3.1363,
+      "step": 38870
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6191946864128113,
+      "learning_rate": 0.00017655305933593124,
+      "loss": 2.8576,
+      "step": 38871
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6468594074249268,
+      "learning_rate": 0.00017653902718941464,
+      "loss": 3.2429,
+      "step": 38872
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.653252124786377,
+      "learning_rate": 0.00017652499536807052,
+      "loss": 2.8514,
+      "step": 38873
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6690567135810852,
+      "learning_rate": 0.00017651096387193567,
+      "loss": 2.8927,
+      "step": 38874
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6499602794647217,
+      "learning_rate": 0.0001764969327010473,
+      "loss": 2.9986,
+      "step": 38875
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6645433902740479,
+      "learning_rate": 0.00017648290185544213,
+      "loss": 3.0087,
+      "step": 38876
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6643459796905518,
+      "learning_rate": 0.00017646887133515734,
+      "loss": 3.1536,
+      "step": 38877
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6513760685920715,
+      "learning_rate": 0.00017645484114022972,
+      "loss": 3.0292,
+      "step": 38878
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.646967887878418,
+      "learning_rate": 0.00017644081127069622,
+      "loss": 3.1265,
+      "step": 38879
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7216665744781494,
+      "learning_rate": 0.00017642678172659395,
+      "loss": 3.0569,
+      "step": 38880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6408677101135254,
+      "learning_rate": 0.00017641275250795968,
+      "loss": 2.9765,
+      "step": 38881
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6516188979148865,
+      "learning_rate": 0.00017639872361483052,
+      "loss": 3.033,
+      "step": 38882
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6692624688148499,
+      "learning_rate": 0.0001763846950472432,
+      "loss": 2.9756,
+      "step": 38883
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6817802786827087,
+      "learning_rate": 0.00017637066680523483,
+      "loss": 2.9432,
+      "step": 38884
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7394009828567505,
+      "learning_rate": 0.00017635663888884243,
+      "loss": 3.1279,
+      "step": 38885
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6396210789680481,
+      "learning_rate": 0.00017634261129810273,
+      "loss": 3.2632,
+      "step": 38886
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6615631580352783,
+      "learning_rate": 0.0001763285840330529,
+      "loss": 3.0701,
+      "step": 38887
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6404348611831665,
+      "learning_rate": 0.00017631455709372971,
+      "loss": 2.9918,
+      "step": 38888
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.667752742767334,
+      "learning_rate": 0.00017630053048017016,
+      "loss": 3.0932,
+      "step": 38889
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6676562428474426,
+      "learning_rate": 0.0001762865041924113,
+      "loss": 2.9182,
+      "step": 38890
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6496162414550781,
+      "learning_rate": 0.00017627247823048987,
+      "loss": 3.0129,
+      "step": 38891
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.650679886341095,
+      "learning_rate": 0.00017625845259444304,
+      "loss": 2.9894,
+      "step": 38892
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6663594245910645,
+      "learning_rate": 0.00017624442728430748,
+      "loss": 3.0855,
+      "step": 38893
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6415731310844421,
+      "learning_rate": 0.00017623040230012037,
+      "loss": 3.0113,
+      "step": 38894
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6369303464889526,
+      "learning_rate": 0.0001762163776419186,
+      "loss": 2.9257,
+      "step": 38895
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6577563285827637,
+      "learning_rate": 0.00017620235330973897,
+      "loss": 3.0859,
+      "step": 38896
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6328701972961426,
+      "learning_rate": 0.00017618832930361858,
+      "loss": 3.0281,
+      "step": 38897
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6667453050613403,
+      "learning_rate": 0.00017617430562359422,
+      "loss": 3.0712,
+      "step": 38898
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6224763989448547,
+      "learning_rate": 0.00017616028226970297,
+      "loss": 2.9339,
+      "step": 38899
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6647245287895203,
+      "learning_rate": 0.0001761462592419818,
+      "loss": 3.0157,
+      "step": 38900
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6523852944374084,
+      "learning_rate": 0.0001761322365404674,
+      "loss": 2.9646,
+      "step": 38901
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6578008532524109,
+      "learning_rate": 0.00017611821416519693,
+      "loss": 2.9051,
+      "step": 38902
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6416773200035095,
+      "learning_rate": 0.0001761041921162072,
+      "loss": 3.0441,
+      "step": 38903
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6548252701759338,
+      "learning_rate": 0.00017609017039353517,
+      "loss": 2.8382,
+      "step": 38904
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.641315758228302,
+      "learning_rate": 0.00017607614899721786,
+      "loss": 2.9971,
+      "step": 38905
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6452323794364929,
+      "learning_rate": 0.00017606212792729207,
+      "loss": 3.0848,
+      "step": 38906
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6650744080543518,
+      "learning_rate": 0.00017604810718379487,
+      "loss": 3.0076,
+      "step": 38907
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.678641140460968,
+      "learning_rate": 0.00017603408676676295,
+      "loss": 3.0404,
+      "step": 38908
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6717094779014587,
+      "learning_rate": 0.00017602006667623345,
+      "loss": 3.0307,
+      "step": 38909
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6648962497711182,
+      "learning_rate": 0.0001760060469122433,
+      "loss": 2.8707,
+      "step": 38910
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6424950361251831,
+      "learning_rate": 0.00017599202747482925,
+      "loss": 2.8513,
+      "step": 38911
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6827686429023743,
+      "learning_rate": 0.00017597800836402844,
+      "loss": 2.8907,
+      "step": 38912
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6990754008293152,
+      "learning_rate": 0.0001759639895798776,
+      "loss": 3.1935,
+      "step": 38913
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6267073154449463,
+      "learning_rate": 0.00017594997112241375,
+      "loss": 2.9517,
+      "step": 38914
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6684610843658447,
+      "learning_rate": 0.00017593595299167386,
+      "loss": 2.6714,
+      "step": 38915
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.65615314245224,
+      "learning_rate": 0.0001759219351876947,
+      "loss": 3.2431,
+      "step": 38916
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6717365980148315,
+      "learning_rate": 0.00017590791771051343,
+      "loss": 3.0229,
+      "step": 38917
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6720296740531921,
+      "learning_rate": 0.0001758939005601667,
+      "loss": 2.9914,
+      "step": 38918
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6723449230194092,
+      "learning_rate": 0.00017587988373669156,
+      "loss": 3.0294,
+      "step": 38919
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6474607586860657,
+      "learning_rate": 0.000175865867240125,
+      "loss": 2.9674,
+      "step": 38920
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6441729068756104,
+      "learning_rate": 0.00017585185107050374,
+      "loss": 3.0136,
+      "step": 38921
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6469467878341675,
+      "learning_rate": 0.00017583783522786496,
+      "loss": 2.8793,
+      "step": 38922
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6387615203857422,
+      "learning_rate": 0.00017582381971224527,
+      "loss": 3.1108,
+      "step": 38923
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6527659296989441,
+      "learning_rate": 0.0001758098045236818,
+      "loss": 2.997,
+      "step": 38924
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6545872092247009,
+      "learning_rate": 0.00017579578966221145,
+      "loss": 3.3368,
+      "step": 38925
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6568055152893066,
+      "learning_rate": 0.00017578177512787103,
+      "loss": 3.0761,
+      "step": 38926
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6849158406257629,
+      "learning_rate": 0.00017576776092069759,
+      "loss": 2.8588,
+      "step": 38927
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6461156606674194,
+      "learning_rate": 0.00017575374704072787,
+      "loss": 2.8396,
+      "step": 38928
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6694230437278748,
+      "learning_rate": 0.00017573973348799883,
+      "loss": 3.0169,
+      "step": 38929
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6535854935646057,
+      "learning_rate": 0.00017572572026254755,
+      "loss": 2.9841,
+      "step": 38930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6744319200515747,
+      "learning_rate": 0.00017571170736441072,
+      "loss": 3.0615,
+      "step": 38931
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6252499222755432,
+      "learning_rate": 0.00017569769479362543,
+      "loss": 3.0025,
+      "step": 38932
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6410253047943115,
+      "learning_rate": 0.00017568368255022835,
+      "loss": 3.1207,
+      "step": 38933
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6542544960975647,
+      "learning_rate": 0.00017566967063425658,
+      "loss": 2.8341,
+      "step": 38934
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6271516680717468,
+      "learning_rate": 0.00017565565904574702,
+      "loss": 3.0428,
+      "step": 38935
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6435152888298035,
+      "learning_rate": 0.00017564164778473642,
+      "loss": 3.0798,
+      "step": 38936
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6573494672775269,
+      "learning_rate": 0.00017562763685126194,
+      "loss": 3.2102,
+      "step": 38937
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6944133639335632,
+      "learning_rate": 0.00017561362624536015,
+      "loss": 2.9844,
+      "step": 38938
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7348315119743347,
+      "learning_rate": 0.0001755996159670682,
+      "loss": 3.2201,
+      "step": 38939
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6667076945304871,
+      "learning_rate": 0.00017558560601642302,
+      "loss": 3.1203,
+      "step": 38940
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9448343515396118,
+      "learning_rate": 0.00017557159639346124,
+      "loss": 2.9531,
+      "step": 38941
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6497033834457397,
+      "learning_rate": 0.00017555758709822004,
+      "loss": 2.9215,
+      "step": 38942
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2949033975601196,
+      "learning_rate": 0.00017554357813073622,
+      "loss": 3.0897,
+      "step": 38943
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6681045293807983,
+      "learning_rate": 0.00017552956949104656,
+      "loss": 3.1463,
+      "step": 38944
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6465238332748413,
+      "learning_rate": 0.0001755155611791882,
+      "loss": 2.9325,
+      "step": 38945
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6648427844047546,
+      "learning_rate": 0.00017550155319519777,
+      "loss": 2.8387,
+      "step": 38946
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8117325305938721,
+      "learning_rate": 0.00017548754553911241,
+      "loss": 2.6738,
+      "step": 38947
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.629001796245575,
+      "learning_rate": 0.00017547353821096873,
+      "loss": 3.0694,
+      "step": 38948
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6560392379760742,
+      "learning_rate": 0.00017545953121080388,
+      "loss": 2.8574,
+      "step": 38949
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.636073887348175,
+      "learning_rate": 0.00017544552453865474,
+      "loss": 3.1828,
+      "step": 38950
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6364843845367432,
+      "learning_rate": 0.0001754315181945579,
+      "loss": 2.8929,
+      "step": 38951
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7276064157485962,
+      "learning_rate": 0.00017541751217855066,
+      "loss": 2.9348,
+      "step": 38952
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.67525315284729,
+      "learning_rate": 0.00017540350649066966,
+      "loss": 2.9005,
+      "step": 38953
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6328374743461609,
+      "learning_rate": 0.00017538950113095192,
+      "loss": 2.9466,
+      "step": 38954
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6840808987617493,
+      "learning_rate": 0.0001753754960994341,
+      "loss": 2.9267,
+      "step": 38955
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.635502278804779,
+      "learning_rate": 0.00017536149139615322,
+      "loss": 3.1029,
+      "step": 38956
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6700057983398438,
+      "learning_rate": 0.00017534748702114638,
+      "loss": 2.954,
+      "step": 38957
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6268248558044434,
+      "learning_rate": 0.0001753334829744501,
+      "loss": 2.9197,
+      "step": 38958
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6551121473312378,
+      "learning_rate": 0.00017531947925610155,
+      "loss": 3.0124,
+      "step": 38959
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7056483030319214,
+      "learning_rate": 0.00017530547586613738,
+      "loss": 3.1388,
+      "step": 38960
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6477718949317932,
+      "learning_rate": 0.00017529147280459466,
+      "loss": 3.0884,
+      "step": 38961
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6173779964447021,
+      "learning_rate": 0.00017527747007151022,
+      "loss": 3.0833,
+      "step": 38962
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6646998524665833,
+      "learning_rate": 0.00017526346766692083,
+      "loss": 3.027,
+      "step": 38963
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7051180005073547,
+      "learning_rate": 0.0001752494655908636,
+      "loss": 2.7579,
+      "step": 38964
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6498520374298096,
+      "learning_rate": 0.00017523546384337515,
+      "loss": 2.9913,
+      "step": 38965
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6855219602584839,
+      "learning_rate": 0.00017522146242449245,
+      "loss": 2.9673,
+      "step": 38966
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6545772552490234,
+      "learning_rate": 0.00017520746133425252,
+      "loss": 3.0786,
+      "step": 38967
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7005255222320557,
+      "learning_rate": 0.00017519346057269207,
+      "loss": 2.9772,
+      "step": 38968
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.652550220489502,
+      "learning_rate": 0.0001751794601398481,
+      "loss": 3.0452,
+      "step": 38969
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.626100480556488,
+      "learning_rate": 0.00017516546003575726,
+      "loss": 3.0872,
+      "step": 38970
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6244769096374512,
+      "learning_rate": 0.00017515146026045664,
+      "loss": 2.9593,
+      "step": 38971
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6298187375068665,
+      "learning_rate": 0.0001751374608139831,
+      "loss": 3.0651,
+      "step": 38972
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.611535906791687,
+      "learning_rate": 0.00017512346169637336,
+      "loss": 3.2176,
+      "step": 38973
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6949870586395264,
+      "learning_rate": 0.00017510946290766446,
+      "loss": 3.2544,
+      "step": 38974
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6494085192680359,
+      "learning_rate": 0.00017509546444789315,
+      "loss": 2.9234,
+      "step": 38975
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.656135618686676,
+      "learning_rate": 0.00017508146631709643,
+      "loss": 3.0353,
+      "step": 38976
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6190617084503174,
+      "learning_rate": 0.00017506746851531107,
+      "loss": 2.99,
+      "step": 38977
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6905009746551514,
+      "learning_rate": 0.0001750534710425739,
+      "loss": 2.7008,
+      "step": 38978
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6204220056533813,
+      "learning_rate": 0.0001750394738989219,
+      "loss": 3.1247,
+      "step": 38979
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6508306264877319,
+      "learning_rate": 0.00017502547708439178,
+      "loss": 3.0195,
+      "step": 38980
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.684338390827179,
+      "learning_rate": 0.00017501148059902047,
+      "loss": 2.9336,
+      "step": 38981
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6646551489830017,
+      "learning_rate": 0.000174997484442845,
+      "loss": 2.8238,
+      "step": 38982
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7016298770904541,
+      "learning_rate": 0.000174983488615902,
+      "loss": 3.1572,
+      "step": 38983
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7458086013793945,
+      "learning_rate": 0.00017496949311822854,
+      "loss": 2.8436,
+      "step": 38984
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6394898891448975,
+      "learning_rate": 0.00017495549794986138,
+      "loss": 2.989,
+      "step": 38985
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6666737198829651,
+      "learning_rate": 0.00017494150311083724,
+      "loss": 2.8933,
+      "step": 38986
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6851913332939148,
+      "learning_rate": 0.00017492750860119324,
+      "loss": 2.8082,
+      "step": 38987
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.674731433391571,
+      "learning_rate": 0.00017491351442096599,
+      "loss": 3.1715,
+      "step": 38988
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6982241868972778,
+      "learning_rate": 0.0001748995205701926,
+      "loss": 2.9834,
+      "step": 38989
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7044297456741333,
+      "learning_rate": 0.00017488552704890965,
+      "loss": 2.8811,
+      "step": 38990
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6905117630958557,
+      "learning_rate": 0.00017487153385715426,
+      "loss": 2.897,
+      "step": 38991
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6385732889175415,
+      "learning_rate": 0.00017485754099496318,
+      "loss": 2.971,
+      "step": 38992
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6887711882591248,
+      "learning_rate": 0.0001748435484623732,
+      "loss": 2.9601,
+      "step": 38993
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6429010033607483,
+      "learning_rate": 0.00017482955625942125,
+      "loss": 3.0313,
+      "step": 38994
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6536113023757935,
+      "learning_rate": 0.00017481556438614411,
+      "loss": 2.9513,
+      "step": 38995
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6555978655815125,
+      "learning_rate": 0.00017480157284257864,
+      "loss": 2.8433,
+      "step": 38996
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6542037725448608,
+      "learning_rate": 0.00017478758162876184,
+      "loss": 2.9936,
+      "step": 38997
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6809813976287842,
+      "learning_rate": 0.00017477359074473034,
+      "loss": 2.9191,
+      "step": 38998
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6719657182693481,
+      "learning_rate": 0.00017475960019052122,
+      "loss": 2.9595,
+      "step": 38999
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6527001857757568,
+      "learning_rate": 0.00017474560996617126,
+      "loss": 3.1502,
+      "step": 39000
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6487961411476135,
+      "learning_rate": 0.0001747316200717171,
+      "loss": 2.6611,
+      "step": 39001
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.689109742641449,
+      "learning_rate": 0.00017471763050719584,
+      "loss": 2.9463,
+      "step": 39002
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6380165815353394,
+      "learning_rate": 0.00017470364127264413,
+      "loss": 3.0582,
+      "step": 39003
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7141815423965454,
+      "learning_rate": 0.00017468965236809906,
+      "loss": 3.0516,
+      "step": 39004
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6605910658836365,
+      "learning_rate": 0.0001746756637935972,
+      "loss": 3.1249,
+      "step": 39005
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6353509426116943,
+      "learning_rate": 0.0001746616755491756,
+      "loss": 3.0085,
+      "step": 39006
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6613842248916626,
+      "learning_rate": 0.0001746476876348711,
+      "loss": 2.9168,
+      "step": 39007
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6803551316261292,
+      "learning_rate": 0.0001746337000507203,
+      "loss": 2.9971,
+      "step": 39008
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6482704877853394,
+      "learning_rate": 0.00017461971279676027,
+      "loss": 3.0769,
+      "step": 39009
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6631468534469604,
+      "learning_rate": 0.00017460572587302777,
+      "loss": 3.2677,
+      "step": 39010
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.740536630153656,
+      "learning_rate": 0.0001745917392795596,
+      "loss": 2.9764,
+      "step": 39011
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6036262512207031,
+      "learning_rate": 0.00017457775301639276,
+      "loss": 3.0704,
+      "step": 39012
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6554529666900635,
+      "learning_rate": 0.00017456376708356404,
+      "loss": 2.9901,
+      "step": 39013
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6793158650398254,
+      "learning_rate": 0.00017454978148111016,
+      "loss": 3.0852,
+      "step": 39014
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6710419654846191,
+      "learning_rate": 0.00017453579620906794,
+      "loss": 3.0756,
+      "step": 39015
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.65449458360672,
+      "learning_rate": 0.00017452181126747425,
+      "loss": 3.0163,
+      "step": 39016
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6442614793777466,
+      "learning_rate": 0.0001745078266563661,
+      "loss": 2.8513,
+      "step": 39017
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6554681658744812,
+      "learning_rate": 0.00017449384237578004,
+      "loss": 3.0563,
+      "step": 39018
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6386138796806335,
+      "learning_rate": 0.00017447985842575315,
+      "loss": 3.106,
+      "step": 39019
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6400095820426941,
+      "learning_rate": 0.0001744658748063221,
+      "loss": 3.0464,
+      "step": 39020
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6746090650558472,
+      "learning_rate": 0.00017445189151752387,
+      "loss": 3.0485,
+      "step": 39021
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6967078447341919,
+      "learning_rate": 0.0001744379085593952,
+      "loss": 2.9496,
+      "step": 39022
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6654866337776184,
+      "learning_rate": 0.00017442392593197274,
+      "loss": 3.1855,
+      "step": 39023
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6376835107803345,
+      "learning_rate": 0.00017440994363529365,
+      "loss": 2.8765,
+      "step": 39024
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6771338582038879,
+      "learning_rate": 0.00017439596166939452,
+      "loss": 2.9306,
+      "step": 39025
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6553162932395935,
+      "learning_rate": 0.0001743819800343122,
+      "loss": 2.962,
+      "step": 39026
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6516581773757935,
+      "learning_rate": 0.00017436799873008372,
+      "loss": 2.978,
+      "step": 39027
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6631935834884644,
+      "learning_rate": 0.00017435401775674574,
+      "loss": 3.1225,
+      "step": 39028
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6429182291030884,
+      "learning_rate": 0.00017434003711433508,
+      "loss": 2.9737,
+      "step": 39029
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.634573757648468,
+      "learning_rate": 0.00017432605680288847,
+      "loss": 3.0406,
+      "step": 39030
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.613707423210144,
+      "learning_rate": 0.00017431207682244295,
+      "loss": 3.0347,
+      "step": 39031
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.681470513343811,
+      "learning_rate": 0.0001742980971730351,
+      "loss": 3.0185,
+      "step": 39032
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6601608395576477,
+      "learning_rate": 0.00017428411785470194,
+      "loss": 2.944,
+      "step": 39033
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6461837291717529,
+      "learning_rate": 0.00017427013886748028,
+      "loss": 3.1707,
+      "step": 39034
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6573340892791748,
+      "learning_rate": 0.0001742561602114069,
+      "loss": 3.0512,
+      "step": 39035
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6389563083648682,
+      "learning_rate": 0.00017424218188651859,
+      "loss": 2.9478,
+      "step": 39036
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6320285797119141,
+      "learning_rate": 0.00017422820389285202,
+      "loss": 3.2002,
+      "step": 39037
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6769739985466003,
+      "learning_rate": 0.0001742142262304442,
+      "loss": 3.0265,
+      "step": 39038
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6444475054740906,
+      "learning_rate": 0.000174200248899332,
+      "loss": 2.9721,
+      "step": 39039
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6511723399162292,
+      "learning_rate": 0.00017418627189955198,
+      "loss": 3.2104,
+      "step": 39040
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8678764700889587,
+      "learning_rate": 0.00017417229523114126,
+      "loss": 3.0062,
+      "step": 39041
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6624598503112793,
+      "learning_rate": 0.0001741583188941364,
+      "loss": 3.0018,
+      "step": 39042
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6774799823760986,
+      "learning_rate": 0.00017414434288857438,
+      "loss": 2.7333,
+      "step": 39043
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6489653587341309,
+      "learning_rate": 0.00017413036721449198,
+      "loss": 3.112,
+      "step": 39044
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6598437428474426,
+      "learning_rate": 0.00017411639187192582,
+      "loss": 2.9259,
+      "step": 39045
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6542210578918457,
+      "learning_rate": 0.00017410241686091298,
+      "loss": 2.8695,
+      "step": 39046
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6994691491127014,
+      "learning_rate": 0.00017408844218149008,
+      "loss": 2.8518,
+      "step": 39047
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6199400424957275,
+      "learning_rate": 0.00017407446783369393,
+      "loss": 2.9526,
+      "step": 39048
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6711297631263733,
+      "learning_rate": 0.00017406049381756152,
+      "loss": 2.8745,
+      "step": 39049
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6305173635482788,
+      "learning_rate": 0.00017404652013312954,
+      "loss": 3.0848,
+      "step": 39050
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6580938100814819,
+      "learning_rate": 0.0001740325467804348,
+      "loss": 3.0753,
+      "step": 39051
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6921206116676331,
+      "learning_rate": 0.00017401857375951395,
+      "loss": 3.0142,
+      "step": 39052
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0677298307418823,
+      "learning_rate": 0.00017400460107040394,
+      "loss": 3.0512,
+      "step": 39053
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6595743298530579,
+      "learning_rate": 0.0001739906287131417,
+      "loss": 3.0075,
+      "step": 39054
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6712070107460022,
+      "learning_rate": 0.00017397665668776374,
+      "loss": 3.1811,
+      "step": 39055
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6308397650718689,
+      "learning_rate": 0.00017396268499430718,
+      "loss": 3.1166,
+      "step": 39056
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6080089807510376,
+      "learning_rate": 0.0001739487136328086,
+      "loss": 3.0898,
+      "step": 39057
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6288923025131226,
+      "learning_rate": 0.00017393474260330477,
+      "loss": 3.0053,
+      "step": 39058
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6430457830429077,
+      "learning_rate": 0.00017392077190583267,
+      "loss": 3.2299,
+      "step": 39059
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6252223253250122,
+      "learning_rate": 0.00017390680154042888,
+      "loss": 3.121,
+      "step": 39060
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6593741774559021,
+      "learning_rate": 0.00017389283150713038,
+      "loss": 3.0993,
+      "step": 39061
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6887229681015015,
+      "learning_rate": 0.0001738788618059738,
+      "loss": 3.1088,
+      "step": 39062
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6626426577568054,
+      "learning_rate": 0.00017386489243699608,
+      "loss": 3.0802,
+      "step": 39063
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.663589358329773,
+      "learning_rate": 0.00017385092340023404,
+      "loss": 2.8885,
+      "step": 39064
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6585755944252014,
+      "learning_rate": 0.00017383695469572435,
+      "loss": 2.9546,
+      "step": 39065
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6396164894104004,
+      "learning_rate": 0.00017382298632350387,
+      "loss": 3.1609,
+      "step": 39066
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.647235095500946,
+      "learning_rate": 0.00017380901828360925,
+      "loss": 2.9741,
+      "step": 39067
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6416546702384949,
+      "learning_rate": 0.0001737950505760774,
+      "loss": 3.097,
+      "step": 39068
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.697792649269104,
+      "learning_rate": 0.00017378108320094518,
+      "loss": 2.6795,
+      "step": 39069
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.716894805431366,
+      "learning_rate": 0.0001737671161582492,
+      "loss": 2.7536,
+      "step": 39070
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6366348266601562,
+      "learning_rate": 0.00017375314944802648,
+      "loss": 2.8962,
+      "step": 39071
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6648399829864502,
+      "learning_rate": 0.00017373918307031364,
+      "loss": 3.1113,
+      "step": 39072
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6838628649711609,
+      "learning_rate": 0.0001737252170251474,
+      "loss": 3.1721,
+      "step": 39073
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6086000204086304,
+      "learning_rate": 0.0001737112513125647,
+      "loss": 3.0741,
+      "step": 39074
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6315998435020447,
+      "learning_rate": 0.0001736972859326022,
+      "loss": 3.0657,
+      "step": 39075
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6525732278823853,
+      "learning_rate": 0.0001736833208852968,
+      "loss": 2.9191,
+      "step": 39076
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6616672873497009,
+      "learning_rate": 0.00017366935617068517,
+      "loss": 2.8436,
+      "step": 39077
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6813806295394897,
+      "learning_rate": 0.00017365539178880422,
+      "loss": 3.0425,
+      "step": 39078
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1355561017990112,
+      "learning_rate": 0.00017364142773969066,
+      "loss": 2.9847,
+      "step": 39079
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6255960464477539,
+      "learning_rate": 0.00017362746402338114,
+      "loss": 2.9503,
+      "step": 39080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6622676253318787,
+      "learning_rate": 0.0001736135006399127,
+      "loss": 3.4366,
+      "step": 39081
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6947815418243408,
+      "learning_rate": 0.00017359953758932183,
+      "loss": 2.9106,
+      "step": 39082
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6500281095504761,
+      "learning_rate": 0.00017358557487164548,
+      "loss": 2.6224,
+      "step": 39083
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6997030973434448,
+      "learning_rate": 0.0001735716124869205,
+      "loss": 2.9147,
+      "step": 39084
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6823716759681702,
+      "learning_rate": 0.0001735576504351834,
+      "loss": 2.9929,
+      "step": 39085
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6616767644882202,
+      "learning_rate": 0.0001735436887164713,
+      "loss": 2.9272,
+      "step": 39086
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6855137348175049,
+      "learning_rate": 0.00017352972733082075,
+      "loss": 2.9099,
+      "step": 39087
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6969751715660095,
+      "learning_rate": 0.00017351576627826847,
+      "loss": 2.7265,
+      "step": 39088
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6577869057655334,
+      "learning_rate": 0.00017350180555885143,
+      "loss": 3.2084,
+      "step": 39089
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6708926558494568,
+      "learning_rate": 0.00017348784517260614,
+      "loss": 3.1078,
+      "step": 39090
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6796287894248962,
+      "learning_rate": 0.00017347388511956965,
+      "loss": 2.8497,
+      "step": 39091
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6603464484214783,
+      "learning_rate": 0.00017345992539977847,
+      "loss": 2.8234,
+      "step": 39092
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.635272741317749,
+      "learning_rate": 0.00017344596601326964,
+      "loss": 3.1138,
+      "step": 39093
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6617568731307983,
+      "learning_rate": 0.00017343200696007979,
+      "loss": 3.1461,
+      "step": 39094
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6393948793411255,
+      "learning_rate": 0.0001734180482402455,
+      "loss": 2.8271,
+      "step": 39095
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6940094232559204,
+      "learning_rate": 0.0001734040898538039,
+      "loss": 3.0738,
+      "step": 39096
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6491406559944153,
+      "learning_rate": 0.00017339013180079136,
+      "loss": 3.0714,
+      "step": 39097
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6679297089576721,
+      "learning_rate": 0.0001733761740812449,
+      "loss": 2.9234,
+      "step": 39098
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6698212027549744,
+      "learning_rate": 0.00017336221669520135,
+      "loss": 2.956,
+      "step": 39099
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6277477145195007,
+      "learning_rate": 0.00017334825964269722,
+      "loss": 2.9522,
+      "step": 39100
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.63001549243927,
+      "learning_rate": 0.00017333430292376965,
+      "loss": 3.1925,
+      "step": 39101
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6419739127159119,
+      "learning_rate": 0.0001733203465384549,
+      "loss": 2.9653,
+      "step": 39102
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6369063258171082,
+      "learning_rate": 0.00017330639048678995,
+      "loss": 2.8996,
+      "step": 39103
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7089871764183044,
+      "learning_rate": 0.0001732924347688117,
+      "loss": 2.9241,
+      "step": 39104
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.639457106590271,
+      "learning_rate": 0.00017327847938455673,
+      "loss": 2.9672,
+      "step": 39105
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.641934335231781,
+      "learning_rate": 0.00017326452433406194,
+      "loss": 2.8944,
+      "step": 39106
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6629417538642883,
+      "learning_rate": 0.0001732505696173639,
+      "loss": 2.8033,
+      "step": 39107
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6630560159683228,
+      "learning_rate": 0.00017323661523449958,
+      "loss": 2.9384,
+      "step": 39108
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6467246413230896,
+      "learning_rate": 0.00017322266118550557,
+      "loss": 2.9888,
+      "step": 39109
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6524175405502319,
+      "learning_rate": 0.00017320870747041862,
+      "loss": 3.2388,
+      "step": 39110
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6555929183959961,
+      "learning_rate": 0.00017319475408927564,
+      "loss": 3.0857,
+      "step": 39111
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6451212167739868,
+      "learning_rate": 0.00017318080104211314,
+      "loss": 2.8654,
+      "step": 39112
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6508936285972595,
+      "learning_rate": 0.00017316684832896815,
+      "loss": 2.9725,
+      "step": 39113
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6641062498092651,
+      "learning_rate": 0.0001731528959498771,
+      "loss": 2.8809,
+      "step": 39114
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6545625329017639,
+      "learning_rate": 0.0001731389439048771,
+      "loss": 2.9962,
+      "step": 39115
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.652595579624176,
+      "learning_rate": 0.00017312499219400461,
+      "loss": 2.9298,
+      "step": 39116
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6414437890052795,
+      "learning_rate": 0.00017311104081729642,
+      "loss": 2.9284,
+      "step": 39117
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7035130262374878,
+      "learning_rate": 0.00017309708977478942,
+      "loss": 3.058,
+      "step": 39118
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6662630438804626,
+      "learning_rate": 0.00017308313906652016,
+      "loss": 3.0701,
+      "step": 39119
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.645362377166748,
+      "learning_rate": 0.0001730691886925255,
+      "loss": 3.1298,
+      "step": 39120
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6399428844451904,
+      "learning_rate": 0.0001730552386528423,
+      "loss": 3.0839,
+      "step": 39121
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6200828552246094,
+      "learning_rate": 0.000173041288947507,
+      "loss": 2.9407,
+      "step": 39122
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6302321553230286,
+      "learning_rate": 0.00017302733957655662,
+      "loss": 2.9237,
+      "step": 39123
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6495705246925354,
+      "learning_rate": 0.00017301339054002785,
+      "loss": 2.8885,
+      "step": 39124
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.634657621383667,
+      "learning_rate": 0.00017299944183795724,
+      "loss": 3.1106,
+      "step": 39125
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6378446817398071,
+      "learning_rate": 0.00017298549347038178,
+      "loss": 2.7874,
+      "step": 39126
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6395211815834045,
+      "learning_rate": 0.00017297154543733796,
+      "loss": 3.0792,
+      "step": 39127
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6783758997917175,
+      "learning_rate": 0.0001729575977388627,
+      "loss": 2.8279,
+      "step": 39128
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6913485527038574,
+      "learning_rate": 0.00017294365037499263,
+      "loss": 3.101,
+      "step": 39129
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.659443199634552,
+      "learning_rate": 0.00017292970334576466,
+      "loss": 3.1545,
+      "step": 39130
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6611900329589844,
+      "learning_rate": 0.00017291575665121534,
+      "loss": 2.865,
+      "step": 39131
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6406236290931702,
+      "learning_rate": 0.0001729018102913814,
+      "loss": 3.0973,
+      "step": 39132
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6783467531204224,
+      "learning_rate": 0.0001728878642662997,
+      "loss": 3.074,
+      "step": 39133
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6730901002883911,
+      "learning_rate": 0.0001728739185760068,
+      "loss": 2.8599,
+      "step": 39134
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6598025560379028,
+      "learning_rate": 0.00017285997322053954,
+      "loss": 3.1268,
+      "step": 39135
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6816617846488953,
+      "learning_rate": 0.00017284602819993476,
+      "loss": 2.944,
+      "step": 39136
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6511579155921936,
+      "learning_rate": 0.00017283208351422908,
+      "loss": 3.1893,
+      "step": 39137
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.693321943283081,
+      "learning_rate": 0.00017281813916345917,
+      "loss": 3.0119,
+      "step": 39138
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6485716700553894,
+      "learning_rate": 0.00017280419514766174,
+      "loss": 2.7844,
+      "step": 39139
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6522268056869507,
+      "learning_rate": 0.00017279025146687353,
+      "loss": 2.9277,
+      "step": 39140
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6479056477546692,
+      "learning_rate": 0.0001727763081211315,
+      "loss": 2.686,
+      "step": 39141
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6812885999679565,
+      "learning_rate": 0.00017276236511047202,
+      "loss": 2.8859,
+      "step": 39142
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6709456443786621,
+      "learning_rate": 0.0001727484224349321,
+      "loss": 2.9349,
+      "step": 39143
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6877416372299194,
+      "learning_rate": 0.00017273448009454825,
+      "loss": 2.9836,
+      "step": 39144
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6465352177619934,
+      "learning_rate": 0.00017272053808935743,
+      "loss": 3.0815,
+      "step": 39145
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6705036163330078,
+      "learning_rate": 0.0001727065964193962,
+      "loss": 3.1177,
+      "step": 39146
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6689993739128113,
+      "learning_rate": 0.00017269265508470112,
+      "loss": 3.0918,
+      "step": 39147
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6450504660606384,
+      "learning_rate": 0.00017267871408530928,
+      "loss": 3.0432,
+      "step": 39148
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6703768968582153,
+      "learning_rate": 0.00017266477342125702,
+      "loss": 3.0998,
+      "step": 39149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6024968028068542,
+      "learning_rate": 0.00017265083309258127,
+      "loss": 2.9523,
+      "step": 39150
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6327059268951416,
+      "learning_rate": 0.00017263689309931886,
+      "loss": 3.127,
+      "step": 39151
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.709504246711731,
+      "learning_rate": 0.00017262295344150634,
+      "loss": 3.1062,
+      "step": 39152
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6775631308555603,
+      "learning_rate": 0.00017260901411918043,
+      "loss": 3.0224,
+      "step": 39153
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6679512858390808,
+      "learning_rate": 0.00017259507513237777,
+      "loss": 2.8959,
+      "step": 39154
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7322613596916199,
+      "learning_rate": 0.00017258113648113517,
+      "loss": 2.9774,
+      "step": 39155
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6525643467903137,
+      "learning_rate": 0.00017256719816548943,
+      "loss": 3.0671,
+      "step": 39156
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.662835419178009,
+      "learning_rate": 0.00017255326018547707,
+      "loss": 2.9745,
+      "step": 39157
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7078630328178406,
+      "learning_rate": 0.000172539322541135,
+      "loss": 3.0263,
+      "step": 39158
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6778189539909363,
+      "learning_rate": 0.0001725253852324998,
+      "loss": 3.1786,
+      "step": 39159
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7059077024459839,
+      "learning_rate": 0.00017251144825960812,
+      "loss": 2.7726,
+      "step": 39160
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6346893310546875,
+      "learning_rate": 0.00017249751162249686,
+      "loss": 2.9413,
+      "step": 39161
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6974074840545654,
+      "learning_rate": 0.00017248357532120248,
+      "loss": 3.048,
+      "step": 39162
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6645182967185974,
+      "learning_rate": 0.00017246963935576196,
+      "loss": 2.7881,
+      "step": 39163
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6341907978057861,
+      "learning_rate": 0.0001724557037262117,
+      "loss": 3.1263,
+      "step": 39164
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.625954806804657,
+      "learning_rate": 0.00017244176843258862,
+      "loss": 2.7248,
+      "step": 39165
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6558056473731995,
+      "learning_rate": 0.0001724278334749295,
+      "loss": 3.1032,
+      "step": 39166
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6719792485237122,
+      "learning_rate": 0.00017241389885327087,
+      "loss": 2.9548,
+      "step": 39167
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.682371199131012,
+      "learning_rate": 0.00017239996456764952,
+      "loss": 3.1351,
+      "step": 39168
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6656419634819031,
+      "learning_rate": 0.00017238603061810195,
+      "loss": 2.8981,
+      "step": 39169
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6694563031196594,
+      "learning_rate": 0.00017237209700466502,
+      "loss": 3.2168,
+      "step": 39170
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7101410627365112,
+      "learning_rate": 0.00017235816372737558,
+      "loss": 2.9831,
+      "step": 39171
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6762426495552063,
+      "learning_rate": 0.00017234423078627003,
+      "loss": 2.8293,
+      "step": 39172
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.617826521396637,
+      "learning_rate": 0.00017233029818138532,
+      "loss": 2.8919,
+      "step": 39173
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7168938517570496,
+      "learning_rate": 0.00017231636591275807,
+      "loss": 2.9602,
+      "step": 39174
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6657389998435974,
+      "learning_rate": 0.00017230243398042475,
+      "loss": 2.9613,
+      "step": 39175
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6456219553947449,
+      "learning_rate": 0.0001722885023844224,
+      "loss": 2.8773,
+      "step": 39176
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6605064868927002,
+      "learning_rate": 0.00017227457112478745,
+      "loss": 3.088,
+      "step": 39177
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6815346479415894,
+      "learning_rate": 0.0001722606402015568,
+      "loss": 3.0484,
+      "step": 39178
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6612982153892517,
+      "learning_rate": 0.0001722467096147669,
+      "loss": 3.0271,
+      "step": 39179
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6247538328170776,
+      "learning_rate": 0.00017223277936445475,
+      "loss": 2.8698,
+      "step": 39180
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6820695996284485,
+      "learning_rate": 0.00017221884945065687,
+      "loss": 3.0734,
+      "step": 39181
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6383649110794067,
+      "learning_rate": 0.0001722049198734098,
+      "loss": 3.1281,
+      "step": 39182
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6495553255081177,
+      "learning_rate": 0.00017219099063275052,
+      "loss": 3.0078,
+      "step": 39183
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6677046418190002,
+      "learning_rate": 0.0001721770617287154,
+      "loss": 3.0541,
+      "step": 39184
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6571451425552368,
+      "learning_rate": 0.00017216313316134135,
+      "loss": 2.964,
+      "step": 39185
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6236145496368408,
+      "learning_rate": 0.0001721492049306651,
+      "loss": 2.8809,
+      "step": 39186
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6534735560417175,
+      "learning_rate": 0.00017213527703672314,
+      "loss": 2.8744,
+      "step": 39187
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6529717445373535,
+      "learning_rate": 0.00017212134947955235,
+      "loss": 3.0686,
+      "step": 39188
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6746391654014587,
+      "learning_rate": 0.00017210742225918935,
+      "loss": 3.1402,
+      "step": 39189
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6508035659790039,
+      "learning_rate": 0.00017209349537567072,
+      "loss": 2.9857,
+      "step": 39190
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6782659888267517,
+      "learning_rate": 0.00017207956882903314,
+      "loss": 3.0049,
+      "step": 39191
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6732063293457031,
+      "learning_rate": 0.00017206564261931334,
+      "loss": 2.904,
+      "step": 39192
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.719319760799408,
+      "learning_rate": 0.00017205171674654815,
+      "loss": 3.0394,
+      "step": 39193
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6435275673866272,
+      "learning_rate": 0.00017203779121077397,
+      "loss": 2.911,
+      "step": 39194
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6958181858062744,
+      "learning_rate": 0.00017202386601202776,
+      "loss": 3.1103,
+      "step": 39195
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6404264569282532,
+      "learning_rate": 0.00017200994115034606,
+      "loss": 2.9457,
+      "step": 39196
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6635984778404236,
+      "learning_rate": 0.00017199601662576543,
+      "loss": 2.8922,
+      "step": 39197
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6762363910675049,
+      "learning_rate": 0.00017198209243832278,
+      "loss": 3.0213,
+      "step": 39198
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6661695241928101,
+      "learning_rate": 0.0001719681685880545,
+      "loss": 2.809,
+      "step": 39199
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6384485363960266,
+      "learning_rate": 0.00017195424507499755,
+      "loss": 3.1867,
+      "step": 39200
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6842610836029053,
+      "learning_rate": 0.0001719403218991884,
+      "loss": 2.9495,
+      "step": 39201
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6691260933876038,
+      "learning_rate": 0.00017192639906066386,
+      "loss": 3.0114,
+      "step": 39202
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6943384408950806,
+      "learning_rate": 0.00017191247655946057,
+      "loss": 3.1103,
+      "step": 39203
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6369804739952087,
+      "learning_rate": 0.00017189855439561503,
+      "loss": 3.121,
+      "step": 39204
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6395381689071655,
+      "learning_rate": 0.00017188463256916417,
+      "loss": 2.9585,
+      "step": 39205
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.621781051158905,
+      "learning_rate": 0.00017187071108014447,
+      "loss": 2.9129,
+      "step": 39206
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6474804282188416,
+      "learning_rate": 0.00017185678992859258,
+      "loss": 2.9731,
+      "step": 39207
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6578860282897949,
+      "learning_rate": 0.0001718428691145454,
+      "loss": 2.8956,
+      "step": 39208
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7093460559844971,
+      "learning_rate": 0.00017182894863803928,
+      "loss": 2.9726,
+      "step": 39209
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6357724070549011,
+      "learning_rate": 0.00017181502849911121,
+      "loss": 3.2573,
+      "step": 39210
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6507198214530945,
+      "learning_rate": 0.0001718011086977977,
+      "loss": 2.9724,
+      "step": 39211
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6744953393936157,
+      "learning_rate": 0.00017178718923413524,
+      "loss": 3.127,
+      "step": 39212
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6369718313217163,
+      "learning_rate": 0.00017177327010816076,
+      "loss": 2.9405,
+      "step": 39213
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6532459259033203,
+      "learning_rate": 0.00017175935131991067,
+      "loss": 2.9442,
+      "step": 39214
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6640715003013611,
+      "learning_rate": 0.00017174543286942192,
+      "loss": 2.7625,
+      "step": 39215
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7218378782272339,
+      "learning_rate": 0.0001717315147567309,
+      "loss": 3.116,
+      "step": 39216
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7001231908798218,
+      "learning_rate": 0.0001717175969818745,
+      "loss": 2.9236,
+      "step": 39217
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6841039657592773,
+      "learning_rate": 0.00017170367954488926,
+      "loss": 2.7602,
+      "step": 39218
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.67069011926651,
+      "learning_rate": 0.00017168976244581174,
+      "loss": 3.0251,
+      "step": 39219
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6532872319221497,
+      "learning_rate": 0.0001716758456846788,
+      "loss": 3.0348,
+      "step": 39220
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6450393795967102,
+      "learning_rate": 0.00017166192926152684,
+      "loss": 3.1042,
+      "step": 39221
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6509761214256287,
+      "learning_rate": 0.0001716480131763927,
+      "loss": 2.8918,
+      "step": 39222
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6896138191223145,
+      "learning_rate": 0.00017163409742931303,
+      "loss": 3.0114,
+      "step": 39223
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6828722953796387,
+      "learning_rate": 0.00017162018202032442,
+      "loss": 3.0185,
+      "step": 39224
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6410359144210815,
+      "learning_rate": 0.0001716062669494636,
+      "loss": 3.0876,
+      "step": 39225
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6558801531791687,
+      "learning_rate": 0.0001715923522167672,
+      "loss": 2.8841,
+      "step": 39226
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6510255336761475,
+      "learning_rate": 0.00017157843782227169,
+      "loss": 2.9658,
+      "step": 39227
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6774110794067383,
+      "learning_rate": 0.00017156452376601398,
+      "loss": 3.0839,
+      "step": 39228
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6347213983535767,
+      "learning_rate": 0.0001715506100480305,
+      "loss": 3.1037,
+      "step": 39229
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6718534827232361,
+      "learning_rate": 0.00017153669666835809,
+      "loss": 3.0421,
+      "step": 39230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.674842894077301,
+      "learning_rate": 0.00017152278362703318,
+      "loss": 3.0798,
+      "step": 39231
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6837745308876038,
+      "learning_rate": 0.00017150887092409265,
+      "loss": 2.7996,
+      "step": 39232
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6717110276222229,
+      "learning_rate": 0.00017149495855957303,
+      "loss": 2.842,
+      "step": 39233
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6553992629051208,
+      "learning_rate": 0.00017148104653351087,
+      "loss": 2.9463,
+      "step": 39234
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.702616810798645,
+      "learning_rate": 0.00017146713484594297,
+      "loss": 2.8664,
+      "step": 39235
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6626030206680298,
+      "learning_rate": 0.0001714532234969058,
+      "loss": 3.052,
+      "step": 39236
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6390082240104675,
+      "learning_rate": 0.00017143931248643607,
+      "loss": 3.0915,
+      "step": 39237
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6931836009025574,
+      "learning_rate": 0.0001714254018145706,
+      "loss": 3.2053,
+      "step": 39238
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6288201212882996,
+      "learning_rate": 0.0001714114914813459,
+      "loss": 3.1003,
+      "step": 39239
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6410542130470276,
+      "learning_rate": 0.00017139758148679854,
+      "loss": 3.0281,
+      "step": 39240
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.664252758026123,
+      "learning_rate": 0.0001713836718309651,
+      "loss": 3.0345,
+      "step": 39241
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6826432943344116,
+      "learning_rate": 0.00017136976251388232,
+      "loss": 3.0711,
+      "step": 39242
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6907158493995667,
+      "learning_rate": 0.00017135585353558695,
+      "loss": 2.7451,
+      "step": 39243
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6534987092018127,
+      "learning_rate": 0.00017134194489611538,
+      "loss": 3.0869,
+      "step": 39244
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6228570342063904,
+      "learning_rate": 0.0001713280365955045,
+      "loss": 3.0009,
+      "step": 39245
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.697776198387146,
+      "learning_rate": 0.00017131412863379068,
+      "loss": 3.175,
+      "step": 39246
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6710581183433533,
+      "learning_rate": 0.0001713002210110108,
+      "loss": 2.8232,
+      "step": 39247
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7085484862327576,
+      "learning_rate": 0.00017128631372720138,
+      "loss": 2.9369,
+      "step": 39248
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.666691780090332,
+      "learning_rate": 0.00017127240678239884,
+      "loss": 2.9581,
+      "step": 39249
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.686436653137207,
+      "learning_rate": 0.0001712585001766402,
+      "loss": 2.9945,
+      "step": 39250
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6662260890007019,
+      "learning_rate": 0.00017124459390996174,
+      "loss": 2.924,
+      "step": 39251
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6701951026916504,
+      "learning_rate": 0.0001712306879824003,
+      "loss": 3.0694,
+      "step": 39252
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6563501358032227,
+      "learning_rate": 0.00017121678239399252,
+      "loss": 3.0597,
+      "step": 39253
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6329666972160339,
+      "learning_rate": 0.00017120287714477494,
+      "loss": 2.7156,
+      "step": 39254
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6433852314949036,
+      "learning_rate": 0.0001711889722347842,
+      "loss": 2.9309,
+      "step": 39255
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7456541657447815,
+      "learning_rate": 0.00017117506766405684,
+      "loss": 2.9767,
+      "step": 39256
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6476452350616455,
+      "learning_rate": 0.0001711611634326295,
+      "loss": 2.9284,
+      "step": 39257
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6435794830322266,
+      "learning_rate": 0.00017114725954053904,
+      "loss": 3.1301,
+      "step": 39258
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6286764144897461,
+      "learning_rate": 0.0001711333559878217,
+      "loss": 3.2672,
+      "step": 39259
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6440817713737488,
+      "learning_rate": 0.00017111945277451445,
+      "loss": 3.0443,
+      "step": 39260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6571938991546631,
+      "learning_rate": 0.0001711055499006538,
+      "loss": 2.9511,
+      "step": 39261
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7274107933044434,
+      "learning_rate": 0.00017109164736627618,
+      "loss": 2.9802,
+      "step": 39262
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6378130912780762,
+      "learning_rate": 0.00017107774517141847,
+      "loss": 3.0211,
+      "step": 39263
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6885344386100769,
+      "learning_rate": 0.00017106384331611705,
+      "loss": 2.9844,
+      "step": 39264
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6548222303390503,
+      "learning_rate": 0.0001710499418004088,
+      "loss": 3.0318,
+      "step": 39265
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.681775689125061,
+      "learning_rate": 0.00017103604062433,
+      "loss": 2.7903,
+      "step": 39266
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.674473226070404,
+      "learning_rate": 0.00017102213978791763,
+      "loss": 3.0038,
+      "step": 39267
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6563646793365479,
+      "learning_rate": 0.00017100823929120797,
+      "loss": 2.8295,
+      "step": 39268
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6491327881813049,
+      "learning_rate": 0.00017099433913423791,
+      "loss": 3.0033,
+      "step": 39269
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6606160402297974,
+      "learning_rate": 0.00017098043931704394,
+      "loss": 3.0782,
+      "step": 39270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6784059405326843,
+      "learning_rate": 0.00017096653983966252,
+      "loss": 3.1224,
+      "step": 39271
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6741899251937866,
+      "learning_rate": 0.00017095264070213055,
+      "loss": 2.8379,
+      "step": 39272
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6609475016593933,
+      "learning_rate": 0.00017093874190448434,
+      "loss": 3.0115,
+      "step": 39273
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7126678228378296,
+      "learning_rate": 0.00017092484344676068,
+      "loss": 2.9805,
+      "step": 39274
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6484171152114868,
+      "learning_rate": 0.00017091094532899624,
+      "loss": 2.8915,
+      "step": 39275
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6500241160392761,
+      "learning_rate": 0.0001708970475512275,
+      "loss": 2.8879,
+      "step": 39276
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6842443943023682,
+      "learning_rate": 0.00017088315011349115,
+      "loss": 3.0103,
+      "step": 39277
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6493595838546753,
+      "learning_rate": 0.00017086925301582358,
+      "loss": 2.9825,
+      "step": 39278
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6276535391807556,
+      "learning_rate": 0.00017085535625826156,
+      "loss": 3.0221,
+      "step": 39279
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.693079948425293,
+      "learning_rate": 0.00017084145984084175,
+      "loss": 3.1203,
+      "step": 39280
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6312422156333923,
+      "learning_rate": 0.0001708275637636006,
+      "loss": 3.001,
+      "step": 39281
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6259576082229614,
+      "learning_rate": 0.0001708136680265749,
+      "loss": 2.9753,
+      "step": 39282
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6311910152435303,
+      "learning_rate": 0.00017079977262980116,
+      "loss": 3.1335,
+      "step": 39283
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6822161674499512,
+      "learning_rate": 0.0001707858775733158,
+      "loss": 2.8774,
+      "step": 39284
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7114952802658081,
+      "learning_rate": 0.0001707719828571557,
+      "loss": 3.1555,
+      "step": 39285
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6649879813194275,
+      "learning_rate": 0.00017075808848135722,
+      "loss": 3.0921,
+      "step": 39286
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6248159408569336,
+      "learning_rate": 0.00017074419444595716,
+      "loss": 2.9811,
+      "step": 39287
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6573299169540405,
+      "learning_rate": 0.00017073030075099191,
+      "loss": 2.8717,
+      "step": 39288
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7020032405853271,
+      "learning_rate": 0.00017071640739649814,
+      "loss": 2.8755,
+      "step": 39289
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6383627653121948,
+      "learning_rate": 0.00017070251438251264,
+      "loss": 2.9567,
+      "step": 39290
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7252497673034668,
+      "learning_rate": 0.00017068862170907181,
+      "loss": 2.9716,
+      "step": 39291
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.666100025177002,
+      "learning_rate": 0.00017067472937621229,
+      "loss": 3.1238,
+      "step": 39292
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6566705107688904,
+      "learning_rate": 0.00017066083738397048,
+      "loss": 3.0706,
+      "step": 39293
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.657951295375824,
+      "learning_rate": 0.00017064694573238316,
+      "loss": 3.0127,
+      "step": 39294
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6560387015342712,
+      "learning_rate": 0.000170633054421487,
+      "loss": 2.9246,
+      "step": 39295
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6837169528007507,
+      "learning_rate": 0.00017061916345131834,
+      "loss": 3.0309,
+      "step": 39296
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7252517938613892,
+      "learning_rate": 0.00017060527282191406,
+      "loss": 2.6591,
+      "step": 39297
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6759359240531921,
+      "learning_rate": 0.00017059138253331055,
+      "loss": 3.1497,
+      "step": 39298
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.631263256072998,
+      "learning_rate": 0.00017057749258554434,
+      "loss": 3.1716,
+      "step": 39299
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6345106959342957,
+      "learning_rate": 0.0001705636029786522,
+      "loss": 2.8964,
+      "step": 39300
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6319267153739929,
+      "learning_rate": 0.00017054971371267054,
+      "loss": 3.0213,
+      "step": 39301
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6083059906959534,
+      "learning_rate": 0.00017053582478763614,
+      "loss": 2.8726,
+      "step": 39302
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6584345698356628,
+      "learning_rate": 0.0001705219362035853,
+      "loss": 2.9741,
+      "step": 39303
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.720427930355072,
+      "learning_rate": 0.0001705080479605549,
+      "loss": 2.8017,
+      "step": 39304
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6475998759269714,
+      "learning_rate": 0.00017049416005858142,
+      "loss": 3.1804,
+      "step": 39305
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6432875394821167,
+      "learning_rate": 0.00017048027249770122,
+      "loss": 2.8589,
+      "step": 39306
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.645972490310669,
+      "learning_rate": 0.0001704663852779512,
+      "loss": 2.9266,
+      "step": 39307
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6561703681945801,
+      "learning_rate": 0.00017045249839936768,
+      "loss": 3.0237,
+      "step": 39308
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6930022239685059,
+      "learning_rate": 0.00017043861186198732,
+      "loss": 2.9589,
+      "step": 39309
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6722618937492371,
+      "learning_rate": 0.00017042472566584688,
+      "loss": 3.1136,
+      "step": 39310
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6967858672142029,
+      "learning_rate": 0.00017041083981098263,
+      "loss": 2.8958,
+      "step": 39311
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.649520218372345,
+      "learning_rate": 0.00017039695429743143,
+      "loss": 3.0567,
+      "step": 39312
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6798179745674133,
+      "learning_rate": 0.0001703830691252297,
+      "loss": 3.0234,
+      "step": 39313
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6915568709373474,
+      "learning_rate": 0.00017036918429441387,
+      "loss": 2.9978,
+      "step": 39314
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6442030072212219,
+      "learning_rate": 0.0001703552998050208,
+      "loss": 3.2237,
+      "step": 39315
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6549793481826782,
+      "learning_rate": 0.00017034141565708683,
+      "loss": 3.2327,
+      "step": 39316
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6851274371147156,
+      "learning_rate": 0.0001703275318506487,
+      "loss": 2.8293,
+      "step": 39317
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6663837432861328,
+      "learning_rate": 0.0001703136483857428,
+      "loss": 2.8694,
+      "step": 39318
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6201865077018738,
+      "learning_rate": 0.00017029976526240588,
+      "loss": 2.9923,
+      "step": 39319
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6380475759506226,
+      "learning_rate": 0.00017028588248067445,
+      "loss": 2.9472,
+      "step": 39320
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6493070721626282,
+      "learning_rate": 0.0001702720000405849,
+      "loss": 3.1009,
+      "step": 39321
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6758432388305664,
+      "learning_rate": 0.00017025811794217408,
+      "loss": 2.8572,
+      "step": 39322
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6773945093154907,
+      "learning_rate": 0.00017024423618547824,
+      "loss": 3.0127,
+      "step": 39323
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6368323564529419,
+      "learning_rate": 0.00017023035477053416,
+      "loss": 2.8353,
+      "step": 39324
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6623914241790771,
+      "learning_rate": 0.00017021647369737844,
+      "loss": 3.0751,
+      "step": 39325
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6415942311286926,
+      "learning_rate": 0.00017020259296604743,
+      "loss": 2.9105,
+      "step": 39326
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6625025868415833,
+      "learning_rate": 0.00017018871257657808,
+      "loss": 2.912,
+      "step": 39327
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6630984544754028,
+      "learning_rate": 0.00017017483252900641,
+      "loss": 2.9823,
+      "step": 39328
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.637005090713501,
+      "learning_rate": 0.00017016095282336926,
+      "loss": 3.0265,
+      "step": 39329
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6559372544288635,
+      "learning_rate": 0.0001701470734597033,
+      "loss": 3.1208,
+      "step": 39330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.683228075504303,
+      "learning_rate": 0.0001701331944380448,
+      "loss": 3.2746,
+      "step": 39331
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6639457941055298,
+      "learning_rate": 0.00017011931575843062,
+      "loss": 3.0369,
+      "step": 39332
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.691205620765686,
+      "learning_rate": 0.00017010543742089707,
+      "loss": 2.921,
+      "step": 39333
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6341986656188965,
+      "learning_rate": 0.00017009155942548088,
+      "loss": 3.1622,
+      "step": 39334
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6385431289672852,
+      "learning_rate": 0.00017007768177221857,
+      "loss": 2.8519,
+      "step": 39335
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6719568967819214,
+      "learning_rate": 0.00017006380446114646,
+      "loss": 2.9594,
+      "step": 39336
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6813459396362305,
+      "learning_rate": 0.00017004992749230144,
+      "loss": 2.9986,
+      "step": 39337
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7117205262184143,
+      "learning_rate": 0.0001700360508657198,
+      "loss": 3.0113,
+      "step": 39338
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6474804878234863,
+      "learning_rate": 0.00017002217458143813,
+      "loss": 2.8304,
+      "step": 39339
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6792690753936768,
+      "learning_rate": 0.00017000829863949316,
+      "loss": 2.9919,
+      "step": 39340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6472276449203491,
+      "learning_rate": 0.00016999442303992137,
+      "loss": 3.0074,
+      "step": 39341
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6135537028312683,
+      "learning_rate": 0.0001699805477827592,
+      "loss": 2.929,
+      "step": 39342
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6506457924842834,
+      "learning_rate": 0.00016996667286804318,
+      "loss": 2.8383,
+      "step": 39343
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6891106963157654,
+      "learning_rate": 0.00016995279829580998,
+      "loss": 2.7531,
+      "step": 39344
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6155516505241394,
+      "learning_rate": 0.000169938924066096,
+      "loss": 2.9584,
+      "step": 39345
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6519812941551208,
+      "learning_rate": 0.00016992505017893782,
+      "loss": 3.0768,
+      "step": 39346
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6716668605804443,
+      "learning_rate": 0.0001699111766343722,
+      "loss": 2.9624,
+      "step": 39347
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6471678614616394,
+      "learning_rate": 0.00016989730343243536,
+      "loss": 3.1556,
+      "step": 39348
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6281684637069702,
+      "learning_rate": 0.0001698834305731641,
+      "loss": 3.1132,
+      "step": 39349
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.621414840221405,
+      "learning_rate": 0.00016986955805659487,
+      "loss": 2.9695,
+      "step": 39350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6490262746810913,
+      "learning_rate": 0.000169855685882764,
+      "loss": 2.8509,
+      "step": 39351
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.687820553779602,
+      "learning_rate": 0.00016984181405170837,
+      "loss": 3.0534,
+      "step": 39352
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6453794240951538,
+      "learning_rate": 0.00016982794256346424,
+      "loss": 3.0,
+      "step": 39353
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7020411491394043,
+      "learning_rate": 0.00016981407141806836,
+      "loss": 3.0087,
+      "step": 39354
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6691149473190308,
+      "learning_rate": 0.00016980020061555702,
+      "loss": 2.934,
+      "step": 39355
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6700564622879028,
+      "learning_rate": 0.00016978633015596708,
+      "loss": 2.7317,
+      "step": 39356
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6754282712936401,
+      "learning_rate": 0.0001697724600393348,
+      "loss": 3.1237,
+      "step": 39357
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6140692830085754,
+      "learning_rate": 0.00016975859026569672,
+      "loss": 3.0405,
+      "step": 39358
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6934306025505066,
+      "learning_rate": 0.00016974472083508958,
+      "loss": 3.0548,
+      "step": 39359
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7280018925666809,
+      "learning_rate": 0.00016973085174754962,
+      "loss": 2.8646,
+      "step": 39360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.664756178855896,
+      "learning_rate": 0.00016971698300311357,
+      "loss": 3.0039,
+      "step": 39361
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6578002572059631,
+      "learning_rate": 0.000169703114601818,
+      "loss": 2.6559,
+      "step": 39362
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.661143958568573,
+      "learning_rate": 0.0001696892465436994,
+      "loss": 2.5943,
+      "step": 39363
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6856162548065186,
+      "learning_rate": 0.00016967537882879417,
+      "loss": 3.0467,
+      "step": 39364
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.661816418170929,
+      "learning_rate": 0.00016966151145713886,
+      "loss": 3.0191,
+      "step": 39365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.665028989315033,
+      "learning_rate": 0.00016964764442877001,
+      "loss": 2.9423,
+      "step": 39366
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.702292799949646,
+      "learning_rate": 0.00016963377774372433,
+      "loss": 3.2149,
+      "step": 39367
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6974702477455139,
+      "learning_rate": 0.00016961991140203802,
+      "loss": 2.791,
+      "step": 39368
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6458693146705627,
+      "learning_rate": 0.0001696060454037479,
+      "loss": 3.1692,
+      "step": 39369
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6239611506462097,
+      "learning_rate": 0.0001695921797488903,
+      "loss": 3.1647,
+      "step": 39370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6828367710113525,
+      "learning_rate": 0.00016957831443750184,
+      "loss": 2.7982,
+      "step": 39371
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6554055213928223,
+      "learning_rate": 0.00016956444946961905,
+      "loss": 2.841,
+      "step": 39372
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6682385206222534,
+      "learning_rate": 0.00016955058484527825,
+      "loss": 3.153,
+      "step": 39373
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6697953939437866,
+      "learning_rate": 0.00016953672056451626,
+      "loss": 3.0903,
+      "step": 39374
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.693271279335022,
+      "learning_rate": 0.0001695228566273693,
+      "loss": 2.9609,
+      "step": 39375
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6608006358146667,
+      "learning_rate": 0.000169508993033874,
+      "loss": 3.1086,
+      "step": 39376
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6983053088188171,
+      "learning_rate": 0.00016949512978406704,
+      "loss": 2.9174,
+      "step": 39377
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6728823184967041,
+      "learning_rate": 0.00016948126687798483,
+      "loss": 3.0248,
+      "step": 39378
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6789221167564392,
+      "learning_rate": 0.00016946740431566376,
+      "loss": 3.0388,
+      "step": 39379
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.619369387626648,
+      "learning_rate": 0.00016945354209714036,
+      "loss": 3.1008,
+      "step": 39380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6840798258781433,
+      "learning_rate": 0.0001694396802224512,
+      "loss": 3.2137,
+      "step": 39381
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6583610773086548,
+      "learning_rate": 0.00016942581869163292,
+      "loss": 3.0252,
+      "step": 39382
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6744231581687927,
+      "learning_rate": 0.00016941195750472176,
+      "loss": 2.9969,
+      "step": 39383
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.00016939809666175452,
+      "loss": 3.0379,
+      "step": 39384
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6821282505989075,
+      "learning_rate": 0.00016938423616276755,
+      "loss": 2.6843,
+      "step": 39385
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6183244585990906,
+      "learning_rate": 0.00016937037600779729,
+      "loss": 2.9519,
+      "step": 39386
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6722292304039001,
+      "learning_rate": 0.00016935651619688036,
+      "loss": 2.8854,
+      "step": 39387
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6663751006126404,
+      "learning_rate": 0.00016934265673005313,
+      "loss": 2.9347,
+      "step": 39388
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.694520890712738,
+      "learning_rate": 0.00016932879760735235,
+      "loss": 2.9421,
+      "step": 39389
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6110134124755859,
+      "learning_rate": 0.0001693149388288142,
+      "loss": 3.0589,
+      "step": 39390
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6420994400978088,
+      "learning_rate": 0.00016930108039447538,
+      "loss": 3.1698,
+      "step": 39391
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6498850584030151,
+      "learning_rate": 0.00016928722230437248,
+      "loss": 3.1342,
+      "step": 39392
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7132246494293213,
+      "learning_rate": 0.00016927336455854186,
+      "loss": 2.945,
+      "step": 39393
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6706508994102478,
+      "learning_rate": 0.00016925950715702005,
+      "loss": 2.9076,
+      "step": 39394
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7032055854797363,
+      "learning_rate": 0.00016924565009984342,
+      "loss": 2.9207,
+      "step": 39395
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6941987872123718,
+      "learning_rate": 0.0001692317933870486,
+      "loss": 3.0327,
+      "step": 39396
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6361653208732605,
+      "learning_rate": 0.00016921793701867217,
+      "loss": 3.0947,
+      "step": 39397
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6204395890235901,
+      "learning_rate": 0.0001692040809947504,
+      "loss": 2.8356,
+      "step": 39398
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6345608830451965,
+      "learning_rate": 0.00016919022531532,
+      "loss": 3.0054,
+      "step": 39399
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9609873294830322,
+      "learning_rate": 0.00016917636998041742,
+      "loss": 2.9836,
+      "step": 39400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6671732664108276,
+      "learning_rate": 0.000169162514990079,
+      "loss": 2.8382,
+      "step": 39401
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6544909477233887,
+      "learning_rate": 0.0001691486603443414,
+      "loss": 2.9315,
+      "step": 39402
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6705605387687683,
+      "learning_rate": 0.00016913480604324095,
+      "loss": 2.9288,
+      "step": 39403
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6991368532180786,
+      "learning_rate": 0.00016912095208681437,
+      "loss": 2.7712,
+      "step": 39404
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7298339009284973,
+      "learning_rate": 0.00016910709847509786,
+      "loss": 3.005,
+      "step": 39405
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6603250503540039,
+      "learning_rate": 0.00016909324520812823,
+      "loss": 2.8994,
+      "step": 39406
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6446770429611206,
+      "learning_rate": 0.00016907939228594175,
+      "loss": 2.9513,
+      "step": 39407
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6649962067604065,
+      "learning_rate": 0.00016906553970857488,
+      "loss": 2.9068,
+      "step": 39408
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6251447796821594,
+      "learning_rate": 0.00016905168747606428,
+      "loss": 2.8336,
+      "step": 39409
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6825475692749023,
+      "learning_rate": 0.00016903783558844617,
+      "loss": 2.9488,
+      "step": 39410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6691312789916992,
+      "learning_rate": 0.00016902398404575726,
+      "loss": 3.0109,
+      "step": 39411
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6389507055282593,
+      "learning_rate": 0.00016901013284803407,
+      "loss": 3.1017,
+      "step": 39412
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6700393557548523,
+      "learning_rate": 0.00016899628199531282,
+      "loss": 2.999,
+      "step": 39413
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6409572958946228,
+      "learning_rate": 0.00016898243148763032,
+      "loss": 2.8796,
+      "step": 39414
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6346601843833923,
+      "learning_rate": 0.00016896858132502287,
+      "loss": 3.0772,
+      "step": 39415
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6301149129867554,
+      "learning_rate": 0.00016895473150752683,
+      "loss": 3.0398,
+      "step": 39416
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6977820992469788,
+      "learning_rate": 0.00016894088203517894,
+      "loss": 2.8781,
+      "step": 39417
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9036363363265991,
+      "learning_rate": 0.00016892703290801543,
+      "loss": 2.9435,
+      "step": 39418
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7055711150169373,
+      "learning_rate": 0.00016891318412607298,
+      "loss": 2.9065,
+      "step": 39419
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6767593026161194,
+      "learning_rate": 0.00016889933568938788,
+      "loss": 2.8684,
+      "step": 39420
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6799156665802002,
+      "learning_rate": 0.0001688854875979968,
+      "loss": 3.161,
+      "step": 39421
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.615311324596405,
+      "learning_rate": 0.00016887163985193615,
+      "loss": 3.001,
+      "step": 39422
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6541116833686829,
+      "learning_rate": 0.00016885779245124223,
+      "loss": 3.0344,
+      "step": 39423
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6408389806747437,
+      "learning_rate": 0.00016884394539595172,
+      "loss": 2.9967,
+      "step": 39424
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7208709716796875,
+      "learning_rate": 0.00016883009868610093,
+      "loss": 2.9828,
+      "step": 39425
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6646652221679688,
+      "learning_rate": 0.00016881625232172652,
+      "loss": 2.9533,
+      "step": 39426
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.637306809425354,
+      "learning_rate": 0.00016880240630286477,
+      "loss": 3.0291,
+      "step": 39427
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6594432592391968,
+      "learning_rate": 0.0001687885606295522,
+      "loss": 3.0415,
+      "step": 39428
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6402209401130676,
+      "learning_rate": 0.0001687747153018256,
+      "loss": 2.9242,
+      "step": 39429
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6318802237510681,
+      "learning_rate": 0.00016876087031972084,
+      "loss": 2.9576,
+      "step": 39430
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6425923109054565,
+      "learning_rate": 0.00016874702568327489,
+      "loss": 2.804,
+      "step": 39431
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6498662829399109,
+      "learning_rate": 0.00016873318139252382,
+      "loss": 2.924,
+      "step": 39432
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6536316871643066,
+      "learning_rate": 0.0001687193374475043,
+      "loss": 2.885,
+      "step": 39433
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6774493455886841,
+      "learning_rate": 0.0001687054938482529,
+      "loss": 2.8245,
+      "step": 39434
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6602688431739807,
+      "learning_rate": 0.00016869165059480584,
+      "loss": 3.3949,
+      "step": 39435
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6157000660896301,
+      "learning_rate": 0.00016867780768719983,
+      "loss": 2.9687,
+      "step": 39436
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6612973213195801,
+      "learning_rate": 0.00016866396512547124,
+      "loss": 2.9654,
+      "step": 39437
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6561480760574341,
+      "learning_rate": 0.0001686501229096563,
+      "loss": 2.9176,
+      "step": 39438
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.680400550365448,
+      "learning_rate": 0.0001686362810397918,
+      "loss": 3.068,
+      "step": 39439
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6502717733383179,
+      "learning_rate": 0.00016862243951591395,
+      "loss": 3.0433,
+      "step": 39440
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.641892671585083,
+      "learning_rate": 0.00016860859833805943,
+      "loss": 2.891,
+      "step": 39441
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6740272641181946,
+      "learning_rate": 0.00016859475750626444,
+      "loss": 3.0464,
+      "step": 39442
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6247941851615906,
+      "learning_rate": 0.0001685809170205657,
+      "loss": 2.8856,
+      "step": 39443
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6555472016334534,
+      "learning_rate": 0.00016856707688099956,
+      "loss": 3.1368,
+      "step": 39444
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6383205056190491,
+      "learning_rate": 0.0001685532370876023,
+      "loss": 2.6563,
+      "step": 39445
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6372672319412231,
+      "learning_rate": 0.0001685393976404106,
+      "loss": 3.1764,
+      "step": 39446
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6690751910209656,
+      "learning_rate": 0.00016852555853946074,
+      "loss": 2.9395,
+      "step": 39447
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6416040658950806,
+      "learning_rate": 0.00016851171978478927,
+      "loss": 2.9809,
+      "step": 39448
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6781452894210815,
+      "learning_rate": 0.0001684978813764327,
+      "loss": 3.1393,
+      "step": 39449
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6450480818748474,
+      "learning_rate": 0.00016848404331442732,
+      "loss": 3.0979,
+      "step": 39450
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6464423537254333,
+      "learning_rate": 0.0001684702055988098,
+      "loss": 3.0349,
+      "step": 39451
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6569921374320984,
+      "learning_rate": 0.00016845636822961638,
+      "loss": 2.9834,
+      "step": 39452
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6588374972343445,
+      "learning_rate": 0.0001684425312068835,
+      "loss": 3.0806,
+      "step": 39453
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6816110014915466,
+      "learning_rate": 0.00016842869453064776,
+      "loss": 2.9424,
+      "step": 39454
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6436490416526794,
+      "learning_rate": 0.00016841485820094543,
+      "loss": 2.9879,
+      "step": 39455
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6433058381080627,
+      "learning_rate": 0.00016840102221781316,
+      "loss": 2.932,
+      "step": 39456
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6444736123085022,
+      "learning_rate": 0.00016838718658128714,
+      "loss": 2.9911,
+      "step": 39457
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6469634175300598,
+      "learning_rate": 0.00016837335129140404,
+      "loss": 2.8933,
+      "step": 39458
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6529709100723267,
+      "learning_rate": 0.00016835951634820023,
+      "loss": 3.0116,
+      "step": 39459
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6344585418701172,
+      "learning_rate": 0.00016834568175171197,
+      "loss": 2.9716,
+      "step": 39460
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6618320941925049,
+      "learning_rate": 0.00016833184750197596,
+      "loss": 3.0455,
+      "step": 39461
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6651561260223389,
+      "learning_rate": 0.00016831801359902846,
+      "loss": 3.1178,
+      "step": 39462
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6415659785270691,
+      "learning_rate": 0.00016830418004290594,
+      "loss": 2.7613,
+      "step": 39463
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7086910605430603,
+      "learning_rate": 0.00016829034683364498,
+      "loss": 2.8569,
+      "step": 39464
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.619654655456543,
+      "learning_rate": 0.00016827651397128188,
+      "loss": 3.1485,
+      "step": 39465
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6941916942596436,
+      "learning_rate": 0.00016826268145585314,
+      "loss": 2.7932,
+      "step": 39466
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7001242637634277,
+      "learning_rate": 0.00016824884928739496,
+      "loss": 2.9649,
+      "step": 39467
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6473196148872375,
+      "learning_rate": 0.00016823501746594403,
+      "loss": 3.0718,
+      "step": 39468
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6944223642349243,
+      "learning_rate": 0.00016822118599153676,
+      "loss": 3.0691,
+      "step": 39469
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6287515759468079,
+      "learning_rate": 0.00016820735486420944,
+      "loss": 2.9734,
+      "step": 39470
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.699953556060791,
+      "learning_rate": 0.00016819352408399869,
+      "loss": 2.9848,
+      "step": 39471
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6306780576705933,
+      "learning_rate": 0.00016817969365094074,
+      "loss": 3.1997,
+      "step": 39472
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6510525345802307,
+      "learning_rate": 0.0001681658635650722,
+      "loss": 2.9396,
+      "step": 39473
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6423694491386414,
+      "learning_rate": 0.0001681520338264294,
+      "loss": 3.058,
+      "step": 39474
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.662390410900116,
+      "learning_rate": 0.00016813820443504866,
+      "loss": 3.0601,
+      "step": 39475
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.653954267501831,
+      "learning_rate": 0.0001681243753909666,
+      "loss": 3.0233,
+      "step": 39476
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6505160331726074,
+      "learning_rate": 0.00016811054669421948,
+      "loss": 3.0404,
+      "step": 39477
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6412315368652344,
+      "learning_rate": 0.00016809671834484377,
+      "loss": 2.9671,
+      "step": 39478
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.646365225315094,
+      "learning_rate": 0.00016808289034287604,
+      "loss": 2.9477,
+      "step": 39479
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6499340534210205,
+      "learning_rate": 0.00016806906268835263,
+      "loss": 3.2125,
+      "step": 39480
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6421015858650208,
+      "learning_rate": 0.00016805523538130986,
+      "loss": 3.1327,
+      "step": 39481
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6436095833778381,
+      "learning_rate": 0.00016804140842178412,
+      "loss": 2.9494,
+      "step": 39482
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.658585786819458,
+      "learning_rate": 0.0001680275818098119,
+      "loss": 3.0283,
+      "step": 39483
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6562331318855286,
+      "learning_rate": 0.00016801375554542977,
+      "loss": 3.0036,
+      "step": 39484
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7266846299171448,
+      "learning_rate": 0.00016799992962867385,
+      "loss": 2.9681,
+      "step": 39485
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6242222785949707,
+      "learning_rate": 0.00016798610405958087,
+      "loss": 2.9437,
+      "step": 39486
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6533987522125244,
+      "learning_rate": 0.00016797227883818706,
+      "loss": 2.7359,
+      "step": 39487
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6742711067199707,
+      "learning_rate": 0.00016795845396452876,
+      "loss": 2.8229,
+      "step": 39488
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6334719657897949,
+      "learning_rate": 0.00016794462943864257,
+      "loss": 2.9316,
+      "step": 39489
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.651542067527771,
+      "learning_rate": 0.00016793080526056472,
+      "loss": 3.0136,
+      "step": 39490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6780474185943604,
+      "learning_rate": 0.00016791698143033183,
+      "loss": 2.9013,
+      "step": 39491
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.674872100353241,
+      "learning_rate": 0.00016790315794798006,
+      "loss": 2.8134,
+      "step": 39492
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.00016788933481354595,
+      "loss": 2.7779,
+      "step": 39493
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6219632625579834,
+      "learning_rate": 0.00016787551202706604,
+      "loss": 2.8627,
+      "step": 39494
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.640094518661499,
+      "learning_rate": 0.0001678616895885766,
+      "loss": 2.8384,
+      "step": 39495
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7231557965278625,
+      "learning_rate": 0.000167847867498114,
+      "loss": 2.9999,
+      "step": 39496
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6574937701225281,
+      "learning_rate": 0.0001678340457557146,
+      "loss": 2.9708,
+      "step": 39497
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6397494077682495,
+      "learning_rate": 0.000167820224361415,
+      "loss": 2.8948,
+      "step": 39498
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6591821908950806,
+      "learning_rate": 0.0001678064033152514,
+      "loss": 3.1144,
+      "step": 39499
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.684015691280365,
+      "learning_rate": 0.00016779258261726029,
+      "loss": 3.0855,
+      "step": 39500
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6491175293922424,
+      "learning_rate": 0.00016777876226747816,
+      "loss": 2.8436,
+      "step": 39501
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6679362058639526,
+      "learning_rate": 0.0001677649422659413,
+      "loss": 2.9078,
+      "step": 39502
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6936571598052979,
+      "learning_rate": 0.00016775112261268617,
+      "loss": 2.9691,
+      "step": 39503
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.673439621925354,
+      "learning_rate": 0.000167737303307749,
+      "loss": 3.0569,
+      "step": 39504
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6793358325958252,
+      "learning_rate": 0.00016772348435116637,
+      "loss": 3.1337,
+      "step": 39505
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6524600982666016,
+      "learning_rate": 0.00016770966574297465,
+      "loss": 2.9261,
+      "step": 39506
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6856192946434021,
+      "learning_rate": 0.00016769584748321014,
+      "loss": 3.0568,
+      "step": 39507
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6691726446151733,
+      "learning_rate": 0.00016768202957190943,
+      "loss": 2.7727,
+      "step": 39508
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7057440876960754,
+      "learning_rate": 0.00016766821200910883,
+      "loss": 2.9153,
+      "step": 39509
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6541669368743896,
+      "learning_rate": 0.00016765439479484448,
+      "loss": 3.16,
+      "step": 39510
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7234159708023071,
+      "learning_rate": 0.00016764057792915317,
+      "loss": 3.1711,
+      "step": 39511
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6421957015991211,
+      "learning_rate": 0.00016762676141207096,
+      "loss": 2.8255,
+      "step": 39512
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6890547871589661,
+      "learning_rate": 0.0001676129452436345,
+      "loss": 2.9754,
+      "step": 39513
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6445096731185913,
+      "learning_rate": 0.00016759912942388,
+      "loss": 2.8582,
+      "step": 39514
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6753875017166138,
+      "learning_rate": 0.00016758531395284383,
+      "loss": 3.0514,
+      "step": 39515
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6762891411781311,
+      "learning_rate": 0.00016757149883056264,
+      "loss": 2.9444,
+      "step": 39516
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6587990522384644,
+      "learning_rate": 0.0001675576840570726,
+      "loss": 2.8884,
+      "step": 39517
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6049191951751709,
+      "learning_rate": 0.00016754386963241014,
+      "loss": 2.928,
+      "step": 39518
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6751249432563782,
+      "learning_rate": 0.00016753005555661152,
+      "loss": 3.0188,
+      "step": 39519
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6588065028190613,
+      "learning_rate": 0.00016751624182971325,
+      "loss": 3.018,
+      "step": 39520
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6722527742385864,
+      "learning_rate": 0.00016750242845175182,
+      "loss": 2.9835,
+      "step": 39521
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7164931893348694,
+      "learning_rate": 0.00016748861542276332,
+      "loss": 2.8443,
+      "step": 39522
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6754207015037537,
+      "learning_rate": 0.0001674748027427845,
+      "loss": 3.1071,
+      "step": 39523
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6811302304267883,
+      "learning_rate": 0.00016746099041185152,
+      "loss": 2.8322,
+      "step": 39524
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6523797512054443,
+      "learning_rate": 0.00016744717843000066,
+      "loss": 3.0216,
+      "step": 39525
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6112352013587952,
+      "learning_rate": 0.0001674333667972685,
+      "loss": 2.9525,
+      "step": 39526
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6537315249443054,
+      "learning_rate": 0.0001674195555136913,
+      "loss": 3.0354,
+      "step": 39527
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6480567455291748,
+      "learning_rate": 0.00016740574457930552,
+      "loss": 3.0474,
+      "step": 39528
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6240382194519043,
+      "learning_rate": 0.00016739193399414743,
+      "loss": 2.7532,
+      "step": 39529
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6805428266525269,
+      "learning_rate": 0.00016737812375825357,
+      "loss": 2.878,
+      "step": 39530
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6385943293571472,
+      "learning_rate": 0.00016736431387166017,
+      "loss": 3.1757,
+      "step": 39531
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6469088196754456,
+      "learning_rate": 0.00016735050433440353,
+      "loss": 2.8245,
+      "step": 39532
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6487658619880676,
+      "learning_rate": 0.0001673366951465203,
+      "loss": 2.864,
+      "step": 39533
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6648675799369812,
+      "learning_rate": 0.0001673228863080465,
+      "loss": 3.1449,
+      "step": 39534
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6737173199653625,
+      "learning_rate": 0.00016730907781901873,
+      "loss": 2.8128,
+      "step": 39535
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6540577411651611,
+      "learning_rate": 0.00016729526967947344,
+      "loss": 3.0712,
+      "step": 39536
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6505504250526428,
+      "learning_rate": 0.00016728146188944672,
+      "loss": 3.0605,
+      "step": 39537
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6246081590652466,
+      "learning_rate": 0.0001672676544489752,
+      "loss": 2.902,
+      "step": 39538
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6178304553031921,
+      "learning_rate": 0.00016725384735809517,
+      "loss": 2.738,
+      "step": 39539
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.711046040058136,
+      "learning_rate": 0.00016724004061684282,
+      "loss": 3.0829,
+      "step": 39540
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6916102766990662,
+      "learning_rate": 0.00016722623422525478,
+      "loss": 3.0951,
+      "step": 39541
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6540066599845886,
+      "learning_rate": 0.00016721242818336712,
+      "loss": 3.0883,
+      "step": 39542
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6461518406867981,
+      "learning_rate": 0.00016719862249121655,
+      "loss": 3.0277,
+      "step": 39543
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6972729563713074,
+      "learning_rate": 0.00016718481714883912,
+      "loss": 3.252,
+      "step": 39544
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6396310329437256,
+      "learning_rate": 0.00016717101215627143,
+      "loss": 2.5932,
+      "step": 39545
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7706670165061951,
+      "learning_rate": 0.0001671572075135497,
+      "loss": 3.0133,
+      "step": 39546
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6633245944976807,
+      "learning_rate": 0.00016714340322071023,
+      "loss": 3.0178,
+      "step": 39547
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7238604426383972,
+      "learning_rate": 0.00016712959927778957,
+      "loss": 2.921,
+      "step": 39548
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6844163537025452,
+      "learning_rate": 0.0001671157956848239,
+      "loss": 3.0155,
+      "step": 39549
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6485677361488342,
+      "learning_rate": 0.0001671019924418496,
+      "loss": 3.1059,
+      "step": 39550
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6759480237960815,
+      "learning_rate": 0.0001670881895489032,
+      "loss": 3.2116,
+      "step": 39551
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.641191303730011,
+      "learning_rate": 0.00016707438700602084,
+      "loss": 2.8688,
+      "step": 39552
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6632412672042847,
+      "learning_rate": 0.00016706058481323908,
+      "loss": 3.114,
+      "step": 39553
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6805461049079895,
+      "learning_rate": 0.0001670467829705941,
+      "loss": 2.9444,
+      "step": 39554
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6704729199409485,
+      "learning_rate": 0.00016703298147812223,
+      "loss": 2.851,
+      "step": 39555
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.625964879989624,
+      "learning_rate": 0.00016701918033586,
+      "loss": 3.2429,
+      "step": 39556
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6714993119239807,
+      "learning_rate": 0.00016700537954384356,
+      "loss": 2.9443,
+      "step": 39557
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6405186057090759,
+      "learning_rate": 0.0001669915791021095,
+      "loss": 3.0493,
+      "step": 39558
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6585462093353271,
+      "learning_rate": 0.00016697777901069385,
+      "loss": 3.0023,
+      "step": 39559
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7218492031097412,
+      "learning_rate": 0.00016696397926963325,
+      "loss": 3.137,
+      "step": 39560
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6385161280632019,
+      "learning_rate": 0.00016695017987896398,
+      "loss": 3.008,
+      "step": 39561
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6418231725692749,
+      "learning_rate": 0.00016693638083872217,
+      "loss": 3.023,
+      "step": 39562
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6520888805389404,
+      "learning_rate": 0.00016692258214894446,
+      "loss": 3.0538,
+      "step": 39563
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6636407375335693,
+      "learning_rate": 0.0001669087838096669,
+      "loss": 2.8663,
+      "step": 39564
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6697338223457336,
+      "learning_rate": 0.000166894985820926,
+      "loss": 2.9433,
+      "step": 39565
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6582183241844177,
+      "learning_rate": 0.00016688118818275824,
+      "loss": 3.0544,
+      "step": 39566
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6985874176025391,
+      "learning_rate": 0.0001668673908951998,
+      "loss": 3.1551,
+      "step": 39567
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6799558997154236,
+      "learning_rate": 0.00016685359395828702,
+      "loss": 2.9644,
+      "step": 39568
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6577656865119934,
+      "learning_rate": 0.00016683979737205614,
+      "loss": 2.9913,
+      "step": 39569
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6054019331932068,
+      "learning_rate": 0.0001668260011365436,
+      "loss": 2.8949,
+      "step": 39570
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7262248992919922,
+      "learning_rate": 0.00016681220525178585,
+      "loss": 3.1803,
+      "step": 39571
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6668470501899719,
+      "learning_rate": 0.00016679840971781904,
+      "loss": 3.1008,
+      "step": 39572
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6440438032150269,
+      "learning_rate": 0.00016678461453467963,
+      "loss": 2.9656,
+      "step": 39573
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6357288956642151,
+      "learning_rate": 0.00016677081970240386,
+      "loss": 2.9421,
+      "step": 39574
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6600980162620544,
+      "learning_rate": 0.00016675702522102822,
+      "loss": 2.661,
+      "step": 39575
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6919565796852112,
+      "learning_rate": 0.0001667432310905889,
+      "loss": 2.8373,
+      "step": 39576
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6861438751220703,
+      "learning_rate": 0.0001667294373111222,
+      "loss": 2.9764,
+      "step": 39577
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7029724717140198,
+      "learning_rate": 0.00016671564388266456,
+      "loss": 2.9436,
+      "step": 39578
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6634342670440674,
+      "learning_rate": 0.0001667018508052522,
+      "loss": 2.8775,
+      "step": 39579
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6349776387214661,
+      "learning_rate": 0.0001666880580789216,
+      "loss": 2.9675,
+      "step": 39580
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6573511362075806,
+      "learning_rate": 0.00016667426570370888,
+      "loss": 2.9222,
+      "step": 39581
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6723518967628479,
+      "learning_rate": 0.00016666047367965058,
+      "loss": 2.9447,
+      "step": 39582
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6813273429870605,
+      "learning_rate": 0.00016664668200678297,
+      "loss": 3.0502,
+      "step": 39583
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6408233046531677,
+      "learning_rate": 0.0001666328906851422,
+      "loss": 2.913,
+      "step": 39584
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6882518529891968,
+      "learning_rate": 0.00016661909971476489,
+      "loss": 2.9673,
+      "step": 39585
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7611137628555298,
+      "learning_rate": 0.00016660530909568704,
+      "loss": 3.0111,
+      "step": 39586
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6440632343292236,
+      "learning_rate": 0.00016659151882794516,
+      "loss": 3.0178,
+      "step": 39587
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6395631432533264,
+      "learning_rate": 0.00016657772891157563,
+      "loss": 2.963,
+      "step": 39588
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6646182537078857,
+      "learning_rate": 0.0001665639393466147,
+      "loss": 2.8769,
+      "step": 39589
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6622651815414429,
+      "learning_rate": 0.00016655015013309867,
+      "loss": 3.1089,
+      "step": 39590
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6579053997993469,
+      "learning_rate": 0.00016653636127106374,
+      "loss": 2.8252,
+      "step": 39591
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.635835587978363,
+      "learning_rate": 0.00016652257276054638,
+      "loss": 3.1497,
+      "step": 39592
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.645119309425354,
+      "learning_rate": 0.00016650878460158295,
+      "loss": 3.0745,
+      "step": 39593
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.707358717918396,
+      "learning_rate": 0.0001664949967942096,
+      "loss": 3.0034,
+      "step": 39594
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6341613531112671,
+      "learning_rate": 0.00016648120933846289,
+      "loss": 2.9241,
+      "step": 39595
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6709751486778259,
+      "learning_rate": 0.00016646742223437879,
+      "loss": 3.0487,
+      "step": 39596
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6443012356758118,
+      "learning_rate": 0.00016645363548199396,
+      "loss": 2.947,
+      "step": 39597
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6468132734298706,
+      "learning_rate": 0.0001664398490813445,
+      "loss": 3.1464,
+      "step": 39598
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6579627394676208,
+      "learning_rate": 0.0001664260630324667,
+      "loss": 2.9348,
+      "step": 39599
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6067564487457275,
+      "learning_rate": 0.00016641227733539703,
+      "loss": 2.9836,
+      "step": 39600
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6814645528793335,
+      "learning_rate": 0.00016639849199017164,
+      "loss": 2.9497,
+      "step": 39601
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7532250285148621,
+      "learning_rate": 0.00016638470699682688,
+      "loss": 2.8528,
+      "step": 39602
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6621189713478088,
+      "learning_rate": 0.00016637092235539925,
+      "loss": 3.103,
+      "step": 39603
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6254897713661194,
+      "learning_rate": 0.00016635713806592483,
+      "loss": 2.6625,
+      "step": 39604
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6650421023368835,
+      "learning_rate": 0.00016634335412844,
+      "loss": 3.0827,
+      "step": 39605
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6702773571014404,
+      "learning_rate": 0.00016632957054298097,
+      "loss": 3.0373,
+      "step": 39606
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.629437267780304,
+      "learning_rate": 0.0001663157873095841,
+      "loss": 2.8015,
+      "step": 39607
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6859908103942871,
+      "learning_rate": 0.00016630200442828582,
+      "loss": 2.8849,
+      "step": 39608
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6508818864822388,
+      "learning_rate": 0.00016628822189912226,
+      "loss": 2.7454,
+      "step": 39609
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6694782972335815,
+      "learning_rate": 0.00016627443972212986,
+      "loss": 2.9539,
+      "step": 39610
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6411961317062378,
+      "learning_rate": 0.00016626065789734486,
+      "loss": 2.9037,
+      "step": 39611
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6573426127433777,
+      "learning_rate": 0.00016624687642480344,
+      "loss": 2.8916,
+      "step": 39612
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7193799614906311,
+      "learning_rate": 0.00016623309530454215,
+      "loss": 2.843,
+      "step": 39613
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6922988891601562,
+      "learning_rate": 0.00016621931453659697,
+      "loss": 3.0807,
+      "step": 39614
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6599312424659729,
+      "learning_rate": 0.00016620553412100453,
+      "loss": 2.8812,
+      "step": 39615
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6752668619155884,
+      "learning_rate": 0.00016619175405780081,
+      "loss": 2.9444,
+      "step": 39616
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6617905497550964,
+      "learning_rate": 0.00016617797434702233,
+      "loss": 3.1259,
+      "step": 39617
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6991008520126343,
+      "learning_rate": 0.00016616419498870533,
+      "loss": 3.0066,
+      "step": 39618
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6415068507194519,
+      "learning_rate": 0.00016615041598288613,
+      "loss": 3.1282,
+      "step": 39619
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6914513111114502,
+      "learning_rate": 0.000166136637329601,
+      "loss": 3.2745,
+      "step": 39620
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6138527989387512,
+      "learning_rate": 0.00016612285902888605,
+      "loss": 2.8569,
+      "step": 39621
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6552730798721313,
+      "learning_rate": 0.00016610908108077776,
+      "loss": 3.088,
+      "step": 39622
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6678380370140076,
+      "learning_rate": 0.00016609530348531246,
+      "loss": 3.0937,
+      "step": 39623
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7249351143836975,
+      "learning_rate": 0.00016608152624252628,
+      "loss": 3.1583,
+      "step": 39624
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6721065044403076,
+      "learning_rate": 0.00016606774935245568,
+      "loss": 3.0651,
+      "step": 39625
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6632776260375977,
+      "learning_rate": 0.00016605397281513684,
+      "loss": 2.9947,
+      "step": 39626
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6905990839004517,
+      "learning_rate": 0.00016604019663060595,
+      "loss": 2.9802,
+      "step": 39627
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7240660190582275,
+      "learning_rate": 0.00016602642079889955,
+      "loss": 2.9532,
+      "step": 39628
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.653408944606781,
+      "learning_rate": 0.00016601264532005363,
+      "loss": 3.0149,
+      "step": 39629
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.665780246257782,
+      "learning_rate": 0.00016599887019410473,
+      "loss": 2.9089,
+      "step": 39630
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6445152163505554,
+      "learning_rate": 0.0001659850954210889,
+      "loss": 2.9931,
+      "step": 39631
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6995826363563538,
+      "learning_rate": 0.00016597132100104264,
+      "loss": 3.13,
+      "step": 39632
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6662886738777161,
+      "learning_rate": 0.00016595754693400215,
+      "loss": 3.0491,
+      "step": 39633
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6708129048347473,
+      "learning_rate": 0.00016594377322000356,
+      "loss": 2.9691,
+      "step": 39634
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6616761088371277,
+      "learning_rate": 0.00016592999985908343,
+      "loss": 3.0653,
+      "step": 39635
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6500592827796936,
+      "learning_rate": 0.0001659162268512777,
+      "loss": 3.0368,
+      "step": 39636
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6823159456253052,
+      "learning_rate": 0.00016590245419662283,
+      "loss": 2.9157,
+      "step": 39637
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6613301634788513,
+      "learning_rate": 0.00016588868189515522,
+      "loss": 2.8774,
+      "step": 39638
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.647834062576294,
+      "learning_rate": 0.00016587490994691092,
+      "loss": 2.8103,
+      "step": 39639
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6654465198516846,
+      "learning_rate": 0.00016586113835192635,
+      "loss": 2.9277,
+      "step": 39640
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6587596535682678,
+      "learning_rate": 0.0001658473671102378,
+      "loss": 2.9219,
+      "step": 39641
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6656188368797302,
+      "learning_rate": 0.00016583359622188132,
+      "loss": 3.0531,
+      "step": 39642
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6441434025764465,
+      "learning_rate": 0.00016581982568689342,
+      "loss": 2.9493,
+      "step": 39643
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6490876078605652,
+      "learning_rate": 0.00016580605550531018,
+      "loss": 2.9725,
+      "step": 39644
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7419745326042175,
+      "learning_rate": 0.00016579228567716807,
+      "loss": 2.88,
+      "step": 39645
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6507675051689148,
+      "learning_rate": 0.00016577851620250313,
+      "loss": 2.8218,
+      "step": 39646
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6721502542495728,
+      "learning_rate": 0.00016576474708135188,
+      "loss": 3.1349,
+      "step": 39647
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6784391403198242,
+      "learning_rate": 0.00016575097831375045,
+      "loss": 2.9116,
+      "step": 39648
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6822816133499146,
+      "learning_rate": 0.000165737209899735,
+      "loss": 2.8589,
+      "step": 39649
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6557987332344055,
+      "learning_rate": 0.000165723441839342,
+      "loss": 2.8979,
+      "step": 39650
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6441841125488281,
+      "learning_rate": 0.00016570967413260748,
+      "loss": 2.9015,
+      "step": 39651
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7314832210540771,
+      "learning_rate": 0.00016569590677956785,
+      "loss": 3.1157,
+      "step": 39652
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6173034310340881,
+      "learning_rate": 0.00016568213978025947,
+      "loss": 3.1966,
+      "step": 39653
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6204674243927002,
+      "learning_rate": 0.00016566837313471838,
+      "loss": 3.0183,
+      "step": 39654
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6612196564674377,
+      "learning_rate": 0.00016565460684298116,
+      "loss": 2.8712,
+      "step": 39655
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6848688721656799,
+      "learning_rate": 0.00016564084090508362,
+      "loss": 3.0089,
+      "step": 39656
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6412416696548462,
+      "learning_rate": 0.00016562707532106235,
+      "loss": 2.9572,
+      "step": 39657
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.667161226272583,
+      "learning_rate": 0.0001656133100909534,
+      "loss": 2.8977,
+      "step": 39658
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.639665961265564,
+      "learning_rate": 0.00016559954521479315,
+      "loss": 3.0027,
+      "step": 39659
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6585216522216797,
+      "learning_rate": 0.00016558578069261791,
+      "loss": 3.0025,
+      "step": 39660
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.637984573841095,
+      "learning_rate": 0.00016557201652446374,
+      "loss": 2.8542,
+      "step": 39661
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6432386040687561,
+      "learning_rate": 0.00016555825271036715,
+      "loss": 2.9786,
+      "step": 39662
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6286998391151428,
+      "learning_rate": 0.00016554448925036426,
+      "loss": 3.0735,
+      "step": 39663
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6470546722412109,
+      "learning_rate": 0.00016553072614449118,
+      "loss": 2.8628,
+      "step": 39664
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7382664084434509,
+      "learning_rate": 0.0001655169633927844,
+      "loss": 3.0028,
+      "step": 39665
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6897470355033875,
+      "learning_rate": 0.0001655032009952799,
+      "loss": 2.8576,
+      "step": 39666
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6223475933074951,
+      "learning_rate": 0.00016548943895201428,
+      "loss": 2.8861,
+      "step": 39667
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.721520721912384,
+      "learning_rate": 0.0001654756772630234,
+      "loss": 3.0925,
+      "step": 39668
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6690992116928101,
+      "learning_rate": 0.00016546191592834385,
+      "loss": 2.8354,
+      "step": 39669
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6852384209632874,
+      "learning_rate": 0.00016544815494801173,
+      "loss": 3.0092,
+      "step": 39670
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6654887795448303,
+      "learning_rate": 0.00016543439432206312,
+      "loss": 3.1952,
+      "step": 39671
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6750345826148987,
+      "learning_rate": 0.00016542063405053457,
+      "loss": 3.0411,
+      "step": 39672
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.695560872554779,
+      "learning_rate": 0.00016540687413346203,
+      "loss": 3.0145,
+      "step": 39673
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7053586840629578,
+      "learning_rate": 0.0001653931145708819,
+      "loss": 3.1015,
+      "step": 39674
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6435714960098267,
+      "learning_rate": 0.0001653793553628305,
+      "loss": 3.0385,
+      "step": 39675
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6890304684638977,
+      "learning_rate": 0.00016536559650934385,
+      "loss": 2.8383,
+      "step": 39676
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7017620205879211,
+      "learning_rate": 0.00016535183801045847,
+      "loss": 2.8856,
+      "step": 39677
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.638668417930603,
+      "learning_rate": 0.0001653380798662104,
+      "loss": 2.9055,
+      "step": 39678
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7142661809921265,
+      "learning_rate": 0.0001653243220766358,
+      "loss": 2.8482,
+      "step": 39679
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6465545892715454,
+      "learning_rate": 0.0001653105646417711,
+      "loss": 2.9118,
+      "step": 39680
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.640198826789856,
+      "learning_rate": 0.00016529680756165237,
+      "loss": 3.0155,
+      "step": 39681
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6553422212600708,
+      "learning_rate": 0.00016528305083631604,
+      "loss": 3.0283,
+      "step": 39682
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.842480480670929,
+      "learning_rate": 0.0001652692944657981,
+      "loss": 2.7033,
+      "step": 39683
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6969196200370789,
+      "learning_rate": 0.000165255538450135,
+      "loss": 3.1925,
+      "step": 39684
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6522579193115234,
+      "learning_rate": 0.00016524178278936295,
+      "loss": 2.8948,
+      "step": 39685
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.659083902835846,
+      "learning_rate": 0.00016522802748351797,
+      "loss": 2.8768,
+      "step": 39686
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6438480615615845,
+      "learning_rate": 0.00016521427253263652,
+      "loss": 2.9011,
+      "step": 39687
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6511509418487549,
+      "learning_rate": 0.0001652005179367546,
+      "loss": 2.9839,
+      "step": 39688
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6866702437400818,
+      "learning_rate": 0.0001651867636959086,
+      "loss": 2.9139,
+      "step": 39689
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6492428183555603,
+      "learning_rate": 0.00016517300981013484,
+      "loss": 3.0238,
+      "step": 39690
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.629295825958252,
+      "learning_rate": 0.00016515925627946942,
+      "loss": 3.1701,
+      "step": 39691
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7377083897590637,
+      "learning_rate": 0.0001651455031039486,
+      "loss": 2.8515,
+      "step": 39692
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6295761466026306,
+      "learning_rate": 0.00016513175028360844,
+      "loss": 2.862,
+      "step": 39693
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6240883469581604,
+      "learning_rate": 0.00016511799781848528,
+      "loss": 2.7688,
+      "step": 39694
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7190060615539551,
+      "learning_rate": 0.00016510424570861544,
+      "loss": 3.069,
+      "step": 39695
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6325370073318481,
+      "learning_rate": 0.00016509049395403496,
+      "loss": 2.8299,
+      "step": 39696
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6780902743339539,
+      "learning_rate": 0.00016507674255478028,
+      "loss": 2.9697,
+      "step": 39697
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6456710696220398,
+      "learning_rate": 0.0001650629915108874,
+      "loss": 2.8723,
+      "step": 39698
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6725703477859497,
+      "learning_rate": 0.00016504924082239272,
+      "loss": 3.1925,
+      "step": 39699
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6326077580451965,
+      "learning_rate": 0.00016503549048933237,
+      "loss": 2.9816,
+      "step": 39700
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6344588994979858,
+      "learning_rate": 0.00016502174051174245,
+      "loss": 3.0964,
+      "step": 39701
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6674083471298218,
+      "learning_rate": 0.00016500799088965942,
+      "loss": 2.851,
+      "step": 39702
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6438904404640198,
+      "learning_rate": 0.0001649942416231192,
+      "loss": 3.1094,
+      "step": 39703
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6310093402862549,
+      "learning_rate": 0.00016498049271215822,
+      "loss": 3.0114,
+      "step": 39704
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6457886099815369,
+      "learning_rate": 0.0001649667441568127,
+      "loss": 3.0389,
+      "step": 39705
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6540348529815674,
+      "learning_rate": 0.00016495299595711886,
+      "loss": 3.1004,
+      "step": 39706
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6192497611045837,
+      "learning_rate": 0.00016493924811311278,
+      "loss": 2.7377,
+      "step": 39707
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6560617685317993,
+      "learning_rate": 0.00016492550062483064,
+      "loss": 2.9241,
+      "step": 39708
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6480932831764221,
+      "learning_rate": 0.0001649117534923087,
+      "loss": 3.131,
+      "step": 39709
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6678999662399292,
+      "learning_rate": 0.00016489800671558335,
+      "loss": 3.0523,
+      "step": 39710
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7164120674133301,
+      "learning_rate": 0.0001648842602946905,
+      "loss": 2.8345,
+      "step": 39711
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6553846001625061,
+      "learning_rate": 0.00016487051422966667,
+      "loss": 2.9484,
+      "step": 39712
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6383706331253052,
+      "learning_rate": 0.00016485676852054785,
+      "loss": 2.9157,
+      "step": 39713
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6358631253242493,
+      "learning_rate": 0.00016484302316737022,
+      "loss": 3.0168,
+      "step": 39714
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7204211354255676,
+      "learning_rate": 0.00016482927817017018,
+      "loss": 3.0729,
+      "step": 39715
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7235838770866394,
+      "learning_rate": 0.00016481553352898365,
+      "loss": 3.097,
+      "step": 39716
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6387816667556763,
+      "learning_rate": 0.0001648017892438471,
+      "loss": 3.0294,
+      "step": 39717
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.71495121717453,
+      "learning_rate": 0.00016478804531479653,
+      "loss": 3.0884,
+      "step": 39718
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6762590408325195,
+      "learning_rate": 0.00016477430174186822,
+      "loss": 3.1445,
+      "step": 39719
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6747094392776489,
+      "learning_rate": 0.0001647605585250985,
+      "loss": 2.7726,
+      "step": 39720
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6876216530799866,
+      "learning_rate": 0.00016474681566452343,
+      "loss": 2.9254,
+      "step": 39721
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.662525475025177,
+      "learning_rate": 0.00016473307316017922,
+      "loss": 3.0107,
+      "step": 39722
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6243906617164612,
+      "learning_rate": 0.00016471933101210194,
+      "loss": 2.8323,
+      "step": 39723
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6680139303207397,
+      "learning_rate": 0.00016470558922032793,
+      "loss": 2.8206,
+      "step": 39724
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6394197940826416,
+      "learning_rate": 0.00016469184778489346,
+      "loss": 3.0246,
+      "step": 39725
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7584770917892456,
+      "learning_rate": 0.00016467810670583446,
+      "loss": 2.7716,
+      "step": 39726
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7045025825500488,
+      "learning_rate": 0.0001646643659831875,
+      "loss": 2.8362,
+      "step": 39727
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.676650881767273,
+      "learning_rate": 0.00016465062561698846,
+      "loss": 3.054,
+      "step": 39728
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.686996340751648,
+      "learning_rate": 0.00016463688560727354,
+      "loss": 2.9678,
+      "step": 39729
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6655346155166626,
+      "learning_rate": 0.00016462314595407914,
+      "loss": 3.0299,
+      "step": 39730
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6884318590164185,
+      "learning_rate": 0.0001646094066574412,
+      "loss": 2.9813,
+      "step": 39731
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6549413204193115,
+      "learning_rate": 0.00016459566771739612,
+      "loss": 2.9558,
+      "step": 39732
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6976496577262878,
+      "learning_rate": 0.0001645819291339799,
+      "loss": 2.9014,
+      "step": 39733
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6469190716743469,
+      "learning_rate": 0.00016456819090722894,
+      "loss": 2.9572,
+      "step": 39734
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6241649985313416,
+      "learning_rate": 0.0001645544530371793,
+      "loss": 3.1763,
+      "step": 39735
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6479203701019287,
+      "learning_rate": 0.000164540715523867,
+      "loss": 2.9041,
+      "step": 39736
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7182705402374268,
+      "learning_rate": 0.00016452697836732858,
+      "loss": 3.0069,
+      "step": 39737
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6855570077896118,
+      "learning_rate": 0.00016451324156759986,
+      "loss": 3.1234,
+      "step": 39738
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7441180944442749,
+      "learning_rate": 0.0001644995051247173,
+      "loss": 3.099,
+      "step": 39739
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7113296985626221,
+      "learning_rate": 0.00016448576903871687,
+      "loss": 3.0428,
+      "step": 39740
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6470918655395508,
+      "learning_rate": 0.00016447203330963482,
+      "loss": 2.9745,
+      "step": 39741
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6484912037849426,
+      "learning_rate": 0.0001644582979375075,
+      "loss": 3.0841,
+      "step": 39742
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7016591429710388,
+      "learning_rate": 0.00016444456292237093,
+      "loss": 3.1074,
+      "step": 39743
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7112533450126648,
+      "learning_rate": 0.00016443082826426126,
+      "loss": 2.7582,
+      "step": 39744
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6694428324699402,
+      "learning_rate": 0.00016441709396321463,
+      "loss": 2.7622,
+      "step": 39745
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6349228024482727,
+      "learning_rate": 0.00016440336001926722,
+      "loss": 2.8594,
+      "step": 39746
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6667150855064392,
+      "learning_rate": 0.00016438962643245544,
+      "loss": 3.1511,
+      "step": 39747
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6422046422958374,
+      "learning_rate": 0.00016437589320281515,
+      "loss": 3.1107,
+      "step": 39748
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.644203782081604,
+      "learning_rate": 0.00016436216033038274,
+      "loss": 2.8401,
+      "step": 39749
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6971120834350586,
+      "learning_rate": 0.00016434842781519433,
+      "loss": 2.9795,
+      "step": 39750
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6946253180503845,
+      "learning_rate": 0.00016433469565728593,
+      "loss": 2.9588,
+      "step": 39751
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6759042143821716,
+      "learning_rate": 0.00016432096385669397,
+      "loss": 3.0513,
+      "step": 39752
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8178485035896301,
+      "learning_rate": 0.00016430723241345432,
+      "loss": 2.9891,
+      "step": 39753
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6548961997032166,
+      "learning_rate": 0.00016429350132760346,
+      "loss": 3.034,
+      "step": 39754
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6240988969802856,
+      "learning_rate": 0.00016427977059917727,
+      "loss": 2.8678,
+      "step": 39755
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6791014671325684,
+      "learning_rate": 0.00016426604022821205,
+      "loss": 2.9517,
+      "step": 39756
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6903157234191895,
+      "learning_rate": 0.00016425231021474418,
+      "loss": 3.2275,
+      "step": 39757
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6479774117469788,
+      "learning_rate": 0.00016423858055880937,
+      "loss": 2.8358,
+      "step": 39758
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6902576088905334,
+      "learning_rate": 0.0001642248512604441,
+      "loss": 3.2466,
+      "step": 39759
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6335325837135315,
+      "learning_rate": 0.00016421112231968437,
+      "loss": 2.871,
+      "step": 39760
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7053777575492859,
+      "learning_rate": 0.00016419739373656642,
+      "loss": 2.9095,
+      "step": 39761
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6629794239997864,
+      "learning_rate": 0.00016418366551112644,
+      "loss": 3.159,
+      "step": 39762
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6280484199523926,
+      "learning_rate": 0.0001641699376434005,
+      "loss": 3.0098,
+      "step": 39763
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6917252540588379,
+      "learning_rate": 0.00016415621013342492,
+      "loss": 2.9122,
+      "step": 39764
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6806678771972656,
+      "learning_rate": 0.00016414248298123572,
+      "loss": 3.1373,
+      "step": 39765
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6609867215156555,
+      "learning_rate": 0.00016412875618686894,
+      "loss": 2.8685,
+      "step": 39766
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6640313267707825,
+      "learning_rate": 0.00016411502975036097,
+      "loss": 2.9706,
+      "step": 39767
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6441121697425842,
+      "learning_rate": 0.0001641013036717478,
+      "loss": 2.926,
+      "step": 39768
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6373971104621887,
+      "learning_rate": 0.00016408757795106574,
+      "loss": 3.1379,
+      "step": 39769
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6475521326065063,
+      "learning_rate": 0.0001640738525883507,
+      "loss": 3.017,
+      "step": 39770
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6653847694396973,
+      "learning_rate": 0.0001640601275836391,
+      "loss": 2.8866,
+      "step": 39771
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6433156132698059,
+      "learning_rate": 0.000164046402936967,
+      "loss": 2.9023,
+      "step": 39772
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6815069913864136,
+      "learning_rate": 0.00016403267864837035,
+      "loss": 3.0265,
+      "step": 39773
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6718235015869141,
+      "learning_rate": 0.00016401895471788557,
+      "loss": 3.1506,
+      "step": 39774
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7037975788116455,
+      "learning_rate": 0.00016400523114554858,
+      "loss": 3.1061,
+      "step": 39775
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6948851346969604,
+      "learning_rate": 0.00016399150793139567,
+      "loss": 2.9012,
+      "step": 39776
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6527612209320068,
+      "learning_rate": 0.00016397778507546307,
+      "loss": 2.7146,
+      "step": 39777
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6541974544525146,
+      "learning_rate": 0.00016396406257778663,
+      "loss": 2.8992,
+      "step": 39778
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6501400470733643,
+      "learning_rate": 0.00016395034043840284,
+      "loss": 3.1708,
+      "step": 39779
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6790065169334412,
+      "learning_rate": 0.00016393661865734765,
+      "loss": 3.1035,
+      "step": 39780
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7167571187019348,
+      "learning_rate": 0.0001639228972346571,
+      "loss": 2.9389,
+      "step": 39781
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6675640344619751,
+      "learning_rate": 0.00016390917617036758,
+      "loss": 2.938,
+      "step": 39782
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6446592211723328,
+      "learning_rate": 0.00016389545546451498,
+      "loss": 3.2109,
+      "step": 39783
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.637033998966217,
+      "learning_rate": 0.00016388173511713565,
+      "loss": 2.8193,
+      "step": 39784
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6496434807777405,
+      "learning_rate": 0.00016386801512826554,
+      "loss": 3.1458,
+      "step": 39785
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6477523446083069,
+      "learning_rate": 0.000163854295497941,
+      "loss": 3.1229,
+      "step": 39786
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6604776382446289,
+      "learning_rate": 0.00016384057622619803,
+      "loss": 2.9036,
+      "step": 39787
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6297851800918579,
+      "learning_rate": 0.00016382685731307267,
+      "loss": 2.9685,
+      "step": 39788
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.713854968547821,
+      "learning_rate": 0.0001638131387586013,
+      "loss": 3.1751,
+      "step": 39789
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6805860996246338,
+      "learning_rate": 0.00016379942056281977,
+      "loss": 2.9272,
+      "step": 39790
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6449320912361145,
+      "learning_rate": 0.00016378570272576436,
+      "loss": 2.8766,
+      "step": 39791
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6588577628135681,
+      "learning_rate": 0.00016377198524747128,
+      "loss": 2.8883,
+      "step": 39792
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6518809795379639,
+      "learning_rate": 0.00016375826812797662,
+      "loss": 2.8129,
+      "step": 39793
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6627069711685181,
+      "learning_rate": 0.00016374455136731642,
+      "loss": 3.0616,
+      "step": 39794
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6738892197608948,
+      "learning_rate": 0.0001637308349655267,
+      "loss": 3.2886,
+      "step": 39795
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6399411559104919,
+      "learning_rate": 0.00016371711892264386,
+      "loss": 3.0394,
+      "step": 39796
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6393530964851379,
+      "learning_rate": 0.00016370340323870392,
+      "loss": 2.8555,
+      "step": 39797
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6676591038703918,
+      "learning_rate": 0.0001636896879137429,
+      "loss": 3.174,
+      "step": 39798
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6630427241325378,
+      "learning_rate": 0.0001636759729477971,
+      "loss": 2.9689,
+      "step": 39799
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6797800064086914,
+      "learning_rate": 0.00016366225834090246,
+      "loss": 2.9354,
+      "step": 39800
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6970747709274292,
+      "learning_rate": 0.0001636485440930953,
+      "loss": 3.0008,
+      "step": 39801
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7164403796195984,
+      "learning_rate": 0.00016363483020441166,
+      "loss": 3.0449,
+      "step": 39802
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6394767165184021,
+      "learning_rate": 0.00016362111667488747,
+      "loss": 2.8067,
+      "step": 39803
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6600726842880249,
+      "learning_rate": 0.00016360740350455916,
+      "loss": 3.053,
+      "step": 39804
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6320900321006775,
+      "learning_rate": 0.00016359369069346257,
+      "loss": 3.0765,
+      "step": 39805
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7160202264785767,
+      "learning_rate": 0.00016357997824163395,
+      "loss": 2.9636,
+      "step": 39806
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6760608553886414,
+      "learning_rate": 0.00016356626614910956,
+      "loss": 3.1093,
+      "step": 39807
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6356624960899353,
+      "learning_rate": 0.0001635525544159253,
+      "loss": 3.0642,
+      "step": 39808
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6428079009056091,
+      "learning_rate": 0.0001635388430421174,
+      "loss": 3.1223,
+      "step": 39809
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7246979475021362,
+      "learning_rate": 0.00016352513202772182,
+      "loss": 2.7984,
+      "step": 39810
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6196163892745972,
+      "learning_rate": 0.00016351142137277483,
+      "loss": 3.017,
+      "step": 39811
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7229598164558411,
+      "learning_rate": 0.00016349771107731245,
+      "loss": 2.9938,
+      "step": 39812
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6564931869506836,
+      "learning_rate": 0.00016348400114137078,
+      "loss": 3.0232,
+      "step": 39813
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7086764574050903,
+      "learning_rate": 0.00016347029156498608,
+      "loss": 2.7794,
+      "step": 39814
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6638086438179016,
+      "learning_rate": 0.0001634565823481944,
+      "loss": 3.0231,
+      "step": 39815
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6503745317459106,
+      "learning_rate": 0.00016344287349103178,
+      "loss": 2.9724,
+      "step": 39816
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6424645781517029,
+      "learning_rate": 0.00016342916499353428,
+      "loss": 3.037,
+      "step": 39817
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6524776816368103,
+      "learning_rate": 0.00016341545685573798,
+      "loss": 2.8168,
+      "step": 39818
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6896033883094788,
+      "learning_rate": 0.00016340174907767926,
+      "loss": 2.7819,
+      "step": 39819
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6743185520172119,
+      "learning_rate": 0.00016338804165939394,
+      "loss": 2.8043,
+      "step": 39820
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7410269379615784,
+      "learning_rate": 0.0001633743346009183,
+      "loss": 2.8985,
+      "step": 39821
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6511000990867615,
+      "learning_rate": 0.00016336062790228828,
+      "loss": 2.9387,
+      "step": 39822
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6686780452728271,
+      "learning_rate": 0.0001633469215635402,
+      "loss": 3.1086,
+      "step": 39823
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6792110800743103,
+      "learning_rate": 0.00016333321558471,
+      "loss": 2.9515,
+      "step": 39824
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6629397869110107,
+      "learning_rate": 0.00016331950996583373,
+      "loss": 2.9703,
+      "step": 39825
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6876334547996521,
+      "learning_rate": 0.00016330580470694762,
+      "loss": 3.1131,
+      "step": 39826
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6542200446128845,
+      "learning_rate": 0.00016329209980808765,
+      "loss": 3.0926,
+      "step": 39827
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7039780616760254,
+      "learning_rate": 0.00016327839526928995,
+      "loss": 2.9963,
+      "step": 39828
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6578562259674072,
+      "learning_rate": 0.00016326469109059077,
+      "loss": 3.1772,
+      "step": 39829
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6808101534843445,
+      "learning_rate": 0.00016325098727202608,
+      "loss": 3.1833,
+      "step": 39830
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.650033175945282,
+      "learning_rate": 0.00016323728381363196,
+      "loss": 2.9292,
+      "step": 39831
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7056677341461182,
+      "learning_rate": 0.00016322358071544442,
+      "loss": 2.8928,
+      "step": 39832
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6719111204147339,
+      "learning_rate": 0.00016320987797749966,
+      "loss": 2.8892,
+      "step": 39833
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6775871515274048,
+      "learning_rate": 0.00016319617559983386,
+      "loss": 2.929,
+      "step": 39834
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6379905343055725,
+      "learning_rate": 0.00016318247358248285,
+      "loss": 3.0039,
+      "step": 39835
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.636997640132904,
+      "learning_rate": 0.00016316877192548303,
+      "loss": 3.059,
+      "step": 39836
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6647437810897827,
+      "learning_rate": 0.00016315507062887033,
+      "loss": 2.9804,
+      "step": 39837
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6169809103012085,
+      "learning_rate": 0.00016314136969268069,
+      "loss": 3.0341,
+      "step": 39838
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.661453366279602,
+      "learning_rate": 0.00016312766911695046,
+      "loss": 3.0617,
+      "step": 39839
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6447640657424927,
+      "learning_rate": 0.00016311396890171553,
+      "loss": 3.1025,
+      "step": 39840
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6600822806358337,
+      "learning_rate": 0.00016310026904701213,
+      "loss": 3.0792,
+      "step": 39841
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6973288655281067,
+      "learning_rate": 0.00016308656955287617,
+      "loss": 2.8341,
+      "step": 39842
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.656122088432312,
+      "learning_rate": 0.0001630728704193439,
+      "loss": 3.031,
+      "step": 39843
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6374558210372925,
+      "learning_rate": 0.0001630591716464514,
+      "loss": 3.0751,
+      "step": 39844
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6601612567901611,
+      "learning_rate": 0.00016304547323423469,
+      "loss": 3.0094,
+      "step": 39845
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7332333326339722,
+      "learning_rate": 0.00016303177518272983,
+      "loss": 2.8679,
+      "step": 39846
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6448028683662415,
+      "learning_rate": 0.00016301807749197284,
+      "loss": 3.2066,
+      "step": 39847
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.660399317741394,
+      "learning_rate": 0.00016300438016199982,
+      "loss": 3.1289,
+      "step": 39848
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5987576246261597,
+      "learning_rate": 0.00016299068319284706,
+      "loss": 3.0243,
+      "step": 39849
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6372739672660828,
+      "learning_rate": 0.0001629769865845503,
+      "loss": 2.5937,
+      "step": 39850
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7394205927848816,
+      "learning_rate": 0.00016296329033714597,
+      "loss": 3.0907,
+      "step": 39851
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6508176922798157,
+      "learning_rate": 0.0001629495944506699,
+      "loss": 3.2754,
+      "step": 39852
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6771491765975952,
+      "learning_rate": 0.00016293589892515817,
+      "loss": 3.012,
+      "step": 39853
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6909971833229065,
+      "learning_rate": 0.000162922203760647,
+      "loss": 3.0311,
+      "step": 39854
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7127150297164917,
+      "learning_rate": 0.00016290850895717222,
+      "loss": 3.0109,
+      "step": 39855
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6183626055717468,
+      "learning_rate": 0.00016289481451477015,
+      "loss": 2.9736,
+      "step": 39856
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6582019329071045,
+      "learning_rate": 0.0001628811204334767,
+      "loss": 2.9166,
+      "step": 39857
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5948113203048706,
+      "learning_rate": 0.00016286742671332803,
+      "loss": 3.0286,
+      "step": 39858
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6852336525917053,
+      "learning_rate": 0.00016285373335436022,
+      "loss": 2.9953,
+      "step": 39859
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7192272543907166,
+      "learning_rate": 0.00016284004035660915,
+      "loss": 2.9652,
+      "step": 39860
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.705686092376709,
+      "learning_rate": 0.00016282634772011112,
+      "loss": 2.9451,
+      "step": 39861
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6376373767852783,
+      "learning_rate": 0.00016281265544490202,
+      "loss": 3.0674,
+      "step": 39862
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.680194079875946,
+      "learning_rate": 0.00016279896353101793,
+      "loss": 3.0108,
+      "step": 39863
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6736658811569214,
+      "learning_rate": 0.00016278527197849507,
+      "loss": 2.8899,
+      "step": 39864
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.688944935798645,
+      "learning_rate": 0.00016277158078736932,
+      "loss": 2.7236,
+      "step": 39865
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6319748163223267,
+      "learning_rate": 0.00016275788995767692,
+      "loss": 2.7577,
+      "step": 39866
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.683836817741394,
+      "learning_rate": 0.00016274419948945383,
+      "loss": 2.9514,
+      "step": 39867
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6992816925048828,
+      "learning_rate": 0.00016273050938273598,
+      "loss": 3.1241,
+      "step": 39868
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6931037306785583,
+      "learning_rate": 0.00016271681963755968,
+      "loss": 3.0657,
+      "step": 39869
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7147294282913208,
+      "learning_rate": 0.0001627031302539607,
+      "loss": 2.9349,
+      "step": 39870
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6773857474327087,
+      "learning_rate": 0.0001626894412319754,
+      "loss": 2.9131,
+      "step": 39871
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6557199954986572,
+      "learning_rate": 0.0001626757525716396,
+      "loss": 2.7906,
+      "step": 39872
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6949388980865479,
+      "learning_rate": 0.00016266206427298954,
+      "loss": 3.2067,
+      "step": 39873
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6508885025978088,
+      "learning_rate": 0.00016264837633606113,
+      "loss": 2.8906,
+      "step": 39874
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6729174852371216,
+      "learning_rate": 0.00016263468876089038,
+      "loss": 3.0934,
+      "step": 39875
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6540240049362183,
+      "learning_rate": 0.00016262100154751353,
+      "loss": 2.9799,
+      "step": 39876
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7319003939628601,
+      "learning_rate": 0.00016260731469596639,
+      "loss": 3.0345,
+      "step": 39877
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6470528841018677,
+      "learning_rate": 0.00016259362820628513,
+      "loss": 2.9739,
+      "step": 39878
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6552753448486328,
+      "learning_rate": 0.000162579942078506,
+      "loss": 3.0781,
+      "step": 39879
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6793324947357178,
+      "learning_rate": 0.00016256625631266467,
+      "loss": 3.0156,
+      "step": 39880
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.636117160320282,
+      "learning_rate": 0.00016255257090879753,
+      "loss": 3.1659,
+      "step": 39881
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6674087643623352,
+      "learning_rate": 0.00016253888586694045,
+      "loss": 3.0635,
+      "step": 39882
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7054731249809265,
+      "learning_rate": 0.00016252520118712935,
+      "loss": 3.1278,
+      "step": 39883
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6451898813247681,
+      "learning_rate": 0.00016251151686940055,
+      "loss": 2.7129,
+      "step": 39884
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6893364787101746,
+      "learning_rate": 0.00016249783291378986,
+      "loss": 3.0023,
+      "step": 39885
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6900562047958374,
+      "learning_rate": 0.0001624841493203335,
+      "loss": 3.0193,
+      "step": 39886
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6447057723999023,
+      "learning_rate": 0.00016247046608906728,
+      "loss": 3.0807,
+      "step": 39887
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7077726125717163,
+      "learning_rate": 0.00016245678322002755,
+      "loss": 2.8504,
+      "step": 39888
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6641325354576111,
+      "learning_rate": 0.00016244310071325017,
+      "loss": 3.2085,
+      "step": 39889
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.701665461063385,
+      "learning_rate": 0.00016242941856877107,
+      "loss": 3.1182,
+      "step": 39890
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6562961339950562,
+      "learning_rate": 0.00016241573678662656,
+      "loss": 3.1035,
+      "step": 39891
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6483310461044312,
+      "learning_rate": 0.00016240205536685235,
+      "loss": 2.9854,
+      "step": 39892
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6495265364646912,
+      "learning_rate": 0.0001623883743094848,
+      "loss": 3.0443,
+      "step": 39893
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7036017179489136,
+      "learning_rate": 0.00016237469361455963,
+      "loss": 3.0209,
+      "step": 39894
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.674313485622406,
+      "learning_rate": 0.00016236101328211317,
+      "loss": 2.8208,
+      "step": 39895
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7025102972984314,
+      "learning_rate": 0.00016234733331218132,
+      "loss": 2.8201,
+      "step": 39896
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.66330885887146,
+      "learning_rate": 0.00016233365370479997,
+      "loss": 3.1881,
+      "step": 39897
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6553804874420166,
+      "learning_rate": 0.0001623199744600054,
+      "loss": 2.9553,
+      "step": 39898
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6467947959899902,
+      "learning_rate": 0.00016230629557783342,
+      "loss": 2.9607,
+      "step": 39899
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6393236517906189,
+      "learning_rate": 0.00016229261705832015,
+      "loss": 3.1014,
+      "step": 39900
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.675190269947052,
+      "learning_rate": 0.00016227893890150173,
+      "loss": 2.8303,
+      "step": 39901
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6858446002006531,
+      "learning_rate": 0.00016226526110741396,
+      "loss": 2.9271,
+      "step": 39902
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6574386954307556,
+      "learning_rate": 0.00016225158367609314,
+      "loss": 2.8188,
+      "step": 39903
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6161906123161316,
+      "learning_rate": 0.0001622379066075751,
+      "loss": 2.8077,
+      "step": 39904
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6860578060150146,
+      "learning_rate": 0.0001622242299018958,
+      "loss": 2.8483,
+      "step": 39905
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7148755788803101,
+      "learning_rate": 0.00016221055355909148,
+      "loss": 3.0416,
+      "step": 39906
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7170261740684509,
+      "learning_rate": 0.00016219687757919793,
+      "loss": 2.8434,
+      "step": 39907
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6350681185722351,
+      "learning_rate": 0.00016218320196225138,
+      "loss": 3.034,
+      "step": 39908
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6371099948883057,
+      "learning_rate": 0.00016216952670828762,
+      "loss": 3.0398,
+      "step": 39909
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6577742695808411,
+      "learning_rate": 0.00016215585181734296,
+      "loss": 2.7016,
+      "step": 39910
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6610209345817566,
+      "learning_rate": 0.00016214217728945324,
+      "loss": 3.0579,
+      "step": 39911
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.685881495475769,
+      "learning_rate": 0.00016212850312465437,
+      "loss": 2.9577,
+      "step": 39912
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6734350919723511,
+      "learning_rate": 0.00016211482932298264,
+      "loss": 3.0301,
+      "step": 39913
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6929466128349304,
+      "learning_rate": 0.00016210115588447378,
+      "loss": 3.0674,
+      "step": 39914
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6628326177597046,
+      "learning_rate": 0.00016208748280916393,
+      "loss": 3.0554,
+      "step": 39915
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6858869791030884,
+      "learning_rate": 0.00016207381009708923,
+      "loss": 2.9002,
+      "step": 39916
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6722764372825623,
+      "learning_rate": 0.00016206013774828557,
+      "loss": 2.8665,
+      "step": 39917
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6647232174873352,
+      "learning_rate": 0.00016204646576278893,
+      "loss": 3.0932,
+      "step": 39918
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7160334587097168,
+      "learning_rate": 0.00016203279414063525,
+      "loss": 2.9176,
+      "step": 39919
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6342455148696899,
+      "learning_rate": 0.00016201912288186067,
+      "loss": 3.0347,
+      "step": 39920
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.711457371711731,
+      "learning_rate": 0.00016200545198650126,
+      "loss": 3.0407,
+      "step": 39921
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6820720434188843,
+      "learning_rate": 0.00016199178145459277,
+      "loss": 3.1249,
+      "step": 39922
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7002881765365601,
+      "learning_rate": 0.00016197811128617153,
+      "loss": 2.8607,
+      "step": 39923
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7007981538772583,
+      "learning_rate": 0.0001619644414812733,
+      "loss": 2.8763,
+      "step": 39924
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6570713520050049,
+      "learning_rate": 0.0001619507720399342,
+      "loss": 2.885,
+      "step": 39925
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6478700637817383,
+      "learning_rate": 0.00016193710296219025,
+      "loss": 3.041,
+      "step": 39926
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7194265127182007,
+      "learning_rate": 0.0001619234342480773,
+      "loss": 2.9688,
+      "step": 39927
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6922836303710938,
+      "learning_rate": 0.00016190976589763151,
+      "loss": 2.9779,
+      "step": 39928
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6822351217269897,
+      "learning_rate": 0.00016189609791088878,
+      "loss": 2.9213,
+      "step": 39929
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6374616622924805,
+      "learning_rate": 0.00016188243028788507,
+      "loss": 2.8969,
+      "step": 39930
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6212733387947083,
+      "learning_rate": 0.00016186876302865663,
+      "loss": 2.8603,
+      "step": 39931
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6491883397102356,
+      "learning_rate": 0.00016185509613323928,
+      "loss": 3.0147,
+      "step": 39932
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6715843677520752,
+      "learning_rate": 0.000161841429601669,
+      "loss": 2.8747,
+      "step": 39933
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.658258318901062,
+      "learning_rate": 0.0001618277634339817,
+      "loss": 2.9536,
+      "step": 39934
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7133314609527588,
+      "learning_rate": 0.00016181409763021346,
+      "loss": 3.0629,
+      "step": 39935
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6304369568824768,
+      "learning_rate": 0.00016180043219040041,
+      "loss": 2.8939,
+      "step": 39936
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6943387985229492,
+      "learning_rate": 0.00016178676711457832,
+      "loss": 3.0713,
+      "step": 39937
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6555782556533813,
+      "learning_rate": 0.00016177310240278342,
+      "loss": 2.8656,
+      "step": 39938
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.634103000164032,
+      "learning_rate": 0.00016175943805505152,
+      "loss": 2.9851,
+      "step": 39939
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6928660869598389,
+      "learning_rate": 0.00016174577407141857,
+      "loss": 2.9063,
+      "step": 39940
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6522265076637268,
+      "learning_rate": 0.0001617321104519208,
+      "loss": 2.973,
+      "step": 39941
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7050772905349731,
+      "learning_rate": 0.00016171844719659384,
+      "loss": 2.8929,
+      "step": 39942
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6689801216125488,
+      "learning_rate": 0.000161704784305474,
+      "loss": 3.0081,
+      "step": 39943
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7276343107223511,
+      "learning_rate": 0.0001616911217785971,
+      "loss": 2.9564,
+      "step": 39944
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6313931345939636,
+      "learning_rate": 0.00016167745961599914,
+      "loss": 3.0023,
+      "step": 39945
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6859260201454163,
+      "learning_rate": 0.0001616637978177162,
+      "loss": 2.778,
+      "step": 39946
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6386987566947937,
+      "learning_rate": 0.00016165013638378423,
+      "loss": 2.9645,
+      "step": 39947
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6548264026641846,
+      "learning_rate": 0.0001616364753142392,
+      "loss": 2.9473,
+      "step": 39948
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6422974467277527,
+      "learning_rate": 0.00016162281460911694,
+      "loss": 2.829,
+      "step": 39949
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6441436409950256,
+      "learning_rate": 0.0001616091542684535,
+      "loss": 3.045,
+      "step": 39950
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6431066989898682,
+      "learning_rate": 0.0001615954942922851,
+      "loss": 3.079,
+      "step": 39951
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6600309610366821,
+      "learning_rate": 0.00016158183468064738,
+      "loss": 3.1532,
+      "step": 39952
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7169448733329773,
+      "learning_rate": 0.00016156817543357656,
+      "loss": 2.9717,
+      "step": 39953
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6597325801849365,
+      "learning_rate": 0.00016155451655110856,
+      "loss": 3.0926,
+      "step": 39954
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6279834508895874,
+      "learning_rate": 0.0001615408580332792,
+      "loss": 2.9351,
+      "step": 39955
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6826247572898865,
+      "learning_rate": 0.00016152719988012467,
+      "loss": 2.9439,
+      "step": 39956
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6647608876228333,
+      "learning_rate": 0.00016151354209168074,
+      "loss": 2.9722,
+      "step": 39957
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6797239780426025,
+      "learning_rate": 0.00016149988466798365,
+      "loss": 3.0933,
+      "step": 39958
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7346055507659912,
+      "learning_rate": 0.000161486227609069,
+      "loss": 2.8231,
+      "step": 39959
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6545577049255371,
+      "learning_rate": 0.00016147257091497316,
+      "loss": 3.0375,
+      "step": 39960
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6826615929603577,
+      "learning_rate": 0.0001614589145857319,
+      "loss": 2.7554,
+      "step": 39961
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6645894646644592,
+      "learning_rate": 0.00016144525862138106,
+      "loss": 2.9856,
+      "step": 39962
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6471953392028809,
+      "learning_rate": 0.00016143160302195682,
+      "loss": 3.0498,
+      "step": 39963
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6736137866973877,
+      "learning_rate": 0.00016141794778749502,
+      "loss": 3.0682,
+      "step": 39964
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.65671706199646,
+      "learning_rate": 0.00016140429291803166,
+      "loss": 2.9904,
+      "step": 39965
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6748507618904114,
+      "learning_rate": 0.00016139063841360285,
+      "loss": 3.1008,
+      "step": 39966
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6879149079322815,
+      "learning_rate": 0.00016137698427424427,
+      "loss": 2.9373,
+      "step": 39967
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.706725001335144,
+      "learning_rate": 0.00016136333049999215,
+      "loss": 3.0366,
+      "step": 39968
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6533642411231995,
+      "learning_rate": 0.00016134967709088236,
+      "loss": 2.8464,
+      "step": 39969
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6878400444984436,
+      "learning_rate": 0.0001613360240469508,
+      "loss": 3.0078,
+      "step": 39970
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6259313225746155,
+      "learning_rate": 0.00016132237136823338,
+      "loss": 2.9485,
+      "step": 39971
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6699492335319519,
+      "learning_rate": 0.00016130871905476615,
+      "loss": 3.141,
+      "step": 39972
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.653860867023468,
+      "learning_rate": 0.0001612950671065852,
+      "loss": 3.0009,
+      "step": 39973
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6694608330726624,
+      "learning_rate": 0.00016128141552372618,
+      "loss": 3.0055,
+      "step": 39974
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6927410960197449,
+      "learning_rate": 0.00016126776430622536,
+      "loss": 3.1213,
+      "step": 39975
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.690608561038971,
+      "learning_rate": 0.00016125411345411855,
+      "loss": 2.9176,
+      "step": 39976
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6918655633926392,
+      "learning_rate": 0.00016124046296744159,
+      "loss": 2.9613,
+      "step": 39977
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6636469960212708,
+      "learning_rate": 0.00016122681284623064,
+      "loss": 2.9775,
+      "step": 39978
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6595686078071594,
+      "learning_rate": 0.00016121316309052145,
+      "loss": 2.8676,
+      "step": 39979
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6494026184082031,
+      "learning_rate": 0.00016119951370035021,
+      "loss": 2.8945,
+      "step": 39980
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7089259624481201,
+      "learning_rate": 0.0001611858646757526,
+      "loss": 2.8564,
+      "step": 39981
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6717854142189026,
+      "learning_rate": 0.0001611722160167647,
+      "loss": 2.9878,
+      "step": 39982
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6851207613945007,
+      "learning_rate": 0.00016115856772342275,
+      "loss": 3.0027,
+      "step": 39983
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6713508367538452,
+      "learning_rate": 0.00016114491979576215,
+      "loss": 3.0956,
+      "step": 39984
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6420844793319702,
+      "learning_rate": 0.00016113127223381919,
+      "loss": 3.0019,
+      "step": 39985
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.63764488697052,
+      "learning_rate": 0.00016111762503762965,
+      "loss": 3.0084,
+      "step": 39986
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6774821877479553,
+      "learning_rate": 0.00016110397820722953,
+      "loss": 2.9804,
+      "step": 39987
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6736724376678467,
+      "learning_rate": 0.00016109033174265498,
+      "loss": 3.0381,
+      "step": 39988
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6819289922714233,
+      "learning_rate": 0.00016107668564394157,
+      "loss": 3.1965,
+      "step": 39989
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7004780769348145,
+      "learning_rate": 0.0001610630399111256,
+      "loss": 2.8297,
+      "step": 39990
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6639523506164551,
+      "learning_rate": 0.00016104939454424284,
+      "loss": 3.05,
+      "step": 39991
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7015445828437805,
+      "learning_rate": 0.00016103574954332906,
+      "loss": 2.9532,
+      "step": 39992
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6690757274627686,
+      "learning_rate": 0.00016102210490842058,
+      "loss": 2.9515,
+      "step": 39993
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6284964680671692,
+      "learning_rate": 0.00016100846063955295,
+      "loss": 3.1455,
+      "step": 39994
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7172853350639343,
+      "learning_rate": 0.00016099481673676242,
+      "loss": 2.9555,
+      "step": 39995
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6635323166847229,
+      "learning_rate": 0.00016098117320008465,
+      "loss": 3.0923,
+      "step": 39996
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6090884208679199,
+      "learning_rate": 0.00016096753002955587,
+      "loss": 3.099,
+      "step": 39997
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6727035045623779,
+      "learning_rate": 0.00016095388722521184,
+      "loss": 3.0681,
+      "step": 39998
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6885789036750793,
+      "learning_rate": 0.0001609402447870884,
+      "loss": 3.3607,
+      "step": 39999
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6628068685531616,
+      "learning_rate": 0.0001609266027152217,
+      "loss": 2.766,
+      "step": 40000
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7326607704162598,
+      "learning_rate": 0.00016091296100964746,
+      "loss": 2.8427,
+      "step": 40001
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.637006938457489,
+      "learning_rate": 0.00016089931967040171,
+      "loss": 2.9066,
+      "step": 40002
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6497381329536438,
+      "learning_rate": 0.0001608856786975205,
+      "loss": 2.8787,
+      "step": 40003
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6747108101844788,
+      "learning_rate": 0.00016087203809103952,
+      "loss": 3.0503,
+      "step": 40004
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6333727240562439,
+      "learning_rate": 0.00016085839785099493,
+      "loss": 2.9565,
+      "step": 40005
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6521469950675964,
+      "learning_rate": 0.00016084475797742252,
+      "loss": 3.0972,
+      "step": 40006
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7146298885345459,
+      "learning_rate": 0.00016083111847035815,
+      "loss": 2.9459,
+      "step": 40007
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6354973912239075,
+      "learning_rate": 0.00016081747932983792,
+      "loss": 3.2163,
+      "step": 40008
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6594139337539673,
+      "learning_rate": 0.00016080384055589756,
+      "loss": 2.9049,
+      "step": 40009
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.693379282951355,
+      "learning_rate": 0.00016079020214857325,
+      "loss": 2.8049,
+      "step": 40010
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6748722791671753,
+      "learning_rate": 0.0001607765641079006,
+      "loss": 2.7994,
+      "step": 40011
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6617879867553711,
+      "learning_rate": 0.0001607629264339158,
+      "loss": 2.9751,
+      "step": 40012
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6831246018409729,
+      "learning_rate": 0.00016074928912665468,
+      "loss": 2.9856,
+      "step": 40013
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6496304869651794,
+      "learning_rate": 0.000160735652186153,
+      "loss": 2.932,
+      "step": 40014
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6654090881347656,
+      "learning_rate": 0.0001607220156124469,
+      "loss": 3.2564,
+      "step": 40015
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.690848171710968,
+      "learning_rate": 0.00016070837940557214,
+      "loss": 3.0275,
+      "step": 40016
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6424733400344849,
+      "learning_rate": 0.0001606947435655647,
+      "loss": 3.1051,
+      "step": 40017
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6729825735092163,
+      "learning_rate": 0.00016068110809246058,
+      "loss": 3.0315,
+      "step": 40018
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6387492418289185,
+      "learning_rate": 0.00016066747298629558,
+      "loss": 3.0497,
+      "step": 40019
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6648398637771606,
+      "learning_rate": 0.0001606538382471057,
+      "loss": 2.9517,
+      "step": 40020
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6861499547958374,
+      "learning_rate": 0.00016064020387492666,
+      "loss": 2.9324,
+      "step": 40021
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6802994608879089,
+      "learning_rate": 0.00016062656986979448,
+      "loss": 2.9477,
+      "step": 40022
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6521588563919067,
+      "learning_rate": 0.00016061293623174523,
+      "loss": 2.8447,
+      "step": 40023
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7014820575714111,
+      "learning_rate": 0.00016059930296081457,
+      "loss": 2.9435,
+      "step": 40024
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.634397029876709,
+      "learning_rate": 0.00016058567005703864,
+      "loss": 3.1462,
+      "step": 40025
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6457260847091675,
+      "learning_rate": 0.00016057203752045307,
+      "loss": 2.9698,
+      "step": 40026
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7065685391426086,
+      "learning_rate": 0.00016055840535109407,
+      "loss": 3.1251,
+      "step": 40027
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6713890433311462,
+      "learning_rate": 0.00016054477354899739,
+      "loss": 2.9537,
+      "step": 40028
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.674758791923523,
+      "learning_rate": 0.0001605311421141988,
+      "loss": 3.1935,
+      "step": 40029
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6733911633491516,
+      "learning_rate": 0.00016051751104673447,
+      "loss": 3.0881,
+      "step": 40030
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6857780814170837,
+      "learning_rate": 0.00016050388034664006,
+      "loss": 2.8133,
+      "step": 40031
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6858083605766296,
+      "learning_rate": 0.00016049025001395158,
+      "loss": 2.9045,
+      "step": 40032
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6884641647338867,
+      "learning_rate": 0.00016047662004870506,
+      "loss": 3.1941,
+      "step": 40033
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6528140902519226,
+      "learning_rate": 0.00016046299045093624,
+      "loss": 2.9926,
+      "step": 40034
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.644972026348114,
+      "learning_rate": 0.00016044936122068106,
+      "loss": 2.9634,
+      "step": 40035
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6678783297538757,
+      "learning_rate": 0.00016043573235797532,
+      "loss": 3.021,
+      "step": 40036
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6511356830596924,
+      "learning_rate": 0.00016042210386285497,
+      "loss": 2.9231,
+      "step": 40037
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6782459020614624,
+      "learning_rate": 0.00016040847573535607,
+      "loss": 2.848,
+      "step": 40038
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7153217792510986,
+      "learning_rate": 0.00016039484797551425,
+      "loss": 3.042,
+      "step": 40039
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6678678393363953,
+      "learning_rate": 0.00016038122058336567,
+      "loss": 3.0906,
+      "step": 40040
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6697278618812561,
+      "learning_rate": 0.00016036759355894604,
+      "loss": 3.0417,
+      "step": 40041
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6855822801589966,
+      "learning_rate": 0.00016035396690229124,
+      "loss": 3.2378,
+      "step": 40042
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6451258659362793,
+      "learning_rate": 0.00016034034061343728,
+      "loss": 3.1821,
+      "step": 40043
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8559314012527466,
+      "learning_rate": 0.00016032671469241988,
+      "loss": 2.936,
+      "step": 40044
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6715162396430969,
+      "learning_rate": 0.00016031308913927515,
+      "loss": 3.0956,
+      "step": 40045
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6763110160827637,
+      "learning_rate": 0.00016029946395403877,
+      "loss": 3.0766,
+      "step": 40046
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6643348336219788,
+      "learning_rate": 0.0001602858391367468,
+      "loss": 2.9967,
+      "step": 40047
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6329568028450012,
+      "learning_rate": 0.00016027221468743497,
+      "loss": 3.013,
+      "step": 40048
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6494516134262085,
+      "learning_rate": 0.0001602585906061393,
+      "loss": 2.789,
+      "step": 40049
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6577889919281006,
+      "learning_rate": 0.00016024496689289564,
+      "loss": 3.0301,
+      "step": 40050
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7097561359405518,
+      "learning_rate": 0.0001602313435477397,
+      "loss": 2.8324,
+      "step": 40051
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.652508556842804,
+      "learning_rate": 0.0001602177205707076,
+      "loss": 2.8232,
+      "step": 40052
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6659618020057678,
+      "learning_rate": 0.00016020409796183507,
+      "loss": 2.7219,
+      "step": 40053
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6810920834541321,
+      "learning_rate": 0.000160190475721158,
+      "loss": 2.9999,
+      "step": 40054
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7163436412811279,
+      "learning_rate": 0.00016017685384871243,
+      "loss": 3.0742,
+      "step": 40055
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6159772276878357,
+      "learning_rate": 0.00016016323234453413,
+      "loss": 2.757,
+      "step": 40056
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.642678439617157,
+      "learning_rate": 0.00016014961120865894,
+      "loss": 2.9309,
+      "step": 40057
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7379891276359558,
+      "learning_rate": 0.0001601359904411227,
+      "loss": 2.9141,
+      "step": 40058
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6562783122062683,
+      "learning_rate": 0.0001601223700419613,
+      "loss": 3.2553,
+      "step": 40059
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6276502013206482,
+      "learning_rate": 0.00016010875001121075,
+      "loss": 3.117,
+      "step": 40060
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.681545615196228,
+      "learning_rate": 0.00016009513034890674,
+      "loss": 3.1442,
+      "step": 40061
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6715508699417114,
+      "learning_rate": 0.00016008151105508536,
+      "loss": 2.8528,
+      "step": 40062
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6921969652175903,
+      "learning_rate": 0.00016006789212978233,
+      "loss": 3.1951,
+      "step": 40063
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.635017454624176,
+      "learning_rate": 0.00016005427357303347,
+      "loss": 3.0769,
+      "step": 40064
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6426366567611694,
+      "learning_rate": 0.00016004065538487478,
+      "loss": 3.112,
+      "step": 40065
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6538386940956116,
+      "learning_rate": 0.00016002703756534195,
+      "loss": 3.0287,
+      "step": 40066
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6401711702346802,
+      "learning_rate": 0.00016001342011447113,
+      "loss": 2.9856,
+      "step": 40067
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6818058490753174,
+      "learning_rate": 0.00015999980303229788,
+      "loss": 2.9975,
+      "step": 40068
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6866037845611572,
+      "learning_rate": 0.0001599861863188582,
+      "loss": 2.8041,
+      "step": 40069
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6289970874786377,
+      "learning_rate": 0.00015997256997418808,
+      "loss": 2.9064,
+      "step": 40070
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6188867688179016,
+      "learning_rate": 0.0001599589539983233,
+      "loss": 2.975,
+      "step": 40071
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6899058818817139,
+      "learning_rate": 0.00015994533839129961,
+      "loss": 3.1588,
+      "step": 40072
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6458503603935242,
+      "learning_rate": 0.00015993172315315285,
+      "loss": 2.9297,
+      "step": 40073
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6466934084892273,
+      "learning_rate": 0.000159918108283919,
+      "loss": 2.9095,
+      "step": 40074
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6431629657745361,
+      "learning_rate": 0.000159904493783634,
+      "loss": 3.019,
+      "step": 40075
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6387059688568115,
+      "learning_rate": 0.00015989087965233347,
+      "loss": 2.7465,
+      "step": 40076
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6597829461097717,
+      "learning_rate": 0.00015987726589005352,
+      "loss": 3.1271,
+      "step": 40077
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.673991858959198,
+      "learning_rate": 0.00015986365249682992,
+      "loss": 3.0424,
+      "step": 40078
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6489351391792297,
+      "learning_rate": 0.00015985003947269834,
+      "loss": 2.8951,
+      "step": 40079
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7270581722259521,
+      "learning_rate": 0.00015983642681769494,
+      "loss": 2.8268,
+      "step": 40080
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7210774421691895,
+      "learning_rate": 0.00015982281453185522,
+      "loss": 3.0683,
+      "step": 40081
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6519688963890076,
+      "learning_rate": 0.0001598092026152154,
+      "loss": 2.9316,
+      "step": 40082
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6710745692253113,
+      "learning_rate": 0.00015979559106781103,
+      "loss": 3.1201,
+      "step": 40083
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6843429803848267,
+      "learning_rate": 0.00015978197988967811,
+      "loss": 2.9192,
+      "step": 40084
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6874869465827942,
+      "learning_rate": 0.00015976836908085268,
+      "loss": 2.9876,
+      "step": 40085
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6737939715385437,
+      "learning_rate": 0.00015975475864137017,
+      "loss": 2.9837,
+      "step": 40086
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.69536292552948,
+      "learning_rate": 0.00015974114857126675,
+      "loss": 3.025,
+      "step": 40087
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6652324795722961,
+      "learning_rate": 0.00015972753887057799,
+      "loss": 2.9949,
+      "step": 40088
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6829466223716736,
+      "learning_rate": 0.0001597139295393399,
+      "loss": 3.1127,
+      "step": 40089
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7316030859947205,
+      "learning_rate": 0.0001597003205775885,
+      "loss": 2.9742,
+      "step": 40090
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6916874051094055,
+      "learning_rate": 0.0001596867119853593,
+      "loss": 3.2263,
+      "step": 40091
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.647091805934906,
+      "learning_rate": 0.00015967310376268837,
+      "loss": 3.0349,
+      "step": 40092
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6750726699829102,
+      "learning_rate": 0.00015965949590961156,
+      "loss": 2.8189,
+      "step": 40093
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6377853751182556,
+      "learning_rate": 0.00015964588842616446,
+      "loss": 2.7829,
+      "step": 40094
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7172527313232422,
+      "learning_rate": 0.00015963228131238324,
+      "loss": 3.0597,
+      "step": 40095
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6895819902420044,
+      "learning_rate": 0.00015961867456830339,
+      "loss": 3.0861,
+      "step": 40096
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7082463502883911,
+      "learning_rate": 0.0001596050681939611,
+      "loss": 2.7711,
+      "step": 40097
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6395615339279175,
+      "learning_rate": 0.00015959146218939193,
+      "loss": 3.0708,
+      "step": 40098
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6931965351104736,
+      "learning_rate": 0.00015957785655463193,
+      "loss": 2.9829,
+      "step": 40099
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6493892073631287,
+      "learning_rate": 0.00015956425128971682,
+      "loss": 3.0342,
+      "step": 40100
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6747897267341614,
+      "learning_rate": 0.0001595506463946823,
+      "loss": 2.8114,
+      "step": 40101
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6698330044746399,
+      "learning_rate": 0.00015953704186956452,
+      "loss": 3.0373,
+      "step": 40102
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6742802262306213,
+      "learning_rate": 0.00015952343771439903,
+      "loss": 3.0902,
+      "step": 40103
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6375805735588074,
+      "learning_rate": 0.00015950983392922175,
+      "loss": 3.0517,
+      "step": 40104
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6356774568557739,
+      "learning_rate": 0.0001594962305140686,
+      "loss": 2.8752,
+      "step": 40105
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6412781476974487,
+      "learning_rate": 0.00015948262746897526,
+      "loss": 3.0477,
+      "step": 40106
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6381820440292358,
+      "learning_rate": 0.00015946902479397777,
+      "loss": 3.0045,
+      "step": 40107
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6312680840492249,
+      "learning_rate": 0.0001594554224891118,
+      "loss": 2.9117,
+      "step": 40108
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6366028785705566,
+      "learning_rate": 0.0001594418205544131,
+      "loss": 3.0584,
+      "step": 40109
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6530911922454834,
+      "learning_rate": 0.0001594282189899177,
+      "loss": 2.8788,
+      "step": 40110
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.653728187084198,
+      "learning_rate": 0.0001594146177956612,
+      "loss": 2.8951,
+      "step": 40111
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6730470061302185,
+      "learning_rate": 0.0001594010169716797,
+      "loss": 3.071,
+      "step": 40112
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6829031705856323,
+      "learning_rate": 0.00015938741651800868,
+      "loss": 2.9737,
+      "step": 40113
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6463053226470947,
+      "learning_rate": 0.00015937381643468427,
+      "loss": 3.0131,
+      "step": 40114
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6664642095565796,
+      "learning_rate": 0.00015936021672174216,
+      "loss": 2.7908,
+      "step": 40115
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6486683487892151,
+      "learning_rate": 0.00015934661737921808,
+      "loss": 3.1284,
+      "step": 40116
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6821164488792419,
+      "learning_rate": 0.00015933301840714804,
+      "loss": 3.0977,
+      "step": 40117
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6829363107681274,
+      "learning_rate": 0.00015931941980556767,
+      "loss": 3.0408,
+      "step": 40118
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7041226625442505,
+      "learning_rate": 0.00015930582157451284,
+      "loss": 2.9775,
+      "step": 40119
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6732624769210815,
+      "learning_rate": 0.00015929222371401956,
+      "loss": 2.9952,
+      "step": 40120
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6260468363761902,
+      "learning_rate": 0.00015927862622412345,
+      "loss": 3.1095,
+      "step": 40121
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6741329431533813,
+      "learning_rate": 0.00015926502910486033,
+      "loss": 2.9427,
+      "step": 40122
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6579768657684326,
+      "learning_rate": 0.00015925143235626595,
+      "loss": 3.0081,
+      "step": 40123
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6937235593795776,
+      "learning_rate": 0.00015923783597837635,
+      "loss": 3.1918,
+      "step": 40124
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6266592144966125,
+      "learning_rate": 0.00015922423997122705,
+      "loss": 2.9933,
+      "step": 40125
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6419720649719238,
+      "learning_rate": 0.00015921064433485404,
+      "loss": 2.8718,
+      "step": 40126
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7935446500778198,
+      "learning_rate": 0.00015919704906929317,
+      "loss": 2.76,
+      "step": 40127
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6469537615776062,
+      "learning_rate": 0.00015918345417458008,
+      "loss": 2.8866,
+      "step": 40128
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6699641346931458,
+      "learning_rate": 0.00015916985965075078,
+      "loss": 2.925,
+      "step": 40129
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6339029669761658,
+      "learning_rate": 0.000159156265497841,
+      "loss": 2.9374,
+      "step": 40130
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6614867448806763,
+      "learning_rate": 0.00015914267171588637,
+      "loss": 2.8457,
+      "step": 40131
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6942468881607056,
+      "learning_rate": 0.00015912907830492292,
+      "loss": 3.162,
+      "step": 40132
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6900768280029297,
+      "learning_rate": 0.00015911548526498627,
+      "loss": 3.1054,
+      "step": 40133
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6508978009223938,
+      "learning_rate": 0.00015910189259611248,
+      "loss": 2.7635,
+      "step": 40134
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6943615674972534,
+      "learning_rate": 0.00015908830029833702,
+      "loss": 3.0908,
+      "step": 40135
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6539590954780579,
+      "learning_rate": 0.00015907470837169598,
+      "loss": 2.8507,
+      "step": 40136
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6708890795707703,
+      "learning_rate": 0.0001590611168162251,
+      "loss": 3.1577,
+      "step": 40137
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7016001343727112,
+      "learning_rate": 0.00015904752563195994,
+      "loss": 2.9599,
+      "step": 40138
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.716665506362915,
+      "learning_rate": 0.00015903393481893658,
+      "loss": 3.1311,
+      "step": 40139
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6903090476989746,
+      "learning_rate": 0.0001590203443771906,
+      "loss": 2.9423,
+      "step": 40140
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6860828995704651,
+      "learning_rate": 0.0001590067543067579,
+      "loss": 3.0349,
+      "step": 40141
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6501972675323486,
+      "learning_rate": 0.00015899316460767437,
+      "loss": 2.9293,
+      "step": 40142
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6735811233520508,
+      "learning_rate": 0.00015897957527997574,
+      "loss": 3.2349,
+      "step": 40143
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6617088317871094,
+      "learning_rate": 0.00015896598632369774,
+      "loss": 2.9299,
+      "step": 40144
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6840248703956604,
+      "learning_rate": 0.00015895239773887611,
+      "loss": 3.0315,
+      "step": 40145
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6260735988616943,
+      "learning_rate": 0.0001589388095255467,
+      "loss": 2.7975,
+      "step": 40146
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7066246867179871,
+      "learning_rate": 0.00015892522168374547,
+      "loss": 2.9681,
+      "step": 40147
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6458545327186584,
+      "learning_rate": 0.00015891163421350788,
+      "loss": 2.8748,
+      "step": 40148
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6603613495826721,
+      "learning_rate": 0.00015889804711487007,
+      "loss": 3.029,
+      "step": 40149
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6358199715614319,
+      "learning_rate": 0.00015888446038786746,
+      "loss": 2.9619,
+      "step": 40150
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7179542779922485,
+      "learning_rate": 0.00015887087403253618,
+      "loss": 2.9119,
+      "step": 40151
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6448338627815247,
+      "learning_rate": 0.00015885728804891186,
+      "loss": 2.9725,
+      "step": 40152
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7053864002227783,
+      "learning_rate": 0.00015884370243703018,
+      "loss": 2.6094,
+      "step": 40153
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6579211354255676,
+      "learning_rate": 0.0001588301171969271,
+      "loss": 2.9604,
+      "step": 40154
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6514232754707336,
+      "learning_rate": 0.00015881653232863822,
+      "loss": 3.1176,
+      "step": 40155
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6942654848098755,
+      "learning_rate": 0.00015880294783219943,
+      "loss": 3.1528,
+      "step": 40156
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7441583871841431,
+      "learning_rate": 0.0001587893637076466,
+      "loss": 2.9195,
+      "step": 40157
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6600908041000366,
+      "learning_rate": 0.00015877577995501542,
+      "loss": 2.9988,
+      "step": 40158
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6638275980949402,
+      "learning_rate": 0.00015876219657434164,
+      "loss": 2.9348,
+      "step": 40159
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6445525288581848,
+      "learning_rate": 0.00015874861356566096,
+      "loss": 2.8587,
+      "step": 40160
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6753540635108948,
+      "learning_rate": 0.00015873503092900924,
+      "loss": 2.8209,
+      "step": 40161
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6722790002822876,
+      "learning_rate": 0.0001587214486644224,
+      "loss": 2.9628,
+      "step": 40162
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7291383147239685,
+      "learning_rate": 0.00015870786677193592,
+      "loss": 3.0806,
+      "step": 40163
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6721978187561035,
+      "learning_rate": 0.00015869428525158588,
+      "loss": 3.0379,
+      "step": 40164
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7075223922729492,
+      "learning_rate": 0.00015868070410340787,
+      "loss": 3.0957,
+      "step": 40165
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6854174733161926,
+      "learning_rate": 0.00015866712332743757,
+      "loss": 2.9606,
+      "step": 40166
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.678521454334259,
+      "learning_rate": 0.00015865354292371098,
+      "loss": 2.8023,
+      "step": 40167
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6438807249069214,
+      "learning_rate": 0.0001586399628922637,
+      "loss": 3.0427,
+      "step": 40168
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6650844216346741,
+      "learning_rate": 0.00015862638323313157,
+      "loss": 3.129,
+      "step": 40169
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6512045860290527,
+      "learning_rate": 0.00015861280394635028,
+      "loss": 3.1556,
+      "step": 40170
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6742256879806519,
+      "learning_rate": 0.00015859922503195563,
+      "loss": 2.9698,
+      "step": 40171
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6796380877494812,
+      "learning_rate": 0.00015858564648998355,
+      "loss": 2.9838,
+      "step": 40172
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6988964080810547,
+      "learning_rate": 0.00015857206832046967,
+      "loss": 2.9046,
+      "step": 40173
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6292441487312317,
+      "learning_rate": 0.00015855849052344974,
+      "loss": 3.0492,
+      "step": 40174
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6308088898658752,
+      "learning_rate": 0.0001585449130989594,
+      "loss": 2.9864,
+      "step": 40175
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6453452706336975,
+      "learning_rate": 0.00015853133604703452,
+      "loss": 3.0101,
+      "step": 40176
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7138879299163818,
+      "learning_rate": 0.00015851775936771102,
+      "loss": 3.0329,
+      "step": 40177
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.687667191028595,
+      "learning_rate": 0.00015850418306102436,
+      "loss": 2.7237,
+      "step": 40178
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.720391571521759,
+      "learning_rate": 0.0001584906071270106,
+      "loss": 3.1241,
+      "step": 40179
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6851097345352173,
+      "learning_rate": 0.0001584770315657053,
+      "loss": 2.8208,
+      "step": 40180
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6699183583259583,
+      "learning_rate": 0.00015846345637714418,
+      "loss": 3.0279,
+      "step": 40181
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6366994976997375,
+      "learning_rate": 0.00015844988156136323,
+      "loss": 2.9347,
+      "step": 40182
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6660268306732178,
+      "learning_rate": 0.00015843630711839788,
+      "loss": 2.9707,
+      "step": 40183
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6552741527557373,
+      "learning_rate": 0.00015842273304828415,
+      "loss": 2.951,
+      "step": 40184
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7053703665733337,
+      "learning_rate": 0.0001584091593510576,
+      "loss": 2.9394,
+      "step": 40185
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6609983444213867,
+      "learning_rate": 0.00015839558602675423,
+      "loss": 2.9316,
+      "step": 40186
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6549779772758484,
+      "learning_rate": 0.00015838201307540959,
+      "loss": 2.8942,
+      "step": 40187
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6527225971221924,
+      "learning_rate": 0.0001583684404970594,
+      "loss": 3.155,
+      "step": 40188
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6407042145729065,
+      "learning_rate": 0.00015835486829173954,
+      "loss": 2.8456,
+      "step": 40189
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6975271701812744,
+      "learning_rate": 0.00015834129645948564,
+      "loss": 2.7692,
+      "step": 40190
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6264335513114929,
+      "learning_rate": 0.00015832772500033344,
+      "loss": 2.8132,
+      "step": 40191
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6382396817207336,
+      "learning_rate": 0.00015831415391431893,
+      "loss": 2.8311,
+      "step": 40192
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6645261645317078,
+      "learning_rate": 0.0001583005832014775,
+      "loss": 3.2165,
+      "step": 40193
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6498206257820129,
+      "learning_rate": 0.00015828701286184517,
+      "loss": 2.9257,
+      "step": 40194
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6160268187522888,
+      "learning_rate": 0.00015827344289545765,
+      "loss": 2.9067,
+      "step": 40195
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6547241806983948,
+      "learning_rate": 0.00015825987330235044,
+      "loss": 2.8953,
+      "step": 40196
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6759021282196045,
+      "learning_rate": 0.00015824630408255955,
+      "loss": 2.8691,
+      "step": 40197
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6542566418647766,
+      "learning_rate": 0.0001582327352361205,
+      "loss": 2.8604,
+      "step": 40198
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6945884227752686,
+      "learning_rate": 0.0001582191667630693,
+      "loss": 3.0368,
+      "step": 40199
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6422370076179504,
+      "learning_rate": 0.00015820559866344138,
+      "loss": 2.9877,
+      "step": 40200
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6469019651412964,
+      "learning_rate": 0.00015819203093727271,
+      "loss": 3.0606,
+      "step": 40201
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6708106398582458,
+      "learning_rate": 0.000158178463584599,
+      "loss": 3.2249,
+      "step": 40202
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6726407408714294,
+      "learning_rate": 0.00015816489660545579,
+      "loss": 2.8245,
+      "step": 40203
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6494171619415283,
+      "learning_rate": 0.00015815132999987905,
+      "loss": 3.1015,
+      "step": 40204
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6671391725540161,
+      "learning_rate": 0.00015813776376790426,
+      "loss": 2.937,
+      "step": 40205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6737003922462463,
+      "learning_rate": 0.00015812419790956744,
+      "loss": 2.9281,
+      "step": 40206
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6515318155288696,
+      "learning_rate": 0.00015811063242490407,
+      "loss": 3.1345,
+      "step": 40207
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6995930671691895,
+      "learning_rate": 0.00015809706731395002,
+      "loss": 2.8716,
+      "step": 40208
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6816954612731934,
+      "learning_rate": 0.00015808350257674113,
+      "loss": 3.0883,
+      "step": 40209
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6604844331741333,
+      "learning_rate": 0.00015806993821331283,
+      "loss": 3.192,
+      "step": 40210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6579017043113708,
+      "learning_rate": 0.0001580563742237011,
+      "loss": 2.839,
+      "step": 40211
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6625229716300964,
+      "learning_rate": 0.00015804281060794142,
+      "loss": 2.9405,
+      "step": 40212
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7139385938644409,
+      "learning_rate": 0.00015802924736606965,
+      "loss": 3.0133,
+      "step": 40213
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6507391333580017,
+      "learning_rate": 0.00015801568449812166,
+      "loss": 3.032,
+      "step": 40214
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6921162009239197,
+      "learning_rate": 0.00015800212200413292,
+      "loss": 2.862,
+      "step": 40215
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6766071319580078,
+      "learning_rate": 0.00015798855988413937,
+      "loss": 3.1456,
+      "step": 40216
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7535398602485657,
+      "learning_rate": 0.00015797499813817665,
+      "loss": 2.9585,
+      "step": 40217
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6730259656906128,
+      "learning_rate": 0.0001579614367662803,
+      "loss": 3.04,
+      "step": 40218
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6672115921974182,
+      "learning_rate": 0.0001579478757684863,
+      "loss": 2.9337,
+      "step": 40219
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.708581805229187,
+      "learning_rate": 0.00015793431514483016,
+      "loss": 2.9796,
+      "step": 40220
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6547756791114807,
+      "learning_rate": 0.00015792075489534782,
+      "loss": 2.9422,
+      "step": 40221
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6528579592704773,
+      "learning_rate": 0.0001579071950200747,
+      "loss": 2.9878,
+      "step": 40222
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6749711036682129,
+      "learning_rate": 0.00015789363551904685,
+      "loss": 2.8666,
+      "step": 40223
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6884815692901611,
+      "learning_rate": 0.0001578800763922998,
+      "loss": 2.9652,
+      "step": 40224
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6515948176383972,
+      "learning_rate": 0.00015786651763986912,
+      "loss": 2.9444,
+      "step": 40225
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7087914347648621,
+      "learning_rate": 0.00015785295926179087,
+      "loss": 2.9341,
+      "step": 40226
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6495019197463989,
+      "learning_rate": 0.0001578394012581004,
+      "loss": 3.0155,
+      "step": 40227
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6784089207649231,
+      "learning_rate": 0.00015782584362883362,
+      "loss": 3.0786,
+      "step": 40228
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6523367166519165,
+      "learning_rate": 0.00015781228637402629,
+      "loss": 2.9048,
+      "step": 40229
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6989542245864868,
+      "learning_rate": 0.00015779872949371395,
+      "loss": 3.0083,
+      "step": 40230
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6683775186538696,
+      "learning_rate": 0.00015778517298793249,
+      "loss": 2.8116,
+      "step": 40231
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7025153636932373,
+      "learning_rate": 0.0001577716168567175,
+      "loss": 3.2109,
+      "step": 40232
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6500304937362671,
+      "learning_rate": 0.0001577580611001046,
+      "loss": 3.2659,
+      "step": 40233
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6912034749984741,
+      "learning_rate": 0.0001577445057181297,
+      "loss": 2.8978,
+      "step": 40234
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6490848660469055,
+      "learning_rate": 0.00015773095071082825,
+      "loss": 3.2104,
+      "step": 40235
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7083659172058105,
+      "learning_rate": 0.00015771739607823627,
+      "loss": 2.9851,
+      "step": 40236
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6967908143997192,
+      "learning_rate": 0.00015770384182038914,
+      "loss": 3.0441,
+      "step": 40237
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7174261808395386,
+      "learning_rate": 0.00015769028793732282,
+      "loss": 3.1049,
+      "step": 40238
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6764041185379028,
+      "learning_rate": 0.0001576767344290729,
+      "loss": 2.8842,
+      "step": 40239
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6803184747695923,
+      "learning_rate": 0.00015766318129567498,
+      "loss": 3.1289,
+      "step": 40240
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6668773293495178,
+      "learning_rate": 0.0001576496285371649,
+      "loss": 3.0398,
+      "step": 40241
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.639971137046814,
+      "learning_rate": 0.00015763607615357821,
+      "loss": 3.1722,
+      "step": 40242
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6707262396812439,
+      "learning_rate": 0.00015762252414495073,
+      "loss": 2.9338,
+      "step": 40243
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6912310123443604,
+      "learning_rate": 0.00015760897251131822,
+      "loss": 2.9795,
+      "step": 40244
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6759024858474731,
+      "learning_rate": 0.0001575954212527163,
+      "loss": 3.0455,
+      "step": 40245
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6769490242004395,
+      "learning_rate": 0.00015758187036918054,
+      "loss": 2.9084,
+      "step": 40246
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7240374684333801,
+      "learning_rate": 0.0001575683198607467,
+      "loss": 3.026,
+      "step": 40247
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7014404535293579,
+      "learning_rate": 0.0001575547697274505,
+      "loss": 2.8883,
+      "step": 40248
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6260035634040833,
+      "learning_rate": 0.0001575412199693277,
+      "loss": 2.9092,
+      "step": 40249
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6681224703788757,
+      "learning_rate": 0.00015752767058641376,
+      "loss": 3.019,
+      "step": 40250
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6676934957504272,
+      "learning_rate": 0.00015751412157874467,
+      "loss": 3.0937,
+      "step": 40251
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7086468935012817,
+      "learning_rate": 0.00015750057294635584,
+      "loss": 2.9019,
+      "step": 40252
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6498782634735107,
+      "learning_rate": 0.00015748702468928322,
+      "loss": 2.9223,
+      "step": 40253
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6463537216186523,
+      "learning_rate": 0.0001574734768075623,
+      "loss": 2.8543,
+      "step": 40254
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.66627037525177,
+      "learning_rate": 0.00015745992930122873,
+      "loss": 2.9977,
+      "step": 40255
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6671886444091797,
+      "learning_rate": 0.0001574463821703184,
+      "loss": 3.0417,
+      "step": 40256
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6478363275527954,
+      "learning_rate": 0.00015743283541486673,
+      "loss": 2.8345,
+      "step": 40257
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7030870914459229,
+      "learning_rate": 0.00015741928903490954,
+      "loss": 2.8857,
+      "step": 40258
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6763714551925659,
+      "learning_rate": 0.0001574057430304826,
+      "loss": 3.0909,
+      "step": 40259
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6961510181427002,
+      "learning_rate": 0.0001573921974016215,
+      "loss": 2.933,
+      "step": 40260
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6534474492073059,
+      "learning_rate": 0.00015737865214836193,
+      "loss": 3.1245,
+      "step": 40261
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6820818185806274,
+      "learning_rate": 0.0001573651072707394,
+      "loss": 3.09,
+      "step": 40262
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6295596361160278,
+      "learning_rate": 0.00015735156276878976,
+      "loss": 3.1183,
+      "step": 40263
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6853962540626526,
+      "learning_rate": 0.00015733801864254874,
+      "loss": 3.0549,
+      "step": 40264
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7549799084663391,
+      "learning_rate": 0.0001573244748920518,
+      "loss": 3.0955,
+      "step": 40265
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6778276562690735,
+      "learning_rate": 0.00015731093151733484,
+      "loss": 2.9153,
+      "step": 40266
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6821315884590149,
+      "learning_rate": 0.00015729738851843344,
+      "loss": 3.0041,
+      "step": 40267
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7044394016265869,
+      "learning_rate": 0.00015728384589538315,
+      "loss": 2.8476,
+      "step": 40268
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6207550764083862,
+      "learning_rate": 0.00015727030364821984,
+      "loss": 3.1545,
+      "step": 40269
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6819803714752197,
+      "learning_rate": 0.000157256761776979,
+      "loss": 3.0553,
+      "step": 40270
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6631947159767151,
+      "learning_rate": 0.00015724322028169645,
+      "loss": 2.9441,
+      "step": 40271
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6724246144294739,
+      "learning_rate": 0.0001572296791624077,
+      "loss": 3.1259,
+      "step": 40272
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6821426749229431,
+      "learning_rate": 0.00015721613841914852,
+      "loss": 3.0238,
+      "step": 40273
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6199890971183777,
+      "learning_rate": 0.0001572025980519546,
+      "loss": 3.0075,
+      "step": 40274
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.637385368347168,
+      "learning_rate": 0.00015718905806086158,
+      "loss": 2.9653,
+      "step": 40275
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6946399211883545,
+      "learning_rate": 0.00015717551844590512,
+      "loss": 3.0097,
+      "step": 40276
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6899637579917908,
+      "learning_rate": 0.00015716197920712072,
+      "loss": 2.9472,
+      "step": 40277
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6917797923088074,
+      "learning_rate": 0.00015714844034454432,
+      "loss": 2.9553,
+      "step": 40278
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6806402206420898,
+      "learning_rate": 0.00015713490185821128,
+      "loss": 3.078,
+      "step": 40279
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6591494679450989,
+      "learning_rate": 0.00015712136374815744,
+      "loss": 3.1675,
+      "step": 40280
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.632714033126831,
+      "learning_rate": 0.00015710782601441854,
+      "loss": 2.9648,
+      "step": 40281
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6618544459342957,
+      "learning_rate": 0.00015709428865703012,
+      "loss": 3.0405,
+      "step": 40282
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6480012536048889,
+      "learning_rate": 0.00015708075167602785,
+      "loss": 2.8349,
+      "step": 40283
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6463007926940918,
+      "learning_rate": 0.00015706721507144727,
+      "loss": 3.0346,
+      "step": 40284
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.652336597442627,
+      "learning_rate": 0.00015705367884332413,
+      "loss": 3.1047,
+      "step": 40285
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6886484026908875,
+      "learning_rate": 0.0001570401429916942,
+      "loss": 2.8904,
+      "step": 40286
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.632122814655304,
+      "learning_rate": 0.00015702660751659288,
+      "loss": 3.155,
+      "step": 40287
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7023600339889526,
+      "learning_rate": 0.00015701307241805613,
+      "loss": 2.8844,
+      "step": 40288
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6865475177764893,
+      "learning_rate": 0.00015699953769611938,
+      "loss": 3.0419,
+      "step": 40289
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.676014244556427,
+      "learning_rate": 0.0001569860033508182,
+      "loss": 2.9082,
+      "step": 40290
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6413493752479553,
+      "learning_rate": 0.00015697246938218854,
+      "loss": 2.8572,
+      "step": 40291
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6815982460975647,
+      "learning_rate": 0.00015695893579026576,
+      "loss": 3.0746,
+      "step": 40292
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6308925151824951,
+      "learning_rate": 0.00015694540257508565,
+      "loss": 2.9461,
+      "step": 40293
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7394782304763794,
+      "learning_rate": 0.00015693186973668376,
+      "loss": 2.8717,
+      "step": 40294
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.656547486782074,
+      "learning_rate": 0.00015691833727509577,
+      "loss": 2.8996,
+      "step": 40295
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6732604503631592,
+      "learning_rate": 0.00015690480519035745,
+      "loss": 3.065,
+      "step": 40296
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6596916317939758,
+      "learning_rate": 0.00015689127348250434,
+      "loss": 3.0307,
+      "step": 40297
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7515655159950256,
+      "learning_rate": 0.0001568777421515721,
+      "loss": 2.9068,
+      "step": 40298
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6503599286079407,
+      "learning_rate": 0.0001568642111975962,
+      "loss": 2.922,
+      "step": 40299
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6916700601577759,
+      "learning_rate": 0.00015685068062061243,
+      "loss": 2.9162,
+      "step": 40300
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6582286357879639,
+      "learning_rate": 0.00015683715042065654,
+      "loss": 2.8721,
+      "step": 40301
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6865876317024231,
+      "learning_rate": 0.00015682362059776393,
+      "loss": 3.0664,
+      "step": 40302
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7114717960357666,
+      "learning_rate": 0.00015681009115197048,
+      "loss": 3.087,
+      "step": 40303
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6616129875183105,
+      "learning_rate": 0.00015679656208331165,
+      "loss": 3.0272,
+      "step": 40304
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6594287753105164,
+      "learning_rate": 0.00015678303339182305,
+      "loss": 2.9298,
+      "step": 40305
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7083821892738342,
+      "learning_rate": 0.00015676950507754044,
+      "loss": 3.2278,
+      "step": 40306
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6950168013572693,
+      "learning_rate": 0.00015675597714049925,
+      "loss": 2.9107,
+      "step": 40307
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7014042139053345,
+      "learning_rate": 0.00015674244958073543,
+      "loss": 2.892,
+      "step": 40308
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7078503370285034,
+      "learning_rate": 0.00015672892239828432,
+      "loss": 2.8923,
+      "step": 40309
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6410806179046631,
+      "learning_rate": 0.00015671539559318162,
+      "loss": 3.0353,
+      "step": 40310
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6679813265800476,
+      "learning_rate": 0.00015670186916546318,
+      "loss": 2.9274,
+      "step": 40311
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6931245923042297,
+      "learning_rate": 0.00015668834311516428,
+      "loss": 2.9186,
+      "step": 40312
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6886323094367981,
+      "learning_rate": 0.00015667481744232082,
+      "loss": 3.0251,
+      "step": 40313
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6314588785171509,
+      "learning_rate": 0.00015666129214696815,
+      "loss": 3.0817,
+      "step": 40314
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6671131253242493,
+      "learning_rate": 0.00015664776722914205,
+      "loss": 2.987,
+      "step": 40315
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6521415114402771,
+      "learning_rate": 0.00015663424268887828,
+      "loss": 2.884,
+      "step": 40316
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6604424118995667,
+      "learning_rate": 0.00015662071852621222,
+      "loss": 2.767,
+      "step": 40317
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6475251317024231,
+      "learning_rate": 0.00015660719474117968,
+      "loss": 3.0083,
+      "step": 40318
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6570774912834167,
+      "learning_rate": 0.0001565936713338162,
+      "loss": 3.0535,
+      "step": 40319
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6820668578147888,
+      "learning_rate": 0.00015658014830415728,
+      "loss": 2.9305,
+      "step": 40320
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6360170841217041,
+      "learning_rate": 0.00015656662565223878,
+      "loss": 3.0399,
+      "step": 40321
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6604784727096558,
+      "learning_rate": 0.00015655310337809604,
+      "loss": 3.2472,
+      "step": 40322
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.694609522819519,
+      "learning_rate": 0.00015653958148176492,
+      "loss": 2.9491,
+      "step": 40323
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6777598857879639,
+      "learning_rate": 0.00015652605996328085,
+      "loss": 3.0202,
+      "step": 40324
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6691681742668152,
+      "learning_rate": 0.00015651253882267965,
+      "loss": 3.0187,
+      "step": 40325
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6447973251342773,
+      "learning_rate": 0.0001564990180599968,
+      "loss": 3.2181,
+      "step": 40326
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6347460746765137,
+      "learning_rate": 0.00015648549767526777,
+      "loss": 3.0601,
+      "step": 40327
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6752090454101562,
+      "learning_rate": 0.00015647197766852848,
+      "loss": 2.9086,
+      "step": 40328
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6812394261360168,
+      "learning_rate": 0.00015645845803981425,
+      "loss": 2.9608,
+      "step": 40329
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7588211894035339,
+      "learning_rate": 0.00015644493878916077,
+      "loss": 3.021,
+      "step": 40330
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6755926012992859,
+      "learning_rate": 0.00015643141991660388,
+      "loss": 3.0217,
+      "step": 40331
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6614903211593628,
+      "learning_rate": 0.00015641790142217882,
+      "loss": 3.1655,
+      "step": 40332
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6844372749328613,
+      "learning_rate": 0.00015640438330592158,
+      "loss": 2.9864,
+      "step": 40333
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6596974730491638,
+      "learning_rate": 0.00015639086556786747,
+      "loss": 3.099,
+      "step": 40334
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6417086720466614,
+      "learning_rate": 0.00015637734820805212,
+      "loss": 3.0785,
+      "step": 40335
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6517664194107056,
+      "learning_rate": 0.00015636383122651125,
+      "loss": 2.8524,
+      "step": 40336
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7032948136329651,
+      "learning_rate": 0.00015635031462328034,
+      "loss": 3.0513,
+      "step": 40337
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.666313886642456,
+      "learning_rate": 0.0001563367983983952,
+      "loss": 3.2258,
+      "step": 40338
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6794551610946655,
+      "learning_rate": 0.0001563232825518911,
+      "loss": 3.0646,
+      "step": 40339
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6857566833496094,
+      "learning_rate": 0.00015630976708380396,
+      "loss": 3.1632,
+      "step": 40340
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6492341756820679,
+      "learning_rate": 0.00015629625199416926,
+      "loss": 3.1526,
+      "step": 40341
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6422709226608276,
+      "learning_rate": 0.00015628273728302241,
+      "loss": 3.0216,
+      "step": 40342
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.677402138710022,
+      "learning_rate": 0.00015626922295039936,
+      "loss": 2.9903,
+      "step": 40343
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6274300217628479,
+      "learning_rate": 0.00015625570899633533,
+      "loss": 2.9704,
+      "step": 40344
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6816664934158325,
+      "learning_rate": 0.00015624219542086613,
+      "loss": 3.1336,
+      "step": 40345
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.719647228717804,
+      "learning_rate": 0.00015622868222402747,
+      "loss": 3.0212,
+      "step": 40346
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7077556252479553,
+      "learning_rate": 0.00015621516940585474,
+      "loss": 2.945,
+      "step": 40347
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7448094487190247,
+      "learning_rate": 0.0001562016569663836,
+      "loss": 2.854,
+      "step": 40348
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6531862616539001,
+      "learning_rate": 0.0001561881449056495,
+      "loss": 3.0116,
+      "step": 40349
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6494091749191284,
+      "learning_rate": 0.00015617463322368814,
+      "loss": 2.9924,
+      "step": 40350
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6503402590751648,
+      "learning_rate": 0.00015616112192053526,
+      "loss": 3.1069,
+      "step": 40351
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7476397752761841,
+      "learning_rate": 0.00015614761099622617,
+      "loss": 2.9541,
+      "step": 40352
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6769456267356873,
+      "learning_rate": 0.0001561341004507967,
+      "loss": 2.9051,
+      "step": 40353
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6509860157966614,
+      "learning_rate": 0.00015612059028428224,
+      "loss": 2.997,
+      "step": 40354
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6968262195587158,
+      "learning_rate": 0.0001561070804967185,
+      "loss": 2.9891,
+      "step": 40355
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6519584655761719,
+      "learning_rate": 0.0001560935710881411,
+      "loss": 3.1037,
+      "step": 40356
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6796095967292786,
+      "learning_rate": 0.00015608006205858539,
+      "loss": 2.8397,
+      "step": 40357
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6914886832237244,
+      "learning_rate": 0.00015606655340808718,
+      "loss": 3.1213,
+      "step": 40358
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6436863541603088,
+      "learning_rate": 0.00015605304513668187,
+      "loss": 3.0416,
+      "step": 40359
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7205463647842407,
+      "learning_rate": 0.00015603953724440528,
+      "loss": 3.1417,
+      "step": 40360
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6410875916481018,
+      "learning_rate": 0.00015602602973129273,
+      "loss": 2.9679,
+      "step": 40361
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6640878915786743,
+      "learning_rate": 0.00015601252259737997,
+      "loss": 3.0664,
+      "step": 40362
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6714800000190735,
+      "learning_rate": 0.00015599901584270257,
+      "loss": 2.942,
+      "step": 40363
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6468148231506348,
+      "learning_rate": 0.0001559855094672959,
+      "loss": 2.7303,
+      "step": 40364
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6648983955383301,
+      "learning_rate": 0.00015597200347119581,
+      "loss": 3.0801,
+      "step": 40365
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.784553050994873,
+      "learning_rate": 0.00015595849785443762,
+      "loss": 2.8801,
+      "step": 40366
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6907336711883545,
+      "learning_rate": 0.00015594499261705703,
+      "loss": 2.9111,
+      "step": 40367
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6566920876502991,
+      "learning_rate": 0.00015593148775908973,
+      "loss": 2.9668,
+      "step": 40368
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6529011130332947,
+      "learning_rate": 0.00015591798328057113,
+      "loss": 2.9668,
+      "step": 40369
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6959037184715271,
+      "learning_rate": 0.00015590447918153687,
+      "loss": 3.1346,
+      "step": 40370
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6820028424263,
+      "learning_rate": 0.00015589097546202235,
+      "loss": 2.9913,
+      "step": 40371
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.688714325428009,
+      "learning_rate": 0.0001558774721220633,
+      "loss": 3.0301,
+      "step": 40372
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6635189056396484,
+      "learning_rate": 0.00015586396916169532,
+      "loss": 3.0752,
+      "step": 40373
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6321013569831848,
+      "learning_rate": 0.00015585046658095383,
+      "loss": 3.0512,
+      "step": 40374
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6665242910385132,
+      "learning_rate": 0.00015583696437987459,
+      "loss": 3.0112,
+      "step": 40375
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6853665709495544,
+      "learning_rate": 0.00015582346255849287,
+      "loss": 3.0421,
+      "step": 40376
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6680829524993896,
+      "learning_rate": 0.00015580996111684456,
+      "loss": 3.0632,
+      "step": 40377
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6497752070426941,
+      "learning_rate": 0.00015579646005496505,
+      "loss": 2.9086,
+      "step": 40378
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6857454776763916,
+      "learning_rate": 0.00015578295937288979,
+      "loss": 2.7439,
+      "step": 40379
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6618149280548096,
+      "learning_rate": 0.00015576945907065456,
+      "loss": 2.9654,
+      "step": 40380
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6831642389297485,
+      "learning_rate": 0.00015575595914829474,
+      "loss": 3.0545,
+      "step": 40381
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6822869777679443,
+      "learning_rate": 0.00015574245960584595,
+      "loss": 3.1841,
+      "step": 40382
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6751419305801392,
+      "learning_rate": 0.00015572896044334386,
+      "loss": 2.9026,
+      "step": 40383
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7170292735099792,
+      "learning_rate": 0.00015571546166082397,
+      "loss": 3.2591,
+      "step": 40384
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6834322214126587,
+      "learning_rate": 0.00015570196325832175,
+      "loss": 2.9228,
+      "step": 40385
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.675579309463501,
+      "learning_rate": 0.00015568846523587263,
+      "loss": 3.0748,
+      "step": 40386
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.694878101348877,
+      "learning_rate": 0.00015567496759351238,
+      "loss": 3.1112,
+      "step": 40387
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6923894882202148,
+      "learning_rate": 0.00015566147033127656,
+      "loss": 3.0667,
+      "step": 40388
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7370680570602417,
+      "learning_rate": 0.00015564797344920058,
+      "loss": 3.0021,
+      "step": 40389
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7134600281715393,
+      "learning_rate": 0.00015563447694732016,
+      "loss": 3.0047,
+      "step": 40390
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6903583407402039,
+      "learning_rate": 0.00015562098082567071,
+      "loss": 3.044,
+      "step": 40391
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6875481605529785,
+      "learning_rate": 0.00015560748508428775,
+      "loss": 2.8929,
+      "step": 40392
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6734123826026917,
+      "learning_rate": 0.00015559398972320694,
+      "loss": 2.7966,
+      "step": 40393
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6842463612556458,
+      "learning_rate": 0.0001555804947424637,
+      "loss": 2.912,
+      "step": 40394
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6658815741539001,
+      "learning_rate": 0.00015556700014209376,
+      "loss": 3.0623,
+      "step": 40395
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6950148344039917,
+      "learning_rate": 0.00015555350592213237,
+      "loss": 2.833,
+      "step": 40396
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6506776809692383,
+      "learning_rate": 0.0001555400120826153,
+      "loss": 2.842,
+      "step": 40397
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6653673052787781,
+      "learning_rate": 0.00015552651862357818,
+      "loss": 3.0188,
+      "step": 40398
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6891602873802185,
+      "learning_rate": 0.00015551302554505633,
+      "loss": 2.92,
+      "step": 40399
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6852928400039673,
+      "learning_rate": 0.00015549953284708543,
+      "loss": 3.1038,
+      "step": 40400
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6566252708435059,
+      "learning_rate": 0.00015548604052970078,
+      "loss": 3.0613,
+      "step": 40401
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7153254151344299,
+      "learning_rate": 0.00015547254859293812,
+      "loss": 2.9753,
+      "step": 40402
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6642997860908508,
+      "learning_rate": 0.00015545905703683304,
+      "loss": 2.9404,
+      "step": 40403
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6532772779464722,
+      "learning_rate": 0.0001554455658614209,
+      "loss": 2.8544,
+      "step": 40404
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6967506408691406,
+      "learning_rate": 0.00015543207506673744,
+      "loss": 2.8222,
+      "step": 40405
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6587045788764954,
+      "learning_rate": 0.00015541858465281807,
+      "loss": 2.8928,
+      "step": 40406
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.668938398361206,
+      "learning_rate": 0.00015540509461969818,
+      "loss": 3.1886,
+      "step": 40407
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6699063181877136,
+      "learning_rate": 0.00015539160496741357,
+      "loss": 3.013,
+      "step": 40408
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6921393275260925,
+      "learning_rate": 0.00015537811569599954,
+      "loss": 2.9297,
+      "step": 40409
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7287542223930359,
+      "learning_rate": 0.0001553646268054918,
+      "loss": 3.1772,
+      "step": 40410
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6881520748138428,
+      "learning_rate": 0.0001553511382959257,
+      "loss": 2.9042,
+      "step": 40411
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6746994256973267,
+      "learning_rate": 0.000155337650167337,
+      "loss": 3.0089,
+      "step": 40412
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6475083827972412,
+      "learning_rate": 0.00015532416241976108,
+      "loss": 2.9789,
+      "step": 40413
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7340616583824158,
+      "learning_rate": 0.00015531067505323334,
+      "loss": 2.9834,
+      "step": 40414
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6629500389099121,
+      "learning_rate": 0.00015529718806778955,
+      "loss": 3.1256,
+      "step": 40415
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6916855573654175,
+      "learning_rate": 0.00015528370146346504,
+      "loss": 2.9015,
+      "step": 40416
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6919158697128296,
+      "learning_rate": 0.0001552702152402954,
+      "loss": 3.1538,
+      "step": 40417
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6587050557136536,
+      "learning_rate": 0.00015525672939831623,
+      "loss": 2.9421,
+      "step": 40418
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6316267848014832,
+      "learning_rate": 0.00015524324393756292,
+      "loss": 2.9828,
+      "step": 40419
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6622289419174194,
+      "learning_rate": 0.0001552297588580711,
+      "loss": 2.9457,
+      "step": 40420
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.656522274017334,
+      "learning_rate": 0.0001552162741598763,
+      "loss": 2.9718,
+      "step": 40421
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6395564079284668,
+      "learning_rate": 0.0001552027898430138,
+      "loss": 3.034,
+      "step": 40422
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6881017684936523,
+      "learning_rate": 0.00015518930590751942,
+      "loss": 3.075,
+      "step": 40423
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7126860022544861,
+      "learning_rate": 0.0001551758223534284,
+      "loss": 2.8778,
+      "step": 40424
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6271448135375977,
+      "learning_rate": 0.00015516233918077657,
+      "loss": 3.032,
+      "step": 40425
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6470649838447571,
+      "learning_rate": 0.0001551488563895991,
+      "loss": 2.8064,
+      "step": 40426
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6680401563644409,
+      "learning_rate": 0.0001551353739799318,
+      "loss": 2.9065,
+      "step": 40427
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.716007649898529,
+      "learning_rate": 0.00015512189195181003,
+      "loss": 3.0163,
+      "step": 40428
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6913284063339233,
+      "learning_rate": 0.0001551084103052692,
+      "loss": 3.1034,
+      "step": 40429
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.689062774181366,
+      "learning_rate": 0.00015509492904034505,
+      "loss": 3.1922,
+      "step": 40430
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6667220592498779,
+      "learning_rate": 0.0001550814481570729,
+      "loss": 2.9831,
+      "step": 40431
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.665783166885376,
+      "learning_rate": 0.00015506796765548823,
+      "loss": 2.9346,
+      "step": 40432
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6425278186798096,
+      "learning_rate": 0.00015505448753562682,
+      "loss": 2.8453,
+      "step": 40433
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6473648548126221,
+      "learning_rate": 0.00015504100779752385,
+      "loss": 2.9768,
+      "step": 40434
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6833198070526123,
+      "learning_rate": 0.00015502752844121512,
+      "loss": 2.8831,
+      "step": 40435
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9269675016403198,
+      "learning_rate": 0.00015501404946673597,
+      "loss": 3.0654,
+      "step": 40436
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7202673554420471,
+      "learning_rate": 0.0001550005708741219,
+      "loss": 3.1508,
+      "step": 40437
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6838274598121643,
+      "learning_rate": 0.0001549870926634083,
+      "loss": 2.8832,
+      "step": 40438
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6578678488731384,
+      "learning_rate": 0.0001549736148346308,
+      "loss": 3.1325,
+      "step": 40439
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6588073968887329,
+      "learning_rate": 0.00015496013738782501,
+      "loss": 2.9813,
+      "step": 40440
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6366355419158936,
+      "learning_rate": 0.00015494666032302617,
+      "loss": 2.9525,
+      "step": 40441
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6619384288787842,
+      "learning_rate": 0.00015493318364027007,
+      "loss": 2.9329,
+      "step": 40442
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6706056594848633,
+      "learning_rate": 0.00015491970733959198,
+      "loss": 2.832,
+      "step": 40443
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.638305127620697,
+      "learning_rate": 0.00015490623142102737,
+      "loss": 2.9308,
+      "step": 40444
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6442788243293762,
+      "learning_rate": 0.00015489275588461196,
+      "loss": 3.0546,
+      "step": 40445
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6786911487579346,
+      "learning_rate": 0.00015487928073038094,
+      "loss": 3.0076,
+      "step": 40446
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6867267489433289,
+      "learning_rate": 0.00015486580595837013,
+      "loss": 2.8399,
+      "step": 40447
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6981363296508789,
+      "learning_rate": 0.0001548523315686147,
+      "loss": 2.8704,
+      "step": 40448
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6754472851753235,
+      "learning_rate": 0.00015483885756115045,
+      "loss": 3.1064,
+      "step": 40449
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7413714528083801,
+      "learning_rate": 0.0001548253839360127,
+      "loss": 2.9747,
+      "step": 40450
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6990171670913696,
+      "learning_rate": 0.0001548119106932368,
+      "loss": 2.9777,
+      "step": 40451
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7089271545410156,
+      "learning_rate": 0.00015479843783285856,
+      "loss": 2.9411,
+      "step": 40452
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6640129089355469,
+      "learning_rate": 0.00015478496535491315,
+      "loss": 3.1461,
+      "step": 40453
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6691185832023621,
+      "learning_rate": 0.00015477149325943617,
+      "loss": 3.0974,
+      "step": 40454
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6627618074417114,
+      "learning_rate": 0.00015475802154646327,
+      "loss": 2.9525,
+      "step": 40455
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7101806998252869,
+      "learning_rate": 0.00015474455021602967,
+      "loss": 2.8795,
+      "step": 40456
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6481982469558716,
+      "learning_rate": 0.00015473107926817107,
+      "loss": 3.1016,
+      "step": 40457
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6829726696014404,
+      "learning_rate": 0.00015471760870292285,
+      "loss": 3.1465,
+      "step": 40458
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.708656370639801,
+      "learning_rate": 0.00015470413852032038,
+      "loss": 2.91,
+      "step": 40459
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6740055084228516,
+      "learning_rate": 0.0001546906687203994,
+      "loss": 2.9464,
+      "step": 40460
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6535217761993408,
+      "learning_rate": 0.00015467719930319508,
+      "loss": 3.3205,
+      "step": 40461
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.656358540058136,
+      "learning_rate": 0.00015466373026874313,
+      "loss": 3.0558,
+      "step": 40462
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6297536492347717,
+      "learning_rate": 0.00015465026161707888,
+      "loss": 3.3642,
+      "step": 40463
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6395627856254578,
+      "learning_rate": 0.00015463679334823795,
+      "loss": 2.8474,
+      "step": 40464
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6379209756851196,
+      "learning_rate": 0.00015462332546225578,
+      "loss": 3.1087,
+      "step": 40465
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7360820770263672,
+      "learning_rate": 0.00015460985795916766,
+      "loss": 2.8954,
+      "step": 40466
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6527162790298462,
+      "learning_rate": 0.00015459639083900928,
+      "loss": 3.1293,
+      "step": 40467
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6693764328956604,
+      "learning_rate": 0.0001545829241018159,
+      "loss": 2.9546,
+      "step": 40468
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7010257244110107,
+      "learning_rate": 0.00015456945774762316,
+      "loss": 2.8822,
+      "step": 40469
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6681190133094788,
+      "learning_rate": 0.00015455599177646655,
+      "loss": 2.9926,
+      "step": 40470
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7095162868499756,
+      "learning_rate": 0.0001545425261883815,
+      "loss": 3.2781,
+      "step": 40471
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6761109828948975,
+      "learning_rate": 0.0001545290609834034,
+      "loss": 2.8825,
+      "step": 40472
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6583576202392578,
+      "learning_rate": 0.00015451559616156766,
+      "loss": 2.9249,
+      "step": 40473
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6421191692352295,
+      "learning_rate": 0.0001545021317229099,
+      "loss": 2.8429,
+      "step": 40474
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6621305346488953,
+      "learning_rate": 0.00015448866766746552,
+      "loss": 2.9998,
+      "step": 40475
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.708996057510376,
+      "learning_rate": 0.00015447520399526998,
+      "loss": 2.8448,
+      "step": 40476
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6437864899635315,
+      "learning_rate": 0.00015446174070635882,
+      "loss": 2.9074,
+      "step": 40477
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6887125372886658,
+      "learning_rate": 0.00015444827780076726,
+      "loss": 3.0311,
+      "step": 40478
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6327663660049438,
+      "learning_rate": 0.00015443481527853113,
+      "loss": 3.1015,
+      "step": 40479
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6891899704933167,
+      "learning_rate": 0.00015442135313968564,
+      "loss": 2.9709,
+      "step": 40480
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6350082159042358,
+      "learning_rate": 0.00015440789138426614,
+      "loss": 3.0892,
+      "step": 40481
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6630797386169434,
+      "learning_rate": 0.00015439443001230842,
+      "loss": 3.0736,
+      "step": 40482
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6615853309631348,
+      "learning_rate": 0.0001543809690238476,
+      "loss": 3.1135,
+      "step": 40483
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.69512939453125,
+      "learning_rate": 0.00015436750841891925,
+      "loss": 3.0358,
+      "step": 40484
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7393690943717957,
+      "learning_rate": 0.000154354048197559,
+      "loss": 2.8721,
+      "step": 40485
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6448101997375488,
+      "learning_rate": 0.0001543405883598022,
+      "loss": 2.8879,
+      "step": 40486
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6300211548805237,
+      "learning_rate": 0.00015432712890568422,
+      "loss": 2.9113,
+      "step": 40487
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7020927667617798,
+      "learning_rate": 0.00015431366983524041,
+      "loss": 2.9078,
+      "step": 40488
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.717994213104248,
+      "learning_rate": 0.00015430021114850635,
+      "loss": 2.9872,
+      "step": 40489
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6843698620796204,
+      "learning_rate": 0.00015428675284551765,
+      "loss": 2.9769,
+      "step": 40490
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6766223907470703,
+      "learning_rate": 0.00015427329492630944,
+      "loss": 2.8667,
+      "step": 40491
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6386144161224365,
+      "learning_rate": 0.00015425983739091747,
+      "loss": 2.972,
+      "step": 40492
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6815845966339111,
+      "learning_rate": 0.00015424638023937707,
+      "loss": 3.0052,
+      "step": 40493
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6865285038948059,
+      "learning_rate": 0.00015423292347172348,
+      "loss": 2.8614,
+      "step": 40494
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8084060549736023,
+      "learning_rate": 0.00015421946708799248,
+      "loss": 2.7633,
+      "step": 40495
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6883880496025085,
+      "learning_rate": 0.00015420601108821918,
+      "loss": 3.018,
+      "step": 40496
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6464810967445374,
+      "learning_rate": 0.00015419255547243936,
+      "loss": 2.8539,
+      "step": 40497
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6759091019630432,
+      "learning_rate": 0.00015417910024068815,
+      "loss": 2.9933,
+      "step": 40498
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7044882774353027,
+      "learning_rate": 0.0001541656453930011,
+      "loss": 3.1197,
+      "step": 40499
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6467495560646057,
+      "learning_rate": 0.0001541521909294138,
+      "loss": 2.8981,
+      "step": 40500
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6994920372962952,
+      "learning_rate": 0.00015413873684996156,
+      "loss": 3.0469,
+      "step": 40501
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6590608954429626,
+      "learning_rate": 0.00015412528315467987,
+      "loss": 2.7399,
+      "step": 40502
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6317247748374939,
+      "learning_rate": 0.0001541118298436039,
+      "loss": 2.947,
+      "step": 40503
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7404329180717468,
+      "learning_rate": 0.0001540983769167694,
+      "loss": 2.9812,
+      "step": 40504
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6499910950660706,
+      "learning_rate": 0.00015408492437421175,
+      "loss": 2.9587,
+      "step": 40505
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6785842180252075,
+      "learning_rate": 0.0001540714722159662,
+      "loss": 2.9186,
+      "step": 40506
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7388772368431091,
+      "learning_rate": 0.00015405802044206844,
+      "loss": 2.8404,
+      "step": 40507
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6410542726516724,
+      "learning_rate": 0.00015404456905255377,
+      "loss": 2.897,
+      "step": 40508
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6782169342041016,
+      "learning_rate": 0.0001540311180474575,
+      "loss": 3.1239,
+      "step": 40509
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6322621703147888,
+      "learning_rate": 0.00015401766742681528,
+      "loss": 2.882,
+      "step": 40510
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6843157410621643,
+      "learning_rate": 0.0001540042171906623,
+      "loss": 2.9253,
+      "step": 40511
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6755958795547485,
+      "learning_rate": 0.00015399076733903425,
+      "loss": 2.7574,
+      "step": 40512
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6259028911590576,
+      "learning_rate": 0.00015397731787196632,
+      "loss": 2.7801,
+      "step": 40513
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6746711134910583,
+      "learning_rate": 0.0001539638687894941,
+      "loss": 2.9839,
+      "step": 40514
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.639659583568573,
+      "learning_rate": 0.000153950420091653,
+      "loss": 3.1256,
+      "step": 40515
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6220987439155579,
+      "learning_rate": 0.00015393697177847824,
+      "loss": 2.8792,
+      "step": 40516
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6759223937988281,
+      "learning_rate": 0.0001539235238500055,
+      "loss": 2.9302,
+      "step": 40517
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6941930055618286,
+      "learning_rate": 0.00015391007630627,
+      "loss": 2.8693,
+      "step": 40518
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6453860402107239,
+      "learning_rate": 0.00015389662914730732,
+      "loss": 2.9384,
+      "step": 40519
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6453847885131836,
+      "learning_rate": 0.0001538831823731527,
+      "loss": 2.9717,
+      "step": 40520
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6334241032600403,
+      "learning_rate": 0.00015386973598384166,
+      "loss": 2.8373,
+      "step": 40521
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6756218075752258,
+      "learning_rate": 0.00015385628997940972,
+      "loss": 2.9777,
+      "step": 40522
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6967214345932007,
+      "learning_rate": 0.00015384284435989218,
+      "loss": 2.9146,
+      "step": 40523
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.692237138748169,
+      "learning_rate": 0.00015382939912532447,
+      "loss": 2.9344,
+      "step": 40524
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6731816530227661,
+      "learning_rate": 0.00015381595427574189,
+      "loss": 3.0359,
+      "step": 40525
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6681119203567505,
+      "learning_rate": 0.00015380250981117993,
+      "loss": 2.9214,
+      "step": 40526
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6464305520057678,
+      "learning_rate": 0.00015378906573167413,
+      "loss": 2.9634,
+      "step": 40527
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6641029119491577,
+      "learning_rate": 0.0001537756220372597,
+      "loss": 3.0965,
+      "step": 40528
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6519691944122314,
+      "learning_rate": 0.00015376217872797228,
+      "loss": 3.0628,
+      "step": 40529
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7187452912330627,
+      "learning_rate": 0.00015374873580384707,
+      "loss": 3.1994,
+      "step": 40530
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.666256308555603,
+      "learning_rate": 0.0001537352932649195,
+      "loss": 2.9878,
+      "step": 40531
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6944901347160339,
+      "learning_rate": 0.0001537218511112251,
+      "loss": 3.0209,
+      "step": 40532
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7129674553871155,
+      "learning_rate": 0.00015370840934279906,
+      "loss": 3.0064,
+      "step": 40533
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6722531914710999,
+      "learning_rate": 0.00015369496795967705,
+      "loss": 2.8155,
+      "step": 40534
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6533281803131104,
+      "learning_rate": 0.0001536815269618942,
+      "loss": 2.9369,
+      "step": 40535
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6529192328453064,
+      "learning_rate": 0.00015366808634948605,
+      "loss": 2.8809,
+      "step": 40536
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6997420787811279,
+      "learning_rate": 0.00015365464612248825,
+      "loss": 3.0539,
+      "step": 40537
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7066969275474548,
+      "learning_rate": 0.00015364120628093573,
+      "loss": 2.9932,
+      "step": 40538
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6391503810882568,
+      "learning_rate": 0.0001536277668248642,
+      "loss": 2.9198,
+      "step": 40539
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6769759058952332,
+      "learning_rate": 0.00015361432775430885,
+      "loss": 2.9239,
+      "step": 40540
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6748432517051697,
+      "learning_rate": 0.0001536008890693052,
+      "loss": 3.0284,
+      "step": 40541
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7206935882568359,
+      "learning_rate": 0.00015358745076988873,
+      "loss": 2.8153,
+      "step": 40542
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6426916122436523,
+      "learning_rate": 0.0001535740128560946,
+      "loss": 2.8724,
+      "step": 40543
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7100978493690491,
+      "learning_rate": 0.00015356057532795855,
+      "loss": 2.7467,
+      "step": 40544
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6432538032531738,
+      "learning_rate": 0.00015354713818551568,
+      "loss": 2.8214,
+      "step": 40545
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6779876351356506,
+      "learning_rate": 0.00015353370142880132,
+      "loss": 2.8472,
+      "step": 40546
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6484413743019104,
+      "learning_rate": 0.0001535202650578512,
+      "loss": 2.9393,
+      "step": 40547
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6245802640914917,
+      "learning_rate": 0.00015350682907270033,
+      "loss": 3.1935,
+      "step": 40548
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6747826337814331,
+      "learning_rate": 0.00015349339347338442,
+      "loss": 3.0041,
+      "step": 40549
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6607273817062378,
+      "learning_rate": 0.0001534799582599386,
+      "loss": 2.8044,
+      "step": 40550
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6659206748008728,
+      "learning_rate": 0.0001534665234323985,
+      "loss": 2.9159,
+      "step": 40551
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6552886962890625,
+      "learning_rate": 0.00015345308899079936,
+      "loss": 2.8379,
+      "step": 40552
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6883482933044434,
+      "learning_rate": 0.0001534396549351765,
+      "loss": 2.922,
+      "step": 40553
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6916592121124268,
+      "learning_rate": 0.00015342622126556546,
+      "loss": 2.9854,
+      "step": 40554
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6768868565559387,
+      "learning_rate": 0.00015341278798200141,
+      "loss": 2.969,
+      "step": 40555
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6460238695144653,
+      "learning_rate": 0.0001533993550845199,
+      "loss": 3.0113,
+      "step": 40556
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7118356823921204,
+      "learning_rate": 0.00015338592257315638,
+      "loss": 2.9523,
+      "step": 40557
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6956257224082947,
+      "learning_rate": 0.00015337249044794602,
+      "loss": 2.9509,
+      "step": 40558
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.702347457408905,
+      "learning_rate": 0.0001533590587089244,
+      "loss": 2.9354,
+      "step": 40559
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6793229579925537,
+      "learning_rate": 0.0001533456273561268,
+      "loss": 2.9578,
+      "step": 40560
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.738370418548584,
+      "learning_rate": 0.00015333219638958849,
+      "loss": 3.0186,
+      "step": 40561
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7036491632461548,
+      "learning_rate": 0.00015331876580934505,
+      "loss": 3.0268,
+      "step": 40562
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7223859429359436,
+      "learning_rate": 0.00015330533561543164,
+      "loss": 2.8532,
+      "step": 40563
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6854945421218872,
+      "learning_rate": 0.00015329190580788387,
+      "loss": 3.0173,
+      "step": 40564
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6289524435997009,
+      "learning_rate": 0.00015327847638673688,
+      "loss": 2.9122,
+      "step": 40565
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7152692675590515,
+      "learning_rate": 0.00015326504735202622,
+      "loss": 2.9671,
+      "step": 40566
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6731276512145996,
+      "learning_rate": 0.0001532516187037872,
+      "loss": 3.0716,
+      "step": 40567
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6647647619247437,
+      "learning_rate": 0.00015323819044205507,
+      "loss": 2.9302,
+      "step": 40568
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7299318909645081,
+      "learning_rate": 0.0001532247625668654,
+      "loss": 2.6701,
+      "step": 40569
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7116023302078247,
+      "learning_rate": 0.0001532113350782534,
+      "loss": 2.8473,
+      "step": 40570
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6508423089981079,
+      "learning_rate": 0.0001531979079762544,
+      "loss": 3.0764,
+      "step": 40571
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6736328601837158,
+      "learning_rate": 0.00015318448126090407,
+      "loss": 3.047,
+      "step": 40572
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6313000917434692,
+      "learning_rate": 0.00015317105493223746,
+      "loss": 2.8944,
+      "step": 40573
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6281668543815613,
+      "learning_rate": 0.0001531576289902901,
+      "loss": 2.7681,
+      "step": 40574
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6827503442764282,
+      "learning_rate": 0.00015314420343509716,
+      "loss": 3.1714,
+      "step": 40575
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6493374109268188,
+      "learning_rate": 0.00015313077826669414,
+      "loss": 3.0646,
+      "step": 40576
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.702488899230957,
+      "learning_rate": 0.00015311735348511648,
+      "loss": 2.9156,
+      "step": 40577
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6838111281394958,
+      "learning_rate": 0.00015310392909039932,
+      "loss": 2.7314,
+      "step": 40578
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7085728645324707,
+      "learning_rate": 0.0001530905050825783,
+      "loss": 2.9128,
+      "step": 40579
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6542535424232483,
+      "learning_rate": 0.0001530770814616884,
+      "loss": 3.0594,
+      "step": 40580
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6909346580505371,
+      "learning_rate": 0.0001530636582277654,
+      "loss": 2.9486,
+      "step": 40581
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6850855350494385,
+      "learning_rate": 0.0001530502353808444,
+      "loss": 3.1318,
+      "step": 40582
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.633678138256073,
+      "learning_rate": 0.0001530368129209607,
+      "loss": 3.0352,
+      "step": 40583
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6679724454879761,
+      "learning_rate": 0.0001530233908481499,
+      "loss": 3.1599,
+      "step": 40584
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7710044384002686,
+      "learning_rate": 0.00015300996916244703,
+      "loss": 3.0995,
+      "step": 40585
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6416032314300537,
+      "learning_rate": 0.00015299654786388766,
+      "loss": 2.9725,
+      "step": 40586
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6761838793754578,
+      "learning_rate": 0.0001529831269525072,
+      "loss": 3.2538,
+      "step": 40587
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6844536066055298,
+      "learning_rate": 0.00015296970642834088,
+      "loss": 2.9308,
+      "step": 40588
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6797305941581726,
+      "learning_rate": 0.000152956286291424,
+      "loss": 2.9528,
+      "step": 40589
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6660833954811096,
+      "learning_rate": 0.00015294286654179192,
+      "loss": 3.0079,
+      "step": 40590
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6612690091133118,
+      "learning_rate": 0.0001529294471794801,
+      "loss": 3.0193,
+      "step": 40591
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6777284145355225,
+      "learning_rate": 0.00015291602820452374,
+      "loss": 3.0346,
+      "step": 40592
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.678779125213623,
+      "learning_rate": 0.00015290260961695824,
+      "loss": 2.9215,
+      "step": 40593
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6552702784538269,
+      "learning_rate": 0.00015288919141681905,
+      "loss": 3.193,
+      "step": 40594
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.66253662109375,
+      "learning_rate": 0.00015287577360414146,
+      "loss": 2.8081,
+      "step": 40595
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6897073984146118,
+      "learning_rate": 0.00015286235617896073,
+      "loss": 3.0626,
+      "step": 40596
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6485527753829956,
+      "learning_rate": 0.0001528489391413121,
+      "loss": 3.1438,
+      "step": 40597
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6479106545448303,
+      "learning_rate": 0.00015283552249123112,
+      "loss": 3.1592,
+      "step": 40598
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6787962317466736,
+      "learning_rate": 0.00015282210622875308,
+      "loss": 3.2376,
+      "step": 40599
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.656581699848175,
+      "learning_rate": 0.00015280869035391324,
+      "loss": 3.0471,
+      "step": 40600
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6656479835510254,
+      "learning_rate": 0.00015279527486674707,
+      "loss": 2.8886,
+      "step": 40601
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6412723660469055,
+      "learning_rate": 0.00015278185976728972,
+      "loss": 3.0208,
+      "step": 40602
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7223705649375916,
+      "learning_rate": 0.0001527684450555767,
+      "loss": 3.1058,
+      "step": 40603
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7059385180473328,
+      "learning_rate": 0.00015275503073164327,
+      "loss": 3.0413,
+      "step": 40604
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7070484161376953,
+      "learning_rate": 0.00015274161679552464,
+      "loss": 3.0564,
+      "step": 40605
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6791709661483765,
+      "learning_rate": 0.00015272820324725635,
+      "loss": 2.8612,
+      "step": 40606
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6438567638397217,
+      "learning_rate": 0.00015271479008687359,
+      "loss": 2.9881,
+      "step": 40607
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.631389856338501,
+      "learning_rate": 0.00015270137731441164,
+      "loss": 3.0258,
+      "step": 40608
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7261651754379272,
+      "learning_rate": 0.00015268796492990603,
+      "loss": 2.9647,
+      "step": 40609
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6476674675941467,
+      "learning_rate": 0.00015267455293339205,
+      "loss": 2.9719,
+      "step": 40610
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6464186310768127,
+      "learning_rate": 0.00015266114132490488,
+      "loss": 3.1635,
+      "step": 40611
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6697189211845398,
+      "learning_rate": 0.0001526477301044798,
+      "loss": 2.9862,
+      "step": 40612
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6742683053016663,
+      "learning_rate": 0.00015263431927215224,
+      "loss": 2.984,
+      "step": 40613
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6472331881523132,
+      "learning_rate": 0.00015262090882795765,
+      "loss": 3.1692,
+      "step": 40614
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6672264337539673,
+      "learning_rate": 0.0001526074987719311,
+      "loss": 2.9423,
+      "step": 40615
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6128020286560059,
+      "learning_rate": 0.00015259408910410816,
+      "loss": 3.0283,
+      "step": 40616
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.665045976638794,
+      "learning_rate": 0.00015258067982452402,
+      "loss": 2.9884,
+      "step": 40617
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6567105650901794,
+      "learning_rate": 0.00015256727093321387,
+      "loss": 3.0353,
+      "step": 40618
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6515582203865051,
+      "learning_rate": 0.00015255386243021328,
+      "loss": 2.936,
+      "step": 40619
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6716775298118591,
+      "learning_rate": 0.0001525404543155573,
+      "loss": 2.9665,
+      "step": 40620
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.651340901851654,
+      "learning_rate": 0.00015252704658928152,
+      "loss": 3.1269,
+      "step": 40621
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6484601497650146,
+      "learning_rate": 0.00015251363925142098,
+      "loss": 3.0912,
+      "step": 40622
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6475279331207275,
+      "learning_rate": 0.00015250023230201114,
+      "loss": 2.8503,
+      "step": 40623
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.653603732585907,
+      "learning_rate": 0.00015248682574108745,
+      "loss": 2.9839,
+      "step": 40624
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6498973965644836,
+      "learning_rate": 0.00015247341956868505,
+      "loss": 2.8576,
+      "step": 40625
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6715899705886841,
+      "learning_rate": 0.00015246001378483925,
+      "loss": 2.9632,
+      "step": 40626
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6394970417022705,
+      "learning_rate": 0.00015244660838958528,
+      "loss": 2.8658,
+      "step": 40627
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6495617032051086,
+      "learning_rate": 0.00015243320338295856,
+      "loss": 2.9026,
+      "step": 40628
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.710526168346405,
+      "learning_rate": 0.00015241979876499446,
+      "loss": 3.0276,
+      "step": 40629
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6743832230567932,
+      "learning_rate": 0.00015240639453572812,
+      "loss": 2.8014,
+      "step": 40630
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6878013610839844,
+      "learning_rate": 0.00015239299069519504,
+      "loss": 2.8822,
+      "step": 40631
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6774446368217468,
+      "learning_rate": 0.00015237958724343044,
+      "loss": 2.9242,
+      "step": 40632
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6619464159011841,
+      "learning_rate": 0.00015236618418046943,
+      "loss": 3.0985,
+      "step": 40633
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7469950914382935,
+      "learning_rate": 0.00015235278150634764,
+      "loss": 3.0865,
+      "step": 40634
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6729745864868164,
+      "learning_rate": 0.0001523393792211001,
+      "loss": 2.9876,
+      "step": 40635
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.665226936340332,
+      "learning_rate": 0.00015232597732476227,
+      "loss": 2.9907,
+      "step": 40636
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6587421298027039,
+      "learning_rate": 0.00015231257581736933,
+      "loss": 3.2124,
+      "step": 40637
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6516327261924744,
+      "learning_rate": 0.00015229917469895664,
+      "loss": 3.0794,
+      "step": 40638
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.108006477355957,
+      "learning_rate": 0.00015228577396955974,
+      "loss": 2.8191,
+      "step": 40639
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7497507929801941,
+      "learning_rate": 0.0001522723736292134,
+      "loss": 2.9436,
+      "step": 40640
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6455642580986023,
+      "learning_rate": 0.0001522589736779534,
+      "loss": 3.1215,
+      "step": 40641
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6802120804786682,
+      "learning_rate": 0.00015224557411581464,
+      "loss": 2.6826,
+      "step": 40642
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6976883411407471,
+      "learning_rate": 0.0001522321749428326,
+      "loss": 3.0047,
+      "step": 40643
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6384730339050293,
+      "learning_rate": 0.0001522187761590427,
+      "loss": 2.865,
+      "step": 40644
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.635405957698822,
+      "learning_rate": 0.00015220537776448,
+      "loss": 2.9101,
+      "step": 40645
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6480907797813416,
+      "learning_rate": 0.00015219197975918002,
+      "loss": 3.0178,
+      "step": 40646
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6745615005493164,
+      "learning_rate": 0.00015217858214317792,
+      "loss": 2.8982,
+      "step": 40647
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.658542811870575,
+      "learning_rate": 0.0001521651849165088,
+      "loss": 3.2529,
+      "step": 40648
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6688123941421509,
+      "learning_rate": 0.00015215178807920834,
+      "loss": 3.1468,
+      "step": 40649
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6520625352859497,
+      "learning_rate": 0.00015213839163131146,
+      "loss": 3.004,
+      "step": 40650
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6656911969184875,
+      "learning_rate": 0.00015212499557285374,
+      "loss": 3.0078,
+      "step": 40651
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6434037685394287,
+      "learning_rate": 0.00015211159990387015,
+      "loss": 3.0344,
+      "step": 40652
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6860767006874084,
+      "learning_rate": 0.0001520982046243963,
+      "loss": 3.1034,
+      "step": 40653
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6536909341812134,
+      "learning_rate": 0.00015208480973446733,
+      "loss": 3.039,
+      "step": 40654
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6466705203056335,
+      "learning_rate": 0.00015207141523411834,
+      "loss": 3.0428,
+      "step": 40655
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.65459805727005,
+      "learning_rate": 0.00015205802112338492,
+      "loss": 2.9133,
+      "step": 40656
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6901729106903076,
+      "learning_rate": 0.0001520446274023021,
+      "loss": 3.2814,
+      "step": 40657
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6484500765800476,
+      "learning_rate": 0.00015203123407090522,
+      "loss": 2.9905,
+      "step": 40658
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6647442579269409,
+      "learning_rate": 0.00015201784112922975,
+      "loss": 3.1639,
+      "step": 40659
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6574181914329529,
+      "learning_rate": 0.00015200444857731062,
+      "loss": 2.8888,
+      "step": 40660
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6742486953735352,
+      "learning_rate": 0.0001519910564151835,
+      "loss": 2.9154,
+      "step": 40661
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6795637607574463,
+      "learning_rate": 0.00015197766464288337,
+      "loss": 2.9062,
+      "step": 40662
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.642927885055542,
+      "learning_rate": 0.00015196427326044553,
+      "loss": 3.0751,
+      "step": 40663
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6584546566009521,
+      "learning_rate": 0.00015195088226790536,
+      "loss": 3.1254,
+      "step": 40664
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6724053621292114,
+      "learning_rate": 0.00015193749166529804,
+      "loss": 3.3392,
+      "step": 40665
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6559529900550842,
+      "learning_rate": 0.00015192410145265891,
+      "loss": 3.0559,
+      "step": 40666
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6671188473701477,
+      "learning_rate": 0.00015191071163002308,
+      "loss": 3.1581,
+      "step": 40667
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6583215594291687,
+      "learning_rate": 0.0001518973221974261,
+      "loss": 2.9865,
+      "step": 40668
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7388840317726135,
+      "learning_rate": 0.00015188393315490308,
+      "loss": 2.8475,
+      "step": 40669
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6607951521873474,
+      "learning_rate": 0.0001518705445024891,
+      "loss": 3.153,
+      "step": 40670
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6548836827278137,
+      "learning_rate": 0.00015185715624021977,
+      "loss": 2.8508,
+      "step": 40671
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.741969883441925,
+      "learning_rate": 0.00015184376836813003,
+      "loss": 3.0824,
+      "step": 40672
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6918534636497498,
+      "learning_rate": 0.0001518303808862554,
+      "loss": 2.9278,
+      "step": 40673
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.628976583480835,
+      "learning_rate": 0.00015181699379463094,
+      "loss": 2.9431,
+      "step": 40674
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6541244983673096,
+      "learning_rate": 0.0001518036070932921,
+      "loss": 2.7107,
+      "step": 40675
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6655003428459167,
+      "learning_rate": 0.00015179022078227406,
+      "loss": 3.0281,
+      "step": 40676
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6962637901306152,
+      "learning_rate": 0.00015177683486161195,
+      "loss": 3.0287,
+      "step": 40677
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6648634672164917,
+      "learning_rate": 0.00015176344933134117,
+      "loss": 3.0775,
+      "step": 40678
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6377243995666504,
+      "learning_rate": 0.0001517500641914969,
+      "loss": 2.7844,
+      "step": 40679
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6637604236602783,
+      "learning_rate": 0.0001517366794421144,
+      "loss": 3.2034,
+      "step": 40680
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6475324630737305,
+      "learning_rate": 0.00015172329508322908,
+      "loss": 3.0258,
+      "step": 40681
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7287402749061584,
+      "learning_rate": 0.0001517099111148759,
+      "loss": 2.7741,
+      "step": 40682
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6636556386947632,
+      "learning_rate": 0.00015169652753709045,
+      "loss": 2.9621,
+      "step": 40683
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6482304334640503,
+      "learning_rate": 0.0001516831443499078,
+      "loss": 2.9935,
+      "step": 40684
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6574996709823608,
+      "learning_rate": 0.00015166976155336304,
+      "loss": 2.9147,
+      "step": 40685
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6734505891799927,
+      "learning_rate": 0.0001516563791474917,
+      "loss": 3.0778,
+      "step": 40686
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6687676906585693,
+      "learning_rate": 0.00015164299713232886,
+      "loss": 2.9375,
+      "step": 40687
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6934749484062195,
+      "learning_rate": 0.00015162961550790988,
+      "loss": 3.0035,
+      "step": 40688
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7008736729621887,
+      "learning_rate": 0.00015161623427426983,
+      "loss": 2.7388,
+      "step": 40689
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6899448037147522,
+      "learning_rate": 0.00015160285343144417,
+      "loss": 2.8754,
+      "step": 40690
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6953763961791992,
+      "learning_rate": 0.00015158947297946805,
+      "loss": 3.1054,
+      "step": 40691
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7021351456642151,
+      "learning_rate": 0.00015157609291837658,
+      "loss": 2.9845,
+      "step": 40692
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.657197892665863,
+      "learning_rate": 0.00015156271324820523,
+      "loss": 2.8504,
+      "step": 40693
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6748011708259583,
+      "learning_rate": 0.000151549333968989,
+      "loss": 2.7495,
+      "step": 40694
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6754131317138672,
+      "learning_rate": 0.00015153595508076326,
+      "loss": 2.8573,
+      "step": 40695
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6298975944519043,
+      "learning_rate": 0.00015152257658356336,
+      "loss": 2.8821,
+      "step": 40696
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6881123185157776,
+      "learning_rate": 0.0001515091984774244,
+      "loss": 2.8627,
+      "step": 40697
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6599323153495789,
+      "learning_rate": 0.00015149582076238164,
+      "loss": 2.8482,
+      "step": 40698
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.640341579914093,
+      "learning_rate": 0.00015148244343847025,
+      "loss": 3.0448,
+      "step": 40699
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7159940004348755,
+      "learning_rate": 0.00015146906650572546,
+      "loss": 3.1676,
+      "step": 40700
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6907761096954346,
+      "learning_rate": 0.00015145568996418268,
+      "loss": 2.9688,
+      "step": 40701
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6505234837532043,
+      "learning_rate": 0.00015144231381387693,
+      "loss": 2.9647,
+      "step": 40702
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6479827761650085,
+      "learning_rate": 0.00015142893805484363,
+      "loss": 3.0348,
+      "step": 40703
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7323071956634521,
+      "learning_rate": 0.00015141556268711783,
+      "loss": 2.7454,
+      "step": 40704
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6864801049232483,
+      "learning_rate": 0.00015140218771073495,
+      "loss": 2.9794,
+      "step": 40705
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.675094485282898,
+      "learning_rate": 0.00015138881312573007,
+      "loss": 2.9768,
+      "step": 40706
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6760757565498352,
+      "learning_rate": 0.00015137543893213836,
+      "loss": 2.6509,
+      "step": 40707
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6964802145957947,
+      "learning_rate": 0.00015136206512999527,
+      "loss": 2.8603,
+      "step": 40708
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6976985931396484,
+      "learning_rate": 0.00015134869171933577,
+      "loss": 2.8392,
+      "step": 40709
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6337934136390686,
+      "learning_rate": 0.00015133531870019525,
+      "loss": 3.0877,
+      "step": 40710
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6604355573654175,
+      "learning_rate": 0.00015132194607260892,
+      "loss": 2.9733,
+      "step": 40711
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6681331992149353,
+      "learning_rate": 0.00015130857383661203,
+      "loss": 2.9892,
+      "step": 40712
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7403008937835693,
+      "learning_rate": 0.00015129520199223974,
+      "loss": 2.8967,
+      "step": 40713
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6657782196998596,
+      "learning_rate": 0.00015128183053952714,
+      "loss": 2.9276,
+      "step": 40714
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6958048343658447,
+      "learning_rate": 0.00015126845947850955,
+      "loss": 2.9373,
+      "step": 40715
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6533171534538269,
+      "learning_rate": 0.00015125508880922236,
+      "loss": 3.2072,
+      "step": 40716
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6738389730453491,
+      "learning_rate": 0.00015124171853170054,
+      "loss": 3.0335,
+      "step": 40717
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6788296699523926,
+      "learning_rate": 0.0001512283486459795,
+      "loss": 3.0401,
+      "step": 40718
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6417179703712463,
+      "learning_rate": 0.00015121497915209436,
+      "loss": 2.66,
+      "step": 40719
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6341338157653809,
+      "learning_rate": 0.00015120161005008017,
+      "loss": 3.0029,
+      "step": 40720
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6832126975059509,
+      "learning_rate": 0.0001511882413399725,
+      "loss": 2.7799,
+      "step": 40721
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7208033800125122,
+      "learning_rate": 0.0001511748730218062,
+      "loss": 2.8199,
+      "step": 40722
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7194510698318481,
+      "learning_rate": 0.00015116150509561675,
+      "loss": 3.077,
+      "step": 40723
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.665046751499176,
+      "learning_rate": 0.00015114813756143917,
+      "loss": 3.0192,
+      "step": 40724
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6552749872207642,
+      "learning_rate": 0.0001511347704193087,
+      "loss": 3.1042,
+      "step": 40725
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7193647623062134,
+      "learning_rate": 0.00015112140366926074,
+      "loss": 3.11,
+      "step": 40726
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.718508243560791,
+      "learning_rate": 0.00015110803731133038,
+      "loss": 2.9426,
+      "step": 40727
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6804858446121216,
+      "learning_rate": 0.00015109467134555277,
+      "loss": 3.0969,
+      "step": 40728
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6869458556175232,
+      "learning_rate": 0.00015108130577196302,
+      "loss": 3.13,
+      "step": 40729
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.669521152973175,
+      "learning_rate": 0.00015106794059059647,
+      "loss": 2.9521,
+      "step": 40730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6213893294334412,
+      "learning_rate": 0.00015105457580148838,
+      "loss": 3.031,
+      "step": 40731
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6913130879402161,
+      "learning_rate": 0.00015104121140467378,
+      "loss": 2.8968,
+      "step": 40732
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.676311731338501,
+      "learning_rate": 0.00015102784740018803,
+      "loss": 3.1907,
+      "step": 40733
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6928296685218811,
+      "learning_rate": 0.00015101448378806632,
+      "loss": 2.9377,
+      "step": 40734
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6946173906326294,
+      "learning_rate": 0.00015100112056834367,
+      "loss": 3.0254,
+      "step": 40735
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6546354293823242,
+      "learning_rate": 0.0001509877577410555,
+      "loss": 3.2543,
+      "step": 40736
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.707718551158905,
+      "learning_rate": 0.00015097439530623675,
+      "loss": 2.915,
+      "step": 40737
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6598696708679199,
+      "learning_rate": 0.00015096103326392291,
+      "loss": 2.7935,
+      "step": 40738
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.626047670841217,
+      "learning_rate": 0.0001509476716141489,
+      "loss": 3.126,
+      "step": 40739
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6716310977935791,
+      "learning_rate": 0.00015093431035695013,
+      "loss": 3.1383,
+      "step": 40740
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6620728969573975,
+      "learning_rate": 0.0001509209494923617,
+      "loss": 3.0502,
+      "step": 40741
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6665850877761841,
+      "learning_rate": 0.00015090758902041874,
+      "loss": 3.2004,
+      "step": 40742
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6744237542152405,
+      "learning_rate": 0.00015089422894115656,
+      "loss": 3.0869,
+      "step": 40743
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6411026120185852,
+      "learning_rate": 0.00015088086925461018,
+      "loss": 3.0758,
+      "step": 40744
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6751980781555176,
+      "learning_rate": 0.00015086750996081493,
+      "loss": 2.9283,
+      "step": 40745
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6658977270126343,
+      "learning_rate": 0.000150854151059806,
+      "loss": 3.2339,
+      "step": 40746
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6690980195999146,
+      "learning_rate": 0.0001508407925516185,
+      "loss": 2.8654,
+      "step": 40747
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7223593592643738,
+      "learning_rate": 0.0001508274344362877,
+      "loss": 2.8638,
+      "step": 40748
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.673263669013977,
+      "learning_rate": 0.00015081407671384874,
+      "loss": 2.9679,
+      "step": 40749
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.706080436706543,
+      "learning_rate": 0.0001508007193843368,
+      "loss": 3.065,
+      "step": 40750
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6620662212371826,
+      "learning_rate": 0.00015078736244778692,
+      "loss": 2.858,
+      "step": 40751
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6945991516113281,
+      "learning_rate": 0.00015077400590423443,
+      "loss": 2.972,
+      "step": 40752
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6781968474388123,
+      "learning_rate": 0.0001507606497537146,
+      "loss": 2.8857,
+      "step": 40753
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6542848348617554,
+      "learning_rate": 0.00015074729399626243,
+      "loss": 2.7439,
+      "step": 40754
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6352688670158386,
+      "learning_rate": 0.0001507339386319132,
+      "loss": 2.9773,
+      "step": 40755
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6776319742202759,
+      "learning_rate": 0.00015072058366070207,
+      "loss": 2.7851,
+      "step": 40756
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6314222812652588,
+      "learning_rate": 0.00015070722908266409,
+      "loss": 2.9906,
+      "step": 40757
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6974877715110779,
+      "learning_rate": 0.00015069387489783466,
+      "loss": 2.8347,
+      "step": 40758
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6645646691322327,
+      "learning_rate": 0.0001506805211062487,
+      "loss": 2.8755,
+      "step": 40759
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6492670774459839,
+      "learning_rate": 0.00015066716770794164,
+      "loss": 2.9298,
+      "step": 40760
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6646486520767212,
+      "learning_rate": 0.00015065381470294838,
+      "loss": 3.1262,
+      "step": 40761
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6600433588027954,
+      "learning_rate": 0.00015064046209130425,
+      "loss": 3.1513,
+      "step": 40762
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6794731616973877,
+      "learning_rate": 0.0001506271098730445,
+      "loss": 2.994,
+      "step": 40763
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6728107929229736,
+      "learning_rate": 0.00015061375804820422,
+      "loss": 3.0659,
+      "step": 40764
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6461384296417236,
+      "learning_rate": 0.00015060040661681853,
+      "loss": 2.8795,
+      "step": 40765
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6834413409233093,
+      "learning_rate": 0.0001505870555789225,
+      "loss": 2.884,
+      "step": 40766
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6609336733818054,
+      "learning_rate": 0.00015057370493455142,
+      "loss": 2.9974,
+      "step": 40767
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6304715275764465,
+      "learning_rate": 0.00015056035468374054,
+      "loss": 2.9703,
+      "step": 40768
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6766150593757629,
+      "learning_rate": 0.00015054700482652482,
+      "loss": 2.9168,
+      "step": 40769
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6861264705657959,
+      "learning_rate": 0.00015053365536293967,
+      "loss": 2.6705,
+      "step": 40770
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7163991928100586,
+      "learning_rate": 0.00015052030629302007,
+      "loss": 2.8756,
+      "step": 40771
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6539211869239807,
+      "learning_rate": 0.00015050695761680108,
+      "loss": 2.8508,
+      "step": 40772
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6503493785858154,
+      "learning_rate": 0.00015049360933431816,
+      "loss": 2.8387,
+      "step": 40773
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6681285500526428,
+      "learning_rate": 0.00015048026144560617,
+      "loss": 3.1689,
+      "step": 40774
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6761608719825745,
+      "learning_rate": 0.0001504669139507005,
+      "loss": 2.8518,
+      "step": 40775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7086718082427979,
+      "learning_rate": 0.00015045356684963612,
+      "loss": 3.0821,
+      "step": 40776
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6684795618057251,
+      "learning_rate": 0.00015044022014244835,
+      "loss": 3.0127,
+      "step": 40777
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6545731425285339,
+      "learning_rate": 0.00015042687382917224,
+      "loss": 3.1053,
+      "step": 40778
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7030220627784729,
+      "learning_rate": 0.00015041352790984286,
+      "loss": 2.9971,
+      "step": 40779
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6445237994194031,
+      "learning_rate": 0.00015040018238449558,
+      "loss": 2.959,
+      "step": 40780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.702018678188324,
+      "learning_rate": 0.0001503868372531653,
+      "loss": 2.873,
+      "step": 40781
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6564288139343262,
+      "learning_rate": 0.00015037349251588732,
+      "loss": 2.7837,
+      "step": 40782
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.732306957244873,
+      "learning_rate": 0.00015036014817269687,
+      "loss": 3.0326,
+      "step": 40783
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6501794457435608,
+      "learning_rate": 0.00015034680422362884,
+      "loss": 2.9075,
+      "step": 40784
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6539938449859619,
+      "learning_rate": 0.00015033346066871866,
+      "loss": 2.8234,
+      "step": 40785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6749308705329895,
+      "learning_rate": 0.0001503201175080014,
+      "loss": 3.2814,
+      "step": 40786
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6837059855461121,
+      "learning_rate": 0.00015030677474151195,
+      "loss": 2.97,
+      "step": 40787
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6461108922958374,
+      "learning_rate": 0.00015029343236928582,
+      "loss": 3.1164,
+      "step": 40788
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6855670213699341,
+      "learning_rate": 0.00015028009039135781,
+      "loss": 2.8901,
+      "step": 40789
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.683847188949585,
+      "learning_rate": 0.00015026674880776338,
+      "loss": 2.9545,
+      "step": 40790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6807911396026611,
+      "learning_rate": 0.00015025340761853738,
+      "loss": 2.9702,
+      "step": 40791
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.689594566822052,
+      "learning_rate": 0.00015024006682371523,
+      "loss": 3.1826,
+      "step": 40792
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.667519211769104,
+      "learning_rate": 0.00015022672642333188,
+      "loss": 2.9398,
+      "step": 40793
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7207837700843811,
+      "learning_rate": 0.00015021338641742245,
+      "loss": 2.7841,
+      "step": 40794
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.661036491394043,
+      "learning_rate": 0.00015020004680602218,
+      "loss": 3.011,
+      "step": 40795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.72581946849823,
+      "learning_rate": 0.00015018670758916607,
+      "loss": 3.1146,
+      "step": 40796
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6660535335540771,
+      "learning_rate": 0.00015017336876688934,
+      "loss": 3.1452,
+      "step": 40797
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6611814498901367,
+      "learning_rate": 0.00015016003033922726,
+      "loss": 2.895,
+      "step": 40798
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6168580651283264,
+      "learning_rate": 0.00015014669230621483,
+      "loss": 3.0379,
+      "step": 40799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6940125226974487,
+      "learning_rate": 0.00015013335466788712,
+      "loss": 2.9946,
+      "step": 40800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6904090642929077,
+      "learning_rate": 0.00015012001742427923,
+      "loss": 2.9315,
+      "step": 40801
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6492003202438354,
+      "learning_rate": 0.00015010668057542637,
+      "loss": 2.9914,
+      "step": 40802
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.652450442314148,
+      "learning_rate": 0.00015009334412136377,
+      "loss": 3.0156,
+      "step": 40803
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6898772716522217,
+      "learning_rate": 0.00015008000806212635,
+      "loss": 3.0286,
+      "step": 40804
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6315729022026062,
+      "learning_rate": 0.00015006667239774943,
+      "loss": 2.6668,
+      "step": 40805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6902825832366943,
+      "learning_rate": 0.00015005333712826794,
+      "loss": 2.9763,
+      "step": 40806
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6964982151985168,
+      "learning_rate": 0.0001500400022537172,
+      "loss": 2.8922,
+      "step": 40807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6511028409004211,
+      "learning_rate": 0.00015002666777413229,
+      "loss": 3.2217,
+      "step": 40808
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6448655128479004,
+      "learning_rate": 0.0001500133336895481,
+      "loss": 3.1245,
+      "step": 40809
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8458114862442017,
+      "learning_rate": 0.00015000000000000004,
+      "loss": 2.9976,
+      "step": 40810
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6726029515266418,
+      "learning_rate": 0.00014998666670552302,
+      "loss": 2.8374,
+      "step": 40811
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6877138614654541,
+      "learning_rate": 0.00014997333380615223,
+      "loss": 2.915,
+      "step": 40812
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6989501714706421,
+      "learning_rate": 0.00014996000130192293,
+      "loss": 2.9306,
+      "step": 40813
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6564236283302307,
+      "learning_rate": 0.00014994666919287012,
+      "loss": 2.9193,
+      "step": 40814
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6724837422370911,
+      "learning_rate": 0.0001499333374790289,
+      "loss": 2.9177,
+      "step": 40815
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6566126346588135,
+      "learning_rate": 0.00014992000616043428,
+      "loss": 3.009,
+      "step": 40816
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6281498074531555,
+      "learning_rate": 0.00014990667523712145,
+      "loss": 3.0989,
+      "step": 40817
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6152768731117249,
+      "learning_rate": 0.0001498933447091257,
+      "loss": 2.8553,
+      "step": 40818
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.649753749370575,
+      "learning_rate": 0.00014988001457648186,
+      "loss": 2.8566,
+      "step": 40819
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6632745862007141,
+      "learning_rate": 0.0001498666848392253,
+      "loss": 2.9984,
+      "step": 40820
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6726300120353699,
+      "learning_rate": 0.00014985335549739097,
+      "loss": 2.6237,
+      "step": 40821
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6407365202903748,
+      "learning_rate": 0.00014984002655101393,
+      "loss": 2.9663,
+      "step": 40822
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6597623825073242,
+      "learning_rate": 0.00014982669800012944,
+      "loss": 2.8987,
+      "step": 40823
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.689781904220581,
+      "learning_rate": 0.00014981336984477243,
+      "loss": 2.9595,
+      "step": 40824
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6714945435523987,
+      "learning_rate": 0.0001498000420849782,
+      "loss": 2.9263,
+      "step": 40825
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.635680615901947,
+      "learning_rate": 0.00014978671472078167,
+      "loss": 2.8758,
+      "step": 40826
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6559689044952393,
+      "learning_rate": 0.0001497733877522181,
+      "loss": 3.0803,
+      "step": 40827
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6731073260307312,
+      "learning_rate": 0.00014976006117932243,
+      "loss": 2.9924,
+      "step": 40828
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6569121479988098,
+      "learning_rate": 0.00014974673500212996,
+      "loss": 2.9098,
+      "step": 40829
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6949353814125061,
+      "learning_rate": 0.00014973340922067564,
+      "loss": 2.73,
+      "step": 40830
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6641110777854919,
+      "learning_rate": 0.0001497200838349945,
+      "loss": 3.1152,
+      "step": 40831
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7344586253166199,
+      "learning_rate": 0.00014970675884512188,
+      "loss": 2.8915,
+      "step": 40832
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7318583726882935,
+      "learning_rate": 0.00014969343425109258,
+      "loss": 2.968,
+      "step": 40833
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7301783561706543,
+      "learning_rate": 0.00014968011005294187,
+      "loss": 3.0687,
+      "step": 40834
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6690212488174438,
+      "learning_rate": 0.00014966678625070494,
+      "loss": 2.7896,
+      "step": 40835
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6483722925186157,
+      "learning_rate": 0.00014965346284441674,
+      "loss": 2.9243,
+      "step": 40836
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6564923524856567,
+      "learning_rate": 0.00014964013983411237,
+      "loss": 2.853,
+      "step": 40837
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6334251165390015,
+      "learning_rate": 0.00014962681721982686,
+      "loss": 2.8823,
+      "step": 40838
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7064406275749207,
+      "learning_rate": 0.00014961349500159537,
+      "loss": 2.8174,
+      "step": 40839
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7101202607154846,
+      "learning_rate": 0.0001496001731794531,
+      "loss": 3.037,
+      "step": 40840
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.688157320022583,
+      "learning_rate": 0.0001495868517534349,
+      "loss": 2.8154,
+      "step": 40841
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6820224523544312,
+      "learning_rate": 0.0001495735307235761,
+      "loss": 2.9316,
+      "step": 40842
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6670019030570984,
+      "learning_rate": 0.0001495602100899117,
+      "loss": 2.9859,
+      "step": 40843
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7046992778778076,
+      "learning_rate": 0.0001495468898524766,
+      "loss": 2.7433,
+      "step": 40844
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6730843782424927,
+      "learning_rate": 0.00014953357001130617,
+      "loss": 2.7432,
+      "step": 40845
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6858168840408325,
+      "learning_rate": 0.00014952025056643524,
+      "loss": 2.9308,
+      "step": 40846
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6220602989196777,
+      "learning_rate": 0.00014950693151789913,
+      "loss": 2.903,
+      "step": 40847
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6544325947761536,
+      "learning_rate": 0.0001494936128657327,
+      "loss": 3.0168,
+      "step": 40848
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.668447732925415,
+      "learning_rate": 0.0001494802946099711,
+      "loss": 2.9114,
+      "step": 40849
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6595551371574402,
+      "learning_rate": 0.00014946697675064955,
+      "loss": 2.9192,
+      "step": 40850
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6544173359870911,
+      "learning_rate": 0.00014945365928780306,
+      "loss": 3.0946,
+      "step": 40851
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7191558480262756,
+      "learning_rate": 0.0001494403422214666,
+      "loss": 3.0198,
+      "step": 40852
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6780735850334167,
+      "learning_rate": 0.00014942702555167521,
+      "loss": 2.9645,
+      "step": 40853
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6664487719535828,
+      "learning_rate": 0.00014941370927846407,
+      "loss": 2.9704,
+      "step": 40854
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.706179678440094,
+      "learning_rate": 0.00014940039340186835,
+      "loss": 2.8314,
+      "step": 40855
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6947509050369263,
+      "learning_rate": 0.00014938707792192286,
+      "loss": 2.823,
+      "step": 40856
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6411890387535095,
+      "learning_rate": 0.00014937376283866296,
+      "loss": 3.0037,
+      "step": 40857
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7116650938987732,
+      "learning_rate": 0.00014936044815212356,
+      "loss": 2.8026,
+      "step": 40858
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6230390071868896,
+      "learning_rate": 0.00014934713386233968,
+      "loss": 2.948,
+      "step": 40859
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6088177561759949,
+      "learning_rate": 0.00014933381996934655,
+      "loss": 2.8931,
+      "step": 40860
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.68245530128479,
+      "learning_rate": 0.00014932050647317903,
+      "loss": 3.0912,
+      "step": 40861
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6987802386283875,
+      "learning_rate": 0.0001493071933738724,
+      "loss": 2.9147,
+      "step": 40862
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6714006066322327,
+      "learning_rate": 0.0001492938806714615,
+      "loss": 2.9907,
+      "step": 40863
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6982893347740173,
+      "learning_rate": 0.0001492805683659815,
+      "loss": 2.9568,
+      "step": 40864
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6661611795425415,
+      "learning_rate": 0.00014926725645746775,
+      "loss": 2.8846,
+      "step": 40865
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7231001257896423,
+      "learning_rate": 0.0001492539449459548,
+      "loss": 2.9071,
+      "step": 40866
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6181085705757141,
+      "learning_rate": 0.00014924063383147805,
+      "loss": 2.9526,
+      "step": 40867
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6579425930976868,
+      "learning_rate": 0.00014922732311407235,
+      "loss": 2.8463,
+      "step": 40868
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6170904636383057,
+      "learning_rate": 0.00014921401279377285,
+      "loss": 2.8966,
+      "step": 40869
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6523815393447876,
+      "learning_rate": 0.00014920070287061473,
+      "loss": 2.8954,
+      "step": 40870
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7443393468856812,
+      "learning_rate": 0.00014918739334463286,
+      "loss": 2.858,
+      "step": 40871
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6416648030281067,
+      "learning_rate": 0.00014917408421586247,
+      "loss": 2.8031,
+      "step": 40872
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6666995286941528,
+      "learning_rate": 0.0001491607754843385,
+      "loss": 2.951,
+      "step": 40873
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.642076849937439,
+      "learning_rate": 0.0001491474671500959,
+      "loss": 2.7547,
+      "step": 40874
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6719287037849426,
+      "learning_rate": 0.00014913415921316995,
+      "loss": 2.9337,
+      "step": 40875
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6449657678604126,
+      "learning_rate": 0.0001491208516735955,
+      "loss": 2.9809,
+      "step": 40876
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6794679164886475,
+      "learning_rate": 0.00014910754453140775,
+      "loss": 2.9901,
+      "step": 40877
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6926255822181702,
+      "learning_rate": 0.00014909423778664165,
+      "loss": 2.8638,
+      "step": 40878
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6646115779876709,
+      "learning_rate": 0.00014908093143933235,
+      "loss": 2.7086,
+      "step": 40879
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6892146468162537,
+      "learning_rate": 0.00014906762548951485,
+      "loss": 2.8719,
+      "step": 40880
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6423749923706055,
+      "learning_rate": 0.00014905431993722407,
+      "loss": 2.862,
+      "step": 40881
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7053487300872803,
+      "learning_rate": 0.00014904101478249525,
+      "loss": 3.0384,
+      "step": 40882
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6942588090896606,
+      "learning_rate": 0.00014902771002536325,
+      "loss": 2.9993,
+      "step": 40883
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.773780345916748,
+      "learning_rate": 0.0001490144056658632,
+      "loss": 2.9149,
+      "step": 40884
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7698376774787903,
+      "learning_rate": 0.00014900110170403027,
+      "loss": 2.89,
+      "step": 40885
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7158568501472473,
+      "learning_rate": 0.00014898779813989925,
+      "loss": 2.7576,
+      "step": 40886
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.671528697013855,
+      "learning_rate": 0.00014897449497350545,
+      "loss": 2.9737,
+      "step": 40887
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6958116888999939,
+      "learning_rate": 0.0001489611922048837,
+      "loss": 2.8742,
+      "step": 40888
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6262465119361877,
+      "learning_rate": 0.00014894788983406905,
+      "loss": 2.7978,
+      "step": 40889
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7286080718040466,
+      "learning_rate": 0.0001489345878610967,
+      "loss": 2.6667,
+      "step": 40890
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.686791718006134,
+      "learning_rate": 0.0001489212862860014,
+      "loss": 2.9975,
+      "step": 40891
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6629527807235718,
+      "learning_rate": 0.00014890798510881854,
+      "loss": 2.7829,
+      "step": 40892
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6882234215736389,
+      "learning_rate": 0.00014889468432958283,
+      "loss": 2.8711,
+      "step": 40893
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6697384119033813,
+      "learning_rate": 0.00014888138394832955,
+      "loss": 3.0293,
+      "step": 40894
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6356438994407654,
+      "learning_rate": 0.0001488680839650936,
+      "loss": 2.9981,
+      "step": 40895
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6785897612571716,
+      "learning_rate": 0.00014885478437990998,
+      "loss": 2.9045,
+      "step": 40896
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6552438139915466,
+      "learning_rate": 0.0001488414851928138,
+      "loss": 3.0607,
+      "step": 40897
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6709671020507812,
+      "learning_rate": 0.00014882818640384,
+      "loss": 3.0093,
+      "step": 40898
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.632817268371582,
+      "learning_rate": 0.00014881488801302367,
+      "loss": 3.0937,
+      "step": 40899
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6626100540161133,
+      "learning_rate": 0.0001488015900203999,
+      "loss": 2.7311,
+      "step": 40900
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7310886979103088,
+      "learning_rate": 0.0001487882924260037,
+      "loss": 2.8512,
+      "step": 40901
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6690772771835327,
+      "learning_rate": 0.00014877499522986996,
+      "loss": 2.8298,
+      "step": 40902
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6155535578727722,
+      "learning_rate": 0.0001487616984320337,
+      "loss": 2.8708,
+      "step": 40903
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6756638884544373,
+      "learning_rate": 0.00014874840203253016,
+      "loss": 3.0289,
+      "step": 40904
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6889369487762451,
+      "learning_rate": 0.00014873510603139408,
+      "loss": 2.9191,
+      "step": 40905
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6636484265327454,
+      "learning_rate": 0.0001487218104286606,
+      "loss": 2.8135,
+      "step": 40906
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7536406517028809,
+      "learning_rate": 0.0001487085152243649,
+      "loss": 3.1152,
+      "step": 40907
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6910778880119324,
+      "learning_rate": 0.00014869522041854175,
+      "loss": 2.8008,
+      "step": 40908
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7060720324516296,
+      "learning_rate": 0.00014868192601122637,
+      "loss": 3.0552,
+      "step": 40909
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6742306351661682,
+      "learning_rate": 0.0001486686320024537,
+      "loss": 2.7962,
+      "step": 40910
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6626346111297607,
+      "learning_rate": 0.00014865533839225858,
+      "loss": 2.9618,
+      "step": 40911
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6891857385635376,
+      "learning_rate": 0.00014864204518067628,
+      "loss": 2.9439,
+      "step": 40912
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.650107741355896,
+      "learning_rate": 0.00014862875236774164,
+      "loss": 2.9421,
+      "step": 40913
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7103627324104309,
+      "learning_rate": 0.0001486154599534898,
+      "loss": 3.0348,
+      "step": 40914
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6941289901733398,
+      "learning_rate": 0.0001486021679379556,
+      "loss": 3.0479,
+      "step": 40915
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6747994422912598,
+      "learning_rate": 0.00014858887632117424,
+      "loss": 3.063,
+      "step": 40916
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6820676922798157,
+      "learning_rate": 0.00014857558510318072,
+      "loss": 3.0172,
+      "step": 40917
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6499535441398621,
+      "learning_rate": 0.0001485622942840098,
+      "loss": 2.9887,
+      "step": 40918
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6234838962554932,
+      "learning_rate": 0.00014854900386369676,
+      "loss": 2.8712,
+      "step": 40919
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7115194201469421,
+      "learning_rate": 0.00014853571384227637,
+      "loss": 2.9185,
+      "step": 40920
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6853716373443604,
+      "learning_rate": 0.00014852242421978378,
+      "loss": 2.8748,
+      "step": 40921
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7544880509376526,
+      "learning_rate": 0.00014850913499625412,
+      "loss": 2.7981,
+      "step": 40922
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6600387692451477,
+      "learning_rate": 0.00014849584617172222,
+      "loss": 2.9013,
+      "step": 40923
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6780473589897156,
+      "learning_rate": 0.0001484825577462231,
+      "loss": 2.8438,
+      "step": 40924
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.667023241519928,
+      "learning_rate": 0.00014846926971979162,
+      "loss": 2.8672,
+      "step": 40925
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6934083104133606,
+      "learning_rate": 0.00014845598209246294,
+      "loss": 2.7644,
+      "step": 40926
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7000552415847778,
+      "learning_rate": 0.00014844269486427217,
+      "loss": 2.9773,
+      "step": 40927
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6767484545707703,
+      "learning_rate": 0.00014842940803525405,
+      "loss": 2.5985,
+      "step": 40928
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6668387055397034,
+      "learning_rate": 0.00014841612160544382,
+      "loss": 2.9201,
+      "step": 40929
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6447761654853821,
+      "learning_rate": 0.0001484028355748762,
+      "loss": 3.0512,
+      "step": 40930
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6750105619430542,
+      "learning_rate": 0.00014838954994358648,
+      "loss": 3.1484,
+      "step": 40931
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6737284064292908,
+      "learning_rate": 0.00014837626471160952,
+      "loss": 2.8619,
+      "step": 40932
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6524981260299683,
+      "learning_rate": 0.00014836297987898018,
+      "loss": 3.1476,
+      "step": 40933
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6673911809921265,
+      "learning_rate": 0.00014834969544573367,
+      "loss": 2.9744,
+      "step": 40934
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6576171517372131,
+      "learning_rate": 0.00014833641141190476,
+      "loss": 2.8008,
+      "step": 40935
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6706809997558594,
+      "learning_rate": 0.00014832312777752856,
+      "loss": 2.8009,
+      "step": 40936
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6633245348930359,
+      "learning_rate": 0.00014830984454264015,
+      "loss": 2.7447,
+      "step": 40937
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6687329411506653,
+      "learning_rate": 0.00014829656170727443,
+      "loss": 2.8792,
+      "step": 40938
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7036616206169128,
+      "learning_rate": 0.0001482832792714664,
+      "loss": 2.9527,
+      "step": 40939
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6736128330230713,
+      "learning_rate": 0.00014826999723525084,
+      "loss": 2.9944,
+      "step": 40940
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6891912817955017,
+      "learning_rate": 0.00014825671559866294,
+      "loss": 2.9328,
+      "step": 40941
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6677616834640503,
+      "learning_rate": 0.00014824343436173774,
+      "loss": 2.9777,
+      "step": 40942
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6840057969093323,
+      "learning_rate": 0.00014823015352451006,
+      "loss": 2.9738,
+      "step": 40943
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7241817712783813,
+      "learning_rate": 0.00014821687308701502,
+      "loss": 2.8735,
+      "step": 40944
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7008278369903564,
+      "learning_rate": 0.0001482035930492875,
+      "loss": 2.8839,
+      "step": 40945
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.639018714427948,
+      "learning_rate": 0.00014819031341136246,
+      "loss": 2.9546,
+      "step": 40946
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6628701686859131,
+      "learning_rate": 0.000148177034173275,
+      "loss": 2.8829,
+      "step": 40947
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7044912576675415,
+      "learning_rate": 0.00014816375533505992,
+      "loss": 2.7321,
+      "step": 40948
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.658323347568512,
+      "learning_rate": 0.00014815047689675234,
+      "loss": 2.9143,
+      "step": 40949
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6978631615638733,
+      "learning_rate": 0.00014813719885838712,
+      "loss": 2.7587,
+      "step": 40950
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6369495987892151,
+      "learning_rate": 0.00014812392121999927,
+      "loss": 2.9732,
+      "step": 40951
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6941462159156799,
+      "learning_rate": 0.00014811064398162388,
+      "loss": 2.7965,
+      "step": 40952
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6733207106590271,
+      "learning_rate": 0.00014809736714329588,
+      "loss": 2.9032,
+      "step": 40953
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6448726058006287,
+      "learning_rate": 0.00014808409070505012,
+      "loss": 2.8493,
+      "step": 40954
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6968315839767456,
+      "learning_rate": 0.00014807081466692155,
+      "loss": 2.8681,
+      "step": 40955
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6684523820877075,
+      "learning_rate": 0.0001480575390289452,
+      "loss": 2.9887,
+      "step": 40956
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6307798027992249,
+      "learning_rate": 0.0001480442637911562,
+      "loss": 2.8416,
+      "step": 40957
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.660571277141571,
+      "learning_rate": 0.0001480309889535892,
+      "loss": 3.104,
+      "step": 40958
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6901208758354187,
+      "learning_rate": 0.00014801771451627944,
+      "loss": 2.873,
+      "step": 40959
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6825695037841797,
+      "learning_rate": 0.0001480044404792618,
+      "loss": 2.9064,
+      "step": 40960
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6804020404815674,
+      "learning_rate": 0.00014799116684257108,
+      "loss": 2.9686,
+      "step": 40961
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6763026714324951,
+      "learning_rate": 0.00014797789360624248,
+      "loss": 2.9093,
+      "step": 40962
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6673914194107056,
+      "learning_rate": 0.00014796462077031075,
+      "loss": 3.0376,
+      "step": 40963
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6648275852203369,
+      "learning_rate": 0.0001479513483348111,
+      "loss": 2.8259,
+      "step": 40964
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6990496516227722,
+      "learning_rate": 0.00014793807629977817,
+      "loss": 2.86,
+      "step": 40965
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6605084538459778,
+      "learning_rate": 0.0001479248046652471,
+      "loss": 3.1202,
+      "step": 40966
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7304693460464478,
+      "learning_rate": 0.00014791153343125305,
+      "loss": 2.9801,
+      "step": 40967
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7068866491317749,
+      "learning_rate": 0.0001478982625978305,
+      "loss": 2.8674,
+      "step": 40968
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6797716021537781,
+      "learning_rate": 0.00014788499216501481,
+      "loss": 2.8571,
+      "step": 40969
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6691450476646423,
+      "learning_rate": 0.00014787172213284066,
+      "loss": 2.9016,
+      "step": 40970
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6694939732551575,
+      "learning_rate": 0.0001478584525013431,
+      "loss": 2.7443,
+      "step": 40971
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6383887529373169,
+      "learning_rate": 0.00014784518327055722,
+      "loss": 2.8351,
+      "step": 40972
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6808183193206787,
+      "learning_rate": 0.0001478319144405177,
+      "loss": 2.9814,
+      "step": 40973
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6715272665023804,
+      "learning_rate": 0.00014781864601125976,
+      "loss": 3.099,
+      "step": 40974
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7185418009757996,
+      "learning_rate": 0.00014780537798281823,
+      "loss": 2.9109,
+      "step": 40975
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6826639175415039,
+      "learning_rate": 0.00014779211035522794,
+      "loss": 2.8128,
+      "step": 40976
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6765093207359314,
+      "learning_rate": 0.00014777884312852405,
+      "loss": 2.9096,
+      "step": 40977
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.691008448600769,
+      "learning_rate": 0.00014776557630274127,
+      "loss": 2.8658,
+      "step": 40978
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6892204880714417,
+      "learning_rate": 0.00014775230987791477,
+      "loss": 2.9025,
+      "step": 40979
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7367006540298462,
+      "learning_rate": 0.00014773904385407925,
+      "loss": 2.7799,
+      "step": 40980
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7002752423286438,
+      "learning_rate": 0.00014772577823126993,
+      "loss": 2.981,
+      "step": 40981
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6693331003189087,
+      "learning_rate": 0.00014771251300952162,
+      "loss": 2.9569,
+      "step": 40982
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6542279124259949,
+      "learning_rate": 0.00014769924818886908,
+      "loss": 2.99,
+      "step": 40983
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7216241955757141,
+      "learning_rate": 0.00014768598376934758,
+      "loss": 2.9451,
+      "step": 40984
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6984073519706726,
+      "learning_rate": 0.00014767271975099172,
+      "loss": 3.0107,
+      "step": 40985
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6856890320777893,
+      "learning_rate": 0.00014765945613383675,
+      "loss": 2.8309,
+      "step": 40986
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6620368361473083,
+      "learning_rate": 0.00014764619291791732,
+      "loss": 2.8103,
+      "step": 40987
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6587978005409241,
+      "learning_rate": 0.00014763293010326855,
+      "loss": 2.8216,
+      "step": 40988
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6450566649436951,
+      "learning_rate": 0.00014761966768992536,
+      "loss": 3.0106,
+      "step": 40989
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6524011492729187,
+      "learning_rate": 0.00014760640567792265,
+      "loss": 2.8636,
+      "step": 40990
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6809319257736206,
+      "learning_rate": 0.00014759314406729537,
+      "loss": 2.8167,
+      "step": 40991
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6798052787780762,
+      "learning_rate": 0.0001475798828580783,
+      "loss": 2.8203,
+      "step": 40992
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6698510050773621,
+      "learning_rate": 0.00014756662205030646,
+      "loss": 2.9901,
+      "step": 40993
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6708970665931702,
+      "learning_rate": 0.00014755336164401492,
+      "loss": 2.9513,
+      "step": 40994
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7181163430213928,
+      "learning_rate": 0.00014754010163923838,
+      "loss": 2.9335,
+      "step": 40995
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6596902012825012,
+      "learning_rate": 0.00014752684203601197,
+      "loss": 3.0563,
+      "step": 40996
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.681885838508606,
+      "learning_rate": 0.00014751358283437055,
+      "loss": 2.9588,
+      "step": 40997
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6878100037574768,
+      "learning_rate": 0.00014750032403434884,
+      "loss": 2.9884,
+      "step": 40998
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.696635365486145,
+      "learning_rate": 0.0001474870656359821,
+      "loss": 2.9776,
+      "step": 40999
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6515666842460632,
+      "learning_rate": 0.00014747380763930492,
+      "loss": 3.0161,
+      "step": 41000
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.673594057559967,
+      "learning_rate": 0.0001474605500443525,
+      "loss": 3.0754,
+      "step": 41001
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6816129088401794,
+      "learning_rate": 0.00014744729285115954,
+      "loss": 2.928,
+      "step": 41002
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6874986290931702,
+      "learning_rate": 0.00014743403605976113,
+      "loss": 2.9407,
+      "step": 41003
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6564611792564392,
+      "learning_rate": 0.00014742077967019215,
+      "loss": 2.8386,
+      "step": 41004
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6538634300231934,
+      "learning_rate": 0.00014740752368248733,
+      "loss": 2.9536,
+      "step": 41005
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6837107539176941,
+      "learning_rate": 0.00014739426809668186,
+      "loss": 2.7744,
+      "step": 41006
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7023835778236389,
+      "learning_rate": 0.00014738101291281036,
+      "loss": 2.8523,
+      "step": 41007
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6810896396636963,
+      "learning_rate": 0.00014736775813090793,
+      "loss": 3.1852,
+      "step": 41008
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6751614809036255,
+      "learning_rate": 0.0001473545037510096,
+      "loss": 2.9334,
+      "step": 41009
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6512244939804077,
+      "learning_rate": 0.00014734124977315,
+      "loss": 2.8679,
+      "step": 41010
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6602606773376465,
+      "learning_rate": 0.00014732799619736423,
+      "loss": 2.7867,
+      "step": 41011
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6242491006851196,
+      "learning_rate": 0.00014731474302368715,
+      "loss": 3.0989,
+      "step": 41012
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6728277206420898,
+      "learning_rate": 0.00014730149025215356,
+      "loss": 2.9024,
+      "step": 41013
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6889778971672058,
+      "learning_rate": 0.00014728823788279859,
+      "loss": 2.952,
+      "step": 41014
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.709487795829773,
+      "learning_rate": 0.00014727498591565687,
+      "loss": 3.0053,
+      "step": 41015
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7062227129936218,
+      "learning_rate": 0.00014726173435076356,
+      "loss": 2.8784,
+      "step": 41016
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6593536138534546,
+      "learning_rate": 0.00014724848318815333,
+      "loss": 3.0108,
+      "step": 41017
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6188623309135437,
+      "learning_rate": 0.0001472352324278613,
+      "loss": 2.8337,
+      "step": 41018
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6538429260253906,
+      "learning_rate": 0.00014722198206992232,
+      "loss": 2.7436,
+      "step": 41019
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7170898914337158,
+      "learning_rate": 0.00014720873211437107,
+      "loss": 2.916,
+      "step": 41020
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6724970936775208,
+      "learning_rate": 0.00014719548256124276,
+      "loss": 2.9541,
+      "step": 41021
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6659535765647888,
+      "learning_rate": 0.00014718223341057203,
+      "loss": 3.0944,
+      "step": 41022
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6524555683135986,
+      "learning_rate": 0.00014716898466239384,
+      "loss": 3.1875,
+      "step": 41023
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6858758926391602,
+      "learning_rate": 0.00014715573631674328,
+      "loss": 2.8407,
+      "step": 41024
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.64168781042099,
+      "learning_rate": 0.0001471424883736551,
+      "loss": 2.8617,
+      "step": 41025
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6399775743484497,
+      "learning_rate": 0.00014712924083316417,
+      "loss": 3.0222,
+      "step": 41026
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6765655875205994,
+      "learning_rate": 0.0001471159936953053,
+      "loss": 2.8945,
+      "step": 41027
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6615230441093445,
+      "learning_rate": 0.00014710274696011347,
+      "loss": 2.7149,
+      "step": 41028
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.726775586605072,
+      "learning_rate": 0.00014708950062762368,
+      "loss": 2.961,
+      "step": 41029
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7074599862098694,
+      "learning_rate": 0.00014707625469787065,
+      "loss": 3.0504,
+      "step": 41030
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6598540544509888,
+      "learning_rate": 0.0001470630091708894,
+      "loss": 2.9812,
+      "step": 41031
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6536500453948975,
+      "learning_rate": 0.00014704976404671466,
+      "loss": 3.0781,
+      "step": 41032
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7111579775810242,
+      "learning_rate": 0.00014703651932538152,
+      "loss": 3.1258,
+      "step": 41033
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6571623682975769,
+      "learning_rate": 0.00014702327500692474,
+      "loss": 3.0178,
+      "step": 41034
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6660772562026978,
+      "learning_rate": 0.00014701003109137908,
+      "loss": 2.8174,
+      "step": 41035
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6512477993965149,
+      "learning_rate": 0.0001469967875787797,
+      "loss": 3.073,
+      "step": 41036
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6898970603942871,
+      "learning_rate": 0.00014698354446916124,
+      "loss": 2.8279,
+      "step": 41037
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6960706114768982,
+      "learning_rate": 0.00014697030176255864,
+      "loss": 3.0873,
+      "step": 41038
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6520676612854004,
+      "learning_rate": 0.00014695705945900691,
+      "loss": 3.0638,
+      "step": 41039
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.716157853603363,
+      "learning_rate": 0.00014694381755854088,
+      "loss": 2.9452,
+      "step": 41040
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7131884098052979,
+      "learning_rate": 0.00014693057606119533,
+      "loss": 3.0743,
+      "step": 41041
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6574112772941589,
+      "learning_rate": 0.0001469173349670051,
+      "loss": 2.943,
+      "step": 41042
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6817065477371216,
+      "learning_rate": 0.00014690409427600516,
+      "loss": 2.8438,
+      "step": 41043
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6450731754302979,
+      "learning_rate": 0.00014689085398823047,
+      "loss": 3.0053,
+      "step": 41044
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6768467426300049,
+      "learning_rate": 0.0001468776141037157,
+      "loss": 3.0556,
+      "step": 41045
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6716314554214478,
+      "learning_rate": 0.00014686437462249592,
+      "loss": 2.6741,
+      "step": 41046
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6731881499290466,
+      "learning_rate": 0.0001468511355446059,
+      "loss": 2.8936,
+      "step": 41047
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.705536961555481,
+      "learning_rate": 0.00014683789687008043,
+      "loss": 2.8119,
+      "step": 41048
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6849344372749329,
+      "learning_rate": 0.00014682465859895458,
+      "loss": 3.0863,
+      "step": 41049
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7398621439933777,
+      "learning_rate": 0.00014681142073126295,
+      "loss": 2.9127,
+      "step": 41050
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6680430769920349,
+      "learning_rate": 0.00014679818326704072,
+      "loss": 2.9321,
+      "step": 41051
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6883407235145569,
+      "learning_rate": 0.00014678494620632242,
+      "loss": 2.9762,
+      "step": 41052
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6741495132446289,
+      "learning_rate": 0.00014677170954914312,
+      "loss": 2.9253,
+      "step": 41053
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6511622071266174,
+      "learning_rate": 0.00014675847329553772,
+      "loss": 3.0779,
+      "step": 41054
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6725462079048157,
+      "learning_rate": 0.00014674523744554109,
+      "loss": 2.8217,
+      "step": 41055
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.67629075050354,
+      "learning_rate": 0.00014673200199918796,
+      "loss": 2.9472,
+      "step": 41056
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6784157752990723,
+      "learning_rate": 0.0001467187669565131,
+      "loss": 2.8195,
+      "step": 41057
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6392921209335327,
+      "learning_rate": 0.00014670553231755165,
+      "loss": 3.0216,
+      "step": 41058
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6643005013465881,
+      "learning_rate": 0.00014669229808233822,
+      "loss": 3.0051,
+      "step": 41059
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.617713987827301,
+      "learning_rate": 0.00014667906425090775,
+      "loss": 3.0391,
+      "step": 41060
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6412746906280518,
+      "learning_rate": 0.00014666583082329526,
+      "loss": 2.7262,
+      "step": 41061
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6644930243492126,
+      "learning_rate": 0.00014665259779953545,
+      "loss": 2.8943,
+      "step": 41062
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.67530357837677,
+      "learning_rate": 0.00014663936517966315,
+      "loss": 2.7935,
+      "step": 41063
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7180463075637817,
+      "learning_rate": 0.00014662613296371316,
+      "loss": 2.7565,
+      "step": 41064
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6856942176818848,
+      "learning_rate": 0.00014661290115172045,
+      "loss": 2.9684,
+      "step": 41065
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.672465443611145,
+      "learning_rate": 0.00014659966974371993,
+      "loss": 2.8019,
+      "step": 41066
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6677531003952026,
+      "learning_rate": 0.00014658643873974623,
+      "loss": 2.8741,
+      "step": 41067
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6654563546180725,
+      "learning_rate": 0.00014657320813983446,
+      "loss": 2.8794,
+      "step": 41068
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6987980604171753,
+      "learning_rate": 0.00014655997794401932,
+      "loss": 2.9344,
+      "step": 41069
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6206268668174744,
+      "learning_rate": 0.00014654674815233555,
+      "loss": 2.898,
+      "step": 41070
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6681779026985168,
+      "learning_rate": 0.00014653351876481823,
+      "loss": 2.8274,
+      "step": 41071
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6701809167861938,
+      "learning_rate": 0.00014652028978150196,
+      "loss": 2.7613,
+      "step": 41072
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6777552962303162,
+      "learning_rate": 0.00014650706120242186,
+      "loss": 2.967,
+      "step": 41073
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7489084601402283,
+      "learning_rate": 0.00014649383302761247,
+      "loss": 3.0146,
+      "step": 41074
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6484798192977905,
+      "learning_rate": 0.0001464806052571088,
+      "loss": 3.095,
+      "step": 41075
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.716480553150177,
+      "learning_rate": 0.00014646737789094578,
+      "loss": 2.8885,
+      "step": 41076
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6795092821121216,
+      "learning_rate": 0.00014645415092915813,
+      "loss": 2.5886,
+      "step": 41077
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7201026678085327,
+      "learning_rate": 0.00014644092437178074,
+      "loss": 2.8806,
+      "step": 41078
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6745376586914062,
+      "learning_rate": 0.00014642769821884824,
+      "loss": 3.0144,
+      "step": 41079
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7620112895965576,
+      "learning_rate": 0.0001464144724703956,
+      "loss": 2.797,
+      "step": 41080
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6425158381462097,
+      "learning_rate": 0.00014640124712645786,
+      "loss": 2.6399,
+      "step": 41081
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6472249031066895,
+      "learning_rate": 0.00014638802218706955,
+      "loss": 3.0311,
+      "step": 41082
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6940520405769348,
+      "learning_rate": 0.00014637479765226572,
+      "loss": 2.9101,
+      "step": 41083
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6446316242218018,
+      "learning_rate": 0.00014636157352208113,
+      "loss": 2.8199,
+      "step": 41084
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.662503719329834,
+      "learning_rate": 0.00014634834979655042,
+      "loss": 3.034,
+      "step": 41085
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6603543758392334,
+      "learning_rate": 0.00014633512647570877,
+      "loss": 2.9694,
+      "step": 41086
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6480185389518738,
+      "learning_rate": 0.00014632190355959066,
+      "loss": 3.0205,
+      "step": 41087
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.681262195110321,
+      "learning_rate": 0.00014630868104823118,
+      "loss": 3.0011,
+      "step": 41088
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6547122001647949,
+      "learning_rate": 0.000146295458941665,
+      "loss": 2.9701,
+      "step": 41089
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6928819417953491,
+      "learning_rate": 0.00014628223723992694,
+      "loss": 2.9301,
+      "step": 41090
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6732745170593262,
+      "learning_rate": 0.00014626901594305203,
+      "loss": 2.7723,
+      "step": 41091
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6303524971008301,
+      "learning_rate": 0.00014625579505107497,
+      "loss": 2.7936,
+      "step": 41092
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.675657331943512,
+      "learning_rate": 0.00014624257456403054,
+      "loss": 2.9224,
+      "step": 41093
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7712878584861755,
+      "learning_rate": 0.00014622935448195345,
+      "loss": 2.9784,
+      "step": 41094
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6558848023414612,
+      "learning_rate": 0.00014621613480487864,
+      "loss": 2.9069,
+      "step": 41095
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7321421504020691,
+      "learning_rate": 0.00014620291553284106,
+      "loss": 2.8179,
+      "step": 41096
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6873998641967773,
+      "learning_rate": 0.0001461896966658753,
+      "loss": 2.9708,
+      "step": 41097
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7613605856895447,
+      "learning_rate": 0.00014617647820401637,
+      "loss": 3.1689,
+      "step": 41098
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6851860284805298,
+      "learning_rate": 0.000146163260147299,
+      "loss": 3.0292,
+      "step": 41099
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6968518495559692,
+      "learning_rate": 0.0001461500424957579,
+      "loss": 2.7074,
+      "step": 41100
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6626551151275635,
+      "learning_rate": 0.00014613682524942804,
+      "loss": 2.8939,
+      "step": 41101
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6604430079460144,
+      "learning_rate": 0.00014612360840834411,
+      "loss": 2.9987,
+      "step": 41102
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.711830735206604,
+      "learning_rate": 0.00014611039197254106,
+      "loss": 3.0629,
+      "step": 41103
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6499634385108948,
+      "learning_rate": 0.00014609717594205355,
+      "loss": 2.7851,
+      "step": 41104
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6638641953468323,
+      "learning_rate": 0.00014608396031691657,
+      "loss": 2.8837,
+      "step": 41105
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6749254465103149,
+      "learning_rate": 0.00014607074509716478,
+      "loss": 3.0239,
+      "step": 41106
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6623033285140991,
+      "learning_rate": 0.00014605753028283292,
+      "loss": 2.9147,
+      "step": 41107
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7307236790657043,
+      "learning_rate": 0.00014604431587395602,
+      "loss": 2.833,
+      "step": 41108
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6709109544754028,
+      "learning_rate": 0.00014603110187056865,
+      "loss": 2.9349,
+      "step": 41109
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6868018507957458,
+      "learning_rate": 0.00014601788827270574,
+      "loss": 3.0805,
+      "step": 41110
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6887187361717224,
+      "learning_rate": 0.00014600467508040218,
+      "loss": 2.9418,
+      "step": 41111
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6684630513191223,
+      "learning_rate": 0.00014599146229369252,
+      "loss": 2.9893,
+      "step": 41112
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6830692291259766,
+      "learning_rate": 0.00014597824991261185,
+      "loss": 2.9589,
+      "step": 41113
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6974941492080688,
+      "learning_rate": 0.00014596503793719482,
+      "loss": 2.9326,
+      "step": 41114
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.638469934463501,
+      "learning_rate": 0.00014595182636747614,
+      "loss": 2.9861,
+      "step": 41115
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.6490311026573181,
+      "learning_rate": 0.0001459386152034908,
+      "loss": 3.0936,
+      "step": 41116
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6784543991088867,
+      "learning_rate": 0.0001459254044452734,
+      "loss": 3.0954,
+      "step": 41117
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6834042072296143,
+      "learning_rate": 0.00014591219409285894,
+      "loss": 2.914,
+      "step": 41118
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6517180800437927,
+      "learning_rate": 0.00014589898414628198,
+      "loss": 3.13,
+      "step": 41119
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6616392731666565,
+      "learning_rate": 0.00014588577460557755,
+      "loss": 3.1142,
+      "step": 41120
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6413401365280151,
+      "learning_rate": 0.00014587256547078035,
+      "loss": 3.092,
+      "step": 41121
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6451615691184998,
+      "learning_rate": 0.00014585935674192504,
+      "loss": 2.8525,
+      "step": 41122
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6675742268562317,
+      "learning_rate": 0.0001458461484190466,
+      "loss": 3.0727,
+      "step": 41123
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6669958829879761,
+      "learning_rate": 0.00014583294050217967,
+      "loss": 2.8502,
+      "step": 41124
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6869798302650452,
+      "learning_rate": 0.0001458197329913591,
+      "loss": 2.9034,
+      "step": 41125
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6724489331245422,
+      "learning_rate": 0.00014580652588661974,
+      "loss": 2.9271,
+      "step": 41126
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6963558793067932,
+      "learning_rate": 0.00014579331918799634,
+      "loss": 2.8939,
+      "step": 41127
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6525717973709106,
+      "learning_rate": 0.0001457801128955237,
+      "loss": 2.8894,
+      "step": 41128
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6995739936828613,
+      "learning_rate": 0.0001457669070092364,
+      "loss": 2.807,
+      "step": 41129
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6798244714736938,
+      "learning_rate": 0.0001457537015291694,
+      "loss": 3.1188,
+      "step": 41130
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6413742303848267,
+      "learning_rate": 0.0001457404964553576,
+      "loss": 2.8515,
+      "step": 41131
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.68084317445755,
+      "learning_rate": 0.0001457272917878355,
+      "loss": 2.7809,
+      "step": 41132
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6480937004089355,
+      "learning_rate": 0.00014571408752663817,
+      "loss": 2.7539,
+      "step": 41133
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6867110729217529,
+      "learning_rate": 0.00014570088367180006,
+      "loss": 3.0271,
+      "step": 41134
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7035594582557678,
+      "learning_rate": 0.00014568768022335632,
+      "loss": 3.1673,
+      "step": 41135
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6588776707649231,
+      "learning_rate": 0.00014567447718134147,
+      "loss": 2.9375,
+      "step": 41136
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6488538980484009,
+      "learning_rate": 0.00014566127454579026,
+      "loss": 2.9297,
+      "step": 41137
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6975100040435791,
+      "learning_rate": 0.00014564807231673768,
+      "loss": 3.0464,
+      "step": 41138
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6509141325950623,
+      "learning_rate": 0.00014563487049421825,
+      "loss": 2.8602,
+      "step": 41139
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6687818169593811,
+      "learning_rate": 0.00014562166907826696,
+      "loss": 2.992,
+      "step": 41140
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6703136563301086,
+      "learning_rate": 0.00014560846806891837,
+      "loss": 2.8122,
+      "step": 41141
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7211496829986572,
+      "learning_rate": 0.00014559526746620748,
+      "loss": 2.7898,
+      "step": 41142
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6537548303604126,
+      "learning_rate": 0.00014558206727016896,
+      "loss": 2.7898,
+      "step": 41143
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6834998726844788,
+      "learning_rate": 0.00014556886748083743,
+      "loss": 2.9153,
+      "step": 41144
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6743541359901428,
+      "learning_rate": 0.00014555566809824788,
+      "loss": 2.9644,
+      "step": 41145
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6714619994163513,
+      "learning_rate": 0.00014554246912243488,
+      "loss": 2.9662,
+      "step": 41146
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6737746596336365,
+      "learning_rate": 0.0001455292705534333,
+      "loss": 2.7696,
+      "step": 41147
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.654802143573761,
+      "learning_rate": 0.00014551607239127797,
+      "loss": 2.8942,
+      "step": 41148
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6912543177604675,
+      "learning_rate": 0.0001455028746360036,
+      "loss": 3.193,
+      "step": 41149
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6937560439109802,
+      "learning_rate": 0.00014548967728764492,
+      "loss": 2.8186,
+      "step": 41150
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6445648074150085,
+      "learning_rate": 0.00014547648034623657,
+      "loss": 2.8501,
+      "step": 41151
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7112509608268738,
+      "learning_rate": 0.00014546328381181346,
+      "loss": 2.6133,
+      "step": 41152
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6472128629684448,
+      "learning_rate": 0.00014545008768441042,
+      "loss": 2.8535,
+      "step": 41153
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6357482075691223,
+      "learning_rate": 0.00014543689196406197,
+      "loss": 2.9497,
+      "step": 41154
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6605250239372253,
+      "learning_rate": 0.00014542369665080315,
+      "loss": 2.8173,
+      "step": 41155
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6524414420127869,
+      "learning_rate": 0.00014541050174466844,
+      "loss": 3.0099,
+      "step": 41156
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6539895534515381,
+      "learning_rate": 0.0001453973072456928,
+      "loss": 3.0537,
+      "step": 41157
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6876203417778015,
+      "learning_rate": 0.00014538411315391094,
+      "loss": 2.9975,
+      "step": 41158
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6869372725486755,
+      "learning_rate": 0.00014537091946935742,
+      "loss": 3.1229,
+      "step": 41159
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6883721947669983,
+      "learning_rate": 0.0001453577261920673,
+      "loss": 2.8087,
+      "step": 41160
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6697179675102234,
+      "learning_rate": 0.000145344533322075,
+      "loss": 2.9453,
+      "step": 41161
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6436740159988403,
+      "learning_rate": 0.00014533134085941546,
+      "loss": 3.0862,
+      "step": 41162
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6619797945022583,
+      "learning_rate": 0.0001453181488041235,
+      "loss": 3.0743,
+      "step": 41163
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6886985898017883,
+      "learning_rate": 0.0001453049571562338,
+      "loss": 2.8477,
+      "step": 41164
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6293699145317078,
+      "learning_rate": 0.00014529176591578105,
+      "loss": 2.8921,
+      "step": 41165
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7019802331924438,
+      "learning_rate": 0.00014527857508279992,
+      "loss": 2.9149,
+      "step": 41166
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6821876168251038,
+      "learning_rate": 0.00014526538465732522,
+      "loss": 2.7066,
+      "step": 41167
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7083041071891785,
+      "learning_rate": 0.00014525219463939183,
+      "loss": 2.9684,
+      "step": 41168
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7001522183418274,
+      "learning_rate": 0.0001452390050290343,
+      "loss": 2.9329,
+      "step": 41169
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7127074003219604,
+      "learning_rate": 0.00014522581582628754,
+      "loss": 2.958,
+      "step": 41170
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6899229884147644,
+      "learning_rate": 0.00014521262703118617,
+      "loss": 2.9127,
+      "step": 41171
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6654265522956848,
+      "learning_rate": 0.00014519943864376486,
+      "loss": 2.997,
+      "step": 41172
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.670941174030304,
+      "learning_rate": 0.00014518625066405852,
+      "loss": 2.6803,
+      "step": 41173
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6345453858375549,
+      "learning_rate": 0.00014517306309210176,
+      "loss": 3.1015,
+      "step": 41174
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6700384616851807,
+      "learning_rate": 0.0001451598759279294,
+      "loss": 3.0479,
+      "step": 41175
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.686065673828125,
+      "learning_rate": 0.00014514668917157603,
+      "loss": 2.9507,
+      "step": 41176
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6618055701255798,
+      "learning_rate": 0.00014513350282307648,
+      "loss": 2.918,
+      "step": 41177
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7136662006378174,
+      "learning_rate": 0.00014512031688246563,
+      "loss": 3.132,
+      "step": 41178
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6863377690315247,
+      "learning_rate": 0.00014510713134977804,
+      "loss": 2.8091,
+      "step": 41179
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6750915050506592,
+      "learning_rate": 0.00014509394622504843,
+      "loss": 2.9447,
+      "step": 41180
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6594275832176208,
+      "learning_rate": 0.00014508076150831145,
+      "loss": 3.0114,
+      "step": 41181
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6762315630912781,
+      "learning_rate": 0.00014506757719960192,
+      "loss": 2.8729,
+      "step": 41182
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7663244605064392,
+      "learning_rate": 0.0001450543932989547,
+      "loss": 2.8163,
+      "step": 41183
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7097091674804688,
+      "learning_rate": 0.0001450412098064043,
+      "loss": 2.8352,
+      "step": 41184
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7125594615936279,
+      "learning_rate": 0.0001450280267219856,
+      "loss": 3.0027,
+      "step": 41185
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6626073718070984,
+      "learning_rate": 0.00014501484404573327,
+      "loss": 2.8722,
+      "step": 41186
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6956759691238403,
+      "learning_rate": 0.0001450016617776819,
+      "loss": 2.7935,
+      "step": 41187
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.638950765132904,
+      "learning_rate": 0.00014498847991786644,
+      "loss": 2.8794,
+      "step": 41188
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6870553493499756,
+      "learning_rate": 0.00014497529846632136,
+      "loss": 3.0417,
+      "step": 41189
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.681865394115448,
+      "learning_rate": 0.00014496211742308165,
+      "loss": 2.9795,
+      "step": 41190
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6893475651741028,
+      "learning_rate": 0.00014494893678818174,
+      "loss": 2.7124,
+      "step": 41191
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6525934338569641,
+      "learning_rate": 0.00014493575656165654,
+      "loss": 2.8924,
+      "step": 41192
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7462884783744812,
+      "learning_rate": 0.0001449225767435409,
+      "loss": 2.8665,
+      "step": 41193
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.662916898727417,
+      "learning_rate": 0.00014490939733386915,
+      "loss": 3.0731,
+      "step": 41194
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.639997661113739,
+      "learning_rate": 0.00014489621833267628,
+      "loss": 3.1213,
+      "step": 41195
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6598368883132935,
+      "learning_rate": 0.00014488303973999683,
+      "loss": 2.8662,
+      "step": 41196
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6675333380699158,
+      "learning_rate": 0.0001448698615558656,
+      "loss": 2.9091,
+      "step": 41197
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6575494408607483,
+      "learning_rate": 0.0001448566837803174,
+      "loss": 3.143,
+      "step": 41198
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7101870775222778,
+      "learning_rate": 0.00014484350641338672,
+      "loss": 2.8739,
+      "step": 41199
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7055264711380005,
+      "learning_rate": 0.00014483032945510852,
+      "loss": 3.0787,
+      "step": 41200
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6996819376945496,
+      "learning_rate": 0.00014481715290551734,
+      "loss": 3.099,
+      "step": 41201
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.665325939655304,
+      "learning_rate": 0.0001448039767646478,
+      "loss": 2.913,
+      "step": 41202
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6452702879905701,
+      "learning_rate": 0.00014479080103253487,
+      "loss": 2.9902,
+      "step": 41203
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6679407358169556,
+      "learning_rate": 0.00014477762570921294,
+      "loss": 3.0373,
+      "step": 41204
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6822876930236816,
+      "learning_rate": 0.000144764450794717,
+      "loss": 2.9035,
+      "step": 41205
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6947330236434937,
+      "learning_rate": 0.00014475127628908152,
+      "loss": 3.0485,
+      "step": 41206
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6485691070556641,
+      "learning_rate": 0.00014473810219234142,
+      "loss": 2.6635,
+      "step": 41207
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6676596999168396,
+      "learning_rate": 0.00014472492850453126,
+      "loss": 2.921,
+      "step": 41208
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.676518440246582,
+      "learning_rate": 0.00014471175522568568,
+      "loss": 3.0537,
+      "step": 41209
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6761108040809631,
+      "learning_rate": 0.0001446985823558395,
+      "loss": 2.8734,
+      "step": 41210
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6406627893447876,
+      "learning_rate": 0.0001446854098950273,
+      "loss": 3.074,
+      "step": 41211
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6818718314170837,
+      "learning_rate": 0.00014467223784328382,
+      "loss": 3.0591,
+      "step": 41212
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6850602030754089,
+      "learning_rate": 0.00014465906620064387,
+      "loss": 3.0959,
+      "step": 41213
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6739491820335388,
+      "learning_rate": 0.00014464589496714198,
+      "loss": 2.9633,
+      "step": 41214
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7058461308479309,
+      "learning_rate": 0.00014463272414281297,
+      "loss": 2.8375,
+      "step": 41215
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6772422790527344,
+      "learning_rate": 0.00014461955372769149,
+      "loss": 2.969,
+      "step": 41216
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.64620441198349,
+      "learning_rate": 0.00014460638372181217,
+      "loss": 2.8003,
+      "step": 41217
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6821702718734741,
+      "learning_rate": 0.00014459321412520963,
+      "loss": 2.873,
+      "step": 41218
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6443877220153809,
+      "learning_rate": 0.00014458004493791868,
+      "loss": 2.7943,
+      "step": 41219
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.651509165763855,
+      "learning_rate": 0.00014456687615997408,
+      "loss": 2.7859,
+      "step": 41220
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7200280427932739,
+      "learning_rate": 0.0001445537077914103,
+      "loss": 2.7788,
+      "step": 41221
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.71668940782547,
+      "learning_rate": 0.00014454053983226225,
+      "loss": 2.7504,
+      "step": 41222
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6477024555206299,
+      "learning_rate": 0.00014452737228256448,
+      "loss": 2.8924,
+      "step": 41223
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6733539700508118,
+      "learning_rate": 0.00014451420514235158,
+      "loss": 2.9713,
+      "step": 41224
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6498569250106812,
+      "learning_rate": 0.0001445010384116585,
+      "loss": 2.7993,
+      "step": 41225
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.8216848373413086,
+      "learning_rate": 0.00014448787209051962,
+      "loss": 2.8396,
+      "step": 41226
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7104951739311218,
+      "learning_rate": 0.00014447470617896986,
+      "loss": 2.8338,
+      "step": 41227
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6697823405265808,
+      "learning_rate": 0.00014446154067704366,
+      "loss": 2.9435,
+      "step": 41228
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6788064241409302,
+      "learning_rate": 0.00014444837558477596,
+      "loss": 2.8841,
+      "step": 41229
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6955843567848206,
+      "learning_rate": 0.0001444352109022013,
+      "loss": 2.8783,
+      "step": 41230
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6401929259300232,
+      "learning_rate": 0.00014442204662935425,
+      "loss": 2.9695,
+      "step": 41231
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6757379174232483,
+      "learning_rate": 0.0001444088827662697,
+      "loss": 3.1244,
+      "step": 41232
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6907119154930115,
+      "learning_rate": 0.0001443957193129821,
+      "loss": 2.9758,
+      "step": 41233
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.659367561340332,
+      "learning_rate": 0.00014438255626952624,
+      "loss": 2.9205,
+      "step": 41234
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.685149073600769,
+      "learning_rate": 0.00014436939363593685,
+      "loss": 3.0351,
+      "step": 41235
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6743383407592773,
+      "learning_rate": 0.00014435623141224844,
+      "loss": 2.8732,
+      "step": 41236
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7082371115684509,
+      "learning_rate": 0.00014434306959849587,
+      "loss": 2.9676,
+      "step": 41237
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6763203144073486,
+      "learning_rate": 0.0001443299081947137,
+      "loss": 3.0997,
+      "step": 41238
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7034807205200195,
+      "learning_rate": 0.00014431674720093648,
+      "loss": 3.0158,
+      "step": 41239
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.693721354007721,
+      "learning_rate": 0.00014430358661719908,
+      "loss": 2.9763,
+      "step": 41240
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6822949051856995,
+      "learning_rate": 0.000144290426443536,
+      "loss": 2.9631,
+      "step": 41241
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6607871055603027,
+      "learning_rate": 0.00014427726667998203,
+      "loss": 3.0862,
+      "step": 41242
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6698300838470459,
+      "learning_rate": 0.00014426410732657172,
+      "loss": 2.8321,
+      "step": 41243
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6640063524246216,
+      "learning_rate": 0.00014425094838333986,
+      "loss": 2.9378,
+      "step": 41244
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6912232637405396,
+      "learning_rate": 0.000144237789850321,
+      "loss": 3.0132,
+      "step": 41245
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6797363758087158,
+      "learning_rate": 0.00014422463172754973,
+      "loss": 2.7892,
+      "step": 41246
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7383774518966675,
+      "learning_rate": 0.0001442114740150609,
+      "loss": 2.9162,
+      "step": 41247
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7191385626792908,
+      "learning_rate": 0.000144198316712889,
+      "loss": 2.8664,
+      "step": 41248
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.733830988407135,
+      "learning_rate": 0.0001441851598210687,
+      "loss": 2.8591,
+      "step": 41249
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6828484535217285,
+      "learning_rate": 0.00014417200333963486,
+      "loss": 3.1103,
+      "step": 41250
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6779079437255859,
+      "learning_rate": 0.0001441588472686219,
+      "loss": 2.8456,
+      "step": 41251
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6722716093063354,
+      "learning_rate": 0.0001441456916080646,
+      "loss": 2.9776,
+      "step": 41252
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7105075716972351,
+      "learning_rate": 0.00014413253635799743,
+      "loss": 3.1086,
+      "step": 41253
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6589248180389404,
+      "learning_rate": 0.00014411938151845518,
+      "loss": 3.064,
+      "step": 41254
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6822506785392761,
+      "learning_rate": 0.00014410622708947257,
+      "loss": 2.9046,
+      "step": 41255
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7151714563369751,
+      "learning_rate": 0.00014409307307108405,
+      "loss": 2.8397,
+      "step": 41256
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.684058427810669,
+      "learning_rate": 0.0001440799194633245,
+      "loss": 2.7317,
+      "step": 41257
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6461659073829651,
+      "learning_rate": 0.00014406676626622827,
+      "loss": 3.0171,
+      "step": 41258
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6536449790000916,
+      "learning_rate": 0.00014405361347983033,
+      "loss": 3.0362,
+      "step": 41259
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6458353400230408,
+      "learning_rate": 0.00014404046110416517,
+      "loss": 2.7066,
+      "step": 41260
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7059634923934937,
+      "learning_rate": 0.00014402730913926724,
+      "loss": 2.8764,
+      "step": 41261
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6733324527740479,
+      "learning_rate": 0.0001440141575851715,
+      "loss": 2.9203,
+      "step": 41262
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6528270244598389,
+      "learning_rate": 0.0001440010064419124,
+      "loss": 2.8455,
+      "step": 41263
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6702911257743835,
+      "learning_rate": 0.00014398785570952458,
+      "loss": 2.8677,
+      "step": 41264
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.671634316444397,
+      "learning_rate": 0.00014397470538804282,
+      "loss": 3.0517,
+      "step": 41265
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7339351177215576,
+      "learning_rate": 0.00014396155547750166,
+      "loss": 2.8713,
+      "step": 41266
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6775936484336853,
+      "learning_rate": 0.00014394840597793576,
+      "loss": 2.8699,
+      "step": 41267
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6924892663955688,
+      "learning_rate": 0.0001439352568893796,
+      "loss": 3.0556,
+      "step": 41268
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6678718328475952,
+      "learning_rate": 0.0001439221082118679,
+      "loss": 3.0232,
+      "step": 41269
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7233895063400269,
+      "learning_rate": 0.0001439089599454355,
+      "loss": 2.8655,
+      "step": 41270
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6184765100479126,
+      "learning_rate": 0.0001438958120901167,
+      "loss": 3.0533,
+      "step": 41271
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7089718580245972,
+      "learning_rate": 0.00014388266464594645,
+      "loss": 3.1122,
+      "step": 41272
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6720311045646667,
+      "learning_rate": 0.0001438695176129592,
+      "loss": 2.9664,
+      "step": 41273
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.66576087474823,
+      "learning_rate": 0.00014385637099118944,
+      "loss": 2.8472,
+      "step": 41274
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6577184200286865,
+      "learning_rate": 0.0001438432247806721,
+      "loss": 2.7898,
+      "step": 41275
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6298528909683228,
+      "learning_rate": 0.00014383007898144151,
+      "loss": 2.9125,
+      "step": 41276
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6644882559776306,
+      "learning_rate": 0.0001438169335935326,
+      "loss": 2.9001,
+      "step": 41277
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6496797800064087,
+      "learning_rate": 0.00014380378861697968,
+      "loss": 2.8136,
+      "step": 41278
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6571464538574219,
+      "learning_rate": 0.00014379064405181752,
+      "loss": 2.9994,
+      "step": 41279
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6770309209823608,
+      "learning_rate": 0.00014377749989808087,
+      "loss": 2.9402,
+      "step": 41280
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6546334028244019,
+      "learning_rate": 0.0001437643561558042,
+      "loss": 2.9614,
+      "step": 41281
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6517234444618225,
+      "learning_rate": 0.00014375121282502216,
+      "loss": 2.8285,
+      "step": 41282
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6837851405143738,
+      "learning_rate": 0.00014373806990576924,
+      "loss": 3.1049,
+      "step": 41283
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6652602553367615,
+      "learning_rate": 0.00014372492739808017,
+      "loss": 2.958,
+      "step": 41284
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6891589760780334,
+      "learning_rate": 0.0001437117853019897,
+      "loss": 3.0539,
+      "step": 41285
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6521453857421875,
+      "learning_rate": 0.00014369864361753217,
+      "loss": 2.7239,
+      "step": 41286
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.642013669013977,
+      "learning_rate": 0.00014368550234474244,
+      "loss": 3.1218,
+      "step": 41287
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6464647650718689,
+      "learning_rate": 0.00014367236148365504,
+      "loss": 2.8118,
+      "step": 41288
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6746804714202881,
+      "learning_rate": 0.0001436592210343044,
+      "loss": 2.9244,
+      "step": 41289
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6726866364479065,
+      "learning_rate": 0.00014364608099672543,
+      "loss": 2.8907,
+      "step": 41290
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.686701774597168,
+      "learning_rate": 0.00014363294137095247,
+      "loss": 2.9126,
+      "step": 41291
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6961394548416138,
+      "learning_rate": 0.00014361980215702035,
+      "loss": 3.023,
+      "step": 41292
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.662632167339325,
+      "learning_rate": 0.0001436066633549635,
+      "loss": 2.9869,
+      "step": 41293
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6637516617774963,
+      "learning_rate": 0.00014359352496481667,
+      "loss": 2.8747,
+      "step": 41294
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6575713753700256,
+      "learning_rate": 0.00014358038698661442,
+      "loss": 2.8826,
+      "step": 41295
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6427962779998779,
+      "learning_rate": 0.00014356724942039118,
+      "loss": 3.012,
+      "step": 41296
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6798694729804993,
+      "learning_rate": 0.00014355411226618184,
+      "loss": 2.9795,
+      "step": 41297
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.680004894733429,
+      "learning_rate": 0.00014354097552402077,
+      "loss": 3.1077,
+      "step": 41298
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6552472114562988,
+      "learning_rate": 0.00014352783919394273,
+      "loss": 2.9108,
+      "step": 41299
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6746847629547119,
+      "learning_rate": 0.00014351470327598215,
+      "loss": 2.8963,
+      "step": 41300
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6662331223487854,
+      "learning_rate": 0.0001435015677701737,
+      "loss": 2.7933,
+      "step": 41301
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6595388650894165,
+      "learning_rate": 0.00014348843267655219,
+      "loss": 2.7368,
+      "step": 41302
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6785330176353455,
+      "learning_rate": 0.00014347529799515197,
+      "loss": 2.8712,
+      "step": 41303
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6657465696334839,
+      "learning_rate": 0.00014346216372600768,
+      "loss": 2.7114,
+      "step": 41304
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6740917563438416,
+      "learning_rate": 0.00014344902986915386,
+      "loss": 2.8561,
+      "step": 41305
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6666470170021057,
+      "learning_rate": 0.00014343589642462512,
+      "loss": 2.9495,
+      "step": 41306
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.71805739402771,
+      "learning_rate": 0.00014342276339245622,
+      "loss": 3.068,
+      "step": 41307
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7400341033935547,
+      "learning_rate": 0.00014340963077268153,
+      "loss": 3.0277,
+      "step": 41308
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6645949482917786,
+      "learning_rate": 0.00014339649856533583,
+      "loss": 2.9738,
+      "step": 41309
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6728555560112,
+      "learning_rate": 0.00014338336677045365,
+      "loss": 3.0966,
+      "step": 41310
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6942307353019714,
+      "learning_rate": 0.0001433702353880694,
+      "loss": 2.7839,
+      "step": 41311
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.705938458442688,
+      "learning_rate": 0.0001433571044182179,
+      "loss": 2.6229,
+      "step": 41312
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6518526077270508,
+      "learning_rate": 0.00014334397386093356,
+      "loss": 2.8772,
+      "step": 41313
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7186595797538757,
+      "learning_rate": 0.00014333084371625116,
+      "loss": 2.8394,
+      "step": 41314
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6730000376701355,
+      "learning_rate": 0.00014331771398420502,
+      "loss": 2.9375,
+      "step": 41315
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.672404408454895,
+      "learning_rate": 0.00014330458466482986,
+      "loss": 2.8477,
+      "step": 41316
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6947441697120667,
+      "learning_rate": 0.00014329145575816044,
+      "loss": 2.8878,
+      "step": 41317
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7594353556632996,
+      "learning_rate": 0.00014327832726423114,
+      "loss": 3.0445,
+      "step": 41318
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6902020573616028,
+      "learning_rate": 0.00014326519918307653,
+      "loss": 3.1234,
+      "step": 41319
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.641372799873352,
+      "learning_rate": 0.00014325207151473114,
+      "loss": 2.9663,
+      "step": 41320
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6673992276191711,
+      "learning_rate": 0.00014323894425922963,
+      "loss": 2.906,
+      "step": 41321
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6893587708473206,
+      "learning_rate": 0.00014322581741660668,
+      "loss": 2.7704,
+      "step": 41322
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6781730055809021,
+      "learning_rate": 0.00014321269098689664,
+      "loss": 2.6757,
+      "step": 41323
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6808968782424927,
+      "learning_rate": 0.0001431995649701343,
+      "loss": 3.0846,
+      "step": 41324
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.650240421295166,
+      "learning_rate": 0.00014318643936635415,
+      "loss": 2.9279,
+      "step": 41325
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6758670210838318,
+      "learning_rate": 0.0001431733141755906,
+      "loss": 2.9108,
+      "step": 41326
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6735976934432983,
+      "learning_rate": 0.0001431601893978785,
+      "loss": 3.1225,
+      "step": 41327
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6825075745582581,
+      "learning_rate": 0.00014314706503325213,
+      "loss": 2.8253,
+      "step": 41328
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6795275211334229,
+      "learning_rate": 0.00014313394108174637,
+      "loss": 3.0741,
+      "step": 41329
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6708423495292664,
+      "learning_rate": 0.00014312081754339545,
+      "loss": 2.9368,
+      "step": 41330
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7268308997154236,
+      "learning_rate": 0.00014310769441823424,
+      "loss": 2.8793,
+      "step": 41331
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.662429690361023,
+      "learning_rate": 0.00014309457170629718,
+      "loss": 2.9008,
+      "step": 41332
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6782928109169006,
+      "learning_rate": 0.00014308144940761868,
+      "loss": 2.9867,
+      "step": 41333
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6727951765060425,
+      "learning_rate": 0.00014306832752223358,
+      "loss": 3.1139,
+      "step": 41334
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6834380626678467,
+      "learning_rate": 0.00014305520605017617,
+      "loss": 3.0817,
+      "step": 41335
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6775288581848145,
+      "learning_rate": 0.00014304208499148118,
+      "loss": 3.0145,
+      "step": 41336
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6631368398666382,
+      "learning_rate": 0.00014302896434618322,
+      "loss": 3.1274,
+      "step": 41337
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.65370112657547,
+      "learning_rate": 0.00014301584411431664,
+      "loss": 3.0237,
+      "step": 41338
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6937180757522583,
+      "learning_rate": 0.00014300272429591625,
+      "loss": 3.0834,
+      "step": 41339
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6696368455886841,
+      "learning_rate": 0.00014298960489101642,
+      "loss": 3.1994,
+      "step": 41340
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6452273726463318,
+      "learning_rate": 0.00014297648589965169,
+      "loss": 2.9717,
+      "step": 41341
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6574122309684753,
+      "learning_rate": 0.0001429633673218568,
+      "loss": 2.9185,
+      "step": 41342
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6667985320091248,
+      "learning_rate": 0.000142950249157666,
+      "loss": 2.9371,
+      "step": 41343
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6507903337478638,
+      "learning_rate": 0.00014293713140711416,
+      "loss": 2.9564,
+      "step": 41344
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6721150875091553,
+      "learning_rate": 0.0001429240140702356,
+      "loss": 2.9744,
+      "step": 41345
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6516289710998535,
+      "learning_rate": 0.00014291089714706505,
+      "loss": 2.9591,
+      "step": 41346
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6478187441825867,
+      "learning_rate": 0.00014289778063763698,
+      "loss": 2.8334,
+      "step": 41347
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6945951581001282,
+      "learning_rate": 0.0001428846645419858,
+      "loss": 3.1982,
+      "step": 41348
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7314687967300415,
+      "learning_rate": 0.00014287154886014628,
+      "loss": 2.7935,
+      "step": 41349
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7077362537384033,
+      "learning_rate": 0.00014285843359215274,
+      "loss": 2.9902,
+      "step": 41350
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.676292359828949,
+      "learning_rate": 0.00014284531873803987,
+      "loss": 3.0049,
+      "step": 41351
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6695364117622375,
+      "learning_rate": 0.00014283220429784228,
+      "loss": 3.1166,
+      "step": 41352
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7214575409889221,
+      "learning_rate": 0.00014281909027159443,
+      "loss": 2.924,
+      "step": 41353
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6631299257278442,
+      "learning_rate": 0.00014280597665933083,
+      "loss": 3.1442,
+      "step": 41354
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6750709414482117,
+      "learning_rate": 0.00014279286346108593,
+      "loss": 3.065,
+      "step": 41355
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7027854323387146,
+      "learning_rate": 0.0001427797506768944,
+      "loss": 3.1048,
+      "step": 41356
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6556305885314941,
+      "learning_rate": 0.00014276663830679081,
+      "loss": 2.944,
+      "step": 41357
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7109807729721069,
+      "learning_rate": 0.00014275352635080956,
+      "loss": 2.8792,
+      "step": 41358
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7511147260665894,
+      "learning_rate": 0.00014274041480898534,
+      "loss": 2.8699,
+      "step": 41359
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.665418803691864,
+      "learning_rate": 0.0001427273036813525,
+      "loss": 2.9801,
+      "step": 41360
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7250953316688538,
+      "learning_rate": 0.0001427141929679458,
+      "loss": 3.0798,
+      "step": 41361
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6798324584960938,
+      "learning_rate": 0.00014270108266879964,
+      "loss": 3.0348,
+      "step": 41362
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6868711113929749,
+      "learning_rate": 0.00014268797278394843,
+      "loss": 2.9232,
+      "step": 41363
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.683142900466919,
+      "learning_rate": 0.00014267486331342695,
+      "loss": 2.918,
+      "step": 41364
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6508060097694397,
+      "learning_rate": 0.00014266175425726952,
+      "loss": 2.8509,
+      "step": 41365
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6772064566612244,
+      "learning_rate": 0.0001426486456155107,
+      "loss": 2.9454,
+      "step": 41366
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6569616794586182,
+      "learning_rate": 0.00014263553738818517,
+      "loss": 2.8806,
+      "step": 41367
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6555266380310059,
+      "learning_rate": 0.0001426224295753274,
+      "loss": 2.7536,
+      "step": 41368
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6969777345657349,
+      "learning_rate": 0.00014260932217697182,
+      "loss": 3.0595,
+      "step": 41369
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6618245840072632,
+      "learning_rate": 0.00014259621519315292,
+      "loss": 2.9002,
+      "step": 41370
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6819080710411072,
+      "learning_rate": 0.0001425831086239054,
+      "loss": 2.9353,
+      "step": 41371
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.8382446765899658,
+      "learning_rate": 0.00014257000246926352,
+      "loss": 2.6845,
+      "step": 41372
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6515592336654663,
+      "learning_rate": 0.00014255689672926198,
+      "loss": 3.1173,
+      "step": 41373
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6922911405563354,
+      "learning_rate": 0.0001425437914039354,
+      "loss": 2.8563,
+      "step": 41374
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6700411438941956,
+      "learning_rate": 0.00014253068649331817,
+      "loss": 3.0166,
+      "step": 41375
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.666197657585144,
+      "learning_rate": 0.0001425175819974448,
+      "loss": 2.7031,
+      "step": 41376
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7301174402236938,
+      "learning_rate": 0.0001425044779163497,
+      "loss": 2.8902,
+      "step": 41377
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6734825968742371,
+      "learning_rate": 0.0001424913742500675,
+      "loss": 3.0026,
+      "step": 41378
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7192350625991821,
+      "learning_rate": 0.0001424782709986328,
+      "loss": 3.0602,
+      "step": 41379
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7443360090255737,
+      "learning_rate": 0.00014246516816207995,
+      "loss": 2.7758,
+      "step": 41380
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6618134379386902,
+      "learning_rate": 0.00014245206574044358,
+      "loss": 2.6856,
+      "step": 41381
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7043996453285217,
+      "learning_rate": 0.00014243896373375807,
+      "loss": 2.8153,
+      "step": 41382
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7315269708633423,
+      "learning_rate": 0.00014242586214205814,
+      "loss": 2.9694,
+      "step": 41383
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6584864258766174,
+      "learning_rate": 0.0001424127609653781,
+      "loss": 2.7616,
+      "step": 41384
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6800797581672668,
+      "learning_rate": 0.00014239966020375246,
+      "loss": 2.8515,
+      "step": 41385
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7683088183403015,
+      "learning_rate": 0.0001423865598572159,
+      "loss": 2.9446,
+      "step": 41386
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6943177580833435,
+      "learning_rate": 0.00014237345992580266,
+      "loss": 3.2236,
+      "step": 41387
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7042199969291687,
+      "learning_rate": 0.0001423603604095474,
+      "loss": 2.973,
+      "step": 41388
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6742187738418579,
+      "learning_rate": 0.00014234726130848472,
+      "loss": 2.6971,
+      "step": 41389
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6640881299972534,
+      "learning_rate": 0.00014233416262264903,
+      "loss": 2.7863,
+      "step": 41390
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6789493560791016,
+      "learning_rate": 0.0001423210643520748,
+      "loss": 2.9233,
+      "step": 41391
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6761623620986938,
+      "learning_rate": 0.00014230796649679643,
+      "loss": 3.1005,
+      "step": 41392
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6618014574050903,
+      "learning_rate": 0.00014229486905684854,
+      "loss": 3.047,
+      "step": 41393
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7283720970153809,
+      "learning_rate": 0.00014228177203226574,
+      "loss": 3.1616,
+      "step": 41394
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6539528965950012,
+      "learning_rate": 0.00014226867542308226,
+      "loss": 2.833,
+      "step": 41395
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6887137293815613,
+      "learning_rate": 0.00014225557922933286,
+      "loss": 3.0419,
+      "step": 41396
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6629752516746521,
+      "learning_rate": 0.00014224248345105192,
+      "loss": 2.8952,
+      "step": 41397
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6697224378585815,
+      "learning_rate": 0.00014222938808827377,
+      "loss": 3.0586,
+      "step": 41398
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6652210354804993,
+      "learning_rate": 0.00014221629314103317,
+      "loss": 3.0246,
+      "step": 41399
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6590397953987122,
+      "learning_rate": 0.0001422031986093644,
+      "loss": 2.7692,
+      "step": 41400
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7004410028457642,
+      "learning_rate": 0.0001421901044933021,
+      "loss": 2.9138,
+      "step": 41401
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7048691511154175,
+      "learning_rate": 0.00014217701079288062,
+      "loss": 2.7807,
+      "step": 41402
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7057774066925049,
+      "learning_rate": 0.00014216391750813452,
+      "loss": 2.8305,
+      "step": 41403
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6988634467124939,
+      "learning_rate": 0.00014215082463909837,
+      "loss": 2.8967,
+      "step": 41404
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.644032895565033,
+      "learning_rate": 0.00014213773218580658,
+      "loss": 2.9624,
+      "step": 41405
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.659142792224884,
+      "learning_rate": 0.00014212464014829362,
+      "loss": 3.1234,
+      "step": 41406
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7245345115661621,
+      "learning_rate": 0.00014211154852659388,
+      "loss": 3.121,
+      "step": 41407
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6617531180381775,
+      "learning_rate": 0.0001420984573207419,
+      "loss": 2.9157,
+      "step": 41408
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6753281354904175,
+      "learning_rate": 0.00014208536653077236,
+      "loss": 2.8922,
+      "step": 41409
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7002780437469482,
+      "learning_rate": 0.00014207227615671943,
+      "loss": 3.0797,
+      "step": 41410
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7127296328544617,
+      "learning_rate": 0.00014205918619861785,
+      "loss": 3.0361,
+      "step": 41411
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6621199250221252,
+      "learning_rate": 0.00014204609665650198,
+      "loss": 2.8687,
+      "step": 41412
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6912297010421753,
+      "learning_rate": 0.00014203300753040616,
+      "loss": 2.8506,
+      "step": 41413
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6624032258987427,
+      "learning_rate": 0.00014201991882036513,
+      "loss": 2.7342,
+      "step": 41414
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7326284646987915,
+      "learning_rate": 0.00014200683052641307,
+      "loss": 2.9032,
+      "step": 41415
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6524210572242737,
+      "learning_rate": 0.00014199374264858478,
+      "loss": 3.0304,
+      "step": 41416
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6589338183403015,
+      "learning_rate": 0.00014198065518691443,
+      "loss": 2.9891,
+      "step": 41417
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7966728806495667,
+      "learning_rate": 0.00014196756814143665,
+      "loss": 2.8156,
+      "step": 41418
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6796247363090515,
+      "learning_rate": 0.00014195448151218607,
+      "loss": 2.8564,
+      "step": 41419
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.9536327123641968,
+      "learning_rate": 0.00014194139529919675,
+      "loss": 3.0352,
+      "step": 41420
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7161813378334045,
+      "learning_rate": 0.0001419283095025035,
+      "loss": 2.9973,
+      "step": 41421
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7191630005836487,
+      "learning_rate": 0.00014191522412214055,
+      "loss": 2.8155,
+      "step": 41422
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6631558537483215,
+      "learning_rate": 0.00014190213915814246,
+      "loss": 2.923,
+      "step": 41423
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7255247235298157,
+      "learning_rate": 0.00014188905461054384,
+      "loss": 2.9193,
+      "step": 41424
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.733752965927124,
+      "learning_rate": 0.0001418759704793789,
+      "loss": 2.7613,
+      "step": 41425
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7541595697402954,
+      "learning_rate": 0.00014186288676468236,
+      "loss": 2.9079,
+      "step": 41426
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6659170389175415,
+      "learning_rate": 0.00014184980346648848,
+      "loss": 2.9627,
+      "step": 41427
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6909793019294739,
+      "learning_rate": 0.00014183672058483174,
+      "loss": 2.79,
+      "step": 41428
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6981406807899475,
+      "learning_rate": 0.00014182363811974667,
+      "loss": 3.0426,
+      "step": 41429
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6924818754196167,
+      "learning_rate": 0.00014181055607126764,
+      "loss": 2.6948,
+      "step": 41430
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6846038699150085,
+      "learning_rate": 0.00014179747443942927,
+      "loss": 2.9601,
+      "step": 41431
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.722434937953949,
+      "learning_rate": 0.0001417843932242658,
+      "loss": 2.7735,
+      "step": 41432
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6762760877609253,
+      "learning_rate": 0.0001417713124258119,
+      "loss": 3.0094,
+      "step": 41433
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7446743845939636,
+      "learning_rate": 0.0001417582320441019,
+      "loss": 3.1196,
+      "step": 41434
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6979431509971619,
+      "learning_rate": 0.00014174515207917014,
+      "loss": 3.1234,
+      "step": 41435
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6926884055137634,
+      "learning_rate": 0.0001417320725310513,
+      "loss": 3.0273,
+      "step": 41436
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6868926286697388,
+      "learning_rate": 0.00014171899339977963,
+      "loss": 2.8642,
+      "step": 41437
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6802986860275269,
+      "learning_rate": 0.00014170591468538966,
+      "loss": 2.9907,
+      "step": 41438
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7366862893104553,
+      "learning_rate": 0.00014169283638791597,
+      "loss": 2.7683,
+      "step": 41439
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6840645670890808,
+      "learning_rate": 0.00014167975850739277,
+      "loss": 3.0044,
+      "step": 41440
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6888478398323059,
+      "learning_rate": 0.00014166668104385473,
+      "loss": 3.0112,
+      "step": 41441
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6686787009239197,
+      "learning_rate": 0.00014165360399733616,
+      "loss": 2.9337,
+      "step": 41442
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6644048690795898,
+      "learning_rate": 0.0001416405273678714,
+      "loss": 3.073,
+      "step": 41443
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6644816994667053,
+      "learning_rate": 0.0001416274511554952,
+      "loss": 2.9437,
+      "step": 41444
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6657946705818176,
+      "learning_rate": 0.00014161437536024162,
+      "loss": 2.9644,
+      "step": 41445
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6964468955993652,
+      "learning_rate": 0.00014160129998214545,
+      "loss": 3.0717,
+      "step": 41446
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7129657864570618,
+      "learning_rate": 0.00014158822502124084,
+      "loss": 3.0621,
+      "step": 41447
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6289086937904358,
+      "learning_rate": 0.00014157515047756247,
+      "loss": 2.828,
+      "step": 41448
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6725035905838013,
+      "learning_rate": 0.00014156207635114467,
+      "loss": 2.9882,
+      "step": 41449
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6678059101104736,
+      "learning_rate": 0.00014154900264202176,
+      "loss": 2.9306,
+      "step": 41450
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6416679620742798,
+      "learning_rate": 0.0001415359293502284,
+      "loss": 2.7726,
+      "step": 41451
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6881377100944519,
+      "learning_rate": 0.00014152285647579878,
+      "loss": 2.6733,
+      "step": 41452
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6680883765220642,
+      "learning_rate": 0.00014150978401876755,
+      "loss": 2.9299,
+      "step": 41453
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6864650249481201,
+      "learning_rate": 0.000141496711979169,
+      "loss": 2.9543,
+      "step": 41454
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7557768225669861,
+      "learning_rate": 0.00014148364035703763,
+      "loss": 3.0048,
+      "step": 41455
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6464671492576599,
+      "learning_rate": 0.0001414705691524079,
+      "loss": 2.8257,
+      "step": 41456
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6590653657913208,
+      "learning_rate": 0.00014145749836531405,
+      "loss": 2.8427,
+      "step": 41457
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6637300252914429,
+      "learning_rate": 0.00014144442799579075,
+      "loss": 2.8594,
+      "step": 41458
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7155690789222717,
+      "learning_rate": 0.00014143135804387221,
+      "loss": 2.8372,
+      "step": 41459
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6667417287826538,
+      "learning_rate": 0.00014141828850959296,
+      "loss": 2.8662,
+      "step": 41460
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7127690315246582,
+      "learning_rate": 0.00014140521939298755,
+      "loss": 2.9396,
+      "step": 41461
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6322091817855835,
+      "learning_rate": 0.00014139215069409013,
+      "loss": 2.8889,
+      "step": 41462
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6614790558815002,
+      "learning_rate": 0.00014137908241293538,
+      "loss": 2.8967,
+      "step": 41463
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6741194128990173,
+      "learning_rate": 0.0001413660145495576,
+      "loss": 2.9764,
+      "step": 41464
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6606133580207825,
+      "learning_rate": 0.0001413529471039911,
+      "loss": 2.7991,
+      "step": 41465
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6671067476272583,
+      "learning_rate": 0.0001413398800762705,
+      "loss": 3.006,
+      "step": 41466
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6627110242843628,
+      "learning_rate": 0.00014132681346643,
+      "loss": 2.884,
+      "step": 41467
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7146353721618652,
+      "learning_rate": 0.00014131374727450427,
+      "loss": 2.889,
+      "step": 41468
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6611478924751282,
+      "learning_rate": 0.0001413006815005275,
+      "loss": 2.8787,
+      "step": 41469
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6627159118652344,
+      "learning_rate": 0.00014128761614453424,
+      "loss": 3.0298,
+      "step": 41470
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6639991998672485,
+      "learning_rate": 0.0001412745512065589,
+      "loss": 2.9859,
+      "step": 41471
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6625120639801025,
+      "learning_rate": 0.0001412614866866357,
+      "loss": 2.9387,
+      "step": 41472
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6913546919822693,
+      "learning_rate": 0.00014124842258479936,
+      "loss": 2.8833,
+      "step": 41473
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6938081979751587,
+      "learning_rate": 0.00014123535890108398,
+      "loss": 2.8589,
+      "step": 41474
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6782478094100952,
+      "learning_rate": 0.0001412222956355241,
+      "loss": 3.0988,
+      "step": 41475
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6729618906974792,
+      "learning_rate": 0.00014120923278815428,
+      "loss": 2.9759,
+      "step": 41476
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.670555055141449,
+      "learning_rate": 0.00014119617035900875,
+      "loss": 2.8674,
+      "step": 41477
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7138615250587463,
+      "learning_rate": 0.00014118310834812194,
+      "loss": 2.9513,
+      "step": 41478
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6860772967338562,
+      "learning_rate": 0.00014117004675552817,
+      "loss": 2.845,
+      "step": 41479
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7426106333732605,
+      "learning_rate": 0.00014115698558126195,
+      "loss": 2.8868,
+      "step": 41480
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6855692267417908,
+      "learning_rate": 0.00014114392482535772,
+      "loss": 2.9,
+      "step": 41481
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7088014483451843,
+      "learning_rate": 0.00014113086448784974,
+      "loss": 2.9,
+      "step": 41482
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6505365967750549,
+      "learning_rate": 0.0001411178045687726,
+      "loss": 3.0525,
+      "step": 41483
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.702044665813446,
+      "learning_rate": 0.00014110474506816043,
+      "loss": 2.9941,
+      "step": 41484
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6758970022201538,
+      "learning_rate": 0.00014109168598604795,
+      "loss": 3.1789,
+      "step": 41485
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6437978744506836,
+      "learning_rate": 0.00014107862732246935,
+      "loss": 2.6538,
+      "step": 41486
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7227490544319153,
+      "learning_rate": 0.00014106556907745896,
+      "loss": 2.884,
+      "step": 41487
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7027838826179504,
+      "learning_rate": 0.00014105251125105139,
+      "loss": 3.0361,
+      "step": 41488
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6415818333625793,
+      "learning_rate": 0.0001410394538432808,
+      "loss": 2.9468,
+      "step": 41489
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6897194981575012,
+      "learning_rate": 0.0001410263968541817,
+      "loss": 2.9501,
+      "step": 41490
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6428824663162231,
+      "learning_rate": 0.0001410133402837886,
+      "loss": 2.9695,
+      "step": 41491
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7042983770370483,
+      "learning_rate": 0.00014100028413213572,
+      "loss": 2.9618,
+      "step": 41492
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7326884269714355,
+      "learning_rate": 0.00014098722839925754,
+      "loss": 3.2218,
+      "step": 41493
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6782544255256653,
+      "learning_rate": 0.00014097417308518825,
+      "loss": 2.9167,
+      "step": 41494
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6886657476425171,
+      "learning_rate": 0.0001409611181899624,
+      "loss": 2.8998,
+      "step": 41495
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7487680912017822,
+      "learning_rate": 0.00014094806371361447,
+      "loss": 2.9263,
+      "step": 41496
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.654858410358429,
+      "learning_rate": 0.0001409350096561786,
+      "loss": 3.0155,
+      "step": 41497
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6685404181480408,
+      "learning_rate": 0.00014092195601768945,
+      "loss": 3.0298,
+      "step": 41498
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7062793970108032,
+      "learning_rate": 0.0001409089027981812,
+      "loss": 2.883,
+      "step": 41499
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7158371210098267,
+      "learning_rate": 0.00014089584999768821,
+      "loss": 2.9065,
+      "step": 41500
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7162919640541077,
+      "learning_rate": 0.00014088279761624505,
+      "loss": 3.1325,
+      "step": 41501
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6152122616767883,
+      "learning_rate": 0.00014086974565388587,
+      "loss": 2.8914,
+      "step": 41502
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7153500318527222,
+      "learning_rate": 0.00014085669411064523,
+      "loss": 3.1672,
+      "step": 41503
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6505460143089294,
+      "learning_rate": 0.00014084364298655734,
+      "loss": 3.0139,
+      "step": 41504
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7156624794006348,
+      "learning_rate": 0.00014083059228165665,
+      "loss": 2.918,
+      "step": 41505
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.657463550567627,
+      "learning_rate": 0.00014081754199597767,
+      "loss": 3.0337,
+      "step": 41506
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6673892736434937,
+      "learning_rate": 0.00014080449212955468,
+      "loss": 2.8168,
+      "step": 41507
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6680115461349487,
+      "learning_rate": 0.00014079144268242197,
+      "loss": 2.916,
+      "step": 41508
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.676265299320221,
+      "learning_rate": 0.00014077839365461386,
+      "loss": 3.0468,
+      "step": 41509
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6526424288749695,
+      "learning_rate": 0.00014076534504616484,
+      "loss": 2.9184,
+      "step": 41510
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.695165753364563,
+      "learning_rate": 0.00014075229685710935,
+      "loss": 3.1055,
+      "step": 41511
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6782224178314209,
+      "learning_rate": 0.00014073924908748154,
+      "loss": 2.893,
+      "step": 41512
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6610283255577087,
+      "learning_rate": 0.00014072620173731604,
+      "loss": 3.0361,
+      "step": 41513
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6884645223617554,
+      "learning_rate": 0.00014071315480664704,
+      "loss": 3.005,
+      "step": 41514
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7342187166213989,
+      "learning_rate": 0.00014070010829550885,
+      "loss": 2.9996,
+      "step": 41515
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6680518388748169,
+      "learning_rate": 0.000140687062203936,
+      "loss": 3.1241,
+      "step": 41516
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6695300936698914,
+      "learning_rate": 0.00014067401653196268,
+      "loss": 3.0162,
+      "step": 41517
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6923040747642517,
+      "learning_rate": 0.00014066097127962346,
+      "loss": 2.6873,
+      "step": 41518
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7008634209632874,
+      "learning_rate": 0.00014064792644695242,
+      "loss": 2.9065,
+      "step": 41519
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6829542517662048,
+      "learning_rate": 0.0001406348820339841,
+      "loss": 2.8814,
+      "step": 41520
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6533739566802979,
+      "learning_rate": 0.00014062183804075307,
+      "loss": 2.969,
+      "step": 41521
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6787754893302917,
+      "learning_rate": 0.0001406087944672932,
+      "loss": 2.8614,
+      "step": 41522
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6453655958175659,
+      "learning_rate": 0.00014059575131363918,
+      "loss": 2.9535,
+      "step": 41523
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.733506441116333,
+      "learning_rate": 0.00014058270857982523,
+      "loss": 2.9048,
+      "step": 41524
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6687877774238586,
+      "learning_rate": 0.00014056966626588578,
+      "loss": 2.7681,
+      "step": 41525
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6705677509307861,
+      "learning_rate": 0.0001405566243718551,
+      "loss": 3.2083,
+      "step": 41526
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6572078466415405,
+      "learning_rate": 0.00014054358289776754,
+      "loss": 2.8929,
+      "step": 41527
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6767847537994385,
+      "learning_rate": 0.0001405305418436576,
+      "loss": 2.8771,
+      "step": 41528
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6408040523529053,
+      "learning_rate": 0.00014051750120955956,
+      "loss": 2.841,
+      "step": 41529
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6896417140960693,
+      "learning_rate": 0.00014050446099550773,
+      "loss": 2.8297,
+      "step": 41530
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7329106330871582,
+      "learning_rate": 0.00014049142120153632,
+      "loss": 3.0008,
+      "step": 41531
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6867200136184692,
+      "learning_rate": 0.0001404783818276798,
+      "loss": 2.6561,
+      "step": 41532
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6792352199554443,
+      "learning_rate": 0.00014046534287397264,
+      "loss": 3.0706,
+      "step": 41533
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7385411858558655,
+      "learning_rate": 0.00014045230434044895,
+      "loss": 2.5935,
+      "step": 41534
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.662626326084137,
+      "learning_rate": 0.00014043926622714332,
+      "loss": 2.9917,
+      "step": 41535
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.734079897403717,
+      "learning_rate": 0.00014042622853408993,
+      "loss": 3.1046,
+      "step": 41536
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.646392822265625,
+      "learning_rate": 0.0001404131912613231,
+      "loss": 2.8517,
+      "step": 41537
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6550691723823547,
+      "learning_rate": 0.00014040015440887723,
+      "loss": 3.0403,
+      "step": 41538
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7334123849868774,
+      "learning_rate": 0.00014038711797678658,
+      "loss": 2.8179,
+      "step": 41539
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6451386213302612,
+      "learning_rate": 0.00014037408196508562,
+      "loss": 2.9276,
+      "step": 41540
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6706496477127075,
+      "learning_rate": 0.00014036104637380853,
+      "loss": 2.8136,
+      "step": 41541
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6933931112289429,
+      "learning_rate": 0.00014034801120298967,
+      "loss": 3.1359,
+      "step": 41542
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6907314658164978,
+      "learning_rate": 0.00014033497645266356,
+      "loss": 2.8287,
+      "step": 41543
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6742256879806519,
+      "learning_rate": 0.00014032194212286442,
+      "loss": 2.9517,
+      "step": 41544
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6563650369644165,
+      "learning_rate": 0.00014030890821362655,
+      "loss": 2.9116,
+      "step": 41545
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.665621280670166,
+      "learning_rate": 0.00014029587472498415,
+      "loss": 2.9368,
+      "step": 41546
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6880399584770203,
+      "learning_rate": 0.00014028284165697167,
+      "loss": 2.9152,
+      "step": 41547
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7508476376533508,
+      "learning_rate": 0.00014026980900962358,
+      "loss": 3.043,
+      "step": 41548
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7052724957466125,
+      "learning_rate": 0.00014025677678297394,
+      "loss": 3.0261,
+      "step": 41549
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.698532223701477,
+      "learning_rate": 0.00014024374497705736,
+      "loss": 3.0202,
+      "step": 41550
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6969287395477295,
+      "learning_rate": 0.00014023071359190795,
+      "loss": 2.8709,
+      "step": 41551
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7008760571479797,
+      "learning_rate": 0.00014021768262756002,
+      "loss": 2.9338,
+      "step": 41552
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6544531583786011,
+      "learning_rate": 0.00014020465208404808,
+      "loss": 2.8513,
+      "step": 41553
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.659297525882721,
+      "learning_rate": 0.00014019162196140622,
+      "loss": 2.8528,
+      "step": 41554
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6890259385108948,
+      "learning_rate": 0.00014017859225966896,
+      "loss": 2.9265,
+      "step": 41555
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6693836450576782,
+      "learning_rate": 0.00014016556297887043,
+      "loss": 2.9833,
+      "step": 41556
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7449944615364075,
+      "learning_rate": 0.00014015253411904519,
+      "loss": 2.8482,
+      "step": 41557
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6576284766197205,
+      "learning_rate": 0.00014013950568022735,
+      "loss": 2.9817,
+      "step": 41558
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6823240518569946,
+      "learning_rate": 0.0001401264776624512,
+      "loss": 2.7373,
+      "step": 41559
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6630463600158691,
+      "learning_rate": 0.00014011345006575127,
+      "loss": 2.9244,
+      "step": 41560
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6817547082901001,
+      "learning_rate": 0.00014010042289016163,
+      "loss": 3.0798,
+      "step": 41561
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.697632908821106,
+      "learning_rate": 0.00014008739613571668,
+      "loss": 2.9706,
+      "step": 41562
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6645799279212952,
+      "learning_rate": 0.00014007436980245088,
+      "loss": 3.0473,
+      "step": 41563
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7861841320991516,
+      "learning_rate": 0.00014006134389039828,
+      "loss": 2.9287,
+      "step": 41564
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7191126942634583,
+      "learning_rate": 0.00014004831839959346,
+      "loss": 2.8517,
+      "step": 41565
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6533079743385315,
+      "learning_rate": 0.0001400352933300706,
+      "loss": 2.7748,
+      "step": 41566
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6802554130554199,
+      "learning_rate": 0.00014002226868186382,
+      "loss": 2.985,
+      "step": 41567
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6917102932929993,
+      "learning_rate": 0.00014000924445500774,
+      "loss": 2.9208,
+      "step": 41568
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6812888383865356,
+      "learning_rate": 0.0001399962206495364,
+      "loss": 3.2155,
+      "step": 41569
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6770250797271729,
+      "learning_rate": 0.00013998319726548435,
+      "loss": 2.7714,
+      "step": 41570
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6774451732635498,
+      "learning_rate": 0.00013997017430288566,
+      "loss": 2.9362,
+      "step": 41571
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6566124558448792,
+      "learning_rate": 0.00013995715176177482,
+      "loss": 2.9055,
+      "step": 41572
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6735152006149292,
+      "learning_rate": 0.00013994412964218604,
+      "loss": 2.8136,
+      "step": 41573
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6460737586021423,
+      "learning_rate": 0.00013993110794415352,
+      "loss": 2.9369,
+      "step": 41574
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6937167644500732,
+      "learning_rate": 0.00013991808666771177,
+      "loss": 2.94,
+      "step": 41575
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6455183625221252,
+      "learning_rate": 0.00013990506581289483,
+      "loss": 2.7323,
+      "step": 41576
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6627008318901062,
+      "learning_rate": 0.00013989204537973715,
+      "loss": 2.8377,
+      "step": 41577
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6939973831176758,
+      "learning_rate": 0.00013987902536827314,
+      "loss": 2.9573,
+      "step": 41578
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7010552883148193,
+      "learning_rate": 0.00013986600577853696,
+      "loss": 2.984,
+      "step": 41579
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6695321798324585,
+      "learning_rate": 0.0001398529866105629,
+      "loss": 2.967,
+      "step": 41580
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7119990587234497,
+      "learning_rate": 0.0001398399678643851,
+      "loss": 2.8192,
+      "step": 41581
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6729029417037964,
+      "learning_rate": 0.00013982694954003805,
+      "loss": 2.8459,
+      "step": 41582
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6772317290306091,
+      "learning_rate": 0.0001398139316375561,
+      "loss": 2.9779,
+      "step": 41583
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6854050159454346,
+      "learning_rate": 0.0001398009141569733,
+      "loss": 2.9556,
+      "step": 41584
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6610407829284668,
+      "learning_rate": 0.00013978789709832414,
+      "loss": 2.862,
+      "step": 41585
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7166022658348083,
+      "learning_rate": 0.00013977488046164276,
+      "loss": 3.1247,
+      "step": 41586
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.673302948474884,
+      "learning_rate": 0.00013976186424696362,
+      "loss": 2.812,
+      "step": 41587
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6916917562484741,
+      "learning_rate": 0.00013974884845432088,
+      "loss": 3.0233,
+      "step": 41588
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6967342495918274,
+      "learning_rate": 0.0001397358330837487,
+      "loss": 2.9757,
+      "step": 41589
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.741462767124176,
+      "learning_rate": 0.0001397228181352816,
+      "loss": 2.9999,
+      "step": 41590
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7056967616081238,
+      "learning_rate": 0.0001397098036089537,
+      "loss": 2.8734,
+      "step": 41591
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6610565781593323,
+      "learning_rate": 0.00013969678950479926,
+      "loss": 2.9993,
+      "step": 41592
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.653553307056427,
+      "learning_rate": 0.00013968377582285278,
+      "loss": 2.873,
+      "step": 41593
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.724814772605896,
+      "learning_rate": 0.0001396707625631484,
+      "loss": 2.916,
+      "step": 41594
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6634653210639954,
+      "learning_rate": 0.00013965774972572032,
+      "loss": 2.9582,
+      "step": 41595
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6641644835472107,
+      "learning_rate": 0.00013964473731060276,
+      "loss": 3.0106,
+      "step": 41596
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6954382061958313,
+      "learning_rate": 0.00013963172531783013,
+      "loss": 3.1128,
+      "step": 41597
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6707383394241333,
+      "learning_rate": 0.00013961871374743677,
+      "loss": 2.7836,
+      "step": 41598
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6706055402755737,
+      "learning_rate": 0.00013960570259945675,
+      "loss": 2.9707,
+      "step": 41599
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6660712361335754,
+      "learning_rate": 0.00013959269187392457,
+      "loss": 2.9676,
+      "step": 41600
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7226710319519043,
+      "learning_rate": 0.00013957968157087434,
+      "loss": 3.0102,
+      "step": 41601
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.70147705078125,
+      "learning_rate": 0.00013956667169034021,
+      "loss": 3.0778,
+      "step": 41602
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7111939787864685,
+      "learning_rate": 0.00013955366223235679,
+      "loss": 2.958,
+      "step": 41603
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6577525734901428,
+      "learning_rate": 0.00013954065319695796,
+      "loss": 3.0126,
+      "step": 41604
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6963541507720947,
+      "learning_rate": 0.00013952764458417832,
+      "loss": 2.9144,
+      "step": 41605
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6417427659034729,
+      "learning_rate": 0.00013951463639405186,
+      "loss": 2.8289,
+      "step": 41606
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6816616654396057,
+      "learning_rate": 0.00013950162862661305,
+      "loss": 2.8721,
+      "step": 41607
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7183780670166016,
+      "learning_rate": 0.00013948862128189597,
+      "loss": 2.8842,
+      "step": 41608
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6773726940155029,
+      "learning_rate": 0.00013947561435993507,
+      "loss": 2.8267,
+      "step": 41609
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.715185821056366,
+      "learning_rate": 0.00013946260786076452,
+      "loss": 2.8606,
+      "step": 41610
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6540136337280273,
+      "learning_rate": 0.00013944960178441848,
+      "loss": 3.098,
+      "step": 41611
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6640399098396301,
+      "learning_rate": 0.00013943659613093137,
+      "loss": 2.9417,
+      "step": 41612
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6632117629051208,
+      "learning_rate": 0.00013942359090033724,
+      "loss": 2.8215,
+      "step": 41613
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7612461447715759,
+      "learning_rate": 0.00013941058609267048,
+      "loss": 2.7336,
+      "step": 41614
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6554035544395447,
+      "learning_rate": 0.00013939758170796547,
+      "loss": 2.7637,
+      "step": 41615
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6623538732528687,
+      "learning_rate": 0.0001393845777462563,
+      "loss": 3.0458,
+      "step": 41616
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6795579195022583,
+      "learning_rate": 0.00013937157420757722,
+      "loss": 2.8377,
+      "step": 41617
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6572181582450867,
+      "learning_rate": 0.0001393585710919624,
+      "loss": 2.7483,
+      "step": 41618
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6754372715950012,
+      "learning_rate": 0.00013934556839944622,
+      "loss": 2.8756,
+      "step": 41619
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6733306646347046,
+      "learning_rate": 0.00013933256613006298,
+      "loss": 2.8464,
+      "step": 41620
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6594640016555786,
+      "learning_rate": 0.0001393195642838467,
+      "loss": 2.9303,
+      "step": 41621
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6582934260368347,
+      "learning_rate": 0.00013930656286083193,
+      "loss": 2.913,
+      "step": 41622
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6640312671661377,
+      "learning_rate": 0.00013929356186105272,
+      "loss": 2.8359,
+      "step": 41623
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.643187403678894,
+      "learning_rate": 0.0001392805612845432,
+      "loss": 3.2914,
+      "step": 41624
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7015854120254517,
+      "learning_rate": 0.00013926756113133792,
+      "loss": 3.1108,
+      "step": 41625
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6656563878059387,
+      "learning_rate": 0.0001392545614014708,
+      "loss": 2.9735,
+      "step": 41626
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6701093316078186,
+      "learning_rate": 0.00013924156209497638,
+      "loss": 2.7599,
+      "step": 41627
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6650725603103638,
+      "learning_rate": 0.0001392285632118886,
+      "loss": 2.8968,
+      "step": 41628
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6751466989517212,
+      "learning_rate": 0.00013921556475224184,
+      "loss": 2.9247,
+      "step": 41629
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6479591727256775,
+      "learning_rate": 0.00013920256671607047,
+      "loss": 3.0724,
+      "step": 41630
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7332056164741516,
+      "learning_rate": 0.00013918956910340856,
+      "loss": 2.8461,
+      "step": 41631
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.690640926361084,
+      "learning_rate": 0.00013917657191429042,
+      "loss": 2.9417,
+      "step": 41632
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6897100806236267,
+      "learning_rate": 0.0001391635751487501,
+      "loss": 2.9565,
+      "step": 41633
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6536319851875305,
+      "learning_rate": 0.000139150578806822,
+      "loss": 2.6714,
+      "step": 41634
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8222829699516296,
+      "learning_rate": 0.0001391375828885404,
+      "loss": 2.9552,
+      "step": 41635
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7058513164520264,
+      "learning_rate": 0.00013912458739393933,
+      "loss": 2.6434,
+      "step": 41636
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6645085215568542,
+      "learning_rate": 0.00013911159232305327,
+      "loss": 3.0102,
+      "step": 41637
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6806157231330872,
+      "learning_rate": 0.0001390985976759163,
+      "loss": 2.9867,
+      "step": 41638
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7031665444374084,
+      "learning_rate": 0.00013908560345256255,
+      "loss": 3.0195,
+      "step": 41639
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6819475889205933,
+      "learning_rate": 0.00013907260965302648,
+      "loss": 3.0876,
+      "step": 41640
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6787337064743042,
+      "learning_rate": 0.00013905961627734204,
+      "loss": 2.7164,
+      "step": 41641
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7834779024124146,
+      "learning_rate": 0.00013904662332554372,
+      "loss": 2.7447,
+      "step": 41642
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6589922904968262,
+      "learning_rate": 0.0001390336307976655,
+      "loss": 2.8369,
+      "step": 41643
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6925103664398193,
+      "learning_rate": 0.00013902063869374175,
+      "loss": 2.7699,
+      "step": 41644
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6515282392501831,
+      "learning_rate": 0.00013900764701380675,
+      "loss": 2.9382,
+      "step": 41645
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7050473690032959,
+      "learning_rate": 0.0001389946557578946,
+      "loss": 2.8169,
+      "step": 41646
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7375444769859314,
+      "learning_rate": 0.00013898166492603957,
+      "loss": 2.8523,
+      "step": 41647
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8539415597915649,
+      "learning_rate": 0.00013896867451827576,
+      "loss": 2.9165,
+      "step": 41648
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7187871336936951,
+      "learning_rate": 0.0001389556845346374,
+      "loss": 2.6562,
+      "step": 41649
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6538518071174622,
+      "learning_rate": 0.00013894269497515894,
+      "loss": 2.9595,
+      "step": 41650
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6808871030807495,
+      "learning_rate": 0.00013892970583987427,
+      "loss": 2.7846,
+      "step": 41651
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6848717927932739,
+      "learning_rate": 0.0001389167171288179,
+      "loss": 2.8772,
+      "step": 41652
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6991068720817566,
+      "learning_rate": 0.0001389037288420239,
+      "loss": 3.1921,
+      "step": 41653
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.701252818107605,
+      "learning_rate": 0.00013889074097952637,
+      "loss": 2.9685,
+      "step": 41654
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6782150864601135,
+      "learning_rate": 0.00013887775354135971,
+      "loss": 2.807,
+      "step": 41655
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6534834504127502,
+      "learning_rate": 0.00013886476652755794,
+      "loss": 2.9893,
+      "step": 41656
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6911651492118835,
+      "learning_rate": 0.00013885177993815548,
+      "loss": 2.8997,
+      "step": 41657
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6815614700317383,
+      "learning_rate": 0.00013883879377318627,
+      "loss": 2.8842,
+      "step": 41658
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.681822657585144,
+      "learning_rate": 0.00013882580803268482,
+      "loss": 2.9365,
+      "step": 41659
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6272294521331787,
+      "learning_rate": 0.00013881282271668515,
+      "loss": 2.9309,
+      "step": 41660
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7433640360832214,
+      "learning_rate": 0.0001387998378252214,
+      "loss": 2.9243,
+      "step": 41661
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6774015426635742,
+      "learning_rate": 0.00013878685335832789,
+      "loss": 3.0579,
+      "step": 41662
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6565629839897156,
+      "learning_rate": 0.00013877386931603874,
+      "loss": 3.0814,
+      "step": 41663
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6806601285934448,
+      "learning_rate": 0.00013876088569838815,
+      "loss": 2.9186,
+      "step": 41664
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6668409109115601,
+      "learning_rate": 0.0001387479025054105,
+      "loss": 2.8923,
+      "step": 41665
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6895928978919983,
+      "learning_rate": 0.0001387349197371397,
+      "loss": 3.0788,
+      "step": 41666
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.634085476398468,
+      "learning_rate": 0.00013872193739361023,
+      "loss": 2.8866,
+      "step": 41667
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.704136073589325,
+      "learning_rate": 0.0001387089554748561,
+      "loss": 3.0615,
+      "step": 41668
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6691809296607971,
+      "learning_rate": 0.00013869597398091148,
+      "loss": 3.0299,
+      "step": 41669
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6336991190910339,
+      "learning_rate": 0.00013868299291181068,
+      "loss": 2.7573,
+      "step": 41670
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.698936939239502,
+      "learning_rate": 0.00013867001226758774,
+      "loss": 2.8835,
+      "step": 41671
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7142788767814636,
+      "learning_rate": 0.00013865703204827707,
+      "loss": 3.3196,
+      "step": 41672
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.694664716720581,
+      "learning_rate": 0.0001386440522539126,
+      "loss": 2.9352,
+      "step": 41673
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.666140079498291,
+      "learning_rate": 0.00013863107288452877,
+      "loss": 2.9518,
+      "step": 41674
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.722091555595398,
+      "learning_rate": 0.00013861809394015963,
+      "loss": 2.8776,
+      "step": 41675
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6839096546173096,
+      "learning_rate": 0.00013860511542083923,
+      "loss": 3.0636,
+      "step": 41676
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.666094183921814,
+      "learning_rate": 0.000138592137326602,
+      "loss": 2.9686,
+      "step": 41677
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.696448802947998,
+      "learning_rate": 0.00013857915965748195,
+      "loss": 3.0072,
+      "step": 41678
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7537996172904968,
+      "learning_rate": 0.00013856618241351332,
+      "loss": 2.8667,
+      "step": 41679
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6860277652740479,
+      "learning_rate": 0.00013855320559473037,
+      "loss": 2.8746,
+      "step": 41680
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6779218316078186,
+      "learning_rate": 0.00013854022920116725,
+      "loss": 2.9084,
+      "step": 41681
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.680435836315155,
+      "learning_rate": 0.00013852725323285807,
+      "loss": 2.9952,
+      "step": 41682
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6816911697387695,
+      "learning_rate": 0.00013851427768983693,
+      "loss": 2.7457,
+      "step": 41683
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7383108735084534,
+      "learning_rate": 0.0001385013025721382,
+      "loss": 3.1948,
+      "step": 41684
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6533733606338501,
+      "learning_rate": 0.00013848832787979586,
+      "loss": 2.7713,
+      "step": 41685
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6958843469619751,
+      "learning_rate": 0.00013847535361284418,
+      "loss": 3.1565,
+      "step": 41686
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6978946924209595,
+      "learning_rate": 0.00013846237977131742,
+      "loss": 2.8521,
+      "step": 41687
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6730976104736328,
+      "learning_rate": 0.0001384494063552496,
+      "loss": 3.025,
+      "step": 41688
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6826785802841187,
+      "learning_rate": 0.000138436433364675,
+      "loss": 2.9969,
+      "step": 41689
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7159188389778137,
+      "learning_rate": 0.0001384234607996278,
+      "loss": 2.9785,
+      "step": 41690
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6930112838745117,
+      "learning_rate": 0.00013841048866014197,
+      "loss": 2.9677,
+      "step": 41691
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6859667897224426,
+      "learning_rate": 0.00013839751694625196,
+      "loss": 2.9933,
+      "step": 41692
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7143214344978333,
+      "learning_rate": 0.00013838454565799163,
+      "loss": 2.9476,
+      "step": 41693
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7018111944198608,
+      "learning_rate": 0.00013837157479539542,
+      "loss": 2.7747,
+      "step": 41694
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6992251873016357,
+      "learning_rate": 0.00013835860435849727,
+      "loss": 3.0552,
+      "step": 41695
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6739795207977295,
+      "learning_rate": 0.00013834563434733156,
+      "loss": 2.8918,
+      "step": 41696
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7193833589553833,
+      "learning_rate": 0.00013833266476193237,
+      "loss": 3.0882,
+      "step": 41697
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6857161521911621,
+      "learning_rate": 0.0001383196956023337,
+      "loss": 3.0136,
+      "step": 41698
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.73293137550354,
+      "learning_rate": 0.0001383067268685699,
+      "loss": 2.7954,
+      "step": 41699
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6800030469894409,
+      "learning_rate": 0.000138293758560675,
+      "loss": 3.0618,
+      "step": 41700
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6845666766166687,
+      "learning_rate": 0.0001382807906786832,
+      "loss": 3.0205,
+      "step": 41701
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6896492838859558,
+      "learning_rate": 0.00013826782322262881,
+      "loss": 2.8907,
+      "step": 41702
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6447883248329163,
+      "learning_rate": 0.00013825485619254582,
+      "loss": 2.8803,
+      "step": 41703
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6869139671325684,
+      "learning_rate": 0.00013824188958846845,
+      "loss": 3.2243,
+      "step": 41704
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6388272047042847,
+      "learning_rate": 0.00013822892341043067,
+      "loss": 2.7329,
+      "step": 41705
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6573567986488342,
+      "learning_rate": 0.00013821595765846676,
+      "loss": 2.9315,
+      "step": 41706
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6946920156478882,
+      "learning_rate": 0.00013820299233261102,
+      "loss": 2.8361,
+      "step": 41707
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7050734758377075,
+      "learning_rate": 0.00013819002743289732,
+      "loss": 2.8692,
+      "step": 41708
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7093161940574646,
+      "learning_rate": 0.00013817706295936007,
+      "loss": 2.8641,
+      "step": 41709
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6960554122924805,
+      "learning_rate": 0.0001381640989120332,
+      "loss": 2.9397,
+      "step": 41710
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6445857882499695,
+      "learning_rate": 0.00013815113529095104,
+      "loss": 2.9978,
+      "step": 41711
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7119023203849792,
+      "learning_rate": 0.00013813817209614766,
+      "loss": 2.9848,
+      "step": 41712
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6626527309417725,
+      "learning_rate": 0.00013812520932765702,
+      "loss": 2.8887,
+      "step": 41713
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6804422736167908,
+      "learning_rate": 0.00013811224698551358,
+      "loss": 3.0319,
+      "step": 41714
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7170805335044861,
+      "learning_rate": 0.0001380992850697512,
+      "loss": 2.9286,
+      "step": 41715
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6595010161399841,
+      "learning_rate": 0.0001380863235804042,
+      "loss": 2.9433,
+      "step": 41716
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6733608245849609,
+      "learning_rate": 0.0001380733625175067,
+      "loss": 2.8411,
+      "step": 41717
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8234678506851196,
+      "learning_rate": 0.00013806040188109282,
+      "loss": 2.8652,
+      "step": 41718
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7054716348648071,
+      "learning_rate": 0.00013804744167119666,
+      "loss": 2.7808,
+      "step": 41719
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7219303250312805,
+      "learning_rate": 0.00013803448188785228,
+      "loss": 2.8326,
+      "step": 41720
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6587951183319092,
+      "learning_rate": 0.0001380215225310939,
+      "loss": 2.8626,
+      "step": 41721
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6685791611671448,
+      "learning_rate": 0.0001380085636009558,
+      "loss": 2.9384,
+      "step": 41722
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7046728134155273,
+      "learning_rate": 0.00013799560509747185,
+      "loss": 2.9112,
+      "step": 41723
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.1903138160705566,
+      "learning_rate": 0.00013798264702067636,
+      "loss": 2.7223,
+      "step": 41724
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6698146462440491,
+      "learning_rate": 0.00013796968937060345,
+      "loss": 3.0392,
+      "step": 41725
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6656641960144043,
+      "learning_rate": 0.00013795673214728703,
+      "loss": 3.1732,
+      "step": 41726
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6809387803077698,
+      "learning_rate": 0.00013794377535076153,
+      "loss": 3.057,
+      "step": 41727
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7437176704406738,
+      "learning_rate": 0.00013793081898106081,
+      "loss": 3.0012,
+      "step": 41728
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6619541645050049,
+      "learning_rate": 0.0001379178630382193,
+      "loss": 3.0339,
+      "step": 41729
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6481929421424866,
+      "learning_rate": 0.00013790490752227074,
+      "loss": 2.806,
+      "step": 41730
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6587005853652954,
+      "learning_rate": 0.00013789195243324952,
+      "loss": 3.1365,
+      "step": 41731
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6814006567001343,
+      "learning_rate": 0.00013787899777118978,
+      "loss": 2.8502,
+      "step": 41732
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6785389184951782,
+      "learning_rate": 0.00013786604353612558,
+      "loss": 2.901,
+      "step": 41733
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6514621376991272,
+      "learning_rate": 0.00013785308972809102,
+      "loss": 2.7796,
+      "step": 41734
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7052819728851318,
+      "learning_rate": 0.00013784013634712009,
+      "loss": 2.825,
+      "step": 41735
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6751008629798889,
+      "learning_rate": 0.00013782718339324702,
+      "loss": 2.9968,
+      "step": 41736
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6512966752052307,
+      "learning_rate": 0.00013781423086650608,
+      "loss": 2.9316,
+      "step": 41737
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6523333787918091,
+      "learning_rate": 0.00013780127876693112,
+      "loss": 3.0541,
+      "step": 41738
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7050060033798218,
+      "learning_rate": 0.0001377883270945565,
+      "loss": 3.0663,
+      "step": 41739
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7077402472496033,
+      "learning_rate": 0.00013777537584941618,
+      "loss": 3.0436,
+      "step": 41740
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6588230133056641,
+      "learning_rate": 0.0001377624250315442,
+      "loss": 2.858,
+      "step": 41741
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.708586573600769,
+      "learning_rate": 0.00013774947464097488,
+      "loss": 2.8115,
+      "step": 41742
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6874078512191772,
+      "learning_rate": 0.0001377365246777421,
+      "loss": 2.7809,
+      "step": 41743
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6789014935493469,
+      "learning_rate": 0.00013772357514188022,
+      "loss": 3.0268,
+      "step": 41744
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7179741263389587,
+      "learning_rate": 0.00013771062603342308,
+      "loss": 3.13,
+      "step": 41745
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6951627731323242,
+      "learning_rate": 0.00013769767735240493,
+      "loss": 3.0357,
+      "step": 41746
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6996464133262634,
+      "learning_rate": 0.00013768472909886007,
+      "loss": 2.9498,
+      "step": 41747
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6561077237129211,
+      "learning_rate": 0.0001376717812728222,
+      "loss": 2.8745,
+      "step": 41748
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.736611008644104,
+      "learning_rate": 0.00013765883387432568,
+      "loss": 2.8959,
+      "step": 41749
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6801815629005432,
+      "learning_rate": 0.00013764588690340447,
+      "loss": 3.0878,
+      "step": 41750
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6809715032577515,
+      "learning_rate": 0.00013763294036009274,
+      "loss": 2.9666,
+      "step": 41751
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7071965336799622,
+      "learning_rate": 0.00013761999424442473,
+      "loss": 2.9907,
+      "step": 41752
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6758064031600952,
+      "learning_rate": 0.00013760704855643428,
+      "loss": 2.7936,
+      "step": 41753
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6896607875823975,
+      "learning_rate": 0.0001375941032961557,
+      "loss": 2.8758,
+      "step": 41754
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6973819136619568,
+      "learning_rate": 0.000137581158463623,
+      "loss": 2.994,
+      "step": 41755
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6735284924507141,
+      "learning_rate": 0.0001375682140588702,
+      "loss": 2.8614,
+      "step": 41756
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.680656373500824,
+      "learning_rate": 0.00013755527008193154,
+      "loss": 2.9821,
+      "step": 41757
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6982008814811707,
+      "learning_rate": 0.00013754232653284093,
+      "loss": 2.8955,
+      "step": 41758
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.683266282081604,
+      "learning_rate": 0.0001375293834116327,
+      "loss": 3.0358,
+      "step": 41759
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6826072335243225,
+      "learning_rate": 0.00013751644071834066,
+      "loss": 2.7453,
+      "step": 41760
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7050215005874634,
+      "learning_rate": 0.0001375034984529992,
+      "loss": 2.7758,
+      "step": 41761
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7060586214065552,
+      "learning_rate": 0.0001374905566156422,
+      "loss": 2.8637,
+      "step": 41762
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7489438056945801,
+      "learning_rate": 0.0001374776152063037,
+      "loss": 2.8327,
+      "step": 41763
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7340999841690063,
+      "learning_rate": 0.000137464674225018,
+      "loss": 2.963,
+      "step": 41764
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6897082328796387,
+      "learning_rate": 0.00013745173367181899,
+      "loss": 2.8568,
+      "step": 41765
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7091054320335388,
+      "learning_rate": 0.0001374387935467409,
+      "loss": 2.9136,
+      "step": 41766
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7176274657249451,
+      "learning_rate": 0.00013742585384981767,
+      "loss": 2.8918,
+      "step": 41767
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.675040602684021,
+      "learning_rate": 0.00013741291458108342,
+      "loss": 2.8829,
+      "step": 41768
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6755308508872986,
+      "learning_rate": 0.00013739997574057237,
+      "loss": 3.0007,
+      "step": 41769
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.698590874671936,
+      "learning_rate": 0.0001373870373283185,
+      "loss": 3.1754,
+      "step": 41770
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6927210688591003,
+      "learning_rate": 0.00013737409934435588,
+      "loss": 2.8816,
+      "step": 41771
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7321736812591553,
+      "learning_rate": 0.00013736116178871848,
+      "loss": 2.995,
+      "step": 41772
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7068862318992615,
+      "learning_rate": 0.0001373482246614405,
+      "loss": 2.895,
+      "step": 41773
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6535940170288086,
+      "learning_rate": 0.00013733528796255605,
+      "loss": 2.9816,
+      "step": 41774
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6542276740074158,
+      "learning_rate": 0.00013732235169209908,
+      "loss": 2.8163,
+      "step": 41775
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7153056859970093,
+      "learning_rate": 0.0001373094158501038,
+      "loss": 2.7931,
+      "step": 41776
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7129161357879639,
+      "learning_rate": 0.00013729648043660423,
+      "loss": 2.7878,
+      "step": 41777
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.656928300857544,
+      "learning_rate": 0.0001372835454516343,
+      "loss": 2.8457,
+      "step": 41778
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.733238935470581,
+      "learning_rate": 0.00013727061089522833,
+      "loss": 2.9476,
+      "step": 41779
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6839013695716858,
+      "learning_rate": 0.0001372576767674201,
+      "loss": 2.9249,
+      "step": 41780
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7424287796020508,
+      "learning_rate": 0.000137244743068244,
+      "loss": 2.9104,
+      "step": 41781
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7072665691375732,
+      "learning_rate": 0.00013723180979773378,
+      "loss": 2.9528,
+      "step": 41782
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7041286826133728,
+      "learning_rate": 0.00013721887695592372,
+      "loss": 2.7528,
+      "step": 41783
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7332327961921692,
+      "learning_rate": 0.00013720594454284788,
+      "loss": 2.7556,
+      "step": 41784
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6847696900367737,
+      "learning_rate": 0.0001371930125585401,
+      "loss": 2.7119,
+      "step": 41785
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6434175968170166,
+      "learning_rate": 0.00013718008100303472,
+      "loss": 2.8625,
+      "step": 41786
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6727170944213867,
+      "learning_rate": 0.00013716714987636555,
+      "loss": 2.9554,
+      "step": 41787
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6850985884666443,
+      "learning_rate": 0.00013715421917856677,
+      "loss": 3.069,
+      "step": 41788
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6658537983894348,
+      "learning_rate": 0.00013714128890967252,
+      "loss": 2.8858,
+      "step": 41789
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6863417625427246,
+      "learning_rate": 0.00013712835906971667,
+      "loss": 2.8266,
+      "step": 41790
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7340531349182129,
+      "learning_rate": 0.00013711542965873352,
+      "loss": 2.7212,
+      "step": 41791
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7077760696411133,
+      "learning_rate": 0.00013710250067675697,
+      "loss": 2.9836,
+      "step": 41792
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7007636427879333,
+      "learning_rate": 0.000137089572123821,
+      "loss": 2.7947,
+      "step": 41793
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6507523655891418,
+      "learning_rate": 0.0001370766439999598,
+      "loss": 2.7803,
+      "step": 41794
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6822836995124817,
+      "learning_rate": 0.00013706371630520727,
+      "loss": 2.7836,
+      "step": 41795
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.2069634199142456,
+      "learning_rate": 0.00013705078903959768,
+      "loss": 2.9114,
+      "step": 41796
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6869301795959473,
+      "learning_rate": 0.00013703786220316483,
+      "loss": 2.7778,
+      "step": 41797
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6838812828063965,
+      "learning_rate": 0.000137024935795943,
+      "loss": 2.9701,
+      "step": 41798
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6679784655570984,
+      "learning_rate": 0.00013701200981796615,
+      "loss": 2.873,
+      "step": 41799
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6981244683265686,
+      "learning_rate": 0.00013699908426926814,
+      "loss": 3.1318,
+      "step": 41800
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6437106728553772,
+      "learning_rate": 0.00013698615914988334,
+      "loss": 2.8379,
+      "step": 41801
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6813018321990967,
+      "learning_rate": 0.00013697323445984545,
+      "loss": 2.8552,
+      "step": 41802
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7143990993499756,
+      "learning_rate": 0.00013696031019918873,
+      "loss": 2.9512,
+      "step": 41803
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7127072215080261,
+      "learning_rate": 0.00013694738636794727,
+      "loss": 2.966,
+      "step": 41804
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7043535709381104,
+      "learning_rate": 0.000136934462966155,
+      "loss": 2.801,
+      "step": 41805
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6857683062553406,
+      "learning_rate": 0.000136921539993846,
+      "loss": 2.8319,
+      "step": 41806
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6381077766418457,
+      "learning_rate": 0.00013690861745105414,
+      "loss": 3.0228,
+      "step": 41807
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6954429745674133,
+      "learning_rate": 0.00013689569533781364,
+      "loss": 2.9477,
+      "step": 41808
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6456500291824341,
+      "learning_rate": 0.00013688277365415854,
+      "loss": 3.0681,
+      "step": 41809
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6630203127861023,
+      "learning_rate": 0.00013686985240012276,
+      "loss": 2.8573,
+      "step": 41810
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6759008169174194,
+      "learning_rate": 0.00013685693157574046,
+      "loss": 2.7842,
+      "step": 41811
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6748966574668884,
+      "learning_rate": 0.00013684401118104552,
+      "loss": 2.9473,
+      "step": 41812
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7730910778045654,
+      "learning_rate": 0.00013683109121607218,
+      "loss": 2.636,
+      "step": 41813
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6848281025886536,
+      "learning_rate": 0.00013681817168085432,
+      "loss": 2.8989,
+      "step": 41814
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6651986241340637,
+      "learning_rate": 0.0001368052525754259,
+      "loss": 2.7469,
+      "step": 41815
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6602210998535156,
+      "learning_rate": 0.00013679233389982112,
+      "loss": 3.0205,
+      "step": 41816
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7078651189804077,
+      "learning_rate": 0.00013677941565407384,
+      "loss": 2.8248,
+      "step": 41817
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6716862320899963,
+      "learning_rate": 0.00013676649783821813,
+      "loss": 2.8242,
+      "step": 41818
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6786720156669617,
+      "learning_rate": 0.00013675358045228821,
+      "loss": 3.0426,
+      "step": 41819
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6606864929199219,
+      "learning_rate": 0.00013674066349631791,
+      "loss": 3.0192,
+      "step": 41820
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6600707769393921,
+      "learning_rate": 0.00013672774697034127,
+      "loss": 3.1479,
+      "step": 41821
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7069227695465088,
+      "learning_rate": 0.00013671483087439223,
+      "loss": 2.8687,
+      "step": 41822
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6884543895721436,
+      "learning_rate": 0.0001367019152085049,
+      "loss": 2.9547,
+      "step": 41823
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6656718850135803,
+      "learning_rate": 0.0001366889999727134,
+      "loss": 2.9701,
+      "step": 41824
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6968078017234802,
+      "learning_rate": 0.00013667608516705154,
+      "loss": 3.0355,
+      "step": 41825
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7189076542854309,
+      "learning_rate": 0.0001366631707915536,
+      "loss": 2.9283,
+      "step": 41826
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6748219728469849,
+      "learning_rate": 0.00013665025684625335,
+      "loss": 3.0839,
+      "step": 41827
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6655088067054749,
+      "learning_rate": 0.00013663734333118483,
+      "loss": 2.8599,
+      "step": 41828
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6688360571861267,
+      "learning_rate": 0.00013662443024638218,
+      "loss": 3.2349,
+      "step": 41829
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7066716551780701,
+      "learning_rate": 0.00013661151759187927,
+      "loss": 2.9871,
+      "step": 41830
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6709800362586975,
+      "learning_rate": 0.00013659860536771023,
+      "loss": 2.815,
+      "step": 41831
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7102221250534058,
+      "learning_rate": 0.00013658569357390897,
+      "loss": 2.8565,
+      "step": 41832
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6634939312934875,
+      "learning_rate": 0.0001365727822105095,
+      "loss": 2.7641,
+      "step": 41833
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6929799318313599,
+      "learning_rate": 0.000136559871277546,
+      "loss": 2.8275,
+      "step": 41834
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.695609450340271,
+      "learning_rate": 0.00013654696077505232,
+      "loss": 3.0188,
+      "step": 41835
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6726747751235962,
+      "learning_rate": 0.00013653405070306252,
+      "loss": 2.9284,
+      "step": 41836
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7245111465454102,
+      "learning_rate": 0.00013652114106161043,
+      "loss": 2.8936,
+      "step": 41837
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7520313858985901,
+      "learning_rate": 0.00013650823185073035,
+      "loss": 2.9826,
+      "step": 41838
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6968186497688293,
+      "learning_rate": 0.00013649532307045595,
+      "loss": 3.0164,
+      "step": 41839
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7093591094017029,
+      "learning_rate": 0.00013648241472082143,
+      "loss": 3.0614,
+      "step": 41840
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6695505380630493,
+      "learning_rate": 0.0001364695068018609,
+      "loss": 2.957,
+      "step": 41841
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.721322238445282,
+      "learning_rate": 0.0001364565993136082,
+      "loss": 2.7591,
+      "step": 41842
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6672532558441162,
+      "learning_rate": 0.00013644369225609734,
+      "loss": 3.027,
+      "step": 41843
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7743553519248962,
+      "learning_rate": 0.00013643078562936218,
+      "loss": 2.7336,
+      "step": 41844
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7321560978889465,
+      "learning_rate": 0.00013641787943343686,
+      "loss": 2.8884,
+      "step": 41845
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7255355715751648,
+      "learning_rate": 0.0001364049736683555,
+      "loss": 2.8195,
+      "step": 41846
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.721168577671051,
+      "learning_rate": 0.00013639206833415183,
+      "loss": 2.7635,
+      "step": 41847
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6950680613517761,
+      "learning_rate": 0.0001363791634308601,
+      "loss": 2.7786,
+      "step": 41848
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7091295719146729,
+      "learning_rate": 0.00013636625895851413,
+      "loss": 3.0322,
+      "step": 41849
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7242528796195984,
+      "learning_rate": 0.00013635335491714786,
+      "loss": 2.8183,
+      "step": 41850
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7005438804626465,
+      "learning_rate": 0.00013634045130679544,
+      "loss": 2.9768,
+      "step": 41851
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6915015578269958,
+      "learning_rate": 0.00013632754812749066,
+      "loss": 3.1398,
+      "step": 41852
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6812850832939148,
+      "learning_rate": 0.00013631464537926775,
+      "loss": 3.0446,
+      "step": 41853
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.69903165102005,
+      "learning_rate": 0.00013630174306216044,
+      "loss": 2.9758,
+      "step": 41854
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6556075215339661,
+      "learning_rate": 0.00013628884117620284,
+      "loss": 2.8251,
+      "step": 41855
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.9440958499908447,
+      "learning_rate": 0.00013627593972142903,
+      "loss": 2.9879,
+      "step": 41856
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6949568390846252,
+      "learning_rate": 0.00013626303869787292,
+      "loss": 2.7712,
+      "step": 41857
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7072689533233643,
+      "learning_rate": 0.0001362501381055684,
+      "loss": 2.9006,
+      "step": 41858
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7018836140632629,
+      "learning_rate": 0.00013623723794454944,
+      "loss": 3.0257,
+      "step": 41859
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6747689247131348,
+      "learning_rate": 0.00013622433821485006,
+      "loss": 2.9541,
+      "step": 41860
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.704422116279602,
+      "learning_rate": 0.00013621143891650437,
+      "loss": 2.9823,
+      "step": 41861
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6954941749572754,
+      "learning_rate": 0.00013619854004954612,
+      "loss": 3.0837,
+      "step": 41862
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7778325080871582,
+      "learning_rate": 0.00013618564161400947,
+      "loss": 2.797,
+      "step": 41863
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7002326846122742,
+      "learning_rate": 0.00013617274360992832,
+      "loss": 2.8489,
+      "step": 41864
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6892054080963135,
+      "learning_rate": 0.00013615984603733655,
+      "loss": 2.9662,
+      "step": 41865
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6607648730278015,
+      "learning_rate": 0.0001361469488962683,
+      "loss": 2.9692,
+      "step": 41866
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6442403197288513,
+      "learning_rate": 0.00013613405218675735,
+      "loss": 2.8593,
+      "step": 41867
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6695894002914429,
+      "learning_rate": 0.0001361211559088379,
+      "loss": 2.7185,
+      "step": 41868
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.725757360458374,
+      "learning_rate": 0.00013610826006254368,
+      "loss": 2.9114,
+      "step": 41869
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7685546278953552,
+      "learning_rate": 0.00013609536464790871,
+      "loss": 2.8982,
+      "step": 41870
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6293439865112305,
+      "learning_rate": 0.00013608246966496714,
+      "loss": 2.8666,
+      "step": 41871
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7110158205032349,
+      "learning_rate": 0.00013606957511375285,
+      "loss": 2.7713,
+      "step": 41872
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.687812864780426,
+      "learning_rate": 0.00013605668099429967,
+      "loss": 3.1352,
+      "step": 41873
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6826140284538269,
+      "learning_rate": 0.00013604378730664156,
+      "loss": 2.7359,
+      "step": 41874
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6955284476280212,
+      "learning_rate": 0.0001360308940508126,
+      "loss": 2.8839,
+      "step": 41875
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.744838535785675,
+      "learning_rate": 0.00013601800122684676,
+      "loss": 2.8754,
+      "step": 41876
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7015159130096436,
+      "learning_rate": 0.00013600510883477786,
+      "loss": 2.9217,
+      "step": 41877
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6986700296401978,
+      "learning_rate": 0.00013599221687464005,
+      "loss": 2.8404,
+      "step": 41878
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7709964513778687,
+      "learning_rate": 0.00013597932534646718,
+      "loss": 2.8286,
+      "step": 41879
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6718994379043579,
+      "learning_rate": 0.0001359664342502931,
+      "loss": 2.9081,
+      "step": 41880
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7286249399185181,
+      "learning_rate": 0.00013595354358615194,
+      "loss": 2.9856,
+      "step": 41881
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6866543889045715,
+      "learning_rate": 0.00013594065335407752,
+      "loss": 2.8534,
+      "step": 41882
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6790507435798645,
+      "learning_rate": 0.00013592776355410394,
+      "loss": 2.9359,
+      "step": 41883
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6834301948547363,
+      "learning_rate": 0.0001359148741862649,
+      "loss": 2.9917,
+      "step": 41884
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6941182613372803,
+      "learning_rate": 0.00013590198525059466,
+      "loss": 2.926,
+      "step": 41885
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7055644392967224,
+      "learning_rate": 0.00013588909674712698,
+      "loss": 2.7773,
+      "step": 41886
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6956052780151367,
+      "learning_rate": 0.0001358762086758957,
+      "loss": 2.8471,
+      "step": 41887
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6483541131019592,
+      "learning_rate": 0.00013586332103693506,
+      "loss": 3.033,
+      "step": 41888
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6963681578636169,
+      "learning_rate": 0.00013585043383027874,
+      "loss": 3.1238,
+      "step": 41889
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6419037580490112,
+      "learning_rate": 0.00013583754705596076,
+      "loss": 2.8311,
+      "step": 41890
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6917277574539185,
+      "learning_rate": 0.0001358246607140152,
+      "loss": 2.8747,
+      "step": 41891
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7247151732444763,
+      "learning_rate": 0.00013581177480447576,
+      "loss": 2.9386,
+      "step": 41892
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.71748948097229,
+      "learning_rate": 0.00013579888932737665,
+      "loss": 3.047,
+      "step": 41893
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6673287749290466,
+      "learning_rate": 0.00013578600428275166,
+      "loss": 3.0716,
+      "step": 41894
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7329942584037781,
+      "learning_rate": 0.00013577311967063462,
+      "loss": 2.7382,
+      "step": 41895
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.652654767036438,
+      "learning_rate": 0.00013576023549105966,
+      "loss": 3.2103,
+      "step": 41896
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.730176568031311,
+      "learning_rate": 0.00013574735174406055,
+      "loss": 2.8715,
+      "step": 41897
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6587347984313965,
+      "learning_rate": 0.0001357344684296714,
+      "loss": 2.7973,
+      "step": 41898
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6811947822570801,
+      "learning_rate": 0.00013572158554792595,
+      "loss": 2.9729,
+      "step": 41899
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6677283644676208,
+      "learning_rate": 0.0001357087030988583,
+      "loss": 3.0673,
+      "step": 41900
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7097472548484802,
+      "learning_rate": 0.00013569582108250236,
+      "loss": 2.7919,
+      "step": 41901
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6857821345329285,
+      "learning_rate": 0.00013568293949889186,
+      "loss": 3.1361,
+      "step": 41902
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6500460505485535,
+      "learning_rate": 0.000135670058348061,
+      "loss": 2.869,
+      "step": 41903
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6371486186981201,
+      "learning_rate": 0.00013565717763004348,
+      "loss": 2.9958,
+      "step": 41904
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6974950432777405,
+      "learning_rate": 0.0001356442973448733,
+      "loss": 2.9836,
+      "step": 41905
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.727595329284668,
+      "learning_rate": 0.00013563141749258455,
+      "loss": 2.9796,
+      "step": 41906
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.681317150592804,
+      "learning_rate": 0.000135618538073211,
+      "loss": 2.7705,
+      "step": 41907
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6713190674781799,
+      "learning_rate": 0.00013560565908678654,
+      "loss": 2.934,
+      "step": 41908
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.698500394821167,
+      "learning_rate": 0.00013559278053334507,
+      "loss": 2.9041,
+      "step": 41909
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6464803814888,
+      "learning_rate": 0.0001355799024129206,
+      "loss": 2.9814,
+      "step": 41910
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7684870362281799,
+      "learning_rate": 0.00013556702472554712,
+      "loss": 2.9662,
+      "step": 41911
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7648671269416809,
+      "learning_rate": 0.00013555414747125832,
+      "loss": 2.8802,
+      "step": 41912
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6603399515151978,
+      "learning_rate": 0.00013554127065008837,
+      "loss": 3.0558,
+      "step": 41913
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6440830230712891,
+      "learning_rate": 0.00013552839426207093,
+      "loss": 2.8296,
+      "step": 41914
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6936287879943848,
+      "learning_rate": 0.00013551551830724018,
+      "loss": 2.9706,
+      "step": 41915
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6846458911895752,
+      "learning_rate": 0.0001355026427856299,
+      "loss": 3.0088,
+      "step": 41916
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6697417497634888,
+      "learning_rate": 0.00013548976769727386,
+      "loss": 2.8968,
+      "step": 41917
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7135060429573059,
+      "learning_rate": 0.00013547689304220625,
+      "loss": 3.0002,
+      "step": 41918
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.69883131980896,
+      "learning_rate": 0.0001354640188204607,
+      "loss": 2.8384,
+      "step": 41919
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.663644552230835,
+      "learning_rate": 0.00013545114503207142,
+      "loss": 2.6999,
+      "step": 41920
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6672549247741699,
+      "learning_rate": 0.00013543827167707202,
+      "loss": 3.0655,
+      "step": 41921
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7014299035072327,
+      "learning_rate": 0.00013542539875549664,
+      "loss": 2.8712,
+      "step": 41922
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6401073336601257,
+      "learning_rate": 0.00013541252626737908,
+      "loss": 2.9725,
+      "step": 41923
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7074952125549316,
+      "learning_rate": 0.00013539965421275317,
+      "loss": 2.8557,
+      "step": 41924
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6953228712081909,
+      "learning_rate": 0.00013538678259165297,
+      "loss": 2.674,
+      "step": 41925
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6561875939369202,
+      "learning_rate": 0.00013537391140411222,
+      "loss": 3.0387,
+      "step": 41926
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.70853191614151,
+      "learning_rate": 0.00013536104065016492,
+      "loss": 2.6825,
+      "step": 41927
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7176726460456848,
+      "learning_rate": 0.00013534817032984504,
+      "loss": 2.6226,
+      "step": 41928
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7297140955924988,
+      "learning_rate": 0.0001353353004431864,
+      "loss": 2.9366,
+      "step": 41929
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6884056329727173,
+      "learning_rate": 0.00013532243099022287,
+      "loss": 2.9259,
+      "step": 41930
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7293509244918823,
+      "learning_rate": 0.0001353095619709883,
+      "loss": 2.9394,
+      "step": 41931
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7084832191467285,
+      "learning_rate": 0.0001352966933855166,
+      "loss": 3.1906,
+      "step": 41932
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7524933815002441,
+      "learning_rate": 0.00013528382523384183,
+      "loss": 2.9092,
+      "step": 41933
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6961275935173035,
+      "learning_rate": 0.00013527095751599767,
+      "loss": 2.8531,
+      "step": 41934
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6843053698539734,
+      "learning_rate": 0.0001352580902320182,
+      "loss": 2.7244,
+      "step": 41935
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7608031034469604,
+      "learning_rate": 0.0001352452233819371,
+      "loss": 3.1026,
+      "step": 41936
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6482999324798584,
+      "learning_rate": 0.0001352323569657885,
+      "loss": 2.9609,
+      "step": 41937
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6896520853042603,
+      "learning_rate": 0.00013521949098360616,
+      "loss": 2.9405,
+      "step": 41938
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7269614934921265,
+      "learning_rate": 0.00013520662543542385,
+      "loss": 2.7086,
+      "step": 41939
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6921659111976624,
+      "learning_rate": 0.00013519376032127567,
+      "loss": 2.7784,
+      "step": 41940
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6634232401847839,
+      "learning_rate": 0.00013518089564119534,
+      "loss": 3.0583,
+      "step": 41941
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7075629830360413,
+      "learning_rate": 0.00013516803139521677,
+      "loss": 2.8788,
+      "step": 41942
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7123188972473145,
+      "learning_rate": 0.000135155167583374,
+      "loss": 2.802,
+      "step": 41943
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6718761920928955,
+      "learning_rate": 0.00013514230420570085,
+      "loss": 2.9926,
+      "step": 41944
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6959552764892578,
+      "learning_rate": 0.00013512944126223107,
+      "loss": 2.9625,
+      "step": 41945
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7005223035812378,
+      "learning_rate": 0.00013511657875299853,
+      "loss": 2.8976,
+      "step": 41946
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6967735290527344,
+      "learning_rate": 0.00013510371667803717,
+      "loss": 2.9729,
+      "step": 41947
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6776131391525269,
+      "learning_rate": 0.00013509085503738102,
+      "loss": 2.8577,
+      "step": 41948
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6918829679489136,
+      "learning_rate": 0.0001350779938310637,
+      "loss": 3.0947,
+      "step": 41949
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7643651366233826,
+      "learning_rate": 0.00013506513305911933,
+      "loss": 3.176,
+      "step": 41950
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6947302222251892,
+      "learning_rate": 0.00013505227272158163,
+      "loss": 2.9131,
+      "step": 41951
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7408885359764099,
+      "learning_rate": 0.00013503941281848439,
+      "loss": 2.9874,
+      "step": 41952
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7771589159965515,
+      "learning_rate": 0.00013502655334986173,
+      "loss": 3.0083,
+      "step": 41953
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6549042463302612,
+      "learning_rate": 0.00013501369431574725,
+      "loss": 2.8394,
+      "step": 41954
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.687390148639679,
+      "learning_rate": 0.00013500083571617508,
+      "loss": 2.7261,
+      "step": 41955
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6740281581878662,
+      "learning_rate": 0.0001349879775511788,
+      "loss": 2.8224,
+      "step": 41956
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6710662841796875,
+      "learning_rate": 0.00013497511982079246,
+      "loss": 2.709,
+      "step": 41957
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7352544665336609,
+      "learning_rate": 0.00013496226252504998,
+      "loss": 3.0141,
+      "step": 41958
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6972978115081787,
+      "learning_rate": 0.0001349494056639852,
+      "loss": 2.9088,
+      "step": 41959
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6581162214279175,
+      "learning_rate": 0.00013493654923763183,
+      "loss": 2.8458,
+      "step": 41960
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6925591230392456,
+      "learning_rate": 0.00013492369324602375,
+      "loss": 2.9503,
+      "step": 41961
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.663356602191925,
+      "learning_rate": 0.00013491083768919494,
+      "loss": 2.8988,
+      "step": 41962
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7245420813560486,
+      "learning_rate": 0.00013489798256717926,
+      "loss": 2.8092,
+      "step": 41963
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6705074310302734,
+      "learning_rate": 0.00013488512788001043,
+      "loss": 2.8578,
+      "step": 41964
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6748921871185303,
+      "learning_rate": 0.00013487227362772252,
+      "loss": 3.1966,
+      "step": 41965
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.728326141834259,
+      "learning_rate": 0.00013485941981034922,
+      "loss": 2.8332,
+      "step": 41966
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7034401297569275,
+      "learning_rate": 0.00013484656642792436,
+      "loss": 2.9848,
+      "step": 41967
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.68290776014328,
+      "learning_rate": 0.00013483371348048197,
+      "loss": 2.8918,
+      "step": 41968
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6936345100402832,
+      "learning_rate": 0.00013482086096805565,
+      "loss": 2.779,
+      "step": 41969
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7086421251296997,
+      "learning_rate": 0.00013480800889067955,
+      "loss": 2.9097,
+      "step": 41970
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7316939830780029,
+      "learning_rate": 0.00013479515724838722,
+      "loss": 2.8459,
+      "step": 41971
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6890463829040527,
+      "learning_rate": 0.00013478230604121265,
+      "loss": 3.0289,
+      "step": 41972
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6866118907928467,
+      "learning_rate": 0.00013476945526918978,
+      "loss": 2.7847,
+      "step": 41973
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7601082921028137,
+      "learning_rate": 0.00013475660493235237,
+      "loss": 2.8525,
+      "step": 41974
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7042220234870911,
+      "learning_rate": 0.00013474375503073433,
+      "loss": 2.945,
+      "step": 41975
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6664708852767944,
+      "learning_rate": 0.00013473090556436927,
+      "loss": 2.9917,
+      "step": 41976
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6848310232162476,
+      "learning_rate": 0.00013471805653329118,
+      "loss": 2.8884,
+      "step": 41977
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6613977551460266,
+      "learning_rate": 0.00013470520793753407,
+      "loss": 2.9344,
+      "step": 41978
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7326992750167847,
+      "learning_rate": 0.00013469235977713148,
+      "loss": 2.8053,
+      "step": 41979
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6662197113037109,
+      "learning_rate": 0.00013467951205211755,
+      "loss": 2.8492,
+      "step": 41980
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6890004873275757,
+      "learning_rate": 0.00013466666476252595,
+      "loss": 2.9529,
+      "step": 41981
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6676129102706909,
+      "learning_rate": 0.00013465381790839042,
+      "loss": 2.8345,
+      "step": 41982
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.676196813583374,
+      "learning_rate": 0.00013464097148974504,
+      "loss": 3.2066,
+      "step": 41983
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6789147257804871,
+      "learning_rate": 0.0001346281255066234,
+      "loss": 3.0088,
+      "step": 41984
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7471140623092651,
+      "learning_rate": 0.00013461527995905955,
+      "loss": 2.9146,
+      "step": 41985
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6926723718643188,
+      "learning_rate": 0.00013460243484708713,
+      "loss": 2.9489,
+      "step": 41986
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7110167145729065,
+      "learning_rate": 0.00013458959017074017,
+      "loss": 2.8836,
+      "step": 41987
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7335017323493958,
+      "learning_rate": 0.0001345767459300524,
+      "loss": 2.9397,
+      "step": 41988
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7071158289909363,
+      "learning_rate": 0.0001345639021250575,
+      "loss": 2.8792,
+      "step": 41989
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7941440343856812,
+      "learning_rate": 0.0001345510587557896,
+      "loss": 2.9275,
+      "step": 41990
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6561496257781982,
+      "learning_rate": 0.00013453821582228224,
+      "loss": 2.9598,
+      "step": 41991
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7104381918907166,
+      "learning_rate": 0.0001345253733245694,
+      "loss": 2.897,
+      "step": 41992
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.702156662940979,
+      "learning_rate": 0.00013451253126268496,
+      "loss": 2.8701,
+      "step": 41993
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7097292542457581,
+      "learning_rate": 0.00013449968963666253,
+      "loss": 2.7465,
+      "step": 41994
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6726009249687195,
+      "learning_rate": 0.00013448684844653624,
+      "loss": 2.6797,
+      "step": 41995
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6798008680343628,
+      "learning_rate": 0.0001344740076923397,
+      "loss": 2.8157,
+      "step": 41996
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7259616255760193,
+      "learning_rate": 0.00013446116737410676,
+      "loss": 2.9134,
+      "step": 41997
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7127224802970886,
+      "learning_rate": 0.00013444832749187114,
+      "loss": 3.0528,
+      "step": 41998
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6722402572631836,
+      "learning_rate": 0.00013443548804566678,
+      "loss": 2.7478,
+      "step": 41999
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7285507321357727,
+      "learning_rate": 0.0001344226490355276,
+      "loss": 2.9906,
+      "step": 42000
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6551629304885864,
+      "learning_rate": 0.00013440981046148717,
+      "loss": 2.8337,
+      "step": 42001
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.678065836429596,
+      "learning_rate": 0.00013439697232357954,
+      "loss": 2.8061,
+      "step": 42002
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7075507044792175,
+      "learning_rate": 0.00013438413462183842,
+      "loss": 2.7313,
+      "step": 42003
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.658933699131012,
+      "learning_rate": 0.00013437129735629748,
+      "loss": 2.8594,
+      "step": 42004
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6496484279632568,
+      "learning_rate": 0.0001343584605269908,
+      "loss": 3.0021,
+      "step": 42005
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6887893080711365,
+      "learning_rate": 0.00013434562413395195,
+      "loss": 2.803,
+      "step": 42006
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6488357782363892,
+      "learning_rate": 0.00013433278817721496,
+      "loss": 2.9725,
+      "step": 42007
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7380465865135193,
+      "learning_rate": 0.00013431995265681337,
+      "loss": 3.0395,
+      "step": 42008
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.703506350517273,
+      "learning_rate": 0.0001343071175727813,
+      "loss": 3.0208,
+      "step": 42009
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7320290803909302,
+      "learning_rate": 0.00013429428292515238,
+      "loss": 2.8408,
+      "step": 42010
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6865428686141968,
+      "learning_rate": 0.0001342814487139603,
+      "loss": 3.1323,
+      "step": 42011
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6920920014381409,
+      "learning_rate": 0.00013426861493923915,
+      "loss": 2.8396,
+      "step": 42012
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7023778557777405,
+      "learning_rate": 0.00013425578160102238,
+      "loss": 2.9399,
+      "step": 42013
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.648321270942688,
+      "learning_rate": 0.00013424294869934407,
+      "loss": 3.1877,
+      "step": 42014
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6870456337928772,
+      "learning_rate": 0.000134230116234238,
+      "loss": 3.2613,
+      "step": 42015
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7093237638473511,
+      "learning_rate": 0.00013421728420573778,
+      "loss": 2.8811,
+      "step": 42016
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6749858260154724,
+      "learning_rate": 0.00013420445261387747,
+      "loss": 2.7412,
+      "step": 42017
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7406046986579895,
+      "learning_rate": 0.0001341916214586907,
+      "loss": 2.8874,
+      "step": 42018
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6919505596160889,
+      "learning_rate": 0.00013417879074021117,
+      "loss": 2.9656,
+      "step": 42019
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7033610939979553,
+      "learning_rate": 0.00013416596045847293,
+      "loss": 3.0286,
+      "step": 42020
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6647442579269409,
+      "learning_rate": 0.00013415313061350948,
+      "loss": 2.861,
+      "step": 42021
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6930309534072876,
+      "learning_rate": 0.00013414030120535491,
+      "loss": 3.1232,
+      "step": 42022
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6791087985038757,
+      "learning_rate": 0.00013412747223404275,
+      "loss": 2.9181,
+      "step": 42023
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6947936415672302,
+      "learning_rate": 0.00013411464369960706,
+      "loss": 2.7641,
+      "step": 42024
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7209773659706116,
+      "learning_rate": 0.00013410181560208142,
+      "loss": 2.9464,
+      "step": 42025
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7256671190261841,
+      "learning_rate": 0.00013408898794149956,
+      "loss": 2.8287,
+      "step": 42026
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.67732834815979,
+      "learning_rate": 0.00013407616071789552,
+      "loss": 3.0294,
+      "step": 42027
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6837236285209656,
+      "learning_rate": 0.00013406333393130277,
+      "loss": 2.7781,
+      "step": 42028
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6771765351295471,
+      "learning_rate": 0.0001340505075817553,
+      "loss": 2.9683,
+      "step": 42029
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.685924768447876,
+      "learning_rate": 0.00013403768166928698,
+      "loss": 2.9958,
+      "step": 42030
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6353333592414856,
+      "learning_rate": 0.00013402485619393146,
+      "loss": 2.9384,
+      "step": 42031
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.697657585144043,
+      "learning_rate": 0.00013401203115572254,
+      "loss": 2.921,
+      "step": 42032
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6873392462730408,
+      "learning_rate": 0.00013399920655469383,
+      "loss": 2.7485,
+      "step": 42033
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6944345235824585,
+      "learning_rate": 0.0001339863823908793,
+      "loss": 2.8502,
+      "step": 42034
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6621580719947815,
+      "learning_rate": 0.0001339735586643128,
+      "loss": 2.8554,
+      "step": 42035
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6833303570747375,
+      "learning_rate": 0.00013396073537502782,
+      "loss": 3.2196,
+      "step": 42036
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6794368028640747,
+      "learning_rate": 0.00013394791252305848,
+      "loss": 2.8268,
+      "step": 42037
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7149304747581482,
+      "learning_rate": 0.00013393509010843827,
+      "loss": 3.1625,
+      "step": 42038
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6875823736190796,
+      "learning_rate": 0.00013392226813120118,
+      "loss": 2.8485,
+      "step": 42039
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7069306969642639,
+      "learning_rate": 0.00013390944659138085,
+      "loss": 3.0752,
+      "step": 42040
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6851629614830017,
+      "learning_rate": 0.000133896625489011,
+      "loss": 3.107,
+      "step": 42041
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6919379234313965,
+      "learning_rate": 0.00013388380482412554,
+      "loss": 3.0562,
+      "step": 42042
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.724163830280304,
+      "learning_rate": 0.0001338709845967581,
+      "loss": 3.1261,
+      "step": 42043
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6805170178413391,
+      "learning_rate": 0.00013385816480694247,
+      "loss": 2.8018,
+      "step": 42044
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.660524845123291,
+      "learning_rate": 0.0001338453454547126,
+      "loss": 2.8098,
+      "step": 42045
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6695048809051514,
+      "learning_rate": 0.00013383252654010212,
+      "loss": 3.0506,
+      "step": 42046
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6940860748291016,
+      "learning_rate": 0.00013381970806314474,
+      "loss": 2.9447,
+      "step": 42047
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6937212944030762,
+      "learning_rate": 0.0001338068900238742,
+      "loss": 2.954,
+      "step": 42048
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6832905411720276,
+      "learning_rate": 0.0001337940724223243,
+      "loss": 2.8383,
+      "step": 42049
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6810662746429443,
+      "learning_rate": 0.00013378125525852897,
+      "loss": 2.6696,
+      "step": 42050
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6970690488815308,
+      "learning_rate": 0.0001337684385325217,
+      "loss": 3.0258,
+      "step": 42051
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6836695671081543,
+      "learning_rate": 0.00013375562224433646,
+      "loss": 3.0219,
+      "step": 42052
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6802341341972351,
+      "learning_rate": 0.00013374280639400694,
+      "loss": 3.1268,
+      "step": 42053
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6934910416603088,
+      "learning_rate": 0.00013372999098156674,
+      "loss": 2.9171,
+      "step": 42054
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6853800415992737,
+      "learning_rate": 0.00013371717600704986,
+      "loss": 2.9585,
+      "step": 42055
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.9536087512969971,
+      "learning_rate": 0.00013370436147048986,
+      "loss": 2.9071,
+      "step": 42056
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6992372274398804,
+      "learning_rate": 0.00013369154737192065,
+      "loss": 2.9655,
+      "step": 42057
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6866870522499084,
+      "learning_rate": 0.00013367873371137576,
+      "loss": 2.9159,
+      "step": 42058
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6795086860656738,
+      "learning_rate": 0.0001336659204888891,
+      "loss": 2.8051,
+      "step": 42059
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6785724759101868,
+      "learning_rate": 0.00013365310770449454,
+      "loss": 3.0133,
+      "step": 42060
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6523509621620178,
+      "learning_rate": 0.00013364029535822564,
+      "loss": 2.8145,
+      "step": 42061
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7120859026908875,
+      "learning_rate": 0.0001336274834501162,
+      "loss": 3.0528,
+      "step": 42062
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6999312043190002,
+      "learning_rate": 0.00013361467198019984,
+      "loss": 2.9157,
+      "step": 42063
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6902523040771484,
+      "learning_rate": 0.00013360186094851038,
+      "loss": 2.6057,
+      "step": 42064
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6437488198280334,
+      "learning_rate": 0.00013358905035508172,
+      "loss": 2.9131,
+      "step": 42065
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6648966670036316,
+      "learning_rate": 0.0001335762401999474,
+      "loss": 2.6556,
+      "step": 42066
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7151343822479248,
+      "learning_rate": 0.00013356343048314133,
+      "loss": 2.9348,
+      "step": 42067
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6936889290809631,
+      "learning_rate": 0.0001335506212046971,
+      "loss": 2.8895,
+      "step": 42068
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6722537279129028,
+      "learning_rate": 0.00013353781236464846,
+      "loss": 3.094,
+      "step": 42069
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6585366725921631,
+      "learning_rate": 0.00013352500396302927,
+      "loss": 2.9741,
+      "step": 42070
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7267591953277588,
+      "learning_rate": 0.0001335121959998731,
+      "loss": 2.9534,
+      "step": 42071
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6568361520767212,
+      "learning_rate": 0.00013349938847521383,
+      "loss": 2.7674,
+      "step": 42072
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.700718343257904,
+      "learning_rate": 0.00013348658138908503,
+      "loss": 2.9192,
+      "step": 42073
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7025381326675415,
+      "learning_rate": 0.00013347377474152064,
+      "loss": 3.0146,
+      "step": 42074
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7157812714576721,
+      "learning_rate": 0.00013346096853255427,
+      "loss": 2.8323,
+      "step": 42075
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6481377482414246,
+      "learning_rate": 0.00013344816276221957,
+      "loss": 3.0237,
+      "step": 42076
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6517822742462158,
+      "learning_rate": 0.00013343535743055043,
+      "loss": 2.7866,
+      "step": 42077
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6869397759437561,
+      "learning_rate": 0.00013342255253758042,
+      "loss": 2.9609,
+      "step": 42078
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6711219549179077,
+      "learning_rate": 0.00013340974808334345,
+      "loss": 3.0348,
+      "step": 42079
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7152694463729858,
+      "learning_rate": 0.00013339694406787304,
+      "loss": 3.1945,
+      "step": 42080
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7110965251922607,
+      "learning_rate": 0.00013338414049120302,
+      "loss": 2.8687,
+      "step": 42081
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6851297616958618,
+      "learning_rate": 0.0001333713373533672,
+      "loss": 3.0068,
+      "step": 42082
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6575678586959839,
+      "learning_rate": 0.00013335853465439924,
+      "loss": 2.7863,
+      "step": 42083
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6966585516929626,
+      "learning_rate": 0.0001333457323943328,
+      "loss": 2.9452,
+      "step": 42084
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6733703017234802,
+      "learning_rate": 0.0001333329305732015,
+      "loss": 3.0046,
+      "step": 42085
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7381994128227234,
+      "learning_rate": 0.00013332012919103922,
+      "loss": 2.8515,
+      "step": 42086
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6656725406646729,
+      "learning_rate": 0.00013330732824787974,
+      "loss": 2.9327,
+      "step": 42087
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6955260634422302,
+      "learning_rate": 0.00013329452774375657,
+      "loss": 3.0594,
+      "step": 42088
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6607492566108704,
+      "learning_rate": 0.00013328172767870363,
+      "loss": 2.8572,
+      "step": 42089
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6896944046020508,
+      "learning_rate": 0.00013326892805275457,
+      "loss": 2.8783,
+      "step": 42090
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7322929501533508,
+      "learning_rate": 0.00013325612886594293,
+      "loss": 2.9448,
+      "step": 42091
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6637055277824402,
+      "learning_rate": 0.00013324333011830266,
+      "loss": 2.9994,
+      "step": 42092
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.66081303358078,
+      "learning_rate": 0.00013323053180986724,
+      "loss": 2.9542,
+      "step": 42093
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7156060934066772,
+      "learning_rate": 0.00013321773394067064,
+      "loss": 2.8346,
+      "step": 42094
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7147414088249207,
+      "learning_rate": 0.0001332049365107463,
+      "loss": 2.9021,
+      "step": 42095
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6858595013618469,
+      "learning_rate": 0.0001331921395201281,
+      "loss": 2.8172,
+      "step": 42096
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6633896231651306,
+      "learning_rate": 0.00013317934296884978,
+      "loss": 2.9635,
+      "step": 42097
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6843210458755493,
+      "learning_rate": 0.00013316654685694497,
+      "loss": 3.0135,
+      "step": 42098
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7373406887054443,
+      "learning_rate": 0.00013315375118444733,
+      "loss": 2.8179,
+      "step": 42099
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6996890306472778,
+      "learning_rate": 0.00013314095595139057,
+      "loss": 2.9772,
+      "step": 42100
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6776704788208008,
+      "learning_rate": 0.00013312816115780836,
+      "loss": 2.874,
+      "step": 42101
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7402383089065552,
+      "learning_rate": 0.00013311536680373458,
+      "loss": 3.0208,
+      "step": 42102
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6806477308273315,
+      "learning_rate": 0.0001331025728892027,
+      "loss": 2.8178,
+      "step": 42103
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7480725049972534,
+      "learning_rate": 0.00013308977941424663,
+      "loss": 2.759,
+      "step": 42104
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6960058808326721,
+      "learning_rate": 0.00013307698637889996,
+      "loss": 3.0031,
+      "step": 42105
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7045592665672302,
+      "learning_rate": 0.00013306419378319626,
+      "loss": 2.8392,
+      "step": 42106
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6933001279830933,
+      "learning_rate": 0.00013305140162716948,
+      "loss": 2.8758,
+      "step": 42107
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6979026794433594,
+      "learning_rate": 0.00013303860991085306,
+      "loss": 2.8514,
+      "step": 42108
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6984500885009766,
+      "learning_rate": 0.00013302581863428094,
+      "loss": 2.853,
+      "step": 42109
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7042974829673767,
+      "learning_rate": 0.00013301302779748653,
+      "loss": 2.9368,
+      "step": 42110
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6589822173118591,
+      "learning_rate": 0.00013300023740050378,
+      "loss": 2.8031,
+      "step": 42111
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6807518005371094,
+      "learning_rate": 0.0001329874474433663,
+      "loss": 3.1159,
+      "step": 42112
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7369489669799805,
+      "learning_rate": 0.00013297465792610761,
+      "loss": 3.026,
+      "step": 42113
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6795954704284668,
+      "learning_rate": 0.00013296186884876164,
+      "loss": 2.8004,
+      "step": 42114
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6816520690917969,
+      "learning_rate": 0.00013294908021136185,
+      "loss": 2.9879,
+      "step": 42115
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6890973448753357,
+      "learning_rate": 0.00013293629201394203,
+      "loss": 2.9887,
+      "step": 42116
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6854380369186401,
+      "learning_rate": 0.00013292350425653596,
+      "loss": 3.0083,
+      "step": 42117
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7107696533203125,
+      "learning_rate": 0.00013291071693917713,
+      "loss": 2.912,
+      "step": 42118
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6896004676818848,
+      "learning_rate": 0.00013289793006189942,
+      "loss": 2.9846,
+      "step": 42119
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7244950532913208,
+      "learning_rate": 0.0001328851436247364,
+      "loss": 2.9464,
+      "step": 42120
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6931073069572449,
+      "learning_rate": 0.00013287235762772165,
+      "loss": 2.7884,
+      "step": 42121
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6745514869689941,
+      "learning_rate": 0.00013285957207088907,
+      "loss": 2.7589,
+      "step": 42122
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.737249493598938,
+      "learning_rate": 0.00013284678695427206,
+      "loss": 2.9354,
+      "step": 42123
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6523367762565613,
+      "learning_rate": 0.0001328340022779046,
+      "loss": 2.7846,
+      "step": 42124
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6890686750411987,
+      "learning_rate": 0.0001328212180418201,
+      "loss": 3.1741,
+      "step": 42125
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7183402180671692,
+      "learning_rate": 0.00013280843424605239,
+      "loss": 2.9694,
+      "step": 42126
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6588950157165527,
+      "learning_rate": 0.0001327956508906351,
+      "loss": 2.9488,
+      "step": 42127
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6817463040351868,
+      "learning_rate": 0.00013278286797560184,
+      "loss": 2.8579,
+      "step": 42128
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6694334745407104,
+      "learning_rate": 0.00013277008550098638,
+      "loss": 2.9406,
+      "step": 42129
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6861422657966614,
+      "learning_rate": 0.0001327573034668222,
+      "loss": 3.0737,
+      "step": 42130
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6940646171569824,
+      "learning_rate": 0.00013274452187314316,
+      "loss": 2.8372,
+      "step": 42131
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6737503409385681,
+      "learning_rate": 0.0001327317407199829,
+      "loss": 3.0897,
+      "step": 42132
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.692583441734314,
+      "learning_rate": 0.00013271896000737512,
+      "loss": 2.9895,
+      "step": 42133
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7276341915130615,
+      "learning_rate": 0.00013270617973535332,
+      "loss": 2.7632,
+      "step": 42134
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6521527767181396,
+      "learning_rate": 0.0001326933999039512,
+      "loss": 2.7446,
+      "step": 42135
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7015560865402222,
+      "learning_rate": 0.00013268062051320246,
+      "loss": 3.0649,
+      "step": 42136
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6922124028205872,
+      "learning_rate": 0.00013266784156314087,
+      "loss": 2.8697,
+      "step": 42137
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7114979028701782,
+      "learning_rate": 0.00013265506305379983,
+      "loss": 2.9049,
+      "step": 42138
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6223366260528564,
+      "learning_rate": 0.0001326422849852133,
+      "loss": 2.7975,
+      "step": 42139
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7140401005744934,
+      "learning_rate": 0.00013262950735741467,
+      "loss": 3.1222,
+      "step": 42140
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6884499192237854,
+      "learning_rate": 0.00013261673017043782,
+      "loss": 2.8993,
+      "step": 42141
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6920923590660095,
+      "learning_rate": 0.00013260395342431627,
+      "loss": 2.9722,
+      "step": 42142
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.662397563457489,
+      "learning_rate": 0.00013259117711908358,
+      "loss": 3.0986,
+      "step": 42143
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6610293388366699,
+      "learning_rate": 0.00013257840125477363,
+      "loss": 3.0419,
+      "step": 42144
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7399603724479675,
+      "learning_rate": 0.00013256562583141985,
+      "loss": 2.8981,
+      "step": 42145
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.681763768196106,
+      "learning_rate": 0.00013255285084905598,
+      "loss": 2.756,
+      "step": 42146
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7297256588935852,
+      "learning_rate": 0.0001325400763077158,
+      "loss": 3.0475,
+      "step": 42147
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6886538863182068,
+      "learning_rate": 0.0001325273022074328,
+      "loss": 3.1051,
+      "step": 42148
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7075888514518738,
+      "learning_rate": 0.00013251452854824068,
+      "loss": 3.0244,
+      "step": 42149
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6723986864089966,
+      "learning_rate": 0.00013250175533017293,
+      "loss": 3.032,
+      "step": 42150
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6775681972503662,
+      "learning_rate": 0.00013248898255326346,
+      "loss": 2.9657,
+      "step": 42151
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6784015893936157,
+      "learning_rate": 0.00013247621021754562,
+      "loss": 2.8706,
+      "step": 42152
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6836172342300415,
+      "learning_rate": 0.00013246343832305324,
+      "loss": 2.9455,
+      "step": 42153
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7069783806800842,
+      "learning_rate": 0.00013245066686982,
+      "loss": 2.92,
+      "step": 42154
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6457093358039856,
+      "learning_rate": 0.00013243789585787948,
+      "loss": 2.9114,
+      "step": 42155
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7186350226402283,
+      "learning_rate": 0.0001324251252872653,
+      "loss": 2.7841,
+      "step": 42156
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.672428548336029,
+      "learning_rate": 0.00013241235515801093,
+      "loss": 3.0373,
+      "step": 42157
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7136212587356567,
+      "learning_rate": 0.0001323995854701502,
+      "loss": 2.8389,
+      "step": 42158
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6720640659332275,
+      "learning_rate": 0.0001323868162237168,
+      "loss": 2.7359,
+      "step": 42159
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6694479584693909,
+      "learning_rate": 0.0001323740474187442,
+      "loss": 2.7698,
+      "step": 42160
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6907969117164612,
+      "learning_rate": 0.0001323612790552662,
+      "loss": 2.7868,
+      "step": 42161
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6790691018104553,
+      "learning_rate": 0.0001323485111333162,
+      "loss": 3.055,
+      "step": 42162
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7087593674659729,
+      "learning_rate": 0.00013233574365292803,
+      "loss": 2.9253,
+      "step": 42163
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6997819542884827,
+      "learning_rate": 0.0001323229766141353,
+      "loss": 2.972,
+      "step": 42164
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7132028341293335,
+      "learning_rate": 0.00013231021001697146,
+      "loss": 2.8296,
+      "step": 42165
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6946109533309937,
+      "learning_rate": 0.00013229744386147036,
+      "loss": 2.8329,
+      "step": 42166
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7032156586647034,
+      "learning_rate": 0.0001322846781476654,
+      "loss": 2.7633,
+      "step": 42167
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6949759721755981,
+      "learning_rate": 0.0001322719128755903,
+      "loss": 3.0027,
+      "step": 42168
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6636202335357666,
+      "learning_rate": 0.00013225914804527883,
+      "loss": 2.9693,
+      "step": 42169
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6825640797615051,
+      "learning_rate": 0.00013224638365676451,
+      "loss": 3.0983,
+      "step": 42170
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.669562816619873,
+      "learning_rate": 0.0001322336197100809,
+      "loss": 2.9744,
+      "step": 42171
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6761650443077087,
+      "learning_rate": 0.00013222085620526153,
+      "loss": 2.8963,
+      "step": 42172
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7180191278457642,
+      "learning_rate": 0.00013220809314234017,
+      "loss": 2.925,
+      "step": 42173
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6580066084861755,
+      "learning_rate": 0.00013219533052135051,
+      "loss": 2.7424,
+      "step": 42174
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7591055035591125,
+      "learning_rate": 0.0001321825683423259,
+      "loss": 2.7706,
+      "step": 42175
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6858645677566528,
+      "learning_rate": 0.00013216980660530025,
+      "loss": 2.8627,
+      "step": 42176
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7282005548477173,
+      "learning_rate": 0.000132157045310307,
+      "loss": 2.7958,
+      "step": 42177
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6546103954315186,
+      "learning_rate": 0.00013214428445737974,
+      "loss": 2.7244,
+      "step": 42178
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7295477986335754,
+      "learning_rate": 0.00013213152404655218,
+      "loss": 2.9721,
+      "step": 42179
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6619624495506287,
+      "learning_rate": 0.00013211876407785779,
+      "loss": 2.9897,
+      "step": 42180
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7149553298950195,
+      "learning_rate": 0.00013210600455133037,
+      "loss": 2.7235,
+      "step": 42181
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6860978603363037,
+      "learning_rate": 0.00013209324546700334,
+      "loss": 3.0681,
+      "step": 42182
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6987245678901672,
+      "learning_rate": 0.00013208048682491038,
+      "loss": 2.7269,
+      "step": 42183
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6757438778877258,
+      "learning_rate": 0.0001320677286250852,
+      "loss": 2.8652,
+      "step": 42184
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6813808679580688,
+      "learning_rate": 0.0001320549708675613,
+      "loss": 2.8605,
+      "step": 42185
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6611071825027466,
+      "learning_rate": 0.0001320422135523723,
+      "loss": 2.7863,
+      "step": 42186
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6753425598144531,
+      "learning_rate": 0.00013202945667955163,
+      "loss": 2.9375,
+      "step": 42187
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6801412105560303,
+      "learning_rate": 0.00013201670024913308,
+      "loss": 2.9191,
+      "step": 42188
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6811955571174622,
+      "learning_rate": 0.00013200394426115034,
+      "loss": 2.8201,
+      "step": 42189
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6670415997505188,
+      "learning_rate": 0.00013199118871563674,
+      "loss": 2.7303,
+      "step": 42190
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7115445733070374,
+      "learning_rate": 0.00013197843361262613,
+      "loss": 2.8177,
+      "step": 42191
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6806511878967285,
+      "learning_rate": 0.00013196567895215202,
+      "loss": 3.0169,
+      "step": 42192
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6739731431007385,
+      "learning_rate": 0.00013195292473424782,
+      "loss": 2.9967,
+      "step": 42193
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6881250143051147,
+      "learning_rate": 0.0001319401709589474,
+      "loss": 2.8704,
+      "step": 42194
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6715559363365173,
+      "learning_rate": 0.0001319274176262841,
+      "loss": 2.9177,
+      "step": 42195
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6683921813964844,
+      "learning_rate": 0.0001319146647362918,
+      "loss": 2.8413,
+      "step": 42196
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7129725813865662,
+      "learning_rate": 0.00013190191228900378,
+      "loss": 2.8465,
+      "step": 42197
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6918213963508606,
+      "learning_rate": 0.00013188916028445376,
+      "loss": 2.9743,
+      "step": 42198
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7041875720024109,
+      "learning_rate": 0.00013187640872267545,
+      "loss": 2.9971,
+      "step": 42199
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7519069314002991,
+      "learning_rate": 0.00013186365760370233,
+      "loss": 2.8209,
+      "step": 42200
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7009527087211609,
+      "learning_rate": 0.00013185090692756796,
+      "loss": 2.8803,
+      "step": 42201
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7051267623901367,
+      "learning_rate": 0.00013183815669430587,
+      "loss": 2.7405,
+      "step": 42202
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7135884761810303,
+      "learning_rate": 0.00013182540690394968,
+      "loss": 2.9261,
+      "step": 42203
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6404792070388794,
+      "learning_rate": 0.00013181265755653313,
+      "loss": 2.755,
+      "step": 42204
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7060980200767517,
+      "learning_rate": 0.00013179990865208957,
+      "loss": 2.8539,
+      "step": 42205
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6715150475502014,
+      "learning_rate": 0.00013178716019065275,
+      "loss": 2.8695,
+      "step": 42206
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7175034880638123,
+      "learning_rate": 0.0001317744121722562,
+      "loss": 2.8646,
+      "step": 42207
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.669849157333374,
+      "learning_rate": 0.00013176166459693338,
+      "loss": 2.9067,
+      "step": 42208
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6535002589225769,
+      "learning_rate": 0.00013174891746471805,
+      "loss": 2.9545,
+      "step": 42209
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6995859146118164,
+      "learning_rate": 0.00013173617077564357,
+      "loss": 2.938,
+      "step": 42210
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6866672039031982,
+      "learning_rate": 0.00013172342452974378,
+      "loss": 2.8747,
+      "step": 42211
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7818604111671448,
+      "learning_rate": 0.00013171067872705196,
+      "loss": 3.0297,
+      "step": 42212
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6974015831947327,
+      "learning_rate": 0.00013169793336760194,
+      "loss": 2.9315,
+      "step": 42213
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.683866560459137,
+      "learning_rate": 0.00013168518845142716,
+      "loss": 2.9169,
+      "step": 42214
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6737753748893738,
+      "learning_rate": 0.00013167244397856113,
+      "loss": 3.0056,
+      "step": 42215
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7229675054550171,
+      "learning_rate": 0.00013165969994903757,
+      "loss": 3.1931,
+      "step": 42216
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7247565388679504,
+      "learning_rate": 0.00013164695636288986,
+      "loss": 2.9204,
+      "step": 42217
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7432151436805725,
+      "learning_rate": 0.00013163421322015168,
+      "loss": 3.1582,
+      "step": 42218
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6654666662216187,
+      "learning_rate": 0.00013162147052085666,
+      "loss": 3.0593,
+      "step": 42219
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6909993886947632,
+      "learning_rate": 0.00013160872826503816,
+      "loss": 2.7321,
+      "step": 42220
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7081809043884277,
+      "learning_rate": 0.00013159598645273002,
+      "loss": 2.8623,
+      "step": 42221
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6888182163238525,
+      "learning_rate": 0.00013158324508396563,
+      "loss": 2.9768,
+      "step": 42222
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6894894242286682,
+      "learning_rate": 0.00013157050415877843,
+      "loss": 3.0968,
+      "step": 42223
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6939984560012817,
+      "learning_rate": 0.00013155776367720223,
+      "loss": 3.0215,
+      "step": 42224
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7307678461074829,
+      "learning_rate": 0.00013154502363927037,
+      "loss": 2.9241,
+      "step": 42225
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6858737468719482,
+      "learning_rate": 0.00013153228404501658,
+      "loss": 2.9914,
+      "step": 42226
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6769899725914001,
+      "learning_rate": 0.00013151954489447424,
+      "loss": 2.9458,
+      "step": 42227
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6825265884399414,
+      "learning_rate": 0.0001315068061876771,
+      "loss": 2.8933,
+      "step": 42228
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6815954446792603,
+      "learning_rate": 0.00013149406792465859,
+      "loss": 3.093,
+      "step": 42229
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.652065634727478,
+      "learning_rate": 0.00013148133010545214,
+      "loss": 3.0218,
+      "step": 42230
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6865938901901245,
+      "learning_rate": 0.00013146859273009158,
+      "loss": 2.7974,
+      "step": 42231
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6675371527671814,
+      "learning_rate": 0.00013145585579861022,
+      "loss": 2.9922,
+      "step": 42232
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6868575811386108,
+      "learning_rate": 0.0001314431193110418,
+      "loss": 2.9725,
+      "step": 42233
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.689803421497345,
+      "learning_rate": 0.00013143038326741963,
+      "loss": 2.7922,
+      "step": 42234
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7004591822624207,
+      "learning_rate": 0.00013141764766777753,
+      "loss": 2.8264,
+      "step": 42235
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7188933491706848,
+      "learning_rate": 0.00013140491251214888,
+      "loss": 2.9029,
+      "step": 42236
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6429142355918884,
+      "learning_rate": 0.00013139217780056713,
+      "loss": 2.8006,
+      "step": 42237
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6708921194076538,
+      "learning_rate": 0.00013137944353306603,
+      "loss": 2.9435,
+      "step": 42238
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6704015135765076,
+      "learning_rate": 0.00013136670970967897,
+      "loss": 2.9552,
+      "step": 42239
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7032222151756287,
+      "learning_rate": 0.0001313539763304395,
+      "loss": 2.898,
+      "step": 42240
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6727782487869263,
+      "learning_rate": 0.0001313412433953813,
+      "loss": 3.1218,
+      "step": 42241
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7007694244384766,
+      "learning_rate": 0.00013132851090453775,
+      "loss": 3.0368,
+      "step": 42242
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6816314458847046,
+      "learning_rate": 0.00013131577885794248,
+      "loss": 2.9854,
+      "step": 42243
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7471539974212646,
+      "learning_rate": 0.00013130304725562906,
+      "loss": 2.9592,
+      "step": 42244
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6539857983589172,
+      "learning_rate": 0.0001312903160976308,
+      "loss": 2.988,
+      "step": 42245
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6955440044403076,
+      "learning_rate": 0.00013127758538398145,
+      "loss": 3.0305,
+      "step": 42246
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6992368698120117,
+      "learning_rate": 0.0001312648551147144,
+      "loss": 2.9075,
+      "step": 42247
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7383258938789368,
+      "learning_rate": 0.00013125212528986336,
+      "loss": 2.7885,
+      "step": 42248
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7086347341537476,
+      "learning_rate": 0.00013123939590946164,
+      "loss": 3.0402,
+      "step": 42249
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.715691089630127,
+      "learning_rate": 0.00013122666697354296,
+      "loss": 2.9558,
+      "step": 42250
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6848446726799011,
+      "learning_rate": 0.00013121393848214077,
+      "loss": 2.9925,
+      "step": 42251
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.679757297039032,
+      "learning_rate": 0.00013120121043528845,
+      "loss": 2.9478,
+      "step": 42252
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6886526346206665,
+      "learning_rate": 0.00013118848283301981,
+      "loss": 3.0612,
+      "step": 42253
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7555577158927917,
+      "learning_rate": 0.00013117575567536807,
+      "loss": 3.0184,
+      "step": 42254
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6762755513191223,
+      "learning_rate": 0.00013116302896236693,
+      "loss": 2.84,
+      "step": 42255
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6835176944732666,
+      "learning_rate": 0.00013115030269404996,
+      "loss": 2.9008,
+      "step": 42256
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6507018804550171,
+      "learning_rate": 0.00013113757687045063,
+      "loss": 2.7376,
+      "step": 42257
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7082822322845459,
+      "learning_rate": 0.0001311248514916024,
+      "loss": 3.1054,
+      "step": 42258
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6697868704795837,
+      "learning_rate": 0.0001311121265575387,
+      "loss": 2.9395,
+      "step": 42259
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7068183422088623,
+      "learning_rate": 0.00013109940206829316,
+      "loss": 2.9129,
+      "step": 42260
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7108052968978882,
+      "learning_rate": 0.0001310866780238994,
+      "loss": 3.0556,
+      "step": 42261
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7249935865402222,
+      "learning_rate": 0.00013107395442439073,
+      "loss": 3.0242,
+      "step": 42262
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6782382130622864,
+      "learning_rate": 0.00013106123126980083,
+      "loss": 2.988,
+      "step": 42263
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6726512908935547,
+      "learning_rate": 0.00013104850856016304,
+      "loss": 2.9917,
+      "step": 42264
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6596064567565918,
+      "learning_rate": 0.00013103578629551104,
+      "loss": 3.1468,
+      "step": 42265
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6861616969108582,
+      "learning_rate": 0.0001310230644758783,
+      "loss": 2.9792,
+      "step": 42266
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7167417407035828,
+      "learning_rate": 0.0001310103431012982,
+      "loss": 2.8683,
+      "step": 42267
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.671953558921814,
+      "learning_rate": 0.0001309976221718044,
+      "loss": 3.0194,
+      "step": 42268
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6684266924858093,
+      "learning_rate": 0.00013098490168743024,
+      "loss": 2.8826,
+      "step": 42269
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6714996099472046,
+      "learning_rate": 0.00013097218164820934,
+      "loss": 2.8523,
+      "step": 42270
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6615957617759705,
+      "learning_rate": 0.00013095946205417527,
+      "loss": 2.9081,
+      "step": 42271
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7182602882385254,
+      "learning_rate": 0.00013094674290536144,
+      "loss": 2.9164,
+      "step": 42272
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6356480717658997,
+      "learning_rate": 0.00013093402420180137,
+      "loss": 2.974,
+      "step": 42273
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6807072162628174,
+      "learning_rate": 0.00013092130594352845,
+      "loss": 2.9493,
+      "step": 42274
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7285915613174438,
+      "learning_rate": 0.00013090858813057622,
+      "loss": 3.148,
+      "step": 42275
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6464709639549255,
+      "learning_rate": 0.0001308958707629784,
+      "loss": 3.0066,
+      "step": 42276
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7085345387458801,
+      "learning_rate": 0.00013088315384076813,
+      "loss": 2.9503,
+      "step": 42277
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7010959386825562,
+      "learning_rate": 0.00013087043736397923,
+      "loss": 2.9879,
+      "step": 42278
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6488636136054993,
+      "learning_rate": 0.00013085772133264505,
+      "loss": 3.0544,
+      "step": 42279
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6806224584579468,
+      "learning_rate": 0.00013084500574679896,
+      "loss": 3.0751,
+      "step": 42280
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7017746567726135,
+      "learning_rate": 0.00013083229060647473,
+      "loss": 2.8714,
+      "step": 42281
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7165690660476685,
+      "learning_rate": 0.00013081957591170551,
+      "loss": 2.8805,
+      "step": 42282
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6798978447914124,
+      "learning_rate": 0.00013080686166252515,
+      "loss": 2.9131,
+      "step": 42283
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6638495922088623,
+      "learning_rate": 0.00013079414785896676,
+      "loss": 2.9656,
+      "step": 42284
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7176745533943176,
+      "learning_rate": 0.00013078143450106407,
+      "loss": 2.8036,
+      "step": 42285
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6890358328819275,
+      "learning_rate": 0.00013076872158885062,
+      "loss": 2.8643,
+      "step": 42286
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6619595289230347,
+      "learning_rate": 0.0001307560091223598,
+      "loss": 2.8526,
+      "step": 42287
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.682287871837616,
+      "learning_rate": 0.00013074329710162507,
+      "loss": 2.9174,
+      "step": 42288
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6877726912498474,
+      "learning_rate": 0.00013073058552667978,
+      "loss": 3.0925,
+      "step": 42289
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6723368167877197,
+      "learning_rate": 0.0001307178743975576,
+      "loss": 2.8354,
+      "step": 42290
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6675642132759094,
+      "learning_rate": 0.00013070516371429204,
+      "loss": 2.9596,
+      "step": 42291
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6696404218673706,
+      "learning_rate": 0.0001306924534769164,
+      "loss": 2.8843,
+      "step": 42292
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.3090136051177979,
+      "learning_rate": 0.00013067974368546437,
+      "loss": 3.0016,
+      "step": 42293
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6999812722206116,
+      "learning_rate": 0.0001306670343399693,
+      "loss": 2.9312,
+      "step": 42294
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6614927053451538,
+      "learning_rate": 0.00013065432544046455,
+      "loss": 2.9818,
+      "step": 42295
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6659789085388184,
+      "learning_rate": 0.00013064161698698383,
+      "loss": 2.8157,
+      "step": 42296
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6961437463760376,
+      "learning_rate": 0.0001306289089795604,
+      "loss": 2.8556,
+      "step": 42297
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6694318056106567,
+      "learning_rate": 0.00013061620141822792,
+      "loss": 2.9884,
+      "step": 42298
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6987858414649963,
+      "learning_rate": 0.0001306034943030197,
+      "loss": 2.8139,
+      "step": 42299
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6902264952659607,
+      "learning_rate": 0.00013059078763396923,
+      "loss": 2.9773,
+      "step": 42300
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6752839088439941,
+      "learning_rate": 0.00013057808141111026,
+      "loss": 2.8498,
+      "step": 42301
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7169557809829712,
+      "learning_rate": 0.00013056537563447578,
+      "loss": 2.7842,
+      "step": 42302
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7090547680854797,
+      "learning_rate": 0.00013055267030409964,
+      "loss": 3.2104,
+      "step": 42303
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6589310169219971,
+      "learning_rate": 0.00013053996542001503,
+      "loss": 2.7599,
+      "step": 42304
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7081121802330017,
+      "learning_rate": 0.00013052726098225563,
+      "loss": 3.0029,
+      "step": 42305
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6937905550003052,
+      "learning_rate": 0.0001305145569908547,
+      "loss": 3.0007,
+      "step": 42306
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6769243478775024,
+      "learning_rate": 0.0001305018534458458,
+      "loss": 2.8763,
+      "step": 42307
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7453411221504211,
+      "learning_rate": 0.00013048915034726254,
+      "loss": 2.8378,
+      "step": 42308
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.674867570400238,
+      "learning_rate": 0.0001304764476951382,
+      "loss": 3.0571,
+      "step": 42309
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.707789421081543,
+      "learning_rate": 0.00013046374548950632,
+      "loss": 3.1763,
+      "step": 42310
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.00013045104373040015,
+      "loss": 2.9595,
+      "step": 42311
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7053055167198181,
+      "learning_rate": 0.0001304383424178533,
+      "loss": 2.9741,
+      "step": 42312
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6682379841804504,
+      "learning_rate": 0.00013042564155189937,
+      "loss": 2.9207,
+      "step": 42313
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6992611289024353,
+      "learning_rate": 0.00013041294113257153,
+      "loss": 2.9781,
+      "step": 42314
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6636302471160889,
+      "learning_rate": 0.0001304002411599035,
+      "loss": 2.9719,
+      "step": 42315
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6672903299331665,
+      "learning_rate": 0.00013038754163392858,
+      "loss": 3.0387,
+      "step": 42316
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7034769654273987,
+      "learning_rate": 0.0001303748425546801,
+      "loss": 2.8552,
+      "step": 42317
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.665863037109375,
+      "learning_rate": 0.00013036214392219176,
+      "loss": 2.9452,
+      "step": 42318
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6769158244132996,
+      "learning_rate": 0.0001303494457364968,
+      "loss": 2.951,
+      "step": 42319
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6977242827415466,
+      "learning_rate": 0.00013033674799762884,
+      "loss": 2.9524,
+      "step": 42320
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6679743528366089,
+      "learning_rate": 0.00013032405070562112,
+      "loss": 2.8681,
+      "step": 42321
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7165648341178894,
+      "learning_rate": 0.0001303113538605072,
+      "loss": 3.006,
+      "step": 42322
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7313899993896484,
+      "learning_rate": 0.00013029865746232063,
+      "loss": 2.9289,
+      "step": 42323
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6785582304000854,
+      "learning_rate": 0.00013028596151109473,
+      "loss": 2.9195,
+      "step": 42324
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6553201079368591,
+      "learning_rate": 0.00013027326600686294,
+      "loss": 2.9507,
+      "step": 42325
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7384634017944336,
+      "learning_rate": 0.0001302605709496586,
+      "loss": 2.8676,
+      "step": 42326
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6803930401802063,
+      "learning_rate": 0.00013024787633951528,
+      "loss": 2.7254,
+      "step": 42327
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7045084834098816,
+      "learning_rate": 0.0001302351821764665,
+      "loss": 2.9962,
+      "step": 42328
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7233037352561951,
+      "learning_rate": 0.00013022248846054538,
+      "loss": 2.6192,
+      "step": 42329
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7065985798835754,
+      "learning_rate": 0.00013020979519178574,
+      "loss": 3.0464,
+      "step": 42330
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6664899587631226,
+      "learning_rate": 0.00013019710237022078,
+      "loss": 2.9836,
+      "step": 42331
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6708837151527405,
+      "learning_rate": 0.0001301844099958839,
+      "loss": 3.0906,
+      "step": 42332
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6945251822471619,
+      "learning_rate": 0.0001301717180688087,
+      "loss": 2.8813,
+      "step": 42333
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6718724370002747,
+      "learning_rate": 0.00013015902658902837,
+      "loss": 3.0567,
+      "step": 42334
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6948716044425964,
+      "learning_rate": 0.00013014633555657665,
+      "loss": 3.2154,
+      "step": 42335
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6696521639823914,
+      "learning_rate": 0.00013013364497148661,
+      "loss": 2.8059,
+      "step": 42336
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.723652184009552,
+      "learning_rate": 0.000130120954833792,
+      "loss": 2.7653,
+      "step": 42337
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7283807992935181,
+      "learning_rate": 0.00013010826514352615,
+      "loss": 2.9663,
+      "step": 42338
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.7475623488426208,
+      "learning_rate": 0.00013009557590072226,
+      "loss": 3.0521,
+      "step": 42339
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6831005811691284,
+      "learning_rate": 0.00013008288710541406,
+      "loss": 3.1819,
+      "step": 42340
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6755504012107849,
+      "learning_rate": 0.0001300701987576347,
+      "loss": 2.8866,
+      "step": 42341
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6740015149116516,
+      "learning_rate": 0.00013005751085741777,
+      "loss": 2.8911,
+      "step": 42342
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6407496929168701,
+      "learning_rate": 0.00013004482340479672,
+      "loss": 2.9256,
+      "step": 42343
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7576172947883606,
+      "learning_rate": 0.0001300321363998048,
+      "loss": 3.1794,
+      "step": 42344
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6567968726158142,
+      "learning_rate": 0.00013001944984247565,
+      "loss": 2.9828,
+      "step": 42345
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7320488691329956,
+      "learning_rate": 0.00013000676373284254,
+      "loss": 2.9482,
+      "step": 42346
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.646527886390686,
+      "learning_rate": 0.0001299940780709388,
+      "loss": 2.8785,
+      "step": 42347
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6653791069984436,
+      "learning_rate": 0.00012998139285679802,
+      "loss": 3.0511,
+      "step": 42348
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6764574646949768,
+      "learning_rate": 0.00012996870809045346,
+      "loss": 2.9864,
+      "step": 42349
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6396170854568481,
+      "learning_rate": 0.00012995602377193868,
+      "loss": 2.8156,
+      "step": 42350
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6769349575042725,
+      "learning_rate": 0.00012994333990128693,
+      "loss": 2.9659,
+      "step": 42351
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.711395263671875,
+      "learning_rate": 0.00012993065647853177,
+      "loss": 3.0182,
+      "step": 42352
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6921821236610413,
+      "learning_rate": 0.00012991797350370656,
+      "loss": 2.8534,
+      "step": 42353
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.686550498008728,
+      "learning_rate": 0.00012990529097684455,
+      "loss": 2.9083,
+      "step": 42354
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7336958646774292,
+      "learning_rate": 0.00012989260889797934,
+      "loss": 2.7641,
+      "step": 42355
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7120491862297058,
+      "learning_rate": 0.0001298799272671442,
+      "loss": 2.9055,
+      "step": 42356
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6996833086013794,
+      "learning_rate": 0.0001298672460843726,
+      "loss": 2.9461,
+      "step": 42357
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7006560564041138,
+      "learning_rate": 0.000129854565349698,
+      "loss": 2.8888,
+      "step": 42358
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6544038653373718,
+      "learning_rate": 0.0001298418850631538,
+      "loss": 2.8266,
+      "step": 42359
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7305658459663391,
+      "learning_rate": 0.00012982920522477324,
+      "loss": 3.1084,
+      "step": 42360
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6678386330604553,
+      "learning_rate": 0.00012981652583458973,
+      "loss": 2.7814,
+      "step": 42361
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6706693768501282,
+      "learning_rate": 0.00012980384689263675,
+      "loss": 2.9772,
+      "step": 42362
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7006844878196716,
+      "learning_rate": 0.00012979116839894779,
+      "loss": 2.9108,
+      "step": 42363
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7549619078636169,
+      "learning_rate": 0.000129778490353556,
+      "loss": 3.0531,
+      "step": 42364
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6899545192718506,
+      "learning_rate": 0.00012976581275649503,
+      "loss": 2.8675,
+      "step": 42365
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6634476184844971,
+      "learning_rate": 0.00012975313560779805,
+      "loss": 3.0111,
+      "step": 42366
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6710315942764282,
+      "learning_rate": 0.00012974045890749866,
+      "loss": 2.9529,
+      "step": 42367
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6514169573783875,
+      "learning_rate": 0.0001297277826556301,
+      "loss": 2.697,
+      "step": 42368
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6755321025848389,
+      "learning_rate": 0.00012971510685222572,
+      "loss": 3.0741,
+      "step": 42369
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7050230503082275,
+      "learning_rate": 0.00012970243149731904,
+      "loss": 2.6275,
+      "step": 42370
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7091357707977295,
+      "learning_rate": 0.00012968975659094332,
+      "loss": 2.6535,
+      "step": 42371
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6768811941146851,
+      "learning_rate": 0.00012967708213313197,
+      "loss": 3.0096,
+      "step": 42372
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6693670153617859,
+      "learning_rate": 0.00012966440812391854,
+      "loss": 2.8613,
+      "step": 42373
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7181096076965332,
+      "learning_rate": 0.00012965173456333626,
+      "loss": 2.7873,
+      "step": 42374
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7134459614753723,
+      "learning_rate": 0.00012963906145141855,
+      "loss": 2.8422,
+      "step": 42375
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7182124853134155,
+      "learning_rate": 0.00012962638878819864,
+      "loss": 2.9213,
+      "step": 42376
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7274218201637268,
+      "learning_rate": 0.00012961371657371003,
+      "loss": 2.7016,
+      "step": 42377
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6675017476081848,
+      "learning_rate": 0.0001296010448079862,
+      "loss": 2.8717,
+      "step": 42378
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7228912115097046,
+      "learning_rate": 0.00012958837349106032,
+      "loss": 3.0004,
+      "step": 42379
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6356688737869263,
+      "learning_rate": 0.000129575702622966,
+      "loss": 2.5134,
+      "step": 42380
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.663016140460968,
+      "learning_rate": 0.00012956303220373647,
+      "loss": 2.9052,
+      "step": 42381
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6874765157699585,
+      "learning_rate": 0.00012955036223340498,
+      "loss": 3.127,
+      "step": 42382
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7107385396957397,
+      "learning_rate": 0.00012953769271200518,
+      "loss": 2.9775,
+      "step": 42383
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7340335249900818,
+      "learning_rate": 0.00012952502363957014,
+      "loss": 2.8887,
+      "step": 42384
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6798539161682129,
+      "learning_rate": 0.00012951235501613352,
+      "loss": 2.9121,
+      "step": 42385
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7123425006866455,
+      "learning_rate": 0.00012949968684172845,
+      "loss": 3.0914,
+      "step": 42386
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6899136900901794,
+      "learning_rate": 0.00012948701911638845,
+      "loss": 2.9722,
+      "step": 42387
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7115585207939148,
+      "learning_rate": 0.00012947435184014673,
+      "loss": 3.0044,
+      "step": 42388
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6917968392372131,
+      "learning_rate": 0.0001294616850130369,
+      "loss": 3.1324,
+      "step": 42389
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6570953726768494,
+      "learning_rate": 0.00012944901863509213,
+      "loss": 2.9288,
+      "step": 42390
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6744385957717896,
+      "learning_rate": 0.00012943635270634574,
+      "loss": 2.7654,
+      "step": 42391
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7136799693107605,
+      "learning_rate": 0.00012942368722683125,
+      "loss": 2.899,
+      "step": 42392
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6737086772918701,
+      "learning_rate": 0.00012941102219658182,
+      "loss": 2.7251,
+      "step": 42393
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6692909598350525,
+      "learning_rate": 0.00012939835761563095,
+      "loss": 3.0238,
+      "step": 42394
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6992953419685364,
+      "learning_rate": 0.0001293856934840121,
+      "loss": 2.8346,
+      "step": 42395
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6643773913383484,
+      "learning_rate": 0.00012937302980175848,
+      "loss": 2.7622,
+      "step": 42396
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6850830912590027,
+      "learning_rate": 0.00012936036656890345,
+      "loss": 2.9824,
+      "step": 42397
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7364450097084045,
+      "learning_rate": 0.0001293477037854803,
+      "loss": 2.953,
+      "step": 42398
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7071236968040466,
+      "learning_rate": 0.00012933504145152245,
+      "loss": 3.1013,
+      "step": 42399
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7141284942626953,
+      "learning_rate": 0.00012932237956706332,
+      "loss": 2.7942,
+      "step": 42400
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7371499538421631,
+      "learning_rate": 0.00012930971813213612,
+      "loss": 2.533,
+      "step": 42401
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7052841782569885,
+      "learning_rate": 0.0001292970571467744,
+      "loss": 2.8772,
+      "step": 42402
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7750392556190491,
+      "learning_rate": 0.00012928439661101133,
+      "loss": 2.9739,
+      "step": 42403
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7135288715362549,
+      "learning_rate": 0.00012927173652488023,
+      "loss": 2.8134,
+      "step": 42404
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6652043461799622,
+      "learning_rate": 0.00012925907688841464,
+      "loss": 3.063,
+      "step": 42405
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6913604736328125,
+      "learning_rate": 0.00012924641770164762,
+      "loss": 2.7423,
+      "step": 42406
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7233631014823914,
+      "learning_rate": 0.0001292337589646128,
+      "loss": 2.9022,
+      "step": 42407
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6916614174842834,
+      "learning_rate": 0.0001292211006773433,
+      "loss": 2.9954,
+      "step": 42408
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7121450901031494,
+      "learning_rate": 0.00012920844283987256,
+      "loss": 3.0694,
+      "step": 42409
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7093349695205688,
+      "learning_rate": 0.00012919578545223404,
+      "loss": 3.1103,
+      "step": 42410
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6733999848365784,
+      "learning_rate": 0.0001291831285144609,
+      "loss": 2.8649,
+      "step": 42411
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6598990559577942,
+      "learning_rate": 0.00012917047202658656,
+      "loss": 3.0327,
+      "step": 42412
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6923961043357849,
+      "learning_rate": 0.00012915781598864422,
+      "loss": 2.8068,
+      "step": 42413
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6872490048408508,
+      "learning_rate": 0.00012914516040066727,
+      "loss": 2.7399,
+      "step": 42414
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6673179268836975,
+      "learning_rate": 0.00012913250526268923,
+      "loss": 2.9066,
+      "step": 42415
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6778900027275085,
+      "learning_rate": 0.00012911985057474315,
+      "loss": 3.0026,
+      "step": 42416
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7145534157752991,
+      "learning_rate": 0.00012910719633686264,
+      "loss": 2.9472,
+      "step": 42417
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6881458163261414,
+      "learning_rate": 0.00012909454254908084,
+      "loss": 2.8058,
+      "step": 42418
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6922593712806702,
+      "learning_rate": 0.00012908188921143104,
+      "loss": 2.9075,
+      "step": 42419
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6974817514419556,
+      "learning_rate": 0.00012906923632394676,
+      "loss": 3.0196,
+      "step": 42420
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6990817189216614,
+      "learning_rate": 0.0001290565838866611,
+      "loss": 2.8226,
+      "step": 42421
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7017726302146912,
+      "learning_rate": 0.00012904393189960763,
+      "loss": 2.9575,
+      "step": 42422
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7260956168174744,
+      "learning_rate": 0.00012903128036281943,
+      "loss": 2.9157,
+      "step": 42423
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.675983726978302,
+      "learning_rate": 0.00012901862927632994,
+      "loss": 2.8013,
+      "step": 42424
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7136183977127075,
+      "learning_rate": 0.00012900597864017258,
+      "loss": 2.784,
+      "step": 42425
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7095462679862976,
+      "learning_rate": 0.00012899332845438052,
+      "loss": 2.9734,
+      "step": 42426
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6412888765335083,
+      "learning_rate": 0.00012898067871898716,
+      "loss": 3.1996,
+      "step": 42427
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7176571488380432,
+      "learning_rate": 0.00012896802943402564,
+      "loss": 2.721,
+      "step": 42428
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.685215950012207,
+      "learning_rate": 0.00012895538059952947,
+      "loss": 3.2016,
+      "step": 42429
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7202776670455933,
+      "learning_rate": 0.00012894273221553197,
+      "loss": 2.8996,
+      "step": 42430
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6567890644073486,
+      "learning_rate": 0.0001289300842820663,
+      "loss": 3.0631,
+      "step": 42431
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7814381718635559,
+      "learning_rate": 0.00012891743679916597,
+      "loss": 2.7118,
+      "step": 42432
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6643449068069458,
+      "learning_rate": 0.00012890478976686418,
+      "loss": 2.8286,
+      "step": 42433
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6698668003082275,
+      "learning_rate": 0.00012889214318519417,
+      "loss": 2.9327,
+      "step": 42434
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6875276565551758,
+      "learning_rate": 0.00012887949705418938,
+      "loss": 2.9944,
+      "step": 42435
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7161490321159363,
+      "learning_rate": 0.00012886685137388298,
+      "loss": 2.9799,
+      "step": 42436
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6984416246414185,
+      "learning_rate": 0.00012885420614430846,
+      "loss": 2.8483,
+      "step": 42437
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6920896768569946,
+      "learning_rate": 0.00012884156136549893,
+      "loss": 3.0116,
+      "step": 42438
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.704794704914093,
+      "learning_rate": 0.00012882891703748793,
+      "loss": 2.6875,
+      "step": 42439
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7424866557121277,
+      "learning_rate": 0.00012881627316030858,
+      "loss": 2.78,
+      "step": 42440
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6924298405647278,
+      "learning_rate": 0.0001288036297339941,
+      "loss": 2.8718,
+      "step": 42441
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6861541271209717,
+      "learning_rate": 0.000128790986758578,
+      "loss": 2.786,
+      "step": 42442
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.720872700214386,
+      "learning_rate": 0.00012877834423409345,
+      "loss": 2.7509,
+      "step": 42443
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6972219347953796,
+      "learning_rate": 0.00012876570216057375,
+      "loss": 3.0085,
+      "step": 42444
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6819028854370117,
+      "learning_rate": 0.00012875306053805236,
+      "loss": 2.9576,
+      "step": 42445
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6733766794204712,
+      "learning_rate": 0.0001287404193665623,
+      "loss": 2.9113,
+      "step": 42446
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6779769062995911,
+      "learning_rate": 0.00012872777864613718,
+      "loss": 2.7558,
+      "step": 42447
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7019845843315125,
+      "learning_rate": 0.00012871513837681007,
+      "loss": 2.8872,
+      "step": 42448
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7672656178474426,
+      "learning_rate": 0.00012870249855861428,
+      "loss": 3.0762,
+      "step": 42449
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6681340932846069,
+      "learning_rate": 0.00012868985919158322,
+      "loss": 3.0499,
+      "step": 42450
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7144621014595032,
+      "learning_rate": 0.00012867722027574998,
+      "loss": 3.2119,
+      "step": 42451
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.671342134475708,
+      "learning_rate": 0.0001286645818111481,
+      "loss": 3.0014,
+      "step": 42452
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6688801646232605,
+      "learning_rate": 0.00012865194379781063,
+      "loss": 2.8508,
+      "step": 42453
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6864688396453857,
+      "learning_rate": 0.00012863930623577103,
+      "loss": 3.1261,
+      "step": 42454
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7038698792457581,
+      "learning_rate": 0.00012862666912506255,
+      "loss": 2.8566,
+      "step": 42455
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6805946230888367,
+      "learning_rate": 0.00012861403246571836,
+      "loss": 2.8152,
+      "step": 42456
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6786236763000488,
+      "learning_rate": 0.00012860139625777187,
+      "loss": 2.89,
+      "step": 42457
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.694636881351471,
+      "learning_rate": 0.00012858876050125627,
+      "loss": 2.8622,
+      "step": 42458
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7160537838935852,
+      "learning_rate": 0.00012857612519620487,
+      "loss": 2.8614,
+      "step": 42459
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.675896406173706,
+      "learning_rate": 0.00012856349034265105,
+      "loss": 2.8066,
+      "step": 42460
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6733073592185974,
+      "learning_rate": 0.000128550855940628,
+      "loss": 3.0653,
+      "step": 42461
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7059203386306763,
+      "learning_rate": 0.000128538221990169,
+      "loss": 3.1194,
+      "step": 42462
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6924806237220764,
+      "learning_rate": 0.00012852558849130725,
+      "loss": 2.6967,
+      "step": 42463
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6774609684944153,
+      "learning_rate": 0.00012851295544407618,
+      "loss": 3.0024,
+      "step": 42464
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6729218363761902,
+      "learning_rate": 0.0001285003228485089,
+      "loss": 2.94,
+      "step": 42465
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.673554539680481,
+      "learning_rate": 0.0001284876907046387,
+      "loss": 2.9425,
+      "step": 42466
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7064787149429321,
+      "learning_rate": 0.00012847505901249912,
+      "loss": 2.8488,
+      "step": 42467
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6640927791595459,
+      "learning_rate": 0.00012846242777212304,
+      "loss": 2.7895,
+      "step": 42468
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6812401413917542,
+      "learning_rate": 0.00012844979698354406,
+      "loss": 2.8939,
+      "step": 42469
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6853512525558472,
+      "learning_rate": 0.00012843716664679528,
+      "loss": 3.0094,
+      "step": 42470
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.673121988773346,
+      "learning_rate": 0.00012842453676190988,
+      "loss": 2.8918,
+      "step": 42471
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7033911347389221,
+      "learning_rate": 0.00012841190732892135,
+      "loss": 2.8784,
+      "step": 42472
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6943319439888,
+      "learning_rate": 0.00012839927834786268,
+      "loss": 2.8472,
+      "step": 42473
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6493386626243591,
+      "learning_rate": 0.00012838664981876743,
+      "loss": 2.7605,
+      "step": 42474
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6886091232299805,
+      "learning_rate": 0.00012837402174166863,
+      "loss": 2.9289,
+      "step": 42475
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6950169205665588,
+      "learning_rate": 0.0001283613941165997,
+      "loss": 3.0136,
+      "step": 42476
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7558622360229492,
+      "learning_rate": 0.00012834876694359386,
+      "loss": 2.9728,
+      "step": 42477
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7118024230003357,
+      "learning_rate": 0.00012833614022268418,
+      "loss": 2.9034,
+      "step": 42478
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6858274340629578,
+      "learning_rate": 0.0001283235139539042,
+      "loss": 2.7353,
+      "step": 42479
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6761425733566284,
+      "learning_rate": 0.00012831088813728694,
+      "loss": 2.8738,
+      "step": 42480
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.686199426651001,
+      "learning_rate": 0.00012829826277286577,
+      "loss": 2.7441,
+      "step": 42481
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7111159563064575,
+      "learning_rate": 0.00012828563786067398,
+      "loss": 2.8098,
+      "step": 42482
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7450211644172668,
+      "learning_rate": 0.00012827301340074485,
+      "loss": 3.027,
+      "step": 42483
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6911284923553467,
+      "learning_rate": 0.0001282603893931115,
+      "loss": 3.0419,
+      "step": 42484
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6608761548995972,
+      "learning_rate": 0.00012824776583780716,
+      "loss": 2.6919,
+      "step": 42485
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.687088131904602,
+      "learning_rate": 0.00012823514273486511,
+      "loss": 2.9157,
+      "step": 42486
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6786913871765137,
+      "learning_rate": 0.00012822252008431878,
+      "loss": 2.8025,
+      "step": 42487
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6923184394836426,
+      "learning_rate": 0.00012820989788620119,
+      "loss": 2.9774,
+      "step": 42488
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7364439964294434,
+      "learning_rate": 0.0001281972761405457,
+      "loss": 3.0177,
+      "step": 42489
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6883888840675354,
+      "learning_rate": 0.00012818465484738548,
+      "loss": 3.1355,
+      "step": 42490
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6888550519943237,
+      "learning_rate": 0.0001281720340067539,
+      "loss": 2.9592,
+      "step": 42491
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6707346439361572,
+      "learning_rate": 0.0001281594136186841,
+      "loss": 2.9254,
+      "step": 42492
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.695807158946991,
+      "learning_rate": 0.00012814679368320923,
+      "loss": 2.833,
+      "step": 42493
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7045168876647949,
+      "learning_rate": 0.00012813417420036273,
+      "loss": 2.9055,
+      "step": 42494
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6547208428382874,
+      "learning_rate": 0.00012812155517017763,
+      "loss": 2.7095,
+      "step": 42495
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6461189985275269,
+      "learning_rate": 0.0001281089365926873,
+      "loss": 2.9019,
+      "step": 42496
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6801860332489014,
+      "learning_rate": 0.00012809631846792503,
+      "loss": 2.8929,
+      "step": 42497
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.704507052898407,
+      "learning_rate": 0.00012808370079592398,
+      "loss": 2.8624,
+      "step": 42498
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6679760217666626,
+      "learning_rate": 0.00012807108357671738,
+      "loss": 2.8551,
+      "step": 42499
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6944798231124878,
+      "learning_rate": 0.0001280584668103384,
+      "loss": 3.0003,
+      "step": 42500
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7155947089195251,
+      "learning_rate": 0.00012804585049682024,
+      "loss": 2.9066,
+      "step": 42501
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7650623321533203,
+      "learning_rate": 0.00012803323463619638,
+      "loss": 2.8278,
+      "step": 42502
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6742417812347412,
+      "learning_rate": 0.00012802061922849976,
+      "loss": 2.7601,
+      "step": 42503
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6518091559410095,
+      "learning_rate": 0.00012800800427376384,
+      "loss": 3.0368,
+      "step": 42504
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7054563164710999,
+      "learning_rate": 0.00012799538977202173,
+      "loss": 2.8426,
+      "step": 42505
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7084836959838867,
+      "learning_rate": 0.00012798277572330656,
+      "loss": 3.0179,
+      "step": 42506
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6683297753334045,
+      "learning_rate": 0.00012797016212765175,
+      "loss": 2.9062,
+      "step": 42507
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6678385734558105,
+      "learning_rate": 0.00012795754898509035,
+      "loss": 3.106,
+      "step": 42508
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7056717276573181,
+      "learning_rate": 0.00012794493629565576,
+      "loss": 2.9415,
+      "step": 42509
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7119282484054565,
+      "learning_rate": 0.00012793232405938095,
+      "loss": 2.7636,
+      "step": 42510
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.743801474571228,
+      "learning_rate": 0.0001279197122762993,
+      "loss": 2.6962,
+      "step": 42511
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7152933478355408,
+      "learning_rate": 0.0001279071009464441,
+      "loss": 2.87,
+      "step": 42512
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6678538918495178,
+      "learning_rate": 0.00012789449006984853,
+      "loss": 3.0891,
+      "step": 42513
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7106472849845886,
+      "learning_rate": 0.0001278818796465457,
+      "loss": 2.802,
+      "step": 42514
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7291449904441833,
+      "learning_rate": 0.0001278692696765688,
+      "loss": 2.8319,
+      "step": 42515
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6768184304237366,
+      "learning_rate": 0.00012785666015995107,
+      "loss": 2.8618,
+      "step": 42516
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6788813471794128,
+      "learning_rate": 0.00012784405109672595,
+      "loss": 2.8516,
+      "step": 42517
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6908031702041626,
+      "learning_rate": 0.0001278314424869263,
+      "loss": 3.0243,
+      "step": 42518
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7076452374458313,
+      "learning_rate": 0.0001278188343305856,
+      "loss": 2.8712,
+      "step": 42519
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6761187314987183,
+      "learning_rate": 0.00012780622662773698,
+      "loss": 2.963,
+      "step": 42520
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7147707939147949,
+      "learning_rate": 0.0001277936193784135,
+      "loss": 3.1233,
+      "step": 42521
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.0204122066497803,
+      "learning_rate": 0.0001277810125826486,
+      "loss": 2.9935,
+      "step": 42522
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.690827488899231,
+      "learning_rate": 0.00012776840624047526,
+      "loss": 2.8386,
+      "step": 42523
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6465492248535156,
+      "learning_rate": 0.00012775580035192692,
+      "loss": 2.8819,
+      "step": 42524
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6738273501396179,
+      "learning_rate": 0.00012774319491703652,
+      "loss": 2.924,
+      "step": 42525
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6924593448638916,
+      "learning_rate": 0.00012773058993583735,
+      "loss": 3.0896,
+      "step": 42526
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6848966479301453,
+      "learning_rate": 0.00012771798540836283,
+      "loss": 2.8388,
+      "step": 42527
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7153657674789429,
+      "learning_rate": 0.00012770538133464595,
+      "loss": 3.1066,
+      "step": 42528
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6560115814208984,
+      "learning_rate": 0.00012769277771471996,
+      "loss": 2.9721,
+      "step": 42529
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6960185766220093,
+      "learning_rate": 0.0001276801745486179,
+      "loss": 2.9123,
+      "step": 42530
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.64568030834198,
+      "learning_rate": 0.00012766757183637308,
+      "loss": 3.2167,
+      "step": 42531
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.697411060333252,
+      "learning_rate": 0.00012765496957801886,
+      "loss": 2.7463,
+      "step": 42532
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6693733334541321,
+      "learning_rate": 0.00012764236777358816,
+      "loss": 2.99,
+      "step": 42533
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6906144618988037,
+      "learning_rate": 0.00012762976642311438,
+      "loss": 2.9348,
+      "step": 42534
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7639672160148621,
+      "learning_rate": 0.00012761716552663063,
+      "loss": 2.8812,
+      "step": 42535
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6580451130867004,
+      "learning_rate": 0.00012760456508417002,
+      "loss": 2.9266,
+      "step": 42536
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7195175886154175,
+      "learning_rate": 0.00012759196509576587,
+      "loss": 2.6738,
+      "step": 42537
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7359403371810913,
+      "learning_rate": 0.0001275793655614512,
+      "loss": 2.9124,
+      "step": 42538
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7050012946128845,
+      "learning_rate": 0.00012756676648125942,
+      "loss": 2.9858,
+      "step": 42539
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6860030293464661,
+      "learning_rate": 0.00012755416785522347,
+      "loss": 2.7879,
+      "step": 42540
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.667359471321106,
+      "learning_rate": 0.00012754156968337678,
+      "loss": 3.0746,
+      "step": 42541
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6795210242271423,
+      "learning_rate": 0.0001275289719657524,
+      "loss": 2.9643,
+      "step": 42542
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6692383885383606,
+      "learning_rate": 0.0001275163747023834,
+      "loss": 2.9197,
+      "step": 42543
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6974451541900635,
+      "learning_rate": 0.0001275037778933032,
+      "loss": 2.857,
+      "step": 42544
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7106644511222839,
+      "learning_rate": 0.0001274911815385447,
+      "loss": 2.8773,
+      "step": 42545
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7244455814361572,
+      "learning_rate": 0.00012747858563814138,
+      "loss": 2.9333,
+      "step": 42546
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6804713606834412,
+      "learning_rate": 0.00012746599019212613,
+      "loss": 2.7637,
+      "step": 42547
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7047104835510254,
+      "learning_rate": 0.0001274533952005323,
+      "loss": 2.9819,
+      "step": 42548
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6896386742591858,
+      "learning_rate": 0.00012744080066339305,
+      "loss": 2.9393,
+      "step": 42549
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7217565178871155,
+      "learning_rate": 0.00012742820658074157,
+      "loss": 3.0425,
+      "step": 42550
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6903233528137207,
+      "learning_rate": 0.000127415612952611,
+      "loss": 3.207,
+      "step": 42551
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6575881838798523,
+      "learning_rate": 0.00012740301977903434,
+      "loss": 2.8382,
+      "step": 42552
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6848930716514587,
+      "learning_rate": 0.00012739042706004495,
+      "loss": 2.8969,
+      "step": 42553
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6955052614212036,
+      "learning_rate": 0.00012737783479567605,
+      "loss": 3.1267,
+      "step": 42554
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.669148862361908,
+      "learning_rate": 0.00012736524298596061,
+      "loss": 3.1156,
+      "step": 42555
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6611831188201904,
+      "learning_rate": 0.00012735265163093198,
+      "loss": 2.8058,
+      "step": 42556
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6889923810958862,
+      "learning_rate": 0.00012734006073062326,
+      "loss": 2.7144,
+      "step": 42557
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6790809631347656,
+      "learning_rate": 0.00012732747028506745,
+      "loss": 3.0519,
+      "step": 42558
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6864499449729919,
+      "learning_rate": 0.00012731488029429797,
+      "loss": 2.8055,
+      "step": 42559
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6939069628715515,
+      "learning_rate": 0.00012730229075834777,
+      "loss": 2.8274,
+      "step": 42560
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6930176019668579,
+      "learning_rate": 0.00012728970167725023,
+      "loss": 2.8174,
+      "step": 42561
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7337930798530579,
+      "learning_rate": 0.00012727711305103822,
+      "loss": 2.9908,
+      "step": 42562
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6973217725753784,
+      "learning_rate": 0.0001272645248797452,
+      "loss": 3.1223,
+      "step": 42563
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6723378896713257,
+      "learning_rate": 0.00012725193716340414,
+      "loss": 3.0629,
+      "step": 42564
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7197105884552002,
+      "learning_rate": 0.00012723934990204816,
+      "loss": 3.0432,
+      "step": 42565
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6942613124847412,
+      "learning_rate": 0.00012722676309571057,
+      "loss": 2.7,
+      "step": 42566
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6984099745750427,
+      "learning_rate": 0.00012721417674442433,
+      "loss": 3.1964,
+      "step": 42567
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6897585988044739,
+      "learning_rate": 0.0001272015908482227,
+      "loss": 2.9698,
+      "step": 42568
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6685306429862976,
+      "learning_rate": 0.00012718900540713892,
+      "loss": 2.8009,
+      "step": 42569
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7120461463928223,
+      "learning_rate": 0.00012717642042120596,
+      "loss": 2.9525,
+      "step": 42570
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6789481043815613,
+      "learning_rate": 0.00012716383589045715,
+      "loss": 2.6623,
+      "step": 42571
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6555876135826111,
+      "learning_rate": 0.00012715125181492554,
+      "loss": 2.8546,
+      "step": 42572
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7006644010543823,
+      "learning_rate": 0.00012713866819464412,
+      "loss": 2.9681,
+      "step": 42573
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.681387722492218,
+      "learning_rate": 0.0001271260850296463,
+      "loss": 2.846,
+      "step": 42574
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7118346095085144,
+      "learning_rate": 0.00012711350231996502,
+      "loss": 3.0905,
+      "step": 42575
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.673995316028595,
+      "learning_rate": 0.0001271009200656336,
+      "loss": 2.8904,
+      "step": 42576
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6982349157333374,
+      "learning_rate": 0.00012708833826668495,
+      "loss": 2.7196,
+      "step": 42577
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6825549006462097,
+      "learning_rate": 0.00012707575692315248,
+      "loss": 3.1081,
+      "step": 42578
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6906778812408447,
+      "learning_rate": 0.00012706317603506916,
+      "loss": 2.7565,
+      "step": 42579
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6175122857093811,
+      "learning_rate": 0.00012705059560246806,
+      "loss": 3.0223,
+      "step": 42580
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.681134819984436,
+      "learning_rate": 0.00012703801562538252,
+      "loss": 2.924,
+      "step": 42581
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6651497483253479,
+      "learning_rate": 0.00012702543610384546,
+      "loss": 2.8553,
+      "step": 42582
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6877047419548035,
+      "learning_rate": 0.00012701285703789006,
+      "loss": 3.2676,
+      "step": 42583
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.737549901008606,
+      "learning_rate": 0.00012700027842754968,
+      "loss": 2.7642,
+      "step": 42584
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.669161319732666,
+      "learning_rate": 0.0001269877002728572,
+      "loss": 2.8865,
+      "step": 42585
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6651346683502197,
+      "learning_rate": 0.00012697512257384586,
+      "loss": 3.0816,
+      "step": 42586
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6895802021026611,
+      "learning_rate": 0.00012696254533054867,
+      "loss": 3.0483,
+      "step": 42587
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6694716811180115,
+      "learning_rate": 0.0001269499685429988,
+      "loss": 3.1017,
+      "step": 42588
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7255198955535889,
+      "learning_rate": 0.0001269373922112295,
+      "loss": 3.0076,
+      "step": 42589
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6456109881401062,
+      "learning_rate": 0.0001269248163352737,
+      "loss": 2.855,
+      "step": 42590
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6860771775245667,
+      "learning_rate": 0.00012691224091516476,
+      "loss": 3.0306,
+      "step": 42591
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6903136968612671,
+      "learning_rate": 0.00012689966595093557,
+      "loss": 3.0801,
+      "step": 42592
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7179170846939087,
+      "learning_rate": 0.00012688709144261938,
+      "loss": 2.8021,
+      "step": 42593
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6981700658798218,
+      "learning_rate": 0.00012687451739024934,
+      "loss": 2.8115,
+      "step": 42594
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6771155595779419,
+      "learning_rate": 0.0001268619437938584,
+      "loss": 2.8058,
+      "step": 42595
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6560533046722412,
+      "learning_rate": 0.00012684937065347984,
+      "loss": 2.8685,
+      "step": 42596
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6963851451873779,
+      "learning_rate": 0.00012683679796914664,
+      "loss": 2.8669,
+      "step": 42597
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7008418440818787,
+      "learning_rate": 0.00012682422574089198,
+      "loss": 3.0551,
+      "step": 42598
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7024134993553162,
+      "learning_rate": 0.00012681165396874908,
+      "loss": 2.8471,
+      "step": 42599
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7309966683387756,
+      "learning_rate": 0.00012679908265275094,
+      "loss": 2.7808,
+      "step": 42600
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6708288192749023,
+      "learning_rate": 0.00012678651179293072,
+      "loss": 2.6906,
+      "step": 42601
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6789960861206055,
+      "learning_rate": 0.00012677394138932137,
+      "loss": 2.7973,
+      "step": 42602
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7056165337562561,
+      "learning_rate": 0.0001267613714419561,
+      "loss": 3.0738,
+      "step": 42603
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6883794069290161,
+      "learning_rate": 0.0001267488019508682,
+      "loss": 3.0586,
+      "step": 42604
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7074939012527466,
+      "learning_rate": 0.00012673623291609043,
+      "loss": 2.7751,
+      "step": 42605
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7160875797271729,
+      "learning_rate": 0.00012672366433765623,
+      "loss": 2.8472,
+      "step": 42606
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7391530871391296,
+      "learning_rate": 0.00012671109621559856,
+      "loss": 2.7816,
+      "step": 42607
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7099108695983887,
+      "learning_rate": 0.00012669852854995037,
+      "loss": 2.8312,
+      "step": 42608
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6549707651138306,
+      "learning_rate": 0.00012668596134074504,
+      "loss": 2.9045,
+      "step": 42609
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6979371309280396,
+      "learning_rate": 0.00012667339458801542,
+      "loss": 2.9533,
+      "step": 42610
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7130467295646667,
+      "learning_rate": 0.0001266608282917948,
+      "loss": 2.8201,
+      "step": 42611
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.674332857131958,
+      "learning_rate": 0.00012664826245211612,
+      "loss": 2.9441,
+      "step": 42612
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6953020095825195,
+      "learning_rate": 0.0001266356970690126,
+      "loss": 2.7931,
+      "step": 42613
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6988925933837891,
+      "learning_rate": 0.0001266231321425173,
+      "loss": 2.7282,
+      "step": 42614
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7261149287223816,
+      "learning_rate": 0.00012661056767266333,
+      "loss": 2.8631,
+      "step": 42615
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.667811393737793,
+      "learning_rate": 0.0001265980036594838,
+      "loss": 2.9077,
+      "step": 42616
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.694728434085846,
+      "learning_rate": 0.00012658544010301165,
+      "loss": 3.0601,
+      "step": 42617
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7233669757843018,
+      "learning_rate": 0.00012657287700328012,
+      "loss": 2.75,
+      "step": 42618
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7053804993629456,
+      "learning_rate": 0.00012656031436032218,
+      "loss": 3.0824,
+      "step": 42619
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6913226246833801,
+      "learning_rate": 0.000126547752174171,
+      "loss": 2.8967,
+      "step": 42620
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.720045804977417,
+      "learning_rate": 0.00012653519044485975,
+      "loss": 2.7501,
+      "step": 42621
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7346246242523193,
+      "learning_rate": 0.00012652262917242144,
+      "loss": 3.0411,
+      "step": 42622
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6882084012031555,
+      "learning_rate": 0.00012651006835688917,
+      "loss": 2.87,
+      "step": 42623
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6975372433662415,
+      "learning_rate": 0.00012649750799829585,
+      "loss": 2.9828,
+      "step": 42624
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6522185802459717,
+      "learning_rate": 0.00012648494809667469,
+      "loss": 3.0208,
+      "step": 42625
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.709596574306488,
+      "learning_rate": 0.0001264723886520589,
+      "loss": 2.7963,
+      "step": 42626
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6858445405960083,
+      "learning_rate": 0.00012645982966448134,
+      "loss": 3.0234,
+      "step": 42627
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6798964142799377,
+      "learning_rate": 0.00012644727113397532,
+      "loss": 3.0091,
+      "step": 42628
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7389910221099854,
+      "learning_rate": 0.00012643471306057372,
+      "loss": 2.7046,
+      "step": 42629
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6868226528167725,
+      "learning_rate": 0.00012642215544430964,
+      "loss": 2.7799,
+      "step": 42630
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6818183064460754,
+      "learning_rate": 0.00012640959828521628,
+      "loss": 3.1201,
+      "step": 42631
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6521576046943665,
+      "learning_rate": 0.0001263970415833265,
+      "loss": 3.1,
+      "step": 42632
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7379617691040039,
+      "learning_rate": 0.00012638448533867366,
+      "loss": 2.8706,
+      "step": 42633
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7239534258842468,
+      "learning_rate": 0.00012637192955129053,
+      "loss": 2.6417,
+      "step": 42634
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7074881792068481,
+      "learning_rate": 0.00012635937422121035,
+      "loss": 2.987,
+      "step": 42635
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7031037211418152,
+      "learning_rate": 0.00012634681934846624,
+      "loss": 2.9419,
+      "step": 42636
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6728188991546631,
+      "learning_rate": 0.0001263342649330912,
+      "loss": 2.7092,
+      "step": 42637
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7104039788246155,
+      "learning_rate": 0.00012632171097511828,
+      "loss": 2.6515,
+      "step": 42638
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6465256214141846,
+      "learning_rate": 0.00012630915747458043,
+      "loss": 3.0457,
+      "step": 42639
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.690544605255127,
+      "learning_rate": 0.00012629660443151086,
+      "loss": 2.8131,
+      "step": 42640
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6940127611160278,
+      "learning_rate": 0.0001262840518459427,
+      "loss": 2.9941,
+      "step": 42641
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6851858496665955,
+      "learning_rate": 0.0001262714997179088,
+      "loss": 2.9212,
+      "step": 42642
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6655691862106323,
+      "learning_rate": 0.00012625894804744244,
+      "loss": 2.8587,
+      "step": 42643
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.688534140586853,
+      "learning_rate": 0.0001262463968345766,
+      "loss": 3.0816,
+      "step": 42644
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7009671330451965,
+      "learning_rate": 0.00012623384607934415,
+      "loss": 2.8889,
+      "step": 42645
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6391334533691406,
+      "learning_rate": 0.00012622129578177845,
+      "loss": 2.8735,
+      "step": 42646
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6934916973114014,
+      "learning_rate": 0.0001262087459419123,
+      "loss": 2.8965,
+      "step": 42647
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6752188205718994,
+      "learning_rate": 0.00012619619655977897,
+      "loss": 2.9981,
+      "step": 42648
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6940574645996094,
+      "learning_rate": 0.0001261836476354113,
+      "loss": 2.7263,
+      "step": 42649
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7268062829971313,
+      "learning_rate": 0.00012617109916884247,
+      "loss": 3.1159,
+      "step": 42650
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.656769871711731,
+      "learning_rate": 0.0001261585511601056,
+      "loss": 2.9465,
+      "step": 42651
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6630927920341492,
+      "learning_rate": 0.00012614600360923367,
+      "loss": 2.9908,
+      "step": 42652
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6990768909454346,
+      "learning_rate": 0.0001261334565162597,
+      "loss": 2.9062,
+      "step": 42653
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.689720630645752,
+      "learning_rate": 0.00012612090988121661,
+      "loss": 2.9548,
+      "step": 42654
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7280600666999817,
+      "learning_rate": 0.00012610836370413763,
+      "loss": 2.8982,
+      "step": 42655
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6745010614395142,
+      "learning_rate": 0.0001260958179850558,
+      "loss": 2.8903,
+      "step": 42656
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7021997570991516,
+      "learning_rate": 0.00012608327272400406,
+      "loss": 3.0062,
+      "step": 42657
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.648439884185791,
+      "learning_rate": 0.00012607072792101558,
+      "loss": 2.9718,
+      "step": 42658
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7959494590759277,
+      "learning_rate": 0.00012605818357612334,
+      "loss": 2.9918,
+      "step": 42659
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.676099419593811,
+      "learning_rate": 0.00012604563968936027,
+      "loss": 3.0113,
+      "step": 42660
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6749578714370728,
+      "learning_rate": 0.00012603309626075962,
+      "loss": 2.8392,
+      "step": 42661
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6763250231742859,
+      "learning_rate": 0.0001260205532903542,
+      "loss": 3.0267,
+      "step": 42662
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7066740989685059,
+      "learning_rate": 0.00012600801077817728,
+      "loss": 2.9946,
+      "step": 42663
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6909931898117065,
+      "learning_rate": 0.00012599546872426164,
+      "loss": 2.8585,
+      "step": 42664
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6577308177947998,
+      "learning_rate": 0.00012598292712864058,
+      "loss": 2.7868,
+      "step": 42665
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7595558166503906,
+      "learning_rate": 0.000125970385991347,
+      "loss": 2.8711,
+      "step": 42666
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.670626699924469,
+      "learning_rate": 0.0001259578453124138,
+      "loss": 2.9837,
+      "step": 42667
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7183277010917664,
+      "learning_rate": 0.00012594530509187424,
+      "loss": 2.9113,
+      "step": 42668
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6675026416778564,
+      "learning_rate": 0.00012593276532976117,
+      "loss": 3.0022,
+      "step": 42669
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.703376293182373,
+      "learning_rate": 0.00012592022602610768,
+      "loss": 2.9697,
+      "step": 42670
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6627922654151917,
+      "learning_rate": 0.00012590768718094694,
+      "loss": 3.0211,
+      "step": 42671
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.727324366569519,
+      "learning_rate": 0.00012589514879431172,
+      "loss": 2.8978,
+      "step": 42672
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6731564402580261,
+      "learning_rate": 0.00012588261086623527,
+      "loss": 2.9288,
+      "step": 42673
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6762987971305847,
+      "learning_rate": 0.00012587007339675052,
+      "loss": 2.8168,
+      "step": 42674
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7034830451011658,
+      "learning_rate": 0.0001258575363858904,
+      "loss": 2.9776,
+      "step": 42675
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6867435574531555,
+      "learning_rate": 0.00012584499983368807,
+      "loss": 2.8703,
+      "step": 42676
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6966498494148254,
+      "learning_rate": 0.00012583246374017645,
+      "loss": 2.6529,
+      "step": 42677
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6629147529602051,
+      "learning_rate": 0.00012581992810538868,
+      "loss": 2.7213,
+      "step": 42678
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6690661907196045,
+      "learning_rate": 0.00012580739292935759,
+      "loss": 3.1114,
+      "step": 42679
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7012913227081299,
+      "learning_rate": 0.00012579485821211643,
+      "loss": 2.9342,
+      "step": 42680
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7117915749549866,
+      "learning_rate": 0.00012578232395369808,
+      "loss": 2.8775,
+      "step": 42681
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6585233807563782,
+      "learning_rate": 0.00012576979015413542,
+      "loss": 2.899,
+      "step": 42682
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.744594156742096,
+      "learning_rate": 0.00012575725681346174,
+      "loss": 2.9088,
+      "step": 42683
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6832166910171509,
+      "learning_rate": 0.00012574472393170978,
+      "loss": 3.0143,
+      "step": 42684
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7105337977409363,
+      "learning_rate": 0.0001257321915089127,
+      "loss": 3.03,
+      "step": 42685
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6521598100662231,
+      "learning_rate": 0.00012571965954510358,
+      "loss": 2.9901,
+      "step": 42686
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6741447448730469,
+      "learning_rate": 0.00012570712804031535,
+      "loss": 2.9124,
+      "step": 42687
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6789092421531677,
+      "learning_rate": 0.000125694596994581,
+      "loss": 2.9516,
+      "step": 42688
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7190807461738586,
+      "learning_rate": 0.00012568206640793344,
+      "loss": 2.7755,
+      "step": 42689
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7378038763999939,
+      "learning_rate": 0.00012566953628040574,
+      "loss": 2.8331,
+      "step": 42690
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.695372998714447,
+      "learning_rate": 0.00012565700661203107,
+      "loss": 3.0781,
+      "step": 42691
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7172736525535583,
+      "learning_rate": 0.00012564447740284217,
+      "loss": 2.9251,
+      "step": 42692
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6886417865753174,
+      "learning_rate": 0.00012563194865287228,
+      "loss": 2.979,
+      "step": 42693
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7642080187797546,
+      "learning_rate": 0.00012561942036215415,
+      "loss": 2.9911,
+      "step": 42694
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6821017265319824,
+      "learning_rate": 0.00012560689253072102,
+      "loss": 2.8739,
+      "step": 42695
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7164608240127563,
+      "learning_rate": 0.00012559436515860575,
+      "loss": 2.6133,
+      "step": 42696
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6764235496520996,
+      "learning_rate": 0.0001255818382458413,
+      "loss": 3.0949,
+      "step": 42697
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7024563550949097,
+      "learning_rate": 0.0001255693117924608,
+      "loss": 2.678,
+      "step": 42698
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7356938719749451,
+      "learning_rate": 0.00012555678579849707,
+      "loss": 2.7922,
+      "step": 42699
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6670063734054565,
+      "learning_rate": 0.00012554426026398327,
+      "loss": 2.9197,
+      "step": 42700
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.709257185459137,
+      "learning_rate": 0.00012553173518895224,
+      "loss": 2.8767,
+      "step": 42701
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6837866902351379,
+      "learning_rate": 0.00012551921057343715,
+      "loss": 2.9562,
+      "step": 42702
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6681262850761414,
+      "learning_rate": 0.00012550668641747085,
+      "loss": 2.9842,
+      "step": 42703
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7419562339782715,
+      "learning_rate": 0.00012549416272108625,
+      "loss": 2.988,
+      "step": 42704
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7583781480789185,
+      "learning_rate": 0.0001254816394843166,
+      "loss": 3.1105,
+      "step": 42705
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6745335459709167,
+      "learning_rate": 0.00012546911670719457,
+      "loss": 2.9673,
+      "step": 42706
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6698441505432129,
+      "learning_rate": 0.00012545659438975332,
+      "loss": 2.9013,
+      "step": 42707
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7208741903305054,
+      "learning_rate": 0.00012544407253202588,
+      "loss": 2.9384,
+      "step": 42708
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6900094747543335,
+      "learning_rate": 0.0001254315511340452,
+      "loss": 2.9388,
+      "step": 42709
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7146252393722534,
+      "learning_rate": 0.00012541903019584418,
+      "loss": 3.1268,
+      "step": 42710
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6867741942405701,
+      "learning_rate": 0.0001254065097174558,
+      "loss": 2.9237,
+      "step": 42711
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7199037075042725,
+      "learning_rate": 0.00012539398969891298,
+      "loss": 2.8591,
+      "step": 42712
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7059955596923828,
+      "learning_rate": 0.00012538147014024895,
+      "loss": 2.8159,
+      "step": 42713
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6868063807487488,
+      "learning_rate": 0.00012536895104149642,
+      "loss": 3.0604,
+      "step": 42714
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6989514231681824,
+      "learning_rate": 0.00012535643240268852,
+      "loss": 2.9723,
+      "step": 42715
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7568475008010864,
+      "learning_rate": 0.0001253439142238581,
+      "loss": 3.1044,
+      "step": 42716
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.672238290309906,
+      "learning_rate": 0.0001253313965050383,
+      "loss": 2.7832,
+      "step": 42717
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7147806882858276,
+      "learning_rate": 0.000125318879246262,
+      "loss": 2.7591,
+      "step": 42718
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6832359433174133,
+      "learning_rate": 0.000125306362447562,
+      "loss": 2.829,
+      "step": 42719
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6849291324615479,
+      "learning_rate": 0.00012529384610897158,
+      "loss": 3.1068,
+      "step": 42720
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6654471755027771,
+      "learning_rate": 0.0001252813302305234,
+      "loss": 2.8561,
+      "step": 42721
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7064377069473267,
+      "learning_rate": 0.0001252688148122506,
+      "loss": 2.907,
+      "step": 42722
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.665331244468689,
+      "learning_rate": 0.0001252562998541862,
+      "loss": 2.8152,
+      "step": 42723
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.738166332244873,
+      "learning_rate": 0.0001252437853563631,
+      "loss": 2.8205,
+      "step": 42724
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7051920890808105,
+      "learning_rate": 0.0001252312713188142,
+      "loss": 2.7961,
+      "step": 42725
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7235719561576843,
+      "learning_rate": 0.00012521875774157242,
+      "loss": 3.0615,
+      "step": 42726
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7330026626586914,
+      "learning_rate": 0.00012520624462467077,
+      "loss": 2.9727,
+      "step": 42727
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6940485835075378,
+      "learning_rate": 0.00012519373196814238,
+      "loss": 2.731,
+      "step": 42728
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6567136645317078,
+      "learning_rate": 0.00012518121977201993,
+      "loss": 2.971,
+      "step": 42729
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6531281471252441,
+      "learning_rate": 0.0001251687080363366,
+      "loss": 2.893,
+      "step": 42730
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6377910375595093,
+      "learning_rate": 0.00012515619676112526,
+      "loss": 2.8912,
+      "step": 42731
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6578353643417358,
+      "learning_rate": 0.0001251436859464187,
+      "loss": 2.7222,
+      "step": 42732
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6784064173698425,
+      "learning_rate": 0.00012513117559225016,
+      "loss": 2.9927,
+      "step": 42733
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6903968453407288,
+      "learning_rate": 0.00012511866569865236,
+      "loss": 2.8334,
+      "step": 42734
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6765158772468567,
+      "learning_rate": 0.00012510615626565842,
+      "loss": 2.9354,
+      "step": 42735
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7345734238624573,
+      "learning_rate": 0.0001250936472933011,
+      "loss": 2.8646,
+      "step": 42736
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6818332076072693,
+      "learning_rate": 0.00012508113878161348,
+      "loss": 3.0219,
+      "step": 42737
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6421600580215454,
+      "learning_rate": 0.00012506863073062855,
+      "loss": 2.9526,
+      "step": 42738
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7006136178970337,
+      "learning_rate": 0.00012505612314037918,
+      "loss": 2.8891,
+      "step": 42739
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6548375487327576,
+      "learning_rate": 0.00012504361601089833,
+      "loss": 2.9922,
+      "step": 42740
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6888555884361267,
+      "learning_rate": 0.0001250311093422188,
+      "loss": 2.828,
+      "step": 42741
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7467426657676697,
+      "learning_rate": 0.00012501860313437368,
+      "loss": 3.1437,
+      "step": 42742
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6846618056297302,
+      "learning_rate": 0.00012500609738739596,
+      "loss": 2.9356,
+      "step": 42743
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6674962043762207,
+      "learning_rate": 0.00012499359210131842,
+      "loss": 2.9799,
+      "step": 42744
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6744292378425598,
+      "learning_rate": 0.00012498108727617416,
+      "loss": 2.8018,
+      "step": 42745
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6720545291900635,
+      "learning_rate": 0.00012496858291199608,
+      "loss": 2.9251,
+      "step": 42746
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.702659547328949,
+      "learning_rate": 0.00012495607900881692,
+      "loss": 2.9121,
+      "step": 42747
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6851487159729004,
+      "learning_rate": 0.00012494357556666987,
+      "loss": 2.9867,
+      "step": 42748
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6916007995605469,
+      "learning_rate": 0.00012493107258558763,
+      "loss": 2.9223,
+      "step": 42749
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6949058771133423,
+      "learning_rate": 0.0001249185700656034,
+      "loss": 2.7184,
+      "step": 42750
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6698328256607056,
+      "learning_rate": 0.00012490606800674982,
+      "loss": 2.7598,
+      "step": 42751
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.678553581237793,
+      "learning_rate": 0.00012489356640906,
+      "loss": 2.9062,
+      "step": 42752
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6635074615478516,
+      "learning_rate": 0.0001248810652725669,
+      "loss": 3.1156,
+      "step": 42753
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6807247400283813,
+      "learning_rate": 0.00012486856459730336,
+      "loss": 2.9481,
+      "step": 42754
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7227141261100769,
+      "learning_rate": 0.00012485606438330236,
+      "loss": 2.9745,
+      "step": 42755
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7141606211662292,
+      "learning_rate": 0.00012484356463059662,
+      "loss": 2.7561,
+      "step": 42756
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7132400870323181,
+      "learning_rate": 0.00012483106533921925,
+      "loss": 2.8926,
+      "step": 42757
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7027788162231445,
+      "learning_rate": 0.00012481856650920327,
+      "loss": 2.9338,
+      "step": 42758
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6948914527893066,
+      "learning_rate": 0.00012480606814058132,
+      "loss": 2.8984,
+      "step": 42759
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6889869570732117,
+      "learning_rate": 0.0001247935702333866,
+      "loss": 2.8425,
+      "step": 42760
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7095881104469299,
+      "learning_rate": 0.00012478107278765188,
+      "loss": 2.9341,
+      "step": 42761
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7111302018165588,
+      "learning_rate": 0.00012476857580341,
+      "loss": 2.8932,
+      "step": 42762
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6622797846794128,
+      "learning_rate": 0.00012475607928069408,
+      "loss": 2.7765,
+      "step": 42763
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6576592922210693,
+      "learning_rate": 0.00012474358321953681,
+      "loss": 2.9659,
+      "step": 42764
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.68706876039505,
+      "learning_rate": 0.00012473108761997133,
+      "loss": 2.8225,
+      "step": 42765
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7006784081459045,
+      "learning_rate": 0.0001247185924820303,
+      "loss": 2.9659,
+      "step": 42766
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7574266195297241,
+      "learning_rate": 0.00012470609780574688,
+      "loss": 3.0306,
+      "step": 42767
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.678636372089386,
+      "learning_rate": 0.00012469360359115389,
+      "loss": 2.8967,
+      "step": 42768
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7261688113212585,
+      "learning_rate": 0.00012468110983828408,
+      "loss": 2.8648,
+      "step": 42769
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6693586707115173,
+      "learning_rate": 0.00012466861654717058,
+      "loss": 3.0083,
+      "step": 42770
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.699759840965271,
+      "learning_rate": 0.00012465612371784611,
+      "loss": 2.8289,
+      "step": 42771
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6655587553977966,
+      "learning_rate": 0.0001246436313503437,
+      "loss": 2.9072,
+      "step": 42772
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6945595741271973,
+      "learning_rate": 0.00012463113944469626,
+      "loss": 2.9781,
+      "step": 42773
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6992049217224121,
+      "learning_rate": 0.00012461864800093655,
+      "loss": 2.8408,
+      "step": 42774
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6617697477340698,
+      "learning_rate": 0.0001246061570190977,
+      "loss": 2.8496,
+      "step": 42775
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6566224098205566,
+      "learning_rate": 0.00012459366649921247,
+      "loss": 2.8566,
+      "step": 42776
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6914481520652771,
+      "learning_rate": 0.00012458117644131375,
+      "loss": 2.9731,
+      "step": 42777
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6699321866035461,
+      "learning_rate": 0.00012456868684543437,
+      "loss": 2.8621,
+      "step": 42778
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7015001177787781,
+      "learning_rate": 0.00012455619771160732,
+      "loss": 2.9381,
+      "step": 42779
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7836304306983948,
+      "learning_rate": 0.00012454370903986555,
+      "loss": 2.8052,
+      "step": 42780
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6836073994636536,
+      "learning_rate": 0.0001245312208302418,
+      "loss": 2.8882,
+      "step": 42781
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6994935870170593,
+      "learning_rate": 0.00012451873308276913,
+      "loss": 3.035,
+      "step": 42782
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6761261224746704,
+      "learning_rate": 0.00012450624579748035,
+      "loss": 2.9866,
+      "step": 42783
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6878436207771301,
+      "learning_rate": 0.00012449375897440824,
+      "loss": 2.9566,
+      "step": 42784
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6756868958473206,
+      "learning_rate": 0.0001244812726135859,
+      "loss": 2.941,
+      "step": 42785
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.669620156288147,
+      "learning_rate": 0.00012446878671504603,
+      "loss": 2.9263,
+      "step": 42786
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6921489834785461,
+      "learning_rate": 0.00012445630127882166,
+      "loss": 2.9708,
+      "step": 42787
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7075048089027405,
+      "learning_rate": 0.00012444381630494553,
+      "loss": 2.9319,
+      "step": 42788
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7006393074989319,
+      "learning_rate": 0.0001244313317934507,
+      "loss": 2.8783,
+      "step": 42789
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.698784351348877,
+      "learning_rate": 0.00012441884774436993,
+      "loss": 3.0471,
+      "step": 42790
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6695194840431213,
+      "learning_rate": 0.00012440636415773606,
+      "loss": 2.7919,
+      "step": 42791
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6728737354278564,
+      "learning_rate": 0.0001243938810335821,
+      "loss": 2.995,
+      "step": 42792
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6940301656723022,
+      "learning_rate": 0.00012438139837194076,
+      "loss": 2.8124,
+      "step": 42793
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.683421790599823,
+      "learning_rate": 0.00012436891617284503,
+      "loss": 2.7807,
+      "step": 42794
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6776772141456604,
+      "learning_rate": 0.00012435643443632787,
+      "loss": 2.8754,
+      "step": 42795
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6897674202919006,
+      "learning_rate": 0.00012434395316242197,
+      "loss": 3.0021,
+      "step": 42796
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6728845238685608,
+      "learning_rate": 0.00012433147235116037,
+      "loss": 3.2454,
+      "step": 42797
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7249746918678284,
+      "learning_rate": 0.0001243189920025759,
+      "loss": 2.924,
+      "step": 42798
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7298779487609863,
+      "learning_rate": 0.00012430651211670129,
+      "loss": 3.0579,
+      "step": 42799
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.8019329905509949,
+      "learning_rate": 0.00012429403269356956,
+      "loss": 3.0202,
+      "step": 42800
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7995365858078003,
+      "learning_rate": 0.00012428155373321348,
+      "loss": 2.7468,
+      "step": 42801
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.672685980796814,
+      "learning_rate": 0.00012426907523566607,
+      "loss": 2.8585,
+      "step": 42802
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7194468379020691,
+      "learning_rate": 0.00012425659720096,
+      "loss": 2.9569,
+      "step": 42803
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.728124737739563,
+      "learning_rate": 0.0001242441196291283,
+      "loss": 2.766,
+      "step": 42804
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7027937173843384,
+      "learning_rate": 0.00012423164252020377,
+      "loss": 2.9544,
+      "step": 42805
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7222430109977722,
+      "learning_rate": 0.00012421916587421918,
+      "loss": 2.8564,
+      "step": 42806
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.693401038646698,
+      "learning_rate": 0.0001242066896912076,
+      "loss": 2.9747,
+      "step": 42807
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7571882605552673,
+      "learning_rate": 0.00012419421397120162,
+      "loss": 2.9809,
+      "step": 42808
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6618812680244446,
+      "learning_rate": 0.00012418173871423427,
+      "loss": 3.0709,
+      "step": 42809
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6777374148368835,
+      "learning_rate": 0.00012416926392033847,
+      "loss": 2.8907,
+      "step": 42810
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6922290325164795,
+      "learning_rate": 0.00012415678958954703,
+      "loss": 2.8615,
+      "step": 42811
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7021549940109253,
+      "learning_rate": 0.00012414431572189272,
+      "loss": 2.8617,
+      "step": 42812
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7654609084129333,
+      "learning_rate": 0.00012413184231740835,
+      "loss": 3.0118,
+      "step": 42813
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6596476435661316,
+      "learning_rate": 0.00012411936937612686,
+      "loss": 2.9509,
+      "step": 42814
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7117953896522522,
+      "learning_rate": 0.00012410689689808122,
+      "loss": 3.0015,
+      "step": 42815
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7276239991188049,
+      "learning_rate": 0.00012409442488330404,
+      "loss": 2.7217,
+      "step": 42816
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7083224058151245,
+      "learning_rate": 0.0001240819533318284,
+      "loss": 2.9467,
+      "step": 42817
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6544482111930847,
+      "learning_rate": 0.0001240694822436869,
+      "loss": 2.9033,
+      "step": 42818
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6893659234046936,
+      "learning_rate": 0.00012405701161891268,
+      "loss": 2.8934,
+      "step": 42819
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6625657677650452,
+      "learning_rate": 0.0001240445414575384,
+      "loss": 2.988,
+      "step": 42820
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6781535148620605,
+      "learning_rate": 0.00012403207175959688,
+      "loss": 2.9722,
+      "step": 42821
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6784567832946777,
+      "learning_rate": 0.00012401960252512108,
+      "loss": 2.8418,
+      "step": 42822
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6653459668159485,
+      "learning_rate": 0.0001240071337541437,
+      "loss": 2.7502,
+      "step": 42823
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7122622132301331,
+      "learning_rate": 0.00012399466544669763,
+      "loss": 2.9881,
+      "step": 42824
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6495645642280579,
+      "learning_rate": 0.00012398219760281586,
+      "loss": 2.9813,
+      "step": 42825
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6739984154701233,
+      "learning_rate": 0.0001239697302225311,
+      "loss": 2.9362,
+      "step": 42826
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7379757165908813,
+      "learning_rate": 0.00012395726330587619,
+      "loss": 2.8813,
+      "step": 42827
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6480869054794312,
+      "learning_rate": 0.00012394479685288385,
+      "loss": 2.7832,
+      "step": 42828
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7142587304115295,
+      "learning_rate": 0.00012393233086358706,
+      "loss": 2.8355,
+      "step": 42829
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6860468983650208,
+      "learning_rate": 0.00012391986533801872,
+      "loss": 2.9661,
+      "step": 42830
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6586563587188721,
+      "learning_rate": 0.00012390740027621146,
+      "loss": 2.6982,
+      "step": 42831
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6689882278442383,
+      "learning_rate": 0.00012389493567819832,
+      "loss": 3.0087,
+      "step": 42832
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6792545914649963,
+      "learning_rate": 0.00012388247154401205,
+      "loss": 3.1049,
+      "step": 42833
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7196153998374939,
+      "learning_rate": 0.0001238700078736853,
+      "loss": 3.0143,
+      "step": 42834
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6919540762901306,
+      "learning_rate": 0.00012385754466725122,
+      "loss": 2.8527,
+      "step": 42835
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6894993185997009,
+      "learning_rate": 0.00012384508192474233,
+      "loss": 2.8945,
+      "step": 42836
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6774792075157166,
+      "learning_rate": 0.00012383261964619171,
+      "loss": 2.8883,
+      "step": 42837
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6447596549987793,
+      "learning_rate": 0.00012382015783163195,
+      "loss": 3.1129,
+      "step": 42838
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6821197271347046,
+      "learning_rate": 0.000123807696481096,
+      "loss": 2.8552,
+      "step": 42839
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6983603835105896,
+      "learning_rate": 0.00012379523559461674,
+      "loss": 2.7279,
+      "step": 42840
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7073538303375244,
+      "learning_rate": 0.00012378277517222693,
+      "loss": 2.8989,
+      "step": 42841
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6931063532829285,
+      "learning_rate": 0.00012377031521395942,
+      "loss": 2.8065,
+      "step": 42842
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6739084124565125,
+      "learning_rate": 0.00012375785571984682,
+      "loss": 3.0719,
+      "step": 42843
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.8324639797210693,
+      "learning_rate": 0.00012374539668992212,
+      "loss": 2.7407,
+      "step": 42844
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6686299443244934,
+      "learning_rate": 0.00012373293812421827,
+      "loss": 2.8201,
+      "step": 42845
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6707850694656372,
+      "learning_rate": 0.00012372048002276778,
+      "loss": 2.8545,
+      "step": 42846
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6706002354621887,
+      "learning_rate": 0.00012370802238560378,
+      "loss": 2.9133,
+      "step": 42847
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6939883828163147,
+      "learning_rate": 0.0001236955652127589,
+      "loss": 2.7663,
+      "step": 42848
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7302611470222473,
+      "learning_rate": 0.00012368310850426585,
+      "loss": 2.8755,
+      "step": 42849
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6648881435394287,
+      "learning_rate": 0.00012367065226015768,
+      "loss": 3.1254,
+      "step": 42850
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7274592518806458,
+      "learning_rate": 0.00012365819648046697,
+      "loss": 2.9416,
+      "step": 42851
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7390041947364807,
+      "learning_rate": 0.0001236457411652267,
+      "loss": 2.8705,
+      "step": 42852
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7336735129356384,
+      "learning_rate": 0.00012363328631446956,
+      "loss": 2.8143,
+      "step": 42853
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7421209216117859,
+      "learning_rate": 0.00012362083192822847,
+      "loss": 2.9359,
+      "step": 42854
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6754651665687561,
+      "learning_rate": 0.00012360837800653608,
+      "loss": 2.9147,
+      "step": 42855
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6838744282722473,
+      "learning_rate": 0.00012359592454942538,
+      "loss": 2.9293,
+      "step": 42856
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6924825310707092,
+      "learning_rate": 0.00012358347155692906,
+      "loss": 2.908,
+      "step": 42857
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6746190190315247,
+      "learning_rate": 0.00012357101902907983,
+      "loss": 3.1511,
+      "step": 42858
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7157896757125854,
+      "learning_rate": 0.00012355856696591065,
+      "loss": 3.0123,
+      "step": 42859
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6908437609672546,
+      "learning_rate": 0.00012354611536745416,
+      "loss": 2.7298,
+      "step": 42860
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6571826934814453,
+      "learning_rate": 0.00012353366423374323,
+      "loss": 2.7174,
+      "step": 42861
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7128993272781372,
+      "learning_rate": 0.00012352121356481083,
+      "loss": 2.9394,
+      "step": 42862
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7356405258178711,
+      "learning_rate": 0.00012350876336068953,
+      "loss": 2.8878,
+      "step": 42863
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7242211699485779,
+      "learning_rate": 0.0001234963136214122,
+      "loss": 3.0483,
+      "step": 42864
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6854356527328491,
+      "learning_rate": 0.0001234838643470115,
+      "loss": 2.7815,
+      "step": 42865
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7101714015007019,
+      "learning_rate": 0.00012347141553752035,
+      "loss": 3.0021,
+      "step": 42866
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6805014610290527,
+      "learning_rate": 0.0001234589671929716,
+      "loss": 2.8371,
+      "step": 42867
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.70173180103302,
+      "learning_rate": 0.00012344651931339785,
+      "loss": 3.0057,
+      "step": 42868
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6723647117614746,
+      "learning_rate": 0.0001234340718988321,
+      "loss": 2.9304,
+      "step": 42869
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6882205009460449,
+      "learning_rate": 0.00012342162494930704,
+      "loss": 2.8999,
+      "step": 42870
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7191662788391113,
+      "learning_rate": 0.0001234091784648553,
+      "loss": 3.0633,
+      "step": 42871
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6749732494354248,
+      "learning_rate": 0.00012339673244550993,
+      "loss": 2.9349,
+      "step": 42872
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7107117772102356,
+      "learning_rate": 0.00012338428689130348,
+      "loss": 3.0515,
+      "step": 42873
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6865886449813843,
+      "learning_rate": 0.00012337184180226892,
+      "loss": 2.7808,
+      "step": 42874
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6923301219940186,
+      "learning_rate": 0.00012335939717843882,
+      "loss": 2.9724,
+      "step": 42875
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6706522107124329,
+      "learning_rate": 0.00012334695301984607,
+      "loss": 3.0265,
+      "step": 42876
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6823163628578186,
+      "learning_rate": 0.00012333450932652357,
+      "loss": 3.1459,
+      "step": 42877
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7010007500648499,
+      "learning_rate": 0.00012332206609850399,
+      "loss": 2.9557,
+      "step": 42878
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7058342695236206,
+      "learning_rate": 0.00012330962333582005,
+      "loss": 2.85,
+      "step": 42879
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7387443780899048,
+      "learning_rate": 0.00012329718103850447,
+      "loss": 2.7243,
+      "step": 42880
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6735854744911194,
+      "learning_rate": 0.0001232847392065901,
+      "loss": 2.9853,
+      "step": 42881
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7551968693733215,
+      "learning_rate": 0.00012327229784010982,
+      "loss": 2.844,
+      "step": 42882
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6592845916748047,
+      "learning_rate": 0.0001232598569390962,
+      "loss": 2.9419,
+      "step": 42883
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7430188655853271,
+      "learning_rate": 0.00012324741650358221,
+      "loss": 2.8737,
+      "step": 42884
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.683355450630188,
+      "learning_rate": 0.0001232349765336005,
+      "loss": 2.8757,
+      "step": 42885
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7336805462837219,
+      "learning_rate": 0.00012322253702918374,
+      "loss": 2.922,
+      "step": 42886
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7294582724571228,
+      "learning_rate": 0.0001232100979903649,
+      "loss": 2.9502,
+      "step": 42887
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7106903195381165,
+      "learning_rate": 0.0001231976594171765,
+      "loss": 2.7697,
+      "step": 42888
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.702437698841095,
+      "learning_rate": 0.00012318522130965162,
+      "loss": 3.0185,
+      "step": 42889
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.673358142375946,
+      "learning_rate": 0.0001231727836678227,
+      "loss": 3.0657,
+      "step": 42890
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7490993142127991,
+      "learning_rate": 0.00012316034649172272,
+      "loss": 2.8019,
+      "step": 42891
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6476032137870789,
+      "learning_rate": 0.00012314790978138436,
+      "loss": 2.7008,
+      "step": 42892
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6794806122779846,
+      "learning_rate": 0.00012313547353684027,
+      "loss": 2.9429,
+      "step": 42893
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6979109048843384,
+      "learning_rate": 0.0001231230377581234,
+      "loss": 2.7561,
+      "step": 42894
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7019332647323608,
+      "learning_rate": 0.0001231106024452663,
+      "loss": 2.9071,
+      "step": 42895
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7101191878318787,
+      "learning_rate": 0.0001230981675983019,
+      "loss": 3.0134,
+      "step": 42896
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6974600553512573,
+      "learning_rate": 0.00012308573321726288,
+      "loss": 2.9611,
+      "step": 42897
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6884080171585083,
+      "learning_rate": 0.00012307329930218196,
+      "loss": 2.8888,
+      "step": 42898
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7027839422225952,
+      "learning_rate": 0.00012306086585309197,
+      "loss": 3.136,
+      "step": 42899
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.00012304843287002566,
+      "loss": 2.8276,
+      "step": 42900
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7058283686637878,
+      "learning_rate": 0.00012303600035301558,
+      "loss": 2.8065,
+      "step": 42901
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6885665655136108,
+      "learning_rate": 0.00012302356830209477,
+      "loss": 2.69,
+      "step": 42902
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6902278661727905,
+      "learning_rate": 0.00012301113671729565,
+      "loss": 2.8309,
+      "step": 42903
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6766996383666992,
+      "learning_rate": 0.0001229987055986513,
+      "loss": 2.8522,
+      "step": 42904
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7072258591651917,
+      "learning_rate": 0.00012298627494619415,
+      "loss": 2.952,
+      "step": 42905
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7212955355644226,
+      "learning_rate": 0.0001229738447599572,
+      "loss": 2.9739,
+      "step": 42906
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6990088224411011,
+      "learning_rate": 0.0001229614150399731,
+      "loss": 2.8831,
+      "step": 42907
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6565753221511841,
+      "learning_rate": 0.0001229489857862744,
+      "loss": 3.0789,
+      "step": 42908
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7555392384529114,
+      "learning_rate": 0.00012293655699889414,
+      "loss": 2.8962,
+      "step": 42909
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6769063472747803,
+      "learning_rate": 0.0001229241286778648,
+      "loss": 2.825,
+      "step": 42910
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7057018876075745,
+      "learning_rate": 0.00012291170082321924,
+      "loss": 2.8159,
+      "step": 42911
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6804623007774353,
+      "learning_rate": 0.0001228992734349903,
+      "loss": 2.8989,
+      "step": 42912
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6761884093284607,
+      "learning_rate": 0.00012288684651321055,
+      "loss": 2.9359,
+      "step": 42913
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7361788153648376,
+      "learning_rate": 0.00012287442005791275,
+      "loss": 2.9682,
+      "step": 42914
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7586960792541504,
+      "learning_rate": 0.00012286199406912957,
+      "loss": 2.6563,
+      "step": 42915
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7192031145095825,
+      "learning_rate": 0.00012284956854689377,
+      "loss": 2.9592,
+      "step": 42916
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.673794686794281,
+      "learning_rate": 0.00012283714349123825,
+      "loss": 2.9138,
+      "step": 42917
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6666536331176758,
+      "learning_rate": 0.00012282471890219548,
+      "loss": 2.8351,
+      "step": 42918
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6872618794441223,
+      "learning_rate": 0.0001228122947797984,
+      "loss": 3.0766,
+      "step": 42919
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6858128309249878,
+      "learning_rate": 0.00012279987112407954,
+      "loss": 2.9871,
+      "step": 42920
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6915315389633179,
+      "learning_rate": 0.00012278744793507184,
+      "loss": 2.796,
+      "step": 42921
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6912660002708435,
+      "learning_rate": 0.0001227750252128079,
+      "loss": 2.8063,
+      "step": 42922
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7054868340492249,
+      "learning_rate": 0.00012276260295732028,
+      "loss": 2.9651,
+      "step": 42923
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6961238980293274,
+      "learning_rate": 0.00012275018116864202,
+      "loss": 3.0532,
+      "step": 42924
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7195843458175659,
+      "learning_rate": 0.00012273775984680553,
+      "loss": 2.7511,
+      "step": 42925
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6970654129981995,
+      "learning_rate": 0.00012272533899184367,
+      "loss": 3.1432,
+      "step": 42926
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6991729736328125,
+      "learning_rate": 0.00012271291860378925,
+      "loss": 3.0489,
+      "step": 42927
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6705859303474426,
+      "learning_rate": 0.00012270049868267488,
+      "loss": 2.7341,
+      "step": 42928
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.679558277130127,
+      "learning_rate": 0.0001226880792285333,
+      "loss": 2.8421,
+      "step": 42929
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7303120493888855,
+      "learning_rate": 0.00012267566024139705,
+      "loss": 3.0701,
+      "step": 42930
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6934303641319275,
+      "learning_rate": 0.00012266324172129913,
+      "loss": 2.879,
+      "step": 42931
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6617268323898315,
+      "learning_rate": 0.00012265082366827196,
+      "loss": 3.037,
+      "step": 42932
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7124339938163757,
+      "learning_rate": 0.0001226384060823484,
+      "loss": 2.8715,
+      "step": 42933
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.710786759853363,
+      "learning_rate": 0.00012262598896356128,
+      "loss": 3.0071,
+      "step": 42934
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.708827793598175,
+      "learning_rate": 0.00012261357231194313,
+      "loss": 3.0311,
+      "step": 42935
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7814995646476746,
+      "learning_rate": 0.00012260115612752672,
+      "loss": 3.0753,
+      "step": 42936
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6750069260597229,
+      "learning_rate": 0.0001225887404103446,
+      "loss": 2.8407,
+      "step": 42937
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6857645511627197,
+      "learning_rate": 0.0001225763251604296,
+      "loss": 3.0635,
+      "step": 42938
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7252575755119324,
+      "learning_rate": 0.00012256391037781454,
+      "loss": 3.1004,
+      "step": 42939
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.671009361743927,
+      "learning_rate": 0.00012255149606253183,
+      "loss": 3.0177,
+      "step": 42940
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7091543674468994,
+      "learning_rate": 0.0001225390822146145,
+      "loss": 3.0017,
+      "step": 42941
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7315735816955566,
+      "learning_rate": 0.00012252666883409493,
+      "loss": 2.9222,
+      "step": 42942
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7119027376174927,
+      "learning_rate": 0.00012251425592100607,
+      "loss": 2.8792,
+      "step": 42943
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6784539222717285,
+      "learning_rate": 0.0001225018434753805,
+      "loss": 2.9135,
+      "step": 42944
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7480369806289673,
+      "learning_rate": 0.00012248943149725084,
+      "loss": 2.8613,
+      "step": 42945
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7432013154029846,
+      "learning_rate": 0.00012247701998664994,
+      "loss": 2.8941,
+      "step": 42946
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7087968587875366,
+      "learning_rate": 0.0001224646089436103,
+      "loss": 2.9759,
+      "step": 42947
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7181050181388855,
+      "learning_rate": 0.0001224521983681647,
+      "loss": 2.8121,
+      "step": 42948
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6602891087532043,
+      "learning_rate": 0.00012243978826034595,
+      "loss": 3.1781,
+      "step": 42949
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7048685550689697,
+      "learning_rate": 0.00012242737862018662,
+      "loss": 2.9432,
+      "step": 42950
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6855930685997009,
+      "learning_rate": 0.0001224149694477194,
+      "loss": 2.8298,
+      "step": 42951
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6815904378890991,
+      "learning_rate": 0.00012240256074297687,
+      "loss": 2.8817,
+      "step": 42952
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6977430582046509,
+      "learning_rate": 0.00012239015250599183,
+      "loss": 2.9694,
+      "step": 42953
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7615177631378174,
+      "learning_rate": 0.00012237774473679704,
+      "loss": 2.9469,
+      "step": 42954
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6653721332550049,
+      "learning_rate": 0.00012236533743542496,
+      "loss": 2.8976,
+      "step": 42955
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6686670184135437,
+      "learning_rate": 0.00012235293060190855,
+      "loss": 2.8846,
+      "step": 42956
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6878864765167236,
+      "learning_rate": 0.00012234052423628033,
+      "loss": 2.6982,
+      "step": 42957
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.720576822757721,
+      "learning_rate": 0.00012232811833857283,
+      "loss": 2.9867,
+      "step": 42958
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.686820387840271,
+      "learning_rate": 0.00012231571290881898,
+      "loss": 3.0024,
+      "step": 42959
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6706504821777344,
+      "learning_rate": 0.00012230330794705125,
+      "loss": 2.9889,
+      "step": 42960
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.672160804271698,
+      "learning_rate": 0.00012229090345330254,
+      "loss": 3.0674,
+      "step": 42961
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7444672584533691,
+      "learning_rate": 0.00012227849942760528,
+      "loss": 2.9431,
+      "step": 42962
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7115716338157654,
+      "learning_rate": 0.00012226609586999223,
+      "loss": 2.8684,
+      "step": 42963
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.768132746219635,
+      "learning_rate": 0.00012225369278049617,
+      "loss": 2.7876,
+      "step": 42964
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7231113314628601,
+      "learning_rate": 0.0001222412901591497,
+      "loss": 2.8806,
+      "step": 42965
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6965844035148621,
+      "learning_rate": 0.00012222888800598544,
+      "loss": 2.8743,
+      "step": 42966
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6945066452026367,
+      "learning_rate": 0.000122216486321036,
+      "loss": 2.7055,
+      "step": 42967
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.675247073173523,
+      "learning_rate": 0.00012220408510433409,
+      "loss": 2.7753,
+      "step": 42968
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6834036707878113,
+      "learning_rate": 0.00012219168435591252,
+      "loss": 2.7571,
+      "step": 42969
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6711874008178711,
+      "learning_rate": 0.00012217928407580377,
+      "loss": 2.9645,
+      "step": 42970
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6867322325706482,
+      "learning_rate": 0.00012216688426404062,
+      "loss": 2.949,
+      "step": 42971
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.713107168674469,
+      "learning_rate": 0.00012215448492065568,
+      "loss": 2.974,
+      "step": 42972
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6954575777053833,
+      "learning_rate": 0.00012214208604568148,
+      "loss": 2.8547,
+      "step": 42973
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7084396481513977,
+      "learning_rate": 0.00012212968763915095,
+      "loss": 3.0541,
+      "step": 42974
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6738825440406799,
+      "learning_rate": 0.00012211728970109643,
+      "loss": 2.9123,
+      "step": 42975
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.730086624622345,
+      "learning_rate": 0.00012210489223155087,
+      "loss": 3.0752,
+      "step": 42976
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7121165990829468,
+      "learning_rate": 0.00012209249523054667,
+      "loss": 2.9596,
+      "step": 42977
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6749788522720337,
+      "learning_rate": 0.00012208009869811663,
+      "loss": 2.9951,
+      "step": 42978
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7120500206947327,
+      "learning_rate": 0.00012206770263429344,
+      "loss": 2.8199,
+      "step": 42979
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.71989506483078,
+      "learning_rate": 0.00012205530703910967,
+      "loss": 3.0151,
+      "step": 42980
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.698340117931366,
+      "learning_rate": 0.00012204291191259799,
+      "loss": 2.9014,
+      "step": 42981
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6799280047416687,
+      "learning_rate": 0.00012203051725479094,
+      "loss": 2.9977,
+      "step": 42982
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6946273446083069,
+      "learning_rate": 0.00012201812306572124,
+      "loss": 2.6704,
+      "step": 42983
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7306190133094788,
+      "learning_rate": 0.00012200572934542167,
+      "loss": 2.9042,
+      "step": 42984
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7344149947166443,
+      "learning_rate": 0.00012199333609392462,
+      "loss": 2.9512,
+      "step": 42985
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7163786292076111,
+      "learning_rate": 0.00012198094331126299,
+      "loss": 2.9448,
+      "step": 42986
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7278892397880554,
+      "learning_rate": 0.00012196855099746928,
+      "loss": 3.0489,
+      "step": 42987
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7170532941818237,
+      "learning_rate": 0.00012195615915257607,
+      "loss": 2.9299,
+      "step": 42988
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6998884677886963,
+      "learning_rate": 0.00012194376777661615,
+      "loss": 2.6829,
+      "step": 42989
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7183066606521606,
+      "learning_rate": 0.00012193137686962197,
+      "loss": 2.6611,
+      "step": 42990
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6729427576065063,
+      "learning_rate": 0.00012191898643162638,
+      "loss": 3.1633,
+      "step": 42991
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7029131054878235,
+      "learning_rate": 0.00012190659646266182,
+      "loss": 2.9584,
+      "step": 42992
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.709124743938446,
+      "learning_rate": 0.0001218942069627611,
+      "loss": 3.1043,
+      "step": 42993
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.640663743019104,
+      "learning_rate": 0.00012188181793195679,
+      "loss": 3.0283,
+      "step": 42994
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6382255554199219,
+      "learning_rate": 0.00012186942937028135,
+      "loss": 2.9777,
+      "step": 42995
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6668031215667725,
+      "learning_rate": 0.00012185704127776767,
+      "loss": 2.933,
+      "step": 42996
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6796296834945679,
+      "learning_rate": 0.00012184465365444815,
+      "loss": 2.7516,
+      "step": 42997
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6396820545196533,
+      "learning_rate": 0.00012183226650035555,
+      "loss": 2.989,
+      "step": 42998
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6721262335777283,
+      "learning_rate": 0.00012181987981552255,
+      "loss": 3.1257,
+      "step": 42999
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7053427696228027,
+      "learning_rate": 0.0001218074935999816,
+      "loss": 2.7152,
+      "step": 43000
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6978045105934143,
+      "learning_rate": 0.00012179510785376552,
+      "loss": 3.0351,
+      "step": 43001
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6725461483001709,
+      "learning_rate": 0.00012178272257690682,
+      "loss": 2.95,
+      "step": 43002
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7001270651817322,
+      "learning_rate": 0.00012177033776943805,
+      "loss": 2.9442,
+      "step": 43003
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7013981938362122,
+      "learning_rate": 0.00012175795343139199,
+      "loss": 2.8046,
+      "step": 43004
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7114749550819397,
+      "learning_rate": 0.00012174556956280109,
+      "loss": 2.9528,
+      "step": 43005
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.687053918838501,
+      "learning_rate": 0.00012173318616369815,
+      "loss": 3.0319,
+      "step": 43006
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7071757912635803,
+      "learning_rate": 0.00012172080323411557,
+      "loss": 3.0435,
+      "step": 43007
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7466889023780823,
+      "learning_rate": 0.0001217084207740862,
+      "loss": 3.0157,
+      "step": 43008
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6620727777481079,
+      "learning_rate": 0.00012169603878364255,
+      "loss": 2.9545,
+      "step": 43009
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7407063245773315,
+      "learning_rate": 0.0001216836572628171,
+      "loss": 3.0802,
+      "step": 43010
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6958205699920654,
+      "learning_rate": 0.0001216712762116427,
+      "loss": 2.79,
+      "step": 43011
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7473426461219788,
+      "learning_rate": 0.00012165889563015171,
+      "loss": 2.844,
+      "step": 43012
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.701392412185669,
+      "learning_rate": 0.00012164651551837698,
+      "loss": 3.0328,
+      "step": 43013
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7354457974433899,
+      "learning_rate": 0.00012163413587635091,
+      "loss": 3.0237,
+      "step": 43014
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6868544220924377,
+      "learning_rate": 0.0001216217567041063,
+      "loss": 3.0438,
+      "step": 43015
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6519982814788818,
+      "learning_rate": 0.00012160937800167564,
+      "loss": 2.9513,
+      "step": 43016
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6850939393043518,
+      "learning_rate": 0.00012159699976909144,
+      "loss": 2.9938,
+      "step": 43017
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6990240812301636,
+      "learning_rate": 0.00012158462200638655,
+      "loss": 3.0489,
+      "step": 43018
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6919039487838745,
+      "learning_rate": 0.0001215722447135933,
+      "loss": 2.9913,
+      "step": 43019
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6520496606826782,
+      "learning_rate": 0.00012155986789074439,
+      "loss": 2.8828,
+      "step": 43020
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6608403921127319,
+      "learning_rate": 0.00012154749153787257,
+      "loss": 2.7283,
+      "step": 43021
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6845476031303406,
+      "learning_rate": 0.00012153511565501023,
+      "loss": 3.0882,
+      "step": 43022
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6670951247215271,
+      "learning_rate": 0.00012152274024219016,
+      "loss": 2.8529,
+      "step": 43023
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.684021532535553,
+      "learning_rate": 0.0001215103652994448,
+      "loss": 2.8678,
+      "step": 43024
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7129572033882141,
+      "learning_rate": 0.00012149799082680672,
+      "loss": 3.1019,
+      "step": 43025
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6857742667198181,
+      "learning_rate": 0.00012148561682430866,
+      "loss": 2.7703,
+      "step": 43026
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6729403138160706,
+      "learning_rate": 0.00012147324329198302,
+      "loss": 2.7764,
+      "step": 43027
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7263168692588806,
+      "learning_rate": 0.00012146087022986264,
+      "loss": 2.9186,
+      "step": 43028
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6728549003601074,
+      "learning_rate": 0.00012144849763797981,
+      "loss": 2.9232,
+      "step": 43029
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7353836894035339,
+      "learning_rate": 0.00012143612551636741,
+      "loss": 3.0423,
+      "step": 43030
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7096356749534607,
+      "learning_rate": 0.00012142375386505791,
+      "loss": 3.0581,
+      "step": 43031
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6555191874504089,
+      "learning_rate": 0.00012141138268408373,
+      "loss": 2.6584,
+      "step": 43032
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7258883118629456,
+      "learning_rate": 0.00012139901197347775,
+      "loss": 2.9691,
+      "step": 43033
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6590479016304016,
+      "learning_rate": 0.00012138664173327224,
+      "loss": 2.7755,
+      "step": 43034
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6529909372329712,
+      "learning_rate": 0.00012137427196349996,
+      "loss": 2.913,
+      "step": 43035
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6840808391571045,
+      "learning_rate": 0.00012136190266419361,
+      "loss": 2.9598,
+      "step": 43036
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6781811714172363,
+      "learning_rate": 0.0001213495338353856,
+      "loss": 2.9957,
+      "step": 43037
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7219123840332031,
+      "learning_rate": 0.00012133716547710852,
+      "loss": 3.1447,
+      "step": 43038
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6630083322525024,
+      "learning_rate": 0.00012132479758939487,
+      "loss": 2.7962,
+      "step": 43039
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6876730918884277,
+      "learning_rate": 0.00012131243017227734,
+      "loss": 2.7907,
+      "step": 43040
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6795483827590942,
+      "learning_rate": 0.00012130006322578856,
+      "loss": 2.9542,
+      "step": 43041
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6933289766311646,
+      "learning_rate": 0.00012128769674996093,
+      "loss": 2.8632,
+      "step": 43042
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7193173170089722,
+      "learning_rate": 0.00012127533074482718,
+      "loss": 2.946,
+      "step": 43043
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7053383588790894,
+      "learning_rate": 0.00012126296521041976,
+      "loss": 2.74,
+      "step": 43044
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6752833127975464,
+      "learning_rate": 0.00012125060014677137,
+      "loss": 2.6228,
+      "step": 43045
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6935940980911255,
+      "learning_rate": 0.00012123823555391449,
+      "loss": 3.0145,
+      "step": 43046
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6865079998970032,
+      "learning_rate": 0.00012122587143188162,
+      "loss": 2.8768,
+      "step": 43047
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6913439631462097,
+      "learning_rate": 0.00012121350778070547,
+      "loss": 2.8929,
+      "step": 43048
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7718340754508972,
+      "learning_rate": 0.00012120114460041844,
+      "loss": 3.1285,
+      "step": 43049
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6865830421447754,
+      "learning_rate": 0.00012118878189105318,
+      "loss": 2.8353,
+      "step": 43050
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6636896729469299,
+      "learning_rate": 0.00012117641965264238,
+      "loss": 2.7779,
+      "step": 43051
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7035448551177979,
+      "learning_rate": 0.00012116405788521848,
+      "loss": 3.0683,
+      "step": 43052
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6768655180931091,
+      "learning_rate": 0.00012115169658881401,
+      "loss": 3.105,
+      "step": 43053
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.688899576663971,
+      "learning_rate": 0.00012113933576346143,
+      "loss": 2.7824,
+      "step": 43054
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7167072892189026,
+      "learning_rate": 0.00012112697540919346,
+      "loss": 2.7641,
+      "step": 43055
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7082327008247375,
+      "learning_rate": 0.0001211146155260427,
+      "loss": 2.829,
+      "step": 43056
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6907978653907776,
+      "learning_rate": 0.0001211022561140415,
+      "loss": 3.096,
+      "step": 43057
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6979928016662598,
+      "learning_rate": 0.00012108989717322266,
+      "loss": 2.9573,
+      "step": 43058
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7088404297828674,
+      "learning_rate": 0.00012107753870361857,
+      "loss": 2.769,
+      "step": 43059
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6992489695549011,
+      "learning_rate": 0.0001210651807052617,
+      "loss": 2.806,
+      "step": 43060
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6478296518325806,
+      "learning_rate": 0.00012105282317818484,
+      "loss": 2.7097,
+      "step": 43061
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6695234775543213,
+      "learning_rate": 0.00012104046612242028,
+      "loss": 3.1377,
+      "step": 43062
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6910408735275269,
+      "learning_rate": 0.0001210281095380008,
+      "loss": 2.951,
+      "step": 43063
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6836959719657898,
+      "learning_rate": 0.00012101575342495874,
+      "loss": 2.9718,
+      "step": 43064
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6675450801849365,
+      "learning_rate": 0.00012100339778332671,
+      "loss": 2.9584,
+      "step": 43065
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6959847807884216,
+      "learning_rate": 0.00012099104261313742,
+      "loss": 2.9065,
+      "step": 43066
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6748732924461365,
+      "learning_rate": 0.00012097868791442327,
+      "loss": 2.9597,
+      "step": 43067
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6882777810096741,
+      "learning_rate": 0.00012096633368721678,
+      "loss": 2.9002,
+      "step": 43068
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7168293595314026,
+      "learning_rate": 0.00012095397993155042,
+      "loss": 2.871,
+      "step": 43069
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.73194420337677,
+      "learning_rate": 0.00012094162664745682,
+      "loss": 2.8233,
+      "step": 43070
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7441113591194153,
+      "learning_rate": 0.00012092927383496862,
+      "loss": 2.9003,
+      "step": 43071
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6556605696678162,
+      "learning_rate": 0.00012091692149411815,
+      "loss": 2.7515,
+      "step": 43072
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6657657623291016,
+      "learning_rate": 0.00012090456962493811,
+      "loss": 2.959,
+      "step": 43073
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6989374756813049,
+      "learning_rate": 0.00012089221822746101,
+      "loss": 3.0809,
+      "step": 43074
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6639280319213867,
+      "learning_rate": 0.00012087986730171923,
+      "loss": 2.8231,
+      "step": 43075
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7291037440299988,
+      "learning_rate": 0.00012086751684774548,
+      "loss": 2.933,
+      "step": 43076
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7148304581642151,
+      "learning_rate": 0.00012085516686557214,
+      "loss": 2.7072,
+      "step": 43077
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.693949818611145,
+      "learning_rate": 0.00012084281735523191,
+      "loss": 2.8575,
+      "step": 43078
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7028672695159912,
+      "learning_rate": 0.00012083046831675711,
+      "loss": 2.8634,
+      "step": 43079
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7147821187973022,
+      "learning_rate": 0.0001208181197501804,
+      "loss": 2.8629,
+      "step": 43080
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.706278383731842,
+      "learning_rate": 0.00012080577165553434,
+      "loss": 2.8998,
+      "step": 43081
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6842054128646851,
+      "learning_rate": 0.00012079342403285143,
+      "loss": 2.9569,
+      "step": 43082
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6698657870292664,
+      "learning_rate": 0.00012078107688216415,
+      "loss": 2.9066,
+      "step": 43083
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6596769690513611,
+      "learning_rate": 0.00012076873020350489,
+      "loss": 2.799,
+      "step": 43084
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7220476865768433,
+      "learning_rate": 0.00012075638399690642,
+      "loss": 2.8281,
+      "step": 43085
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7004108428955078,
+      "learning_rate": 0.00012074403826240103,
+      "loss": 2.7315,
+      "step": 43086
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.68475341796875,
+      "learning_rate": 0.00012073169300002136,
+      "loss": 2.9782,
+      "step": 43087
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7112210988998413,
+      "learning_rate": 0.0001207193482098,
+      "loss": 2.8447,
+      "step": 43088
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6924461722373962,
+      "learning_rate": 0.00012070700389176937,
+      "loss": 2.9229,
+      "step": 43089
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6814031004905701,
+      "learning_rate": 0.000120694660045962,
+      "loss": 2.8308,
+      "step": 43090
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7307980060577393,
+      "learning_rate": 0.00012068231667241027,
+      "loss": 3.0753,
+      "step": 43091
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.729032576084137,
+      "learning_rate": 0.0001206699737711468,
+      "loss": 2.8441,
+      "step": 43092
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6686051487922668,
+      "learning_rate": 0.00012065763134220422,
+      "loss": 2.9681,
+      "step": 43093
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7233748435974121,
+      "learning_rate": 0.00012064528938561483,
+      "loss": 2.8097,
+      "step": 43094
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6904696822166443,
+      "learning_rate": 0.00012063294790141134,
+      "loss": 2.8041,
+      "step": 43095
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6654314994812012,
+      "learning_rate": 0.00012062060688962611,
+      "loss": 2.9394,
+      "step": 43096
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6917867064476013,
+      "learning_rate": 0.0001206082663502916,
+      "loss": 2.9945,
+      "step": 43097
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6887367963790894,
+      "learning_rate": 0.00012059592628344051,
+      "loss": 2.899,
+      "step": 43098
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7017031311988831,
+      "learning_rate": 0.00012058358668910509,
+      "loss": 2.7895,
+      "step": 43099
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6768372654914856,
+      "learning_rate": 0.00012057124756731808,
+      "loss": 2.9466,
+      "step": 43100
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7897763252258301,
+      "learning_rate": 0.0001205589089181118,
+      "loss": 3.1558,
+      "step": 43101
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7289750576019287,
+      "learning_rate": 0.00012054657074151879,
+      "loss": 2.8482,
+      "step": 43102
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6732428073883057,
+      "learning_rate": 0.00012053423303757167,
+      "loss": 2.9046,
+      "step": 43103
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6708196997642517,
+      "learning_rate": 0.00012052189580630288,
+      "loss": 2.9676,
+      "step": 43104
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7545298933982849,
+      "learning_rate": 0.00012050955904774486,
+      "loss": 3.1395,
+      "step": 43105
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7323744893074036,
+      "learning_rate": 0.00012049722276192999,
+      "loss": 2.966,
+      "step": 43106
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6957453489303589,
+      "learning_rate": 0.0001204848869488909,
+      "loss": 2.8968,
+      "step": 43107
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6939908266067505,
+      "learning_rate": 0.0001204725516086602,
+      "loss": 2.7944,
+      "step": 43108
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6847617626190186,
+      "learning_rate": 0.00012046021674127011,
+      "loss": 2.9505,
+      "step": 43109
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7532222867012024,
+      "learning_rate": 0.0001204478823467534,
+      "loss": 2.8696,
+      "step": 43110
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6626277565956116,
+      "learning_rate": 0.00012043554842514238,
+      "loss": 3.0799,
+      "step": 43111
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6772818565368652,
+      "learning_rate": 0.00012042321497646946,
+      "loss": 2.9087,
+      "step": 43112
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6805033683776855,
+      "learning_rate": 0.00012041088200076734,
+      "loss": 2.8525,
+      "step": 43113
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7096461057662964,
+      "learning_rate": 0.00012039854949806833,
+      "loss": 2.9631,
+      "step": 43114
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7710531949996948,
+      "learning_rate": 0.00012038621746840501,
+      "loss": 3.0431,
+      "step": 43115
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6965796947479248,
+      "learning_rate": 0.00012037388591180978,
+      "loss": 2.9314,
+      "step": 43116
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6995751857757568,
+      "learning_rate": 0.00012036155482831526,
+      "loss": 2.9496,
+      "step": 43117
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6960641741752625,
+      "learning_rate": 0.00012034922421795378,
+      "loss": 3.0477,
+      "step": 43118
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6957606673240662,
+      "learning_rate": 0.00012033689408075783,
+      "loss": 2.8089,
+      "step": 43119
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7263481616973877,
+      "learning_rate": 0.00012032456441675999,
+      "loss": 3.0083,
+      "step": 43120
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6750003695487976,
+      "learning_rate": 0.0001203122352259926,
+      "loss": 2.727,
+      "step": 43121
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6566562056541443,
+      "learning_rate": 0.00012029990650848817,
+      "loss": 2.8773,
+      "step": 43122
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6913471221923828,
+      "learning_rate": 0.00012028757826427931,
+      "loss": 3.0402,
+      "step": 43123
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6827965974807739,
+      "learning_rate": 0.00012027525049339829,
+      "loss": 2.7393,
+      "step": 43124
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6656849980354309,
+      "learning_rate": 0.00012026292319587776,
+      "loss": 3.0347,
+      "step": 43125
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7655985355377197,
+      "learning_rate": 0.00012025059637175012,
+      "loss": 2.9697,
+      "step": 43126
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7039027810096741,
+      "learning_rate": 0.00012023827002104767,
+      "loss": 2.8746,
+      "step": 43127
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6682581901550293,
+      "learning_rate": 0.00012022594414380317,
+      "loss": 3.0307,
+      "step": 43128
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6695263981819153,
+      "learning_rate": 0.00012021361874004882,
+      "loss": 2.9126,
+      "step": 43129
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6603823900222778,
+      "learning_rate": 0.00012020129380981729,
+      "loss": 3.0492,
+      "step": 43130
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6570489406585693,
+      "learning_rate": 0.00012018896935314086,
+      "loss": 2.8269,
+      "step": 43131
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7224562764167786,
+      "learning_rate": 0.00012017664537005215,
+      "loss": 2.9248,
+      "step": 43132
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7442463636398315,
+      "learning_rate": 0.00012016432186058359,
+      "loss": 2.8481,
+      "step": 43133
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6883277893066406,
+      "learning_rate": 0.00012015199882476748,
+      "loss": 2.8931,
+      "step": 43134
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6893338561058044,
+      "learning_rate": 0.0001201396762626365,
+      "loss": 3.1737,
+      "step": 43135
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6887253522872925,
+      "learning_rate": 0.00012012735417422288,
+      "loss": 2.9977,
+      "step": 43136
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7243923544883728,
+      "learning_rate": 0.0001201150325595592,
+      "loss": 2.78,
+      "step": 43137
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6988846659660339,
+      "learning_rate": 0.000120102711418678,
+      "loss": 2.8849,
+      "step": 43138
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6805744171142578,
+      "learning_rate": 0.00012009039075161162,
+      "loss": 2.8165,
+      "step": 43139
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7949154376983643,
+      "learning_rate": 0.00012007807055839257,
+      "loss": 2.8517,
+      "step": 43140
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6621174216270447,
+      "learning_rate": 0.00012006575083905312,
+      "loss": 2.9517,
+      "step": 43141
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.702193021774292,
+      "learning_rate": 0.00012005343159362588,
+      "loss": 2.8943,
+      "step": 43142
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6978039145469666,
+      "learning_rate": 0.00012004111282214336,
+      "loss": 2.9092,
+      "step": 43143
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7211534976959229,
+      "learning_rate": 0.0001200287945246378,
+      "loss": 2.7052,
+      "step": 43144
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7075209617614746,
+      "learning_rate": 0.00012001647670114186,
+      "loss": 3.0433,
+      "step": 43145
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7127019166946411,
+      "learning_rate": 0.0001200041593516878,
+      "loss": 3.0556,
+      "step": 43146
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.728064239025116,
+      "learning_rate": 0.00011999184247630822,
+      "loss": 3.0178,
+      "step": 43147
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6802637577056885,
+      "learning_rate": 0.00011997952607503553,
+      "loss": 2.7863,
+      "step": 43148
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.716955840587616,
+      "learning_rate": 0.00011996721014790195,
+      "loss": 2.9714,
+      "step": 43149
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.772565484046936,
+      "learning_rate": 0.00011995489469494026,
+      "loss": 2.9096,
+      "step": 43150
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6997973322868347,
+      "learning_rate": 0.00011994257971618261,
+      "loss": 2.8717,
+      "step": 43151
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7149245142936707,
+      "learning_rate": 0.00011993026521166154,
+      "loss": 2.8153,
+      "step": 43152
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7250232696533203,
+      "learning_rate": 0.0001199179511814096,
+      "loss": 2.9153,
+      "step": 43153
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7074378728866577,
+      "learning_rate": 0.0001199056376254591,
+      "loss": 3.0232,
+      "step": 43154
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7458380460739136,
+      "learning_rate": 0.0001198933245438425,
+      "loss": 2.7417,
+      "step": 43155
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.6658608913421631,
+      "learning_rate": 0.00011988101193659214,
+      "loss": 2.7459,
+      "step": 43156
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.698132336139679,
+      "learning_rate": 0.00011986869980374052,
+      "loss": 3.1057,
+      "step": 43157
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6716579794883728,
+      "learning_rate": 0.00011985638814532016,
+      "loss": 3.0451,
+      "step": 43158
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7294808030128479,
+      "learning_rate": 0.00011984407696136332,
+      "loss": 3.0347,
+      "step": 43159
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7262159585952759,
+      "learning_rate": 0.00011983176625190261,
+      "loss": 2.9116,
+      "step": 43160
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7426044344902039,
+      "learning_rate": 0.00011981945601697036,
+      "loss": 3.1963,
+      "step": 43161
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6831039190292358,
+      "learning_rate": 0.0001198071462565989,
+      "loss": 2.8845,
+      "step": 43162
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.717583954334259,
+      "learning_rate": 0.0001197948369708208,
+      "loss": 2.8779,
+      "step": 43163
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7191320061683655,
+      "learning_rate": 0.00011978252815966833,
+      "loss": 2.7909,
+      "step": 43164
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7239760756492615,
+      "learning_rate": 0.00011977021982317412,
+      "loss": 3.0215,
+      "step": 43165
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7211264967918396,
+      "learning_rate": 0.00011975791196137032,
+      "loss": 3.0887,
+      "step": 43166
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7530267238616943,
+      "learning_rate": 0.00011974560457428965,
+      "loss": 2.6428,
+      "step": 43167
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7176784873008728,
+      "learning_rate": 0.00011973329766196422,
+      "loss": 3.0019,
+      "step": 43168
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6822056174278259,
+      "learning_rate": 0.0001197209912244267,
+      "loss": 3.0091,
+      "step": 43169
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7103027701377869,
+      "learning_rate": 0.00011970868526170942,
+      "loss": 2.9864,
+      "step": 43170
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6895686984062195,
+      "learning_rate": 0.00011969637977384464,
+      "loss": 3.0177,
+      "step": 43171
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7093585729598999,
+      "learning_rate": 0.00011968407476086502,
+      "loss": 3.0156,
+      "step": 43172
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7251325249671936,
+      "learning_rate": 0.0001196717702228027,
+      "loss": 2.7788,
+      "step": 43173
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7053051590919495,
+      "learning_rate": 0.00011965946615969028,
+      "loss": 3.0574,
+      "step": 43174
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6850510239601135,
+      "learning_rate": 0.00011964716257156019,
+      "loss": 3.1772,
+      "step": 43175
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.768746554851532,
+      "learning_rate": 0.00011963485945844477,
+      "loss": 2.9536,
+      "step": 43176
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7009607553482056,
+      "learning_rate": 0.00011962255682037643,
+      "loss": 2.8981,
+      "step": 43177
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.714339017868042,
+      "learning_rate": 0.00011961025465738745,
+      "loss": 2.7134,
+      "step": 43178
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6838600039482117,
+      "learning_rate": 0.00011959795296951032,
+      "loss": 2.9917,
+      "step": 43179
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6880089640617371,
+      "learning_rate": 0.00011958565175677759,
+      "loss": 2.663,
+      "step": 43180
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7251917123794556,
+      "learning_rate": 0.00011957335101922145,
+      "loss": 3.0461,
+      "step": 43181
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7145901918411255,
+      "learning_rate": 0.00011956105075687445,
+      "loss": 3.0669,
+      "step": 43182
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7030085325241089,
+      "learning_rate": 0.00011954875096976883,
+      "loss": 3.0314,
+      "step": 43183
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6625970005989075,
+      "learning_rate": 0.00011953645165793717,
+      "loss": 2.8142,
+      "step": 43184
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.718930184841156,
+      "learning_rate": 0.00011952415282141176,
+      "loss": 2.8112,
+      "step": 43185
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7372531294822693,
+      "learning_rate": 0.0001195118544602249,
+      "loss": 2.9687,
+      "step": 43186
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.714538037776947,
+      "learning_rate": 0.00011949955657440918,
+      "loss": 2.6542,
+      "step": 43187
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6826031804084778,
+      "learning_rate": 0.0001194872591639968,
+      "loss": 2.8859,
+      "step": 43188
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7082632780075073,
+      "learning_rate": 0.0001194749622290202,
+      "loss": 2.933,
+      "step": 43189
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6917764544487,
+      "learning_rate": 0.00011946266576951195,
+      "loss": 3.0133,
+      "step": 43190
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7382910251617432,
+      "learning_rate": 0.00011945036978550428,
+      "loss": 2.9339,
+      "step": 43191
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7105042338371277,
+      "learning_rate": 0.00011943807427702957,
+      "loss": 3.0067,
+      "step": 43192
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6918284296989441,
+      "learning_rate": 0.00011942577924412016,
+      "loss": 3.0175,
+      "step": 43193
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7218431830406189,
+      "learning_rate": 0.00011941348468680847,
+      "loss": 2.8759,
+      "step": 43194
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.695606529712677,
+      "learning_rate": 0.00011940119060512702,
+      "loss": 2.9613,
+      "step": 43195
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7161937952041626,
+      "learning_rate": 0.00011938889699910794,
+      "loss": 2.9276,
+      "step": 43196
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6780910491943359,
+      "learning_rate": 0.00011937660386878388,
+      "loss": 2.9779,
+      "step": 43197
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6835953593254089,
+      "learning_rate": 0.00011936431121418708,
+      "loss": 2.7212,
+      "step": 43198
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7107114791870117,
+      "learning_rate": 0.00011935201903534981,
+      "loss": 3.214,
+      "step": 43199
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6647638082504272,
+      "learning_rate": 0.00011933972733230469,
+      "loss": 2.9332,
+      "step": 43200
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7023340463638306,
+      "learning_rate": 0.00011932743610508383,
+      "loss": 3.1139,
+      "step": 43201
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7224500775337219,
+      "learning_rate": 0.00011931514535371983,
+      "loss": 2.8405,
+      "step": 43202
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6990615725517273,
+      "learning_rate": 0.00011930285507824487,
+      "loss": 2.9217,
+      "step": 43203
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7267166972160339,
+      "learning_rate": 0.0001192905652786914,
+      "loss": 2.9494,
+      "step": 43204
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6758955717086792,
+      "learning_rate": 0.00011927827595509191,
+      "loss": 3.1795,
+      "step": 43205
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7026833295822144,
+      "learning_rate": 0.00011926598710747867,
+      "loss": 3.0531,
+      "step": 43206
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7141755819320679,
+      "learning_rate": 0.00011925369873588404,
+      "loss": 2.865,
+      "step": 43207
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6519131064414978,
+      "learning_rate": 0.00011924141084034027,
+      "loss": 2.8592,
+      "step": 43208
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6646400690078735,
+      "learning_rate": 0.00011922912342087986,
+      "loss": 2.6698,
+      "step": 43209
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7201839685440063,
+      "learning_rate": 0.00011921683647753523,
+      "loss": 2.891,
+      "step": 43210
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7078610062599182,
+      "learning_rate": 0.00011920455001033855,
+      "loss": 2.9393,
+      "step": 43211
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7181376218795776,
+      "learning_rate": 0.0001191922640193224,
+      "loss": 2.7381,
+      "step": 43212
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6957877278327942,
+      "learning_rate": 0.00011917997850451902,
+      "loss": 2.9347,
+      "step": 43213
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7163203358650208,
+      "learning_rate": 0.00011916769346596069,
+      "loss": 3.1533,
+      "step": 43214
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7181594967842102,
+      "learning_rate": 0.00011915540890367997,
+      "loss": 2.8268,
+      "step": 43215
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6955720782279968,
+      "learning_rate": 0.00011914312481770898,
+      "loss": 3.059,
+      "step": 43216
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6729768514633179,
+      "learning_rate": 0.00011913084120808029,
+      "loss": 2.8435,
+      "step": 43217
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7270863056182861,
+      "learning_rate": 0.00011911855807482606,
+      "loss": 2.7089,
+      "step": 43218
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6758031845092773,
+      "learning_rate": 0.00011910627541797882,
+      "loss": 2.683,
+      "step": 43219
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6879879236221313,
+      "learning_rate": 0.00011909399323757088,
+      "loss": 3.0191,
+      "step": 43220
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6507390737533569,
+      "learning_rate": 0.00011908171153363439,
+      "loss": 2.9589,
+      "step": 43221
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7297090291976929,
+      "learning_rate": 0.000119069430306202,
+      "loss": 3.0383,
+      "step": 43222
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6910653114318848,
+      "learning_rate": 0.00011905714955530578,
+      "loss": 2.8401,
+      "step": 43223
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7071283459663391,
+      "learning_rate": 0.00011904486928097823,
+      "loss": 2.9656,
+      "step": 43224
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6900044083595276,
+      "learning_rate": 0.00011903258948325173,
+      "loss": 2.885,
+      "step": 43225
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7246055006980896,
+      "learning_rate": 0.0001190203101621585,
+      "loss": 3.0716,
+      "step": 43226
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6860259175300598,
+      "learning_rate": 0.00011900803131773104,
+      "loss": 3.0308,
+      "step": 43227
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6874629855155945,
+      "learning_rate": 0.00011899575295000155,
+      "loss": 2.9469,
+      "step": 43228
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6858367919921875,
+      "learning_rate": 0.00011898347505900236,
+      "loss": 2.8059,
+      "step": 43229
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6908622980117798,
+      "learning_rate": 0.00011897119764476593,
+      "loss": 2.7786,
+      "step": 43230
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.729853630065918,
+      "learning_rate": 0.00011895892070732445,
+      "loss": 2.9305,
+      "step": 43231
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6964291334152222,
+      "learning_rate": 0.00011894664424671043,
+      "loss": 2.8263,
+      "step": 43232
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7044500112533569,
+      "learning_rate": 0.00011893436826295597,
+      "loss": 2.928,
+      "step": 43233
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7081640958786011,
+      "learning_rate": 0.00011892209275609367,
+      "loss": 2.9927,
+      "step": 43234
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7295190095901489,
+      "learning_rate": 0.00011890981772615573,
+      "loss": 3.0935,
+      "step": 43235
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6798872351646423,
+      "learning_rate": 0.00011889754317317436,
+      "loss": 2.8526,
+      "step": 43236
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7571244835853577,
+      "learning_rate": 0.00011888526909718213,
+      "loss": 2.9359,
+      "step": 43237
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7170717120170593,
+      "learning_rate": 0.00011887299549821114,
+      "loss": 2.925,
+      "step": 43238
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7144536375999451,
+      "learning_rate": 0.00011886072237629378,
+      "loss": 2.9109,
+      "step": 43239
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7104244232177734,
+      "learning_rate": 0.00011884844973146254,
+      "loss": 2.8062,
+      "step": 43240
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.672351062297821,
+      "learning_rate": 0.00011883617756374963,
+      "loss": 2.9023,
+      "step": 43241
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6850356459617615,
+      "learning_rate": 0.00011882390587318737,
+      "loss": 3.0001,
+      "step": 43242
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7461333870887756,
+      "learning_rate": 0.00011881163465980797,
+      "loss": 2.7922,
+      "step": 43243
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6776725053787231,
+      "learning_rate": 0.00011879936392364395,
+      "loss": 3.152,
+      "step": 43244
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7232372760772705,
+      "learning_rate": 0.00011878709366472742,
+      "loss": 2.9213,
+      "step": 43245
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6616763472557068,
+      "learning_rate": 0.00011877482388309084,
+      "loss": 3.0254,
+      "step": 43246
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6551421880722046,
+      "learning_rate": 0.00011876255457876656,
+      "loss": 2.9206,
+      "step": 43247
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.712017834186554,
+      "learning_rate": 0.00011875028575178676,
+      "loss": 2.8227,
+      "step": 43248
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6586719751358032,
+      "learning_rate": 0.00011873801740218392,
+      "loss": 3.1015,
+      "step": 43249
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7056365013122559,
+      "learning_rate": 0.00011872574952999024,
+      "loss": 2.7413,
+      "step": 43250
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6891767978668213,
+      "learning_rate": 0.00011871348213523796,
+      "loss": 2.8436,
+      "step": 43251
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7506184577941895,
+      "learning_rate": 0.0001187012152179596,
+      "loss": 2.9205,
+      "step": 43252
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7506493926048279,
+      "learning_rate": 0.00011868894877818721,
+      "loss": 2.7396,
+      "step": 43253
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7198980450630188,
+      "learning_rate": 0.00011867668281595336,
+      "loss": 2.991,
+      "step": 43254
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6733829975128174,
+      "learning_rate": 0.00011866441733129013,
+      "loss": 2.8639,
+      "step": 43255
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.726628839969635,
+      "learning_rate": 0.00011865215232423003,
+      "loss": 2.8438,
+      "step": 43256
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6809700727462769,
+      "learning_rate": 0.00011863988779480524,
+      "loss": 2.8867,
+      "step": 43257
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6893267631530762,
+      "learning_rate": 0.00011862762374304798,
+      "loss": 3.0118,
+      "step": 43258
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7157814502716064,
+      "learning_rate": 0.00011861536016899074,
+      "loss": 3.1546,
+      "step": 43259
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6700130701065063,
+      "learning_rate": 0.00011860309707266568,
+      "loss": 3.0809,
+      "step": 43260
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6558499932289124,
+      "learning_rate": 0.00011859083445410513,
+      "loss": 2.9951,
+      "step": 43261
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7448561787605286,
+      "learning_rate": 0.0001185785723133415,
+      "loss": 2.9421,
+      "step": 43262
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.716131865978241,
+      "learning_rate": 0.000118566310650407,
+      "loss": 2.95,
+      "step": 43263
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.709511399269104,
+      "learning_rate": 0.00011855404946533395,
+      "loss": 3.1927,
+      "step": 43264
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7204448580741882,
+      "learning_rate": 0.00011854178875815446,
+      "loss": 2.855,
+      "step": 43265
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6846634745597839,
+      "learning_rate": 0.000118529528528901,
+      "loss": 2.9704,
+      "step": 43266
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6869163513183594,
+      "learning_rate": 0.00011851726877760594,
+      "loss": 2.8469,
+      "step": 43267
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7922883033752441,
+      "learning_rate": 0.00011850500950430138,
+      "loss": 2.7902,
+      "step": 43268
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.672174870967865,
+      "learning_rate": 0.00011849275070901979,
+      "loss": 3.0636,
+      "step": 43269
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7000868916511536,
+      "learning_rate": 0.00011848049239179324,
+      "loss": 2.9595,
+      "step": 43270
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6964728236198425,
+      "learning_rate": 0.00011846823455265425,
+      "loss": 2.8609,
+      "step": 43271
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7069092392921448,
+      "learning_rate": 0.00011845597719163501,
+      "loss": 3.1638,
+      "step": 43272
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6975874304771423,
+      "learning_rate": 0.00011844372030876769,
+      "loss": 2.9583,
+      "step": 43273
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6885384917259216,
+      "learning_rate": 0.00011843146390408476,
+      "loss": 2.817,
+      "step": 43274
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7786584496498108,
+      "learning_rate": 0.00011841920797761829,
+      "loss": 2.8851,
+      "step": 43275
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6855915784835815,
+      "learning_rate": 0.0001184069525294007,
+      "loss": 2.7401,
+      "step": 43276
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6739755272865295,
+      "learning_rate": 0.00011839469755946432,
+      "loss": 2.9035,
+      "step": 43277
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6957394480705261,
+      "learning_rate": 0.0001183824430678414,
+      "loss": 3.0632,
+      "step": 43278
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6879928112030029,
+      "learning_rate": 0.00011837018905456415,
+      "loss": 2.8398,
+      "step": 43279
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.701741635799408,
+      "learning_rate": 0.00011835793551966477,
+      "loss": 2.8844,
+      "step": 43280
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6822145581245422,
+      "learning_rate": 0.00011834568246317562,
+      "loss": 2.8215,
+      "step": 43281
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6779221296310425,
+      "learning_rate": 0.00011833342988512909,
+      "loss": 3.0876,
+      "step": 43282
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6709139347076416,
+      "learning_rate": 0.00011832117778555727,
+      "loss": 2.7494,
+      "step": 43283
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7509815096855164,
+      "learning_rate": 0.00011830892616449257,
+      "loss": 2.8125,
+      "step": 43284
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6872658133506775,
+      "learning_rate": 0.00011829667502196722,
+      "loss": 2.7154,
+      "step": 43285
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7320376634597778,
+      "learning_rate": 0.00011828442435801331,
+      "loss": 2.9323,
+      "step": 43286
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6770885586738586,
+      "learning_rate": 0.00011827217417266339,
+      "loss": 2.9723,
+      "step": 43287
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7534797191619873,
+      "learning_rate": 0.00011825992446594948,
+      "loss": 3.0001,
+      "step": 43288
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.720881998538971,
+      "learning_rate": 0.00011824767523790406,
+      "loss": 2.8738,
+      "step": 43289
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6875128149986267,
+      "learning_rate": 0.00011823542648855916,
+      "loss": 3.1074,
+      "step": 43290
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7054754495620728,
+      "learning_rate": 0.00011822317821794718,
+      "loss": 2.8814,
+      "step": 43291
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.647639811038971,
+      "learning_rate": 0.00011821093042610049,
+      "loss": 2.9831,
+      "step": 43292
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6805564165115356,
+      "learning_rate": 0.0001181986831130512,
+      "loss": 2.9677,
+      "step": 43293
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.724385678768158,
+      "learning_rate": 0.0001181864362788316,
+      "loss": 2.9741,
+      "step": 43294
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6952561736106873,
+      "learning_rate": 0.00011817418992347381,
+      "loss": 2.7953,
+      "step": 43295
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7037989497184753,
+      "learning_rate": 0.00011816194404701024,
+      "loss": 2.5987,
+      "step": 43296
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6977366209030151,
+      "learning_rate": 0.0001181496986494732,
+      "loss": 2.6767,
+      "step": 43297
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6811829805374146,
+      "learning_rate": 0.00011813745373089473,
+      "loss": 3.0466,
+      "step": 43298
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6867378354072571,
+      "learning_rate": 0.00011812520929130736,
+      "loss": 3.0597,
+      "step": 43299
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6996347308158875,
+      "learning_rate": 0.00011811296533074315,
+      "loss": 2.8842,
+      "step": 43300
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.661796510219574,
+      "learning_rate": 0.00011810072184923429,
+      "loss": 3.1053,
+      "step": 43301
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6767133474349976,
+      "learning_rate": 0.00011808847884681323,
+      "loss": 2.8145,
+      "step": 43302
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.710540771484375,
+      "learning_rate": 0.00011807623632351201,
+      "loss": 2.9525,
+      "step": 43303
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6866672039031982,
+      "learning_rate": 0.00011806399427936308,
+      "loss": 2.9246,
+      "step": 43304
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6833668947219849,
+      "learning_rate": 0.00011805175271439846,
+      "loss": 3.2081,
+      "step": 43305
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6920410394668579,
+      "learning_rate": 0.00011803951162865048,
+      "loss": 3.0166,
+      "step": 43306
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7049962878227234,
+      "learning_rate": 0.00011802727102215155,
+      "loss": 2.7362,
+      "step": 43307
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.727013349533081,
+      "learning_rate": 0.00011801503089493376,
+      "loss": 2.8944,
+      "step": 43308
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6701471209526062,
+      "learning_rate": 0.00011800279124702936,
+      "loss": 2.7099,
+      "step": 43309
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6716243624687195,
+      "learning_rate": 0.00011799055207847045,
+      "loss": 3.0203,
+      "step": 43310
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7197967171669006,
+      "learning_rate": 0.00011797831338928943,
+      "loss": 2.8381,
+      "step": 43311
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6998694539070129,
+      "learning_rate": 0.00011796607517951858,
+      "loss": 2.9622,
+      "step": 43312
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6807000041007996,
+      "learning_rate": 0.00011795383744918996,
+      "loss": 3.1586,
+      "step": 43313
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6802100539207458,
+      "learning_rate": 0.00011794160019833599,
+      "loss": 2.6665,
+      "step": 43314
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6740386486053467,
+      "learning_rate": 0.00011792936342698882,
+      "loss": 2.8575,
+      "step": 43315
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6891768574714661,
+      "learning_rate": 0.00011791712713518056,
+      "loss": 2.8772,
+      "step": 43316
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7108803391456604,
+      "learning_rate": 0.00011790489132294368,
+      "loss": 3.0195,
+      "step": 43317
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7088941335678101,
+      "learning_rate": 0.00011789265599031012,
+      "loss": 2.92,
+      "step": 43318
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6881635785102844,
+      "learning_rate": 0.00011788042113731239,
+      "loss": 2.8271,
+      "step": 43319
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6677702069282532,
+      "learning_rate": 0.00011786818676398243,
+      "loss": 3.0857,
+      "step": 43320
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7288718223571777,
+      "learning_rate": 0.00011785595287035276,
+      "loss": 3.0319,
+      "step": 43321
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6868347525596619,
+      "learning_rate": 0.00011784371945645546,
+      "loss": 2.5492,
+      "step": 43322
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6867601275444031,
+      "learning_rate": 0.00011783148652232264,
+      "loss": 2.7741,
+      "step": 43323
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.715054988861084,
+      "learning_rate": 0.0001178192540679867,
+      "loss": 3.0581,
+      "step": 43324
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6956353187561035,
+      "learning_rate": 0.0001178070220934797,
+      "loss": 2.9861,
+      "step": 43325
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7079643607139587,
+      "learning_rate": 0.00011779479059883407,
+      "loss": 2.8541,
+      "step": 43326
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.712270200252533,
+      "learning_rate": 0.00011778255958408175,
+      "loss": 2.7609,
+      "step": 43327
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6944162845611572,
+      "learning_rate": 0.00011777032904925512,
+      "loss": 2.9822,
+      "step": 43328
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6619276404380798,
+      "learning_rate": 0.00011775809899438645,
+      "loss": 2.9275,
+      "step": 43329
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7741801738739014,
+      "learning_rate": 0.0001177458694195079,
+      "loss": 2.7809,
+      "step": 43330
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7182871103286743,
+      "learning_rate": 0.00011773364032465167,
+      "loss": 2.8416,
+      "step": 43331
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6957319974899292,
+      "learning_rate": 0.00011772141170984984,
+      "loss": 2.8044,
+      "step": 43332
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7118658423423767,
+      "learning_rate": 0.0001177091835751347,
+      "loss": 2.9776,
+      "step": 43333
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7111032605171204,
+      "learning_rate": 0.00011769695592053861,
+      "loss": 2.8035,
+      "step": 43334
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7072035074234009,
+      "learning_rate": 0.00011768472874609355,
+      "loss": 2.9224,
+      "step": 43335
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6957293748855591,
+      "learning_rate": 0.00011767250205183195,
+      "loss": 2.8217,
+      "step": 43336
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6990187764167786,
+      "learning_rate": 0.00011766027583778586,
+      "loss": 3.1363,
+      "step": 43337
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6784077882766724,
+      "learning_rate": 0.00011764805010398743,
+      "loss": 2.9643,
+      "step": 43338
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6869450211524963,
+      "learning_rate": 0.00011763582485046902,
+      "loss": 2.9682,
+      "step": 43339
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7190410494804382,
+      "learning_rate": 0.00011762360007726266,
+      "loss": 2.8506,
+      "step": 43340
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.667700469493866,
+      "learning_rate": 0.00011761137578440076,
+      "loss": 3.0683,
+      "step": 43341
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7749731540679932,
+      "learning_rate": 0.00011759915197191527,
+      "loss": 2.8538,
+      "step": 43342
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7492234706878662,
+      "learning_rate": 0.00011758692863983863,
+      "loss": 2.7344,
+      "step": 43343
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6971567273139954,
+      "learning_rate": 0.0001175747057882029,
+      "loss": 2.8228,
+      "step": 43344
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6927439570426941,
+      "learning_rate": 0.00011756248341704017,
+      "loss": 2.9262,
+      "step": 43345
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6876089572906494,
+      "learning_rate": 0.00011755026152638288,
+      "loss": 3.01,
+      "step": 43346
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7251725792884827,
+      "learning_rate": 0.00011753804011626297,
+      "loss": 2.8696,
+      "step": 43347
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6793904900550842,
+      "learning_rate": 0.00011752581918671276,
+      "loss": 3.1156,
+      "step": 43348
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7280372381210327,
+      "learning_rate": 0.00011751359873776453,
+      "loss": 3.0312,
+      "step": 43349
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7209813594818115,
+      "learning_rate": 0.00011750137876945021,
+      "loss": 3.1338,
+      "step": 43350
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7237643003463745,
+      "learning_rate": 0.00011748915928180228,
+      "loss": 2.7921,
+      "step": 43351
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7128032445907593,
+      "learning_rate": 0.0001174769402748528,
+      "loss": 2.8238,
+      "step": 43352
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.694153368473053,
+      "learning_rate": 0.0001174647217486338,
+      "loss": 2.8686,
+      "step": 43353
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7112940549850464,
+      "learning_rate": 0.00011745250370317769,
+      "loss": 2.7379,
+      "step": 43354
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7519726753234863,
+      "learning_rate": 0.00011744028613851646,
+      "loss": 2.7854,
+      "step": 43355
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7124344706535339,
+      "learning_rate": 0.00011742806905468248,
+      "loss": 3.0715,
+      "step": 43356
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6832916736602783,
+      "learning_rate": 0.00011741585245170773,
+      "loss": 3.0614,
+      "step": 43357
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7317155599594116,
+      "learning_rate": 0.0001174036363296246,
+      "loss": 3.0663,
+      "step": 43358
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.684130072593689,
+      "learning_rate": 0.00011739142068846512,
+      "loss": 2.7761,
+      "step": 43359
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7288141846656799,
+      "learning_rate": 0.00011737920552826142,
+      "loss": 2.9951,
+      "step": 43360
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7078750133514404,
+      "learning_rate": 0.00011736699084904584,
+      "loss": 2.8456,
+      "step": 43361
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6609907746315002,
+      "learning_rate": 0.00011735477665085035,
+      "loss": 2.8533,
+      "step": 43362
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6616264581680298,
+      "learning_rate": 0.00011734256293370722,
+      "loss": 2.7806,
+      "step": 43363
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7512163519859314,
+      "learning_rate": 0.00011733034969764876,
+      "loss": 2.9013,
+      "step": 43364
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6941211819648743,
+      "learning_rate": 0.00011731813694270699,
+      "loss": 2.8684,
+      "step": 43365
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6857606768608093,
+      "learning_rate": 0.00011730592466891406,
+      "loss": 2.8198,
+      "step": 43366
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6707267165184021,
+      "learning_rate": 0.0001172937128763021,
+      "loss": 2.9227,
+      "step": 43367
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6939967274665833,
+      "learning_rate": 0.00011728150156490332,
+      "loss": 2.8507,
+      "step": 43368
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7025797367095947,
+      "learning_rate": 0.00011726929073475,
+      "loss": 2.8585,
+      "step": 43369
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7331152558326721,
+      "learning_rate": 0.0001172570803858741,
+      "loss": 3.0419,
+      "step": 43370
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.669756293296814,
+      "learning_rate": 0.000117244870518308,
+      "loss": 2.9938,
+      "step": 43371
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6814187169075012,
+      "learning_rate": 0.00011723266113208363,
+      "loss": 2.9681,
+      "step": 43372
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6948679089546204,
+      "learning_rate": 0.00011722045222723333,
+      "loss": 3.1171,
+      "step": 43373
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6597676277160645,
+      "learning_rate": 0.00011720824380378922,
+      "loss": 2.6783,
+      "step": 43374
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6691949367523193,
+      "learning_rate": 0.00011719603586178329,
+      "loss": 2.845,
+      "step": 43375
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6968749165534973,
+      "learning_rate": 0.00011718382840124793,
+      "loss": 3.0667,
+      "step": 43376
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7081393599510193,
+      "learning_rate": 0.0001171716214222151,
+      "loss": 2.9093,
+      "step": 43377
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.677230954170227,
+      "learning_rate": 0.00011715941492471702,
+      "loss": 2.9554,
+      "step": 43378
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6978651285171509,
+      "learning_rate": 0.00011714720890878592,
+      "loss": 2.8337,
+      "step": 43379
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7071506381034851,
+      "learning_rate": 0.00011713500337445392,
+      "loss": 3.0446,
+      "step": 43380
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7174202799797058,
+      "learning_rate": 0.00011712279832175313,
+      "loss": 2.9749,
+      "step": 43381
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7007304430007935,
+      "learning_rate": 0.00011711059375071558,
+      "loss": 3.0169,
+      "step": 43382
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7151933908462524,
+      "learning_rate": 0.00011709838966137354,
+      "loss": 3.0288,
+      "step": 43383
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7130842208862305,
+      "learning_rate": 0.00011708618605375923,
+      "loss": 3.0052,
+      "step": 43384
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7058139443397522,
+      "learning_rate": 0.0001170739829279046,
+      "loss": 2.9758,
+      "step": 43385
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6666308641433716,
+      "learning_rate": 0.000117061780283842,
+      "loss": 2.9686,
+      "step": 43386
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7057749032974243,
+      "learning_rate": 0.00011704957812160347,
+      "loss": 2.9458,
+      "step": 43387
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6779221892356873,
+      "learning_rate": 0.00011703737644122103,
+      "loss": 2.8908,
+      "step": 43388
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.690578818321228,
+      "learning_rate": 0.00011702517524272703,
+      "loss": 3.1712,
+      "step": 43389
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6773284077644348,
+      "learning_rate": 0.00011701297452615342,
+      "loss": 2.9113,
+      "step": 43390
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.700973629951477,
+      "learning_rate": 0.00011700077429153249,
+      "loss": 2.9496,
+      "step": 43391
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6661964654922485,
+      "learning_rate": 0.00011698857453889621,
+      "loss": 2.9452,
+      "step": 43392
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6915501952171326,
+      "learning_rate": 0.00011697637526827682,
+      "loss": 2.8738,
+      "step": 43393
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6811676621437073,
+      "learning_rate": 0.00011696417647970654,
+      "loss": 3.0137,
+      "step": 43394
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6550053358078003,
+      "learning_rate": 0.00011695197817321736,
+      "loss": 2.9058,
+      "step": 43395
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7086590528488159,
+      "learning_rate": 0.00011693978034884147,
+      "loss": 2.7958,
+      "step": 43396
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6640387177467346,
+      "learning_rate": 0.00011692758300661087,
+      "loss": 2.9049,
+      "step": 43397
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7309964299201965,
+      "learning_rate": 0.00011691538614655786,
+      "loss": 2.6874,
+      "step": 43398
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7164502143859863,
+      "learning_rate": 0.0001169031897687144,
+      "loss": 3.0238,
+      "step": 43399
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6886323094367981,
+      "learning_rate": 0.00011689099387311268,
+      "loss": 3.0394,
+      "step": 43400
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6922336220741272,
+      "learning_rate": 0.00011687879845978499,
+      "loss": 2.758,
+      "step": 43401
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6996184587478638,
+      "learning_rate": 0.00011686660352876325,
+      "loss": 2.8422,
+      "step": 43402
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6815118789672852,
+      "learning_rate": 0.00011685440908007968,
+      "loss": 2.984,
+      "step": 43403
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7413270473480225,
+      "learning_rate": 0.00011684221511376619,
+      "loss": 2.9322,
+      "step": 43404
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6742251515388489,
+      "learning_rate": 0.00011683002162985512,
+      "loss": 3.0762,
+      "step": 43405
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6736028790473938,
+      "learning_rate": 0.00011681782862837858,
+      "loss": 2.8401,
+      "step": 43406
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6667304635047913,
+      "learning_rate": 0.00011680563610936853,
+      "loss": 2.9525,
+      "step": 43407
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7072355151176453,
+      "learning_rate": 0.0001167934440728573,
+      "loss": 2.7609,
+      "step": 43408
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6967011094093323,
+      "learning_rate": 0.00011678125251887677,
+      "loss": 2.8879,
+      "step": 43409
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6703258156776428,
+      "learning_rate": 0.00011676906144745923,
+      "loss": 3.1206,
+      "step": 43410
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7139511108398438,
+      "learning_rate": 0.00011675687085863677,
+      "loss": 2.9037,
+      "step": 43411
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7220322489738464,
+      "learning_rate": 0.0001167446807524413,
+      "loss": 2.8709,
+      "step": 43412
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6931965351104736,
+      "learning_rate": 0.00011673249112890518,
+      "loss": 3.021,
+      "step": 43413
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7006539106369019,
+      "learning_rate": 0.00011672030198806031,
+      "loss": 3.0521,
+      "step": 43414
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7006272673606873,
+      "learning_rate": 0.0001167081133299389,
+      "loss": 2.8595,
+      "step": 43415
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7924833297729492,
+      "learning_rate": 0.00011669592515457318,
+      "loss": 2.7673,
+      "step": 43416
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.683752179145813,
+      "learning_rate": 0.00011668373746199508,
+      "loss": 2.8574,
+      "step": 43417
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7187227010726929,
+      "learning_rate": 0.00011667155025223674,
+      "loss": 2.9725,
+      "step": 43418
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6991389393806458,
+      "learning_rate": 0.00011665936352533013,
+      "loss": 3.0743,
+      "step": 43419
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7020309567451477,
+      "learning_rate": 0.00011664717728130751,
+      "loss": 2.9786,
+      "step": 43420
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6862619519233704,
+      "learning_rate": 0.00011663499152020101,
+      "loss": 2.9249,
+      "step": 43421
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6912420392036438,
+      "learning_rate": 0.00011662280624204258,
+      "loss": 2.787,
+      "step": 43422
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7102240920066833,
+      "learning_rate": 0.00011661062144686448,
+      "loss": 2.9573,
+      "step": 43423
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6816980838775635,
+      "learning_rate": 0.0001165984371346987,
+      "loss": 2.8612,
+      "step": 43424
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6912957429885864,
+      "learning_rate": 0.00011658625330557722,
+      "loss": 2.8478,
+      "step": 43425
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.690959095954895,
+      "learning_rate": 0.00011657406995953237,
+      "loss": 2.9166,
+      "step": 43426
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7246711254119873,
+      "learning_rate": 0.00011656188709659603,
+      "loss": 2.7915,
+      "step": 43427
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6712881326675415,
+      "learning_rate": 0.00011654970471680046,
+      "loss": 3.0455,
+      "step": 43428
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7149039506912231,
+      "learning_rate": 0.00011653752282017754,
+      "loss": 2.949,
+      "step": 43429
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7015359401702881,
+      "learning_rate": 0.00011652534140675951,
+      "loss": 2.9687,
+      "step": 43430
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.694905161857605,
+      "learning_rate": 0.00011651316047657853,
+      "loss": 2.9374,
+      "step": 43431
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6754507422447205,
+      "learning_rate": 0.00011650098002966656,
+      "loss": 3.0613,
+      "step": 43432
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.695451021194458,
+      "learning_rate": 0.00011648880006605567,
+      "loss": 3.0479,
+      "step": 43433
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7183724045753479,
+      "learning_rate": 0.00011647662058577789,
+      "loss": 2.8461,
+      "step": 43434
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6836041808128357,
+      "learning_rate": 0.00011646444158886538,
+      "loss": 2.9376,
+      "step": 43435
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6931686401367188,
+      "learning_rate": 0.00011645226307535027,
+      "loss": 3.1012,
+      "step": 43436
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.682621419429779,
+      "learning_rate": 0.00011644008504526452,
+      "loss": 3.041,
+      "step": 43437
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.685973584651947,
+      "learning_rate": 0.00011642790749864034,
+      "loss": 2.728,
+      "step": 43438
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.737464427947998,
+      "learning_rate": 0.0001164157304355097,
+      "loss": 2.8201,
+      "step": 43439
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6804065108299255,
+      "learning_rate": 0.00011640355385590462,
+      "loss": 2.7606,
+      "step": 43440
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7319862842559814,
+      "learning_rate": 0.00011639137775985736,
+      "loss": 2.9008,
+      "step": 43441
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6785258650779724,
+      "learning_rate": 0.00011637920214739974,
+      "loss": 2.857,
+      "step": 43442
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7257917523384094,
+      "learning_rate": 0.0001163670270185641,
+      "loss": 2.9225,
+      "step": 43443
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7215557098388672,
+      "learning_rate": 0.00011635485237338224,
+      "loss": 3.1464,
+      "step": 43444
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7050979733467102,
+      "learning_rate": 0.00011634267821188646,
+      "loss": 2.9606,
+      "step": 43445
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6802178025245667,
+      "learning_rate": 0.00011633050453410873,
+      "loss": 2.981,
+      "step": 43446
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7454336285591125,
+      "learning_rate": 0.00011631833134008099,
+      "loss": 2.9528,
+      "step": 43447
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6547221541404724,
+      "learning_rate": 0.00011630615862983551,
+      "loss": 2.689,
+      "step": 43448
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7083719372749329,
+      "learning_rate": 0.00011629398640340418,
+      "loss": 2.882,
+      "step": 43449
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.72013258934021,
+      "learning_rate": 0.00011628181466081912,
+      "loss": 2.7815,
+      "step": 43450
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6839480400085449,
+      "learning_rate": 0.00011626964340211249,
+      "loss": 2.9452,
+      "step": 43451
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7362170815467834,
+      "learning_rate": 0.00011625747262731617,
+      "loss": 2.8755,
+      "step": 43452
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6892048716545105,
+      "learning_rate": 0.00011624530233646241,
+      "loss": 3.0673,
+      "step": 43453
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7359662652015686,
+      "learning_rate": 0.00011623313252958317,
+      "loss": 3.0499,
+      "step": 43454
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7010781168937683,
+      "learning_rate": 0.00011622096320671038,
+      "loss": 2.8973,
+      "step": 43455
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6811627745628357,
+      "learning_rate": 0.00011620879436787631,
+      "loss": 2.8123,
+      "step": 43456
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7134902477264404,
+      "learning_rate": 0.0001161966260131128,
+      "loss": 2.972,
+      "step": 43457
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7200930714607239,
+      "learning_rate": 0.00011618445814245211,
+      "loss": 2.9097,
+      "step": 43458
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7086115479469299,
+      "learning_rate": 0.00011617229075592606,
+      "loss": 3.1063,
+      "step": 43459
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7229481935501099,
+      "learning_rate": 0.00011616012385356693,
+      "loss": 3.1506,
+      "step": 43460
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6957486271858215,
+      "learning_rate": 0.00011614795743540666,
+      "loss": 2.677,
+      "step": 43461
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7093002200126648,
+      "learning_rate": 0.00011613579150147718,
+      "loss": 2.6256,
+      "step": 43462
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7090989351272583,
+      "learning_rate": 0.00011612362605181073,
+      "loss": 2.9412,
+      "step": 43463
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7614361047744751,
+      "learning_rate": 0.0001161114610864392,
+      "loss": 2.6719,
+      "step": 43464
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7168323993682861,
+      "learning_rate": 0.00011609929660539467,
+      "loss": 2.8601,
+      "step": 43465
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7068440914154053,
+      "learning_rate": 0.0001160871326087093,
+      "loss": 2.883,
+      "step": 43466
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7084940671920776,
+      "learning_rate": 0.000116074969096415,
+      "loss": 2.763,
+      "step": 43467
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6470160484313965,
+      "learning_rate": 0.00011606280606854385,
+      "loss": 2.9537,
+      "step": 43468
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7235172986984253,
+      "learning_rate": 0.00011605064352512779,
+      "loss": 3.1622,
+      "step": 43469
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7104700207710266,
+      "learning_rate": 0.00011603848146619893,
+      "loss": 2.9741,
+      "step": 43470
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7352017164230347,
+      "learning_rate": 0.0001160263198917894,
+      "loss": 2.8902,
+      "step": 43471
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6944370269775391,
+      "learning_rate": 0.000116014158801931,
+      "loss": 2.685,
+      "step": 43472
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6649462580680847,
+      "learning_rate": 0.00011600199819665605,
+      "loss": 2.8065,
+      "step": 43473
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7043724060058594,
+      "learning_rate": 0.0001159898380759963,
+      "loss": 2.8261,
+      "step": 43474
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7225465178489685,
+      "learning_rate": 0.000115977678439984,
+      "loss": 3.0143,
+      "step": 43475
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7245696187019348,
+      "learning_rate": 0.00011596551928865109,
+      "loss": 2.8686,
+      "step": 43476
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7076426148414612,
+      "learning_rate": 0.0001159533606220295,
+      "loss": 2.9664,
+      "step": 43477
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7165865302085876,
+      "learning_rate": 0.00011594120244015143,
+      "loss": 2.9488,
+      "step": 43478
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7248859405517578,
+      "learning_rate": 0.00011592904474304867,
+      "loss": 2.8313,
+      "step": 43479
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6996978521347046,
+      "learning_rate": 0.00011591688753075351,
+      "loss": 2.9924,
+      "step": 43480
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7273635268211365,
+      "learning_rate": 0.00011590473080329775,
+      "loss": 3.0422,
+      "step": 43481
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6897189021110535,
+      "learning_rate": 0.0001158925745607136,
+      "loss": 2.9329,
+      "step": 43482
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.717872142791748,
+      "learning_rate": 0.00011588041880303297,
+      "loss": 3.0231,
+      "step": 43483
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.717965304851532,
+      "learning_rate": 0.00011586826353028777,
+      "loss": 2.9274,
+      "step": 43484
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7113906741142273,
+      "learning_rate": 0.00011585610874251022,
+      "loss": 2.8744,
+      "step": 43485
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6816174387931824,
+      "learning_rate": 0.00011584395443973216,
+      "loss": 3.0101,
+      "step": 43486
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6823785305023193,
+      "learning_rate": 0.00011583180062198567,
+      "loss": 2.9838,
+      "step": 43487
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7035296559333801,
+      "learning_rate": 0.00011581964728930291,
+      "loss": 2.6212,
+      "step": 43488
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.699919581413269,
+      "learning_rate": 0.0001158074944417157,
+      "loss": 2.8144,
+      "step": 43489
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7117913961410522,
+      "learning_rate": 0.00011579534207925614,
+      "loss": 2.9497,
+      "step": 43490
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7389628887176514,
+      "learning_rate": 0.00011578319020195607,
+      "loss": 2.7566,
+      "step": 43491
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6526266932487488,
+      "learning_rate": 0.00011577103880984764,
+      "loss": 2.7454,
+      "step": 43492
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6850768327713013,
+      "learning_rate": 0.00011575888790296294,
+      "loss": 2.7901,
+      "step": 43493
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.70717853307724,
+      "learning_rate": 0.00011574673748133377,
+      "loss": 2.8612,
+      "step": 43494
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6925551295280457,
+      "learning_rate": 0.00011573458754499231,
+      "loss": 2.8711,
+      "step": 43495
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6883793473243713,
+      "learning_rate": 0.00011572243809397042,
+      "loss": 2.872,
+      "step": 43496
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6649243235588074,
+      "learning_rate": 0.00011571028912830027,
+      "loss": 2.8072,
+      "step": 43497
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7801764011383057,
+      "learning_rate": 0.00011569814064801373,
+      "loss": 2.809,
+      "step": 43498
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6581512093544006,
+      "learning_rate": 0.0001156859926531427,
+      "loss": 3.2304,
+      "step": 43499
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6470869779586792,
+      "learning_rate": 0.0001156738451437194,
+      "loss": 3.0452,
+      "step": 43500
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6938720345497131,
+      "learning_rate": 0.00011566169811977564,
+      "loss": 3.0109,
+      "step": 43501
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7173318266868591,
+      "learning_rate": 0.00011564955158134347,
+      "loss": 2.6295,
+      "step": 43502
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6866092085838318,
+      "learning_rate": 0.000115637405528455,
+      "loss": 2.907,
+      "step": 43503
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6871329545974731,
+      "learning_rate": 0.00011562525996114216,
+      "loss": 3.1121,
+      "step": 43504
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7441400289535522,
+      "learning_rate": 0.00011561311487943686,
+      "loss": 2.8122,
+      "step": 43505
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7241978645324707,
+      "learning_rate": 0.00011560097028337107,
+      "loss": 2.8481,
+      "step": 43506
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7289275527000427,
+      "learning_rate": 0.0001155888261729768,
+      "loss": 2.8336,
+      "step": 43507
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6423391103744507,
+      "learning_rate": 0.00011557668254828618,
+      "loss": 2.8059,
+      "step": 43508
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7227512001991272,
+      "learning_rate": 0.00011556453940933098,
+      "loss": 2.8517,
+      "step": 43509
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6960424780845642,
+      "learning_rate": 0.00011555239675614337,
+      "loss": 2.8775,
+      "step": 43510
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7161725163459778,
+      "learning_rate": 0.0001155402545887553,
+      "loss": 2.7589,
+      "step": 43511
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7142117619514465,
+      "learning_rate": 0.00011552811290719856,
+      "loss": 3.1114,
+      "step": 43512
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7251697182655334,
+      "learning_rate": 0.00011551597171150539,
+      "loss": 2.9898,
+      "step": 43513
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6893921494483948,
+      "learning_rate": 0.00011550383100170752,
+      "loss": 2.9428,
+      "step": 43514
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6587451100349426,
+      "learning_rate": 0.0001154916907778372,
+      "loss": 2.6993,
+      "step": 43515
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6924602389335632,
+      "learning_rate": 0.00011547955103992611,
+      "loss": 2.8499,
+      "step": 43516
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6500839591026306,
+      "learning_rate": 0.00011546741178800637,
+      "loss": 3.2,
+      "step": 43517
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7085121273994446,
+      "learning_rate": 0.00011545527302211009,
+      "loss": 2.839,
+      "step": 43518
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6867219805717468,
+      "learning_rate": 0.0001154431347422691,
+      "loss": 2.7854,
+      "step": 43519
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7688367366790771,
+      "learning_rate": 0.00011543099694851534,
+      "loss": 2.9432,
+      "step": 43520
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6721521019935608,
+      "learning_rate": 0.00011541885964088074,
+      "loss": 2.9551,
+      "step": 43521
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6870953440666199,
+      "learning_rate": 0.00011540672281939731,
+      "loss": 2.9556,
+      "step": 43522
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7174733281135559,
+      "learning_rate": 0.00011539458648409718,
+      "loss": 2.8465,
+      "step": 43523
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6979454159736633,
+      "learning_rate": 0.00011538245063501205,
+      "loss": 2.8541,
+      "step": 43524
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7371723651885986,
+      "learning_rate": 0.00011537031527217417,
+      "loss": 3.0364,
+      "step": 43525
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6876466274261475,
+      "learning_rate": 0.0001153581803956153,
+      "loss": 2.9516,
+      "step": 43526
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7251150608062744,
+      "learning_rate": 0.00011534604600536737,
+      "loss": 2.7975,
+      "step": 43527
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7008710503578186,
+      "learning_rate": 0.00011533391210146249,
+      "loss": 2.9521,
+      "step": 43528
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.690662145614624,
+      "learning_rate": 0.00011532177868393248,
+      "loss": 2.7882,
+      "step": 43529
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6789386868476868,
+      "learning_rate": 0.00011530964575280944,
+      "loss": 3.1341,
+      "step": 43530
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6649258732795715,
+      "learning_rate": 0.00011529751330812516,
+      "loss": 2.8537,
+      "step": 43531
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7068399786949158,
+      "learning_rate": 0.00011528538134991167,
+      "loss": 2.5939,
+      "step": 43532
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7076483964920044,
+      "learning_rate": 0.00011527324987820104,
+      "loss": 3.0746,
+      "step": 43533
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6737373471260071,
+      "learning_rate": 0.00011526111889302511,
+      "loss": 3.0245,
+      "step": 43534
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7310186624526978,
+      "learning_rate": 0.00011524898839441589,
+      "loss": 2.9546,
+      "step": 43535
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6874125003814697,
+      "learning_rate": 0.00011523685838240512,
+      "loss": 2.9512,
+      "step": 43536
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7353755831718445,
+      "learning_rate": 0.00011522472885702491,
+      "loss": 2.8711,
+      "step": 43537
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6856021285057068,
+      "learning_rate": 0.00011521259981830732,
+      "loss": 2.8758,
+      "step": 43538
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7210877537727356,
+      "learning_rate": 0.00011520047126628409,
+      "loss": 3.0204,
+      "step": 43539
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6792193055152893,
+      "learning_rate": 0.00011518834320098735,
+      "loss": 2.9862,
+      "step": 43540
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.67461097240448,
+      "learning_rate": 0.00011517621562244893,
+      "loss": 2.9252,
+      "step": 43541
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6909298896789551,
+      "learning_rate": 0.0001151640885307007,
+      "loss": 2.9947,
+      "step": 43542
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.695124626159668,
+      "learning_rate": 0.0001151519619257748,
+      "loss": 2.9083,
+      "step": 43543
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6613640189170837,
+      "learning_rate": 0.00011513983580770294,
+      "loss": 2.9106,
+      "step": 43544
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6796993017196655,
+      "learning_rate": 0.00011512771017651727,
+      "loss": 2.9502,
+      "step": 43545
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6530843377113342,
+      "learning_rate": 0.00011511558503224954,
+      "loss": 2.8476,
+      "step": 43546
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7151798605918884,
+      "learning_rate": 0.0001151034603749319,
+      "loss": 2.8159,
+      "step": 43547
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7040915489196777,
+      "learning_rate": 0.00011509133620459615,
+      "loss": 3.0598,
+      "step": 43548
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7070721983909607,
+      "learning_rate": 0.00011507921252127411,
+      "loss": 2.9802,
+      "step": 43549
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6628305315971375,
+      "learning_rate": 0.00011506708932499795,
+      "loss": 3.0236,
+      "step": 43550
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6968042850494385,
+      "learning_rate": 0.00011505496661579937,
+      "loss": 2.8769,
+      "step": 43551
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7575791478157043,
+      "learning_rate": 0.00011504284439371044,
+      "loss": 2.7383,
+      "step": 43552
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6684389114379883,
+      "learning_rate": 0.00011503072265876314,
+      "loss": 3.0318,
+      "step": 43553
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.67277991771698,
+      "learning_rate": 0.00011501860141098924,
+      "loss": 2.9531,
+      "step": 43554
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6650310754776001,
+      "learning_rate": 0.00011500648065042082,
+      "loss": 3.0134,
+      "step": 43555
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7093574404716492,
+      "learning_rate": 0.00011499436037708975,
+      "loss": 2.7774,
+      "step": 43556
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.701271653175354,
+      "learning_rate": 0.00011498224059102793,
+      "loss": 2.8878,
+      "step": 43557
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6936562657356262,
+      "learning_rate": 0.00011497012129226714,
+      "loss": 2.7935,
+      "step": 43558
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6467651724815369,
+      "learning_rate": 0.00011495800248083947,
+      "loss": 2.8307,
+      "step": 43559
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6694419384002686,
+      "learning_rate": 0.00011494588415677694,
+      "loss": 2.8483,
+      "step": 43560
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7048700451850891,
+      "learning_rate": 0.00011493376632011121,
+      "loss": 2.719,
+      "step": 43561
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6826921701431274,
+      "learning_rate": 0.0001149216489708744,
+      "loss": 2.984,
+      "step": 43562
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7334887385368347,
+      "learning_rate": 0.00011490953210909837,
+      "loss": 3.0855,
+      "step": 43563
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7349691987037659,
+      "learning_rate": 0.00011489741573481491,
+      "loss": 2.7208,
+      "step": 43564
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7117055058479309,
+      "learning_rate": 0.00011488529984805614,
+      "loss": 2.8431,
+      "step": 43565
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7462189197540283,
+      "learning_rate": 0.00011487318444885377,
+      "loss": 3.0923,
+      "step": 43566
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7095943689346313,
+      "learning_rate": 0.00011486106953723987,
+      "loss": 3.0918,
+      "step": 43567
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7356390953063965,
+      "learning_rate": 0.0001148489551132462,
+      "loss": 2.9536,
+      "step": 43568
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6879743337631226,
+      "learning_rate": 0.00011483684117690487,
+      "loss": 2.749,
+      "step": 43569
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6889345645904541,
+      "learning_rate": 0.00011482472772824765,
+      "loss": 2.9549,
+      "step": 43570
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7350010871887207,
+      "learning_rate": 0.00011481261476730634,
+      "loss": 2.9164,
+      "step": 43571
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7565581202507019,
+      "learning_rate": 0.00011480050229411307,
+      "loss": 2.873,
+      "step": 43572
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7308693528175354,
+      "learning_rate": 0.00011478839030869956,
+      "loss": 2.8584,
+      "step": 43573
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7340457439422607,
+      "learning_rate": 0.00011477627881109776,
+      "loss": 3.0429,
+      "step": 43574
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7506102323532104,
+      "learning_rate": 0.00011476416780133972,
+      "loss": 3.0489,
+      "step": 43575
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7139230966567993,
+      "learning_rate": 0.0001147520572794571,
+      "loss": 2.948,
+      "step": 43576
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7269354462623596,
+      "learning_rate": 0.00011473994724548199,
+      "loss": 2.8071,
+      "step": 43577
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7375877499580383,
+      "learning_rate": 0.00011472783769944623,
+      "loss": 2.8066,
+      "step": 43578
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.701732337474823,
+      "learning_rate": 0.00011471572864138158,
+      "loss": 2.7454,
+      "step": 43579
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6951040029525757,
+      "learning_rate": 0.00011470362007132013,
+      "loss": 3.1121,
+      "step": 43580
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.745621919631958,
+      "learning_rate": 0.00011469151198929362,
+      "loss": 2.6036,
+      "step": 43581
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7371377944946289,
+      "learning_rate": 0.00011467940439533406,
+      "loss": 2.7472,
+      "step": 43582
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7030379176139832,
+      "learning_rate": 0.00011466729728947321,
+      "loss": 2.8501,
+      "step": 43583
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6986172795295715,
+      "learning_rate": 0.00011465519067174312,
+      "loss": 2.8457,
+      "step": 43584
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.673969566822052,
+      "learning_rate": 0.0001146430845421756,
+      "loss": 2.8234,
+      "step": 43585
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6397623419761658,
+      "learning_rate": 0.00011463097890080241,
+      "loss": 3.3042,
+      "step": 43586
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7095158696174622,
+      "learning_rate": 0.00011461887374765564,
+      "loss": 2.8524,
+      "step": 43587
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6966813206672668,
+      "learning_rate": 0.000114606769082767,
+      "loss": 3.1101,
+      "step": 43588
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7600648999214172,
+      "learning_rate": 0.00011459466490616845,
+      "loss": 2.838,
+      "step": 43589
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7137221693992615,
+      "learning_rate": 0.00011458256121789194,
+      "loss": 2.8731,
+      "step": 43590
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6784641742706299,
+      "learning_rate": 0.00011457045801796932,
+      "loss": 3.0268,
+      "step": 43591
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7136608362197876,
+      "learning_rate": 0.0001145583553064324,
+      "loss": 2.9876,
+      "step": 43592
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7022787928581238,
+      "learning_rate": 0.00011454625308331298,
+      "loss": 2.7072,
+      "step": 43593
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6970979571342468,
+      "learning_rate": 0.00011453415134864305,
+      "loss": 3.2014,
+      "step": 43594
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6926916241645813,
+      "learning_rate": 0.00011452205010245456,
+      "loss": 2.9012,
+      "step": 43595
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6950446367263794,
+      "learning_rate": 0.00011450994934477922,
+      "loss": 3.2737,
+      "step": 43596
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6955012083053589,
+      "learning_rate": 0.00011449784907564908,
+      "loss": 2.9293,
+      "step": 43597
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6871076226234436,
+      "learning_rate": 0.00011448574929509578,
+      "loss": 3.1272,
+      "step": 43598
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7409434914588928,
+      "learning_rate": 0.0001144736500031514,
+      "loss": 2.9277,
+      "step": 43599
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7080227732658386,
+      "learning_rate": 0.00011446155119984774,
+      "loss": 3.0437,
+      "step": 43600
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7121559977531433,
+      "learning_rate": 0.00011444945288521653,
+      "loss": 2.964,
+      "step": 43601
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7132676839828491,
+      "learning_rate": 0.00011443735505928989,
+      "loss": 2.9648,
+      "step": 43602
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7137494683265686,
+      "learning_rate": 0.0001144252577220994,
+      "loss": 2.8898,
+      "step": 43603
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7248656153678894,
+      "learning_rate": 0.00011441316087367711,
+      "loss": 2.808,
+      "step": 43604
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6689263582229614,
+      "learning_rate": 0.00011440106451405492,
+      "loss": 2.6834,
+      "step": 43605
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6864780783653259,
+      "learning_rate": 0.0001143889686432646,
+      "loss": 2.8517,
+      "step": 43606
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7175418734550476,
+      "learning_rate": 0.000114376873261338,
+      "loss": 3.1024,
+      "step": 43607
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.723504900932312,
+      "learning_rate": 0.00011436477836830692,
+      "loss": 2.971,
+      "step": 43608
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7824511528015137,
+      "learning_rate": 0.00011435268396420327,
+      "loss": 2.944,
+      "step": 43609
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7112360596656799,
+      "learning_rate": 0.00011434059004905903,
+      "loss": 2.7707,
+      "step": 43610
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7221190929412842,
+      "learning_rate": 0.00011432849662290584,
+      "loss": 3.0907,
+      "step": 43611
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6884748339653015,
+      "learning_rate": 0.00011431640368577577,
+      "loss": 2.9186,
+      "step": 43612
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7097779512405396,
+      "learning_rate": 0.00011430431123770052,
+      "loss": 3.1311,
+      "step": 43613
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6831579804420471,
+      "learning_rate": 0.00011429221927871192,
+      "loss": 2.7814,
+      "step": 43614
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7085390686988831,
+      "learning_rate": 0.00011428012780884197,
+      "loss": 2.8956,
+      "step": 43615
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7171388268470764,
+      "learning_rate": 0.0001142680368281223,
+      "loss": 3.1167,
+      "step": 43616
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.679538905620575,
+      "learning_rate": 0.00011425594633658499,
+      "loss": 3.0024,
+      "step": 43617
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7115911245346069,
+      "learning_rate": 0.00011424385633426165,
+      "loss": 2.8164,
+      "step": 43618
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7108883857727051,
+      "learning_rate": 0.00011423176682118425,
+      "loss": 2.9752,
+      "step": 43619
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6663087010383606,
+      "learning_rate": 0.00011421967779738475,
+      "loss": 2.8533,
+      "step": 43620
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6840021014213562,
+      "learning_rate": 0.00011420758926289487,
+      "loss": 3.0553,
+      "step": 43621
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7471802234649658,
+      "learning_rate": 0.00011419550121774641,
+      "loss": 2.944,
+      "step": 43622
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6915409564971924,
+      "learning_rate": 0.00011418341366197113,
+      "loss": 2.7266,
+      "step": 43623
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6648789048194885,
+      "learning_rate": 0.00011417132659560102,
+      "loss": 2.988,
+      "step": 43624
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7293175458908081,
+      "learning_rate": 0.00011415924001866795,
+      "loss": 2.966,
+      "step": 43625
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6726661324501038,
+      "learning_rate": 0.00011414715393120358,
+      "loss": 3.2105,
+      "step": 43626
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6855407357215881,
+      "learning_rate": 0.00011413506833323991,
+      "loss": 2.7334,
+      "step": 43627
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7026200890541077,
+      "learning_rate": 0.00011412298322480873,
+      "loss": 2.9252,
+      "step": 43628
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7490710020065308,
+      "learning_rate": 0.00011411089860594174,
+      "loss": 2.815,
+      "step": 43629
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6966449022293091,
+      "learning_rate": 0.00011409881447667098,
+      "loss": 2.8681,
+      "step": 43630
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7190315127372742,
+      "learning_rate": 0.00011408673083702801,
+      "loss": 3.0971,
+      "step": 43631
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6999360918998718,
+      "learning_rate": 0.00011407464768704497,
+      "loss": 3.0055,
+      "step": 43632
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7241888046264648,
+      "learning_rate": 0.00011406256502675338,
+      "loss": 2.8215,
+      "step": 43633
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6731426119804382,
+      "learning_rate": 0.00011405048285618535,
+      "loss": 3.0833,
+      "step": 43634
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6741295456886292,
+      "learning_rate": 0.00011403840117537243,
+      "loss": 2.9559,
+      "step": 43635
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6998021602630615,
+      "learning_rate": 0.00011402631998434671,
+      "loss": 2.9883,
+      "step": 43636
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7177157402038574,
+      "learning_rate": 0.00011401423928313985,
+      "loss": 2.9022,
+      "step": 43637
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.764556348323822,
+      "learning_rate": 0.00011400215907178359,
+      "loss": 3.133,
+      "step": 43638
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7035207748413086,
+      "learning_rate": 0.00011399007935030994,
+      "loss": 3.065,
+      "step": 43639
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6932657361030579,
+      "learning_rate": 0.00011397800011875052,
+      "loss": 2.9998,
+      "step": 43640
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7348163723945618,
+      "learning_rate": 0.00011396592137713725,
+      "loss": 2.9956,
+      "step": 43641
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6634372472763062,
+      "learning_rate": 0.00011395384312550207,
+      "loss": 2.6726,
+      "step": 43642
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.667530357837677,
+      "learning_rate": 0.00011394176536387664,
+      "loss": 2.8978,
+      "step": 43643
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6972176432609558,
+      "learning_rate": 0.00011392968809229277,
+      "loss": 3.0545,
+      "step": 43644
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7097292542457581,
+      "learning_rate": 0.00011391761131078224,
+      "loss": 2.9076,
+      "step": 43645
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.66300368309021,
+      "learning_rate": 0.00011390553501937687,
+      "loss": 2.765,
+      "step": 43646
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6970599889755249,
+      "learning_rate": 0.00011389345921810862,
+      "loss": 2.7709,
+      "step": 43647
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6801822781562805,
+      "learning_rate": 0.00011388138390700906,
+      "loss": 2.7156,
+      "step": 43648
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6642677187919617,
+      "learning_rate": 0.00011386930908611024,
+      "loss": 2.9296,
+      "step": 43649
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.726485550403595,
+      "learning_rate": 0.00011385723475544383,
+      "loss": 2.9243,
+      "step": 43650
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6999016404151917,
+      "learning_rate": 0.00011384516091504156,
+      "loss": 2.9357,
+      "step": 43651
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6688977479934692,
+      "learning_rate": 0.0001138330875649354,
+      "loss": 2.9612,
+      "step": 43652
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6783860921859741,
+      "learning_rate": 0.00011382101470515694,
+      "loss": 2.9645,
+      "step": 43653
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6920573711395264,
+      "learning_rate": 0.0001138089423357382,
+      "loss": 2.8786,
+      "step": 43654
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6857960820198059,
+      "learning_rate": 0.0001137968704567108,
+      "loss": 2.9031,
+      "step": 43655
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7088309526443481,
+      "learning_rate": 0.00011378479906810658,
+      "loss": 2.9249,
+      "step": 43656
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6725496649742126,
+      "learning_rate": 0.00011377272816995746,
+      "loss": 2.9255,
+      "step": 43657
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7024503350257874,
+      "learning_rate": 0.00011376065776229515,
+      "loss": 3.0291,
+      "step": 43658
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7618658542633057,
+      "learning_rate": 0.00011374858784515143,
+      "loss": 2.8862,
+      "step": 43659
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7083359956741333,
+      "learning_rate": 0.00011373651841855797,
+      "loss": 2.9298,
+      "step": 43660
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7162266373634338,
+      "learning_rate": 0.00011372444948254667,
+      "loss": 3.0221,
+      "step": 43661
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6709370613098145,
+      "learning_rate": 0.00011371238103714944,
+      "loss": 2.74,
+      "step": 43662
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6915350556373596,
+      "learning_rate": 0.00011370031308239783,
+      "loss": 2.9314,
+      "step": 43663
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7247147560119629,
+      "learning_rate": 0.00011368824561832383,
+      "loss": 3.0371,
+      "step": 43664
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.8002133965492249,
+      "learning_rate": 0.00011367617864495916,
+      "loss": 3.008,
+      "step": 43665
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6851048469543457,
+      "learning_rate": 0.00011366411216233548,
+      "loss": 2.9939,
+      "step": 43666
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7082476019859314,
+      "learning_rate": 0.00011365204617048474,
+      "loss": 3.0785,
+      "step": 43667
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.668368399143219,
+      "learning_rate": 0.00011363998066943857,
+      "loss": 2.8655,
+      "step": 43668
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6958811283111572,
+      "learning_rate": 0.00011362791565922892,
+      "loss": 2.7918,
+      "step": 43669
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7040616273880005,
+      "learning_rate": 0.00011361585113988737,
+      "loss": 3.1085,
+      "step": 43670
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7405644655227661,
+      "learning_rate": 0.00011360378711144588,
+      "loss": 2.7158,
+      "step": 43671
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.66157066822052,
+      "learning_rate": 0.00011359172357393618,
+      "loss": 2.8066,
+      "step": 43672
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.8532086610794067,
+      "learning_rate": 0.00011357966052738988,
+      "loss": 2.8688,
+      "step": 43673
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7060077786445618,
+      "learning_rate": 0.00011356759797183896,
+      "loss": 2.921,
+      "step": 43674
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.71628737449646,
+      "learning_rate": 0.00011355553590731503,
+      "loss": 3.0828,
+      "step": 43675
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6801706552505493,
+      "learning_rate": 0.00011354347433384993,
+      "loss": 2.9732,
+      "step": 43676
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6853070259094238,
+      "learning_rate": 0.00011353141325147555,
+      "loss": 2.8799,
+      "step": 43677
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6793110966682434,
+      "learning_rate": 0.00011351935266022343,
+      "loss": 2.8989,
+      "step": 43678
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6548277735710144,
+      "learning_rate": 0.00011350729256012555,
+      "loss": 2.7669,
+      "step": 43679
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7266643047332764,
+      "learning_rate": 0.00011349523295121356,
+      "loss": 2.8731,
+      "step": 43680
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7011723518371582,
+      "learning_rate": 0.00011348317383351915,
+      "loss": 3.0052,
+      "step": 43681
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6877959370613098,
+      "learning_rate": 0.00011347111520707424,
+      "loss": 3.0605,
+      "step": 43682
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7112438678741455,
+      "learning_rate": 0.00011345905707191044,
+      "loss": 3.0148,
+      "step": 43683
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7566985487937927,
+      "learning_rate": 0.00011344699942805966,
+      "loss": 3.1465,
+      "step": 43684
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7064225673675537,
+      "learning_rate": 0.0001134349422755535,
+      "loss": 3.0024,
+      "step": 43685
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6750404238700867,
+      "learning_rate": 0.00011342288561442386,
+      "loss": 2.8038,
+      "step": 43686
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6803438663482666,
+      "learning_rate": 0.00011341082944470248,
+      "loss": 2.9661,
+      "step": 43687
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6766833066940308,
+      "learning_rate": 0.00011339877376642095,
+      "loss": 3.184,
+      "step": 43688
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.70738685131073,
+      "learning_rate": 0.00011338671857961125,
+      "loss": 2.949,
+      "step": 43689
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6981825828552246,
+      "learning_rate": 0.00011337466388430489,
+      "loss": 2.8143,
+      "step": 43690
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6753236651420593,
+      "learning_rate": 0.00011336260968053374,
+      "loss": 3.0315,
+      "step": 43691
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7474436163902283,
+      "learning_rate": 0.0001133505559683297,
+      "loss": 2.7655,
+      "step": 43692
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.717900812625885,
+      "learning_rate": 0.00011333850274772433,
+      "loss": 2.8381,
+      "step": 43693
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6985374093055725,
+      "learning_rate": 0.00011332645001874947,
+      "loss": 2.7779,
+      "step": 43694
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6623815298080444,
+      "learning_rate": 0.00011331439778143669,
+      "loss": 2.7031,
+      "step": 43695
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7060630321502686,
+      "learning_rate": 0.00011330234603581782,
+      "loss": 3.1725,
+      "step": 43696
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6988420486450195,
+      "learning_rate": 0.00011329029478192478,
+      "loss": 2.9677,
+      "step": 43697
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7007870078086853,
+      "learning_rate": 0.00011327824401978907,
+      "loss": 2.9562,
+      "step": 43698
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6870179176330566,
+      "learning_rate": 0.00011326619374944263,
+      "loss": 3.0245,
+      "step": 43699
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7430259585380554,
+      "learning_rate": 0.00011325414397091699,
+      "loss": 2.8442,
+      "step": 43700
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7168817520141602,
+      "learning_rate": 0.0001132420946842441,
+      "loss": 2.5137,
+      "step": 43701
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7056675553321838,
+      "learning_rate": 0.00011323004588945558,
+      "loss": 2.8598,
+      "step": 43702
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6785736680030823,
+      "learning_rate": 0.0001132179975865831,
+      "loss": 2.6708,
+      "step": 43703
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6528010368347168,
+      "learning_rate": 0.00011320594977565854,
+      "loss": 2.8608,
+      "step": 43704
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7285547256469727,
+      "learning_rate": 0.00011319390245671346,
+      "loss": 2.785,
+      "step": 43705
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6859037280082703,
+      "learning_rate": 0.00011318185562977969,
+      "loss": 2.9215,
+      "step": 43706
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.683616578578949,
+      "learning_rate": 0.00011316980929488908,
+      "loss": 2.921,
+      "step": 43707
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7073354125022888,
+      "learning_rate": 0.00011315776345207322,
+      "loss": 3.038,
+      "step": 43708
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6951118111610413,
+      "learning_rate": 0.00011314571810136386,
+      "loss": 3.0371,
+      "step": 43709
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7080778479576111,
+      "learning_rate": 0.00011313367324279263,
+      "loss": 3.0615,
+      "step": 43710
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.712817907333374,
+      "learning_rate": 0.0001131216288763914,
+      "loss": 2.7927,
+      "step": 43711
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7057470679283142,
+      "learning_rate": 0.00011310958500219177,
+      "loss": 3.0071,
+      "step": 43712
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7209771275520325,
+      "learning_rate": 0.00011309754162022551,
+      "loss": 2.9617,
+      "step": 43713
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6734073162078857,
+      "learning_rate": 0.00011308549873052447,
+      "loss": 2.7933,
+      "step": 43714
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6648086905479431,
+      "learning_rate": 0.00011307345633312024,
+      "loss": 2.9428,
+      "step": 43715
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6905327439308167,
+      "learning_rate": 0.00011306141442804456,
+      "loss": 2.8132,
+      "step": 43716
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7069293856620789,
+      "learning_rate": 0.00011304937301532901,
+      "loss": 2.7406,
+      "step": 43717
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6863256096839905,
+      "learning_rate": 0.00011303733209500548,
+      "loss": 2.9026,
+      "step": 43718
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6748086810112,
+      "learning_rate": 0.0001130252916671057,
+      "loss": 2.9089,
+      "step": 43719
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.68429034948349,
+      "learning_rate": 0.0001130132517316612,
+      "loss": 2.9713,
+      "step": 43720
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6415186524391174,
+      "learning_rate": 0.00011300121228870396,
+      "loss": 2.6823,
+      "step": 43721
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6968873143196106,
+      "learning_rate": 0.0001129891733382654,
+      "loss": 2.8784,
+      "step": 43722
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7179439067840576,
+      "learning_rate": 0.00011297713488037748,
+      "loss": 2.8995,
+      "step": 43723
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7167483568191528,
+      "learning_rate": 0.0001129650969150718,
+      "loss": 2.8282,
+      "step": 43724
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7356143593788147,
+      "learning_rate": 0.00011295305944237995,
+      "loss": 2.6649,
+      "step": 43725
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6669628024101257,
+      "learning_rate": 0.00011294102246233385,
+      "loss": 2.8961,
+      "step": 43726
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7157459855079651,
+      "learning_rate": 0.000112928985974965,
+      "loss": 2.9598,
+      "step": 43727
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6857950091362,
+      "learning_rate": 0.00011291694998030517,
+      "loss": 2.9189,
+      "step": 43728
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7263401746749878,
+      "learning_rate": 0.00011290491447838622,
+      "loss": 2.8394,
+      "step": 43729
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7011303901672363,
+      "learning_rate": 0.00011289287946923973,
+      "loss": 2.8701,
+      "step": 43730
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7553688287734985,
+      "learning_rate": 0.00011288084495289737,
+      "loss": 2.8729,
+      "step": 43731
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6972824931144714,
+      "learning_rate": 0.00011286881092939074,
+      "loss": 2.9316,
+      "step": 43732
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7040778398513794,
+      "learning_rate": 0.00011285677739875163,
+      "loss": 2.7414,
+      "step": 43733
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6931003332138062,
+      "learning_rate": 0.00011284474436101188,
+      "loss": 2.7205,
+      "step": 43734
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6897353529930115,
+      "learning_rate": 0.00011283271181620295,
+      "loss": 2.8586,
+      "step": 43735
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.716933012008667,
+      "learning_rate": 0.00011282067976435676,
+      "loss": 3.0455,
+      "step": 43736
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7096711993217468,
+      "learning_rate": 0.00011280864820550474,
+      "loss": 2.9453,
+      "step": 43737
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7447996139526367,
+      "learning_rate": 0.00011279661713967882,
+      "loss": 2.9867,
+      "step": 43738
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6898730993270874,
+      "learning_rate": 0.0001127845865669106,
+      "loss": 2.8042,
+      "step": 43739
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6772977709770203,
+      "learning_rate": 0.00011277255648723163,
+      "loss": 2.9059,
+      "step": 43740
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7092178463935852,
+      "learning_rate": 0.00011276052690067382,
+      "loss": 2.8711,
+      "step": 43741
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6839644312858582,
+      "learning_rate": 0.00011274849780726866,
+      "loss": 2.8759,
+      "step": 43742
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7505874037742615,
+      "learning_rate": 0.00011273646920704792,
+      "loss": 3.0373,
+      "step": 43743
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7028711438179016,
+      "learning_rate": 0.00011272444110004335,
+      "loss": 2.7585,
+      "step": 43744
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7121570706367493,
+      "learning_rate": 0.00011271241348628659,
+      "loss": 2.9666,
+      "step": 43745
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7014894485473633,
+      "learning_rate": 0.00011270038636580928,
+      "loss": 2.9703,
+      "step": 43746
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7833147644996643,
+      "learning_rate": 0.000112688359738643,
+      "loss": 2.7485,
+      "step": 43747
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6829766631126404,
+      "learning_rate": 0.00011267633360481954,
+      "loss": 3.0008,
+      "step": 43748
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7053651809692383,
+      "learning_rate": 0.00011266430796437067,
+      "loss": 3.0635,
+      "step": 43749
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7544901371002197,
+      "learning_rate": 0.00011265228281732784,
+      "loss": 2.9556,
+      "step": 43750
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6905205845832825,
+      "learning_rate": 0.00011264025816372298,
+      "loss": 3.0138,
+      "step": 43751
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7467920780181885,
+      "learning_rate": 0.00011262823400358756,
+      "loss": 2.7328,
+      "step": 43752
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6714316606521606,
+      "learning_rate": 0.00011261621033695327,
+      "loss": 2.9427,
+      "step": 43753
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6609356999397278,
+      "learning_rate": 0.00011260418716385189,
+      "loss": 2.8976,
+      "step": 43754
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.728022038936615,
+      "learning_rate": 0.00011259216448431493,
+      "loss": 2.8679,
+      "step": 43755
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7265163660049438,
+      "learning_rate": 0.00011258014229837424,
+      "loss": 2.9928,
+      "step": 43756
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6698508262634277,
+      "learning_rate": 0.00011256812060606127,
+      "loss": 2.8256,
+      "step": 43757
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7141711115837097,
+      "learning_rate": 0.0001125560994074078,
+      "loss": 2.6912,
+      "step": 43758
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.705557644367218,
+      "learning_rate": 0.0001125440787024456,
+      "loss": 2.8845,
+      "step": 43759
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7180753946304321,
+      "learning_rate": 0.00011253205849120621,
+      "loss": 2.883,
+      "step": 43760
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6521044969558716,
+      "learning_rate": 0.00011252003877372132,
+      "loss": 2.9457,
+      "step": 43761
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7582170367240906,
+      "learning_rate": 0.00011250801955002245,
+      "loss": 3.0808,
+      "step": 43762
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6841397881507874,
+      "learning_rate": 0.00011249600082014135,
+      "loss": 2.8299,
+      "step": 43763
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6913440227508545,
+      "learning_rate": 0.00011248398258410986,
+      "loss": 2.8295,
+      "step": 43764
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6824356317520142,
+      "learning_rate": 0.00011247196484195933,
+      "loss": 2.8794,
+      "step": 43765
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7842134237289429,
+      "learning_rate": 0.00011245994759372164,
+      "loss": 2.8076,
+      "step": 43766
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7321650981903076,
+      "learning_rate": 0.00011244793083942838,
+      "loss": 3.0722,
+      "step": 43767
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.7260737419128418,
+      "learning_rate": 0.00011243591457911107,
+      "loss": 2.994,
+      "step": 43768
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7154833674430847,
+      "learning_rate": 0.00011242389881280158,
+      "loss": 2.985,
+      "step": 43769
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6974570155143738,
+      "learning_rate": 0.00011241188354053131,
+      "loss": 2.9196,
+      "step": 43770
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7746239900588989,
+      "learning_rate": 0.00011239986876233214,
+      "loss": 2.7981,
+      "step": 43771
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6849657297134399,
+      "learning_rate": 0.00011238785447823552,
+      "loss": 2.9164,
+      "step": 43772
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7732822895050049,
+      "learning_rate": 0.0001123758406882733,
+      "loss": 2.8262,
+      "step": 43773
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7015544176101685,
+      "learning_rate": 0.000112363827392477,
+      "loss": 2.9393,
+      "step": 43774
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6607134342193604,
+      "learning_rate": 0.00011235181459087816,
+      "loss": 2.9479,
+      "step": 43775
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7178341150283813,
+      "learning_rate": 0.00011233980228350863,
+      "loss": 2.9219,
+      "step": 43776
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6694148778915405,
+      "learning_rate": 0.00011232779047039983,
+      "loss": 2.7359,
+      "step": 43777
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7360499501228333,
+      "learning_rate": 0.00011231577915158353,
+      "loss": 2.9348,
+      "step": 43778
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.672639787197113,
+      "learning_rate": 0.00011230376832709144,
+      "loss": 2.846,
+      "step": 43779
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7251139879226685,
+      "learning_rate": 0.00011229175799695498,
+      "loss": 3.0,
+      "step": 43780
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6625097393989563,
+      "learning_rate": 0.00011227974816120603,
+      "loss": 2.9168,
+      "step": 43781
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6882855296134949,
+      "learning_rate": 0.0001122677388198761,
+      "loss": 3.1654,
+      "step": 43782
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6938197612762451,
+      "learning_rate": 0.0001122557299729967,
+      "loss": 2.8398,
+      "step": 43783
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.701835572719574,
+      "learning_rate": 0.00011224372162059972,
+      "loss": 3.03,
+      "step": 43784
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6981818675994873,
+      "learning_rate": 0.0001122317137627165,
+      "loss": 2.9271,
+      "step": 43785
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7233262658119202,
+      "learning_rate": 0.00011221970639937891,
+      "loss": 2.7305,
+      "step": 43786
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7583865523338318,
+      "learning_rate": 0.00011220769953061837,
+      "loss": 2.9369,
+      "step": 43787
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6952337622642517,
+      "learning_rate": 0.00011219569315646672,
+      "loss": 2.8671,
+      "step": 43788
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6880810260772705,
+      "learning_rate": 0.00011218368727695546,
+      "loss": 2.9465,
+      "step": 43789
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6928128600120544,
+      "learning_rate": 0.00011217168189211614,
+      "loss": 2.7894,
+      "step": 43790
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6935843825340271,
+      "learning_rate": 0.00011215967700198057,
+      "loss": 2.9268,
+      "step": 43791
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7136497497558594,
+      "learning_rate": 0.00011214767260658011,
+      "loss": 3.0224,
+      "step": 43792
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7205958962440491,
+      "learning_rate": 0.00011213566870594668,
+      "loss": 2.8294,
+      "step": 43793
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.714300811290741,
+      "learning_rate": 0.00011212366530011162,
+      "loss": 3.0795,
+      "step": 43794
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6874080300331116,
+      "learning_rate": 0.00011211166238910677,
+      "loss": 3.0615,
+      "step": 43795
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7105938196182251,
+      "learning_rate": 0.00011209965997296363,
+      "loss": 2.9694,
+      "step": 43796
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7178138494491577,
+      "learning_rate": 0.00011208765805171373,
+      "loss": 3.0575,
+      "step": 43797
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7538326978683472,
+      "learning_rate": 0.0001120756566253889,
+      "loss": 3.0899,
+      "step": 43798
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6940291523933411,
+      "learning_rate": 0.00011206365569402051,
+      "loss": 2.9107,
+      "step": 43799
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.754417896270752,
+      "learning_rate": 0.00011205165525764026,
+      "loss": 2.9723,
+      "step": 43800
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7753744125366211,
+      "learning_rate": 0.00011203965531627991,
+      "loss": 2.9622,
+      "step": 43801
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7098350524902344,
+      "learning_rate": 0.00011202765586997082,
+      "loss": 2.8377,
+      "step": 43802
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7141963839530945,
+      "learning_rate": 0.00011201565691874482,
+      "loss": 2.887,
+      "step": 43803
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6638633608818054,
+      "learning_rate": 0.00011200365846263338,
+      "loss": 2.9423,
+      "step": 43804
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6971076726913452,
+      "learning_rate": 0.00011199166050166801,
+      "loss": 2.8963,
+      "step": 43805
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6292498111724854,
+      "learning_rate": 0.00011197966303588055,
+      "loss": 3.1008,
+      "step": 43806
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.696465790271759,
+      "learning_rate": 0.00011196766606530238,
+      "loss": 2.8679,
+      "step": 43807
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6825982332229614,
+      "learning_rate": 0.00011195566958996526,
+      "loss": 3.0,
+      "step": 43808
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7092751264572144,
+      "learning_rate": 0.00011194367360990063,
+      "loss": 3.0249,
+      "step": 43809
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7213112115859985,
+      "learning_rate": 0.0001119316781251403,
+      "loss": 2.9623,
+      "step": 43810
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6883493065834045,
+      "learning_rate": 0.0001119196831357157,
+      "loss": 3.0142,
+      "step": 43811
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6896858215332031,
+      "learning_rate": 0.00011190768864165837,
+      "loss": 2.941,
+      "step": 43812
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7189692258834839,
+      "learning_rate": 0.00011189569464300007,
+      "loss": 2.9459,
+      "step": 43813
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.67086261510849,
+      "learning_rate": 0.00011188370113977226,
+      "loss": 3.2535,
+      "step": 43814
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6902649998664856,
+      "learning_rate": 0.00011187170813200654,
+      "loss": 3.2002,
+      "step": 43815
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6927818655967712,
+      "learning_rate": 0.00011185971561973463,
+      "loss": 2.8208,
+      "step": 43816
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.728071391582489,
+      "learning_rate": 0.00011184772360298804,
+      "loss": 3.154,
+      "step": 43817
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6964516043663025,
+      "learning_rate": 0.00011183573208179832,
+      "loss": 3.2181,
+      "step": 43818
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7113494277000427,
+      "learning_rate": 0.00011182374105619696,
+      "loss": 2.8732,
+      "step": 43819
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7100608944892883,
+      "learning_rate": 0.00011181175052621568,
+      "loss": 3.012,
+      "step": 43820
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6832837462425232,
+      "learning_rate": 0.00011179976049188609,
+      "loss": 2.7346,
+      "step": 43821
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7172902226448059,
+      "learning_rate": 0.00011178777095323966,
+      "loss": 3.0578,
+      "step": 43822
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7123684883117676,
+      "learning_rate": 0.00011177578191030807,
+      "loss": 3.0289,
+      "step": 43823
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7194732427597046,
+      "learning_rate": 0.00011176379336312279,
+      "loss": 2.8659,
+      "step": 43824
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7032181620597839,
+      "learning_rate": 0.00011175180531171553,
+      "loss": 2.9368,
+      "step": 43825
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7210255265235901,
+      "learning_rate": 0.0001117398177561178,
+      "loss": 2.8385,
+      "step": 43826
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7119678258895874,
+      "learning_rate": 0.00011172783069636106,
+      "loss": 2.687,
+      "step": 43827
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6808481812477112,
+      "learning_rate": 0.00011171584413247706,
+      "loss": 2.9508,
+      "step": 43828
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6783161759376526,
+      "learning_rate": 0.00011170385806449719,
+      "loss": 3.0674,
+      "step": 43829
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7140722274780273,
+      "learning_rate": 0.00011169187249245313,
+      "loss": 2.7876,
+      "step": 43830
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6937413215637207,
+      "learning_rate": 0.00011167988741637652,
+      "loss": 3.2024,
+      "step": 43831
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.706088125705719,
+      "learning_rate": 0.00011166790283629889,
+      "loss": 2.8087,
+      "step": 43832
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7085302472114563,
+      "learning_rate": 0.0001116559187522517,
+      "loss": 2.7154,
+      "step": 43833
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6916400790214539,
+      "learning_rate": 0.0001116439351642665,
+      "loss": 2.9563,
+      "step": 43834
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6995882987976074,
+      "learning_rate": 0.00011163195207237493,
+      "loss": 2.7844,
+      "step": 43835
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7213637232780457,
+      "learning_rate": 0.00011161996947660864,
+      "loss": 3.0642,
+      "step": 43836
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6776930689811707,
+      "learning_rate": 0.000111607987376999,
+      "loss": 3.0639,
+      "step": 43837
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6928814649581909,
+      "learning_rate": 0.00011159600577357778,
+      "loss": 2.9717,
+      "step": 43838
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6714714169502258,
+      "learning_rate": 0.0001115840246663764,
+      "loss": 2.9155,
+      "step": 43839
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6734461784362793,
+      "learning_rate": 0.00011157204405542631,
+      "loss": 3.2625,
+      "step": 43840
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7229616641998291,
+      "learning_rate": 0.00011156006394075935,
+      "loss": 2.9042,
+      "step": 43841
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7451503276824951,
+      "learning_rate": 0.00011154808432240677,
+      "loss": 2.969,
+      "step": 43842
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6939261555671692,
+      "learning_rate": 0.00011153610520040039,
+      "loss": 2.8367,
+      "step": 43843
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7914009690284729,
+      "learning_rate": 0.00011152412657477155,
+      "loss": 2.8234,
+      "step": 43844
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7722774744033813,
+      "learning_rate": 0.00011151214844555186,
+      "loss": 2.8591,
+      "step": 43845
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.1138639450073242,
+      "learning_rate": 0.00011150017081277302,
+      "loss": 3.0576,
+      "step": 43846
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7328840494155884,
+      "learning_rate": 0.00011148819367646643,
+      "loss": 2.8833,
+      "step": 43847
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6928589344024658,
+      "learning_rate": 0.00011147621703666368,
+      "loss": 2.7493,
+      "step": 43848
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7033262848854065,
+      "learning_rate": 0.00011146424089339616,
+      "loss": 2.7993,
+      "step": 43849
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7417920827865601,
+      "learning_rate": 0.0001114522652466956,
+      "loss": 3.0088,
+      "step": 43850
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7107601761817932,
+      "learning_rate": 0.00011144029009659352,
+      "loss": 2.8515,
+      "step": 43851
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6659914255142212,
+      "learning_rate": 0.00011142831544312139,
+      "loss": 2.9079,
+      "step": 43852
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6738490462303162,
+      "learning_rate": 0.00011141634128631082,
+      "loss": 2.926,
+      "step": 43853
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.708010733127594,
+      "learning_rate": 0.00011140436762619336,
+      "loss": 2.9225,
+      "step": 43854
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7281856536865234,
+      "learning_rate": 0.00011139239446280039,
+      "loss": 3.0498,
+      "step": 43855
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7190080881118774,
+      "learning_rate": 0.00011138042179616364,
+      "loss": 2.6951,
+      "step": 43856
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7010225057601929,
+      "learning_rate": 0.00011136844962631449,
+      "loss": 2.8787,
+      "step": 43857
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7194762229919434,
+      "learning_rate": 0.00011135647795328461,
+      "loss": 2.806,
+      "step": 43858
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7051863670349121,
+      "learning_rate": 0.0001113445067771054,
+      "loss": 2.9298,
+      "step": 43859
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7123686671257019,
+      "learning_rate": 0.00011133253609780841,
+      "loss": 2.8497,
+      "step": 43860
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7194857597351074,
+      "learning_rate": 0.00011132056591542535,
+      "loss": 2.9762,
+      "step": 43861
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.705207884311676,
+      "learning_rate": 0.00011130859622998755,
+      "loss": 2.7528,
+      "step": 43862
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6921209692955017,
+      "learning_rate": 0.00011129662704152663,
+      "loss": 2.9349,
+      "step": 43863
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6830700635910034,
+      "learning_rate": 0.00011128465835007397,
+      "loss": 2.8637,
+      "step": 43864
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7572240829467773,
+      "learning_rate": 0.00011127269015566133,
+      "loss": 3.0668,
+      "step": 43865
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6755383014678955,
+      "learning_rate": 0.00011126072245831996,
+      "loss": 2.814,
+      "step": 43866
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7219648361206055,
+      "learning_rate": 0.00011124875525808156,
+      "loss": 2.7501,
+      "step": 43867
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7269349098205566,
+      "learning_rate": 0.00011123678855497768,
+      "loss": 2.7667,
+      "step": 43868
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7479811310768127,
+      "learning_rate": 0.0001112248223490398,
+      "loss": 2.907,
+      "step": 43869
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7241057753562927,
+      "learning_rate": 0.00011121285664029938,
+      "loss": 3.0555,
+      "step": 43870
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6962974071502686,
+      "learning_rate": 0.00011120089142878789,
+      "loss": 2.985,
+      "step": 43871
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7270846366882324,
+      "learning_rate": 0.0001111889267145369,
+      "loss": 2.7828,
+      "step": 43872
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7334160208702087,
+      "learning_rate": 0.00011117696249757801,
+      "loss": 2.902,
+      "step": 43873
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7237550616264343,
+      "learning_rate": 0.00011116499877794262,
+      "loss": 3.0554,
+      "step": 43874
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7188262343406677,
+      "learning_rate": 0.00011115303555566233,
+      "loss": 2.7096,
+      "step": 43875
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7280434370040894,
+      "learning_rate": 0.00011114107283076862,
+      "loss": 2.7683,
+      "step": 43876
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6606218218803406,
+      "learning_rate": 0.0001111291106032929,
+      "loss": 2.9093,
+      "step": 43877
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7308887839317322,
+      "learning_rate": 0.00011111714887326684,
+      "loss": 2.7953,
+      "step": 43878
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7169768214225769,
+      "learning_rate": 0.00011110518764072174,
+      "loss": 2.6644,
+      "step": 43879
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7009234428405762,
+      "learning_rate": 0.00011109322690568934,
+      "loss": 2.9879,
+      "step": 43880
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7171613574028015,
+      "learning_rate": 0.00011108126666820094,
+      "loss": 3.0013,
+      "step": 43881
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7147977352142334,
+      "learning_rate": 0.0001110693069282881,
+      "loss": 2.7512,
+      "step": 43882
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7143003344535828,
+      "learning_rate": 0.00011105734768598244,
+      "loss": 2.9142,
+      "step": 43883
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6658389568328857,
+      "learning_rate": 0.00011104538894131537,
+      "loss": 2.8317,
+      "step": 43884
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7165668606758118,
+      "learning_rate": 0.00011103343069431838,
+      "loss": 2.9532,
+      "step": 43885
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7112611532211304,
+      "learning_rate": 0.00011102147294502287,
+      "loss": 3.2125,
+      "step": 43886
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6863123178482056,
+      "learning_rate": 0.00011100951569346045,
+      "loss": 2.9475,
+      "step": 43887
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7303819060325623,
+      "learning_rate": 0.00011099755893966266,
+      "loss": 2.982,
+      "step": 43888
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7144728899002075,
+      "learning_rate": 0.00011098560268366087,
+      "loss": 2.8729,
+      "step": 43889
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6913188099861145,
+      "learning_rate": 0.00011097364692548669,
+      "loss": 2.8669,
+      "step": 43890
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7074322700500488,
+      "learning_rate": 0.00011096169166517153,
+      "loss": 3.1753,
+      "step": 43891
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6860098242759705,
+      "learning_rate": 0.00011094973690274684,
+      "loss": 2.9712,
+      "step": 43892
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6915674209594727,
+      "learning_rate": 0.00011093778263824422,
+      "loss": 2.9525,
+      "step": 43893
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7016196846961975,
+      "learning_rate": 0.00011092582887169501,
+      "loss": 2.9581,
+      "step": 43894
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6921491622924805,
+      "learning_rate": 0.00011091387560313088,
+      "loss": 2.9924,
+      "step": 43895
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7028470635414124,
+      "learning_rate": 0.00011090192283258312,
+      "loss": 2.7785,
+      "step": 43896
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6804261803627014,
+      "learning_rate": 0.0001108899705600834,
+      "loss": 2.8979,
+      "step": 43897
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7549912929534912,
+      "learning_rate": 0.00011087801878566308,
+      "loss": 3.1259,
+      "step": 43898
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6933797001838684,
+      "learning_rate": 0.00011086606750935358,
+      "loss": 2.9454,
+      "step": 43899
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6788550615310669,
+      "learning_rate": 0.00011085411673118654,
+      "loss": 2.9116,
+      "step": 43900
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7214063405990601,
+      "learning_rate": 0.00011084216645119328,
+      "loss": 2.8206,
+      "step": 43901
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6881868839263916,
+      "learning_rate": 0.00011083021666940533,
+      "loss": 2.8979,
+      "step": 43902
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7121565937995911,
+      "learning_rate": 0.00011081826738585428,
+      "loss": 2.8872,
+      "step": 43903
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6824859380722046,
+      "learning_rate": 0.00011080631860057144,
+      "loss": 2.7616,
+      "step": 43904
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7560449242591858,
+      "learning_rate": 0.00011079437031358844,
+      "loss": 3.0213,
+      "step": 43905
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7062668204307556,
+      "learning_rate": 0.00011078242252493663,
+      "loss": 3.0686,
+      "step": 43906
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6976759433746338,
+      "learning_rate": 0.00011077047523464738,
+      "loss": 2.8067,
+      "step": 43907
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7655328512191772,
+      "learning_rate": 0.00011075852844275242,
+      "loss": 2.8811,
+      "step": 43908
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6743826270103455,
+      "learning_rate": 0.00011074658214928298,
+      "loss": 2.7917,
+      "step": 43909
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6853476762771606,
+      "learning_rate": 0.00011073463635427068,
+      "loss": 3.049,
+      "step": 43910
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6936709880828857,
+      "learning_rate": 0.00011072269105774686,
+      "loss": 3.0572,
+      "step": 43911
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7399135231971741,
+      "learning_rate": 0.00011071074625974316,
+      "loss": 2.7455,
+      "step": 43912
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6544350385665894,
+      "learning_rate": 0.00011069880196029089,
+      "loss": 2.8913,
+      "step": 43913
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7409467101097107,
+      "learning_rate": 0.00011068685815942146,
+      "loss": 2.8284,
+      "step": 43914
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7109572291374207,
+      "learning_rate": 0.00011067491485716651,
+      "loss": 2.8065,
+      "step": 43915
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6747780442237854,
+      "learning_rate": 0.00011066297205355728,
+      "loss": 2.9514,
+      "step": 43916
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7135562896728516,
+      "learning_rate": 0.00011065102974862537,
+      "loss": 2.9687,
+      "step": 43917
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6812368035316467,
+      "learning_rate": 0.00011063908794240228,
+      "loss": 2.9524,
+      "step": 43918
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.8200016617774963,
+      "learning_rate": 0.00011062714663491941,
+      "loss": 2.8971,
+      "step": 43919
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6595801115036011,
+      "learning_rate": 0.00011061520582620818,
+      "loss": 3.0807,
+      "step": 43920
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6876057982444763,
+      "learning_rate": 0.00011060326551629993,
+      "loss": 2.9015,
+      "step": 43921
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6769374012947083,
+      "learning_rate": 0.00011059132570522627,
+      "loss": 2.9048,
+      "step": 43922
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7245048880577087,
+      "learning_rate": 0.00011057938639301868,
+      "loss": 3.0612,
+      "step": 43923
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6851451396942139,
+      "learning_rate": 0.00011056744757970844,
+      "loss": 2.8888,
+      "step": 43924
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7699195742607117,
+      "learning_rate": 0.00011055550926532718,
+      "loss": 2.7047,
+      "step": 43925
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6714342832565308,
+      "learning_rate": 0.00011054357144990615,
+      "loss": 3.0598,
+      "step": 43926
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7293652892112732,
+      "learning_rate": 0.00011053163413347701,
+      "loss": 2.7219,
+      "step": 43927
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7357663512229919,
+      "learning_rate": 0.00011051969731607106,
+      "loss": 3.0263,
+      "step": 43928
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6910032033920288,
+      "learning_rate": 0.00011050776099771966,
+      "loss": 2.9287,
+      "step": 43929
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6952933669090271,
+      "learning_rate": 0.00011049582517845445,
+      "loss": 2.846,
+      "step": 43930
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.742231547832489,
+      "learning_rate": 0.00011048388985830668,
+      "loss": 2.9032,
+      "step": 43931
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7026122212409973,
+      "learning_rate": 0.00011047195503730786,
+      "loss": 2.7614,
+      "step": 43932
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7984775304794312,
+      "learning_rate": 0.00011046002071548954,
+      "loss": 2.631,
+      "step": 43933
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6924632787704468,
+      "learning_rate": 0.00011044808689288306,
+      "loss": 2.9778,
+      "step": 43934
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7308608889579773,
+      "learning_rate": 0.00011043615356951982,
+      "loss": 2.9257,
+      "step": 43935
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.719944953918457,
+      "learning_rate": 0.00011042422074543118,
+      "loss": 2.8804,
+      "step": 43936
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7189018130302429,
+      "learning_rate": 0.00011041228842064864,
+      "loss": 2.9876,
+      "step": 43937
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7342965006828308,
+      "learning_rate": 0.00011040035659520375,
+      "loss": 3.0549,
+      "step": 43938
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7213689684867859,
+      "learning_rate": 0.00011038842526912776,
+      "loss": 2.8142,
+      "step": 43939
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7366634011268616,
+      "learning_rate": 0.00011037649444245225,
+      "loss": 2.757,
+      "step": 43940
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7220673561096191,
+      "learning_rate": 0.00011036456411520853,
+      "loss": 2.5639,
+      "step": 43941
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.705657422542572,
+      "learning_rate": 0.00011035263428742799,
+      "loss": 2.9537,
+      "step": 43942
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7454379200935364,
+      "learning_rate": 0.00011034070495914219,
+      "loss": 2.7032,
+      "step": 43943
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7317204475402832,
+      "learning_rate": 0.00011032877613038238,
+      "loss": 2.9766,
+      "step": 43944
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7271674871444702,
+      "learning_rate": 0.00011031684780118016,
+      "loss": 2.9546,
+      "step": 43945
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6818316578865051,
+      "learning_rate": 0.00011030491997156679,
+      "loss": 2.8576,
+      "step": 43946
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6842644810676575,
+      "learning_rate": 0.00011029299264157382,
+      "loss": 3.1527,
+      "step": 43947
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6735297441482544,
+      "learning_rate": 0.00011028106581123253,
+      "loss": 3.0781,
+      "step": 43948
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.745730996131897,
+      "learning_rate": 0.00011026913948057448,
+      "loss": 2.9958,
+      "step": 43949
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7029087543487549,
+      "learning_rate": 0.000110257213649631,
+      "loss": 2.845,
+      "step": 43950
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6922839879989624,
+      "learning_rate": 0.0001102452883184334,
+      "loss": 2.9185,
+      "step": 43951
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7484878301620483,
+      "learning_rate": 0.00011023336348701332,
+      "loss": 3.0122,
+      "step": 43952
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7015834450721741,
+      "learning_rate": 0.00011022143915540189,
+      "loss": 2.9061,
+      "step": 43953
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7120810151100159,
+      "learning_rate": 0.0001102095153236307,
+      "loss": 2.984,
+      "step": 43954
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.730353832244873,
+      "learning_rate": 0.00011019759199173123,
+      "loss": 2.9169,
+      "step": 43955
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7234374284744263,
+      "learning_rate": 0.00011018566915973477,
+      "loss": 3.1308,
+      "step": 43956
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6735356450080872,
+      "learning_rate": 0.0001101737468276727,
+      "loss": 2.9635,
+      "step": 43957
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6654012799263,
+      "learning_rate": 0.00011016182499557638,
+      "loss": 2.8069,
+      "step": 43958
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6770622134208679,
+      "learning_rate": 0.00011014990366347725,
+      "loss": 2.9357,
+      "step": 43959
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7623130083084106,
+      "learning_rate": 0.00011013798283140686,
+      "loss": 3.0011,
+      "step": 43960
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7139075398445129,
+      "learning_rate": 0.00011012606249939635,
+      "loss": 3.1044,
+      "step": 43961
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7361212372779846,
+      "learning_rate": 0.00011011414266747738,
+      "loss": 2.9142,
+      "step": 43962
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6977233290672302,
+      "learning_rate": 0.00011010222333568108,
+      "loss": 3.1171,
+      "step": 43963
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6813346743583679,
+      "learning_rate": 0.0001100903045040391,
+      "loss": 3.0365,
+      "step": 43964
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7580142021179199,
+      "learning_rate": 0.00011007838617258272,
+      "loss": 2.8071,
+      "step": 43965
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6768600940704346,
+      "learning_rate": 0.0001100664683413432,
+      "loss": 2.8438,
+      "step": 43966
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6694023013114929,
+      "learning_rate": 0.00011005455101035216,
+      "loss": 2.9374,
+      "step": 43967
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6759893894195557,
+      "learning_rate": 0.00011004263417964081,
+      "loss": 2.9976,
+      "step": 43968
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6973206996917725,
+      "learning_rate": 0.00011003071784924054,
+      "loss": 3.0073,
+      "step": 43969
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7266998291015625,
+      "learning_rate": 0.00011001880201918295,
+      "loss": 2.9469,
+      "step": 43970
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6822577118873596,
+      "learning_rate": 0.00011000688668949927,
+      "loss": 2.8839,
+      "step": 43971
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.7668481469154358,
+      "learning_rate": 0.00010999497186022086,
+      "loss": 3.0304,
+      "step": 43972
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7106690406799316,
+      "learning_rate": 0.00010998305753137901,
+      "loss": 2.7812,
+      "step": 43973
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6828577518463135,
+      "learning_rate": 0.00010997114370300523,
+      "loss": 2.889,
+      "step": 43974
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6851549744606018,
+      "learning_rate": 0.00010995923037513101,
+      "loss": 2.9489,
+      "step": 43975
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6917469501495361,
+      "learning_rate": 0.00010994731754778749,
+      "loss": 3.2144,
+      "step": 43976
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6860583424568176,
+      "learning_rate": 0.00010993540522100626,
+      "loss": 2.9111,
+      "step": 43977
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6750058531761169,
+      "learning_rate": 0.0001099234933948186,
+      "loss": 2.7888,
+      "step": 43978
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6400216817855835,
+      "learning_rate": 0.00010991158206925578,
+      "loss": 3.1329,
+      "step": 43979
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6475121378898621,
+      "learning_rate": 0.00010989967124434934,
+      "loss": 2.9708,
+      "step": 43980
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6716324687004089,
+      "learning_rate": 0.0001098877609201305,
+      "loss": 2.8048,
+      "step": 43981
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7722020745277405,
+      "learning_rate": 0.00010987585109663082,
+      "loss": 2.9072,
+      "step": 43982
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.750815212726593,
+      "learning_rate": 0.00010986394177388144,
+      "loss": 2.8756,
+      "step": 43983
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.725193977355957,
+      "learning_rate": 0.00010985203295191385,
+      "loss": 2.9399,
+      "step": 43984
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7670903205871582,
+      "learning_rate": 0.00010984012463075953,
+      "loss": 2.6659,
+      "step": 43985
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6852924823760986,
+      "learning_rate": 0.0001098282168104497,
+      "loss": 2.96,
+      "step": 43986
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7253130078315735,
+      "learning_rate": 0.00010981630949101578,
+      "loss": 2.5994,
+      "step": 43987
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7172092795372009,
+      "learning_rate": 0.00010980440267248896,
+      "loss": 3.1027,
+      "step": 43988
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7548342347145081,
+      "learning_rate": 0.00010979249635490075,
+      "loss": 2.9589,
+      "step": 43989
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7267433404922485,
+      "learning_rate": 0.0001097805905382826,
+      "loss": 2.8408,
+      "step": 43990
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6913394927978516,
+      "learning_rate": 0.00010976868522266569,
+      "loss": 2.9435,
+      "step": 43991
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7468941807746887,
+      "learning_rate": 0.00010975678040808153,
+      "loss": 2.8856,
+      "step": 43992
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7527220249176025,
+      "learning_rate": 0.00010974487609456138,
+      "loss": 2.7722,
+      "step": 43993
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8277668952941895,
+      "learning_rate": 0.00010973297228213654,
+      "loss": 2.9391,
+      "step": 43994
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7322959899902344,
+      "learning_rate": 0.00010972106897083851,
+      "loss": 2.8405,
+      "step": 43995
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6588741540908813,
+      "learning_rate": 0.0001097091661606985,
+      "loss": 2.8649,
+      "step": 43996
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7032271027565002,
+      "learning_rate": 0.00010969726385174799,
+      "loss": 2.939,
+      "step": 43997
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.666032075881958,
+      "learning_rate": 0.00010968536204401817,
+      "loss": 3.0517,
+      "step": 43998
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6766948699951172,
+      "learning_rate": 0.00010967346073754061,
+      "loss": 2.9951,
+      "step": 43999
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6832386255264282,
+      "learning_rate": 0.0001096615599323465,
+      "loss": 2.7958,
+      "step": 44000
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6713371276855469,
+      "learning_rate": 0.00010964965962846708,
+      "loss": 2.9533,
+      "step": 44001
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7049140930175781,
+      "learning_rate": 0.00010963775982593396,
+      "loss": 3.0323,
+      "step": 44002
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7148593068122864,
+      "learning_rate": 0.00010962586052477825,
+      "loss": 3.0128,
+      "step": 44003
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7361739277839661,
+      "learning_rate": 0.00010961396172503137,
+      "loss": 2.8242,
+      "step": 44004
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6860854029655457,
+      "learning_rate": 0.00010960206342672477,
+      "loss": 2.8884,
+      "step": 44005
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8418459892272949,
+      "learning_rate": 0.00010959016562988962,
+      "loss": 3.0021,
+      "step": 44006
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.728326678276062,
+      "learning_rate": 0.00010957826833455737,
+      "loss": 2.8175,
+      "step": 44007
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7612557411193848,
+      "learning_rate": 0.00010956637154075938,
+      "loss": 2.8144,
+      "step": 44008
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7064254283905029,
+      "learning_rate": 0.00010955447524852677,
+      "loss": 2.8047,
+      "step": 44009
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.694352924823761,
+      "learning_rate": 0.00010954257945789114,
+      "loss": 3.0043,
+      "step": 44010
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6741226315498352,
+      "learning_rate": 0.00010953068416888359,
+      "loss": 2.9749,
+      "step": 44011
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6707633137702942,
+      "learning_rate": 0.00010951878938153567,
+      "loss": 3.0853,
+      "step": 44012
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7317514419555664,
+      "learning_rate": 0.00010950689509587851,
+      "loss": 2.8133,
+      "step": 44013
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6689852476119995,
+      "learning_rate": 0.00010949500131194364,
+      "loss": 2.7656,
+      "step": 44014
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7170653939247131,
+      "learning_rate": 0.00010948310802976224,
+      "loss": 2.8662,
+      "step": 44015
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6306023597717285,
+      "learning_rate": 0.0001094712152493656,
+      "loss": 2.8672,
+      "step": 44016
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6976908445358276,
+      "learning_rate": 0.00010945932297078516,
+      "loss": 3.2326,
+      "step": 44017
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6834633946418762,
+      "learning_rate": 0.00010944743119405214,
+      "loss": 2.9629,
+      "step": 44018
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.666874885559082,
+      "learning_rate": 0.00010943553991919788,
+      "loss": 2.8406,
+      "step": 44019
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6879183650016785,
+      "learning_rate": 0.00010942364914625387,
+      "loss": 2.7367,
+      "step": 44020
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6888954043388367,
+      "learning_rate": 0.00010941175887525128,
+      "loss": 3.0009,
+      "step": 44021
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7139040231704712,
+      "learning_rate": 0.00010939986910622141,
+      "loss": 2.979,
+      "step": 44022
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7036026120185852,
+      "learning_rate": 0.00010938797983919553,
+      "loss": 2.9986,
+      "step": 44023
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7085335850715637,
+      "learning_rate": 0.00010937609107420512,
+      "loss": 2.8668,
+      "step": 44024
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7103490233421326,
+      "learning_rate": 0.0001093642028112813,
+      "loss": 2.8923,
+      "step": 44025
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6850371956825256,
+      "learning_rate": 0.00010935231505045547,
+      "loss": 2.8388,
+      "step": 44026
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6872725486755371,
+      "learning_rate": 0.00010934042779175905,
+      "loss": 3.0184,
+      "step": 44027
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6997257471084595,
+      "learning_rate": 0.00010932854103522318,
+      "loss": 3.055,
+      "step": 44028
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6553372144699097,
+      "learning_rate": 0.00010931665478087933,
+      "loss": 2.7495,
+      "step": 44029
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7009562253952026,
+      "learning_rate": 0.0001093047690287587,
+      "loss": 3.0277,
+      "step": 44030
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7105560302734375,
+      "learning_rate": 0.00010929288377889249,
+      "loss": 3.132,
+      "step": 44031
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6748614311218262,
+      "learning_rate": 0.00010928099903131225,
+      "loss": 3.0997,
+      "step": 44032
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7473457455635071,
+      "learning_rate": 0.00010926911478604903,
+      "loss": 2.9778,
+      "step": 44033
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6980936527252197,
+      "learning_rate": 0.00010925723104313439,
+      "loss": 3.0761,
+      "step": 44034
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6893765330314636,
+      "learning_rate": 0.00010924534780259937,
+      "loss": 2.7932,
+      "step": 44035
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6904683113098145,
+      "learning_rate": 0.0001092334650644755,
+      "loss": 2.8992,
+      "step": 44036
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7253763675689697,
+      "learning_rate": 0.00010922158282879398,
+      "loss": 2.9312,
+      "step": 44037
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6974086165428162,
+      "learning_rate": 0.00010920970109558597,
+      "loss": 2.9157,
+      "step": 44038
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7278842329978943,
+      "learning_rate": 0.00010919781986488302,
+      "loss": 2.8189,
+      "step": 44039
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.73302161693573,
+      "learning_rate": 0.00010918593913671615,
+      "loss": 2.7879,
+      "step": 44040
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7178779244422913,
+      "learning_rate": 0.00010917405891111684,
+      "loss": 2.8211,
+      "step": 44041
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7002609372138977,
+      "learning_rate": 0.00010916217918811642,
+      "loss": 2.9447,
+      "step": 44042
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7161980867385864,
+      "learning_rate": 0.00010915029996774607,
+      "loss": 2.8354,
+      "step": 44043
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7740674614906311,
+      "learning_rate": 0.0001091384212500371,
+      "loss": 3.0059,
+      "step": 44044
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7472088932991028,
+      "learning_rate": 0.00010912654303502072,
+      "loss": 2.5457,
+      "step": 44045
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6775591373443604,
+      "learning_rate": 0.00010911466532272827,
+      "loss": 2.7964,
+      "step": 44046
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6801954507827759,
+      "learning_rate": 0.00010910278811319119,
+      "loss": 2.9941,
+      "step": 44047
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7444169521331787,
+      "learning_rate": 0.00010909091140644049,
+      "loss": 3.0355,
+      "step": 44048
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7006654739379883,
+      "learning_rate": 0.0001090790352025077,
+      "loss": 2.8252,
+      "step": 44049
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7231435775756836,
+      "learning_rate": 0.00010906715950142389,
+      "loss": 2.8966,
+      "step": 44050
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6741910576820374,
+      "learning_rate": 0.00010905528430322054,
+      "loss": 2.9011,
+      "step": 44051
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6673889756202698,
+      "learning_rate": 0.0001090434096079288,
+      "loss": 2.8311,
+      "step": 44052
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7044496536254883,
+      "learning_rate": 0.00010903153541557988,
+      "loss": 2.8997,
+      "step": 44053
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7047674655914307,
+      "learning_rate": 0.00010901966172620527,
+      "loss": 2.9994,
+      "step": 44054
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7237603664398193,
+      "learning_rate": 0.00010900778853983599,
+      "loss": 2.6291,
+      "step": 44055
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8270739316940308,
+      "learning_rate": 0.00010899591585650345,
+      "loss": 2.7419,
+      "step": 44056
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7122185230255127,
+      "learning_rate": 0.000108984043676239,
+      "loss": 3.0659,
+      "step": 44057
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6983858346939087,
+      "learning_rate": 0.00010897217199907381,
+      "loss": 3.0083,
+      "step": 44058
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7358377575874329,
+      "learning_rate": 0.00010896030082503915,
+      "loss": 2.6745,
+      "step": 44059
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7250675559043884,
+      "learning_rate": 0.00010894843015416622,
+      "loss": 2.9184,
+      "step": 44060
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7548567056655884,
+      "learning_rate": 0.00010893655998648634,
+      "loss": 3.1155,
+      "step": 44061
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7145891189575195,
+      "learning_rate": 0.00010892469032203088,
+      "loss": 2.9655,
+      "step": 44062
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7194769382476807,
+      "learning_rate": 0.00010891282116083093,
+      "loss": 2.9013,
+      "step": 44063
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6655535101890564,
+      "learning_rate": 0.00010890095250291792,
+      "loss": 3.0525,
+      "step": 44064
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7517229318618774,
+      "learning_rate": 0.00010888908434832294,
+      "loss": 3.1682,
+      "step": 44065
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6540136337280273,
+      "learning_rate": 0.00010887721669707742,
+      "loss": 2.9677,
+      "step": 44066
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7223973274230957,
+      "learning_rate": 0.00010886534954921253,
+      "loss": 2.8192,
+      "step": 44067
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7570539116859436,
+      "learning_rate": 0.00010885348290475943,
+      "loss": 3.0616,
+      "step": 44068
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6754083037376404,
+      "learning_rate": 0.00010884161676374957,
+      "loss": 3.1191,
+      "step": 44069
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6810846328735352,
+      "learning_rate": 0.00010882975112621401,
+      "loss": 3.0448,
+      "step": 44070
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7092759609222412,
+      "learning_rate": 0.00010881788599218406,
+      "loss": 2.939,
+      "step": 44071
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6637600064277649,
+      "learning_rate": 0.00010880602136169116,
+      "loss": 2.9407,
+      "step": 44072
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7038546800613403,
+      "learning_rate": 0.00010879415723476636,
+      "loss": 3.0425,
+      "step": 44073
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7373028993606567,
+      "learning_rate": 0.00010878229361144098,
+      "loss": 3.0257,
+      "step": 44074
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6334795355796814,
+      "learning_rate": 0.00010877043049174608,
+      "loss": 2.9988,
+      "step": 44075
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.716715395450592,
+      "learning_rate": 0.0001087585678757131,
+      "loss": 2.8645,
+      "step": 44076
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7330484986305237,
+      "learning_rate": 0.00010874670576337335,
+      "loss": 3.0347,
+      "step": 44077
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6826630234718323,
+      "learning_rate": 0.00010873484415475789,
+      "loss": 3.0281,
+      "step": 44078
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6680867671966553,
+      "learning_rate": 0.00010872298304989811,
+      "loss": 3.0422,
+      "step": 44079
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6701593995094299,
+      "learning_rate": 0.00010871112244882519,
+      "loss": 2.7085,
+      "step": 44080
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7055572271347046,
+      "learning_rate": 0.00010869926235157028,
+      "loss": 3.1402,
+      "step": 44081
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6829777359962463,
+      "learning_rate": 0.00010868740275816478,
+      "loss": 2.7673,
+      "step": 44082
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6810120344161987,
+      "learning_rate": 0.00010867554366863972,
+      "loss": 3.0536,
+      "step": 44083
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6956943869590759,
+      "learning_rate": 0.00010866368508302658,
+      "loss": 2.9712,
+      "step": 44084
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7180123329162598,
+      "learning_rate": 0.00010865182700135638,
+      "loss": 2.8563,
+      "step": 44085
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6888797283172607,
+      "learning_rate": 0.00010863996942366042,
+      "loss": 3.0218,
+      "step": 44086
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6845893263816833,
+      "learning_rate": 0.00010862811234997007,
+      "loss": 2.9983,
+      "step": 44087
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7172102928161621,
+      "learning_rate": 0.00010861625578031642,
+      "loss": 2.9244,
+      "step": 44088
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7032397985458374,
+      "learning_rate": 0.00010860439971473072,
+      "loss": 2.9089,
+      "step": 44089
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7030555605888367,
+      "learning_rate": 0.00010859254415324413,
+      "loss": 2.8213,
+      "step": 44090
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7154149413108826,
+      "learning_rate": 0.0001085806890958879,
+      "loss": 2.6665,
+      "step": 44091
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6783580780029297,
+      "learning_rate": 0.00010856883454269342,
+      "loss": 2.9993,
+      "step": 44092
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7258152365684509,
+      "learning_rate": 0.0001085569804936917,
+      "loss": 2.9013,
+      "step": 44093
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6536148190498352,
+      "learning_rate": 0.00010854512694891413,
+      "loss": 2.8165,
+      "step": 44094
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.712685763835907,
+      "learning_rate": 0.00010853327390839184,
+      "loss": 2.9026,
+      "step": 44095
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6730918288230896,
+      "learning_rate": 0.00010852142137215597,
+      "loss": 2.9592,
+      "step": 44096
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6709897518157959,
+      "learning_rate": 0.00010850956934023795,
+      "loss": 2.5844,
+      "step": 44097
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6992678642272949,
+      "learning_rate": 0.00010849771781266872,
+      "loss": 2.9257,
+      "step": 44098
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7160495519638062,
+      "learning_rate": 0.0001084858667894798,
+      "loss": 3.0139,
+      "step": 44099
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6836722493171692,
+      "learning_rate": 0.00010847401627070212,
+      "loss": 3.0412,
+      "step": 44100
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6511483192443848,
+      "learning_rate": 0.00010846216625636712,
+      "loss": 2.8481,
+      "step": 44101
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6964370012283325,
+      "learning_rate": 0.00010845031674650589,
+      "loss": 2.901,
+      "step": 44102
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7739323377609253,
+      "learning_rate": 0.0001084384677411496,
+      "loss": 3.06,
+      "step": 44103
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7379111051559448,
+      "learning_rate": 0.0001084266192403296,
+      "loss": 2.8302,
+      "step": 44104
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7733993530273438,
+      "learning_rate": 0.00010841477124407694,
+      "loss": 2.986,
+      "step": 44105
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7553689479827881,
+      "learning_rate": 0.00010840292375242298,
+      "loss": 2.9365,
+      "step": 44106
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6967439651489258,
+      "learning_rate": 0.00010839107676539875,
+      "loss": 2.9238,
+      "step": 44107
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6885391473770142,
+      "learning_rate": 0.00010837923028303555,
+      "loss": 2.9053,
+      "step": 44108
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6458328366279602,
+      "learning_rate": 0.00010836738430536466,
+      "loss": 2.945,
+      "step": 44109
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7221814393997192,
+      "learning_rate": 0.00010835553883241723,
+      "loss": 2.827,
+      "step": 44110
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6976704001426697,
+      "learning_rate": 0.00010834369386422441,
+      "loss": 2.8029,
+      "step": 44111
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7037103176116943,
+      "learning_rate": 0.0001083318494008173,
+      "loss": 2.8437,
+      "step": 44112
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7122330069541931,
+      "learning_rate": 0.00010832000544222726,
+      "loss": 2.9479,
+      "step": 44113
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.705687403678894,
+      "learning_rate": 0.0001083081619884855,
+      "loss": 2.8805,
+      "step": 44114
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6933003067970276,
+      "learning_rate": 0.00010829631903962304,
+      "loss": 2.9338,
+      "step": 44115
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7246118783950806,
+      "learning_rate": 0.00010828447659567132,
+      "loss": 2.9069,
+      "step": 44116
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6804153323173523,
+      "learning_rate": 0.00010827263465666137,
+      "loss": 2.8831,
+      "step": 44117
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7856549024581909,
+      "learning_rate": 0.00010826079322262427,
+      "loss": 2.8996,
+      "step": 44118
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6795721650123596,
+      "learning_rate": 0.00010824895229359146,
+      "loss": 2.716,
+      "step": 44119
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7031411528587341,
+      "learning_rate": 0.00010823711186959392,
+      "loss": 2.9412,
+      "step": 44120
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6899999976158142,
+      "learning_rate": 0.00010822527195066301,
+      "loss": 3.0925,
+      "step": 44121
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7404148578643799,
+      "learning_rate": 0.00010821343253682977,
+      "loss": 2.9588,
+      "step": 44122
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7039032578468323,
+      "learning_rate": 0.0001082015936281255,
+      "loss": 2.9895,
+      "step": 44123
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6816439032554626,
+      "learning_rate": 0.0001081897552245813,
+      "loss": 2.707,
+      "step": 44124
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7398577928543091,
+      "learning_rate": 0.00010817791732622829,
+      "loss": 3.0734,
+      "step": 44125
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7134485244750977,
+      "learning_rate": 0.00010816607993309785,
+      "loss": 2.9215,
+      "step": 44126
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7148016691207886,
+      "learning_rate": 0.0001081542430452209,
+      "loss": 3.0477,
+      "step": 44127
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7097327709197998,
+      "learning_rate": 0.00010814240666262879,
+      "loss": 2.9591,
+      "step": 44128
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7393547892570496,
+      "learning_rate": 0.00010813057078535272,
+      "loss": 3.0054,
+      "step": 44129
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7232438325881958,
+      "learning_rate": 0.00010811873541342372,
+      "loss": 2.7291,
+      "step": 44130
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8301907181739807,
+      "learning_rate": 0.00010810690054687315,
+      "loss": 2.7839,
+      "step": 44131
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.751732349395752,
+      "learning_rate": 0.00010809506618573207,
+      "loss": 2.8497,
+      "step": 44132
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7149835228919983,
+      "learning_rate": 0.00010808323233003154,
+      "loss": 3.0302,
+      "step": 44133
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7186267375946045,
+      "learning_rate": 0.00010807139897980295,
+      "loss": 2.8241,
+      "step": 44134
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6884257197380066,
+      "learning_rate": 0.00010805956613507726,
+      "loss": 2.8621,
+      "step": 44135
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7444910407066345,
+      "learning_rate": 0.00010804773379588582,
+      "loss": 2.8246,
+      "step": 44136
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7724918127059937,
+      "learning_rate": 0.00010803590196225963,
+      "loss": 2.9742,
+      "step": 44137
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.707662045955658,
+      "learning_rate": 0.00010802407063423001,
+      "loss": 2.7956,
+      "step": 44138
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7242353558540344,
+      "learning_rate": 0.00010801223981182806,
+      "loss": 3.0574,
+      "step": 44139
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6966133713722229,
+      "learning_rate": 0.0001080004094950848,
+      "loss": 3.0594,
+      "step": 44140
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7180776000022888,
+      "learning_rate": 0.00010798857968403162,
+      "loss": 2.9338,
+      "step": 44141
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.97182697057724,
+      "learning_rate": 0.00010797675037869947,
+      "loss": 2.9121,
+      "step": 44142
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7087554335594177,
+      "learning_rate": 0.0001079649215791196,
+      "loss": 2.6782,
+      "step": 44143
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6923799514770508,
+      "learning_rate": 0.00010795309328532327,
+      "loss": 2.8516,
+      "step": 44144
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6832737922668457,
+      "learning_rate": 0.00010794126549734155,
+      "loss": 2.8526,
+      "step": 44145
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6955265402793884,
+      "learning_rate": 0.00010792943821520554,
+      "loss": 2.892,
+      "step": 44146
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7082833051681519,
+      "learning_rate": 0.00010791761143894635,
+      "loss": 2.7952,
+      "step": 44147
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6620471477508545,
+      "learning_rate": 0.00010790578516859517,
+      "loss": 3.1079,
+      "step": 44148
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7037109136581421,
+      "learning_rate": 0.00010789395940418331,
+      "loss": 3.0704,
+      "step": 44149
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6921175122261047,
+      "learning_rate": 0.0001078821341457417,
+      "loss": 2.9341,
+      "step": 44150
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7180284261703491,
+      "learning_rate": 0.00010787030939330165,
+      "loss": 2.9196,
+      "step": 44151
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6754325032234192,
+      "learning_rate": 0.00010785848514689414,
+      "loss": 2.8187,
+      "step": 44152
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7477900981903076,
+      "learning_rate": 0.00010784666140655051,
+      "loss": 3.0184,
+      "step": 44153
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7051811814308167,
+      "learning_rate": 0.0001078348381723018,
+      "loss": 2.8238,
+      "step": 44154
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7260898947715759,
+      "learning_rate": 0.00010782301544417903,
+      "loss": 2.8263,
+      "step": 44155
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7143333554267883,
+      "learning_rate": 0.00010781119322221354,
+      "loss": 2.773,
+      "step": 44156
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7197613716125488,
+      "learning_rate": 0.00010779937150643628,
+      "loss": 2.9153,
+      "step": 44157
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7276310324668884,
+      "learning_rate": 0.00010778755029687852,
+      "loss": 2.8371,
+      "step": 44158
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6611834764480591,
+      "learning_rate": 0.00010777572959357143,
+      "loss": 3.0825,
+      "step": 44159
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6871948838233948,
+      "learning_rate": 0.0001077639093965461,
+      "loss": 3.0084,
+      "step": 44160
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7019148468971252,
+      "learning_rate": 0.00010775208970583359,
+      "loss": 2.9931,
+      "step": 44161
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6876431703567505,
+      "learning_rate": 0.000107740270521465,
+      "loss": 3.141,
+      "step": 44162
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.667335033416748,
+      "learning_rate": 0.00010772845184347154,
+      "loss": 2.7994,
+      "step": 44163
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6756302118301392,
+      "learning_rate": 0.00010771663367188445,
+      "loss": 3.0605,
+      "step": 44164
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7222525477409363,
+      "learning_rate": 0.00010770481600673463,
+      "loss": 3.2013,
+      "step": 44165
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7120800018310547,
+      "learning_rate": 0.0001076929988480534,
+      "loss": 2.8396,
+      "step": 44166
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.698549211025238,
+      "learning_rate": 0.00010768118219587182,
+      "loss": 2.9708,
+      "step": 44167
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6790973544120789,
+      "learning_rate": 0.00010766936605022088,
+      "loss": 3.1835,
+      "step": 44168
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6756714582443237,
+      "learning_rate": 0.00010765755041113194,
+      "loss": 2.6655,
+      "step": 44169
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7276606559753418,
+      "learning_rate": 0.00010764573527863587,
+      "loss": 2.8426,
+      "step": 44170
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6843422055244446,
+      "learning_rate": 0.00010763392065276404,
+      "loss": 2.9493,
+      "step": 44171
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7065037488937378,
+      "learning_rate": 0.00010762210653354733,
+      "loss": 2.8149,
+      "step": 44172
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7013691067695618,
+      "learning_rate": 0.00010761029292101693,
+      "loss": 2.811,
+      "step": 44173
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.727316677570343,
+      "learning_rate": 0.00010759847981520412,
+      "loss": 3.0262,
+      "step": 44174
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7056801319122314,
+      "learning_rate": 0.0001075866672161399,
+      "loss": 2.8486,
+      "step": 44175
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6833690404891968,
+      "learning_rate": 0.00010757485512385535,
+      "loss": 2.8004,
+      "step": 44176
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6740131974220276,
+      "learning_rate": 0.00010756304353838152,
+      "loss": 2.829,
+      "step": 44177
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7866957187652588,
+      "learning_rate": 0.00010755123245974969,
+      "loss": 2.8718,
+      "step": 44178
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6585379838943481,
+      "learning_rate": 0.00010753942188799077,
+      "loss": 2.9546,
+      "step": 44179
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7182008624076843,
+      "learning_rate": 0.00010752761182313597,
+      "loss": 2.8891,
+      "step": 44180
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.695152223110199,
+      "learning_rate": 0.0001075158022652165,
+      "loss": 3.0113,
+      "step": 44181
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7504631280899048,
+      "learning_rate": 0.00010750399321426336,
+      "loss": 2.8999,
+      "step": 44182
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6874749064445496,
+      "learning_rate": 0.00010749218467030767,
+      "loss": 2.7791,
+      "step": 44183
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.725240170955658,
+      "learning_rate": 0.00010748037663338039,
+      "loss": 2.7461,
+      "step": 44184
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7044048309326172,
+      "learning_rate": 0.00010746856910351275,
+      "loss": 2.9411,
+      "step": 44185
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7227499485015869,
+      "learning_rate": 0.00010745676208073597,
+      "loss": 2.9253,
+      "step": 44186
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.700607180595398,
+      "learning_rate": 0.0001074449555650809,
+      "loss": 2.886,
+      "step": 44187
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6638893485069275,
+      "learning_rate": 0.00010743314955657889,
+      "loss": 2.8618,
+      "step": 44188
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7097645401954651,
+      "learning_rate": 0.00010742134405526078,
+      "loss": 3.0554,
+      "step": 44189
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7432897686958313,
+      "learning_rate": 0.0001074095390611579,
+      "loss": 3.0488,
+      "step": 44190
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7031722068786621,
+      "learning_rate": 0.00010739773457430122,
+      "loss": 2.9038,
+      "step": 44191
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6727100610733032,
+      "learning_rate": 0.00010738593059472174,
+      "loss": 2.8709,
+      "step": 44192
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7058899402618408,
+      "learning_rate": 0.00010737412712245074,
+      "loss": 3.1592,
+      "step": 44193
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6950834393501282,
+      "learning_rate": 0.00010736232415751913,
+      "loss": 3.0163,
+      "step": 44194
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7071797251701355,
+      "learning_rate": 0.00010735052169995809,
+      "loss": 2.9139,
+      "step": 44195
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6854180097579956,
+      "learning_rate": 0.00010733871974979878,
+      "loss": 2.8134,
+      "step": 44196
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7189562916755676,
+      "learning_rate": 0.00010732691830707221,
+      "loss": 2.8558,
+      "step": 44197
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6695606112480164,
+      "learning_rate": 0.00010731511737180947,
+      "loss": 3.0807,
+      "step": 44198
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6888101696968079,
+      "learning_rate": 0.00010730331694404153,
+      "loss": 3.014,
+      "step": 44199
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6768573522567749,
+      "learning_rate": 0.00010729151702379955,
+      "loss": 2.9749,
+      "step": 44200
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7207431197166443,
+      "learning_rate": 0.00010727971761111472,
+      "loss": 3.0053,
+      "step": 44201
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7181470394134521,
+      "learning_rate": 0.00010726791870601795,
+      "loss": 2.881,
+      "step": 44202
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7218087911605835,
+      "learning_rate": 0.00010725612030854048,
+      "loss": 2.86,
+      "step": 44203
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7341431379318237,
+      "learning_rate": 0.00010724432241871329,
+      "loss": 2.9529,
+      "step": 44204
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7279463410377502,
+      "learning_rate": 0.00010723252503656737,
+      "loss": 2.9883,
+      "step": 44205
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6712480187416077,
+      "learning_rate": 0.00010722072816213396,
+      "loss": 2.8739,
+      "step": 44206
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6901949048042297,
+      "learning_rate": 0.00010720893179544397,
+      "loss": 2.8589,
+      "step": 44207
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7036940455436707,
+      "learning_rate": 0.00010719713593652866,
+      "loss": 2.8149,
+      "step": 44208
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7346253991127014,
+      "learning_rate": 0.00010718534058541885,
+      "loss": 2.9119,
+      "step": 44209
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7171415686607361,
+      "learning_rate": 0.00010717354574214577,
+      "loss": 3.0225,
+      "step": 44210
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7323898673057556,
+      "learning_rate": 0.00010716175140674055,
+      "loss": 2.8658,
+      "step": 44211
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7688388824462891,
+      "learning_rate": 0.00010714995757923418,
+      "loss": 2.8879,
+      "step": 44212
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6723164319992065,
+      "learning_rate": 0.00010713816425965768,
+      "loss": 2.9502,
+      "step": 44213
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6992418766021729,
+      "learning_rate": 0.00010712637144804207,
+      "loss": 2.9745,
+      "step": 44214
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6746936440467834,
+      "learning_rate": 0.00010711457914441846,
+      "loss": 2.9747,
+      "step": 44215
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7169411182403564,
+      "learning_rate": 0.00010710278734881804,
+      "loss": 3.0951,
+      "step": 44216
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7070459127426147,
+      "learning_rate": 0.00010709099606127165,
+      "loss": 2.8509,
+      "step": 44217
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6970311999320984,
+      "learning_rate": 0.00010707920528181053,
+      "loss": 2.991,
+      "step": 44218
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7649372816085815,
+      "learning_rate": 0.00010706741501046566,
+      "loss": 3.022,
+      "step": 44219
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.727436900138855,
+      "learning_rate": 0.00010705562524726802,
+      "loss": 2.9837,
+      "step": 44220
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6991956830024719,
+      "learning_rate": 0.0001070438359922488,
+      "loss": 2.7914,
+      "step": 44221
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6980074048042297,
+      "learning_rate": 0.0001070320472454389,
+      "loss": 3.0317,
+      "step": 44222
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6773894429206848,
+      "learning_rate": 0.00010702025900686954,
+      "loss": 2.8222,
+      "step": 44223
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7050642967224121,
+      "learning_rate": 0.00010700847127657159,
+      "loss": 2.8748,
+      "step": 44224
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7063111662864685,
+      "learning_rate": 0.00010699668405457625,
+      "loss": 3.1069,
+      "step": 44225
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7139179706573486,
+      "learning_rate": 0.00010698489734091454,
+      "loss": 2.855,
+      "step": 44226
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6964961886405945,
+      "learning_rate": 0.00010697311113561733,
+      "loss": 3.0488,
+      "step": 44227
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6961595416069031,
+      "learning_rate": 0.00010696132543871591,
+      "loss": 2.9491,
+      "step": 44228
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6952822208404541,
+      "learning_rate": 0.00010694954025024115,
+      "loss": 3.0702,
+      "step": 44229
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7170188426971436,
+      "learning_rate": 0.00010693775557022407,
+      "loss": 2.9606,
+      "step": 44230
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7045404314994812,
+      "learning_rate": 0.00010692597139869595,
+      "loss": 2.8441,
+      "step": 44231
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7429175972938538,
+      "learning_rate": 0.00010691418773568752,
+      "loss": 3.0192,
+      "step": 44232
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7014785408973694,
+      "learning_rate": 0.00010690240458123008,
+      "loss": 2.8359,
+      "step": 44233
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7238576412200928,
+      "learning_rate": 0.00010689062193535456,
+      "loss": 2.9959,
+      "step": 44234
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.701926052570343,
+      "learning_rate": 0.00010687883979809183,
+      "loss": 2.9024,
+      "step": 44235
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7221980094909668,
+      "learning_rate": 0.00010686705816947323,
+      "loss": 2.9919,
+      "step": 44236
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7259619235992432,
+      "learning_rate": 0.0001068552770495295,
+      "loss": 2.8936,
+      "step": 44237
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7003749012947083,
+      "learning_rate": 0.00010684349643829188,
+      "loss": 3.1794,
+      "step": 44238
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7119761109352112,
+      "learning_rate": 0.00010683171633579124,
+      "loss": 2.933,
+      "step": 44239
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7327340841293335,
+      "learning_rate": 0.00010681993674205877,
+      "loss": 2.743,
+      "step": 44240
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7414251565933228,
+      "learning_rate": 0.00010680815765712543,
+      "loss": 2.885,
+      "step": 44241
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.683208703994751,
+      "learning_rate": 0.0001067963790810221,
+      "loss": 2.9057,
+      "step": 44242
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.711431086063385,
+      "learning_rate": 0.00010678460101378003,
+      "loss": 2.8266,
+      "step": 44243
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.711889386177063,
+      "learning_rate": 0.00010677282345543005,
+      "loss": 2.8303,
+      "step": 44244
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7022669315338135,
+      "learning_rate": 0.00010676104640600324,
+      "loss": 2.8831,
+      "step": 44245
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.684972882270813,
+      "learning_rate": 0.00010674926986553074,
+      "loss": 2.8915,
+      "step": 44246
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6765041947364807,
+      "learning_rate": 0.0001067374938340435,
+      "loss": 2.8127,
+      "step": 44247
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6731730103492737,
+      "learning_rate": 0.00010672571831157251,
+      "loss": 2.9058,
+      "step": 44248
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7448628544807434,
+      "learning_rate": 0.00010671394329814864,
+      "loss": 2.6898,
+      "step": 44249
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7296499609947205,
+      "learning_rate": 0.00010670216879380308,
+      "loss": 3.0543,
+      "step": 44250
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7346038222312927,
+      "learning_rate": 0.00010669039479856691,
+      "loss": 2.8413,
+      "step": 44251
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6547524929046631,
+      "learning_rate": 0.00010667862131247092,
+      "loss": 2.7705,
+      "step": 44252
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.719100832939148,
+      "learning_rate": 0.00010666684833554635,
+      "loss": 3.0006,
+      "step": 44253
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7397176027297974,
+      "learning_rate": 0.00010665507586782399,
+      "loss": 2.7199,
+      "step": 44254
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7125775814056396,
+      "learning_rate": 0.00010664330390933505,
+      "loss": 2.9314,
+      "step": 44255
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7293382287025452,
+      "learning_rate": 0.00010663153246011042,
+      "loss": 3.1108,
+      "step": 44256
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6914783120155334,
+      "learning_rate": 0.00010661976152018101,
+      "loss": 2.8427,
+      "step": 44257
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7073983550071716,
+      "learning_rate": 0.00010660799108957806,
+      "loss": 2.8991,
+      "step": 44258
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7683893442153931,
+      "learning_rate": 0.00010659622116833233,
+      "loss": 2.8923,
+      "step": 44259
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7236352562904358,
+      "learning_rate": 0.00010658445175647506,
+      "loss": 2.9805,
+      "step": 44260
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7174046635627747,
+      "learning_rate": 0.00010657268285403697,
+      "loss": 3.0995,
+      "step": 44261
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6961384415626526,
+      "learning_rate": 0.00010656091446104936,
+      "loss": 2.8501,
+      "step": 44262
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7074950933456421,
+      "learning_rate": 0.00010654914657754304,
+      "loss": 2.8524,
+      "step": 44263
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6755015850067139,
+      "learning_rate": 0.00010653737920354896,
+      "loss": 2.8667,
+      "step": 44264
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.673994779586792,
+      "learning_rate": 0.00010652561233909828,
+      "loss": 2.8819,
+      "step": 44265
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6857726573944092,
+      "learning_rate": 0.00010651384598422178,
+      "loss": 2.9646,
+      "step": 44266
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7020795941352844,
+      "learning_rate": 0.00010650208013895057,
+      "loss": 2.888,
+      "step": 44267
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7138441801071167,
+      "learning_rate": 0.00010649031480331577,
+      "loss": 2.9728,
+      "step": 44268
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6965253949165344,
+      "learning_rate": 0.00010647854997734824,
+      "loss": 2.8346,
+      "step": 44269
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6731846332550049,
+      "learning_rate": 0.00010646678566107895,
+      "loss": 2.8683,
+      "step": 44270
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7317405939102173,
+      "learning_rate": 0.0001064550218545388,
+      "loss": 2.9809,
+      "step": 44271
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7305527925491333,
+      "learning_rate": 0.00010644325855775889,
+      "loss": 2.9701,
+      "step": 44272
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6821311712265015,
+      "learning_rate": 0.00010643149577077027,
+      "loss": 2.8188,
+      "step": 44273
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7498180866241455,
+      "learning_rate": 0.00010641973349360374,
+      "loss": 3.0598,
+      "step": 44274
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.731899619102478,
+      "learning_rate": 0.00010640797172629051,
+      "loss": 2.8892,
+      "step": 44275
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7006483674049377,
+      "learning_rate": 0.00010639621046886128,
+      "loss": 2.86,
+      "step": 44276
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7303549647331238,
+      "learning_rate": 0.00010638444972134728,
+      "loss": 2.597,
+      "step": 44277
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6863207817077637,
+      "learning_rate": 0.00010637268948377939,
+      "loss": 2.6882,
+      "step": 44278
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.66311115026474,
+      "learning_rate": 0.00010636092975618847,
+      "loss": 2.9305,
+      "step": 44279
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7255011796951294,
+      "learning_rate": 0.00010634917053860569,
+      "loss": 2.8173,
+      "step": 44280
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6959855556488037,
+      "learning_rate": 0.00010633741183106184,
+      "loss": 2.8846,
+      "step": 44281
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6742451190948486,
+      "learning_rate": 0.00010632565363358797,
+      "loss": 3.0126,
+      "step": 44282
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7093662023544312,
+      "learning_rate": 0.00010631389594621515,
+      "loss": 3.0639,
+      "step": 44283
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6777709126472473,
+      "learning_rate": 0.00010630213876897423,
+      "loss": 2.9066,
+      "step": 44284
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.700195848941803,
+      "learning_rate": 0.00010629038210189623,
+      "loss": 2.9751,
+      "step": 44285
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6865321397781372,
+      "learning_rate": 0.00010627862594501196,
+      "loss": 3.0077,
+      "step": 44286
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6771736741065979,
+      "learning_rate": 0.0001062668702983525,
+      "loss": 2.7006,
+      "step": 44287
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7320778965950012,
+      "learning_rate": 0.00010625511516194895,
+      "loss": 2.9485,
+      "step": 44288
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7206733226776123,
+      "learning_rate": 0.00010624336053583202,
+      "loss": 2.8812,
+      "step": 44289
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6806963682174683,
+      "learning_rate": 0.00010623160642003292,
+      "loss": 2.8639,
+      "step": 44290
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7328031659126282,
+      "learning_rate": 0.00010621985281458233,
+      "loss": 2.9097,
+      "step": 44291
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7521789073944092,
+      "learning_rate": 0.00010620809971951148,
+      "loss": 2.605,
+      "step": 44292
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7122011780738831,
+      "learning_rate": 0.00010619634713485122,
+      "loss": 2.9234,
+      "step": 44293
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7520004510879517,
+      "learning_rate": 0.00010618459506063235,
+      "loss": 2.8301,
+      "step": 44294
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7129939794540405,
+      "learning_rate": 0.00010617284349688607,
+      "loss": 2.9049,
+      "step": 44295
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6909685134887695,
+      "learning_rate": 0.0001061610924436431,
+      "loss": 2.9319,
+      "step": 44296
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6826851963996887,
+      "learning_rate": 0.00010614934190093454,
+      "loss": 2.7563,
+      "step": 44297
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6831250190734863,
+      "learning_rate": 0.00010613759186879138,
+      "loss": 2.9022,
+      "step": 44298
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6965422034263611,
+      "learning_rate": 0.00010612584234724452,
+      "loss": 3.158,
+      "step": 44299
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7100363969802856,
+      "learning_rate": 0.00010611409333632486,
+      "loss": 2.6984,
+      "step": 44300
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7082723379135132,
+      "learning_rate": 0.00010610234483606325,
+      "loss": 2.8544,
+      "step": 44301
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7065490484237671,
+      "learning_rate": 0.00010609059684649076,
+      "loss": 3.014,
+      "step": 44302
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7121196389198303,
+      "learning_rate": 0.00010607884936763841,
+      "loss": 2.8571,
+      "step": 44303
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7432886958122253,
+      "learning_rate": 0.00010606710239953697,
+      "loss": 2.9936,
+      "step": 44304
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6981472969055176,
+      "learning_rate": 0.00010605535594221757,
+      "loss": 2.8229,
+      "step": 44305
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7373912334442139,
+      "learning_rate": 0.000106043609995711,
+      "loss": 2.8332,
+      "step": 44306
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7022235989570618,
+      "learning_rate": 0.00010603186456004816,
+      "loss": 2.783,
+      "step": 44307
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7700810432434082,
+      "learning_rate": 0.00010602011963526013,
+      "loss": 2.9577,
+      "step": 44308
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7263349294662476,
+      "learning_rate": 0.0001060083752213777,
+      "loss": 2.8884,
+      "step": 44309
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6605125665664673,
+      "learning_rate": 0.00010599663131843197,
+      "loss": 2.9842,
+      "step": 44310
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.680346667766571,
+      "learning_rate": 0.00010598488792645367,
+      "loss": 3.0116,
+      "step": 44311
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.696643054485321,
+      "learning_rate": 0.00010597314504547386,
+      "loss": 2.5903,
+      "step": 44312
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7104796171188354,
+      "learning_rate": 0.00010596140267552351,
+      "loss": 3.0219,
+      "step": 44313
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.8690118193626404,
+      "learning_rate": 0.00010594966081663352,
+      "loss": 2.9789,
+      "step": 44314
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7163686156272888,
+      "learning_rate": 0.00010593791946883473,
+      "loss": 3.0463,
+      "step": 44315
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7209365367889404,
+      "learning_rate": 0.00010592617863215806,
+      "loss": 2.8777,
+      "step": 44316
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6892098188400269,
+      "learning_rate": 0.00010591443830663447,
+      "loss": 2.7619,
+      "step": 44317
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7479181289672852,
+      "learning_rate": 0.00010590269849229497,
+      "loss": 2.9923,
+      "step": 44318
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6602400541305542,
+      "learning_rate": 0.00010589095918917032,
+      "loss": 3.0618,
+      "step": 44319
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6800590753555298,
+      "learning_rate": 0.00010587922039729162,
+      "loss": 2.837,
+      "step": 44320
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6915212869644165,
+      "learning_rate": 0.00010586748211668975,
+      "loss": 3.046,
+      "step": 44321
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6551668643951416,
+      "learning_rate": 0.00010585574434739542,
+      "loss": 2.8772,
+      "step": 44322
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7143257856369019,
+      "learning_rate": 0.00010584400708943979,
+      "loss": 2.9301,
+      "step": 44323
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6705986857414246,
+      "learning_rate": 0.00010583227034285357,
+      "loss": 2.8245,
+      "step": 44324
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6597686409950256,
+      "learning_rate": 0.00010582053410766792,
+      "loss": 2.9585,
+      "step": 44325
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7265187501907349,
+      "learning_rate": 0.00010580879838391349,
+      "loss": 2.9596,
+      "step": 44326
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7158097624778748,
+      "learning_rate": 0.00010579706317162145,
+      "loss": 2.7455,
+      "step": 44327
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7624508142471313,
+      "learning_rate": 0.0001057853284708225,
+      "loss": 2.8151,
+      "step": 44328
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7447002530097961,
+      "learning_rate": 0.00010577359428154759,
+      "loss": 2.5754,
+      "step": 44329
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6778799295425415,
+      "learning_rate": 0.00010576186060382771,
+      "loss": 2.855,
+      "step": 44330
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7103135585784912,
+      "learning_rate": 0.0001057501274376936,
+      "loss": 2.9367,
+      "step": 44331
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6705018281936646,
+      "learning_rate": 0.0001057383947831763,
+      "loss": 3.1099,
+      "step": 44332
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7252798676490784,
+      "learning_rate": 0.00010572666264030677,
+      "loss": 2.8609,
+      "step": 44333
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7287251353263855,
+      "learning_rate": 0.00010571493100911573,
+      "loss": 2.8493,
+      "step": 44334
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7141810059547424,
+      "learning_rate": 0.00010570319988963424,
+      "loss": 2.855,
+      "step": 44335
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7226595878601074,
+      "learning_rate": 0.00010569146928189316,
+      "loss": 2.6597,
+      "step": 44336
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7063793540000916,
+      "learning_rate": 0.00010567973918592334,
+      "loss": 3.0757,
+      "step": 44337
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7134846448898315,
+      "learning_rate": 0.0001056680096017556,
+      "loss": 2.8873,
+      "step": 44338
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7262305021286011,
+      "learning_rate": 0.00010565628052942091,
+      "loss": 3.1055,
+      "step": 44339
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.729193389415741,
+      "learning_rate": 0.00010564455196895029,
+      "loss": 2.9053,
+      "step": 44340
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7212259769439697,
+      "learning_rate": 0.00010563282392037443,
+      "loss": 2.9384,
+      "step": 44341
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7601874470710754,
+      "learning_rate": 0.00010562109638372436,
+      "loss": 2.7878,
+      "step": 44342
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6938914656639099,
+      "learning_rate": 0.00010560936935903093,
+      "loss": 3.0421,
+      "step": 44343
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7167030572891235,
+      "learning_rate": 0.00010559764284632493,
+      "loss": 3.0963,
+      "step": 44344
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7385327219963074,
+      "learning_rate": 0.0001055859168456374,
+      "loss": 3.0804,
+      "step": 44345
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7111912369728088,
+      "learning_rate": 0.00010557419135699906,
+      "loss": 3.0677,
+      "step": 44346
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7527398467063904,
+      "learning_rate": 0.000105562466380441,
+      "loss": 3.0916,
+      "step": 44347
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6765745282173157,
+      "learning_rate": 0.00010555074191599387,
+      "loss": 2.9669,
+      "step": 44348
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6760923266410828,
+      "learning_rate": 0.00010553901796368876,
+      "loss": 2.8139,
+      "step": 44349
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7211750149726868,
+      "learning_rate": 0.00010552729452355644,
+      "loss": 2.7989,
+      "step": 44350
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6986032128334045,
+      "learning_rate": 0.00010551557159562769,
+      "loss": 3.0181,
+      "step": 44351
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7039467692375183,
+      "learning_rate": 0.00010550384917993362,
+      "loss": 2.9076,
+      "step": 44352
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7151457667350769,
+      "learning_rate": 0.00010549212727650487,
+      "loss": 3.0128,
+      "step": 44353
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7260016202926636,
+      "learning_rate": 0.00010548040588537244,
+      "loss": 3.0107,
+      "step": 44354
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7048295140266418,
+      "learning_rate": 0.00010546868500656728,
+      "loss": 2.9732,
+      "step": 44355
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6705439686775208,
+      "learning_rate": 0.00010545696464012006,
+      "loss": 2.736,
+      "step": 44356
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6685308814048767,
+      "learning_rate": 0.00010544524478606185,
+      "loss": 3.0948,
+      "step": 44357
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7032225728034973,
+      "learning_rate": 0.00010543352544442342,
+      "loss": 2.8816,
+      "step": 44358
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7694846987724304,
+      "learning_rate": 0.00010542180661523555,
+      "loss": 3.0432,
+      "step": 44359
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6599303483963013,
+      "learning_rate": 0.00010541008829852929,
+      "loss": 2.9444,
+      "step": 44360
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7349076271057129,
+      "learning_rate": 0.00010539837049433531,
+      "loss": 2.981,
+      "step": 44361
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7268679141998291,
+      "learning_rate": 0.00010538665320268466,
+      "loss": 2.9341,
+      "step": 44362
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6871877312660217,
+      "learning_rate": 0.00010537493642360802,
+      "loss": 2.8963,
+      "step": 44363
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.669917643070221,
+      "learning_rate": 0.00010536322015713642,
+      "loss": 2.921,
+      "step": 44364
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7450718283653259,
+      "learning_rate": 0.00010535150440330068,
+      "loss": 2.6976,
+      "step": 44365
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6828457713127136,
+      "learning_rate": 0.00010533978916213148,
+      "loss": 2.8831,
+      "step": 44366
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7049870491027832,
+      "learning_rate": 0.00010532807443365994,
+      "loss": 2.9459,
+      "step": 44367
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6955314874649048,
+      "learning_rate": 0.00010531636021791664,
+      "loss": 2.8951,
+      "step": 44368
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6996759176254272,
+      "learning_rate": 0.00010530464651493263,
+      "loss": 2.9312,
+      "step": 44369
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7178577184677124,
+      "learning_rate": 0.00010529293332473879,
+      "loss": 2.7958,
+      "step": 44370
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6655074954032898,
+      "learning_rate": 0.00010528122064736591,
+      "loss": 2.9293,
+      "step": 44371
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6931262016296387,
+      "learning_rate": 0.00010526950848284482,
+      "loss": 2.9426,
+      "step": 44372
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6952087879180908,
+      "learning_rate": 0.00010525779683120621,
+      "loss": 2.921,
+      "step": 44373
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6865946054458618,
+      "learning_rate": 0.00010524608569248113,
+      "loss": 2.8322,
+      "step": 44374
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6755927801132202,
+      "learning_rate": 0.00010523437506670052,
+      "loss": 2.7113,
+      "step": 44375
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.704673171043396,
+      "learning_rate": 0.00010522266495389491,
+      "loss": 2.8718,
+      "step": 44376
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6965922713279724,
+      "learning_rate": 0.00010521095535409547,
+      "loss": 2.8379,
+      "step": 44377
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7056841850280762,
+      "learning_rate": 0.00010519924626733275,
+      "loss": 3.0354,
+      "step": 44378
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7823092937469482,
+      "learning_rate": 0.00010518753769363782,
+      "loss": 3.1314,
+      "step": 44379
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6921338438987732,
+      "learning_rate": 0.00010517582963304146,
+      "loss": 3.0036,
+      "step": 44380
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6734832525253296,
+      "learning_rate": 0.00010516412208557434,
+      "loss": 2.9865,
+      "step": 44381
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7287729382514954,
+      "learning_rate": 0.00010515241505126754,
+      "loss": 3.0678,
+      "step": 44382
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6677972674369812,
+      "learning_rate": 0.00010514070853015169,
+      "loss": 2.9314,
+      "step": 44383
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7038114070892334,
+      "learning_rate": 0.00010512900252225766,
+      "loss": 3.0285,
+      "step": 44384
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6778184175491333,
+      "learning_rate": 0.00010511729702761648,
+      "loss": 2.8647,
+      "step": 44385
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7425260543823242,
+      "learning_rate": 0.00010510559204625882,
+      "loss": 2.9822,
+      "step": 44386
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.720942497253418,
+      "learning_rate": 0.00010509388757821555,
+      "loss": 2.9104,
+      "step": 44387
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7403482794761658,
+      "learning_rate": 0.00010508218362351735,
+      "loss": 2.881,
+      "step": 44388
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6942797303199768,
+      "learning_rate": 0.00010507048018219515,
+      "loss": 3.1106,
+      "step": 44389
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7015599012374878,
+      "learning_rate": 0.00010505877725427991,
+      "loss": 2.8412,
+      "step": 44390
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7615519165992737,
+      "learning_rate": 0.00010504707483980221,
+      "loss": 2.9333,
+      "step": 44391
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.711105465888977,
+      "learning_rate": 0.00010503537293879313,
+      "loss": 3.13,
+      "step": 44392
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7053470015525818,
+      "learning_rate": 0.00010502367155128332,
+      "loss": 2.9103,
+      "step": 44393
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6969516277313232,
+      "learning_rate": 0.00010501197067730357,
+      "loss": 2.8205,
+      "step": 44394
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7566748857498169,
+      "learning_rate": 0.00010500027031688482,
+      "loss": 3.0093,
+      "step": 44395
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6897950172424316,
+      "learning_rate": 0.00010498857047005777,
+      "loss": 3.0615,
+      "step": 44396
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7079964280128479,
+      "learning_rate": 0.00010497687113685337,
+      "loss": 3.1859,
+      "step": 44397
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.72109454870224,
+      "learning_rate": 0.00010496517231730228,
+      "loss": 2.4967,
+      "step": 44398
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7014322876930237,
+      "learning_rate": 0.00010495347401143538,
+      "loss": 3.1057,
+      "step": 44399
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7019551396369934,
+      "learning_rate": 0.00010494177621928358,
+      "loss": 2.9177,
+      "step": 44400
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.721088171005249,
+      "learning_rate": 0.00010493007894087762,
+      "loss": 3.1077,
+      "step": 44401
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6936663389205933,
+      "learning_rate": 0.00010491838217624829,
+      "loss": 2.9189,
+      "step": 44402
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7555628418922424,
+      "learning_rate": 0.0001049066859254263,
+      "loss": 2.906,
+      "step": 44403
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6784551739692688,
+      "learning_rate": 0.00010489499018844253,
+      "loss": 2.9672,
+      "step": 44404
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7072699069976807,
+      "learning_rate": 0.00010488329496532791,
+      "loss": 2.966,
+      "step": 44405
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7025425434112549,
+      "learning_rate": 0.00010487160025611305,
+      "loss": 2.966,
+      "step": 44406
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7028836607933044,
+      "learning_rate": 0.00010485990606082895,
+      "loss": 2.991,
+      "step": 44407
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7300164103507996,
+      "learning_rate": 0.0001048482123795063,
+      "loss": 2.8152,
+      "step": 44408
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6703711748123169,
+      "learning_rate": 0.0001048365192121758,
+      "loss": 2.9551,
+      "step": 44409
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7038151621818542,
+      "learning_rate": 0.00010482482655886841,
+      "loss": 2.5322,
+      "step": 44410
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7217896580696106,
+      "learning_rate": 0.0001048131344196148,
+      "loss": 2.7781,
+      "step": 44411
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7087219953536987,
+      "learning_rate": 0.00010480144279444595,
+      "loss": 2.8338,
+      "step": 44412
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7262712121009827,
+      "learning_rate": 0.0001047897516833924,
+      "loss": 3.0148,
+      "step": 44413
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6909366846084595,
+      "learning_rate": 0.0001047780610864852,
+      "loss": 2.9738,
+      "step": 44414
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6847735047340393,
+      "learning_rate": 0.0001047663710037549,
+      "loss": 2.733,
+      "step": 44415
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7341592907905579,
+      "learning_rate": 0.00010475468143523248,
+      "loss": 2.8791,
+      "step": 44416
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7717563509941101,
+      "learning_rate": 0.0001047429923809487,
+      "loss": 2.9478,
+      "step": 44417
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6956798434257507,
+      "learning_rate": 0.00010473130384093417,
+      "loss": 2.9691,
+      "step": 44418
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7189202904701233,
+      "learning_rate": 0.00010471961581521993,
+      "loss": 3.0525,
+      "step": 44419
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6681233048439026,
+      "learning_rate": 0.0001047079283038365,
+      "loss": 2.8649,
+      "step": 44420
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7311338186264038,
+      "learning_rate": 0.00010469624130681485,
+      "loss": 3.015,
+      "step": 44421
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.722417950630188,
+      "learning_rate": 0.00010468455482418577,
+      "loss": 2.8841,
+      "step": 44422
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6731986403465271,
+      "learning_rate": 0.00010467286885598003,
+      "loss": 2.7704,
+      "step": 44423
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6961910724639893,
+      "learning_rate": 0.00010466118340222833,
+      "loss": 2.8648,
+      "step": 44424
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7090938687324524,
+      "learning_rate": 0.00010464949846296138,
+      "loss": 2.8517,
+      "step": 44425
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.690933108329773,
+      "learning_rate": 0.00010463781403821004,
+      "loss": 2.8032,
+      "step": 44426
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6868634819984436,
+      "learning_rate": 0.00010462613012800524,
+      "loss": 3.0002,
+      "step": 44427
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6949028968811035,
+      "learning_rate": 0.00010461444673237747,
+      "loss": 2.784,
+      "step": 44428
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7235555648803711,
+      "learning_rate": 0.00010460276385135778,
+      "loss": 2.958,
+      "step": 44429
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7306112051010132,
+      "learning_rate": 0.00010459108148497681,
+      "loss": 2.9012,
+      "step": 44430
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7331579327583313,
+      "learning_rate": 0.0001045793996332652,
+      "loss": 2.8728,
+      "step": 44431
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7387495636940002,
+      "learning_rate": 0.00010456771829625396,
+      "loss": 2.9061,
+      "step": 44432
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6881783604621887,
+      "learning_rate": 0.0001045560374739736,
+      "loss": 2.9365,
+      "step": 44433
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7136433720588684,
+      "learning_rate": 0.00010454435716645518,
+      "loss": 2.8599,
+      "step": 44434
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6872485280036926,
+      "learning_rate": 0.00010453267737372915,
+      "loss": 2.7754,
+      "step": 44435
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7080912590026855,
+      "learning_rate": 0.00010452099809582646,
+      "loss": 3.0945,
+      "step": 44436
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7158118486404419,
+      "learning_rate": 0.00010450931933277795,
+      "loss": 2.9047,
+      "step": 44437
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6851445436477661,
+      "learning_rate": 0.00010449764108461425,
+      "loss": 2.8758,
+      "step": 44438
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7490003705024719,
+      "learning_rate": 0.00010448596335136615,
+      "loss": 2.8528,
+      "step": 44439
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7051994800567627,
+      "learning_rate": 0.00010447428613306428,
+      "loss": 3.0044,
+      "step": 44440
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7388154864311218,
+      "learning_rate": 0.00010446260942973953,
+      "loss": 3.0324,
+      "step": 44441
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7008410096168518,
+      "learning_rate": 0.00010445093324142272,
+      "loss": 3.035,
+      "step": 44442
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6961836218833923,
+      "learning_rate": 0.00010443925756814442,
+      "loss": 3.0496,
+      "step": 44443
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7032221555709839,
+      "learning_rate": 0.00010442758240993557,
+      "loss": 2.9266,
+      "step": 44444
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7082571387290955,
+      "learning_rate": 0.00010441590776682686,
+      "loss": 3.0056,
+      "step": 44445
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7039412260055542,
+      "learning_rate": 0.00010440423363884888,
+      "loss": 2.8115,
+      "step": 44446
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6998090744018555,
+      "learning_rate": 0.00010439256002603261,
+      "loss": 3.0467,
+      "step": 44447
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7542231678962708,
+      "learning_rate": 0.0001043808869284086,
+      "loss": 2.9567,
+      "step": 44448
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7332549095153809,
+      "learning_rate": 0.00010436921434600778,
+      "loss": 2.8982,
+      "step": 44449
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6957454681396484,
+      "learning_rate": 0.00010435754227886072,
+      "loss": 2.8315,
+      "step": 44450
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.715060293674469,
+      "learning_rate": 0.00010434587072699832,
+      "loss": 2.8012,
+      "step": 44451
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7155557870864868,
+      "learning_rate": 0.00010433419969045122,
+      "loss": 3.2461,
+      "step": 44452
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7036327123641968,
+      "learning_rate": 0.00010432252916925012,
+      "loss": 3.0977,
+      "step": 44453
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7026028037071228,
+      "learning_rate": 0.00010431085916342589,
+      "loss": 2.854,
+      "step": 44454
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7156589031219482,
+      "learning_rate": 0.00010429918967300912,
+      "loss": 2.8383,
+      "step": 44455
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7051274180412292,
+      "learning_rate": 0.00010428752069803063,
+      "loss": 2.9757,
+      "step": 44456
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6463907957077026,
+      "learning_rate": 0.00010427585223852122,
+      "loss": 2.9478,
+      "step": 44457
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7040029764175415,
+      "learning_rate": 0.00010426418429451144,
+      "loss": 2.6731,
+      "step": 44458
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7455715537071228,
+      "learning_rate": 0.00010425251686603225,
+      "loss": 2.8721,
+      "step": 44459
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7634180784225464,
+      "learning_rate": 0.00010424084995311426,
+      "loss": 3.066,
+      "step": 44460
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7320479154586792,
+      "learning_rate": 0.0001042291835557881,
+      "loss": 2.9601,
+      "step": 44461
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7387556433677673,
+      "learning_rate": 0.00010421751767408468,
+      "loss": 2.9843,
+      "step": 44462
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7048659324645996,
+      "learning_rate": 0.00010420585230803452,
+      "loss": 2.8313,
+      "step": 44463
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7122004628181458,
+      "learning_rate": 0.00010419418745766858,
+      "loss": 2.9771,
+      "step": 44464
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7229511141777039,
+      "learning_rate": 0.0001041825231230174,
+      "loss": 2.9935,
+      "step": 44465
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7182523608207703,
+      "learning_rate": 0.00010417085930411185,
+      "loss": 2.9483,
+      "step": 44466
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7179908156394958,
+      "learning_rate": 0.00010415919600098255,
+      "loss": 2.85,
+      "step": 44467
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7596323490142822,
+      "learning_rate": 0.00010414753321366015,
+      "loss": 2.7197,
+      "step": 44468
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7082833051681519,
+      "learning_rate": 0.00010413587094217552,
+      "loss": 2.9153,
+      "step": 44469
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6668033003807068,
+      "learning_rate": 0.00010412420918655925,
+      "loss": 3.0079,
+      "step": 44470
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7020968794822693,
+      "learning_rate": 0.00010411254794684212,
+      "loss": 2.8986,
+      "step": 44471
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6882535815238953,
+      "learning_rate": 0.00010410088722305493,
+      "loss": 2.9372,
+      "step": 44472
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7956342101097107,
+      "learning_rate": 0.0001040892270152283,
+      "loss": 3.0606,
+      "step": 44473
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7361034750938416,
+      "learning_rate": 0.00010407756732339293,
+      "loss": 2.9101,
+      "step": 44474
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7044380903244019,
+      "learning_rate": 0.00010406590814757946,
+      "loss": 2.798,
+      "step": 44475
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6986120343208313,
+      "learning_rate": 0.00010405424948781869,
+      "loss": 2.7251,
+      "step": 44476
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.712506890296936,
+      "learning_rate": 0.0001040425913441414,
+      "loss": 2.9249,
+      "step": 44477
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7093831896781921,
+      "learning_rate": 0.00010403093371657807,
+      "loss": 2.7835,
+      "step": 44478
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7020789384841919,
+      "learning_rate": 0.00010401927660515972,
+      "loss": 2.7917,
+      "step": 44479
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.70029616355896,
+      "learning_rate": 0.00010400762000991675,
+      "loss": 3.0307,
+      "step": 44480
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7338650226593018,
+      "learning_rate": 0.00010399596393088007,
+      "loss": 3.0209,
+      "step": 44481
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.700697124004364,
+      "learning_rate": 0.00010398430836808032,
+      "loss": 2.859,
+      "step": 44482
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6929804086685181,
+      "learning_rate": 0.00010397265332154807,
+      "loss": 2.814,
+      "step": 44483
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6898903250694275,
+      "learning_rate": 0.00010396099879131422,
+      "loss": 2.8862,
+      "step": 44484
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7237740755081177,
+      "learning_rate": 0.00010394934477740928,
+      "loss": 3.033,
+      "step": 44485
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6766157150268555,
+      "learning_rate": 0.00010393769127986403,
+      "loss": 2.9624,
+      "step": 44486
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7159266471862793,
+      "learning_rate": 0.00010392603829870928,
+      "loss": 2.8887,
+      "step": 44487
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7595884799957275,
+      "learning_rate": 0.0001039143858339756,
+      "loss": 3.0751,
+      "step": 44488
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6963126063346863,
+      "learning_rate": 0.00010390273388569367,
+      "loss": 2.8867,
+      "step": 44489
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.810042142868042,
+      "learning_rate": 0.00010389108245389412,
+      "loss": 2.8458,
+      "step": 44490
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7496387362480164,
+      "learning_rate": 0.00010387943153860781,
+      "loss": 2.8515,
+      "step": 44491
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7184833884239197,
+      "learning_rate": 0.00010386778113986525,
+      "loss": 2.8776,
+      "step": 44492
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7305948734283447,
+      "learning_rate": 0.00010385613125769719,
+      "loss": 2.8932,
+      "step": 44493
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7090479135513306,
+      "learning_rate": 0.00010384448189213446,
+      "loss": 2.8305,
+      "step": 44494
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7010915279388428,
+      "learning_rate": 0.00010383283304320758,
+      "loss": 2.9246,
+      "step": 44495
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7132809162139893,
+      "learning_rate": 0.0001038211847109473,
+      "loss": 2.831,
+      "step": 44496
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7011138796806335,
+      "learning_rate": 0.00010380953689538413,
+      "loss": 2.9277,
+      "step": 44497
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7759160995483398,
+      "learning_rate": 0.00010379788959654889,
+      "loss": 2.8512,
+      "step": 44498
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7896947264671326,
+      "learning_rate": 0.00010378624281447236,
+      "loss": 2.8438,
+      "step": 44499
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7089138627052307,
+      "learning_rate": 0.000103774596549185,
+      "loss": 2.9216,
+      "step": 44500
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7258244752883911,
+      "learning_rate": 0.00010376295080071769,
+      "loss": 2.8647,
+      "step": 44501
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7142013311386108,
+      "learning_rate": 0.00010375130556910088,
+      "loss": 2.9314,
+      "step": 44502
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7016758322715759,
+      "learning_rate": 0.00010373966085436548,
+      "loss": 3.0253,
+      "step": 44503
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7166877388954163,
+      "learning_rate": 0.00010372801665654205,
+      "loss": 2.858,
+      "step": 44504
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6986833214759827,
+      "learning_rate": 0.00010371637297566112,
+      "loss": 3.0418,
+      "step": 44505
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7260124683380127,
+      "learning_rate": 0.00010370472981175361,
+      "loss": 2.9055,
+      "step": 44506
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7071568369865417,
+      "learning_rate": 0.00010369308716484999,
+      "loss": 2.8121,
+      "step": 44507
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6688234210014343,
+      "learning_rate": 0.00010368144503498097,
+      "loss": 3.0533,
+      "step": 44508
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6707281470298767,
+      "learning_rate": 0.00010366980342217734,
+      "loss": 3.061,
+      "step": 44509
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7021904587745667,
+      "learning_rate": 0.00010365816232646965,
+      "loss": 2.7661,
+      "step": 44510
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7200233936309814,
+      "learning_rate": 0.0001036465217478886,
+      "loss": 3.0572,
+      "step": 44511
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7074561715126038,
+      "learning_rate": 0.00010363488168646473,
+      "loss": 3.0943,
+      "step": 44512
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7289097309112549,
+      "learning_rate": 0.00010362324214222876,
+      "loss": 2.9338,
+      "step": 44513
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7148504257202148,
+      "learning_rate": 0.00010361160311521148,
+      "loss": 3.0845,
+      "step": 44514
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6821744441986084,
+      "learning_rate": 0.00010359996460544335,
+      "loss": 3.1103,
+      "step": 44515
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6762775182723999,
+      "learning_rate": 0.00010358832661295524,
+      "loss": 2.9369,
+      "step": 44516
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7053170204162598,
+      "learning_rate": 0.00010357668913777756,
+      "loss": 2.8441,
+      "step": 44517
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7255358099937439,
+      "learning_rate": 0.00010356505217994118,
+      "loss": 2.8484,
+      "step": 44518
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6944178938865662,
+      "learning_rate": 0.00010355341573947665,
+      "loss": 2.8011,
+      "step": 44519
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.722078800201416,
+      "learning_rate": 0.00010354177981641449,
+      "loss": 2.8922,
+      "step": 44520
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7471203207969666,
+      "learning_rate": 0.00010353014441078558,
+      "loss": 2.8489,
+      "step": 44521
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6949812173843384,
+      "learning_rate": 0.00010351850952262036,
+      "loss": 2.863,
+      "step": 44522
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7270025610923767,
+      "learning_rate": 0.0001035068751519496,
+      "loss": 3.0245,
+      "step": 44523
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7315390110015869,
+      "learning_rate": 0.00010349524129880399,
+      "loss": 2.8959,
+      "step": 44524
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.747600793838501,
+      "learning_rate": 0.00010348360796321411,
+      "loss": 2.7812,
+      "step": 44525
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7093079686164856,
+      "learning_rate": 0.00010347197514521058,
+      "loss": 2.6684,
+      "step": 44526
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7230992317199707,
+      "learning_rate": 0.00010346034284482392,
+      "loss": 2.9073,
+      "step": 44527
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6705207824707031,
+      "learning_rate": 0.00010344871106208492,
+      "loss": 2.9985,
+      "step": 44528
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6426035761833191,
+      "learning_rate": 0.00010343707979702429,
+      "loss": 2.997,
+      "step": 44529
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7323707938194275,
+      "learning_rate": 0.00010342544904967247,
+      "loss": 2.7661,
+      "step": 44530
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7027085423469543,
+      "learning_rate": 0.00010341381882006027,
+      "loss": 2.7511,
+      "step": 44531
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6900554299354553,
+      "learning_rate": 0.00010340218910821823,
+      "loss": 2.9398,
+      "step": 44532
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7302149534225464,
+      "learning_rate": 0.00010339055991417688,
+      "loss": 3.0751,
+      "step": 44533
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7446348071098328,
+      "learning_rate": 0.0001033789312379671,
+      "loss": 2.897,
+      "step": 44534
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7004778385162354,
+      "learning_rate": 0.00010336730307961925,
+      "loss": 2.9385,
+      "step": 44535
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6848006844520569,
+      "learning_rate": 0.00010335567543916417,
+      "loss": 3.0406,
+      "step": 44536
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6743120551109314,
+      "learning_rate": 0.00010334404831663232,
+      "loss": 2.8368,
+      "step": 44537
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6830146908760071,
+      "learning_rate": 0.00010333242171205436,
+      "loss": 2.7874,
+      "step": 44538
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6831412315368652,
+      "learning_rate": 0.00010332079562546108,
+      "loss": 2.5951,
+      "step": 44539
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7036148309707642,
+      "learning_rate": 0.00010330917005688297,
+      "loss": 2.7791,
+      "step": 44540
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7623158693313599,
+      "learning_rate": 0.00010329754500635067,
+      "loss": 2.9707,
+      "step": 44541
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6715494394302368,
+      "learning_rate": 0.00010328592047389464,
+      "loss": 3.103,
+      "step": 44542
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7432985305786133,
+      "learning_rate": 0.00010327429645954567,
+      "loss": 2.717,
+      "step": 44543
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7097046375274658,
+      "learning_rate": 0.00010326267296333445,
+      "loss": 3.0353,
+      "step": 44544
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7098416090011597,
+      "learning_rate": 0.00010325104998529136,
+      "loss": 2.8349,
+      "step": 44545
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.719925582408905,
+      "learning_rate": 0.00010323942752544725,
+      "loss": 2.8611,
+      "step": 44546
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7523378133773804,
+      "learning_rate": 0.0001032278055838326,
+      "loss": 2.8775,
+      "step": 44547
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7218157649040222,
+      "learning_rate": 0.00010321618416047796,
+      "loss": 2.8731,
+      "step": 44548
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6982775926589966,
+      "learning_rate": 0.00010320456325541412,
+      "loss": 2.6777,
+      "step": 44549
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7226266264915466,
+      "learning_rate": 0.0001031929428686715,
+      "loss": 3.0359,
+      "step": 44550
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7298018932342529,
+      "learning_rate": 0.00010318132300028087,
+      "loss": 2.7433,
+      "step": 44551
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.725264310836792,
+      "learning_rate": 0.00010316970365027267,
+      "loss": 2.873,
+      "step": 44552
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.70024573802948,
+      "learning_rate": 0.00010315808481867765,
+      "loss": 3.1367,
+      "step": 44553
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6710588335990906,
+      "learning_rate": 0.0001031464665055264,
+      "loss": 2.8933,
+      "step": 44554
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7196180820465088,
+      "learning_rate": 0.00010313484871084935,
+      "loss": 2.9405,
+      "step": 44555
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6892144083976746,
+      "learning_rate": 0.00010312323143467734,
+      "loss": 2.879,
+      "step": 44556
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7082641124725342,
+      "learning_rate": 0.00010311161467704073,
+      "loss": 2.9238,
+      "step": 44557
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.709845244884491,
+      "learning_rate": 0.00010309999843797025,
+      "loss": 2.761,
+      "step": 44558
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7715518474578857,
+      "learning_rate": 0.00010308838271749657,
+      "loss": 2.8075,
+      "step": 44559
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6767800450325012,
+      "learning_rate": 0.0001030767675156501,
+      "loss": 2.8401,
+      "step": 44560
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.74344402551651,
+      "learning_rate": 0.00010306515283246161,
+      "loss": 2.967,
+      "step": 44561
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7046286463737488,
+      "learning_rate": 0.00010305353866796163,
+      "loss": 3.0008,
+      "step": 44562
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6904906034469604,
+      "learning_rate": 0.00010304192502218059,
+      "loss": 2.8009,
+      "step": 44563
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7006396055221558,
+      "learning_rate": 0.00010303031189514934,
+      "loss": 2.9043,
+      "step": 44564
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6934382915496826,
+      "learning_rate": 0.00010301869928689824,
+      "loss": 2.9064,
+      "step": 44565
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7191388010978699,
+      "learning_rate": 0.00010300708719745807,
+      "loss": 2.9023,
+      "step": 44566
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6918665170669556,
+      "learning_rate": 0.00010299547562685924,
+      "loss": 2.9055,
+      "step": 44567
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7260489463806152,
+      "learning_rate": 0.00010298386457513247,
+      "loss": 3.0175,
+      "step": 44568
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.688124418258667,
+      "learning_rate": 0.00010297225404230834,
+      "loss": 2.9031,
+      "step": 44569
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6864445805549622,
+      "learning_rate": 0.00010296064402841724,
+      "loss": 2.9554,
+      "step": 44570
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6903675198554993,
+      "learning_rate": 0.00010294903453349,
+      "loss": 2.8524,
+      "step": 44571
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6836926937103271,
+      "learning_rate": 0.00010293742555755695,
+      "loss": 3.106,
+      "step": 44572
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6697653532028198,
+      "learning_rate": 0.00010292581710064892,
+      "loss": 2.6941,
+      "step": 44573
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7098059058189392,
+      "learning_rate": 0.00010291420916279625,
+      "loss": 2.8763,
+      "step": 44574
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6917982697486877,
+      "learning_rate": 0.00010290260174402974,
+      "loss": 2.9143,
+      "step": 44575
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7418469786643982,
+      "learning_rate": 0.00010289099484437982,
+      "loss": 2.8845,
+      "step": 44576
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6864647269248962,
+      "learning_rate": 0.00010287938846387701,
+      "loss": 2.9388,
+      "step": 44577
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6718887686729431,
+      "learning_rate": 0.00010286778260255203,
+      "loss": 3.0018,
+      "step": 44578
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7265290021896362,
+      "learning_rate": 0.00010285617726043529,
+      "loss": 2.9683,
+      "step": 44579
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6988957524299622,
+      "learning_rate": 0.00010284457243755743,
+      "loss": 2.8479,
+      "step": 44580
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7348703145980835,
+      "learning_rate": 0.00010283296813394911,
+      "loss": 2.7909,
+      "step": 44581
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7384020090103149,
+      "learning_rate": 0.00010282136434964072,
+      "loss": 2.8118,
+      "step": 44582
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7310593724250793,
+      "learning_rate": 0.00010280976108466301,
+      "loss": 2.8275,
+      "step": 44583
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7117740511894226,
+      "learning_rate": 0.00010279815833904645,
+      "loss": 2.8532,
+      "step": 44584
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6952208280563354,
+      "learning_rate": 0.0001027865561128215,
+      "loss": 2.6818,
+      "step": 44585
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6998794078826904,
+      "learning_rate": 0.00010277495440601887,
+      "loss": 2.7961,
+      "step": 44586
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.71541827917099,
+      "learning_rate": 0.00010276335321866896,
+      "loss": 3.0561,
+      "step": 44587
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7112535834312439,
+      "learning_rate": 0.00010275175255080251,
+      "loss": 2.9963,
+      "step": 44588
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7849178314208984,
+      "learning_rate": 0.00010274015240244992,
+      "loss": 2.8191,
+      "step": 44589
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7260420918464661,
+      "learning_rate": 0.00010272855277364188,
+      "loss": 2.6095,
+      "step": 44590
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6618551015853882,
+      "learning_rate": 0.00010271695366440885,
+      "loss": 2.9215,
+      "step": 44591
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6712572574615479,
+      "learning_rate": 0.00010270535507478135,
+      "loss": 2.9213,
+      "step": 44592
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7594893574714661,
+      "learning_rate": 0.00010269375700479002,
+      "loss": 2.7903,
+      "step": 44593
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7108433246612549,
+      "learning_rate": 0.00010268215945446529,
+      "loss": 2.9075,
+      "step": 44594
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.725123941898346,
+      "learning_rate": 0.00010267056242383777,
+      "loss": 2.88,
+      "step": 44595
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6654282808303833,
+      "learning_rate": 0.0001026589659129381,
+      "loss": 2.8843,
+      "step": 44596
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.751349687576294,
+      "learning_rate": 0.00010264736992179675,
+      "loss": 3.0375,
+      "step": 44597
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7495384216308594,
+      "learning_rate": 0.00010263577445044425,
+      "loss": 2.9641,
+      "step": 44598
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7475329041481018,
+      "learning_rate": 0.00010262417949891104,
+      "loss": 2.908,
+      "step": 44599
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6864625811576843,
+      "learning_rate": 0.00010261258506722772,
+      "loss": 2.9052,
+      "step": 44600
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6639381647109985,
+      "learning_rate": 0.00010260099115542499,
+      "loss": 2.8416,
+      "step": 44601
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7066051959991455,
+      "learning_rate": 0.00010258939776353316,
+      "loss": 2.7455,
+      "step": 44602
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6839228868484497,
+      "learning_rate": 0.00010257780489158295,
+      "loss": 3.0094,
+      "step": 44603
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7170974016189575,
+      "learning_rate": 0.00010256621253960474,
+      "loss": 2.9012,
+      "step": 44604
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7041524052619934,
+      "learning_rate": 0.00010255462070762921,
+      "loss": 3.0658,
+      "step": 44605
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6791936159133911,
+      "learning_rate": 0.0001025430293956868,
+      "loss": 2.9157,
+      "step": 44606
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7103440165519714,
+      "learning_rate": 0.00010253143860380793,
+      "loss": 2.9272,
+      "step": 44607
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6542874574661255,
+      "learning_rate": 0.00010251984833202337,
+      "loss": 2.9294,
+      "step": 44608
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7594596147537231,
+      "learning_rate": 0.00010250825858036346,
+      "loss": 2.8874,
+      "step": 44609
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7474169731140137,
+      "learning_rate": 0.00010249666934885874,
+      "loss": 2.8206,
+      "step": 44610
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7913607954978943,
+      "learning_rate": 0.0001024850806375399,
+      "loss": 2.8971,
+      "step": 44611
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6927233934402466,
+      "learning_rate": 0.00010247349244643734,
+      "loss": 2.7587,
+      "step": 44612
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6824236512184143,
+      "learning_rate": 0.0001024619047755816,
+      "loss": 2.9516,
+      "step": 44613
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7336830496788025,
+      "learning_rate": 0.00010245031762500306,
+      "loss": 2.9245,
+      "step": 44614
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7069776654243469,
+      "learning_rate": 0.0001024387309947324,
+      "loss": 2.7558,
+      "step": 44615
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6687659621238708,
+      "learning_rate": 0.00010242714488480018,
+      "loss": 2.8424,
+      "step": 44616
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7121585607528687,
+      "learning_rate": 0.00010241555929523673,
+      "loss": 2.8238,
+      "step": 44617
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6747063398361206,
+      "learning_rate": 0.00010240397422607277,
+      "loss": 3.1629,
+      "step": 44618
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7251124978065491,
+      "learning_rate": 0.00010239238967733865,
+      "loss": 2.8919,
+      "step": 44619
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6380995512008667,
+      "learning_rate": 0.000102380805649065,
+      "loss": 2.9561,
+      "step": 44620
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7634221315383911,
+      "learning_rate": 0.0001023692221412823,
+      "loss": 2.9855,
+      "step": 44621
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6780490875244141,
+      "learning_rate": 0.00010235763915402094,
+      "loss": 2.9611,
+      "step": 44622
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.715461254119873,
+      "learning_rate": 0.0001023460566873116,
+      "loss": 2.8342,
+      "step": 44623
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6985108852386475,
+      "learning_rate": 0.0001023344747411846,
+      "loss": 2.9235,
+      "step": 44624
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7491456866264343,
+      "learning_rate": 0.00010232289331567057,
+      "loss": 2.9031,
+      "step": 44625
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.709112823009491,
+      "learning_rate": 0.00010231131241080009,
+      "loss": 2.8829,
+      "step": 44626
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6923708319664001,
+      "learning_rate": 0.00010229973202660358,
+      "loss": 3.0084,
+      "step": 44627
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7466208934783936,
+      "learning_rate": 0.00010228815216311152,
+      "loss": 2.9558,
+      "step": 44628
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7512304782867432,
+      "learning_rate": 0.00010227657282035431,
+      "loss": 2.763,
+      "step": 44629
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7179058790206909,
+      "learning_rate": 0.00010226499399836256,
+      "loss": 2.9326,
+      "step": 44630
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7273489236831665,
+      "learning_rate": 0.00010225341569716686,
+      "loss": 2.8466,
+      "step": 44631
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6925496459007263,
+      "learning_rate": 0.00010224183791679751,
+      "loss": 3.0426,
+      "step": 44632
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7936264276504517,
+      "learning_rate": 0.00010223026065728517,
+      "loss": 2.8253,
+      "step": 44633
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7013174295425415,
+      "learning_rate": 0.00010221868391866031,
+      "loss": 2.9166,
+      "step": 44634
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7175893187522888,
+      "learning_rate": 0.00010220710770095326,
+      "loss": 2.8768,
+      "step": 44635
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6898190975189209,
+      "learning_rate": 0.00010219553200419469,
+      "loss": 2.7969,
+      "step": 44636
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7262145280838013,
+      "learning_rate": 0.00010218395682841494,
+      "loss": 2.6368,
+      "step": 44637
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.693634033203125,
+      "learning_rate": 0.00010217238217364471,
+      "loss": 3.0816,
+      "step": 44638
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7244294285774231,
+      "learning_rate": 0.0001021608080399142,
+      "loss": 2.8939,
+      "step": 44639
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6989253163337708,
+      "learning_rate": 0.00010214923442725407,
+      "loss": 2.8912,
+      "step": 44640
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.694697380065918,
+      "learning_rate": 0.00010213766133569488,
+      "loss": 2.85,
+      "step": 44641
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7070184350013733,
+      "learning_rate": 0.000102126088765267,
+      "loss": 3.0055,
+      "step": 44642
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7489998936653137,
+      "learning_rate": 0.00010211451671600095,
+      "loss": 2.6904,
+      "step": 44643
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7626986503601074,
+      "learning_rate": 0.00010210294518792708,
+      "loss": 2.8428,
+      "step": 44644
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6796693801879883,
+      "learning_rate": 0.00010209137418107604,
+      "loss": 3.1733,
+      "step": 44645
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7073202133178711,
+      "learning_rate": 0.00010207980369547814,
+      "loss": 3.0606,
+      "step": 44646
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7209928035736084,
+      "learning_rate": 0.00010206823373116391,
+      "loss": 2.6608,
+      "step": 44647
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6928178668022156,
+      "learning_rate": 0.00010205666428816403,
+      "loss": 2.8192,
+      "step": 44648
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.8096620440483093,
+      "learning_rate": 0.00010204509536650876,
+      "loss": 2.9148,
+      "step": 44649
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.75314861536026,
+      "learning_rate": 0.00010203352696622862,
+      "loss": 2.6021,
+      "step": 44650
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7307797074317932,
+      "learning_rate": 0.00010202195908735396,
+      "loss": 2.9043,
+      "step": 44651
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7347580790519714,
+      "learning_rate": 0.00010201039172991536,
+      "loss": 2.7061,
+      "step": 44652
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7171032428741455,
+      "learning_rate": 0.00010199882489394339,
+      "loss": 3.0612,
+      "step": 44653
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7112158536911011,
+      "learning_rate": 0.0001019872585794683,
+      "loss": 2.8196,
+      "step": 44654
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6790067553520203,
+      "learning_rate": 0.0001019756927865208,
+      "loss": 2.8819,
+      "step": 44655
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7215171456336975,
+      "learning_rate": 0.00010196412751513117,
+      "loss": 2.9303,
+      "step": 44656
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7309430241584778,
+      "learning_rate": 0.00010195256276532984,
+      "loss": 3.0757,
+      "step": 44657
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7545703649520874,
+      "learning_rate": 0.00010194099853714742,
+      "loss": 2.9345,
+      "step": 44658
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7092612981796265,
+      "learning_rate": 0.00010192943483061423,
+      "loss": 2.8953,
+      "step": 44659
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6743860840797424,
+      "learning_rate": 0.00010191787164576088,
+      "loss": 3.0049,
+      "step": 44660
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.676878035068512,
+      "learning_rate": 0.00010190630898261759,
+      "loss": 2.8814,
+      "step": 44661
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7018594741821289,
+      "learning_rate": 0.000101894746841215,
+      "loss": 2.926,
+      "step": 44662
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6919565200805664,
+      "learning_rate": 0.00010188318522158362,
+      "loss": 3.0665,
+      "step": 44663
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7357789278030396,
+      "learning_rate": 0.00010187162412375378,
+      "loss": 2.9906,
+      "step": 44664
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7537810802459717,
+      "learning_rate": 0.00010186006354775598,
+      "loss": 2.9191,
+      "step": 44665
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7341231107711792,
+      "learning_rate": 0.00010184850349362051,
+      "loss": 3.0268,
+      "step": 44666
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6830435991287231,
+      "learning_rate": 0.00010183694396137797,
+      "loss": 2.8872,
+      "step": 44667
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7135305404663086,
+      "learning_rate": 0.00010182538495105886,
+      "loss": 2.8633,
+      "step": 44668
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7441238164901733,
+      "learning_rate": 0.00010181382646269347,
+      "loss": 2.9731,
+      "step": 44669
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7147581577301025,
+      "learning_rate": 0.00010180226849631244,
+      "loss": 2.7636,
+      "step": 44670
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7436352968215942,
+      "learning_rate": 0.00010179071105194603,
+      "loss": 2.9936,
+      "step": 44671
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7712357640266418,
+      "learning_rate": 0.0001017791541296247,
+      "loss": 2.7682,
+      "step": 44672
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6808838248252869,
+      "learning_rate": 0.00010176759772937898,
+      "loss": 2.7931,
+      "step": 44673
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6870922446250916,
+      "learning_rate": 0.0001017560418512392,
+      "loss": 2.9253,
+      "step": 44674
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7315462827682495,
+      "learning_rate": 0.0001017444864952359,
+      "loss": 3.0074,
+      "step": 44675
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7336083054542542,
+      "learning_rate": 0.00010173293166139942,
+      "loss": 2.8301,
+      "step": 44676
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7012789845466614,
+      "learning_rate": 0.00010172137734976032,
+      "loss": 2.7756,
+      "step": 44677
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.704544723033905,
+      "learning_rate": 0.00010170982356034896,
+      "loss": 2.9755,
+      "step": 44678
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6943780183792114,
+      "learning_rate": 0.00010169827029319562,
+      "loss": 2.9111,
+      "step": 44679
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7300577759742737,
+      "learning_rate": 0.00010168671754833104,
+      "loss": 2.8051,
+      "step": 44680
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7285476922988892,
+      "learning_rate": 0.00010167516532578535,
+      "loss": 2.6524,
+      "step": 44681
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7168903946876526,
+      "learning_rate": 0.00010166361362558908,
+      "loss": 2.9434,
+      "step": 44682
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6996946930885315,
+      "learning_rate": 0.00010165206244777282,
+      "loss": 2.8507,
+      "step": 44683
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6721283197402954,
+      "learning_rate": 0.00010164051179236674,
+      "loss": 2.8241,
+      "step": 44684
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6588842272758484,
+      "learning_rate": 0.00010162896165940148,
+      "loss": 2.8318,
+      "step": 44685
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6902130842208862,
+      "learning_rate": 0.00010161741204890734,
+      "loss": 2.8502,
+      "step": 44686
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6781286001205444,
+      "learning_rate": 0.00010160586296091468,
+      "loss": 3.0355,
+      "step": 44687
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7113543748855591,
+      "learning_rate": 0.00010159431439545409,
+      "loss": 2.9687,
+      "step": 44688
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.737820029258728,
+      "learning_rate": 0.00010158276635255579,
+      "loss": 2.955,
+      "step": 44689
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7128449082374573,
+      "learning_rate": 0.0001015712188322504,
+      "loss": 2.8879,
+      "step": 44690
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6655691862106323,
+      "learning_rate": 0.0001015596718345681,
+      "loss": 2.86,
+      "step": 44691
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7450001239776611,
+      "learning_rate": 0.00010154812535953957,
+      "loss": 2.9949,
+      "step": 44692
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6946448087692261,
+      "learning_rate": 0.0001015365794071951,
+      "loss": 3.2081,
+      "step": 44693
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6991720199584961,
+      "learning_rate": 0.00010152503397756493,
+      "loss": 2.9359,
+      "step": 44694
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7270272970199585,
+      "learning_rate": 0.00010151348907067978,
+      "loss": 2.7,
+      "step": 44695
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6899583339691162,
+      "learning_rate": 0.00010150194468656975,
+      "loss": 3.1077,
+      "step": 44696
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7620866894721985,
+      "learning_rate": 0.00010149040082526541,
+      "loss": 2.9878,
+      "step": 44697
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6977541446685791,
+      "learning_rate": 0.0001014788574867973,
+      "loss": 2.9698,
+      "step": 44698
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7355969548225403,
+      "learning_rate": 0.00010146731467119562,
+      "loss": 2.9087,
+      "step": 44699
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6769136190414429,
+      "learning_rate": 0.00010145577237849085,
+      "loss": 2.988,
+      "step": 44700
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6845611333847046,
+      "learning_rate": 0.00010144423060871322,
+      "loss": 2.7466,
+      "step": 44701
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6977527141571045,
+      "learning_rate": 0.00010143268936189332,
+      "loss": 2.8652,
+      "step": 44702
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7107285261154175,
+      "learning_rate": 0.00010142114863806158,
+      "loss": 3.1307,
+      "step": 44703
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7218692898750305,
+      "learning_rate": 0.00010140960843724822,
+      "loss": 2.9057,
+      "step": 44704
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6744379997253418,
+      "learning_rate": 0.00010139806875948385,
+      "loss": 2.8445,
+      "step": 44705
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7107676267623901,
+      "learning_rate": 0.00010138652960479864,
+      "loss": 3.0127,
+      "step": 44706
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7380881309509277,
+      "learning_rate": 0.00010137499097322315,
+      "loss": 2.743,
+      "step": 44707
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6708388924598694,
+      "learning_rate": 0.00010136345286478777,
+      "loss": 2.9925,
+      "step": 44708
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7307834625244141,
+      "learning_rate": 0.0001013519152795227,
+      "loss": 2.9303,
+      "step": 44709
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7044564485549927,
+      "learning_rate": 0.00010134037821745853,
+      "loss": 2.8851,
+      "step": 44710
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7049198746681213,
+      "learning_rate": 0.00010132884167862551,
+      "loss": 2.8379,
+      "step": 44711
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7499628067016602,
+      "learning_rate": 0.00010131730566305404,
+      "loss": 2.8606,
+      "step": 44712
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7331987023353577,
+      "learning_rate": 0.00010130577017077468,
+      "loss": 3.0568,
+      "step": 44713
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7273868322372437,
+      "learning_rate": 0.00010129423520181768,
+      "loss": 2.8243,
+      "step": 44714
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7468082308769226,
+      "learning_rate": 0.00010128270075621343,
+      "loss": 2.9944,
+      "step": 44715
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6693267226219177,
+      "learning_rate": 0.00010127116683399219,
+      "loss": 2.9631,
+      "step": 44716
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7424618005752563,
+      "learning_rate": 0.00010125963343518446,
+      "loss": 3.0344,
+      "step": 44717
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6935816407203674,
+      "learning_rate": 0.00010124810055982069,
+      "loss": 3.0167,
+      "step": 44718
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7153097987174988,
+      "learning_rate": 0.00010123656820793108,
+      "loss": 3.0504,
+      "step": 44719
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7567762136459351,
+      "learning_rate": 0.0001012250363795462,
+      "loss": 2.9657,
+      "step": 44720
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7075486183166504,
+      "learning_rate": 0.00010121350507469632,
+      "loss": 2.9507,
+      "step": 44721
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7259413599967957,
+      "learning_rate": 0.00010120197429341169,
+      "loss": 2.8704,
+      "step": 44722
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6765598654747009,
+      "learning_rate": 0.00010119044403572293,
+      "loss": 2.7903,
+      "step": 44723
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6906706690788269,
+      "learning_rate": 0.00010117891430166018,
+      "loss": 2.7619,
+      "step": 44724
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7371125221252441,
+      "learning_rate": 0.00010116738509125397,
+      "loss": 2.7714,
+      "step": 44725
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.711427628993988,
+      "learning_rate": 0.00010115585640453453,
+      "loss": 2.9452,
+      "step": 44726
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6924580931663513,
+      "learning_rate": 0.00010114432824153239,
+      "loss": 2.8359,
+      "step": 44727
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7518722414970398,
+      "learning_rate": 0.00010113280060227773,
+      "loss": 2.8871,
+      "step": 44728
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.671915590763092,
+      "learning_rate": 0.00010112127348680109,
+      "loss": 3.0578,
+      "step": 44729
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7346581220626831,
+      "learning_rate": 0.00010110974689513275,
+      "loss": 2.9824,
+      "step": 44730
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6971009373664856,
+      "learning_rate": 0.00010109822082730295,
+      "loss": 2.8534,
+      "step": 44731
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7431304454803467,
+      "learning_rate": 0.00010108669528334225,
+      "loss": 2.9716,
+      "step": 44732
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7102420330047607,
+      "learning_rate": 0.00010107517026328083,
+      "loss": 3.0343,
+      "step": 44733
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7316977381706238,
+      "learning_rate": 0.0001010636457671491,
+      "loss": 3.0913,
+      "step": 44734
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.705828845500946,
+      "learning_rate": 0.00010105212179497757,
+      "loss": 3.0188,
+      "step": 44735
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7134692072868347,
+      "learning_rate": 0.00010104059834679643,
+      "loss": 3.1093,
+      "step": 44736
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7043414115905762,
+      "learning_rate": 0.00010102907542263608,
+      "loss": 2.8323,
+      "step": 44737
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.694769561290741,
+      "learning_rate": 0.00010101755302252678,
+      "loss": 2.843,
+      "step": 44738
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7042750120162964,
+      "learning_rate": 0.00010100603114649892,
+      "loss": 2.9738,
+      "step": 44739
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7265387177467346,
+      "learning_rate": 0.000100994509794583,
+      "loss": 2.8053,
+      "step": 44740
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.743010401725769,
+      "learning_rate": 0.00010098298896680912,
+      "loss": 2.7329,
+      "step": 44741
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6783542633056641,
+      "learning_rate": 0.00010097146866320784,
+      "loss": 3.1741,
+      "step": 44742
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6887816190719604,
+      "learning_rate": 0.00010095994888380932,
+      "loss": 2.9669,
+      "step": 44743
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6920436024665833,
+      "learning_rate": 0.00010094842962864411,
+      "loss": 3.0444,
+      "step": 44744
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6976600289344788,
+      "learning_rate": 0.0001009369108977424,
+      "loss": 3.287,
+      "step": 44745
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7294341921806335,
+      "learning_rate": 0.00010092539269113444,
+      "loss": 2.784,
+      "step": 44746
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.687296986579895,
+      "learning_rate": 0.00010091387500885083,
+      "loss": 2.8358,
+      "step": 44747
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7157872319221497,
+      "learning_rate": 0.00010090235785092164,
+      "loss": 2.8321,
+      "step": 44748
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6645101308822632,
+      "learning_rate": 0.0001008908412173773,
+      "loss": 3.1118,
+      "step": 44749
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7363380193710327,
+      "learning_rate": 0.00010087932510824829,
+      "loss": 2.7542,
+      "step": 44750
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.732443630695343,
+      "learning_rate": 0.00010086780952356481,
+      "loss": 3.1261,
+      "step": 44751
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7466281652450562,
+      "learning_rate": 0.00010085629446335718,
+      "loss": 2.8356,
+      "step": 44752
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.705704391002655,
+      "learning_rate": 0.00010084477992765566,
+      "loss": 2.7202,
+      "step": 44753
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6960504055023193,
+      "learning_rate": 0.00010083326591649065,
+      "loss": 2.9017,
+      "step": 44754
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7246660590171814,
+      "learning_rate": 0.0001008217524298926,
+      "loss": 2.9993,
+      "step": 44755
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7061540484428406,
+      "learning_rate": 0.00010081023946789163,
+      "loss": 2.8255,
+      "step": 44756
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7059864401817322,
+      "learning_rate": 0.00010079872703051826,
+      "loss": 2.7661,
+      "step": 44757
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7149134278297424,
+      "learning_rate": 0.00010078721511780266,
+      "loss": 2.8085,
+      "step": 44758
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6905004978179932,
+      "learning_rate": 0.00010077570372977513,
+      "loss": 2.9624,
+      "step": 44759
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7025719285011292,
+      "learning_rate": 0.00010076419286646614,
+      "loss": 2.9586,
+      "step": 44760
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7020415663719177,
+      "learning_rate": 0.00010075268252790586,
+      "loss": 2.8406,
+      "step": 44761
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7688156366348267,
+      "learning_rate": 0.00010074117271412473,
+      "loss": 2.9296,
+      "step": 44762
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7104856967926025,
+      "learning_rate": 0.0001007296634251529,
+      "loss": 2.875,
+      "step": 44763
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7477237582206726,
+      "learning_rate": 0.0001007181546610208,
+      "loss": 2.8623,
+      "step": 44764
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7128356099128723,
+      "learning_rate": 0.00010070664642175885,
+      "loss": 2.9696,
+      "step": 44765
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.730711042881012,
+      "learning_rate": 0.0001006951387073972,
+      "loss": 2.933,
+      "step": 44766
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7141094207763672,
+      "learning_rate": 0.0001006836315179662,
+      "loss": 2.9074,
+      "step": 44767
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6910660266876221,
+      "learning_rate": 0.00010067212485349609,
+      "loss": 2.9853,
+      "step": 44768
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6767820715904236,
+      "learning_rate": 0.00010066061871401721,
+      "loss": 2.9947,
+      "step": 44769
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7095887660980225,
+      "learning_rate": 0.00010064911309956001,
+      "loss": 2.7702,
+      "step": 44770
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7199665904045105,
+      "learning_rate": 0.00010063760801015457,
+      "loss": 2.8923,
+      "step": 44771
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7285208702087402,
+      "learning_rate": 0.0001006261034458314,
+      "loss": 2.9582,
+      "step": 44772
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7214688658714294,
+      "learning_rate": 0.00010061459940662073,
+      "loss": 2.7663,
+      "step": 44773
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7294226288795471,
+      "learning_rate": 0.00010060309589255273,
+      "loss": 3.0265,
+      "step": 44774
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7086969614028931,
+      "learning_rate": 0.00010059159290365788,
+      "loss": 2.8846,
+      "step": 44775
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7389560341835022,
+      "learning_rate": 0.00010058009043996628,
+      "loss": 2.8932,
+      "step": 44776
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6759814620018005,
+      "learning_rate": 0.00010056858850150849,
+      "loss": 2.9601,
+      "step": 44777
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7065625786781311,
+      "learning_rate": 0.00010055708708831451,
+      "loss": 2.806,
+      "step": 44778
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7095101475715637,
+      "learning_rate": 0.00010054558620041492,
+      "loss": 2.9055,
+      "step": 44779
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.791003942489624,
+      "learning_rate": 0.00010053408583783984,
+      "loss": 2.9966,
+      "step": 44780
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7229596376419067,
+      "learning_rate": 0.00010052258600061953,
+      "loss": 2.7862,
+      "step": 44781
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7416842579841614,
+      "learning_rate": 0.0001005110866887844,
+      "loss": 2.8895,
+      "step": 44782
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.737678050994873,
+      "learning_rate": 0.0001004995879023646,
+      "loss": 3.0674,
+      "step": 44783
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6982100605964661,
+      "learning_rate": 0.00010048808964139045,
+      "loss": 2.7872,
+      "step": 44784
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7631065845489502,
+      "learning_rate": 0.00010047659190589243,
+      "loss": 2.8958,
+      "step": 44785
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6911045908927917,
+      "learning_rate": 0.00010046509469590054,
+      "loss": 2.7337,
+      "step": 44786
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6925402283668518,
+      "learning_rate": 0.00010045359801144526,
+      "loss": 2.8405,
+      "step": 44787
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7005817890167236,
+      "learning_rate": 0.00010044210185255685,
+      "loss": 3.0554,
+      "step": 44788
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7038406133651733,
+      "learning_rate": 0.0001004306062192654,
+      "loss": 2.8066,
+      "step": 44789
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7171939015388489,
+      "learning_rate": 0.00010041911111160143,
+      "loss": 2.9208,
+      "step": 44790
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.757029116153717,
+      "learning_rate": 0.00010040761652959505,
+      "loss": 2.892,
+      "step": 44791
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.703034520149231,
+      "learning_rate": 0.00010039612247327666,
+      "loss": 2.908,
+      "step": 44792
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7032923698425293,
+      "learning_rate": 0.00010038462894267639,
+      "loss": 2.9547,
+      "step": 44793
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7094815373420715,
+      "learning_rate": 0.00010037313593782467,
+      "loss": 2.9108,
+      "step": 44794
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6859133839607239,
+      "learning_rate": 0.00010036164345875168,
+      "loss": 2.9368,
+      "step": 44795
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7258820533752441,
+      "learning_rate": 0.00010035015150548763,
+      "loss": 2.901,
+      "step": 44796
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7605149149894714,
+      "learning_rate": 0.00010033866007806296,
+      "loss": 2.7665,
+      "step": 44797
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7210118174552917,
+      "learning_rate": 0.00010032716917650771,
+      "loss": 3.0509,
+      "step": 44798
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7114245891571045,
+      "learning_rate": 0.0001003156788008523,
+      "loss": 2.6478,
+      "step": 44799
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7188319563865662,
+      "learning_rate": 0.00010030418895112701,
+      "loss": 2.9248,
+      "step": 44800
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7011023759841919,
+      "learning_rate": 0.00010029269962736209,
+      "loss": 2.8442,
+      "step": 44801
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7070204019546509,
+      "learning_rate": 0.00010028121082958777,
+      "loss": 2.9462,
+      "step": 44802
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7259615659713745,
+      "learning_rate": 0.0001002697225578342,
+      "loss": 3.1676,
+      "step": 44803
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6928755640983582,
+      "learning_rate": 0.00010025823481213183,
+      "loss": 2.977,
+      "step": 44804
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7410005331039429,
+      "learning_rate": 0.00010024674759251074,
+      "loss": 3.0249,
+      "step": 44805
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6871789693832397,
+      "learning_rate": 0.00010023526089900125,
+      "loss": 2.867,
+      "step": 44806
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6890696883201599,
+      "learning_rate": 0.00010022377473163378,
+      "loss": 3.096,
+      "step": 44807
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6946797966957092,
+      "learning_rate": 0.00010021228909043831,
+      "loss": 2.9074,
+      "step": 44808
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7045637369155884,
+      "learning_rate": 0.0001002008039754453,
+      "loss": 3.0301,
+      "step": 44809
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7553611397743225,
+      "learning_rate": 0.00010018931938668497,
+      "loss": 2.8607,
+      "step": 44810
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6659156680107117,
+      "learning_rate": 0.00010017783532418738,
+      "loss": 2.957,
+      "step": 44811
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7131342887878418,
+      "learning_rate": 0.00010016635178798304,
+      "loss": 2.9252,
+      "step": 44812
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.741779625415802,
+      "learning_rate": 0.00010015486877810194,
+      "loss": 2.9722,
+      "step": 44813
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6962339282035828,
+      "learning_rate": 0.0001001433862945746,
+      "loss": 2.8588,
+      "step": 44814
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6637634038925171,
+      "learning_rate": 0.00010013190433743097,
+      "loss": 2.912,
+      "step": 44815
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6877065896987915,
+      "learning_rate": 0.00010012042290670158,
+      "loss": 3.0355,
+      "step": 44816
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7660030126571655,
+      "learning_rate": 0.00010010894200241648,
+      "loss": 3.0438,
+      "step": 44817
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.711549699306488,
+      "learning_rate": 0.00010009746162460592,
+      "loss": 2.8984,
+      "step": 44818
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7448240518569946,
+      "learning_rate": 0.00010008598177330022,
+      "loss": 2.9206,
+      "step": 44819
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6869280934333801,
+      "learning_rate": 0.00010007450244852951,
+      "loss": 2.9734,
+      "step": 44820
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.8840466141700745,
+      "learning_rate": 0.00010006302365032404,
+      "loss": 2.6196,
+      "step": 44821
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7302159070968628,
+      "learning_rate": 0.00010005154537871423,
+      "loss": 2.9981,
+      "step": 44822
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7159656882286072,
+      "learning_rate": 0.00010004006763373014,
+      "loss": 2.7305,
+      "step": 44823
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7011760473251343,
+      "learning_rate": 0.00010002859041540202,
+      "loss": 2.9915,
+      "step": 44824
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.8024114370346069,
+      "learning_rate": 0.00010001711372376005,
+      "loss": 2.9443,
+      "step": 44825
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7106544971466064,
+      "learning_rate": 0.00010000563755883449,
+      "loss": 2.8404,
+      "step": 44826
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7113576531410217,
+      "learning_rate": 9.999416192065568e-05,
+      "loss": 3.0292,
+      "step": 44827
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7601901888847351,
+      "learning_rate": 9.998268680925368e-05,
+      "loss": 2.8893,
+      "step": 44828
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.721528172492981,
+      "learning_rate": 9.997121222465886e-05,
+      "loss": 2.8252,
+      "step": 44829
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7221820950508118,
+      "learning_rate": 9.99597381669013e-05,
+      "loss": 2.986,
+      "step": 44830
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.733727216720581,
+      "learning_rate": 9.994826463601136e-05,
+      "loss": 2.9532,
+      "step": 44831
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6860153675079346,
+      "learning_rate": 9.993679163201921e-05,
+      "loss": 2.9451,
+      "step": 44832
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7270806431770325,
+      "learning_rate": 9.992531915495494e-05,
+      "loss": 2.8611,
+      "step": 44833
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6891951560974121,
+      "learning_rate": 9.9913847204849e-05,
+      "loss": 3.1389,
+      "step": 44834
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7794495224952698,
+      "learning_rate": 9.990237578173135e-05,
+      "loss": 2.7893,
+      "step": 44835
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.707525908946991,
+      "learning_rate": 9.989090488563233e-05,
+      "loss": 2.9123,
+      "step": 44836
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.788293719291687,
+      "learning_rate": 9.987943451658228e-05,
+      "loss": 2.8435,
+      "step": 44837
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7396550178527832,
+      "learning_rate": 9.98679646746113e-05,
+      "loss": 2.9739,
+      "step": 44838
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7474350929260254,
+      "learning_rate": 9.985649535974954e-05,
+      "loss": 2.8834,
+      "step": 44839
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7349193096160889,
+      "learning_rate": 9.984502657202716e-05,
+      "loss": 2.9247,
+      "step": 44840
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6847119927406311,
+      "learning_rate": 9.983355831147446e-05,
+      "loss": 3.1132,
+      "step": 44841
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7022256255149841,
+      "learning_rate": 9.982209057812177e-05,
+      "loss": 2.9059,
+      "step": 44842
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7146331667900085,
+      "learning_rate": 9.981062337199905e-05,
+      "loss": 2.7936,
+      "step": 44843
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7405462265014648,
+      "learning_rate": 9.979915669313672e-05,
+      "loss": 2.8458,
+      "step": 44844
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7254295945167542,
+      "learning_rate": 9.97876905415648e-05,
+      "loss": 3.0363,
+      "step": 44845
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6980922222137451,
+      "learning_rate": 9.977622491731364e-05,
+      "loss": 3.0195,
+      "step": 44846
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6992781758308411,
+      "learning_rate": 9.97647598204134e-05,
+      "loss": 3.0519,
+      "step": 44847
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7055832147598267,
+      "learning_rate": 9.975329525089412e-05,
+      "loss": 2.8358,
+      "step": 44848
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6944383382797241,
+      "learning_rate": 9.974183120878624e-05,
+      "loss": 3.0116,
+      "step": 44849
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.714631199836731,
+      "learning_rate": 9.973036769411972e-05,
+      "loss": 3.0383,
+      "step": 44850
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7210114598274231,
+      "learning_rate": 9.971890470692489e-05,
+      "loss": 2.9004,
+      "step": 44851
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7096450924873352,
+      "learning_rate": 9.970744224723201e-05,
+      "loss": 2.8675,
+      "step": 44852
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6983294486999512,
+      "learning_rate": 9.969598031507119e-05,
+      "loss": 3.1334,
+      "step": 44853
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6857281923294067,
+      "learning_rate": 9.968451891047259e-05,
+      "loss": 2.967,
+      "step": 44854
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6746247410774231,
+      "learning_rate": 9.967305803346633e-05,
+      "loss": 3.0097,
+      "step": 44855
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7361533045768738,
+      "learning_rate": 9.966159768408267e-05,
+      "loss": 2.8268,
+      "step": 44856
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7003144025802612,
+      "learning_rate": 9.965013786235192e-05,
+      "loss": 3.0137,
+      "step": 44857
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7251600623130798,
+      "learning_rate": 9.963867856830404e-05,
+      "loss": 2.7261,
+      "step": 44858
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7403432130813599,
+      "learning_rate": 9.962721980196942e-05,
+      "loss": 3.0198,
+      "step": 44859
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7124742865562439,
+      "learning_rate": 9.961576156337814e-05,
+      "loss": 2.8851,
+      "step": 44860
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6854257583618164,
+      "learning_rate": 9.960430385256029e-05,
+      "loss": 3.0093,
+      "step": 44861
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7400795221328735,
+      "learning_rate": 9.95928466695462e-05,
+      "loss": 3.0098,
+      "step": 44862
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7203598022460938,
+      "learning_rate": 9.958139001436593e-05,
+      "loss": 2.8234,
+      "step": 44863
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7296319007873535,
+      "learning_rate": 9.956993388704977e-05,
+      "loss": 3.0328,
+      "step": 44864
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7356207966804504,
+      "learning_rate": 9.955847828762778e-05,
+      "loss": 2.9695,
+      "step": 44865
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7014438509941101,
+      "learning_rate": 9.954702321613015e-05,
+      "loss": 3.0264,
+      "step": 44866
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.740602970123291,
+      "learning_rate": 9.953556867258717e-05,
+      "loss": 2.8827,
+      "step": 44867
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.699821412563324,
+      "learning_rate": 9.952411465702893e-05,
+      "loss": 2.8705,
+      "step": 44868
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7270916104316711,
+      "learning_rate": 9.95126611694856e-05,
+      "loss": 3.002,
+      "step": 44869
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6713517904281616,
+      "learning_rate": 9.950120820998721e-05,
+      "loss": 2.8885,
+      "step": 44870
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7152098417282104,
+      "learning_rate": 9.948975577856409e-05,
+      "loss": 3.0077,
+      "step": 44871
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.703411340713501,
+      "learning_rate": 9.947830387524644e-05,
+      "loss": 2.8513,
+      "step": 44872
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7669411301612854,
+      "learning_rate": 9.946685250006427e-05,
+      "loss": 3.0245,
+      "step": 44873
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6936809420585632,
+      "learning_rate": 9.945540165304789e-05,
+      "loss": 2.8511,
+      "step": 44874
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7249215841293335,
+      "learning_rate": 9.944395133422741e-05,
+      "loss": 2.9093,
+      "step": 44875
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7058155536651611,
+      "learning_rate": 9.943250154363283e-05,
+      "loss": 3.0376,
+      "step": 44876
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7191633582115173,
+      "learning_rate": 9.942105228129456e-05,
+      "loss": 3.0574,
+      "step": 44877
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7496932744979858,
+      "learning_rate": 9.940960354724255e-05,
+      "loss": 2.8689,
+      "step": 44878
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7125669121742249,
+      "learning_rate": 9.939815534150711e-05,
+      "loss": 3.145,
+      "step": 44879
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7709475159645081,
+      "learning_rate": 9.938670766411825e-05,
+      "loss": 2.8035,
+      "step": 44880
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.705838680267334,
+      "learning_rate": 9.937526051510629e-05,
+      "loss": 2.9948,
+      "step": 44881
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7062419056892395,
+      "learning_rate": 9.936381389450129e-05,
+      "loss": 2.7527,
+      "step": 44882
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7100111246109009,
+      "learning_rate": 9.935236780233325e-05,
+      "loss": 2.9641,
+      "step": 44883
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6903637051582336,
+      "learning_rate": 9.934092223863261e-05,
+      "loss": 2.9546,
+      "step": 44884
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6916725039482117,
+      "learning_rate": 9.932947720342921e-05,
+      "loss": 2.8971,
+      "step": 44885
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7668691277503967,
+      "learning_rate": 9.931803269675351e-05,
+      "loss": 2.7209,
+      "step": 44886
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6978892087936401,
+      "learning_rate": 9.930658871863533e-05,
+      "loss": 2.9635,
+      "step": 44887
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7086914777755737,
+      "learning_rate": 9.929514526910499e-05,
+      "loss": 3.0682,
+      "step": 44888
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7077818512916565,
+      "learning_rate": 9.928370234819274e-05,
+      "loss": 2.9384,
+      "step": 44889
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6918936967849731,
+      "learning_rate": 9.927225995592855e-05,
+      "loss": 2.7194,
+      "step": 44890
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7205008864402771,
+      "learning_rate": 9.926081809234262e-05,
+      "loss": 2.9189,
+      "step": 44891
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7491716146469116,
+      "learning_rate": 9.924937675746493e-05,
+      "loss": 3.0693,
+      "step": 44892
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7079654932022095,
+      "learning_rate": 9.923793595132577e-05,
+      "loss": 3.0083,
+      "step": 44893
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7141336798667908,
+      "learning_rate": 9.922649567395537e-05,
+      "loss": 2.8574,
+      "step": 44894
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7455933690071106,
+      "learning_rate": 9.921505592538362e-05,
+      "loss": 3.0319,
+      "step": 44895
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7107555270195007,
+      "learning_rate": 9.920361670564087e-05,
+      "loss": 2.7566,
+      "step": 44896
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7711759209632874,
+      "learning_rate": 9.919217801475715e-05,
+      "loss": 2.9126,
+      "step": 44897
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6815950870513916,
+      "learning_rate": 9.918073985276248e-05,
+      "loss": 2.7761,
+      "step": 44898
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7206303477287292,
+      "learning_rate": 9.916930221968724e-05,
+      "loss": 3.123,
+      "step": 44899
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.675313413143158,
+      "learning_rate": 9.915786511556123e-05,
+      "loss": 2.9108,
+      "step": 44900
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7314799427986145,
+      "learning_rate": 9.91464285404149e-05,
+      "loss": 2.7628,
+      "step": 44901
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7563438415527344,
+      "learning_rate": 9.913499249427812e-05,
+      "loss": 2.858,
+      "step": 44902
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7086213231086731,
+      "learning_rate": 9.912355697718119e-05,
+      "loss": 2.8109,
+      "step": 44903
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7429339289665222,
+      "learning_rate": 9.911212198915415e-05,
+      "loss": 2.8593,
+      "step": 44904
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6543605327606201,
+      "learning_rate": 9.910068753022706e-05,
+      "loss": 2.8576,
+      "step": 44905
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7347574830055237,
+      "learning_rate": 9.908925360043015e-05,
+      "loss": 2.725,
+      "step": 44906
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7006809115409851,
+      "learning_rate": 9.907782019979343e-05,
+      "loss": 2.8389,
+      "step": 44907
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7191192507743835,
+      "learning_rate": 9.906638732834701e-05,
+      "loss": 2.9768,
+      "step": 44908
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7019370794296265,
+      "learning_rate": 9.905495498612118e-05,
+      "loss": 2.7632,
+      "step": 44909
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7217749357223511,
+      "learning_rate": 9.904352317314583e-05,
+      "loss": 2.9252,
+      "step": 44910
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7480356097221375,
+      "learning_rate": 9.903209188945127e-05,
+      "loss": 2.9351,
+      "step": 44911
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7302222847938538,
+      "learning_rate": 9.902066113506747e-05,
+      "loss": 2.8074,
+      "step": 44912
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7058594226837158,
+      "learning_rate": 9.90092309100245e-05,
+      "loss": 2.976,
+      "step": 44913
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7320786118507385,
+      "learning_rate": 9.899780121435262e-05,
+      "loss": 2.8483,
+      "step": 44914
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7021784782409668,
+      "learning_rate": 9.898637204808176e-05,
+      "loss": 2.9878,
+      "step": 44915
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7117753028869629,
+      "learning_rate": 9.89749434112422e-05,
+      "loss": 2.871,
+      "step": 44916
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7381766438484192,
+      "learning_rate": 9.896351530386385e-05,
+      "loss": 2.8505,
+      "step": 44917
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.69048011302948,
+      "learning_rate": 9.895208772597705e-05,
+      "loss": 3.1488,
+      "step": 44918
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7255758047103882,
+      "learning_rate": 9.894066067761171e-05,
+      "loss": 2.8735,
+      "step": 44919
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6982675790786743,
+      "learning_rate": 9.89292341587979e-05,
+      "loss": 2.9529,
+      "step": 44920
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7159181237220764,
+      "learning_rate": 9.891780816956587e-05,
+      "loss": 2.9456,
+      "step": 44921
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6875530481338501,
+      "learning_rate": 9.890638270994553e-05,
+      "loss": 2.8369,
+      "step": 44922
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6837921142578125,
+      "learning_rate": 9.889495777996709e-05,
+      "loss": 3.0856,
+      "step": 44923
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7514594793319702,
+      "learning_rate": 9.888353337966075e-05,
+      "loss": 3.0671,
+      "step": 44924
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7067429423332214,
+      "learning_rate": 9.887210950905643e-05,
+      "loss": 2.7875,
+      "step": 44925
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7050120830535889,
+      "learning_rate": 9.88606861681843e-05,
+      "loss": 3.0002,
+      "step": 44926
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7361913919448853,
+      "learning_rate": 9.884926335707431e-05,
+      "loss": 2.9435,
+      "step": 44927
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7067899107933044,
+      "learning_rate": 9.883784107575665e-05,
+      "loss": 2.8466,
+      "step": 44928
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7761995792388916,
+      "learning_rate": 9.882641932426149e-05,
+      "loss": 2.9587,
+      "step": 44929
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7403863668441772,
+      "learning_rate": 9.881499810261871e-05,
+      "loss": 2.7738,
+      "step": 44930
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6781747341156006,
+      "learning_rate": 9.880357741085865e-05,
+      "loss": 2.973,
+      "step": 44931
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7273198366165161,
+      "learning_rate": 9.879215724901109e-05,
+      "loss": 2.8573,
+      "step": 44932
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7114270925521851,
+      "learning_rate": 9.878073761710639e-05,
+      "loss": 2.7882,
+      "step": 44933
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7011170387268066,
+      "learning_rate": 9.876931851517454e-05,
+      "loss": 2.8851,
+      "step": 44934
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7164783477783203,
+      "learning_rate": 9.875789994324544e-05,
+      "loss": 3.0467,
+      "step": 44935
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7922382950782776,
+      "learning_rate": 9.874648190134941e-05,
+      "loss": 3.0482,
+      "step": 44936
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6847321391105652,
+      "learning_rate": 9.87350643895163e-05,
+      "loss": 2.9396,
+      "step": 44937
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7641184329986572,
+      "learning_rate": 9.872364740777632e-05,
+      "loss": 3.0284,
+      "step": 44938
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7687472105026245,
+      "learning_rate": 9.871223095615959e-05,
+      "loss": 3.0097,
+      "step": 44939
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7540838122367859,
+      "learning_rate": 9.87008150346961e-05,
+      "loss": 2.9874,
+      "step": 44940
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.8029380440711975,
+      "learning_rate": 9.868939964341597e-05,
+      "loss": 2.6693,
+      "step": 44941
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7017350792884827,
+      "learning_rate": 9.86779847823491e-05,
+      "loss": 3.0033,
+      "step": 44942
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7199008464813232,
+      "learning_rate": 9.866657045152564e-05,
+      "loss": 3.0994,
+      "step": 44943
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6676777005195618,
+      "learning_rate": 9.865515665097581e-05,
+      "loss": 2.7228,
+      "step": 44944
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6894500851631165,
+      "learning_rate": 9.864374338072945e-05,
+      "loss": 2.9413,
+      "step": 44945
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7256870269775391,
+      "learning_rate": 9.863233064081682e-05,
+      "loss": 2.7942,
+      "step": 44946
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7502337098121643,
+      "learning_rate": 9.862091843126777e-05,
+      "loss": 3.1109,
+      "step": 44947
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6998251080513,
+      "learning_rate": 9.860950675211255e-05,
+      "loss": 2.8859,
+      "step": 44948
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7343108057975769,
+      "learning_rate": 9.859809560338115e-05,
+      "loss": 3.0134,
+      "step": 44949
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7494150996208191,
+      "learning_rate": 9.85866849851035e-05,
+      "loss": 2.7913,
+      "step": 44950
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6747055649757385,
+      "learning_rate": 9.857527489730987e-05,
+      "loss": 2.9572,
+      "step": 44951
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6779135465621948,
+      "learning_rate": 9.856386534003011e-05,
+      "loss": 2.8614,
+      "step": 44952
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.713498592376709,
+      "learning_rate": 9.855245631329432e-05,
+      "loss": 2.8165,
+      "step": 44953
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6696012616157532,
+      "learning_rate": 9.854104781713272e-05,
+      "loss": 3.0592,
+      "step": 44954
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7587345242500305,
+      "learning_rate": 9.852963985157525e-05,
+      "loss": 2.8299,
+      "step": 44955
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7578355669975281,
+      "learning_rate": 9.851823241665188e-05,
+      "loss": 2.9793,
+      "step": 44956
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6964207291603088,
+      "learning_rate": 9.850682551239265e-05,
+      "loss": 2.9437,
+      "step": 44957
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.8269992470741272,
+      "learning_rate": 9.849541913882774e-05,
+      "loss": 2.9062,
+      "step": 44958
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6761069893836975,
+      "learning_rate": 9.848401329598705e-05,
+      "loss": 2.9209,
+      "step": 44959
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7033849954605103,
+      "learning_rate": 9.847260798390064e-05,
+      "loss": 2.974,
+      "step": 44960
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7039138078689575,
+      "learning_rate": 9.846120320259873e-05,
+      "loss": 2.9274,
+      "step": 44961
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6891793608665466,
+      "learning_rate": 9.844979895211121e-05,
+      "loss": 2.7909,
+      "step": 44962
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6782719492912292,
+      "learning_rate": 9.843839523246815e-05,
+      "loss": 2.8842,
+      "step": 44963
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7017290592193604,
+      "learning_rate": 9.842699204369944e-05,
+      "loss": 3.038,
+      "step": 44964
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.8234224915504456,
+      "learning_rate": 9.841558938583525e-05,
+      "loss": 2.8788,
+      "step": 44965
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.720818817615509,
+      "learning_rate": 9.840418725890569e-05,
+      "loss": 2.7221,
+      "step": 44966
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7300722599029541,
+      "learning_rate": 9.839278566294063e-05,
+      "loss": 2.9979,
+      "step": 44967
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6792693138122559,
+      "learning_rate": 9.838138459797026e-05,
+      "loss": 2.7892,
+      "step": 44968
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7436376214027405,
+      "learning_rate": 9.83699840640244e-05,
+      "loss": 3.0367,
+      "step": 44969
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6573380827903748,
+      "learning_rate": 9.835858406113333e-05,
+      "loss": 2.9167,
+      "step": 44970
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6874247789382935,
+      "learning_rate": 9.834718458932692e-05,
+      "loss": 2.7693,
+      "step": 44971
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7144591808319092,
+      "learning_rate": 9.833578564863514e-05,
+      "loss": 2.7311,
+      "step": 44972
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.72472083568573,
+      "learning_rate": 9.832438723908814e-05,
+      "loss": 3.0349,
+      "step": 44973
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7377344369888306,
+      "learning_rate": 9.831298936071586e-05,
+      "loss": 3.0525,
+      "step": 44974
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7162564992904663,
+      "learning_rate": 9.830159201354828e-05,
+      "loss": 2.935,
+      "step": 44975
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7077290415763855,
+      "learning_rate": 9.829019519761564e-05,
+      "loss": 3.0306,
+      "step": 44976
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6880072951316833,
+      "learning_rate": 9.827879891294777e-05,
+      "loss": 2.8846,
+      "step": 44977
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7291051149368286,
+      "learning_rate": 9.826740315957476e-05,
+      "loss": 2.8189,
+      "step": 44978
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7197936177253723,
+      "learning_rate": 9.825600793752647e-05,
+      "loss": 2.816,
+      "step": 44979
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6987482309341431,
+      "learning_rate": 9.824461324683304e-05,
+      "loss": 2.7839,
+      "step": 44980
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7558214664459229,
+      "learning_rate": 9.823321908752454e-05,
+      "loss": 2.8651,
+      "step": 44981
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7327513694763184,
+      "learning_rate": 9.822182545963082e-05,
+      "loss": 2.8579,
+      "step": 44982
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7057175040245056,
+      "learning_rate": 9.821043236318209e-05,
+      "loss": 2.9509,
+      "step": 44983
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7285510301589966,
+      "learning_rate": 9.819903979820825e-05,
+      "loss": 2.9622,
+      "step": 44984
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6786232590675354,
+      "learning_rate": 9.81876477647392e-05,
+      "loss": 3.0846,
+      "step": 44985
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7109321355819702,
+      "learning_rate": 9.817625626280511e-05,
+      "loss": 2.8528,
+      "step": 44986
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7196544408798218,
+      "learning_rate": 9.816486529243587e-05,
+      "loss": 2.7905,
+      "step": 44987
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6925801634788513,
+      "learning_rate": 9.815347485366162e-05,
+      "loss": 2.8368,
+      "step": 44988
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.687707245349884,
+      "learning_rate": 9.814208494651214e-05,
+      "loss": 2.8255,
+      "step": 44989
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7053709030151367,
+      "learning_rate": 9.81306955710176e-05,
+      "loss": 3.0476,
+      "step": 44990
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.6974628567695618,
+      "learning_rate": 9.811930672720803e-05,
+      "loss": 2.739,
+      "step": 44991
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.7128167152404785,
+      "learning_rate": 9.810791841511336e-05,
+      "loss": 3.0011,
+      "step": 44992
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.694808304309845,
+      "learning_rate": 9.809653063476358e-05,
+      "loss": 2.8174,
+      "step": 44993
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7266348600387573,
+      "learning_rate": 9.808514338618856e-05,
+      "loss": 2.8004,
+      "step": 44994
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6718068718910217,
+      "learning_rate": 9.807375666941846e-05,
+      "loss": 2.9927,
+      "step": 44995
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7475922107696533,
+      "learning_rate": 9.806237048448329e-05,
+      "loss": 2.8641,
+      "step": 44996
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6909314393997192,
+      "learning_rate": 9.805098483141287e-05,
+      "loss": 2.8757,
+      "step": 44997
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7345043420791626,
+      "learning_rate": 9.803959971023742e-05,
+      "loss": 2.8709,
+      "step": 44998
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7070412635803223,
+      "learning_rate": 9.802821512098677e-05,
+      "loss": 2.997,
+      "step": 44999
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6929289698600769,
+      "learning_rate": 9.801683106369082e-05,
+      "loss": 3.0065,
+      "step": 45000
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7302191257476807,
+      "learning_rate": 9.800544753837978e-05,
+      "loss": 2.8856,
+      "step": 45001
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7192659378051758,
+      "learning_rate": 9.799406454508343e-05,
+      "loss": 2.8797,
+      "step": 45002
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.757664680480957,
+      "learning_rate": 9.798268208383192e-05,
+      "loss": 2.4703,
+      "step": 45003
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7271679043769836,
+      "learning_rate": 9.797130015465507e-05,
+      "loss": 2.98,
+      "step": 45004
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7153617739677429,
+      "learning_rate": 9.7959918757583e-05,
+      "loss": 2.992,
+      "step": 45005
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7023801207542419,
+      "learning_rate": 9.794853789264564e-05,
+      "loss": 2.8245,
+      "step": 45006
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7436354160308838,
+      "learning_rate": 9.793715755987283e-05,
+      "loss": 2.9209,
+      "step": 45007
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7024301886558533,
+      "learning_rate": 9.792577775929476e-05,
+      "loss": 3.0383,
+      "step": 45008
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.680385410785675,
+      "learning_rate": 9.791439849094124e-05,
+      "loss": 2.8592,
+      "step": 45009
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6959872245788574,
+      "learning_rate": 9.790301975484228e-05,
+      "loss": 2.9236,
+      "step": 45010
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7033017873764038,
+      "learning_rate": 9.789164155102795e-05,
+      "loss": 3.054,
+      "step": 45011
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6920191645622253,
+      "learning_rate": 9.788026387952807e-05,
+      "loss": 2.8266,
+      "step": 45012
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6858896017074585,
+      "learning_rate": 9.786888674037276e-05,
+      "loss": 2.9281,
+      "step": 45013
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6978284120559692,
+      "learning_rate": 9.785751013359189e-05,
+      "loss": 2.9375,
+      "step": 45014
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7071654200553894,
+      "learning_rate": 9.784613405921535e-05,
+      "loss": 2.8444,
+      "step": 45015
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6944267749786377,
+      "learning_rate": 9.78347585172733e-05,
+      "loss": 2.8455,
+      "step": 45016
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7331633567810059,
+      "learning_rate": 9.782338350779548e-05,
+      "loss": 3.0644,
+      "step": 45017
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7539132237434387,
+      "learning_rate": 9.781200903081206e-05,
+      "loss": 2.8949,
+      "step": 45018
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7396433353424072,
+      "learning_rate": 9.780063508635279e-05,
+      "loss": 2.938,
+      "step": 45019
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7210110425949097,
+      "learning_rate": 9.778926167444781e-05,
+      "loss": 2.6836,
+      "step": 45020
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7514802813529968,
+      "learning_rate": 9.777788879512702e-05,
+      "loss": 2.9036,
+      "step": 45021
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7779740691184998,
+      "learning_rate": 9.776651644842026e-05,
+      "loss": 2.8719,
+      "step": 45022
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7061024308204651,
+      "learning_rate": 9.775514463435767e-05,
+      "loss": 2.7755,
+      "step": 45023
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7684993743896484,
+      "learning_rate": 9.7743773352969e-05,
+      "loss": 2.855,
+      "step": 45024
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7269738912582397,
+      "learning_rate": 9.773240260428431e-05,
+      "loss": 3.2446,
+      "step": 45025
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6850230693817139,
+      "learning_rate": 9.772103238833364e-05,
+      "loss": 3.0956,
+      "step": 45026
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6893491148948669,
+      "learning_rate": 9.770966270514684e-05,
+      "loss": 2.9326,
+      "step": 45027
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7849996089935303,
+      "learning_rate": 9.769829355475388e-05,
+      "loss": 2.8876,
+      "step": 45028
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7221559882164001,
+      "learning_rate": 9.768692493718457e-05,
+      "loss": 2.9922,
+      "step": 45029
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6811572909355164,
+      "learning_rate": 9.767555685246896e-05,
+      "loss": 2.9712,
+      "step": 45030
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7301697731018066,
+      "learning_rate": 9.766418930063709e-05,
+      "loss": 3.1497,
+      "step": 45031
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7387075424194336,
+      "learning_rate": 9.765282228171869e-05,
+      "loss": 2.731,
+      "step": 45032
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7403630018234253,
+      "learning_rate": 9.764145579574394e-05,
+      "loss": 3.0009,
+      "step": 45033
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7094658613204956,
+      "learning_rate": 9.763008984274254e-05,
+      "loss": 2.7886,
+      "step": 45034
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6980571150779724,
+      "learning_rate": 9.761872442274464e-05,
+      "loss": 3.1573,
+      "step": 45035
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7270064949989319,
+      "learning_rate": 9.760735953578006e-05,
+      "loss": 2.9224,
+      "step": 45036
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.8115057349205017,
+      "learning_rate": 9.759599518187867e-05,
+      "loss": 3.1933,
+      "step": 45037
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7091026306152344,
+      "learning_rate": 9.758463136107055e-05,
+      "loss": 2.6673,
+      "step": 45038
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7223027348518372,
+      "learning_rate": 9.757326807338548e-05,
+      "loss": 2.7039,
+      "step": 45039
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7187199592590332,
+      "learning_rate": 9.756190531885353e-05,
+      "loss": 2.971,
+      "step": 45040
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7661718726158142,
+      "learning_rate": 9.755054309750451e-05,
+      "loss": 2.9357,
+      "step": 45041
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.710060179233551,
+      "learning_rate": 9.753918140936846e-05,
+      "loss": 3.0332,
+      "step": 45042
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6724956035614014,
+      "learning_rate": 9.752782025447523e-05,
+      "loss": 3.0016,
+      "step": 45043
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6956879496574402,
+      "learning_rate": 9.751645963285469e-05,
+      "loss": 3.0643,
+      "step": 45044
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7180279493331909,
+      "learning_rate": 9.750509954453689e-05,
+      "loss": 3.1198,
+      "step": 45045
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6832389235496521,
+      "learning_rate": 9.749373998955163e-05,
+      "loss": 2.9694,
+      "step": 45046
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6939116716384888,
+      "learning_rate": 9.748238096792886e-05,
+      "loss": 2.8727,
+      "step": 45047
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6981992721557617,
+      "learning_rate": 9.747102247969862e-05,
+      "loss": 2.8948,
+      "step": 45048
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7490467429161072,
+      "learning_rate": 9.745966452489072e-05,
+      "loss": 2.6799,
+      "step": 45049
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7155380845069885,
+      "learning_rate": 9.74483071035351e-05,
+      "loss": 2.9145,
+      "step": 45050
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7536515593528748,
+      "learning_rate": 9.743695021566157e-05,
+      "loss": 2.81,
+      "step": 45051
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6787039041519165,
+      "learning_rate": 9.742559386130012e-05,
+      "loss": 2.8627,
+      "step": 45052
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7021150588989258,
+      "learning_rate": 9.741423804048077e-05,
+      "loss": 2.9647,
+      "step": 45053
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7255040407180786,
+      "learning_rate": 9.740288275323321e-05,
+      "loss": 3.0502,
+      "step": 45054
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7342783212661743,
+      "learning_rate": 9.739152799958758e-05,
+      "loss": 2.8024,
+      "step": 45055
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6887534260749817,
+      "learning_rate": 9.73801737795736e-05,
+      "loss": 2.7358,
+      "step": 45056
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6920632123947144,
+      "learning_rate": 9.736882009322131e-05,
+      "loss": 2.8555,
+      "step": 45057
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7355381846427917,
+      "learning_rate": 9.735746694056055e-05,
+      "loss": 3.0263,
+      "step": 45058
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7083019018173218,
+      "learning_rate": 9.734611432162114e-05,
+      "loss": 2.8918,
+      "step": 45059
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7241811156272888,
+      "learning_rate": 9.733476223643317e-05,
+      "loss": 2.8173,
+      "step": 45060
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7107488512992859,
+      "learning_rate": 9.73234106850263e-05,
+      "loss": 2.9891,
+      "step": 45061
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7031468152999878,
+      "learning_rate": 9.731205966743058e-05,
+      "loss": 2.7615,
+      "step": 45062
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6482835412025452,
+      "learning_rate": 9.730070918367598e-05,
+      "loss": 2.8392,
+      "step": 45063
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.696492612361908,
+      "learning_rate": 9.728935923379234e-05,
+      "loss": 2.9115,
+      "step": 45064
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6701128482818604,
+      "learning_rate": 9.727800981780946e-05,
+      "loss": 2.6987,
+      "step": 45065
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6930135488510132,
+      "learning_rate": 9.726666093575723e-05,
+      "loss": 3.1348,
+      "step": 45066
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7013523578643799,
+      "learning_rate": 9.725531258766555e-05,
+      "loss": 2.9139,
+      "step": 45067
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6763277649879456,
+      "learning_rate": 9.72439647735645e-05,
+      "loss": 2.9529,
+      "step": 45068
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7029493451118469,
+      "learning_rate": 9.723261749348373e-05,
+      "loss": 2.7961,
+      "step": 45069
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6882089376449585,
+      "learning_rate": 9.72212707474533e-05,
+      "loss": 2.9743,
+      "step": 45070
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7075545191764832,
+      "learning_rate": 9.720992453550292e-05,
+      "loss": 3.0006,
+      "step": 45071
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7261462211608887,
+      "learning_rate": 9.719857885766266e-05,
+      "loss": 2.9622,
+      "step": 45072
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7145310044288635,
+      "learning_rate": 9.718723371396234e-05,
+      "loss": 2.9799,
+      "step": 45073
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7117692828178406,
+      "learning_rate": 9.71758891044317e-05,
+      "loss": 2.9716,
+      "step": 45074
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7104676961898804,
+      "learning_rate": 9.716454502910082e-05,
+      "loss": 3.0304,
+      "step": 45075
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7376676797866821,
+      "learning_rate": 9.715320148799942e-05,
+      "loss": 2.8945,
+      "step": 45076
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7351451516151428,
+      "learning_rate": 9.714185848115746e-05,
+      "loss": 2.8031,
+      "step": 45077
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7405024766921997,
+      "learning_rate": 9.713051600860486e-05,
+      "loss": 2.973,
+      "step": 45078
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7368142604827881,
+      "learning_rate": 9.711917407037147e-05,
+      "loss": 3.0478,
+      "step": 45079
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.756643533706665,
+      "learning_rate": 9.71078326664871e-05,
+      "loss": 3.123,
+      "step": 45080
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.742498517036438,
+      "learning_rate": 9.709649179698156e-05,
+      "loss": 2.9582,
+      "step": 45081
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7527225017547607,
+      "learning_rate": 9.708515146188482e-05,
+      "loss": 2.8509,
+      "step": 45082
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7525281310081482,
+      "learning_rate": 9.707381166122685e-05,
+      "loss": 3.034,
+      "step": 45083
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.69659823179245,
+      "learning_rate": 9.70624723950373e-05,
+      "loss": 2.8755,
+      "step": 45084
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7091747522354126,
+      "learning_rate": 9.705113366334622e-05,
+      "loss": 2.8933,
+      "step": 45085
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6765292882919312,
+      "learning_rate": 9.70397954661834e-05,
+      "loss": 2.8283,
+      "step": 45086
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7405532002449036,
+      "learning_rate": 9.702845780357864e-05,
+      "loss": 2.9788,
+      "step": 45087
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7477160692214966,
+      "learning_rate": 9.701712067556192e-05,
+      "loss": 2.8926,
+      "step": 45088
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7491059303283691,
+      "learning_rate": 9.700578408216296e-05,
+      "loss": 2.7266,
+      "step": 45089
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6882579326629639,
+      "learning_rate": 9.69944480234118e-05,
+      "loss": 3.1125,
+      "step": 45090
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7268614768981934,
+      "learning_rate": 9.698311249933807e-05,
+      "loss": 2.9157,
+      "step": 45091
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7116838693618774,
+      "learning_rate": 9.697177750997175e-05,
+      "loss": 2.7934,
+      "step": 45092
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.718450129032135,
+      "learning_rate": 9.696044305534281e-05,
+      "loss": 2.9242,
+      "step": 45093
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7193343043327332,
+      "learning_rate": 9.694910913548099e-05,
+      "loss": 2.7036,
+      "step": 45094
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6873836517333984,
+      "learning_rate": 9.693777575041613e-05,
+      "loss": 2.9068,
+      "step": 45095
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7232233881950378,
+      "learning_rate": 9.692644290017801e-05,
+      "loss": 3.0087,
+      "step": 45096
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6899810433387756,
+      "learning_rate": 9.691511058479653e-05,
+      "loss": 3.0144,
+      "step": 45097
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7038057446479797,
+      "learning_rate": 9.690377880430168e-05,
+      "loss": 2.8663,
+      "step": 45098
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6957258582115173,
+      "learning_rate": 9.689244755872308e-05,
+      "loss": 3.0446,
+      "step": 45099
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7041094899177551,
+      "learning_rate": 9.688111684809078e-05,
+      "loss": 2.7151,
+      "step": 45100
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.668713390827179,
+      "learning_rate": 9.686978667243453e-05,
+      "loss": 3.0839,
+      "step": 45101
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7322125434875488,
+      "learning_rate": 9.685845703178408e-05,
+      "loss": 2.8446,
+      "step": 45102
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7117242217063904,
+      "learning_rate": 9.684712792616946e-05,
+      "loss": 3.0491,
+      "step": 45103
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.724531352519989,
+      "learning_rate": 9.683579935562029e-05,
+      "loss": 2.8113,
+      "step": 45104
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7037317752838135,
+      "learning_rate": 9.682447132016664e-05,
+      "loss": 2.9621,
+      "step": 45105
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7283707857131958,
+      "learning_rate": 9.681314381983811e-05,
+      "loss": 3.037,
+      "step": 45106
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6818432807922363,
+      "learning_rate": 9.680181685466477e-05,
+      "loss": 2.7621,
+      "step": 45107
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6957379579544067,
+      "learning_rate": 9.679049042467636e-05,
+      "loss": 2.7626,
+      "step": 45108
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6879173517227173,
+      "learning_rate": 9.677916452990255e-05,
+      "loss": 2.795,
+      "step": 45109
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6849760413169861,
+      "learning_rate": 9.676783917037343e-05,
+      "loss": 2.8795,
+      "step": 45110
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7054252028465271,
+      "learning_rate": 9.67565143461186e-05,
+      "loss": 2.7672,
+      "step": 45111
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6849786043167114,
+      "learning_rate": 9.674519005716809e-05,
+      "loss": 2.8216,
+      "step": 45112
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6915918588638306,
+      "learning_rate": 9.673386630355158e-05,
+      "loss": 2.9443,
+      "step": 45113
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6977018117904663,
+      "learning_rate": 9.672254308529891e-05,
+      "loss": 2.933,
+      "step": 45114
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7735402584075928,
+      "learning_rate": 9.671122040244004e-05,
+      "loss": 2.7578,
+      "step": 45115
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7309532165527344,
+      "learning_rate": 9.669989825500467e-05,
+      "loss": 2.9269,
+      "step": 45116
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7018609046936035,
+      "learning_rate": 9.668857664302269e-05,
+      "loss": 2.8062,
+      "step": 45117
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7004704475402832,
+      "learning_rate": 9.667725556652372e-05,
+      "loss": 2.9491,
+      "step": 45118
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7916359305381775,
+      "learning_rate": 9.666593502553776e-05,
+      "loss": 3.0852,
+      "step": 45119
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7221834659576416,
+      "learning_rate": 9.665461502009468e-05,
+      "loss": 2.9854,
+      "step": 45120
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.706676721572876,
+      "learning_rate": 9.664329555022415e-05,
+      "loss": 2.9534,
+      "step": 45121
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7531946301460266,
+      "learning_rate": 9.663197661595609e-05,
+      "loss": 2.8679,
+      "step": 45122
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7091394066810608,
+      "learning_rate": 9.66206582173203e-05,
+      "loss": 3.028,
+      "step": 45123
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7693554162979126,
+      "learning_rate": 9.660934035434644e-05,
+      "loss": 2.9596,
+      "step": 45124
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7340413928031921,
+      "learning_rate": 9.659802302706452e-05,
+      "loss": 2.7628,
+      "step": 45125
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7072880864143372,
+      "learning_rate": 9.65867062355042e-05,
+      "loss": 2.7449,
+      "step": 45126
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6786409020423889,
+      "learning_rate": 9.657538997969542e-05,
+      "loss": 2.7573,
+      "step": 45127
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6778995394706726,
+      "learning_rate": 9.656407425966782e-05,
+      "loss": 3.0114,
+      "step": 45128
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7381721138954163,
+      "learning_rate": 9.655275907545139e-05,
+      "loss": 2.7733,
+      "step": 45129
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7340525984764099,
+      "learning_rate": 9.654144442707583e-05,
+      "loss": 2.7722,
+      "step": 45130
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.72030109167099,
+      "learning_rate": 9.653013031457089e-05,
+      "loss": 2.9553,
+      "step": 45131
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7066547870635986,
+      "learning_rate": 9.65188167379665e-05,
+      "loss": 2.9542,
+      "step": 45132
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7587066292762756,
+      "learning_rate": 9.650750369729228e-05,
+      "loss": 2.8834,
+      "step": 45133
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7313148379325867,
+      "learning_rate": 9.649619119257813e-05,
+      "loss": 2.7555,
+      "step": 45134
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.713611900806427,
+      "learning_rate": 9.648487922385397e-05,
+      "loss": 2.8716,
+      "step": 45135
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6982900500297546,
+      "learning_rate": 9.647356779114934e-05,
+      "loss": 2.9958,
+      "step": 45136
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7607917189598083,
+      "learning_rate": 9.646225689449428e-05,
+      "loss": 2.9409,
+      "step": 45137
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6986991763114929,
+      "learning_rate": 9.645094653391848e-05,
+      "loss": 2.8637,
+      "step": 45138
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7355843782424927,
+      "learning_rate": 9.643963670945159e-05,
+      "loss": 2.914,
+      "step": 45139
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.746160089969635,
+      "learning_rate": 9.642832742112363e-05,
+      "loss": 2.9648,
+      "step": 45140
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.718352198600769,
+      "learning_rate": 9.641701866896418e-05,
+      "loss": 2.7866,
+      "step": 45141
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6826595067977905,
+      "learning_rate": 9.640571045300323e-05,
+      "loss": 2.7788,
+      "step": 45142
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7046639919281006,
+      "learning_rate": 9.639440277327033e-05,
+      "loss": 2.8692,
+      "step": 45143
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6994836926460266,
+      "learning_rate": 9.638309562979551e-05,
+      "loss": 2.9952,
+      "step": 45144
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7048764824867249,
+      "learning_rate": 9.637178902260843e-05,
+      "loss": 3.0218,
+      "step": 45145
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6403487920761108,
+      "learning_rate": 9.636048295173873e-05,
+      "loss": 2.8368,
+      "step": 45146
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7118167281150818,
+      "learning_rate": 9.634917741721648e-05,
+      "loss": 2.7272,
+      "step": 45147
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7441112399101257,
+      "learning_rate": 9.633787241907117e-05,
+      "loss": 3.164,
+      "step": 45148
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.707645833492279,
+      "learning_rate": 9.632656795733271e-05,
+      "loss": 3.0041,
+      "step": 45149
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6987390518188477,
+      "learning_rate": 9.631526403203098e-05,
+      "loss": 2.8839,
+      "step": 45150
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7065211534500122,
+      "learning_rate": 9.630396064319565e-05,
+      "loss": 2.8892,
+      "step": 45151
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7638340592384338,
+      "learning_rate": 9.629265779085647e-05,
+      "loss": 2.8871,
+      "step": 45152
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7217139005661011,
+      "learning_rate": 9.628135547504315e-05,
+      "loss": 3.0046,
+      "step": 45153
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7055973410606384,
+      "learning_rate": 9.627005369578553e-05,
+      "loss": 2.738,
+      "step": 45154
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7258812785148621,
+      "learning_rate": 9.625875245311345e-05,
+      "loss": 3.0541,
+      "step": 45155
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7114151120185852,
+      "learning_rate": 9.624745174705655e-05,
+      "loss": 3.0669,
+      "step": 45156
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6896607875823975,
+      "learning_rate": 9.623615157764473e-05,
+      "loss": 2.8723,
+      "step": 45157
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7319231033325195,
+      "learning_rate": 9.622485194490757e-05,
+      "loss": 3.1262,
+      "step": 45158
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6564500331878662,
+      "learning_rate": 9.621355284887502e-05,
+      "loss": 2.8374,
+      "step": 45159
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7160139083862305,
+      "learning_rate": 9.620225428957677e-05,
+      "loss": 2.93,
+      "step": 45160
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7266965508460999,
+      "learning_rate": 9.619095626704244e-05,
+      "loss": 2.8838,
+      "step": 45161
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7377114295959473,
+      "learning_rate": 9.617965878130203e-05,
+      "loss": 2.9046,
+      "step": 45162
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7144932746887207,
+      "learning_rate": 9.616836183238504e-05,
+      "loss": 3.0129,
+      "step": 45163
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6912056803703308,
+      "learning_rate": 9.61570654203214e-05,
+      "loss": 2.7707,
+      "step": 45164
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.730238139629364,
+      "learning_rate": 9.614576954514088e-05,
+      "loss": 3.1059,
+      "step": 45165
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6824841499328613,
+      "learning_rate": 9.613447420687319e-05,
+      "loss": 2.7708,
+      "step": 45166
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7248892188072205,
+      "learning_rate": 9.612317940554806e-05,
+      "loss": 2.8777,
+      "step": 45167
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7106010913848877,
+      "learning_rate": 9.611188514119515e-05,
+      "loss": 2.8532,
+      "step": 45168
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7336381077766418,
+      "learning_rate": 9.610059141384427e-05,
+      "loss": 2.8722,
+      "step": 45169
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7006831765174866,
+      "learning_rate": 9.60892982235253e-05,
+      "loss": 2.8112,
+      "step": 45170
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7395011782646179,
+      "learning_rate": 9.607800557026777e-05,
+      "loss": 2.9595,
+      "step": 45171
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.730022668838501,
+      "learning_rate": 9.606671345410164e-05,
+      "loss": 3.002,
+      "step": 45172
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7525355815887451,
+      "learning_rate": 9.605542187505643e-05,
+      "loss": 2.8871,
+      "step": 45173
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6975340247154236,
+      "learning_rate": 9.60441308331621e-05,
+      "loss": 2.8554,
+      "step": 45174
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7521911859512329,
+      "learning_rate": 9.603284032844828e-05,
+      "loss": 2.772,
+      "step": 45175
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7508547306060791,
+      "learning_rate": 9.60215503609446e-05,
+      "loss": 2.8397,
+      "step": 45176
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7319634556770325,
+      "learning_rate": 9.601026093068098e-05,
+      "loss": 2.9338,
+      "step": 45177
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6809556484222412,
+      "learning_rate": 9.5998972037687e-05,
+      "loss": 2.8386,
+      "step": 45178
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7330763936042786,
+      "learning_rate": 9.598768368199245e-05,
+      "loss": 2.6866,
+      "step": 45179
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7155676484107971,
+      "learning_rate": 9.59763958636272e-05,
+      "loss": 2.8636,
+      "step": 45180
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6808676719665527,
+      "learning_rate": 9.596510858262085e-05,
+      "loss": 2.9295,
+      "step": 45181
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7002042531967163,
+      "learning_rate": 9.595382183900312e-05,
+      "loss": 3.0021,
+      "step": 45182
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7323654294013977,
+      "learning_rate": 9.594253563280364e-05,
+      "loss": 2.8457,
+      "step": 45183
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.677535891532898,
+      "learning_rate": 9.593124996405228e-05,
+      "loss": 2.9339,
+      "step": 45184
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7651351094245911,
+      "learning_rate": 9.591996483277882e-05,
+      "loss": 2.7807,
+      "step": 45185
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6927827000617981,
+      "learning_rate": 9.590868023901283e-05,
+      "loss": 2.9166,
+      "step": 45186
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.731319010257721,
+      "learning_rate": 9.589739618278415e-05,
+      "loss": 2.9789,
+      "step": 45187
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7109827995300293,
+      "learning_rate": 9.588611266412247e-05,
+      "loss": 2.8897,
+      "step": 45188
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7135148048400879,
+      "learning_rate": 9.58748296830574e-05,
+      "loss": 2.9604,
+      "step": 45189
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6991270184516907,
+      "learning_rate": 9.586354723961883e-05,
+      "loss": 3.079,
+      "step": 45190
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.692098081111908,
+      "learning_rate": 9.585226533383627e-05,
+      "loss": 2.877,
+      "step": 45191
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7232987880706787,
+      "learning_rate": 9.584098396573969e-05,
+      "loss": 3.0033,
+      "step": 45192
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7077509760856628,
+      "learning_rate": 9.582970313535857e-05,
+      "loss": 3.0322,
+      "step": 45193
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7513511776924133,
+      "learning_rate": 9.581842284272279e-05,
+      "loss": 2.9617,
+      "step": 45194
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7315356135368347,
+      "learning_rate": 9.58071430878619e-05,
+      "loss": 3.0344,
+      "step": 45195
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.745606005191803,
+      "learning_rate": 9.57958638708058e-05,
+      "loss": 2.9354,
+      "step": 45196
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7538384795188904,
+      "learning_rate": 9.578458519158409e-05,
+      "loss": 3.0113,
+      "step": 45197
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7272838354110718,
+      "learning_rate": 9.577330705022637e-05,
+      "loss": 2.9533,
+      "step": 45198
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6927639842033386,
+      "learning_rate": 9.576202944676254e-05,
+      "loss": 3.0826,
+      "step": 45199
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6876212358474731,
+      "learning_rate": 9.575075238122217e-05,
+      "loss": 2.7837,
+      "step": 45200
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.721787691116333,
+      "learning_rate": 9.573947585363499e-05,
+      "loss": 2.7393,
+      "step": 45201
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7061600685119629,
+      "learning_rate": 9.572819986403079e-05,
+      "loss": 2.8766,
+      "step": 45202
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7322124242782593,
+      "learning_rate": 9.571692441243925e-05,
+      "loss": 2.9802,
+      "step": 45203
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7068262696266174,
+      "learning_rate": 9.570564949888998e-05,
+      "loss": 3.033,
+      "step": 45204
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7690457105636597,
+      "learning_rate": 9.569437512341261e-05,
+      "loss": 2.9735,
+      "step": 45205
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7033405303955078,
+      "learning_rate": 9.568310128603696e-05,
+      "loss": 3.0367,
+      "step": 45206
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7443763613700867,
+      "learning_rate": 9.567182798679278e-05,
+      "loss": 2.7348,
+      "step": 45207
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7035591006278992,
+      "learning_rate": 9.56605552257096e-05,
+      "loss": 2.8969,
+      "step": 45208
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.693112313747406,
+      "learning_rate": 9.564928300281726e-05,
+      "loss": 2.8667,
+      "step": 45209
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7365559339523315,
+      "learning_rate": 9.563801131814539e-05,
+      "loss": 2.7897,
+      "step": 45210
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7358099818229675,
+      "learning_rate": 9.56267401717236e-05,
+      "loss": 3.0661,
+      "step": 45211
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7144678235054016,
+      "learning_rate": 9.56154695635817e-05,
+      "loss": 3.0905,
+      "step": 45212
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7409480810165405,
+      "learning_rate": 9.560419949374927e-05,
+      "loss": 3.028,
+      "step": 45213
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7363123297691345,
+      "learning_rate": 9.55929299622561e-05,
+      "loss": 2.9172,
+      "step": 45214
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7069172263145447,
+      "learning_rate": 9.558166096913173e-05,
+      "loss": 2.9742,
+      "step": 45215
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7248828411102295,
+      "learning_rate": 9.55703925144059e-05,
+      "loss": 3.0391,
+      "step": 45216
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7333267331123352,
+      "learning_rate": 9.555912459810845e-05,
+      "loss": 2.8933,
+      "step": 45217
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6944749355316162,
+      "learning_rate": 9.554785722026892e-05,
+      "loss": 2.8961,
+      "step": 45218
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7157734036445618,
+      "learning_rate": 9.553659038091695e-05,
+      "loss": 2.8081,
+      "step": 45219
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.743134081363678,
+      "learning_rate": 9.552532408008216e-05,
+      "loss": 2.7779,
+      "step": 45220
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6870113611221313,
+      "learning_rate": 9.551405831779431e-05,
+      "loss": 2.8549,
+      "step": 45221
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7241684198379517,
+      "learning_rate": 9.550279309408321e-05,
+      "loss": 2.9022,
+      "step": 45222
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6814631819725037,
+      "learning_rate": 9.549152840897827e-05,
+      "loss": 2.82,
+      "step": 45223
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6824125647544861,
+      "learning_rate": 9.54802642625094e-05,
+      "loss": 3.0936,
+      "step": 45224
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7024969458580017,
+      "learning_rate": 9.546900065470615e-05,
+      "loss": 2.9814,
+      "step": 45225
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7450183629989624,
+      "learning_rate": 9.545773758559809e-05,
+      "loss": 2.8562,
+      "step": 45226
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6997722387313843,
+      "learning_rate": 9.544647505521508e-05,
+      "loss": 3.0313,
+      "step": 45227
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7680618166923523,
+      "learning_rate": 9.543521306358657e-05,
+      "loss": 3.1848,
+      "step": 45228
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7047203779220581,
+      "learning_rate": 9.542395161074246e-05,
+      "loss": 2.8618,
+      "step": 45229
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6786448359489441,
+      "learning_rate": 9.541269069671221e-05,
+      "loss": 2.9918,
+      "step": 45230
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7172592878341675,
+      "learning_rate": 9.540143032152564e-05,
+      "loss": 3.0713,
+      "step": 45231
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7199307680130005,
+      "learning_rate": 9.539017048521233e-05,
+      "loss": 3.0211,
+      "step": 45232
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7008538842201233,
+      "learning_rate": 9.537891118780181e-05,
+      "loss": 2.8901,
+      "step": 45233
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6756240129470825,
+      "learning_rate": 9.536765242932396e-05,
+      "loss": 2.9633,
+      "step": 45234
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6890457272529602,
+      "learning_rate": 9.535639420980826e-05,
+      "loss": 2.9048,
+      "step": 45235
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6927510499954224,
+      "learning_rate": 9.534513652928442e-05,
+      "loss": 2.8731,
+      "step": 45236
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6908006072044373,
+      "learning_rate": 9.53338793877822e-05,
+      "loss": 2.7577,
+      "step": 45237
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7486088871955872,
+      "learning_rate": 9.532262278533107e-05,
+      "loss": 2.8153,
+      "step": 45238
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7346619963645935,
+      "learning_rate": 9.531136672196082e-05,
+      "loss": 2.8707,
+      "step": 45239
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7260370254516602,
+      "learning_rate": 9.530011119770107e-05,
+      "loss": 2.7412,
+      "step": 45240
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.743890106678009,
+      "learning_rate": 9.528885621258134e-05,
+      "loss": 2.8874,
+      "step": 45241
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7786975502967834,
+      "learning_rate": 9.527760176663143e-05,
+      "loss": 2.8397,
+      "step": 45242
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6872302293777466,
+      "learning_rate": 9.526634785988087e-05,
+      "loss": 2.9666,
+      "step": 45243
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6927504539489746,
+      "learning_rate": 9.52550944923594e-05,
+      "loss": 2.873,
+      "step": 45244
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7176833152770996,
+      "learning_rate": 9.524384166409653e-05,
+      "loss": 3.0863,
+      "step": 45245
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.71924889087677,
+      "learning_rate": 9.523258937512204e-05,
+      "loss": 2.7674,
+      "step": 45246
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6952587962150574,
+      "learning_rate": 9.522133762546555e-05,
+      "loss": 2.9272,
+      "step": 45247
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7094781398773193,
+      "learning_rate": 9.521008641515653e-05,
+      "loss": 2.8632,
+      "step": 45248
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7094677686691284,
+      "learning_rate": 9.519883574422481e-05,
+      "loss": 2.9484,
+      "step": 45249
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6779662370681763,
+      "learning_rate": 9.518758561269988e-05,
+      "loss": 2.7069,
+      "step": 45250
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6990089416503906,
+      "learning_rate": 9.517633602061141e-05,
+      "loss": 3.1154,
+      "step": 45251
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7468245625495911,
+      "learning_rate": 9.516508696798914e-05,
+      "loss": 3.0644,
+      "step": 45252
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7292611598968506,
+      "learning_rate": 9.515383845486263e-05,
+      "loss": 3.1084,
+      "step": 45253
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6730436682701111,
+      "learning_rate": 9.514259048126147e-05,
+      "loss": 2.9853,
+      "step": 45254
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6685360074043274,
+      "learning_rate": 9.513134304721521e-05,
+      "loss": 2.9323,
+      "step": 45255
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7453835606575012,
+      "learning_rate": 9.512009615275356e-05,
+      "loss": 2.8677,
+      "step": 45256
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7116197347640991,
+      "learning_rate": 9.510884979790624e-05,
+      "loss": 2.7648,
+      "step": 45257
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7128415703773499,
+      "learning_rate": 9.50976039827027e-05,
+      "loss": 2.8006,
+      "step": 45258
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6778522729873657,
+      "learning_rate": 9.508635870717272e-05,
+      "loss": 3.0609,
+      "step": 45259
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6794968843460083,
+      "learning_rate": 9.507511397134574e-05,
+      "loss": 2.8762,
+      "step": 45260
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6920288801193237,
+      "learning_rate": 9.506386977525156e-05,
+      "loss": 2.7519,
+      "step": 45261
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7245926856994629,
+      "learning_rate": 9.50526261189197e-05,
+      "loss": 2.9135,
+      "step": 45262
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7115115523338318,
+      "learning_rate": 9.50413830023797e-05,
+      "loss": 3.035,
+      "step": 45263
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6681340336799622,
+      "learning_rate": 9.503014042566135e-05,
+      "loss": 2.8618,
+      "step": 45264
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6481229662895203,
+      "learning_rate": 9.501889838879407e-05,
+      "loss": 2.9743,
+      "step": 45265
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7262416481971741,
+      "learning_rate": 9.500765689180754e-05,
+      "loss": 2.9994,
+      "step": 45266
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7244082093238831,
+      "learning_rate": 9.499641593473152e-05,
+      "loss": 2.7771,
+      "step": 45267
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7363994121551514,
+      "learning_rate": 9.498517551759548e-05,
+      "loss": 3.0425,
+      "step": 45268
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7266939878463745,
+      "learning_rate": 9.497393564042901e-05,
+      "loss": 2.9354,
+      "step": 45269
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7804268002510071,
+      "learning_rate": 9.496269630326165e-05,
+      "loss": 2.9353,
+      "step": 45270
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6953659653663635,
+      "learning_rate": 9.495145750612319e-05,
+      "loss": 3.292,
+      "step": 45271
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6974464654922485,
+      "learning_rate": 9.494021924904302e-05,
+      "loss": 2.9793,
+      "step": 45272
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7132477760314941,
+      "learning_rate": 9.492898153205085e-05,
+      "loss": 2.9696,
+      "step": 45273
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6814945340156555,
+      "learning_rate": 9.491774435517637e-05,
+      "loss": 2.5446,
+      "step": 45274
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6983672976493835,
+      "learning_rate": 9.490650771844899e-05,
+      "loss": 2.9157,
+      "step": 45275
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7214673161506653,
+      "learning_rate": 9.48952716218986e-05,
+      "loss": 2.794,
+      "step": 45276
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7465963363647461,
+      "learning_rate": 9.48840360655544e-05,
+      "loss": 2.991,
+      "step": 45277
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7039017677307129,
+      "learning_rate": 9.487280104944616e-05,
+      "loss": 2.845,
+      "step": 45278
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.737330436706543,
+      "learning_rate": 9.486156657360357e-05,
+      "loss": 2.8189,
+      "step": 45279
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6810418367385864,
+      "learning_rate": 9.485033263805607e-05,
+      "loss": 3.0818,
+      "step": 45280
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6977468729019165,
+      "learning_rate": 9.483909924283338e-05,
+      "loss": 2.782,
+      "step": 45281
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7160059213638306,
+      "learning_rate": 9.482786638796496e-05,
+      "loss": 2.9945,
+      "step": 45282
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7127458453178406,
+      "learning_rate": 9.48166340734805e-05,
+      "loss": 2.7842,
+      "step": 45283
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6878350973129272,
+      "learning_rate": 9.480540229940956e-05,
+      "loss": 2.7996,
+      "step": 45284
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6775646209716797,
+      "learning_rate": 9.47941710657816e-05,
+      "loss": 2.926,
+      "step": 45285
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7190422415733337,
+      "learning_rate": 9.478294037262642e-05,
+      "loss": 2.9772,
+      "step": 45286
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6985235214233398,
+      "learning_rate": 9.477171021997335e-05,
+      "loss": 3.0118,
+      "step": 45287
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7282534241676331,
+      "learning_rate": 9.476048060785213e-05,
+      "loss": 2.8941,
+      "step": 45288
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7488526701927185,
+      "learning_rate": 9.474925153629238e-05,
+      "loss": 3.0031,
+      "step": 45289
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7149317860603333,
+      "learning_rate": 9.473802300532361e-05,
+      "loss": 3.0255,
+      "step": 45290
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6947160959243774,
+      "learning_rate": 9.472679501497538e-05,
+      "loss": 2.9155,
+      "step": 45291
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7043709754943848,
+      "learning_rate": 9.471556756527716e-05,
+      "loss": 2.9663,
+      "step": 45292
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7300408482551575,
+      "learning_rate": 9.470434065625865e-05,
+      "loss": 2.7279,
+      "step": 45293
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6901246309280396,
+      "learning_rate": 9.469311428794948e-05,
+      "loss": 2.9823,
+      "step": 45294
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.719524621963501,
+      "learning_rate": 9.468188846037904e-05,
+      "loss": 3.1144,
+      "step": 45295
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7352761030197144,
+      "learning_rate": 9.467066317357707e-05,
+      "loss": 2.8261,
+      "step": 45296
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7071006894111633,
+      "learning_rate": 9.4659438427573e-05,
+      "loss": 3.0008,
+      "step": 45297
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7212498188018799,
+      "learning_rate": 9.46482142223965e-05,
+      "loss": 2.8313,
+      "step": 45298
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7842340469360352,
+      "learning_rate": 9.463699055807713e-05,
+      "loss": 2.8263,
+      "step": 45299
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6890537142753601,
+      "learning_rate": 9.462576743464426e-05,
+      "loss": 2.9741,
+      "step": 45300
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7267663478851318,
+      "learning_rate": 9.461454485212774e-05,
+      "loss": 2.9529,
+      "step": 45301
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7003971934318542,
+      "learning_rate": 9.460332281055684e-05,
+      "loss": 2.9145,
+      "step": 45302
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7068154215812683,
+      "learning_rate": 9.45921013099613e-05,
+      "loss": 2.8808,
+      "step": 45303
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7120148539543152,
+      "learning_rate": 9.45808803503707e-05,
+      "loss": 2.8749,
+      "step": 45304
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6946321725845337,
+      "learning_rate": 9.456965993181453e-05,
+      "loss": 2.9578,
+      "step": 45305
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.699981689453125,
+      "learning_rate": 9.455844005432234e-05,
+      "loss": 2.8174,
+      "step": 45306
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6788482666015625,
+      "learning_rate": 9.45472207179236e-05,
+      "loss": 2.8667,
+      "step": 45307
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7016308307647705,
+      "learning_rate": 9.453600192264792e-05,
+      "loss": 2.7427,
+      "step": 45308
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7483291625976562,
+      "learning_rate": 9.452478366852498e-05,
+      "loss": 2.8937,
+      "step": 45309
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6895776987075806,
+      "learning_rate": 9.451356595558412e-05,
+      "loss": 2.9125,
+      "step": 45310
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7158215641975403,
+      "learning_rate": 9.450234878385505e-05,
+      "loss": 2.9937,
+      "step": 45311
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7149104475975037,
+      "learning_rate": 9.449113215336727e-05,
+      "loss": 2.8569,
+      "step": 45312
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7052308320999146,
+      "learning_rate": 9.447991606415021e-05,
+      "loss": 3.0051,
+      "step": 45313
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7073119878768921,
+      "learning_rate": 9.446870051623358e-05,
+      "loss": 2.9697,
+      "step": 45314
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7047192454338074,
+      "learning_rate": 9.445748550964675e-05,
+      "loss": 3.0161,
+      "step": 45315
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7001875042915344,
+      "learning_rate": 9.444627104441944e-05,
+      "loss": 2.9102,
+      "step": 45316
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6942550539970398,
+      "learning_rate": 9.443505712058098e-05,
+      "loss": 3.0716,
+      "step": 45317
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6810586452484131,
+      "learning_rate": 9.4423843738161e-05,
+      "loss": 2.7442,
+      "step": 45318
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.688390851020813,
+      "learning_rate": 9.441263089718919e-05,
+      "loss": 2.7613,
+      "step": 45319
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7119620442390442,
+      "learning_rate": 9.440141859769494e-05,
+      "loss": 3.0136,
+      "step": 45320
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7164089679718018,
+      "learning_rate": 9.439020683970778e-05,
+      "loss": 2.7685,
+      "step": 45321
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7110692858695984,
+      "learning_rate": 9.437899562325715e-05,
+      "loss": 2.7599,
+      "step": 45322
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.8134690523147583,
+      "learning_rate": 9.436778494837264e-05,
+      "loss": 2.9939,
+      "step": 45323
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7323058843612671,
+      "learning_rate": 9.435657481508393e-05,
+      "loss": 2.9119,
+      "step": 45324
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6736330389976501,
+      "learning_rate": 9.434536522342033e-05,
+      "loss": 3.0363,
+      "step": 45325
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.710964560508728,
+      "learning_rate": 9.433415617341156e-05,
+      "loss": 2.7434,
+      "step": 45326
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6646226644515991,
+      "learning_rate": 9.432294766508705e-05,
+      "loss": 3.0675,
+      "step": 45327
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7395647168159485,
+      "learning_rate": 9.43117396984762e-05,
+      "loss": 2.9635,
+      "step": 45328
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.691923975944519,
+      "learning_rate": 9.430053227360875e-05,
+      "loss": 2.7976,
+      "step": 45329
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6718361377716064,
+      "learning_rate": 9.428932539051399e-05,
+      "loss": 2.9561,
+      "step": 45330
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7138594388961792,
+      "learning_rate": 9.427811904922167e-05,
+      "loss": 3.0096,
+      "step": 45331
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7084429860115051,
+      "learning_rate": 9.426691324976113e-05,
+      "loss": 2.8364,
+      "step": 45332
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7155402898788452,
+      "learning_rate": 9.4255707992162e-05,
+      "loss": 3.0008,
+      "step": 45333
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7018200755119324,
+      "learning_rate": 9.424450327645374e-05,
+      "loss": 2.7881,
+      "step": 45334
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7858231663703918,
+      "learning_rate": 9.423329910266576e-05,
+      "loss": 2.9398,
+      "step": 45335
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.69808030128479,
+      "learning_rate": 9.42220954708278e-05,
+      "loss": 2.9118,
+      "step": 45336
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7413972020149231,
+      "learning_rate": 9.42108923809691e-05,
+      "loss": 2.9831,
+      "step": 45337
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7176880836486816,
+      "learning_rate": 9.419968983311934e-05,
+      "loss": 2.7619,
+      "step": 45338
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7590557336807251,
+      "learning_rate": 9.418848782730805e-05,
+      "loss": 3.0727,
+      "step": 45339
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6953451633453369,
+      "learning_rate": 9.417728636356459e-05,
+      "loss": 3.0103,
+      "step": 45340
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7350430488586426,
+      "learning_rate": 9.416608544191866e-05,
+      "loss": 3.0276,
+      "step": 45341
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6567845344543457,
+      "learning_rate": 9.415488506239964e-05,
+      "loss": 2.8636,
+      "step": 45342
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7473620772361755,
+      "learning_rate": 9.414368522503695e-05,
+      "loss": 2.96,
+      "step": 45343
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7204223871231079,
+      "learning_rate": 9.413248592986025e-05,
+      "loss": 3.0137,
+      "step": 45344
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7221093773841858,
+      "learning_rate": 9.412128717689888e-05,
+      "loss": 3.0327,
+      "step": 45345
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7267528176307678,
+      "learning_rate": 9.41100889661825e-05,
+      "loss": 2.7521,
+      "step": 45346
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7274415493011475,
+      "learning_rate": 9.409889129774043e-05,
+      "loss": 2.9669,
+      "step": 45347
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7493104934692383,
+      "learning_rate": 9.408769417160237e-05,
+      "loss": 2.8927,
+      "step": 45348
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7367200255393982,
+      "learning_rate": 9.407649758779766e-05,
+      "loss": 3.0401,
+      "step": 45349
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6605836749076843,
+      "learning_rate": 9.406530154635574e-05,
+      "loss": 2.9622,
+      "step": 45350
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7259601950645447,
+      "learning_rate": 9.405410604730629e-05,
+      "loss": 3.0617,
+      "step": 45351
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.0172241926193237,
+      "learning_rate": 9.40429110906786e-05,
+      "loss": 2.8292,
+      "step": 45352
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6789811849594116,
+      "learning_rate": 9.403171667650235e-05,
+      "loss": 3.0832,
+      "step": 45353
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7042773962020874,
+      "learning_rate": 9.402052280480677e-05,
+      "loss": 2.8346,
+      "step": 45354
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7092491388320923,
+      "learning_rate": 9.400932947562165e-05,
+      "loss": 2.7202,
+      "step": 45355
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7057832479476929,
+      "learning_rate": 9.399813668897625e-05,
+      "loss": 3.0595,
+      "step": 45356
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.751794159412384,
+      "learning_rate": 9.398694444490006e-05,
+      "loss": 2.8588,
+      "step": 45357
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7302895784378052,
+      "learning_rate": 9.39757527434227e-05,
+      "loss": 2.9267,
+      "step": 45358
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6791434288024902,
+      "learning_rate": 9.396456158457348e-05,
+      "loss": 2.9432,
+      "step": 45359
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7375596761703491,
+      "learning_rate": 9.395337096838191e-05,
+      "loss": 2.9125,
+      "step": 45360
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7245837450027466,
+      "learning_rate": 9.394218089487763e-05,
+      "loss": 2.8705,
+      "step": 45361
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7496141195297241,
+      "learning_rate": 9.393099136408989e-05,
+      "loss": 2.8422,
+      "step": 45362
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7486103773117065,
+      "learning_rate": 9.391980237604834e-05,
+      "loss": 2.9145,
+      "step": 45363
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6848008632659912,
+      "learning_rate": 9.390861393078237e-05,
+      "loss": 2.8837,
+      "step": 45364
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.70891273021698,
+      "learning_rate": 9.389742602832137e-05,
+      "loss": 2.8657,
+      "step": 45365
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7543848752975464,
+      "learning_rate": 9.388623866869498e-05,
+      "loss": 3.0147,
+      "step": 45366
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7381017804145813,
+      "learning_rate": 9.387505185193246e-05,
+      "loss": 2.9008,
+      "step": 45367
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7030778527259827,
+      "learning_rate": 9.38638655780635e-05,
+      "loss": 2.7483,
+      "step": 45368
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6804453730583191,
+      "learning_rate": 9.385267984711733e-05,
+      "loss": 2.8687,
+      "step": 45369
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7367046475410461,
+      "learning_rate": 9.384149465912362e-05,
+      "loss": 2.8755,
+      "step": 45370
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7181659936904907,
+      "learning_rate": 9.383031001411177e-05,
+      "loss": 3.0086,
+      "step": 45371
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7256984114646912,
+      "learning_rate": 9.381912591211108e-05,
+      "loss": 3.0784,
+      "step": 45372
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6880785226821899,
+      "learning_rate": 9.380794235315125e-05,
+      "loss": 3.0513,
+      "step": 45373
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7282678484916687,
+      "learning_rate": 9.379675933726149e-05,
+      "loss": 2.742,
+      "step": 45374
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7389535307884216,
+      "learning_rate": 9.378557686447138e-05,
+      "loss": 2.7666,
+      "step": 45375
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6855918169021606,
+      "learning_rate": 9.377439493481051e-05,
+      "loss": 2.9682,
+      "step": 45376
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6976229548454285,
+      "learning_rate": 9.376321354830815e-05,
+      "loss": 3.0365,
+      "step": 45377
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6974002122879028,
+      "learning_rate": 9.375203270499382e-05,
+      "loss": 2.7856,
+      "step": 45378
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7219595313072205,
+      "learning_rate": 9.374085240489686e-05,
+      "loss": 3.002,
+      "step": 45379
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7123830318450928,
+      "learning_rate": 9.372967264804678e-05,
+      "loss": 2.9279,
+      "step": 45380
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7480422258377075,
+      "learning_rate": 9.371849343447313e-05,
+      "loss": 2.6871,
+      "step": 45381
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7664074897766113,
+      "learning_rate": 9.370731476420521e-05,
+      "loss": 2.9509,
+      "step": 45382
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7201911211013794,
+      "learning_rate": 9.369613663727257e-05,
+      "loss": 2.9775,
+      "step": 45383
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7252894639968872,
+      "learning_rate": 9.368495905370454e-05,
+      "loss": 2.9017,
+      "step": 45384
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7937513589859009,
+      "learning_rate": 9.367378201353072e-05,
+      "loss": 3.3099,
+      "step": 45385
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7466810345649719,
+      "learning_rate": 9.366260551678043e-05,
+      "loss": 2.7903,
+      "step": 45386
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7079269289970398,
+      "learning_rate": 9.365142956348303e-05,
+      "loss": 2.878,
+      "step": 45387
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.75801682472229,
+      "learning_rate": 9.364025415366815e-05,
+      "loss": 2.7751,
+      "step": 45388
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7508972883224487,
+      "learning_rate": 9.362907928736504e-05,
+      "loss": 2.8401,
+      "step": 45389
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6906934380531311,
+      "learning_rate": 9.361790496460321e-05,
+      "loss": 2.8999,
+      "step": 45390
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7488749027252197,
+      "learning_rate": 9.36067311854122e-05,
+      "loss": 2.8891,
+      "step": 45391
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7199647426605225,
+      "learning_rate": 9.359555794982136e-05,
+      "loss": 2.8431,
+      "step": 45392
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7045316100120544,
+      "learning_rate": 9.358438525786006e-05,
+      "loss": 2.8693,
+      "step": 45393
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7193567156791687,
+      "learning_rate": 9.357321310955766e-05,
+      "loss": 3.1406,
+      "step": 45394
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6742414832115173,
+      "learning_rate": 9.356204150494374e-05,
+      "loss": 2.8068,
+      "step": 45395
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7182538509368896,
+      "learning_rate": 9.355087044404772e-05,
+      "loss": 3.1212,
+      "step": 45396
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7040326595306396,
+      "learning_rate": 9.353969992689891e-05,
+      "loss": 3.1036,
+      "step": 45397
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7186934947967529,
+      "learning_rate": 9.352852995352689e-05,
+      "loss": 2.7082,
+      "step": 45398
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7571741342544556,
+      "learning_rate": 9.351736052396088e-05,
+      "loss": 3.0488,
+      "step": 45399
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7321252822875977,
+      "learning_rate": 9.350619163823051e-05,
+      "loss": 2.9676,
+      "step": 45400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7001104950904846,
+      "learning_rate": 9.349502329636505e-05,
+      "loss": 2.819,
+      "step": 45401
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.718051016330719,
+      "learning_rate": 9.34838554983939e-05,
+      "loss": 2.897,
+      "step": 45402
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7028896808624268,
+      "learning_rate": 9.347268824434662e-05,
+      "loss": 2.8919,
+      "step": 45403
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7708454728126526,
+      "learning_rate": 9.346152153425245e-05,
+      "loss": 2.753,
+      "step": 45404
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.712142288684845,
+      "learning_rate": 9.345035536814087e-05,
+      "loss": 2.8239,
+      "step": 45405
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7452940344810486,
+      "learning_rate": 9.34391897460414e-05,
+      "loss": 2.6923,
+      "step": 45406
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7222115993499756,
+      "learning_rate": 9.342802466798334e-05,
+      "loss": 2.8162,
+      "step": 45407
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7443071603775024,
+      "learning_rate": 9.341686013399609e-05,
+      "loss": 3.0949,
+      "step": 45408
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7208560109138489,
+      "learning_rate": 9.340569614410902e-05,
+      "loss": 3.0639,
+      "step": 45409
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7204750180244446,
+      "learning_rate": 9.339453269835158e-05,
+      "loss": 2.9431,
+      "step": 45410
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6994853615760803,
+      "learning_rate": 9.338336979675323e-05,
+      "loss": 2.8151,
+      "step": 45411
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6930583119392395,
+      "learning_rate": 9.337220743934328e-05,
+      "loss": 3.2522,
+      "step": 45412
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.783711314201355,
+      "learning_rate": 9.336104562615123e-05,
+      "loss": 2.9764,
+      "step": 45413
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.766595721244812,
+      "learning_rate": 9.334988435720643e-05,
+      "loss": 2.9608,
+      "step": 45414
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7293518781661987,
+      "learning_rate": 9.333872363253817e-05,
+      "loss": 3.0567,
+      "step": 45415
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6588782668113708,
+      "learning_rate": 9.332756345217604e-05,
+      "loss": 3.0559,
+      "step": 45416
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7530074715614319,
+      "learning_rate": 9.331640381614923e-05,
+      "loss": 2.9012,
+      "step": 45417
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.740515410900116,
+      "learning_rate": 9.330524472448735e-05,
+      "loss": 2.8191,
+      "step": 45418
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7402000427246094,
+      "learning_rate": 9.329408617721953e-05,
+      "loss": 2.8501,
+      "step": 45419
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7017174959182739,
+      "learning_rate": 9.328292817437535e-05,
+      "loss": 3.007,
+      "step": 45420
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7643221616744995,
+      "learning_rate": 9.327177071598425e-05,
+      "loss": 3.0573,
+      "step": 45421
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7307689785957336,
+      "learning_rate": 9.326061380207548e-05,
+      "loss": 3.0476,
+      "step": 45422
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7625921368598938,
+      "learning_rate": 9.324945743267852e-05,
+      "loss": 2.6859,
+      "step": 45423
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7725873589515686,
+      "learning_rate": 9.323830160782256e-05,
+      "loss": 2.987,
+      "step": 45424
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7039408683776855,
+      "learning_rate": 9.322714632753721e-05,
+      "loss": 3.0487,
+      "step": 45425
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.725426435470581,
+      "learning_rate": 9.321599159185171e-05,
+      "loss": 2.9296,
+      "step": 45426
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7159402966499329,
+      "learning_rate": 9.320483740079549e-05,
+      "loss": 2.8903,
+      "step": 45427
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7357923984527588,
+      "learning_rate": 9.319368375439798e-05,
+      "loss": 2.6201,
+      "step": 45428
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7285069227218628,
+      "learning_rate": 9.318253065268856e-05,
+      "loss": 2.908,
+      "step": 45429
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7228646278381348,
+      "learning_rate": 9.317137809569651e-05,
+      "loss": 3.1464,
+      "step": 45430
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6842017769813538,
+      "learning_rate": 9.316022608345115e-05,
+      "loss": 3.0632,
+      "step": 45431
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7215100526809692,
+      "learning_rate": 9.314907461598198e-05,
+      "loss": 2.8692,
+      "step": 45432
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6963567137718201,
+      "learning_rate": 9.31379236933184e-05,
+      "loss": 2.8567,
+      "step": 45433
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6841099262237549,
+      "learning_rate": 9.312677331548965e-05,
+      "loss": 2.6787,
+      "step": 45434
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6942437291145325,
+      "learning_rate": 9.311562348252523e-05,
+      "loss": 3.0439,
+      "step": 45435
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7410614490509033,
+      "learning_rate": 9.310447419445446e-05,
+      "loss": 2.9797,
+      "step": 45436
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.685966432094574,
+      "learning_rate": 9.30933254513066e-05,
+      "loss": 2.8331,
+      "step": 45437
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7045655846595764,
+      "learning_rate": 9.30821772531112e-05,
+      "loss": 2.9347,
+      "step": 45438
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6830521821975708,
+      "learning_rate": 9.307102959989741e-05,
+      "loss": 2.9615,
+      "step": 45439
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6648380756378174,
+      "learning_rate": 9.305988249169481e-05,
+      "loss": 3.034,
+      "step": 45440
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7092429995536804,
+      "learning_rate": 9.304873592853253e-05,
+      "loss": 2.9286,
+      "step": 45441
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7085046172142029,
+      "learning_rate": 9.303758991044007e-05,
+      "loss": 2.8802,
+      "step": 45442
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7178999781608582,
+      "learning_rate": 9.302644443744689e-05,
+      "loss": 2.7642,
+      "step": 45443
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7461972236633301,
+      "learning_rate": 9.301529950958218e-05,
+      "loss": 2.9643,
+      "step": 45444
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7783269286155701,
+      "learning_rate": 9.300415512687538e-05,
+      "loss": 3.0405,
+      "step": 45445
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7068384885787964,
+      "learning_rate": 9.299301128935568e-05,
+      "loss": 2.7719,
+      "step": 45446
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7588903903961182,
+      "learning_rate": 9.298186799705251e-05,
+      "loss": 2.9524,
+      "step": 45447
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7695575952529907,
+      "learning_rate": 9.297072524999541e-05,
+      "loss": 2.719,
+      "step": 45448
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6897878646850586,
+      "learning_rate": 9.295958304821344e-05,
+      "loss": 2.7618,
+      "step": 45449
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7157047390937805,
+      "learning_rate": 9.294844139173622e-05,
+      "loss": 2.966,
+      "step": 45450
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7025874257087708,
+      "learning_rate": 9.293730028059295e-05,
+      "loss": 3.0808,
+      "step": 45451
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6616024374961853,
+      "learning_rate": 9.292615971481287e-05,
+      "loss": 3.149,
+      "step": 45452
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6998037099838257,
+      "learning_rate": 9.291501969442551e-05,
+      "loss": 2.9832,
+      "step": 45453
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7034969925880432,
+      "learning_rate": 9.290388021946005e-05,
+      "loss": 2.8652,
+      "step": 45454
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6883959174156189,
+      "learning_rate": 9.289274128994603e-05,
+      "loss": 2.9851,
+      "step": 45455
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6983670592308044,
+      "learning_rate": 9.288160290591255e-05,
+      "loss": 2.6393,
+      "step": 45456
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7119463086128235,
+      "learning_rate": 9.287046506738919e-05,
+      "loss": 2.8454,
+      "step": 45457
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6975045204162598,
+      "learning_rate": 9.285932777440517e-05,
+      "loss": 2.897,
+      "step": 45458
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7082361578941345,
+      "learning_rate": 9.284819102698967e-05,
+      "loss": 2.7897,
+      "step": 45459
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6975535154342651,
+      "learning_rate": 9.283705482517231e-05,
+      "loss": 2.9471,
+      "step": 45460
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6530376076698303,
+      "learning_rate": 9.282591916898216e-05,
+      "loss": 2.8358,
+      "step": 45461
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7379580140113831,
+      "learning_rate": 9.281478405844866e-05,
+      "loss": 2.7979,
+      "step": 45462
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7067075967788696,
+      "learning_rate": 9.280364949360125e-05,
+      "loss": 2.9221,
+      "step": 45463
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7226753234863281,
+      "learning_rate": 9.279251547446906e-05,
+      "loss": 2.8154,
+      "step": 45464
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.72322678565979,
+      "learning_rate": 9.278138200108161e-05,
+      "loss": 2.9727,
+      "step": 45465
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6826564073562622,
+      "learning_rate": 9.277024907346812e-05,
+      "loss": 2.9562,
+      "step": 45466
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7021982669830322,
+      "learning_rate": 9.27591166916578e-05,
+      "loss": 2.8354,
+      "step": 45467
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7087491154670715,
+      "learning_rate": 9.274798485568019e-05,
+      "loss": 2.7002,
+      "step": 45468
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7043088674545288,
+      "learning_rate": 9.273685356556437e-05,
+      "loss": 2.8718,
+      "step": 45469
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7019631266593933,
+      "learning_rate": 9.272572282133992e-05,
+      "loss": 2.7755,
+      "step": 45470
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6537641882896423,
+      "learning_rate": 9.271459262303594e-05,
+      "loss": 3.0271,
+      "step": 45471
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7400332689285278,
+      "learning_rate": 9.270346297068192e-05,
+      "loss": 2.8694,
+      "step": 45472
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7204238772392273,
+      "learning_rate": 9.269233386430705e-05,
+      "loss": 2.9346,
+      "step": 45473
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6749469637870789,
+      "learning_rate": 9.26812053039406e-05,
+      "loss": 2.7155,
+      "step": 45474
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6849311590194702,
+      "learning_rate": 9.267007728961205e-05,
+      "loss": 2.9678,
+      "step": 45475
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6702030301094055,
+      "learning_rate": 9.265894982135054e-05,
+      "loss": 2.946,
+      "step": 45476
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7221550345420837,
+      "learning_rate": 9.264782289918542e-05,
+      "loss": 3.0065,
+      "step": 45477
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6980705857276917,
+      "learning_rate": 9.263669652314615e-05,
+      "loss": 3.0259,
+      "step": 45478
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7406960725784302,
+      "learning_rate": 9.26255706932619e-05,
+      "loss": 2.8508,
+      "step": 45479
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7291260957717896,
+      "learning_rate": 9.261444540956201e-05,
+      "loss": 2.8585,
+      "step": 45480
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7162569165229797,
+      "learning_rate": 9.260332067207563e-05,
+      "loss": 3.0759,
+      "step": 45481
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6985812783241272,
+      "learning_rate": 9.25921964808322e-05,
+      "loss": 2.9614,
+      "step": 45482
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7387300133705139,
+      "learning_rate": 9.258107283586111e-05,
+      "loss": 2.8777,
+      "step": 45483
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7236722111701965,
+      "learning_rate": 9.256994973719148e-05,
+      "loss": 2.9205,
+      "step": 45484
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.8212042450904846,
+      "learning_rate": 9.255882718485272e-05,
+      "loss": 2.83,
+      "step": 45485
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7353643774986267,
+      "learning_rate": 9.254770517887403e-05,
+      "loss": 2.8415,
+      "step": 45486
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6906113624572754,
+      "learning_rate": 9.253658371928487e-05,
+      "loss": 2.7953,
+      "step": 45487
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7319288849830627,
+      "learning_rate": 9.25254628061144e-05,
+      "loss": 3.0537,
+      "step": 45488
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6931775212287903,
+      "learning_rate": 9.251434243939183e-05,
+      "loss": 2.9983,
+      "step": 45489
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6803140640258789,
+      "learning_rate": 9.250322261914666e-05,
+      "loss": 2.9426,
+      "step": 45490
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7066940665245056,
+      "learning_rate": 9.249210334540798e-05,
+      "loss": 2.713,
+      "step": 45491
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6657136082649231,
+      "learning_rate": 9.248098461820513e-05,
+      "loss": 2.6789,
+      "step": 45492
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7160326242446899,
+      "learning_rate": 9.246986643756755e-05,
+      "loss": 2.6241,
+      "step": 45493
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7386571764945984,
+      "learning_rate": 9.245874880352441e-05,
+      "loss": 3.0217,
+      "step": 45494
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7494421601295471,
+      "learning_rate": 9.244763171610494e-05,
+      "loss": 2.9368,
+      "step": 45495
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7182496786117554,
+      "learning_rate": 9.24365151753384e-05,
+      "loss": 3.0068,
+      "step": 45496
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7699613571166992,
+      "learning_rate": 9.242539918125416e-05,
+      "loss": 2.7448,
+      "step": 45497
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6990972757339478,
+      "learning_rate": 9.241428373388152e-05,
+      "loss": 2.9623,
+      "step": 45498
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7023851275444031,
+      "learning_rate": 9.240316883324964e-05,
+      "loss": 2.8577,
+      "step": 45499
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.710388720035553,
+      "learning_rate": 9.239205447938793e-05,
+      "loss": 2.885,
+      "step": 45500
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7227920889854431,
+      "learning_rate": 9.238094067232551e-05,
+      "loss": 2.8459,
+      "step": 45501
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7624646425247192,
+      "learning_rate": 9.236982741209191e-05,
+      "loss": 2.9227,
+      "step": 45502
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7066624164581299,
+      "learning_rate": 9.235871469871607e-05,
+      "loss": 3.1932,
+      "step": 45503
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7189279198646545,
+      "learning_rate": 9.23476025322274e-05,
+      "loss": 2.8258,
+      "step": 45504
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6847929954528809,
+      "learning_rate": 9.233649091265529e-05,
+      "loss": 2.6352,
+      "step": 45505
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7308363318443298,
+      "learning_rate": 9.232537984002878e-05,
+      "loss": 3.0242,
+      "step": 45506
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7426416873931885,
+      "learning_rate": 9.231426931437734e-05,
+      "loss": 2.9488,
+      "step": 45507
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6900115609169006,
+      "learning_rate": 9.230315933573005e-05,
+      "loss": 2.8308,
+      "step": 45508
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7146220207214355,
+      "learning_rate": 9.229204990411637e-05,
+      "loss": 2.9156,
+      "step": 45509
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.733957052230835,
+      "learning_rate": 9.228094101956546e-05,
+      "loss": 2.8853,
+      "step": 45510
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7220903635025024,
+      "learning_rate": 9.226983268210649e-05,
+      "loss": 2.9089,
+      "step": 45511
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7509836554527283,
+      "learning_rate": 9.225872489176888e-05,
+      "loss": 2.9585,
+      "step": 45512
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7305395603179932,
+      "learning_rate": 9.224761764858172e-05,
+      "loss": 3.0472,
+      "step": 45513
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7344761490821838,
+      "learning_rate": 9.223651095257435e-05,
+      "loss": 2.9322,
+      "step": 45514
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.693290650844574,
+      "learning_rate": 9.222540480377613e-05,
+      "loss": 2.8171,
+      "step": 45515
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6867750287055969,
+      "learning_rate": 9.22142992022162e-05,
+      "loss": 3.0071,
+      "step": 45516
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7052892446517944,
+      "learning_rate": 9.220319414792382e-05,
+      "loss": 2.9504,
+      "step": 45517
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7113419771194458,
+      "learning_rate": 9.219208964092812e-05,
+      "loss": 2.9487,
+      "step": 45518
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7192926406860352,
+      "learning_rate": 9.218098568125849e-05,
+      "loss": 2.9342,
+      "step": 45519
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7447081208229065,
+      "learning_rate": 9.216988226894423e-05,
+      "loss": 2.9044,
+      "step": 45520
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7151831388473511,
+      "learning_rate": 9.215877940401441e-05,
+      "loss": 3.0041,
+      "step": 45521
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7142324447631836,
+      "learning_rate": 9.214767708649845e-05,
+      "loss": 2.8919,
+      "step": 45522
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7118008732795715,
+      "learning_rate": 9.213657531642541e-05,
+      "loss": 2.7913,
+      "step": 45523
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7367550730705261,
+      "learning_rate": 9.212547409382473e-05,
+      "loss": 3.1136,
+      "step": 45524
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7201369404792786,
+      "learning_rate": 9.211437341872557e-05,
+      "loss": 2.8379,
+      "step": 45525
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7369980216026306,
+      "learning_rate": 9.2103273291157e-05,
+      "loss": 2.9005,
+      "step": 45526
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7533488869667053,
+      "learning_rate": 9.209217371114852e-05,
+      "loss": 3.1198,
+      "step": 45527
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6949545741081238,
+      "learning_rate": 9.208107467872914e-05,
+      "loss": 2.8966,
+      "step": 45528
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7416513562202454,
+      "learning_rate": 9.20699761939282e-05,
+      "loss": 2.7839,
+      "step": 45529
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6859050989151001,
+      "learning_rate": 9.205887825677504e-05,
+      "loss": 2.9379,
+      "step": 45530
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7521802186965942,
+      "learning_rate": 9.204778086729875e-05,
+      "loss": 3.0114,
+      "step": 45531
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6823815703392029,
+      "learning_rate": 9.203668402552857e-05,
+      "loss": 2.9262,
+      "step": 45532
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7390176057815552,
+      "learning_rate": 9.202558773149368e-05,
+      "loss": 2.6721,
+      "step": 45533
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7297099232673645,
+      "learning_rate": 9.201449198522334e-05,
+      "loss": 2.9628,
+      "step": 45534
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7210890054702759,
+      "learning_rate": 9.20033967867469e-05,
+      "loss": 2.8604,
+      "step": 45535
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7243456244468689,
+      "learning_rate": 9.199230213609342e-05,
+      "loss": 2.7529,
+      "step": 45536
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7083402872085571,
+      "learning_rate": 9.198120803329225e-05,
+      "loss": 3.0748,
+      "step": 45537
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7190839648246765,
+      "learning_rate": 9.197011447837253e-05,
+      "loss": 2.7782,
+      "step": 45538
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7521587014198303,
+      "learning_rate": 9.19590214713634e-05,
+      "loss": 3.3249,
+      "step": 45539
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7044076323509216,
+      "learning_rate": 9.194792901229429e-05,
+      "loss": 2.9068,
+      "step": 45540
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7235004901885986,
+      "learning_rate": 9.19368371011942e-05,
+      "loss": 3.2096,
+      "step": 45541
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.728836715221405,
+      "learning_rate": 9.19257457380925e-05,
+      "loss": 2.7914,
+      "step": 45542
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7259882688522339,
+      "learning_rate": 9.191465492301827e-05,
+      "loss": 2.8201,
+      "step": 45543
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6856926083564758,
+      "learning_rate": 9.190356465600081e-05,
+      "loss": 3.0713,
+      "step": 45544
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7389464974403381,
+      "learning_rate": 9.189247493706939e-05,
+      "loss": 3.0653,
+      "step": 45545
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7572897672653198,
+      "learning_rate": 9.188138576625314e-05,
+      "loss": 3.0166,
+      "step": 45546
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7614259123802185,
+      "learning_rate": 9.187029714358124e-05,
+      "loss": 2.9215,
+      "step": 45547
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.8179017305374146,
+      "learning_rate": 9.185920906908288e-05,
+      "loss": 2.6157,
+      "step": 45548
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7034313082695007,
+      "learning_rate": 9.184812154278726e-05,
+      "loss": 3.0541,
+      "step": 45549
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7329689264297485,
+      "learning_rate": 9.183703456472373e-05,
+      "loss": 3.0281,
+      "step": 45550
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7217639088630676,
+      "learning_rate": 9.18259481349213e-05,
+      "loss": 2.6217,
+      "step": 45551
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7059627771377563,
+      "learning_rate": 9.181486225340938e-05,
+      "loss": 2.8865,
+      "step": 45552
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.726000964641571,
+      "learning_rate": 9.180377692021702e-05,
+      "loss": 2.8841,
+      "step": 45553
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7023661136627197,
+      "learning_rate": 9.179269213537334e-05,
+      "loss": 3.087,
+      "step": 45554
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.736512303352356,
+      "learning_rate": 9.178160789890776e-05,
+      "loss": 2.9646,
+      "step": 45555
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7114843130111694,
+      "learning_rate": 9.177052421084923e-05,
+      "loss": 2.8579,
+      "step": 45556
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7629355192184448,
+      "learning_rate": 9.175944107122718e-05,
+      "loss": 2.6514,
+      "step": 45557
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7430128455162048,
+      "learning_rate": 9.17483584800706e-05,
+      "loss": 2.9601,
+      "step": 45558
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6909064650535583,
+      "learning_rate": 9.173727643740882e-05,
+      "loss": 2.8679,
+      "step": 45559
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6793957948684692,
+      "learning_rate": 9.172619494327101e-05,
+      "loss": 2.9662,
+      "step": 45560
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6956231594085693,
+      "learning_rate": 9.171511399768621e-05,
+      "loss": 2.8303,
+      "step": 45561
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7205344438552856,
+      "learning_rate": 9.170403360068381e-05,
+      "loss": 2.8042,
+      "step": 45562
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7068600058555603,
+      "learning_rate": 9.16929537522928e-05,
+      "loss": 2.7959,
+      "step": 45563
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7382459044456482,
+      "learning_rate": 9.168187445254247e-05,
+      "loss": 3.0018,
+      "step": 45564
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7958078980445862,
+      "learning_rate": 9.167079570146207e-05,
+      "loss": 2.7408,
+      "step": 45565
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7165841460227966,
+      "learning_rate": 9.165971749908063e-05,
+      "loss": 2.958,
+      "step": 45566
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7124565839767456,
+      "learning_rate": 9.164863984542746e-05,
+      "loss": 2.9686,
+      "step": 45567
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7308372855186462,
+      "learning_rate": 9.163756274053171e-05,
+      "loss": 2.7848,
+      "step": 45568
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6912724375724792,
+      "learning_rate": 9.162648618442239e-05,
+      "loss": 2.6554,
+      "step": 45569
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7064422965049744,
+      "learning_rate": 9.161541017712893e-05,
+      "loss": 2.9808,
+      "step": 45570
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7386999130249023,
+      "learning_rate": 9.160433471868026e-05,
+      "loss": 2.8824,
+      "step": 45571
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7071724534034729,
+      "learning_rate": 9.159325980910578e-05,
+      "loss": 3.0607,
+      "step": 45572
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.734996497631073,
+      "learning_rate": 9.158218544843442e-05,
+      "loss": 2.6605,
+      "step": 45573
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7159255743026733,
+      "learning_rate": 9.15711116366956e-05,
+      "loss": 3.0333,
+      "step": 45574
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6946538686752319,
+      "learning_rate": 9.156003837391834e-05,
+      "loss": 3.0226,
+      "step": 45575
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.678464949131012,
+      "learning_rate": 9.154896566013172e-05,
+      "loss": 2.7583,
+      "step": 45576
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7494169473648071,
+      "learning_rate": 9.153789349536511e-05,
+      "loss": 3.1025,
+      "step": 45577
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7178239822387695,
+      "learning_rate": 9.152682187964749e-05,
+      "loss": 3.0465,
+      "step": 45578
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7179162502288818,
+      "learning_rate": 9.151575081300807e-05,
+      "loss": 2.8249,
+      "step": 45579
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6873793005943298,
+      "learning_rate": 9.150468029547616e-05,
+      "loss": 3.0525,
+      "step": 45580
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7262808680534363,
+      "learning_rate": 9.14936103270808e-05,
+      "loss": 2.9266,
+      "step": 45581
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6978574991226196,
+      "learning_rate": 9.148254090785113e-05,
+      "loss": 2.8802,
+      "step": 45582
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7047426104545593,
+      "learning_rate": 9.147147203781622e-05,
+      "loss": 2.6446,
+      "step": 45583
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7362146377563477,
+      "learning_rate": 9.14604037170054e-05,
+      "loss": 2.7547,
+      "step": 45584
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6840211153030396,
+      "learning_rate": 9.14493359454477e-05,
+      "loss": 3.0048,
+      "step": 45585
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.727241575717926,
+      "learning_rate": 9.143826872317224e-05,
+      "loss": 2.8566,
+      "step": 45586
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6769251823425293,
+      "learning_rate": 9.14272020502084e-05,
+      "loss": 2.8878,
+      "step": 45587
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7173831462860107,
+      "learning_rate": 9.141613592658504e-05,
+      "loss": 2.9605,
+      "step": 45588
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7770647406578064,
+      "learning_rate": 9.140507035233154e-05,
+      "loss": 2.8595,
+      "step": 45589
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7519656419754028,
+      "learning_rate": 9.139400532747694e-05,
+      "loss": 2.6777,
+      "step": 45590
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7570061087608337,
+      "learning_rate": 9.138294085205031e-05,
+      "loss": 2.9145,
+      "step": 45591
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7038795351982117,
+      "learning_rate": 9.137187692608091e-05,
+      "loss": 2.9259,
+      "step": 45592
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7403584718704224,
+      "learning_rate": 9.136081354959777e-05,
+      "loss": 3.0187,
+      "step": 45593
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6944155693054199,
+      "learning_rate": 9.134975072263022e-05,
+      "loss": 2.928,
+      "step": 45594
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7204017043113708,
+      "learning_rate": 9.133868844520712e-05,
+      "loss": 2.8462,
+      "step": 45595
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7419741153717041,
+      "learning_rate": 9.132762671735789e-05,
+      "loss": 3.0493,
+      "step": 45596
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6750164031982422,
+      "learning_rate": 9.13165655391115e-05,
+      "loss": 2.957,
+      "step": 45597
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7366441488265991,
+      "learning_rate": 9.130550491049705e-05,
+      "loss": 2.8784,
+      "step": 45598
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.8148265480995178,
+      "learning_rate": 9.129444483154379e-05,
+      "loss": 3.0984,
+      "step": 45599
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7063391804695129,
+      "learning_rate": 9.128338530228074e-05,
+      "loss": 2.8162,
+      "step": 45600
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7022308707237244,
+      "learning_rate": 9.127232632273708e-05,
+      "loss": 2.8945,
+      "step": 45601
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6840649247169495,
+      "learning_rate": 9.126126789294203e-05,
+      "loss": 2.8795,
+      "step": 45602
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7260749340057373,
+      "learning_rate": 9.125021001292462e-05,
+      "loss": 2.9185,
+      "step": 45603
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7081303596496582,
+      "learning_rate": 9.1239152682714e-05,
+      "loss": 3.0542,
+      "step": 45604
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6989736557006836,
+      "learning_rate": 9.122809590233916e-05,
+      "loss": 3.1102,
+      "step": 45605
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6952615976333618,
+      "learning_rate": 9.121703967182934e-05,
+      "loss": 2.8599,
+      "step": 45606
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.714263916015625,
+      "learning_rate": 9.120598399121375e-05,
+      "loss": 2.8885,
+      "step": 45607
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7635734677314758,
+      "learning_rate": 9.119492886052133e-05,
+      "loss": 2.9962,
+      "step": 45608
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.690173864364624,
+      "learning_rate": 9.118387427978137e-05,
+      "loss": 2.9021,
+      "step": 45609
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6887124180793762,
+      "learning_rate": 9.117282024902282e-05,
+      "loss": 2.9197,
+      "step": 45610
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6885898113250732,
+      "learning_rate": 9.116176676827496e-05,
+      "loss": 2.9658,
+      "step": 45611
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6705561280250549,
+      "learning_rate": 9.115071383756681e-05,
+      "loss": 3.082,
+      "step": 45612
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6776615977287292,
+      "learning_rate": 9.113966145692737e-05,
+      "loss": 3.0067,
+      "step": 45613
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7082581520080566,
+      "learning_rate": 9.112860962638598e-05,
+      "loss": 3.1007,
+      "step": 45614
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7481296062469482,
+      "learning_rate": 9.111755834597155e-05,
+      "loss": 2.9351,
+      "step": 45615
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6857460737228394,
+      "learning_rate": 9.110650761571325e-05,
+      "loss": 2.7641,
+      "step": 45616
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7014007568359375,
+      "learning_rate": 9.109545743564032e-05,
+      "loss": 2.897,
+      "step": 45617
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7194677591323853,
+      "learning_rate": 9.108440780578172e-05,
+      "loss": 2.8158,
+      "step": 45618
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7104588150978088,
+      "learning_rate": 9.107335872616661e-05,
+      "loss": 2.8991,
+      "step": 45619
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7148900628089905,
+      "learning_rate": 9.106231019682396e-05,
+      "loss": 2.6667,
+      "step": 45620
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7613111734390259,
+      "learning_rate": 9.105126221778299e-05,
+      "loss": 2.9804,
+      "step": 45621
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7180856466293335,
+      "learning_rate": 9.104021478907289e-05,
+      "loss": 3.0473,
+      "step": 45622
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7277432680130005,
+      "learning_rate": 9.102916791072252e-05,
+      "loss": 3.0207,
+      "step": 45623
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6842195987701416,
+      "learning_rate": 9.101812158276121e-05,
+      "loss": 2.8484,
+      "step": 45624
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.719687819480896,
+      "learning_rate": 9.100707580521785e-05,
+      "loss": 3.1122,
+      "step": 45625
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7382856607437134,
+      "learning_rate": 9.099603057812173e-05,
+      "loss": 2.9134,
+      "step": 45626
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7554395198822021,
+      "learning_rate": 9.098498590150183e-05,
+      "loss": 2.8051,
+      "step": 45627
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7509669661521912,
+      "learning_rate": 9.097394177538717e-05,
+      "loss": 2.9297,
+      "step": 45628
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.0624964237213135,
+      "learning_rate": 9.096289819980701e-05,
+      "loss": 2.9979,
+      "step": 45629
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6812724471092224,
+      "learning_rate": 9.095185517479025e-05,
+      "loss": 2.8224,
+      "step": 45630
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7406055927276611,
+      "learning_rate": 9.094081270036608e-05,
+      "loss": 2.8994,
+      "step": 45631
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7246912717819214,
+      "learning_rate": 9.092977077656366e-05,
+      "loss": 2.7739,
+      "step": 45632
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7108069658279419,
+      "learning_rate": 9.091872940341199e-05,
+      "loss": 2.8253,
+      "step": 45633
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7075560688972473,
+      "learning_rate": 9.090768858094015e-05,
+      "loss": 3.2586,
+      "step": 45634
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7471787333488464,
+      "learning_rate": 9.089664830917713e-05,
+      "loss": 2.915,
+      "step": 45635
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6982598304748535,
+      "learning_rate": 9.088560858815208e-05,
+      "loss": 2.8928,
+      "step": 45636
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6782969236373901,
+      "learning_rate": 9.087456941789421e-05,
+      "loss": 2.8622,
+      "step": 45637
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7105681896209717,
+      "learning_rate": 9.086353079843238e-05,
+      "loss": 2.9933,
+      "step": 45638
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.719509482383728,
+      "learning_rate": 9.085249272979582e-05,
+      "loss": 2.9826,
+      "step": 45639
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7250756025314331,
+      "learning_rate": 9.084145521201357e-05,
+      "loss": 2.8474,
+      "step": 45640
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7311588525772095,
+      "learning_rate": 9.083041824511455e-05,
+      "loss": 2.9898,
+      "step": 45641
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7223825454711914,
+      "learning_rate": 9.081938182912809e-05,
+      "loss": 2.9584,
+      "step": 45642
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7354487180709839,
+      "learning_rate": 9.080834596408301e-05,
+      "loss": 2.9169,
+      "step": 45643
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7054993510246277,
+      "learning_rate": 9.079731065000858e-05,
+      "loss": 2.7169,
+      "step": 45644
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7050856351852417,
+      "learning_rate": 9.07862758869337e-05,
+      "loss": 2.8571,
+      "step": 45645
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7472595572471619,
+      "learning_rate": 9.077524167488746e-05,
+      "loss": 3.0456,
+      "step": 45646
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7568950653076172,
+      "learning_rate": 9.076420801389909e-05,
+      "loss": 2.7929,
+      "step": 45647
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6783512830734253,
+      "learning_rate": 9.075317490399752e-05,
+      "loss": 2.9669,
+      "step": 45648
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.706605851650238,
+      "learning_rate": 9.074214234521182e-05,
+      "loss": 2.7641,
+      "step": 45649
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6669334173202515,
+      "learning_rate": 9.073111033757094e-05,
+      "loss": 2.7432,
+      "step": 45650
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7224730253219604,
+      "learning_rate": 9.072007888110404e-05,
+      "loss": 2.7118,
+      "step": 45651
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7125197052955627,
+      "learning_rate": 9.070904797584029e-05,
+      "loss": 2.9566,
+      "step": 45652
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7195077538490295,
+      "learning_rate": 9.069801762180853e-05,
+      "loss": 2.6613,
+      "step": 45653
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7205254435539246,
+      "learning_rate": 9.0686987819038e-05,
+      "loss": 3.0001,
+      "step": 45654
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7503374218940735,
+      "learning_rate": 9.067595856755765e-05,
+      "loss": 2.7024,
+      "step": 45655
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6984966397285461,
+      "learning_rate": 9.066492986739647e-05,
+      "loss": 2.7526,
+      "step": 45656
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6726536154747009,
+      "learning_rate": 9.065390171858366e-05,
+      "loss": 2.9738,
+      "step": 45657
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6979069113731384,
+      "learning_rate": 9.064287412114808e-05,
+      "loss": 2.9208,
+      "step": 45658
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.695800244808197,
+      "learning_rate": 9.063184707511899e-05,
+      "loss": 2.9246,
+      "step": 45659
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7155253887176514,
+      "learning_rate": 9.06208205805252e-05,
+      "loss": 2.8638,
+      "step": 45660
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6801446676254272,
+      "learning_rate": 9.060979463739598e-05,
+      "loss": 2.9644,
+      "step": 45661
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6838302612304688,
+      "learning_rate": 9.059876924576029e-05,
+      "loss": 2.9361,
+      "step": 45662
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7186485528945923,
+      "learning_rate": 9.0587744405647e-05,
+      "loss": 2.8708,
+      "step": 45663
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.664668083190918,
+      "learning_rate": 9.057672011708541e-05,
+      "loss": 2.7418,
+      "step": 45664
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7575511336326599,
+      "learning_rate": 9.056569638010436e-05,
+      "loss": 2.9639,
+      "step": 45665
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7134124040603638,
+      "learning_rate": 9.055467319473303e-05,
+      "loss": 2.6629,
+      "step": 45666
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7106820344924927,
+      "learning_rate": 9.054365056100027e-05,
+      "loss": 2.8322,
+      "step": 45667
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6775752902030945,
+      "learning_rate": 9.053262847893526e-05,
+      "loss": 2.8653,
+      "step": 45668
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7096874713897705,
+      "learning_rate": 9.052160694856706e-05,
+      "loss": 3.0051,
+      "step": 45669
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7215758562088013,
+      "learning_rate": 9.051058596992463e-05,
+      "loss": 2.9697,
+      "step": 45670
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7285184860229492,
+      "learning_rate": 9.049956554303702e-05,
+      "loss": 2.8607,
+      "step": 45671
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7512555122375488,
+      "learning_rate": 9.048854566793314e-05,
+      "loss": 3.0024,
+      "step": 45672
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7152148485183716,
+      "learning_rate": 9.04775263446421e-05,
+      "loss": 3.0288,
+      "step": 45673
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7303699851036072,
+      "learning_rate": 9.046650757319304e-05,
+      "loss": 2.9171,
+      "step": 45674
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6994890570640564,
+      "learning_rate": 9.045548935361475e-05,
+      "loss": 2.7558,
+      "step": 45675
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7585562467575073,
+      "learning_rate": 9.044447168593649e-05,
+      "loss": 3.004,
+      "step": 45676
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7422113418579102,
+      "learning_rate": 9.043345457018718e-05,
+      "loss": 2.9295,
+      "step": 45677
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7175257802009583,
+      "learning_rate": 9.042243800639569e-05,
+      "loss": 2.9137,
+      "step": 45678
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7064406275749207,
+      "learning_rate": 9.041142199459127e-05,
+      "loss": 2.9965,
+      "step": 45679
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7073319554328918,
+      "learning_rate": 9.040040653480272e-05,
+      "loss": 2.8982,
+      "step": 45680
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7129002213478088,
+      "learning_rate": 9.038939162705926e-05,
+      "loss": 2.8788,
+      "step": 45681
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7113178968429565,
+      "learning_rate": 9.037837727138974e-05,
+      "loss": 3.0197,
+      "step": 45682
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7069679498672485,
+      "learning_rate": 9.036736346782331e-05,
+      "loss": 2.7209,
+      "step": 45683
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7247236371040344,
+      "learning_rate": 9.035635021638889e-05,
+      "loss": 2.9684,
+      "step": 45684
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.685911238193512,
+      "learning_rate": 9.034533751711539e-05,
+      "loss": 3.061,
+      "step": 45685
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7022624611854553,
+      "learning_rate": 9.033432537003204e-05,
+      "loss": 2.9441,
+      "step": 45686
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7100083827972412,
+      "learning_rate": 9.032331377516764e-05,
+      "loss": 2.7339,
+      "step": 45687
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7229495048522949,
+      "learning_rate": 9.031230273255124e-05,
+      "loss": 2.8123,
+      "step": 45688
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7110562920570374,
+      "learning_rate": 9.030129224221201e-05,
+      "loss": 2.9297,
+      "step": 45689
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6981791257858276,
+      "learning_rate": 9.029028230417871e-05,
+      "loss": 3.0289,
+      "step": 45690
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7141596674919128,
+      "learning_rate": 9.027927291848055e-05,
+      "loss": 2.9254,
+      "step": 45691
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7503494620323181,
+      "learning_rate": 9.026826408514641e-05,
+      "loss": 2.9531,
+      "step": 45692
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.729244589805603,
+      "learning_rate": 9.025725580420521e-05,
+      "loss": 2.9759,
+      "step": 45693
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7077925801277161,
+      "learning_rate": 9.024624807568615e-05,
+      "loss": 2.9944,
+      "step": 45694
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7241457104682922,
+      "learning_rate": 9.023524089961797e-05,
+      "loss": 2.8535,
+      "step": 45695
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7095908522605896,
+      "learning_rate": 9.022423427602991e-05,
+      "loss": 2.9944,
+      "step": 45696
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.717788815498352,
+      "learning_rate": 9.021322820495074e-05,
+      "loss": 2.9462,
+      "step": 45697
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6879000663757324,
+      "learning_rate": 9.020222268640968e-05,
+      "loss": 2.8645,
+      "step": 45698
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7336751222610474,
+      "learning_rate": 9.019121772043554e-05,
+      "loss": 2.8715,
+      "step": 45699
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7165769338607788,
+      "learning_rate": 9.01802133070573e-05,
+      "loss": 3.0437,
+      "step": 45700
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7212401032447815,
+      "learning_rate": 9.016920944630407e-05,
+      "loss": 2.9006,
+      "step": 45701
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7075793147087097,
+      "learning_rate": 9.015820613820468e-05,
+      "loss": 2.9832,
+      "step": 45702
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6998143196105957,
+      "learning_rate": 9.014720338278818e-05,
+      "loss": 2.8956,
+      "step": 45703
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6937516927719116,
+      "learning_rate": 9.013620118008365e-05,
+      "loss": 3.1962,
+      "step": 45704
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6879575252532959,
+      "learning_rate": 9.012519953011999e-05,
+      "loss": 2.9001,
+      "step": 45705
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7310060858726501,
+      "learning_rate": 9.011419843292615e-05,
+      "loss": 2.6941,
+      "step": 45706
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6938087344169617,
+      "learning_rate": 9.010319788853102e-05,
+      "loss": 2.8176,
+      "step": 45707
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7372106313705444,
+      "learning_rate": 9.009219789696369e-05,
+      "loss": 2.8348,
+      "step": 45708
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7054558992385864,
+      "learning_rate": 9.00811984582532e-05,
+      "loss": 2.8556,
+      "step": 45709
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.709180474281311,
+      "learning_rate": 9.007019957242831e-05,
+      "loss": 2.8494,
+      "step": 45710
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.714429497718811,
+      "learning_rate": 9.005920123951824e-05,
+      "loss": 3.0477,
+      "step": 45711
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7273450493812561,
+      "learning_rate": 9.004820345955173e-05,
+      "loss": 2.9587,
+      "step": 45712
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7173202633857727,
+      "learning_rate": 9.003720623255795e-05,
+      "loss": 3.0399,
+      "step": 45713
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7252642512321472,
+      "learning_rate": 9.002620955856574e-05,
+      "loss": 3.1725,
+      "step": 45714
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.707578182220459,
+      "learning_rate": 9.0015213437604e-05,
+      "loss": 2.9012,
+      "step": 45715
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7207184433937073,
+      "learning_rate": 9.000421786970188e-05,
+      "loss": 2.5106,
+      "step": 45716
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7439934611320496,
+      "learning_rate": 8.999322285488814e-05,
+      "loss": 2.9038,
+      "step": 45717
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6846615076065063,
+      "learning_rate": 8.998222839319182e-05,
+      "loss": 2.8526,
+      "step": 45718
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6808404326438904,
+      "learning_rate": 8.997123448464199e-05,
+      "loss": 3.0216,
+      "step": 45719
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6881184577941895,
+      "learning_rate": 8.99602411292675e-05,
+      "loss": 2.9798,
+      "step": 45720
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7790664434432983,
+      "learning_rate": 8.994924832709732e-05,
+      "loss": 3.0434,
+      "step": 45721
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7797252535820007,
+      "learning_rate": 8.993825607816031e-05,
+      "loss": 2.7815,
+      "step": 45722
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6707017421722412,
+      "learning_rate": 8.99272643824855e-05,
+      "loss": 3.0048,
+      "step": 45723
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7407448887825012,
+      "learning_rate": 8.991627324010193e-05,
+      "loss": 2.8102,
+      "step": 45724
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7850496768951416,
+      "learning_rate": 8.990528265103839e-05,
+      "loss": 3.0061,
+      "step": 45725
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7100532054901123,
+      "learning_rate": 8.989429261532398e-05,
+      "loss": 2.9508,
+      "step": 45726
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7451736927032471,
+      "learning_rate": 8.988330313298748e-05,
+      "loss": 2.9667,
+      "step": 45727
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7049230337142944,
+      "learning_rate": 8.9872314204058e-05,
+      "loss": 2.787,
+      "step": 45728
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6660589575767517,
+      "learning_rate": 8.986132582856442e-05,
+      "loss": 2.7769,
+      "step": 45729
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7338511347770691,
+      "learning_rate": 8.985033800653558e-05,
+      "loss": 2.8149,
+      "step": 45730
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6853067278862,
+      "learning_rate": 8.983935073800059e-05,
+      "loss": 2.7641,
+      "step": 45731
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7548957467079163,
+      "learning_rate": 8.98283640229882e-05,
+      "loss": 2.7663,
+      "step": 45732
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7389004826545715,
+      "learning_rate": 8.981737786152748e-05,
+      "loss": 3.224,
+      "step": 45733
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6959095001220703,
+      "learning_rate": 8.98063922536474e-05,
+      "loss": 2.9059,
+      "step": 45734
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7141698002815247,
+      "learning_rate": 8.979540719937684e-05,
+      "loss": 3.022,
+      "step": 45735
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7193446159362793,
+      "learning_rate": 8.978442269874473e-05,
+      "loss": 2.8698,
+      "step": 45736
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6984773874282837,
+      "learning_rate": 8.977343875177988e-05,
+      "loss": 2.8872,
+      "step": 45737
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7044937610626221,
+      "learning_rate": 8.976245535851145e-05,
+      "loss": 2.9768,
+      "step": 45738
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.732440173625946,
+      "learning_rate": 8.975147251896816e-05,
+      "loss": 2.8359,
+      "step": 45739
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7215022444725037,
+      "learning_rate": 8.974049023317898e-05,
+      "loss": 2.7151,
+      "step": 45740
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.741974949836731,
+      "learning_rate": 8.972950850117303e-05,
+      "loss": 2.7282,
+      "step": 45741
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.702601432800293,
+      "learning_rate": 8.971852732297908e-05,
+      "loss": 2.9357,
+      "step": 45742
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.690036952495575,
+      "learning_rate": 8.970754669862605e-05,
+      "loss": 2.7958,
+      "step": 45743
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7078272700309753,
+      "learning_rate": 8.969656662814278e-05,
+      "loss": 2.6961,
+      "step": 45744
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7023754119873047,
+      "learning_rate": 8.968558711155828e-05,
+      "loss": 2.9431,
+      "step": 45745
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7015004754066467,
+      "learning_rate": 8.967460814890154e-05,
+      "loss": 2.9241,
+      "step": 45746
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7229611277580261,
+      "learning_rate": 8.966362974020135e-05,
+      "loss": 2.9218,
+      "step": 45747
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7046184539794922,
+      "learning_rate": 8.965265188548675e-05,
+      "loss": 2.7128,
+      "step": 45748
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6686332821846008,
+      "learning_rate": 8.964167458478649e-05,
+      "loss": 2.9473,
+      "step": 45749
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7126154899597168,
+      "learning_rate": 8.963069783812966e-05,
+      "loss": 3.0076,
+      "step": 45750
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.720814049243927,
+      "learning_rate": 8.961972164554509e-05,
+      "loss": 2.8959,
+      "step": 45751
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.767537534236908,
+      "learning_rate": 8.960874600706163e-05,
+      "loss": 2.9654,
+      "step": 45752
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7737765312194824,
+      "learning_rate": 8.95977709227083e-05,
+      "loss": 2.9991,
+      "step": 45753
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7451054453849792,
+      "learning_rate": 8.958679639251383e-05,
+      "loss": 2.76,
+      "step": 45754
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.682572066783905,
+      "learning_rate": 8.957582241650729e-05,
+      "loss": 3.0767,
+      "step": 45755
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7428975701332092,
+      "learning_rate": 8.956484899471761e-05,
+      "loss": 2.8986,
+      "step": 45756
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7675819993019104,
+      "learning_rate": 8.955387612717362e-05,
+      "loss": 2.911,
+      "step": 45757
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7058717012405396,
+      "learning_rate": 8.954290381390422e-05,
+      "loss": 2.8263,
+      "step": 45758
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7053318023681641,
+      "learning_rate": 8.953193205493819e-05,
+      "loss": 2.8086,
+      "step": 45759
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7443513870239258,
+      "learning_rate": 8.952096085030458e-05,
+      "loss": 2.9749,
+      "step": 45760
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6907855868339539,
+      "learning_rate": 8.950999020003236e-05,
+      "loss": 2.86,
+      "step": 45761
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6709179282188416,
+      "learning_rate": 8.949902010415016e-05,
+      "loss": 2.8877,
+      "step": 45762
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6978246569633484,
+      "learning_rate": 8.948805056268718e-05,
+      "loss": 2.9097,
+      "step": 45763
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7394630908966064,
+      "learning_rate": 8.947708157567213e-05,
+      "loss": 2.9971,
+      "step": 45764
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7391385436058044,
+      "learning_rate": 8.946611314313386e-05,
+      "loss": 2.8428,
+      "step": 45765
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7328847646713257,
+      "learning_rate": 8.945514526510141e-05,
+      "loss": 2.82,
+      "step": 45766
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7323940992355347,
+      "learning_rate": 8.944417794160352e-05,
+      "loss": 3.0405,
+      "step": 45767
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7202035784721375,
+      "learning_rate": 8.94332111726692e-05,
+      "loss": 2.8026,
+      "step": 45768
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6832515597343445,
+      "learning_rate": 8.942224495832721e-05,
+      "loss": 2.8618,
+      "step": 45769
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6816558241844177,
+      "learning_rate": 8.941127929860645e-05,
+      "loss": 2.9306,
+      "step": 45770
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7403445839881897,
+      "learning_rate": 8.9400314193536e-05,
+      "loss": 3.0726,
+      "step": 45771
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6996352672576904,
+      "learning_rate": 8.938934964314456e-05,
+      "loss": 2.7873,
+      "step": 45772
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7047386169433594,
+      "learning_rate": 8.937838564746107e-05,
+      "loss": 2.8053,
+      "step": 45773
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7159958481788635,
+      "learning_rate": 8.936742220651427e-05,
+      "loss": 2.9682,
+      "step": 45774
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6776139736175537,
+      "learning_rate": 8.935645932033311e-05,
+      "loss": 2.88,
+      "step": 45775
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6902870535850525,
+      "learning_rate": 8.934549698894664e-05,
+      "loss": 2.8853,
+      "step": 45776
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7082247138023376,
+      "learning_rate": 8.933453521238348e-05,
+      "loss": 2.9993,
+      "step": 45777
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.694507360458374,
+      "learning_rate": 8.93235739906727e-05,
+      "loss": 2.9914,
+      "step": 45778
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6993120908737183,
+      "learning_rate": 8.931261332384308e-05,
+      "loss": 2.9088,
+      "step": 45779
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.701582670211792,
+      "learning_rate": 8.930165321192341e-05,
+      "loss": 2.9265,
+      "step": 45780
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7343215942382812,
+      "learning_rate": 8.92906936549427e-05,
+      "loss": 2.856,
+      "step": 45781
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7673470973968506,
+      "learning_rate": 8.927973465292967e-05,
+      "loss": 3.0379,
+      "step": 45782
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7404628396034241,
+      "learning_rate": 8.926877620591339e-05,
+      "loss": 3.0584,
+      "step": 45783
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7033378481864929,
+      "learning_rate": 8.925781831392246e-05,
+      "loss": 2.9548,
+      "step": 45784
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.70482337474823,
+      "learning_rate": 8.9246860976986e-05,
+      "loss": 2.8364,
+      "step": 45785
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6887717247009277,
+      "learning_rate": 8.923590419513273e-05,
+      "loss": 3.0289,
+      "step": 45786
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7255544066429138,
+      "learning_rate": 8.922494796839143e-05,
+      "loss": 2.9807,
+      "step": 45787
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6836867928504944,
+      "learning_rate": 8.921399229679118e-05,
+      "loss": 2.8207,
+      "step": 45788
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.707114040851593,
+      "learning_rate": 8.920303718036058e-05,
+      "loss": 2.8818,
+      "step": 45789
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7001112699508667,
+      "learning_rate": 8.919208261912862e-05,
+      "loss": 3.0064,
+      "step": 45790
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7056055068969727,
+      "learning_rate": 8.918112861312422e-05,
+      "loss": 2.915,
+      "step": 45791
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.71812504529953,
+      "learning_rate": 8.91701751623761e-05,
+      "loss": 2.6304,
+      "step": 45792
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7017490267753601,
+      "learning_rate": 8.915922226691323e-05,
+      "loss": 3.0009,
+      "step": 45793
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6928110718727112,
+      "learning_rate": 8.914826992676439e-05,
+      "loss": 2.8037,
+      "step": 45794
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7178500890731812,
+      "learning_rate": 8.913731814195836e-05,
+      "loss": 3.0088,
+      "step": 45795
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7326123118400574,
+      "learning_rate": 8.91263669125241e-05,
+      "loss": 2.7809,
+      "step": 45796
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7544444799423218,
+      "learning_rate": 8.911541623849035e-05,
+      "loss": 2.9505,
+      "step": 45797
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6959039568901062,
+      "learning_rate": 8.910446611988609e-05,
+      "loss": 2.9881,
+      "step": 45798
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7174462676048279,
+      "learning_rate": 8.909351655674e-05,
+      "loss": 2.8899,
+      "step": 45799
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7137072682380676,
+      "learning_rate": 8.908256754908108e-05,
+      "loss": 2.8453,
+      "step": 45800
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7283577919006348,
+      "learning_rate": 8.907161909693806e-05,
+      "loss": 2.7934,
+      "step": 45801
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.732901930809021,
+      "learning_rate": 8.906067120033975e-05,
+      "loss": 2.6958,
+      "step": 45802
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7311907410621643,
+      "learning_rate": 8.904972385931509e-05,
+      "loss": 2.8987,
+      "step": 45803
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6583136320114136,
+      "learning_rate": 8.903877707389279e-05,
+      "loss": 2.809,
+      "step": 45804
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6969159245491028,
+      "learning_rate": 8.902783084410174e-05,
+      "loss": 2.8369,
+      "step": 45805
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7024007439613342,
+      "learning_rate": 8.901688516997091e-05,
+      "loss": 2.8592,
+      "step": 45806
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.8098918795585632,
+      "learning_rate": 8.900594005152896e-05,
+      "loss": 2.9321,
+      "step": 45807
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7050246000289917,
+      "learning_rate": 8.899499548880476e-05,
+      "loss": 2.8934,
+      "step": 45808
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7347972393035889,
+      "learning_rate": 8.898405148182708e-05,
+      "loss": 2.9272,
+      "step": 45809
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7350515127182007,
+      "learning_rate": 8.897310803062475e-05,
+      "loss": 2.971,
+      "step": 45810
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7553649544715881,
+      "learning_rate": 8.896216513522676e-05,
+      "loss": 3.0858,
+      "step": 45811
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.736858069896698,
+      "learning_rate": 8.895122279566168e-05,
+      "loss": 2.8214,
+      "step": 45812
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7918192744255066,
+      "learning_rate": 8.894028101195859e-05,
+      "loss": 2.7361,
+      "step": 45813
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.70150226354599,
+      "learning_rate": 8.89293397841461e-05,
+      "loss": 2.9136,
+      "step": 45814
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7298532724380493,
+      "learning_rate": 8.89183991122532e-05,
+      "loss": 2.9162,
+      "step": 45815
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6884034276008606,
+      "learning_rate": 8.890745899630857e-05,
+      "loss": 2.7884,
+      "step": 45816
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7185069918632507,
+      "learning_rate": 8.889651943634101e-05,
+      "loss": 2.9132,
+      "step": 45817
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7185012698173523,
+      "learning_rate": 8.888558043237946e-05,
+      "loss": 2.8284,
+      "step": 45818
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6928901672363281,
+      "learning_rate": 8.88746419844526e-05,
+      "loss": 2.8727,
+      "step": 45819
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7991043329238892,
+      "learning_rate": 8.886370409258935e-05,
+      "loss": 3.0365,
+      "step": 45820
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7154514193534851,
+      "learning_rate": 8.885276675681842e-05,
+      "loss": 2.7826,
+      "step": 45821
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6965875625610352,
+      "learning_rate": 8.884182997716874e-05,
+      "loss": 3.0469,
+      "step": 45822
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7708733677864075,
+      "learning_rate": 8.883089375366904e-05,
+      "loss": 2.8216,
+      "step": 45823
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7230860590934753,
+      "learning_rate": 8.881995808634801e-05,
+      "loss": 3.0006,
+      "step": 45824
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6626743674278259,
+      "learning_rate": 8.880902297523471e-05,
+      "loss": 2.7648,
+      "step": 45825
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7058641910552979,
+      "learning_rate": 8.879808842035769e-05,
+      "loss": 2.9872,
+      "step": 45826
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7745085954666138,
+      "learning_rate": 8.878715442174585e-05,
+      "loss": 2.9538,
+      "step": 45827
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6953856945037842,
+      "learning_rate": 8.877622097942807e-05,
+      "loss": 2.8034,
+      "step": 45828
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.8223050236701965,
+      "learning_rate": 8.876528809343302e-05,
+      "loss": 2.8538,
+      "step": 45829
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7568581700325012,
+      "learning_rate": 8.87543557637897e-05,
+      "loss": 2.9885,
+      "step": 45830
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7025486826896667,
+      "learning_rate": 8.874342399052658e-05,
+      "loss": 2.6614,
+      "step": 45831
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7281057834625244,
+      "learning_rate": 8.873249277367262e-05,
+      "loss": 2.9158,
+      "step": 45832
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7436516284942627,
+      "learning_rate": 8.872156211325673e-05,
+      "loss": 2.8681,
+      "step": 45833
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6978209018707275,
+      "learning_rate": 8.871063200930747e-05,
+      "loss": 2.8703,
+      "step": 45834
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7288390398025513,
+      "learning_rate": 8.869970246185382e-05,
+      "loss": 2.9508,
+      "step": 45835
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6683169007301331,
+      "learning_rate": 8.868877347092445e-05,
+      "loss": 3.0472,
+      "step": 45836
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7435875535011292,
+      "learning_rate": 8.867784503654823e-05,
+      "loss": 3.1275,
+      "step": 45837
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.751097559928894,
+      "learning_rate": 8.86669171587539e-05,
+      "loss": 2.6619,
+      "step": 45838
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7168517708778381,
+      "learning_rate": 8.865598983757017e-05,
+      "loss": 2.9837,
+      "step": 45839
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7347090840339661,
+      "learning_rate": 8.864506307302596e-05,
+      "loss": 2.6833,
+      "step": 45840
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6629171967506409,
+      "learning_rate": 8.863413686514986e-05,
+      "loss": 2.9058,
+      "step": 45841
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7456681132316589,
+      "learning_rate": 8.862321121397079e-05,
+      "loss": 2.7433,
+      "step": 45842
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6898133158683777,
+      "learning_rate": 8.861228611951762e-05,
+      "loss": 2.9031,
+      "step": 45843
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7045638561248779,
+      "learning_rate": 8.860136158181898e-05,
+      "loss": 2.8671,
+      "step": 45844
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7033824324607849,
+      "learning_rate": 8.859043760090368e-05,
+      "loss": 3.0791,
+      "step": 45845
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7060011029243469,
+      "learning_rate": 8.857951417680036e-05,
+      "loss": 2.7681,
+      "step": 45846
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6827572584152222,
+      "learning_rate": 8.856859130953795e-05,
+      "loss": 3.0132,
+      "step": 45847
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.725695788860321,
+      "learning_rate": 8.855766899914527e-05,
+      "loss": 3.0672,
+      "step": 45848
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7227092981338501,
+      "learning_rate": 8.854674724565086e-05,
+      "loss": 3.0037,
+      "step": 45849
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6998618245124817,
+      "learning_rate": 8.853582604908374e-05,
+      "loss": 2.7627,
+      "step": 45850
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7095021605491638,
+      "learning_rate": 8.852490540947249e-05,
+      "loss": 2.8435,
+      "step": 45851
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7027034163475037,
+      "learning_rate": 8.8513985326846e-05,
+      "loss": 2.8237,
+      "step": 45852
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7380695939064026,
+      "learning_rate": 8.850306580123298e-05,
+      "loss": 2.7772,
+      "step": 45853
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7337902784347534,
+      "learning_rate": 8.849214683266208e-05,
+      "loss": 3.0559,
+      "step": 45854
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7061793804168701,
+      "learning_rate": 8.848122842116227e-05,
+      "loss": 2.8721,
+      "step": 45855
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7373761534690857,
+      "learning_rate": 8.847031056676211e-05,
+      "loss": 2.9133,
+      "step": 45856
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7317122220993042,
+      "learning_rate": 8.84593932694904e-05,
+      "loss": 2.997,
+      "step": 45857
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7210045456886292,
+      "learning_rate": 8.844847652937609e-05,
+      "loss": 2.9035,
+      "step": 45858
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7218835949897766,
+      "learning_rate": 8.843756034644773e-05,
+      "loss": 2.761,
+      "step": 45859
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6628503799438477,
+      "learning_rate": 8.842664472073415e-05,
+      "loss": 2.9027,
+      "step": 45860
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.694086492061615,
+      "learning_rate": 8.841572965226396e-05,
+      "loss": 2.9911,
+      "step": 45861
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7303391098976135,
+      "learning_rate": 8.840481514106602e-05,
+      "loss": 2.8838,
+      "step": 45862
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7094843983650208,
+      "learning_rate": 8.839390118716915e-05,
+      "loss": 3.0449,
+      "step": 45863
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6973925828933716,
+      "learning_rate": 8.838298779060198e-05,
+      "loss": 2.9713,
+      "step": 45864
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7205259799957275,
+      "learning_rate": 8.837207495139336e-05,
+      "loss": 2.8643,
+      "step": 45865
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6511806845664978,
+      "learning_rate": 8.836116266957193e-05,
+      "loss": 2.7828,
+      "step": 45866
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7268171906471252,
+      "learning_rate": 8.83502509451664e-05,
+      "loss": 2.982,
+      "step": 45867
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.751723051071167,
+      "learning_rate": 8.833933977820569e-05,
+      "loss": 3.0784,
+      "step": 45868
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7350054979324341,
+      "learning_rate": 8.83284291687183e-05,
+      "loss": 2.9919,
+      "step": 45869
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7295348644256592,
+      "learning_rate": 8.83175191167332e-05,
+      "loss": 2.9076,
+      "step": 45870
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7354061603546143,
+      "learning_rate": 8.830660962227892e-05,
+      "loss": 2.8714,
+      "step": 45871
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7122185230255127,
+      "learning_rate": 8.82957006853843e-05,
+      "loss": 2.8408,
+      "step": 45872
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7535485029220581,
+      "learning_rate": 8.828479230607814e-05,
+      "loss": 3.0789,
+      "step": 45873
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7035789489746094,
+      "learning_rate": 8.827388448438908e-05,
+      "loss": 2.9393,
+      "step": 45874
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7172368764877319,
+      "learning_rate": 8.826297722034589e-05,
+      "loss": 2.8124,
+      "step": 45875
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7242939472198486,
+      "learning_rate": 8.825207051397717e-05,
+      "loss": 2.8569,
+      "step": 45876
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7067150473594666,
+      "learning_rate": 8.824116436531174e-05,
+      "loss": 2.7216,
+      "step": 45877
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6949864029884338,
+      "learning_rate": 8.823025877437839e-05,
+      "loss": 3.1062,
+      "step": 45878
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6859537363052368,
+      "learning_rate": 8.821935374120572e-05,
+      "loss": 2.8185,
+      "step": 45879
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7593329548835754,
+      "learning_rate": 8.82084492658226e-05,
+      "loss": 2.8668,
+      "step": 45880
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6818985342979431,
+      "learning_rate": 8.819754534825766e-05,
+      "loss": 2.8889,
+      "step": 45881
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7531813383102417,
+      "learning_rate": 8.818664198853954e-05,
+      "loss": 2.6414,
+      "step": 45882
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6858770847320557,
+      "learning_rate": 8.817573918669716e-05,
+      "loss": 2.6542,
+      "step": 45883
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7132046222686768,
+      "learning_rate": 8.8164836942759e-05,
+      "loss": 2.9236,
+      "step": 45884
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7180655598640442,
+      "learning_rate": 8.815393525675397e-05,
+      "loss": 2.9675,
+      "step": 45885
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6883565187454224,
+      "learning_rate": 8.814303412871063e-05,
+      "loss": 2.8287,
+      "step": 45886
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7425944209098816,
+      "learning_rate": 8.813213355865783e-05,
+      "loss": 2.9131,
+      "step": 45887
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.736083447933197,
+      "learning_rate": 8.812123354662427e-05,
+      "loss": 2.8409,
+      "step": 45888
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6757725477218628,
+      "learning_rate": 8.811033409263844e-05,
+      "loss": 2.7647,
+      "step": 45889
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7013656497001648,
+      "learning_rate": 8.809943519672936e-05,
+      "loss": 2.9786,
+      "step": 45890
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7240048050880432,
+      "learning_rate": 8.808853685892545e-05,
+      "loss": 3.0001,
+      "step": 45891
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7245268821716309,
+      "learning_rate": 8.807763907925568e-05,
+      "loss": 2.9762,
+      "step": 45892
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7503712177276611,
+      "learning_rate": 8.806674185774852e-05,
+      "loss": 2.8701,
+      "step": 45893
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7144389152526855,
+      "learning_rate": 8.805584519443276e-05,
+      "loss": 2.8848,
+      "step": 45894
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6713013052940369,
+      "learning_rate": 8.804494908933723e-05,
+      "loss": 3.0602,
+      "step": 45895
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7255938053131104,
+      "learning_rate": 8.803405354249049e-05,
+      "loss": 2.8985,
+      "step": 45896
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7103670835494995,
+      "learning_rate": 8.802315855392127e-05,
+      "loss": 2.865,
+      "step": 45897
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7415536046028137,
+      "learning_rate": 8.801226412365818e-05,
+      "loss": 2.9041,
+      "step": 45898
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7270449995994568,
+      "learning_rate": 8.800137025172998e-05,
+      "loss": 2.9277,
+      "step": 45899
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7240244150161743,
+      "learning_rate": 8.799047693816545e-05,
+      "loss": 2.9637,
+      "step": 45900
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6903491616249084,
+      "learning_rate": 8.79795841829931e-05,
+      "loss": 2.8616,
+      "step": 45901
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6921778917312622,
+      "learning_rate": 8.796869198624182e-05,
+      "loss": 3.0011,
+      "step": 45902
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7207080125808716,
+      "learning_rate": 8.795780034794024e-05,
+      "loss": 2.7592,
+      "step": 45903
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7446100115776062,
+      "learning_rate": 8.794690926811687e-05,
+      "loss": 3.0125,
+      "step": 45904
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7436581254005432,
+      "learning_rate": 8.793601874680061e-05,
+      "loss": 2.8783,
+      "step": 45905
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.69056236743927,
+      "learning_rate": 8.792512878402001e-05,
+      "loss": 2.9237,
+      "step": 45906
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.743360698223114,
+      "learning_rate": 8.791423937980387e-05,
+      "loss": 2.9432,
+      "step": 45907
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7039749026298523,
+      "learning_rate": 8.790335053418074e-05,
+      "loss": 3.0534,
+      "step": 45908
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6992817521095276,
+      "learning_rate": 8.789246224717945e-05,
+      "loss": 3.0331,
+      "step": 45909
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6970164179801941,
+      "learning_rate": 8.78815745188286e-05,
+      "loss": 2.9252,
+      "step": 45910
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7451713681221008,
+      "learning_rate": 8.787068734915673e-05,
+      "loss": 3.0809,
+      "step": 45911
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7480039000511169,
+      "learning_rate": 8.78598007381928e-05,
+      "loss": 3.0333,
+      "step": 45912
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7085638046264648,
+      "learning_rate": 8.784891468596517e-05,
+      "loss": 2.881,
+      "step": 45913
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.728363573551178,
+      "learning_rate": 8.783802919250271e-05,
+      "loss": 2.8696,
+      "step": 45914
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.708055317401886,
+      "learning_rate": 8.782714425783414e-05,
+      "loss": 3.2101,
+      "step": 45915
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6821580529212952,
+      "learning_rate": 8.781625988198792e-05,
+      "loss": 3.1762,
+      "step": 45916
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7033885717391968,
+      "learning_rate": 8.780537606499297e-05,
+      "loss": 2.8277,
+      "step": 45917
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7111839652061462,
+      "learning_rate": 8.779449280687778e-05,
+      "loss": 2.9337,
+      "step": 45918
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7263917326927185,
+      "learning_rate": 8.778361010767099e-05,
+      "loss": 2.8357,
+      "step": 45919
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7069695591926575,
+      "learning_rate": 8.777272796740139e-05,
+      "loss": 2.7324,
+      "step": 45920
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7394420504570007,
+      "learning_rate": 8.77618463860975e-05,
+      "loss": 2.7066,
+      "step": 45921
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7445680499076843,
+      "learning_rate": 8.77509653637882e-05,
+      "loss": 2.8956,
+      "step": 45922
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7384760975837708,
+      "learning_rate": 8.774008490050185e-05,
+      "loss": 2.944,
+      "step": 45923
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7003813982009888,
+      "learning_rate": 8.772920499626741e-05,
+      "loss": 2.7802,
+      "step": 45924
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7218019366264343,
+      "learning_rate": 8.771832565111335e-05,
+      "loss": 2.6604,
+      "step": 45925
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7279051542282104,
+      "learning_rate": 8.770744686506828e-05,
+      "loss": 2.8343,
+      "step": 45926
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6893950700759888,
+      "learning_rate": 8.769656863816106e-05,
+      "loss": 2.986,
+      "step": 45927
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7431288361549377,
+      "learning_rate": 8.768569097042011e-05,
+      "loss": 2.7472,
+      "step": 45928
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7378048896789551,
+      "learning_rate": 8.767481386187418e-05,
+      "loss": 2.8564,
+      "step": 45929
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7334486842155457,
+      "learning_rate": 8.766393731255205e-05,
+      "loss": 2.9231,
+      "step": 45930
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7250813841819763,
+      "learning_rate": 8.76530613224822e-05,
+      "loss": 2.9032,
+      "step": 45931
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7708786725997925,
+      "learning_rate": 8.764218589169338e-05,
+      "loss": 2.8288,
+      "step": 45932
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7491968274116516,
+      "learning_rate": 8.763131102021402e-05,
+      "loss": 2.9354,
+      "step": 45933
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7650907039642334,
+      "learning_rate": 8.762043670807293e-05,
+      "loss": 2.8662,
+      "step": 45934
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7074927091598511,
+      "learning_rate": 8.760956295529887e-05,
+      "loss": 3.0082,
+      "step": 45935
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6994934678077698,
+      "learning_rate": 8.75986897619202e-05,
+      "loss": 2.7367,
+      "step": 45936
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7439029812812805,
+      "learning_rate": 8.758781712796584e-05,
+      "loss": 2.9358,
+      "step": 45937
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.737886369228363,
+      "learning_rate": 8.757694505346416e-05,
+      "loss": 2.7807,
+      "step": 45938
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6919266581535339,
+      "learning_rate": 8.756607353844405e-05,
+      "loss": 2.9541,
+      "step": 45939
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6991779804229736,
+      "learning_rate": 8.7555202582934e-05,
+      "loss": 2.8848,
+      "step": 45940
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7198300361633301,
+      "learning_rate": 8.754433218696258e-05,
+      "loss": 2.7478,
+      "step": 45941
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7473375201225281,
+      "learning_rate": 8.753346235055857e-05,
+      "loss": 2.7545,
+      "step": 45942
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6938808560371399,
+      "learning_rate": 8.752259307375046e-05,
+      "loss": 2.8858,
+      "step": 45943
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6992427110671997,
+      "learning_rate": 8.751172435656695e-05,
+      "loss": 3.0062,
+      "step": 45944
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7342805862426758,
+      "learning_rate": 8.750085619903672e-05,
+      "loss": 3.0506,
+      "step": 45945
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.714984655380249,
+      "learning_rate": 8.748998860118838e-05,
+      "loss": 2.9983,
+      "step": 45946
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6858196258544922,
+      "learning_rate": 8.74791215630505e-05,
+      "loss": 2.8878,
+      "step": 45947
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7082827687263489,
+      "learning_rate": 8.74682550846516e-05,
+      "loss": 2.883,
+      "step": 45948
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7259243726730347,
+      "learning_rate": 8.745738916602044e-05,
+      "loss": 2.6689,
+      "step": 45949
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7247135639190674,
+      "learning_rate": 8.744652380718568e-05,
+      "loss": 2.9188,
+      "step": 45950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7118974924087524,
+      "learning_rate": 8.743565900817582e-05,
+      "loss": 2.8574,
+      "step": 45951
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6941021680831909,
+      "learning_rate": 8.742479476901958e-05,
+      "loss": 3.0137,
+      "step": 45952
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7223904728889465,
+      "learning_rate": 8.741393108974545e-05,
+      "loss": 2.9941,
+      "step": 45953
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7473703026771545,
+      "learning_rate": 8.740306797038217e-05,
+      "loss": 2.8769,
+      "step": 45954
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6923856735229492,
+      "learning_rate": 8.739220541095831e-05,
+      "loss": 3.0033,
+      "step": 45955
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7007562518119812,
+      "learning_rate": 8.73813434115024e-05,
+      "loss": 3.1408,
+      "step": 45956
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7167068123817444,
+      "learning_rate": 8.737048197204317e-05,
+      "loss": 2.8446,
+      "step": 45957
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7058669328689575,
+      "learning_rate": 8.73596210926091e-05,
+      "loss": 2.9255,
+      "step": 45958
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7336747646331787,
+      "learning_rate": 8.734876077322887e-05,
+      "loss": 2.907,
+      "step": 45959
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6712983250617981,
+      "learning_rate": 8.733790101393114e-05,
+      "loss": 2.8126,
+      "step": 45960
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7041860818862915,
+      "learning_rate": 8.732704181474448e-05,
+      "loss": 2.9029,
+      "step": 45961
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.739032506942749,
+      "learning_rate": 8.731618317569742e-05,
+      "loss": 2.9452,
+      "step": 45962
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.790059506893158,
+      "learning_rate": 8.730532509681853e-05,
+      "loss": 3.131,
+      "step": 45963
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6874712109565735,
+      "learning_rate": 8.729446757813651e-05,
+      "loss": 2.8963,
+      "step": 45964
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.718464732170105,
+      "learning_rate": 8.728361061967998e-05,
+      "loss": 2.9917,
+      "step": 45965
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7371983528137207,
+      "learning_rate": 8.727275422147743e-05,
+      "loss": 3.0787,
+      "step": 45966
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.689892590045929,
+      "learning_rate": 8.726189838355756e-05,
+      "loss": 2.9387,
+      "step": 45967
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7095820903778076,
+      "learning_rate": 8.725104310594893e-05,
+      "loss": 2.8798,
+      "step": 45968
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7040510773658752,
+      "learning_rate": 8.724018838867999e-05,
+      "loss": 2.8971,
+      "step": 45969
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.713518500328064,
+      "learning_rate": 8.722933423177956e-05,
+      "loss": 2.9733,
+      "step": 45970
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6998233199119568,
+      "learning_rate": 8.721848063527601e-05,
+      "loss": 2.6608,
+      "step": 45971
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6755211353302002,
+      "learning_rate": 8.720762759919814e-05,
+      "loss": 2.9071,
+      "step": 45972
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7184854745864868,
+      "learning_rate": 8.719677512357432e-05,
+      "loss": 2.8504,
+      "step": 45973
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.704601526260376,
+      "learning_rate": 8.718592320843335e-05,
+      "loss": 2.9019,
+      "step": 45974
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7312195301055908,
+      "learning_rate": 8.717507185380359e-05,
+      "loss": 2.9244,
+      "step": 45975
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7099519371986389,
+      "learning_rate": 8.716422105971385e-05,
+      "loss": 2.9385,
+      "step": 45976
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7298969626426697,
+      "learning_rate": 8.715337082619256e-05,
+      "loss": 2.9582,
+      "step": 45977
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7228435277938843,
+      "learning_rate": 8.714252115326825e-05,
+      "loss": 3.0337,
+      "step": 45978
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7032173275947571,
+      "learning_rate": 8.713167204096967e-05,
+      "loss": 2.8777,
+      "step": 45979
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7508836388587952,
+      "learning_rate": 8.712082348932518e-05,
+      "loss": 2.8243,
+      "step": 45980
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7287681102752686,
+      "learning_rate": 8.71099754983635e-05,
+      "loss": 3.0163,
+      "step": 45981
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7005704045295715,
+      "learning_rate": 8.709912806811327e-05,
+      "loss": 2.8342,
+      "step": 45982
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7210054993629456,
+      "learning_rate": 8.708828119860295e-05,
+      "loss": 3.0389,
+      "step": 45983
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.693412184715271,
+      "learning_rate": 8.70774348898611e-05,
+      "loss": 2.8852,
+      "step": 45984
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.687518835067749,
+      "learning_rate": 8.706658914191625e-05,
+      "loss": 2.9454,
+      "step": 45985
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7263759970664978,
+      "learning_rate": 8.7055743954797e-05,
+      "loss": 2.9181,
+      "step": 45986
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7387471199035645,
+      "learning_rate": 8.704489932853203e-05,
+      "loss": 2.8456,
+      "step": 45987
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6722941398620605,
+      "learning_rate": 8.703405526314975e-05,
+      "loss": 2.8528,
+      "step": 45988
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6813567876815796,
+      "learning_rate": 8.702321175867884e-05,
+      "loss": 2.9437,
+      "step": 45989
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7118014693260193,
+      "learning_rate": 8.701236881514784e-05,
+      "loss": 2.8783,
+      "step": 45990
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7127276659011841,
+      "learning_rate": 8.700152643258516e-05,
+      "loss": 3.0367,
+      "step": 45991
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.8109259605407715,
+      "learning_rate": 8.699068461101955e-05,
+      "loss": 2.8818,
+      "step": 45992
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.689415693283081,
+      "learning_rate": 8.697984335047939e-05,
+      "loss": 3.0599,
+      "step": 45993
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7123958468437195,
+      "learning_rate": 8.696900265099343e-05,
+      "loss": 2.8822,
+      "step": 45994
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7250252366065979,
+      "learning_rate": 8.695816251259002e-05,
+      "loss": 2.9265,
+      "step": 45995
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7128042578697205,
+      "learning_rate": 8.694732293529783e-05,
+      "loss": 3.0612,
+      "step": 45996
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7381935119628906,
+      "learning_rate": 8.693648391914543e-05,
+      "loss": 2.7554,
+      "step": 45997
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6966440677642822,
+      "learning_rate": 8.692564546416137e-05,
+      "loss": 2.7047,
+      "step": 45998
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7341095209121704,
+      "learning_rate": 8.691480757037414e-05,
+      "loss": 3.1269,
+      "step": 45999
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6964755654335022,
+      "learning_rate": 8.690397023781219e-05,
+      "loss": 3.019,
+      "step": 46000
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7326573729515076,
+      "learning_rate": 8.689313346650419e-05,
+      "loss": 2.7295,
+      "step": 46001
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7106379866600037,
+      "learning_rate": 8.688229725647876e-05,
+      "loss": 2.8008,
+      "step": 46002
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6738712787628174,
+      "learning_rate": 8.687146160776424e-05,
+      "loss": 2.7775,
+      "step": 46003
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6668511033058167,
+      "learning_rate": 8.686062652038939e-05,
+      "loss": 2.9989,
+      "step": 46004
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.8261511921882629,
+      "learning_rate": 8.684979199438259e-05,
+      "loss": 2.7222,
+      "step": 46005
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 8.683895802977236e-05,
+      "loss": 3.1088,
+      "step": 46006
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7278913259506226,
+      "learning_rate": 8.682812462658736e-05,
+      "loss": 3.029,
+      "step": 46007
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6950763463973999,
+      "learning_rate": 8.681729178485595e-05,
+      "loss": 2.7216,
+      "step": 46008
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.753512442111969,
+      "learning_rate": 8.68064595046069e-05,
+      "loss": 2.9456,
+      "step": 46009
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7178856730461121,
+      "learning_rate": 8.679562778586848e-05,
+      "loss": 2.7979,
+      "step": 46010
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6799273490905762,
+      "learning_rate": 8.678479662866944e-05,
+      "loss": 3.0293,
+      "step": 46011
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7092909812927246,
+      "learning_rate": 8.677396603303821e-05,
+      "loss": 2.8555,
+      "step": 46012
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.744310736656189,
+      "learning_rate": 8.676313599900325e-05,
+      "loss": 3.0313,
+      "step": 46013
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7400776743888855,
+      "learning_rate": 8.675230652659322e-05,
+      "loss": 2.9776,
+      "step": 46014
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7001795768737793,
+      "learning_rate": 8.674147761583649e-05,
+      "loss": 3.0375,
+      "step": 46015
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7090204954147339,
+      "learning_rate": 8.673064926676166e-05,
+      "loss": 3.1105,
+      "step": 46016
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6803956031799316,
+      "learning_rate": 8.671982147939733e-05,
+      "loss": 2.9111,
+      "step": 46017
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6881393790245056,
+      "learning_rate": 8.670899425377191e-05,
+      "loss": 2.9803,
+      "step": 46018
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7794377207756042,
+      "learning_rate": 8.669816758991397e-05,
+      "loss": 3.0788,
+      "step": 46019
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6823427677154541,
+      "learning_rate": 8.668734148785205e-05,
+      "loss": 2.7918,
+      "step": 46020
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7123182415962219,
+      "learning_rate": 8.667651594761452e-05,
+      "loss": 2.896,
+      "step": 46021
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7293609976768494,
+      "learning_rate": 8.666569096923007e-05,
+      "loss": 3.039,
+      "step": 46022
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7326741218566895,
+      "learning_rate": 8.665486655272708e-05,
+      "loss": 3.0871,
+      "step": 46023
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7077974677085876,
+      "learning_rate": 8.664404269813416e-05,
+      "loss": 2.9412,
+      "step": 46024
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7263396382331848,
+      "learning_rate": 8.663321940547971e-05,
+      "loss": 2.8853,
+      "step": 46025
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6902568340301514,
+      "learning_rate": 8.662239667479238e-05,
+      "loss": 2.8761,
+      "step": 46026
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6834216713905334,
+      "learning_rate": 8.661157450610058e-05,
+      "loss": 2.987,
+      "step": 46027
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7192767262458801,
+      "learning_rate": 8.660075289943274e-05,
+      "loss": 3.0016,
+      "step": 46028
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.711889922618866,
+      "learning_rate": 8.658993185481752e-05,
+      "loss": 2.9948,
+      "step": 46029
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7222273349761963,
+      "learning_rate": 8.657911137228328e-05,
+      "loss": 2.9268,
+      "step": 46030
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7341220378875732,
+      "learning_rate": 8.656829145185858e-05,
+      "loss": 2.8919,
+      "step": 46031
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6676339507102966,
+      "learning_rate": 8.655747209357203e-05,
+      "loss": 2.9097,
+      "step": 46032
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7094964385032654,
+      "learning_rate": 8.654665329745199e-05,
+      "loss": 3.0597,
+      "step": 46033
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7654356360435486,
+      "learning_rate": 8.653583506352702e-05,
+      "loss": 2.7896,
+      "step": 46034
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7312438488006592,
+      "learning_rate": 8.652501739182546e-05,
+      "loss": 2.9281,
+      "step": 46035
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6991698741912842,
+      "learning_rate": 8.651420028237594e-05,
+      "loss": 2.9008,
+      "step": 46036
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.725242018699646,
+      "learning_rate": 8.650338373520702e-05,
+      "loss": 2.9847,
+      "step": 46037
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6978802680969238,
+      "learning_rate": 8.6492567750347e-05,
+      "loss": 2.6138,
+      "step": 46038
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7490013241767883,
+      "learning_rate": 8.648175232782458e-05,
+      "loss": 2.8943,
+      "step": 46039
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7092497944831848,
+      "learning_rate": 8.647093746766799e-05,
+      "loss": 2.9318,
+      "step": 46040
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.709816575050354,
+      "learning_rate": 8.6460123169906e-05,
+      "loss": 2.881,
+      "step": 46041
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7483521699905396,
+      "learning_rate": 8.644930943456693e-05,
+      "loss": 2.7731,
+      "step": 46042
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7091911435127258,
+      "learning_rate": 8.643849626167917e-05,
+      "loss": 2.8516,
+      "step": 46043
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7004270553588867,
+      "learning_rate": 8.642768365127141e-05,
+      "loss": 2.8595,
+      "step": 46044
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.703554093837738,
+      "learning_rate": 8.641687160337195e-05,
+      "loss": 2.5378,
+      "step": 46045
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.68243408203125,
+      "learning_rate": 8.640606011800935e-05,
+      "loss": 2.8221,
+      "step": 46046
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.743057906627655,
+      "learning_rate": 8.639524919521213e-05,
+      "loss": 2.9218,
+      "step": 46047
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6928690075874329,
+      "learning_rate": 8.638443883500877e-05,
+      "loss": 2.9012,
+      "step": 46048
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7314662337303162,
+      "learning_rate": 8.637362903742768e-05,
+      "loss": 2.9544,
+      "step": 46049
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6911783218383789,
+      "learning_rate": 8.63628198024972e-05,
+      "loss": 2.8013,
+      "step": 46050
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8123417496681213,
+      "learning_rate": 8.635201113024607e-05,
+      "loss": 2.9374,
+      "step": 46051
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7197860479354858,
+      "learning_rate": 8.634120302070254e-05,
+      "loss": 3.0019,
+      "step": 46052
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7315776944160461,
+      "learning_rate": 8.633039547389511e-05,
+      "loss": 2.7753,
+      "step": 46053
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.790367841720581,
+      "learning_rate": 8.631958848985241e-05,
+      "loss": 2.9611,
+      "step": 46054
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8130670189857483,
+      "learning_rate": 8.630878206860275e-05,
+      "loss": 2.861,
+      "step": 46055
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7119998931884766,
+      "learning_rate": 8.629797621017475e-05,
+      "loss": 2.7987,
+      "step": 46056
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7341967821121216,
+      "learning_rate": 8.62871709145966e-05,
+      "loss": 2.9332,
+      "step": 46057
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7142411470413208,
+      "learning_rate": 8.62763661818969e-05,
+      "loss": 2.8448,
+      "step": 46058
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7042397260665894,
+      "learning_rate": 8.626556201210424e-05,
+      "loss": 2.8789,
+      "step": 46059
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7835875153541565,
+      "learning_rate": 8.625475840524681e-05,
+      "loss": 2.9927,
+      "step": 46060
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.718367338180542,
+      "learning_rate": 8.624395536135336e-05,
+      "loss": 3.0031,
+      "step": 46061
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.742847204208374,
+      "learning_rate": 8.623315288045207e-05,
+      "loss": 2.9724,
+      "step": 46062
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7116962671279907,
+      "learning_rate": 8.62223509625716e-05,
+      "loss": 2.7999,
+      "step": 46063
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7158562541007996,
+      "learning_rate": 8.621154960774035e-05,
+      "loss": 3.0854,
+      "step": 46064
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7016698122024536,
+      "learning_rate": 8.620074881598662e-05,
+      "loss": 2.6261,
+      "step": 46065
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7218732833862305,
+      "learning_rate": 8.618994858733906e-05,
+      "loss": 3.0268,
+      "step": 46066
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7430708408355713,
+      "learning_rate": 8.617914892182592e-05,
+      "loss": 3.0253,
+      "step": 46067
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7524659633636475,
+      "learning_rate": 8.616834981947577e-05,
+      "loss": 2.8653,
+      "step": 46068
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6865372061729431,
+      "learning_rate": 8.615755128031714e-05,
+      "loss": 3.1231,
+      "step": 46069
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7529311180114746,
+      "learning_rate": 8.614675330437835e-05,
+      "loss": 2.8919,
+      "step": 46070
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7249378561973572,
+      "learning_rate": 8.613595589168787e-05,
+      "loss": 2.9534,
+      "step": 46071
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7141491770744324,
+      "learning_rate": 8.612515904227401e-05,
+      "loss": 2.8534,
+      "step": 46072
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6884422302246094,
+      "learning_rate": 8.611436275616534e-05,
+      "loss": 2.8794,
+      "step": 46073
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7267909646034241,
+      "learning_rate": 8.610356703339033e-05,
+      "loss": 2.7915,
+      "step": 46074
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7484807968139648,
+      "learning_rate": 8.609277187397728e-05,
+      "loss": 2.9957,
+      "step": 46075
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7652937769889832,
+      "learning_rate": 8.60819772779548e-05,
+      "loss": 2.8617,
+      "step": 46076
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6917968392372131,
+      "learning_rate": 8.607118324535114e-05,
+      "loss": 2.7486,
+      "step": 46077
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7388394474983215,
+      "learning_rate": 8.606038977619492e-05,
+      "loss": 2.768,
+      "step": 46078
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7591552734375,
+      "learning_rate": 8.604959687051444e-05,
+      "loss": 2.7493,
+      "step": 46079
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7432824969291687,
+      "learning_rate": 8.603880452833805e-05,
+      "loss": 2.7448,
+      "step": 46080
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6959530115127563,
+      "learning_rate": 8.602801274969436e-05,
+      "loss": 3.0574,
+      "step": 46081
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7385079264640808,
+      "learning_rate": 8.601722153461162e-05,
+      "loss": 2.8571,
+      "step": 46082
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6802938580513,
+      "learning_rate": 8.600643088311833e-05,
+      "loss": 2.9388,
+      "step": 46083
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7278597950935364,
+      "learning_rate": 8.599564079524302e-05,
+      "loss": 2.9041,
+      "step": 46084
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7153709530830383,
+      "learning_rate": 8.598485127101399e-05,
+      "loss": 2.9392,
+      "step": 46085
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7356995344161987,
+      "learning_rate": 8.59740623104597e-05,
+      "loss": 3.0901,
+      "step": 46086
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7170118689537048,
+      "learning_rate": 8.596327391360843e-05,
+      "loss": 3.0913,
+      "step": 46087
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7780619859695435,
+      "learning_rate": 8.595248608048869e-05,
+      "loss": 2.8865,
+      "step": 46088
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7607010006904602,
+      "learning_rate": 8.594169881112902e-05,
+      "loss": 2.9598,
+      "step": 46089
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7569159269332886,
+      "learning_rate": 8.593091210555764e-05,
+      "loss": 2.8838,
+      "step": 46090
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.738321840763092,
+      "learning_rate": 8.592012596380311e-05,
+      "loss": 2.94,
+      "step": 46091
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.722771406173706,
+      "learning_rate": 8.59093403858938e-05,
+      "loss": 3.0879,
+      "step": 46092
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7251681089401245,
+      "learning_rate": 8.589855537185794e-05,
+      "loss": 3.1147,
+      "step": 46093
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7316486239433289,
+      "learning_rate": 8.58877709217242e-05,
+      "loss": 2.9141,
+      "step": 46094
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6975620985031128,
+      "learning_rate": 8.587698703552078e-05,
+      "loss": 2.7911,
+      "step": 46095
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.69795161485672,
+      "learning_rate": 8.586620371327625e-05,
+      "loss": 2.9119,
+      "step": 46096
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7358713746070862,
+      "learning_rate": 8.585542095501883e-05,
+      "loss": 2.9688,
+      "step": 46097
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.692922830581665,
+      "learning_rate": 8.584463876077703e-05,
+      "loss": 3.0083,
+      "step": 46098
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7178481817245483,
+      "learning_rate": 8.58338571305793e-05,
+      "loss": 2.887,
+      "step": 46099
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6919596195220947,
+      "learning_rate": 8.582307606445402e-05,
+      "loss": 2.746,
+      "step": 46100
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7563260197639465,
+      "learning_rate": 8.581229556242951e-05,
+      "loss": 2.7533,
+      "step": 46101
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7244362831115723,
+      "learning_rate": 8.58015156245341e-05,
+      "loss": 3.0494,
+      "step": 46102
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7256781458854675,
+      "learning_rate": 8.579073625079627e-05,
+      "loss": 2.8626,
+      "step": 46103
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7332637310028076,
+      "learning_rate": 8.577995744124451e-05,
+      "loss": 2.8991,
+      "step": 46104
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7425779700279236,
+      "learning_rate": 8.576917919590704e-05,
+      "loss": 2.7907,
+      "step": 46105
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7489483952522278,
+      "learning_rate": 8.57584015148124e-05,
+      "loss": 2.9985,
+      "step": 46106
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7290831208229065,
+      "learning_rate": 8.57476243979889e-05,
+      "loss": 2.969,
+      "step": 46107
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.752694308757782,
+      "learning_rate": 8.573684784546485e-05,
+      "loss": 3.069,
+      "step": 46108
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7758665084838867,
+      "learning_rate": 8.572607185726879e-05,
+      "loss": 2.8742,
+      "step": 46109
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7011808753013611,
+      "learning_rate": 8.571529643342891e-05,
+      "loss": 3.0316,
+      "step": 46110
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7500031590461731,
+      "learning_rate": 8.570452157397378e-05,
+      "loss": 3.0522,
+      "step": 46111
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7314960956573486,
+      "learning_rate": 8.569374727893164e-05,
+      "loss": 3.0971,
+      "step": 46112
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6992918252944946,
+      "learning_rate": 8.568297354833102e-05,
+      "loss": 2.9469,
+      "step": 46113
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7275790572166443,
+      "learning_rate": 8.567220038220016e-05,
+      "loss": 3.1835,
+      "step": 46114
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7153680920600891,
+      "learning_rate": 8.566142778056743e-05,
+      "loss": 2.79,
+      "step": 46115
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7057393193244934,
+      "learning_rate": 8.565065574346132e-05,
+      "loss": 3.0257,
+      "step": 46116
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7389050126075745,
+      "learning_rate": 8.563988427091005e-05,
+      "loss": 2.9956,
+      "step": 46117
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.69362473487854,
+      "learning_rate": 8.562911336294204e-05,
+      "loss": 2.6957,
+      "step": 46118
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7017712593078613,
+      "learning_rate": 8.56183430195858e-05,
+      "loss": 2.9043,
+      "step": 46119
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7296519875526428,
+      "learning_rate": 8.560757324086951e-05,
+      "loss": 3.0865,
+      "step": 46120
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7155781388282776,
+      "learning_rate": 8.559680402682168e-05,
+      "loss": 2.9978,
+      "step": 46121
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7151504158973694,
+      "learning_rate": 8.558603537747062e-05,
+      "loss": 2.8498,
+      "step": 46122
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7091836929321289,
+      "learning_rate": 8.557526729284459e-05,
+      "loss": 3.0795,
+      "step": 46123
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.701084315776825,
+      "learning_rate": 8.55644997729721e-05,
+      "loss": 2.8647,
+      "step": 46124
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7470046281814575,
+      "learning_rate": 8.55537328178814e-05,
+      "loss": 2.8312,
+      "step": 46125
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7091386318206787,
+      "learning_rate": 8.554296642760095e-05,
+      "loss": 2.9283,
+      "step": 46126
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7808635234832764,
+      "learning_rate": 8.5532200602159e-05,
+      "loss": 2.9434,
+      "step": 46127
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6947294473648071,
+      "learning_rate": 8.552143534158403e-05,
+      "loss": 2.8923,
+      "step": 46128
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6899562478065491,
+      "learning_rate": 8.551067064590432e-05,
+      "loss": 2.7862,
+      "step": 46129
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6616258025169373,
+      "learning_rate": 8.54999065151481e-05,
+      "loss": 2.9572,
+      "step": 46130
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7294912338256836,
+      "learning_rate": 8.548914294934399e-05,
+      "loss": 3.0817,
+      "step": 46131
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7399985194206238,
+      "learning_rate": 8.547837994852007e-05,
+      "loss": 2.8683,
+      "step": 46132
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6928227543830872,
+      "learning_rate": 8.546761751270492e-05,
+      "loss": 2.9985,
+      "step": 46133
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7888341546058655,
+      "learning_rate": 8.545685564192669e-05,
+      "loss": 2.8663,
+      "step": 46134
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7424514889717102,
+      "learning_rate": 8.544609433621388e-05,
+      "loss": 2.8363,
+      "step": 46135
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7457790970802307,
+      "learning_rate": 8.543533359559479e-05,
+      "loss": 2.7649,
+      "step": 46136
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7245312333106995,
+      "learning_rate": 8.542457342009765e-05,
+      "loss": 2.641,
+      "step": 46137
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.761008620262146,
+      "learning_rate": 8.541381380975096e-05,
+      "loss": 2.9057,
+      "step": 46138
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7968191504478455,
+      "learning_rate": 8.54030547645829e-05,
+      "loss": 2.6337,
+      "step": 46139
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7402268648147583,
+      "learning_rate": 8.53922962846219e-05,
+      "loss": 3.0397,
+      "step": 46140
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8710418939590454,
+      "learning_rate": 8.538153836989635e-05,
+      "loss": 2.8891,
+      "step": 46141
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.734683096408844,
+      "learning_rate": 8.537078102043447e-05,
+      "loss": 2.929,
+      "step": 46142
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7593229413032532,
+      "learning_rate": 8.536002423626475e-05,
+      "loss": 2.8161,
+      "step": 46143
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6786648035049438,
+      "learning_rate": 8.534926801741538e-05,
+      "loss": 2.9832,
+      "step": 46144
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7301282286643982,
+      "learning_rate": 8.533851236391463e-05,
+      "loss": 2.9853,
+      "step": 46145
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.725745439529419,
+      "learning_rate": 8.532775727579104e-05,
+      "loss": 3.1637,
+      "step": 46146
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7195970416069031,
+      "learning_rate": 8.531700275307274e-05,
+      "loss": 2.8625,
+      "step": 46147
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7576825618743896,
+      "learning_rate": 8.530624879578824e-05,
+      "loss": 2.8941,
+      "step": 46148
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7363550066947937,
+      "learning_rate": 8.529549540396564e-05,
+      "loss": 3.004,
+      "step": 46149
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6844567656517029,
+      "learning_rate": 8.528474257763346e-05,
+      "loss": 2.9282,
+      "step": 46150
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6960411667823792,
+      "learning_rate": 8.527399031682e-05,
+      "loss": 3.001,
+      "step": 46151
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7222846150398254,
+      "learning_rate": 8.52632386215534e-05,
+      "loss": 2.987,
+      "step": 46152
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7160438895225525,
+      "learning_rate": 8.525248749186219e-05,
+      "loss": 2.9166,
+      "step": 46153
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7318093776702881,
+      "learning_rate": 8.52417369277745e-05,
+      "loss": 2.9424,
+      "step": 46154
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7212334275245667,
+      "learning_rate": 8.523098692931877e-05,
+      "loss": 2.9443,
+      "step": 46155
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7121982574462891,
+      "learning_rate": 8.522023749652338e-05,
+      "loss": 2.9703,
+      "step": 46156
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.722030520439148,
+      "learning_rate": 8.520948862941646e-05,
+      "loss": 3.0426,
+      "step": 46157
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7229368090629578,
+      "learning_rate": 8.519874032802658e-05,
+      "loss": 3.1051,
+      "step": 46158
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7376280426979065,
+      "learning_rate": 8.518799259238167e-05,
+      "loss": 2.8847,
+      "step": 46159
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6587594151496887,
+      "learning_rate": 8.517724542251026e-05,
+      "loss": 2.7757,
+      "step": 46160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7172147631645203,
+      "learning_rate": 8.516649881844077e-05,
+      "loss": 2.6424,
+      "step": 46161
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7306764125823975,
+      "learning_rate": 8.515575278020125e-05,
+      "loss": 2.9843,
+      "step": 46162
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7316793203353882,
+      "learning_rate": 8.514500730782023e-05,
+      "loss": 2.9263,
+      "step": 46163
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8195810317993164,
+      "learning_rate": 8.513426240132584e-05,
+      "loss": 2.9318,
+      "step": 46164
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7335458397865295,
+      "learning_rate": 8.51235180607465e-05,
+      "loss": 2.8091,
+      "step": 46165
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6977335810661316,
+      "learning_rate": 8.511277428611048e-05,
+      "loss": 3.0965,
+      "step": 46166
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6865370273590088,
+      "learning_rate": 8.510203107744598e-05,
+      "loss": 2.9652,
+      "step": 46167
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.737516462802887,
+      "learning_rate": 8.509128843478145e-05,
+      "loss": 2.9361,
+      "step": 46168
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7313123345375061,
+      "learning_rate": 8.508054635814501e-05,
+      "loss": 2.9974,
+      "step": 46169
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7545174360275269,
+      "learning_rate": 8.506980484756507e-05,
+      "loss": 3.1408,
+      "step": 46170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7121573090553284,
+      "learning_rate": 8.505906390306994e-05,
+      "loss": 2.9658,
+      "step": 46171
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7106088399887085,
+      "learning_rate": 8.504832352468795e-05,
+      "loss": 2.9336,
+      "step": 46172
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.741375744342804,
+      "learning_rate": 8.503758371244724e-05,
+      "loss": 3.0173,
+      "step": 46173
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7445472478866577,
+      "learning_rate": 8.502684446637608e-05,
+      "loss": 2.7716,
+      "step": 46174
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7355408668518066,
+      "learning_rate": 8.501610578650289e-05,
+      "loss": 2.6886,
+      "step": 46175
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7312442660331726,
+      "learning_rate": 8.500536767285595e-05,
+      "loss": 2.8622,
+      "step": 46176
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7817408442497253,
+      "learning_rate": 8.49946301254634e-05,
+      "loss": 2.7271,
+      "step": 46177
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7041735649108887,
+      "learning_rate": 8.498389314435372e-05,
+      "loss": 2.8909,
+      "step": 46178
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7182589173316956,
+      "learning_rate": 8.497315672955498e-05,
+      "loss": 2.8677,
+      "step": 46179
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6795207262039185,
+      "learning_rate": 8.496242088109569e-05,
+      "loss": 2.7094,
+      "step": 46180
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6875749826431274,
+      "learning_rate": 8.495168559900397e-05,
+      "loss": 2.8307,
+      "step": 46181
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7328691482543945,
+      "learning_rate": 8.494095088330803e-05,
+      "loss": 2.9532,
+      "step": 46182
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7125970721244812,
+      "learning_rate": 8.493021673403635e-05,
+      "loss": 2.7698,
+      "step": 46183
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7266592383384705,
+      "learning_rate": 8.491948315121696e-05,
+      "loss": 2.8341,
+      "step": 46184
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7378901243209839,
+      "learning_rate": 8.490875013487831e-05,
+      "loss": 2.9183,
+      "step": 46185
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7208519577980042,
+      "learning_rate": 8.489801768504868e-05,
+      "loss": 3.2133,
+      "step": 46186
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7134647369384766,
+      "learning_rate": 8.488728580175625e-05,
+      "loss": 2.9837,
+      "step": 46187
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7949679493904114,
+      "learning_rate": 8.487655448502934e-05,
+      "loss": 2.9484,
+      "step": 46188
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7748767137527466,
+      "learning_rate": 8.48658237348961e-05,
+      "loss": 2.7965,
+      "step": 46189
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7083140015602112,
+      "learning_rate": 8.485509355138485e-05,
+      "loss": 2.912,
+      "step": 46190
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6907854676246643,
+      "learning_rate": 8.484436393452399e-05,
+      "loss": 2.6598,
+      "step": 46191
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7154836058616638,
+      "learning_rate": 8.483363488434156e-05,
+      "loss": 2.9983,
+      "step": 46192
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6884499788284302,
+      "learning_rate": 8.482290640086602e-05,
+      "loss": 2.947,
+      "step": 46193
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.723860502243042,
+      "learning_rate": 8.481217848412554e-05,
+      "loss": 2.8889,
+      "step": 46194
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7943236827850342,
+      "learning_rate": 8.480145113414828e-05,
+      "loss": 2.9106,
+      "step": 46195
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7377872467041016,
+      "learning_rate": 8.479072435096265e-05,
+      "loss": 2.8235,
+      "step": 46196
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6710457801818848,
+      "learning_rate": 8.477999813459678e-05,
+      "loss": 2.9226,
+      "step": 46197
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6965027451515198,
+      "learning_rate": 8.476927248507904e-05,
+      "loss": 3.1496,
+      "step": 46198
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6677283048629761,
+      "learning_rate": 8.475854740243755e-05,
+      "loss": 2.8667,
+      "step": 46199
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7114673256874084,
+      "learning_rate": 8.474782288670057e-05,
+      "loss": 3.0478,
+      "step": 46200
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.734982967376709,
+      "learning_rate": 8.473709893789654e-05,
+      "loss": 3.0154,
+      "step": 46201
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7224012613296509,
+      "learning_rate": 8.472637555605356e-05,
+      "loss": 2.8927,
+      "step": 46202
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7512854337692261,
+      "learning_rate": 8.471565274119984e-05,
+      "loss": 2.7395,
+      "step": 46203
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7039549946784973,
+      "learning_rate": 8.470493049336361e-05,
+      "loss": 2.8335,
+      "step": 46204
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7275938987731934,
+      "learning_rate": 8.469420881257322e-05,
+      "loss": 2.8,
+      "step": 46205
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6578214168548584,
+      "learning_rate": 8.468348769885678e-05,
+      "loss": 2.8258,
+      "step": 46206
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6866016387939453,
+      "learning_rate": 8.467276715224257e-05,
+      "loss": 2.9636,
+      "step": 46207
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7433063983917236,
+      "learning_rate": 8.466204717275894e-05,
+      "loss": 3.011,
+      "step": 46208
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6856465935707092,
+      "learning_rate": 8.465132776043407e-05,
+      "loss": 2.7571,
+      "step": 46209
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6921495795249939,
+      "learning_rate": 8.464060891529612e-05,
+      "loss": 2.729,
+      "step": 46210
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7133631706237793,
+      "learning_rate": 8.46298906373733e-05,
+      "loss": 2.9614,
+      "step": 46211
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6963069438934326,
+      "learning_rate": 8.461917292669389e-05,
+      "loss": 2.9202,
+      "step": 46212
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7098976373672485,
+      "learning_rate": 8.46084557832862e-05,
+      "loss": 2.7825,
+      "step": 46213
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6936521530151367,
+      "learning_rate": 8.459773920717831e-05,
+      "loss": 2.7475,
+      "step": 46214
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6678090691566467,
+      "learning_rate": 8.458702319839861e-05,
+      "loss": 2.8444,
+      "step": 46215
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8158767223358154,
+      "learning_rate": 8.457630775697522e-05,
+      "loss": 2.8724,
+      "step": 46216
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7068122625350952,
+      "learning_rate": 8.456559288293632e-05,
+      "loss": 2.9663,
+      "step": 46217
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7318935990333557,
+      "learning_rate": 8.455487857631023e-05,
+      "loss": 2.9523,
+      "step": 46218
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7318170666694641,
+      "learning_rate": 8.454416483712505e-05,
+      "loss": 2.8784,
+      "step": 46219
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.718914270401001,
+      "learning_rate": 8.45334516654092e-05,
+      "loss": 2.9895,
+      "step": 46220
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7228567600250244,
+      "learning_rate": 8.452273906119064e-05,
+      "loss": 2.9328,
+      "step": 46221
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7042717933654785,
+      "learning_rate": 8.451202702449773e-05,
+      "loss": 3.196,
+      "step": 46222
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7152907848358154,
+      "learning_rate": 8.450131555535876e-05,
+      "loss": 2.9061,
+      "step": 46223
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7358282208442688,
+      "learning_rate": 8.449060465380188e-05,
+      "loss": 2.9795,
+      "step": 46224
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6989845037460327,
+      "learning_rate": 8.447989431985521e-05,
+      "loss": 2.791,
+      "step": 46225
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7169927954673767,
+      "learning_rate": 8.446918455354699e-05,
+      "loss": 2.9684,
+      "step": 46226
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7589333653450012,
+      "learning_rate": 8.44584753549054e-05,
+      "loss": 2.7864,
+      "step": 46227
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7092520594596863,
+      "learning_rate": 8.444776672395886e-05,
+      "loss": 3.1328,
+      "step": 46228
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7001926898956299,
+      "learning_rate": 8.443705866073526e-05,
+      "loss": 2.773,
+      "step": 46229
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.713153600692749,
+      "learning_rate": 8.44263511652631e-05,
+      "loss": 2.9185,
+      "step": 46230
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7044378519058228,
+      "learning_rate": 8.441564423757045e-05,
+      "loss": 2.8194,
+      "step": 46231
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7447625398635864,
+      "learning_rate": 8.44049378776854e-05,
+      "loss": 2.9084,
+      "step": 46232
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7260729074478149,
+      "learning_rate": 8.439423208563633e-05,
+      "loss": 2.9366,
+      "step": 46233
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6695955991744995,
+      "learning_rate": 8.438352686145124e-05,
+      "loss": 2.9983,
+      "step": 46234
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7658150792121887,
+      "learning_rate": 8.43728222051586e-05,
+      "loss": 2.9357,
+      "step": 46235
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6993694305419922,
+      "learning_rate": 8.436211811678632e-05,
+      "loss": 2.9622,
+      "step": 46236
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7462490200996399,
+      "learning_rate": 8.435141459636281e-05,
+      "loss": 2.8533,
+      "step": 46237
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7425763607025146,
+      "learning_rate": 8.43407116439162e-05,
+      "loss": 2.9425,
+      "step": 46238
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7130029201507568,
+      "learning_rate": 8.433000925947452e-05,
+      "loss": 2.857,
+      "step": 46239
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6981950402259827,
+      "learning_rate": 8.43193074430662e-05,
+      "loss": 2.7975,
+      "step": 46240
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6925346255302429,
+      "learning_rate": 8.430860619471921e-05,
+      "loss": 2.9858,
+      "step": 46241
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6922650933265686,
+      "learning_rate": 8.429790551446186e-05,
+      "loss": 2.8014,
+      "step": 46242
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7049052119255066,
+      "learning_rate": 8.42872054023224e-05,
+      "loss": 3.0081,
+      "step": 46243
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7104154825210571,
+      "learning_rate": 8.427650585832884e-05,
+      "loss": 2.9192,
+      "step": 46244
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6965845227241516,
+      "learning_rate": 8.426580688250952e-05,
+      "loss": 3.0567,
+      "step": 46245
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6756129264831543,
+      "learning_rate": 8.425510847489257e-05,
+      "loss": 2.8244,
+      "step": 46246
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7112818956375122,
+      "learning_rate": 8.424441063550603e-05,
+      "loss": 2.77,
+      "step": 46247
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7068428993225098,
+      "learning_rate": 8.42337133643783e-05,
+      "loss": 2.8858,
+      "step": 46248
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6985403299331665,
+      "learning_rate": 8.422301666153734e-05,
+      "loss": 2.8649,
+      "step": 46249
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6854354739189148,
+      "learning_rate": 8.421232052701152e-05,
+      "loss": 2.9009,
+      "step": 46250
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6568443775177002,
+      "learning_rate": 8.42016249608288e-05,
+      "loss": 2.7745,
+      "step": 46251
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7257750034332275,
+      "learning_rate": 8.419092996301761e-05,
+      "loss": 2.9674,
+      "step": 46252
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7708278894424438,
+      "learning_rate": 8.418023553360596e-05,
+      "loss": 2.837,
+      "step": 46253
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7032803893089294,
+      "learning_rate": 8.416954167262193e-05,
+      "loss": 3.0264,
+      "step": 46254
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6957308650016785,
+      "learning_rate": 8.415884838009388e-05,
+      "loss": 2.9592,
+      "step": 46255
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7218675017356873,
+      "learning_rate": 8.414815565604983e-05,
+      "loss": 2.8956,
+      "step": 46256
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7233446836471558,
+      "learning_rate": 8.413746350051797e-05,
+      "loss": 2.9225,
+      "step": 46257
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7115861177444458,
+      "learning_rate": 8.412677191352656e-05,
+      "loss": 3.0288,
+      "step": 46258
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7438410520553589,
+      "learning_rate": 8.41160808951037e-05,
+      "loss": 2.8894,
+      "step": 46259
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.700964629650116,
+      "learning_rate": 8.410539044527753e-05,
+      "loss": 3.0023,
+      "step": 46260
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7393155694007874,
+      "learning_rate": 8.409470056407614e-05,
+      "loss": 2.9852,
+      "step": 46261
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6867033243179321,
+      "learning_rate": 8.408401125152776e-05,
+      "loss": 2.9322,
+      "step": 46262
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6977315545082092,
+      "learning_rate": 8.407332250766063e-05,
+      "loss": 2.9155,
+      "step": 46263
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6820677518844604,
+      "learning_rate": 8.40626343325027e-05,
+      "loss": 2.9321,
+      "step": 46264
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7445458173751831,
+      "learning_rate": 8.405194672608233e-05,
+      "loss": 2.9113,
+      "step": 46265
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6989564895629883,
+      "learning_rate": 8.40412596884275e-05,
+      "loss": 3.0236,
+      "step": 46266
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7086830735206604,
+      "learning_rate": 8.403057321956648e-05,
+      "loss": 2.9824,
+      "step": 46267
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7059872150421143,
+      "learning_rate": 8.401988731952741e-05,
+      "loss": 3.0183,
+      "step": 46268
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7030812501907349,
+      "learning_rate": 8.400920198833829e-05,
+      "loss": 3.0954,
+      "step": 46269
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7751390337944031,
+      "learning_rate": 8.399851722602743e-05,
+      "loss": 2.9932,
+      "step": 46270
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7409400343894958,
+      "learning_rate": 8.398783303262283e-05,
+      "loss": 2.9746,
+      "step": 46271
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7673764228820801,
+      "learning_rate": 8.39771494081527e-05,
+      "loss": 2.8363,
+      "step": 46272
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7108684182167053,
+      "learning_rate": 8.396646635264528e-05,
+      "loss": 3.0562,
+      "step": 46273
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6982866525650024,
+      "learning_rate": 8.395578386612861e-05,
+      "loss": 2.8194,
+      "step": 46274
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7467199563980103,
+      "learning_rate": 8.394510194863085e-05,
+      "loss": 3.0077,
+      "step": 46275
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7095677256584167,
+      "learning_rate": 8.393442060017998e-05,
+      "loss": 2.9502,
+      "step": 46276
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7236535549163818,
+      "learning_rate": 8.392373982080428e-05,
+      "loss": 2.7709,
+      "step": 46277
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7553972005844116,
+      "learning_rate": 8.391305961053197e-05,
+      "loss": 2.9195,
+      "step": 46278
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7126523852348328,
+      "learning_rate": 8.390237996939098e-05,
+      "loss": 2.9921,
+      "step": 46279
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.743773341178894,
+      "learning_rate": 8.389170089740962e-05,
+      "loss": 2.9643,
+      "step": 46280
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7017213106155396,
+      "learning_rate": 8.388102239461582e-05,
+      "loss": 3.0562,
+      "step": 46281
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7224516868591309,
+      "learning_rate": 8.387034446103794e-05,
+      "loss": 2.9265,
+      "step": 46282
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7359697818756104,
+      "learning_rate": 8.385966709670395e-05,
+      "loss": 3.0036,
+      "step": 46283
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6996152400970459,
+      "learning_rate": 8.384899030164189e-05,
+      "loss": 2.9987,
+      "step": 46284
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6811954975128174,
+      "learning_rate": 8.383831407588011e-05,
+      "loss": 3.0215,
+      "step": 46285
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.746880829334259,
+      "learning_rate": 8.382763841944654e-05,
+      "loss": 3.0656,
+      "step": 46286
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7244267463684082,
+      "learning_rate": 8.381696333236944e-05,
+      "loss": 2.9721,
+      "step": 46287
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7403231263160706,
+      "learning_rate": 8.380628881467674e-05,
+      "loss": 3.0218,
+      "step": 46288
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7562906742095947,
+      "learning_rate": 8.379561486639679e-05,
+      "loss": 2.9035,
+      "step": 46289
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7179818749427795,
+      "learning_rate": 8.378494148755753e-05,
+      "loss": 2.8786,
+      "step": 46290
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7321277856826782,
+      "learning_rate": 8.377426867818708e-05,
+      "loss": 2.9989,
+      "step": 46291
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7414536476135254,
+      "learning_rate": 8.376359643831369e-05,
+      "loss": 2.8159,
+      "step": 46292
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7076948285102844,
+      "learning_rate": 8.375292476796524e-05,
+      "loss": 2.7403,
+      "step": 46293
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7004036903381348,
+      "learning_rate": 8.374225366717002e-05,
+      "loss": 2.9799,
+      "step": 46294
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.728317379951477,
+      "learning_rate": 8.373158313595612e-05,
+      "loss": 2.6187,
+      "step": 46295
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7136837840080261,
+      "learning_rate": 8.372091317435164e-05,
+      "loss": 2.8687,
+      "step": 46296
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6955955028533936,
+      "learning_rate": 8.371024378238464e-05,
+      "loss": 2.9601,
+      "step": 46297
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6891588568687439,
+      "learning_rate": 8.369957496008314e-05,
+      "loss": 2.9733,
+      "step": 46298
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7407911419868469,
+      "learning_rate": 8.368890670747535e-05,
+      "loss": 2.8449,
+      "step": 46299
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7350627183914185,
+      "learning_rate": 8.367823902458941e-05,
+      "loss": 2.957,
+      "step": 46300
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7319565415382385,
+      "learning_rate": 8.366757191145328e-05,
+      "loss": 2.77,
+      "step": 46301
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7195742130279541,
+      "learning_rate": 8.365690536809525e-05,
+      "loss": 2.8073,
+      "step": 46302
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7700269818305969,
+      "learning_rate": 8.364623939454318e-05,
+      "loss": 2.8723,
+      "step": 46303
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7323468923568726,
+      "learning_rate": 8.363557399082535e-05,
+      "loss": 2.8773,
+      "step": 46304
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7363598346710205,
+      "learning_rate": 8.362490915696981e-05,
+      "loss": 2.843,
+      "step": 46305
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7405996322631836,
+      "learning_rate": 8.36142448930045e-05,
+      "loss": 3.0222,
+      "step": 46306
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7685582041740417,
+      "learning_rate": 8.360358119895773e-05,
+      "loss": 2.8072,
+      "step": 46307
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7079715728759766,
+      "learning_rate": 8.35929180748574e-05,
+      "loss": 2.8948,
+      "step": 46308
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7493276000022888,
+      "learning_rate": 8.358225552073166e-05,
+      "loss": 2.8799,
+      "step": 46309
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7022292613983154,
+      "learning_rate": 8.357159353660871e-05,
+      "loss": 3.0506,
+      "step": 46310
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7002536058425903,
+      "learning_rate": 8.356093212251652e-05,
+      "loss": 2.9324,
+      "step": 46311
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7278270721435547,
+      "learning_rate": 8.35502712784832e-05,
+      "loss": 2.9016,
+      "step": 46312
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6652935147285461,
+      "learning_rate": 8.353961100453668e-05,
+      "loss": 3.0654,
+      "step": 46313
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7217055559158325,
+      "learning_rate": 8.352895130070518e-05,
+      "loss": 2.75,
+      "step": 46314
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7330880165100098,
+      "learning_rate": 8.351829216701685e-05,
+      "loss": 2.9563,
+      "step": 46315
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6977697014808655,
+      "learning_rate": 8.350763360349961e-05,
+      "loss": 3.0273,
+      "step": 46316
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7105409502983093,
+      "learning_rate": 8.349697561018167e-05,
+      "loss": 2.9189,
+      "step": 46317
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7222777009010315,
+      "learning_rate": 8.348631818709103e-05,
+      "loss": 2.8138,
+      "step": 46318
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6798412203788757,
+      "learning_rate": 8.347566133425568e-05,
+      "loss": 2.9272,
+      "step": 46319
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6820101737976074,
+      "learning_rate": 8.346500505170385e-05,
+      "loss": 3.1816,
+      "step": 46320
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7206324934959412,
+      "learning_rate": 8.345434933946345e-05,
+      "loss": 2.8773,
+      "step": 46321
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7084947228431702,
+      "learning_rate": 8.344369419756268e-05,
+      "loss": 2.8789,
+      "step": 46322
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7819315195083618,
+      "learning_rate": 8.343303962602948e-05,
+      "loss": 2.7798,
+      "step": 46323
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7215809226036072,
+      "learning_rate": 8.342238562489192e-05,
+      "loss": 3.1057,
+      "step": 46324
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7359743118286133,
+      "learning_rate": 8.341173219417826e-05,
+      "loss": 2.7669,
+      "step": 46325
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7675909996032715,
+      "learning_rate": 8.340107933391641e-05,
+      "loss": 2.9143,
+      "step": 46326
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7441685199737549,
+      "learning_rate": 8.339042704413443e-05,
+      "loss": 2.8385,
+      "step": 46327
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6832141280174255,
+      "learning_rate": 8.337977532486024e-05,
+      "loss": 2.7445,
+      "step": 46328
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6876572966575623,
+      "learning_rate": 8.336912417612208e-05,
+      "loss": 3.1471,
+      "step": 46329
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6860120892524719,
+      "learning_rate": 8.335847359794805e-05,
+      "loss": 3.0647,
+      "step": 46330
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6835752725601196,
+      "learning_rate": 8.334782359036599e-05,
+      "loss": 3.1026,
+      "step": 46331
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7607629299163818,
+      "learning_rate": 8.333717415340419e-05,
+      "loss": 2.8308,
+      "step": 46332
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7346097826957703,
+      "learning_rate": 8.332652528709054e-05,
+      "loss": 2.8343,
+      "step": 46333
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7202624678611755,
+      "learning_rate": 8.331587699145303e-05,
+      "loss": 2.9422,
+      "step": 46334
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7167713642120361,
+      "learning_rate": 8.330522926651992e-05,
+      "loss": 2.8753,
+      "step": 46335
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6617588400840759,
+      "learning_rate": 8.329458211231902e-05,
+      "loss": 2.9843,
+      "step": 46336
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7064581513404846,
+      "learning_rate": 8.32839355288786e-05,
+      "loss": 2.9314,
+      "step": 46337
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.670749306678772,
+      "learning_rate": 8.327328951622646e-05,
+      "loss": 2.8856,
+      "step": 46338
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6679714918136597,
+      "learning_rate": 8.326264407439088e-05,
+      "loss": 2.9836,
+      "step": 46339
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.729651153087616,
+      "learning_rate": 8.325199920339978e-05,
+      "loss": 2.982,
+      "step": 46340
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.706098735332489,
+      "learning_rate": 8.324135490328112e-05,
+      "loss": 2.766,
+      "step": 46341
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7295530438423157,
+      "learning_rate": 8.32307111740631e-05,
+      "loss": 3.0328,
+      "step": 46342
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6685080528259277,
+      "learning_rate": 8.322006801577356e-05,
+      "loss": 2.9178,
+      "step": 46343
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6723880171775818,
+      "learning_rate": 8.320942542844065e-05,
+      "loss": 2.8969,
+      "step": 46344
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.741836428642273,
+      "learning_rate": 8.319878341209248e-05,
+      "loss": 3.0116,
+      "step": 46345
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7528204321861267,
+      "learning_rate": 8.318814196675689e-05,
+      "loss": 2.9499,
+      "step": 46346
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7128376364707947,
+      "learning_rate": 8.317750109246213e-05,
+      "loss": 2.8801,
+      "step": 46347
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7809441685676575,
+      "learning_rate": 8.316686078923608e-05,
+      "loss": 2.8204,
+      "step": 46348
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7417147755622864,
+      "learning_rate": 8.315622105710672e-05,
+      "loss": 2.7649,
+      "step": 46349
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6948042511940002,
+      "learning_rate": 8.31455818961022e-05,
+      "loss": 2.8864,
+      "step": 46350
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.757095456123352,
+      "learning_rate": 8.313494330625041e-05,
+      "loss": 2.9817,
+      "step": 46351
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6737290024757385,
+      "learning_rate": 8.312430528757954e-05,
+      "loss": 2.9095,
+      "step": 46352
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7008237838745117,
+      "learning_rate": 8.31136678401174e-05,
+      "loss": 2.9619,
+      "step": 46353
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7046443223953247,
+      "learning_rate": 8.310303096389224e-05,
+      "loss": 2.9833,
+      "step": 46354
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7089089751243591,
+      "learning_rate": 8.309239465893191e-05,
+      "loss": 2.67,
+      "step": 46355
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7592144012451172,
+      "learning_rate": 8.30817589252644e-05,
+      "loss": 2.9825,
+      "step": 46356
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6998030543327332,
+      "learning_rate": 8.307112376291789e-05,
+      "loss": 2.8194,
+      "step": 46357
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.714889407157898,
+      "learning_rate": 8.306048917192021e-05,
+      "loss": 2.9081,
+      "step": 46358
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7006431818008423,
+      "learning_rate": 8.304985515229943e-05,
+      "loss": 3.1693,
+      "step": 46359
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7032368183135986,
+      "learning_rate": 8.303922170408366e-05,
+      "loss": 2.8906,
+      "step": 46360
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7039076089859009,
+      "learning_rate": 8.302858882730085e-05,
+      "loss": 3.0944,
+      "step": 46361
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7378221750259399,
+      "learning_rate": 8.301795652197898e-05,
+      "loss": 2.7746,
+      "step": 46362
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.738791823387146,
+      "learning_rate": 8.300732478814594e-05,
+      "loss": 2.9973,
+      "step": 46363
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7050439715385437,
+      "learning_rate": 8.299669362582995e-05,
+      "loss": 2.8897,
+      "step": 46364
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7334348559379578,
+      "learning_rate": 8.298606303505883e-05,
+      "loss": 2.8901,
+      "step": 46365
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6942351460456848,
+      "learning_rate": 8.297543301586062e-05,
+      "loss": 3.1832,
+      "step": 46366
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7496381998062134,
+      "learning_rate": 8.296480356826347e-05,
+      "loss": 2.9803,
+      "step": 46367
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7410102486610413,
+      "learning_rate": 8.295417469229514e-05,
+      "loss": 2.7448,
+      "step": 46368
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6778762936592102,
+      "learning_rate": 8.294354638798386e-05,
+      "loss": 2.9249,
+      "step": 46369
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7150377035140991,
+      "learning_rate": 8.293291865535748e-05,
+      "loss": 2.9564,
+      "step": 46370
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7311868071556091,
+      "learning_rate": 8.292229149444394e-05,
+      "loss": 2.8055,
+      "step": 46371
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7008099555969238,
+      "learning_rate": 8.291166490527143e-05,
+      "loss": 2.923,
+      "step": 46372
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7243220210075378,
+      "learning_rate": 8.290103888786768e-05,
+      "loss": 3.0309,
+      "step": 46373
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7139156460762024,
+      "learning_rate": 8.289041344226091e-05,
+      "loss": 3.1218,
+      "step": 46374
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7168537378311157,
+      "learning_rate": 8.287978856847894e-05,
+      "loss": 2.9758,
+      "step": 46375
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7514629364013672,
+      "learning_rate": 8.286916426654988e-05,
+      "loss": 2.6922,
+      "step": 46376
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7429746985435486,
+      "learning_rate": 8.285854053650167e-05,
+      "loss": 3.0634,
+      "step": 46377
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.8082495331764221,
+      "learning_rate": 8.284791737836219e-05,
+      "loss": 2.968,
+      "step": 46378
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7452201843261719,
+      "learning_rate": 8.283729479215959e-05,
+      "loss": 2.9215,
+      "step": 46379
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7076332569122314,
+      "learning_rate": 8.282667277792168e-05,
+      "loss": 2.6926,
+      "step": 46380
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6897568106651306,
+      "learning_rate": 8.281605133567652e-05,
+      "loss": 2.9806,
+      "step": 46381
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7367916703224182,
+      "learning_rate": 8.280543046545215e-05,
+      "loss": 2.9092,
+      "step": 46382
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7310504913330078,
+      "learning_rate": 8.27948101672764e-05,
+      "loss": 2.8276,
+      "step": 46383
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7102162837982178,
+      "learning_rate": 8.278419044117749e-05,
+      "loss": 2.9182,
+      "step": 46384
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6854374408721924,
+      "learning_rate": 8.277357128718305e-05,
+      "loss": 2.9456,
+      "step": 46385
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7474464178085327,
+      "learning_rate": 8.276295270532119e-05,
+      "loss": 2.8981,
+      "step": 46386
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7687355875968933,
+      "learning_rate": 8.275233469562004e-05,
+      "loss": 2.7403,
+      "step": 46387
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7094236612319946,
+      "learning_rate": 8.274171725810729e-05,
+      "loss": 2.9686,
+      "step": 46388
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7075121402740479,
+      "learning_rate": 8.273110039281117e-05,
+      "loss": 2.8354,
+      "step": 46389
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.724186360836029,
+      "learning_rate": 8.272048409975942e-05,
+      "loss": 2.7629,
+      "step": 46390
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7304971218109131,
+      "learning_rate": 8.27098683789802e-05,
+      "loss": 2.7759,
+      "step": 46391
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7718026041984558,
+      "learning_rate": 8.269925323050134e-05,
+      "loss": 3.0894,
+      "step": 46392
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7265807390213013,
+      "learning_rate": 8.268863865435074e-05,
+      "loss": 2.9327,
+      "step": 46393
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6995274424552917,
+      "learning_rate": 8.267802465055656e-05,
+      "loss": 2.6844,
+      "step": 46394
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7589938044548035,
+      "learning_rate": 8.266741121914653e-05,
+      "loss": 2.9637,
+      "step": 46395
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7459881901741028,
+      "learning_rate": 8.265679836014872e-05,
+      "loss": 2.7266,
+      "step": 46396
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7490134239196777,
+      "learning_rate": 8.264618607359115e-05,
+      "loss": 2.9801,
+      "step": 46397
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7336806058883667,
+      "learning_rate": 8.263557435950172e-05,
+      "loss": 2.8788,
+      "step": 46398
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6905733942985535,
+      "learning_rate": 8.262496321790831e-05,
+      "loss": 2.9608,
+      "step": 46399
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6890795230865479,
+      "learning_rate": 8.261435264883887e-05,
+      "loss": 2.9883,
+      "step": 46400
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7028673887252808,
+      "learning_rate": 8.260374265232135e-05,
+      "loss": 2.8123,
+      "step": 46401
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7058315873146057,
+      "learning_rate": 8.259313322838386e-05,
+      "loss": 2.807,
+      "step": 46402
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7144296169281006,
+      "learning_rate": 8.258252437705411e-05,
+      "loss": 2.912,
+      "step": 46403
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7268756628036499,
+      "learning_rate": 8.257191609836024e-05,
+      "loss": 2.9637,
+      "step": 46404
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6951913237571716,
+      "learning_rate": 8.256130839233001e-05,
+      "loss": 2.7703,
+      "step": 46405
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7235651016235352,
+      "learning_rate": 8.255070125899151e-05,
+      "loss": 3.0491,
+      "step": 46406
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6981080770492554,
+      "learning_rate": 8.254009469837264e-05,
+      "loss": 2.9714,
+      "step": 46407
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6944175362586975,
+      "learning_rate": 8.252948871050122e-05,
+      "loss": 2.8881,
+      "step": 46408
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7283467054367065,
+      "learning_rate": 8.251888329540536e-05,
+      "loss": 2.8444,
+      "step": 46409
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7321710586547852,
+      "learning_rate": 8.250827845311283e-05,
+      "loss": 3.0853,
+      "step": 46410
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.747199535369873,
+      "learning_rate": 8.249767418365162e-05,
+      "loss": 3.0333,
+      "step": 46411
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.693293571472168,
+      "learning_rate": 8.248707048704975e-05,
+      "loss": 2.949,
+      "step": 46412
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6886857151985168,
+      "learning_rate": 8.247646736333511e-05,
+      "loss": 3.0812,
+      "step": 46413
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7032137513160706,
+      "learning_rate": 8.24658648125356e-05,
+      "loss": 2.7775,
+      "step": 46414
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7268944382667542,
+      "learning_rate": 8.2455262834679e-05,
+      "loss": 2.8439,
+      "step": 46415
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6935317516326904,
+      "learning_rate": 8.244466142979341e-05,
+      "loss": 2.8731,
+      "step": 46416
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7060828804969788,
+      "learning_rate": 8.243406059790676e-05,
+      "loss": 2.8517,
+      "step": 46417
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.6857595443725586,
+      "learning_rate": 8.242346033904688e-05,
+      "loss": 2.9907,
+      "step": 46418
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7835713624954224,
+      "learning_rate": 8.24128606532418e-05,
+      "loss": 2.6536,
+      "step": 46419
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7131385803222656,
+      "learning_rate": 8.240226154051936e-05,
+      "loss": 2.8273,
+      "step": 46420
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7091208100318909,
+      "learning_rate": 8.23916630009074e-05,
+      "loss": 2.8953,
+      "step": 46421
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7581696510314941,
+      "learning_rate": 8.238106503443402e-05,
+      "loss": 2.8396,
+      "step": 46422
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7144935131072998,
+      "learning_rate": 8.237046764112694e-05,
+      "loss": 2.9896,
+      "step": 46423
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7232292890548706,
+      "learning_rate": 8.235987082101427e-05,
+      "loss": 2.5879,
+      "step": 46424
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7325698137283325,
+      "learning_rate": 8.234927457412371e-05,
+      "loss": 2.9548,
+      "step": 46425
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7658261656761169,
+      "learning_rate": 8.233867890048327e-05,
+      "loss": 2.8471,
+      "step": 46426
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7133269906044006,
+      "learning_rate": 8.232808380012095e-05,
+      "loss": 3.0446,
+      "step": 46427
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7498300075531006,
+      "learning_rate": 8.231748927306456e-05,
+      "loss": 2.9269,
+      "step": 46428
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7016955614089966,
+      "learning_rate": 8.230689531934203e-05,
+      "loss": 3.0732,
+      "step": 46429
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6857497692108154,
+      "learning_rate": 8.229630193898111e-05,
+      "loss": 2.9937,
+      "step": 46430
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7271170020103455,
+      "learning_rate": 8.228570913200987e-05,
+      "loss": 2.8245,
+      "step": 46431
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.720839262008667,
+      "learning_rate": 8.227511689845627e-05,
+      "loss": 3.0504,
+      "step": 46432
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.720252513885498,
+      "learning_rate": 8.2264525238348e-05,
+      "loss": 3.0355,
+      "step": 46433
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7447952032089233,
+      "learning_rate": 8.225393415171316e-05,
+      "loss": 3.0458,
+      "step": 46434
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.720052182674408,
+      "learning_rate": 8.224334363857954e-05,
+      "loss": 2.916,
+      "step": 46435
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7244362235069275,
+      "learning_rate": 8.223275369897497e-05,
+      "loss": 2.7844,
+      "step": 46436
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7059257626533508,
+      "learning_rate": 8.222216433292751e-05,
+      "loss": 3.0789,
+      "step": 46437
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6963223218917847,
+      "learning_rate": 8.221157554046486e-05,
+      "loss": 2.9243,
+      "step": 46438
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7097539305686951,
+      "learning_rate": 8.22009873216151e-05,
+      "loss": 2.9694,
+      "step": 46439
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.70842045545578,
+      "learning_rate": 8.219039967640594e-05,
+      "loss": 2.9949,
+      "step": 46440
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.730398952960968,
+      "learning_rate": 8.217981260486542e-05,
+      "loss": 2.9813,
+      "step": 46441
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7620899081230164,
+      "learning_rate": 8.216922610702136e-05,
+      "loss": 2.8141,
+      "step": 46442
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7119383811950684,
+      "learning_rate": 8.215864018290154e-05,
+      "loss": 2.92,
+      "step": 46443
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7197368144989014,
+      "learning_rate": 8.214805483253403e-05,
+      "loss": 2.9015,
+      "step": 46444
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7236267924308777,
+      "learning_rate": 8.213747005594655e-05,
+      "loss": 2.8312,
+      "step": 46445
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.743288516998291,
+      "learning_rate": 8.21268858531671e-05,
+      "loss": 3.0393,
+      "step": 46446
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6969639658927917,
+      "learning_rate": 8.211630222422343e-05,
+      "loss": 3.0242,
+      "step": 46447
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.9204838275909424,
+      "learning_rate": 8.210571916914348e-05,
+      "loss": 2.927,
+      "step": 46448
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7029668092727661,
+      "learning_rate": 8.209513668795523e-05,
+      "loss": 2.8216,
+      "step": 46449
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7012711763381958,
+      "learning_rate": 8.208455478068646e-05,
+      "loss": 2.8378,
+      "step": 46450
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7693458199501038,
+      "learning_rate": 8.207397344736502e-05,
+      "loss": 2.926,
+      "step": 46451
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6860461831092834,
+      "learning_rate": 8.20633926880187e-05,
+      "loss": 2.8357,
+      "step": 46452
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7062203288078308,
+      "learning_rate": 8.20528125026755e-05,
+      "loss": 3.0731,
+      "step": 46453
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7307056188583374,
+      "learning_rate": 8.20422328913633e-05,
+      "loss": 2.6463,
+      "step": 46454
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6875963807106018,
+      "learning_rate": 8.203165385410981e-05,
+      "loss": 2.8983,
+      "step": 46455
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8253306150436401,
+      "learning_rate": 8.202107539094311e-05,
+      "loss": 2.8164,
+      "step": 46456
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.703727662563324,
+      "learning_rate": 8.201049750189094e-05,
+      "loss": 2.9208,
+      "step": 46457
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7636400461196899,
+      "learning_rate": 8.19999201869811e-05,
+      "loss": 2.9006,
+      "step": 46458
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6992866396903992,
+      "learning_rate": 8.198934344624156e-05,
+      "loss": 3.0271,
+      "step": 46459
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7500672340393066,
+      "learning_rate": 8.197876727970006e-05,
+      "loss": 2.8297,
+      "step": 46460
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.682694137096405,
+      "learning_rate": 8.196819168738465e-05,
+      "loss": 2.9394,
+      "step": 46461
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6699866056442261,
+      "learning_rate": 8.195761666932291e-05,
+      "loss": 2.9927,
+      "step": 46462
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7636361718177795,
+      "learning_rate": 8.194704222554297e-05,
+      "loss": 2.6734,
+      "step": 46463
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8084840774536133,
+      "learning_rate": 8.193646835607256e-05,
+      "loss": 2.8364,
+      "step": 46464
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7584612965583801,
+      "learning_rate": 8.192589506093943e-05,
+      "loss": 2.8917,
+      "step": 46465
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8246785998344421,
+      "learning_rate": 8.191532234017162e-05,
+      "loss": 2.7962,
+      "step": 46466
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7840319871902466,
+      "learning_rate": 8.190475019379678e-05,
+      "loss": 2.8299,
+      "step": 46467
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7328280806541443,
+      "learning_rate": 8.189417862184287e-05,
+      "loss": 2.7687,
+      "step": 46468
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7843512892723083,
+      "learning_rate": 8.18836076243378e-05,
+      "loss": 3.0041,
+      "step": 46469
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7056668400764465,
+      "learning_rate": 8.187303720130922e-05,
+      "loss": 3.0747,
+      "step": 46470
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7091785669326782,
+      "learning_rate": 8.186246735278518e-05,
+      "loss": 2.9689,
+      "step": 46471
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7240222096443176,
+      "learning_rate": 8.185189807879345e-05,
+      "loss": 2.8462,
+      "step": 46472
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6857366561889648,
+      "learning_rate": 8.18413293793617e-05,
+      "loss": 3.0774,
+      "step": 46473
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7089368104934692,
+      "learning_rate": 8.183076125451803e-05,
+      "loss": 2.8356,
+      "step": 46474
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8050333261489868,
+      "learning_rate": 8.182019370429005e-05,
+      "loss": 3.0691,
+      "step": 46475
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.740153431892395,
+      "learning_rate": 8.180962672870577e-05,
+      "loss": 2.796,
+      "step": 46476
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7142683267593384,
+      "learning_rate": 8.179906032779283e-05,
+      "loss": 2.8724,
+      "step": 46477
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6894727349281311,
+      "learning_rate": 8.178849450157932e-05,
+      "loss": 2.6895,
+      "step": 46478
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7132574915885925,
+      "learning_rate": 8.177792925009286e-05,
+      "loss": 2.925,
+      "step": 46479
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7323726415634155,
+      "learning_rate": 8.176736457336129e-05,
+      "loss": 2.7728,
+      "step": 46480
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7239104509353638,
+      "learning_rate": 8.175680047141254e-05,
+      "loss": 2.7761,
+      "step": 46481
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8582285046577454,
+      "learning_rate": 8.174623694427433e-05,
+      "loss": 3.2315,
+      "step": 46482
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6943944692611694,
+      "learning_rate": 8.173567399197448e-05,
+      "loss": 2.8219,
+      "step": 46483
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7097330689430237,
+      "learning_rate": 8.172511161454095e-05,
+      "loss": 2.7027,
+      "step": 46484
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7035298347473145,
+      "learning_rate": 8.17145498120015e-05,
+      "loss": 2.9617,
+      "step": 46485
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7039844393730164,
+      "learning_rate": 8.17039885843839e-05,
+      "loss": 2.9092,
+      "step": 46486
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7377235889434814,
+      "learning_rate": 8.16934279317159e-05,
+      "loss": 2.912,
+      "step": 46487
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6866671442985535,
+      "learning_rate": 8.168286785402537e-05,
+      "loss": 2.9288,
+      "step": 46488
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7029548287391663,
+      "learning_rate": 8.167230835134027e-05,
+      "loss": 2.9683,
+      "step": 46489
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7336953282356262,
+      "learning_rate": 8.16617494236882e-05,
+      "loss": 2.921,
+      "step": 46490
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6844274997711182,
+      "learning_rate": 8.165119107109714e-05,
+      "loss": 2.8364,
+      "step": 46491
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7402933835983276,
+      "learning_rate": 8.164063329359473e-05,
+      "loss": 2.6783,
+      "step": 46492
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6989611983299255,
+      "learning_rate": 8.163007609120897e-05,
+      "loss": 3.0188,
+      "step": 46493
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7093426585197449,
+      "learning_rate": 8.161951946396756e-05,
+      "loss": 2.6546,
+      "step": 46494
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7372518181800842,
+      "learning_rate": 8.160896341189821e-05,
+      "loss": 2.7915,
+      "step": 46495
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7127835154533386,
+      "learning_rate": 8.159840793502891e-05,
+      "loss": 2.8474,
+      "step": 46496
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7472658157348633,
+      "learning_rate": 8.158785303338731e-05,
+      "loss": 2.912,
+      "step": 46497
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7016239166259766,
+      "learning_rate": 8.157729870700125e-05,
+      "loss": 2.835,
+      "step": 46498
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7203468680381775,
+      "learning_rate": 8.156674495589862e-05,
+      "loss": 2.7075,
+      "step": 46499
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6911539435386658,
+      "learning_rate": 8.155619178010717e-05,
+      "loss": 2.7875,
+      "step": 46500
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8295294046401978,
+      "learning_rate": 8.154563917965465e-05,
+      "loss": 2.7513,
+      "step": 46501
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.715740978717804,
+      "learning_rate": 8.15350871545688e-05,
+      "loss": 2.8735,
+      "step": 46502
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7473269104957581,
+      "learning_rate": 8.152453570487749e-05,
+      "loss": 2.6743,
+      "step": 46503
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7766650319099426,
+      "learning_rate": 8.151398483060859e-05,
+      "loss": 2.8402,
+      "step": 46504
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6803699731826782,
+      "learning_rate": 8.150343453178971e-05,
+      "loss": 2.6261,
+      "step": 46505
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.9094350934028625,
+      "learning_rate": 8.149288480844884e-05,
+      "loss": 2.9338,
+      "step": 46506
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7277110815048218,
+      "learning_rate": 8.148233566061352e-05,
+      "loss": 3.097,
+      "step": 46507
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7063420414924622,
+      "learning_rate": 8.147178708831181e-05,
+      "loss": 2.9711,
+      "step": 46508
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7325363159179688,
+      "learning_rate": 8.146123909157137e-05,
+      "loss": 2.838,
+      "step": 46509
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6835830211639404,
+      "learning_rate": 8.145069167041982e-05,
+      "loss": 2.8725,
+      "step": 46510
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7027117609977722,
+      "learning_rate": 8.144014482488523e-05,
+      "loss": 2.8241,
+      "step": 46511
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7053887248039246,
+      "learning_rate": 8.142959855499511e-05,
+      "loss": 2.895,
+      "step": 46512
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6964938044548035,
+      "learning_rate": 8.141905286077736e-05,
+      "loss": 3.0054,
+      "step": 46513
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7192620038986206,
+      "learning_rate": 8.140850774225988e-05,
+      "loss": 2.8054,
+      "step": 46514
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7804059386253357,
+      "learning_rate": 8.139796319947032e-05,
+      "loss": 2.9574,
+      "step": 46515
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6990054845809937,
+      "learning_rate": 8.138741923243643e-05,
+      "loss": 2.7702,
+      "step": 46516
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7165387272834778,
+      "learning_rate": 8.137687584118594e-05,
+      "loss": 2.841,
+      "step": 46517
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7471354603767395,
+      "learning_rate": 8.136633302574678e-05,
+      "loss": 2.9077,
+      "step": 46518
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7257309556007385,
+      "learning_rate": 8.13557907861465e-05,
+      "loss": 3.0477,
+      "step": 46519
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7215765118598938,
+      "learning_rate": 8.1345249122413e-05,
+      "loss": 2.9011,
+      "step": 46520
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6833771467208862,
+      "learning_rate": 8.133470803457413e-05,
+      "loss": 2.7987,
+      "step": 46521
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7473302483558655,
+      "learning_rate": 8.132416752265754e-05,
+      "loss": 2.707,
+      "step": 46522
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7858054637908936,
+      "learning_rate": 8.131362758669103e-05,
+      "loss": 3.0084,
+      "step": 46523
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7041542530059814,
+      "learning_rate": 8.130308822670223e-05,
+      "loss": 2.665,
+      "step": 46524
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.705814778804779,
+      "learning_rate": 8.1292549442719e-05,
+      "loss": 3.0473,
+      "step": 46525
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7002612352371216,
+      "learning_rate": 8.128201123476923e-05,
+      "loss": 2.7564,
+      "step": 46526
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.746374249458313,
+      "learning_rate": 8.127147360288043e-05,
+      "loss": 2.9071,
+      "step": 46527
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7429247498512268,
+      "learning_rate": 8.126093654708056e-05,
+      "loss": 2.9663,
+      "step": 46528
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7674092650413513,
+      "learning_rate": 8.125040006739722e-05,
+      "loss": 3.1434,
+      "step": 46529
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6966801285743713,
+      "learning_rate": 8.123986416385831e-05,
+      "loss": 2.9157,
+      "step": 46530
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7209636569023132,
+      "learning_rate": 8.122932883649147e-05,
+      "loss": 3.0265,
+      "step": 46531
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7358286380767822,
+      "learning_rate": 8.121879408532439e-05,
+      "loss": 3.1313,
+      "step": 46532
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7346293926239014,
+      "learning_rate": 8.120825991038502e-05,
+      "loss": 2.9752,
+      "step": 46533
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7059187889099121,
+      "learning_rate": 8.119772631170088e-05,
+      "loss": 2.8991,
+      "step": 46534
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6938578486442566,
+      "learning_rate": 8.118719328929983e-05,
+      "loss": 2.9862,
+      "step": 46535
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7114361524581909,
+      "learning_rate": 8.117666084320967e-05,
+      "loss": 2.7991,
+      "step": 46536
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7124288082122803,
+      "learning_rate": 8.116612897345807e-05,
+      "loss": 3.0468,
+      "step": 46537
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7127500772476196,
+      "learning_rate": 8.115559768007278e-05,
+      "loss": 2.8632,
+      "step": 46538
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6895161867141724,
+      "learning_rate": 8.114506696308147e-05,
+      "loss": 2.7845,
+      "step": 46539
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7293110489845276,
+      "learning_rate": 8.113453682251188e-05,
+      "loss": 2.8978,
+      "step": 46540
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7473793029785156,
+      "learning_rate": 8.112400725839193e-05,
+      "loss": 3.1251,
+      "step": 46541
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.710602343082428,
+      "learning_rate": 8.111347827074912e-05,
+      "loss": 2.9912,
+      "step": 46542
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6810015439987183,
+      "learning_rate": 8.110294985961136e-05,
+      "loss": 2.8958,
+      "step": 46543
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6719366312026978,
+      "learning_rate": 8.10924220250063e-05,
+      "loss": 3.0995,
+      "step": 46544
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6884929537773132,
+      "learning_rate": 8.108189476696162e-05,
+      "loss": 3.2298,
+      "step": 46545
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7371322512626648,
+      "learning_rate": 8.107136808550513e-05,
+      "loss": 2.9997,
+      "step": 46546
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7723472118377686,
+      "learning_rate": 8.106084198066448e-05,
+      "loss": 2.9791,
+      "step": 46547
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7253080606460571,
+      "learning_rate": 8.105031645246752e-05,
+      "loss": 2.6226,
+      "step": 46548
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7322795987129211,
+      "learning_rate": 8.103979150094178e-05,
+      "loss": 2.8257,
+      "step": 46549
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7000114321708679,
+      "learning_rate": 8.102926712611508e-05,
+      "loss": 3.0231,
+      "step": 46550
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7282766699790955,
+      "learning_rate": 8.101874332801526e-05,
+      "loss": 3.0256,
+      "step": 46551
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.702883243560791,
+      "learning_rate": 8.100822010666995e-05,
+      "loss": 2.8617,
+      "step": 46552
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7281079292297363,
+      "learning_rate": 8.099769746210681e-05,
+      "loss": 3.0817,
+      "step": 46553
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7449455261230469,
+      "learning_rate": 8.098717539435349e-05,
+      "loss": 2.7433,
+      "step": 46554
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7251495122909546,
+      "learning_rate": 8.09766539034378e-05,
+      "loss": 2.6756,
+      "step": 46555
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7186842560768127,
+      "learning_rate": 8.096613298938758e-05,
+      "loss": 2.9822,
+      "step": 46556
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7262799739837646,
+      "learning_rate": 8.095561265223028e-05,
+      "loss": 2.6999,
+      "step": 46557
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7726730108261108,
+      "learning_rate": 8.094509289199384e-05,
+      "loss": 2.7446,
+      "step": 46558
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7000957727432251,
+      "learning_rate": 8.093457370870588e-05,
+      "loss": 2.8547,
+      "step": 46559
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7143601179122925,
+      "learning_rate": 8.092405510239402e-05,
+      "loss": 2.8717,
+      "step": 46560
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6854216456413269,
+      "learning_rate": 8.09135370730861e-05,
+      "loss": 2.9581,
+      "step": 46561
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.745839536190033,
+      "learning_rate": 8.090301962080968e-05,
+      "loss": 2.9961,
+      "step": 46562
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7243121862411499,
+      "learning_rate": 8.089250274559262e-05,
+      "loss": 2.821,
+      "step": 46563
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7288719415664673,
+      "learning_rate": 8.088198644746244e-05,
+      "loss": 2.6618,
+      "step": 46564
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7389212250709534,
+      "learning_rate": 8.087147072644704e-05,
+      "loss": 2.9932,
+      "step": 46565
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7060454487800598,
+      "learning_rate": 8.086095558257402e-05,
+      "loss": 2.9091,
+      "step": 46566
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7329028248786926,
+      "learning_rate": 8.085044101587097e-05,
+      "loss": 2.9153,
+      "step": 46567
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7609493136405945,
+      "learning_rate": 8.083992702636578e-05,
+      "loss": 3.0979,
+      "step": 46568
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7355934381484985,
+      "learning_rate": 8.082941361408593e-05,
+      "loss": 2.9956,
+      "step": 46569
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6936449408531189,
+      "learning_rate": 8.081890077905925e-05,
+      "loss": 3.0962,
+      "step": 46570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7607849836349487,
+      "learning_rate": 8.08083885213135e-05,
+      "loss": 2.7487,
+      "step": 46571
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7081700563430786,
+      "learning_rate": 8.079787684087614e-05,
+      "loss": 2.7324,
+      "step": 46572
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7279213070869446,
+      "learning_rate": 8.07873657377751e-05,
+      "loss": 2.8963,
+      "step": 46573
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7351059317588806,
+      "learning_rate": 8.077685521203795e-05,
+      "loss": 2.8639,
+      "step": 46574
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7259153127670288,
+      "learning_rate": 8.07663452636923e-05,
+      "loss": 2.8075,
+      "step": 46575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7282455563545227,
+      "learning_rate": 8.0755835892766e-05,
+      "loss": 3.0483,
+      "step": 46576
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6987950801849365,
+      "learning_rate": 8.074532709928651e-05,
+      "loss": 2.9277,
+      "step": 46577
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7334350347518921,
+      "learning_rate": 8.073481888328174e-05,
+      "loss": 2.9942,
+      "step": 46578
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7089470624923706,
+      "learning_rate": 8.072431124477917e-05,
+      "loss": 2.9821,
+      "step": 46579
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7162338495254517,
+      "learning_rate": 8.071380418380663e-05,
+      "loss": 2.7657,
+      "step": 46580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7498306035995483,
+      "learning_rate": 8.070329770039175e-05,
+      "loss": 2.8501,
+      "step": 46581
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.735768735408783,
+      "learning_rate": 8.069279179456209e-05,
+      "loss": 2.9595,
+      "step": 46582
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7566409111022949,
+      "learning_rate": 8.068228646634549e-05,
+      "loss": 2.9093,
+      "step": 46583
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7465990781784058,
+      "learning_rate": 8.067178171576945e-05,
+      "loss": 2.9164,
+      "step": 46584
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7334000468254089,
+      "learning_rate": 8.066127754286173e-05,
+      "loss": 3.0872,
+      "step": 46585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8011319041252136,
+      "learning_rate": 8.06507739476501e-05,
+      "loss": 2.9013,
+      "step": 46586
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.729567289352417,
+      "learning_rate": 8.064027093016208e-05,
+      "loss": 2.8262,
+      "step": 46587
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7324111461639404,
+      "learning_rate": 8.062976849042539e-05,
+      "loss": 2.7336,
+      "step": 46588
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7233032584190369,
+      "learning_rate": 8.06192666284676e-05,
+      "loss": 2.4499,
+      "step": 46589
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7346892952919006,
+      "learning_rate": 8.060876534431642e-05,
+      "loss": 2.9368,
+      "step": 46590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7054365277290344,
+      "learning_rate": 8.059826463799961e-05,
+      "loss": 2.6252,
+      "step": 46591
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7693349719047546,
+      "learning_rate": 8.058776450954467e-05,
+      "loss": 2.9058,
+      "step": 46592
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7086447477340698,
+      "learning_rate": 8.057726495897941e-05,
+      "loss": 2.8654,
+      "step": 46593
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7039918303489685,
+      "learning_rate": 8.056676598633134e-05,
+      "loss": 2.985,
+      "step": 46594
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.722930371761322,
+      "learning_rate": 8.055626759162824e-05,
+      "loss": 2.9315,
+      "step": 46595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7084254026412964,
+      "learning_rate": 8.054576977489771e-05,
+      "loss": 2.7143,
+      "step": 46596
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7326437830924988,
+      "learning_rate": 8.053527253616731e-05,
+      "loss": 2.91,
+      "step": 46597
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.709077000617981,
+      "learning_rate": 8.052477587546486e-05,
+      "loss": 2.8769,
+      "step": 46598
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7226983308792114,
+      "learning_rate": 8.051427979281779e-05,
+      "loss": 2.8338,
+      "step": 46599
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7126448154449463,
+      "learning_rate": 8.050378428825398e-05,
+      "loss": 2.8205,
+      "step": 46600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6979407668113708,
+      "learning_rate": 8.049328936180087e-05,
+      "loss": 2.9157,
+      "step": 46601
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7206785678863525,
+      "learning_rate": 8.048279501348624e-05,
+      "loss": 2.8154,
+      "step": 46602
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.711652934551239,
+      "learning_rate": 8.047230124333771e-05,
+      "loss": 3.0502,
+      "step": 46603
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.0295687913894653,
+      "learning_rate": 8.04618080513828e-05,
+      "loss": 2.8277,
+      "step": 46604
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7558762431144714,
+      "learning_rate": 8.045131543764933e-05,
+      "loss": 2.846,
+      "step": 46605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7087622284889221,
+      "learning_rate": 8.044082340216474e-05,
+      "loss": 2.8821,
+      "step": 46606
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7025753259658813,
+      "learning_rate": 8.043033194495678e-05,
+      "loss": 3.1409,
+      "step": 46607
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6761602163314819,
+      "learning_rate": 8.041984106605316e-05,
+      "loss": 2.7138,
+      "step": 46608
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7491973042488098,
+      "learning_rate": 8.040935076548135e-05,
+      "loss": 2.8269,
+      "step": 46609
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7301766276359558,
+      "learning_rate": 8.039886104326921e-05,
+      "loss": 3.1797,
+      "step": 46610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6868095993995667,
+      "learning_rate": 8.0388371899444e-05,
+      "loss": 3.0877,
+      "step": 46611
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7199649214744568,
+      "learning_rate": 8.037788333403357e-05,
+      "loss": 3.0416,
+      "step": 46612
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6923564076423645,
+      "learning_rate": 8.036739534706563e-05,
+      "loss": 2.9795,
+      "step": 46613
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7084708213806152,
+      "learning_rate": 8.03569079385676e-05,
+      "loss": 2.8675,
+      "step": 46614
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7493816614151001,
+      "learning_rate": 8.03464211085673e-05,
+      "loss": 3.133,
+      "step": 46615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6935444474220276,
+      "learning_rate": 8.033593485709214e-05,
+      "loss": 2.8476,
+      "step": 46616
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7323094606399536,
+      "learning_rate": 8.032544918416996e-05,
+      "loss": 3.0348,
+      "step": 46617
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7361232042312622,
+      "learning_rate": 8.03149640898283e-05,
+      "loss": 2.812,
+      "step": 46618
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7444538474082947,
+      "learning_rate": 8.030447957409464e-05,
+      "loss": 2.874,
+      "step": 46619
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.718134343624115,
+      "learning_rate": 8.029399563699674e-05,
+      "loss": 2.8746,
+      "step": 46620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7226566076278687,
+      "learning_rate": 8.028351227856213e-05,
+      "loss": 3.0168,
+      "step": 46621
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.844632089138031,
+      "learning_rate": 8.027302949881845e-05,
+      "loss": 2.8895,
+      "step": 46622
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7833714485168457,
+      "learning_rate": 8.026254729779344e-05,
+      "loss": 2.9055,
+      "step": 46623
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7003939747810364,
+      "learning_rate": 8.025206567551455e-05,
+      "loss": 3.0071,
+      "step": 46624
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6972513794898987,
+      "learning_rate": 8.024158463200946e-05,
+      "loss": 2.9896,
+      "step": 46625
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7376452684402466,
+      "learning_rate": 8.023110416730564e-05,
+      "loss": 2.8573,
+      "step": 46626
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.720854640007019,
+      "learning_rate": 8.022062428143076e-05,
+      "loss": 3.1533,
+      "step": 46627
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7451248168945312,
+      "learning_rate": 8.02101449744126e-05,
+      "loss": 3.0563,
+      "step": 46628
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7125359773635864,
+      "learning_rate": 8.019966624627852e-05,
+      "loss": 2.8611,
+      "step": 46629
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7607391476631165,
+      "learning_rate": 8.018918809705628e-05,
+      "loss": 2.8573,
+      "step": 46630
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7929643988609314,
+      "learning_rate": 8.017871052677335e-05,
+      "loss": 2.8047,
+      "step": 46631
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7148974537849426,
+      "learning_rate": 8.016823353545749e-05,
+      "loss": 2.8755,
+      "step": 46632
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.8018413782119751,
+      "learning_rate": 8.015775712313616e-05,
+      "loss": 2.7893,
+      "step": 46633
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6894913911819458,
+      "learning_rate": 8.014728128983693e-05,
+      "loss": 2.7284,
+      "step": 46634
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7120516896247864,
+      "learning_rate": 8.013680603558754e-05,
+      "loss": 2.9376,
+      "step": 46635
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7336519360542297,
+      "learning_rate": 8.012633136041537e-05,
+      "loss": 2.8246,
+      "step": 46636
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7027100920677185,
+      "learning_rate": 8.011585726434813e-05,
+      "loss": 2.955,
+      "step": 46637
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7373955249786377,
+      "learning_rate": 8.010538374741355e-05,
+      "loss": 2.8527,
+      "step": 46638
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6577633619308472,
+      "learning_rate": 8.0094910809639e-05,
+      "loss": 2.8324,
+      "step": 46639
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7087864279747009,
+      "learning_rate": 8.008443845105216e-05,
+      "loss": 2.7932,
+      "step": 46640
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7115554809570312,
+      "learning_rate": 8.007396667168052e-05,
+      "loss": 2.8594,
+      "step": 46641
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7160695791244507,
+      "learning_rate": 8.006349547155171e-05,
+      "loss": 2.9263,
+      "step": 46642
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7285076379776001,
+      "learning_rate": 8.005302485069339e-05,
+      "loss": 2.9813,
+      "step": 46643
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7814637422561646,
+      "learning_rate": 8.004255480913302e-05,
+      "loss": 2.8582,
+      "step": 46644
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6893579959869385,
+      "learning_rate": 8.00320853468983e-05,
+      "loss": 2.6703,
+      "step": 46645
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.8030416965484619,
+      "learning_rate": 8.002161646401671e-05,
+      "loss": 2.9652,
+      "step": 46646
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7509509325027466,
+      "learning_rate": 8.00111481605158e-05,
+      "loss": 2.639,
+      "step": 46647
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6878439784049988,
+      "learning_rate": 8.000068043642325e-05,
+      "loss": 3.2898,
+      "step": 46648
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6887226104736328,
+      "learning_rate": 7.999021329176649e-05,
+      "loss": 2.7212,
+      "step": 46649
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7231206297874451,
+      "learning_rate": 7.997974672657325e-05,
+      "loss": 2.9418,
+      "step": 46650
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7704378366470337,
+      "learning_rate": 7.996928074087092e-05,
+      "loss": 2.7953,
+      "step": 46651
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6910778284072876,
+      "learning_rate": 7.995881533468717e-05,
+      "loss": 3.2411,
+      "step": 46652
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6970378160476685,
+      "learning_rate": 7.994835050804966e-05,
+      "loss": 2.6902,
+      "step": 46653
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6877171993255615,
+      "learning_rate": 7.993788626098583e-05,
+      "loss": 2.701,
+      "step": 46654
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7191632390022278,
+      "learning_rate": 7.992742259352324e-05,
+      "loss": 2.7879,
+      "step": 46655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7293527722358704,
+      "learning_rate": 7.991695950568939e-05,
+      "loss": 2.9878,
+      "step": 46656
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6775935292243958,
+      "learning_rate": 7.99064969975119e-05,
+      "loss": 2.8923,
+      "step": 46657
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7595655918121338,
+      "learning_rate": 7.989603506901842e-05,
+      "loss": 2.995,
+      "step": 46658
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6996396780014038,
+      "learning_rate": 7.988557372023637e-05,
+      "loss": 2.8157,
+      "step": 46659
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7048630714416504,
+      "learning_rate": 7.987511295119342e-05,
+      "loss": 2.8022,
+      "step": 46660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7288792133331299,
+      "learning_rate": 7.986465276191708e-05,
+      "loss": 2.8384,
+      "step": 46661
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6870715618133545,
+      "learning_rate": 7.985419315243476e-05,
+      "loss": 3.0466,
+      "step": 46662
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7056323289871216,
+      "learning_rate": 7.984373412277422e-05,
+      "loss": 2.8514,
+      "step": 46663
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6987048387527466,
+      "learning_rate": 7.983327567296284e-05,
+      "loss": 3.0196,
+      "step": 46664
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7563616037368774,
+      "learning_rate": 7.982281780302834e-05,
+      "loss": 2.9289,
+      "step": 46665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.684380054473877,
+      "learning_rate": 7.981236051299806e-05,
+      "loss": 2.8782,
+      "step": 46666
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.715981662273407,
+      "learning_rate": 7.980190380289973e-05,
+      "loss": 2.9129,
+      "step": 46667
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7292078733444214,
+      "learning_rate": 7.97914476727608e-05,
+      "loss": 2.8948,
+      "step": 46668
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.791476845741272,
+      "learning_rate": 7.978099212260877e-05,
+      "loss": 2.6953,
+      "step": 46669
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7197558879852295,
+      "learning_rate": 7.977053715247127e-05,
+      "loss": 2.9594,
+      "step": 46670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7770285606384277,
+      "learning_rate": 7.976008276237572e-05,
+      "loss": 2.7839,
+      "step": 46671
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6870531439781189,
+      "learning_rate": 7.97496289523498e-05,
+      "loss": 2.8836,
+      "step": 46672
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7081472277641296,
+      "learning_rate": 7.973917572242091e-05,
+      "loss": 2.8042,
+      "step": 46673
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.702302098274231,
+      "learning_rate": 7.972872307261661e-05,
+      "loss": 2.8638,
+      "step": 46674
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7121188044548035,
+      "learning_rate": 7.971827100296456e-05,
+      "loss": 2.8277,
+      "step": 46675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7616391777992249,
+      "learning_rate": 7.970781951349218e-05,
+      "loss": 2.8267,
+      "step": 46676
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7238693237304688,
+      "learning_rate": 7.969736860422704e-05,
+      "loss": 3.0476,
+      "step": 46677
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7182580828666687,
+      "learning_rate": 7.968691827519653e-05,
+      "loss": 3.1638,
+      "step": 46678
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.708895742893219,
+      "learning_rate": 7.967646852642827e-05,
+      "loss": 3.0584,
+      "step": 46679
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7667748332023621,
+      "learning_rate": 7.96660193579499e-05,
+      "loss": 2.935,
+      "step": 46680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7279300093650818,
+      "learning_rate": 7.965557076978871e-05,
+      "loss": 2.8778,
+      "step": 46681
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6909590363502502,
+      "learning_rate": 7.964512276197244e-05,
+      "loss": 2.6265,
+      "step": 46682
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7029200792312622,
+      "learning_rate": 7.963467533452851e-05,
+      "loss": 2.8785,
+      "step": 46683
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6875868439674377,
+      "learning_rate": 7.962422848748434e-05,
+      "loss": 2.9,
+      "step": 46684
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7104182243347168,
+      "learning_rate": 7.961378222086765e-05,
+      "loss": 2.7144,
+      "step": 46685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7271749973297119,
+      "learning_rate": 7.960333653470575e-05,
+      "loss": 2.947,
+      "step": 46686
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7475045919418335,
+      "learning_rate": 7.959289142902633e-05,
+      "loss": 2.7681,
+      "step": 46687
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7369611263275146,
+      "learning_rate": 7.958244690385669e-05,
+      "loss": 3.0237,
+      "step": 46688
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7487080097198486,
+      "learning_rate": 7.95720029592246e-05,
+      "loss": 2.9306,
+      "step": 46689
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7130888104438782,
+      "learning_rate": 7.956155959515741e-05,
+      "loss": 2.7723,
+      "step": 46690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7896283268928528,
+      "learning_rate": 7.955111681168256e-05,
+      "loss": 2.7895,
+      "step": 46691
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.783572256565094,
+      "learning_rate": 7.954067460882776e-05,
+      "loss": 2.797,
+      "step": 46692
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7980788350105286,
+      "learning_rate": 7.953023298662029e-05,
+      "loss": 2.9751,
+      "step": 46693
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7563785314559937,
+      "learning_rate": 7.951979194508772e-05,
+      "loss": 2.9569,
+      "step": 46694
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7462531328201294,
+      "learning_rate": 7.950935148425772e-05,
+      "loss": 2.9621,
+      "step": 46695
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7288581132888794,
+      "learning_rate": 7.949891160415755e-05,
+      "loss": 2.8099,
+      "step": 46696
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7490408420562744,
+      "learning_rate": 7.948847230481491e-05,
+      "loss": 2.7926,
+      "step": 46697
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.726197361946106,
+      "learning_rate": 7.947803358625716e-05,
+      "loss": 2.9515,
+      "step": 46698
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7250626087188721,
+      "learning_rate": 7.946759544851178e-05,
+      "loss": 2.9783,
+      "step": 46699
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7216305732727051,
+      "learning_rate": 7.945715789160637e-05,
+      "loss": 2.7268,
+      "step": 46700
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7879276275634766,
+      "learning_rate": 7.944672091556829e-05,
+      "loss": 2.9204,
+      "step": 46701
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7333400845527649,
+      "learning_rate": 7.94362845204252e-05,
+      "loss": 2.8761,
+      "step": 46702
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7444015145301819,
+      "learning_rate": 7.942584870620439e-05,
+      "loss": 3.0149,
+      "step": 46703
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6867544651031494,
+      "learning_rate": 7.941541347293353e-05,
+      "loss": 2.848,
+      "step": 46704
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7290630340576172,
+      "learning_rate": 7.940497882064e-05,
+      "loss": 3.0744,
+      "step": 46705
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7640529870986938,
+      "learning_rate": 7.939454474935122e-05,
+      "loss": 3.1576,
+      "step": 46706
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7068827748298645,
+      "learning_rate": 7.938411125909487e-05,
+      "loss": 2.9971,
+      "step": 46707
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6828535199165344,
+      "learning_rate": 7.937367834989821e-05,
+      "loss": 2.9971,
+      "step": 46708
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6947302222251892,
+      "learning_rate": 7.93632460217888e-05,
+      "loss": 2.9175,
+      "step": 46709
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7065187096595764,
+      "learning_rate": 7.935281427479425e-05,
+      "loss": 3.0409,
+      "step": 46710
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7438076734542847,
+      "learning_rate": 7.934238310894183e-05,
+      "loss": 2.8061,
+      "step": 46711
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7637118101119995,
+      "learning_rate": 7.933195252425929e-05,
+      "loss": 2.8936,
+      "step": 46712
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7260427474975586,
+      "learning_rate": 7.932152252077374e-05,
+      "loss": 2.9526,
+      "step": 46713
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7055556774139404,
+      "learning_rate": 7.931109309851283e-05,
+      "loss": 2.802,
+      "step": 46714
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7367199659347534,
+      "learning_rate": 7.93006642575041e-05,
+      "loss": 3.0472,
+      "step": 46715
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6921574473381042,
+      "learning_rate": 7.929023599777484e-05,
+      "loss": 2.6497,
+      "step": 46716
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6976795792579651,
+      "learning_rate": 7.927980831935276e-05,
+      "loss": 2.7202,
+      "step": 46717
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7396517992019653,
+      "learning_rate": 7.926938122226508e-05,
+      "loss": 3.1099,
+      "step": 46718
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7003223299980164,
+      "learning_rate": 7.925895470653943e-05,
+      "loss": 2.8682,
+      "step": 46719
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7250968217849731,
+      "learning_rate": 7.924852877220324e-05,
+      "loss": 3.0162,
+      "step": 46720
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7336358428001404,
+      "learning_rate": 7.923810341928385e-05,
+      "loss": 2.7529,
+      "step": 46721
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6997026205062866,
+      "learning_rate": 7.92276786478089e-05,
+      "loss": 2.8156,
+      "step": 46722
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7014487981796265,
+      "learning_rate": 7.921725445780567e-05,
+      "loss": 3.0013,
+      "step": 46723
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6987916827201843,
+      "learning_rate": 7.920683084930166e-05,
+      "loss": 3.029,
+      "step": 46724
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7134968042373657,
+      "learning_rate": 7.919640782232451e-05,
+      "loss": 2.8053,
+      "step": 46725
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7232719659805298,
+      "learning_rate": 7.91859853769015e-05,
+      "loss": 3.2245,
+      "step": 46726
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.718622624874115,
+      "learning_rate": 7.917556351306011e-05,
+      "loss": 2.8381,
+      "step": 46727
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.708051860332489,
+      "learning_rate": 7.916514223082768e-05,
+      "loss": 2.9364,
+      "step": 46728
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7392677068710327,
+      "learning_rate": 7.91547215302318e-05,
+      "loss": 3.0071,
+      "step": 46729
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7383646368980408,
+      "learning_rate": 7.914430141129994e-05,
+      "loss": 2.6716,
+      "step": 46730
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6965837478637695,
+      "learning_rate": 7.913388187405943e-05,
+      "loss": 2.9672,
+      "step": 46731
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7411499619483948,
+      "learning_rate": 7.912346291853783e-05,
+      "loss": 2.9199,
+      "step": 46732
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7892962098121643,
+      "learning_rate": 7.911304454476244e-05,
+      "loss": 2.9208,
+      "step": 46733
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6945703029632568,
+      "learning_rate": 7.910262675276087e-05,
+      "loss": 2.8022,
+      "step": 46734
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7150418162345886,
+      "learning_rate": 7.909220954256046e-05,
+      "loss": 2.8295,
+      "step": 46735
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6972780823707581,
+      "learning_rate": 7.90817929141886e-05,
+      "loss": 2.9666,
+      "step": 46736
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7024328708648682,
+      "learning_rate": 7.907137686767281e-05,
+      "loss": 2.7697,
+      "step": 46737
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7076515555381775,
+      "learning_rate": 7.906096140304046e-05,
+      "loss": 2.7878,
+      "step": 46738
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7264235615730286,
+      "learning_rate": 7.905054652031897e-05,
+      "loss": 2.7891,
+      "step": 46739
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7383031845092773,
+      "learning_rate": 7.904013221953594e-05,
+      "loss": 3.013,
+      "step": 46740
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7196172475814819,
+      "learning_rate": 7.90297185007187e-05,
+      "loss": 2.5622,
+      "step": 46741
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7266930341720581,
+      "learning_rate": 7.901930536389461e-05,
+      "loss": 3.0448,
+      "step": 46742
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6897668242454529,
+      "learning_rate": 7.900889280909106e-05,
+      "loss": 2.7908,
+      "step": 46743
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7381284832954407,
+      "learning_rate": 7.899848083633556e-05,
+      "loss": 2.889,
+      "step": 46744
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6909975409507751,
+      "learning_rate": 7.898806944565563e-05,
+      "loss": 3.0759,
+      "step": 46745
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.791771411895752,
+      "learning_rate": 7.897765863707848e-05,
+      "loss": 2.8085,
+      "step": 46746
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7384874224662781,
+      "learning_rate": 7.896724841063176e-05,
+      "loss": 2.9682,
+      "step": 46747
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7547340989112854,
+      "learning_rate": 7.895683876634274e-05,
+      "loss": 2.9999,
+      "step": 46748
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7127857208251953,
+      "learning_rate": 7.894642970423878e-05,
+      "loss": 2.5953,
+      "step": 46749
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7140387296676636,
+      "learning_rate": 7.893602122434748e-05,
+      "loss": 2.8312,
+      "step": 46750
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6861268877983093,
+      "learning_rate": 7.892561332669609e-05,
+      "loss": 2.9191,
+      "step": 46751
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.731544017791748,
+      "learning_rate": 7.891520601131213e-05,
+      "loss": 2.9862,
+      "step": 46752
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7233723402023315,
+      "learning_rate": 7.89047992782229e-05,
+      "loss": 3.1886,
+      "step": 46753
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7004035115242004,
+      "learning_rate": 7.889439312745598e-05,
+      "loss": 2.8099,
+      "step": 46754
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7136520743370056,
+      "learning_rate": 7.888398755903857e-05,
+      "loss": 2.9018,
+      "step": 46755
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7034738659858704,
+      "learning_rate": 7.88735825729983e-05,
+      "loss": 2.9447,
+      "step": 46756
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7353273630142212,
+      "learning_rate": 7.886317816936242e-05,
+      "loss": 2.9035,
+      "step": 46757
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7724609971046448,
+      "learning_rate": 7.885277434815828e-05,
+      "loss": 2.8167,
+      "step": 46758
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.9952621459960938,
+      "learning_rate": 7.884237110941348e-05,
+      "loss": 3.0053,
+      "step": 46759
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7721501588821411,
+      "learning_rate": 7.883196845315521e-05,
+      "loss": 2.9342,
+      "step": 46760
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7446574568748474,
+      "learning_rate": 7.882156637941098e-05,
+      "loss": 2.8327,
+      "step": 46761
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7871111631393433,
+      "learning_rate": 7.881116488820827e-05,
+      "loss": 2.8406,
+      "step": 46762
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7120261788368225,
+      "learning_rate": 7.880076397957438e-05,
+      "loss": 2.7964,
+      "step": 46763
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.8707037568092346,
+      "learning_rate": 7.879036365353672e-05,
+      "loss": 2.9292,
+      "step": 46764
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.699736475944519,
+      "learning_rate": 7.877996391012256e-05,
+      "loss": 2.8143,
+      "step": 46765
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.731170117855072,
+      "learning_rate": 7.876956474935941e-05,
+      "loss": 2.8974,
+      "step": 46766
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7486166954040527,
+      "learning_rate": 7.875916617127475e-05,
+      "loss": 3.0931,
+      "step": 46767
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6976475119590759,
+      "learning_rate": 7.874876817589572e-05,
+      "loss": 3.0062,
+      "step": 46768
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.76833176612854,
+      "learning_rate": 7.873837076325001e-05,
+      "loss": 2.9005,
+      "step": 46769
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7090509533882141,
+      "learning_rate": 7.872797393336482e-05,
+      "loss": 3.1187,
+      "step": 46770
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7460590600967407,
+      "learning_rate": 7.871757768626749e-05,
+      "loss": 2.9877,
+      "step": 46771
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7121726870536804,
+      "learning_rate": 7.870718202198555e-05,
+      "loss": 2.9237,
+      "step": 46772
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7376371026039124,
+      "learning_rate": 7.869678694054622e-05,
+      "loss": 2.9882,
+      "step": 46773
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7384995818138123,
+      "learning_rate": 7.868639244197704e-05,
+      "loss": 2.9219,
+      "step": 46774
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.8270297646522522,
+      "learning_rate": 7.867599852630523e-05,
+      "loss": 2.8947,
+      "step": 46775
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.765021026134491,
+      "learning_rate": 7.866560519355825e-05,
+      "loss": 2.8757,
+      "step": 46776
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6961572170257568,
+      "learning_rate": 7.865521244376353e-05,
+      "loss": 3.0606,
+      "step": 46777
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7141311764717102,
+      "learning_rate": 7.864482027694842e-05,
+      "loss": 2.6212,
+      "step": 46778
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6754100322723389,
+      "learning_rate": 7.863442869314024e-05,
+      "loss": 2.7227,
+      "step": 46779
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7049227952957153,
+      "learning_rate": 7.862403769236628e-05,
+      "loss": 3.0977,
+      "step": 46780
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.780094563961029,
+      "learning_rate": 7.861364727465399e-05,
+      "loss": 2.8449,
+      "step": 46781
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6903862357139587,
+      "learning_rate": 7.860325744003085e-05,
+      "loss": 3.1181,
+      "step": 46782
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6869261860847473,
+      "learning_rate": 7.859286818852401e-05,
+      "loss": 2.9011,
+      "step": 46783
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7356374859809875,
+      "learning_rate": 7.858247952016106e-05,
+      "loss": 3.0236,
+      "step": 46784
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6988884210586548,
+      "learning_rate": 7.857209143496925e-05,
+      "loss": 2.8257,
+      "step": 46785
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7464091777801514,
+      "learning_rate": 7.856170393297582e-05,
+      "loss": 2.5629,
+      "step": 46786
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7155094742774963,
+      "learning_rate": 7.855131701420836e-05,
+      "loss": 3.0864,
+      "step": 46787
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7148899435997009,
+      "learning_rate": 7.854093067869401e-05,
+      "loss": 2.9552,
+      "step": 46788
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7054211497306824,
+      "learning_rate": 7.853054492646033e-05,
+      "loss": 3.0602,
+      "step": 46789
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7065744996070862,
+      "learning_rate": 7.852015975753447e-05,
+      "loss": 2.9686,
+      "step": 46790
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7648019194602966,
+      "learning_rate": 7.850977517194397e-05,
+      "loss": 3.0381,
+      "step": 46791
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7037829756736755,
+      "learning_rate": 7.849939116971608e-05,
+      "loss": 3.0974,
+      "step": 46792
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7240031957626343,
+      "learning_rate": 7.84890077508781e-05,
+      "loss": 3.1276,
+      "step": 46793
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7773902416229248,
+      "learning_rate": 7.847862491545755e-05,
+      "loss": 2.8614,
+      "step": 46794
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.708374559879303,
+      "learning_rate": 7.846824266348153e-05,
+      "loss": 2.7868,
+      "step": 46795
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7300131916999817,
+      "learning_rate": 7.845786099497757e-05,
+      "loss": 2.8318,
+      "step": 46796
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7228390574455261,
+      "learning_rate": 7.844747990997305e-05,
+      "loss": 3.0077,
+      "step": 46797
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6896500587463379,
+      "learning_rate": 7.843709940849513e-05,
+      "loss": 2.915,
+      "step": 46798
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7400826811790466,
+      "learning_rate": 7.842671949057134e-05,
+      "loss": 3.0408,
+      "step": 46799
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7104637622833252,
+      "learning_rate": 7.841634015622891e-05,
+      "loss": 2.9581,
+      "step": 46800
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7338810563087463,
+      "learning_rate": 7.840596140549514e-05,
+      "loss": 3.0433,
+      "step": 46801
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.727415919303894,
+      "learning_rate": 7.83955832383975e-05,
+      "loss": 3.0346,
+      "step": 46802
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7342788577079773,
+      "learning_rate": 7.838520565496314e-05,
+      "loss": 2.829,
+      "step": 46803
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7318736910820007,
+      "learning_rate": 7.837482865521959e-05,
+      "loss": 3.0128,
+      "step": 46804
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7315552830696106,
+      "learning_rate": 7.836445223919401e-05,
+      "loss": 3.0684,
+      "step": 46805
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7073324918746948,
+      "learning_rate": 7.835407640691388e-05,
+      "loss": 2.7697,
+      "step": 46806
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.659433126449585,
+      "learning_rate": 7.834370115840647e-05,
+      "loss": 2.8127,
+      "step": 46807
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7416350245475769,
+      "learning_rate": 7.833332649369902e-05,
+      "loss": 2.9297,
+      "step": 46808
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6925668120384216,
+      "learning_rate": 7.832295241281898e-05,
+      "loss": 2.7095,
+      "step": 46809
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7669311761856079,
+      "learning_rate": 7.831257891579358e-05,
+      "loss": 2.9366,
+      "step": 46810
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7601991891860962,
+      "learning_rate": 7.830220600265014e-05,
+      "loss": 2.8255,
+      "step": 46811
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6947373747825623,
+      "learning_rate": 7.829183367341613e-05,
+      "loss": 2.9179,
+      "step": 46812
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7191194295883179,
+      "learning_rate": 7.828146192811877e-05,
+      "loss": 2.9629,
+      "step": 46813
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.753840982913971,
+      "learning_rate": 7.827109076678533e-05,
+      "loss": 2.7047,
+      "step": 46814
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7168121933937073,
+      "learning_rate": 7.826072018944312e-05,
+      "loss": 3.094,
+      "step": 46815
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.75309157371521,
+      "learning_rate": 7.825035019611948e-05,
+      "loss": 2.8804,
+      "step": 46816
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6898667216300964,
+      "learning_rate": 7.82399807868418e-05,
+      "loss": 3.1949,
+      "step": 46817
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7027104496955872,
+      "learning_rate": 7.82296119616373e-05,
+      "loss": 2.7947,
+      "step": 46818
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7549039125442505,
+      "learning_rate": 7.821924372053338e-05,
+      "loss": 3.0059,
+      "step": 46819
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7107914090156555,
+      "learning_rate": 7.82088760635572e-05,
+      "loss": 3.0119,
+      "step": 46820
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7330605387687683,
+      "learning_rate": 7.819850899073625e-05,
+      "loss": 3.0567,
+      "step": 46821
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7199442982673645,
+      "learning_rate": 7.818814250209775e-05,
+      "loss": 2.9784,
+      "step": 46822
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7049921751022339,
+      "learning_rate": 7.81777765976689e-05,
+      "loss": 3.0305,
+      "step": 46823
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7271491289138794,
+      "learning_rate": 7.816741127747716e-05,
+      "loss": 2.9861,
+      "step": 46824
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7731443643569946,
+      "learning_rate": 7.815704654154971e-05,
+      "loss": 2.9895,
+      "step": 46825
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7125991582870483,
+      "learning_rate": 7.81466823899139e-05,
+      "loss": 3.0968,
+      "step": 46826
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7187132239341736,
+      "learning_rate": 7.813631882259708e-05,
+      "loss": 2.8701,
+      "step": 46827
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.7413823008537292,
+      "learning_rate": 7.812595583962653e-05,
+      "loss": 2.9285,
+      "step": 46828
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6879110932350159,
+      "learning_rate": 7.811559344102953e-05,
+      "loss": 3.0651,
+      "step": 46829
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6981056332588196,
+      "learning_rate": 7.810523162683323e-05,
+      "loss": 2.7413,
+      "step": 46830
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7995371222496033,
+      "learning_rate": 7.809487039706517e-05,
+      "loss": 2.7898,
+      "step": 46831
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6955015063285828,
+      "learning_rate": 7.808450975175239e-05,
+      "loss": 2.8881,
+      "step": 46832
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7273989319801331,
+      "learning_rate": 7.807414969092233e-05,
+      "loss": 3.0188,
+      "step": 46833
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7289535403251648,
+      "learning_rate": 7.806379021460233e-05,
+      "loss": 2.9289,
+      "step": 46834
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7082308530807495,
+      "learning_rate": 7.805343132281951e-05,
+      "loss": 2.8786,
+      "step": 46835
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.736067533493042,
+      "learning_rate": 7.804307301560131e-05,
+      "loss": 3.1124,
+      "step": 46836
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.751034677028656,
+      "learning_rate": 7.803271529297496e-05,
+      "loss": 2.9974,
+      "step": 46837
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7050460577011108,
+      "learning_rate": 7.802235815496762e-05,
+      "loss": 2.7239,
+      "step": 46838
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7277799844741821,
+      "learning_rate": 7.801200160160676e-05,
+      "loss": 3.0167,
+      "step": 46839
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7088861465454102,
+      "learning_rate": 7.800164563291947e-05,
+      "loss": 2.8003,
+      "step": 46840
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6924530267715454,
+      "learning_rate": 7.79912902489332e-05,
+      "loss": 2.7845,
+      "step": 46841
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7400920391082764,
+      "learning_rate": 7.798093544967509e-05,
+      "loss": 3.0177,
+      "step": 46842
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7346185445785522,
+      "learning_rate": 7.797058123517253e-05,
+      "loss": 3.1688,
+      "step": 46843
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7445531487464905,
+      "learning_rate": 7.796022760545272e-05,
+      "loss": 2.8967,
+      "step": 46844
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7266702651977539,
+      "learning_rate": 7.794987456054288e-05,
+      "loss": 2.9966,
+      "step": 46845
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7435703277587891,
+      "learning_rate": 7.793952210047039e-05,
+      "loss": 2.9924,
+      "step": 46846
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7759212255477905,
+      "learning_rate": 7.792917022526239e-05,
+      "loss": 2.6906,
+      "step": 46847
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7034785747528076,
+      "learning_rate": 7.791881893494622e-05,
+      "loss": 3.0356,
+      "step": 46848
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7417486906051636,
+      "learning_rate": 7.790846822954922e-05,
+      "loss": 2.8477,
+      "step": 46849
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.73434978723526,
+      "learning_rate": 7.78981181090986e-05,
+      "loss": 2.7031,
+      "step": 46850
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7167014479637146,
+      "learning_rate": 7.788776857362157e-05,
+      "loss": 2.9902,
+      "step": 46851
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.717901885509491,
+      "learning_rate": 7.787741962314533e-05,
+      "loss": 3.0099,
+      "step": 46852
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7119569778442383,
+      "learning_rate": 7.78670712576972e-05,
+      "loss": 2.8717,
+      "step": 46853
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7578343152999878,
+      "learning_rate": 7.785672347730459e-05,
+      "loss": 2.9561,
+      "step": 46854
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.741181492805481,
+      "learning_rate": 7.784637628199448e-05,
+      "loss": 2.934,
+      "step": 46855
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7272793054580688,
+      "learning_rate": 7.783602967179438e-05,
+      "loss": 2.8121,
+      "step": 46856
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7096351981163025,
+      "learning_rate": 7.78256836467313e-05,
+      "loss": 3.0762,
+      "step": 46857
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.703946590423584,
+      "learning_rate": 7.781533820683272e-05,
+      "loss": 2.9331,
+      "step": 46858
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7070743441581726,
+      "learning_rate": 7.780499335212576e-05,
+      "loss": 2.8825,
+      "step": 46859
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7188315987586975,
+      "learning_rate": 7.779464908263762e-05,
+      "loss": 2.9131,
+      "step": 46860
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6855785846710205,
+      "learning_rate": 7.778430539839569e-05,
+      "loss": 2.9938,
+      "step": 46861
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7094498872756958,
+      "learning_rate": 7.777396229942705e-05,
+      "loss": 2.9944,
+      "step": 46862
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7395783066749573,
+      "learning_rate": 7.776361978575902e-05,
+      "loss": 2.9292,
+      "step": 46863
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.729633629322052,
+      "learning_rate": 7.775327785741894e-05,
+      "loss": 2.8902,
+      "step": 46864
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7409253716468811,
+      "learning_rate": 7.774293651443394e-05,
+      "loss": 3.0206,
+      "step": 46865
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7665625810623169,
+      "learning_rate": 7.773259575683131e-05,
+      "loss": 2.5951,
+      "step": 46866
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7642672657966614,
+      "learning_rate": 7.772225558463813e-05,
+      "loss": 2.8159,
+      "step": 46867
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7062388062477112,
+      "learning_rate": 7.771191599788174e-05,
+      "loss": 2.9574,
+      "step": 46868
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7327240705490112,
+      "learning_rate": 7.770157699658947e-05,
+      "loss": 2.9558,
+      "step": 46869
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7051011323928833,
+      "learning_rate": 7.76912385807884e-05,
+      "loss": 2.7563,
+      "step": 46870
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.721517026424408,
+      "learning_rate": 7.768090075050588e-05,
+      "loss": 2.9567,
+      "step": 46871
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7337429523468018,
+      "learning_rate": 7.767056350576911e-05,
+      "loss": 2.8228,
+      "step": 46872
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7145420908927917,
+      "learning_rate": 7.76602268466052e-05,
+      "loss": 2.8212,
+      "step": 46873
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.717994749546051,
+      "learning_rate": 7.764989077304153e-05,
+      "loss": 2.808,
+      "step": 46874
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6848964095115662,
+      "learning_rate": 7.763955528510516e-05,
+      "loss": 3.0177,
+      "step": 46875
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7062221765518188,
+      "learning_rate": 7.762922038282352e-05,
+      "loss": 2.721,
+      "step": 46876
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7325926423072815,
+      "learning_rate": 7.761888606622362e-05,
+      "loss": 2.8309,
+      "step": 46877
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7448124289512634,
+      "learning_rate": 7.760855233533279e-05,
+      "loss": 2.8649,
+      "step": 46878
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6935537457466125,
+      "learning_rate": 7.759821919017831e-05,
+      "loss": 2.824,
+      "step": 46879
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7377654314041138,
+      "learning_rate": 7.75878866307873e-05,
+      "loss": 2.9616,
+      "step": 46880
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7462741732597351,
+      "learning_rate": 7.757755465718703e-05,
+      "loss": 2.763,
+      "step": 46881
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7333121299743652,
+      "learning_rate": 7.756722326940455e-05,
+      "loss": 3.0852,
+      "step": 46882
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7244988083839417,
+      "learning_rate": 7.755689246746721e-05,
+      "loss": 2.975,
+      "step": 46883
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7320109605789185,
+      "learning_rate": 7.754656225140229e-05,
+      "loss": 2.8481,
+      "step": 46884
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6962860226631165,
+      "learning_rate": 7.753623262123683e-05,
+      "loss": 2.9392,
+      "step": 46885
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.761470377445221,
+      "learning_rate": 7.75259035769982e-05,
+      "loss": 3.0863,
+      "step": 46886
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6684213876724243,
+      "learning_rate": 7.751557511871356e-05,
+      "loss": 2.7991,
+      "step": 46887
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6986576318740845,
+      "learning_rate": 7.750524724640996e-05,
+      "loss": 2.8605,
+      "step": 46888
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7003403902053833,
+      "learning_rate": 7.74949199601148e-05,
+      "loss": 2.8998,
+      "step": 46889
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7219340205192566,
+      "learning_rate": 7.748459325985512e-05,
+      "loss": 2.9807,
+      "step": 46890
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7780990600585938,
+      "learning_rate": 7.747426714565828e-05,
+      "loss": 2.9845,
+      "step": 46891
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7375266551971436,
+      "learning_rate": 7.74639416175513e-05,
+      "loss": 2.8539,
+      "step": 46892
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.702833890914917,
+      "learning_rate": 7.745361667556159e-05,
+      "loss": 2.7213,
+      "step": 46893
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7389039397239685,
+      "learning_rate": 7.744329231971621e-05,
+      "loss": 3.0499,
+      "step": 46894
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7284427285194397,
+      "learning_rate": 7.743296855004224e-05,
+      "loss": 2.8602,
+      "step": 46895
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7277424335479736,
+      "learning_rate": 7.742264536656713e-05,
+      "loss": 2.7597,
+      "step": 46896
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7396055459976196,
+      "learning_rate": 7.741232276931785e-05,
+      "loss": 2.9429,
+      "step": 46897
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7248284220695496,
+      "learning_rate": 7.740200075832164e-05,
+      "loss": 2.803,
+      "step": 46898
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7068817019462585,
+      "learning_rate": 7.739167933360584e-05,
+      "loss": 2.8822,
+      "step": 46899
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.724687933921814,
+      "learning_rate": 7.73813584951974e-05,
+      "loss": 2.7615,
+      "step": 46900
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7012940645217896,
+      "learning_rate": 7.73710382431237e-05,
+      "loss": 3.0116,
+      "step": 46901
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.706241250038147,
+      "learning_rate": 7.736071857741188e-05,
+      "loss": 3.0091,
+      "step": 46902
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.8327338099479675,
+      "learning_rate": 7.735039949808897e-05,
+      "loss": 2.8047,
+      "step": 46903
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7665473818778992,
+      "learning_rate": 7.734008100518233e-05,
+      "loss": 2.9935,
+      "step": 46904
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6795613169670105,
+      "learning_rate": 7.732976309871901e-05,
+      "loss": 2.8637,
+      "step": 46905
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7590937614440918,
+      "learning_rate": 7.731944577872629e-05,
+      "loss": 2.9044,
+      "step": 46906
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6913984417915344,
+      "learning_rate": 7.730912904523123e-05,
+      "loss": 2.8735,
+      "step": 46907
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7234129309654236,
+      "learning_rate": 7.729881289826116e-05,
+      "loss": 2.7942,
+      "step": 46908
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7211498618125916,
+      "learning_rate": 7.728849733784313e-05,
+      "loss": 3.0061,
+      "step": 46909
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7252943515777588,
+      "learning_rate": 7.72781823640043e-05,
+      "loss": 2.9296,
+      "step": 46910
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7195742130279541,
+      "learning_rate": 7.726786797677191e-05,
+      "loss": 3.0563,
+      "step": 46911
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7221367955207825,
+      "learning_rate": 7.7257554176173e-05,
+      "loss": 3.0016,
+      "step": 46912
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7329986095428467,
+      "learning_rate": 7.724724096223496e-05,
+      "loss": 3.0884,
+      "step": 46913
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7828996181488037,
+      "learning_rate": 7.723692833498471e-05,
+      "loss": 2.7201,
+      "step": 46914
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7411604523658752,
+      "learning_rate": 7.722661629444959e-05,
+      "loss": 2.9872,
+      "step": 46915
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7232510447502136,
+      "learning_rate": 7.72163048406567e-05,
+      "loss": 2.959,
+      "step": 46916
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7410550713539124,
+      "learning_rate": 7.720599397363309e-05,
+      "loss": 2.7832,
+      "step": 46917
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6873605251312256,
+      "learning_rate": 7.719568369340611e-05,
+      "loss": 2.9664,
+      "step": 46918
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6807610392570496,
+      "learning_rate": 7.718537400000275e-05,
+      "loss": 2.8011,
+      "step": 46919
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7145941257476807,
+      "learning_rate": 7.71750648934502e-05,
+      "loss": 2.9894,
+      "step": 46920
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6965793371200562,
+      "learning_rate": 7.716475637377575e-05,
+      "loss": 3.0861,
+      "step": 46921
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7429308295249939,
+      "learning_rate": 7.715444844100637e-05,
+      "loss": 2.9534,
+      "step": 46922
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7118332386016846,
+      "learning_rate": 7.714414109516937e-05,
+      "loss": 3.0844,
+      "step": 46923
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7275873422622681,
+      "learning_rate": 7.71338343362918e-05,
+      "loss": 2.7865,
+      "step": 46924
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.779687762260437,
+      "learning_rate": 7.712352816440072e-05,
+      "loss": 3.0541,
+      "step": 46925
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7450523376464844,
+      "learning_rate": 7.711322257952349e-05,
+      "loss": 2.8133,
+      "step": 46926
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.725862979888916,
+      "learning_rate": 7.7102917581687e-05,
+      "loss": 2.8848,
+      "step": 46927
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.705909788608551,
+      "learning_rate": 7.709261317091869e-05,
+      "loss": 2.7166,
+      "step": 46928
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7430562376976013,
+      "learning_rate": 7.70823093472454e-05,
+      "loss": 2.7686,
+      "step": 46929
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.717370867729187,
+      "learning_rate": 7.70720061106945e-05,
+      "loss": 2.9506,
+      "step": 46930
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7363516688346863,
+      "learning_rate": 7.706170346129303e-05,
+      "loss": 2.9871,
+      "step": 46931
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7221070528030396,
+      "learning_rate": 7.705140139906806e-05,
+      "loss": 2.7575,
+      "step": 46932
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6913543939590454,
+      "learning_rate": 7.704109992404684e-05,
+      "loss": 3.0035,
+      "step": 46933
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7119285464286804,
+      "learning_rate": 7.703079903625639e-05,
+      "loss": 2.9111,
+      "step": 46934
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7075133323669434,
+      "learning_rate": 7.702049873572392e-05,
+      "loss": 2.7502,
+      "step": 46935
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.717917263507843,
+      "learning_rate": 7.701019902247661e-05,
+      "loss": 2.8735,
+      "step": 46936
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.735043466091156,
+      "learning_rate": 7.699989989654145e-05,
+      "loss": 2.7548,
+      "step": 46937
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7249195575714111,
+      "learning_rate": 7.698960135794581e-05,
+      "loss": 2.8316,
+      "step": 46938
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7049363255500793,
+      "learning_rate": 7.697930340671641e-05,
+      "loss": 3.0058,
+      "step": 46939
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7325100302696228,
+      "learning_rate": 7.696900604288066e-05,
+      "loss": 2.9678,
+      "step": 46940
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7083897590637207,
+      "learning_rate": 7.695870926646573e-05,
+      "loss": 2.813,
+      "step": 46941
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7134562730789185,
+      "learning_rate": 7.69484130774985e-05,
+      "loss": 3.0571,
+      "step": 46942
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7213225364685059,
+      "learning_rate": 7.693811747600633e-05,
+      "loss": 2.9821,
+      "step": 46943
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6950119137763977,
+      "learning_rate": 7.692782246201615e-05,
+      "loss": 2.9263,
+      "step": 46944
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6881087422370911,
+      "learning_rate": 7.691752803555524e-05,
+      "loss": 2.8272,
+      "step": 46945
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7488217353820801,
+      "learning_rate": 7.69072341966506e-05,
+      "loss": 2.9328,
+      "step": 46946
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7348080277442932,
+      "learning_rate": 7.68969409453293e-05,
+      "loss": 2.8666,
+      "step": 46947
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.738717257976532,
+      "learning_rate": 7.688664828161861e-05,
+      "loss": 2.9959,
+      "step": 46948
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7138335108757019,
+      "learning_rate": 7.687635620554549e-05,
+      "loss": 3.0209,
+      "step": 46949
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6965306997299194,
+      "learning_rate": 7.686606471713708e-05,
+      "loss": 2.7597,
+      "step": 46950
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.702285885810852,
+      "learning_rate": 7.685577381642059e-05,
+      "loss": 2.8853,
+      "step": 46951
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7388809323310852,
+      "learning_rate": 7.684548350342305e-05,
+      "loss": 2.8097,
+      "step": 46952
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7757939696311951,
+      "learning_rate": 7.683519377817156e-05,
+      "loss": 2.7593,
+      "step": 46953
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7448767423629761,
+      "learning_rate": 7.682490464069315e-05,
+      "loss": 2.9256,
+      "step": 46954
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7056995630264282,
+      "learning_rate": 7.6814616091015e-05,
+      "loss": 2.8403,
+      "step": 46955
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6833332180976868,
+      "learning_rate": 7.680432812916427e-05,
+      "loss": 2.839,
+      "step": 46956
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6968466639518738,
+      "learning_rate": 7.679404075516786e-05,
+      "loss": 2.8719,
+      "step": 46957
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7767067551612854,
+      "learning_rate": 7.678375396905314e-05,
+      "loss": 3.0657,
+      "step": 46958
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.690322995185852,
+      "learning_rate": 7.677346777084691e-05,
+      "loss": 2.9365,
+      "step": 46959
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7153850197792053,
+      "learning_rate": 7.676318216057654e-05,
+      "loss": 2.8756,
+      "step": 46960
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6876060962677002,
+      "learning_rate": 7.675289713826896e-05,
+      "loss": 2.8716,
+      "step": 46961
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6872727870941162,
+      "learning_rate": 7.674261270395119e-05,
+      "loss": 2.7886,
+      "step": 46962
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7221353054046631,
+      "learning_rate": 7.673232885765053e-05,
+      "loss": 2.7449,
+      "step": 46963
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7340434789657593,
+      "learning_rate": 7.672204559939381e-05,
+      "loss": 2.7857,
+      "step": 46964
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7158043384552002,
+      "learning_rate": 7.671176292920828e-05,
+      "loss": 2.9561,
+      "step": 46965
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.68252032995224,
+      "learning_rate": 7.67014808471211e-05,
+      "loss": 2.8707,
+      "step": 46966
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7790015339851379,
+      "learning_rate": 7.669119935315923e-05,
+      "loss": 2.9822,
+      "step": 46967
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7351877689361572,
+      "learning_rate": 7.668091844734975e-05,
+      "loss": 2.7752,
+      "step": 46968
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7354260683059692,
+      "learning_rate": 7.667063812971968e-05,
+      "loss": 2.8619,
+      "step": 46969
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7240676283836365,
+      "learning_rate": 7.666035840029615e-05,
+      "loss": 2.9733,
+      "step": 46970
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6971290111541748,
+      "learning_rate": 7.665007925910635e-05,
+      "loss": 2.8701,
+      "step": 46971
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7421733140945435,
+      "learning_rate": 7.663980070617721e-05,
+      "loss": 2.8046,
+      "step": 46972
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7057068347930908,
+      "learning_rate": 7.66295227415359e-05,
+      "loss": 2.9185,
+      "step": 46973
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.704230010509491,
+      "learning_rate": 7.661924536520943e-05,
+      "loss": 2.961,
+      "step": 46974
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7011248469352722,
+      "learning_rate": 7.66089685772248e-05,
+      "loss": 3.1366,
+      "step": 46975
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7493485808372498,
+      "learning_rate": 7.659869237760925e-05,
+      "loss": 2.9313,
+      "step": 46976
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7271580696105957,
+      "learning_rate": 7.658841676638963e-05,
+      "loss": 2.659,
+      "step": 46977
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7556811571121216,
+      "learning_rate": 7.657814174359323e-05,
+      "loss": 2.7887,
+      "step": 46978
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.786780595779419,
+      "learning_rate": 7.656786730924692e-05,
+      "loss": 2.8375,
+      "step": 46979
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7101176381111145,
+      "learning_rate": 7.655759346337788e-05,
+      "loss": 3.0459,
+      "step": 46980
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.753944456577301,
+      "learning_rate": 7.654732020601318e-05,
+      "loss": 3.0886,
+      "step": 46981
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6981664896011353,
+      "learning_rate": 7.653704753717983e-05,
+      "loss": 2.9098,
+      "step": 46982
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7053401470184326,
+      "learning_rate": 7.652677545690493e-05,
+      "loss": 2.8841,
+      "step": 46983
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7201017737388611,
+      "learning_rate": 7.65165039652154e-05,
+      "loss": 2.8831,
+      "step": 46984
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7298591732978821,
+      "learning_rate": 7.650623306213845e-05,
+      "loss": 2.8965,
+      "step": 46985
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7369044423103333,
+      "learning_rate": 7.649596274770098e-05,
+      "loss": 2.7928,
+      "step": 46986
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7387498021125793,
+      "learning_rate": 7.648569302193014e-05,
+      "loss": 3.0365,
+      "step": 46987
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7449101209640503,
+      "learning_rate": 7.647542388485308e-05,
+      "loss": 2.7906,
+      "step": 46988
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7396485805511475,
+      "learning_rate": 7.64651553364967e-05,
+      "loss": 2.9116,
+      "step": 46989
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6997805237770081,
+      "learning_rate": 7.64548873768881e-05,
+      "loss": 2.776,
+      "step": 46990
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6954943537712097,
+      "learning_rate": 7.644462000605423e-05,
+      "loss": 2.9855,
+      "step": 46991
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7475723028182983,
+      "learning_rate": 7.64343532240222e-05,
+      "loss": 3.1449,
+      "step": 46992
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7167398929595947,
+      "learning_rate": 7.642408703081911e-05,
+      "loss": 2.9414,
+      "step": 46993
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7235413789749146,
+      "learning_rate": 7.64138214264719e-05,
+      "loss": 3.0369,
+      "step": 46994
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7205052971839905,
+      "learning_rate": 7.640355641100773e-05,
+      "loss": 2.9086,
+      "step": 46995
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7066383957862854,
+      "learning_rate": 7.639329198445354e-05,
+      "loss": 3.0573,
+      "step": 46996
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7013208270072937,
+      "learning_rate": 7.638302814683632e-05,
+      "loss": 2.8161,
+      "step": 46997
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7191861271858215,
+      "learning_rate": 7.637276489818325e-05,
+      "loss": 2.9867,
+      "step": 46998
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7596316337585449,
+      "learning_rate": 7.63625022385212e-05,
+      "loss": 2.9672,
+      "step": 46999
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7377591729164124,
+      "learning_rate": 7.635224016787734e-05,
+      "loss": 3.1616,
+      "step": 47000
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6754733324050903,
+      "learning_rate": 7.634197868627858e-05,
+      "loss": 2.9541,
+      "step": 47001
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6994279026985168,
+      "learning_rate": 7.633171779375199e-05,
+      "loss": 2.7826,
+      "step": 47002
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7444949746131897,
+      "learning_rate": 7.632145749032468e-05,
+      "loss": 2.913,
+      "step": 47003
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7129213213920593,
+      "learning_rate": 7.631119777602365e-05,
+      "loss": 2.9534,
+      "step": 47004
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7305833101272583,
+      "learning_rate": 7.630093865087584e-05,
+      "loss": 2.8632,
+      "step": 47005
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7327616214752197,
+      "learning_rate": 7.629068011490822e-05,
+      "loss": 2.9634,
+      "step": 47006
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7405668497085571,
+      "learning_rate": 7.628042216814789e-05,
+      "loss": 3.079,
+      "step": 47007
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.721222996711731,
+      "learning_rate": 7.627016481062196e-05,
+      "loss": 3.0155,
+      "step": 47008
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7441813945770264,
+      "learning_rate": 7.625990804235729e-05,
+      "loss": 2.9115,
+      "step": 47009
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7018464803695679,
+      "learning_rate": 7.624965186338106e-05,
+      "loss": 2.7319,
+      "step": 47010
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7066758275032043,
+      "learning_rate": 7.623939627372015e-05,
+      "loss": 2.9433,
+      "step": 47011
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7435897588729858,
+      "learning_rate": 7.622914127340156e-05,
+      "loss": 3.0037,
+      "step": 47012
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7094926238059998,
+      "learning_rate": 7.62188868624524e-05,
+      "loss": 2.9102,
+      "step": 47013
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7312250733375549,
+      "learning_rate": 7.620863304089955e-05,
+      "loss": 2.9911,
+      "step": 47014
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7396236658096313,
+      "learning_rate": 7.619837980877019e-05,
+      "loss": 3.1673,
+      "step": 47015
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7736823558807373,
+      "learning_rate": 7.61881271660911e-05,
+      "loss": 2.6794,
+      "step": 47016
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7479914426803589,
+      "learning_rate": 7.617787511288953e-05,
+      "loss": 3.0262,
+      "step": 47017
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7253764271736145,
+      "learning_rate": 7.616762364919236e-05,
+      "loss": 2.7085,
+      "step": 47018
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6933542490005493,
+      "learning_rate": 7.615737277502649e-05,
+      "loss": 3.026,
+      "step": 47019
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7153028845787048,
+      "learning_rate": 7.614712249041912e-05,
+      "loss": 3.0809,
+      "step": 47020
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7277600765228271,
+      "learning_rate": 7.613687279539706e-05,
+      "loss": 3.0096,
+      "step": 47021
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7183275818824768,
+      "learning_rate": 7.612662368998736e-05,
+      "loss": 2.9682,
+      "step": 47022
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7380486726760864,
+      "learning_rate": 7.611637517421717e-05,
+      "loss": 2.8966,
+      "step": 47023
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7355697154998779,
+      "learning_rate": 7.61061272481133e-05,
+      "loss": 3.1274,
+      "step": 47024
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6961773633956909,
+      "learning_rate": 7.609587991170284e-05,
+      "loss": 2.9941,
+      "step": 47025
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7682431936264038,
+      "learning_rate": 7.608563316501276e-05,
+      "loss": 2.8109,
+      "step": 47026
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7720531225204468,
+      "learning_rate": 7.607538700806994e-05,
+      "loss": 2.9825,
+      "step": 47027
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.725392758846283,
+      "learning_rate": 7.606514144090154e-05,
+      "loss": 3.0161,
+      "step": 47028
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7036164999008179,
+      "learning_rate": 7.605489646353437e-05,
+      "loss": 2.7043,
+      "step": 47029
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.734400749206543,
+      "learning_rate": 7.604465207599561e-05,
+      "loss": 3.0003,
+      "step": 47030
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7298409342765808,
+      "learning_rate": 7.603440827831203e-05,
+      "loss": 2.6084,
+      "step": 47031
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7258648872375488,
+      "learning_rate": 7.60241650705108e-05,
+      "loss": 2.769,
+      "step": 47032
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.676993191242218,
+      "learning_rate": 7.601392245261883e-05,
+      "loss": 2.8381,
+      "step": 47033
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7960457801818848,
+      "learning_rate": 7.600368042466298e-05,
+      "loss": 2.8848,
+      "step": 47034
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7140025496482849,
+      "learning_rate": 7.599343898667042e-05,
+      "loss": 2.9983,
+      "step": 47035
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.707834005355835,
+      "learning_rate": 7.598319813866794e-05,
+      "loss": 2.9884,
+      "step": 47036
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.8292667269706726,
+      "learning_rate": 7.59729578806826e-05,
+      "loss": 2.8562,
+      "step": 47037
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6989453434944153,
+      "learning_rate": 7.596271821274148e-05,
+      "loss": 2.7451,
+      "step": 47038
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7435256838798523,
+      "learning_rate": 7.595247913487135e-05,
+      "loss": 2.8466,
+      "step": 47039
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7675233483314514,
+      "learning_rate": 7.594224064709946e-05,
+      "loss": 2.978,
+      "step": 47040
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7073809504508972,
+      "learning_rate": 7.593200274945238e-05,
+      "loss": 2.8363,
+      "step": 47041
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7317488789558411,
+      "learning_rate": 7.592176544195729e-05,
+      "loss": 2.9328,
+      "step": 47042
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7460283041000366,
+      "learning_rate": 7.591152872464123e-05,
+      "loss": 2.8801,
+      "step": 47043
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7187976241111755,
+      "learning_rate": 7.5901292597531e-05,
+      "loss": 2.9518,
+      "step": 47044
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7074181437492371,
+      "learning_rate": 7.589105706065371e-05,
+      "loss": 2.8221,
+      "step": 47045
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6958547830581665,
+      "learning_rate": 7.588082211403617e-05,
+      "loss": 2.9221,
+      "step": 47046
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7111318707466125,
+      "learning_rate": 7.58705877577055e-05,
+      "loss": 2.8711,
+      "step": 47047
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6999523639678955,
+      "learning_rate": 7.586035399168856e-05,
+      "loss": 2.839,
+      "step": 47048
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.726325511932373,
+      "learning_rate": 7.585012081601222e-05,
+      "loss": 2.83,
+      "step": 47049
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7136778235435486,
+      "learning_rate": 7.583988823070362e-05,
+      "loss": 2.9703,
+      "step": 47050
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7415741086006165,
+      "learning_rate": 7.582965623578954e-05,
+      "loss": 2.8102,
+      "step": 47051
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.83095782995224,
+      "learning_rate": 7.581942483129697e-05,
+      "loss": 2.9474,
+      "step": 47052
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7691175937652588,
+      "learning_rate": 7.580919401725303e-05,
+      "loss": 2.9403,
+      "step": 47053
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7467166185379028,
+      "learning_rate": 7.579896379368449e-05,
+      "loss": 3.0011,
+      "step": 47054
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6988757252693176,
+      "learning_rate": 7.578873416061833e-05,
+      "loss": 3.056,
+      "step": 47055
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.666217029094696,
+      "learning_rate": 7.577850511808139e-05,
+      "loss": 2.89,
+      "step": 47056
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7455030083656311,
+      "learning_rate": 7.576827666610075e-05,
+      "loss": 2.9968,
+      "step": 47057
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.742010235786438,
+      "learning_rate": 7.575804880470338e-05,
+      "loss": 3.046,
+      "step": 47058
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7573196291923523,
+      "learning_rate": 7.574782153391608e-05,
+      "loss": 2.8716,
+      "step": 47059
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.725332498550415,
+      "learning_rate": 7.573759485376593e-05,
+      "loss": 2.8598,
+      "step": 47060
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7262852787971497,
+      "learning_rate": 7.572736876427973e-05,
+      "loss": 2.8723,
+      "step": 47061
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6897485852241516,
+      "learning_rate": 7.571714326548454e-05,
+      "loss": 2.8583,
+      "step": 47062
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6986570954322815,
+      "learning_rate": 7.570691835740725e-05,
+      "loss": 2.6931,
+      "step": 47063
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7163300514221191,
+      "learning_rate": 7.569669404007471e-05,
+      "loss": 3.034,
+      "step": 47064
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7340909838676453,
+      "learning_rate": 7.568647031351396e-05,
+      "loss": 2.7613,
+      "step": 47065
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.75508713722229,
+      "learning_rate": 7.567624717775179e-05,
+      "loss": 3.004,
+      "step": 47066
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6939140558242798,
+      "learning_rate": 7.566602463281528e-05,
+      "loss": 2.7456,
+      "step": 47067
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7123984694480896,
+      "learning_rate": 7.565580267873124e-05,
+      "loss": 2.9987,
+      "step": 47068
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7077834010124207,
+      "learning_rate": 7.56455813155267e-05,
+      "loss": 2.9786,
+      "step": 47069
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7424761652946472,
+      "learning_rate": 7.563536054322856e-05,
+      "loss": 2.8784,
+      "step": 47070
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7440283894538879,
+      "learning_rate": 7.562514036186359e-05,
+      "loss": 2.7986,
+      "step": 47071
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7975735068321228,
+      "learning_rate": 7.56149207714589e-05,
+      "loss": 2.9229,
+      "step": 47072
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6961278915405273,
+      "learning_rate": 7.560470177204124e-05,
+      "loss": 2.8992,
+      "step": 47073
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6985488533973694,
+      "learning_rate": 7.55944833636376e-05,
+      "loss": 2.6921,
+      "step": 47074
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7156649827957153,
+      "learning_rate": 7.558426554627499e-05,
+      "loss": 2.8174,
+      "step": 47075
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.776537299156189,
+      "learning_rate": 7.557404831998028e-05,
+      "loss": 2.9881,
+      "step": 47076
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7213156819343567,
+      "learning_rate": 7.55638316847803e-05,
+      "loss": 2.8593,
+      "step": 47077
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7126505374908447,
+      "learning_rate": 7.55536156407019e-05,
+      "loss": 2.7945,
+      "step": 47078
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7194722890853882,
+      "learning_rate": 7.554340018777207e-05,
+      "loss": 2.8476,
+      "step": 47079
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.718202531337738,
+      "learning_rate": 7.553318532601785e-05,
+      "loss": 2.6935,
+      "step": 47080
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7163863778114319,
+      "learning_rate": 7.552297105546588e-05,
+      "loss": 3.0641,
+      "step": 47081
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7556397914886475,
+      "learning_rate": 7.551275737614335e-05,
+      "loss": 2.9546,
+      "step": 47082
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7361882925033569,
+      "learning_rate": 7.55025442880769e-05,
+      "loss": 2.9487,
+      "step": 47083
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7101148962974548,
+      "learning_rate": 7.549233179129362e-05,
+      "loss": 2.9951,
+      "step": 47084
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7225397229194641,
+      "learning_rate": 7.548211988582035e-05,
+      "loss": 3.103,
+      "step": 47085
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7092863917350769,
+      "learning_rate": 7.547190857168387e-05,
+      "loss": 2.7833,
+      "step": 47086
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6932755708694458,
+      "learning_rate": 7.546169784891127e-05,
+      "loss": 2.8817,
+      "step": 47087
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6869648098945618,
+      "learning_rate": 7.545148771752925e-05,
+      "loss": 2.8556,
+      "step": 47088
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7577425837516785,
+      "learning_rate": 7.54412781775648e-05,
+      "loss": 2.88,
+      "step": 47089
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7419918179512024,
+      "learning_rate": 7.543106922904488e-05,
+      "loss": 2.9787,
+      "step": 47090
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7026674747467041,
+      "learning_rate": 7.542086087199631e-05,
+      "loss": 2.8453,
+      "step": 47091
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7323837280273438,
+      "learning_rate": 7.541065310644597e-05,
+      "loss": 2.9609,
+      "step": 47092
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6933581233024597,
+      "learning_rate": 7.540044593242067e-05,
+      "loss": 3.0251,
+      "step": 47093
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7326129078865051,
+      "learning_rate": 7.539023934994734e-05,
+      "loss": 2.8222,
+      "step": 47094
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7666764259338379,
+      "learning_rate": 7.5380033359053e-05,
+      "loss": 2.9111,
+      "step": 47095
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7760982513427734,
+      "learning_rate": 7.536982795976436e-05,
+      "loss": 2.896,
+      "step": 47096
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6980563998222351,
+      "learning_rate": 7.535962315210843e-05,
+      "loss": 2.9076,
+      "step": 47097
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.730495810508728,
+      "learning_rate": 7.534941893611201e-05,
+      "loss": 3.0254,
+      "step": 47098
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6933460235595703,
+      "learning_rate": 7.533921531180191e-05,
+      "loss": 2.8862,
+      "step": 47099
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7039830088615417,
+      "learning_rate": 7.532901227920517e-05,
+      "loss": 2.9391,
+      "step": 47100
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7359967827796936,
+      "learning_rate": 7.531880983834845e-05,
+      "loss": 2.8176,
+      "step": 47101
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6860753893852234,
+      "learning_rate": 7.530860798925888e-05,
+      "loss": 2.8287,
+      "step": 47102
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7534465789794922,
+      "learning_rate": 7.529840673196309e-05,
+      "loss": 2.7622,
+      "step": 47103
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7103568315505981,
+      "learning_rate": 7.528820606648805e-05,
+      "loss": 3.1858,
+      "step": 47104
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6781308054924011,
+      "learning_rate": 7.527800599286073e-05,
+      "loss": 3.0361,
+      "step": 47105
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7126346230506897,
+      "learning_rate": 7.526780651110788e-05,
+      "loss": 2.8852,
+      "step": 47106
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7078046798706055,
+      "learning_rate": 7.525760762125636e-05,
+      "loss": 2.8481,
+      "step": 47107
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7195985317230225,
+      "learning_rate": 7.524740932333298e-05,
+      "loss": 2.8783,
+      "step": 47108
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7203580737113953,
+      "learning_rate": 7.523721161736468e-05,
+      "loss": 2.9165,
+      "step": 47109
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7243421673774719,
+      "learning_rate": 7.522701450337837e-05,
+      "loss": 2.8335,
+      "step": 47110
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7248387336730957,
+      "learning_rate": 7.521681798140076e-05,
+      "loss": 2.9192,
+      "step": 47111
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7272186875343323,
+      "learning_rate": 7.520662205145888e-05,
+      "loss": 2.9508,
+      "step": 47112
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6788175702095032,
+      "learning_rate": 7.519642671357951e-05,
+      "loss": 2.9344,
+      "step": 47113
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7410783767700195,
+      "learning_rate": 7.518623196778939e-05,
+      "loss": 2.9244,
+      "step": 47114
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7222927212715149,
+      "learning_rate": 7.517603781411558e-05,
+      "loss": 2.7248,
+      "step": 47115
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7629684805870056,
+      "learning_rate": 7.516584425258471e-05,
+      "loss": 3.0434,
+      "step": 47116
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.762963056564331,
+      "learning_rate": 7.515565128322382e-05,
+      "loss": 3.1039,
+      "step": 47117
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7189714908599854,
+      "learning_rate": 7.51454589060596e-05,
+      "loss": 2.8823,
+      "step": 47118
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7012630105018616,
+      "learning_rate": 7.513526712111904e-05,
+      "loss": 2.8067,
+      "step": 47119
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.734362006187439,
+      "learning_rate": 7.512507592842892e-05,
+      "loss": 2.7973,
+      "step": 47120
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7629452347755432,
+      "learning_rate": 7.511488532801598e-05,
+      "loss": 3.0645,
+      "step": 47121
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7741067409515381,
+      "learning_rate": 7.510469531990723e-05,
+      "loss": 2.7464,
+      "step": 47122
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7644525170326233,
+      "learning_rate": 7.509450590412936e-05,
+      "loss": 2.8866,
+      "step": 47123
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7327432036399841,
+      "learning_rate": 7.508431708070927e-05,
+      "loss": 2.83,
+      "step": 47124
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6900554299354553,
+      "learning_rate": 7.507412884967387e-05,
+      "loss": 2.9412,
+      "step": 47125
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.742673933506012,
+      "learning_rate": 7.506394121104988e-05,
+      "loss": 2.7973,
+      "step": 47126
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7525250315666199,
+      "learning_rate": 7.505375416486423e-05,
+      "loss": 2.7082,
+      "step": 47127
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6825203895568848,
+      "learning_rate": 7.50435677111437e-05,
+      "loss": 2.8795,
+      "step": 47128
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6943187117576599,
+      "learning_rate": 7.503338184991506e-05,
+      "loss": 3.0387,
+      "step": 47129
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.781815767288208,
+      "learning_rate": 7.502319658120526e-05,
+      "loss": 3.0341,
+      "step": 47130
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7488953471183777,
+      "learning_rate": 7.5013011905041e-05,
+      "loss": 2.9886,
+      "step": 47131
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7378432750701904,
+      "learning_rate": 7.500282782144923e-05,
+      "loss": 2.9618,
+      "step": 47132
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6865012049674988,
+      "learning_rate": 7.499264433045663e-05,
+      "loss": 2.6925,
+      "step": 47133
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7527883648872375,
+      "learning_rate": 7.498246143209022e-05,
+      "loss": 2.9324,
+      "step": 47134
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7522280216217041,
+      "learning_rate": 7.497227912637666e-05,
+      "loss": 3.007,
+      "step": 47135
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7576622366905212,
+      "learning_rate": 7.496209741334274e-05,
+      "loss": 2.9856,
+      "step": 47136
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7035020589828491,
+      "learning_rate": 7.49519162930154e-05,
+      "loss": 3.0091,
+      "step": 47137
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7081121206283569,
+      "learning_rate": 7.494173576542137e-05,
+      "loss": 2.7836,
+      "step": 47138
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7172082662582397,
+      "learning_rate": 7.493155583058747e-05,
+      "loss": 2.775,
+      "step": 47139
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7473821043968201,
+      "learning_rate": 7.492137648854062e-05,
+      "loss": 3.1286,
+      "step": 47140
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7130926847457886,
+      "learning_rate": 7.491119773930754e-05,
+      "loss": 2.9354,
+      "step": 47141
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7745637893676758,
+      "learning_rate": 7.490101958291504e-05,
+      "loss": 2.7904,
+      "step": 47142
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6880003213882446,
+      "learning_rate": 7.489084201938985e-05,
+      "loss": 2.8231,
+      "step": 47143
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7113720178604126,
+      "learning_rate": 7.488066504875897e-05,
+      "loss": 2.9742,
+      "step": 47144
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.710329532623291,
+      "learning_rate": 7.4870488671049e-05,
+      "loss": 3.0068,
+      "step": 47145
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7468894720077515,
+      "learning_rate": 7.486031288628682e-05,
+      "loss": 2.9647,
+      "step": 47146
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7187853455543518,
+      "learning_rate": 7.485013769449935e-05,
+      "loss": 2.9264,
+      "step": 47147
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6992723941802979,
+      "learning_rate": 7.483996309571319e-05,
+      "loss": 2.8219,
+      "step": 47148
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7436049580574036,
+      "learning_rate": 7.482978908995532e-05,
+      "loss": 2.9799,
+      "step": 47149
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7207313776016235,
+      "learning_rate": 7.481961567725246e-05,
+      "loss": 2.915,
+      "step": 47150
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7421787977218628,
+      "learning_rate": 7.480944285763127e-05,
+      "loss": 2.8734,
+      "step": 47151
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.73797607421875,
+      "learning_rate": 7.479927063111879e-05,
+      "loss": 2.9639,
+      "step": 47152
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7333713173866272,
+      "learning_rate": 7.478909899774161e-05,
+      "loss": 2.8489,
+      "step": 47153
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.717434823513031,
+      "learning_rate": 7.477892795752669e-05,
+      "loss": 2.7659,
+      "step": 47154
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.715705156326294,
+      "learning_rate": 7.476875751050063e-05,
+      "loss": 2.7922,
+      "step": 47155
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7075499892234802,
+      "learning_rate": 7.475858765669038e-05,
+      "loss": 3.0544,
+      "step": 47156
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7804867625236511,
+      "learning_rate": 7.474841839612269e-05,
+      "loss": 2.7703,
+      "step": 47157
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7036874890327454,
+      "learning_rate": 7.473824972882422e-05,
+      "loss": 2.846,
+      "step": 47158
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7310692071914673,
+      "learning_rate": 7.472808165482195e-05,
+      "loss": 2.9657,
+      "step": 47159
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7222773432731628,
+      "learning_rate": 7.471791417414246e-05,
+      "loss": 3.0682,
+      "step": 47160
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6930948495864868,
+      "learning_rate": 7.47077472868126e-05,
+      "loss": 2.6749,
+      "step": 47161
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.748211681842804,
+      "learning_rate": 7.46975809928593e-05,
+      "loss": 2.9393,
+      "step": 47162
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7243354916572571,
+      "learning_rate": 7.468741529230911e-05,
+      "loss": 2.8963,
+      "step": 47163
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7261446714401245,
+      "learning_rate": 7.4677250185189e-05,
+      "loss": 2.9313,
+      "step": 47164
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7376651167869568,
+      "learning_rate": 7.466708567152565e-05,
+      "loss": 2.7539,
+      "step": 47165
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7748557925224304,
+      "learning_rate": 7.465692175134575e-05,
+      "loss": 2.8347,
+      "step": 47166
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6955622434616089,
+      "learning_rate": 7.46467584246762e-05,
+      "loss": 2.6828,
+      "step": 47167
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7150453329086304,
+      "learning_rate": 7.463659569154367e-05,
+      "loss": 2.9824,
+      "step": 47168
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7358688712120056,
+      "learning_rate": 7.462643355197506e-05,
+      "loss": 2.9723,
+      "step": 47169
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7066740393638611,
+      "learning_rate": 7.461627200599694e-05,
+      "loss": 2.6598,
+      "step": 47170
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7047446370124817,
+      "learning_rate": 7.46061110536363e-05,
+      "loss": 2.9277,
+      "step": 47171
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7581452131271362,
+      "learning_rate": 7.459595069491976e-05,
+      "loss": 2.9114,
+      "step": 47172
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7496849894523621,
+      "learning_rate": 7.4585790929874e-05,
+      "loss": 2.9983,
+      "step": 47173
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7440602779388428,
+      "learning_rate": 7.4575631758526e-05,
+      "loss": 3.0909,
+      "step": 47174
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7057591080665588,
+      "learning_rate": 7.456547318090234e-05,
+      "loss": 2.9252,
+      "step": 47175
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7352858781814575,
+      "learning_rate": 7.455531519702978e-05,
+      "loss": 2.806,
+      "step": 47176
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7194220423698425,
+      "learning_rate": 7.454515780693527e-05,
+      "loss": 2.9611,
+      "step": 47177
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7085871696472168,
+      "learning_rate": 7.453500101064542e-05,
+      "loss": 3.1392,
+      "step": 47178
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7426517605781555,
+      "learning_rate": 7.452484480818694e-05,
+      "loss": 2.7845,
+      "step": 47179
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.671974241733551,
+      "learning_rate": 7.451468919958657e-05,
+      "loss": 2.9442,
+      "step": 47180
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6885994672775269,
+      "learning_rate": 7.450453418487112e-05,
+      "loss": 2.8299,
+      "step": 47181
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 7.44943797640674e-05,
+      "loss": 3.0384,
+      "step": 47182
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7851859927177429,
+      "learning_rate": 7.448422593720203e-05,
+      "loss": 2.8663,
+      "step": 47183
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.704609751701355,
+      "learning_rate": 7.447407270430185e-05,
+      "loss": 2.9252,
+      "step": 47184
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.705119252204895,
+      "learning_rate": 7.446392006539351e-05,
+      "loss": 3.0968,
+      "step": 47185
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.739789605140686,
+      "learning_rate": 7.445376802050387e-05,
+      "loss": 2.7645,
+      "step": 47186
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7058214545249939,
+      "learning_rate": 7.444361656965956e-05,
+      "loss": 2.7778,
+      "step": 47187
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6922550201416016,
+      "learning_rate": 7.443346571288733e-05,
+      "loss": 3.1309,
+      "step": 47188
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7196448445320129,
+      "learning_rate": 7.4423315450214e-05,
+      "loss": 2.9757,
+      "step": 47189
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7384520173072815,
+      "learning_rate": 7.441316578166617e-05,
+      "loss": 2.8736,
+      "step": 47190
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.8098287582397461,
+      "learning_rate": 7.440301670727062e-05,
+      "loss": 2.8157,
+      "step": 47191
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7096481919288635,
+      "learning_rate": 7.439286822705423e-05,
+      "loss": 2.9116,
+      "step": 47192
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7278345823287964,
+      "learning_rate": 7.438272034104356e-05,
+      "loss": 2.9966,
+      "step": 47193
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7431257367134094,
+      "learning_rate": 7.437257304926541e-05,
+      "loss": 2.9264,
+      "step": 47194
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7055538892745972,
+      "learning_rate": 7.436242635174638e-05,
+      "loss": 2.9575,
+      "step": 47195
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7002086639404297,
+      "learning_rate": 7.43522802485133e-05,
+      "loss": 2.7348,
+      "step": 47196
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7327378988265991,
+      "learning_rate": 7.434213473959299e-05,
+      "loss": 2.9672,
+      "step": 47197
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6708829402923584,
+      "learning_rate": 7.4331989825012e-05,
+      "loss": 2.9387,
+      "step": 47198
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6982355117797852,
+      "learning_rate": 7.432184550479715e-05,
+      "loss": 2.7249,
+      "step": 47199
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.719652533531189,
+      "learning_rate": 7.431170177897514e-05,
+      "loss": 3.0177,
+      "step": 47200
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7651442885398865,
+      "learning_rate": 7.430155864757261e-05,
+      "loss": 3.0903,
+      "step": 47201
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7121278047561646,
+      "learning_rate": 7.42914161106164e-05,
+      "loss": 2.9088,
+      "step": 47202
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6944317817687988,
+      "learning_rate": 7.42812741681331e-05,
+      "loss": 3.1477,
+      "step": 47203
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7282336950302124,
+      "learning_rate": 7.427113282014955e-05,
+      "loss": 2.9578,
+      "step": 47204
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6893202662467957,
+      "learning_rate": 7.426099206669234e-05,
+      "loss": 2.8417,
+      "step": 47205
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7046981453895569,
+      "learning_rate": 7.425085190778818e-05,
+      "loss": 2.9142,
+      "step": 47206
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.694952130317688,
+      "learning_rate": 7.424071234346395e-05,
+      "loss": 3.0216,
+      "step": 47207
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7959489822387695,
+      "learning_rate": 7.423057337374623e-05,
+      "loss": 2.7588,
+      "step": 47208
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.8016543388366699,
+      "learning_rate": 7.422043499866174e-05,
+      "loss": 2.9142,
+      "step": 47209
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7541260719299316,
+      "learning_rate": 7.421029721823706e-05,
+      "loss": 2.8306,
+      "step": 47210
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7044775485992432,
+      "learning_rate": 7.420016003249904e-05,
+      "loss": 2.8552,
+      "step": 47211
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.704594075679779,
+      "learning_rate": 7.419002344147441e-05,
+      "loss": 3.0789,
+      "step": 47212
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7283633351325989,
+      "learning_rate": 7.417988744518968e-05,
+      "loss": 2.8004,
+      "step": 47213
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.9223310351371765,
+      "learning_rate": 7.416975204367182e-05,
+      "loss": 2.9323,
+      "step": 47214
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7180284857749939,
+      "learning_rate": 7.41596172369473e-05,
+      "loss": 3.002,
+      "step": 47215
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7672765254974365,
+      "learning_rate": 7.414948302504284e-05,
+      "loss": 3.1372,
+      "step": 47216
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7458218932151794,
+      "learning_rate": 7.413934940798525e-05,
+      "loss": 2.8517,
+      "step": 47217
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6984786987304688,
+      "learning_rate": 7.412921638580106e-05,
+      "loss": 2.946,
+      "step": 47218
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7315978407859802,
+      "learning_rate": 7.411908395851715e-05,
+      "loss": 2.9058,
+      "step": 47219
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7283531427383423,
+      "learning_rate": 7.410895212616e-05,
+      "loss": 3.0075,
+      "step": 47220
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7277526259422302,
+      "learning_rate": 7.409882088875649e-05,
+      "loss": 2.9993,
+      "step": 47221
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7374258637428284,
+      "learning_rate": 7.408869024633321e-05,
+      "loss": 2.9424,
+      "step": 47222
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7135753035545349,
+      "learning_rate": 7.407856019891673e-05,
+      "loss": 2.9005,
+      "step": 47223
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7380639910697937,
+      "learning_rate": 7.406843074653395e-05,
+      "loss": 3.0794,
+      "step": 47224
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7215073108673096,
+      "learning_rate": 7.405830188921134e-05,
+      "loss": 3.0077,
+      "step": 47225
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7313400506973267,
+      "learning_rate": 7.404817362697579e-05,
+      "loss": 3.0572,
+      "step": 47226
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7082253694534302,
+      "learning_rate": 7.403804595985377e-05,
+      "loss": 2.9114,
+      "step": 47227
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7084292769432068,
+      "learning_rate": 7.402791888787204e-05,
+      "loss": 2.8432,
+      "step": 47228
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6946665048599243,
+      "learning_rate": 7.401779241105739e-05,
+      "loss": 2.9464,
+      "step": 47229
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.697638988494873,
+      "learning_rate": 7.400766652943636e-05,
+      "loss": 2.7472,
+      "step": 47230
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7209494709968567,
+      "learning_rate": 7.399754124303568e-05,
+      "loss": 2.576,
+      "step": 47231
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7583011984825134,
+      "learning_rate": 7.398741655188184e-05,
+      "loss": 3.1185,
+      "step": 47232
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.8109094500541687,
+      "learning_rate": 7.39772924560017e-05,
+      "loss": 3.1152,
+      "step": 47233
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7156438231468201,
+      "learning_rate": 7.396716895542191e-05,
+      "loss": 2.9722,
+      "step": 47234
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.712462842464447,
+      "learning_rate": 7.395704605016903e-05,
+      "loss": 2.8642,
+      "step": 47235
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6899236440658569,
+      "learning_rate": 7.394692374026989e-05,
+      "loss": 2.9877,
+      "step": 47236
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7041913270950317,
+      "learning_rate": 7.393680202575106e-05,
+      "loss": 2.9676,
+      "step": 47237
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7307323217391968,
+      "learning_rate": 7.392668090663905e-05,
+      "loss": 2.9229,
+      "step": 47238
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7059488892555237,
+      "learning_rate": 7.39165603829608e-05,
+      "loss": 3.1652,
+      "step": 47239
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.723430335521698,
+      "learning_rate": 7.39064404547427e-05,
+      "loss": 2.8094,
+      "step": 47240
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.700806200504303,
+      "learning_rate": 7.38963211220116e-05,
+      "loss": 2.7546,
+      "step": 47241
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7575515508651733,
+      "learning_rate": 7.388620238479399e-05,
+      "loss": 2.9442,
+      "step": 47242
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7537017464637756,
+      "learning_rate": 7.387608424311671e-05,
+      "loss": 2.818,
+      "step": 47243
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7906556129455566,
+      "learning_rate": 7.38659666970063e-05,
+      "loss": 2.6717,
+      "step": 47244
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7230514287948608,
+      "learning_rate": 7.385584974648932e-05,
+      "loss": 2.937,
+      "step": 47245
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7493695020675659,
+      "learning_rate": 7.384573339159261e-05,
+      "loss": 2.8982,
+      "step": 47246
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6845947504043579,
+      "learning_rate": 7.383561763234262e-05,
+      "loss": 2.7624,
+      "step": 47247
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6852810382843018,
+      "learning_rate": 7.382550246876609e-05,
+      "loss": 2.8006,
+      "step": 47248
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7320659756660461,
+      "learning_rate": 7.381538790088974e-05,
+      "loss": 2.9585,
+      "step": 47249
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.704412579536438,
+      "learning_rate": 7.380527392874004e-05,
+      "loss": 2.7925,
+      "step": 47250
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7940717935562134,
+      "learning_rate": 7.379516055234381e-05,
+      "loss": 3.0233,
+      "step": 47251
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7513147592544556,
+      "learning_rate": 7.378504777172757e-05,
+      "loss": 3.0405,
+      "step": 47252
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7274273633956909,
+      "learning_rate": 7.377493558691793e-05,
+      "loss": 3.1427,
+      "step": 47253
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7024219632148743,
+      "learning_rate": 7.376482399794164e-05,
+      "loss": 2.915,
+      "step": 47254
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7163192629814148,
+      "learning_rate": 7.375471300482517e-05,
+      "loss": 3.0311,
+      "step": 47255
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.685631513595581,
+      "learning_rate": 7.374460260759531e-05,
+      "loss": 2.7272,
+      "step": 47256
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7799549102783203,
+      "learning_rate": 7.373449280627858e-05,
+      "loss": 2.7492,
+      "step": 47257
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6964437365531921,
+      "learning_rate": 7.372438360090174e-05,
+      "loss": 2.9227,
+      "step": 47258
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7885956764221191,
+      "learning_rate": 7.371427499149127e-05,
+      "loss": 2.8981,
+      "step": 47259
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7466039657592773,
+      "learning_rate": 7.370416697807384e-05,
+      "loss": 2.7457,
+      "step": 47260
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7249606251716614,
+      "learning_rate": 7.369405956067612e-05,
+      "loss": 2.8658,
+      "step": 47261
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7615867853164673,
+      "learning_rate": 7.368395273932459e-05,
+      "loss": 2.8922,
+      "step": 47262
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6936567425727844,
+      "learning_rate": 7.367384651404602e-05,
+      "loss": 2.9543,
+      "step": 47263
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7578475475311279,
+      "learning_rate": 7.366374088486705e-05,
+      "loss": 2.954,
+      "step": 47264
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7116115689277649,
+      "learning_rate": 7.365363585181413e-05,
+      "loss": 3.0167,
+      "step": 47265
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7062137722969055,
+      "learning_rate": 7.364353141491413e-05,
+      "loss": 2.9352,
+      "step": 47266
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7953941226005554,
+      "learning_rate": 7.363342757419338e-05,
+      "loss": 2.9842,
+      "step": 47267
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7345401644706726,
+      "learning_rate": 7.362332432967857e-05,
+      "loss": 2.936,
+      "step": 47268
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.768124520778656,
+      "learning_rate": 7.361322168139647e-05,
+      "loss": 2.9616,
+      "step": 47269
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7298167943954468,
+      "learning_rate": 7.360311962937347e-05,
+      "loss": 3.0135,
+      "step": 47270
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6858828067779541,
+      "learning_rate": 7.35930181736364e-05,
+      "loss": 3.1113,
+      "step": 47271
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7418141961097717,
+      "learning_rate": 7.358291731421165e-05,
+      "loss": 2.8592,
+      "step": 47272
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7051644921302795,
+      "learning_rate": 7.357281705112598e-05,
+      "loss": 2.9234,
+      "step": 47273
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7438302040100098,
+      "learning_rate": 7.356271738440597e-05,
+      "loss": 2.9264,
+      "step": 47274
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.700322151184082,
+      "learning_rate": 7.355261831407807e-05,
+      "loss": 2.9011,
+      "step": 47275
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7566295862197876,
+      "learning_rate": 7.354251984016907e-05,
+      "loss": 3.0113,
+      "step": 47276
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7644765377044678,
+      "learning_rate": 7.353242196270545e-05,
+      "loss": 2.7803,
+      "step": 47277
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7297467589378357,
+      "learning_rate": 7.352232468171381e-05,
+      "loss": 2.927,
+      "step": 47278
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7533126473426819,
+      "learning_rate": 7.351222799722087e-05,
+      "loss": 2.7635,
+      "step": 47279
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7412834763526917,
+      "learning_rate": 7.350213190925316e-05,
+      "loss": 2.8367,
+      "step": 47280
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7042602300643921,
+      "learning_rate": 7.349203641783722e-05,
+      "loss": 2.9655,
+      "step": 47281
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7818167209625244,
+      "learning_rate": 7.348194152299956e-05,
+      "loss": 2.9304,
+      "step": 47282
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6732924580574036,
+      "learning_rate": 7.34718472247669e-05,
+      "loss": 2.804,
+      "step": 47283
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7412101030349731,
+      "learning_rate": 7.346175352316587e-05,
+      "loss": 2.9378,
+      "step": 47284
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7184247374534607,
+      "learning_rate": 7.345166041822291e-05,
+      "loss": 2.9313,
+      "step": 47285
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7510634064674377,
+      "learning_rate": 7.344156790996476e-05,
+      "loss": 2.8656,
+      "step": 47286
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6889684200286865,
+      "learning_rate": 7.343147599841783e-05,
+      "loss": 2.844,
+      "step": 47287
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.688102126121521,
+      "learning_rate": 7.342138468360885e-05,
+      "loss": 2.9992,
+      "step": 47288
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7358808517456055,
+      "learning_rate": 7.341129396556437e-05,
+      "loss": 3.0776,
+      "step": 47289
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7751854062080383,
+      "learning_rate": 7.340120384431084e-05,
+      "loss": 2.8425,
+      "step": 47290
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7157254815101624,
+      "learning_rate": 7.339111431987501e-05,
+      "loss": 2.9434,
+      "step": 47291
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.703119695186615,
+      "learning_rate": 7.338102539228328e-05,
+      "loss": 2.9188,
+      "step": 47292
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6870105862617493,
+      "learning_rate": 7.337093706156232e-05,
+      "loss": 3.0188,
+      "step": 47293
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.753438413143158,
+      "learning_rate": 7.336084932773878e-05,
+      "loss": 2.8878,
+      "step": 47294
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.715949535369873,
+      "learning_rate": 7.335076219083915e-05,
+      "loss": 2.9588,
+      "step": 47295
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6934475898742676,
+      "learning_rate": 7.334067565089001e-05,
+      "loss": 2.8501,
+      "step": 47296
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.696337878704071,
+      "learning_rate": 7.333058970791781e-05,
+      "loss": 2.8632,
+      "step": 47297
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7058655023574829,
+      "learning_rate": 7.332050436194928e-05,
+      "loss": 3.0071,
+      "step": 47298
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7588328719139099,
+      "learning_rate": 7.331041961301086e-05,
+      "loss": 2.8746,
+      "step": 47299
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7144933938980103,
+      "learning_rate": 7.330033546112915e-05,
+      "loss": 2.7623,
+      "step": 47300
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7363554239273071,
+      "learning_rate": 7.329025190633081e-05,
+      "loss": 2.9335,
+      "step": 47301
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7441504597663879,
+      "learning_rate": 7.328016894864232e-05,
+      "loss": 2.7166,
+      "step": 47302
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7541495561599731,
+      "learning_rate": 7.327008658809025e-05,
+      "loss": 2.8355,
+      "step": 47303
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7827197909355164,
+      "learning_rate": 7.326000482470102e-05,
+      "loss": 3.0587,
+      "step": 47304
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7094131112098694,
+      "learning_rate": 7.324992365850133e-05,
+      "loss": 2.8584,
+      "step": 47305
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7611328959465027,
+      "learning_rate": 7.323984308951778e-05,
+      "loss": 2.8228,
+      "step": 47306
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6947973370552063,
+      "learning_rate": 7.322976311777675e-05,
+      "loss": 2.6941,
+      "step": 47307
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7067661881446838,
+      "learning_rate": 7.321968374330495e-05,
+      "loss": 3.1039,
+      "step": 47308
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7562273144721985,
+      "learning_rate": 7.320960496612881e-05,
+      "loss": 2.8796,
+      "step": 47309
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7289214134216309,
+      "learning_rate": 7.319952678627499e-05,
+      "loss": 3.0467,
+      "step": 47310
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7128106951713562,
+      "learning_rate": 7.318944920376997e-05,
+      "loss": 2.882,
+      "step": 47311
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7181450128555298,
+      "learning_rate": 7.31793722186402e-05,
+      "loss": 3.0642,
+      "step": 47312
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7351670265197754,
+      "learning_rate": 7.316929583091238e-05,
+      "loss": 2.862,
+      "step": 47313
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.693118691444397,
+      "learning_rate": 7.315922004061292e-05,
+      "loss": 2.9532,
+      "step": 47314
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7117541432380676,
+      "learning_rate": 7.314914484776841e-05,
+      "loss": 2.8341,
+      "step": 47315
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.690302312374115,
+      "learning_rate": 7.313907025240549e-05,
+      "loss": 2.7515,
+      "step": 47316
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7544472217559814,
+      "learning_rate": 7.312899625455057e-05,
+      "loss": 2.8819,
+      "step": 47317
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7129111886024475,
+      "learning_rate": 7.311892285423025e-05,
+      "loss": 2.878,
+      "step": 47318
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7087661623954773,
+      "learning_rate": 7.31088500514709e-05,
+      "loss": 2.8482,
+      "step": 47319
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7653520703315735,
+      "learning_rate": 7.30987778462992e-05,
+      "loss": 2.7622,
+      "step": 47320
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6924532055854797,
+      "learning_rate": 7.308870623874171e-05,
+      "loss": 2.8243,
+      "step": 47321
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7128939032554626,
+      "learning_rate": 7.307863522882483e-05,
+      "loss": 2.8937,
+      "step": 47322
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6976691484451294,
+      "learning_rate": 7.306856481657526e-05,
+      "loss": 2.7315,
+      "step": 47323
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7400538921356201,
+      "learning_rate": 7.30584950020194e-05,
+      "loss": 2.7653,
+      "step": 47324
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.734775960445404,
+      "learning_rate": 7.30484257851837e-05,
+      "loss": 3.0132,
+      "step": 47325
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7255701422691345,
+      "learning_rate": 7.303835716609482e-05,
+      "loss": 2.882,
+      "step": 47326
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7620897889137268,
+      "learning_rate": 7.302828914477918e-05,
+      "loss": 2.9522,
+      "step": 47327
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.8050826787948608,
+      "learning_rate": 7.301822172126344e-05,
+      "loss": 2.9349,
+      "step": 47328
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.737463116645813,
+      "learning_rate": 7.300815489557394e-05,
+      "loss": 2.7589,
+      "step": 47329
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7167901992797852,
+      "learning_rate": 7.299808866773722e-05,
+      "loss": 3.015,
+      "step": 47330
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7121860384941101,
+      "learning_rate": 7.298802303777999e-05,
+      "loss": 2.9854,
+      "step": 47331
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7176076769828796,
+      "learning_rate": 7.297795800572858e-05,
+      "loss": 2.8822,
+      "step": 47332
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7200085520744324,
+      "learning_rate": 7.296789357160958e-05,
+      "loss": 3.0043,
+      "step": 47333
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7072357535362244,
+      "learning_rate": 7.295782973544932e-05,
+      "loss": 2.8592,
+      "step": 47334
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7582478523254395,
+      "learning_rate": 7.294776649727446e-05,
+      "loss": 3.0922,
+      "step": 47335
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.767278254032135,
+      "learning_rate": 7.29377038571116e-05,
+      "loss": 2.7644,
+      "step": 47336
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7298552989959717,
+      "learning_rate": 7.2927641814987e-05,
+      "loss": 2.8319,
+      "step": 47337
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6883513331413269,
+      "learning_rate": 7.291758037092739e-05,
+      "loss": 3.0519,
+      "step": 47338
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7653185725212097,
+      "learning_rate": 7.290751952495918e-05,
+      "loss": 3.0538,
+      "step": 47339
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7106703519821167,
+      "learning_rate": 7.289745927710878e-05,
+      "loss": 2.7821,
+      "step": 47340
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7490907311439514,
+      "learning_rate": 7.288739962740285e-05,
+      "loss": 3.0965,
+      "step": 47341
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7440788149833679,
+      "learning_rate": 7.287734057586768e-05,
+      "loss": 2.9511,
+      "step": 47342
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7107495665550232,
+      "learning_rate": 7.286728212252997e-05,
+      "loss": 2.9284,
+      "step": 47343
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7320402264595032,
+      "learning_rate": 7.285722426741608e-05,
+      "loss": 2.8895,
+      "step": 47344
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7153285145759583,
+      "learning_rate": 7.284716701055263e-05,
+      "loss": 2.7128,
+      "step": 47345
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7777250409126282,
+      "learning_rate": 7.283711035196602e-05,
+      "loss": 2.8695,
+      "step": 47346
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7311616539955139,
+      "learning_rate": 7.282705429168263e-05,
+      "loss": 3.0246,
+      "step": 47347
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7358239889144897,
+      "learning_rate": 7.281699882972916e-05,
+      "loss": 2.912,
+      "step": 47348
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7160769701004028,
+      "learning_rate": 7.280694396613194e-05,
+      "loss": 3.0558,
+      "step": 47349
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.8619295358657837,
+      "learning_rate": 7.279688970091744e-05,
+      "loss": 2.6971,
+      "step": 47350
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7436091303825378,
+      "learning_rate": 7.278683603411235e-05,
+      "loss": 2.9886,
+      "step": 47351
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7118484973907471,
+      "learning_rate": 7.277678296574288e-05,
+      "loss": 2.9102,
+      "step": 47352
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7110477089881897,
+      "learning_rate": 7.276673049583576e-05,
+      "loss": 2.9047,
+      "step": 47353
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7068123817443848,
+      "learning_rate": 7.275667862441731e-05,
+      "loss": 3.0401,
+      "step": 47354
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7033482789993286,
+      "learning_rate": 7.274662735151396e-05,
+      "loss": 2.8322,
+      "step": 47355
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.707513689994812,
+      "learning_rate": 7.273657667715235e-05,
+      "loss": 2.9297,
+      "step": 47356
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7551565170288086,
+      "learning_rate": 7.272652660135877e-05,
+      "loss": 3.0123,
+      "step": 47357
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7359175682067871,
+      "learning_rate": 7.271647712415987e-05,
+      "loss": 2.9357,
+      "step": 47358
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6989636421203613,
+      "learning_rate": 7.270642824558192e-05,
+      "loss": 2.9436,
+      "step": 47359
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7528616189956665,
+      "learning_rate": 7.269637996565159e-05,
+      "loss": 2.9145,
+      "step": 47360
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7683088183403015,
+      "learning_rate": 7.268633228439526e-05,
+      "loss": 2.9134,
+      "step": 47361
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7808603048324585,
+      "learning_rate": 7.267628520183927e-05,
+      "loss": 2.794,
+      "step": 47362
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.763262927532196,
+      "learning_rate": 7.266623871801032e-05,
+      "loss": 2.8442,
+      "step": 47363
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7707550525665283,
+      "learning_rate": 7.265619283293464e-05,
+      "loss": 2.8606,
+      "step": 47364
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7410865426063538,
+      "learning_rate": 7.26461475466388e-05,
+      "loss": 2.995,
+      "step": 47365
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7402563095092773,
+      "learning_rate": 7.263610285914933e-05,
+      "loss": 2.8527,
+      "step": 47366
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.774781346321106,
+      "learning_rate": 7.26260587704925e-05,
+      "loss": 2.9038,
+      "step": 47367
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7119985222816467,
+      "learning_rate": 7.261601528069507e-05,
+      "loss": 2.8396,
+      "step": 47368
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.8101449608802795,
+      "learning_rate": 7.260597238978316e-05,
+      "loss": 2.9007,
+      "step": 47369
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7611187696456909,
+      "learning_rate": 7.259593009778329e-05,
+      "loss": 2.9946,
+      "step": 47370
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7523656487464905,
+      "learning_rate": 7.258588840472209e-05,
+      "loss": 2.9665,
+      "step": 47371
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7050994634628296,
+      "learning_rate": 7.257584731062583e-05,
+      "loss": 2.9543,
+      "step": 47372
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7177386283874512,
+      "learning_rate": 7.256580681552106e-05,
+      "loss": 3.003,
+      "step": 47373
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7362361550331116,
+      "learning_rate": 7.255576691943413e-05,
+      "loss": 2.6465,
+      "step": 47374
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7049803733825684,
+      "learning_rate": 7.25457276223916e-05,
+      "loss": 2.9968,
+      "step": 47375
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6966225504875183,
+      "learning_rate": 7.253568892441988e-05,
+      "loss": 2.8923,
+      "step": 47376
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.733306884765625,
+      "learning_rate": 7.252565082554526e-05,
+      "loss": 2.816,
+      "step": 47377
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7428738474845886,
+      "learning_rate": 7.251561332579441e-05,
+      "loss": 2.5529,
+      "step": 47378
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7241402268409729,
+      "learning_rate": 7.250557642519354e-05,
+      "loss": 2.6333,
+      "step": 47379
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7124062776565552,
+      "learning_rate": 7.249554012376931e-05,
+      "loss": 2.9414,
+      "step": 47380
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.721976101398468,
+      "learning_rate": 7.248550442154797e-05,
+      "loss": 2.9294,
+      "step": 47381
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7031937837600708,
+      "learning_rate": 7.247546931855609e-05,
+      "loss": 2.6558,
+      "step": 47382
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7368708848953247,
+      "learning_rate": 7.246543481482002e-05,
+      "loss": 2.6632,
+      "step": 47383
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7131219506263733,
+      "learning_rate": 7.245540091036614e-05,
+      "loss": 2.7965,
+      "step": 47384
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7883298993110657,
+      "learning_rate": 7.244536760522103e-05,
+      "loss": 2.7718,
+      "step": 47385
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.8104693293571472,
+      "learning_rate": 7.243533489941094e-05,
+      "loss": 2.874,
+      "step": 47386
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7419458031654358,
+      "learning_rate": 7.242530279296239e-05,
+      "loss": 2.8306,
+      "step": 47387
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7081753015518188,
+      "learning_rate": 7.241527128590185e-05,
+      "loss": 2.8082,
+      "step": 47388
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7221062183380127,
+      "learning_rate": 7.240524037825561e-05,
+      "loss": 2.83,
+      "step": 47389
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.736663281917572,
+      "learning_rate": 7.239521007005026e-05,
+      "loss": 2.8742,
+      "step": 47390
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7576841711997986,
+      "learning_rate": 7.238518036131214e-05,
+      "loss": 2.9401,
+      "step": 47391
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7570251822471619,
+      "learning_rate": 7.237515125206755e-05,
+      "loss": 2.7961,
+      "step": 47392
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7038759589195251,
+      "learning_rate": 7.236512274234308e-05,
+      "loss": 2.9395,
+      "step": 47393
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.714281439781189,
+      "learning_rate": 7.235509483216501e-05,
+      "loss": 3.0216,
+      "step": 47394
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7493951916694641,
+      "learning_rate": 7.234506752155986e-05,
+      "loss": 2.8871,
+      "step": 47395
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7389973998069763,
+      "learning_rate": 7.233504081055393e-05,
+      "loss": 3.0943,
+      "step": 47396
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7300322651863098,
+      "learning_rate": 7.232501469917377e-05,
+      "loss": 2.9865,
+      "step": 47397
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6961638331413269,
+      "learning_rate": 7.231498918744567e-05,
+      "loss": 2.7805,
+      "step": 47398
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7474333047866821,
+      "learning_rate": 7.230496427539601e-05,
+      "loss": 2.7427,
+      "step": 47399
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7698484063148499,
+      "learning_rate": 7.229493996305135e-05,
+      "loss": 2.8524,
+      "step": 47400
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7011622190475464,
+      "learning_rate": 7.228491625043791e-05,
+      "loss": 3.0087,
+      "step": 47401
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7040664553642273,
+      "learning_rate": 7.227489313758214e-05,
+      "loss": 2.7225,
+      "step": 47402
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7392282485961914,
+      "learning_rate": 7.226487062451061e-05,
+      "loss": 2.9377,
+      "step": 47403
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.697232723236084,
+      "learning_rate": 7.225484871124953e-05,
+      "loss": 2.9972,
+      "step": 47404
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6810755133628845,
+      "learning_rate": 7.224482739782538e-05,
+      "loss": 3.0003,
+      "step": 47405
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7662976384162903,
+      "learning_rate": 7.223480668426444e-05,
+      "loss": 2.6729,
+      "step": 47406
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7100536227226257,
+      "learning_rate": 7.222478657059318e-05,
+      "loss": 2.8443,
+      "step": 47407
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7035341262817383,
+      "learning_rate": 7.221476705683807e-05,
+      "loss": 2.8649,
+      "step": 47408
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7688108086585999,
+      "learning_rate": 7.220474814302536e-05,
+      "loss": 2.8819,
+      "step": 47409
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7461357116699219,
+      "learning_rate": 7.219472982918155e-05,
+      "loss": 2.9177,
+      "step": 47410
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.727735161781311,
+      "learning_rate": 7.218471211533293e-05,
+      "loss": 2.9527,
+      "step": 47411
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7366343140602112,
+      "learning_rate": 7.217469500150602e-05,
+      "loss": 2.886,
+      "step": 47412
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.76343834400177,
+      "learning_rate": 7.21646784877271e-05,
+      "loss": 2.9228,
+      "step": 47413
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7279701828956604,
+      "learning_rate": 7.215466257402252e-05,
+      "loss": 2.9934,
+      "step": 47414
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7330734133720398,
+      "learning_rate": 7.214464726041873e-05,
+      "loss": 2.8277,
+      "step": 47415
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7413004636764526,
+      "learning_rate": 7.213463254694206e-05,
+      "loss": 2.9595,
+      "step": 47416
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7532314658164978,
+      "learning_rate": 7.212461843361888e-05,
+      "loss": 2.9631,
+      "step": 47417
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7230085134506226,
+      "learning_rate": 7.211460492047571e-05,
+      "loss": 3.1592,
+      "step": 47418
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7556701302528381,
+      "learning_rate": 7.21045920075388e-05,
+      "loss": 2.9073,
+      "step": 47419
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.713424801826477,
+      "learning_rate": 7.209457969483454e-05,
+      "loss": 3.1115,
+      "step": 47420
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7252991199493408,
+      "learning_rate": 7.20845679823892e-05,
+      "loss": 2.8228,
+      "step": 47421
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7200642228126526,
+      "learning_rate": 7.207455687022924e-05,
+      "loss": 2.9238,
+      "step": 47422
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7238421440124512,
+      "learning_rate": 7.206454635838115e-05,
+      "loss": 2.826,
+      "step": 47423
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7494496703147888,
+      "learning_rate": 7.205453644687107e-05,
+      "loss": 2.9615,
+      "step": 47424
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7229695916175842,
+      "learning_rate": 7.204452713572555e-05,
+      "loss": 2.6852,
+      "step": 47425
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7765418887138367,
+      "learning_rate": 7.20345184249709e-05,
+      "loss": 2.8497,
+      "step": 47426
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7109969258308411,
+      "learning_rate": 7.202451031463334e-05,
+      "loss": 2.8812,
+      "step": 47427
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7664215564727783,
+      "learning_rate": 7.201450280473943e-05,
+      "loss": 2.7231,
+      "step": 47428
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7280341982841492,
+      "learning_rate": 7.20044958953154e-05,
+      "loss": 3.0021,
+      "step": 47429
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7474103569984436,
+      "learning_rate": 7.199448958638768e-05,
+      "loss": 2.6722,
+      "step": 47430
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.8022735714912415,
+      "learning_rate": 7.198448387798253e-05,
+      "loss": 2.6483,
+      "step": 47431
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7711005210876465,
+      "learning_rate": 7.19744787701264e-05,
+      "loss": 2.7537,
+      "step": 47432
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7397340536117554,
+      "learning_rate": 7.196447426284565e-05,
+      "loss": 2.9846,
+      "step": 47433
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7473747730255127,
+      "learning_rate": 7.195447035616661e-05,
+      "loss": 2.7482,
+      "step": 47434
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7164973020553589,
+      "learning_rate": 7.194446705011561e-05,
+      "loss": 3.0895,
+      "step": 47435
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7285258769989014,
+      "learning_rate": 7.19344643447189e-05,
+      "loss": 2.9992,
+      "step": 47436
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.741519033908844,
+      "learning_rate": 7.19244622400029e-05,
+      "loss": 2.8107,
+      "step": 47437
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7468116283416748,
+      "learning_rate": 7.191446073599408e-05,
+      "loss": 3.1146,
+      "step": 47438
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7286897301673889,
+      "learning_rate": 7.19044598327186e-05,
+      "loss": 2.855,
+      "step": 47439
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7384865880012512,
+      "learning_rate": 7.189445953020292e-05,
+      "loss": 2.6907,
+      "step": 47440
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7932009696960449,
+      "learning_rate": 7.188445982847336e-05,
+      "loss": 2.9254,
+      "step": 47441
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7377317547798157,
+      "learning_rate": 7.187446072755613e-05,
+      "loss": 2.8892,
+      "step": 47442
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7107809782028198,
+      "learning_rate": 7.18644622274778e-05,
+      "loss": 2.9628,
+      "step": 47443
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8373969197273254,
+      "learning_rate": 7.185446432826444e-05,
+      "loss": 2.7945,
+      "step": 47444
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7064989805221558,
+      "learning_rate": 7.184446702994262e-05,
+      "loss": 2.7455,
+      "step": 47445
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7400098443031311,
+      "learning_rate": 7.183447033253845e-05,
+      "loss": 2.8944,
+      "step": 47446
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8077271580696106,
+      "learning_rate": 7.182447423607852e-05,
+      "loss": 3.0493,
+      "step": 47447
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6956033706665039,
+      "learning_rate": 7.181447874058895e-05,
+      "loss": 2.9979,
+      "step": 47448
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6707293391227722,
+      "learning_rate": 7.180448384609606e-05,
+      "loss": 2.9002,
+      "step": 47449
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6920158863067627,
+      "learning_rate": 7.179448955262634e-05,
+      "loss": 3.0506,
+      "step": 47450
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7085862159729004,
+      "learning_rate": 7.17844958602059e-05,
+      "loss": 2.8566,
+      "step": 47451
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7076865434646606,
+      "learning_rate": 7.177450276886129e-05,
+      "loss": 2.936,
+      "step": 47452
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7326858043670654,
+      "learning_rate": 7.176451027861863e-05,
+      "loss": 3.1567,
+      "step": 47453
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6912264823913574,
+      "learning_rate": 7.175451838950433e-05,
+      "loss": 3.0702,
+      "step": 47454
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7292463779449463,
+      "learning_rate": 7.174452710154475e-05,
+      "loss": 2.7248,
+      "step": 47455
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7355402112007141,
+      "learning_rate": 7.173453641476622e-05,
+      "loss": 2.8516,
+      "step": 47456
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7186945080757141,
+      "learning_rate": 7.172454632919493e-05,
+      "loss": 2.7671,
+      "step": 47457
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7160497903823853,
+      "learning_rate": 7.171455684485721e-05,
+      "loss": 2.8474,
+      "step": 47458
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7210008502006531,
+      "learning_rate": 7.170456796177939e-05,
+      "loss": 2.6995,
+      "step": 47459
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7707409858703613,
+      "learning_rate": 7.16945796799879e-05,
+      "loss": 2.9242,
+      "step": 47460
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6895835399627686,
+      "learning_rate": 7.168459199950884e-05,
+      "loss": 2.9707,
+      "step": 47461
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.734785795211792,
+      "learning_rate": 7.167460492036871e-05,
+      "loss": 2.9447,
+      "step": 47462
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7238161563873291,
+      "learning_rate": 7.166461844259376e-05,
+      "loss": 2.8899,
+      "step": 47463
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6985582709312439,
+      "learning_rate": 7.165463256621016e-05,
+      "loss": 3.0099,
+      "step": 47464
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8546037673950195,
+      "learning_rate": 7.164464729124439e-05,
+      "loss": 2.9665,
+      "step": 47465
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7167295813560486,
+      "learning_rate": 7.163466261772261e-05,
+      "loss": 2.8678,
+      "step": 47466
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7737851142883301,
+      "learning_rate": 7.162467854567125e-05,
+      "loss": 2.9525,
+      "step": 47467
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.741790771484375,
+      "learning_rate": 7.161469507511642e-05,
+      "loss": 3.0993,
+      "step": 47468
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7221540808677673,
+      "learning_rate": 7.160471220608466e-05,
+      "loss": 2.7828,
+      "step": 47469
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7722229957580566,
+      "learning_rate": 7.159472993860212e-05,
+      "loss": 2.8201,
+      "step": 47470
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7124035358428955,
+      "learning_rate": 7.1584748272695e-05,
+      "loss": 2.8623,
+      "step": 47471
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6828311681747437,
+      "learning_rate": 7.157476720838981e-05,
+      "loss": 3.045,
+      "step": 47472
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6699220538139343,
+      "learning_rate": 7.156478674571261e-05,
+      "loss": 2.8361,
+      "step": 47473
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7215057015419006,
+      "learning_rate": 7.15548068846898e-05,
+      "loss": 3.0085,
+      "step": 47474
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7264130115509033,
+      "learning_rate": 7.154482762534776e-05,
+      "loss": 2.7652,
+      "step": 47475
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7632395029067993,
+      "learning_rate": 7.15348489677126e-05,
+      "loss": 2.9227,
+      "step": 47476
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7025119662284851,
+      "learning_rate": 7.152487091181074e-05,
+      "loss": 2.9366,
+      "step": 47477
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.700484037399292,
+      "learning_rate": 7.151489345766842e-05,
+      "loss": 2.8368,
+      "step": 47478
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7083252668380737,
+      "learning_rate": 7.150491660531179e-05,
+      "loss": 2.9546,
+      "step": 47479
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7263261675834656,
+      "learning_rate": 7.149494035476732e-05,
+      "loss": 3.1148,
+      "step": 47480
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.755137026309967,
+      "learning_rate": 7.148496470606114e-05,
+      "loss": 2.9635,
+      "step": 47481
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7463118433952332,
+      "learning_rate": 7.147498965921965e-05,
+      "loss": 2.8403,
+      "step": 47482
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6622539162635803,
+      "learning_rate": 7.1465015214269e-05,
+      "loss": 2.8063,
+      "step": 47483
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7079895734786987,
+      "learning_rate": 7.145504137123555e-05,
+      "loss": 2.8559,
+      "step": 47484
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6947416067123413,
+      "learning_rate": 7.144506813014557e-05,
+      "loss": 2.9942,
+      "step": 47485
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7773357629776001,
+      "learning_rate": 7.143509549102524e-05,
+      "loss": 3.0241,
+      "step": 47486
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7051486372947693,
+      "learning_rate": 7.142512345390093e-05,
+      "loss": 2.9468,
+      "step": 47487
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7427360415458679,
+      "learning_rate": 7.141515201879877e-05,
+      "loss": 2.8745,
+      "step": 47488
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6730073094367981,
+      "learning_rate": 7.14051811857451e-05,
+      "loss": 2.7274,
+      "step": 47489
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6964417695999146,
+      "learning_rate": 7.139521095476629e-05,
+      "loss": 2.8322,
+      "step": 47490
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6890613436698914,
+      "learning_rate": 7.13852413258884e-05,
+      "loss": 3.0825,
+      "step": 47491
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7256914973258972,
+      "learning_rate": 7.137527229913796e-05,
+      "loss": 2.9446,
+      "step": 47492
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6925685405731201,
+      "learning_rate": 7.136530387454089e-05,
+      "loss": 2.8681,
+      "step": 47493
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.712538480758667,
+      "learning_rate": 7.135533605212363e-05,
+      "loss": 3.152,
+      "step": 47494
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7189764380455017,
+      "learning_rate": 7.134536883191244e-05,
+      "loss": 2.7111,
+      "step": 47495
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.73213130235672,
+      "learning_rate": 7.133540221393349e-05,
+      "loss": 2.8043,
+      "step": 47496
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.704254150390625,
+      "learning_rate": 7.132543619821318e-05,
+      "loss": 2.9035,
+      "step": 47497
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7130454778671265,
+      "learning_rate": 7.131547078477757e-05,
+      "loss": 2.6674,
+      "step": 47498
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6928410530090332,
+      "learning_rate": 7.130550597365308e-05,
+      "loss": 2.9636,
+      "step": 47499
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6888012886047363,
+      "learning_rate": 7.129554176486587e-05,
+      "loss": 2.921,
+      "step": 47500
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7133302688598633,
+      "learning_rate": 7.12855781584421e-05,
+      "loss": 2.9925,
+      "step": 47501
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7603508234024048,
+      "learning_rate": 7.127561515440818e-05,
+      "loss": 2.8541,
+      "step": 47502
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8669589757919312,
+      "learning_rate": 7.126565275279019e-05,
+      "loss": 2.982,
+      "step": 47503
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7202988862991333,
+      "learning_rate": 7.125569095361447e-05,
+      "loss": 2.8785,
+      "step": 47504
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6969466805458069,
+      "learning_rate": 7.124572975690731e-05,
+      "loss": 3.0048,
+      "step": 47505
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7168153524398804,
+      "learning_rate": 7.123576916269487e-05,
+      "loss": 2.8723,
+      "step": 47506
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7444748282432556,
+      "learning_rate": 7.122580917100338e-05,
+      "loss": 3.053,
+      "step": 47507
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7564586997032166,
+      "learning_rate": 7.121584978185903e-05,
+      "loss": 2.9999,
+      "step": 47508
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7514041066169739,
+      "learning_rate": 7.120589099528807e-05,
+      "loss": 2.8877,
+      "step": 47509
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7307299971580505,
+      "learning_rate": 7.119593281131684e-05,
+      "loss": 3.0657,
+      "step": 47510
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7431744933128357,
+      "learning_rate": 7.118597522997142e-05,
+      "loss": 2.7934,
+      "step": 47511
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7438200116157532,
+      "learning_rate": 7.117601825127817e-05,
+      "loss": 2.9412,
+      "step": 47512
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7108175754547119,
+      "learning_rate": 7.11660618752632e-05,
+      "loss": 2.8302,
+      "step": 47513
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6957178711891174,
+      "learning_rate": 7.115610610195285e-05,
+      "loss": 3.1017,
+      "step": 47514
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7372037768363953,
+      "learning_rate": 7.114615093137326e-05,
+      "loss": 2.8933,
+      "step": 47515
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7208225131034851,
+      "learning_rate": 7.113619636355061e-05,
+      "loss": 3.0632,
+      "step": 47516
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7117692232131958,
+      "learning_rate": 7.112624239851123e-05,
+      "loss": 2.95,
+      "step": 47517
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7726485133171082,
+      "learning_rate": 7.11162890362812e-05,
+      "loss": 3.0097,
+      "step": 47518
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7577906847000122,
+      "learning_rate": 7.110633627688684e-05,
+      "loss": 2.875,
+      "step": 47519
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7112172245979309,
+      "learning_rate": 7.109638412035439e-05,
+      "loss": 2.795,
+      "step": 47520
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7526218891143799,
+      "learning_rate": 7.108643256671002e-05,
+      "loss": 3.0337,
+      "step": 47521
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.883800208568573,
+      "learning_rate": 7.107648161597993e-05,
+      "loss": 2.7492,
+      "step": 47522
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7418673634529114,
+      "learning_rate": 7.106653126819027e-05,
+      "loss": 2.9126,
+      "step": 47523
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7279163002967834,
+      "learning_rate": 7.10565815233673e-05,
+      "loss": 3.0511,
+      "step": 47524
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7238751649856567,
+      "learning_rate": 7.104663238153731e-05,
+      "loss": 2.8631,
+      "step": 47525
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7138600945472717,
+      "learning_rate": 7.103668384272632e-05,
+      "loss": 2.7983,
+      "step": 47526
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7913941144943237,
+      "learning_rate": 7.102673590696074e-05,
+      "loss": 3.039,
+      "step": 47527
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7563286423683167,
+      "learning_rate": 7.101678857426669e-05,
+      "loss": 2.974,
+      "step": 47528
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.744836151599884,
+      "learning_rate": 7.100684184467025e-05,
+      "loss": 2.9618,
+      "step": 47529
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7190981507301331,
+      "learning_rate": 7.099689571819778e-05,
+      "loss": 3.0592,
+      "step": 47530
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7641801834106445,
+      "learning_rate": 7.098695019487535e-05,
+      "loss": 2.817,
+      "step": 47531
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7573862075805664,
+      "learning_rate": 7.097700527472934e-05,
+      "loss": 2.9914,
+      "step": 47532
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7357966899871826,
+      "learning_rate": 7.096706095778567e-05,
+      "loss": 2.8585,
+      "step": 47533
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8040800094604492,
+      "learning_rate": 7.095711724407081e-05,
+      "loss": 3.1379,
+      "step": 47534
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7585517764091492,
+      "learning_rate": 7.094717413361075e-05,
+      "loss": 2.8646,
+      "step": 47535
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7271699905395508,
+      "learning_rate": 7.09372316264318e-05,
+      "loss": 3.0099,
+      "step": 47536
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7077714800834656,
+      "learning_rate": 7.092728972256013e-05,
+      "loss": 2.8668,
+      "step": 47537
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7180951237678528,
+      "learning_rate": 7.09173484220218e-05,
+      "loss": 2.758,
+      "step": 47538
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7497140765190125,
+      "learning_rate": 7.090740772484315e-05,
+      "loss": 2.8124,
+      "step": 47539
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7486867904663086,
+      "learning_rate": 7.089746763105021e-05,
+      "loss": 2.8029,
+      "step": 47540
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6781291961669922,
+      "learning_rate": 7.088752814066926e-05,
+      "loss": 3.086,
+      "step": 47541
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7236211895942688,
+      "learning_rate": 7.087758925372658e-05,
+      "loss": 2.9152,
+      "step": 47542
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7155327200889587,
+      "learning_rate": 7.08676509702482e-05,
+      "loss": 2.8893,
+      "step": 47543
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7060009837150574,
+      "learning_rate": 7.085771329026036e-05,
+      "loss": 2.9783,
+      "step": 47544
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7284157872200012,
+      "learning_rate": 7.084777621378908e-05,
+      "loss": 2.9028,
+      "step": 47545
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7081487774848938,
+      "learning_rate": 7.083783974086068e-05,
+      "loss": 2.9875,
+      "step": 47546
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7967720627784729,
+      "learning_rate": 7.082790387150137e-05,
+      "loss": 2.9191,
+      "step": 47547
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6838148832321167,
+      "learning_rate": 7.081796860573721e-05,
+      "loss": 2.9943,
+      "step": 47548
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7207189798355103,
+      "learning_rate": 7.080803394359448e-05,
+      "loss": 2.8106,
+      "step": 47549
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7479190826416016,
+      "learning_rate": 7.079809988509926e-05,
+      "loss": 2.8428,
+      "step": 47550
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6582784652709961,
+      "learning_rate": 7.078816643027767e-05,
+      "loss": 2.7417,
+      "step": 47551
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7573397755622864,
+      "learning_rate": 7.077823357915602e-05,
+      "loss": 3.0219,
+      "step": 47552
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7158548831939697,
+      "learning_rate": 7.076830133176032e-05,
+      "loss": 2.9151,
+      "step": 47553
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7018261551856995,
+      "learning_rate": 7.075836968811687e-05,
+      "loss": 2.9753,
+      "step": 47554
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7017635703086853,
+      "learning_rate": 7.074843864825164e-05,
+      "loss": 3.038,
+      "step": 47555
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7433605194091797,
+      "learning_rate": 7.073850821219094e-05,
+      "loss": 2.915,
+      "step": 47556
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7094408273696899,
+      "learning_rate": 7.072857837996098e-05,
+      "loss": 2.9466,
+      "step": 47557
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7520463466644287,
+      "learning_rate": 7.071864915158777e-05,
+      "loss": 2.8484,
+      "step": 47558
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7429606318473816,
+      "learning_rate": 7.070872052709754e-05,
+      "loss": 2.9665,
+      "step": 47559
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7459293603897095,
+      "learning_rate": 7.069879250651634e-05,
+      "loss": 3.1393,
+      "step": 47560
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7327176928520203,
+      "learning_rate": 7.06888650898704e-05,
+      "loss": 2.6332,
+      "step": 47561
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7204429507255554,
+      "learning_rate": 7.06789382771859e-05,
+      "loss": 2.6896,
+      "step": 47562
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.751361608505249,
+      "learning_rate": 7.066901206848888e-05,
+      "loss": 3.0084,
+      "step": 47563
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7837120294570923,
+      "learning_rate": 7.065908646380565e-05,
+      "loss": 2.9443,
+      "step": 47564
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7149581909179688,
+      "learning_rate": 7.064916146316221e-05,
+      "loss": 3.048,
+      "step": 47565
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7094987630844116,
+      "learning_rate": 7.063923706658469e-05,
+      "loss": 2.8596,
+      "step": 47566
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7471241354942322,
+      "learning_rate": 7.062931327409933e-05,
+      "loss": 2.9456,
+      "step": 47567
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8007349967956543,
+      "learning_rate": 7.061939008573216e-05,
+      "loss": 3.0527,
+      "step": 47568
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7094486355781555,
+      "learning_rate": 7.060946750150944e-05,
+      "loss": 2.8813,
+      "step": 47569
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6963364481925964,
+      "learning_rate": 7.059954552145718e-05,
+      "loss": 2.9723,
+      "step": 47570
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.737665057182312,
+      "learning_rate": 7.058962414560161e-05,
+      "loss": 2.8979,
+      "step": 47571
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7348478436470032,
+      "learning_rate": 7.057970337396884e-05,
+      "loss": 3.1241,
+      "step": 47572
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7251835465431213,
+      "learning_rate": 7.05697832065849e-05,
+      "loss": 2.9297,
+      "step": 47573
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7356771230697632,
+      "learning_rate": 7.055986364347608e-05,
+      "loss": 2.9944,
+      "step": 47574
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7503240704536438,
+      "learning_rate": 7.05499446846683e-05,
+      "loss": 2.7325,
+      "step": 47575
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7172363996505737,
+      "learning_rate": 7.054002633018785e-05,
+      "loss": 2.8792,
+      "step": 47576
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7425442934036255,
+      "learning_rate": 7.05301085800609e-05,
+      "loss": 2.9969,
+      "step": 47577
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7278176546096802,
+      "learning_rate": 7.052019143431336e-05,
+      "loss": 2.8364,
+      "step": 47578
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.711143970489502,
+      "learning_rate": 7.051027489297157e-05,
+      "loss": 2.8685,
+      "step": 47579
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7150108218193054,
+      "learning_rate": 7.050035895606156e-05,
+      "loss": 2.9239,
+      "step": 47580
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7470396161079407,
+      "learning_rate": 7.049044362360938e-05,
+      "loss": 2.9947,
+      "step": 47581
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6868398189544678,
+      "learning_rate": 7.048052889564125e-05,
+      "loss": 3.0457,
+      "step": 47582
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7090058326721191,
+      "learning_rate": 7.047061477218312e-05,
+      "loss": 2.9331,
+      "step": 47583
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7475082278251648,
+      "learning_rate": 7.046070125326135e-05,
+      "loss": 2.9428,
+      "step": 47584
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7163534164428711,
+      "learning_rate": 7.045078833890183e-05,
+      "loss": 2.9627,
+      "step": 47585
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7448176741600037,
+      "learning_rate": 7.044087602913084e-05,
+      "loss": 2.9439,
+      "step": 47586
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7112734317779541,
+      "learning_rate": 7.043096432397437e-05,
+      "loss": 2.8171,
+      "step": 47587
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7507283687591553,
+      "learning_rate": 7.042105322345847e-05,
+      "loss": 2.9742,
+      "step": 47588
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.749320924282074,
+      "learning_rate": 7.041114272760948e-05,
+      "loss": 2.8345,
+      "step": 47589
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7169381976127625,
+      "learning_rate": 7.040123283645322e-05,
+      "loss": 3.0269,
+      "step": 47590
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7252839207649231,
+      "learning_rate": 7.039132355001592e-05,
+      "loss": 2.8734,
+      "step": 47591
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7459837794303894,
+      "learning_rate": 7.03814148683238e-05,
+      "loss": 2.8828,
+      "step": 47592
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7205169200897217,
+      "learning_rate": 7.037150679140276e-05,
+      "loss": 2.8298,
+      "step": 47593
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7566838264465332,
+      "learning_rate": 7.036159931927911e-05,
+      "loss": 2.6596,
+      "step": 47594
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7399532198905945,
+      "learning_rate": 7.035169245197867e-05,
+      "loss": 2.791,
+      "step": 47595
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.744898796081543,
+      "learning_rate": 7.034178618952769e-05,
+      "loss": 2.9397,
+      "step": 47596
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.741045355796814,
+      "learning_rate": 7.03318805319523e-05,
+      "loss": 2.9292,
+      "step": 47597
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7274682521820068,
+      "learning_rate": 7.032197547927847e-05,
+      "loss": 2.7797,
+      "step": 47598
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7332300543785095,
+      "learning_rate": 7.031207103153241e-05,
+      "loss": 2.758,
+      "step": 47599
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7493497729301453,
+      "learning_rate": 7.030216718874011e-05,
+      "loss": 3.1175,
+      "step": 47600
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7280975580215454,
+      "learning_rate": 7.029226395092774e-05,
+      "loss": 2.8282,
+      "step": 47601
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7699424028396606,
+      "learning_rate": 7.028236131812138e-05,
+      "loss": 2.8258,
+      "step": 47602
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7214636206626892,
+      "learning_rate": 7.027245929034695e-05,
+      "loss": 2.7958,
+      "step": 47603
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7424951195716858,
+      "learning_rate": 7.026255786763075e-05,
+      "loss": 2.8481,
+      "step": 47604
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7581061124801636,
+      "learning_rate": 7.025265704999867e-05,
+      "loss": 2.9243,
+      "step": 47605
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6965128779411316,
+      "learning_rate": 7.024275683747688e-05,
+      "loss": 2.958,
+      "step": 47606
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7387304902076721,
+      "learning_rate": 7.023285723009155e-05,
+      "loss": 3.0201,
+      "step": 47607
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7237225770950317,
+      "learning_rate": 7.022295822786861e-05,
+      "loss": 2.7551,
+      "step": 47608
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6971107721328735,
+      "learning_rate": 7.021305983083423e-05,
+      "loss": 2.8338,
+      "step": 47609
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7018718719482422,
+      "learning_rate": 7.020316203901432e-05,
+      "loss": 2.6294,
+      "step": 47610
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.678143322467804,
+      "learning_rate": 7.019326485243513e-05,
+      "loss": 3.0442,
+      "step": 47611
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.718937873840332,
+      "learning_rate": 7.018336827112257e-05,
+      "loss": 2.7917,
+      "step": 47612
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.701008677482605,
+      "learning_rate": 7.01734722951028e-05,
+      "loss": 2.7951,
+      "step": 47613
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7457568645477295,
+      "learning_rate": 7.016357692440193e-05,
+      "loss": 2.8648,
+      "step": 47614
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6575247049331665,
+      "learning_rate": 7.015368215904593e-05,
+      "loss": 2.987,
+      "step": 47615
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7346352338790894,
+      "learning_rate": 7.014378799906095e-05,
+      "loss": 2.8045,
+      "step": 47616
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7044388055801392,
+      "learning_rate": 7.013389444447301e-05,
+      "loss": 2.7003,
+      "step": 47617
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7355045676231384,
+      "learning_rate": 7.012400149530804e-05,
+      "loss": 2.9629,
+      "step": 47618
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7335948348045349,
+      "learning_rate": 7.011410915159232e-05,
+      "loss": 2.8067,
+      "step": 47619
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7035436630249023,
+      "learning_rate": 7.010421741335168e-05,
+      "loss": 2.7938,
+      "step": 47620
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6967812180519104,
+      "learning_rate": 7.009432628061243e-05,
+      "loss": 2.8902,
+      "step": 47621
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7366414070129395,
+      "learning_rate": 7.008443575340032e-05,
+      "loss": 2.6877,
+      "step": 47622
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7091039419174194,
+      "learning_rate": 7.00745458317417e-05,
+      "loss": 2.7761,
+      "step": 47623
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7306919693946838,
+      "learning_rate": 7.006465651566247e-05,
+      "loss": 2.782,
+      "step": 47624
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6974079012870789,
+      "learning_rate": 7.005476780518856e-05,
+      "loss": 2.9357,
+      "step": 47625
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7919893860816956,
+      "learning_rate": 7.004487970034627e-05,
+      "loss": 2.9349,
+      "step": 47626
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7445873022079468,
+      "learning_rate": 7.00349922011614e-05,
+      "loss": 3.0683,
+      "step": 47627
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7282187342643738,
+      "learning_rate": 7.00251053076601e-05,
+      "loss": 2.9457,
+      "step": 47628
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6826276779174805,
+      "learning_rate": 7.001521901986848e-05,
+      "loss": 2.7284,
+      "step": 47629
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7542269825935364,
+      "learning_rate": 7.000533333781256e-05,
+      "loss": 2.8766,
+      "step": 47630
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.763077974319458,
+      "learning_rate": 6.999544826151829e-05,
+      "loss": 2.846,
+      "step": 47631
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7759633660316467,
+      "learning_rate": 6.99855637910117e-05,
+      "loss": 3.0532,
+      "step": 47632
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7288686037063599,
+      "learning_rate": 6.997567992631883e-05,
+      "loss": 3.0587,
+      "step": 47633
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7107670307159424,
+      "learning_rate": 6.996579666746584e-05,
+      "loss": 2.9452,
+      "step": 47634
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7164334654808044,
+      "learning_rate": 6.995591401447858e-05,
+      "loss": 3.1213,
+      "step": 47635
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7293843030929565,
+      "learning_rate": 6.994603196738326e-05,
+      "loss": 2.9488,
+      "step": 47636
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7001025676727295,
+      "learning_rate": 6.993615052620576e-05,
+      "loss": 2.8525,
+      "step": 47637
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.736518919467926,
+      "learning_rate": 6.992626969097221e-05,
+      "loss": 3.0508,
+      "step": 47638
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6913434267044067,
+      "learning_rate": 6.991638946170861e-05,
+      "loss": 2.8179,
+      "step": 47639
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6732978820800781,
+      "learning_rate": 6.990650983844087e-05,
+      "loss": 3.0415,
+      "step": 47640
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7510814070701599,
+      "learning_rate": 6.989663082119518e-05,
+      "loss": 2.9317,
+      "step": 47641
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6988676190376282,
+      "learning_rate": 6.98867524099974e-05,
+      "loss": 2.9012,
+      "step": 47642
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7889525890350342,
+      "learning_rate": 6.987687460487362e-05,
+      "loss": 2.7416,
+      "step": 47643
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6934266686439514,
+      "learning_rate": 6.986699740584998e-05,
+      "loss": 3.0119,
+      "step": 47644
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7091248631477356,
+      "learning_rate": 6.985712081295235e-05,
+      "loss": 2.8714,
+      "step": 47645
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7630372047424316,
+      "learning_rate": 6.984724482620678e-05,
+      "loss": 3.0798,
+      "step": 47646
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7124559879302979,
+      "learning_rate": 6.983736944563919e-05,
+      "loss": 3.0748,
+      "step": 47647
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7879320979118347,
+      "learning_rate": 6.982749467127566e-05,
+      "loss": 2.7061,
+      "step": 47648
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7572959065437317,
+      "learning_rate": 6.981762050314232e-05,
+      "loss": 3.1158,
+      "step": 47649
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7393097877502441,
+      "learning_rate": 6.980774694126495e-05,
+      "loss": 2.8247,
+      "step": 47650
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7403393387794495,
+      "learning_rate": 6.979787398566979e-05,
+      "loss": 2.862,
+      "step": 47651
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7622966170310974,
+      "learning_rate": 6.978800163638268e-05,
+      "loss": 2.9045,
+      "step": 47652
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7208401560783386,
+      "learning_rate": 6.977812989342964e-05,
+      "loss": 2.8656,
+      "step": 47653
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7213475108146667,
+      "learning_rate": 6.976825875683674e-05,
+      "loss": 2.9971,
+      "step": 47654
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7142321467399597,
+      "learning_rate": 6.975838822662987e-05,
+      "loss": 3.0197,
+      "step": 47655
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8058083057403564,
+      "learning_rate": 6.97485183028352e-05,
+      "loss": 2.7902,
+      "step": 47656
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7557923197746277,
+      "learning_rate": 6.973864898547852e-05,
+      "loss": 2.9234,
+      "step": 47657
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6770955324172974,
+      "learning_rate": 6.97287802745859e-05,
+      "loss": 2.6823,
+      "step": 47658
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7257001399993896,
+      "learning_rate": 6.971891217018349e-05,
+      "loss": 2.9746,
+      "step": 47659
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.665259063243866,
+      "learning_rate": 6.97090446722971e-05,
+      "loss": 2.6029,
+      "step": 47660
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7421854734420776,
+      "learning_rate": 6.96991777809528e-05,
+      "loss": 2.9455,
+      "step": 47661
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7707141041755676,
+      "learning_rate": 6.968931149617643e-05,
+      "loss": 3.1785,
+      "step": 47662
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7522193789482117,
+      "learning_rate": 6.967944581799409e-05,
+      "loss": 2.8248,
+      "step": 47663
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.747096836566925,
+      "learning_rate": 6.966958074643189e-05,
+      "loss": 3.0049,
+      "step": 47664
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7807623744010925,
+      "learning_rate": 6.965971628151558e-05,
+      "loss": 2.9797,
+      "step": 47665
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6890551447868347,
+      "learning_rate": 6.964985242327132e-05,
+      "loss": 2.7737,
+      "step": 47666
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.744841456413269,
+      "learning_rate": 6.963998917172501e-05,
+      "loss": 2.9702,
+      "step": 47667
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7255067825317383,
+      "learning_rate": 6.963012652690257e-05,
+      "loss": 2.9437,
+      "step": 47668
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7375694513320923,
+      "learning_rate": 6.962026448883013e-05,
+      "loss": 2.8523,
+      "step": 47669
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7139278054237366,
+      "learning_rate": 6.961040305753347e-05,
+      "loss": 2.7384,
+      "step": 47670
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.74039626121521,
+      "learning_rate": 6.960054223303875e-05,
+      "loss": 2.9185,
+      "step": 47671
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7546448707580566,
+      "learning_rate": 6.95906820153718e-05,
+      "loss": 2.917,
+      "step": 47672
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.726020872592926,
+      "learning_rate": 6.958082240455871e-05,
+      "loss": 2.8537,
+      "step": 47673
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7269747257232666,
+      "learning_rate": 6.95709634006254e-05,
+      "loss": 2.9355,
+      "step": 47674
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7227620482444763,
+      "learning_rate": 6.956110500359776e-05,
+      "loss": 2.8675,
+      "step": 47675
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6997947096824646,
+      "learning_rate": 6.955124721350187e-05,
+      "loss": 2.8877,
+      "step": 47676
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7475396990776062,
+      "learning_rate": 6.954139003036357e-05,
+      "loss": 2.9018,
+      "step": 47677
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6795568466186523,
+      "learning_rate": 6.953153345420892e-05,
+      "loss": 2.901,
+      "step": 47678
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7249543070793152,
+      "learning_rate": 6.952167748506393e-05,
+      "loss": 2.9914,
+      "step": 47679
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.715927243232727,
+      "learning_rate": 6.951182212295438e-05,
+      "loss": 2.8587,
+      "step": 47680
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7298700213432312,
+      "learning_rate": 6.950196736790645e-05,
+      "loss": 2.942,
+      "step": 47681
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7658237814903259,
+      "learning_rate": 6.949211321994597e-05,
+      "loss": 3.0521,
+      "step": 47682
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7058979272842407,
+      "learning_rate": 6.948225967909878e-05,
+      "loss": 2.8072,
+      "step": 47683
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6835278868675232,
+      "learning_rate": 6.947240674539109e-05,
+      "loss": 3.063,
+      "step": 47684
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7149368524551392,
+      "learning_rate": 6.946255441884858e-05,
+      "loss": 3.0687,
+      "step": 47685
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7184211611747742,
+      "learning_rate": 6.945270269949744e-05,
+      "loss": 2.9289,
+      "step": 47686
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8042710423469543,
+      "learning_rate": 6.944285158736344e-05,
+      "loss": 2.8712,
+      "step": 47687
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7624470591545105,
+      "learning_rate": 6.943300108247268e-05,
+      "loss": 2.86,
+      "step": 47688
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6996687650680542,
+      "learning_rate": 6.9423151184851e-05,
+      "loss": 2.894,
+      "step": 47689
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7101437449455261,
+      "learning_rate": 6.941330189452431e-05,
+      "loss": 2.7634,
+      "step": 47690
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7409912943840027,
+      "learning_rate": 6.940345321151867e-05,
+      "loss": 2.7299,
+      "step": 47691
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7612147331237793,
+      "learning_rate": 6.939360513585988e-05,
+      "loss": 2.9949,
+      "step": 47692
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7287157773971558,
+      "learning_rate": 6.938375766757401e-05,
+      "loss": 2.9987,
+      "step": 47693
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7364184260368347,
+      "learning_rate": 6.937391080668691e-05,
+      "loss": 2.9102,
+      "step": 47694
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7477836608886719,
+      "learning_rate": 6.936406455322458e-05,
+      "loss": 2.9471,
+      "step": 47695
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.746062159538269,
+      "learning_rate": 6.935421890721293e-05,
+      "loss": 2.8956,
+      "step": 47696
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7618799209594727,
+      "learning_rate": 6.934437386867779e-05,
+      "loss": 3.0421,
+      "step": 47697
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6989330053329468,
+      "learning_rate": 6.933452943764527e-05,
+      "loss": 3.0442,
+      "step": 47698
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7096387147903442,
+      "learning_rate": 6.932468561414115e-05,
+      "loss": 2.7441,
+      "step": 47699
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7569087147712708,
+      "learning_rate": 6.931484239819134e-05,
+      "loss": 2.8035,
+      "step": 47700
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7354052662849426,
+      "learning_rate": 6.930499978982198e-05,
+      "loss": 3.0321,
+      "step": 47701
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.723621129989624,
+      "learning_rate": 6.929515778905878e-05,
+      "loss": 2.9771,
+      "step": 47702
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7257896661758423,
+      "learning_rate": 6.928531639592778e-05,
+      "loss": 2.5832,
+      "step": 47703
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7249346375465393,
+      "learning_rate": 6.927547561045487e-05,
+      "loss": 2.8028,
+      "step": 47704
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7552078366279602,
+      "learning_rate": 6.926563543266587e-05,
+      "loss": 2.9345,
+      "step": 47705
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7495518922805786,
+      "learning_rate": 6.925579586258689e-05,
+      "loss": 2.7828,
+      "step": 47706
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6926343441009521,
+      "learning_rate": 6.92459569002436e-05,
+      "loss": 3.0711,
+      "step": 47707
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7301257252693176,
+      "learning_rate": 6.923611854566219e-05,
+      "loss": 2.9068,
+      "step": 47708
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7383841276168823,
+      "learning_rate": 6.922628079886829e-05,
+      "loss": 2.9509,
+      "step": 47709
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7082839608192444,
+      "learning_rate": 6.921644365988807e-05,
+      "loss": 2.8345,
+      "step": 47710
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8540667295455933,
+      "learning_rate": 6.920660712874732e-05,
+      "loss": 2.9104,
+      "step": 47711
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6855905652046204,
+      "learning_rate": 6.919677120547189e-05,
+      "loss": 3.1228,
+      "step": 47712
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7041099667549133,
+      "learning_rate": 6.91869358900878e-05,
+      "loss": 2.6503,
+      "step": 47713
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7083460688591003,
+      "learning_rate": 6.91771011826208e-05,
+      "loss": 2.8676,
+      "step": 47714
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7578673362731934,
+      "learning_rate": 6.916726708309693e-05,
+      "loss": 3.0593,
+      "step": 47715
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7105020880699158,
+      "learning_rate": 6.915743359154212e-05,
+      "loss": 2.7786,
+      "step": 47716
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7636972665786743,
+      "learning_rate": 6.91476007079821e-05,
+      "loss": 2.8917,
+      "step": 47717
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7262598276138306,
+      "learning_rate": 6.9137768432443e-05,
+      "loss": 3.1578,
+      "step": 47718
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7343432307243347,
+      "learning_rate": 6.912793676495057e-05,
+      "loss": 3.0025,
+      "step": 47719
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7269200086593628,
+      "learning_rate": 6.911810570553063e-05,
+      "loss": 2.8975,
+      "step": 47720
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7389751672744751,
+      "learning_rate": 6.910827525420923e-05,
+      "loss": 3.0294,
+      "step": 47721
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7205168604850769,
+      "learning_rate": 6.909844541101213e-05,
+      "loss": 2.8921,
+      "step": 47722
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7263239622116089,
+      "learning_rate": 6.908861617596538e-05,
+      "loss": 2.9286,
+      "step": 47723
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7304479479789734,
+      "learning_rate": 6.907878754909467e-05,
+      "loss": 2.722,
+      "step": 47724
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7433699369430542,
+      "learning_rate": 6.906895953042611e-05,
+      "loss": 2.9203,
+      "step": 47725
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7258157730102539,
+      "learning_rate": 6.905913211998546e-05,
+      "loss": 2.8464,
+      "step": 47726
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7196770906448364,
+      "learning_rate": 6.904930531779852e-05,
+      "loss": 2.9407,
+      "step": 47727
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.736857533454895,
+      "learning_rate": 6.903947912389133e-05,
+      "loss": 2.902,
+      "step": 47728
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8262715935707092,
+      "learning_rate": 6.902965353828963e-05,
+      "loss": 3.0228,
+      "step": 47729
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.740009069442749,
+      "learning_rate": 6.901982856101938e-05,
+      "loss": 2.985,
+      "step": 47730
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7470229864120483,
+      "learning_rate": 6.901000419210652e-05,
+      "loss": 2.9201,
+      "step": 47731
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.705930769443512,
+      "learning_rate": 6.900018043157686e-05,
+      "loss": 2.4926,
+      "step": 47732
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7490159869194031,
+      "learning_rate": 6.899035727945624e-05,
+      "loss": 2.8409,
+      "step": 47733
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7825806736946106,
+      "learning_rate": 6.898053473577052e-05,
+      "loss": 2.9334,
+      "step": 47734
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7251452207565308,
+      "learning_rate": 6.897071280054558e-05,
+      "loss": 2.864,
+      "step": 47735
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7280852794647217,
+      "learning_rate": 6.896089147380743e-05,
+      "loss": 2.8302,
+      "step": 47736
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7135215401649475,
+      "learning_rate": 6.895107075558173e-05,
+      "loss": 2.8002,
+      "step": 47737
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7429882884025574,
+      "learning_rate": 6.894125064589455e-05,
+      "loss": 2.7959,
+      "step": 47738
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7092627882957458,
+      "learning_rate": 6.893143114477152e-05,
+      "loss": 2.8633,
+      "step": 47739
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7626802921295166,
+      "learning_rate": 6.892161225223875e-05,
+      "loss": 2.9394,
+      "step": 47740
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7585441470146179,
+      "learning_rate": 6.891179396832196e-05,
+      "loss": 2.8403,
+      "step": 47741
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6958228945732117,
+      "learning_rate": 6.890197629304694e-05,
+      "loss": 2.9464,
+      "step": 47742
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7759091854095459,
+      "learning_rate": 6.889215922643975e-05,
+      "loss": 2.838,
+      "step": 47743
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7096752524375916,
+      "learning_rate": 6.888234276852604e-05,
+      "loss": 2.9345,
+      "step": 47744
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7344189882278442,
+      "learning_rate": 6.887252691933175e-05,
+      "loss": 3.2115,
+      "step": 47745
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7086279392242432,
+      "learning_rate": 6.886271167888284e-05,
+      "loss": 2.7383,
+      "step": 47746
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7072312235832214,
+      "learning_rate": 6.885289704720507e-05,
+      "loss": 2.9756,
+      "step": 47747
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7190287709236145,
+      "learning_rate": 6.884308302432428e-05,
+      "loss": 2.8416,
+      "step": 47748
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7287060618400574,
+      "learning_rate": 6.883326961026621e-05,
+      "loss": 2.9435,
+      "step": 47749
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7083621621131897,
+      "learning_rate": 6.882345680505684e-05,
+      "loss": 2.9013,
+      "step": 47750
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7850791215896606,
+      "learning_rate": 6.881364460872209e-05,
+      "loss": 2.8539,
+      "step": 47751
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7175219655036926,
+      "learning_rate": 6.880383302128761e-05,
+      "loss": 2.8696,
+      "step": 47752
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7923054099082947,
+      "learning_rate": 6.879402204277945e-05,
+      "loss": 2.7013,
+      "step": 47753
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7685131430625916,
+      "learning_rate": 6.878421167322329e-05,
+      "loss": 2.782,
+      "step": 47754
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6883244514465332,
+      "learning_rate": 6.877440191264497e-05,
+      "loss": 2.882,
+      "step": 47755
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7355750203132629,
+      "learning_rate": 6.876459276107043e-05,
+      "loss": 2.6442,
+      "step": 47756
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7222840189933777,
+      "learning_rate": 6.875478421852539e-05,
+      "loss": 2.9603,
+      "step": 47757
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7285621762275696,
+      "learning_rate": 6.874497628503582e-05,
+      "loss": 2.9548,
+      "step": 47758
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7584347724914551,
+      "learning_rate": 6.87351689606274e-05,
+      "loss": 2.9329,
+      "step": 47759
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8179945945739746,
+      "learning_rate": 6.872536224532602e-05,
+      "loss": 3.0256,
+      "step": 47760
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7475566267967224,
+      "learning_rate": 6.871555613915764e-05,
+      "loss": 3.0588,
+      "step": 47761
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6885706186294556,
+      "learning_rate": 6.870575064214795e-05,
+      "loss": 3.1584,
+      "step": 47762
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.713989794254303,
+      "learning_rate": 6.869594575432282e-05,
+      "loss": 2.9286,
+      "step": 47763
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7147015929222107,
+      "learning_rate": 6.868614147570795e-05,
+      "loss": 2.9192,
+      "step": 47764
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.718229353427887,
+      "learning_rate": 6.867633780632934e-05,
+      "loss": 2.8925,
+      "step": 47765
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7605919241905212,
+      "learning_rate": 6.866653474621268e-05,
+      "loss": 2.8551,
+      "step": 47766
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7661386728286743,
+      "learning_rate": 6.865673229538384e-05,
+      "loss": 2.9573,
+      "step": 47767
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7024096846580505,
+      "learning_rate": 6.864693045386872e-05,
+      "loss": 2.832,
+      "step": 47768
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7051783800125122,
+      "learning_rate": 6.863712922169305e-05,
+      "loss": 2.7672,
+      "step": 47769
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7214155793190002,
+      "learning_rate": 6.862732859888267e-05,
+      "loss": 2.6812,
+      "step": 47770
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7348127365112305,
+      "learning_rate": 6.861752858546328e-05,
+      "loss": 2.9103,
+      "step": 47771
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6963437795639038,
+      "learning_rate": 6.860772918146082e-05,
+      "loss": 2.7701,
+      "step": 47772
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.733690619468689,
+      "learning_rate": 6.859793038690113e-05,
+      "loss": 2.8199,
+      "step": 47773
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6885663270950317,
+      "learning_rate": 6.858813220180988e-05,
+      "loss": 2.8728,
+      "step": 47774
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.708145797252655,
+      "learning_rate": 6.857833462621303e-05,
+      "loss": 2.7329,
+      "step": 47775
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.71966552734375,
+      "learning_rate": 6.85685376601363e-05,
+      "loss": 2.8923,
+      "step": 47776
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7468180656433105,
+      "learning_rate": 6.855874130360546e-05,
+      "loss": 2.9768,
+      "step": 47777
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7484171986579895,
+      "learning_rate": 6.85489455566464e-05,
+      "loss": 2.7614,
+      "step": 47778
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7327077388763428,
+      "learning_rate": 6.853915041928479e-05,
+      "loss": 3.0613,
+      "step": 47779
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7350683212280273,
+      "learning_rate": 6.85293558915466e-05,
+      "loss": 2.9889,
+      "step": 47780
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7374870181083679,
+      "learning_rate": 6.851956197345747e-05,
+      "loss": 2.9741,
+      "step": 47781
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7582040429115295,
+      "learning_rate": 6.850976866504325e-05,
+      "loss": 2.8146,
+      "step": 47782
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6711180806159973,
+      "learning_rate": 6.849997596632983e-05,
+      "loss": 3.099,
+      "step": 47783
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7463443279266357,
+      "learning_rate": 6.849018387734293e-05,
+      "loss": 2.9372,
+      "step": 47784
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.74971604347229,
+      "learning_rate": 6.848039239810832e-05,
+      "loss": 2.8886,
+      "step": 47785
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7736210227012634,
+      "learning_rate": 6.847060152865173e-05,
+      "loss": 2.7884,
+      "step": 47786
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8005310297012329,
+      "learning_rate": 6.846081126899903e-05,
+      "loss": 2.9747,
+      "step": 47787
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7372254133224487,
+      "learning_rate": 6.845102161917604e-05,
+      "loss": 2.6831,
+      "step": 47788
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7566565275192261,
+      "learning_rate": 6.844123257920845e-05,
+      "loss": 3.0529,
+      "step": 47789
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.726959764957428,
+      "learning_rate": 6.843144414912214e-05,
+      "loss": 2.8369,
+      "step": 47790
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7393621206283569,
+      "learning_rate": 6.842165632894286e-05,
+      "loss": 3.0986,
+      "step": 47791
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7415285110473633,
+      "learning_rate": 6.841186911869628e-05,
+      "loss": 2.9621,
+      "step": 47792
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7591820955276489,
+      "learning_rate": 6.840208251840834e-05,
+      "loss": 2.9591,
+      "step": 47793
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7299678325653076,
+      "learning_rate": 6.839229652810467e-05,
+      "loss": 2.8123,
+      "step": 47794
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.727067232131958,
+      "learning_rate": 6.838251114781118e-05,
+      "loss": 3.0317,
+      "step": 47795
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.721146285533905,
+      "learning_rate": 6.837272637755351e-05,
+      "loss": 2.9183,
+      "step": 47796
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7278393507003784,
+      "learning_rate": 6.836294221735761e-05,
+      "loss": 2.8804,
+      "step": 47797
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6947075724601746,
+      "learning_rate": 6.835315866724911e-05,
+      "loss": 2.8544,
+      "step": 47798
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7457790970802307,
+      "learning_rate": 6.834337572725372e-05,
+      "loss": 2.9784,
+      "step": 47799
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7459518313407898,
+      "learning_rate": 6.83335933973974e-05,
+      "loss": 2.8344,
+      "step": 47800
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7571932673454285,
+      "learning_rate": 6.832381167770574e-05,
+      "loss": 2.968,
+      "step": 47801
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6742973327636719,
+      "learning_rate": 6.831403056820455e-05,
+      "loss": 2.9231,
+      "step": 47802
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7140032052993774,
+      "learning_rate": 6.83042500689197e-05,
+      "loss": 2.7132,
+      "step": 47803
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7056100368499756,
+      "learning_rate": 6.82944701798768e-05,
+      "loss": 2.9182,
+      "step": 47804
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7132856845855713,
+      "learning_rate": 6.828469090110172e-05,
+      "loss": 2.9506,
+      "step": 47805
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7402212619781494,
+      "learning_rate": 6.827491223262018e-05,
+      "loss": 2.9935,
+      "step": 47806
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7678213119506836,
+      "learning_rate": 6.826513417445785e-05,
+      "loss": 2.8589,
+      "step": 47807
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6965996623039246,
+      "learning_rate": 6.825535672664063e-05,
+      "loss": 2.8132,
+      "step": 47808
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7514899969100952,
+      "learning_rate": 6.824557988919414e-05,
+      "loss": 2.8645,
+      "step": 47809
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7812216281890869,
+      "learning_rate": 6.823580366214428e-05,
+      "loss": 2.6188,
+      "step": 47810
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7255864143371582,
+      "learning_rate": 6.822602804551659e-05,
+      "loss": 2.8784,
+      "step": 47811
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7047956585884094,
+      "learning_rate": 6.821625303933703e-05,
+      "loss": 2.9166,
+      "step": 47812
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6879226565361023,
+      "learning_rate": 6.820647864363128e-05,
+      "loss": 2.9923,
+      "step": 47813
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7026329636573792,
+      "learning_rate": 6.819670485842495e-05,
+      "loss": 2.7337,
+      "step": 47814
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7578959465026855,
+      "learning_rate": 6.818693168374398e-05,
+      "loss": 2.8791,
+      "step": 47815
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6991091966629028,
+      "learning_rate": 6.817715911961391e-05,
+      "loss": 2.6325,
+      "step": 47816
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.769070029258728,
+      "learning_rate": 6.81673871660606e-05,
+      "loss": 3.1683,
+      "step": 47817
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.72699373960495,
+      "learning_rate": 6.815761582310986e-05,
+      "loss": 2.7933,
+      "step": 47818
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.756587028503418,
+      "learning_rate": 6.814784509078728e-05,
+      "loss": 2.7654,
+      "step": 47819
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7062783241271973,
+      "learning_rate": 6.813807496911882e-05,
+      "loss": 2.9737,
+      "step": 47820
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.708519458770752,
+      "learning_rate": 6.812830545812985e-05,
+      "loss": 2.8675,
+      "step": 47821
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7568124532699585,
+      "learning_rate": 6.811853655784633e-05,
+      "loss": 2.9477,
+      "step": 47822
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7141517400741577,
+      "learning_rate": 6.810876826829403e-05,
+      "loss": 2.8733,
+      "step": 47823
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.732783854007721,
+      "learning_rate": 6.80990005894985e-05,
+      "loss": 2.923,
+      "step": 47824
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7289860248565674,
+      "learning_rate": 6.808923352148567e-05,
+      "loss": 2.9246,
+      "step": 47825
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7409937977790833,
+      "learning_rate": 6.80794670642811e-05,
+      "loss": 3.007,
+      "step": 47826
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7329813838005066,
+      "learning_rate": 6.806970121791062e-05,
+      "loss": 2.8461,
+      "step": 47827
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6916979551315308,
+      "learning_rate": 6.805993598239998e-05,
+      "loss": 2.9873,
+      "step": 47828
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7569551467895508,
+      "learning_rate": 6.805017135777469e-05,
+      "loss": 2.7445,
+      "step": 47829
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7574961185455322,
+      "learning_rate": 6.804040734406071e-05,
+      "loss": 2.7435,
+      "step": 47830
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7416138648986816,
+      "learning_rate": 6.803064394128355e-05,
+      "loss": 2.873,
+      "step": 47831
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7477595210075378,
+      "learning_rate": 6.802088114946908e-05,
+      "loss": 2.9572,
+      "step": 47832
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6934207677841187,
+      "learning_rate": 6.801111896864301e-05,
+      "loss": 2.7564,
+      "step": 47833
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7345800995826721,
+      "learning_rate": 6.800135739883103e-05,
+      "loss": 2.9987,
+      "step": 47834
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7357404232025146,
+      "learning_rate": 6.799159644005882e-05,
+      "loss": 2.8571,
+      "step": 47835
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7066752314567566,
+      "learning_rate": 6.7981836092352e-05,
+      "loss": 2.9571,
+      "step": 47836
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.766917884349823,
+      "learning_rate": 6.797207635573636e-05,
+      "loss": 2.8954,
+      "step": 47837
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8036574721336365,
+      "learning_rate": 6.796231723023772e-05,
+      "loss": 2.9591,
+      "step": 47838
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7077456116676331,
+      "learning_rate": 6.795255871588157e-05,
+      "loss": 2.9171,
+      "step": 47839
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7191286087036133,
+      "learning_rate": 6.794280081269385e-05,
+      "loss": 2.8068,
+      "step": 47840
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6867917776107788,
+      "learning_rate": 6.79330435207e-05,
+      "loss": 2.9191,
+      "step": 47841
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7825203537940979,
+      "learning_rate": 6.792328683992598e-05,
+      "loss": 2.9092,
+      "step": 47842
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7153095602989197,
+      "learning_rate": 6.791353077039733e-05,
+      "loss": 2.7737,
+      "step": 47843
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.768951416015625,
+      "learning_rate": 6.790377531213969e-05,
+      "loss": 2.8843,
+      "step": 47844
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8321551084518433,
+      "learning_rate": 6.789402046517896e-05,
+      "loss": 2.6498,
+      "step": 47845
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.726223886013031,
+      "learning_rate": 6.788426622954059e-05,
+      "loss": 2.914,
+      "step": 47846
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7852296233177185,
+      "learning_rate": 6.787451260525045e-05,
+      "loss": 2.9952,
+      "step": 47847
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7488097548484802,
+      "learning_rate": 6.786475959233414e-05,
+      "loss": 3.0178,
+      "step": 47848
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.7254324555397034,
+      "learning_rate": 6.785500719081742e-05,
+      "loss": 3.0343,
+      "step": 47849
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7108405828475952,
+      "learning_rate": 6.784525540072596e-05,
+      "loss": 2.789,
+      "step": 47850
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7349480986595154,
+      "learning_rate": 6.783550422208534e-05,
+      "loss": 2.6636,
+      "step": 47851
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7456988096237183,
+      "learning_rate": 6.782575365492138e-05,
+      "loss": 2.9112,
+      "step": 47852
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7272082567214966,
+      "learning_rate": 6.781600369925962e-05,
+      "loss": 3.1821,
+      "step": 47853
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7798593640327454,
+      "learning_rate": 6.780625435512585e-05,
+      "loss": 2.9526,
+      "step": 47854
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7535591721534729,
+      "learning_rate": 6.77965056225458e-05,
+      "loss": 2.8622,
+      "step": 47855
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.680900514125824,
+      "learning_rate": 6.778675750154506e-05,
+      "loss": 2.9729,
+      "step": 47856
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7286996245384216,
+      "learning_rate": 6.777700999214928e-05,
+      "loss": 3.0362,
+      "step": 47857
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.772244930267334,
+      "learning_rate": 6.776726309438408e-05,
+      "loss": 3.0197,
+      "step": 47858
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7076226472854614,
+      "learning_rate": 6.775751680827525e-05,
+      "loss": 3.0025,
+      "step": 47859
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7190698385238647,
+      "learning_rate": 6.774777113384848e-05,
+      "loss": 3.1247,
+      "step": 47860
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7115904688835144,
+      "learning_rate": 6.773802607112934e-05,
+      "loss": 2.7755,
+      "step": 47861
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7206274271011353,
+      "learning_rate": 6.772828162014359e-05,
+      "loss": 2.8821,
+      "step": 47862
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7294838428497314,
+      "learning_rate": 6.771853778091678e-05,
+      "loss": 2.8908,
+      "step": 47863
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.728641927242279,
+      "learning_rate": 6.77087945534747e-05,
+      "loss": 2.984,
+      "step": 47864
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6940865516662598,
+      "learning_rate": 6.769905193784295e-05,
+      "loss": 2.7064,
+      "step": 47865
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7145792245864868,
+      "learning_rate": 6.768930993404712e-05,
+      "loss": 2.7834,
+      "step": 47866
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7666870951652527,
+      "learning_rate": 6.767956854211302e-05,
+      "loss": 2.8824,
+      "step": 47867
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6863735318183899,
+      "learning_rate": 6.766982776206616e-05,
+      "loss": 2.7799,
+      "step": 47868
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.714529275894165,
+      "learning_rate": 6.766008759393222e-05,
+      "loss": 2.8986,
+      "step": 47869
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6909911036491394,
+      "learning_rate": 6.7650348037737e-05,
+      "loss": 2.6996,
+      "step": 47870
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6930615901947021,
+      "learning_rate": 6.764060909350604e-05,
+      "loss": 3.0113,
+      "step": 47871
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7450976967811584,
+      "learning_rate": 6.7630870761265e-05,
+      "loss": 2.7526,
+      "step": 47872
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7528846263885498,
+      "learning_rate": 6.762113304103947e-05,
+      "loss": 3.0688,
+      "step": 47873
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7500634789466858,
+      "learning_rate": 6.761139593285512e-05,
+      "loss": 2.8791,
+      "step": 47874
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7247843146324158,
+      "learning_rate": 6.760165943673774e-05,
+      "loss": 3.054,
+      "step": 47875
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6777768731117249,
+      "learning_rate": 6.759192355271272e-05,
+      "loss": 3.0018,
+      "step": 47876
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7233495116233826,
+      "learning_rate": 6.758218828080598e-05,
+      "loss": 2.7997,
+      "step": 47877
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7728328108787537,
+      "learning_rate": 6.757245362104303e-05,
+      "loss": 3.1007,
+      "step": 47878
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6918376684188843,
+      "learning_rate": 6.756271957344938e-05,
+      "loss": 2.9147,
+      "step": 47879
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6865652799606323,
+      "learning_rate": 6.755298613805092e-05,
+      "loss": 3.0058,
+      "step": 47880
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7141489386558533,
+      "learning_rate": 6.754325331487308e-05,
+      "loss": 2.8366,
+      "step": 47881
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7318657636642456,
+      "learning_rate": 6.753352110394163e-05,
+      "loss": 2.7465,
+      "step": 47882
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7256183624267578,
+      "learning_rate": 6.752378950528206e-05,
+      "loss": 2.975,
+      "step": 47883
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.8105762600898743,
+      "learning_rate": 6.751405851892009e-05,
+      "loss": 2.629,
+      "step": 47884
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7484055757522583,
+      "learning_rate": 6.750432814488145e-05,
+      "loss": 2.9615,
+      "step": 47885
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7243873476982117,
+      "learning_rate": 6.749459838319164e-05,
+      "loss": 2.8934,
+      "step": 47886
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7047001719474792,
+      "learning_rate": 6.748486923387636e-05,
+      "loss": 2.9622,
+      "step": 47887
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6941709518432617,
+      "learning_rate": 6.747514069696104e-05,
+      "loss": 3.0624,
+      "step": 47888
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7410936951637268,
+      "learning_rate": 6.74654127724715e-05,
+      "loss": 2.6276,
+      "step": 47889
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.74365234375,
+      "learning_rate": 6.745568546043338e-05,
+      "loss": 3.0654,
+      "step": 47890
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7380276322364807,
+      "learning_rate": 6.744595876087211e-05,
+      "loss": 2.9712,
+      "step": 47891
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6733323335647583,
+      "learning_rate": 6.743623267381355e-05,
+      "loss": 2.6858,
+      "step": 47892
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7377191781997681,
+      "learning_rate": 6.742650719928317e-05,
+      "loss": 2.7803,
+      "step": 47893
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7515687942504883,
+      "learning_rate": 6.741678233730654e-05,
+      "loss": 3.0201,
+      "step": 47894
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7286888360977173,
+      "learning_rate": 6.740705808790943e-05,
+      "loss": 3.0113,
+      "step": 47895
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7062298059463501,
+      "learning_rate": 6.739733445111728e-05,
+      "loss": 2.7799,
+      "step": 47896
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7720447778701782,
+      "learning_rate": 6.738761142695588e-05,
+      "loss": 2.9522,
+      "step": 47897
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.714799165725708,
+      "learning_rate": 6.737788901545067e-05,
+      "loss": 2.951,
+      "step": 47898
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7219805717468262,
+      "learning_rate": 6.736816721662737e-05,
+      "loss": 2.8175,
+      "step": 47899
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7763769626617432,
+      "learning_rate": 6.735844603051158e-05,
+      "loss": 2.9363,
+      "step": 47900
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7317289113998413,
+      "learning_rate": 6.73487254571288e-05,
+      "loss": 2.9334,
+      "step": 47901
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7248978018760681,
+      "learning_rate": 6.733900549650475e-05,
+      "loss": 3.0164,
+      "step": 47902
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7333472967147827,
+      "learning_rate": 6.732928614866494e-05,
+      "loss": 3.0241,
+      "step": 47903
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7930324673652649,
+      "learning_rate": 6.731956741363498e-05,
+      "loss": 2.8516,
+      "step": 47904
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7575118541717529,
+      "learning_rate": 6.730984929144058e-05,
+      "loss": 2.876,
+      "step": 47905
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7359734773635864,
+      "learning_rate": 6.730013178210717e-05,
+      "loss": 2.8377,
+      "step": 47906
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7023991942405701,
+      "learning_rate": 6.729041488566054e-05,
+      "loss": 2.9254,
+      "step": 47907
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7352596521377563,
+      "learning_rate": 6.728069860212616e-05,
+      "loss": 2.9139,
+      "step": 47908
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6935974359512329,
+      "learning_rate": 6.727098293152952e-05,
+      "loss": 2.9124,
+      "step": 47909
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7225365042686462,
+      "learning_rate": 6.726126787389645e-05,
+      "loss": 2.8963,
+      "step": 47910
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7281026840209961,
+      "learning_rate": 6.725155342925228e-05,
+      "loss": 2.9505,
+      "step": 47911
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7165417075157166,
+      "learning_rate": 6.724183959762285e-05,
+      "loss": 2.8035,
+      "step": 47912
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7453783750534058,
+      "learning_rate": 6.72321263790335e-05,
+      "loss": 2.9435,
+      "step": 47913
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7814623117446899,
+      "learning_rate": 6.722241377351006e-05,
+      "loss": 2.9833,
+      "step": 47914
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7536753416061401,
+      "learning_rate": 6.721270178107794e-05,
+      "loss": 2.8194,
+      "step": 47915
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7346058487892151,
+      "learning_rate": 6.72029904017627e-05,
+      "loss": 2.7661,
+      "step": 47916
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7435902953147888,
+      "learning_rate": 6.719327963559005e-05,
+      "loss": 2.7599,
+      "step": 47917
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7150677442550659,
+      "learning_rate": 6.718356948258543e-05,
+      "loss": 2.984,
+      "step": 47918
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7524111270904541,
+      "learning_rate": 6.717385994277446e-05,
+      "loss": 2.838,
+      "step": 47919
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7796128988265991,
+      "learning_rate": 6.716415101618284e-05,
+      "loss": 2.7455,
+      "step": 47920
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7386764883995056,
+      "learning_rate": 6.715444270283598e-05,
+      "loss": 2.9067,
+      "step": 47921
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7747188806533813,
+      "learning_rate": 6.71447350027596e-05,
+      "loss": 2.8656,
+      "step": 47922
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7116460800170898,
+      "learning_rate": 6.713502791597906e-05,
+      "loss": 2.8489,
+      "step": 47923
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7225497961044312,
+      "learning_rate": 6.712532144252014e-05,
+      "loss": 2.9257,
+      "step": 47924
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7185220122337341,
+      "learning_rate": 6.711561558240818e-05,
+      "loss": 2.9328,
+      "step": 47925
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7107573747634888,
+      "learning_rate": 6.710591033566888e-05,
+      "loss": 2.8896,
+      "step": 47926
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6854578852653503,
+      "learning_rate": 6.709620570232787e-05,
+      "loss": 3.0585,
+      "step": 47927
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7601986527442932,
+      "learning_rate": 6.708650168241055e-05,
+      "loss": 2.7619,
+      "step": 47928
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7219326496124268,
+      "learning_rate": 6.707679827594266e-05,
+      "loss": 3.0518,
+      "step": 47929
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7219486236572266,
+      "learning_rate": 6.706709548294963e-05,
+      "loss": 2.8814,
+      "step": 47930
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7470270395278931,
+      "learning_rate": 6.705739330345699e-05,
+      "loss": 2.8936,
+      "step": 47931
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7633888125419617,
+      "learning_rate": 6.70476917374904e-05,
+      "loss": 3.121,
+      "step": 47932
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7289944291114807,
+      "learning_rate": 6.703799078507527e-05,
+      "loss": 2.9113,
+      "step": 47933
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7318068742752075,
+      "learning_rate": 6.702829044623736e-05,
+      "loss": 2.9518,
+      "step": 47934
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6930985450744629,
+      "learning_rate": 6.701859072100197e-05,
+      "loss": 3.0098,
+      "step": 47935
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.691349983215332,
+      "learning_rate": 6.700889160939488e-05,
+      "loss": 3.0021,
+      "step": 47936
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6962461471557617,
+      "learning_rate": 6.699919311144151e-05,
+      "loss": 2.9648,
+      "step": 47937
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6997219324111938,
+      "learning_rate": 6.698949522716737e-05,
+      "loss": 3.0215,
+      "step": 47938
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7151573896408081,
+      "learning_rate": 6.69797979565981e-05,
+      "loss": 2.9138,
+      "step": 47939
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7079306840896606,
+      "learning_rate": 6.697010129975917e-05,
+      "loss": 2.7507,
+      "step": 47940
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7613853216171265,
+      "learning_rate": 6.69604052566761e-05,
+      "loss": 2.7584,
+      "step": 47941
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7670363187789917,
+      "learning_rate": 6.695070982737456e-05,
+      "loss": 2.9046,
+      "step": 47942
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7608156204223633,
+      "learning_rate": 6.694101501187993e-05,
+      "loss": 2.9756,
+      "step": 47943
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7187494039535522,
+      "learning_rate": 6.693132081021788e-05,
+      "loss": 2.8245,
+      "step": 47944
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.766250729560852,
+      "learning_rate": 6.69216272224139e-05,
+      "loss": 2.7639,
+      "step": 47945
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7323248386383057,
+      "learning_rate": 6.69119342484934e-05,
+      "loss": 2.975,
+      "step": 47946
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7261918187141418,
+      "learning_rate": 6.690224188848208e-05,
+      "loss": 3.0653,
+      "step": 47947
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7363904118537903,
+      "learning_rate": 6.689255014240533e-05,
+      "loss": 2.8412,
+      "step": 47948
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7276012301445007,
+      "learning_rate": 6.688285901028883e-05,
+      "loss": 3.0029,
+      "step": 47949
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7301309108734131,
+      "learning_rate": 6.68731684921579e-05,
+      "loss": 2.9247,
+      "step": 47950
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7242274880409241,
+      "learning_rate": 6.686347858803827e-05,
+      "loss": 2.767,
+      "step": 47951
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7706395387649536,
+      "learning_rate": 6.685378929795541e-05,
+      "loss": 3.0477,
+      "step": 47952
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7164594531059265,
+      "learning_rate": 6.684410062193467e-05,
+      "loss": 2.8957,
+      "step": 47953
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7162006497383118,
+      "learning_rate": 6.683441256000182e-05,
+      "loss": 2.9765,
+      "step": 47954
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7521647214889526,
+      "learning_rate": 6.682472511218213e-05,
+      "loss": 2.8028,
+      "step": 47955
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7329918146133423,
+      "learning_rate": 6.681503827850124e-05,
+      "loss": 3.0094,
+      "step": 47956
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7082628011703491,
+      "learning_rate": 6.680535205898476e-05,
+      "loss": 3.1608,
+      "step": 47957
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.721556544303894,
+      "learning_rate": 6.679566645365813e-05,
+      "loss": 2.7341,
+      "step": 47958
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7148202657699585,
+      "learning_rate": 6.67859814625468e-05,
+      "loss": 3.0916,
+      "step": 47959
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7291754484176636,
+      "learning_rate": 6.677629708567624e-05,
+      "loss": 2.8334,
+      "step": 47960
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.8292950391769409,
+      "learning_rate": 6.676661332307201e-05,
+      "loss": 3.0277,
+      "step": 47961
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7232643365859985,
+      "learning_rate": 6.675693017475972e-05,
+      "loss": 2.8574,
+      "step": 47962
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7282116413116455,
+      "learning_rate": 6.674724764076471e-05,
+      "loss": 3.191,
+      "step": 47963
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7802262902259827,
+      "learning_rate": 6.673756572111268e-05,
+      "loss": 2.9908,
+      "step": 47964
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.732752799987793,
+      "learning_rate": 6.672788441582886e-05,
+      "loss": 2.8625,
+      "step": 47965
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7601721286773682,
+      "learning_rate": 6.671820372493901e-05,
+      "loss": 2.9282,
+      "step": 47966
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7560071349143982,
+      "learning_rate": 6.670852364846855e-05,
+      "loss": 3.0195,
+      "step": 47967
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.694574236869812,
+      "learning_rate": 6.669884418644281e-05,
+      "loss": 3.183,
+      "step": 47968
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7665233016014099,
+      "learning_rate": 6.668916533888752e-05,
+      "loss": 3.1757,
+      "step": 47969
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7764652371406555,
+      "learning_rate": 6.667948710582798e-05,
+      "loss": 2.9517,
+      "step": 47970
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7054910063743591,
+      "learning_rate": 6.666980948728978e-05,
+      "loss": 2.9967,
+      "step": 47971
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7566725611686707,
+      "learning_rate": 6.666013248329845e-05,
+      "loss": 2.9945,
+      "step": 47972
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7435179948806763,
+      "learning_rate": 6.665045609387944e-05,
+      "loss": 2.9586,
+      "step": 47973
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6983252167701721,
+      "learning_rate": 6.664078031905822e-05,
+      "loss": 2.8248,
+      "step": 47974
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7251109480857849,
+      "learning_rate": 6.663110515886021e-05,
+      "loss": 3.0567,
+      "step": 47975
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7957937717437744,
+      "learning_rate": 6.662143061331093e-05,
+      "loss": 2.6457,
+      "step": 47976
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7234058380126953,
+      "learning_rate": 6.661175668243596e-05,
+      "loss": 3.1194,
+      "step": 47977
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7288504242897034,
+      "learning_rate": 6.660208336626066e-05,
+      "loss": 2.9923,
+      "step": 47978
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7363245487213135,
+      "learning_rate": 6.659241066481063e-05,
+      "loss": 3.0909,
+      "step": 47979
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7586546540260315,
+      "learning_rate": 6.658273857811123e-05,
+      "loss": 2.919,
+      "step": 47980
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7015671730041504,
+      "learning_rate": 6.657306710618794e-05,
+      "loss": 2.8089,
+      "step": 47981
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7294623851776123,
+      "learning_rate": 6.656339624906633e-05,
+      "loss": 3.0357,
+      "step": 47982
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7198174595832825,
+      "learning_rate": 6.655372600677171e-05,
+      "loss": 2.9374,
+      "step": 47983
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7496002316474915,
+      "learning_rate": 6.654405637932976e-05,
+      "loss": 2.7852,
+      "step": 47984
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.71015465259552,
+      "learning_rate": 6.653438736676575e-05,
+      "loss": 2.9953,
+      "step": 47985
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7362428307533264,
+      "learning_rate": 6.65247189691052e-05,
+      "loss": 2.6275,
+      "step": 47986
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7148375511169434,
+      "learning_rate": 6.651505118637374e-05,
+      "loss": 2.952,
+      "step": 47987
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7447728514671326,
+      "learning_rate": 6.650538401859665e-05,
+      "loss": 3.0321,
+      "step": 47988
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.716418981552124,
+      "learning_rate": 6.649571746579946e-05,
+      "loss": 2.7179,
+      "step": 47989
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7445757389068604,
+      "learning_rate": 6.648605152800753e-05,
+      "loss": 2.9258,
+      "step": 47990
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7147215008735657,
+      "learning_rate": 6.647638620524642e-05,
+      "loss": 2.9995,
+      "step": 47991
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7629702091217041,
+      "learning_rate": 6.646672149754161e-05,
+      "loss": 2.941,
+      "step": 47992
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6876111030578613,
+      "learning_rate": 6.645705740491847e-05,
+      "loss": 2.9911,
+      "step": 47993
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7215240597724915,
+      "learning_rate": 6.644739392740255e-05,
+      "loss": 2.8185,
+      "step": 47994
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7500731348991394,
+      "learning_rate": 6.643773106501925e-05,
+      "loss": 2.9694,
+      "step": 47995
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7356647253036499,
+      "learning_rate": 6.642806881779392e-05,
+      "loss": 2.8889,
+      "step": 47996
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7396929860115051,
+      "learning_rate": 6.641840718575221e-05,
+      "loss": 2.9491,
+      "step": 47997
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7047430276870728,
+      "learning_rate": 6.640874616891938e-05,
+      "loss": 2.9185,
+      "step": 47998
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7342384457588196,
+      "learning_rate": 6.639908576732105e-05,
+      "loss": 2.8062,
+      "step": 47999
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7048220634460449,
+      "learning_rate": 6.638942598098245e-05,
+      "loss": 3.0256,
+      "step": 48000
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6992772817611694,
+      "learning_rate": 6.637976680992926e-05,
+      "loss": 2.8008,
+      "step": 48001
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7305018305778503,
+      "learning_rate": 6.63701082541868e-05,
+      "loss": 2.9617,
+      "step": 48002
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7258499264717102,
+      "learning_rate": 6.636045031378041e-05,
+      "loss": 2.9443,
+      "step": 48003
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7828282713890076,
+      "learning_rate": 6.635079298873572e-05,
+      "loss": 3.0035,
+      "step": 48004
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7275944948196411,
+      "learning_rate": 6.634113627907798e-05,
+      "loss": 2.8979,
+      "step": 48005
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7850326299667358,
+      "learning_rate": 6.633148018483281e-05,
+      "loss": 3.0463,
+      "step": 48006
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7134331464767456,
+      "learning_rate": 6.63218247060255e-05,
+      "loss": 2.8976,
+      "step": 48007
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7655074596405029,
+      "learning_rate": 6.631216984268149e-05,
+      "loss": 2.9853,
+      "step": 48008
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7371658682823181,
+      "learning_rate": 6.630251559482633e-05,
+      "loss": 2.8503,
+      "step": 48009
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7096714973449707,
+      "learning_rate": 6.629286196248539e-05,
+      "loss": 2.905,
+      "step": 48010
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7324158549308777,
+      "learning_rate": 6.628320894568407e-05,
+      "loss": 3.0072,
+      "step": 48011
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7133394479751587,
+      "learning_rate": 6.627355654444774e-05,
+      "loss": 2.8081,
+      "step": 48012
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7191671133041382,
+      "learning_rate": 6.626390475880183e-05,
+      "loss": 2.7956,
+      "step": 48013
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7937895655632019,
+      "learning_rate": 6.625425358877194e-05,
+      "loss": 2.8923,
+      "step": 48014
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7353552579879761,
+      "learning_rate": 6.624460303438325e-05,
+      "loss": 2.9375,
+      "step": 48015
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7080419063568115,
+      "learning_rate": 6.623495309566138e-05,
+      "loss": 2.9092,
+      "step": 48016
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7076489925384521,
+      "learning_rate": 6.622530377263164e-05,
+      "loss": 3.1517,
+      "step": 48017
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7452599406242371,
+      "learning_rate": 6.621565506531939e-05,
+      "loss": 2.9966,
+      "step": 48018
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7066365480422974,
+      "learning_rate": 6.620600697375019e-05,
+      "loss": 2.7327,
+      "step": 48019
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6905827522277832,
+      "learning_rate": 6.619635949794929e-05,
+      "loss": 3.0837,
+      "step": 48020
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7231424450874329,
+      "learning_rate": 6.618671263794226e-05,
+      "loss": 3.1329,
+      "step": 48021
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7269807457923889,
+      "learning_rate": 6.617706639375437e-05,
+      "loss": 2.7666,
+      "step": 48022
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7062513828277588,
+      "learning_rate": 6.616742076541117e-05,
+      "loss": 2.7635,
+      "step": 48023
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7632284760475159,
+      "learning_rate": 6.615777575293797e-05,
+      "loss": 2.9912,
+      "step": 48024
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7120453715324402,
+      "learning_rate": 6.614813135636008e-05,
+      "loss": 3.2382,
+      "step": 48025
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7771921157836914,
+      "learning_rate": 6.61384875757031e-05,
+      "loss": 2.8899,
+      "step": 48026
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.71280837059021,
+      "learning_rate": 6.612884441099229e-05,
+      "loss": 2.8302,
+      "step": 48027
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7356342077255249,
+      "learning_rate": 6.611920186225305e-05,
+      "loss": 3.1114,
+      "step": 48028
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7350482940673828,
+      "learning_rate": 6.610955992951092e-05,
+      "loss": 3.127,
+      "step": 48029
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7893431782722473,
+      "learning_rate": 6.60999186127911e-05,
+      "loss": 2.8438,
+      "step": 48030
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.716373085975647,
+      "learning_rate": 6.609027791211919e-05,
+      "loss": 2.901,
+      "step": 48031
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7196996212005615,
+      "learning_rate": 6.608063782752045e-05,
+      "loss": 2.7733,
+      "step": 48032
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7472556829452515,
+      "learning_rate": 6.607099835902023e-05,
+      "loss": 3.0575,
+      "step": 48033
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7139104604721069,
+      "learning_rate": 6.606135950664403e-05,
+      "loss": 2.9093,
+      "step": 48034
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.8291552662849426,
+      "learning_rate": 6.605172127041715e-05,
+      "loss": 2.989,
+      "step": 48035
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7301540970802307,
+      "learning_rate": 6.604208365036507e-05,
+      "loss": 2.89,
+      "step": 48036
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7189569473266602,
+      "learning_rate": 6.603244664651301e-05,
+      "loss": 2.8514,
+      "step": 48037
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7034281492233276,
+      "learning_rate": 6.602281025888656e-05,
+      "loss": 2.8714,
+      "step": 48038
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7544010281562805,
+      "learning_rate": 6.601317448751101e-05,
+      "loss": 2.8995,
+      "step": 48039
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6748170256614685,
+      "learning_rate": 6.600353933241164e-05,
+      "loss": 2.6243,
+      "step": 48040
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.685340940952301,
+      "learning_rate": 6.599390479361398e-05,
+      "loss": 2.7344,
+      "step": 48041
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6887692213058472,
+      "learning_rate": 6.598427087114326e-05,
+      "loss": 3.0116,
+      "step": 48042
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7316964864730835,
+      "learning_rate": 6.597463756502492e-05,
+      "loss": 3.0474,
+      "step": 48043
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7430610060691833,
+      "learning_rate": 6.596500487528447e-05,
+      "loss": 2.7318,
+      "step": 48044
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.71198570728302,
+      "learning_rate": 6.595537280194704e-05,
+      "loss": 2.9316,
+      "step": 48045
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.720703661441803,
+      "learning_rate": 6.594574134503822e-05,
+      "loss": 3.1858,
+      "step": 48046
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7150319218635559,
+      "learning_rate": 6.593611050458325e-05,
+      "loss": 2.8819,
+      "step": 48047
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7347413897514343,
+      "learning_rate": 6.592648028060746e-05,
+      "loss": 3.1495,
+      "step": 48048
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7172079682350159,
+      "learning_rate": 6.591685067313634e-05,
+      "loss": 2.856,
+      "step": 48049
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7229471802711487,
+      "learning_rate": 6.59072216821951e-05,
+      "loss": 2.9738,
+      "step": 48050
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7183575630187988,
+      "learning_rate": 6.589759330780926e-05,
+      "loss": 3.0105,
+      "step": 48051
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7318481802940369,
+      "learning_rate": 6.588796555000403e-05,
+      "loss": 2.8139,
+      "step": 48052
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7228243350982666,
+      "learning_rate": 6.587833840880493e-05,
+      "loss": 2.8365,
+      "step": 48053
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7609462738037109,
+      "learning_rate": 6.586871188423724e-05,
+      "loss": 2.6898,
+      "step": 48054
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7056893110275269,
+      "learning_rate": 6.585908597632622e-05,
+      "loss": 2.9007,
+      "step": 48055
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7319220304489136,
+      "learning_rate": 6.584946068509737e-05,
+      "loss": 2.9661,
+      "step": 48056
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7054177522659302,
+      "learning_rate": 6.583983601057593e-05,
+      "loss": 2.7722,
+      "step": 48057
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7039554119110107,
+      "learning_rate": 6.583021195278726e-05,
+      "loss": 2.7155,
+      "step": 48058
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7091907858848572,
+      "learning_rate": 6.582058851175683e-05,
+      "loss": 3.0179,
+      "step": 48059
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7543073892593384,
+      "learning_rate": 6.581096568750993e-05,
+      "loss": 2.9976,
+      "step": 48060
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7862129211425781,
+      "learning_rate": 6.580134348007183e-05,
+      "loss": 2.6822,
+      "step": 48061
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7358943223953247,
+      "learning_rate": 6.579172188946789e-05,
+      "loss": 2.8732,
+      "step": 48062
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7497488260269165,
+      "learning_rate": 6.578210091572344e-05,
+      "loss": 2.9566,
+      "step": 48063
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6996098160743713,
+      "learning_rate": 6.577248055886396e-05,
+      "loss": 3.085,
+      "step": 48064
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7074849009513855,
+      "learning_rate": 6.576286081891462e-05,
+      "loss": 2.998,
+      "step": 48065
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7271658182144165,
+      "learning_rate": 6.575324169590086e-05,
+      "loss": 3.0539,
+      "step": 48066
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8061689138412476,
+      "learning_rate": 6.574362318984792e-05,
+      "loss": 2.6798,
+      "step": 48067
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7243710160255432,
+      "learning_rate": 6.573400530078129e-05,
+      "loss": 2.8028,
+      "step": 48068
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7632570862770081,
+      "learning_rate": 6.57243880287262e-05,
+      "loss": 2.8476,
+      "step": 48069
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7301486134529114,
+      "learning_rate": 6.571477137370788e-05,
+      "loss": 2.9998,
+      "step": 48070
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7374549508094788,
+      "learning_rate": 6.570515533575188e-05,
+      "loss": 3.0035,
+      "step": 48071
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7524304986000061,
+      "learning_rate": 6.569553991488332e-05,
+      "loss": 2.8548,
+      "step": 48072
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.706081748008728,
+      "learning_rate": 6.568592511112759e-05,
+      "loss": 2.8905,
+      "step": 48073
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7375027537345886,
+      "learning_rate": 6.567631092451014e-05,
+      "loss": 3.1871,
+      "step": 48074
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7398857474327087,
+      "learning_rate": 6.566669735505619e-05,
+      "loss": 2.9136,
+      "step": 48075
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7311112880706787,
+      "learning_rate": 6.565708440279106e-05,
+      "loss": 2.6949,
+      "step": 48076
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7611480355262756,
+      "learning_rate": 6.564747206773996e-05,
+      "loss": 3.016,
+      "step": 48077
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7483656406402588,
+      "learning_rate": 6.563786034992843e-05,
+      "loss": 3.0053,
+      "step": 48078
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7387386560440063,
+      "learning_rate": 6.562824924938162e-05,
+      "loss": 2.9809,
+      "step": 48079
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.745112955570221,
+      "learning_rate": 6.561863876612484e-05,
+      "loss": 2.9586,
+      "step": 48080
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7196276783943176,
+      "learning_rate": 6.560902890018357e-05,
+      "loss": 3.2109,
+      "step": 48081
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7283993363380432,
+      "learning_rate": 6.559941965158297e-05,
+      "loss": 3.1374,
+      "step": 48082
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7114156484603882,
+      "learning_rate": 6.558981102034844e-05,
+      "loss": 3.015,
+      "step": 48083
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7603095769882202,
+      "learning_rate": 6.55802030065051e-05,
+      "loss": 2.9397,
+      "step": 48084
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6935713291168213,
+      "learning_rate": 6.557059561007843e-05,
+      "loss": 3.1052,
+      "step": 48085
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6886460185050964,
+      "learning_rate": 6.556098883109378e-05,
+      "loss": 2.9214,
+      "step": 48086
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7370249032974243,
+      "learning_rate": 6.555138266957627e-05,
+      "loss": 2.9907,
+      "step": 48087
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7281854748725891,
+      "learning_rate": 6.554177712555138e-05,
+      "loss": 2.7215,
+      "step": 48088
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7895329594612122,
+      "learning_rate": 6.553217219904424e-05,
+      "loss": 2.8622,
+      "step": 48089
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7070068717002869,
+      "learning_rate": 6.552256789008031e-05,
+      "loss": 2.8635,
+      "step": 48090
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8132027387619019,
+      "learning_rate": 6.551296419868481e-05,
+      "loss": 2.7608,
+      "step": 48091
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7146351933479309,
+      "learning_rate": 6.5503361124883e-05,
+      "loss": 2.5896,
+      "step": 48092
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7069380879402161,
+      "learning_rate": 6.549375866870023e-05,
+      "loss": 3.038,
+      "step": 48093
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7371593713760376,
+      "learning_rate": 6.548415683016169e-05,
+      "loss": 2.9843,
+      "step": 48094
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7395793795585632,
+      "learning_rate": 6.547455560929278e-05,
+      "loss": 2.6887,
+      "step": 48095
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7200589179992676,
+      "learning_rate": 6.546495500611882e-05,
+      "loss": 2.8007,
+      "step": 48096
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7023578882217407,
+      "learning_rate": 6.545535502066505e-05,
+      "loss": 2.632,
+      "step": 48097
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7165631055831909,
+      "learning_rate": 6.544575565295672e-05,
+      "loss": 2.9052,
+      "step": 48098
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7894178032875061,
+      "learning_rate": 6.543615690301903e-05,
+      "loss": 2.7053,
+      "step": 48099
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6898087859153748,
+      "learning_rate": 6.54265587708774e-05,
+      "loss": 2.7166,
+      "step": 48100
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7402461171150208,
+      "learning_rate": 6.541696125655711e-05,
+      "loss": 3.0452,
+      "step": 48101
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7221176028251648,
+      "learning_rate": 6.540736436008333e-05,
+      "loss": 2.9999,
+      "step": 48102
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7226662039756775,
+      "learning_rate": 6.539776808148148e-05,
+      "loss": 2.8164,
+      "step": 48103
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7799202799797058,
+      "learning_rate": 6.538817242077676e-05,
+      "loss": 2.9798,
+      "step": 48104
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7280610203742981,
+      "learning_rate": 6.537857737799438e-05,
+      "loss": 2.9594,
+      "step": 48105
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7227134108543396,
+      "learning_rate": 6.536898295315975e-05,
+      "loss": 2.6752,
+      "step": 48106
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6850351691246033,
+      "learning_rate": 6.535938914629797e-05,
+      "loss": 2.8372,
+      "step": 48107
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.716296911239624,
+      "learning_rate": 6.534979595743449e-05,
+      "loss": 2.9671,
+      "step": 48108
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6905850172042847,
+      "learning_rate": 6.534020338659439e-05,
+      "loss": 2.728,
+      "step": 48109
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7352511882781982,
+      "learning_rate": 6.533061143380304e-05,
+      "loss": 2.9374,
+      "step": 48110
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7677817940711975,
+      "learning_rate": 6.53210200990858e-05,
+      "loss": 2.7527,
+      "step": 48111
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7257785201072693,
+      "learning_rate": 6.531142938246781e-05,
+      "loss": 3.1291,
+      "step": 48112
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7301912903785706,
+      "learning_rate": 6.530183928397435e-05,
+      "loss": 2.7511,
+      "step": 48113
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7411414384841919,
+      "learning_rate": 6.529224980363062e-05,
+      "loss": 2.822,
+      "step": 48114
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7773602604866028,
+      "learning_rate": 6.528266094146191e-05,
+      "loss": 2.8123,
+      "step": 48115
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7311223149299622,
+      "learning_rate": 6.527307269749358e-05,
+      "loss": 2.9162,
+      "step": 48116
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7010335922241211,
+      "learning_rate": 6.526348507175073e-05,
+      "loss": 2.9099,
+      "step": 48117
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7506341338157654,
+      "learning_rate": 6.525389806425875e-05,
+      "loss": 2.963,
+      "step": 48118
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7527403235435486,
+      "learning_rate": 6.524431167504283e-05,
+      "loss": 2.9887,
+      "step": 48119
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7092116475105286,
+      "learning_rate": 6.523472590412814e-05,
+      "loss": 2.7487,
+      "step": 48120
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7055248618125916,
+      "learning_rate": 6.522514075154006e-05,
+      "loss": 3.0072,
+      "step": 48121
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7026886940002441,
+      "learning_rate": 6.521555621730372e-05,
+      "loss": 2.7419,
+      "step": 48122
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6752901673316956,
+      "learning_rate": 6.52059723014445e-05,
+      "loss": 2.918,
+      "step": 48123
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.715085506439209,
+      "learning_rate": 6.519638900398747e-05,
+      "loss": 2.9607,
+      "step": 48124
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7532249093055725,
+      "learning_rate": 6.518680632495805e-05,
+      "loss": 2.8491,
+      "step": 48125
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7319445610046387,
+      "learning_rate": 6.517722426438139e-05,
+      "loss": 2.9336,
+      "step": 48126
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7512108683586121,
+      "learning_rate": 6.516764282228264e-05,
+      "loss": 2.6582,
+      "step": 48127
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7304036021232605,
+      "learning_rate": 6.515806199868722e-05,
+      "loss": 3.0088,
+      "step": 48128
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7679119110107422,
+      "learning_rate": 6.514848179362017e-05,
+      "loss": 2.9008,
+      "step": 48129
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6933061480522156,
+      "learning_rate": 6.513890220710682e-05,
+      "loss": 3.1265,
+      "step": 48130
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7610794305801392,
+      "learning_rate": 6.512932323917251e-05,
+      "loss": 2.8423,
+      "step": 48131
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7362642288208008,
+      "learning_rate": 6.511974488984226e-05,
+      "loss": 2.924,
+      "step": 48132
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7452098727226257,
+      "learning_rate": 6.511016715914147e-05,
+      "loss": 2.9748,
+      "step": 48133
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7530220150947571,
+      "learning_rate": 6.51005900470953e-05,
+      "loss": 3.0002,
+      "step": 48134
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8055080771446228,
+      "learning_rate": 6.50910135537289e-05,
+      "loss": 3.0089,
+      "step": 48135
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.722628653049469,
+      "learning_rate": 6.508143767906761e-05,
+      "loss": 2.9678,
+      "step": 48136
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7451878190040588,
+      "learning_rate": 6.507186242313657e-05,
+      "loss": 2.8849,
+      "step": 48137
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7400360107421875,
+      "learning_rate": 6.50622877859611e-05,
+      "loss": 2.9511,
+      "step": 48138
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7950756549835205,
+      "learning_rate": 6.505271376756625e-05,
+      "loss": 2.7903,
+      "step": 48139
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7754967212677002,
+      "learning_rate": 6.504314036797742e-05,
+      "loss": 2.8494,
+      "step": 48140
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.719846785068512,
+      "learning_rate": 6.503356758721974e-05,
+      "loss": 3.1789,
+      "step": 48141
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7460191249847412,
+      "learning_rate": 6.502399542531835e-05,
+      "loss": 2.7791,
+      "step": 48142
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6921855211257935,
+      "learning_rate": 6.501442388229862e-05,
+      "loss": 2.8102,
+      "step": 48143
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8616728782653809,
+      "learning_rate": 6.500485295818557e-05,
+      "loss": 2.8678,
+      "step": 48144
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7126127481460571,
+      "learning_rate": 6.499528265300453e-05,
+      "loss": 3.0098,
+      "step": 48145
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.723812460899353,
+      "learning_rate": 6.498571296678075e-05,
+      "loss": 2.8521,
+      "step": 48146
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7311473488807678,
+      "learning_rate": 6.497614389953932e-05,
+      "loss": 3.0363,
+      "step": 48147
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7247294783592224,
+      "learning_rate": 6.496657545130565e-05,
+      "loss": 2.7482,
+      "step": 48148
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7272924184799194,
+      "learning_rate": 6.495700762210459e-05,
+      "loss": 3.0696,
+      "step": 48149
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7352733016014099,
+      "learning_rate": 6.494744041196158e-05,
+      "loss": 2.9417,
+      "step": 48150
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7280297875404358,
+      "learning_rate": 6.493787382090186e-05,
+      "loss": 3.0833,
+      "step": 48151
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6772411465644836,
+      "learning_rate": 6.492830784895042e-05,
+      "loss": 3.0068,
+      "step": 48152
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7198385000228882,
+      "learning_rate": 6.491874249613267e-05,
+      "loss": 3.0322,
+      "step": 48153
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7675883173942566,
+      "learning_rate": 6.490917776247364e-05,
+      "loss": 3.009,
+      "step": 48154
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7238621115684509,
+      "learning_rate": 6.489961364799868e-05,
+      "loss": 2.8473,
+      "step": 48155
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.804047703742981,
+      "learning_rate": 6.489005015273288e-05,
+      "loss": 2.7081,
+      "step": 48156
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.688144862651825,
+      "learning_rate": 6.488048727670136e-05,
+      "loss": 3.1035,
+      "step": 48157
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8125393390655518,
+      "learning_rate": 6.487092501992946e-05,
+      "loss": 2.8867,
+      "step": 48158
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7027606964111328,
+      "learning_rate": 6.486136338244222e-05,
+      "loss": 3.0313,
+      "step": 48159
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7894521355628967,
+      "learning_rate": 6.485180236426499e-05,
+      "loss": 2.9952,
+      "step": 48160
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7138534188270569,
+      "learning_rate": 6.484224196542276e-05,
+      "loss": 2.9025,
+      "step": 48161
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7329240441322327,
+      "learning_rate": 6.48326821859409e-05,
+      "loss": 2.9142,
+      "step": 48162
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7437616586685181,
+      "learning_rate": 6.482312302584448e-05,
+      "loss": 2.9068,
+      "step": 48163
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7317550182342529,
+      "learning_rate": 6.481356448515863e-05,
+      "loss": 2.7661,
+      "step": 48164
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7197153568267822,
+      "learning_rate": 6.480400656390866e-05,
+      "loss": 2.8229,
+      "step": 48165
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7194134593009949,
+      "learning_rate": 6.479444926211959e-05,
+      "loss": 2.7384,
+      "step": 48166
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7750094532966614,
+      "learning_rate": 6.478489257981667e-05,
+      "loss": 2.7421,
+      "step": 48167
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7340207695960999,
+      "learning_rate": 6.477533651702517e-05,
+      "loss": 2.7373,
+      "step": 48168
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7263909578323364,
+      "learning_rate": 6.476578107377008e-05,
+      "loss": 2.9162,
+      "step": 48169
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7772607207298279,
+      "learning_rate": 6.475622625007674e-05,
+      "loss": 3.024,
+      "step": 48170
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7200969457626343,
+      "learning_rate": 6.47466720459702e-05,
+      "loss": 3.0136,
+      "step": 48171
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7096529006958008,
+      "learning_rate": 6.473711846147557e-05,
+      "loss": 3.0276,
+      "step": 48172
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7910388708114624,
+      "learning_rate": 6.47275654966182e-05,
+      "loss": 2.8153,
+      "step": 48173
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.753176212310791,
+      "learning_rate": 6.471801315142304e-05,
+      "loss": 2.9071,
+      "step": 48174
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7686089277267456,
+      "learning_rate": 6.470846142591545e-05,
+      "loss": 2.9368,
+      "step": 48175
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7298809885978699,
+      "learning_rate": 6.46989103201204e-05,
+      "loss": 2.8194,
+      "step": 48176
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7550878524780273,
+      "learning_rate": 6.46893598340632e-05,
+      "loss": 3.235,
+      "step": 48177
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7033806443214417,
+      "learning_rate": 6.467980996776898e-05,
+      "loss": 2.869,
+      "step": 48178
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.743836522102356,
+      "learning_rate": 6.467026072126275e-05,
+      "loss": 2.7897,
+      "step": 48179
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7496225833892822,
+      "learning_rate": 6.466071209456984e-05,
+      "loss": 2.9425,
+      "step": 48180
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7606794238090515,
+      "learning_rate": 6.465116408771526e-05,
+      "loss": 2.9058,
+      "step": 48181
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7309315204620361,
+      "learning_rate": 6.46416167007242e-05,
+      "loss": 2.6442,
+      "step": 48182
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7935462594032288,
+      "learning_rate": 6.463206993362191e-05,
+      "loss": 2.9268,
+      "step": 48183
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7384037971496582,
+      "learning_rate": 6.462252378643345e-05,
+      "loss": 3.1264,
+      "step": 48184
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7172885537147522,
+      "learning_rate": 6.461297825918394e-05,
+      "loss": 2.9865,
+      "step": 48185
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7741020321846008,
+      "learning_rate": 6.460343335189849e-05,
+      "loss": 2.7657,
+      "step": 48186
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.742149293422699,
+      "learning_rate": 6.45938890646023e-05,
+      "loss": 2.8609,
+      "step": 48187
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6724493503570557,
+      "learning_rate": 6.458434539732058e-05,
+      "loss": 2.9769,
+      "step": 48188
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7558503746986389,
+      "learning_rate": 6.457480235007829e-05,
+      "loss": 2.8232,
+      "step": 48189
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7640053033828735,
+      "learning_rate": 6.456525992290075e-05,
+      "loss": 2.8353,
+      "step": 48190
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7217065095901489,
+      "learning_rate": 6.455571811581293e-05,
+      "loss": 2.814,
+      "step": 48191
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7169857025146484,
+      "learning_rate": 6.454617692884014e-05,
+      "loss": 2.9869,
+      "step": 48192
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7590023279190063,
+      "learning_rate": 6.453663636200738e-05,
+      "loss": 2.8972,
+      "step": 48193
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6975011825561523,
+      "learning_rate": 6.45270964153397e-05,
+      "loss": 3.0994,
+      "step": 48194
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7145510911941528,
+      "learning_rate": 6.451755708886243e-05,
+      "loss": 2.8702,
+      "step": 48195
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7827848196029663,
+      "learning_rate": 6.450801838260055e-05,
+      "loss": 3.2263,
+      "step": 48196
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7232081890106201,
+      "learning_rate": 6.449848029657918e-05,
+      "loss": 2.9435,
+      "step": 48197
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7153977155685425,
+      "learning_rate": 6.448894283082358e-05,
+      "loss": 2.8311,
+      "step": 48198
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7476628422737122,
+      "learning_rate": 6.44794059853588e-05,
+      "loss": 2.9139,
+      "step": 48199
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7741516828536987,
+      "learning_rate": 6.446986976020994e-05,
+      "loss": 2.8474,
+      "step": 48200
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7252960801124573,
+      "learning_rate": 6.4460334155402e-05,
+      "loss": 3.0896,
+      "step": 48201
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.711409866809845,
+      "learning_rate": 6.445079917096024e-05,
+      "loss": 2.9098,
+      "step": 48202
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6946399211883545,
+      "learning_rate": 6.44412648069098e-05,
+      "loss": 2.8107,
+      "step": 48203
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7085345983505249,
+      "learning_rate": 6.443173106327568e-05,
+      "loss": 3.0131,
+      "step": 48204
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7308792471885681,
+      "learning_rate": 6.44221979400831e-05,
+      "loss": 3.0106,
+      "step": 48205
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7140591144561768,
+      "learning_rate": 6.441266543735712e-05,
+      "loss": 3.0825,
+      "step": 48206
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7082121968269348,
+      "learning_rate": 6.440313355512277e-05,
+      "loss": 2.9419,
+      "step": 48207
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7402740120887756,
+      "learning_rate": 6.43936022934053e-05,
+      "loss": 3.0304,
+      "step": 48208
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7371035814285278,
+      "learning_rate": 6.438407165222964e-05,
+      "loss": 2.9194,
+      "step": 48209
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.695787250995636,
+      "learning_rate": 6.437454163162108e-05,
+      "loss": 3.0194,
+      "step": 48210
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7401152849197388,
+      "learning_rate": 6.436501223160456e-05,
+      "loss": 3.0295,
+      "step": 48211
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7056068778038025,
+      "learning_rate": 6.435548345220523e-05,
+      "loss": 3.2774,
+      "step": 48212
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7429165840148926,
+      "learning_rate": 6.434595529344832e-05,
+      "loss": 2.6779,
+      "step": 48213
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7436327338218689,
+      "learning_rate": 6.43364277553588e-05,
+      "loss": 2.8903,
+      "step": 48214
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7075293064117432,
+      "learning_rate": 6.432690083796177e-05,
+      "loss": 2.9532,
+      "step": 48215
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.728111982345581,
+      "learning_rate": 6.431737454128224e-05,
+      "loss": 2.8952,
+      "step": 48216
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7516396045684814,
+      "learning_rate": 6.43078488653454e-05,
+      "loss": 3.0249,
+      "step": 48217
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7346088290214539,
+      "learning_rate": 6.42983238101764e-05,
+      "loss": 2.7594,
+      "step": 48218
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7349608540534973,
+      "learning_rate": 6.428879937580015e-05,
+      "loss": 2.7937,
+      "step": 48219
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7432407140731812,
+      "learning_rate": 6.427927556224196e-05,
+      "loss": 2.8822,
+      "step": 48220
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7252591848373413,
+      "learning_rate": 6.426975236952675e-05,
+      "loss": 2.7845,
+      "step": 48221
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6931779384613037,
+      "learning_rate": 6.42602297976796e-05,
+      "loss": 2.9773,
+      "step": 48222
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7488669157028198,
+      "learning_rate": 6.425070784672571e-05,
+      "loss": 2.9075,
+      "step": 48223
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7043687105178833,
+      "learning_rate": 6.424118651668998e-05,
+      "loss": 3.07,
+      "step": 48224
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.726883053779602,
+      "learning_rate": 6.423166580759771e-05,
+      "loss": 2.6815,
+      "step": 48225
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7561427354812622,
+      "learning_rate": 6.422214571947376e-05,
+      "loss": 2.7455,
+      "step": 48226
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6870282292366028,
+      "learning_rate": 6.421262625234337e-05,
+      "loss": 2.6591,
+      "step": 48227
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7265474200248718,
+      "learning_rate": 6.420310740623154e-05,
+      "loss": 2.7049,
+      "step": 48228
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7331658601760864,
+      "learning_rate": 6.419358918116329e-05,
+      "loss": 2.8969,
+      "step": 48229
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7479391098022461,
+      "learning_rate": 6.418407157716381e-05,
+      "loss": 3.0111,
+      "step": 48230
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.738805890083313,
+      "learning_rate": 6.417455459425804e-05,
+      "loss": 2.7992,
+      "step": 48231
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7408674955368042,
+      "learning_rate": 6.416503823247115e-05,
+      "loss": 2.7176,
+      "step": 48232
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7549987435340881,
+      "learning_rate": 6.415552249182812e-05,
+      "loss": 2.8767,
+      "step": 48233
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7115577459335327,
+      "learning_rate": 6.414600737235402e-05,
+      "loss": 2.8056,
+      "step": 48234
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7305412292480469,
+      "learning_rate": 6.413649287407406e-05,
+      "loss": 2.9721,
+      "step": 48235
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7211993336677551,
+      "learning_rate": 6.412697899701315e-05,
+      "loss": 2.8724,
+      "step": 48236
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7803027629852295,
+      "learning_rate": 6.411746574119642e-05,
+      "loss": 2.7537,
+      "step": 48237
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7594745755195618,
+      "learning_rate": 6.410795310664878e-05,
+      "loss": 3.07,
+      "step": 48238
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7324116230010986,
+      "learning_rate": 6.409844109339542e-05,
+      "loss": 2.8141,
+      "step": 48239
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6955828070640564,
+      "learning_rate": 6.408892970146143e-05,
+      "loss": 2.7801,
+      "step": 48240
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7122442722320557,
+      "learning_rate": 6.407941893087172e-05,
+      "loss": 2.8966,
+      "step": 48241
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7142869234085083,
+      "learning_rate": 6.40699087816515e-05,
+      "loss": 2.817,
+      "step": 48242
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7723212838172913,
+      "learning_rate": 6.40603992538257e-05,
+      "loss": 2.9792,
+      "step": 48243
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7571322917938232,
+      "learning_rate": 6.405089034741937e-05,
+      "loss": 2.9604,
+      "step": 48244
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.730701744556427,
+      "learning_rate": 6.404138206245765e-05,
+      "loss": 2.9751,
+      "step": 48245
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7998728156089783,
+      "learning_rate": 6.403187439896541e-05,
+      "loss": 3.1012,
+      "step": 48246
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6990532279014587,
+      "learning_rate": 6.402236735696793e-05,
+      "loss": 2.839,
+      "step": 48247
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7441544532775879,
+      "learning_rate": 6.401286093648998e-05,
+      "loss": 2.9951,
+      "step": 48248
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6886500716209412,
+      "learning_rate": 6.400335513755679e-05,
+      "loss": 2.8154,
+      "step": 48249
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7198939323425293,
+      "learning_rate": 6.399384996019349e-05,
+      "loss": 3.056,
+      "step": 48250
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7159644961357117,
+      "learning_rate": 6.398434540442479e-05,
+      "loss": 2.8903,
+      "step": 48251
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.733924388885498,
+      "learning_rate": 6.3974841470276e-05,
+      "loss": 2.9423,
+      "step": 48252
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7536724209785461,
+      "learning_rate": 6.396533815777197e-05,
+      "loss": 2.8155,
+      "step": 48253
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7304123044013977,
+      "learning_rate": 6.395583546693781e-05,
+      "loss": 2.8958,
+      "step": 48254
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7245292663574219,
+      "learning_rate": 6.394633339779865e-05,
+      "loss": 2.9736,
+      "step": 48255
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7507905960083008,
+      "learning_rate": 6.393683195037931e-05,
+      "loss": 2.8121,
+      "step": 48256
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7602225542068481,
+      "learning_rate": 6.392733112470502e-05,
+      "loss": 3.0037,
+      "step": 48257
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7352206707000732,
+      "learning_rate": 6.391783092080069e-05,
+      "loss": 2.929,
+      "step": 48258
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7542200684547424,
+      "learning_rate": 6.390833133869132e-05,
+      "loss": 2.9124,
+      "step": 48259
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7620871663093567,
+      "learning_rate": 6.389883237840201e-05,
+      "loss": 2.9583,
+      "step": 48260
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.756596028804779,
+      "learning_rate": 6.388933403995768e-05,
+      "loss": 2.8351,
+      "step": 48261
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7041810750961304,
+      "learning_rate": 6.387983632338346e-05,
+      "loss": 2.9144,
+      "step": 48262
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7106823921203613,
+      "learning_rate": 6.387033922870425e-05,
+      "loss": 3.1492,
+      "step": 48263
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7551729083061218,
+      "learning_rate": 6.386084275594519e-05,
+      "loss": 2.929,
+      "step": 48264
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7914819717407227,
+      "learning_rate": 6.385134690513122e-05,
+      "loss": 2.8709,
+      "step": 48265
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7370805144309998,
+      "learning_rate": 6.384185167628726e-05,
+      "loss": 2.7838,
+      "step": 48266
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7235316634178162,
+      "learning_rate": 6.383235706943852e-05,
+      "loss": 3.0178,
+      "step": 48267
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7227551341056824,
+      "learning_rate": 6.382286308460982e-05,
+      "loss": 3.1076,
+      "step": 48268
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7112234830856323,
+      "learning_rate": 6.381336972182624e-05,
+      "loss": 2.9721,
+      "step": 48269
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7669709920883179,
+      "learning_rate": 6.380387698111288e-05,
+      "loss": 2.8578,
+      "step": 48270
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7495376467704773,
+      "learning_rate": 6.379438486249457e-05,
+      "loss": 2.8566,
+      "step": 48271
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6867067813873291,
+      "learning_rate": 6.378489336599644e-05,
+      "loss": 3.2184,
+      "step": 48272
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6908702254295349,
+      "learning_rate": 6.377540249164348e-05,
+      "loss": 2.8861,
+      "step": 48273
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7597622871398926,
+      "learning_rate": 6.376591223946054e-05,
+      "loss": 2.673,
+      "step": 48274
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7196870446205139,
+      "learning_rate": 6.375642260947278e-05,
+      "loss": 2.8766,
+      "step": 48275
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7505701780319214,
+      "learning_rate": 6.374693360170512e-05,
+      "loss": 2.9577,
+      "step": 48276
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7584325671195984,
+      "learning_rate": 6.373744521618261e-05,
+      "loss": 2.9858,
+      "step": 48277
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7458747625350952,
+      "learning_rate": 6.372795745293013e-05,
+      "loss": 3.0763,
+      "step": 48278
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.79670649766922,
+      "learning_rate": 6.371847031197283e-05,
+      "loss": 2.8207,
+      "step": 48279
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.67808997631073,
+      "learning_rate": 6.37089837933356e-05,
+      "loss": 2.705,
+      "step": 48280
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.741486132144928,
+      "learning_rate": 6.369949789704332e-05,
+      "loss": 2.8786,
+      "step": 48281
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7459058165550232,
+      "learning_rate": 6.369001262312118e-05,
+      "loss": 3.1757,
+      "step": 48282
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6977839469909668,
+      "learning_rate": 6.368052797159402e-05,
+      "loss": 2.8612,
+      "step": 48283
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7027153968811035,
+      "learning_rate": 6.367104394248684e-05,
+      "loss": 2.9531,
+      "step": 48284
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6924808621406555,
+      "learning_rate": 6.366156053582475e-05,
+      "loss": 2.8736,
+      "step": 48285
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7258365750312805,
+      "learning_rate": 6.365207775163259e-05,
+      "loss": 2.7555,
+      "step": 48286
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7599617838859558,
+      "learning_rate": 6.36425955899354e-05,
+      "loss": 2.9788,
+      "step": 48287
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6839301586151123,
+      "learning_rate": 6.363311405075803e-05,
+      "loss": 3.2282,
+      "step": 48288
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7343278527259827,
+      "learning_rate": 6.362363313412557e-05,
+      "loss": 2.9432,
+      "step": 48289
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.739058256149292,
+      "learning_rate": 6.361415284006301e-05,
+      "loss": 2.7162,
+      "step": 48290
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7455694675445557,
+      "learning_rate": 6.360467316859522e-05,
+      "loss": 3.0939,
+      "step": 48291
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7573563456535339,
+      "learning_rate": 6.359519411974733e-05,
+      "loss": 2.9621,
+      "step": 48292
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.8001168370246887,
+      "learning_rate": 6.358571569354408e-05,
+      "loss": 2.9751,
+      "step": 48293
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7187392115592957,
+      "learning_rate": 6.357623789001064e-05,
+      "loss": 2.984,
+      "step": 48294
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6924291849136353,
+      "learning_rate": 6.356676070917192e-05,
+      "loss": 2.9977,
+      "step": 48295
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7191535830497742,
+      "learning_rate": 6.355728415105276e-05,
+      "loss": 2.8086,
+      "step": 48296
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7379903793334961,
+      "learning_rate": 6.354780821567828e-05,
+      "loss": 2.7732,
+      "step": 48297
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7109218239784241,
+      "learning_rate": 6.353833290307329e-05,
+      "loss": 2.6506,
+      "step": 48298
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7022459506988525,
+      "learning_rate": 6.35288582132628e-05,
+      "loss": 2.8524,
+      "step": 48299
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.734695553779602,
+      "learning_rate": 6.351938414627191e-05,
+      "loss": 2.8649,
+      "step": 48300
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7235419750213623,
+      "learning_rate": 6.350991070212544e-05,
+      "loss": 2.6928,
+      "step": 48301
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7387106418609619,
+      "learning_rate": 6.350043788084835e-05,
+      "loss": 2.7351,
+      "step": 48302
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.709895133972168,
+      "learning_rate": 6.349096568246549e-05,
+      "loss": 2.8434,
+      "step": 48303
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.757611870765686,
+      "learning_rate": 6.348149410700192e-05,
+      "loss": 2.9994,
+      "step": 48304
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6948244571685791,
+      "learning_rate": 6.347202315448265e-05,
+      "loss": 2.7651,
+      "step": 48305
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7108848094940186,
+      "learning_rate": 6.346255282493248e-05,
+      "loss": 2.6992,
+      "step": 48306
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7500197291374207,
+      "learning_rate": 6.34530831183765e-05,
+      "loss": 2.9149,
+      "step": 48307
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7581573724746704,
+      "learning_rate": 6.34436140348396e-05,
+      "loss": 2.8101,
+      "step": 48308
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7715035676956177,
+      "learning_rate": 6.343414557434658e-05,
+      "loss": 2.795,
+      "step": 48309
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7194180488586426,
+      "learning_rate": 6.342467773692258e-05,
+      "loss": 2.9186,
+      "step": 48310
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7068958282470703,
+      "learning_rate": 6.341521052259236e-05,
+      "loss": 2.9039,
+      "step": 48311
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7456514239311218,
+      "learning_rate": 6.340574393138104e-05,
+      "loss": 2.9809,
+      "step": 48312
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7299491763114929,
+      "learning_rate": 6.339627796331338e-05,
+      "loss": 3.1068,
+      "step": 48313
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7144024968147278,
+      "learning_rate": 6.338681261841447e-05,
+      "loss": 2.7794,
+      "step": 48314
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7453769445419312,
+      "learning_rate": 6.337734789670907e-05,
+      "loss": 2.8954,
+      "step": 48315
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7227847576141357,
+      "learning_rate": 6.336788379822226e-05,
+      "loss": 2.9263,
+      "step": 48316
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7394989132881165,
+      "learning_rate": 6.335842032297892e-05,
+      "loss": 2.9357,
+      "step": 48317
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7643879652023315,
+      "learning_rate": 6.334895747100388e-05,
+      "loss": 2.8486,
+      "step": 48318
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.710767388343811,
+      "learning_rate": 6.333949524232222e-05,
+      "loss": 2.8892,
+      "step": 48319
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.743543803691864,
+      "learning_rate": 6.33300336369587e-05,
+      "loss": 3.076,
+      "step": 48320
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7929369211196899,
+      "learning_rate": 6.332057265493834e-05,
+      "loss": 2.9037,
+      "step": 48321
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7415796518325806,
+      "learning_rate": 6.33111122962861e-05,
+      "loss": 2.8008,
+      "step": 48322
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7029473185539246,
+      "learning_rate": 6.330165256102687e-05,
+      "loss": 2.8482,
+      "step": 48323
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7240558862686157,
+      "learning_rate": 6.32921934491855e-05,
+      "loss": 2.8949,
+      "step": 48324
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.9161472320556641,
+      "learning_rate": 6.328273496078686e-05,
+      "loss": 2.7527,
+      "step": 48325
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7548016309738159,
+      "learning_rate": 6.327327709585597e-05,
+      "loss": 2.7239,
+      "step": 48326
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7531619071960449,
+      "learning_rate": 6.326381985441776e-05,
+      "loss": 2.9115,
+      "step": 48327
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7124982476234436,
+      "learning_rate": 6.3254363236497e-05,
+      "loss": 2.8327,
+      "step": 48328
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7740418910980225,
+      "learning_rate": 6.324490724211879e-05,
+      "loss": 2.811,
+      "step": 48329
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7585882544517517,
+      "learning_rate": 6.32354518713079e-05,
+      "loss": 2.7411,
+      "step": 48330
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6965351104736328,
+      "learning_rate": 6.322599712408922e-05,
+      "loss": 2.8907,
+      "step": 48331
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7372415065765381,
+      "learning_rate": 6.321654300048774e-05,
+      "loss": 3.0206,
+      "step": 48332
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7417186498641968,
+      "learning_rate": 6.320708950052827e-05,
+      "loss": 2.8534,
+      "step": 48333
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.701483964920044,
+      "learning_rate": 6.31976366242358e-05,
+      "loss": 2.8217,
+      "step": 48334
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7164797186851501,
+      "learning_rate": 6.318818437163514e-05,
+      "loss": 2.8596,
+      "step": 48335
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.727675199508667,
+      "learning_rate": 6.31787327427512e-05,
+      "loss": 2.7906,
+      "step": 48336
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6889267563819885,
+      "learning_rate": 6.3169281737609e-05,
+      "loss": 3.2159,
+      "step": 48337
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7056227326393127,
+      "learning_rate": 6.315983135623331e-05,
+      "loss": 2.9742,
+      "step": 48338
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7440009713172913,
+      "learning_rate": 6.315038159864905e-05,
+      "loss": 2.872,
+      "step": 48339
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7045519948005676,
+      "learning_rate": 6.314093246488101e-05,
+      "loss": 3.1583,
+      "step": 48340
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7586000561714172,
+      "learning_rate": 6.313148395495418e-05,
+      "loss": 2.7775,
+      "step": 48341
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7050657272338867,
+      "learning_rate": 6.312203606889356e-05,
+      "loss": 2.7034,
+      "step": 48342
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.734698474407196,
+      "learning_rate": 6.311258880672376e-05,
+      "loss": 2.8888,
+      "step": 48343
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7230560183525085,
+      "learning_rate": 6.310314216846992e-05,
+      "loss": 2.8722,
+      "step": 48344
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7460218667984009,
+      "learning_rate": 6.309369615415681e-05,
+      "loss": 2.943,
+      "step": 48345
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7174545526504517,
+      "learning_rate": 6.308425076380925e-05,
+      "loss": 2.6023,
+      "step": 48346
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7263779044151306,
+      "learning_rate": 6.307480599745225e-05,
+      "loss": 2.9191,
+      "step": 48347
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7491353154182434,
+      "learning_rate": 6.306536185511053e-05,
+      "loss": 2.9063,
+      "step": 48348
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.691290557384491,
+      "learning_rate": 6.30559183368091e-05,
+      "loss": 3.0822,
+      "step": 48349
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.776136577129364,
+      "learning_rate": 6.304647544257273e-05,
+      "loss": 2.7441,
+      "step": 48350
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7215666174888611,
+      "learning_rate": 6.30370331724264e-05,
+      "loss": 2.9793,
+      "step": 48351
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7477503418922424,
+      "learning_rate": 6.302759152639496e-05,
+      "loss": 3.0098,
+      "step": 48352
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7336341142654419,
+      "learning_rate": 6.301815050450313e-05,
+      "loss": 2.9931,
+      "step": 48353
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.8228724598884583,
+      "learning_rate": 6.300871010677596e-05,
+      "loss": 2.9006,
+      "step": 48354
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7225062847137451,
+      "learning_rate": 6.299927033323817e-05,
+      "loss": 3.0748,
+      "step": 48355
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7133581638336182,
+      "learning_rate": 6.298983118391467e-05,
+      "loss": 2.8641,
+      "step": 48356
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7188670635223389,
+      "learning_rate": 6.298039265883044e-05,
+      "loss": 2.9304,
+      "step": 48357
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6973915100097656,
+      "learning_rate": 6.297095475801014e-05,
+      "loss": 3.0409,
+      "step": 48358
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7095557451248169,
+      "learning_rate": 6.296151748147883e-05,
+      "loss": 3.026,
+      "step": 48359
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.75076824426651,
+      "learning_rate": 6.295208082926126e-05,
+      "loss": 3.1059,
+      "step": 48360
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7055838704109192,
+      "learning_rate": 6.294264480138222e-05,
+      "loss": 2.9436,
+      "step": 48361
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7464839220046997,
+      "learning_rate": 6.29332093978667e-05,
+      "loss": 3.007,
+      "step": 48362
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.715093731880188,
+      "learning_rate": 6.29237746187394e-05,
+      "loss": 2.8428,
+      "step": 48363
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7014065384864807,
+      "learning_rate": 6.291434046402533e-05,
+      "loss": 3.0259,
+      "step": 48364
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7219024896621704,
+      "learning_rate": 6.290490693374915e-05,
+      "loss": 3.1076,
+      "step": 48365
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7014488577842712,
+      "learning_rate": 6.289547402793594e-05,
+      "loss": 2.9713,
+      "step": 48366
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.703105092048645,
+      "learning_rate": 6.288604174661041e-05,
+      "loss": 2.9675,
+      "step": 48367
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7211623191833496,
+      "learning_rate": 6.287661008979732e-05,
+      "loss": 2.8565,
+      "step": 48368
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7292293310165405,
+      "learning_rate": 6.286717905752169e-05,
+      "loss": 3.0596,
+      "step": 48369
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7422485947608948,
+      "learning_rate": 6.285774864980819e-05,
+      "loss": 2.8302,
+      "step": 48370
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7382919192314148,
+      "learning_rate": 6.284831886668175e-05,
+      "loss": 2.8995,
+      "step": 48371
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7146931886672974,
+      "learning_rate": 6.283888970816726e-05,
+      "loss": 3.0275,
+      "step": 48372
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6943666934967041,
+      "learning_rate": 6.282946117428942e-05,
+      "loss": 2.9928,
+      "step": 48373
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.730255126953125,
+      "learning_rate": 6.282003326507323e-05,
+      "loss": 2.9221,
+      "step": 48374
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6991233825683594,
+      "learning_rate": 6.281060598054342e-05,
+      "loss": 2.6633,
+      "step": 48375
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7602810859680176,
+      "learning_rate": 6.280117932072474e-05,
+      "loss": 2.8916,
+      "step": 48376
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6797572374343872,
+      "learning_rate": 6.279175328564217e-05,
+      "loss": 2.6735,
+      "step": 48377
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.8254128694534302,
+      "learning_rate": 6.278232787532041e-05,
+      "loss": 2.8041,
+      "step": 48378
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.701127827167511,
+      "learning_rate": 6.277290308978445e-05,
+      "loss": 2.8629,
+      "step": 48379
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7156503796577454,
+      "learning_rate": 6.276347892905887e-05,
+      "loss": 2.8143,
+      "step": 48380
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7461564540863037,
+      "learning_rate": 6.275405539316876e-05,
+      "loss": 3.0213,
+      "step": 48381
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7387633323669434,
+      "learning_rate": 6.274463248213877e-05,
+      "loss": 3.003,
+      "step": 48382
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7231044173240662,
+      "learning_rate": 6.27352101959937e-05,
+      "loss": 2.9757,
+      "step": 48383
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7560230493545532,
+      "learning_rate": 6.27257885347585e-05,
+      "loss": 2.792,
+      "step": 48384
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7416335344314575,
+      "learning_rate": 6.271636749845783e-05,
+      "loss": 2.8567,
+      "step": 48385
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7156395316123962,
+      "learning_rate": 6.270694708711656e-05,
+      "loss": 2.8991,
+      "step": 48386
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7483652830123901,
+      "learning_rate": 6.269752730075962e-05,
+      "loss": 2.8864,
+      "step": 48387
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7325050234794617,
+      "learning_rate": 6.268810813941173e-05,
+      "loss": 2.7133,
+      "step": 48388
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7214424014091492,
+      "learning_rate": 6.26786896030977e-05,
+      "loss": 2.7828,
+      "step": 48389
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7492168545722961,
+      "learning_rate": 6.266927169184223e-05,
+      "loss": 2.9989,
+      "step": 48390
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7547978758811951,
+      "learning_rate": 6.265985440567029e-05,
+      "loss": 2.6112,
+      "step": 48391
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6953510642051697,
+      "learning_rate": 6.265043774460659e-05,
+      "loss": 3.0589,
+      "step": 48392
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7603529095649719,
+      "learning_rate": 6.264102170867588e-05,
+      "loss": 2.8438,
+      "step": 48393
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6982311606407166,
+      "learning_rate": 6.263160629790318e-05,
+      "loss": 2.8759,
+      "step": 48394
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6933879852294922,
+      "learning_rate": 6.262219151231305e-05,
+      "loss": 3.0471,
+      "step": 48395
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7062197923660278,
+      "learning_rate": 6.261277735193047e-05,
+      "loss": 2.9002,
+      "step": 48396
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7573047280311584,
+      "learning_rate": 6.260336381678011e-05,
+      "loss": 2.892,
+      "step": 48397
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6713040471076965,
+      "learning_rate": 6.259395090688678e-05,
+      "loss": 2.9715,
+      "step": 48398
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7490809559822083,
+      "learning_rate": 6.258453862227534e-05,
+      "loss": 2.9012,
+      "step": 48399
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7417430877685547,
+      "learning_rate": 6.257512696297047e-05,
+      "loss": 2.9564,
+      "step": 48400
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7003105878829956,
+      "learning_rate": 6.256571592899708e-05,
+      "loss": 2.9267,
+      "step": 48401
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7170599102973938,
+      "learning_rate": 6.255630552037983e-05,
+      "loss": 2.8603,
+      "step": 48402
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7321606874465942,
+      "learning_rate": 6.254689573714363e-05,
+      "loss": 3.1191,
+      "step": 48403
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7242743372917175,
+      "learning_rate": 6.253748657931326e-05,
+      "loss": 2.9785,
+      "step": 48404
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7323653101921082,
+      "learning_rate": 6.252807804691333e-05,
+      "loss": 2.8328,
+      "step": 48405
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7143465280532837,
+      "learning_rate": 6.251867013996884e-05,
+      "loss": 2.9296,
+      "step": 48406
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7798712849617004,
+      "learning_rate": 6.250926285850437e-05,
+      "loss": 3.0471,
+      "step": 48407
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7423224449157715,
+      "learning_rate": 6.249985620254479e-05,
+      "loss": 2.9293,
+      "step": 48408
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7109093070030212,
+      "learning_rate": 6.249045017211499e-05,
+      "loss": 2.9326,
+      "step": 48409
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7001209259033203,
+      "learning_rate": 6.248104476723963e-05,
+      "loss": 2.6705,
+      "step": 48410
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7019423842430115,
+      "learning_rate": 6.247163998794349e-05,
+      "loss": 2.7414,
+      "step": 48411
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7206318378448486,
+      "learning_rate": 6.246223583425122e-05,
+      "loss": 2.8257,
+      "step": 48412
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7256973385810852,
+      "learning_rate": 6.245283230618773e-05,
+      "loss": 2.7982,
+      "step": 48413
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7365603446960449,
+      "learning_rate": 6.244342940377787e-05,
+      "loss": 2.8841,
+      "step": 48414
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7689672708511353,
+      "learning_rate": 6.243402712704618e-05,
+      "loss": 2.956,
+      "step": 48415
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7122086882591248,
+      "learning_rate": 6.242462547601762e-05,
+      "loss": 2.8686,
+      "step": 48416
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7387104034423828,
+      "learning_rate": 6.241522445071683e-05,
+      "loss": 3.0019,
+      "step": 48417
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7708308100700378,
+      "learning_rate": 6.240582405116865e-05,
+      "loss": 2.8456,
+      "step": 48418
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7489545345306396,
+      "learning_rate": 6.239642427739783e-05,
+      "loss": 2.8216,
+      "step": 48419
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7343196272850037,
+      "learning_rate": 6.238702512942904e-05,
+      "loss": 2.7546,
+      "step": 48420
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7246786952018738,
+      "learning_rate": 6.237762660728712e-05,
+      "loss": 3.0248,
+      "step": 48421
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7109408378601074,
+      "learning_rate": 6.236822871099678e-05,
+      "loss": 2.9897,
+      "step": 48422
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7401118278503418,
+      "learning_rate": 6.235883144058275e-05,
+      "loss": 2.7363,
+      "step": 48423
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6865191459655762,
+      "learning_rate": 6.234943479606993e-05,
+      "loss": 2.7305,
+      "step": 48424
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7721907496452332,
+      "learning_rate": 6.234003877748295e-05,
+      "loss": 2.9704,
+      "step": 48425
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.728862464427948,
+      "learning_rate": 6.233064338484656e-05,
+      "loss": 2.7578,
+      "step": 48426
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7266650199890137,
+      "learning_rate": 6.232124861818543e-05,
+      "loss": 2.904,
+      "step": 48427
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.698026716709137,
+      "learning_rate": 6.231185447752444e-05,
+      "loss": 3.1348,
+      "step": 48428
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7453986406326294,
+      "learning_rate": 6.23024609628883e-05,
+      "loss": 2.982,
+      "step": 48429
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7807963490486145,
+      "learning_rate": 6.229306807430168e-05,
+      "loss": 3.0589,
+      "step": 48430
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7143594622612,
+      "learning_rate": 6.228367581178945e-05,
+      "loss": 2.7484,
+      "step": 48431
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7836489081382751,
+      "learning_rate": 6.227428417537629e-05,
+      "loss": 2.9194,
+      "step": 48432
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.702145516872406,
+      "learning_rate": 6.226489316508678e-05,
+      "loss": 2.9083,
+      "step": 48433
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.8490208387374878,
+      "learning_rate": 6.225550278094591e-05,
+      "loss": 2.8837,
+      "step": 48434
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6956592798233032,
+      "learning_rate": 6.224611302297821e-05,
+      "loss": 3.056,
+      "step": 48435
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7090912461280823,
+      "learning_rate": 6.223672389120855e-05,
+      "loss": 2.9451,
+      "step": 48436
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7443285584449768,
+      "learning_rate": 6.222733538566153e-05,
+      "loss": 2.851,
+      "step": 48437
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7666028738021851,
+      "learning_rate": 6.221794750636195e-05,
+      "loss": 2.8243,
+      "step": 48438
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7153770923614502,
+      "learning_rate": 6.220856025333462e-05,
+      "loss": 2.7963,
+      "step": 48439
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7110601663589478,
+      "learning_rate": 6.219917362660419e-05,
+      "loss": 2.9492,
+      "step": 48440
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7466623187065125,
+      "learning_rate": 6.218978762619533e-05,
+      "loss": 2.8231,
+      "step": 48441
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.746893584728241,
+      "learning_rate": 6.218040225213277e-05,
+      "loss": 2.8501,
+      "step": 48442
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7358802556991577,
+      "learning_rate": 6.217101750444124e-05,
+      "loss": 2.9031,
+      "step": 48443
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6966966986656189,
+      "learning_rate": 6.216163338314555e-05,
+      "loss": 3.0072,
+      "step": 48444
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7609187960624695,
+      "learning_rate": 6.215224988827027e-05,
+      "loss": 2.7928,
+      "step": 48445
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6976915001869202,
+      "learning_rate": 6.214286701984028e-05,
+      "loss": 3.0634,
+      "step": 48446
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7734736204147339,
+      "learning_rate": 6.213348477788017e-05,
+      "loss": 3.026,
+      "step": 48447
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7737905383110046,
+      "learning_rate": 6.212410316241465e-05,
+      "loss": 2.9261,
+      "step": 48448
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7344973683357239,
+      "learning_rate": 6.21147221734685e-05,
+      "loss": 2.6841,
+      "step": 48449
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.687893807888031,
+      "learning_rate": 6.210534181106634e-05,
+      "loss": 2.7757,
+      "step": 48450
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7061976194381714,
+      "learning_rate": 6.209596207523299e-05,
+      "loss": 3.0322,
+      "step": 48451
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7552989721298218,
+      "learning_rate": 6.2086582965993e-05,
+      "loss": 2.8587,
+      "step": 48452
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7478323578834534,
+      "learning_rate": 6.207720448337127e-05,
+      "loss": 2.93,
+      "step": 48453
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7599266767501831,
+      "learning_rate": 6.206782662739236e-05,
+      "loss": 2.9545,
+      "step": 48454
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.752623975276947,
+      "learning_rate": 6.205844939808096e-05,
+      "loss": 2.79,
+      "step": 48455
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7089788913726807,
+      "learning_rate": 6.204907279546188e-05,
+      "loss": 2.9757,
+      "step": 48456
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7725157737731934,
+      "learning_rate": 6.203969681955967e-05,
+      "loss": 3.0366,
+      "step": 48457
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6831510663032532,
+      "learning_rate": 6.203032147039909e-05,
+      "loss": 3.068,
+      "step": 48458
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7839120626449585,
+      "learning_rate": 6.202094674800492e-05,
+      "loss": 2.8344,
+      "step": 48459
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7446629405021667,
+      "learning_rate": 6.20115726524017e-05,
+      "loss": 2.8361,
+      "step": 48460
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7064789533615112,
+      "learning_rate": 6.200219918361431e-05,
+      "loss": 2.8209,
+      "step": 48461
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7748939394950867,
+      "learning_rate": 6.199282634166726e-05,
+      "loss": 2.9435,
+      "step": 48462
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7227396368980408,
+      "learning_rate": 6.198345412658529e-05,
+      "loss": 2.6958,
+      "step": 48463
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.705480694770813,
+      "learning_rate": 6.197408253839314e-05,
+      "loss": 2.9662,
+      "step": 48464
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.704855740070343,
+      "learning_rate": 6.196471157711535e-05,
+      "loss": 3.0984,
+      "step": 48465
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8089683651924133,
+      "learning_rate": 6.19553412427768e-05,
+      "loss": 2.8041,
+      "step": 48466
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7625295519828796,
+      "learning_rate": 6.194597153540199e-05,
+      "loss": 2.6505,
+      "step": 48467
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7611942887306213,
+      "learning_rate": 6.193660245501575e-05,
+      "loss": 2.988,
+      "step": 48468
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7407697439193726,
+      "learning_rate": 6.192723400164268e-05,
+      "loss": 2.9857,
+      "step": 48469
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7640385627746582,
+      "learning_rate": 6.191786617530739e-05,
+      "loss": 2.7547,
+      "step": 48470
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7157742977142334,
+      "learning_rate": 6.190849897603472e-05,
+      "loss": 2.9381,
+      "step": 48471
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7664361596107483,
+      "learning_rate": 6.189913240384911e-05,
+      "loss": 2.8453,
+      "step": 48472
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7323439121246338,
+      "learning_rate": 6.188976645877548e-05,
+      "loss": 2.9364,
+      "step": 48473
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7110058069229126,
+      "learning_rate": 6.188040114083829e-05,
+      "loss": 2.8488,
+      "step": 48474
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7112416625022888,
+      "learning_rate": 6.187103645006228e-05,
+      "loss": 2.9664,
+      "step": 48475
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7495659589767456,
+      "learning_rate": 6.186167238647232e-05,
+      "loss": 2.6364,
+      "step": 48476
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7033055424690247,
+      "learning_rate": 6.18523089500927e-05,
+      "loss": 2.9183,
+      "step": 48477
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6921057105064392,
+      "learning_rate": 6.184294614094835e-05,
+      "loss": 2.8459,
+      "step": 48478
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7195352911949158,
+      "learning_rate": 6.183358395906376e-05,
+      "loss": 2.8092,
+      "step": 48479
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6981593370437622,
+      "learning_rate": 6.182422240446372e-05,
+      "loss": 2.9155,
+      "step": 48480
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.715376079082489,
+      "learning_rate": 6.181486147717287e-05,
+      "loss": 2.7584,
+      "step": 48481
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7249025106430054,
+      "learning_rate": 6.180550117721577e-05,
+      "loss": 2.8825,
+      "step": 48482
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7154632806777954,
+      "learning_rate": 6.179614150461722e-05,
+      "loss": 2.695,
+      "step": 48483
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7087334394454956,
+      "learning_rate": 6.178678245940179e-05,
+      "loss": 2.7371,
+      "step": 48484
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7212601900100708,
+      "learning_rate": 6.177742404159404e-05,
+      "loss": 2.6432,
+      "step": 48485
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7712526321411133,
+      "learning_rate": 6.176806625121882e-05,
+      "loss": 2.7397,
+      "step": 48486
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7613638639450073,
+      "learning_rate": 6.175870908830056e-05,
+      "loss": 2.9963,
+      "step": 48487
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7476872205734253,
+      "learning_rate": 6.174935255286408e-05,
+      "loss": 2.7339,
+      "step": 48488
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7133546471595764,
+      "learning_rate": 6.17399966449339e-05,
+      "loss": 2.8263,
+      "step": 48489
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.736723005771637,
+      "learning_rate": 6.173064136453477e-05,
+      "loss": 2.7688,
+      "step": 48490
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7491676807403564,
+      "learning_rate": 6.172128671169128e-05,
+      "loss": 3.0156,
+      "step": 48491
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7248450517654419,
+      "learning_rate": 6.171193268642799e-05,
+      "loss": 2.8164,
+      "step": 48492
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7055536508560181,
+      "learning_rate": 6.170257928876969e-05,
+      "loss": 2.9714,
+      "step": 48493
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7106114029884338,
+      "learning_rate": 6.169322651874085e-05,
+      "loss": 3.0877,
+      "step": 48494
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7218091487884521,
+      "learning_rate": 6.168387437636619e-05,
+      "loss": 2.9154,
+      "step": 48495
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7130762338638306,
+      "learning_rate": 6.167452286167042e-05,
+      "loss": 2.9306,
+      "step": 48496
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7284539937973022,
+      "learning_rate": 6.1665171974678e-05,
+      "loss": 2.9702,
+      "step": 48497
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.736548125743866,
+      "learning_rate": 6.165582171541375e-05,
+      "loss": 2.7348,
+      "step": 48498
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7026635408401489,
+      "learning_rate": 6.164647208390217e-05,
+      "loss": 3.0233,
+      "step": 48499
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7796565890312195,
+      "learning_rate": 6.163712308016788e-05,
+      "loss": 2.9025,
+      "step": 48500
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8084689378738403,
+      "learning_rate": 6.162777470423555e-05,
+      "loss": 2.9667,
+      "step": 48501
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7033407092094421,
+      "learning_rate": 6.161842695612974e-05,
+      "loss": 2.8283,
+      "step": 48502
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7982221841812134,
+      "learning_rate": 6.160907983587521e-05,
+      "loss": 3.0055,
+      "step": 48503
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7438850402832031,
+      "learning_rate": 6.159973334349637e-05,
+      "loss": 2.9091,
+      "step": 48504
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7206811904907227,
+      "learning_rate": 6.159038747901804e-05,
+      "loss": 2.7498,
+      "step": 48505
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6850907206535339,
+      "learning_rate": 6.158104224246477e-05,
+      "loss": 2.9421,
+      "step": 48506
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7406187057495117,
+      "learning_rate": 6.157169763386102e-05,
+      "loss": 2.954,
+      "step": 48507
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.728126585483551,
+      "learning_rate": 6.156235365323165e-05,
+      "loss": 2.7771,
+      "step": 48508
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6951186656951904,
+      "learning_rate": 6.15530103006011e-05,
+      "loss": 2.8559,
+      "step": 48509
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7429779767990112,
+      "learning_rate": 6.154366757599399e-05,
+      "loss": 2.92,
+      "step": 48510
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6777859330177307,
+      "learning_rate": 6.153432547943504e-05,
+      "loss": 2.9665,
+      "step": 48511
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7383310794830322,
+      "learning_rate": 6.15249840109488e-05,
+      "loss": 2.7535,
+      "step": 48512
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7504093050956726,
+      "learning_rate": 6.151564317055986e-05,
+      "loss": 2.9476,
+      "step": 48513
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6900618672370911,
+      "learning_rate": 6.150630295829275e-05,
+      "loss": 2.7624,
+      "step": 48514
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7341283559799194,
+      "learning_rate": 6.149696337417213e-05,
+      "loss": 2.8167,
+      "step": 48515
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7598667144775391,
+      "learning_rate": 6.148762441822266e-05,
+      "loss": 2.7693,
+      "step": 48516
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7878901362419128,
+      "learning_rate": 6.147828609046882e-05,
+      "loss": 2.7556,
+      "step": 48517
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7257072329521179,
+      "learning_rate": 6.146894839093535e-05,
+      "loss": 3.006,
+      "step": 48518
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7379465103149414,
+      "learning_rate": 6.145961131964671e-05,
+      "loss": 3.0419,
+      "step": 48519
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7503682374954224,
+      "learning_rate": 6.145027487662762e-05,
+      "loss": 2.8252,
+      "step": 48520
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7141533493995667,
+      "learning_rate": 6.144093906190256e-05,
+      "loss": 2.9226,
+      "step": 48521
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7183959484100342,
+      "learning_rate": 6.14316038754961e-05,
+      "loss": 2.8882,
+      "step": 48522
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7660282254219055,
+      "learning_rate": 6.142226931743296e-05,
+      "loss": 2.8312,
+      "step": 48523
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7023894190788269,
+      "learning_rate": 6.141293538773758e-05,
+      "loss": 2.787,
+      "step": 48524
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7554915547370911,
+      "learning_rate": 6.140360208643459e-05,
+      "loss": 2.9541,
+      "step": 48525
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6967014670372009,
+      "learning_rate": 6.139426941354866e-05,
+      "loss": 2.9179,
+      "step": 48526
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7923035025596619,
+      "learning_rate": 6.138493736910433e-05,
+      "loss": 3.1075,
+      "step": 48527
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7242387533187866,
+      "learning_rate": 6.137560595312614e-05,
+      "loss": 2.7576,
+      "step": 48528
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7367933392524719,
+      "learning_rate": 6.136627516563861e-05,
+      "loss": 2.8662,
+      "step": 48529
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7342579364776611,
+      "learning_rate": 6.135694500666638e-05,
+      "loss": 2.8216,
+      "step": 48530
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7151744365692139,
+      "learning_rate": 6.134761547623413e-05,
+      "loss": 2.8982,
+      "step": 48531
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7614980340003967,
+      "learning_rate": 6.133828657436622e-05,
+      "loss": 2.969,
+      "step": 48532
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7628613114356995,
+      "learning_rate": 6.132895830108744e-05,
+      "loss": 2.946,
+      "step": 48533
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7301518321037292,
+      "learning_rate": 6.131963065642221e-05,
+      "loss": 2.8227,
+      "step": 48534
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7378628253936768,
+      "learning_rate": 6.13103036403951e-05,
+      "loss": 3.0836,
+      "step": 48535
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7157490253448486,
+      "learning_rate": 6.130097725303078e-05,
+      "loss": 2.8048,
+      "step": 48536
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7131938338279724,
+      "learning_rate": 6.129165149435364e-05,
+      "loss": 3.134,
+      "step": 48537
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7120147943496704,
+      "learning_rate": 6.128232636438846e-05,
+      "loss": 2.9152,
+      "step": 48538
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8592355847358704,
+      "learning_rate": 6.127300186315962e-05,
+      "loss": 2.9719,
+      "step": 48539
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7270697951316833,
+      "learning_rate": 6.126367799069172e-05,
+      "loss": 2.8289,
+      "step": 48540
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7252556681632996,
+      "learning_rate": 6.125435474700944e-05,
+      "loss": 2.7769,
+      "step": 48541
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8012216687202454,
+      "learning_rate": 6.124503213213723e-05,
+      "loss": 2.9533,
+      "step": 48542
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7498360872268677,
+      "learning_rate": 6.123571014609967e-05,
+      "loss": 2.8852,
+      "step": 48543
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7274046540260315,
+      "learning_rate": 6.122638878892121e-05,
+      "loss": 3.0165,
+      "step": 48544
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.713111400604248,
+      "learning_rate": 6.12170680606266e-05,
+      "loss": 2.8686,
+      "step": 48545
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7824342250823975,
+      "learning_rate": 6.120774796124017e-05,
+      "loss": 2.973,
+      "step": 48546
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7419819235801697,
+      "learning_rate": 6.119842849078656e-05,
+      "loss": 2.9885,
+      "step": 48547
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7952620983123779,
+      "learning_rate": 6.118910964929042e-05,
+      "loss": 2.8178,
+      "step": 48548
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7857388854026794,
+      "learning_rate": 6.117979143677623e-05,
+      "loss": 2.9158,
+      "step": 48549
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7016808986663818,
+      "learning_rate": 6.117047385326851e-05,
+      "loss": 3.0688,
+      "step": 48550
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7236056923866272,
+      "learning_rate": 6.116115689879168e-05,
+      "loss": 3.1444,
+      "step": 48551
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7185097336769104,
+      "learning_rate": 6.11518405733704e-05,
+      "loss": 2.6668,
+      "step": 48552
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7458245754241943,
+      "learning_rate": 6.114252487702931e-05,
+      "loss": 2.8227,
+      "step": 48553
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7336084246635437,
+      "learning_rate": 6.113320980979275e-05,
+      "loss": 2.7382,
+      "step": 48554
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.750745415687561,
+      "learning_rate": 6.112389537168544e-05,
+      "loss": 2.87,
+      "step": 48555
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7281879186630249,
+      "learning_rate": 6.11145815627318e-05,
+      "loss": 2.9631,
+      "step": 48556
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.777651309967041,
+      "learning_rate": 6.110526838295631e-05,
+      "loss": 2.7747,
+      "step": 48557
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7897119522094727,
+      "learning_rate": 6.109595583238363e-05,
+      "loss": 2.9692,
+      "step": 48558
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7822614312171936,
+      "learning_rate": 6.108664391103815e-05,
+      "loss": 2.7721,
+      "step": 48559
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7523203492164612,
+      "learning_rate": 6.107733261894457e-05,
+      "loss": 2.9797,
+      "step": 48560
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.747348964214325,
+      "learning_rate": 6.10680219561272e-05,
+      "loss": 3.1613,
+      "step": 48561
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7030299305915833,
+      "learning_rate": 6.105871192261068e-05,
+      "loss": 2.8633,
+      "step": 48562
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.703755259513855,
+      "learning_rate": 6.104940251841966e-05,
+      "loss": 2.968,
+      "step": 48563
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7231656908988953,
+      "learning_rate": 6.104009374357846e-05,
+      "loss": 2.7655,
+      "step": 48564
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7645540833473206,
+      "learning_rate": 6.10307855981117e-05,
+      "loss": 3.0025,
+      "step": 48565
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7383845448493958,
+      "learning_rate": 6.1021478082043765e-05,
+      "loss": 2.9483,
+      "step": 48566
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7293666005134583,
+      "learning_rate": 6.101217119539927e-05,
+      "loss": 3.0159,
+      "step": 48567
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.736205518245697,
+      "learning_rate": 6.100286493820281e-05,
+      "loss": 2.806,
+      "step": 48568
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7195806503295898,
+      "learning_rate": 6.099355931047872e-05,
+      "loss": 2.8859,
+      "step": 48569
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7205501198768616,
+      "learning_rate": 6.098425431225169e-05,
+      "loss": 2.6709,
+      "step": 48570
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7264590263366699,
+      "learning_rate": 6.097494994354612e-05,
+      "loss": 2.8121,
+      "step": 48571
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7377135753631592,
+      "learning_rate": 6.0965646204386445e-05,
+      "loss": 2.7602,
+      "step": 48572
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7292772531509399,
+      "learning_rate": 6.0956343094797355e-05,
+      "loss": 2.9107,
+      "step": 48573
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8081817626953125,
+      "learning_rate": 6.0947040614803144e-05,
+      "loss": 2.7616,
+      "step": 48574
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6972966194152832,
+      "learning_rate": 6.093773876442852e-05,
+      "loss": 2.8046,
+      "step": 48575
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6994813084602356,
+      "learning_rate": 6.09284375436978e-05,
+      "loss": 2.872,
+      "step": 48576
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7522056102752686,
+      "learning_rate": 6.091913695263557e-05,
+      "loss": 3.0694,
+      "step": 48577
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7025183439254761,
+      "learning_rate": 6.0909836991266506e-05,
+      "loss": 2.9151,
+      "step": 48578
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7216269373893738,
+      "learning_rate": 6.090053765961471e-05,
+      "loss": 2.8595,
+      "step": 48579
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7327861189842224,
+      "learning_rate": 6.089123895770498e-05,
+      "loss": 2.7038,
+      "step": 48580
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7535680532455444,
+      "learning_rate": 6.088194088556165e-05,
+      "loss": 2.866,
+      "step": 48581
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7392609119415283,
+      "learning_rate": 6.0872643443209246e-05,
+      "loss": 2.8061,
+      "step": 48582
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7233937382698059,
+      "learning_rate": 6.0863346630672336e-05,
+      "loss": 2.8853,
+      "step": 48583
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7154392600059509,
+      "learning_rate": 6.085405044797531e-05,
+      "loss": 3.0673,
+      "step": 48584
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7913565039634705,
+      "learning_rate": 6.084475489514274e-05,
+      "loss": 3.0256,
+      "step": 48585
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7132130861282349,
+      "learning_rate": 6.0835459972199086e-05,
+      "loss": 2.9848,
+      "step": 48586
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7644349932670593,
+      "learning_rate": 6.082616567916871e-05,
+      "loss": 2.9009,
+      "step": 48587
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7143843173980713,
+      "learning_rate": 6.081687201607628e-05,
+      "loss": 2.6485,
+      "step": 48588
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6977376341819763,
+      "learning_rate": 6.080757898294606e-05,
+      "loss": 2.8017,
+      "step": 48589
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7115483283996582,
+      "learning_rate": 6.0798286579802765e-05,
+      "loss": 2.9167,
+      "step": 48590
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7304779887199402,
+      "learning_rate": 6.0788994806670656e-05,
+      "loss": 2.9317,
+      "step": 48591
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7648419141769409,
+      "learning_rate": 6.0779703663574364e-05,
+      "loss": 3.0281,
+      "step": 48592
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7392321825027466,
+      "learning_rate": 6.077041315053828e-05,
+      "loss": 2.9541,
+      "step": 48593
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8053700923919678,
+      "learning_rate": 6.076112326758684e-05,
+      "loss": 3.0026,
+      "step": 48594
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7006090879440308,
+      "learning_rate": 6.075183401474464e-05,
+      "loss": 3.0143,
+      "step": 48595
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7133437395095825,
+      "learning_rate": 6.0742545392035945e-05,
+      "loss": 2.8935,
+      "step": 48596
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7549411654472351,
+      "learning_rate": 6.073325739948538e-05,
+      "loss": 2.9046,
+      "step": 48597
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6961774826049805,
+      "learning_rate": 6.072397003711742e-05,
+      "loss": 2.9923,
+      "step": 48598
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8232783675193787,
+      "learning_rate": 6.0714683304956415e-05,
+      "loss": 2.8209,
+      "step": 48599
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7344974875450134,
+      "learning_rate": 6.070539720302694e-05,
+      "loss": 2.8822,
+      "step": 48600
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7219202518463135,
+      "learning_rate": 6.0696111731353414e-05,
+      "loss": 2.8987,
+      "step": 48601
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7357993721961975,
+      "learning_rate": 6.0686826889960185e-05,
+      "loss": 2.9248,
+      "step": 48602
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7212379574775696,
+      "learning_rate": 6.0677542678871894e-05,
+      "loss": 2.7127,
+      "step": 48603
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7509791254997253,
+      "learning_rate": 6.066825909811279e-05,
+      "loss": 2.7341,
+      "step": 48604
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6835814714431763,
+      "learning_rate": 6.065897614770753e-05,
+      "loss": 2.8247,
+      "step": 48605
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7493883371353149,
+      "learning_rate": 6.0649693827680376e-05,
+      "loss": 2.8594,
+      "step": 48606
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7076842784881592,
+      "learning_rate": 6.064041213805596e-05,
+      "loss": 2.963,
+      "step": 48607
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7497107982635498,
+      "learning_rate": 6.063113107885862e-05,
+      "loss": 2.8824,
+      "step": 48608
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7138265371322632,
+      "learning_rate": 6.0621850650112735e-05,
+      "loss": 2.7468,
+      "step": 48609
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7438095211982727,
+      "learning_rate": 6.061257085184292e-05,
+      "loss": 2.7966,
+      "step": 48610
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7051562666893005,
+      "learning_rate": 6.060329168407343e-05,
+      "loss": 2.8157,
+      "step": 48611
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7326208353042603,
+      "learning_rate": 6.059401314682877e-05,
+      "loss": 2.7358,
+      "step": 48612
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7127295732498169,
+      "learning_rate": 6.058473524013352e-05,
+      "loss": 2.9006,
+      "step": 48613
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7416107654571533,
+      "learning_rate": 6.0575457964012e-05,
+      "loss": 2.9238,
+      "step": 48614
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7051231265068054,
+      "learning_rate": 6.0566181318488614e-05,
+      "loss": 2.9534,
+      "step": 48615
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6944036483764648,
+      "learning_rate": 6.0556905303587786e-05,
+      "loss": 2.7596,
+      "step": 48616
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7137592434883118,
+      "learning_rate": 6.054762991933396e-05,
+      "loss": 2.7521,
+      "step": 48617
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7165379524230957,
+      "learning_rate": 6.053835516575168e-05,
+      "loss": 2.7684,
+      "step": 48618
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7815676331520081,
+      "learning_rate": 6.052908104286524e-05,
+      "loss": 2.6846,
+      "step": 48619
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7360538840293884,
+      "learning_rate": 6.0519807550699164e-05,
+      "loss": 2.9516,
+      "step": 48620
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7272911667823792,
+      "learning_rate": 6.0510534689277746e-05,
+      "loss": 2.9062,
+      "step": 48621
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.761579155921936,
+      "learning_rate": 6.050126245862557e-05,
+      "loss": 2.8845,
+      "step": 48622
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7632530331611633,
+      "learning_rate": 6.049199085876697e-05,
+      "loss": 2.9801,
+      "step": 48623
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8465824723243713,
+      "learning_rate": 6.04827198897263e-05,
+      "loss": 2.775,
+      "step": 48624
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7626355290412903,
+      "learning_rate": 6.047344955152812e-05,
+      "loss": 2.7857,
+      "step": 48625
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7368438839912415,
+      "learning_rate": 6.0464179844196693e-05,
+      "loss": 2.8184,
+      "step": 48626
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.72593754529953,
+      "learning_rate": 6.04549107677566e-05,
+      "loss": 2.9202,
+      "step": 48627
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7318313717842102,
+      "learning_rate": 6.04456423222321e-05,
+      "loss": 2.9295,
+      "step": 48628
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7129902243614197,
+      "learning_rate": 6.043637450764777e-05,
+      "loss": 2.8795,
+      "step": 48629
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7956576943397522,
+      "learning_rate": 6.042710732402791e-05,
+      "loss": 2.9245,
+      "step": 48630
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7019463777542114,
+      "learning_rate": 6.0417840771396854e-05,
+      "loss": 2.6454,
+      "step": 48631
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7243613600730896,
+      "learning_rate": 6.04085748497792e-05,
+      "loss": 2.916,
+      "step": 48632
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6990789771080017,
+      "learning_rate": 6.039930955919914e-05,
+      "loss": 3.0681,
+      "step": 48633
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7699520587921143,
+      "learning_rate": 6.039004489968121e-05,
+      "loss": 2.863,
+      "step": 48634
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7257212400436401,
+      "learning_rate": 6.038078087124983e-05,
+      "loss": 2.7365,
+      "step": 48635
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7219493985176086,
+      "learning_rate": 6.037151747392939e-05,
+      "loss": 2.9616,
+      "step": 48636
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7039251327514648,
+      "learning_rate": 6.036225470774425e-05,
+      "loss": 2.5781,
+      "step": 48637
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7816781401634216,
+      "learning_rate": 6.035299257271874e-05,
+      "loss": 2.8161,
+      "step": 48638
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7119601368904114,
+      "learning_rate": 6.034373106887733e-05,
+      "loss": 2.9124,
+      "step": 48639
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7373980283737183,
+      "learning_rate": 6.033447019624448e-05,
+      "loss": 2.7964,
+      "step": 48640
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6862812042236328,
+      "learning_rate": 6.032520995484442e-05,
+      "loss": 2.9101,
+      "step": 48641
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7756862640380859,
+      "learning_rate": 6.031595034470171e-05,
+      "loss": 2.986,
+      "step": 48642
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6969619393348694,
+      "learning_rate": 6.0306691365840596e-05,
+      "loss": 2.9996,
+      "step": 48643
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7498837113380432,
+      "learning_rate": 6.029743301828559e-05,
+      "loss": 2.8987,
+      "step": 48644
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7080926895141602,
+      "learning_rate": 6.028817530206104e-05,
+      "loss": 2.9105,
+      "step": 48645
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7312246561050415,
+      "learning_rate": 6.027891821719122e-05,
+      "loss": 3.199,
+      "step": 48646
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7515241503715515,
+      "learning_rate": 6.026966176370065e-05,
+      "loss": 2.9646,
+      "step": 48647
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7634444832801819,
+      "learning_rate": 6.026040594161358e-05,
+      "loss": 3.049,
+      "step": 48648
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7122647166252136,
+      "learning_rate": 6.025115075095448e-05,
+      "loss": 2.6741,
+      "step": 48649
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6937204003334045,
+      "learning_rate": 6.0241896191747774e-05,
+      "loss": 2.8746,
+      "step": 48650
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7191386222839355,
+      "learning_rate": 6.023264226401777e-05,
+      "loss": 2.843,
+      "step": 48651
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7549066543579102,
+      "learning_rate": 6.022338896778882e-05,
+      "loss": 2.9481,
+      "step": 48652
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.731148898601532,
+      "learning_rate": 6.021413630308527e-05,
+      "loss": 2.8964,
+      "step": 48653
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7206429243087769,
+      "learning_rate": 6.020488426993153e-05,
+      "loss": 2.709,
+      "step": 48654
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7050144076347351,
+      "learning_rate": 6.019563286835205e-05,
+      "loss": 2.7941,
+      "step": 48655
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7599301338195801,
+      "learning_rate": 6.0186382098371047e-05,
+      "loss": 2.9327,
+      "step": 48656
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6911978721618652,
+      "learning_rate": 6.0177131960013056e-05,
+      "loss": 2.8945,
+      "step": 48657
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7535004019737244,
+      "learning_rate": 6.016788245330231e-05,
+      "loss": 2.9582,
+      "step": 48658
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7646125555038452,
+      "learning_rate": 6.015863357826314e-05,
+      "loss": 3.0724,
+      "step": 48659
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7803003787994385,
+      "learning_rate": 6.014938533492006e-05,
+      "loss": 2.959,
+      "step": 48660
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.730233371257782,
+      "learning_rate": 6.014013772329727e-05,
+      "loss": 2.8426,
+      "step": 48661
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7188946008682251,
+      "learning_rate": 6.0130890743419225e-05,
+      "loss": 3.0968,
+      "step": 48662
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7240670323371887,
+      "learning_rate": 6.012164439531022e-05,
+      "loss": 2.8007,
+      "step": 48663
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7316474318504333,
+      "learning_rate": 6.011239867899463e-05,
+      "loss": 2.9304,
+      "step": 48664
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8154338598251343,
+      "learning_rate": 6.010315359449688e-05,
+      "loss": 2.9659,
+      "step": 48665
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7354592084884644,
+      "learning_rate": 6.0093909141841266e-05,
+      "loss": 2.9925,
+      "step": 48666
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7725946307182312,
+      "learning_rate": 6.008466532105212e-05,
+      "loss": 2.8038,
+      "step": 48667
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7260931134223938,
+      "learning_rate": 6.0075422132153705e-05,
+      "loss": 2.8687,
+      "step": 48668
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7305506467819214,
+      "learning_rate": 6.0066179575170445e-05,
+      "loss": 2.9901,
+      "step": 48669
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8036696910858154,
+      "learning_rate": 6.005693765012679e-05,
+      "loss": 2.9703,
+      "step": 48670
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7172324061393738,
+      "learning_rate": 6.004769635704691e-05,
+      "loss": 2.9001,
+      "step": 48671
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7303609251976013,
+      "learning_rate": 6.0038455695955265e-05,
+      "loss": 3.0238,
+      "step": 48672
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7004274725914001,
+      "learning_rate": 6.0029215666876194e-05,
+      "loss": 2.9343,
+      "step": 48673
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7272935509681702,
+      "learning_rate": 6.001997626983388e-05,
+      "loss": 2.738,
+      "step": 48674
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7348398566246033,
+      "learning_rate": 6.001073750485283e-05,
+      "loss": 2.7692,
+      "step": 48675
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7049981951713562,
+      "learning_rate": 6.000149937195724e-05,
+      "loss": 2.9259,
+      "step": 48676
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7831222414970398,
+      "learning_rate": 5.999226187117161e-05,
+      "loss": 2.91,
+      "step": 48677
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7655361294746399,
+      "learning_rate": 5.998302500252009e-05,
+      "loss": 2.6398,
+      "step": 48678
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7228807210922241,
+      "learning_rate": 5.997378876602713e-05,
+      "loss": 2.789,
+      "step": 48679
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7596110105514526,
+      "learning_rate": 5.996455316171708e-05,
+      "loss": 2.8773,
+      "step": 48680
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7604156136512756,
+      "learning_rate": 5.995531818961408e-05,
+      "loss": 2.9434,
+      "step": 48681
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7096068859100342,
+      "learning_rate": 5.9946083849742645e-05,
+      "loss": 2.9145,
+      "step": 48682
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7317915558815002,
+      "learning_rate": 5.993685014212696e-05,
+      "loss": 2.9156,
+      "step": 48683
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6918458342552185,
+      "learning_rate": 5.992761706679143e-05,
+      "loss": 2.7525,
+      "step": 48684
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7611396908760071,
+      "learning_rate": 5.991838462376042e-05,
+      "loss": 2.9744,
+      "step": 48685
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7494409680366516,
+      "learning_rate": 5.990915281305809e-05,
+      "loss": 3.0139,
+      "step": 48686
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7255983352661133,
+      "learning_rate": 5.9899921634708936e-05,
+      "loss": 3.0039,
+      "step": 48687
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7247331738471985,
+      "learning_rate": 5.9890691088737165e-05,
+      "loss": 2.9438,
+      "step": 48688
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7208533883094788,
+      "learning_rate": 5.988146117516702e-05,
+      "loss": 2.7187,
+      "step": 48689
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7259299755096436,
+      "learning_rate": 5.987223189402299e-05,
+      "loss": 3.0192,
+      "step": 48690
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7051576375961304,
+      "learning_rate": 5.986300324532922e-05,
+      "loss": 3.1708,
+      "step": 48691
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8053131699562073,
+      "learning_rate": 5.985377522911016e-05,
+      "loss": 2.8649,
+      "step": 48692
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7644606828689575,
+      "learning_rate": 5.984454784538994e-05,
+      "loss": 2.9055,
+      "step": 48693
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7201364636421204,
+      "learning_rate": 5.9835321094193035e-05,
+      "loss": 2.8487,
+      "step": 48694
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7057961821556091,
+      "learning_rate": 5.9826094975543705e-05,
+      "loss": 3.0138,
+      "step": 48695
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7159346342086792,
+      "learning_rate": 5.9816869489466135e-05,
+      "loss": 2.6607,
+      "step": 48696
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7882915735244751,
+      "learning_rate": 5.9807644635984775e-05,
+      "loss": 2.9877,
+      "step": 48697
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7171709537506104,
+      "learning_rate": 5.9798420415123774e-05,
+      "loss": 2.8216,
+      "step": 48698
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7261692881584167,
+      "learning_rate": 5.97891968269075e-05,
+      "loss": 2.7935,
+      "step": 48699
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7348390221595764,
+      "learning_rate": 5.977997387136032e-05,
+      "loss": 3.0635,
+      "step": 48700
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7532333731651306,
+      "learning_rate": 5.97707515485064e-05,
+      "loss": 2.9131,
+      "step": 48701
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6886923313140869,
+      "learning_rate": 5.976152985837026e-05,
+      "loss": 2.7078,
+      "step": 48702
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7542856335639954,
+      "learning_rate": 5.975230880097581e-05,
+      "loss": 2.6983,
+      "step": 48703
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7375133037567139,
+      "learning_rate": 5.974308837634767e-05,
+      "loss": 2.6483,
+      "step": 48704
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7276037335395813,
+      "learning_rate": 5.973386858450988e-05,
+      "loss": 2.8235,
+      "step": 48705
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7167425751686096,
+      "learning_rate": 5.972464942548686e-05,
+      "loss": 2.6771,
+      "step": 48706
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.760288655757904,
+      "learning_rate": 5.9715430899302954e-05,
+      "loss": 2.7104,
+      "step": 48707
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7757988572120667,
+      "learning_rate": 5.970621300598227e-05,
+      "loss": 2.8529,
+      "step": 48708
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7403610348701477,
+      "learning_rate": 5.969699574554927e-05,
+      "loss": 2.9158,
+      "step": 48709
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7176191210746765,
+      "learning_rate": 5.968777911802812e-05,
+      "loss": 3.0029,
+      "step": 48710
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.711616039276123,
+      "learning_rate": 5.9678563123443045e-05,
+      "loss": 3.1245,
+      "step": 48711
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7264795303344727,
+      "learning_rate": 5.9669347761818455e-05,
+      "loss": 2.7855,
+      "step": 48712
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7248842120170593,
+      "learning_rate": 5.966013303317846e-05,
+      "loss": 2.8838,
+      "step": 48713
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7030981183052063,
+      "learning_rate": 5.965091893754751e-05,
+      "loss": 3.0203,
+      "step": 48714
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7812050580978394,
+      "learning_rate": 5.964170547494972e-05,
+      "loss": 2.9395,
+      "step": 48715
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.762490451335907,
+      "learning_rate": 5.9632492645409467e-05,
+      "loss": 2.9752,
+      "step": 48716
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.68095463514328,
+      "learning_rate": 5.962328044895097e-05,
+      "loss": 2.803,
+      "step": 48717
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6975252628326416,
+      "learning_rate": 5.961406888559844e-05,
+      "loss": 3.1397,
+      "step": 48718
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7371811866760254,
+      "learning_rate": 5.960485795537626e-05,
+      "loss": 2.9046,
+      "step": 48719
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7280109524726868,
+      "learning_rate": 5.959564765830854e-05,
+      "loss": 2.8761,
+      "step": 48720
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7430052161216736,
+      "learning_rate": 5.95864379944196e-05,
+      "loss": 2.8337,
+      "step": 48721
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7345473170280457,
+      "learning_rate": 5.957722896373381e-05,
+      "loss": 2.9135,
+      "step": 48722
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7026997208595276,
+      "learning_rate": 5.9568020566275235e-05,
+      "loss": 2.9278,
+      "step": 48723
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7376667857170105,
+      "learning_rate": 5.9558812802068335e-05,
+      "loss": 2.7952,
+      "step": 48724
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7143561840057373,
+      "learning_rate": 5.95496056711372e-05,
+      "loss": 2.7486,
+      "step": 48725
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7290409803390503,
+      "learning_rate": 5.954039917350608e-05,
+      "loss": 2.697,
+      "step": 48726
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7321444749832153,
+      "learning_rate": 5.953119330919935e-05,
+      "loss": 2.8251,
+      "step": 48727
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7561767101287842,
+      "learning_rate": 5.95219880782411e-05,
+      "loss": 2.7957,
+      "step": 48728
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.750032365322113,
+      "learning_rate": 5.95127834806557e-05,
+      "loss": 3.2731,
+      "step": 48729
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7444902658462524,
+      "learning_rate": 5.95035795164673e-05,
+      "loss": 2.8988,
+      "step": 48730
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.735755205154419,
+      "learning_rate": 5.9494376185700256e-05,
+      "loss": 2.9861,
+      "step": 48731
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7328507900238037,
+      "learning_rate": 5.948517348837874e-05,
+      "loss": 2.9542,
+      "step": 48732
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7095339894294739,
+      "learning_rate": 5.947597142452693e-05,
+      "loss": 2.7841,
+      "step": 48733
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7547805309295654,
+      "learning_rate": 5.946676999416917e-05,
+      "loss": 2.9635,
+      "step": 48734
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6976374387741089,
+      "learning_rate": 5.945756919732956e-05,
+      "loss": 2.8442,
+      "step": 48735
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7196562886238098,
+      "learning_rate": 5.944836903403246e-05,
+      "loss": 3.1681,
+      "step": 48736
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7531615495681763,
+      "learning_rate": 5.94391695043021e-05,
+      "loss": 3.1516,
+      "step": 48737
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7510145306587219,
+      "learning_rate": 5.942997060816268e-05,
+      "loss": 2.8634,
+      "step": 48738
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7500221133232117,
+      "learning_rate": 5.9420772345638423e-05,
+      "loss": 3.1042,
+      "step": 48739
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7888630032539368,
+      "learning_rate": 5.9411574716753474e-05,
+      "loss": 2.7464,
+      "step": 48740
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7947269082069397,
+      "learning_rate": 5.9402377721532144e-05,
+      "loss": 2.9241,
+      "step": 48741
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6809054613113403,
+      "learning_rate": 5.939318135999872e-05,
+      "loss": 2.7824,
+      "step": 48742
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.785391628742218,
+      "learning_rate": 5.938398563217728e-05,
+      "loss": 2.9763,
+      "step": 48743
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7565441131591797,
+      "learning_rate": 5.9374790538092166e-05,
+      "loss": 3.02,
+      "step": 48744
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.719491720199585,
+      "learning_rate": 5.93655960777675e-05,
+      "loss": 3.204,
+      "step": 48745
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7319931983947754,
+      "learning_rate": 5.9356402251227585e-05,
+      "loss": 2.8652,
+      "step": 48746
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7423975467681885,
+      "learning_rate": 5.9347209058496615e-05,
+      "loss": 2.956,
+      "step": 48747
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7304463386535645,
+      "learning_rate": 5.9338016499598725e-05,
+      "loss": 2.9522,
+      "step": 48748
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7549775838851929,
+      "learning_rate": 5.932882457455821e-05,
+      "loss": 2.9967,
+      "step": 48749
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6921327710151672,
+      "learning_rate": 5.93196332833992e-05,
+      "loss": 2.9852,
+      "step": 48750
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7499932646751404,
+      "learning_rate": 5.931044262614596e-05,
+      "loss": 3.0477,
+      "step": 48751
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7360295057296753,
+      "learning_rate": 5.9301252602822794e-05,
+      "loss": 3.2046,
+      "step": 48752
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.70384281873703,
+      "learning_rate": 5.929206321345378e-05,
+      "loss": 2.6769,
+      "step": 48753
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7233741879463196,
+      "learning_rate": 5.9282874458063136e-05,
+      "loss": 2.8938,
+      "step": 48754
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7333195805549622,
+      "learning_rate": 5.9273686336675007e-05,
+      "loss": 2.8446,
+      "step": 48755
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7375016808509827,
+      "learning_rate": 5.9264498849313644e-05,
+      "loss": 3.0102,
+      "step": 48756
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7300297617912292,
+      "learning_rate": 5.925531199600336e-05,
+      "loss": 2.8862,
+      "step": 48757
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7252567410469055,
+      "learning_rate": 5.924612577676817e-05,
+      "loss": 2.9107,
+      "step": 48758
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7813132405281067,
+      "learning_rate": 5.923694019163242e-05,
+      "loss": 2.8536,
+      "step": 48759
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.714250385761261,
+      "learning_rate": 5.9227755240620254e-05,
+      "loss": 2.9715,
+      "step": 48760
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7643177509307861,
+      "learning_rate": 5.921857092375576e-05,
+      "loss": 2.942,
+      "step": 48761
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7428061366081238,
+      "learning_rate": 5.920938724106328e-05,
+      "loss": 2.9505,
+      "step": 48762
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7188575267791748,
+      "learning_rate": 5.9200204192566856e-05,
+      "loss": 2.9623,
+      "step": 48763
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7388699650764465,
+      "learning_rate": 5.9191021778290817e-05,
+      "loss": 3.0572,
+      "step": 48764
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7034398913383484,
+      "learning_rate": 5.918183999825923e-05,
+      "loss": 2.8113,
+      "step": 48765
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7972329258918762,
+      "learning_rate": 5.917265885249631e-05,
+      "loss": 3.0816,
+      "step": 48766
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7589764595031738,
+      "learning_rate": 5.916347834102635e-05,
+      "loss": 2.9365,
+      "step": 48767
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7065830826759338,
+      "learning_rate": 5.915429846387342e-05,
+      "loss": 2.9464,
+      "step": 48768
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8015071153640747,
+      "learning_rate": 5.914511922106173e-05,
+      "loss": 3.0396,
+      "step": 48769
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7204930186271667,
+      "learning_rate": 5.913594061261534e-05,
+      "loss": 2.7337,
+      "step": 48770
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7478088140487671,
+      "learning_rate": 5.9126762638558556e-05,
+      "loss": 2.9955,
+      "step": 48771
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7554129362106323,
+      "learning_rate": 5.911758529891559e-05,
+      "loss": 2.9007,
+      "step": 48772
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7005829811096191,
+      "learning_rate": 5.9108408593710465e-05,
+      "loss": 2.9987,
+      "step": 48773
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7382482290267944,
+      "learning_rate": 5.909923252296752e-05,
+      "loss": 3.1134,
+      "step": 48774
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8164120316505432,
+      "learning_rate": 5.909005708671084e-05,
+      "loss": 2.9471,
+      "step": 48775
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.706451416015625,
+      "learning_rate": 5.9080882284964506e-05,
+      "loss": 3.1401,
+      "step": 48776
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7515252232551575,
+      "learning_rate": 5.907170811775283e-05,
+      "loss": 2.8985,
+      "step": 48777
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6926350593566895,
+      "learning_rate": 5.906253458509982e-05,
+      "loss": 2.9885,
+      "step": 48778
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7306018471717834,
+      "learning_rate": 5.90533616870298e-05,
+      "loss": 2.8483,
+      "step": 48779
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6878507733345032,
+      "learning_rate": 5.904418942356678e-05,
+      "loss": 2.9312,
+      "step": 48780
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.728437602519989,
+      "learning_rate": 5.9035017794735107e-05,
+      "loss": 2.9679,
+      "step": 48781
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.710863471031189,
+      "learning_rate": 5.902584680055876e-05,
+      "loss": 2.9908,
+      "step": 48782
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7090634703636169,
+      "learning_rate": 5.901667644106193e-05,
+      "loss": 3.095,
+      "step": 48783
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7385855913162231,
+      "learning_rate": 5.900750671626885e-05,
+      "loss": 3.1334,
+      "step": 48784
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.705529510974884,
+      "learning_rate": 5.899833762620354e-05,
+      "loss": 2.8821,
+      "step": 48785
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.716135561466217,
+      "learning_rate": 5.8989169170890316e-05,
+      "loss": 2.8941,
+      "step": 48786
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.690018355846405,
+      "learning_rate": 5.898000135035316e-05,
+      "loss": 2.8815,
+      "step": 48787
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7214192748069763,
+      "learning_rate": 5.8970834164616273e-05,
+      "loss": 2.8851,
+      "step": 48788
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8130916357040405,
+      "learning_rate": 5.8961667613703923e-05,
+      "loss": 2.9957,
+      "step": 48789
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7414829730987549,
+      "learning_rate": 5.8952501697640145e-05,
+      "loss": 2.6515,
+      "step": 48790
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7804091572761536,
+      "learning_rate": 5.894333641644908e-05,
+      "loss": 2.7691,
+      "step": 48791
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7167255878448486,
+      "learning_rate": 5.8934171770154824e-05,
+      "loss": 2.9098,
+      "step": 48792
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7627856135368347,
+      "learning_rate": 5.892500775878155e-05,
+      "loss": 3.1166,
+      "step": 48793
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7267736792564392,
+      "learning_rate": 5.8915844382353474e-05,
+      "loss": 3.0016,
+      "step": 48794
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7814797163009644,
+      "learning_rate": 5.89066816408946e-05,
+      "loss": 2.6712,
+      "step": 48795
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7293899059295654,
+      "learning_rate": 5.889751953442919e-05,
+      "loss": 2.9074,
+      "step": 48796
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7417458891868591,
+      "learning_rate": 5.888835806298132e-05,
+      "loss": 2.8068,
+      "step": 48797
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.742914080619812,
+      "learning_rate": 5.8879197226575034e-05,
+      "loss": 2.6278,
+      "step": 48798
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7378799915313721,
+      "learning_rate": 5.887003702523462e-05,
+      "loss": 2.7636,
+      "step": 48799
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7445825338363647,
+      "learning_rate": 5.886087745898405e-05,
+      "loss": 2.9709,
+      "step": 48800
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7399541139602661,
+      "learning_rate": 5.88517185278476e-05,
+      "loss": 2.699,
+      "step": 48801
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7600640654563904,
+      "learning_rate": 5.884256023184924e-05,
+      "loss": 2.9231,
+      "step": 48802
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7569655776023865,
+      "learning_rate": 5.883340257101312e-05,
+      "loss": 3.065,
+      "step": 48803
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7483676671981812,
+      "learning_rate": 5.8824245545363615e-05,
+      "loss": 2.8574,
+      "step": 48804
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7173334360122681,
+      "learning_rate": 5.8815089154924456e-05,
+      "loss": 2.865,
+      "step": 48805
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7077659964561462,
+      "learning_rate": 5.8805933399719993e-05,
+      "loss": 2.8537,
+      "step": 48806
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7615037560462952,
+      "learning_rate": 5.879677827977425e-05,
+      "loss": 2.7053,
+      "step": 48807
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7203567028045654,
+      "learning_rate": 5.878762379511133e-05,
+      "loss": 2.8675,
+      "step": 48808
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7756044864654541,
+      "learning_rate": 5.877846994575548e-05,
+      "loss": 2.9412,
+      "step": 48809
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7094483971595764,
+      "learning_rate": 5.876931673173062e-05,
+      "loss": 2.9699,
+      "step": 48810
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7442003488540649,
+      "learning_rate": 5.8760164153061064e-05,
+      "loss": 2.912,
+      "step": 48811
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7222998142242432,
+      "learning_rate": 5.8751012209770786e-05,
+      "loss": 2.8013,
+      "step": 48812
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7189565300941467,
+      "learning_rate": 5.8741860901883844e-05,
+      "loss": 2.8706,
+      "step": 48813
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7589866518974304,
+      "learning_rate": 5.8732710229424507e-05,
+      "loss": 3.0814,
+      "step": 48814
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7047926783561707,
+      "learning_rate": 5.8723560192416665e-05,
+      "loss": 2.7251,
+      "step": 48815
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7266496419906616,
+      "learning_rate": 5.8714410790884626e-05,
+      "loss": 2.8967,
+      "step": 48816
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7001505494117737,
+      "learning_rate": 5.8705262024852304e-05,
+      "loss": 2.843,
+      "step": 48817
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7681334018707275,
+      "learning_rate": 5.869611389434399e-05,
+      "loss": 2.7958,
+      "step": 48818
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7098401784896851,
+      "learning_rate": 5.868696639938365e-05,
+      "loss": 2.9353,
+      "step": 48819
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7518061995506287,
+      "learning_rate": 5.867781953999531e-05,
+      "loss": 2.8312,
+      "step": 48820
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7705138325691223,
+      "learning_rate": 5.866867331620324e-05,
+      "loss": 2.8373,
+      "step": 48821
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7576525211334229,
+      "learning_rate": 5.865952772803136e-05,
+      "loss": 2.8809,
+      "step": 48822
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7420570850372314,
+      "learning_rate": 5.8650382775503855e-05,
+      "loss": 3.1835,
+      "step": 48823
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7061996459960938,
+      "learning_rate": 5.8641238458644835e-05,
+      "loss": 2.9064,
+      "step": 48824
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7823624610900879,
+      "learning_rate": 5.8632094777478275e-05,
+      "loss": 2.9696,
+      "step": 48825
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7167471647262573,
+      "learning_rate": 5.8622951732028435e-05,
+      "loss": 3.052,
+      "step": 48826
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7196730375289917,
+      "learning_rate": 5.861380932231925e-05,
+      "loss": 2.752,
+      "step": 48827
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.758525013923645,
+      "learning_rate": 5.8604667548374776e-05,
+      "loss": 2.8304,
+      "step": 48828
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7166337370872498,
+      "learning_rate": 5.859552641021922e-05,
+      "loss": 3.0448,
+      "step": 48829
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6951048970222473,
+      "learning_rate": 5.8586385907876534e-05,
+      "loss": 2.9832,
+      "step": 48830
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7193121314048767,
+      "learning_rate": 5.8577246041370896e-05,
+      "loss": 2.9533,
+      "step": 48831
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7775723934173584,
+      "learning_rate": 5.856810681072626e-05,
+      "loss": 2.7201,
+      "step": 48832
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7279043793678284,
+      "learning_rate": 5.855896821596683e-05,
+      "loss": 2.7716,
+      "step": 48833
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7554843425750732,
+      "learning_rate": 5.854983025711662e-05,
+      "loss": 2.8317,
+      "step": 48834
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7403421401977539,
+      "learning_rate": 5.854069293419963e-05,
+      "loss": 3.1781,
+      "step": 48835
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.70306396484375,
+      "learning_rate": 5.8531556247240054e-05,
+      "loss": 2.8058,
+      "step": 48836
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.716789960861206,
+      "learning_rate": 5.852242019626181e-05,
+      "loss": 2.9193,
+      "step": 48837
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7575057148933411,
+      "learning_rate": 5.851328478128902e-05,
+      "loss": 2.7854,
+      "step": 48838
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7853378057479858,
+      "learning_rate": 5.850415000234585e-05,
+      "loss": 3.17,
+      "step": 48839
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7358184456825256,
+      "learning_rate": 5.8495015859456294e-05,
+      "loss": 2.8608,
+      "step": 48840
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6891388893127441,
+      "learning_rate": 5.8485882352644366e-05,
+      "loss": 2.8855,
+      "step": 48841
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7379217743873596,
+      "learning_rate": 5.847674948193407e-05,
+      "loss": 2.6099,
+      "step": 48842
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7153098583221436,
+      "learning_rate": 5.846761724734953e-05,
+      "loss": 2.8372,
+      "step": 48843
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.750810980796814,
+      "learning_rate": 5.845848564891489e-05,
+      "loss": 2.8569,
+      "step": 48844
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7509940266609192,
+      "learning_rate": 5.8449354686654025e-05,
+      "loss": 2.6827,
+      "step": 48845
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7342125773429871,
+      "learning_rate": 5.844022436059116e-05,
+      "loss": 2.7864,
+      "step": 48846
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7348741292953491,
+      "learning_rate": 5.8431094670750165e-05,
+      "loss": 2.8741,
+      "step": 48847
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7125906348228455,
+      "learning_rate": 5.842196561715526e-05,
+      "loss": 3.0068,
+      "step": 48848
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7036176323890686,
+      "learning_rate": 5.84128371998304e-05,
+      "loss": 2.9852,
+      "step": 48849
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7853854298591614,
+      "learning_rate": 5.8403709418799546e-05,
+      "loss": 2.9464,
+      "step": 48850
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7350966930389404,
+      "learning_rate": 5.8394582274086935e-05,
+      "loss": 2.9554,
+      "step": 48851
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7737547755241394,
+      "learning_rate": 5.838545576571639e-05,
+      "loss": 2.9423,
+      "step": 48852
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7212133407592773,
+      "learning_rate": 5.837632989371205e-05,
+      "loss": 2.8444,
+      "step": 48853
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7580508589744568,
+      "learning_rate": 5.836720465809808e-05,
+      "loss": 2.9049,
+      "step": 48854
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7183226943016052,
+      "learning_rate": 5.8358080058898326e-05,
+      "loss": 2.894,
+      "step": 48855
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7060414552688599,
+      "learning_rate": 5.8348956096136945e-05,
+      "loss": 2.6954,
+      "step": 48856
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7525982856750488,
+      "learning_rate": 5.833983276983778e-05,
+      "loss": 3.1094,
+      "step": 48857
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7521215677261353,
+      "learning_rate": 5.833071008002508e-05,
+      "loss": 2.8246,
+      "step": 48858
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7773438096046448,
+      "learning_rate": 5.832158802672272e-05,
+      "loss": 2.7105,
+      "step": 48859
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7317736148834229,
+      "learning_rate": 5.8312466609954756e-05,
+      "loss": 3.0221,
+      "step": 48860
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7350138425827026,
+      "learning_rate": 5.830334582974533e-05,
+      "loss": 2.9349,
+      "step": 48861
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7256360650062561,
+      "learning_rate": 5.829422568611839e-05,
+      "loss": 2.9305,
+      "step": 48862
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7844482660293579,
+      "learning_rate": 5.828510617909795e-05,
+      "loss": 3.0334,
+      "step": 48863
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7349040508270264,
+      "learning_rate": 5.827598730870793e-05,
+      "loss": 2.9445,
+      "step": 48864
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7633810639381409,
+      "learning_rate": 5.8266869074972436e-05,
+      "loss": 2.8846,
+      "step": 48865
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7389031052589417,
+      "learning_rate": 5.825775147791555e-05,
+      "loss": 2.833,
+      "step": 48866
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7971253395080566,
+      "learning_rate": 5.8248634517561156e-05,
+      "loss": 2.8842,
+      "step": 48867
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7472830414772034,
+      "learning_rate": 5.82395181939334e-05,
+      "loss": 2.6718,
+      "step": 48868
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.0377117395401,
+      "learning_rate": 5.8230402507056165e-05,
+      "loss": 2.8771,
+      "step": 48869
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7521938681602478,
+      "learning_rate": 5.822128745695359e-05,
+      "loss": 2.8167,
+      "step": 48870
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7492688298225403,
+      "learning_rate": 5.821217304364963e-05,
+      "loss": 2.8992,
+      "step": 48871
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7289814949035645,
+      "learning_rate": 5.820305926716817e-05,
+      "loss": 2.9214,
+      "step": 48872
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6818594336509705,
+      "learning_rate": 5.819394612753342e-05,
+      "loss": 2.7029,
+      "step": 48873
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7078396677970886,
+      "learning_rate": 5.818483362476919e-05,
+      "loss": 2.6411,
+      "step": 48874
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7174808979034424,
+      "learning_rate": 5.8175721758899566e-05,
+      "loss": 2.9228,
+      "step": 48875
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7881730794906616,
+      "learning_rate": 5.8166610529948624e-05,
+      "loss": 2.882,
+      "step": 48876
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7044768929481506,
+      "learning_rate": 5.8157499937940275e-05,
+      "loss": 2.976,
+      "step": 48877
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.72385573387146,
+      "learning_rate": 5.814838998289855e-05,
+      "loss": 2.8273,
+      "step": 48878
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.68699049949646,
+      "learning_rate": 5.813928066484734e-05,
+      "loss": 2.8484,
+      "step": 48879
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7593944072723389,
+      "learning_rate": 5.813017198381071e-05,
+      "loss": 2.8658,
+      "step": 48880
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7795584201812744,
+      "learning_rate": 5.8121063939812776e-05,
+      "loss": 2.9373,
+      "step": 48881
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7256165742874146,
+      "learning_rate": 5.811195653287727e-05,
+      "loss": 2.8412,
+      "step": 48882
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8968254923820496,
+      "learning_rate": 5.8102849763028444e-05,
+      "loss": 2.9702,
+      "step": 48883
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7178112864494324,
+      "learning_rate": 5.8093743630290125e-05,
+      "loss": 2.5675,
+      "step": 48884
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.75639408826828,
+      "learning_rate": 5.808463813468628e-05,
+      "loss": 2.9394,
+      "step": 48885
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7796924114227295,
+      "learning_rate": 5.807553327624099e-05,
+      "loss": 2.9314,
+      "step": 48886
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7454410195350647,
+      "learning_rate": 5.806642905497811e-05,
+      "loss": 3.0443,
+      "step": 48887
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7198060750961304,
+      "learning_rate": 5.805732547092179e-05,
+      "loss": 2.7605,
+      "step": 48888
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6760872006416321,
+      "learning_rate": 5.8048222524095833e-05,
+      "loss": 2.7732,
+      "step": 48889
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7360389828681946,
+      "learning_rate": 5.8039120214524293e-05,
+      "loss": 2.6977,
+      "step": 48890
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7301698327064514,
+      "learning_rate": 5.803001854223122e-05,
+      "loss": 2.8464,
+      "step": 48891
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7200692296028137,
+      "learning_rate": 5.8020917507240516e-05,
+      "loss": 3.0045,
+      "step": 48892
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7441954016685486,
+      "learning_rate": 5.8011817109576145e-05,
+      "loss": 3.1825,
+      "step": 48893
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7359301447868347,
+      "learning_rate": 5.800271734926197e-05,
+      "loss": 2.9192,
+      "step": 48894
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7263439893722534,
+      "learning_rate": 5.799361822632209e-05,
+      "loss": 2.9108,
+      "step": 48895
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7272346615791321,
+      "learning_rate": 5.798451974078051e-05,
+      "loss": 2.8907,
+      "step": 48896
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7894495129585266,
+      "learning_rate": 5.7975421892661045e-05,
+      "loss": 2.9208,
+      "step": 48897
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8710970282554626,
+      "learning_rate": 5.796632468198784e-05,
+      "loss": 3.0931,
+      "step": 48898
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6832615733146667,
+      "learning_rate": 5.795722810878475e-05,
+      "loss": 2.9985,
+      "step": 48899
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7081145644187927,
+      "learning_rate": 5.794813217307565e-05,
+      "loss": 3.0453,
+      "step": 48900
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7711806893348694,
+      "learning_rate": 5.793903687488468e-05,
+      "loss": 2.8482,
+      "step": 48901
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8075990676879883,
+      "learning_rate": 5.79299422142356e-05,
+      "loss": 3.1117,
+      "step": 48902
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7430427670478821,
+      "learning_rate": 5.7920848191152555e-05,
+      "loss": 2.9316,
+      "step": 48903
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7735027074813843,
+      "learning_rate": 5.791175480565936e-05,
+      "loss": 2.8317,
+      "step": 48904
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7572428584098816,
+      "learning_rate": 5.790266205778006e-05,
+      "loss": 2.9534,
+      "step": 48905
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.721071720123291,
+      "learning_rate": 5.789356994753853e-05,
+      "loss": 2.9628,
+      "step": 48906
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7726892828941345,
+      "learning_rate": 5.788447847495869e-05,
+      "loss": 2.8635,
+      "step": 48907
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7207630276679993,
+      "learning_rate": 5.787538764006463e-05,
+      "loss": 2.9928,
+      "step": 48908
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7308686971664429,
+      "learning_rate": 5.786629744288012e-05,
+      "loss": 2.8482,
+      "step": 48909
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8075633645057678,
+      "learning_rate": 5.785720788342917e-05,
+      "loss": 2.8302,
+      "step": 48910
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7197324633598328,
+      "learning_rate": 5.78481189617358e-05,
+      "loss": 2.9422,
+      "step": 48911
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7496452927589417,
+      "learning_rate": 5.783903067782383e-05,
+      "loss": 2.7668,
+      "step": 48912
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7587379813194275,
+      "learning_rate": 5.7829943031717295e-05,
+      "loss": 2.8343,
+      "step": 48913
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7361343502998352,
+      "learning_rate": 5.782085602344012e-05,
+      "loss": 2.9595,
+      "step": 48914
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6897303462028503,
+      "learning_rate": 5.7811769653016117e-05,
+      "loss": 2.8756,
+      "step": 48915
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7227274179458618,
+      "learning_rate": 5.7802683920469364e-05,
+      "loss": 3.046,
+      "step": 48916
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7711227536201477,
+      "learning_rate": 5.779359882582364e-05,
+      "loss": 3.0059,
+      "step": 48917
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6948781609535217,
+      "learning_rate": 5.7784514369103064e-05,
+      "loss": 2.9202,
+      "step": 48918
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7455137968063354,
+      "learning_rate": 5.777543055033138e-05,
+      "loss": 2.9468,
+      "step": 48919
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.738844096660614,
+      "learning_rate": 5.776634736953267e-05,
+      "loss": 3.0334,
+      "step": 48920
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7826026082038879,
+      "learning_rate": 5.7757264826730786e-05,
+      "loss": 2.8642,
+      "step": 48921
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7317241430282593,
+      "learning_rate": 5.774818292194957e-05,
+      "loss": 2.8831,
+      "step": 48922
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.745648980140686,
+      "learning_rate": 5.773910165521311e-05,
+      "loss": 3.1187,
+      "step": 48923
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7066932916641235,
+      "learning_rate": 5.7730021026545107e-05,
+      "loss": 2.9324,
+      "step": 48924
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.719154417514801,
+      "learning_rate": 5.772094103596965e-05,
+      "loss": 2.9116,
+      "step": 48925
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7419759631156921,
+      "learning_rate": 5.7711861683510654e-05,
+      "loss": 2.7544,
+      "step": 48926
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7109057307243347,
+      "learning_rate": 5.770278296919193e-05,
+      "loss": 2.6735,
+      "step": 48927
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6973155736923218,
+      "learning_rate": 5.769370489303749e-05,
+      "loss": 2.9472,
+      "step": 48928
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7412685751914978,
+      "learning_rate": 5.768462745507123e-05,
+      "loss": 2.96,
+      "step": 48929
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.726198136806488,
+      "learning_rate": 5.767555065531695e-05,
+      "loss": 2.9192,
+      "step": 48930
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7647451162338257,
+      "learning_rate": 5.76664744937987e-05,
+      "loss": 2.9987,
+      "step": 48931
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.790251612663269,
+      "learning_rate": 5.7657398970540226e-05,
+      "loss": 2.8102,
+      "step": 48932
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7141204476356506,
+      "learning_rate": 5.764832408556561e-05,
+      "loss": 3.3217,
+      "step": 48933
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.742522120475769,
+      "learning_rate": 5.76392498388986e-05,
+      "loss": 2.9367,
+      "step": 48934
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6904557347297668,
+      "learning_rate": 5.763017623056325e-05,
+      "loss": 2.5379,
+      "step": 48935
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6816965937614441,
+      "learning_rate": 5.762110326058337e-05,
+      "loss": 2.8829,
+      "step": 48936
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7373006343841553,
+      "learning_rate": 5.761203092898277e-05,
+      "loss": 2.9315,
+      "step": 48937
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7109895348548889,
+      "learning_rate": 5.76029592357855e-05,
+      "loss": 2.756,
+      "step": 48938
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7374069690704346,
+      "learning_rate": 5.75938881810153e-05,
+      "loss": 3.042,
+      "step": 48939
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7400507926940918,
+      "learning_rate": 5.758481776469623e-05,
+      "loss": 2.9827,
+      "step": 48940
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7280579209327698,
+      "learning_rate": 5.757574798685204e-05,
+      "loss": 2.8857,
+      "step": 48941
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7616007924079895,
+      "learning_rate": 5.756667884750673e-05,
+      "loss": 2.8854,
+      "step": 48942
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7113364338874817,
+      "learning_rate": 5.755761034668413e-05,
+      "loss": 3.0651,
+      "step": 48943
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7248356938362122,
+      "learning_rate": 5.754854248440807e-05,
+      "loss": 2.927,
+      "step": 48944
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7483928799629211,
+      "learning_rate": 5.753947526070254e-05,
+      "loss": 2.9758,
+      "step": 48945
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7303197383880615,
+      "learning_rate": 5.753040867559132e-05,
+      "loss": 3.0986,
+      "step": 48946
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7678343653678894,
+      "learning_rate": 5.752134272909833e-05,
+      "loss": 2.8695,
+      "step": 48947
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7396532893180847,
+      "learning_rate": 5.751227742124751e-05,
+      "loss": 3.1046,
+      "step": 48948
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7055127620697021,
+      "learning_rate": 5.7503212752062646e-05,
+      "loss": 2.9769,
+      "step": 48949
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8863729238510132,
+      "learning_rate": 5.7494148721567676e-05,
+      "loss": 2.8523,
+      "step": 48950
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7113443613052368,
+      "learning_rate": 5.748508532978649e-05,
+      "loss": 2.8839,
+      "step": 48951
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7774294018745422,
+      "learning_rate": 5.747602257674284e-05,
+      "loss": 2.8566,
+      "step": 48952
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.747187614440918,
+      "learning_rate": 5.746696046246073e-05,
+      "loss": 2.8786,
+      "step": 48953
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6938170790672302,
+      "learning_rate": 5.745789898696388e-05,
+      "loss": 2.8971,
+      "step": 48954
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7477014064788818,
+      "learning_rate": 5.7448838150276366e-05,
+      "loss": 2.9451,
+      "step": 48955
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7510203123092651,
+      "learning_rate": 5.7439777952421805e-05,
+      "loss": 2.918,
+      "step": 48956
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7017959356307983,
+      "learning_rate": 5.7430718393424315e-05,
+      "loss": 2.7877,
+      "step": 48957
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7510301470756531,
+      "learning_rate": 5.7421659473307605e-05,
+      "loss": 2.9053,
+      "step": 48958
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6991977095603943,
+      "learning_rate": 5.7412601192095454e-05,
+      "loss": 3.0594,
+      "step": 48959
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6977912783622742,
+      "learning_rate": 5.740354354981195e-05,
+      "loss": 3.0323,
+      "step": 48960
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7294617295265198,
+      "learning_rate": 5.73944865464807e-05,
+      "loss": 3.0129,
+      "step": 48961
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7957404255867004,
+      "learning_rate": 5.738543018212573e-05,
+      "loss": 2.7911,
+      "step": 48962
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.850426971912384,
+      "learning_rate": 5.73763744567709e-05,
+      "loss": 2.9213,
+      "step": 48963
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7113354802131653,
+      "learning_rate": 5.7367319370440014e-05,
+      "loss": 2.8616,
+      "step": 48964
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.77390456199646,
+      "learning_rate": 5.735826492315687e-05,
+      "loss": 3.1204,
+      "step": 48965
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.734718382358551,
+      "learning_rate": 5.7349211114945325e-05,
+      "loss": 2.8761,
+      "step": 48966
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9392869472503662,
+      "learning_rate": 5.734015794582926e-05,
+      "loss": 2.9756,
+      "step": 48967
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7067123055458069,
+      "learning_rate": 5.733110541583258e-05,
+      "loss": 3.0634,
+      "step": 48968
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7297945618629456,
+      "learning_rate": 5.732205352497901e-05,
+      "loss": 2.9535,
+      "step": 48969
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7064141631126404,
+      "learning_rate": 5.731300227329249e-05,
+      "loss": 3.0467,
+      "step": 48970
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7072996497154236,
+      "learning_rate": 5.730395166079677e-05,
+      "loss": 2.9136,
+      "step": 48971
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.749560534954071,
+      "learning_rate": 5.7294901687515795e-05,
+      "loss": 2.8613,
+      "step": 48972
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.735333263874054,
+      "learning_rate": 5.7285852353473315e-05,
+      "loss": 2.983,
+      "step": 48973
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7418922185897827,
+      "learning_rate": 5.727680365869315e-05,
+      "loss": 3.1744,
+      "step": 48974
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7562124133110046,
+      "learning_rate": 5.726775560319924e-05,
+      "loss": 2.8761,
+      "step": 48975
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.724221408367157,
+      "learning_rate": 5.725870818701527e-05,
+      "loss": 2.9732,
+      "step": 48976
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7300118207931519,
+      "learning_rate": 5.724966141016515e-05,
+      "loss": 2.77,
+      "step": 48977
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7261303663253784,
+      "learning_rate": 5.7240615272672763e-05,
+      "loss": 2.919,
+      "step": 48978
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7926436066627502,
+      "learning_rate": 5.723156977456189e-05,
+      "loss": 2.8964,
+      "step": 48979
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7375580072402954,
+      "learning_rate": 5.7222524915856314e-05,
+      "loss": 3.0317,
+      "step": 48980
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7428630590438843,
+      "learning_rate": 5.721348069657985e-05,
+      "loss": 2.9936,
+      "step": 48981
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.766548216342926,
+      "learning_rate": 5.720443711675631e-05,
+      "loss": 2.9699,
+      "step": 48982
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7440727949142456,
+      "learning_rate": 5.719539417640967e-05,
+      "loss": 3.1029,
+      "step": 48983
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7492101192474365,
+      "learning_rate": 5.718635187556355e-05,
+      "loss": 2.7294,
+      "step": 48984
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8655734658241272,
+      "learning_rate": 5.71773102142419e-05,
+      "loss": 2.683,
+      "step": 48985
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7333561182022095,
+      "learning_rate": 5.71682691924685e-05,
+      "loss": 2.9099,
+      "step": 48986
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7671554684638977,
+      "learning_rate": 5.715922881026705e-05,
+      "loss": 2.9234,
+      "step": 48987
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6992800235748291,
+      "learning_rate": 5.715018906766156e-05,
+      "loss": 3.0187,
+      "step": 48988
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.707787036895752,
+      "learning_rate": 5.714114996467565e-05,
+      "loss": 2.8687,
+      "step": 48989
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7215185165405273,
+      "learning_rate": 5.713211150133328e-05,
+      "loss": 2.97,
+      "step": 48990
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7288063168525696,
+      "learning_rate": 5.7123073677658126e-05,
+      "loss": 3.0844,
+      "step": 48991
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6853630542755127,
+      "learning_rate": 5.7114036493674045e-05,
+      "loss": 3.0357,
+      "step": 48992
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.694296658039093,
+      "learning_rate": 5.710499994940491e-05,
+      "loss": 3.081,
+      "step": 48993
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7616432905197144,
+      "learning_rate": 5.7095964044874496e-05,
+      "loss": 3.1458,
+      "step": 48994
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7581666707992554,
+      "learning_rate": 5.708692878010653e-05,
+      "loss": 2.9142,
+      "step": 48995
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8747384548187256,
+      "learning_rate": 5.7077894155124795e-05,
+      "loss": 2.6311,
+      "step": 48996
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.71066814661026,
+      "learning_rate": 5.706886016995309e-05,
+      "loss": 2.8955,
+      "step": 48997
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7173435688018799,
+      "learning_rate": 5.7059826824615377e-05,
+      "loss": 2.8613,
+      "step": 48998
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7304638028144836,
+      "learning_rate": 5.705079411913522e-05,
+      "loss": 2.5475,
+      "step": 48999
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7528974413871765,
+      "learning_rate": 5.704176205353662e-05,
+      "loss": 2.7217,
+      "step": 49000
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7631033062934875,
+      "learning_rate": 5.703273062784321e-05,
+      "loss": 2.8355,
+      "step": 49001
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.734296441078186,
+      "learning_rate": 5.702369984207882e-05,
+      "loss": 2.9648,
+      "step": 49002
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6912224888801575,
+      "learning_rate": 5.701466969626728e-05,
+      "loss": 3.0517,
+      "step": 49003
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7426257133483887,
+      "learning_rate": 5.700564019043224e-05,
+      "loss": 2.8531,
+      "step": 49004
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.795962393283844,
+      "learning_rate": 5.6996611324597695e-05,
+      "loss": 2.8197,
+      "step": 49005
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7349714040756226,
+      "learning_rate": 5.698758309878721e-05,
+      "loss": 2.938,
+      "step": 49006
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7070464491844177,
+      "learning_rate": 5.697855551302475e-05,
+      "loss": 3.0683,
+      "step": 49007
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7317219972610474,
+      "learning_rate": 5.696952856733401e-05,
+      "loss": 2.8491,
+      "step": 49008
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7292014956474304,
+      "learning_rate": 5.696050226173865e-05,
+      "loss": 2.8261,
+      "step": 49009
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7460183501243591,
+      "learning_rate": 5.695147659626265e-05,
+      "loss": 2.7796,
+      "step": 49010
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7135501503944397,
+      "learning_rate": 5.6942451570929614e-05,
+      "loss": 3.0784,
+      "step": 49011
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7707333564758301,
+      "learning_rate": 5.6933427185763466e-05,
+      "loss": 3.0009,
+      "step": 49012
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7286562323570251,
+      "learning_rate": 5.692440344078779e-05,
+      "loss": 2.8597,
+      "step": 49013
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7598409056663513,
+      "learning_rate": 5.6915380336026494e-05,
+      "loss": 3.0185,
+      "step": 49014
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7345313429832458,
+      "learning_rate": 5.6906357871503326e-05,
+      "loss": 3.0258,
+      "step": 49015
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7436714172363281,
+      "learning_rate": 5.689733604724207e-05,
+      "loss": 2.8548,
+      "step": 49016
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.750116765499115,
+      "learning_rate": 5.6888314863266436e-05,
+      "loss": 2.9594,
+      "step": 49017
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7049277424812317,
+      "learning_rate": 5.687929431960011e-05,
+      "loss": 2.9449,
+      "step": 49018
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7214593291282654,
+      "learning_rate": 5.6870274416266934e-05,
+      "loss": 3.0394,
+      "step": 49019
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7183932662010193,
+      "learning_rate": 5.6861255153290765e-05,
+      "loss": 2.8671,
+      "step": 49020
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7463213801383972,
+      "learning_rate": 5.685223653069514e-05,
+      "loss": 2.714,
+      "step": 49021
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7598058581352234,
+      "learning_rate": 5.684321854850401e-05,
+      "loss": 2.9408,
+      "step": 49022
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7609604597091675,
+      "learning_rate": 5.683420120674106e-05,
+      "loss": 2.9313,
+      "step": 49023
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7309617400169373,
+      "learning_rate": 5.682518450542996e-05,
+      "loss": 2.8476,
+      "step": 49024
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7466161251068115,
+      "learning_rate": 5.6816168444594576e-05,
+      "loss": 2.8262,
+      "step": 49025
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7562272548675537,
+      "learning_rate": 5.680715302425855e-05,
+      "loss": 2.8261,
+      "step": 49026
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7695786952972412,
+      "learning_rate": 5.6798138244445745e-05,
+      "loss": 2.9201,
+      "step": 49027
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.686572790145874,
+      "learning_rate": 5.678912410517976e-05,
+      "loss": 2.8701,
+      "step": 49028
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.746677815914154,
+      "learning_rate": 5.678011060648441e-05,
+      "loss": 2.8989,
+      "step": 49029
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7117769122123718,
+      "learning_rate": 5.677109774838363e-05,
+      "loss": 2.8207,
+      "step": 49030
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7121254205703735,
+      "learning_rate": 5.676208553090078e-05,
+      "loss": 3.1207,
+      "step": 49031
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7127178907394409,
+      "learning_rate": 5.6753073954059866e-05,
+      "loss": 3.032,
+      "step": 49032
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7273410558700562,
+      "learning_rate": 5.6744063017884447e-05,
+      "loss": 2.8643,
+      "step": 49033
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7432453632354736,
+      "learning_rate": 5.673505272239839e-05,
+      "loss": 2.9869,
+      "step": 49034
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7287410497665405,
+      "learning_rate": 5.6726043067625447e-05,
+      "loss": 2.8154,
+      "step": 49035
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7052610516548157,
+      "learning_rate": 5.6717034053589205e-05,
+      "loss": 2.8844,
+      "step": 49036
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7654694318771362,
+      "learning_rate": 5.670802568031354e-05,
+      "loss": 2.9412,
+      "step": 49037
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7241032123565674,
+      "learning_rate": 5.6699017947822137e-05,
+      "loss": 2.8518,
+      "step": 49038
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7288097143173218,
+      "learning_rate": 5.66900108561386e-05,
+      "loss": 2.8373,
+      "step": 49039
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7443707585334778,
+      "learning_rate": 5.6681004405286826e-05,
+      "loss": 2.9401,
+      "step": 49040
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7623165249824524,
+      "learning_rate": 5.667199859529038e-05,
+      "loss": 2.8009,
+      "step": 49041
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7050398588180542,
+      "learning_rate": 5.666299342617315e-05,
+      "loss": 2.7906,
+      "step": 49042
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7741838097572327,
+      "learning_rate": 5.665398889795869e-05,
+      "loss": 2.944,
+      "step": 49043
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7164863348007202,
+      "learning_rate": 5.6644985010670877e-05,
+      "loss": 2.9287,
+      "step": 49044
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7640129327774048,
+      "learning_rate": 5.663598176433332e-05,
+      "loss": 2.988,
+      "step": 49045
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7587370872497559,
+      "learning_rate": 5.662697915896967e-05,
+      "loss": 2.884,
+      "step": 49046
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8077144026756287,
+      "learning_rate": 5.66179771946038e-05,
+      "loss": 2.9191,
+      "step": 49047
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7830314636230469,
+      "learning_rate": 5.660897587125928e-05,
+      "loss": 2.935,
+      "step": 49048
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7073858976364136,
+      "learning_rate": 5.6599975188959835e-05,
+      "loss": 3.1283,
+      "step": 49049
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7274441719055176,
+      "learning_rate": 5.659097514772932e-05,
+      "loss": 2.8923,
+      "step": 49050
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7369108200073242,
+      "learning_rate": 5.658197574759126e-05,
+      "loss": 2.8339,
+      "step": 49051
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7142553925514221,
+      "learning_rate": 5.6572976988569486e-05,
+      "loss": 2.8431,
+      "step": 49052
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7000463604927063,
+      "learning_rate": 5.656397887068765e-05,
+      "loss": 2.8695,
+      "step": 49053
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.693121612071991,
+      "learning_rate": 5.655498139396938e-05,
+      "loss": 2.7841,
+      "step": 49054
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6943217515945435,
+      "learning_rate": 5.654598455843851e-05,
+      "loss": 3.0532,
+      "step": 49055
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7242630124092102,
+      "learning_rate": 5.653698836411855e-05,
+      "loss": 2.8341,
+      "step": 49056
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7549852132797241,
+      "learning_rate": 5.6527992811033427e-05,
+      "loss": 2.8954,
+      "step": 49057
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6854625940322876,
+      "learning_rate": 5.6518997899206616e-05,
+      "loss": 3.1429,
+      "step": 49058
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7364606857299805,
+      "learning_rate": 5.6510003628662006e-05,
+      "loss": 2.7754,
+      "step": 49059
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7241211533546448,
+      "learning_rate": 5.6501009999423164e-05,
+      "loss": 3.0467,
+      "step": 49060
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7505494952201843,
+      "learning_rate": 5.649201701151371e-05,
+      "loss": 3.0433,
+      "step": 49061
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7099612355232239,
+      "learning_rate": 5.648302466495753e-05,
+      "loss": 2.716,
+      "step": 49062
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7239779233932495,
+      "learning_rate": 5.6474032959778104e-05,
+      "loss": 2.9557,
+      "step": 49063
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7321217656135559,
+      "learning_rate": 5.6465041895999205e-05,
+      "loss": 2.7211,
+      "step": 49064
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7070171236991882,
+      "learning_rate": 5.645605147364459e-05,
+      "loss": 2.7646,
+      "step": 49065
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6989107131958008,
+      "learning_rate": 5.644706169273787e-05,
+      "loss": 2.9561,
+      "step": 49066
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.730031430721283,
+      "learning_rate": 5.6438072553302725e-05,
+      "loss": 2.9178,
+      "step": 49067
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7238811254501343,
+      "learning_rate": 5.642908405536274e-05,
+      "loss": 2.7931,
+      "step": 49068
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7353603839874268,
+      "learning_rate": 5.6420096198941654e-05,
+      "loss": 2.9666,
+      "step": 49069
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.738623321056366,
+      "learning_rate": 5.6411108984063255e-05,
+      "loss": 2.9291,
+      "step": 49070
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7306110858917236,
+      "learning_rate": 5.640212241075103e-05,
+      "loss": 2.9059,
+      "step": 49071
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7487077713012695,
+      "learning_rate": 5.639313647902881e-05,
+      "loss": 2.8363,
+      "step": 49072
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7644553184509277,
+      "learning_rate": 5.638415118892009e-05,
+      "loss": 2.9736,
+      "step": 49073
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.27792489528656,
+      "learning_rate": 5.6375166540448744e-05,
+      "loss": 3.1355,
+      "step": 49074
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7529338598251343,
+      "learning_rate": 5.636618253363829e-05,
+      "loss": 2.8665,
+      "step": 49075
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7581132650375366,
+      "learning_rate": 5.635719916851238e-05,
+      "loss": 3.0295,
+      "step": 49076
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7204800248146057,
+      "learning_rate": 5.634821644509479e-05,
+      "loss": 2.7356,
+      "step": 49077
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7828012704849243,
+      "learning_rate": 5.6339234363408994e-05,
+      "loss": 2.927,
+      "step": 49078
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.786764919757843,
+      "learning_rate": 5.633025292347878e-05,
+      "loss": 2.9383,
+      "step": 49079
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7644099593162537,
+      "learning_rate": 5.6321272125327865e-05,
+      "loss": 2.9283,
+      "step": 49080
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7182053923606873,
+      "learning_rate": 5.631229196897983e-05,
+      "loss": 2.9578,
+      "step": 49081
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7207154631614685,
+      "learning_rate": 5.6303312454458316e-05,
+      "loss": 3.0525,
+      "step": 49082
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.773795485496521,
+      "learning_rate": 5.6294333581786875e-05,
+      "loss": 3.0164,
+      "step": 49083
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7547546625137329,
+      "learning_rate": 5.628535535098929e-05,
+      "loss": 3.0012,
+      "step": 49084
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7342237830162048,
+      "learning_rate": 5.6276377762089243e-05,
+      "loss": 2.9923,
+      "step": 49085
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7141954898834229,
+      "learning_rate": 5.626740081511024e-05,
+      "loss": 2.6876,
+      "step": 49086
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7377126216888428,
+      "learning_rate": 5.6258424510076074e-05,
+      "loss": 2.8332,
+      "step": 49087
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7284415364265442,
+      "learning_rate": 5.624944884701028e-05,
+      "loss": 3.057,
+      "step": 49088
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7273825407028198,
+      "learning_rate": 5.624047382593655e-05,
+      "loss": 2.7143,
+      "step": 49089
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7405761480331421,
+      "learning_rate": 5.6231499446878424e-05,
+      "loss": 2.8364,
+      "step": 49090
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7068529725074768,
+      "learning_rate": 5.622252570985959e-05,
+      "loss": 2.8806,
+      "step": 49091
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6906842589378357,
+      "learning_rate": 5.621355261490382e-05,
+      "loss": 2.7804,
+      "step": 49092
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7543202638626099,
+      "learning_rate": 5.620458016203454e-05,
+      "loss": 2.6974,
+      "step": 49093
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.735177218914032,
+      "learning_rate": 5.619560835127556e-05,
+      "loss": 3.0508,
+      "step": 49094
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.733551025390625,
+      "learning_rate": 5.618663718265032e-05,
+      "loss": 2.8695,
+      "step": 49095
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7386725544929504,
+      "learning_rate": 5.617766665618267e-05,
+      "loss": 2.9166,
+      "step": 49096
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7156366109848022,
+      "learning_rate": 5.61686967718961e-05,
+      "loss": 2.7209,
+      "step": 49097
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7398567199707031,
+      "learning_rate": 5.615972752981419e-05,
+      "loss": 2.716,
+      "step": 49098
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7204150557518005,
+      "learning_rate": 5.6150758929960715e-05,
+      "loss": 3.0253,
+      "step": 49099
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7159655690193176,
+      "learning_rate": 5.614179097235913e-05,
+      "loss": 2.8584,
+      "step": 49100
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6986108422279358,
+      "learning_rate": 5.61328236570331e-05,
+      "loss": 2.7455,
+      "step": 49101
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7308990955352783,
+      "learning_rate": 5.612385698400639e-05,
+      "loss": 3.1152,
+      "step": 49102
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7399376034736633,
+      "learning_rate": 5.611489095330251e-05,
+      "loss": 2.8173,
+      "step": 49103
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7102044820785522,
+      "learning_rate": 5.6105925564945066e-05,
+      "loss": 2.8044,
+      "step": 49104
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7224224805831909,
+      "learning_rate": 5.609696081895758e-05,
+      "loss": 2.9648,
+      "step": 49105
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6727746725082397,
+      "learning_rate": 5.608799671536377e-05,
+      "loss": 2.7387,
+      "step": 49106
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7503066062927246,
+      "learning_rate": 5.607903325418731e-05,
+      "loss": 3.0684,
+      "step": 49107
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7269359827041626,
+      "learning_rate": 5.6070070435451685e-05,
+      "loss": 2.9476,
+      "step": 49108
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7300078868865967,
+      "learning_rate": 5.6061108259180575e-05,
+      "loss": 2.7939,
+      "step": 49109
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7930198311805725,
+      "learning_rate": 5.60521467253976e-05,
+      "loss": 2.9471,
+      "step": 49110
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7446059584617615,
+      "learning_rate": 5.604318583412622e-05,
+      "loss": 2.9892,
+      "step": 49111
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6801463961601257,
+      "learning_rate": 5.6034225585390245e-05,
+      "loss": 2.8705,
+      "step": 49112
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.82508385181427,
+      "learning_rate": 5.6025265979213074e-05,
+      "loss": 2.7941,
+      "step": 49113
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7679817080497742,
+      "learning_rate": 5.6016307015618454e-05,
+      "loss": 2.8535,
+      "step": 49114
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7149144411087036,
+      "learning_rate": 5.600734869462987e-05,
+      "loss": 2.7853,
+      "step": 49115
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7261170148849487,
+      "learning_rate": 5.599839101627097e-05,
+      "loss": 2.8658,
+      "step": 49116
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7425785660743713,
+      "learning_rate": 5.5989433980565435e-05,
+      "loss": 2.9014,
+      "step": 49117
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7709639072418213,
+      "learning_rate": 5.5980477587536775e-05,
+      "loss": 2.8676,
+      "step": 49118
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6940927505493164,
+      "learning_rate": 5.5971521837208546e-05,
+      "loss": 2.9155,
+      "step": 49119
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.747685968875885,
+      "learning_rate": 5.5962566729604317e-05,
+      "loss": 2.6554,
+      "step": 49120
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7228460907936096,
+      "learning_rate": 5.595361226474768e-05,
+      "loss": 2.7611,
+      "step": 49121
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6861898303031921,
+      "learning_rate": 5.5944658442662385e-05,
+      "loss": 2.8041,
+      "step": 49122
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7472571134567261,
+      "learning_rate": 5.593570526337176e-05,
+      "loss": 2.7697,
+      "step": 49123
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7464053630828857,
+      "learning_rate": 5.592675272689965e-05,
+      "loss": 2.8793,
+      "step": 49124
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6809940338134766,
+      "learning_rate": 5.591780083326948e-05,
+      "loss": 2.857,
+      "step": 49125
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6927462220191956,
+      "learning_rate": 5.5908849582504764e-05,
+      "loss": 2.9232,
+      "step": 49126
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7227660417556763,
+      "learning_rate": 5.589989897462923e-05,
+      "loss": 2.784,
+      "step": 49127
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7501949071884155,
+      "learning_rate": 5.589094900966632e-05,
+      "loss": 2.7866,
+      "step": 49128
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7532203197479248,
+      "learning_rate": 5.5881999687639746e-05,
+      "loss": 2.9745,
+      "step": 49129
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7279804944992065,
+      "learning_rate": 5.587305100857292e-05,
+      "loss": 2.9465,
+      "step": 49130
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7034523487091064,
+      "learning_rate": 5.58641029724895e-05,
+      "loss": 2.9068,
+      "step": 49131
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7209697961807251,
+      "learning_rate": 5.585515557941322e-05,
+      "loss": 2.7805,
+      "step": 49132
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7370447516441345,
+      "learning_rate": 5.584620882936731e-05,
+      "loss": 2.8805,
+      "step": 49133
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7712350487709045,
+      "learning_rate": 5.5837262722375574e-05,
+      "loss": 2.7915,
+      "step": 49134
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7178522944450378,
+      "learning_rate": 5.58283172584614e-05,
+      "loss": 2.9186,
+      "step": 49135
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.730476438999176,
+      "learning_rate": 5.581937243764846e-05,
+      "loss": 2.9216,
+      "step": 49136
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7608109712600708,
+      "learning_rate": 5.5810428259960375e-05,
+      "loss": 3.1027,
+      "step": 49137
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7009885907173157,
+      "learning_rate": 5.580148472542055e-05,
+      "loss": 2.7478,
+      "step": 49138
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7496047019958496,
+      "learning_rate": 5.579254183405272e-05,
+      "loss": 2.9551,
+      "step": 49139
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7258723974227905,
+      "learning_rate": 5.578359958588028e-05,
+      "loss": 2.7385,
+      "step": 49140
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7414917945861816,
+      "learning_rate": 5.5774657980926815e-05,
+      "loss": 2.8253,
+      "step": 49141
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7586607933044434,
+      "learning_rate": 5.576571701921595e-05,
+      "loss": 2.8642,
+      "step": 49142
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7755255699157715,
+      "learning_rate": 5.575677670077112e-05,
+      "loss": 2.964,
+      "step": 49143
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7838589549064636,
+      "learning_rate": 5.574783702561598e-05,
+      "loss": 2.9531,
+      "step": 49144
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7484715580940247,
+      "learning_rate": 5.573889799377398e-05,
+      "loss": 2.9606,
+      "step": 49145
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7210723757743835,
+      "learning_rate": 5.572995960526879e-05,
+      "loss": 2.9659,
+      "step": 49146
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7425600290298462,
+      "learning_rate": 5.572102186012387e-05,
+      "loss": 3.0856,
+      "step": 49147
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7232694625854492,
+      "learning_rate": 5.571208475836269e-05,
+      "loss": 2.9543,
+      "step": 49148
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.713165819644928,
+      "learning_rate": 5.570314830000895e-05,
+      "loss": 2.8556,
+      "step": 49149
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.741438627243042,
+      "learning_rate": 5.569421248508601e-05,
+      "loss": 3.0215,
+      "step": 49150
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7061321139335632,
+      "learning_rate": 5.56852773136175e-05,
+      "loss": 3.0182,
+      "step": 49151
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7206674814224243,
+      "learning_rate": 5.5676342785627045e-05,
+      "loss": 2.8039,
+      "step": 49152
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7030647397041321,
+      "learning_rate": 5.566740890113799e-05,
+      "loss": 3.02,
+      "step": 49153
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7327060699462891,
+      "learning_rate": 5.565847566017403e-05,
+      "loss": 2.7408,
+      "step": 49154
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7196124196052551,
+      "learning_rate": 5.5649543062758617e-05,
+      "loss": 2.8843,
+      "step": 49155
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7554773092269897,
+      "learning_rate": 5.5640611108915204e-05,
+      "loss": 2.8299,
+      "step": 49156
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7063097357749939,
+      "learning_rate": 5.5631679798667495e-05,
+      "loss": 3.0298,
+      "step": 49157
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7428682446479797,
+      "learning_rate": 5.5622749132038815e-05,
+      "loss": 3.0282,
+      "step": 49158
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7211458086967468,
+      "learning_rate": 5.561381910905287e-05,
+      "loss": 2.7996,
+      "step": 49159
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7217656373977661,
+      "learning_rate": 5.5604889729733016e-05,
+      "loss": 2.6599,
+      "step": 49160
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7122716903686523,
+      "learning_rate": 5.55959609941029e-05,
+      "loss": 2.7535,
+      "step": 49161
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7155596613883972,
+      "learning_rate": 5.558703290218599e-05,
+      "loss": 2.8666,
+      "step": 49162
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7553215622901917,
+      "learning_rate": 5.557810545400575e-05,
+      "loss": 2.9545,
+      "step": 49163
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.727124810218811,
+      "learning_rate": 5.556917864958579e-05,
+      "loss": 2.8262,
+      "step": 49164
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7256839871406555,
+      "learning_rate": 5.556025248894952e-05,
+      "loss": 2.8445,
+      "step": 49165
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7141228318214417,
+      "learning_rate": 5.555132697212049e-05,
+      "loss": 2.7558,
+      "step": 49166
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7409408688545227,
+      "learning_rate": 5.5542402099122287e-05,
+      "loss": 2.7731,
+      "step": 49167
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7171137928962708,
+      "learning_rate": 5.553347786997835e-05,
+      "loss": 3.0352,
+      "step": 49168
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7588046789169312,
+      "learning_rate": 5.5524554284712196e-05,
+      "loss": 3.0276,
+      "step": 49169
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7045204639434814,
+      "learning_rate": 5.5515631343347244e-05,
+      "loss": 2.807,
+      "step": 49170
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7897858619689941,
+      "learning_rate": 5.550670904590714e-05,
+      "loss": 2.7827,
+      "step": 49171
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7354269623756409,
+      "learning_rate": 5.5497787392415227e-05,
+      "loss": 2.6785,
+      "step": 49172
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7657710313796997,
+      "learning_rate": 5.548886638289506e-05,
+      "loss": 2.7895,
+      "step": 49173
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.789874255657196,
+      "learning_rate": 5.547994601737028e-05,
+      "loss": 2.8425,
+      "step": 49174
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7321118712425232,
+      "learning_rate": 5.547102629586417e-05,
+      "loss": 2.8132,
+      "step": 49175
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7470834851264954,
+      "learning_rate": 5.546210721840039e-05,
+      "loss": 2.9483,
+      "step": 49176
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.725573718547821,
+      "learning_rate": 5.5453188785002366e-05,
+      "loss": 3.0163,
+      "step": 49177
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7329528331756592,
+      "learning_rate": 5.5444270995693496e-05,
+      "loss": 2.7606,
+      "step": 49178
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6969563364982605,
+      "learning_rate": 5.543535385049742e-05,
+      "loss": 2.8289,
+      "step": 49179
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7398406267166138,
+      "learning_rate": 5.542643734943749e-05,
+      "loss": 3.0499,
+      "step": 49180
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7166543006896973,
+      "learning_rate": 5.541752149253735e-05,
+      "loss": 2.9627,
+      "step": 49181
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7646118998527527,
+      "learning_rate": 5.540860627982029e-05,
+      "loss": 2.8877,
+      "step": 49182
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7465002536773682,
+      "learning_rate": 5.5399691711309944e-05,
+      "loss": 2.8294,
+      "step": 49183
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7415258288383484,
+      "learning_rate": 5.539077778702977e-05,
+      "loss": 3.1178,
+      "step": 49184
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7385164499282837,
+      "learning_rate": 5.538186450700312e-05,
+      "loss": 3.0409,
+      "step": 49185
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7214295864105225,
+      "learning_rate": 5.537295187125366e-05,
+      "loss": 2.9575,
+      "step": 49186
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.742912769317627,
+      "learning_rate": 5.5364039879804646e-05,
+      "loss": 2.7598,
+      "step": 49187
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.726211428642273,
+      "learning_rate": 5.535512853267969e-05,
+      "loss": 3.0765,
+      "step": 49188
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6831627488136292,
+      "learning_rate": 5.534621782990232e-05,
+      "loss": 2.8655,
+      "step": 49189
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7041841745376587,
+      "learning_rate": 5.5337307771495933e-05,
+      "loss": 2.9758,
+      "step": 49190
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7825841307640076,
+      "learning_rate": 5.5328398357483986e-05,
+      "loss": 2.8501,
+      "step": 49191
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.795807957649231,
+      "learning_rate": 5.5319489587889886e-05,
+      "loss": 2.7646,
+      "step": 49192
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7698444724082947,
+      "learning_rate": 5.5310581462737155e-05,
+      "loss": 2.7806,
+      "step": 49193
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6966370940208435,
+      "learning_rate": 5.5301673982049314e-05,
+      "loss": 3.0312,
+      "step": 49194
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7309635877609253,
+      "learning_rate": 5.5292767145849726e-05,
+      "loss": 2.7085,
+      "step": 49195
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7802844643592834,
+      "learning_rate": 5.528386095416195e-05,
+      "loss": 2.9802,
+      "step": 49196
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7086724638938904,
+      "learning_rate": 5.5274955407009305e-05,
+      "loss": 3.0243,
+      "step": 49197
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.8560823798179626,
+      "learning_rate": 5.5266050504415424e-05,
+      "loss": 2.8772,
+      "step": 49198
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6955791115760803,
+      "learning_rate": 5.525714624640367e-05,
+      "loss": 2.7175,
+      "step": 49199
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7265035510063171,
+      "learning_rate": 5.524824263299738e-05,
+      "loss": 2.8838,
+      "step": 49200
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7440124750137329,
+      "learning_rate": 5.523933966422025e-05,
+      "loss": 2.9098,
+      "step": 49201
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7409399151802063,
+      "learning_rate": 5.523043734009548e-05,
+      "loss": 2.7065,
+      "step": 49202
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7265484929084778,
+      "learning_rate": 5.522153566064662e-05,
+      "loss": 3.1762,
+      "step": 49203
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7463249564170837,
+      "learning_rate": 5.521263462589722e-05,
+      "loss": 2.8447,
+      "step": 49204
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7473812699317932,
+      "learning_rate": 5.520373423587064e-05,
+      "loss": 2.7834,
+      "step": 49205
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.725105881690979,
+      "learning_rate": 5.51948344905903e-05,
+      "loss": 3.1029,
+      "step": 49206
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.694555938243866,
+      "learning_rate": 5.518593539007957e-05,
+      "loss": 2.9168,
+      "step": 49207
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7044192552566528,
+      "learning_rate": 5.5177036934362e-05,
+      "loss": 2.7009,
+      "step": 49208
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7090144753456116,
+      "learning_rate": 5.516813912346102e-05,
+      "loss": 3.0191,
+      "step": 49209
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7357432842254639,
+      "learning_rate": 5.515924195740002e-05,
+      "loss": 2.8309,
+      "step": 49210
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6832718253135681,
+      "learning_rate": 5.5150345436202495e-05,
+      "loss": 3.1344,
+      "step": 49211
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6962417960166931,
+      "learning_rate": 5.514144955989187e-05,
+      "loss": 2.9699,
+      "step": 49212
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7619498372077942,
+      "learning_rate": 5.513255432849144e-05,
+      "loss": 2.7511,
+      "step": 49213
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7212022542953491,
+      "learning_rate": 5.512365974202482e-05,
+      "loss": 2.8064,
+      "step": 49214
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7095093131065369,
+      "learning_rate": 5.511476580051528e-05,
+      "loss": 2.7405,
+      "step": 49215
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7469050884246826,
+      "learning_rate": 5.510587250398638e-05,
+      "loss": 2.773,
+      "step": 49216
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.708609938621521,
+      "learning_rate": 5.509697985246143e-05,
+      "loss": 2.9376,
+      "step": 49217
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7691136598587036,
+      "learning_rate": 5.508808784596388e-05,
+      "loss": 2.9028,
+      "step": 49218
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7040303945541382,
+      "learning_rate": 5.5079196484517274e-05,
+      "loss": 2.8717,
+      "step": 49219
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7086603045463562,
+      "learning_rate": 5.507030576814493e-05,
+      "loss": 2.9845,
+      "step": 49220
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7217570543289185,
+      "learning_rate": 5.5061415696870223e-05,
+      "loss": 3.0093,
+      "step": 49221
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7315422892570496,
+      "learning_rate": 5.505252627071655e-05,
+      "loss": 2.9514,
+      "step": 49222
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7391554117202759,
+      "learning_rate": 5.5043637489707406e-05,
+      "loss": 2.9218,
+      "step": 49223
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7177310585975647,
+      "learning_rate": 5.503474935386621e-05,
+      "loss": 2.8227,
+      "step": 49224
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7659426331520081,
+      "learning_rate": 5.502586186321629e-05,
+      "loss": 2.9178,
+      "step": 49225
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7270904183387756,
+      "learning_rate": 5.5016975017781184e-05,
+      "loss": 2.9833,
+      "step": 49226
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6957828998565674,
+      "learning_rate": 5.500808881758421e-05,
+      "loss": 2.883,
+      "step": 49227
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7378817200660706,
+      "learning_rate": 5.49992032626487e-05,
+      "loss": 2.8653,
+      "step": 49228
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.750215470790863,
+      "learning_rate": 5.499031835299823e-05,
+      "loss": 2.5859,
+      "step": 49229
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.790286123752594,
+      "learning_rate": 5.4981434088656026e-05,
+      "loss": 2.9305,
+      "step": 49230
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7604691386222839,
+      "learning_rate": 5.497255046964565e-05,
+      "loss": 2.8586,
+      "step": 49231
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7047723531723022,
+      "learning_rate": 5.496366749599034e-05,
+      "loss": 2.9883,
+      "step": 49232
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7287726998329163,
+      "learning_rate": 5.4954785167713645e-05,
+      "loss": 2.9042,
+      "step": 49233
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7202035188674927,
+      "learning_rate": 5.494590348483888e-05,
+      "loss": 2.8477,
+      "step": 49234
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7223737239837646,
+      "learning_rate": 5.4937022447389365e-05,
+      "loss": 2.9645,
+      "step": 49235
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7385289669036865,
+      "learning_rate": 5.4928142055388645e-05,
+      "loss": 2.9676,
+      "step": 49236
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7490689754486084,
+      "learning_rate": 5.491926230885997e-05,
+      "loss": 2.8101,
+      "step": 49237
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7428318858146667,
+      "learning_rate": 5.491038320782678e-05,
+      "loss": 2.9541,
+      "step": 49238
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7282332181930542,
+      "learning_rate": 5.490150475231257e-05,
+      "loss": 3.0235,
+      "step": 49239
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7052958011627197,
+      "learning_rate": 5.4892626942340533e-05,
+      "loss": 2.8107,
+      "step": 49240
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7355846166610718,
+      "learning_rate": 5.488374977793421e-05,
+      "loss": 3.0651,
+      "step": 49241
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7115058898925781,
+      "learning_rate": 5.4874873259116924e-05,
+      "loss": 2.8074,
+      "step": 49242
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6894238591194153,
+      "learning_rate": 5.4865997385911976e-05,
+      "loss": 2.9663,
+      "step": 49243
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7079063653945923,
+      "learning_rate": 5.48571221583429e-05,
+      "loss": 2.9716,
+      "step": 49244
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7443049550056458,
+      "learning_rate": 5.484824757643289e-05,
+      "loss": 2.9997,
+      "step": 49245
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7488800883293152,
+      "learning_rate": 5.483937364020552e-05,
+      "loss": 2.8637,
+      "step": 49246
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6989515423774719,
+      "learning_rate": 5.483050034968397e-05,
+      "loss": 2.9274,
+      "step": 49247
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.66061931848526,
+      "learning_rate": 5.4821627704891756e-05,
+      "loss": 2.9987,
+      "step": 49248
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7440872192382812,
+      "learning_rate": 5.4812755705852194e-05,
+      "loss": 3.2342,
+      "step": 49249
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7412754893302917,
+      "learning_rate": 5.4803884352588566e-05,
+      "loss": 2.899,
+      "step": 49250
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7123976945877075,
+      "learning_rate": 5.479501364512442e-05,
+      "loss": 2.9107,
+      "step": 49251
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7113731503486633,
+      "learning_rate": 5.478614358348293e-05,
+      "loss": 3.1027,
+      "step": 49252
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7375507354736328,
+      "learning_rate": 5.4777274167687625e-05,
+      "loss": 2.7018,
+      "step": 49253
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7448770999908447,
+      "learning_rate": 5.4768405397761674e-05,
+      "loss": 2.9676,
+      "step": 49254
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7443416714668274,
+      "learning_rate": 5.4759537273728595e-05,
+      "loss": 2.904,
+      "step": 49255
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7311747670173645,
+      "learning_rate": 5.475066979561174e-05,
+      "loss": 3.1093,
+      "step": 49256
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.728496253490448,
+      "learning_rate": 5.474180296343444e-05,
+      "loss": 3.0179,
+      "step": 49257
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7192690968513489,
+      "learning_rate": 5.4732936777220035e-05,
+      "loss": 2.9779,
+      "step": 49258
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7940871715545654,
+      "learning_rate": 5.472407123699176e-05,
+      "loss": 2.894,
+      "step": 49259
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7260642647743225,
+      "learning_rate": 5.471520634277313e-05,
+      "loss": 2.9742,
+      "step": 49260
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7162463665008545,
+      "learning_rate": 5.470634209458746e-05,
+      "loss": 2.9621,
+      "step": 49261
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7385729551315308,
+      "learning_rate": 5.469747849245803e-05,
+      "loss": 2.6355,
+      "step": 49262
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7240334153175354,
+      "learning_rate": 5.468861553640833e-05,
+      "loss": 2.7548,
+      "step": 49263
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7283338308334351,
+      "learning_rate": 5.46797532264616e-05,
+      "loss": 2.7219,
+      "step": 49264
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.8645117282867432,
+      "learning_rate": 5.467089156264109e-05,
+      "loss": 2.8017,
+      "step": 49265
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7759827375411987,
+      "learning_rate": 5.46620305449703e-05,
+      "loss": 2.8121,
+      "step": 49266
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7253652811050415,
+      "learning_rate": 5.465317017347244e-05,
+      "loss": 2.9753,
+      "step": 49267
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.720482349395752,
+      "learning_rate": 5.4644310448171e-05,
+      "loss": 3.0358,
+      "step": 49268
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7371129393577576,
+      "learning_rate": 5.463545136908915e-05,
+      "loss": 2.7949,
+      "step": 49269
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6884015202522278,
+      "learning_rate": 5.462659293625038e-05,
+      "loss": 2.8906,
+      "step": 49270
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7280858159065247,
+      "learning_rate": 5.4617735149677925e-05,
+      "loss": 3.058,
+      "step": 49271
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7321854829788208,
+      "learning_rate": 5.4608878009395084e-05,
+      "loss": 3.049,
+      "step": 49272
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7545171976089478,
+      "learning_rate": 5.4600021515425284e-05,
+      "loss": 2.9722,
+      "step": 49273
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7591923475265503,
+      "learning_rate": 5.45911656677917e-05,
+      "loss": 2.9183,
+      "step": 49274
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.6920126676559448,
+      "learning_rate": 5.458231046651779e-05,
+      "loss": 2.8652,
+      "step": 49275
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7270497679710388,
+      "learning_rate": 5.457345591162692e-05,
+      "loss": 2.8887,
+      "step": 49276
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.8276441693305969,
+      "learning_rate": 5.456460200314224e-05,
+      "loss": 2.7831,
+      "step": 49277
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7535173892974854,
+      "learning_rate": 5.455574874108725e-05,
+      "loss": 3.0215,
+      "step": 49278
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7337012887001038,
+      "learning_rate": 5.4546896125485173e-05,
+      "loss": 2.7542,
+      "step": 49279
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7610710859298706,
+      "learning_rate": 5.453804415635927e-05,
+      "loss": 3.0278,
+      "step": 49280
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.751594066619873,
+      "learning_rate": 5.4529192833732996e-05,
+      "loss": 3.1184,
+      "step": 49281
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6849843859672546,
+      "learning_rate": 5.452034215762948e-05,
+      "loss": 3.1234,
+      "step": 49282
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7445322871208191,
+      "learning_rate": 5.4511492128072255e-05,
+      "loss": 2.7937,
+      "step": 49283
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7667579054832458,
+      "learning_rate": 5.4502642745084425e-05,
+      "loss": 2.9009,
+      "step": 49284
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6802947521209717,
+      "learning_rate": 5.449379400868944e-05,
+      "loss": 3.005,
+      "step": 49285
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7073135375976562,
+      "learning_rate": 5.4484945918910585e-05,
+      "loss": 2.8548,
+      "step": 49286
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7205443978309631,
+      "learning_rate": 5.4476098475771033e-05,
+      "loss": 2.8307,
+      "step": 49287
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7851844429969788,
+      "learning_rate": 5.446725167929428e-05,
+      "loss": 2.8409,
+      "step": 49288
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7031289339065552,
+      "learning_rate": 5.445840552950346e-05,
+      "loss": 2.8974,
+      "step": 49289
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7477833032608032,
+      "learning_rate": 5.4449560026421924e-05,
+      "loss": 2.9867,
+      "step": 49290
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7260494828224182,
+      "learning_rate": 5.444071517007306e-05,
+      "loss": 2.6814,
+      "step": 49291
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7753791809082031,
+      "learning_rate": 5.443187096048011e-05,
+      "loss": 2.9645,
+      "step": 49292
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.8131752610206604,
+      "learning_rate": 5.4423027397666354e-05,
+      "loss": 3.0435,
+      "step": 49293
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7920990586280823,
+      "learning_rate": 5.441418448165501e-05,
+      "loss": 2.957,
+      "step": 49294
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7970216274261475,
+      "learning_rate": 5.4405342212469427e-05,
+      "loss": 2.7669,
+      "step": 49295
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7318688631057739,
+      "learning_rate": 5.4396500590132984e-05,
+      "loss": 2.9577,
+      "step": 49296
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7441763281822205,
+      "learning_rate": 5.4387659614668834e-05,
+      "loss": 2.9414,
+      "step": 49297
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7256457209587097,
+      "learning_rate": 5.437881928610035e-05,
+      "loss": 2.8141,
+      "step": 49298
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.790198564529419,
+      "learning_rate": 5.436997960445075e-05,
+      "loss": 3.0438,
+      "step": 49299
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7521136999130249,
+      "learning_rate": 5.4361140569743423e-05,
+      "loss": 2.976,
+      "step": 49300
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7699940800666809,
+      "learning_rate": 5.435230218200157e-05,
+      "loss": 2.8781,
+      "step": 49301
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7283733487129211,
+      "learning_rate": 5.4343464441248384e-05,
+      "loss": 2.7156,
+      "step": 49302
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7702507376670837,
+      "learning_rate": 5.433462734750734e-05,
+      "loss": 3.0198,
+      "step": 49303
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.300162672996521,
+      "learning_rate": 5.432579090080152e-05,
+      "loss": 2.9121,
+      "step": 49304
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7907041907310486,
+      "learning_rate": 5.431695510115428e-05,
+      "loss": 2.9838,
+      "step": 49305
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7288923859596252,
+      "learning_rate": 5.4308119948588994e-05,
+      "loss": 2.9276,
+      "step": 49306
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7126811742782593,
+      "learning_rate": 5.4299285443128806e-05,
+      "loss": 2.9427,
+      "step": 49307
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7134640216827393,
+      "learning_rate": 5.429045158479703e-05,
+      "loss": 2.8017,
+      "step": 49308
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7535879015922546,
+      "learning_rate": 5.428161837361686e-05,
+      "loss": 2.7213,
+      "step": 49309
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6850473880767822,
+      "learning_rate": 5.4272785809611606e-05,
+      "loss": 2.9667,
+      "step": 49310
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7917988896369934,
+      "learning_rate": 5.4263953892804614e-05,
+      "loss": 2.9065,
+      "step": 49311
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7340681552886963,
+      "learning_rate": 5.4255122623218994e-05,
+      "loss": 2.9701,
+      "step": 49312
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7131487131118774,
+      "learning_rate": 5.4246292000878164e-05,
+      "loss": 2.864,
+      "step": 49313
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7326993942260742,
+      "learning_rate": 5.423746202580528e-05,
+      "loss": 2.9615,
+      "step": 49314
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7834832668304443,
+      "learning_rate": 5.422863269802358e-05,
+      "loss": 3.0211,
+      "step": 49315
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7300680875778198,
+      "learning_rate": 5.421980401755645e-05,
+      "loss": 2.9014,
+      "step": 49316
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7191560864448547,
+      "learning_rate": 5.421097598442693e-05,
+      "loss": 2.6949,
+      "step": 49317
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7470296025276184,
+      "learning_rate": 5.420214859865851e-05,
+      "loss": 2.8543,
+      "step": 49318
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7371562719345093,
+      "learning_rate": 5.419332186027424e-05,
+      "loss": 2.9266,
+      "step": 49319
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7162860631942749,
+      "learning_rate": 5.418449576929746e-05,
+      "loss": 2.8709,
+      "step": 49320
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7437654137611389,
+      "learning_rate": 5.41756703257515e-05,
+      "loss": 2.857,
+      "step": 49321
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7650477290153503,
+      "learning_rate": 5.416684552965949e-05,
+      "loss": 2.6571,
+      "step": 49322
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.743986189365387,
+      "learning_rate": 5.415802138104468e-05,
+      "loss": 2.8046,
+      "step": 49323
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7592489719390869,
+      "learning_rate": 5.4149197879930296e-05,
+      "loss": 2.8845,
+      "step": 49324
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.700624406337738,
+      "learning_rate": 5.414037502633965e-05,
+      "loss": 2.8208,
+      "step": 49325
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7299220561981201,
+      "learning_rate": 5.4131552820295875e-05,
+      "loss": 2.7412,
+      "step": 49326
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.756610095500946,
+      "learning_rate": 5.412273126182227e-05,
+      "loss": 3.0103,
+      "step": 49327
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7145278453826904,
+      "learning_rate": 5.411391035094218e-05,
+      "loss": 2.8637,
+      "step": 49328
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.749764621257782,
+      "learning_rate": 5.410509008767868e-05,
+      "loss": 2.8035,
+      "step": 49329
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7430508136749268,
+      "learning_rate": 5.409627047205509e-05,
+      "loss": 2.7404,
+      "step": 49330
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7532656788825989,
+      "learning_rate": 5.4087451504094494e-05,
+      "loss": 3.022,
+      "step": 49331
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7438516020774841,
+      "learning_rate": 5.407863318382023e-05,
+      "loss": 3.084,
+      "step": 49332
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7108733654022217,
+      "learning_rate": 5.4069815511255585e-05,
+      "loss": 2.985,
+      "step": 49333
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7077670693397522,
+      "learning_rate": 5.406099848642368e-05,
+      "loss": 3.0003,
+      "step": 49334
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7985681891441345,
+      "learning_rate": 5.405218210934782e-05,
+      "loss": 2.78,
+      "step": 49335
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7880358695983887,
+      "learning_rate": 5.404336638005116e-05,
+      "loss": 3.0255,
+      "step": 49336
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7047920227050781,
+      "learning_rate": 5.403455129855687e-05,
+      "loss": 3.151,
+      "step": 49337
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7189116477966309,
+      "learning_rate": 5.4025736864888346e-05,
+      "loss": 2.9223,
+      "step": 49338
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7386056780815125,
+      "learning_rate": 5.4016923079068565e-05,
+      "loss": 2.9541,
+      "step": 49339
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7559934854507446,
+      "learning_rate": 5.400810994112097e-05,
+      "loss": 2.9204,
+      "step": 49340
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7456420063972473,
+      "learning_rate": 5.3999297451068614e-05,
+      "loss": 3.1086,
+      "step": 49341
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.748192310333252,
+      "learning_rate": 5.399048560893474e-05,
+      "loss": 2.7161,
+      "step": 49342
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7148739099502563,
+      "learning_rate": 5.398167441474267e-05,
+      "loss": 2.9892,
+      "step": 49343
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7266228795051575,
+      "learning_rate": 5.397286386851552e-05,
+      "loss": 3.0108,
+      "step": 49344
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7317888736724854,
+      "learning_rate": 5.396405397027649e-05,
+      "loss": 3.0233,
+      "step": 49345
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7267348766326904,
+      "learning_rate": 5.3955244720048704e-05,
+      "loss": 2.9689,
+      "step": 49346
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7102034687995911,
+      "learning_rate": 5.394643611785549e-05,
+      "loss": 2.6048,
+      "step": 49347
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7675157785415649,
+      "learning_rate": 5.393762816372004e-05,
+      "loss": 3.0444,
+      "step": 49348
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7861278653144836,
+      "learning_rate": 5.392882085766548e-05,
+      "loss": 2.9381,
+      "step": 49349
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7278153300285339,
+      "learning_rate": 5.392001419971512e-05,
+      "loss": 2.6851,
+      "step": 49350
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7110613584518433,
+      "learning_rate": 5.391120818989205e-05,
+      "loss": 3.0432,
+      "step": 49351
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7356806993484497,
+      "learning_rate": 5.390240282821945e-05,
+      "loss": 2.8609,
+      "step": 49352
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.73133784532547,
+      "learning_rate": 5.389359811472066e-05,
+      "loss": 2.9498,
+      "step": 49353
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7367465496063232,
+      "learning_rate": 5.388479404941863e-05,
+      "loss": 2.9129,
+      "step": 49354
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7416877746582031,
+      "learning_rate": 5.387599063233682e-05,
+      "loss": 2.9267,
+      "step": 49355
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7233571410179138,
+      "learning_rate": 5.386718786349819e-05,
+      "loss": 3.1255,
+      "step": 49356
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7607924342155457,
+      "learning_rate": 5.3858385742926016e-05,
+      "loss": 2.7879,
+      "step": 49357
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7684453129768372,
+      "learning_rate": 5.384958427064362e-05,
+      "loss": 2.8697,
+      "step": 49358
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7534241676330566,
+      "learning_rate": 5.3840783446673895e-05,
+      "loss": 2.7628,
+      "step": 49359
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7305625677108765,
+      "learning_rate": 5.383198327104026e-05,
+      "loss": 2.663,
+      "step": 49360
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7438086867332458,
+      "learning_rate": 5.3823183743765716e-05,
+      "loss": 2.9783,
+      "step": 49361
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7273472547531128,
+      "learning_rate": 5.381438486487356e-05,
+      "loss": 2.9193,
+      "step": 49362
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7327690124511719,
+      "learning_rate": 5.3805586634387e-05,
+      "loss": 2.7964,
+      "step": 49363
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7488052248954773,
+      "learning_rate": 5.379678905232905e-05,
+      "loss": 2.9159,
+      "step": 49364
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7673367857933044,
+      "learning_rate": 5.378799211872306e-05,
+      "loss": 3.1029,
+      "step": 49365
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7909415364265442,
+      "learning_rate": 5.37791958335921e-05,
+      "loss": 3.0563,
+      "step": 49366
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7198272347450256,
+      "learning_rate": 5.377040019695931e-05,
+      "loss": 2.782,
+      "step": 49367
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7427800893783569,
+      "learning_rate": 5.3761605208847944e-05,
+      "loss": 3.2383,
+      "step": 49368
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6806737184524536,
+      "learning_rate": 5.375281086928104e-05,
+      "loss": 2.7696,
+      "step": 49369
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7873852252960205,
+      "learning_rate": 5.374401717828193e-05,
+      "loss": 2.9428,
+      "step": 49370
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7423635721206665,
+      "learning_rate": 5.373522413587363e-05,
+      "loss": 2.8635,
+      "step": 49371
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7548320293426514,
+      "learning_rate": 5.37264317420794e-05,
+      "loss": 3.1002,
+      "step": 49372
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.711924135684967,
+      "learning_rate": 5.371763999692238e-05,
+      "loss": 3.0354,
+      "step": 49373
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.773473858833313,
+      "learning_rate": 5.370884890042558e-05,
+      "loss": 2.8397,
+      "step": 49374
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7292789220809937,
+      "learning_rate": 5.370005845261239e-05,
+      "loss": 2.9119,
+      "step": 49375
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7399142980575562,
+      "learning_rate": 5.369126865350576e-05,
+      "loss": 2.8673,
+      "step": 49376
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7260851263999939,
+      "learning_rate": 5.368247950312889e-05,
+      "loss": 2.8445,
+      "step": 49377
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7297760248184204,
+      "learning_rate": 5.367369100150508e-05,
+      "loss": 2.8716,
+      "step": 49378
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7921993136405945,
+      "learning_rate": 5.366490314865727e-05,
+      "loss": 3.0418,
+      "step": 49379
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7311543226242065,
+      "learning_rate": 5.36561159446088e-05,
+      "loss": 2.768,
+      "step": 49380
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7326648831367493,
+      "learning_rate": 5.364732938938266e-05,
+      "loss": 3.0984,
+      "step": 49381
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6894707083702087,
+      "learning_rate": 5.363854348300201e-05,
+      "loss": 2.6443,
+      "step": 49382
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.754602313041687,
+      "learning_rate": 5.362975822549007e-05,
+      "loss": 2.7576,
+      "step": 49383
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7371578216552734,
+      "learning_rate": 5.362097361686988e-05,
+      "loss": 2.9871,
+      "step": 49384
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7002536058425903,
+      "learning_rate": 5.36121896571647e-05,
+      "loss": 2.982,
+      "step": 49385
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7025123238563538,
+      "learning_rate": 5.3603406346397504e-05,
+      "loss": 2.8782,
+      "step": 49386
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7461484670639038,
+      "learning_rate": 5.359462368459162e-05,
+      "loss": 3.0057,
+      "step": 49387
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7868021130561829,
+      "learning_rate": 5.3585841671770036e-05,
+      "loss": 2.9907,
+      "step": 49388
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7514649033546448,
+      "learning_rate": 5.3577060307955867e-05,
+      "loss": 3.0199,
+      "step": 49389
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7672540545463562,
+      "learning_rate": 5.356827959317239e-05,
+      "loss": 2.8189,
+      "step": 49390
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6891894340515137,
+      "learning_rate": 5.355949952744253e-05,
+      "loss": 2.785,
+      "step": 49391
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.775546669960022,
+      "learning_rate": 5.355072011078953e-05,
+      "loss": 2.8854,
+      "step": 49392
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7490105628967285,
+      "learning_rate": 5.35419413432366e-05,
+      "loss": 2.9474,
+      "step": 49393
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7415432333946228,
+      "learning_rate": 5.3533163224806755e-05,
+      "loss": 2.5371,
+      "step": 49394
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7412639260292053,
+      "learning_rate": 5.352438575552312e-05,
+      "loss": 2.828,
+      "step": 49395
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7198470234870911,
+      "learning_rate": 5.351560893540873e-05,
+      "loss": 2.9522,
+      "step": 49396
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7310846447944641,
+      "learning_rate": 5.3506832764486776e-05,
+      "loss": 3.1216,
+      "step": 49397
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.776674747467041,
+      "learning_rate": 5.34980572427805e-05,
+      "loss": 2.9169,
+      "step": 49398
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7521768808364868,
+      "learning_rate": 5.348928237031279e-05,
+      "loss": 3.0317,
+      "step": 49399
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7370387315750122,
+      "learning_rate": 5.348050814710696e-05,
+      "loss": 3.0164,
+      "step": 49400
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.727357804775238,
+      "learning_rate": 5.347173457318595e-05,
+      "loss": 2.8261,
+      "step": 49401
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.697250247001648,
+      "learning_rate": 5.346296164857305e-05,
+      "loss": 2.782,
+      "step": 49402
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7319405674934387,
+      "learning_rate": 5.3454189373291246e-05,
+      "loss": 2.8101,
+      "step": 49403
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7242304682731628,
+      "learning_rate": 5.344541774736357e-05,
+      "loss": 2.9944,
+      "step": 49404
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7132266163825989,
+      "learning_rate": 5.3436646770813286e-05,
+      "loss": 2.9694,
+      "step": 49405
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7501279711723328,
+      "learning_rate": 5.342787644366337e-05,
+      "loss": 2.7354,
+      "step": 49406
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6961253881454468,
+      "learning_rate": 5.3419106765937025e-05,
+      "loss": 2.8499,
+      "step": 49407
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7209061980247498,
+      "learning_rate": 5.3410337737657216e-05,
+      "loss": 3.1587,
+      "step": 49408
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.718574047088623,
+      "learning_rate": 5.340156935884722e-05,
+      "loss": 2.867,
+      "step": 49409
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.8040580153465271,
+      "learning_rate": 5.3392801629530015e-05,
+      "loss": 2.8241,
+      "step": 49410
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.717680037021637,
+      "learning_rate": 5.338403454972865e-05,
+      "loss": 3.0446,
+      "step": 49411
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7457809448242188,
+      "learning_rate": 5.337526811946634e-05,
+      "loss": 2.5612,
+      "step": 49412
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7153359055519104,
+      "learning_rate": 5.336650233876604e-05,
+      "loss": 3.0441,
+      "step": 49413
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7195101976394653,
+      "learning_rate": 5.3357737207650886e-05,
+      "loss": 3.1782,
+      "step": 49414
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6793602705001831,
+      "learning_rate": 5.3348972726144104e-05,
+      "loss": 2.7251,
+      "step": 49415
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7900977730751038,
+      "learning_rate": 5.334020889426863e-05,
+      "loss": 2.7028,
+      "step": 49416
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7003040909767151,
+      "learning_rate": 5.333144571204756e-05,
+      "loss": 2.7757,
+      "step": 49417
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7215965986251831,
+      "learning_rate": 5.332268317950393e-05,
+      "loss": 2.8256,
+      "step": 49418
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6919942498207092,
+      "learning_rate": 5.33139212966609e-05,
+      "loss": 2.7628,
+      "step": 49419
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7243634462356567,
+      "learning_rate": 5.3305160063541576e-05,
+      "loss": 2.8523,
+      "step": 49420
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7183544635772705,
+      "learning_rate": 5.329639948016891e-05,
+      "loss": 2.7497,
+      "step": 49421
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7528842687606812,
+      "learning_rate": 5.3287639546566144e-05,
+      "loss": 2.9952,
+      "step": 49422
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7461656332015991,
+      "learning_rate": 5.327888026275616e-05,
+      "loss": 2.9245,
+      "step": 49423
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7598857879638672,
+      "learning_rate": 5.327012162876219e-05,
+      "loss": 2.8832,
+      "step": 49424
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7612981200218201,
+      "learning_rate": 5.3261363644607255e-05,
+      "loss": 3.0593,
+      "step": 49425
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 5.325260631031433e-05,
+      "loss": 2.6926,
+      "step": 49426
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7351194620132446,
+      "learning_rate": 5.324384962590661e-05,
+      "loss": 3.1384,
+      "step": 49427
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7707248330116272,
+      "learning_rate": 5.323509359140702e-05,
+      "loss": 2.9092,
+      "step": 49428
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7522650957107544,
+      "learning_rate": 5.322633820683874e-05,
+      "loss": 2.8542,
+      "step": 49429
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7015370726585388,
+      "learning_rate": 5.321758347222481e-05,
+      "loss": 2.7529,
+      "step": 49430
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.733875036239624,
+      "learning_rate": 5.320882938758833e-05,
+      "loss": 2.9341,
+      "step": 49431
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7439164519309998,
+      "learning_rate": 5.320007595295226e-05,
+      "loss": 3.0875,
+      "step": 49432
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7210975289344788,
+      "learning_rate": 5.319132316833963e-05,
+      "loss": 2.8665,
+      "step": 49433
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7494742274284363,
+      "learning_rate": 5.3182571033773543e-05,
+      "loss": 2.8854,
+      "step": 49434
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7517364025115967,
+      "learning_rate": 5.3173819549277164e-05,
+      "loss": 2.7901,
+      "step": 49435
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.8002448678016663,
+      "learning_rate": 5.316506871487336e-05,
+      "loss": 2.639,
+      "step": 49436
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6921734809875488,
+      "learning_rate": 5.3156318530585315e-05,
+      "loss": 2.8351,
+      "step": 49437
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7180572152137756,
+      "learning_rate": 5.314756899643605e-05,
+      "loss": 2.8778,
+      "step": 49438
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7430548667907715,
+      "learning_rate": 5.313882011244848e-05,
+      "loss": 3.0747,
+      "step": 49439
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7170864939689636,
+      "learning_rate": 5.313007187864584e-05,
+      "loss": 2.9257,
+      "step": 49440
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7517858743667603,
+      "learning_rate": 5.312132429505099e-05,
+      "loss": 2.9093,
+      "step": 49441
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7107861042022705,
+      "learning_rate": 5.311257736168715e-05,
+      "loss": 2.7721,
+      "step": 49442
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7427153587341309,
+      "learning_rate": 5.310383107857718e-05,
+      "loss": 2.8472,
+      "step": 49443
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7173393964767456,
+      "learning_rate": 5.309508544574419e-05,
+      "loss": 2.7931,
+      "step": 49444
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7084386944770813,
+      "learning_rate": 5.308634046321131e-05,
+      "loss": 2.6603,
+      "step": 49445
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6978672742843628,
+      "learning_rate": 5.307759613100152e-05,
+      "loss": 2.8571,
+      "step": 49446
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7057198286056519,
+      "learning_rate": 5.3068852449137775e-05,
+      "loss": 2.7171,
+      "step": 49447
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7628961801528931,
+      "learning_rate": 5.306010941764313e-05,
+      "loss": 2.8952,
+      "step": 49448
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7418630123138428,
+      "learning_rate": 5.305136703654058e-05,
+      "loss": 2.7643,
+      "step": 49449
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7383913993835449,
+      "learning_rate": 5.304262530585329e-05,
+      "loss": 2.8741,
+      "step": 49450
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7368167042732239,
+      "learning_rate": 5.303388422560414e-05,
+      "loss": 2.8344,
+      "step": 49451
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7100529670715332,
+      "learning_rate": 5.302514379581627e-05,
+      "loss": 2.9651,
+      "step": 49452
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7233177423477173,
+      "learning_rate": 5.301640401651267e-05,
+      "loss": 3.1435,
+      "step": 49453
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7448151111602783,
+      "learning_rate": 5.3007664887716216e-05,
+      "loss": 2.6896,
+      "step": 49454
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7152411341667175,
+      "learning_rate": 5.2998926409450126e-05,
+      "loss": 2.9082,
+      "step": 49455
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7502825260162354,
+      "learning_rate": 5.299018858173725e-05,
+      "loss": 2.7657,
+      "step": 49456
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7130440473556519,
+      "learning_rate": 5.298145140460076e-05,
+      "loss": 2.8697,
+      "step": 49457
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7099788188934326,
+      "learning_rate": 5.297271487806352e-05,
+      "loss": 2.6939,
+      "step": 49458
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.706955075263977,
+      "learning_rate": 5.2963979002148596e-05,
+      "loss": 2.7304,
+      "step": 49459
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7011268734931946,
+      "learning_rate": 5.295524377687914e-05,
+      "loss": 2.7672,
+      "step": 49460
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7620493769645691,
+      "learning_rate": 5.294650920227791e-05,
+      "loss": 2.7686,
+      "step": 49461
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7569233775138855,
+      "learning_rate": 5.293777527836808e-05,
+      "loss": 2.7607,
+      "step": 49462
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7677346467971802,
+      "learning_rate": 5.292904200517253e-05,
+      "loss": 2.9383,
+      "step": 49463
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7182645201683044,
+      "learning_rate": 5.292030938271434e-05,
+      "loss": 2.9959,
+      "step": 49464
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7732892036437988,
+      "learning_rate": 5.291157741101657e-05,
+      "loss": 2.8637,
+      "step": 49465
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7590673565864563,
+      "learning_rate": 5.290284609010209e-05,
+      "loss": 2.7361,
+      "step": 49466
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6932297945022583,
+      "learning_rate": 5.289411541999406e-05,
+      "loss": 3.0343,
+      "step": 49467
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7510533332824707,
+      "learning_rate": 5.288538540071534e-05,
+      "loss": 2.8191,
+      "step": 49468
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7724784016609192,
+      "learning_rate": 5.287665603228887e-05,
+      "loss": 2.7771,
+      "step": 49469
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7106216549873352,
+      "learning_rate": 5.2867927314737815e-05,
+      "loss": 2.9812,
+      "step": 49470
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7474266290664673,
+      "learning_rate": 5.2859199248084994e-05,
+      "loss": 2.8478,
+      "step": 49471
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7339901924133301,
+      "learning_rate": 5.2850471832353594e-05,
+      "loss": 2.9416,
+      "step": 49472
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7034941911697388,
+      "learning_rate": 5.284174506756636e-05,
+      "loss": 2.9021,
+      "step": 49473
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7442716956138611,
+      "learning_rate": 5.28330189537465e-05,
+      "loss": 2.9255,
+      "step": 49474
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7037724852561951,
+      "learning_rate": 5.28242934909169e-05,
+      "loss": 2.6829,
+      "step": 49475
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7233594655990601,
+      "learning_rate": 5.281556867910048e-05,
+      "loss": 2.8918,
+      "step": 49476
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7807363867759705,
+      "learning_rate": 5.280684451832032e-05,
+      "loss": 2.9158,
+      "step": 49477
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7810481786727905,
+      "learning_rate": 5.279812100859934e-05,
+      "loss": 2.9131,
+      "step": 49478
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7510912418365479,
+      "learning_rate": 5.278939814996057e-05,
+      "loss": 2.8349,
+      "step": 49479
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.8030551671981812,
+      "learning_rate": 5.2780675942426874e-05,
+      "loss": 2.7635,
+      "step": 49480
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7197742462158203,
+      "learning_rate": 5.2771954386021296e-05,
+      "loss": 3.1214,
+      "step": 49481
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7838296890258789,
+      "learning_rate": 5.276323348076689e-05,
+      "loss": 3.0325,
+      "step": 49482
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.777555525302887,
+      "learning_rate": 5.2754513226686556e-05,
+      "loss": 2.9526,
+      "step": 49483
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7960125207901001,
+      "learning_rate": 5.274579362380322e-05,
+      "loss": 2.8986,
+      "step": 49484
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7355430126190186,
+      "learning_rate": 5.2737074672139837e-05,
+      "loss": 2.8928,
+      "step": 49485
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7454689741134644,
+      "learning_rate": 5.272835637171941e-05,
+      "loss": 2.8384,
+      "step": 49486
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7542171478271484,
+      "learning_rate": 5.2719638722564986e-05,
+      "loss": 3.036,
+      "step": 49487
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7559950947761536,
+      "learning_rate": 5.271092172469935e-05,
+      "loss": 2.6835,
+      "step": 49488
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6909935474395752,
+      "learning_rate": 5.270220537814562e-05,
+      "loss": 2.8508,
+      "step": 49489
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7282992005348206,
+      "learning_rate": 5.269348968292671e-05,
+      "loss": 2.9179,
+      "step": 49490
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7650557160377502,
+      "learning_rate": 5.268477463906548e-05,
+      "loss": 2.8272,
+      "step": 49491
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.744831383228302,
+      "learning_rate": 5.267606024658502e-05,
+      "loss": 2.8916,
+      "step": 49492
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7628011107444763,
+      "learning_rate": 5.266734650550813e-05,
+      "loss": 2.6101,
+      "step": 49493
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7256711721420288,
+      "learning_rate": 5.265863341585794e-05,
+      "loss": 3.0817,
+      "step": 49494
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7441546320915222,
+      "learning_rate": 5.2649920977657225e-05,
+      "loss": 2.9964,
+      "step": 49495
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7593109607696533,
+      "learning_rate": 5.264120919092911e-05,
+      "loss": 2.8387,
+      "step": 49496
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.722559928894043,
+      "learning_rate": 5.2632498055696424e-05,
+      "loss": 3.0698,
+      "step": 49497
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7306622862815857,
+      "learning_rate": 5.2623787571982044e-05,
+      "loss": 2.8901,
+      "step": 49498
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7256033420562744,
+      "learning_rate": 5.261507773980912e-05,
+      "loss": 2.958,
+      "step": 49499
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7559966444969177,
+      "learning_rate": 5.260636855920036e-05,
+      "loss": 2.9003,
+      "step": 49500
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7539312839508057,
+      "learning_rate": 5.2597660030178825e-05,
+      "loss": 2.7591,
+      "step": 49501
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7615699172019958,
+      "learning_rate": 5.258895215276752e-05,
+      "loss": 2.8271,
+      "step": 49502
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.720807671546936,
+      "learning_rate": 5.258024492698919e-05,
+      "loss": 2.739,
+      "step": 49503
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.8670095801353455,
+      "learning_rate": 5.2571538352866994e-05,
+      "loss": 2.8535,
+      "step": 49504
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.708129346370697,
+      "learning_rate": 5.256283243042374e-05,
+      "loss": 3.1879,
+      "step": 49505
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6828888058662415,
+      "learning_rate": 5.255412715968228e-05,
+      "loss": 3.0061,
+      "step": 49506
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.722131073474884,
+      "learning_rate": 5.254542254066569e-05,
+      "loss": 3.0957,
+      "step": 49507
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7353353500366211,
+      "learning_rate": 5.2536718573396796e-05,
+      "loss": 2.8735,
+      "step": 49508
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7344322204589844,
+      "learning_rate": 5.2528015257898605e-05,
+      "loss": 2.8081,
+      "step": 49509
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7135079503059387,
+      "learning_rate": 5.251931259419393e-05,
+      "loss": 2.9412,
+      "step": 49510
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7717560529708862,
+      "learning_rate": 5.251061058230582e-05,
+      "loss": 2.8145,
+      "step": 49511
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7272468209266663,
+      "learning_rate": 5.250190922225717e-05,
+      "loss": 2.9457,
+      "step": 49512
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7593507170677185,
+      "learning_rate": 5.2493208514070736e-05,
+      "loss": 2.7668,
+      "step": 49513
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7987231612205505,
+      "learning_rate": 5.2484508457769657e-05,
+      "loss": 2.7785,
+      "step": 49514
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6929014325141907,
+      "learning_rate": 5.247580905337667e-05,
+      "loss": 2.7896,
+      "step": 49515
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7104371190071106,
+      "learning_rate": 5.246711030091476e-05,
+      "loss": 2.8161,
+      "step": 49516
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6814081072807312,
+      "learning_rate": 5.245841220040691e-05,
+      "loss": 2.8394,
+      "step": 49517
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.691145658493042,
+      "learning_rate": 5.244971475187599e-05,
+      "loss": 2.9412,
+      "step": 49518
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7448225021362305,
+      "learning_rate": 5.244101795534487e-05,
+      "loss": 2.8151,
+      "step": 49519
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7341607213020325,
+      "learning_rate": 5.243232181083638e-05,
+      "loss": 2.8329,
+      "step": 49520
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7958085536956787,
+      "learning_rate": 5.24236263183735e-05,
+      "loss": 2.9727,
+      "step": 49521
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7446455359458923,
+      "learning_rate": 5.2414931477979225e-05,
+      "loss": 3.0168,
+      "step": 49522
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7200234532356262,
+      "learning_rate": 5.24062372896763e-05,
+      "loss": 2.8772,
+      "step": 49523
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7179722785949707,
+      "learning_rate": 5.239754375348777e-05,
+      "loss": 2.9839,
+      "step": 49524
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7151455879211426,
+      "learning_rate": 5.2388850869436385e-05,
+      "loss": 2.7769,
+      "step": 49525
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7526031732559204,
+      "learning_rate": 5.238015863754519e-05,
+      "loss": 3.0104,
+      "step": 49526
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7519762516021729,
+      "learning_rate": 5.2371467057837003e-05,
+      "loss": 2.9088,
+      "step": 49527
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7097587585449219,
+      "learning_rate": 5.236277613033461e-05,
+      "loss": 2.985,
+      "step": 49528
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6919777989387512,
+      "learning_rate": 5.235408585506111e-05,
+      "loss": 2.8978,
+      "step": 49529
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7104206681251526,
+      "learning_rate": 5.2345396232039205e-05,
+      "loss": 2.9985,
+      "step": 49530
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.721882164478302,
+      "learning_rate": 5.2336707261291865e-05,
+      "loss": 2.9761,
+      "step": 49531
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7098190784454346,
+      "learning_rate": 5.2328018942842044e-05,
+      "loss": 2.9406,
+      "step": 49532
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7680677771568298,
+      "learning_rate": 5.231933127671258e-05,
+      "loss": 2.9672,
+      "step": 49533
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7131537795066833,
+      "learning_rate": 5.2310644262926305e-05,
+      "loss": 2.9661,
+      "step": 49534
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7221238017082214,
+      "learning_rate": 5.2301957901506054e-05,
+      "loss": 2.9696,
+      "step": 49535
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7687680125236511,
+      "learning_rate": 5.229327219247478e-05,
+      "loss": 2.8797,
+      "step": 49536
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7497595548629761,
+      "learning_rate": 5.228458713585543e-05,
+      "loss": 2.8141,
+      "step": 49537
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7552316784858704,
+      "learning_rate": 5.2275902731670716e-05,
+      "loss": 2.9196,
+      "step": 49538
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7658446431159973,
+      "learning_rate": 5.2267218979943695e-05,
+      "loss": 2.9684,
+      "step": 49539
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7868543267250061,
+      "learning_rate": 5.225853588069714e-05,
+      "loss": 2.8374,
+      "step": 49540
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.72367262840271,
+      "learning_rate": 5.224985343395384e-05,
+      "loss": 2.9543,
+      "step": 49541
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7303585410118103,
+      "learning_rate": 5.22411716397368e-05,
+      "loss": 2.9499,
+      "step": 49542
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7236695885658264,
+      "learning_rate": 5.223249049806878e-05,
+      "loss": 3.0774,
+      "step": 49543
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7127900123596191,
+      "learning_rate": 5.222381000897279e-05,
+      "loss": 3.1424,
+      "step": 49544
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7575745582580566,
+      "learning_rate": 5.2215130172471476e-05,
+      "loss": 3.0335,
+      "step": 49545
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7566746473312378,
+      "learning_rate": 5.220645098858786e-05,
+      "loss": 2.8638,
+      "step": 49546
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7185924053192139,
+      "learning_rate": 5.219777245734484e-05,
+      "loss": 2.9908,
+      "step": 49547
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7872166633605957,
+      "learning_rate": 5.218909457876515e-05,
+      "loss": 2.8108,
+      "step": 49548
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7263684868812561,
+      "learning_rate": 5.2180417352871736e-05,
+      "loss": 2.7144,
+      "step": 49549
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7238753437995911,
+      "learning_rate": 5.2171740779687346e-05,
+      "loss": 2.9509,
+      "step": 49550
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7120121121406555,
+      "learning_rate": 5.216306485923485e-05,
+      "loss": 2.8628,
+      "step": 49551
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6943619251251221,
+      "learning_rate": 5.2154389591537236e-05,
+      "loss": 2.7111,
+      "step": 49552
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.709274411201477,
+      "learning_rate": 5.214571497661723e-05,
+      "loss": 2.7601,
+      "step": 49553
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7480226755142212,
+      "learning_rate": 5.2137041014497726e-05,
+      "loss": 2.9873,
+      "step": 49554
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7200178503990173,
+      "learning_rate": 5.2128367705201594e-05,
+      "loss": 2.7593,
+      "step": 49555
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7735962867736816,
+      "learning_rate": 5.211969504875153e-05,
+      "loss": 2.8678,
+      "step": 49556
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7781702876091003,
+      "learning_rate": 5.2111023045170565e-05,
+      "loss": 2.9205,
+      "step": 49557
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7568464875221252,
+      "learning_rate": 5.21023516944814e-05,
+      "loss": 3.0284,
+      "step": 49558
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7191801071166992,
+      "learning_rate": 5.2093680996707e-05,
+      "loss": 2.9578,
+      "step": 49559
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7512852549552917,
+      "learning_rate": 5.208501095187003e-05,
+      "loss": 3.0233,
+      "step": 49560
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7531700134277344,
+      "learning_rate": 5.207634155999355e-05,
+      "loss": 2.6399,
+      "step": 49561
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7530140280723572,
+      "learning_rate": 5.2067672821100234e-05,
+      "loss": 3.0416,
+      "step": 49562
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7887686491012573,
+      "learning_rate": 5.2059004735212904e-05,
+      "loss": 3.0472,
+      "step": 49563
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7277945280075073,
+      "learning_rate": 5.20503373023545e-05,
+      "loss": 2.7531,
+      "step": 49564
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7439764142036438,
+      "learning_rate": 5.2041670522547715e-05,
+      "loss": 2.9304,
+      "step": 49565
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.737322986125946,
+      "learning_rate": 5.203300439581556e-05,
+      "loss": 3.0708,
+      "step": 49566
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7382588982582092,
+      "learning_rate": 5.2024338922180606e-05,
+      "loss": 2.8939,
+      "step": 49567
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.729706883430481,
+      "learning_rate": 5.201567410166588e-05,
+      "loss": 3.2045,
+      "step": 49568
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7282545566558838,
+      "learning_rate": 5.20070099342942e-05,
+      "loss": 2.8105,
+      "step": 49569
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7177003026008606,
+      "learning_rate": 5.19983464200883e-05,
+      "loss": 2.7949,
+      "step": 49570
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7702353000640869,
+      "learning_rate": 5.1989683559071074e-05,
+      "loss": 2.9669,
+      "step": 49571
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7214787602424622,
+      "learning_rate": 5.198102135126516e-05,
+      "loss": 2.8121,
+      "step": 49572
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.8282303214073181,
+      "learning_rate": 5.1972359796693554e-05,
+      "loss": 2.9693,
+      "step": 49573
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7337009906768799,
+      "learning_rate": 5.196369889537906e-05,
+      "loss": 2.7557,
+      "step": 49574
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7773966789245605,
+      "learning_rate": 5.195503864734439e-05,
+      "loss": 2.9181,
+      "step": 49575
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6771313548088074,
+      "learning_rate": 5.194637905261247e-05,
+      "loss": 2.6734,
+      "step": 49576
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7496398091316223,
+      "learning_rate": 5.1937720111206085e-05,
+      "loss": 2.9345,
+      "step": 49577
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7166499495506287,
+      "learning_rate": 5.19290618231479e-05,
+      "loss": 2.7787,
+      "step": 49578
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7843281626701355,
+      "learning_rate": 5.192040418846091e-05,
+      "loss": 2.7845,
+      "step": 49579
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7141745686531067,
+      "learning_rate": 5.1911747207167754e-05,
+      "loss": 2.7912,
+      "step": 49580
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7347038388252258,
+      "learning_rate": 5.1903090879291384e-05,
+      "loss": 2.7822,
+      "step": 49581
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7614521980285645,
+      "learning_rate": 5.189443520485442e-05,
+      "loss": 2.749,
+      "step": 49582
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7512968182563782,
+      "learning_rate": 5.188578018387981e-05,
+      "loss": 2.9849,
+      "step": 49583
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7517346143722534,
+      "learning_rate": 5.1877125816390365e-05,
+      "loss": 2.7312,
+      "step": 49584
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.728527307510376,
+      "learning_rate": 5.1868472102408834e-05,
+      "loss": 2.9731,
+      "step": 49585
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7840355634689331,
+      "learning_rate": 5.185981904195796e-05,
+      "loss": 2.8932,
+      "step": 49586
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7409562468528748,
+      "learning_rate": 5.18511666350605e-05,
+      "loss": 3.0215,
+      "step": 49587
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.756843626499176,
+      "learning_rate": 5.184251488173934e-05,
+      "loss": 3.0411,
+      "step": 49588
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7006279826164246,
+      "learning_rate": 5.1833863782017284e-05,
+      "loss": 2.7748,
+      "step": 49589
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.708868145942688,
+      "learning_rate": 5.182521333591698e-05,
+      "loss": 2.8136,
+      "step": 49590
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.779094934463501,
+      "learning_rate": 5.181656354346142e-05,
+      "loss": 2.9222,
+      "step": 49591
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7329817414283752,
+      "learning_rate": 5.1807914404673244e-05,
+      "loss": 2.7675,
+      "step": 49592
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7352889180183411,
+      "learning_rate": 5.1799265919575175e-05,
+      "loss": 2.6653,
+      "step": 49593
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.737707793712616,
+      "learning_rate": 5.179061808819015e-05,
+      "loss": 3.0523,
+      "step": 49594
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.737191915512085,
+      "learning_rate": 5.178197091054079e-05,
+      "loss": 3.004,
+      "step": 49595
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7531353831291199,
+      "learning_rate": 5.177332438665005e-05,
+      "loss": 3.0624,
+      "step": 49596
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6949726939201355,
+      "learning_rate": 5.17646785165405e-05,
+      "loss": 2.8186,
+      "step": 49597
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6904205083847046,
+      "learning_rate": 5.17560333002351e-05,
+      "loss": 3.0604,
+      "step": 49598
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7312195301055908,
+      "learning_rate": 5.174738873775652e-05,
+      "loss": 3.1245,
+      "step": 49599
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7542881965637207,
+      "learning_rate": 5.173874482912749e-05,
+      "loss": 2.9555,
+      "step": 49600
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7179239988327026,
+      "learning_rate": 5.173010157437088e-05,
+      "loss": 2.8191,
+      "step": 49601
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7660733461380005,
+      "learning_rate": 5.172145897350933e-05,
+      "loss": 2.7249,
+      "step": 49602
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7874742150306702,
+      "learning_rate": 5.171281702656568e-05,
+      "loss": 3.0061,
+      "step": 49603
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7039852738380432,
+      "learning_rate": 5.170417573356276e-05,
+      "loss": 3.0712,
+      "step": 49604
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7054070830345154,
+      "learning_rate": 5.169553509452319e-05,
+      "loss": 2.7834,
+      "step": 49605
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7984621524810791,
+      "learning_rate": 5.168689510946985e-05,
+      "loss": 2.8207,
+      "step": 49606
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7172672748565674,
+      "learning_rate": 5.167825577842546e-05,
+      "loss": 2.8539,
+      "step": 49607
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7304677367210388,
+      "learning_rate": 5.1669617101412663e-05,
+      "loss": 2.7716,
+      "step": 49608
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6986484527587891,
+      "learning_rate": 5.166097907845438e-05,
+      "loss": 3.0376,
+      "step": 49609
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7038609981536865,
+      "learning_rate": 5.165234170957319e-05,
+      "loss": 2.8795,
+      "step": 49610
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7226280570030212,
+      "learning_rate": 5.164370499479204e-05,
+      "loss": 2.8304,
+      "step": 49611
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7109252214431763,
+      "learning_rate": 5.16350689341335e-05,
+      "loss": 2.8575,
+      "step": 49612
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7385297417640686,
+      "learning_rate": 5.162643352762045e-05,
+      "loss": 2.8261,
+      "step": 49613
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7628830671310425,
+      "learning_rate": 5.1617798775275576e-05,
+      "loss": 2.9239,
+      "step": 49614
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.787331759929657,
+      "learning_rate": 5.160916467712154e-05,
+      "loss": 3.0273,
+      "step": 49615
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7409013509750366,
+      "learning_rate": 5.160053123318125e-05,
+      "loss": 2.9011,
+      "step": 49616
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7271203994750977,
+      "learning_rate": 5.1591898443477263e-05,
+      "loss": 2.9749,
+      "step": 49617
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7120565176010132,
+      "learning_rate": 5.158326630803242e-05,
+      "loss": 2.8932,
+      "step": 49618
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.706852376461029,
+      "learning_rate": 5.1574634826869544e-05,
+      "loss": 3.0749,
+      "step": 49619
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7524295449256897,
+      "learning_rate": 5.156600400001124e-05,
+      "loss": 2.855,
+      "step": 49620
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7389339208602905,
+      "learning_rate": 5.1557373827480264e-05,
+      "loss": 2.967,
+      "step": 49621
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7299748063087463,
+      "learning_rate": 5.1548744309299295e-05,
+      "loss": 2.9577,
+      "step": 49622
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7077644467353821,
+      "learning_rate": 5.154011544549108e-05,
+      "loss": 2.8618,
+      "step": 49623
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7535231113433838,
+      "learning_rate": 5.153148723607851e-05,
+      "loss": 2.6616,
+      "step": 49624
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7145181894302368,
+      "learning_rate": 5.1522859681084084e-05,
+      "loss": 2.9405,
+      "step": 49625
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7331188917160034,
+      "learning_rate": 5.151423278053069e-05,
+      "loss": 2.9301,
+      "step": 49626
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.756357729434967,
+      "learning_rate": 5.150560653444093e-05,
+      "loss": 2.965,
+      "step": 49627
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7148869633674622,
+      "learning_rate": 5.149698094283765e-05,
+      "loss": 2.9109,
+      "step": 49628
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7142993211746216,
+      "learning_rate": 5.148835600574349e-05,
+      "loss": 3.0988,
+      "step": 49629
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.8061341047286987,
+      "learning_rate": 5.14797317231811e-05,
+      "loss": 2.7739,
+      "step": 49630
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7073236107826233,
+      "learning_rate": 5.147110809517334e-05,
+      "loss": 2.8595,
+      "step": 49631
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7803508043289185,
+      "learning_rate": 5.146248512174278e-05,
+      "loss": 2.7204,
+      "step": 49632
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.758151113986969,
+      "learning_rate": 5.1453862802912195e-05,
+      "loss": 2.9361,
+      "step": 49633
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7642468214035034,
+      "learning_rate": 5.1445241138704395e-05,
+      "loss": 2.6602,
+      "step": 49634
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7350394129753113,
+      "learning_rate": 5.143662012914197e-05,
+      "loss": 2.734,
+      "step": 49635
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6950120329856873,
+      "learning_rate": 5.1427999774247664e-05,
+      "loss": 2.6326,
+      "step": 49636
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7212783694267273,
+      "learning_rate": 5.141938007404409e-05,
+      "loss": 2.8121,
+      "step": 49637
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7310404777526855,
+      "learning_rate": 5.14107610285541e-05,
+      "loss": 2.9614,
+      "step": 49638
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7168389558792114,
+      "learning_rate": 5.1402142637800246e-05,
+      "loss": 2.9321,
+      "step": 49639
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7117984294891357,
+      "learning_rate": 5.139352490180534e-05,
+      "loss": 2.7888,
+      "step": 49640
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7920997738838196,
+      "learning_rate": 5.138490782059209e-05,
+      "loss": 2.9202,
+      "step": 49641
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6791305541992188,
+      "learning_rate": 5.1376291394183154e-05,
+      "loss": 2.5108,
+      "step": 49642
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7199430465698242,
+      "learning_rate": 5.136767562260121e-05,
+      "loss": 2.7794,
+      "step": 49643
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6989941596984863,
+      "learning_rate": 5.135906050586888e-05,
+      "loss": 2.9884,
+      "step": 49644
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7715273499488831,
+      "learning_rate": 5.135044604400893e-05,
+      "loss": 2.8518,
+      "step": 49645
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7331072688102722,
+      "learning_rate": 5.134183223704412e-05,
+      "loss": 2.9443,
+      "step": 49646
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7313644886016846,
+      "learning_rate": 5.1333219084997036e-05,
+      "loss": 2.8292,
+      "step": 49647
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7467599511146545,
+      "learning_rate": 5.1324606587890416e-05,
+      "loss": 3.0896,
+      "step": 49648
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7694347500801086,
+      "learning_rate": 5.131599474574688e-05,
+      "loss": 3.0131,
+      "step": 49649
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7463330626487732,
+      "learning_rate": 5.130738355858924e-05,
+      "loss": 3.0468,
+      "step": 49650
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7481557130813599,
+      "learning_rate": 5.1298773026440056e-05,
+      "loss": 2.9252,
+      "step": 49651
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.705471932888031,
+      "learning_rate": 5.129016314932199e-05,
+      "loss": 2.9475,
+      "step": 49652
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7567241191864014,
+      "learning_rate": 5.128155392725785e-05,
+      "loss": 2.8548,
+      "step": 49653
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.6947323083877563,
+      "learning_rate": 5.127294536027012e-05,
+      "loss": 2.675,
+      "step": 49654
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7445532083511353,
+      "learning_rate": 5.126433744838161e-05,
+      "loss": 3.0083,
+      "step": 49655
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7622706294059753,
+      "learning_rate": 5.1255730191615055e-05,
+      "loss": 2.7358,
+      "step": 49656
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7756778597831726,
+      "learning_rate": 5.1247123589993e-05,
+      "loss": 3.0063,
+      "step": 49657
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7175281047821045,
+      "learning_rate": 5.123851764353817e-05,
+      "loss": 2.9401,
+      "step": 49658
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.710747480392456,
+      "learning_rate": 5.122991235227314e-05,
+      "loss": 2.8863,
+      "step": 49659
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7191362977027893,
+      "learning_rate": 5.122130771622063e-05,
+      "loss": 2.8623,
+      "step": 49660
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7421351075172424,
+      "learning_rate": 5.121270373540338e-05,
+      "loss": 2.8252,
+      "step": 49661
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7567232847213745,
+      "learning_rate": 5.120410040984395e-05,
+      "loss": 3.1095,
+      "step": 49662
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7147611975669861,
+      "learning_rate": 5.1195497739565086e-05,
+      "loss": 2.9548,
+      "step": 49663
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7359359264373779,
+      "learning_rate": 5.11868957245894e-05,
+      "loss": 2.8819,
+      "step": 49664
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7486942410469055,
+      "learning_rate": 5.1178294364939464e-05,
+      "loss": 3.0448,
+      "step": 49665
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7085438966751099,
+      "learning_rate": 5.116969366063811e-05,
+      "loss": 2.9634,
+      "step": 49666
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7525193095207214,
+      "learning_rate": 5.1161093611707816e-05,
+      "loss": 2.8993,
+      "step": 49667
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.760326623916626,
+      "learning_rate": 5.115249421817136e-05,
+      "loss": 3.0144,
+      "step": 49668
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.75190669298172,
+      "learning_rate": 5.114389548005129e-05,
+      "loss": 2.6683,
+      "step": 49669
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7118992805480957,
+      "learning_rate": 5.113529739737029e-05,
+      "loss": 3.0528,
+      "step": 49670
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7461121082305908,
+      "learning_rate": 5.112669997015111e-05,
+      "loss": 2.8172,
+      "step": 49671
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7285375595092773,
+      "learning_rate": 5.1118103198416325e-05,
+      "loss": 2.8323,
+      "step": 49672
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7130274176597595,
+      "learning_rate": 5.110950708218853e-05,
+      "loss": 2.9635,
+      "step": 49673
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7329024076461792,
+      "learning_rate": 5.110091162149033e-05,
+      "loss": 2.9495,
+      "step": 49674
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7448126077651978,
+      "learning_rate": 5.1092316816344404e-05,
+      "loss": 3.0704,
+      "step": 49675
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7353721857070923,
+      "learning_rate": 5.108372266677351e-05,
+      "loss": 2.918,
+      "step": 49676
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7445089817047119,
+      "learning_rate": 5.107512917280012e-05,
+      "loss": 3.0014,
+      "step": 49677
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7183898091316223,
+      "learning_rate": 5.1066536334447e-05,
+      "loss": 2.6663,
+      "step": 49678
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7457320690155029,
+      "learning_rate": 5.105794415173671e-05,
+      "loss": 2.7268,
+      "step": 49679
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7721786499023438,
+      "learning_rate": 5.104935262469181e-05,
+      "loss": 2.8341,
+      "step": 49680
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7069118022918701,
+      "learning_rate": 5.1040761753335044e-05,
+      "loss": 2.6973,
+      "step": 49681
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7151609659194946,
+      "learning_rate": 5.103217153768897e-05,
+      "loss": 2.901,
+      "step": 49682
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7430471777915955,
+      "learning_rate": 5.1023581977776296e-05,
+      "loss": 3.03,
+      "step": 49683
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7141035199165344,
+      "learning_rate": 5.1014993073619536e-05,
+      "loss": 2.9409,
+      "step": 49684
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7425801753997803,
+      "learning_rate": 5.100640482524134e-05,
+      "loss": 2.962,
+      "step": 49685
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.8909529447555542,
+      "learning_rate": 5.099781723266453e-05,
+      "loss": 2.8787,
+      "step": 49686
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7250303030014038,
+      "learning_rate": 5.098923029591138e-05,
+      "loss": 2.8883,
+      "step": 49687
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7350665926933289,
+      "learning_rate": 5.098064401500477e-05,
+      "loss": 2.9089,
+      "step": 49688
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7405259013175964,
+      "learning_rate": 5.0972058389967116e-05,
+      "loss": 2.9799,
+      "step": 49689
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7162747383117676,
+      "learning_rate": 5.096347342082114e-05,
+      "loss": 2.8799,
+      "step": 49690
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7429664731025696,
+      "learning_rate": 5.095488910758956e-05,
+      "loss": 2.8375,
+      "step": 49691
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7161949872970581,
+      "learning_rate": 5.094630545029475e-05,
+      "loss": 2.8939,
+      "step": 49692
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7380734086036682,
+      "learning_rate": 5.093772244895952e-05,
+      "loss": 2.928,
+      "step": 49693
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.778752326965332,
+      "learning_rate": 5.092914010360642e-05,
+      "loss": 2.9385,
+      "step": 49694
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6871793866157532,
+      "learning_rate": 5.092055841425795e-05,
+      "loss": 2.8888,
+      "step": 49695
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7245429158210754,
+      "learning_rate": 5.091197738093684e-05,
+      "loss": 3.1834,
+      "step": 49696
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7372999787330627,
+      "learning_rate": 5.0903397003665616e-05,
+      "loss": 2.8991,
+      "step": 49697
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7384405732154846,
+      "learning_rate": 5.089481728246695e-05,
+      "loss": 3.0387,
+      "step": 49698
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7918789386749268,
+      "learning_rate": 5.088623821736333e-05,
+      "loss": 2.6822,
+      "step": 49699
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6995186805725098,
+      "learning_rate": 5.087765980837747e-05,
+      "loss": 2.7499,
+      "step": 49700
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7654358148574829,
+      "learning_rate": 5.0869082055531915e-05,
+      "loss": 2.7592,
+      "step": 49701
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7654218077659607,
+      "learning_rate": 5.0860504958849194e-05,
+      "loss": 2.7973,
+      "step": 49702
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7214524149894714,
+      "learning_rate": 5.0851928518352e-05,
+      "loss": 2.9906,
+      "step": 49703
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.737037718296051,
+      "learning_rate": 5.084335273406284e-05,
+      "loss": 2.8326,
+      "step": 49704
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6899082660675049,
+      "learning_rate": 5.083477760600431e-05,
+      "loss": 3.0577,
+      "step": 49705
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.717755138874054,
+      "learning_rate": 5.0826203134199095e-05,
+      "loss": 2.9758,
+      "step": 49706
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7414624094963074,
+      "learning_rate": 5.081762931866965e-05,
+      "loss": 2.791,
+      "step": 49707
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7816479802131653,
+      "learning_rate": 5.0809056159438686e-05,
+      "loss": 2.9598,
+      "step": 49708
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7575608491897583,
+      "learning_rate": 5.080048365652872e-05,
+      "loss": 2.9766,
+      "step": 49709
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7834826707839966,
+      "learning_rate": 5.079191180996224e-05,
+      "loss": 2.8922,
+      "step": 49710
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7255894541740417,
+      "learning_rate": 5.078334061976195e-05,
+      "loss": 3.0187,
+      "step": 49711
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7802708745002747,
+      "learning_rate": 5.0774770085950336e-05,
+      "loss": 2.9521,
+      "step": 49712
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6934137940406799,
+      "learning_rate": 5.076620020855008e-05,
+      "loss": 2.8263,
+      "step": 49713
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7279984951019287,
+      "learning_rate": 5.0757630987583596e-05,
+      "loss": 2.8758,
+      "step": 49714
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7556265592575073,
+      "learning_rate": 5.074906242307364e-05,
+      "loss": 2.8469,
+      "step": 49715
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.784926176071167,
+      "learning_rate": 5.074049451504265e-05,
+      "loss": 2.8924,
+      "step": 49716
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7134363651275635,
+      "learning_rate": 5.0731927263513185e-05,
+      "loss": 2.9432,
+      "step": 49717
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7337852120399475,
+      "learning_rate": 5.072336066850792e-05,
+      "loss": 3.0742,
+      "step": 49718
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7417206168174744,
+      "learning_rate": 5.071479473004927e-05,
+      "loss": 2.9438,
+      "step": 49719
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7676438689231873,
+      "learning_rate": 5.0706229448159954e-05,
+      "loss": 3.0447,
+      "step": 49720
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7735446691513062,
+      "learning_rate": 5.069766482286238e-05,
+      "loss": 3.1604,
+      "step": 49721
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7267362475395203,
+      "learning_rate": 5.0689100854179245e-05,
+      "loss": 2.8411,
+      "step": 49722
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7745874524116516,
+      "learning_rate": 5.068053754213305e-05,
+      "loss": 3.0884,
+      "step": 49723
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7192590236663818,
+      "learning_rate": 5.067197488674625e-05,
+      "loss": 2.8488,
+      "step": 49724
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.766343891620636,
+      "learning_rate": 5.066341288804155e-05,
+      "loss": 2.883,
+      "step": 49725
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7043959498405457,
+      "learning_rate": 5.0654851546041356e-05,
+      "loss": 2.8619,
+      "step": 49726
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7229441404342651,
+      "learning_rate": 5.064629086076832e-05,
+      "loss": 2.9009,
+      "step": 49727
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.71528559923172,
+      "learning_rate": 5.063773083224502e-05,
+      "loss": 2.8358,
+      "step": 49728
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7294989824295044,
+      "learning_rate": 5.062917146049388e-05,
+      "loss": 2.5765,
+      "step": 49729
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7297892570495605,
+      "learning_rate": 5.062061274553759e-05,
+      "loss": 3.1941,
+      "step": 49730
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7526649832725525,
+      "learning_rate": 5.061205468739861e-05,
+      "loss": 2.9326,
+      "step": 49731
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7116498351097107,
+      "learning_rate": 5.06034972860994e-05,
+      "loss": 2.7313,
+      "step": 49732
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7177486419677734,
+      "learning_rate": 5.0594940541662664e-05,
+      "loss": 2.9048,
+      "step": 49733
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7500801682472229,
+      "learning_rate": 5.058638445411077e-05,
+      "loss": 2.8435,
+      "step": 49734
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7371991872787476,
+      "learning_rate": 5.0577829023466446e-05,
+      "loss": 2.8987,
+      "step": 49735
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7661231756210327,
+      "learning_rate": 5.056927424975199e-05,
+      "loss": 2.7352,
+      "step": 49736
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7188680171966553,
+      "learning_rate": 5.056072013299016e-05,
+      "loss": 3.0446,
+      "step": 49737
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7141361832618713,
+      "learning_rate": 5.055216667320341e-05,
+      "loss": 3.023,
+      "step": 49738
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7152678966522217,
+      "learning_rate": 5.0543613870414144e-05,
+      "loss": 2.9097,
+      "step": 49739
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7155481576919556,
+      "learning_rate": 5.0535061724645086e-05,
+      "loss": 2.9771,
+      "step": 49740
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7778782844543457,
+      "learning_rate": 5.052651023591858e-05,
+      "loss": 3.011,
+      "step": 49741
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6794896125793457,
+      "learning_rate": 5.0517959404257215e-05,
+      "loss": 2.8614,
+      "step": 49742
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7529333233833313,
+      "learning_rate": 5.050940922968363e-05,
+      "loss": 3.0088,
+      "step": 49743
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7487618923187256,
+      "learning_rate": 5.0500859712220245e-05,
+      "loss": 2.8297,
+      "step": 49744
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.735675573348999,
+      "learning_rate": 5.0492310851889575e-05,
+      "loss": 2.8424,
+      "step": 49745
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7913566827774048,
+      "learning_rate": 5.048376264871406e-05,
+      "loss": 2.986,
+      "step": 49746
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.668773353099823,
+      "learning_rate": 5.0475215102716306e-05,
+      "loss": 2.8925,
+      "step": 49747
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7151376008987427,
+      "learning_rate": 5.046666821391887e-05,
+      "loss": 2.7141,
+      "step": 49748
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7672746777534485,
+      "learning_rate": 5.0458121982344144e-05,
+      "loss": 2.8902,
+      "step": 49749
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6590734124183655,
+      "learning_rate": 5.044957640801475e-05,
+      "loss": 2.8578,
+      "step": 49750
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7381081581115723,
+      "learning_rate": 5.044103149095309e-05,
+      "loss": 2.8413,
+      "step": 49751
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7295042872428894,
+      "learning_rate": 5.043248723118182e-05,
+      "loss": 2.9015,
+      "step": 49752
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7408165335655212,
+      "learning_rate": 5.042394362872332e-05,
+      "loss": 2.8261,
+      "step": 49753
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7612532377243042,
+      "learning_rate": 5.041540068360004e-05,
+      "loss": 2.8644,
+      "step": 49754
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7137171030044556,
+      "learning_rate": 5.0406858395834656e-05,
+      "loss": 3.0074,
+      "step": 49755
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6950783133506775,
+      "learning_rate": 5.03983167654495e-05,
+      "loss": 2.9107,
+      "step": 49756
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7441567778587341,
+      "learning_rate": 5.0389775792467134e-05,
+      "loss": 2.7601,
+      "step": 49757
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7424207925796509,
+      "learning_rate": 5.038123547691012e-05,
+      "loss": 2.8487,
+      "step": 49758
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6980047225952148,
+      "learning_rate": 5.03726958188009e-05,
+      "loss": 2.7768,
+      "step": 49759
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7469981908798218,
+      "learning_rate": 5.036415681816196e-05,
+      "loss": 2.9057,
+      "step": 49760
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7088938355445862,
+      "learning_rate": 5.0355618475015716e-05,
+      "loss": 2.9294,
+      "step": 49761
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7196033000946045,
+      "learning_rate": 5.034708078938474e-05,
+      "loss": 2.9308,
+      "step": 49762
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7394540309906006,
+      "learning_rate": 5.033854376129156e-05,
+      "loss": 2.7985,
+      "step": 49763
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7375257611274719,
+      "learning_rate": 5.033000739075852e-05,
+      "loss": 2.968,
+      "step": 49764
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7190216779708862,
+      "learning_rate": 5.0321471677808304e-05,
+      "loss": 2.9031,
+      "step": 49765
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.769286572933197,
+      "learning_rate": 5.031293662246325e-05,
+      "loss": 2.5801,
+      "step": 49766
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7621033191680908,
+      "learning_rate": 5.030440222474582e-05,
+      "loss": 2.91,
+      "step": 49767
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7350282669067383,
+      "learning_rate": 5.0295868484678595e-05,
+      "loss": 2.8721,
+      "step": 49768
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7469178438186646,
+      "learning_rate": 5.028733540228391e-05,
+      "loss": 2.9342,
+      "step": 49769
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8052636384963989,
+      "learning_rate": 5.027880297758443e-05,
+      "loss": 3.0468,
+      "step": 49770
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7492589354515076,
+      "learning_rate": 5.027027121060243e-05,
+      "loss": 2.9264,
+      "step": 49771
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7447794675827026,
+      "learning_rate": 5.026174010136046e-05,
+      "loss": 2.9722,
+      "step": 49772
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7605386972427368,
+      "learning_rate": 5.0253209649881096e-05,
+      "loss": 2.9998,
+      "step": 49773
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7599433660507202,
+      "learning_rate": 5.024467985618673e-05,
+      "loss": 2.7689,
+      "step": 49774
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7454763650894165,
+      "learning_rate": 5.023615072029977e-05,
+      "loss": 2.8439,
+      "step": 49775
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7649785876274109,
+      "learning_rate": 5.022762224224267e-05,
+      "loss": 2.7798,
+      "step": 49776
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7098384499549866,
+      "learning_rate": 5.021909442203794e-05,
+      "loss": 2.8681,
+      "step": 49777
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7371013760566711,
+      "learning_rate": 5.0210567259708135e-05,
+      "loss": 2.8648,
+      "step": 49778
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7675368785858154,
+      "learning_rate": 5.020204075527553e-05,
+      "loss": 2.9071,
+      "step": 49779
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7284151315689087,
+      "learning_rate": 5.0193514908762775e-05,
+      "loss": 2.9625,
+      "step": 49780
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7707260251045227,
+      "learning_rate": 5.0184989720192194e-05,
+      "loss": 2.9693,
+      "step": 49781
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7367619276046753,
+      "learning_rate": 5.017646518958619e-05,
+      "loss": 2.9056,
+      "step": 49782
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7275775074958801,
+      "learning_rate": 5.0167941316967417e-05,
+      "loss": 2.9327,
+      "step": 49783
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7671332359313965,
+      "learning_rate": 5.015941810235809e-05,
+      "loss": 2.7362,
+      "step": 49784
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7026022672653198,
+      "learning_rate": 5.015089554578087e-05,
+      "loss": 2.8569,
+      "step": 49785
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7949316501617432,
+      "learning_rate": 5.014237364725801e-05,
+      "loss": 2.9605,
+      "step": 49786
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7388706207275391,
+      "learning_rate": 5.013385240681216e-05,
+      "loss": 2.7446,
+      "step": 49787
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7240619659423828,
+      "learning_rate": 5.012533182446561e-05,
+      "loss": 2.7338,
+      "step": 49788
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7435773015022278,
+      "learning_rate": 5.011681190024082e-05,
+      "loss": 2.9415,
+      "step": 49789
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7612718939781189,
+      "learning_rate": 5.0108292634160294e-05,
+      "loss": 2.857,
+      "step": 49790
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7587111592292786,
+      "learning_rate": 5.0099774026246366e-05,
+      "loss": 3.0447,
+      "step": 49791
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7427108287811279,
+      "learning_rate": 5.009125607652163e-05,
+      "loss": 2.682,
+      "step": 49792
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.75941002368927,
+      "learning_rate": 5.008273878500836e-05,
+      "loss": 2.8607,
+      "step": 49793
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7439859509468079,
+      "learning_rate": 5.007422215172905e-05,
+      "loss": 2.8351,
+      "step": 49794
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.758802056312561,
+      "learning_rate": 5.006570617670619e-05,
+      "loss": 2.8682,
+      "step": 49795
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7315456867218018,
+      "learning_rate": 5.005719085996216e-05,
+      "loss": 3.1513,
+      "step": 49796
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7400312423706055,
+      "learning_rate": 5.004867620151941e-05,
+      "loss": 2.9039,
+      "step": 49797
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.766927182674408,
+      "learning_rate": 5.0040162201400256e-05,
+      "loss": 3.0392,
+      "step": 49798
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.756879985332489,
+      "learning_rate": 5.003164885962722e-05,
+      "loss": 2.956,
+      "step": 49799
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7322831153869629,
+      "learning_rate": 5.00231361762228e-05,
+      "loss": 2.8099,
+      "step": 49800
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7429162859916687,
+      "learning_rate": 5.0014624151209224e-05,
+      "loss": 2.9539,
+      "step": 49801
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7556778192520142,
+      "learning_rate": 5.000611278460911e-05,
+      "loss": 3.0639,
+      "step": 49802
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7347615957260132,
+      "learning_rate": 4.99976020764448e-05,
+      "loss": 2.9699,
+      "step": 49803
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7194696068763733,
+      "learning_rate": 4.998909202673858e-05,
+      "loss": 2.9908,
+      "step": 49804
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7030249238014221,
+      "learning_rate": 4.998058263551306e-05,
+      "loss": 2.8154,
+      "step": 49805
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7510625123977661,
+      "learning_rate": 4.997207390279052e-05,
+      "loss": 2.9402,
+      "step": 49806
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7080065011978149,
+      "learning_rate": 4.9963565828593455e-05,
+      "loss": 2.8348,
+      "step": 49807
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7457221150398254,
+      "learning_rate": 4.9955058412944205e-05,
+      "loss": 3.1089,
+      "step": 49808
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7274998426437378,
+      "learning_rate": 4.994655165586519e-05,
+      "loss": 2.8142,
+      "step": 49809
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7139880657196045,
+      "learning_rate": 4.993804555737888e-05,
+      "loss": 2.9572,
+      "step": 49810
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7105646133422852,
+      "learning_rate": 4.9929540117507674e-05,
+      "loss": 2.8571,
+      "step": 49811
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6922122240066528,
+      "learning_rate": 4.9921035336273904e-05,
+      "loss": 2.9435,
+      "step": 49812
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7425985932350159,
+      "learning_rate": 4.991253121369993e-05,
+      "loss": 2.8171,
+      "step": 49813
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7572745084762573,
+      "learning_rate": 4.990402774980824e-05,
+      "loss": 3.0096,
+      "step": 49814
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7443933486938477,
+      "learning_rate": 4.989552494462127e-05,
+      "loss": 2.6377,
+      "step": 49815
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7220988273620605,
+      "learning_rate": 4.9887022798161266e-05,
+      "loss": 3.0297,
+      "step": 49816
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7042877078056335,
+      "learning_rate": 4.987852131045079e-05,
+      "loss": 2.9561,
+      "step": 49817
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7358666062355042,
+      "learning_rate": 4.987002048151215e-05,
+      "loss": 3.0252,
+      "step": 49818
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7317989468574524,
+      "learning_rate": 4.986152031136766e-05,
+      "loss": 2.9775,
+      "step": 49819
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7053791284561157,
+      "learning_rate": 4.985302080003988e-05,
+      "loss": 2.7853,
+      "step": 49820
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7373185753822327,
+      "learning_rate": 4.984452194755101e-05,
+      "loss": 2.8829,
+      "step": 49821
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7272002100944519,
+      "learning_rate": 4.98360237539236e-05,
+      "loss": 3.0424,
+      "step": 49822
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7341743111610413,
+      "learning_rate": 4.9827526219179914e-05,
+      "loss": 2.8273,
+      "step": 49823
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7472624778747559,
+      "learning_rate": 4.981902934334242e-05,
+      "loss": 2.8973,
+      "step": 49824
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7138078808784485,
+      "learning_rate": 4.9810533126433475e-05,
+      "loss": 2.9605,
+      "step": 49825
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7111098766326904,
+      "learning_rate": 4.980203756847532e-05,
+      "loss": 3.0122,
+      "step": 49826
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7603220343589783,
+      "learning_rate": 4.979354266949057e-05,
+      "loss": 2.9321,
+      "step": 49827
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7600682973861694,
+      "learning_rate": 4.978504842950138e-05,
+      "loss": 2.6835,
+      "step": 49828
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7622936367988586,
+      "learning_rate": 4.9776554848530205e-05,
+      "loss": 2.6772,
+      "step": 49829
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7368950843811035,
+      "learning_rate": 4.976806192659951e-05,
+      "loss": 2.8547,
+      "step": 49830
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7676935195922852,
+      "learning_rate": 4.975956966373153e-05,
+      "loss": 2.8766,
+      "step": 49831
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7402030229568481,
+      "learning_rate": 4.975107805994872e-05,
+      "loss": 2.8607,
+      "step": 49832
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.745424211025238,
+      "learning_rate": 4.974258711527346e-05,
+      "loss": 2.6656,
+      "step": 49833
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.756466269493103,
+      "learning_rate": 4.973409682972794e-05,
+      "loss": 2.926,
+      "step": 49834
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7675303816795349,
+      "learning_rate": 4.972560720333473e-05,
+      "loss": 2.8321,
+      "step": 49835
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.758357584476471,
+      "learning_rate": 4.971711823611606e-05,
+      "loss": 3.0498,
+      "step": 49836
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8041175603866577,
+      "learning_rate": 4.970862992809436e-05,
+      "loss": 2.9132,
+      "step": 49837
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7109980583190918,
+      "learning_rate": 4.9700142279291933e-05,
+      "loss": 2.7802,
+      "step": 49838
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.785557746887207,
+      "learning_rate": 4.9691655289731205e-05,
+      "loss": 3.0044,
+      "step": 49839
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.726588249206543,
+      "learning_rate": 4.968316895943448e-05,
+      "loss": 2.7941,
+      "step": 49840
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7485399842262268,
+      "learning_rate": 4.967468328842406e-05,
+      "loss": 2.9152,
+      "step": 49841
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.740860104560852,
+      "learning_rate": 4.96661982767224e-05,
+      "loss": 2.8871,
+      "step": 49842
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.790346622467041,
+      "learning_rate": 4.965771392435174e-05,
+      "loss": 3.1134,
+      "step": 49843
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7316777110099792,
+      "learning_rate": 4.964923023133448e-05,
+      "loss": 3.0394,
+      "step": 49844
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7692133784294128,
+      "learning_rate": 4.9640747197693007e-05,
+      "loss": 2.8604,
+      "step": 49845
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7087564468383789,
+      "learning_rate": 4.963226482344968e-05,
+      "loss": 2.8347,
+      "step": 49846
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7064667344093323,
+      "learning_rate": 4.962378310862672e-05,
+      "loss": 3.0958,
+      "step": 49847
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7500705122947693,
+      "learning_rate": 4.961530205324651e-05,
+      "loss": 2.8577,
+      "step": 49848
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7233101725578308,
+      "learning_rate": 4.960682165733137e-05,
+      "loss": 2.8902,
+      "step": 49849
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7733365893363953,
+      "learning_rate": 4.959834192090377e-05,
+      "loss": 2.8184,
+      "step": 49850
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7586425542831421,
+      "learning_rate": 4.9589862843985886e-05,
+      "loss": 2.9315,
+      "step": 49851
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6858609914779663,
+      "learning_rate": 4.958138442660016e-05,
+      "loss": 2.6727,
+      "step": 49852
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6984794735908508,
+      "learning_rate": 4.957290666876881e-05,
+      "loss": 2.9782,
+      "step": 49853
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7984606027603149,
+      "learning_rate": 4.956442957051429e-05,
+      "loss": 3.0055,
+      "step": 49854
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7043806314468384,
+      "learning_rate": 4.9555953131858875e-05,
+      "loss": 2.9892,
+      "step": 49855
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.732420802116394,
+      "learning_rate": 4.9547477352824815e-05,
+      "loss": 2.8946,
+      "step": 49856
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7292214632034302,
+      "learning_rate": 4.9539002233434535e-05,
+      "loss": 2.8447,
+      "step": 49857
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7352430820465088,
+      "learning_rate": 4.95305277737103e-05,
+      "loss": 2.9061,
+      "step": 49858
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7342036962509155,
+      "learning_rate": 4.952205397367441e-05,
+      "loss": 3.0595,
+      "step": 49859
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6796708106994629,
+      "learning_rate": 4.951358083334931e-05,
+      "loss": 3.1384,
+      "step": 49860
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7156011462211609,
+      "learning_rate": 4.950510835275724e-05,
+      "loss": 2.8325,
+      "step": 49861
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7859975099563599,
+      "learning_rate": 4.9496636531920484e-05,
+      "loss": 2.9611,
+      "step": 49862
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7750837206840515,
+      "learning_rate": 4.948816537086133e-05,
+      "loss": 3.0251,
+      "step": 49863
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7656334638595581,
+      "learning_rate": 4.9479694869602123e-05,
+      "loss": 2.9598,
+      "step": 49864
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.720266580581665,
+      "learning_rate": 4.9471225028165285e-05,
+      "loss": 2.8096,
+      "step": 49865
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7152171730995178,
+      "learning_rate": 4.946275584657292e-05,
+      "loss": 3.0708,
+      "step": 49866
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7371180653572083,
+      "learning_rate": 4.945428732484752e-05,
+      "loss": 2.7161,
+      "step": 49867
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7790018916130066,
+      "learning_rate": 4.9445819463011294e-05,
+      "loss": 2.8841,
+      "step": 49868
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7312780022621155,
+      "learning_rate": 4.9437352261086594e-05,
+      "loss": 3.1128,
+      "step": 49869
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7291393876075745,
+      "learning_rate": 4.94288857190956e-05,
+      "loss": 3.0187,
+      "step": 49870
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7577651739120483,
+      "learning_rate": 4.94204198370607e-05,
+      "loss": 2.9457,
+      "step": 49871
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7718584537506104,
+      "learning_rate": 4.941195461500424e-05,
+      "loss": 2.8328,
+      "step": 49872
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7502276301383972,
+      "learning_rate": 4.940349005294839e-05,
+      "loss": 2.9141,
+      "step": 49873
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7148169279098511,
+      "learning_rate": 4.939502615091562e-05,
+      "loss": 2.9375,
+      "step": 49874
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7954051494598389,
+      "learning_rate": 4.938656290892803e-05,
+      "loss": 2.9624,
+      "step": 49875
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7510395646095276,
+      "learning_rate": 4.937810032700811e-05,
+      "loss": 2.9211,
+      "step": 49876
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7023428678512573,
+      "learning_rate": 4.9369638405178e-05,
+      "loss": 3.2698,
+      "step": 49877
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7281398773193359,
+      "learning_rate": 4.936117714345996e-05,
+      "loss": 2.9131,
+      "step": 49878
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7942187190055847,
+      "learning_rate": 4.9352716541876436e-05,
+      "loss": 2.9639,
+      "step": 49879
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.70033860206604,
+      "learning_rate": 4.934425660044951e-05,
+      "loss": 2.9204,
+      "step": 49880
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7631633281707764,
+      "learning_rate": 4.933579731920162e-05,
+      "loss": 3.0097,
+      "step": 49881
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7145242691040039,
+      "learning_rate": 4.932733869815503e-05,
+      "loss": 2.8329,
+      "step": 49882
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7550216913223267,
+      "learning_rate": 4.9318880737332015e-05,
+      "loss": 2.8755,
+      "step": 49883
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7773282527923584,
+      "learning_rate": 4.931042343675482e-05,
+      "loss": 2.9673,
+      "step": 49884
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7082733511924744,
+      "learning_rate": 4.9301966796445636e-05,
+      "loss": 2.9873,
+      "step": 49885
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6855611801147461,
+      "learning_rate": 4.9293510816426807e-05,
+      "loss": 2.7685,
+      "step": 49886
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.696873664855957,
+      "learning_rate": 4.928505549672072e-05,
+      "loss": 3.0811,
+      "step": 49887
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7408643364906311,
+      "learning_rate": 4.927660083734948e-05,
+      "loss": 2.7622,
+      "step": 49888
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7321268916130066,
+      "learning_rate": 4.926814683833547e-05,
+      "loss": 2.9631,
+      "step": 49889
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7376499176025391,
+      "learning_rate": 4.925969349970093e-05,
+      "loss": 3.2182,
+      "step": 49890
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6759738922119141,
+      "learning_rate": 4.925124082146804e-05,
+      "loss": 2.8273,
+      "step": 49891
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7222961783409119,
+      "learning_rate": 4.924278880365917e-05,
+      "loss": 2.7853,
+      "step": 49892
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7377529740333557,
+      "learning_rate": 4.923433744629647e-05,
+      "loss": 3.1511,
+      "step": 49893
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8030790686607361,
+      "learning_rate": 4.922588674940231e-05,
+      "loss": 2.9492,
+      "step": 49894
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7189480066299438,
+      "learning_rate": 4.921743671299888e-05,
+      "loss": 2.8765,
+      "step": 49895
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7487554550170898,
+      "learning_rate": 4.920898733710843e-05,
+      "loss": 2.9909,
+      "step": 49896
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7432091236114502,
+      "learning_rate": 4.92005386217533e-05,
+      "loss": 2.8594,
+      "step": 49897
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7807390093803406,
+      "learning_rate": 4.919209056695572e-05,
+      "loss": 3.0344,
+      "step": 49898
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7431802749633789,
+      "learning_rate": 4.918364317273786e-05,
+      "loss": 2.9692,
+      "step": 49899
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7242401242256165,
+      "learning_rate": 4.917519643912198e-05,
+      "loss": 2.6837,
+      "step": 49900
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7620343565940857,
+      "learning_rate": 4.9166750366130345e-05,
+      "loss": 2.7063,
+      "step": 49901
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7530914545059204,
+      "learning_rate": 4.9158304953785286e-05,
+      "loss": 2.923,
+      "step": 49902
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7380638122558594,
+      "learning_rate": 4.914986020210892e-05,
+      "loss": 3.0402,
+      "step": 49903
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7419988512992859,
+      "learning_rate": 4.914141611112361e-05,
+      "loss": 2.7586,
+      "step": 49904
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7261704802513123,
+      "learning_rate": 4.913297268085152e-05,
+      "loss": 3.0975,
+      "step": 49905
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7934512495994568,
+      "learning_rate": 4.912452991131483e-05,
+      "loss": 2.8156,
+      "step": 49906
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7128746509552002,
+      "learning_rate": 4.911608780253593e-05,
+      "loss": 2.8872,
+      "step": 49907
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7485001683235168,
+      "learning_rate": 4.910764635453689e-05,
+      "loss": 2.9185,
+      "step": 49908
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7733572721481323,
+      "learning_rate": 4.909920556734009e-05,
+      "loss": 2.8178,
+      "step": 49909
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7357202172279358,
+      "learning_rate": 4.909076544096763e-05,
+      "loss": 2.7962,
+      "step": 49910
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.1617584228515625,
+      "learning_rate": 4.9082325975441805e-05,
+      "loss": 3.1216,
+      "step": 49911
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7278109788894653,
+      "learning_rate": 4.907388717078498e-05,
+      "loss": 3.0274,
+      "step": 49912
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7125959992408752,
+      "learning_rate": 4.9065449027019125e-05,
+      "loss": 2.8246,
+      "step": 49913
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8183368444442749,
+      "learning_rate": 4.9057011544166625e-05,
+      "loss": 2.6864,
+      "step": 49914
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7317843437194824,
+      "learning_rate": 4.90485747222496e-05,
+      "loss": 2.7996,
+      "step": 49915
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8040397763252258,
+      "learning_rate": 4.904013856129036e-05,
+      "loss": 3.1255,
+      "step": 49916
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6812474131584167,
+      "learning_rate": 4.903170306131113e-05,
+      "loss": 2.9811,
+      "step": 49917
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7134155631065369,
+      "learning_rate": 4.9023268222334054e-05,
+      "loss": 2.9236,
+      "step": 49918
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.730156421661377,
+      "learning_rate": 4.901483404438145e-05,
+      "loss": 3.0052,
+      "step": 49919
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7486832737922668,
+      "learning_rate": 4.9006400527475455e-05,
+      "loss": 2.9662,
+      "step": 49920
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7389925122261047,
+      "learning_rate": 4.8997967671638225e-05,
+      "loss": 2.913,
+      "step": 49921
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7639115452766418,
+      "learning_rate": 4.898953547689212e-05,
+      "loss": 2.9612,
+      "step": 49922
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8156792521476746,
+      "learning_rate": 4.89811039432592e-05,
+      "loss": 3.0135,
+      "step": 49923
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7284063696861267,
+      "learning_rate": 4.8972673070761836e-05,
+      "loss": 2.9221,
+      "step": 49924
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7258008122444153,
+      "learning_rate": 4.8964242859422033e-05,
+      "loss": 2.9088,
+      "step": 49925
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7349685430526733,
+      "learning_rate": 4.8955813309262204e-05,
+      "loss": 3.1959,
+      "step": 49926
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7146064043045044,
+      "learning_rate": 4.894738442030444e-05,
+      "loss": 2.7013,
+      "step": 49927
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.75479656457901,
+      "learning_rate": 4.8938956192570847e-05,
+      "loss": 3.0321,
+      "step": 49928
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7491357922554016,
+      "learning_rate": 4.8930528626083846e-05,
+      "loss": 2.6871,
+      "step": 49929
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7528855800628662,
+      "learning_rate": 4.8922101720865384e-05,
+      "loss": 2.7284,
+      "step": 49930
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7925986647605896,
+      "learning_rate": 4.891367547693784e-05,
+      "loss": 2.9621,
+      "step": 49931
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7412816882133484,
+      "learning_rate": 4.890524989432341e-05,
+      "loss": 2.8603,
+      "step": 49932
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7685806751251221,
+      "learning_rate": 4.889682497304412e-05,
+      "loss": 2.7904,
+      "step": 49933
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7178759574890137,
+      "learning_rate": 4.8888400713122376e-05,
+      "loss": 2.9155,
+      "step": 49934
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7180420160293579,
+      "learning_rate": 4.887997711458025e-05,
+      "loss": 2.8599,
+      "step": 49935
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7158687710762024,
+      "learning_rate": 4.887155417743985e-05,
+      "loss": 2.684,
+      "step": 49936
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7684953212738037,
+      "learning_rate": 4.886313190172354e-05,
+      "loss": 2.7244,
+      "step": 49937
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7468008399009705,
+      "learning_rate": 4.885471028745332e-05,
+      "loss": 2.9189,
+      "step": 49938
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7421904802322388,
+      "learning_rate": 4.8846289334651554e-05,
+      "loss": 2.9611,
+      "step": 49939
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7222840785980225,
+      "learning_rate": 4.883786904334021e-05,
+      "loss": 2.7801,
+      "step": 49940
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7169114351272583,
+      "learning_rate": 4.882944941354169e-05,
+      "loss": 3.0289,
+      "step": 49941
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7277195453643799,
+      "learning_rate": 4.882103044527805e-05,
+      "loss": 2.8314,
+      "step": 49942
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7513947486877441,
+      "learning_rate": 4.8812612138571426e-05,
+      "loss": 2.7394,
+      "step": 49943
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7872393727302551,
+      "learning_rate": 4.880419449344406e-05,
+      "loss": 2.8735,
+      "step": 49944
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8093016743659973,
+      "learning_rate": 4.8795777509918064e-05,
+      "loss": 2.8194,
+      "step": 49945
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7250414490699768,
+      "learning_rate": 4.878736118801566e-05,
+      "loss": 2.9011,
+      "step": 49946
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7930007576942444,
+      "learning_rate": 4.877894552775906e-05,
+      "loss": 2.9038,
+      "step": 49947
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8064557909965515,
+      "learning_rate": 4.877053052917035e-05,
+      "loss": 2.7648,
+      "step": 49948
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7449289560317993,
+      "learning_rate": 4.876211619227174e-05,
+      "loss": 2.8408,
+      "step": 49949
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7603309154510498,
+      "learning_rate": 4.875370251708528e-05,
+      "loss": 2.8804,
+      "step": 49950
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7044733762741089,
+      "learning_rate": 4.874528950363325e-05,
+      "loss": 2.9399,
+      "step": 49951
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7044036388397217,
+      "learning_rate": 4.873687715193775e-05,
+      "loss": 2.7616,
+      "step": 49952
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7298387885093689,
+      "learning_rate": 4.8728465462020935e-05,
+      "loss": 2.9534,
+      "step": 49953
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7330809831619263,
+      "learning_rate": 4.872005443390508e-05,
+      "loss": 3.0583,
+      "step": 49954
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7165080904960632,
+      "learning_rate": 4.8711644067612145e-05,
+      "loss": 2.9493,
+      "step": 49955
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7864888310432434,
+      "learning_rate": 4.870323436316446e-05,
+      "loss": 2.9212,
+      "step": 49956
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7288803458213806,
+      "learning_rate": 4.8694825320584097e-05,
+      "loss": 2.8663,
+      "step": 49957
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7244741320610046,
+      "learning_rate": 4.8686416939893114e-05,
+      "loss": 2.874,
+      "step": 49958
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7267864346504211,
+      "learning_rate": 4.8678009221113826e-05,
+      "loss": 2.8159,
+      "step": 49959
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7464987635612488,
+      "learning_rate": 4.866960216426821e-05,
+      "loss": 2.8817,
+      "step": 49960
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7408061027526855,
+      "learning_rate": 4.866119576937858e-05,
+      "loss": 2.9639,
+      "step": 49961
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7390897870063782,
+      "learning_rate": 4.8652790036466925e-05,
+      "loss": 3.0278,
+      "step": 49962
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7427120804786682,
+      "learning_rate": 4.86443849655555e-05,
+      "loss": 3.0713,
+      "step": 49963
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7576953172683716,
+      "learning_rate": 4.86359805566664e-05,
+      "loss": 2.905,
+      "step": 49964
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7212951183319092,
+      "learning_rate": 4.8627576809821654e-05,
+      "loss": 3.0014,
+      "step": 49965
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.761062741279602,
+      "learning_rate": 4.861917372504361e-05,
+      "loss": 2.7335,
+      "step": 49966
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7463256120681763,
+      "learning_rate": 4.861077130235419e-05,
+      "loss": 2.7925,
+      "step": 49967
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7250059843063354,
+      "learning_rate": 4.86023695417756e-05,
+      "loss": 2.9038,
+      "step": 49968
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7397086024284363,
+      "learning_rate": 4.859396844333009e-05,
+      "loss": 3.0888,
+      "step": 49969
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7244906425476074,
+      "learning_rate": 4.858556800703964e-05,
+      "loss": 3.0914,
+      "step": 49970
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7587701082229614,
+      "learning_rate": 4.857716823292644e-05,
+      "loss": 2.8498,
+      "step": 49971
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7259935736656189,
+      "learning_rate": 4.856876912101254e-05,
+      "loss": 2.7492,
+      "step": 49972
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7055765390396118,
+      "learning_rate": 4.856037067132012e-05,
+      "loss": 3.1046,
+      "step": 49973
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7304884195327759,
+      "learning_rate": 4.8551972883871313e-05,
+      "loss": 2.8807,
+      "step": 49974
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7568566203117371,
+      "learning_rate": 4.854357575868819e-05,
+      "loss": 2.9917,
+      "step": 49975
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7316765189170837,
+      "learning_rate": 4.8535179295792957e-05,
+      "loss": 3.0772,
+      "step": 49976
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7427469491958618,
+      "learning_rate": 4.8526783495207565e-05,
+      "loss": 2.8457,
+      "step": 49977
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7722180485725403,
+      "learning_rate": 4.851838835695433e-05,
+      "loss": 3.0185,
+      "step": 49978
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7619324922561646,
+      "learning_rate": 4.850999388105527e-05,
+      "loss": 2.7384,
+      "step": 49979
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7218577861785889,
+      "learning_rate": 4.850160006753239e-05,
+      "loss": 2.9204,
+      "step": 49980
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7448925375938416,
+      "learning_rate": 4.8493206916407976e-05,
+      "loss": 2.8308,
+      "step": 49981
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7064295411109924,
+      "learning_rate": 4.8484814427703986e-05,
+      "loss": 3.0217,
+      "step": 49982
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7662560939788818,
+      "learning_rate": 4.8476422601442564e-05,
+      "loss": 2.6942,
+      "step": 49983
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7750674486160278,
+      "learning_rate": 4.846803143764593e-05,
+      "loss": 2.761,
+      "step": 49984
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7387879490852356,
+      "learning_rate": 4.845964093633612e-05,
+      "loss": 2.8892,
+      "step": 49985
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7031064033508301,
+      "learning_rate": 4.845125109753516e-05,
+      "loss": 3.0094,
+      "step": 49986
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7086855173110962,
+      "learning_rate": 4.844286192126513e-05,
+      "loss": 2.8456,
+      "step": 49987
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7467719912528992,
+      "learning_rate": 4.843447340754821e-05,
+      "loss": 2.914,
+      "step": 49988
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7186830639839172,
+      "learning_rate": 4.8426085556406523e-05,
+      "loss": 3.1394,
+      "step": 49989
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7500671744346619,
+      "learning_rate": 4.841769836786205e-05,
+      "loss": 2.966,
+      "step": 49990
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7453409433364868,
+      "learning_rate": 4.840931184193703e-05,
+      "loss": 2.7501,
+      "step": 49991
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7155488133430481,
+      "learning_rate": 4.840092597865342e-05,
+      "loss": 2.8779,
+      "step": 49992
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7144157290458679,
+      "learning_rate": 4.8392540778033295e-05,
+      "loss": 3.0127,
+      "step": 49993
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7175836563110352,
+      "learning_rate": 4.838415624009888e-05,
+      "loss": 2.9094,
+      "step": 49994
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7162435054779053,
+      "learning_rate": 4.837577236487209e-05,
+      "loss": 2.8378,
+      "step": 49995
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7388350963592529,
+      "learning_rate": 4.8367389152375166e-05,
+      "loss": 2.8836,
+      "step": 49996
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7505469918251038,
+      "learning_rate": 4.835900660263003e-05,
+      "loss": 2.8015,
+      "step": 49997
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.734264612197876,
+      "learning_rate": 4.835062471565887e-05,
+      "loss": 2.9549,
+      "step": 49998
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.753314733505249,
+      "learning_rate": 4.834224349148379e-05,
+      "loss": 2.9851,
+      "step": 49999
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7283154726028442,
+      "learning_rate": 4.833386293012681e-05,
+      "loss": 2.9575,
+      "step": 50000
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6950429081916809,
+      "learning_rate": 4.832548303160998e-05,
+      "loss": 3.0719,
+      "step": 50001
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7324820756912231,
+      "learning_rate": 4.831710379595535e-05,
+      "loss": 2.906,
+      "step": 50002
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7295659780502319,
+      "learning_rate": 4.8308725223185e-05,
+      "loss": 2.8802,
+      "step": 50003
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7294290661811829,
+      "learning_rate": 4.830034731332111e-05,
+      "loss": 3.0532,
+      "step": 50004
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.764670729637146,
+      "learning_rate": 4.82919700663856e-05,
+      "loss": 3.0021,
+      "step": 50005
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7712165713310242,
+      "learning_rate": 4.828359348240069e-05,
+      "loss": 2.8191,
+      "step": 50006
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.749573290348053,
+      "learning_rate": 4.827521756138831e-05,
+      "loss": 3.0938,
+      "step": 50007
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7555610537528992,
+      "learning_rate": 4.8266842303370534e-05,
+      "loss": 2.7916,
+      "step": 50008
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.716448187828064,
+      "learning_rate": 4.8258467708369495e-05,
+      "loss": 3.0179,
+      "step": 50009
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7162739038467407,
+      "learning_rate": 4.825009377640712e-05,
+      "loss": 2.8407,
+      "step": 50010
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7964804768562317,
+      "learning_rate": 4.824172050750564e-05,
+      "loss": 2.8375,
+      "step": 50011
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7051910161972046,
+      "learning_rate": 4.823334790168693e-05,
+      "loss": 2.7792,
+      "step": 50012
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7237557768821716,
+      "learning_rate": 4.822497595897315e-05,
+      "loss": 2.9245,
+      "step": 50013
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7709342837333679,
+      "learning_rate": 4.821660467938645e-05,
+      "loss": 2.8221,
+      "step": 50014
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7040632367134094,
+      "learning_rate": 4.8208234062948626e-05,
+      "loss": 2.7416,
+      "step": 50015
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6828323602676392,
+      "learning_rate": 4.819986410968192e-05,
+      "loss": 2.9322,
+      "step": 50016
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7632225751876831,
+      "learning_rate": 4.819149481960825e-05,
+      "loss": 2.7827,
+      "step": 50017
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7432723641395569,
+      "learning_rate": 4.8183126192749686e-05,
+      "loss": 2.7461,
+      "step": 50018
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7697473764419556,
+      "learning_rate": 4.81747582291284e-05,
+      "loss": 2.8574,
+      "step": 50019
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6835653185844421,
+      "learning_rate": 4.8166390928766286e-05,
+      "loss": 2.9083,
+      "step": 50020
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6899628639221191,
+      "learning_rate": 4.815802429168547e-05,
+      "loss": 2.7771,
+      "step": 50021
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7158491015434265,
+      "learning_rate": 4.814965831790797e-05,
+      "loss": 2.8758,
+      "step": 50022
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7541047930717468,
+      "learning_rate": 4.8141293007455727e-05,
+      "loss": 2.8541,
+      "step": 50023
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7383518815040588,
+      "learning_rate": 4.8132928360350934e-05,
+      "loss": 2.932,
+      "step": 50024
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.736765444278717,
+      "learning_rate": 4.8124564376615424e-05,
+      "loss": 2.6686,
+      "step": 50025
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7256408929824829,
+      "learning_rate": 4.811620105627143e-05,
+      "loss": 3.1262,
+      "step": 50026
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7852066159248352,
+      "learning_rate": 4.810783839934082e-05,
+      "loss": 2.8989,
+      "step": 50027
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7440248131752014,
+      "learning_rate": 4.809947640584576e-05,
+      "loss": 2.8755,
+      "step": 50028
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7258433699607849,
+      "learning_rate": 4.809111507580819e-05,
+      "loss": 2.9438,
+      "step": 50029
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7557896375656128,
+      "learning_rate": 4.808275440925005e-05,
+      "loss": 2.8328,
+      "step": 50030
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7530323266983032,
+      "learning_rate": 4.8074394406193564e-05,
+      "loss": 2.973,
+      "step": 50031
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7720298767089844,
+      "learning_rate": 4.8066035066660544e-05,
+      "loss": 2.9985,
+      "step": 50032
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7010929584503174,
+      "learning_rate": 4.8057676390673184e-05,
+      "loss": 2.9103,
+      "step": 50033
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8532752990722656,
+      "learning_rate": 4.804931837825332e-05,
+      "loss": 2.9676,
+      "step": 50034
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7763757705688477,
+      "learning_rate": 4.804096102942311e-05,
+      "loss": 2.7373,
+      "step": 50035
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7673393487930298,
+      "learning_rate": 4.803260434420456e-05,
+      "loss": 2.762,
+      "step": 50036
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.722378134727478,
+      "learning_rate": 4.802424832261963e-05,
+      "loss": 3.0049,
+      "step": 50037
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7699034810066223,
+      "learning_rate": 4.8015892964690364e-05,
+      "loss": 3.0195,
+      "step": 50038
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8004751801490784,
+      "learning_rate": 4.800753827043864e-05,
+      "loss": 2.8241,
+      "step": 50039
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7232542037963867,
+      "learning_rate": 4.799918423988658e-05,
+      "loss": 2.8742,
+      "step": 50040
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.778252363204956,
+      "learning_rate": 4.799083087305623e-05,
+      "loss": 2.7134,
+      "step": 50041
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7834492325782776,
+      "learning_rate": 4.798247816996948e-05,
+      "loss": 2.9848,
+      "step": 50042
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7058895230293274,
+      "learning_rate": 4.7974126130648425e-05,
+      "loss": 3.0141,
+      "step": 50043
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7730516791343689,
+      "learning_rate": 4.7965774755115035e-05,
+      "loss": 2.8361,
+      "step": 50044
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7547017931938171,
+      "learning_rate": 4.795742404339122e-05,
+      "loss": 2.7368,
+      "step": 50045
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.747089684009552,
+      "learning_rate": 4.79490739954991e-05,
+      "loss": 2.7331,
+      "step": 50046
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7834062576293945,
+      "learning_rate": 4.794072461146055e-05,
+      "loss": 2.8744,
+      "step": 50047
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7755687236785889,
+      "learning_rate": 4.793237589129772e-05,
+      "loss": 2.6205,
+      "step": 50048
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7240215539932251,
+      "learning_rate": 4.7924027835032385e-05,
+      "loss": 3.0052,
+      "step": 50049
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7895364165306091,
+      "learning_rate": 4.7915680442686745e-05,
+      "loss": 3.0121,
+      "step": 50050
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7438759207725525,
+      "learning_rate": 4.790733371428267e-05,
+      "loss": 2.6353,
+      "step": 50051
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7612919211387634,
+      "learning_rate": 4.789898764984211e-05,
+      "loss": 2.974,
+      "step": 50052
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7200536727905273,
+      "learning_rate": 4.7890642249387154e-05,
+      "loss": 2.8291,
+      "step": 50053
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.744355320930481,
+      "learning_rate": 4.788229751293965e-05,
+      "loss": 2.8287,
+      "step": 50054
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7230464816093445,
+      "learning_rate": 4.787395344052167e-05,
+      "loss": 2.8453,
+      "step": 50055
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7152355909347534,
+      "learning_rate": 4.7865610032155246e-05,
+      "loss": 2.9215,
+      "step": 50056
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7902154922485352,
+      "learning_rate": 4.785726728786221e-05,
+      "loss": 2.7708,
+      "step": 50057
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7618465423583984,
+      "learning_rate": 4.784892520766466e-05,
+      "loss": 2.8228,
+      "step": 50058
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8009930849075317,
+      "learning_rate": 4.784058379158453e-05,
+      "loss": 2.6846,
+      "step": 50059
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7363435626029968,
+      "learning_rate": 4.7832243039643714e-05,
+      "loss": 2.7056,
+      "step": 50060
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7347774505615234,
+      "learning_rate": 4.782390295186429e-05,
+      "loss": 2.8223,
+      "step": 50061
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7615084648132324,
+      "learning_rate": 4.7815563528268105e-05,
+      "loss": 3.0836,
+      "step": 50062
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7269819378852844,
+      "learning_rate": 4.780722476887725e-05,
+      "loss": 2.7853,
+      "step": 50063
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7375494837760925,
+      "learning_rate": 4.779888667371356e-05,
+      "loss": 2.589,
+      "step": 50064
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7187495827674866,
+      "learning_rate": 4.779054924279917e-05,
+      "loss": 2.8879,
+      "step": 50065
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7238545417785645,
+      "learning_rate": 4.778221247615591e-05,
+      "loss": 2.8817,
+      "step": 50066
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7386276125907898,
+      "learning_rate": 4.7773876373805676e-05,
+      "loss": 2.8928,
+      "step": 50067
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.720185399055481,
+      "learning_rate": 4.776554093577061e-05,
+      "loss": 2.9978,
+      "step": 50068
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7330771684646606,
+      "learning_rate": 4.775720616207249e-05,
+      "loss": 2.6761,
+      "step": 50069
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6842424273490906,
+      "learning_rate": 4.774887205273335e-05,
+      "loss": 2.7894,
+      "step": 50070
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7490212321281433,
+      "learning_rate": 4.774053860777518e-05,
+      "loss": 2.9307,
+      "step": 50071
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7413821220397949,
+      "learning_rate": 4.773220582721988e-05,
+      "loss": 2.7725,
+      "step": 50072
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6934046149253845,
+      "learning_rate": 4.7723873711089436e-05,
+      "loss": 2.8426,
+      "step": 50073
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7559138536453247,
+      "learning_rate": 4.7715542259405645e-05,
+      "loss": 2.9685,
+      "step": 50074
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7740827798843384,
+      "learning_rate": 4.770721147219061e-05,
+      "loss": 3.0002,
+      "step": 50075
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7439709305763245,
+      "learning_rate": 4.7698881349466265e-05,
+      "loss": 3.0176,
+      "step": 50076
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7715969085693359,
+      "learning_rate": 4.7690551891254436e-05,
+      "loss": 2.862,
+      "step": 50077
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7283157110214233,
+      "learning_rate": 4.76822230975772e-05,
+      "loss": 3.0025,
+      "step": 50078
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7790409326553345,
+      "learning_rate": 4.7673894968456375e-05,
+      "loss": 2.9668,
+      "step": 50079
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7363762259483337,
+      "learning_rate": 4.766556750391404e-05,
+      "loss": 2.8291,
+      "step": 50080
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.723741888999939,
+      "learning_rate": 4.765724070397201e-05,
+      "loss": 3.0666,
+      "step": 50081
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7611827254295349,
+      "learning_rate": 4.7648914568652176e-05,
+      "loss": 2.9559,
+      "step": 50082
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7597978115081787,
+      "learning_rate": 4.764058909797661e-05,
+      "loss": 2.7526,
+      "step": 50083
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7658380270004272,
+      "learning_rate": 4.763226429196707e-05,
+      "loss": 2.6695,
+      "step": 50084
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7255105972290039,
+      "learning_rate": 4.76239401506456e-05,
+      "loss": 2.7215,
+      "step": 50085
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.715186595916748,
+      "learning_rate": 4.761561667403418e-05,
+      "loss": 2.8727,
+      "step": 50086
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7413427233695984,
+      "learning_rate": 4.760729386215464e-05,
+      "loss": 2.8856,
+      "step": 50087
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7388500571250916,
+      "learning_rate": 4.759897171502891e-05,
+      "loss": 3.0116,
+      "step": 50088
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6811955571174622,
+      "learning_rate": 4.759065023267885e-05,
+      "loss": 3.0865,
+      "step": 50089
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7523310780525208,
+      "learning_rate": 4.758232941512644e-05,
+      "loss": 2.9596,
+      "step": 50090
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7621850371360779,
+      "learning_rate": 4.757400926239366e-05,
+      "loss": 2.8558,
+      "step": 50091
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.697986900806427,
+      "learning_rate": 4.75656897745023e-05,
+      "loss": 2.8994,
+      "step": 50092
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.722909688949585,
+      "learning_rate": 4.7557370951474404e-05,
+      "loss": 2.8762,
+      "step": 50093
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7483739852905273,
+      "learning_rate": 4.754905279333182e-05,
+      "loss": 2.8685,
+      "step": 50094
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7431113123893738,
+      "learning_rate": 4.754073530009633e-05,
+      "loss": 2.8286,
+      "step": 50095
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.750974178314209,
+      "learning_rate": 4.7532418471790055e-05,
+      "loss": 3.0905,
+      "step": 50096
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7384655475616455,
+      "learning_rate": 4.752410230843474e-05,
+      "loss": 2.9023,
+      "step": 50097
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7771185636520386,
+      "learning_rate": 4.75157868100524e-05,
+      "loss": 2.6056,
+      "step": 50098
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7554793357849121,
+      "learning_rate": 4.750747197666482e-05,
+      "loss": 2.7112,
+      "step": 50099
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7633035182952881,
+      "learning_rate": 4.749915780829399e-05,
+      "loss": 2.9715,
+      "step": 50100
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7684073448181152,
+      "learning_rate": 4.749084430496184e-05,
+      "loss": 3.0109,
+      "step": 50101
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.8110712170600891,
+      "learning_rate": 4.74825314666902e-05,
+      "loss": 2.9253,
+      "step": 50102
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6965272426605225,
+      "learning_rate": 4.747421929350098e-05,
+      "loss": 2.757,
+      "step": 50103
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.834354817867279,
+      "learning_rate": 4.7465907785416005e-05,
+      "loss": 2.9619,
+      "step": 50104
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7369504570960999,
+      "learning_rate": 4.7457596942457274e-05,
+      "loss": 3.0316,
+      "step": 50105
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.727476954460144,
+      "learning_rate": 4.744928676464659e-05,
+      "loss": 2.924,
+      "step": 50106
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7420288920402527,
+      "learning_rate": 4.744097725200585e-05,
+      "loss": 2.7727,
+      "step": 50107
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7241715788841248,
+      "learning_rate": 4.7432668404557054e-05,
+      "loss": 2.7901,
+      "step": 50108
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.70522540807724,
+      "learning_rate": 4.7424360222322e-05,
+      "loss": 2.6263,
+      "step": 50109
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7506316304206848,
+      "learning_rate": 4.7416052705322584e-05,
+      "loss": 3.0343,
+      "step": 50110
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7315349578857422,
+      "learning_rate": 4.740774585358058e-05,
+      "loss": 3.0413,
+      "step": 50111
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7280415892601013,
+      "learning_rate": 4.739943966711798e-05,
+      "loss": 2.9838,
+      "step": 50112
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7337613701820374,
+      "learning_rate": 4.739113414595672e-05,
+      "loss": 3.1778,
+      "step": 50113
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7724540829658508,
+      "learning_rate": 4.73828292901185e-05,
+      "loss": 2.7373,
+      "step": 50114
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.728657603263855,
+      "learning_rate": 4.737452509962538e-05,
+      "loss": 2.6146,
+      "step": 50115
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.774933397769928,
+      "learning_rate": 4.736622157449913e-05,
+      "loss": 2.9025,
+      "step": 50116
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7137971520423889,
+      "learning_rate": 4.735791871476159e-05,
+      "loss": 2.8057,
+      "step": 50117
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.723777711391449,
+      "learning_rate": 4.734961652043471e-05,
+      "loss": 2.9628,
+      "step": 50118
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7265045642852783,
+      "learning_rate": 4.734131499154027e-05,
+      "loss": 2.6844,
+      "step": 50119
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7293531894683838,
+      "learning_rate": 4.733301412810022e-05,
+      "loss": 2.9703,
+      "step": 50120
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7541471719741821,
+      "learning_rate": 4.7324713930136346e-05,
+      "loss": 2.9009,
+      "step": 50121
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7486446499824524,
+      "learning_rate": 4.7316414397670535e-05,
+      "loss": 2.8286,
+      "step": 50122
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7592146396636963,
+      "learning_rate": 4.730811553072472e-05,
+      "loss": 2.7649,
+      "step": 50123
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7211746573448181,
+      "learning_rate": 4.72998173293207e-05,
+      "loss": 2.611,
+      "step": 50124
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7064350843429565,
+      "learning_rate": 4.729151979348032e-05,
+      "loss": 2.9046,
+      "step": 50125
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7288532257080078,
+      "learning_rate": 4.7283222923225385e-05,
+      "loss": 2.9871,
+      "step": 50126
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7138639092445374,
+      "learning_rate": 4.727492671857782e-05,
+      "loss": 3.0573,
+      "step": 50127
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7234009504318237,
+      "learning_rate": 4.726663117955951e-05,
+      "loss": 2.9079,
+      "step": 50128
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7255566716194153,
+      "learning_rate": 4.725833630619217e-05,
+      "loss": 2.8949,
+      "step": 50129
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7464993596076965,
+      "learning_rate": 4.725004209849784e-05,
+      "loss": 2.8752,
+      "step": 50130
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7366446256637573,
+      "learning_rate": 4.724174855649822e-05,
+      "loss": 2.8406,
+      "step": 50131
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7235312461853027,
+      "learning_rate": 4.723345568021514e-05,
+      "loss": 2.7292,
+      "step": 50132
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7639561891555786,
+      "learning_rate": 4.7225163469670555e-05,
+      "loss": 2.93,
+      "step": 50133
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7084791660308838,
+      "learning_rate": 4.721687192488616e-05,
+      "loss": 3.0532,
+      "step": 50134
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7618387937545776,
+      "learning_rate": 4.720858104588399e-05,
+      "loss": 2.9025,
+      "step": 50135
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7836553454399109,
+      "learning_rate": 4.720029083268566e-05,
+      "loss": 2.9431,
+      "step": 50136
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7372074127197266,
+      "learning_rate": 4.7192001285313095e-05,
+      "loss": 2.4869,
+      "step": 50137
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7809687852859497,
+      "learning_rate": 4.718371240378824e-05,
+      "loss": 2.9196,
+      "step": 50138
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7471933960914612,
+      "learning_rate": 4.717542418813285e-05,
+      "loss": 3.0387,
+      "step": 50139
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7411444187164307,
+      "learning_rate": 4.7167136638368694e-05,
+      "loss": 3.055,
+      "step": 50140
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7536376118659973,
+      "learning_rate": 4.7158849754517604e-05,
+      "loss": 2.8807,
+      "step": 50141
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7287087440490723,
+      "learning_rate": 4.715056353660142e-05,
+      "loss": 3.0737,
+      "step": 50142
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7162652611732483,
+      "learning_rate": 4.7142277984642084e-05,
+      "loss": 3.0043,
+      "step": 50143
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7836141586303711,
+      "learning_rate": 4.7133993098661226e-05,
+      "loss": 2.9177,
+      "step": 50144
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7386676669120789,
+      "learning_rate": 4.712570887868086e-05,
+      "loss": 2.6577,
+      "step": 50145
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7532510757446289,
+      "learning_rate": 4.711742532472269e-05,
+      "loss": 2.8556,
+      "step": 50146
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7517036199569702,
+      "learning_rate": 4.710914243680851e-05,
+      "loss": 2.9074,
+      "step": 50147
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7926651239395142,
+      "learning_rate": 4.710086021496024e-05,
+      "loss": 2.6681,
+      "step": 50148
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7204487919807434,
+      "learning_rate": 4.709257865919954e-05,
+      "loss": 2.8429,
+      "step": 50149
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7931594252586365,
+      "learning_rate": 4.708429776954838e-05,
+      "loss": 2.6764,
+      "step": 50150
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7327009439468384,
+      "learning_rate": 4.7076017546028474e-05,
+      "loss": 2.8861,
+      "step": 50151
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7598114609718323,
+      "learning_rate": 4.7067737988661706e-05,
+      "loss": 2.5865,
+      "step": 50152
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7243731617927551,
+      "learning_rate": 4.705945909746984e-05,
+      "loss": 2.872,
+      "step": 50153
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7335036396980286,
+      "learning_rate": 4.705118087247461e-05,
+      "loss": 2.701,
+      "step": 50154
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7911424040794373,
+      "learning_rate": 4.7042903313697966e-05,
+      "loss": 2.7696,
+      "step": 50155
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6929256916046143,
+      "learning_rate": 4.703462642116158e-05,
+      "loss": 2.9583,
+      "step": 50156
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7362402081489563,
+      "learning_rate": 4.7026350194887276e-05,
+      "loss": 3.0049,
+      "step": 50157
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6989345550537109,
+      "learning_rate": 4.7018074634896966e-05,
+      "loss": 3.0146,
+      "step": 50158
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7882124781608582,
+      "learning_rate": 4.700979974121226e-05,
+      "loss": 2.9058,
+      "step": 50159
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7394413948059082,
+      "learning_rate": 4.7001525513855155e-05,
+      "loss": 3.002,
+      "step": 50160
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7787472009658813,
+      "learning_rate": 4.699325195284732e-05,
+      "loss": 2.9563,
+      "step": 50161
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7208383083343506,
+      "learning_rate": 4.698497905821052e-05,
+      "loss": 2.9109,
+      "step": 50162
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7691134214401245,
+      "learning_rate": 4.6976706829966646e-05,
+      "loss": 2.6999,
+      "step": 50163
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7439152598381042,
+      "learning_rate": 4.696843526813737e-05,
+      "loss": 2.7271,
+      "step": 50164
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7308996915817261,
+      "learning_rate": 4.69601643727446e-05,
+      "loss": 2.6876,
+      "step": 50165
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.8340305685997009,
+      "learning_rate": 4.695189414380999e-05,
+      "loss": 3.0734,
+      "step": 50166
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6861405968666077,
+      "learning_rate": 4.694362458135544e-05,
+      "loss": 2.9448,
+      "step": 50167
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7236627340316772,
+      "learning_rate": 4.693535568540272e-05,
+      "loss": 2.7754,
+      "step": 50168
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7470681071281433,
+      "learning_rate": 4.6927087455973466e-05,
+      "loss": 2.9247,
+      "step": 50169
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7589225172996521,
+      "learning_rate": 4.6918819893089667e-05,
+      "loss": 2.902,
+      "step": 50170
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7107172608375549,
+      "learning_rate": 4.69105529967729e-05,
+      "loss": 2.9717,
+      "step": 50171
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7193456888198853,
+      "learning_rate": 4.690228676704502e-05,
+      "loss": 2.9345,
+      "step": 50172
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7701574563980103,
+      "learning_rate": 4.689402120392787e-05,
+      "loss": 2.6803,
+      "step": 50173
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7131604552268982,
+      "learning_rate": 4.688575630744317e-05,
+      "loss": 3.0103,
+      "step": 50174
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7473611831665039,
+      "learning_rate": 4.68774920776127e-05,
+      "loss": 2.9097,
+      "step": 50175
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7232372164726257,
+      "learning_rate": 4.6869228514458076e-05,
+      "loss": 2.8485,
+      "step": 50176
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7372998595237732,
+      "learning_rate": 4.6860965618001223e-05,
+      "loss": 2.696,
+      "step": 50177
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7132263779640198,
+      "learning_rate": 4.6852703388263946e-05,
+      "loss": 2.7469,
+      "step": 50178
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7559736967086792,
+      "learning_rate": 4.684444182526787e-05,
+      "loss": 2.8729,
+      "step": 50179
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7392177581787109,
+      "learning_rate": 4.6836180929034826e-05,
+      "loss": 2.8812,
+      "step": 50180
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7340937256813049,
+      "learning_rate": 4.6827920699586555e-05,
+      "loss": 2.896,
+      "step": 50181
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7307325005531311,
+      "learning_rate": 4.681966113694485e-05,
+      "loss": 2.8654,
+      "step": 50182
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7325087785720825,
+      "learning_rate": 4.6811402241131445e-05,
+      "loss": 2.8743,
+      "step": 50183
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7375662326812744,
+      "learning_rate": 4.680314401216798e-05,
+      "loss": 2.886,
+      "step": 50184
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7653541564941406,
+      "learning_rate": 4.679488645007641e-05,
+      "loss": 2.9494,
+      "step": 50185
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7421068549156189,
+      "learning_rate": 4.678662955487831e-05,
+      "loss": 3.1609,
+      "step": 50186
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7914893627166748,
+      "learning_rate": 4.6778373326595534e-05,
+      "loss": 3.062,
+      "step": 50187
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7131171226501465,
+      "learning_rate": 4.677011776524973e-05,
+      "loss": 2.5862,
+      "step": 50188
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7471457123756409,
+      "learning_rate": 4.6761862870862776e-05,
+      "loss": 3.0483,
+      "step": 50189
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7291505932807922,
+      "learning_rate": 4.6753608643456365e-05,
+      "loss": 2.9675,
+      "step": 50190
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7441790103912354,
+      "learning_rate": 4.674535508305211e-05,
+      "loss": 2.9056,
+      "step": 50191
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7517664432525635,
+      "learning_rate": 4.6737102189671916e-05,
+      "loss": 2.8592,
+      "step": 50192
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7586925029754639,
+      "learning_rate": 4.6728849963337415e-05,
+      "loss": 2.7071,
+      "step": 50193
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.754487931728363,
+      "learning_rate": 4.672059840407034e-05,
+      "loss": 3.0674,
+      "step": 50194
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7265577912330627,
+      "learning_rate": 4.6712347511892556e-05,
+      "loss": 2.9569,
+      "step": 50195
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7128158807754517,
+      "learning_rate": 4.67040972868257e-05,
+      "loss": 3.096,
+      "step": 50196
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6750921607017517,
+      "learning_rate": 4.66958477288915e-05,
+      "loss": 2.7767,
+      "step": 50197
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.724058210849762,
+      "learning_rate": 4.668759883811163e-05,
+      "loss": 2.9174,
+      "step": 50198
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7566876411437988,
+      "learning_rate": 4.667935061450787e-05,
+      "loss": 2.9974,
+      "step": 50199
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7634623050689697,
+      "learning_rate": 4.667110305810201e-05,
+      "loss": 2.8993,
+      "step": 50200
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7499482035636902,
+      "learning_rate": 4.666285616891563e-05,
+      "loss": 3.0086,
+      "step": 50201
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7458711862564087,
+      "learning_rate": 4.6654609946970614e-05,
+      "loss": 3.0176,
+      "step": 50202
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7508730292320251,
+      "learning_rate": 4.664636439228854e-05,
+      "loss": 2.8624,
+      "step": 50203
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7427207231521606,
+      "learning_rate": 4.663811950489122e-05,
+      "loss": 2.8796,
+      "step": 50204
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7459037899971008,
+      "learning_rate": 4.6629875284800385e-05,
+      "loss": 2.821,
+      "step": 50205
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6995728611946106,
+      "learning_rate": 4.662163173203758e-05,
+      "loss": 2.9301,
+      "step": 50206
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7462338805198669,
+      "learning_rate": 4.661338884662471e-05,
+      "loss": 2.9296,
+      "step": 50207
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7074584364891052,
+      "learning_rate": 4.6605146628583365e-05,
+      "loss": 2.956,
+      "step": 50208
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.745572030544281,
+      "learning_rate": 4.6596905077935266e-05,
+      "loss": 2.9507,
+      "step": 50209
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7482078075408936,
+      "learning_rate": 4.658866419470225e-05,
+      "loss": 2.9588,
+      "step": 50210
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6879667043685913,
+      "learning_rate": 4.658042397890591e-05,
+      "loss": 2.7677,
+      "step": 50211
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7026304602622986,
+      "learning_rate": 4.657218443056796e-05,
+      "loss": 3.1308,
+      "step": 50212
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7168485522270203,
+      "learning_rate": 4.656394554971002e-05,
+      "loss": 2.6956,
+      "step": 50213
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7557949423789978,
+      "learning_rate": 4.655570733635389e-05,
+      "loss": 2.959,
+      "step": 50214
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7124577760696411,
+      "learning_rate": 4.654746979052133e-05,
+      "loss": 2.9676,
+      "step": 50215
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7355839014053345,
+      "learning_rate": 4.653923291223386e-05,
+      "loss": 3.0602,
+      "step": 50216
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7148059606552124,
+      "learning_rate": 4.653099670151336e-05,
+      "loss": 2.8388,
+      "step": 50217
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7521902918815613,
+      "learning_rate": 4.6522761158381407e-05,
+      "loss": 2.7998,
+      "step": 50218
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7478179335594177,
+      "learning_rate": 4.651452628285969e-05,
+      "loss": 3.0467,
+      "step": 50219
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7289862036705017,
+      "learning_rate": 4.6506292074969955e-05,
+      "loss": 2.8849,
+      "step": 50220
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7552449703216553,
+      "learning_rate": 4.649805853473382e-05,
+      "loss": 2.9719,
+      "step": 50221
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7331597805023193,
+      "learning_rate": 4.648982566217307e-05,
+      "loss": 3.1323,
+      "step": 50222
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7304408550262451,
+      "learning_rate": 4.648159345730927e-05,
+      "loss": 2.7518,
+      "step": 50223
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7600448727607727,
+      "learning_rate": 4.647336192016413e-05,
+      "loss": 2.8241,
+      "step": 50224
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7154588103294373,
+      "learning_rate": 4.646513105075949e-05,
+      "loss": 2.7548,
+      "step": 50225
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.729921281337738,
+      "learning_rate": 4.645690084911685e-05,
+      "loss": 2.94,
+      "step": 50226
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7225162386894226,
+      "learning_rate": 4.644867131525795e-05,
+      "loss": 2.8544,
+      "step": 50227
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7441264390945435,
+      "learning_rate": 4.6440442449204386e-05,
+      "loss": 3.0495,
+      "step": 50228
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7257991433143616,
+      "learning_rate": 4.643221425097792e-05,
+      "loss": 2.7045,
+      "step": 50229
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7100118398666382,
+      "learning_rate": 4.642398672060026e-05,
+      "loss": 2.7428,
+      "step": 50230
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7242577075958252,
+      "learning_rate": 4.6415759858092926e-05,
+      "loss": 2.9569,
+      "step": 50231
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7217210531234741,
+      "learning_rate": 4.6407533663477766e-05,
+      "loss": 3.0893,
+      "step": 50232
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7589887380599976,
+      "learning_rate": 4.6399308136776336e-05,
+      "loss": 2.7432,
+      "step": 50233
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6991437077522278,
+      "learning_rate": 4.639108327801028e-05,
+      "loss": 3.0003,
+      "step": 50234
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7275975942611694,
+      "learning_rate": 4.638285908720137e-05,
+      "loss": 2.8634,
+      "step": 50235
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7588894367218018,
+      "learning_rate": 4.637463556437111e-05,
+      "loss": 3.0073,
+      "step": 50236
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7242416739463806,
+      "learning_rate": 4.6366412709541336e-05,
+      "loss": 3.0063,
+      "step": 50237
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7257980704307556,
+      "learning_rate": 4.635819052273353e-05,
+      "loss": 3.0214,
+      "step": 50238
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7471466064453125,
+      "learning_rate": 4.6349969003969444e-05,
+      "loss": 2.9178,
+      "step": 50239
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.746959924697876,
+      "learning_rate": 4.6341748153270864e-05,
+      "loss": 2.9513,
+      "step": 50240
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7836167812347412,
+      "learning_rate": 4.633352797065917e-05,
+      "loss": 2.9061,
+      "step": 50241
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.764286994934082,
+      "learning_rate": 4.632530845615621e-05,
+      "loss": 2.8445,
+      "step": 50242
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7559465765953064,
+      "learning_rate": 4.63170896097835e-05,
+      "loss": 2.8471,
+      "step": 50243
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7415547966957092,
+      "learning_rate": 4.630887143156276e-05,
+      "loss": 2.8428,
+      "step": 50244
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7760792970657349,
+      "learning_rate": 4.6300653921515665e-05,
+      "loss": 2.8229,
+      "step": 50245
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7538564801216125,
+      "learning_rate": 4.6292437079663766e-05,
+      "loss": 2.8987,
+      "step": 50246
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7254053354263306,
+      "learning_rate": 4.6284220906028854e-05,
+      "loss": 2.9147,
+      "step": 50247
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.808498203754425,
+      "learning_rate": 4.627600540063248e-05,
+      "loss": 2.7537,
+      "step": 50248
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7465435862541199,
+      "learning_rate": 4.626779056349617e-05,
+      "loss": 2.7917,
+      "step": 50249
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7259182929992676,
+      "learning_rate": 4.625957639464177e-05,
+      "loss": 2.9146,
+      "step": 50250
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7807119488716125,
+      "learning_rate": 4.6251362894090715e-05,
+      "loss": 2.967,
+      "step": 50251
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.744609534740448,
+      "learning_rate": 4.624315006186482e-05,
+      "loss": 2.8336,
+      "step": 50252
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.703451931476593,
+      "learning_rate": 4.623493789798557e-05,
+      "loss": 2.8774,
+      "step": 50253
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.8264244198799133,
+      "learning_rate": 4.622672640247471e-05,
+      "loss": 2.7383,
+      "step": 50254
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.736638069152832,
+      "learning_rate": 4.621851557535383e-05,
+      "loss": 2.7002,
+      "step": 50255
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7538456320762634,
+      "learning_rate": 4.621030541664444e-05,
+      "loss": 2.9702,
+      "step": 50256
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7488062977790833,
+      "learning_rate": 4.6202095926368366e-05,
+      "loss": 2.812,
+      "step": 50257
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7549282908439636,
+      "learning_rate": 4.619388710454704e-05,
+      "loss": 2.9381,
+      "step": 50258
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7125292420387268,
+      "learning_rate": 4.618567895120223e-05,
+      "loss": 3.0659,
+      "step": 50259
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.8093187808990479,
+      "learning_rate": 4.617747146635544e-05,
+      "loss": 2.9393,
+      "step": 50260
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7467921376228333,
+      "learning_rate": 4.616926465002836e-05,
+      "loss": 2.9444,
+      "step": 50261
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.774636447429657,
+      "learning_rate": 4.616105850224263e-05,
+      "loss": 2.9592,
+      "step": 50262
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7182961702346802,
+      "learning_rate": 4.615285302301981e-05,
+      "loss": 2.8498,
+      "step": 50263
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7876618504524231,
+      "learning_rate": 4.614464821238154e-05,
+      "loss": 2.8896,
+      "step": 50264
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7740433812141418,
+      "learning_rate": 4.613644407034934e-05,
+      "loss": 3.0115,
+      "step": 50265
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7827932834625244,
+      "learning_rate": 4.612824059694486e-05,
+      "loss": 3.0513,
+      "step": 50266
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7410049438476562,
+      "learning_rate": 4.6120037792189854e-05,
+      "loss": 2.7407,
+      "step": 50267
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7162790298461914,
+      "learning_rate": 4.61118356561057e-05,
+      "loss": 2.7981,
+      "step": 50268
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7440741658210754,
+      "learning_rate": 4.6103634188714214e-05,
+      "loss": 2.9791,
+      "step": 50269
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.715707540512085,
+      "learning_rate": 4.6095433390036876e-05,
+      "loss": 2.86,
+      "step": 50270
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7865996956825256,
+      "learning_rate": 4.60872332600952e-05,
+      "loss": 2.8264,
+      "step": 50271
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7152213454246521,
+      "learning_rate": 4.607903379891098e-05,
+      "loss": 2.9668,
+      "step": 50272
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7056474089622498,
+      "learning_rate": 4.6070835006505655e-05,
+      "loss": 2.9386,
+      "step": 50273
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7383698225021362,
+      "learning_rate": 4.606263688290095e-05,
+      "loss": 3.0749,
+      "step": 50274
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7201040387153625,
+      "learning_rate": 4.6054439428118306e-05,
+      "loss": 2.8478,
+      "step": 50275
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7062273025512695,
+      "learning_rate": 4.6046242642179474e-05,
+      "loss": 2.9832,
+      "step": 50276
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.687692403793335,
+      "learning_rate": 4.6038046525105974e-05,
+      "loss": 2.6567,
+      "step": 50277
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7824928760528564,
+      "learning_rate": 4.602985107691929e-05,
+      "loss": 2.8693,
+      "step": 50278
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7675195932388306,
+      "learning_rate": 4.60216562976412e-05,
+      "loss": 2.7285,
+      "step": 50279
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7547996640205383,
+      "learning_rate": 4.6013462187293085e-05,
+      "loss": 2.709,
+      "step": 50280
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7081446647644043,
+      "learning_rate": 4.600526874589666e-05,
+      "loss": 2.9257,
+      "step": 50281
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7581326961517334,
+      "learning_rate": 4.599707597347352e-05,
+      "loss": 2.8561,
+      "step": 50282
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6948836445808411,
+      "learning_rate": 4.5988883870045135e-05,
+      "loss": 2.9192,
+      "step": 50283
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7415758371353149,
+      "learning_rate": 4.598069243563323e-05,
+      "loss": 2.7314,
+      "step": 50284
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7500677704811096,
+      "learning_rate": 4.5972501670259275e-05,
+      "loss": 2.8871,
+      "step": 50285
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.739080548286438,
+      "learning_rate": 4.59643115739448e-05,
+      "loss": 3.08,
+      "step": 50286
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7174320816993713,
+      "learning_rate": 4.595612214671151e-05,
+      "loss": 3.0523,
+      "step": 50287
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7595369219779968,
+      "learning_rate": 4.5947933388580835e-05,
+      "loss": 3.0033,
+      "step": 50288
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.741030216217041,
+      "learning_rate": 4.593974529957448e-05,
+      "loss": 2.7456,
+      "step": 50289
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7448769807815552,
+      "learning_rate": 4.5931557879713865e-05,
+      "loss": 2.9838,
+      "step": 50290
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.744935154914856,
+      "learning_rate": 4.5923371129020704e-05,
+      "loss": 2.8322,
+      "step": 50291
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7401701211929321,
+      "learning_rate": 4.591518504751648e-05,
+      "loss": 3.0576,
+      "step": 50292
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.748890221118927,
+      "learning_rate": 4.5906999635222706e-05,
+      "loss": 2.9098,
+      "step": 50293
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7550206780433655,
+      "learning_rate": 4.589881489216104e-05,
+      "loss": 2.846,
+      "step": 50294
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7691561579704285,
+      "learning_rate": 4.5890630818352926e-05,
+      "loss": 2.8187,
+      "step": 50295
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7871925234794617,
+      "learning_rate": 4.5882447413820015e-05,
+      "loss": 2.7289,
+      "step": 50296
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7003697752952576,
+      "learning_rate": 4.587426467858388e-05,
+      "loss": 2.8819,
+      "step": 50297
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7725687026977539,
+      "learning_rate": 4.586608261266602e-05,
+      "loss": 2.8305,
+      "step": 50298
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7272656559944153,
+      "learning_rate": 4.585790121608798e-05,
+      "loss": 3.1106,
+      "step": 50299
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7602006196975708,
+      "learning_rate": 4.584972048887127e-05,
+      "loss": 2.934,
+      "step": 50300
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7348072528839111,
+      "learning_rate": 4.584154043103747e-05,
+      "loss": 3.0664,
+      "step": 50301
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.752991795539856,
+      "learning_rate": 4.583336104260821e-05,
+      "loss": 2.8116,
+      "step": 50302
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7089395523071289,
+      "learning_rate": 4.582518232360489e-05,
+      "loss": 2.9561,
+      "step": 50303
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7611316442489624,
+      "learning_rate": 4.5817004274049205e-05,
+      "loss": 2.9413,
+      "step": 50304
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7629949450492859,
+      "learning_rate": 4.5808826893962535e-05,
+      "loss": 2.7534,
+      "step": 50305
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7246274948120117,
+      "learning_rate": 4.580065018336656e-05,
+      "loss": 2.9361,
+      "step": 50306
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7618169784545898,
+      "learning_rate": 4.5792474142282766e-05,
+      "loss": 2.8229,
+      "step": 50307
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7503147721290588,
+      "learning_rate": 4.5784298770732595e-05,
+      "loss": 3.0257,
+      "step": 50308
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.721960723400116,
+      "learning_rate": 4.577612406873774e-05,
+      "loss": 2.9242,
+      "step": 50309
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7781165242195129,
+      "learning_rate": 4.576795003631954e-05,
+      "loss": 2.9042,
+      "step": 50310
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7347601056098938,
+      "learning_rate": 4.575977667349966e-05,
+      "loss": 2.8163,
+      "step": 50311
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7000023722648621,
+      "learning_rate": 4.575160398029967e-05,
+      "loss": 2.8964,
+      "step": 50312
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7539365887641907,
+      "learning_rate": 4.574343195674102e-05,
+      "loss": 3.0921,
+      "step": 50313
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6888923645019531,
+      "learning_rate": 4.573526060284523e-05,
+      "loss": 2.8024,
+      "step": 50314
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7311072945594788,
+      "learning_rate": 4.572708991863374e-05,
+      "loss": 2.6267,
+      "step": 50315
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7336209416389465,
+      "learning_rate": 4.571891990412822e-05,
+      "loss": 2.8586,
+      "step": 50316
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7051305174827576,
+      "learning_rate": 4.5710750559350126e-05,
+      "loss": 2.9322,
+      "step": 50317
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7536783814430237,
+      "learning_rate": 4.5702581884320964e-05,
+      "loss": 2.7619,
+      "step": 50318
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7377327680587769,
+      "learning_rate": 4.569441387906231e-05,
+      "loss": 2.7176,
+      "step": 50319
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7616892457008362,
+      "learning_rate": 4.568624654359561e-05,
+      "loss": 2.8448,
+      "step": 50320
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7389867305755615,
+      "learning_rate": 4.567807987794234e-05,
+      "loss": 2.8724,
+      "step": 50321
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7348322868347168,
+      "learning_rate": 4.566991388212413e-05,
+      "loss": 3.0046,
+      "step": 50322
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7657172679901123,
+      "learning_rate": 4.5661748556162357e-05,
+      "loss": 2.9771,
+      "step": 50323
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6982909440994263,
+      "learning_rate": 4.5653583900078636e-05,
+      "loss": 2.8583,
+      "step": 50324
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8317862153053284,
+      "learning_rate": 4.564541991389439e-05,
+      "loss": 2.9129,
+      "step": 50325
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6974571943283081,
+      "learning_rate": 4.563725659763112e-05,
+      "loss": 2.9483,
+      "step": 50326
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.813891589641571,
+      "learning_rate": 4.5629093951310456e-05,
+      "loss": 2.9055,
+      "step": 50327
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7852758765220642,
+      "learning_rate": 4.5620931974953815e-05,
+      "loss": 2.8643,
+      "step": 50328
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7360036373138428,
+      "learning_rate": 4.5612770668582664e-05,
+      "loss": 2.8992,
+      "step": 50329
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7240647077560425,
+      "learning_rate": 4.560461003221844e-05,
+      "loss": 2.815,
+      "step": 50330
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7547604441642761,
+      "learning_rate": 4.559645006588274e-05,
+      "loss": 3.0039,
+      "step": 50331
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7343550324440002,
+      "learning_rate": 4.558829076959706e-05,
+      "loss": 2.8534,
+      "step": 50332
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7282590866088867,
+      "learning_rate": 4.5580132143382805e-05,
+      "loss": 3.1069,
+      "step": 50333
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7644495964050293,
+      "learning_rate": 4.557197418726161e-05,
+      "loss": 3.1148,
+      "step": 50334
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7662924528121948,
+      "learning_rate": 4.556381690125487e-05,
+      "loss": 3.0473,
+      "step": 50335
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7156445980072021,
+      "learning_rate": 4.555566028538398e-05,
+      "loss": 2.9062,
+      "step": 50336
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7530838251113892,
+      "learning_rate": 4.554750433967058e-05,
+      "loss": 2.9819,
+      "step": 50337
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7620644569396973,
+      "learning_rate": 4.5539349064136034e-05,
+      "loss": 2.9154,
+      "step": 50338
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7103702425956726,
+      "learning_rate": 4.5531194458801914e-05,
+      "loss": 3.012,
+      "step": 50339
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6881940960884094,
+      "learning_rate": 4.55230405236896e-05,
+      "loss": 3.1089,
+      "step": 50340
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7752044796943665,
+      "learning_rate": 4.5514887258820686e-05,
+      "loss": 2.8773,
+      "step": 50341
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7745306491851807,
+      "learning_rate": 4.550673466421658e-05,
+      "loss": 2.8716,
+      "step": 50342
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7594026923179626,
+      "learning_rate": 4.5498582739898705e-05,
+      "loss": 2.9974,
+      "step": 50343
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7362309694290161,
+      "learning_rate": 4.549043148588863e-05,
+      "loss": 2.9818,
+      "step": 50344
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7302587032318115,
+      "learning_rate": 4.548228090220773e-05,
+      "loss": 2.8843,
+      "step": 50345
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7178745865821838,
+      "learning_rate": 4.547413098887755e-05,
+      "loss": 2.8736,
+      "step": 50346
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.753461480140686,
+      "learning_rate": 4.546598174591951e-05,
+      "loss": 3.1034,
+      "step": 50347
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7480427026748657,
+      "learning_rate": 4.545783317335504e-05,
+      "loss": 3.1151,
+      "step": 50348
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7292770147323608,
+      "learning_rate": 4.544968527120576e-05,
+      "loss": 2.8363,
+      "step": 50349
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7527563571929932,
+      "learning_rate": 4.5441538039492985e-05,
+      "loss": 2.9367,
+      "step": 50350
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7489085793495178,
+      "learning_rate": 4.543339147823824e-05,
+      "loss": 2.9372,
+      "step": 50351
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7443061470985413,
+      "learning_rate": 4.5425245587462855e-05,
+      "loss": 2.9903,
+      "step": 50352
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.73963463306427,
+      "learning_rate": 4.5417100367188365e-05,
+      "loss": 2.977,
+      "step": 50353
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7527986168861389,
+      "learning_rate": 4.5408955817436344e-05,
+      "loss": 2.7493,
+      "step": 50354
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7716648578643799,
+      "learning_rate": 4.540081193822804e-05,
+      "loss": 2.9532,
+      "step": 50355
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6952173113822937,
+      "learning_rate": 4.539266872958508e-05,
+      "loss": 3.0827,
+      "step": 50356
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7592951059341431,
+      "learning_rate": 4.538452619152884e-05,
+      "loss": 2.9578,
+      "step": 50357
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7521546483039856,
+      "learning_rate": 4.537638432408066e-05,
+      "loss": 2.7848,
+      "step": 50358
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7431459426879883,
+      "learning_rate": 4.536824312726217e-05,
+      "loss": 2.8052,
+      "step": 50359
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7520620822906494,
+      "learning_rate": 4.536010260109465e-05,
+      "loss": 2.8657,
+      "step": 50360
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7569798827171326,
+      "learning_rate": 4.535196274559968e-05,
+      "loss": 2.9279,
+      "step": 50361
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.771931529045105,
+      "learning_rate": 4.5343823560798574e-05,
+      "loss": 2.8402,
+      "step": 50362
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7233810424804688,
+      "learning_rate": 4.533568504671282e-05,
+      "loss": 2.8706,
+      "step": 50363
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6947945952415466,
+      "learning_rate": 4.53275472033639e-05,
+      "loss": 2.771,
+      "step": 50364
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7521064281463623,
+      "learning_rate": 4.531941003077326e-05,
+      "loss": 2.7515,
+      "step": 50365
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7488537430763245,
+      "learning_rate": 4.531127352896226e-05,
+      "loss": 2.7538,
+      "step": 50366
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.78842693567276,
+      "learning_rate": 4.5303137697952266e-05,
+      "loss": 3.043,
+      "step": 50367
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7508882284164429,
+      "learning_rate": 4.5295002537764804e-05,
+      "loss": 2.9645,
+      "step": 50368
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8009318113327026,
+      "learning_rate": 4.528686804842136e-05,
+      "loss": 2.9572,
+      "step": 50369
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7622132897377014,
+      "learning_rate": 4.5278734229943204e-05,
+      "loss": 2.6743,
+      "step": 50370
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7067914605140686,
+      "learning_rate": 4.52706010823519e-05,
+      "loss": 2.8212,
+      "step": 50371
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7103236317634583,
+      "learning_rate": 4.5262468605668823e-05,
+      "loss": 2.9323,
+      "step": 50372
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7678020596504211,
+      "learning_rate": 4.525433679991529e-05,
+      "loss": 3.0528,
+      "step": 50373
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7323145866394043,
+      "learning_rate": 4.5246205665112924e-05,
+      "loss": 2.7942,
+      "step": 50374
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.713549792766571,
+      "learning_rate": 4.52380752012829e-05,
+      "loss": 2.9964,
+      "step": 50375
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7168814539909363,
+      "learning_rate": 4.5229945408446865e-05,
+      "loss": 2.8764,
+      "step": 50376
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7501139044761658,
+      "learning_rate": 4.522181628662605e-05,
+      "loss": 2.8077,
+      "step": 50377
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7667758464813232,
+      "learning_rate": 4.5213687835841983e-05,
+      "loss": 2.8093,
+      "step": 50378
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7523639798164368,
+      "learning_rate": 4.520556005611603e-05,
+      "loss": 2.8804,
+      "step": 50379
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7287237048149109,
+      "learning_rate": 4.5197432947469536e-05,
+      "loss": 3.0469,
+      "step": 50380
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7388063073158264,
+      "learning_rate": 4.518930650992405e-05,
+      "loss": 2.8228,
+      "step": 50381
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.751380980014801,
+      "learning_rate": 4.518118074350079e-05,
+      "loss": 2.8182,
+      "step": 50382
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7500567436218262,
+      "learning_rate": 4.517305564822127e-05,
+      "loss": 3.1013,
+      "step": 50383
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7376142740249634,
+      "learning_rate": 4.516493122410697e-05,
+      "loss": 2.8591,
+      "step": 50384
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6814765334129333,
+      "learning_rate": 4.515680747117911e-05,
+      "loss": 2.7155,
+      "step": 50385
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7675392031669617,
+      "learning_rate": 4.5148684389459244e-05,
+      "loss": 2.9702,
+      "step": 50386
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7462222576141357,
+      "learning_rate": 4.514056197896868e-05,
+      "loss": 2.7139,
+      "step": 50387
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7849350571632385,
+      "learning_rate": 4.5132440239728774e-05,
+      "loss": 3.1364,
+      "step": 50388
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7008315324783325,
+      "learning_rate": 4.512431917176107e-05,
+      "loss": 3.0296,
+      "step": 50389
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7495970726013184,
+      "learning_rate": 4.5116198775086756e-05,
+      "loss": 2.5677,
+      "step": 50390
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.732743501663208,
+      "learning_rate": 4.510807904972737e-05,
+      "loss": 2.8116,
+      "step": 50391
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7251850962638855,
+      "learning_rate": 4.509995999570422e-05,
+      "loss": 2.9006,
+      "step": 50392
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7376589775085449,
+      "learning_rate": 4.509184161303876e-05,
+      "loss": 3.0105,
+      "step": 50393
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7272580862045288,
+      "learning_rate": 4.508372390175235e-05,
+      "loss": 2.7939,
+      "step": 50394
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7777386903762817,
+      "learning_rate": 4.5075606861866285e-05,
+      "loss": 2.7377,
+      "step": 50395
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7213007211685181,
+      "learning_rate": 4.5067490493402066e-05,
+      "loss": 2.7687,
+      "step": 50396
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.707280158996582,
+      "learning_rate": 4.505937479638092e-05,
+      "loss": 2.8246,
+      "step": 50397
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.74385666847229,
+      "learning_rate": 4.505125977082436e-05,
+      "loss": 2.7228,
+      "step": 50398
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.738721489906311,
+      "learning_rate": 4.5043145416753755e-05,
+      "loss": 2.8539,
+      "step": 50399
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7245712876319885,
+      "learning_rate": 4.5035031734190476e-05,
+      "loss": 3.0549,
+      "step": 50400
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7531771063804626,
+      "learning_rate": 4.5026918723155826e-05,
+      "loss": 2.9992,
+      "step": 50401
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7361901998519897,
+      "learning_rate": 4.501880638367112e-05,
+      "loss": 2.7009,
+      "step": 50402
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7059527039527893,
+      "learning_rate": 4.5010694715757826e-05,
+      "loss": 2.8706,
+      "step": 50403
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7453323602676392,
+      "learning_rate": 4.500258371943737e-05,
+      "loss": 2.6934,
+      "step": 50404
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7383242845535278,
+      "learning_rate": 4.4994473394730944e-05,
+      "loss": 2.8522,
+      "step": 50405
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.723354697227478,
+      "learning_rate": 4.498636374166007e-05,
+      "loss": 2.944,
+      "step": 50406
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7948548793792725,
+      "learning_rate": 4.4978254760245955e-05,
+      "loss": 2.9672,
+      "step": 50407
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7579402923583984,
+      "learning_rate": 4.4970146450510124e-05,
+      "loss": 2.704,
+      "step": 50408
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7572686672210693,
+      "learning_rate": 4.496203881247385e-05,
+      "loss": 2.9605,
+      "step": 50409
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7329517006874084,
+      "learning_rate": 4.49539318461584e-05,
+      "loss": 2.8146,
+      "step": 50410
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.744154691696167,
+      "learning_rate": 4.49458255515853e-05,
+      "loss": 2.7126,
+      "step": 50411
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7456027269363403,
+      "learning_rate": 4.493771992877572e-05,
+      "loss": 2.9384,
+      "step": 50412
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7537271976470947,
+      "learning_rate": 4.492961497775109e-05,
+      "loss": 2.9007,
+      "step": 50413
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7661435008049011,
+      "learning_rate": 4.4921510698532846e-05,
+      "loss": 2.7748,
+      "step": 50414
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7212929129600525,
+      "learning_rate": 4.4913407091142265e-05,
+      "loss": 3.0509,
+      "step": 50415
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7772074341773987,
+      "learning_rate": 4.490530415560065e-05,
+      "loss": 2.6768,
+      "step": 50416
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7107312083244324,
+      "learning_rate": 4.48972018919293e-05,
+      "loss": 3.0139,
+      "step": 50417
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7618693709373474,
+      "learning_rate": 4.488910030014973e-05,
+      "loss": 2.992,
+      "step": 50418
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7590107321739197,
+      "learning_rate": 4.4880999380283075e-05,
+      "loss": 2.7462,
+      "step": 50419
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7364644408226013,
+      "learning_rate": 4.487289913235077e-05,
+      "loss": 2.9629,
+      "step": 50420
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.726629376411438,
+      "learning_rate": 4.486479955637422e-05,
+      "loss": 3.1113,
+      "step": 50421
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7420285940170288,
+      "learning_rate": 4.4856700652374714e-05,
+      "loss": 2.9697,
+      "step": 50422
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7878907322883606,
+      "learning_rate": 4.484860242037352e-05,
+      "loss": 2.7718,
+      "step": 50423
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7129849791526794,
+      "learning_rate": 4.484050486039195e-05,
+      "loss": 3.0591,
+      "step": 50424
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7504441142082214,
+      "learning_rate": 4.4832407972451354e-05,
+      "loss": 2.9215,
+      "step": 50425
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7380017042160034,
+      "learning_rate": 4.4824311756573204e-05,
+      "loss": 2.9334,
+      "step": 50426
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7509652376174927,
+      "learning_rate": 4.481621621277861e-05,
+      "loss": 3.0772,
+      "step": 50427
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7229263186454773,
+      "learning_rate": 4.4808121341089054e-05,
+      "loss": 2.8744,
+      "step": 50428
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7127883434295654,
+      "learning_rate": 4.480002714152575e-05,
+      "loss": 2.7404,
+      "step": 50429
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.77323979139328,
+      "learning_rate": 4.479193361411012e-05,
+      "loss": 3.0917,
+      "step": 50430
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7353416085243225,
+      "learning_rate": 4.478384075886341e-05,
+      "loss": 2.8433,
+      "step": 50431
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7446988821029663,
+      "learning_rate": 4.477574857580687e-05,
+      "loss": 3.1125,
+      "step": 50432
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7732658386230469,
+      "learning_rate": 4.476765706496198e-05,
+      "loss": 2.7603,
+      "step": 50433
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7612114548683167,
+      "learning_rate": 4.475956622634986e-05,
+      "loss": 3.1181,
+      "step": 50434
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.9173521399497986,
+      "learning_rate": 4.475147605999196e-05,
+      "loss": 2.8608,
+      "step": 50435
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7881463766098022,
+      "learning_rate": 4.47433865659096e-05,
+      "loss": 2.8996,
+      "step": 50436
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7628815174102783,
+      "learning_rate": 4.473529774412405e-05,
+      "loss": 2.9975,
+      "step": 50437
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.710648238658905,
+      "learning_rate": 4.472720959465658e-05,
+      "loss": 3.011,
+      "step": 50438
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.705659806728363,
+      "learning_rate": 4.471912211752845e-05,
+      "loss": 3.1302,
+      "step": 50439
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7531771659851074,
+      "learning_rate": 4.471103531276102e-05,
+      "loss": 2.7541,
+      "step": 50440
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7266751527786255,
+      "learning_rate": 4.470294918037565e-05,
+      "loss": 2.8924,
+      "step": 50441
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7597355246543884,
+      "learning_rate": 4.4694863720393536e-05,
+      "loss": 2.7383,
+      "step": 50442
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7361400723457336,
+      "learning_rate": 4.4686778932836043e-05,
+      "loss": 2.8306,
+      "step": 50443
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7326222062110901,
+      "learning_rate": 4.467869481772448e-05,
+      "loss": 2.7957,
+      "step": 50444
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7179027199745178,
+      "learning_rate": 4.467061137508e-05,
+      "loss": 2.7548,
+      "step": 50445
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7314981818199158,
+      "learning_rate": 4.466252860492409e-05,
+      "loss": 2.7318,
+      "step": 50446
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8546863794326782,
+      "learning_rate": 4.465444650727785e-05,
+      "loss": 2.7946,
+      "step": 50447
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7847939729690552,
+      "learning_rate": 4.4646365082162715e-05,
+      "loss": 2.8562,
+      "step": 50448
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7865023016929626,
+      "learning_rate": 4.463828432959986e-05,
+      "loss": 2.7398,
+      "step": 50449
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7689087390899658,
+      "learning_rate": 4.463020424961064e-05,
+      "loss": 2.9209,
+      "step": 50450
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7183415293693542,
+      "learning_rate": 4.462212484221636e-05,
+      "loss": 2.8917,
+      "step": 50451
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7046708464622498,
+      "learning_rate": 4.461404610743824e-05,
+      "loss": 3.1845,
+      "step": 50452
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7757652401924133,
+      "learning_rate": 4.4605968045297566e-05,
+      "loss": 2.8422,
+      "step": 50453
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.702682614326477,
+      "learning_rate": 4.459789065581559e-05,
+      "loss": 2.8507,
+      "step": 50454
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7367978692054749,
+      "learning_rate": 4.45898139390136e-05,
+      "loss": 2.7709,
+      "step": 50455
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7395363450050354,
+      "learning_rate": 4.458173789491293e-05,
+      "loss": 2.6215,
+      "step": 50456
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8202866315841675,
+      "learning_rate": 4.457366252353477e-05,
+      "loss": 2.9662,
+      "step": 50457
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.738206148147583,
+      "learning_rate": 4.4565587824900475e-05,
+      "loss": 2.843,
+      "step": 50458
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6997504830360413,
+      "learning_rate": 4.455751379903129e-05,
+      "loss": 2.8662,
+      "step": 50459
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7250627875328064,
+      "learning_rate": 4.454944044594836e-05,
+      "loss": 2.9364,
+      "step": 50460
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7618042230606079,
+      "learning_rate": 4.4541367765673105e-05,
+      "loss": 2.7884,
+      "step": 50461
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7377805113792419,
+      "learning_rate": 4.453329575822665e-05,
+      "loss": 2.7566,
+      "step": 50462
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7210468649864197,
+      "learning_rate": 4.4525224423630434e-05,
+      "loss": 2.8037,
+      "step": 50463
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7345592975616455,
+      "learning_rate": 4.451715376190551e-05,
+      "loss": 2.8834,
+      "step": 50464
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7215968370437622,
+      "learning_rate": 4.4509083773073226e-05,
+      "loss": 2.9943,
+      "step": 50465
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7119701504707336,
+      "learning_rate": 4.4501014457154935e-05,
+      "loss": 2.8041,
+      "step": 50466
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7098177075386047,
+      "learning_rate": 4.449294581417182e-05,
+      "loss": 2.8293,
+      "step": 50467
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7481569647789001,
+      "learning_rate": 4.448487784414506e-05,
+      "loss": 2.802,
+      "step": 50468
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7269555330276489,
+      "learning_rate": 4.447681054709594e-05,
+      "loss": 2.7692,
+      "step": 50469
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6905206441879272,
+      "learning_rate": 4.446874392304568e-05,
+      "loss": 2.6578,
+      "step": 50470
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7605257630348206,
+      "learning_rate": 4.446067797201569e-05,
+      "loss": 2.8615,
+      "step": 50471
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7663816213607788,
+      "learning_rate": 4.445261269402699e-05,
+      "loss": 2.7005,
+      "step": 50472
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7191436290740967,
+      "learning_rate": 4.4444548089101e-05,
+      "loss": 3.0061,
+      "step": 50473
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7263677716255188,
+      "learning_rate": 4.443648415725889e-05,
+      "loss": 2.8994,
+      "step": 50474
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7229388952255249,
+      "learning_rate": 4.442842089852182e-05,
+      "loss": 2.9245,
+      "step": 50475
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8069564700126648,
+      "learning_rate": 4.442035831291117e-05,
+      "loss": 2.7624,
+      "step": 50476
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7363615036010742,
+      "learning_rate": 4.4412296400448064e-05,
+      "loss": 2.8038,
+      "step": 50477
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7362646460533142,
+      "learning_rate": 4.4404235161153846e-05,
+      "loss": 2.9683,
+      "step": 50478
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7199697494506836,
+      "learning_rate": 4.439617459504959e-05,
+      "loss": 2.852,
+      "step": 50479
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7283482551574707,
+      "learning_rate": 4.43881147021567e-05,
+      "loss": 2.7507,
+      "step": 50480
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7187907695770264,
+      "learning_rate": 4.43800554824963e-05,
+      "loss": 3.07,
+      "step": 50481
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.711829423904419,
+      "learning_rate": 4.4371996936089595e-05,
+      "loss": 2.7205,
+      "step": 50482
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7250786423683167,
+      "learning_rate": 4.436393906295792e-05,
+      "loss": 2.7172,
+      "step": 50483
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8321327567100525,
+      "learning_rate": 4.4355881863122367e-05,
+      "loss": 2.9453,
+      "step": 50484
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7267419099807739,
+      "learning_rate": 4.4347825336604216e-05,
+      "loss": 2.6185,
+      "step": 50485
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7733017802238464,
+      "learning_rate": 4.4339769483424783e-05,
+      "loss": 3.1191,
+      "step": 50486
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.75758957862854,
+      "learning_rate": 4.433171430360508e-05,
+      "loss": 2.7548,
+      "step": 50487
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7199147939682007,
+      "learning_rate": 4.432365979716653e-05,
+      "loss": 2.8798,
+      "step": 50488
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.723355233669281,
+      "learning_rate": 4.431560596413025e-05,
+      "loss": 2.8854,
+      "step": 50489
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7314903140068054,
+      "learning_rate": 4.430755280451739e-05,
+      "loss": 2.8459,
+      "step": 50490
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8310138583183289,
+      "learning_rate": 4.429950031834932e-05,
+      "loss": 3.1573,
+      "step": 50491
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.728308916091919,
+      "learning_rate": 4.4291448505647043e-05,
+      "loss": 3.0318,
+      "step": 50492
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.776800274848938,
+      "learning_rate": 4.4283397366431964e-05,
+      "loss": 2.8144,
+      "step": 50493
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7904463410377502,
+      "learning_rate": 4.427534690072516e-05,
+      "loss": 2.8312,
+      "step": 50494
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.734760046005249,
+      "learning_rate": 4.426729710854793e-05,
+      "loss": 3.018,
+      "step": 50495
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7029054760932922,
+      "learning_rate": 4.425924798992141e-05,
+      "loss": 2.8469,
+      "step": 50496
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7326534390449524,
+      "learning_rate": 4.4251199544866755e-05,
+      "loss": 2.8889,
+      "step": 50497
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7710874676704407,
+      "learning_rate": 4.424315177340528e-05,
+      "loss": 2.9888,
+      "step": 50498
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7983696460723877,
+      "learning_rate": 4.423510467555804e-05,
+      "loss": 3.0389,
+      "step": 50499
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7880064249038696,
+      "learning_rate": 4.4227058251346416e-05,
+      "loss": 2.8939,
+      "step": 50500
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7095093131065369,
+      "learning_rate": 4.421901250079138e-05,
+      "loss": 2.9274,
+      "step": 50501
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.732288658618927,
+      "learning_rate": 4.4210967423914366e-05,
+      "loss": 3.0009,
+      "step": 50502
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7357195615768433,
+      "learning_rate": 4.420292302073638e-05,
+      "loss": 3.0076,
+      "step": 50503
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6846479773521423,
+      "learning_rate": 4.419487929127864e-05,
+      "loss": 2.9429,
+      "step": 50504
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7166612148284912,
+      "learning_rate": 4.41868362355624e-05,
+      "loss": 2.7909,
+      "step": 50505
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7130656242370605,
+      "learning_rate": 4.417879385360876e-05,
+      "loss": 2.7799,
+      "step": 50506
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7403449416160583,
+      "learning_rate": 4.417075214543893e-05,
+      "loss": 3.0642,
+      "step": 50507
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7145542502403259,
+      "learning_rate": 4.416271111107417e-05,
+      "loss": 2.9387,
+      "step": 50508
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7323039174079895,
+      "learning_rate": 4.41546707505355e-05,
+      "loss": 2.8822,
+      "step": 50509
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7286449074745178,
+      "learning_rate": 4.414663106384428e-05,
+      "loss": 3.0583,
+      "step": 50510
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.773391842842102,
+      "learning_rate": 4.4138592051021595e-05,
+      "loss": 2.8229,
+      "step": 50511
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7383174300193787,
+      "learning_rate": 4.413055371208854e-05,
+      "loss": 3.0271,
+      "step": 50512
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7910652160644531,
+      "learning_rate": 4.4122516047066457e-05,
+      "loss": 2.7217,
+      "step": 50513
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.874261200428009,
+      "learning_rate": 4.4114479055976316e-05,
+      "loss": 3.0323,
+      "step": 50514
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7225984930992126,
+      "learning_rate": 4.41064427388395e-05,
+      "loss": 3.0949,
+      "step": 50515
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7266954779624939,
+      "learning_rate": 4.4098407095676956e-05,
+      "loss": 2.8574,
+      "step": 50516
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7139513492584229,
+      "learning_rate": 4.409037212651004e-05,
+      "loss": 2.6369,
+      "step": 50517
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7322560548782349,
+      "learning_rate": 4.408233783135984e-05,
+      "loss": 2.8167,
+      "step": 50518
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7208927869796753,
+      "learning_rate": 4.407430421024746e-05,
+      "loss": 2.8963,
+      "step": 50519
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7421464323997498,
+      "learning_rate": 4.406627126319415e-05,
+      "loss": 2.9859,
+      "step": 50520
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7248895764350891,
+      "learning_rate": 4.4058238990220975e-05,
+      "loss": 2.9614,
+      "step": 50521
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7308017611503601,
+      "learning_rate": 4.405020739134914e-05,
+      "loss": 2.8708,
+      "step": 50522
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7517200708389282,
+      "learning_rate": 4.404217646659985e-05,
+      "loss": 2.8794,
+      "step": 50523
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7376325726509094,
+      "learning_rate": 4.403414621599421e-05,
+      "loss": 2.8069,
+      "step": 50524
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7713701128959656,
+      "learning_rate": 4.4026116639553364e-05,
+      "loss": 2.7789,
+      "step": 50525
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7685970067977905,
+      "learning_rate": 4.401808773729842e-05,
+      "loss": 2.8286,
+      "step": 50526
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.753248393535614,
+      "learning_rate": 4.401005950925054e-05,
+      "loss": 2.903,
+      "step": 50527
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7071647047996521,
+      "learning_rate": 4.4002031955430994e-05,
+      "loss": 2.7273,
+      "step": 50528
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7305782437324524,
+      "learning_rate": 4.399400507586074e-05,
+      "loss": 3.0772,
+      "step": 50529
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7409292459487915,
+      "learning_rate": 4.398597887056109e-05,
+      "loss": 2.8982,
+      "step": 50530
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7586793303489685,
+      "learning_rate": 4.397795333955303e-05,
+      "loss": 3.0227,
+      "step": 50531
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7480147480964661,
+      "learning_rate": 4.396992848285781e-05,
+      "loss": 2.5625,
+      "step": 50532
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7570845484733582,
+      "learning_rate": 4.3961904300496574e-05,
+      "loss": 2.9181,
+      "step": 50533
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7564980983734131,
+      "learning_rate": 4.3953880792490285e-05,
+      "loss": 2.9893,
+      "step": 50534
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7372446060180664,
+      "learning_rate": 4.394585795886031e-05,
+      "loss": 2.7992,
+      "step": 50535
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.789084792137146,
+      "learning_rate": 4.3937835799627577e-05,
+      "loss": 2.991,
+      "step": 50536
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7723736763000488,
+      "learning_rate": 4.392981431481329e-05,
+      "loss": 2.9124,
+      "step": 50537
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7522666454315186,
+      "learning_rate": 4.3921793504438685e-05,
+      "loss": 2.9372,
+      "step": 50538
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.765546977519989,
+      "learning_rate": 4.391377336852476e-05,
+      "loss": 2.8857,
+      "step": 50539
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8160582780838013,
+      "learning_rate": 4.390575390709271e-05,
+      "loss": 3.0308,
+      "step": 50540
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7867605686187744,
+      "learning_rate": 4.3897735120163544e-05,
+      "loss": 2.9236,
+      "step": 50541
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7485517263412476,
+      "learning_rate": 4.3889717007758416e-05,
+      "loss": 3.0409,
+      "step": 50542
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7289761900901794,
+      "learning_rate": 4.388169956989861e-05,
+      "loss": 3.1161,
+      "step": 50543
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7518919110298157,
+      "learning_rate": 4.3873682806605004e-05,
+      "loss": 2.7525,
+      "step": 50544
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.758308470249176,
+      "learning_rate": 4.3865666717898915e-05,
+      "loss": 2.7692,
+      "step": 50545
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7173359394073486,
+      "learning_rate": 4.385765130380137e-05,
+      "loss": 2.7888,
+      "step": 50546
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7279924154281616,
+      "learning_rate": 4.384963656433337e-05,
+      "loss": 2.8149,
+      "step": 50547
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.754176139831543,
+      "learning_rate": 4.384162249951624e-05,
+      "loss": 2.9872,
+      "step": 50548
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7126150727272034,
+      "learning_rate": 4.383360910937089e-05,
+      "loss": 2.9256,
+      "step": 50549
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7775080800056458,
+      "learning_rate": 4.3825596393918586e-05,
+      "loss": 2.9169,
+      "step": 50550
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7692570686340332,
+      "learning_rate": 4.381758435318029e-05,
+      "loss": 2.8355,
+      "step": 50551
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7409884929656982,
+      "learning_rate": 4.380957298717717e-05,
+      "loss": 2.9527,
+      "step": 50552
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7356708645820618,
+      "learning_rate": 4.3801562295930395e-05,
+      "loss": 2.7726,
+      "step": 50553
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7332990169525146,
+      "learning_rate": 4.379355227946102e-05,
+      "loss": 2.7626,
+      "step": 50554
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7492383718490601,
+      "learning_rate": 4.37855429377901e-05,
+      "loss": 2.7641,
+      "step": 50555
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7193925976753235,
+      "learning_rate": 4.377753427093867e-05,
+      "loss": 2.7948,
+      "step": 50556
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7237381339073181,
+      "learning_rate": 4.3769526278927935e-05,
+      "loss": 2.8049,
+      "step": 50557
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7913342118263245,
+      "learning_rate": 4.376151896177903e-05,
+      "loss": 2.9314,
+      "step": 50558
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6835290789604187,
+      "learning_rate": 4.3753512319512876e-05,
+      "loss": 3.2759,
+      "step": 50559
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7239914536476135,
+      "learning_rate": 4.3745506352150715e-05,
+      "loss": 2.7385,
+      "step": 50560
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7153903841972351,
+      "learning_rate": 4.373750105971361e-05,
+      "loss": 3.0137,
+      "step": 50561
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7405564785003662,
+      "learning_rate": 4.37294964422225e-05,
+      "loss": 2.8258,
+      "step": 50562
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7112692594528198,
+      "learning_rate": 4.3721492499698674e-05,
+      "loss": 2.9587,
+      "step": 50563
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7458570003509521,
+      "learning_rate": 4.3713489232163014e-05,
+      "loss": 2.8849,
+      "step": 50564
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7017235159873962,
+      "learning_rate": 4.37054866396368e-05,
+      "loss": 2.9045,
+      "step": 50565
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7555688619613647,
+      "learning_rate": 4.3697484722140917e-05,
+      "loss": 2.9832,
+      "step": 50566
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7248188853263855,
+      "learning_rate": 4.368948347969655e-05,
+      "loss": 2.955,
+      "step": 50567
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7523334622383118,
+      "learning_rate": 4.368148291232492e-05,
+      "loss": 2.9627,
+      "step": 50568
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7347005009651184,
+      "learning_rate": 4.3673483020046764e-05,
+      "loss": 3.0475,
+      "step": 50569
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7141953110694885,
+      "learning_rate": 4.366548380288337e-05,
+      "loss": 2.8616,
+      "step": 50570
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7329511046409607,
+      "learning_rate": 4.365748526085573e-05,
+      "loss": 2.9409,
+      "step": 50571
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7447880506515503,
+      "learning_rate": 4.364948739398498e-05,
+      "loss": 2.8136,
+      "step": 50572
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7083245515823364,
+      "learning_rate": 4.3641490202292086e-05,
+      "loss": 2.616,
+      "step": 50573
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7109681367874146,
+      "learning_rate": 4.363349368579818e-05,
+      "loss": 3.0928,
+      "step": 50574
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8247482180595398,
+      "learning_rate": 4.362549784452436e-05,
+      "loss": 2.8209,
+      "step": 50575
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7620562314987183,
+      "learning_rate": 4.3617502678491666e-05,
+      "loss": 3.0653,
+      "step": 50576
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7446070313453674,
+      "learning_rate": 4.360950818772112e-05,
+      "loss": 2.795,
+      "step": 50577
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7512825727462769,
+      "learning_rate": 4.36015143722337e-05,
+      "loss": 2.978,
+      "step": 50578
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7243783473968506,
+      "learning_rate": 4.3593521232050564e-05,
+      "loss": 3.0133,
+      "step": 50579
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7371456027030945,
+      "learning_rate": 4.358552876719282e-05,
+      "loss": 2.8551,
+      "step": 50580
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7351969480514526,
+      "learning_rate": 4.357753697768136e-05,
+      "loss": 3.0309,
+      "step": 50581
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7740240693092346,
+      "learning_rate": 4.356954586353739e-05,
+      "loss": 2.8982,
+      "step": 50582
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7251624464988708,
+      "learning_rate": 4.356155542478187e-05,
+      "loss": 2.8026,
+      "step": 50583
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.734462559223175,
+      "learning_rate": 4.3553565661435804e-05,
+      "loss": 2.9494,
+      "step": 50584
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7287883758544922,
+      "learning_rate": 4.354557657352039e-05,
+      "loss": 2.782,
+      "step": 50585
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7678456902503967,
+      "learning_rate": 4.3537588161056466e-05,
+      "loss": 2.7814,
+      "step": 50586
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.722035825252533,
+      "learning_rate": 4.352960042406526e-05,
+      "loss": 2.8701,
+      "step": 50587
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7033929824829102,
+      "learning_rate": 4.352161336256768e-05,
+      "loss": 2.9547,
+      "step": 50588
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7072238326072693,
+      "learning_rate": 4.351362697658477e-05,
+      "loss": 2.9044,
+      "step": 50589
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7303126454353333,
+      "learning_rate": 4.350564126613769e-05,
+      "loss": 2.7136,
+      "step": 50590
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7483873963356018,
+      "learning_rate": 4.349765623124739e-05,
+      "loss": 2.8554,
+      "step": 50591
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.722395658493042,
+      "learning_rate": 4.3489671871934915e-05,
+      "loss": 2.8171,
+      "step": 50592
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.71732497215271,
+      "learning_rate": 4.3481688188221185e-05,
+      "loss": 2.9246,
+      "step": 50593
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7207403182983398,
+      "learning_rate": 4.347370518012731e-05,
+      "loss": 3.0359,
+      "step": 50594
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7453340291976929,
+      "learning_rate": 4.346572284767442e-05,
+      "loss": 2.9454,
+      "step": 50595
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7322558164596558,
+      "learning_rate": 4.345774119088339e-05,
+      "loss": 3.0738,
+      "step": 50596
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7502358555793762,
+      "learning_rate": 4.344976020977533e-05,
+      "loss": 2.9023,
+      "step": 50597
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7384856939315796,
+      "learning_rate": 4.3441779904371234e-05,
+      "loss": 2.8681,
+      "step": 50598
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7567711472511292,
+      "learning_rate": 4.343380027469208e-05,
+      "loss": 3.0825,
+      "step": 50599
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7449147701263428,
+      "learning_rate": 4.342582132075895e-05,
+      "loss": 2.8255,
+      "step": 50600
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7349591255187988,
+      "learning_rate": 4.341784304259276e-05,
+      "loss": 2.8695,
+      "step": 50601
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7392820119857788,
+      "learning_rate": 4.340986544021466e-05,
+      "loss": 3.0477,
+      "step": 50602
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7246425747871399,
+      "learning_rate": 4.3401888513645545e-05,
+      "loss": 2.7809,
+      "step": 50603
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7254135608673096,
+      "learning_rate": 4.339391226290655e-05,
+      "loss": 2.8379,
+      "step": 50604
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7831563353538513,
+      "learning_rate": 4.3385936688018564e-05,
+      "loss": 2.6841,
+      "step": 50605
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7377023100852966,
+      "learning_rate": 4.33779617890026e-05,
+      "loss": 2.8574,
+      "step": 50606
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7295764088630676,
+      "learning_rate": 4.336998756587975e-05,
+      "loss": 2.7794,
+      "step": 50607
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7548102736473083,
+      "learning_rate": 4.336201401867089e-05,
+      "loss": 2.8338,
+      "step": 50608
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7400689721107483,
+      "learning_rate": 4.33540411473971e-05,
+      "loss": 3.0272,
+      "step": 50609
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7324529886245728,
+      "learning_rate": 4.334606895207944e-05,
+      "loss": 2.8919,
+      "step": 50610
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7176007628440857,
+      "learning_rate": 4.333809743273878e-05,
+      "loss": 2.9282,
+      "step": 50611
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7532137632369995,
+      "learning_rate": 4.333012658939625e-05,
+      "loss": 2.8807,
+      "step": 50612
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7730857133865356,
+      "learning_rate": 4.332215642207276e-05,
+      "loss": 2.9552,
+      "step": 50613
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7338061332702637,
+      "learning_rate": 4.3314186930789226e-05,
+      "loss": 2.8655,
+      "step": 50614
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7536339163780212,
+      "learning_rate": 4.33062181155668e-05,
+      "loss": 2.9174,
+      "step": 50615
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7680534720420837,
+      "learning_rate": 4.3298249976426336e-05,
+      "loss": 2.8697,
+      "step": 50616
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7779417634010315,
+      "learning_rate": 4.329028251338897e-05,
+      "loss": 2.6885,
+      "step": 50617
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7388231754302979,
+      "learning_rate": 4.3282315726475505e-05,
+      "loss": 2.8754,
+      "step": 50618
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7245052456855774,
+      "learning_rate": 4.327434961570707e-05,
+      "loss": 2.9758,
+      "step": 50619
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7409709692001343,
+      "learning_rate": 4.3266384181104595e-05,
+      "loss": 2.9113,
+      "step": 50620
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7566462755203247,
+      "learning_rate": 4.3258419422688995e-05,
+      "loss": 3.0348,
+      "step": 50621
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7550995349884033,
+      "learning_rate": 4.325045534048135e-05,
+      "loss": 2.9427,
+      "step": 50622
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6849644184112549,
+      "learning_rate": 4.324249193450254e-05,
+      "loss": 2.708,
+      "step": 50623
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7459685206413269,
+      "learning_rate": 4.3234529204773624e-05,
+      "loss": 2.933,
+      "step": 50624
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.753688633441925,
+      "learning_rate": 4.3226567151315585e-05,
+      "loss": 2.771,
+      "step": 50625
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7236641645431519,
+      "learning_rate": 4.321860577414933e-05,
+      "loss": 2.9226,
+      "step": 50626
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7358285188674927,
+      "learning_rate": 4.321064507329588e-05,
+      "loss": 2.8284,
+      "step": 50627
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.769458532333374,
+      "learning_rate": 4.3202685048776085e-05,
+      "loss": 2.8737,
+      "step": 50628
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.721544086933136,
+      "learning_rate": 4.3194725700611e-05,
+      "loss": 2.9255,
+      "step": 50629
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7584481239318848,
+      "learning_rate": 4.3186767028821666e-05,
+      "loss": 2.8509,
+      "step": 50630
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8179458975791931,
+      "learning_rate": 4.317880903342887e-05,
+      "loss": 2.7925,
+      "step": 50631
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7520755529403687,
+      "learning_rate": 4.317085171445377e-05,
+      "loss": 2.8761,
+      "step": 50632
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7519469261169434,
+      "learning_rate": 4.31628950719171e-05,
+      "loss": 2.8604,
+      "step": 50633
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7431280016899109,
+      "learning_rate": 4.3154939105840056e-05,
+      "loss": 2.6998,
+      "step": 50634
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7596693634986877,
+      "learning_rate": 4.314698381624342e-05,
+      "loss": 3.019,
+      "step": 50635
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6919360756874084,
+      "learning_rate": 4.313902920314817e-05,
+      "loss": 2.8299,
+      "step": 50636
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7272117733955383,
+      "learning_rate": 4.313107526657532e-05,
+      "loss": 3.0207,
+      "step": 50637
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7418824434280396,
+      "learning_rate": 4.312312200654573e-05,
+      "loss": 2.9491,
+      "step": 50638
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.759490966796875,
+      "learning_rate": 4.3115169423080406e-05,
+      "loss": 2.7287,
+      "step": 50639
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7105665802955627,
+      "learning_rate": 4.310721751620034e-05,
+      "loss": 2.8817,
+      "step": 50640
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6908241510391235,
+      "learning_rate": 4.309926628592645e-05,
+      "loss": 2.9844,
+      "step": 50641
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7713239789009094,
+      "learning_rate": 4.309131573227964e-05,
+      "loss": 2.9163,
+      "step": 50642
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7205983400344849,
+      "learning_rate": 4.308336585528077e-05,
+      "loss": 2.8435,
+      "step": 50643
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7385979294776917,
+      "learning_rate": 4.307541665495089e-05,
+      "loss": 2.8786,
+      "step": 50644
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.753772497177124,
+      "learning_rate": 4.306746813131098e-05,
+      "loss": 2.8506,
+      "step": 50645
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7099587321281433,
+      "learning_rate": 4.305952028438186e-05,
+      "loss": 2.9241,
+      "step": 50646
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7350571155548096,
+      "learning_rate": 4.305157311418457e-05,
+      "loss": 3.0212,
+      "step": 50647
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7650042176246643,
+      "learning_rate": 4.304362662073997e-05,
+      "loss": 3.1281,
+      "step": 50648
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.761756181716919,
+      "learning_rate": 4.303568080406901e-05,
+      "loss": 2.9067,
+      "step": 50649
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7702816724777222,
+      "learning_rate": 4.302773566419256e-05,
+      "loss": 3.1063,
+      "step": 50650
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7798253893852234,
+      "learning_rate": 4.301979120113159e-05,
+      "loss": 2.8491,
+      "step": 50651
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7569634318351746,
+      "learning_rate": 4.301184741490711e-05,
+      "loss": 2.9733,
+      "step": 50652
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7624911665916443,
+      "learning_rate": 4.3003904305539896e-05,
+      "loss": 3.0518,
+      "step": 50653
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7225568294525146,
+      "learning_rate": 4.2995961873051e-05,
+      "loss": 2.8497,
+      "step": 50654
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6899399757385254,
+      "learning_rate": 4.298802011746121e-05,
+      "loss": 2.7573,
+      "step": 50655
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7465593218803406,
+      "learning_rate": 4.298007903879157e-05,
+      "loss": 2.9092,
+      "step": 50656
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7861872315406799,
+      "learning_rate": 4.297213863706294e-05,
+      "loss": 2.9574,
+      "step": 50657
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7941358685493469,
+      "learning_rate": 4.296419891229619e-05,
+      "loss": 3.0628,
+      "step": 50658
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7803980112075806,
+      "learning_rate": 4.295625986451231e-05,
+      "loss": 2.8888,
+      "step": 50659
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7576730847358704,
+      "learning_rate": 4.294832149373212e-05,
+      "loss": 2.9702,
+      "step": 50660
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6931632161140442,
+      "learning_rate": 4.294038379997656e-05,
+      "loss": 2.9009,
+      "step": 50661
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7237907648086548,
+      "learning_rate": 4.293244678326666e-05,
+      "loss": 3.0382,
+      "step": 50662
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7643592953681946,
+      "learning_rate": 4.292451044362319e-05,
+      "loss": 3.0474,
+      "step": 50663
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7706605195999146,
+      "learning_rate": 4.2916574781067116e-05,
+      "loss": 2.8414,
+      "step": 50664
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7509461045265198,
+      "learning_rate": 4.2908639795619204e-05,
+      "loss": 2.8932,
+      "step": 50665
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7358827590942383,
+      "learning_rate": 4.290070548730048e-05,
+      "loss": 2.8797,
+      "step": 50666
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7475077509880066,
+      "learning_rate": 4.289277185613189e-05,
+      "loss": 2.8333,
+      "step": 50667
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7222772240638733,
+      "learning_rate": 4.288483890213416e-05,
+      "loss": 2.7275,
+      "step": 50668
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7573405504226685,
+      "learning_rate": 4.287690662532839e-05,
+      "loss": 2.9758,
+      "step": 50669
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7598530054092407,
+      "learning_rate": 4.286897502573534e-05,
+      "loss": 3.048,
+      "step": 50670
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7809165716171265,
+      "learning_rate": 4.286104410337586e-05,
+      "loss": 3.0113,
+      "step": 50671
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7310870885848999,
+      "learning_rate": 4.2853113858270974e-05,
+      "loss": 2.9604,
+      "step": 50672
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7275199294090271,
+      "learning_rate": 4.2845184290441415e-05,
+      "loss": 2.9085,
+      "step": 50673
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7422723174095154,
+      "learning_rate": 4.283725539990821e-05,
+      "loss": 2.789,
+      "step": 50674
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7516480088233948,
+      "learning_rate": 4.282932718669214e-05,
+      "loss": 2.8959,
+      "step": 50675
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7623327374458313,
+      "learning_rate": 4.2821399650814126e-05,
+      "loss": 2.8064,
+      "step": 50676
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7669196724891663,
+      "learning_rate": 4.281347279229511e-05,
+      "loss": 2.8621,
+      "step": 50677
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.726617157459259,
+      "learning_rate": 4.2805546611155916e-05,
+      "loss": 2.616,
+      "step": 50678
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7164513468742371,
+      "learning_rate": 4.279762110741742e-05,
+      "loss": 2.7553,
+      "step": 50679
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7366183400154114,
+      "learning_rate": 4.2789696281100417e-05,
+      "loss": 3.0931,
+      "step": 50680
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7067437171936035,
+      "learning_rate": 4.278177213222584e-05,
+      "loss": 2.9771,
+      "step": 50681
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8154669404029846,
+      "learning_rate": 4.2773848660814656e-05,
+      "loss": 2.7634,
+      "step": 50682
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7058912515640259,
+      "learning_rate": 4.2765925866887574e-05,
+      "loss": 2.9507,
+      "step": 50683
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7300341129302979,
+      "learning_rate": 4.275800375046561e-05,
+      "loss": 3.0809,
+      "step": 50684
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7966801524162292,
+      "learning_rate": 4.275008231156954e-05,
+      "loss": 2.898,
+      "step": 50685
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7171053290367126,
+      "learning_rate": 4.27421615502202e-05,
+      "loss": 3.0457,
+      "step": 50686
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7475465536117554,
+      "learning_rate": 4.2734241466438554e-05,
+      "loss": 2.9942,
+      "step": 50687
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.738580584526062,
+      "learning_rate": 4.272632206024533e-05,
+      "loss": 3.0752,
+      "step": 50688
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7871452569961548,
+      "learning_rate": 4.2718403331661547e-05,
+      "loss": 3.007,
+      "step": 50689
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.745241105556488,
+      "learning_rate": 4.2710485280707873e-05,
+      "loss": 2.8626,
+      "step": 50690
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7506486773490906,
+      "learning_rate": 4.2702567907405305e-05,
+      "loss": 2.9813,
+      "step": 50691
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7289913296699524,
+      "learning_rate": 4.269465121177469e-05,
+      "loss": 2.7847,
+      "step": 50692
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7110291123390198,
+      "learning_rate": 4.268673519383687e-05,
+      "loss": 3.036,
+      "step": 50693
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.773071825504303,
+      "learning_rate": 4.2678819853612646e-05,
+      "loss": 2.926,
+      "step": 50694
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7590256333351135,
+      "learning_rate": 4.267090519112285e-05,
+      "loss": 2.9054,
+      "step": 50695
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7588886022567749,
+      "learning_rate": 4.266299120638834e-05,
+      "loss": 2.8203,
+      "step": 50696
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7508629560470581,
+      "learning_rate": 4.265507789943007e-05,
+      "loss": 2.8658,
+      "step": 50697
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7774369120597839,
+      "learning_rate": 4.264716527026871e-05,
+      "loss": 3.0411,
+      "step": 50698
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7264341115951538,
+      "learning_rate": 4.263925331892526e-05,
+      "loss": 2.9546,
+      "step": 50699
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7322808504104614,
+      "learning_rate": 4.2631342045420523e-05,
+      "loss": 2.8151,
+      "step": 50700
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7444933652877808,
+      "learning_rate": 4.2623431449775194e-05,
+      "loss": 2.7928,
+      "step": 50701
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7420282959938049,
+      "learning_rate": 4.261552153201028e-05,
+      "loss": 2.8152,
+      "step": 50702
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7798295021057129,
+      "learning_rate": 4.2607612292146496e-05,
+      "loss": 2.8232,
+      "step": 50703
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7416540384292603,
+      "learning_rate": 4.259970373020477e-05,
+      "loss": 2.9673,
+      "step": 50704
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6930776238441467,
+      "learning_rate": 4.2591795846205844e-05,
+      "loss": 2.8342,
+      "step": 50705
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7511429786682129,
+      "learning_rate": 4.2583888640170674e-05,
+      "loss": 2.8983,
+      "step": 50706
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7975016236305237,
+      "learning_rate": 4.257598211211997e-05,
+      "loss": 2.861,
+      "step": 50707
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7396607995033264,
+      "learning_rate": 4.256807626207451e-05,
+      "loss": 2.6752,
+      "step": 50708
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7415522336959839,
+      "learning_rate": 4.256017109005526e-05,
+      "loss": 2.7366,
+      "step": 50709
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7048690915107727,
+      "learning_rate": 4.2552266596082926e-05,
+      "loss": 2.9888,
+      "step": 50710
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.787742555141449,
+      "learning_rate": 4.2544362780178366e-05,
+      "loss": 2.7388,
+      "step": 50711
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8210556507110596,
+      "learning_rate": 4.253645964236249e-05,
+      "loss": 2.6602,
+      "step": 50712
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7322614789009094,
+      "learning_rate": 4.252855718265594e-05,
+      "loss": 2.8288,
+      "step": 50713
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7463746070861816,
+      "learning_rate": 4.252065540107972e-05,
+      "loss": 2.9142,
+      "step": 50714
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7110394835472107,
+      "learning_rate": 4.25127542976545e-05,
+      "loss": 2.7829,
+      "step": 50715
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7307591438293457,
+      "learning_rate": 4.250485387240109e-05,
+      "loss": 3.1279,
+      "step": 50716
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7428984045982361,
+      "learning_rate": 4.249695412534039e-05,
+      "loss": 2.9116,
+      "step": 50717
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7223466634750366,
+      "learning_rate": 4.24890550564931e-05,
+      "loss": 2.9612,
+      "step": 50718
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7388598322868347,
+      "learning_rate": 4.2481156665880154e-05,
+      "loss": 3.0812,
+      "step": 50719
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7969038486480713,
+      "learning_rate": 4.247325895352219e-05,
+      "loss": 2.5446,
+      "step": 50720
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7594483494758606,
+      "learning_rate": 4.2465361919440165e-05,
+      "loss": 2.7923,
+      "step": 50721
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7183117270469666,
+      "learning_rate": 4.2457465563654824e-05,
+      "loss": 2.7408,
+      "step": 50722
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7206244468688965,
+      "learning_rate": 4.244956988618692e-05,
+      "loss": 3.0161,
+      "step": 50723
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.781730592250824,
+      "learning_rate": 4.24416748870573e-05,
+      "loss": 2.9401,
+      "step": 50724
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7326677441596985,
+      "learning_rate": 4.243378056628668e-05,
+      "loss": 2.9432,
+      "step": 50725
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7561776041984558,
+      "learning_rate": 4.242588692389595e-05,
+      "loss": 2.7459,
+      "step": 50726
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7506656646728516,
+      "learning_rate": 4.241799395990588e-05,
+      "loss": 2.7,
+      "step": 50727
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7250350713729858,
+      "learning_rate": 4.241010167433727e-05,
+      "loss": 2.8735,
+      "step": 50728
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7116358876228333,
+      "learning_rate": 4.24022100672109e-05,
+      "loss": 2.829,
+      "step": 50729
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7642216682434082,
+      "learning_rate": 4.239431913854744e-05,
+      "loss": 2.8395,
+      "step": 50730
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7662079334259033,
+      "learning_rate": 4.238642888836782e-05,
+      "loss": 2.9226,
+      "step": 50731
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7560353875160217,
+      "learning_rate": 4.237853931669272e-05,
+      "loss": 2.8221,
+      "step": 50732
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7635805010795593,
+      "learning_rate": 4.237065042354295e-05,
+      "loss": 2.9538,
+      "step": 50733
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.750133752822876,
+      "learning_rate": 4.236276220893937e-05,
+      "loss": 3.0837,
+      "step": 50734
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7176488041877747,
+      "learning_rate": 4.2354874672902664e-05,
+      "loss": 2.914,
+      "step": 50735
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7769431471824646,
+      "learning_rate": 4.234698781545367e-05,
+      "loss": 3.0444,
+      "step": 50736
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7291898727416992,
+      "learning_rate": 4.2339101636613116e-05,
+      "loss": 2.7426,
+      "step": 50737
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7305982112884521,
+      "learning_rate": 4.2331216136401716e-05,
+      "loss": 2.8365,
+      "step": 50738
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7421076893806458,
+      "learning_rate": 4.232333131484038e-05,
+      "loss": 2.9467,
+      "step": 50739
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7870148420333862,
+      "learning_rate": 4.2315447171949735e-05,
+      "loss": 2.9623,
+      "step": 50740
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7365769147872925,
+      "learning_rate": 4.230756370775066e-05,
+      "loss": 3.0139,
+      "step": 50741
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7033193707466125,
+      "learning_rate": 4.2299680922263835e-05,
+      "loss": 2.8967,
+      "step": 50742
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7226850390434265,
+      "learning_rate": 4.229179881551008e-05,
+      "loss": 2.9138,
+      "step": 50743
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.763777494430542,
+      "learning_rate": 4.228391738751015e-05,
+      "loss": 2.7401,
+      "step": 50744
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7105336785316467,
+      "learning_rate": 4.227603663828472e-05,
+      "loss": 3.0466,
+      "step": 50745
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7253413200378418,
+      "learning_rate": 4.226815656785467e-05,
+      "loss": 2.976,
+      "step": 50746
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7211529612541199,
+      "learning_rate": 4.226027717624063e-05,
+      "loss": 2.8548,
+      "step": 50747
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7225125432014465,
+      "learning_rate": 4.225239846346341e-05,
+      "loss": 2.841,
+      "step": 50748
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7354315519332886,
+      "learning_rate": 4.224452042954386e-05,
+      "loss": 2.6947,
+      "step": 50749
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7653220891952515,
+      "learning_rate": 4.223664307450263e-05,
+      "loss": 2.9488,
+      "step": 50750
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7226547598838806,
+      "learning_rate": 4.2228766398360434e-05,
+      "loss": 2.9566,
+      "step": 50751
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7502910494804382,
+      "learning_rate": 4.222089040113803e-05,
+      "loss": 2.8887,
+      "step": 50752
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7303741574287415,
+      "learning_rate": 4.221301508285619e-05,
+      "loss": 3.1261,
+      "step": 50753
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7288039326667786,
+      "learning_rate": 4.2205140443535714e-05,
+      "loss": 2.8225,
+      "step": 50754
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7314246296882629,
+      "learning_rate": 4.21972664831972e-05,
+      "loss": 2.9364,
+      "step": 50755
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7128660678863525,
+      "learning_rate": 4.218939320186158e-05,
+      "loss": 2.7165,
+      "step": 50756
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7295336127281189,
+      "learning_rate": 4.218152059954939e-05,
+      "loss": 2.7854,
+      "step": 50757
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7478935122489929,
+      "learning_rate": 4.2173648676281524e-05,
+      "loss": 2.8468,
+      "step": 50758
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7409915328025818,
+      "learning_rate": 4.216577743207863e-05,
+      "loss": 2.678,
+      "step": 50759
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.742350161075592,
+      "learning_rate": 4.2157906866961425e-05,
+      "loss": 2.8212,
+      "step": 50760
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7968260645866394,
+      "learning_rate": 4.215003698095072e-05,
+      "loss": 2.9907,
+      "step": 50761
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.712583601474762,
+      "learning_rate": 4.2142167774067104e-05,
+      "loss": 3.0309,
+      "step": 50762
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7087548971176147,
+      "learning_rate": 4.213429924633143e-05,
+      "loss": 2.7685,
+      "step": 50763
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7460187673568726,
+      "learning_rate": 4.212643139776444e-05,
+      "loss": 2.9495,
+      "step": 50764
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8229163885116577,
+      "learning_rate": 4.211856422838679e-05,
+      "loss": 2.9066,
+      "step": 50765
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7447080612182617,
+      "learning_rate": 4.211069773821922e-05,
+      "loss": 3.06,
+      "step": 50766
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.712235689163208,
+      "learning_rate": 4.2102831927282364e-05,
+      "loss": 2.8577,
+      "step": 50767
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7422438859939575,
+      "learning_rate": 4.209496679559703e-05,
+      "loss": 2.9714,
+      "step": 50768
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7196982502937317,
+      "learning_rate": 4.2087102343183995e-05,
+      "loss": 2.5705,
+      "step": 50769
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7336692214012146,
+      "learning_rate": 4.207923857006379e-05,
+      "loss": 2.9921,
+      "step": 50770
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8158669471740723,
+      "learning_rate": 4.2071375476257316e-05,
+      "loss": 2.8008,
+      "step": 50771
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7556344270706177,
+      "learning_rate": 4.2063513061785236e-05,
+      "loss": 2.9787,
+      "step": 50772
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7272917628288269,
+      "learning_rate": 4.2055651326668126e-05,
+      "loss": 2.8565,
+      "step": 50773
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7603312730789185,
+      "learning_rate": 4.204779027092684e-05,
+      "loss": 2.9942,
+      "step": 50774
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.728800892829895,
+      "learning_rate": 4.203992989458199e-05,
+      "loss": 2.9436,
+      "step": 50775
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7117544412612915,
+      "learning_rate": 4.2032070197654364e-05,
+      "loss": 2.9086,
+      "step": 50776
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7658325433731079,
+      "learning_rate": 4.2024211180164545e-05,
+      "loss": 2.9249,
+      "step": 50777
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6998578906059265,
+      "learning_rate": 4.201635284213335e-05,
+      "loss": 2.9473,
+      "step": 50778
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7814579606056213,
+      "learning_rate": 4.200849518358146e-05,
+      "loss": 2.9595,
+      "step": 50779
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7035899758338928,
+      "learning_rate": 4.200063820452956e-05,
+      "loss": 2.8746,
+      "step": 50780
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7148592472076416,
+      "learning_rate": 4.1992781904998295e-05,
+      "loss": 2.9529,
+      "step": 50781
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7354755401611328,
+      "learning_rate": 4.1984926285008355e-05,
+      "loss": 2.8767,
+      "step": 50782
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.723908543586731,
+      "learning_rate": 4.197707134458046e-05,
+      "loss": 2.9528,
+      "step": 50783
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7492967844009399,
+      "learning_rate": 4.196921708373535e-05,
+      "loss": 3.0366,
+      "step": 50784
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7645527720451355,
+      "learning_rate": 4.1961363502493615e-05,
+      "loss": 2.9825,
+      "step": 50785
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7469053268432617,
+      "learning_rate": 4.1953510600876036e-05,
+      "loss": 2.858,
+      "step": 50786
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7710843682289124,
+      "learning_rate": 4.194565837890327e-05,
+      "loss": 2.8289,
+      "step": 50787
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7673295140266418,
+      "learning_rate": 4.193780683659589e-05,
+      "loss": 3.0426,
+      "step": 50788
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7491339445114136,
+      "learning_rate": 4.192995597397475e-05,
+      "loss": 3.077,
+      "step": 50789
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.756820023059845,
+      "learning_rate": 4.192210579106037e-05,
+      "loss": 3.0091,
+      "step": 50790
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7435303926467896,
+      "learning_rate": 4.1914256287873525e-05,
+      "loss": 2.8404,
+      "step": 50791
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7386417388916016,
+      "learning_rate": 4.190640746443484e-05,
+      "loss": 2.8662,
+      "step": 50792
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7664363980293274,
+      "learning_rate": 4.189855932076497e-05,
+      "loss": 3.0039,
+      "step": 50793
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7032740116119385,
+      "learning_rate": 4.1890711856884786e-05,
+      "loss": 2.7883,
+      "step": 50794
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7009146809577942,
+      "learning_rate": 4.188286507281464e-05,
+      "loss": 2.8372,
+      "step": 50795
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7520615458488464,
+      "learning_rate": 4.187501896857542e-05,
+      "loss": 2.9782,
+      "step": 50796
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.726274847984314,
+      "learning_rate": 4.1867173544187685e-05,
+      "loss": 2.7904,
+      "step": 50797
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7836074233055115,
+      "learning_rate": 4.18593287996721e-05,
+      "loss": 3.0073,
+      "step": 50798
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7328587770462036,
+      "learning_rate": 4.1851484735049425e-05,
+      "loss": 2.7654,
+      "step": 50799
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7041208744049072,
+      "learning_rate": 4.1843641350340205e-05,
+      "loss": 3.1183,
+      "step": 50800
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7885406017303467,
+      "learning_rate": 4.183579864556523e-05,
+      "loss": 3.0173,
+      "step": 50801
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7367026805877686,
+      "learning_rate": 4.182795662074505e-05,
+      "loss": 2.8478,
+      "step": 50802
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7012369632720947,
+      "learning_rate": 4.182011527590031e-05,
+      "loss": 2.9023,
+      "step": 50803
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8011801838874817,
+      "learning_rate": 4.1812274611051765e-05,
+      "loss": 2.8189,
+      "step": 50804
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7242926359176636,
+      "learning_rate": 4.1804434626219896e-05,
+      "loss": 2.8217,
+      "step": 50805
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6962056159973145,
+      "learning_rate": 4.179659532142555e-05,
+      "loss": 2.7622,
+      "step": 50806
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.728546142578125,
+      "learning_rate": 4.1788756696689216e-05,
+      "loss": 3.0023,
+      "step": 50807
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7366439700126648,
+      "learning_rate": 4.1780918752031644e-05,
+      "loss": 2.9501,
+      "step": 50808
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7231490612030029,
+      "learning_rate": 4.177308148747345e-05,
+      "loss": 2.9292,
+      "step": 50809
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7084015011787415,
+      "learning_rate": 4.176524490303521e-05,
+      "loss": 2.782,
+      "step": 50810
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7416843175888062,
+      "learning_rate": 4.175740899873768e-05,
+      "loss": 2.898,
+      "step": 50811
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7785187363624573,
+      "learning_rate": 4.1749573774601354e-05,
+      "loss": 3.0888,
+      "step": 50812
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7746663689613342,
+      "learning_rate": 4.174173923064703e-05,
+      "loss": 2.8559,
+      "step": 50813
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7415159344673157,
+      "learning_rate": 4.173390536689517e-05,
+      "loss": 2.9723,
+      "step": 50814
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.2300888299942017,
+      "learning_rate": 4.172607218336652e-05,
+      "loss": 2.9255,
+      "step": 50815
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7681926488876343,
+      "learning_rate": 4.171823968008177e-05,
+      "loss": 2.791,
+      "step": 50816
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7152865529060364,
+      "learning_rate": 4.171040785706146e-05,
+      "loss": 2.6684,
+      "step": 50817
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7356628179550171,
+      "learning_rate": 4.1702576714326205e-05,
+      "loss": 2.873,
+      "step": 50818
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7573530673980713,
+      "learning_rate": 4.1694746251896606e-05,
+      "loss": 2.8615,
+      "step": 50819
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7217455506324768,
+      "learning_rate": 4.168691646979333e-05,
+      "loss": 2.9256,
+      "step": 50820
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7954509854316711,
+      "learning_rate": 4.1679087368037066e-05,
+      "loss": 2.9018,
+      "step": 50821
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7794967293739319,
+      "learning_rate": 4.16712589466483e-05,
+      "loss": 2.9567,
+      "step": 50822
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.719433605670929,
+      "learning_rate": 4.166343120564781e-05,
+      "loss": 2.7478,
+      "step": 50823
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7607887387275696,
+      "learning_rate": 4.1655604145056085e-05,
+      "loss": 2.9903,
+      "step": 50824
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7895941734313965,
+      "learning_rate": 4.1647777764893745e-05,
+      "loss": 2.8344,
+      "step": 50825
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7224075794219971,
+      "learning_rate": 4.163995206518149e-05,
+      "loss": 3.0332,
+      "step": 50826
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7128440737724304,
+      "learning_rate": 4.1632127045939826e-05,
+      "loss": 2.6762,
+      "step": 50827
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7446962594985962,
+      "learning_rate": 4.162430270718948e-05,
+      "loss": 2.9153,
+      "step": 50828
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7283123731613159,
+      "learning_rate": 4.161647904895092e-05,
+      "loss": 2.8714,
+      "step": 50829
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7481231093406677,
+      "learning_rate": 4.1608656071244897e-05,
+      "loss": 2.9547,
+      "step": 50830
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7285983562469482,
+      "learning_rate": 4.1600833774091945e-05,
+      "loss": 2.8458,
+      "step": 50831
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7572833299636841,
+      "learning_rate": 4.159301215751259e-05,
+      "loss": 2.9878,
+      "step": 50832
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.748000979423523,
+      "learning_rate": 4.1585191221527595e-05,
+      "loss": 2.9091,
+      "step": 50833
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7040547132492065,
+      "learning_rate": 4.157737096615742e-05,
+      "loss": 2.8364,
+      "step": 50834
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7504925727844238,
+      "learning_rate": 4.1569551391422695e-05,
+      "loss": 2.8602,
+      "step": 50835
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7645800113677979,
+      "learning_rate": 4.15617324973441e-05,
+      "loss": 2.9762,
+      "step": 50836
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7223212122917175,
+      "learning_rate": 4.1553914283942094e-05,
+      "loss": 3.0629,
+      "step": 50837
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7501276135444641,
+      "learning_rate": 4.154609675123741e-05,
+      "loss": 2.8408,
+      "step": 50838
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7350544929504395,
+      "learning_rate": 4.153827989925055e-05,
+      "loss": 2.9287,
+      "step": 50839
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7663587331771851,
+      "learning_rate": 4.153046372800208e-05,
+      "loss": 2.9415,
+      "step": 50840
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8215747475624084,
+      "learning_rate": 4.1522648237512666e-05,
+      "loss": 2.8608,
+      "step": 50841
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7529155015945435,
+      "learning_rate": 4.151483342780282e-05,
+      "loss": 2.9685,
+      "step": 50842
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7362189888954163,
+      "learning_rate": 4.1507019298893186e-05,
+      "loss": 2.8494,
+      "step": 50843
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7676265835762024,
+      "learning_rate": 4.149920585080429e-05,
+      "loss": 2.8628,
+      "step": 50844
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7176138758659363,
+      "learning_rate": 4.1491393083556766e-05,
+      "loss": 3.0375,
+      "step": 50845
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7533155083656311,
+      "learning_rate": 4.1483580997171186e-05,
+      "loss": 2.8913,
+      "step": 50846
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7027636170387268,
+      "learning_rate": 4.147576959166802e-05,
+      "loss": 2.9655,
+      "step": 50847
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7691273093223572,
+      "learning_rate": 4.146795886706798e-05,
+      "loss": 2.9198,
+      "step": 50848
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7978218793869019,
+      "learning_rate": 4.146014882339154e-05,
+      "loss": 2.9319,
+      "step": 50849
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7753750085830688,
+      "learning_rate": 4.145233946065929e-05,
+      "loss": 2.8632,
+      "step": 50850
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7168684601783752,
+      "learning_rate": 4.144453077889189e-05,
+      "loss": 3.0417,
+      "step": 50851
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.730107307434082,
+      "learning_rate": 4.143672277810983e-05,
+      "loss": 2.9331,
+      "step": 50852
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7644616365432739,
+      "learning_rate": 4.142891545833369e-05,
+      "loss": 2.7204,
+      "step": 50853
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7631624937057495,
+      "learning_rate": 4.142110881958398e-05,
+      "loss": 3.1044,
+      "step": 50854
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7302795648574829,
+      "learning_rate": 4.1413302861881256e-05,
+      "loss": 2.8666,
+      "step": 50855
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7175984382629395,
+      "learning_rate": 4.140549758524624e-05,
+      "loss": 2.8372,
+      "step": 50856
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7876809239387512,
+      "learning_rate": 4.139769298969927e-05,
+      "loss": 2.9315,
+      "step": 50857
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7984057068824768,
+      "learning_rate": 4.1389889075261105e-05,
+      "loss": 2.756,
+      "step": 50858
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8823890686035156,
+      "learning_rate": 4.138208584195213e-05,
+      "loss": 2.8131,
+      "step": 50859
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7620226144790649,
+      "learning_rate": 4.137428328979302e-05,
+      "loss": 3.0526,
+      "step": 50860
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7054426670074463,
+      "learning_rate": 4.1366481418804266e-05,
+      "loss": 3.0109,
+      "step": 50861
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7472173571586609,
+      "learning_rate": 4.135868022900638e-05,
+      "loss": 2.9188,
+      "step": 50862
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7145726084709167,
+      "learning_rate": 4.1350879720420014e-05,
+      "loss": 2.9618,
+      "step": 50863
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7612037062644958,
+      "learning_rate": 4.1343079893065586e-05,
+      "loss": 2.781,
+      "step": 50864
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7213705778121948,
+      "learning_rate": 4.133528074696372e-05,
+      "loss": 2.8342,
+      "step": 50865
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8010908365249634,
+      "learning_rate": 4.132748228213498e-05,
+      "loss": 2.7692,
+      "step": 50866
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7797710299491882,
+      "learning_rate": 4.1319684498599906e-05,
+      "loss": 3.0265,
+      "step": 50867
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7466559410095215,
+      "learning_rate": 4.131188739637896e-05,
+      "loss": 2.8454,
+      "step": 50868
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7062391638755798,
+      "learning_rate": 4.130409097549268e-05,
+      "loss": 2.9389,
+      "step": 50869
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7370114326477051,
+      "learning_rate": 4.1296295235961627e-05,
+      "loss": 2.7059,
+      "step": 50870
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7448911070823669,
+      "learning_rate": 4.128850017780641e-05,
+      "loss": 3.0902,
+      "step": 50871
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7794041633605957,
+      "learning_rate": 4.128070580104742e-05,
+      "loss": 3.0043,
+      "step": 50872
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7436469197273254,
+      "learning_rate": 4.127291210570536e-05,
+      "loss": 2.7554,
+      "step": 50873
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8578379154205322,
+      "learning_rate": 4.12651190918006e-05,
+      "loss": 2.766,
+      "step": 50874
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7621721029281616,
+      "learning_rate": 4.12573267593537e-05,
+      "loss": 2.7681,
+      "step": 50875
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6907230019569397,
+      "learning_rate": 4.1249535108385246e-05,
+      "loss": 2.9351,
+      "step": 50876
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7239318490028381,
+      "learning_rate": 4.1241744138915676e-05,
+      "loss": 2.8717,
+      "step": 50877
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7087956070899963,
+      "learning_rate": 4.1233953850965616e-05,
+      "loss": 2.7214,
+      "step": 50878
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7738022804260254,
+      "learning_rate": 4.122616424455545e-05,
+      "loss": 2.937,
+      "step": 50879
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7342925071716309,
+      "learning_rate": 4.121837531970575e-05,
+      "loss": 2.8623,
+      "step": 50880
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7471789121627808,
+      "learning_rate": 4.121058707643712e-05,
+      "loss": 2.9401,
+      "step": 50881
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7654462456703186,
+      "learning_rate": 4.120279951476999e-05,
+      "loss": 2.8794,
+      "step": 50882
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6907823085784912,
+      "learning_rate": 4.119501263472489e-05,
+      "loss": 2.8218,
+      "step": 50883
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7497276067733765,
+      "learning_rate": 4.1187226436322265e-05,
+      "loss": 2.9644,
+      "step": 50884
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6869165301322937,
+      "learning_rate": 4.117944091958273e-05,
+      "loss": 3.0914,
+      "step": 50885
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7467650175094604,
+      "learning_rate": 4.117165608452667e-05,
+      "loss": 2.7657,
+      "step": 50886
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7433682084083557,
+      "learning_rate": 4.1163871931174674e-05,
+      "loss": 2.8362,
+      "step": 50887
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7507684826850891,
+      "learning_rate": 4.1156088459547254e-05,
+      "loss": 2.918,
+      "step": 50888
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7188553810119629,
+      "learning_rate": 4.114830566966489e-05,
+      "loss": 2.9672,
+      "step": 50889
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7091402411460876,
+      "learning_rate": 4.114052356154811e-05,
+      "loss": 2.8625,
+      "step": 50890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7486855983734131,
+      "learning_rate": 4.113274213521726e-05,
+      "loss": 2.8445,
+      "step": 50891
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7630000114440918,
+      "learning_rate": 4.112496139069298e-05,
+      "loss": 2.8836,
+      "step": 50892
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8397104740142822,
+      "learning_rate": 4.11171813279958e-05,
+      "loss": 2.9828,
+      "step": 50893
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7878146767616272,
+      "learning_rate": 4.110940194714603e-05,
+      "loss": 3.0389,
+      "step": 50894
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.744600236415863,
+      "learning_rate": 4.1101623248164386e-05,
+      "loss": 2.859,
+      "step": 50895
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7375232577323914,
+      "learning_rate": 4.109384523107122e-05,
+      "loss": 3.0262,
+      "step": 50896
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7445172071456909,
+      "learning_rate": 4.108606789588695e-05,
+      "loss": 2.8573,
+      "step": 50897
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7916731834411621,
+      "learning_rate": 4.107829124263227e-05,
+      "loss": 2.5277,
+      "step": 50898
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7740074396133423,
+      "learning_rate": 4.107051527132741e-05,
+      "loss": 2.8962,
+      "step": 50899
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7415305376052856,
+      "learning_rate": 4.10627399819931e-05,
+      "loss": 3.0435,
+      "step": 50900
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7645403146743774,
+      "learning_rate": 4.105496537464958e-05,
+      "loss": 3.0199,
+      "step": 50901
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7634096741676331,
+      "learning_rate": 4.104719144931749e-05,
+      "loss": 2.9406,
+      "step": 50902
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7687584161758423,
+      "learning_rate": 4.103941820601731e-05,
+      "loss": 2.9617,
+      "step": 50903
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7485208511352539,
+      "learning_rate": 4.1031645644769494e-05,
+      "loss": 2.8768,
+      "step": 50904
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7247251868247986,
+      "learning_rate": 4.102387376559444e-05,
+      "loss": 2.923,
+      "step": 50905
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8152522444725037,
+      "learning_rate": 4.101610256851263e-05,
+      "loss": 2.8562,
+      "step": 50906
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7078531384468079,
+      "learning_rate": 4.100833205354452e-05,
+      "loss": 3.0635,
+      "step": 50907
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7330393195152283,
+      "learning_rate": 4.100056222071073e-05,
+      "loss": 2.6665,
+      "step": 50908
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7819045186042786,
+      "learning_rate": 4.099279307003154e-05,
+      "loss": 3.0064,
+      "step": 50909
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8102908730506897,
+      "learning_rate": 4.098502460152753e-05,
+      "loss": 3.0131,
+      "step": 50910
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7458033561706543,
+      "learning_rate": 4.097725681521912e-05,
+      "loss": 2.8156,
+      "step": 50911
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.750956654548645,
+      "learning_rate": 4.0969489711126726e-05,
+      "loss": 3.1724,
+      "step": 50912
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7348149418830872,
+      "learning_rate": 4.09617232892709e-05,
+      "loss": 2.8127,
+      "step": 50913
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7550942897796631,
+      "learning_rate": 4.095395754967196e-05,
+      "loss": 2.8972,
+      "step": 50914
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7148580551147461,
+      "learning_rate": 4.0946192492350514e-05,
+      "loss": 2.8255,
+      "step": 50915
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7225046157836914,
+      "learning_rate": 4.0938428117326895e-05,
+      "loss": 2.7282,
+      "step": 50916
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7666391134262085,
+      "learning_rate": 4.093066442462157e-05,
+      "loss": 2.7273,
+      "step": 50917
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.746823787689209,
+      "learning_rate": 4.092290141425511e-05,
+      "loss": 2.9481,
+      "step": 50918
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7512428164482117,
+      "learning_rate": 4.091513908624784e-05,
+      "loss": 2.8239,
+      "step": 50919
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7290512919425964,
+      "learning_rate": 4.090737744062023e-05,
+      "loss": 2.7134,
+      "step": 50920
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7385398149490356,
+      "learning_rate": 4.089961647739268e-05,
+      "loss": 3.15,
+      "step": 50921
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7204491496086121,
+      "learning_rate": 4.089185619658566e-05,
+      "loss": 2.7842,
+      "step": 50922
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7641138434410095,
+      "learning_rate": 4.088409659821971e-05,
+      "loss": 2.9864,
+      "step": 50923
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7317721843719482,
+      "learning_rate": 4.087633768231511e-05,
+      "loss": 2.883,
+      "step": 50924
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7656816244125366,
+      "learning_rate": 4.0868579448892415e-05,
+      "loss": 2.813,
+      "step": 50925
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7583822011947632,
+      "learning_rate": 4.086082189797204e-05,
+      "loss": 2.9214,
+      "step": 50926
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7554618716239929,
+      "learning_rate": 4.085306502957427e-05,
+      "loss": 2.8041,
+      "step": 50927
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7482122778892517,
+      "learning_rate": 4.0845308843719756e-05,
+      "loss": 2.8939,
+      "step": 50928
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8274185657501221,
+      "learning_rate": 4.0837553340428784e-05,
+      "loss": 2.7364,
+      "step": 50929
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7026564478874207,
+      "learning_rate": 4.082979851972183e-05,
+      "loss": 2.7929,
+      "step": 50930
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8461477160453796,
+      "learning_rate": 4.082204438161929e-05,
+      "loss": 2.9104,
+      "step": 50931
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7340132594108582,
+      "learning_rate": 4.0814290926141636e-05,
+      "loss": 2.814,
+      "step": 50932
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7171696424484253,
+      "learning_rate": 4.0806538153309284e-05,
+      "loss": 2.979,
+      "step": 50933
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7311767935752869,
+      "learning_rate": 4.079878606314253e-05,
+      "loss": 2.9879,
+      "step": 50934
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7532653212547302,
+      "learning_rate": 4.079103465566198e-05,
+      "loss": 2.8224,
+      "step": 50935
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7105627059936523,
+      "learning_rate": 4.078328393088789e-05,
+      "loss": 2.803,
+      "step": 50936
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7665572166442871,
+      "learning_rate": 4.077553388884074e-05,
+      "loss": 3.1236,
+      "step": 50937
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7285357713699341,
+      "learning_rate": 4.076778452954101e-05,
+      "loss": 2.8497,
+      "step": 50938
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7463434338569641,
+      "learning_rate": 4.0760035853009e-05,
+      "loss": 2.943,
+      "step": 50939
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7962722778320312,
+      "learning_rate": 4.075228785926521e-05,
+      "loss": 3.0252,
+      "step": 50940
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7439616322517395,
+      "learning_rate": 4.0744540548329996e-05,
+      "loss": 2.759,
+      "step": 50941
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7312211990356445,
+      "learning_rate": 4.073679392022371e-05,
+      "loss": 2.6245,
+      "step": 50942
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7195311188697815,
+      "learning_rate": 4.072904797496686e-05,
+      "loss": 3.0517,
+      "step": 50943
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6905331015586853,
+      "learning_rate": 4.072130271257975e-05,
+      "loss": 2.9166,
+      "step": 50944
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7037015557289124,
+      "learning_rate": 4.071355813308291e-05,
+      "loss": 2.9205,
+      "step": 50945
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7677614688873291,
+      "learning_rate": 4.070581423649657e-05,
+      "loss": 2.7968,
+      "step": 50946
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7849667072296143,
+      "learning_rate": 4.069807102284127e-05,
+      "loss": 2.8588,
+      "step": 50947
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7227040529251099,
+      "learning_rate": 4.0690328492137366e-05,
+      "loss": 2.7918,
+      "step": 50948
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8007916212081909,
+      "learning_rate": 4.068258664440518e-05,
+      "loss": 3.1101,
+      "step": 50949
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7214465141296387,
+      "learning_rate": 4.0674845479665185e-05,
+      "loss": 2.6496,
+      "step": 50950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7386294603347778,
+      "learning_rate": 4.066710499793771e-05,
+      "loss": 2.728,
+      "step": 50951
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7470434904098511,
+      "learning_rate": 4.065936519924315e-05,
+      "loss": 2.9676,
+      "step": 50952
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7561085820198059,
+      "learning_rate": 4.065162608360198e-05,
+      "loss": 2.8916,
+      "step": 50953
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7485124468803406,
+      "learning_rate": 4.0643887651034514e-05,
+      "loss": 2.6446,
+      "step": 50954
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7447924017906189,
+      "learning_rate": 4.063614990156112e-05,
+      "loss": 2.8693,
+      "step": 50955
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.729457437992096,
+      "learning_rate": 4.062841283520213e-05,
+      "loss": 2.7354,
+      "step": 50956
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7371611595153809,
+      "learning_rate": 4.062067645197802e-05,
+      "loss": 3.0109,
+      "step": 50957
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7636083960533142,
+      "learning_rate": 4.061294075190914e-05,
+      "loss": 2.6421,
+      "step": 50958
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8881843090057373,
+      "learning_rate": 4.060520573501581e-05,
+      "loss": 2.8784,
+      "step": 50959
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7290064692497253,
+      "learning_rate": 4.059747140131855e-05,
+      "loss": 3.0156,
+      "step": 50960
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7382261157035828,
+      "learning_rate": 4.0589737750837506e-05,
+      "loss": 2.779,
+      "step": 50961
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7628456354141235,
+      "learning_rate": 4.058200478359326e-05,
+      "loss": 2.9542,
+      "step": 50962
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7667835354804993,
+      "learning_rate": 4.05742724996061e-05,
+      "loss": 3.0102,
+      "step": 50963
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7994858026504517,
+      "learning_rate": 4.0566540898896275e-05,
+      "loss": 2.8177,
+      "step": 50964
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7050232887268066,
+      "learning_rate": 4.0558809981484364e-05,
+      "loss": 2.9967,
+      "step": 50965
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7449204325675964,
+      "learning_rate": 4.055107974739053e-05,
+      "loss": 2.8754,
+      "step": 50966
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7724292874336243,
+      "learning_rate": 4.054335019663527e-05,
+      "loss": 2.6654,
+      "step": 50967
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7795135378837585,
+      "learning_rate": 4.0535621329238857e-05,
+      "loss": 2.7718,
+      "step": 50968
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6958447098731995,
+      "learning_rate": 4.052789314522172e-05,
+      "loss": 2.8641,
+      "step": 50969
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7487135529518127,
+      "learning_rate": 4.05201656446042e-05,
+      "loss": 2.8888,
+      "step": 50970
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7429445385932922,
+      "learning_rate": 4.051243882740652e-05,
+      "loss": 2.8184,
+      "step": 50971
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7123727798461914,
+      "learning_rate": 4.050471269364926e-05,
+      "loss": 3.1401,
+      "step": 50972
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7284667491912842,
+      "learning_rate": 4.049698724335254e-05,
+      "loss": 2.7846,
+      "step": 50973
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7588906288146973,
+      "learning_rate": 4.04892624765368e-05,
+      "loss": 2.6116,
+      "step": 50974
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7456479072570801,
+      "learning_rate": 4.0481538393222516e-05,
+      "loss": 2.9805,
+      "step": 50975
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7160006761550903,
+      "learning_rate": 4.047381499342986e-05,
+      "loss": 2.7069,
+      "step": 50976
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7530381083488464,
+      "learning_rate": 4.0466092277179274e-05,
+      "loss": 2.8844,
+      "step": 50977
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7419160604476929,
+      "learning_rate": 4.0458370244490954e-05,
+      "loss": 2.5117,
+      "step": 50978
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7657301425933838,
+      "learning_rate": 4.045064889538536e-05,
+      "loss": 2.5939,
+      "step": 50979
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7660251259803772,
+      "learning_rate": 4.044292822988284e-05,
+      "loss": 2.7927,
+      "step": 50980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8668144941329956,
+      "learning_rate": 4.043520824800368e-05,
+      "loss": 2.9519,
+      "step": 50981
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7728575468063354,
+      "learning_rate": 4.042748894976826e-05,
+      "loss": 3.0094,
+      "step": 50982
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7155167460441589,
+      "learning_rate": 4.04197703351968e-05,
+      "loss": 2.6734,
+      "step": 50983
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7189810276031494,
+      "learning_rate": 4.041205240430981e-05,
+      "loss": 2.9716,
+      "step": 50984
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.728173553943634,
+      "learning_rate": 4.040433515712749e-05,
+      "loss": 2.8533,
+      "step": 50985
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7641592025756836,
+      "learning_rate": 4.039661859367014e-05,
+      "loss": 3.3096,
+      "step": 50986
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7122716903686523,
+      "learning_rate": 4.0388902713958184e-05,
+      "loss": 2.7429,
+      "step": 50987
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7600113153457642,
+      "learning_rate": 4.038118751801184e-05,
+      "loss": 2.947,
+      "step": 50988
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7478308081626892,
+      "learning_rate": 4.0373473005851454e-05,
+      "loss": 2.8651,
+      "step": 50989
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7460908889770508,
+      "learning_rate": 4.036575917749748e-05,
+      "loss": 2.9738,
+      "step": 50990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7135355472564697,
+      "learning_rate": 4.03580460329701e-05,
+      "loss": 2.7612,
+      "step": 50991
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.743721604347229,
+      "learning_rate": 4.0350333572289675e-05,
+      "loss": 2.7419,
+      "step": 50992
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.737909197807312,
+      "learning_rate": 4.034262179547644e-05,
+      "loss": 2.8635,
+      "step": 50993
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7796975374221802,
+      "learning_rate": 4.033491070255075e-05,
+      "loss": 2.8157,
+      "step": 50994
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7635996341705322,
+      "learning_rate": 4.032720029353302e-05,
+      "loss": 2.777,
+      "step": 50995
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7595275640487671,
+      "learning_rate": 4.031949056844338e-05,
+      "loss": 2.8294,
+      "step": 50996
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7468879222869873,
+      "learning_rate": 4.03117815273023e-05,
+      "loss": 2.8661,
+      "step": 50997
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7194596529006958,
+      "learning_rate": 4.0304073170130036e-05,
+      "loss": 3.0833,
+      "step": 50998
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7229351997375488,
+      "learning_rate": 4.029636549694677e-05,
+      "loss": 3.0329,
+      "step": 50999
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7696372270584106,
+      "learning_rate": 4.028865850777295e-05,
+      "loss": 3.0217,
+      "step": 51000
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.730970561504364,
+      "learning_rate": 4.028095220262877e-05,
+      "loss": 2.8736,
+      "step": 51001
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7224681973457336,
+      "learning_rate": 4.0273246581534636e-05,
+      "loss": 2.857,
+      "step": 51002
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7172154188156128,
+      "learning_rate": 4.026554164451073e-05,
+      "loss": 2.9542,
+      "step": 51003
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7283074259757996,
+      "learning_rate": 4.025783739157738e-05,
+      "loss": 2.8613,
+      "step": 51004
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7916626930236816,
+      "learning_rate": 4.0250133822755004e-05,
+      "loss": 3.0261,
+      "step": 51005
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.751471757888794,
+      "learning_rate": 4.0242430938063776e-05,
+      "loss": 2.9181,
+      "step": 51006
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7159141898155212,
+      "learning_rate": 4.023472873752398e-05,
+      "loss": 2.8968,
+      "step": 51007
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.769489049911499,
+      "learning_rate": 4.022702722115584e-05,
+      "loss": 2.9992,
+      "step": 51008
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7926540970802307,
+      "learning_rate": 4.0219326388979736e-05,
+      "loss": 2.8284,
+      "step": 51009
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7663864493370056,
+      "learning_rate": 4.021162624101598e-05,
+      "loss": 2.996,
+      "step": 51010
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8369364142417908,
+      "learning_rate": 4.020392677728473e-05,
+      "loss": 2.7815,
+      "step": 51011
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7207963466644287,
+      "learning_rate": 4.019622799780644e-05,
+      "loss": 3.0512,
+      "step": 51012
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.826167106628418,
+      "learning_rate": 4.018852990260125e-05,
+      "loss": 2.9079,
+      "step": 51013
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7820410132408142,
+      "learning_rate": 4.018083249168942e-05,
+      "loss": 3.0379,
+      "step": 51014
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7185248136520386,
+      "learning_rate": 4.0173135765091325e-05,
+      "loss": 2.8409,
+      "step": 51015
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7619501948356628,
+      "learning_rate": 4.016543972282715e-05,
+      "loss": 2.7795,
+      "step": 51016
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7843639254570007,
+      "learning_rate": 4.0157744364917254e-05,
+      "loss": 2.994,
+      "step": 51017
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7171763777732849,
+      "learning_rate": 4.015004969138178e-05,
+      "loss": 2.8102,
+      "step": 51018
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7367191910743713,
+      "learning_rate": 4.014235570224105e-05,
+      "loss": 2.9537,
+      "step": 51019
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7253427505493164,
+      "learning_rate": 4.013466239751544e-05,
+      "loss": 2.9189,
+      "step": 51020
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7575175166130066,
+      "learning_rate": 4.0126969777225105e-05,
+      "loss": 2.8425,
+      "step": 51021
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7350121140480042,
+      "learning_rate": 4.0119277841390296e-05,
+      "loss": 2.9173,
+      "step": 51022
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7644230723381042,
+      "learning_rate": 4.0111586590031264e-05,
+      "loss": 2.844,
+      "step": 51023
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7468816041946411,
+      "learning_rate": 4.010389602316829e-05,
+      "loss": 3.0427,
+      "step": 51024
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7907638549804688,
+      "learning_rate": 4.0096206140821695e-05,
+      "loss": 2.8312,
+      "step": 51025
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7258133888244629,
+      "learning_rate": 4.00885169430116e-05,
+      "loss": 2.8525,
+      "step": 51026
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7306505441665649,
+      "learning_rate": 4.00808284297584e-05,
+      "loss": 2.9856,
+      "step": 51027
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7286311388015747,
+      "learning_rate": 4.007314060108228e-05,
+      "loss": 3.0388,
+      "step": 51028
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7742034792900085,
+      "learning_rate": 4.006545345700343e-05,
+      "loss": 2.9805,
+      "step": 51029
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7489369511604309,
+      "learning_rate": 4.005776699754217e-05,
+      "loss": 2.8488,
+      "step": 51030
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7013357877731323,
+      "learning_rate": 4.00500812227187e-05,
+      "loss": 2.9688,
+      "step": 51031
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7368605136871338,
+      "learning_rate": 4.0042396132553355e-05,
+      "loss": 2.7995,
+      "step": 51032
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7846304178237915,
+      "learning_rate": 4.003471172706623e-05,
+      "loss": 2.9969,
+      "step": 51033
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7510572075843811,
+      "learning_rate": 4.00270280062777e-05,
+      "loss": 2.8879,
+      "step": 51034
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7406590580940247,
+      "learning_rate": 4.0019344970207976e-05,
+      "loss": 2.8439,
+      "step": 51035
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7419261336326599,
+      "learning_rate": 4.001166261887716e-05,
+      "loss": 2.7751,
+      "step": 51036
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7146857976913452,
+      "learning_rate": 4.000398095230569e-05,
+      "loss": 3.0476,
+      "step": 51037
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7753526568412781,
+      "learning_rate": 3.999629997051359e-05,
+      "loss": 2.6755,
+      "step": 51038
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7410012483596802,
+      "learning_rate": 3.998861967352131e-05,
+      "loss": 2.756,
+      "step": 51039
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7166079878807068,
+      "learning_rate": 3.998094006134885e-05,
+      "loss": 2.9648,
+      "step": 51040
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7411850094795227,
+      "learning_rate": 3.997326113401659e-05,
+      "loss": 2.8957,
+      "step": 51041
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7261947989463806,
+      "learning_rate": 3.996558289154479e-05,
+      "loss": 2.9451,
+      "step": 51042
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7302013635635376,
+      "learning_rate": 3.9957905333953574e-05,
+      "loss": 3.0935,
+      "step": 51043
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7443523406982422,
+      "learning_rate": 3.99502284612632e-05,
+      "loss": 3.1287,
+      "step": 51044
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7418161630630493,
+      "learning_rate": 3.994255227349381e-05,
+      "loss": 2.9928,
+      "step": 51045
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7650214433670044,
+      "learning_rate": 3.9934876770665694e-05,
+      "loss": 2.4832,
+      "step": 51046
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7540137767791748,
+      "learning_rate": 3.9927201952799094e-05,
+      "loss": 2.8556,
+      "step": 51047
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7868987321853638,
+      "learning_rate": 3.991952781991417e-05,
+      "loss": 3.0147,
+      "step": 51048
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6718930006027222,
+      "learning_rate": 3.9911854372031196e-05,
+      "loss": 2.7966,
+      "step": 51049
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.722568690776825,
+      "learning_rate": 3.990418160917036e-05,
+      "loss": 2.8758,
+      "step": 51050
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7367124557495117,
+      "learning_rate": 3.989650953135178e-05,
+      "loss": 2.8975,
+      "step": 51051
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7118638157844543,
+      "learning_rate": 3.988883813859578e-05,
+      "loss": 2.8865,
+      "step": 51052
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7357069849967957,
+      "learning_rate": 3.98811674309225e-05,
+      "loss": 3.1935,
+      "step": 51053
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7383196949958801,
+      "learning_rate": 3.9873497408352194e-05,
+      "loss": 3.1177,
+      "step": 51054
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7383975982666016,
+      "learning_rate": 3.986582807090498e-05,
+      "loss": 2.8114,
+      "step": 51055
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7209492325782776,
+      "learning_rate": 3.9858159418601174e-05,
+      "loss": 2.8342,
+      "step": 51056
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7390090823173523,
+      "learning_rate": 3.98504914514609e-05,
+      "loss": 2.7422,
+      "step": 51057
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7286245226860046,
+      "learning_rate": 3.984282416950433e-05,
+      "loss": 2.655,
+      "step": 51058
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7232625484466553,
+      "learning_rate": 3.983515757275172e-05,
+      "loss": 3.0018,
+      "step": 51059
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7462209463119507,
+      "learning_rate": 3.9827491661223185e-05,
+      "loss": 2.7941,
+      "step": 51060
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7977537512779236,
+      "learning_rate": 3.981982643493898e-05,
+      "loss": 2.9345,
+      "step": 51061
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7521597743034363,
+      "learning_rate": 3.9812161893919356e-05,
+      "loss": 2.7811,
+      "step": 51062
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7260125279426575,
+      "learning_rate": 3.980449803818433e-05,
+      "loss": 2.8128,
+      "step": 51063
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7464675307273865,
+      "learning_rate": 3.9796834867754244e-05,
+      "loss": 3.1766,
+      "step": 51064
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.71709144115448,
+      "learning_rate": 3.9789172382649226e-05,
+      "loss": 2.724,
+      "step": 51065
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7842869162559509,
+      "learning_rate": 3.978151058288939e-05,
+      "loss": 2.8037,
+      "step": 51066
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7320528626441956,
+      "learning_rate": 3.9773849468495044e-05,
+      "loss": 2.8024,
+      "step": 51067
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7437505125999451,
+      "learning_rate": 3.976618903948622e-05,
+      "loss": 2.9678,
+      "step": 51068
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7431596517562866,
+      "learning_rate": 3.9758529295883226e-05,
+      "loss": 2.9204,
+      "step": 51069
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7559123635292053,
+      "learning_rate": 3.9750870237706146e-05,
+      "loss": 2.9171,
+      "step": 51070
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7823855876922607,
+      "learning_rate": 3.974321186497523e-05,
+      "loss": 2.9853,
+      "step": 51071
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7621415853500366,
+      "learning_rate": 3.97355541777106e-05,
+      "loss": 2.9533,
+      "step": 51072
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7120107412338257,
+      "learning_rate": 3.9727897175932377e-05,
+      "loss": 2.8572,
+      "step": 51073
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7418361306190491,
+      "learning_rate": 3.972024085966087e-05,
+      "loss": 2.9235,
+      "step": 51074
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7263070940971375,
+      "learning_rate": 3.9712585228916064e-05,
+      "loss": 2.7967,
+      "step": 51075
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7198815941810608,
+      "learning_rate": 3.970493028371824e-05,
+      "loss": 3.0169,
+      "step": 51076
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7565320134162903,
+      "learning_rate": 3.9697276024087555e-05,
+      "loss": 2.8488,
+      "step": 51077
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7032371759414673,
+      "learning_rate": 3.968962245004419e-05,
+      "loss": 3.0858,
+      "step": 51078
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7568981051445007,
+      "learning_rate": 3.968196956160826e-05,
+      "loss": 3.0574,
+      "step": 51079
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7816786170005798,
+      "learning_rate": 3.967431735879982e-05,
+      "loss": 2.8361,
+      "step": 51080
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7889500260353088,
+      "learning_rate": 3.966666584163916e-05,
+      "loss": 2.8982,
+      "step": 51081
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8262700438499451,
+      "learning_rate": 3.965901501014648e-05,
+      "loss": 2.9173,
+      "step": 51082
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7360769510269165,
+      "learning_rate": 3.965136486434174e-05,
+      "loss": 2.8857,
+      "step": 51083
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7041099071502686,
+      "learning_rate": 3.964371540424529e-05,
+      "loss": 2.8569,
+      "step": 51084
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7271643280982971,
+      "learning_rate": 3.963606662987712e-05,
+      "loss": 2.841,
+      "step": 51085
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7795175313949585,
+      "learning_rate": 3.9628418541257534e-05,
+      "loss": 2.9807,
+      "step": 51086
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7626325488090515,
+      "learning_rate": 3.962077113840656e-05,
+      "loss": 2.9521,
+      "step": 51087
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7495712637901306,
+      "learning_rate": 3.9613124421344275e-05,
+      "loss": 2.8875,
+      "step": 51088
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.960547839009101e-05,
+      "loss": 2.8907,
+      "step": 51089
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7435591816902161,
+      "learning_rate": 3.959783304466673e-05,
+      "loss": 2.9217,
+      "step": 51090
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7713919878005981,
+      "learning_rate": 3.959018838509162e-05,
+      "loss": 2.7192,
+      "step": 51091
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7549655437469482,
+      "learning_rate": 3.958254441138592e-05,
+      "loss": 2.9587,
+      "step": 51092
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7387896776199341,
+      "learning_rate": 3.9574901123569655e-05,
+      "loss": 2.9832,
+      "step": 51093
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7287462949752808,
+      "learning_rate": 3.956725852166302e-05,
+      "loss": 2.7358,
+      "step": 51094
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7182101011276245,
+      "learning_rate": 3.9559616605686e-05,
+      "loss": 2.9946,
+      "step": 51095
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7613502740859985,
+      "learning_rate": 3.955197537565885e-05,
+      "loss": 2.8062,
+      "step": 51096
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7491579055786133,
+      "learning_rate": 3.954433483160174e-05,
+      "loss": 2.9498,
+      "step": 51097
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7342544794082642,
+      "learning_rate": 3.953669497353467e-05,
+      "loss": 2.814,
+      "step": 51098
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7425305843353271,
+      "learning_rate": 3.952905580147787e-05,
+      "loss": 2.8258,
+      "step": 51099
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.73570716381073,
+      "learning_rate": 3.952141731545141e-05,
+      "loss": 2.7309,
+      "step": 51100
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7669528722763062,
+      "learning_rate": 3.951377951547533e-05,
+      "loss": 2.9622,
+      "step": 51101
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8032728433609009,
+      "learning_rate": 3.950614240156992e-05,
+      "loss": 2.6455,
+      "step": 51102
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7362860441207886,
+      "learning_rate": 3.949850597375509e-05,
+      "loss": 2.8745,
+      "step": 51103
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7232388854026794,
+      "learning_rate": 3.9490870232051164e-05,
+      "loss": 2.8219,
+      "step": 51104
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7235947847366333,
+      "learning_rate": 3.948323517647806e-05,
+      "loss": 2.9228,
+      "step": 51105
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7498955726623535,
+      "learning_rate": 3.9475600807055986e-05,
+      "loss": 3.1506,
+      "step": 51106
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7524616122245789,
+      "learning_rate": 3.94679671238051e-05,
+      "loss": 2.9171,
+      "step": 51107
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.707608163356781,
+      "learning_rate": 3.9460334126745455e-05,
+      "loss": 2.8342,
+      "step": 51108
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7663477659225464,
+      "learning_rate": 3.945270181589716e-05,
+      "loss": 2.8276,
+      "step": 51109
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7328577041625977,
+      "learning_rate": 3.94450701912802e-05,
+      "loss": 2.7313,
+      "step": 51110
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7494615912437439,
+      "learning_rate": 3.9437439252914805e-05,
+      "loss": 2.828,
+      "step": 51111
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7223064303398132,
+      "learning_rate": 3.9429809000821145e-05,
+      "loss": 2.8811,
+      "step": 51112
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.746543288230896,
+      "learning_rate": 3.9422179435019106e-05,
+      "loss": 2.8726,
+      "step": 51113
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7731015682220459,
+      "learning_rate": 3.941455055552898e-05,
+      "loss": 2.7396,
+      "step": 51114
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7490095496177673,
+      "learning_rate": 3.9406922362370775e-05,
+      "loss": 2.8321,
+      "step": 51115
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7323755621910095,
+      "learning_rate": 3.9399294855564545e-05,
+      "loss": 2.8937,
+      "step": 51116
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7344057559967041,
+      "learning_rate": 3.9391668035130444e-05,
+      "loss": 2.7432,
+      "step": 51117
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.744623601436615,
+      "learning_rate": 3.938404190108848e-05,
+      "loss": 2.8554,
+      "step": 51118
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7350231409072876,
+      "learning_rate": 3.937641645345888e-05,
+      "loss": 2.9645,
+      "step": 51119
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7532517910003662,
+      "learning_rate": 3.936879169226156e-05,
+      "loss": 2.8945,
+      "step": 51120
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7580457329750061,
+      "learning_rate": 3.936116761751676e-05,
+      "loss": 2.9476,
+      "step": 51121
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7381550073623657,
+      "learning_rate": 3.935354422924445e-05,
+      "loss": 2.6959,
+      "step": 51122
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7573136687278748,
+      "learning_rate": 3.93459215274647e-05,
+      "loss": 2.9742,
+      "step": 51123
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7315176129341125,
+      "learning_rate": 3.933829951219769e-05,
+      "loss": 3.0369,
+      "step": 51124
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8132463693618774,
+      "learning_rate": 3.9330678183463385e-05,
+      "loss": 2.9298,
+      "step": 51125
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7474046945571899,
+      "learning_rate": 3.932305754128196e-05,
+      "loss": 2.8276,
+      "step": 51126
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7960267663002014,
+      "learning_rate": 3.9315437585673356e-05,
+      "loss": 2.8031,
+      "step": 51127
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7314231395721436,
+      "learning_rate": 3.9307818316657744e-05,
+      "loss": 2.7175,
+      "step": 51128
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7347849011421204,
+      "learning_rate": 3.930019973425523e-05,
+      "loss": 2.9909,
+      "step": 51129
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7521358728408813,
+      "learning_rate": 3.92925818384858e-05,
+      "loss": 2.9316,
+      "step": 51130
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7546024322509766,
+      "learning_rate": 3.9284964629369545e-05,
+      "loss": 3.0779,
+      "step": 51131
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7242218255996704,
+      "learning_rate": 3.927734810692643e-05,
+      "loss": 3.0528,
+      "step": 51132
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7637183666229248,
+      "learning_rate": 3.9269732271176624e-05,
+      "loss": 2.8383,
+      "step": 51133
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7487627267837524,
+      "learning_rate": 3.926211712214024e-05,
+      "loss": 3.0692,
+      "step": 51134
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7518205642700195,
+      "learning_rate": 3.925450265983718e-05,
+      "loss": 2.7788,
+      "step": 51135
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7417492270469666,
+      "learning_rate": 3.924688888428765e-05,
+      "loss": 2.9289,
+      "step": 51136
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7587374448776245,
+      "learning_rate": 3.923927579551163e-05,
+      "loss": 3.0257,
+      "step": 51137
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7488877177238464,
+      "learning_rate": 3.923166339352912e-05,
+      "loss": 2.7045,
+      "step": 51138
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.743396520614624,
+      "learning_rate": 3.9224051678360256e-05,
+      "loss": 2.9128,
+      "step": 51139
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.739387571811676,
+      "learning_rate": 3.921644065002502e-05,
+      "loss": 2.8156,
+      "step": 51140
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.771299421787262,
+      "learning_rate": 3.920883030854355e-05,
+      "loss": 2.7683,
+      "step": 51141
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7252259850502014,
+      "learning_rate": 3.9201220653935735e-05,
+      "loss": 2.9843,
+      "step": 51142
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7235883474349976,
+      "learning_rate": 3.919361168622174e-05,
+      "loss": 2.9401,
+      "step": 51143
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.73777174949646,
+      "learning_rate": 3.918600340542163e-05,
+      "loss": 2.9255,
+      "step": 51144
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7344832420349121,
+      "learning_rate": 3.9178395811555396e-05,
+      "loss": 2.7266,
+      "step": 51145
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7370460033416748,
+      "learning_rate": 3.917078890464309e-05,
+      "loss": 2.7727,
+      "step": 51146
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7238647937774658,
+      "learning_rate": 3.916318268470466e-05,
+      "loss": 2.9441,
+      "step": 51147
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7683873176574707,
+      "learning_rate": 3.9155577151760195e-05,
+      "loss": 2.7424,
+      "step": 51148
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7333675026893616,
+      "learning_rate": 3.914797230582981e-05,
+      "loss": 2.9641,
+      "step": 51149
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.789627730846405,
+      "learning_rate": 3.914036814693339e-05,
+      "loss": 2.6447,
+      "step": 51150
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7240229845046997,
+      "learning_rate": 3.913276467509111e-05,
+      "loss": 2.7824,
+      "step": 51151
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7708064317703247,
+      "learning_rate": 3.91251618903229e-05,
+      "loss": 2.9563,
+      "step": 51152
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7075161337852478,
+      "learning_rate": 3.9117559792648764e-05,
+      "loss": 2.9594,
+      "step": 51153
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8024198412895203,
+      "learning_rate": 3.910995838208883e-05,
+      "loss": 2.9412,
+      "step": 51154
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7190766334533691,
+      "learning_rate": 3.910235765866302e-05,
+      "loss": 3.0936,
+      "step": 51155
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7699955701828003,
+      "learning_rate": 3.9094757622391407e-05,
+      "loss": 3.0168,
+      "step": 51156
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7096518874168396,
+      "learning_rate": 3.908715827329394e-05,
+      "loss": 2.9694,
+      "step": 51157
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7942549586296082,
+      "learning_rate": 3.907955961139078e-05,
+      "loss": 2.8103,
+      "step": 51158
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7407145500183105,
+      "learning_rate": 3.90719616367018e-05,
+      "loss": 2.7899,
+      "step": 51159
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7580335736274719,
+      "learning_rate": 3.9064364349247025e-05,
+      "loss": 2.9164,
+      "step": 51160
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7507249116897583,
+      "learning_rate": 3.9056767749046534e-05,
+      "loss": 2.795,
+      "step": 51161
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7054608464241028,
+      "learning_rate": 3.9049171836120255e-05,
+      "loss": 3.0047,
+      "step": 51162
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8000778555870056,
+      "learning_rate": 3.9041576610488226e-05,
+      "loss": 2.8907,
+      "step": 51163
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7749502062797546,
+      "learning_rate": 3.903398207217053e-05,
+      "loss": 2.9108,
+      "step": 51164
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7396555542945862,
+      "learning_rate": 3.902638822118703e-05,
+      "loss": 2.8869,
+      "step": 51165
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7471228241920471,
+      "learning_rate": 3.9018795057557874e-05,
+      "loss": 2.7746,
+      "step": 51166
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7587604522705078,
+      "learning_rate": 3.901120258130297e-05,
+      "loss": 2.8944,
+      "step": 51167
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.729351818561554,
+      "learning_rate": 3.900361079244228e-05,
+      "loss": 3.1015,
+      "step": 51168
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7310697436332703,
+      "learning_rate": 3.899601969099591e-05,
+      "loss": 2.7458,
+      "step": 51169
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7379913330078125,
+      "learning_rate": 3.898842927698371e-05,
+      "loss": 2.7591,
+      "step": 51170
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8379819393157959,
+      "learning_rate": 3.8980839550425846e-05,
+      "loss": 2.8154,
+      "step": 51171
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7568002343177795,
+      "learning_rate": 3.8973250511342127e-05,
+      "loss": 2.9693,
+      "step": 51172
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7051762938499451,
+      "learning_rate": 3.89656621597527e-05,
+      "loss": 2.9149,
+      "step": 51173
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7339165210723877,
+      "learning_rate": 3.895807449567752e-05,
+      "loss": 2.9697,
+      "step": 51174
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7606975436210632,
+      "learning_rate": 3.895048751913643e-05,
+      "loss": 2.8402,
+      "step": 51175
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7482293248176575,
+      "learning_rate": 3.894290123014959e-05,
+      "loss": 3.0584,
+      "step": 51176
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.727558970451355,
+      "learning_rate": 3.8935315628736816e-05,
+      "loss": 2.9286,
+      "step": 51177
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7616907954216003,
+      "learning_rate": 3.892773071491823e-05,
+      "loss": 2.9106,
+      "step": 51178
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7025411128997803,
+      "learning_rate": 3.89201464887138e-05,
+      "loss": 2.7646,
+      "step": 51179
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7629191279411316,
+      "learning_rate": 3.891256295014342e-05,
+      "loss": 3.0237,
+      "step": 51180
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.776381254196167,
+      "learning_rate": 3.8904980099227145e-05,
+      "loss": 3.0137,
+      "step": 51181
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7896276116371155,
+      "learning_rate": 3.8897397935984854e-05,
+      "loss": 2.7922,
+      "step": 51182
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6885989904403687,
+      "learning_rate": 3.888981646043653e-05,
+      "loss": 2.9651,
+      "step": 51183
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7584837675094604,
+      "learning_rate": 3.888223567260226e-05,
+      "loss": 3.0849,
+      "step": 51184
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7259736061096191,
+      "learning_rate": 3.887465557250186e-05,
+      "loss": 2.7324,
+      "step": 51185
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7147319912910461,
+      "learning_rate": 3.886707616015541e-05,
+      "loss": 2.9832,
+      "step": 51186
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7638548016548157,
+      "learning_rate": 3.8859497435582764e-05,
+      "loss": 2.7991,
+      "step": 51187
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7534319162368774,
+      "learning_rate": 3.885191939880404e-05,
+      "loss": 2.9444,
+      "step": 51188
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7515311241149902,
+      "learning_rate": 3.8844342049839065e-05,
+      "loss": 2.7472,
+      "step": 51189
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7729167342185974,
+      "learning_rate": 3.88367653887078e-05,
+      "loss": 2.9857,
+      "step": 51190
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.735460102558136,
+      "learning_rate": 3.8829189415430285e-05,
+      "loss": 2.9408,
+      "step": 51191
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7101142406463623,
+      "learning_rate": 3.882161413002636e-05,
+      "loss": 3.033,
+      "step": 51192
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7463287115097046,
+      "learning_rate": 3.8814039532516005e-05,
+      "loss": 2.7537,
+      "step": 51193
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8116194009780884,
+      "learning_rate": 3.880646562291931e-05,
+      "loss": 3.1006,
+      "step": 51194
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7226468324661255,
+      "learning_rate": 3.87988924012561e-05,
+      "loss": 2.9244,
+      "step": 51195
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7452834248542786,
+      "learning_rate": 3.8791319867546345e-05,
+      "loss": 2.8646,
+      "step": 51196
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.78560471534729,
+      "learning_rate": 3.878374802180994e-05,
+      "loss": 2.895,
+      "step": 51197
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7502028942108154,
+      "learning_rate": 3.8776176864066886e-05,
+      "loss": 2.9412,
+      "step": 51198
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.753572940826416,
+      "learning_rate": 3.8768606394337086e-05,
+      "loss": 2.994,
+      "step": 51199
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7555680871009827,
+      "learning_rate": 3.876103661264048e-05,
+      "loss": 2.7353,
+      "step": 51200
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7278770804405212,
+      "learning_rate": 3.8753467518997094e-05,
+      "loss": 2.8839,
+      "step": 51201
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7377351522445679,
+      "learning_rate": 3.87458991134268e-05,
+      "loss": 2.8566,
+      "step": 51202
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7519879341125488,
+      "learning_rate": 3.873833139594953e-05,
+      "loss": 2.9622,
+      "step": 51203
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7733878493309021,
+      "learning_rate": 3.873076436658512e-05,
+      "loss": 2.902,
+      "step": 51204
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7269855737686157,
+      "learning_rate": 3.872319802535363e-05,
+      "loss": 3.0837,
+      "step": 51205
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7522500157356262,
+      "learning_rate": 3.8715632372275005e-05,
+      "loss": 2.8483,
+      "step": 51206
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7667732834815979,
+      "learning_rate": 3.870806740736907e-05,
+      "loss": 2.8965,
+      "step": 51207
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7667376399040222,
+      "learning_rate": 3.8700503130655836e-05,
+      "loss": 2.6717,
+      "step": 51208
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7345377206802368,
+      "learning_rate": 3.869293954215516e-05,
+      "loss": 2.9421,
+      "step": 51209
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8869428038597107,
+      "learning_rate": 3.868537664188701e-05,
+      "loss": 2.8926,
+      "step": 51210
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.725968062877655,
+      "learning_rate": 3.8677814429871324e-05,
+      "loss": 2.996,
+      "step": 51211
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.805472195148468,
+      "learning_rate": 3.8670252906127905e-05,
+      "loss": 2.8731,
+      "step": 51212
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7062275409698486,
+      "learning_rate": 3.866269207067681e-05,
+      "loss": 2.8816,
+      "step": 51213
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7009338736534119,
+      "learning_rate": 3.865513192353783e-05,
+      "loss": 2.8552,
+      "step": 51214
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7251269817352295,
+      "learning_rate": 3.864757246473094e-05,
+      "loss": 2.9754,
+      "step": 51215
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7357146739959717,
+      "learning_rate": 3.864001369427608e-05,
+      "loss": 3.0632,
+      "step": 51216
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7163044810295105,
+      "learning_rate": 3.863245561219316e-05,
+      "loss": 2.9741,
+      "step": 51217
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7487192153930664,
+      "learning_rate": 3.862489821850204e-05,
+      "loss": 2.9866,
+      "step": 51218
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.727894127368927,
+      "learning_rate": 3.8617341513222554e-05,
+      "loss": 2.9668,
+      "step": 51219
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7058404088020325,
+      "learning_rate": 3.8609785496374714e-05,
+      "loss": 2.8872,
+      "step": 51220
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7830812931060791,
+      "learning_rate": 3.8602230167978434e-05,
+      "loss": 2.5873,
+      "step": 51221
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7016585469245911,
+      "learning_rate": 3.859467552805353e-05,
+      "loss": 2.812,
+      "step": 51222
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7951351404190063,
+      "learning_rate": 3.8587121576619996e-05,
+      "loss": 2.7951,
+      "step": 51223
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7332335114479065,
+      "learning_rate": 3.8579568313697674e-05,
+      "loss": 3.1449,
+      "step": 51224
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7628679275512695,
+      "learning_rate": 3.8572015739306386e-05,
+      "loss": 2.8698,
+      "step": 51225
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7135177850723267,
+      "learning_rate": 3.8564463853466175e-05,
+      "loss": 2.8874,
+      "step": 51226
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7211930751800537,
+      "learning_rate": 3.855691265619677e-05,
+      "loss": 2.8293,
+      "step": 51227
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7237968444824219,
+      "learning_rate": 3.854936214751822e-05,
+      "loss": 2.8799,
+      "step": 51228
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7675337791442871,
+      "learning_rate": 3.854181232745026e-05,
+      "loss": 2.9046,
+      "step": 51229
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7495302557945251,
+      "learning_rate": 3.853426319601286e-05,
+      "loss": 3.2686,
+      "step": 51230
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7274502515792847,
+      "learning_rate": 3.852671475322593e-05,
+      "loss": 2.9237,
+      "step": 51231
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7208870053291321,
+      "learning_rate": 3.851916699910932e-05,
+      "loss": 3.0848,
+      "step": 51232
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7488343715667725,
+      "learning_rate": 3.851161993368291e-05,
+      "loss": 2.658,
+      "step": 51233
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.749234139919281,
+      "learning_rate": 3.850407355696653e-05,
+      "loss": 2.8975,
+      "step": 51234
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7662747502326965,
+      "learning_rate": 3.849652786898005e-05,
+      "loss": 2.8386,
+      "step": 51235
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.779507040977478,
+      "learning_rate": 3.848898286974347e-05,
+      "loss": 2.7845,
+      "step": 51236
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7223421931266785,
+      "learning_rate": 3.848143855927652e-05,
+      "loss": 2.8635,
+      "step": 51237
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7322688102722168,
+      "learning_rate": 3.84738949375992e-05,
+      "loss": 3.0542,
+      "step": 51238
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7349992990493774,
+      "learning_rate": 3.846635200473129e-05,
+      "loss": 2.9684,
+      "step": 51239
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7478016018867493,
+      "learning_rate": 3.845880976069261e-05,
+      "loss": 3.0442,
+      "step": 51240
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7727739214897156,
+      "learning_rate": 3.845126820550317e-05,
+      "loss": 2.7984,
+      "step": 51241
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7362459301948547,
+      "learning_rate": 3.8443727339182695e-05,
+      "loss": 2.8293,
+      "step": 51242
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7158995866775513,
+      "learning_rate": 3.843618716175116e-05,
+      "loss": 2.8788,
+      "step": 51243
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7135204672813416,
+      "learning_rate": 3.8428647673228294e-05,
+      "loss": 2.933,
+      "step": 51244
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7790452241897583,
+      "learning_rate": 3.842110887363404e-05,
+      "loss": 2.9307,
+      "step": 51245
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7330326437950134,
+      "learning_rate": 3.8413570762988325e-05,
+      "loss": 2.6545,
+      "step": 51246
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7503302693367004,
+      "learning_rate": 3.8406033341310884e-05,
+      "loss": 2.8702,
+      "step": 51247
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8307878375053406,
+      "learning_rate": 3.8398496608621644e-05,
+      "loss": 3.0346,
+      "step": 51248
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7150816917419434,
+      "learning_rate": 3.839096056494032e-05,
+      "loss": 2.9994,
+      "step": 51249
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7854032516479492,
+      "learning_rate": 3.8383425210286864e-05,
+      "loss": 2.895,
+      "step": 51250
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7631329298019409,
+      "learning_rate": 3.8375890544681195e-05,
+      "loss": 2.9736,
+      "step": 51251
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7520675659179688,
+      "learning_rate": 3.8368356568143036e-05,
+      "loss": 3.052,
+      "step": 51252
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7277513742446899,
+      "learning_rate": 3.836082328069229e-05,
+      "loss": 2.9834,
+      "step": 51253
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7684758901596069,
+      "learning_rate": 3.8353290682348814e-05,
+      "loss": 2.7606,
+      "step": 51254
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7762331366539001,
+      "learning_rate": 3.834575877313233e-05,
+      "loss": 3.0253,
+      "step": 51255
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7357913851737976,
+      "learning_rate": 3.833822755306285e-05,
+      "loss": 3.0781,
+      "step": 51256
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7422073483467102,
+      "learning_rate": 3.833069702216003e-05,
+      "loss": 2.9321,
+      "step": 51257
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7310648560523987,
+      "learning_rate": 3.832316718044388e-05,
+      "loss": 2.8499,
+      "step": 51258
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6889393925666809,
+      "learning_rate": 3.831563802793409e-05,
+      "loss": 3.0701,
+      "step": 51259
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.1530731916427612,
+      "learning_rate": 3.830810956465058e-05,
+      "loss": 3.111,
+      "step": 51260
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7462144494056702,
+      "learning_rate": 3.830058179061316e-05,
+      "loss": 2.8284,
+      "step": 51261
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7419637441635132,
+      "learning_rate": 3.829305470584158e-05,
+      "loss": 3.0065,
+      "step": 51262
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7542158365249634,
+      "learning_rate": 3.828552831035581e-05,
+      "loss": 2.8055,
+      "step": 51263
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7965158224105835,
+      "learning_rate": 3.82780026041755e-05,
+      "loss": 2.7436,
+      "step": 51264
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.743840217590332,
+      "learning_rate": 3.8270477587320566e-05,
+      "loss": 2.7356,
+      "step": 51265
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7825320959091187,
+      "learning_rate": 3.826295325981091e-05,
+      "loss": 2.7954,
+      "step": 51266
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7897096276283264,
+      "learning_rate": 3.825542962166619e-05,
+      "loss": 2.9194,
+      "step": 51267
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7811319828033447,
+      "learning_rate": 3.8247906672906345e-05,
+      "loss": 2.7623,
+      "step": 51268
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7193589806556702,
+      "learning_rate": 3.824038441355114e-05,
+      "loss": 2.8256,
+      "step": 51269
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7284131050109863,
+      "learning_rate": 3.823286284362031e-05,
+      "loss": 2.8813,
+      "step": 51270
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7225155830383301,
+      "learning_rate": 3.822534196313383e-05,
+      "loss": 2.9283,
+      "step": 51271
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7368038296699524,
+      "learning_rate": 3.8217821772111354e-05,
+      "loss": 2.7434,
+      "step": 51272
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7491246461868286,
+      "learning_rate": 3.821030227057279e-05,
+      "loss": 2.9343,
+      "step": 51273
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7268712520599365,
+      "learning_rate": 3.820278345853788e-05,
+      "loss": 2.9003,
+      "step": 51274
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7777917981147766,
+      "learning_rate": 3.81952653360265e-05,
+      "loss": 3.0695,
+      "step": 51275
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6771141886711121,
+      "learning_rate": 3.818774790305842e-05,
+      "loss": 2.813,
+      "step": 51276
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7285410165786743,
+      "learning_rate": 3.818023115965334e-05,
+      "loss": 2.9205,
+      "step": 51277
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7287654876708984,
+      "learning_rate": 3.817271510583122e-05,
+      "loss": 2.9855,
+      "step": 51278
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7238658666610718,
+      "learning_rate": 3.816519974161171e-05,
+      "loss": 2.6708,
+      "step": 51279
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.756721556186676,
+      "learning_rate": 3.815768506701472e-05,
+      "loss": 2.866,
+      "step": 51280
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7326313853263855,
+      "learning_rate": 3.8150171082059976e-05,
+      "loss": 2.8672,
+      "step": 51281
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7391651272773743,
+      "learning_rate": 3.8142657786767324e-05,
+      "loss": 2.893,
+      "step": 51282
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7335715889930725,
+      "learning_rate": 3.813514518115649e-05,
+      "loss": 2.8543,
+      "step": 51283
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7516505718231201,
+      "learning_rate": 3.812763326524725e-05,
+      "loss": 3.074,
+      "step": 51284
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6861124038696289,
+      "learning_rate": 3.812012203905949e-05,
+      "loss": 2.9163,
+      "step": 51285
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.784110426902771,
+      "learning_rate": 3.8112611502612886e-05,
+      "loss": 2.7284,
+      "step": 51286
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7559488415718079,
+      "learning_rate": 3.810510165592724e-05,
+      "loss": 3.0254,
+      "step": 51287
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7138981223106384,
+      "learning_rate": 3.8097592499022424e-05,
+      "loss": 2.9743,
+      "step": 51288
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7892126441001892,
+      "learning_rate": 3.809008403191806e-05,
+      "loss": 2.8434,
+      "step": 51289
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7221898436546326,
+      "learning_rate": 3.808257625463409e-05,
+      "loss": 3.0125,
+      "step": 51290
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.745223879814148,
+      "learning_rate": 3.807506916719021e-05,
+      "loss": 2.8629,
+      "step": 51291
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7798271775245667,
+      "learning_rate": 3.8067562769606096e-05,
+      "loss": 2.8282,
+      "step": 51292
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7572090029716492,
+      "learning_rate": 3.8060057061901704e-05,
+      "loss": 2.9673,
+      "step": 51293
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7191266417503357,
+      "learning_rate": 3.8052552044096655e-05,
+      "loss": 2.8256,
+      "step": 51294
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7357266545295715,
+      "learning_rate": 3.804504771621082e-05,
+      "loss": 2.7246,
+      "step": 51295
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.742739737033844,
+      "learning_rate": 3.803754407826386e-05,
+      "loss": 3.0129,
+      "step": 51296
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7276623249053955,
+      "learning_rate": 3.8030041130275625e-05,
+      "loss": 2.9153,
+      "step": 51297
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7444546222686768,
+      "learning_rate": 3.80225388722659e-05,
+      "loss": 2.9379,
+      "step": 51298
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.727475643157959,
+      "learning_rate": 3.801503730425429e-05,
+      "loss": 2.8814,
+      "step": 51299
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.714273989200592,
+      "learning_rate": 3.80075364262607e-05,
+      "loss": 2.8059,
+      "step": 51300
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7392892837524414,
+      "learning_rate": 3.80000362383048e-05,
+      "loss": 2.6062,
+      "step": 51301
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7788336277008057,
+      "learning_rate": 3.799253674040638e-05,
+      "loss": 2.6913,
+      "step": 51302
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.76091468334198,
+      "learning_rate": 3.798503793258525e-05,
+      "loss": 2.9406,
+      "step": 51303
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8225963115692139,
+      "learning_rate": 3.7977539814861105e-05,
+      "loss": 2.7359,
+      "step": 51304
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7406080365180969,
+      "learning_rate": 3.797004238725372e-05,
+      "loss": 3.0082,
+      "step": 51305
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7237011194229126,
+      "learning_rate": 3.7962545649782715e-05,
+      "loss": 2.7815,
+      "step": 51306
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7329822778701782,
+      "learning_rate": 3.7955049602467944e-05,
+      "loss": 2.8379,
+      "step": 51307
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7319344282150269,
+      "learning_rate": 3.794755424532919e-05,
+      "loss": 2.7743,
+      "step": 51308
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7189127206802368,
+      "learning_rate": 3.794005957838611e-05,
+      "loss": 2.8306,
+      "step": 51309
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7663130760192871,
+      "learning_rate": 3.7932565601658535e-05,
+      "loss": 2.7893,
+      "step": 51310
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7873907089233398,
+      "learning_rate": 3.792507231516607e-05,
+      "loss": 2.7739,
+      "step": 51311
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7518800497055054,
+      "learning_rate": 3.791757971892859e-05,
+      "loss": 2.6856,
+      "step": 51312
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7844192981719971,
+      "learning_rate": 3.791008781296578e-05,
+      "loss": 2.8565,
+      "step": 51313
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.8025628924369812,
+      "learning_rate": 3.79025965972973e-05,
+      "loss": 2.7685,
+      "step": 51314
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7403668165206909,
+      "learning_rate": 3.789510607194298e-05,
+      "loss": 2.9046,
+      "step": 51315
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7729617953300476,
+      "learning_rate": 3.788761623692246e-05,
+      "loss": 3.0336,
+      "step": 51316
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7558656930923462,
+      "learning_rate": 3.7880127092255505e-05,
+      "loss": 2.9393,
+      "step": 51317
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7228742837905884,
+      "learning_rate": 3.787263863796195e-05,
+      "loss": 3.109,
+      "step": 51318
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7505278587341309,
+      "learning_rate": 3.786515087406137e-05,
+      "loss": 3.0069,
+      "step": 51319
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6907845735549927,
+      "learning_rate": 3.785766380057354e-05,
+      "loss": 2.9476,
+      "step": 51320
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7361884713172913,
+      "learning_rate": 3.7850177417518135e-05,
+      "loss": 2.8551,
+      "step": 51321
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7802038192749023,
+      "learning_rate": 3.7842691724914884e-05,
+      "loss": 3.0691,
+      "step": 51322
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7692674398422241,
+      "learning_rate": 3.783520672278362e-05,
+      "loss": 2.9041,
+      "step": 51323
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7283543348312378,
+      "learning_rate": 3.782772241114388e-05,
+      "loss": 2.753,
+      "step": 51324
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7569485306739807,
+      "learning_rate": 3.7820238790015526e-05,
+      "loss": 3.0877,
+      "step": 51325
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8705662488937378,
+      "learning_rate": 3.781275585941823e-05,
+      "loss": 2.8212,
+      "step": 51326
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7248303890228271,
+      "learning_rate": 3.780527361937159e-05,
+      "loss": 2.9386,
+      "step": 51327
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7549064755439758,
+      "learning_rate": 3.779779206989548e-05,
+      "loss": 2.8461,
+      "step": 51328
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7630690932273865,
+      "learning_rate": 3.7790311211009426e-05,
+      "loss": 2.9711,
+      "step": 51329
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7439910769462585,
+      "learning_rate": 3.778283104273334e-05,
+      "loss": 2.9077,
+      "step": 51330
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7504767179489136,
+      "learning_rate": 3.777535156508671e-05,
+      "loss": 2.7569,
+      "step": 51331
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7491384148597717,
+      "learning_rate": 3.776787277808935e-05,
+      "loss": 2.7003,
+      "step": 51332
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.770693302154541,
+      "learning_rate": 3.7760394681761006e-05,
+      "loss": 2.9364,
+      "step": 51333
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7890000343322754,
+      "learning_rate": 3.775291727612134e-05,
+      "loss": 3.0071,
+      "step": 51334
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.711444616317749,
+      "learning_rate": 3.774544056119e-05,
+      "loss": 2.8482,
+      "step": 51335
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7917689085006714,
+      "learning_rate": 3.773796453698663e-05,
+      "loss": 2.8997,
+      "step": 51336
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6970062255859375,
+      "learning_rate": 3.773048920353099e-05,
+      "loss": 2.9283,
+      "step": 51337
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7180047631263733,
+      "learning_rate": 3.7723014560842825e-05,
+      "loss": 2.8176,
+      "step": 51338
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.743811309337616,
+      "learning_rate": 3.771554060894172e-05,
+      "loss": 3.1697,
+      "step": 51339
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7111158967018127,
+      "learning_rate": 3.7708067347847424e-05,
+      "loss": 2.7527,
+      "step": 51340
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7135360240936279,
+      "learning_rate": 3.770059477757966e-05,
+      "loss": 2.7748,
+      "step": 51341
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.767493724822998,
+      "learning_rate": 3.7693122898157945e-05,
+      "loss": 3.1682,
+      "step": 51342
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7691839337348938,
+      "learning_rate": 3.7685651709602126e-05,
+      "loss": 3.0069,
+      "step": 51343
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7307931780815125,
+      "learning_rate": 3.767818121193176e-05,
+      "loss": 2.963,
+      "step": 51344
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7612271308898926,
+      "learning_rate": 3.767071140516665e-05,
+      "loss": 2.7677,
+      "step": 51345
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7013478875160217,
+      "learning_rate": 3.766324228932633e-05,
+      "loss": 2.7819,
+      "step": 51346
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7199133038520813,
+      "learning_rate": 3.765577386443055e-05,
+      "loss": 2.7997,
+      "step": 51347
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7004144787788391,
+      "learning_rate": 3.764830613049902e-05,
+      "loss": 2.9246,
+      "step": 51348
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7983659505844116,
+      "learning_rate": 3.7640839087551386e-05,
+      "loss": 3.0463,
+      "step": 51349
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7310093641281128,
+      "learning_rate": 3.763337273560728e-05,
+      "loss": 2.8928,
+      "step": 51350
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7364878058433533,
+      "learning_rate": 3.762590707468631e-05,
+      "loss": 2.8745,
+      "step": 51351
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7587478756904602,
+      "learning_rate": 3.7618442104808266e-05,
+      "loss": 2.9023,
+      "step": 51352
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7633194327354431,
+      "learning_rate": 3.761097782599266e-05,
+      "loss": 2.8162,
+      "step": 51353
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7305599451065063,
+      "learning_rate": 3.760351423825928e-05,
+      "loss": 2.9469,
+      "step": 51354
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7444048523902893,
+      "learning_rate": 3.759605134162781e-05,
+      "loss": 3.1084,
+      "step": 51355
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7268261909484863,
+      "learning_rate": 3.758858913611783e-05,
+      "loss": 2.8725,
+      "step": 51356
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7650936841964722,
+      "learning_rate": 3.7581127621748996e-05,
+      "loss": 2.9345,
+      "step": 51357
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7452462315559387,
+      "learning_rate": 3.757366679854089e-05,
+      "loss": 2.8995,
+      "step": 51358
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7164917588233948,
+      "learning_rate": 3.756620666651327e-05,
+      "loss": 2.8016,
+      "step": 51359
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7461237907409668,
+      "learning_rate": 3.75587472256858e-05,
+      "loss": 2.7738,
+      "step": 51360
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7313634753227234,
+      "learning_rate": 3.755128847607802e-05,
+      "loss": 2.7154,
+      "step": 51361
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7086673378944397,
+      "learning_rate": 3.754383041770971e-05,
+      "loss": 3.009,
+      "step": 51362
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7295812368392944,
+      "learning_rate": 3.753637305060044e-05,
+      "loss": 2.7669,
+      "step": 51363
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7354058027267456,
+      "learning_rate": 3.752891637476978e-05,
+      "loss": 3.046,
+      "step": 51364
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7307260632514954,
+      "learning_rate": 3.752146039023753e-05,
+      "loss": 2.8995,
+      "step": 51365
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7760553359985352,
+      "learning_rate": 3.7514005097023125e-05,
+      "loss": 3.0448,
+      "step": 51366
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8252313733100891,
+      "learning_rate": 3.750655049514644e-05,
+      "loss": 2.8753,
+      "step": 51367
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7388860583305359,
+      "learning_rate": 3.7499096584626854e-05,
+      "loss": 2.8755,
+      "step": 51368
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7460933327674866,
+      "learning_rate": 3.7491643365484156e-05,
+      "loss": 3.0417,
+      "step": 51369
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7309091091156006,
+      "learning_rate": 3.748419083773804e-05,
+      "loss": 3.0214,
+      "step": 51370
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7197194695472717,
+      "learning_rate": 3.7476739001408006e-05,
+      "loss": 2.9985,
+      "step": 51371
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7402889132499695,
+      "learning_rate": 3.7469287856513755e-05,
+      "loss": 3.0035,
+      "step": 51372
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7771908640861511,
+      "learning_rate": 3.746183740307476e-05,
+      "loss": 3.1793,
+      "step": 51373
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7580812573432922,
+      "learning_rate": 3.7454387641110804e-05,
+      "loss": 2.6364,
+      "step": 51374
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7813465595245361,
+      "learning_rate": 3.7446938570641515e-05,
+      "loss": 2.84,
+      "step": 51375
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7351908087730408,
+      "learning_rate": 3.7439490191686374e-05,
+      "loss": 2.9712,
+      "step": 51376
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7609416842460632,
+      "learning_rate": 3.743204250426516e-05,
+      "loss": 2.6509,
+      "step": 51377
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7777287364006042,
+      "learning_rate": 3.742459550839743e-05,
+      "loss": 3.0221,
+      "step": 51378
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7442467212677002,
+      "learning_rate": 3.741714920410269e-05,
+      "loss": 2.8488,
+      "step": 51379
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7483687400817871,
+      "learning_rate": 3.740970359140074e-05,
+      "loss": 2.9856,
+      "step": 51380
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7566103935241699,
+      "learning_rate": 3.740225867031103e-05,
+      "loss": 2.9011,
+      "step": 51381
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7553043365478516,
+      "learning_rate": 3.739481444085326e-05,
+      "loss": 2.8707,
+      "step": 51382
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7112729549407959,
+      "learning_rate": 3.7387370903046963e-05,
+      "loss": 2.8141,
+      "step": 51383
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7292284965515137,
+      "learning_rate": 3.737992805691189e-05,
+      "loss": 3.0078,
+      "step": 51384
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7168455123901367,
+      "learning_rate": 3.737248590246749e-05,
+      "loss": 2.8198,
+      "step": 51385
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7240537405014038,
+      "learning_rate": 3.736504443973337e-05,
+      "loss": 2.6507,
+      "step": 51386
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7349284291267395,
+      "learning_rate": 3.7357603668729264e-05,
+      "loss": 2.8739,
+      "step": 51387
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7479743957519531,
+      "learning_rate": 3.7350163589474615e-05,
+      "loss": 2.7535,
+      "step": 51388
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7725386619567871,
+      "learning_rate": 3.734272420198907e-05,
+      "loss": 2.9225,
+      "step": 51389
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7183440923690796,
+      "learning_rate": 3.733528550629232e-05,
+      "loss": 2.6772,
+      "step": 51390
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7492996454238892,
+      "learning_rate": 3.732784750240378e-05,
+      "loss": 2.9754,
+      "step": 51391
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7240472435951233,
+      "learning_rate": 3.732041019034324e-05,
+      "loss": 2.9141,
+      "step": 51392
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7172340750694275,
+      "learning_rate": 3.731297357013018e-05,
+      "loss": 2.8972,
+      "step": 51393
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8641431927680969,
+      "learning_rate": 3.730553764178411e-05,
+      "loss": 2.8425,
+      "step": 51394
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.9124081134796143,
+      "learning_rate": 3.729810240532476e-05,
+      "loss": 2.9483,
+      "step": 51395
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7566990256309509,
+      "learning_rate": 3.7290667860771606e-05,
+      "loss": 2.8396,
+      "step": 51396
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7051795721054077,
+      "learning_rate": 3.728323400814434e-05,
+      "loss": 2.8282,
+      "step": 51397
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7272311449050903,
+      "learning_rate": 3.72758008474624e-05,
+      "loss": 2.8787,
+      "step": 51398
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7363135814666748,
+      "learning_rate": 3.7268368378745486e-05,
+      "loss": 3.0014,
+      "step": 51399
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7577966451644897,
+      "learning_rate": 3.7260936602013136e-05,
+      "loss": 2.8576,
+      "step": 51400
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7400948405265808,
+      "learning_rate": 3.7253505517284876e-05,
+      "loss": 2.6637,
+      "step": 51401
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7425705194473267,
+      "learning_rate": 3.724607512458035e-05,
+      "loss": 3.0053,
+      "step": 51402
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8094008564949036,
+      "learning_rate": 3.7238645423919054e-05,
+      "loss": 2.8479,
+      "step": 51403
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7409995198249817,
+      "learning_rate": 3.723121641532056e-05,
+      "loss": 2.8932,
+      "step": 51404
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7713068723678589,
+      "learning_rate": 3.7223788098804565e-05,
+      "loss": 2.8011,
+      "step": 51405
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7903899550437927,
+      "learning_rate": 3.721636047439054e-05,
+      "loss": 2.9123,
+      "step": 51406
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7373391389846802,
+      "learning_rate": 3.720893354209804e-05,
+      "loss": 2.7982,
+      "step": 51407
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.720302164554596,
+      "learning_rate": 3.7201507301946556e-05,
+      "loss": 2.796,
+      "step": 51408
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7427999973297119,
+      "learning_rate": 3.719408175395573e-05,
+      "loss": 2.8756,
+      "step": 51409
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7941123843193054,
+      "learning_rate": 3.718665689814522e-05,
+      "loss": 3.0277,
+      "step": 51410
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.724496603012085,
+      "learning_rate": 3.717923273453437e-05,
+      "loss": 2.8808,
+      "step": 51411
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7771130204200745,
+      "learning_rate": 3.71718092631429e-05,
+      "loss": 2.6689,
+      "step": 51412
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7540898323059082,
+      "learning_rate": 3.716438648399027e-05,
+      "loss": 3.088,
+      "step": 51413
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7112698554992676,
+      "learning_rate": 3.715696439709609e-05,
+      "loss": 2.8278,
+      "step": 51414
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7136939167976379,
+      "learning_rate": 3.714954300247991e-05,
+      "loss": 2.9853,
+      "step": 51415
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7295254468917847,
+      "learning_rate": 3.714212230016118e-05,
+      "loss": 2.9651,
+      "step": 51416
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7394436597824097,
+      "learning_rate": 3.713470229015956e-05,
+      "loss": 2.8072,
+      "step": 51417
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6955799460411072,
+      "learning_rate": 3.712728297249449e-05,
+      "loss": 2.9469,
+      "step": 51418
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8046622276306152,
+      "learning_rate": 3.7119864347185556e-05,
+      "loss": 3.0347,
+      "step": 51419
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.729836642742157,
+      "learning_rate": 3.7112446414252386e-05,
+      "loss": 2.9362,
+      "step": 51420
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.722966730594635,
+      "learning_rate": 3.710502917371442e-05,
+      "loss": 3.0038,
+      "step": 51421
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7136669158935547,
+      "learning_rate": 3.709761262559124e-05,
+      "loss": 2.904,
+      "step": 51422
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6891632676124573,
+      "learning_rate": 3.709019676990228e-05,
+      "loss": 2.9472,
+      "step": 51423
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7332636117935181,
+      "learning_rate": 3.708278160666711e-05,
+      "loss": 3.063,
+      "step": 51424
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7345249056816101,
+      "learning_rate": 3.707536713590539e-05,
+      "loss": 2.8518,
+      "step": 51425
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7762575745582581,
+      "learning_rate": 3.706795335763649e-05,
+      "loss": 3.0165,
+      "step": 51426
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7118173837661743,
+      "learning_rate": 3.7060540271880044e-05,
+      "loss": 2.721,
+      "step": 51427
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.841128408908844,
+      "learning_rate": 3.705312787865557e-05,
+      "loss": 2.938,
+      "step": 51428
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6889047622680664,
+      "learning_rate": 3.70457161779825e-05,
+      "loss": 2.7814,
+      "step": 51429
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7486089468002319,
+      "learning_rate": 3.7038305169880367e-05,
+      "loss": 3.0047,
+      "step": 51430
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7171392440795898,
+      "learning_rate": 3.703089485436872e-05,
+      "loss": 2.7904,
+      "step": 51431
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7496344447135925,
+      "learning_rate": 3.702348523146718e-05,
+      "loss": 3.0611,
+      "step": 51432
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7369357943534851,
+      "learning_rate": 3.7016076301195084e-05,
+      "loss": 2.9536,
+      "step": 51433
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7333647012710571,
+      "learning_rate": 3.70086680635721e-05,
+      "loss": 3.0673,
+      "step": 51434
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7265602946281433,
+      "learning_rate": 3.7001260518617604e-05,
+      "loss": 3.0387,
+      "step": 51435
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7276657223701477,
+      "learning_rate": 3.6993853666351214e-05,
+      "loss": 2.9921,
+      "step": 51436
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.75041264295578,
+      "learning_rate": 3.6986447506792414e-05,
+      "loss": 2.9,
+      "step": 51437
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.77069091796875,
+      "learning_rate": 3.697904203996066e-05,
+      "loss": 2.801,
+      "step": 51438
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7159631848335266,
+      "learning_rate": 3.6971637265875495e-05,
+      "loss": 2.9552,
+      "step": 51439
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7610557079315186,
+      "learning_rate": 3.6964233184556415e-05,
+      "loss": 2.9516,
+      "step": 51440
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7705284953117371,
+      "learning_rate": 3.695682979602289e-05,
+      "loss": 2.8762,
+      "step": 51441
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7517569065093994,
+      "learning_rate": 3.694942710029452e-05,
+      "loss": 2.9491,
+      "step": 51442
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.774166464805603,
+      "learning_rate": 3.6942025097390705e-05,
+      "loss": 2.8259,
+      "step": 51443
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7174193263053894,
+      "learning_rate": 3.693462378733101e-05,
+      "loss": 2.8627,
+      "step": 51444
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7680938839912415,
+      "learning_rate": 3.692722317013481e-05,
+      "loss": 2.9821,
+      "step": 51445
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7469550371170044,
+      "learning_rate": 3.691982324582167e-05,
+      "loss": 2.8813,
+      "step": 51446
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8036483526229858,
+      "learning_rate": 3.6912424014411156e-05,
+      "loss": 2.7214,
+      "step": 51447
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7748443484306335,
+      "learning_rate": 3.6905025475922634e-05,
+      "loss": 2.9701,
+      "step": 51448
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7516213059425354,
+      "learning_rate": 3.689762763037568e-05,
+      "loss": 3.0216,
+      "step": 51449
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7000873684883118,
+      "learning_rate": 3.6890230477789754e-05,
+      "loss": 2.8373,
+      "step": 51450
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7015297412872314,
+      "learning_rate": 3.688283401818426e-05,
+      "loss": 2.9047,
+      "step": 51451
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.717642068862915,
+      "learning_rate": 3.6875438251578794e-05,
+      "loss": 2.8453,
+      "step": 51452
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7319772243499756,
+      "learning_rate": 3.6868043177992736e-05,
+      "loss": 2.8335,
+      "step": 51453
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7205098867416382,
+      "learning_rate": 3.686064879744567e-05,
+      "loss": 3.0223,
+      "step": 51454
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7658298015594482,
+      "learning_rate": 3.685325510995695e-05,
+      "loss": 2.8972,
+      "step": 51455
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7169014811515808,
+      "learning_rate": 3.684586211554609e-05,
+      "loss": 3.1171,
+      "step": 51456
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7414019703865051,
+      "learning_rate": 3.683846981423267e-05,
+      "loss": 3.0538,
+      "step": 51457
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.772877037525177,
+      "learning_rate": 3.683107820603609e-05,
+      "loss": 2.8904,
+      "step": 51458
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7601258158683777,
+      "learning_rate": 3.682368729097579e-05,
+      "loss": 2.7492,
+      "step": 51459
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7426812052726746,
+      "learning_rate": 3.681629706907115e-05,
+      "loss": 2.7345,
+      "step": 51460
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7161027193069458,
+      "learning_rate": 3.680890754034177e-05,
+      "loss": 3.0118,
+      "step": 51461
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7432463765144348,
+      "learning_rate": 3.6801518704807156e-05,
+      "loss": 2.8489,
+      "step": 51462
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7574390769004822,
+      "learning_rate": 3.679413056248659e-05,
+      "loss": 2.8106,
+      "step": 51463
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7752411961555481,
+      "learning_rate": 3.6786743113399696e-05,
+      "loss": 2.9405,
+      "step": 51464
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7201191782951355,
+      "learning_rate": 3.677935635756586e-05,
+      "loss": 2.943,
+      "step": 51465
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7174774408340454,
+      "learning_rate": 3.67719702950045e-05,
+      "loss": 2.8726,
+      "step": 51466
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7147430777549744,
+      "learning_rate": 3.676458492573515e-05,
+      "loss": 3.0209,
+      "step": 51467
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.757570743560791,
+      "learning_rate": 3.675720024977718e-05,
+      "loss": 3.0395,
+      "step": 51468
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6996948719024658,
+      "learning_rate": 3.674981626715015e-05,
+      "loss": 2.7247,
+      "step": 51469
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8142804503440857,
+      "learning_rate": 3.674243297787335e-05,
+      "loss": 2.7782,
+      "step": 51470
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7229151129722595,
+      "learning_rate": 3.6735050381966346e-05,
+      "loss": 2.7468,
+      "step": 51471
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7174829840660095,
+      "learning_rate": 3.672766847944859e-05,
+      "loss": 2.9069,
+      "step": 51472
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7757089734077454,
+      "learning_rate": 3.67202872703395e-05,
+      "loss": 3.1334,
+      "step": 51473
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7175819873809814,
+      "learning_rate": 3.6712906754658524e-05,
+      "loss": 3.0466,
+      "step": 51474
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.739388108253479,
+      "learning_rate": 3.6705526932424975e-05,
+      "loss": 2.9729,
+      "step": 51475
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7365408539772034,
+      "learning_rate": 3.669814780365842e-05,
+      "loss": 2.8922,
+      "step": 51476
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6910611987113953,
+      "learning_rate": 3.669076936837836e-05,
+      "loss": 2.7283,
+      "step": 51477
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7462716102600098,
+      "learning_rate": 3.668339162660402e-05,
+      "loss": 2.814,
+      "step": 51478
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7815803289413452,
+      "learning_rate": 3.667601457835505e-05,
+      "loss": 2.8589,
+      "step": 51479
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7381653189659119,
+      "learning_rate": 3.666863822365077e-05,
+      "loss": 2.882,
+      "step": 51480
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7210900783538818,
+      "learning_rate": 3.6661262562510566e-05,
+      "loss": 3.0684,
+      "step": 51481
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7968168258666992,
+      "learning_rate": 3.665388759495398e-05,
+      "loss": 3.0299,
+      "step": 51482
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7475718855857849,
+      "learning_rate": 3.664651332100027e-05,
+      "loss": 2.8383,
+      "step": 51483
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8005061149597168,
+      "learning_rate": 3.6639139740669055e-05,
+      "loss": 3.1725,
+      "step": 51484
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7502545118331909,
+      "learning_rate": 3.663176685397962e-05,
+      "loss": 2.7869,
+      "step": 51485
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7680295705795288,
+      "learning_rate": 3.662439466095144e-05,
+      "loss": 2.8294,
+      "step": 51486
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.710451066493988,
+      "learning_rate": 3.661702316160394e-05,
+      "loss": 2.7683,
+      "step": 51487
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6979348063468933,
+      "learning_rate": 3.660965235595644e-05,
+      "loss": 2.9291,
+      "step": 51488
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7553119659423828,
+      "learning_rate": 3.660228224402849e-05,
+      "loss": 2.9579,
+      "step": 51489
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7529929876327515,
+      "learning_rate": 3.659491282583941e-05,
+      "loss": 2.8826,
+      "step": 51490
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7481595277786255,
+      "learning_rate": 3.65875441014086e-05,
+      "loss": 2.8124,
+      "step": 51491
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7778012156486511,
+      "learning_rate": 3.6580176070755576e-05,
+      "loss": 2.7338,
+      "step": 51492
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7474700808525085,
+      "learning_rate": 3.657280873389959e-05,
+      "loss": 3.1009,
+      "step": 51493
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7289611101150513,
+      "learning_rate": 3.656544209086022e-05,
+      "loss": 3.0016,
+      "step": 51494
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.815883219242096,
+      "learning_rate": 3.6558076141656766e-05,
+      "loss": 2.9222,
+      "step": 51495
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7379744648933411,
+      "learning_rate": 3.655071088630855e-05,
+      "loss": 3.0274,
+      "step": 51496
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7309653162956238,
+      "learning_rate": 3.654334632483516e-05,
+      "loss": 2.8887,
+      "step": 51497
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7571339011192322,
+      "learning_rate": 3.6535982457255817e-05,
+      "loss": 2.9202,
+      "step": 51498
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7372366786003113,
+      "learning_rate": 3.6528619283590066e-05,
+      "loss": 3.1448,
+      "step": 51499
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7438984513282776,
+      "learning_rate": 3.652125680385712e-05,
+      "loss": 2.9245,
+      "step": 51500
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7498157024383545,
+      "learning_rate": 3.651389501807658e-05,
+      "loss": 3.1159,
+      "step": 51501
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7399702072143555,
+      "learning_rate": 3.650653392626775e-05,
+      "loss": 3.0412,
+      "step": 51502
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7499619126319885,
+      "learning_rate": 3.6499173528449874e-05,
+      "loss": 2.8588,
+      "step": 51503
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.784655749797821,
+      "learning_rate": 3.649181382464256e-05,
+      "loss": 3.1072,
+      "step": 51504
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.742243230342865,
+      "learning_rate": 3.648445481486504e-05,
+      "loss": 3.0878,
+      "step": 51505
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7311636805534363,
+      "learning_rate": 3.6477096499136735e-05,
+      "loss": 2.9728,
+      "step": 51506
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7376363277435303,
+      "learning_rate": 3.646973887747714e-05,
+      "loss": 2.9388,
+      "step": 51507
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.746273934841156,
+      "learning_rate": 3.646238194990549e-05,
+      "loss": 2.919,
+      "step": 51508
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7192387580871582,
+      "learning_rate": 3.645502571644122e-05,
+      "loss": 2.8624,
+      "step": 51509
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7488130331039429,
+      "learning_rate": 3.644767017710364e-05,
+      "loss": 2.7476,
+      "step": 51510
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7002290487289429,
+      "learning_rate": 3.644031533191223e-05,
+      "loss": 2.8554,
+      "step": 51511
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7030738592147827,
+      "learning_rate": 3.6432961180886246e-05,
+      "loss": 2.9592,
+      "step": 51512
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8064178228378296,
+      "learning_rate": 3.642560772404514e-05,
+      "loss": 2.8433,
+      "step": 51513
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7585741877555847,
+      "learning_rate": 3.641825496140829e-05,
+      "loss": 2.8685,
+      "step": 51514
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7327173352241516,
+      "learning_rate": 3.6410902892994953e-05,
+      "loss": 2.8906,
+      "step": 51515
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7685412168502808,
+      "learning_rate": 3.640355151882468e-05,
+      "loss": 2.6577,
+      "step": 51516
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7944349646568298,
+      "learning_rate": 3.639620083891669e-05,
+      "loss": 2.9258,
+      "step": 51517
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7449418902397156,
+      "learning_rate": 3.6388850853290294e-05,
+      "loss": 2.7102,
+      "step": 51518
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7388597130775452,
+      "learning_rate": 3.638150156196502e-05,
+      "loss": 2.9237,
+      "step": 51519
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7266759872436523,
+      "learning_rate": 3.6374152964960076e-05,
+      "loss": 2.9021,
+      "step": 51520
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7184749841690063,
+      "learning_rate": 3.6366805062294956e-05,
+      "loss": 3.0353,
+      "step": 51521
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7220653891563416,
+      "learning_rate": 3.6359457853988835e-05,
+      "loss": 2.8162,
+      "step": 51522
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7334814667701721,
+      "learning_rate": 3.635211134006124e-05,
+      "loss": 2.9972,
+      "step": 51523
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7480335831642151,
+      "learning_rate": 3.634476552053145e-05,
+      "loss": 2.8407,
+      "step": 51524
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7650389671325684,
+      "learning_rate": 3.633742039541875e-05,
+      "loss": 2.8594,
+      "step": 51525
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7087839841842651,
+      "learning_rate": 3.6330075964742586e-05,
+      "loss": 2.7335,
+      "step": 51526
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7374259829521179,
+      "learning_rate": 3.632273222852222e-05,
+      "loss": 2.954,
+      "step": 51527
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7621423006057739,
+      "learning_rate": 3.6315389186777e-05,
+      "loss": 2.9835,
+      "step": 51528
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7097713351249695,
+      "learning_rate": 3.630804683952637e-05,
+      "loss": 2.8251,
+      "step": 51529
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7970553636550903,
+      "learning_rate": 3.6300705186789627e-05,
+      "loss": 2.8793,
+      "step": 51530
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7120518088340759,
+      "learning_rate": 3.629336422858604e-05,
+      "loss": 2.8642,
+      "step": 51531
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.731629490852356,
+      "learning_rate": 3.628602396493494e-05,
+      "loss": 2.9259,
+      "step": 51532
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7600454688072205,
+      "learning_rate": 3.6278684395855676e-05,
+      "loss": 2.7663,
+      "step": 51533
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7035639882087708,
+      "learning_rate": 3.627134552136769e-05,
+      "loss": 2.8127,
+      "step": 51534
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7026732563972473,
+      "learning_rate": 3.6264007341490145e-05,
+      "loss": 2.805,
+      "step": 51535
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7547394633293152,
+      "learning_rate": 3.6256669856242516e-05,
+      "loss": 2.8502,
+      "step": 51536
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7437660694122314,
+      "learning_rate": 3.624933306564399e-05,
+      "loss": 2.8254,
+      "step": 51537
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7333990335464478,
+      "learning_rate": 3.624199696971403e-05,
+      "loss": 2.9906,
+      "step": 51538
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7599939107894897,
+      "learning_rate": 3.62346615684719e-05,
+      "loss": 2.7959,
+      "step": 51539
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8379149436950684,
+      "learning_rate": 3.622732686193683e-05,
+      "loss": 2.9625,
+      "step": 51540
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7774891257286072,
+      "learning_rate": 3.621999285012828e-05,
+      "loss": 2.7655,
+      "step": 51541
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7588175535202026,
+      "learning_rate": 3.621265953306545e-05,
+      "loss": 2.7889,
+      "step": 51542
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7654353380203247,
+      "learning_rate": 3.62053269107677e-05,
+      "loss": 2.9883,
+      "step": 51543
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.744993269443512,
+      "learning_rate": 3.619799498325441e-05,
+      "loss": 2.7745,
+      "step": 51544
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7357940077781677,
+      "learning_rate": 3.6190663750544833e-05,
+      "loss": 2.6419,
+      "step": 51545
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.737634539604187,
+      "learning_rate": 3.6183333212658294e-05,
+      "loss": 2.9255,
+      "step": 51546
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7712468504905701,
+      "learning_rate": 3.6176003369614e-05,
+      "loss": 3.0007,
+      "step": 51547
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7073559761047363,
+      "learning_rate": 3.616867422143135e-05,
+      "loss": 2.9726,
+      "step": 51548
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7305967807769775,
+      "learning_rate": 3.616134576812971e-05,
+      "loss": 2.8264,
+      "step": 51549
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7160578370094299,
+      "learning_rate": 3.6154018009728216e-05,
+      "loss": 2.9483,
+      "step": 51550
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7831854820251465,
+      "learning_rate": 3.614669094624634e-05,
+      "loss": 2.7699,
+      "step": 51551
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7335165739059448,
+      "learning_rate": 3.61393645777033e-05,
+      "loss": 3.0746,
+      "step": 51552
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8146199584007263,
+      "learning_rate": 3.6132038904118356e-05,
+      "loss": 2.8505,
+      "step": 51553
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7218393087387085,
+      "learning_rate": 3.6124713925510854e-05,
+      "loss": 2.8958,
+      "step": 51554
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7315160036087036,
+      "learning_rate": 3.6117389641900044e-05,
+      "loss": 2.7957,
+      "step": 51555
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7376806735992432,
+      "learning_rate": 3.611006605330527e-05,
+      "loss": 2.8755,
+      "step": 51556
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8134010434150696,
+      "learning_rate": 3.610274315974577e-05,
+      "loss": 2.8359,
+      "step": 51557
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7388611435890198,
+      "learning_rate": 3.6095420961240854e-05,
+      "loss": 2.6661,
+      "step": 51558
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7700594663619995,
+      "learning_rate": 3.608809945780986e-05,
+      "loss": 2.9496,
+      "step": 51559
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7100481986999512,
+      "learning_rate": 3.608077864947202e-05,
+      "loss": 2.7765,
+      "step": 51560
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7788733243942261,
+      "learning_rate": 3.6073458536246616e-05,
+      "loss": 3.0668,
+      "step": 51561
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7740787863731384,
+      "learning_rate": 3.6066139118152894e-05,
+      "loss": 3.0295,
+      "step": 51562
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7355701327323914,
+      "learning_rate": 3.6058820395210105e-05,
+      "loss": 2.6675,
+      "step": 51563
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7832901477813721,
+      "learning_rate": 3.605150236743771e-05,
+      "loss": 2.7939,
+      "step": 51564
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.725982129573822,
+      "learning_rate": 3.604418503485476e-05,
+      "loss": 2.9671,
+      "step": 51565
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7156925797462463,
+      "learning_rate": 3.603686839748072e-05,
+      "loss": 2.908,
+      "step": 51566
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.801988959312439,
+      "learning_rate": 3.602955245533473e-05,
+      "loss": 2.8038,
+      "step": 51567
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7519605159759521,
+      "learning_rate": 3.602223720843607e-05,
+      "loss": 2.9654,
+      "step": 51568
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7388525605201721,
+      "learning_rate": 3.601492265680408e-05,
+      "loss": 3.1603,
+      "step": 51569
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7522304654121399,
+      "learning_rate": 3.600760880045792e-05,
+      "loss": 2.9316,
+      "step": 51570
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7278318405151367,
+      "learning_rate": 3.6000295639416974e-05,
+      "loss": 2.791,
+      "step": 51571
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7324314117431641,
+      "learning_rate": 3.599298317370036e-05,
+      "loss": 3.0109,
+      "step": 51572
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7239487767219543,
+      "learning_rate": 3.598567140332747e-05,
+      "loss": 2.8985,
+      "step": 51573
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.731003999710083,
+      "learning_rate": 3.597836032831754e-05,
+      "loss": 2.9196,
+      "step": 51574
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.746536910533905,
+      "learning_rate": 3.597104994868979e-05,
+      "loss": 2.985,
+      "step": 51575
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7338457703590393,
+      "learning_rate": 3.596374026446348e-05,
+      "loss": 2.9579,
+      "step": 51576
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8143104314804077,
+      "learning_rate": 3.595643127565779e-05,
+      "loss": 2.9791,
+      "step": 51577
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7997439503669739,
+      "learning_rate": 3.594912298229208e-05,
+      "loss": 2.8689,
+      "step": 51578
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.755618691444397,
+      "learning_rate": 3.594181538438562e-05,
+      "loss": 2.9621,
+      "step": 51579
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7650567293167114,
+      "learning_rate": 3.59345084819575e-05,
+      "loss": 2.8828,
+      "step": 51580
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7664803266525269,
+      "learning_rate": 3.5927202275027176e-05,
+      "loss": 3.0211,
+      "step": 51581
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6953228116035461,
+      "learning_rate": 3.591989676361373e-05,
+      "loss": 2.8972,
+      "step": 51582
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7946908473968506,
+      "learning_rate": 3.591259194773641e-05,
+      "loss": 2.993,
+      "step": 51583
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.716802179813385,
+      "learning_rate": 3.5905287827414574e-05,
+      "loss": 2.9496,
+      "step": 51584
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7397177219390869,
+      "learning_rate": 3.589798440266734e-05,
+      "loss": 2.8957,
+      "step": 51585
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7066813707351685,
+      "learning_rate": 3.589068167351401e-05,
+      "loss": 2.8603,
+      "step": 51586
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7754790186882019,
+      "learning_rate": 3.588337963997375e-05,
+      "loss": 3.0405,
+      "step": 51587
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7783805131912231,
+      "learning_rate": 3.587607830206588e-05,
+      "loss": 3.0427,
+      "step": 51588
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7516245245933533,
+      "learning_rate": 3.586877765980961e-05,
+      "loss": 2.9936,
+      "step": 51589
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7606446146965027,
+      "learning_rate": 3.5861477713224096e-05,
+      "loss": 3.0245,
+      "step": 51590
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7204053401947021,
+      "learning_rate": 3.5854178462328654e-05,
+      "loss": 2.9596,
+      "step": 51591
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7473350763320923,
+      "learning_rate": 3.58468799071424e-05,
+      "loss": 2.8518,
+      "step": 51592
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7622993588447571,
+      "learning_rate": 3.583958204768472e-05,
+      "loss": 2.7923,
+      "step": 51593
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7171756029129028,
+      "learning_rate": 3.583228488397464e-05,
+      "loss": 2.9466,
+      "step": 51594
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7305697202682495,
+      "learning_rate": 3.582498841603153e-05,
+      "loss": 2.7989,
+      "step": 51595
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7551934123039246,
+      "learning_rate": 3.581769264387459e-05,
+      "loss": 2.8935,
+      "step": 51596
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7103335857391357,
+      "learning_rate": 3.581039756752302e-05,
+      "loss": 2.8689,
+      "step": 51597
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7509092688560486,
+      "learning_rate": 3.5803103186995985e-05,
+      "loss": 3.0213,
+      "step": 51598
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7707995772361755,
+      "learning_rate": 3.57958095023127e-05,
+      "loss": 2.9597,
+      "step": 51599
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7239026427268982,
+      "learning_rate": 3.578851651349238e-05,
+      "loss": 2.9482,
+      "step": 51600
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.699456512928009,
+      "learning_rate": 3.578122422055435e-05,
+      "loss": 3.0693,
+      "step": 51601
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7883853316307068,
+      "learning_rate": 3.577393262351766e-05,
+      "loss": 2.8765,
+      "step": 51602
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7522447109222412,
+      "learning_rate": 3.576664172240165e-05,
+      "loss": 2.7847,
+      "step": 51603
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8980352282524109,
+      "learning_rate": 3.575935151722541e-05,
+      "loss": 3.0591,
+      "step": 51604
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7412164211273193,
+      "learning_rate": 3.5752062008008163e-05,
+      "loss": 2.887,
+      "step": 51605
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7191179990768433,
+      "learning_rate": 3.574477319476919e-05,
+      "loss": 3.2386,
+      "step": 51606
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8027909398078918,
+      "learning_rate": 3.573748507752755e-05,
+      "loss": 2.7148,
+      "step": 51607
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7330635786056519,
+      "learning_rate": 3.5730197656302585e-05,
+      "loss": 2.9588,
+      "step": 51608
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7735836505889893,
+      "learning_rate": 3.5722910931113345e-05,
+      "loss": 3.1469,
+      "step": 51609
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7595041990280151,
+      "learning_rate": 3.5715624901979154e-05,
+      "loss": 2.7322,
+      "step": 51610
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.683493435382843,
+      "learning_rate": 3.5708339568919164e-05,
+      "loss": 2.8669,
+      "step": 51611
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7193924188613892,
+      "learning_rate": 3.570105493195249e-05,
+      "loss": 2.9912,
+      "step": 51612
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7553349137306213,
+      "learning_rate": 3.5693770991098417e-05,
+      "loss": 3.0392,
+      "step": 51613
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7322878837585449,
+      "learning_rate": 3.5686487746376027e-05,
+      "loss": 2.989,
+      "step": 51614
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7660918235778809,
+      "learning_rate": 3.567920519780454e-05,
+      "loss": 2.9561,
+      "step": 51615
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7109578847885132,
+      "learning_rate": 3.567192334540324e-05,
+      "loss": 2.8681,
+      "step": 51616
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.743213415145874,
+      "learning_rate": 3.566464218919115e-05,
+      "loss": 2.8795,
+      "step": 51617
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7197250723838806,
+      "learning_rate": 3.565736172918762e-05,
+      "loss": 2.8877,
+      "step": 51618
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.692306399345398,
+      "learning_rate": 3.565008196541169e-05,
+      "loss": 2.9454,
+      "step": 51619
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7972225546836853,
+      "learning_rate": 3.564280289788249e-05,
+      "loss": 2.827,
+      "step": 51620
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7207708358764648,
+      "learning_rate": 3.563552452661938e-05,
+      "loss": 2.9103,
+      "step": 51621
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7996771931648254,
+      "learning_rate": 3.562824685164132e-05,
+      "loss": 2.7235,
+      "step": 51622
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6930966973304749,
+      "learning_rate": 3.5620969872967644e-05,
+      "loss": 2.818,
+      "step": 51623
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6936909556388855,
+      "learning_rate": 3.56136935906174e-05,
+      "loss": 2.8292,
+      "step": 51624
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7523632049560547,
+      "learning_rate": 3.560641800460987e-05,
+      "loss": 2.9122,
+      "step": 51625
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7502445578575134,
+      "learning_rate": 3.5599143114964143e-05,
+      "loss": 2.8885,
+      "step": 51626
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7347469329833984,
+      "learning_rate": 3.559186892169934e-05,
+      "loss": 2.8755,
+      "step": 51627
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7211341857910156,
+      "learning_rate": 3.558459542483474e-05,
+      "loss": 3.0235,
+      "step": 51628
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7356860637664795,
+      "learning_rate": 3.557732262438932e-05,
+      "loss": 2.7142,
+      "step": 51629
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7704647779464722,
+      "learning_rate": 3.5570050520382386e-05,
+      "loss": 2.9018,
+      "step": 51630
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6986088156700134,
+      "learning_rate": 3.55627791128331e-05,
+      "loss": 2.9044,
+      "step": 51631
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8135074377059937,
+      "learning_rate": 3.555550840176057e-05,
+      "loss": 2.9279,
+      "step": 51632
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7185907959938049,
+      "learning_rate": 3.554823838718393e-05,
+      "loss": 2.9699,
+      "step": 51633
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7311407327651978,
+      "learning_rate": 3.554096906912226e-05,
+      "loss": 2.9253,
+      "step": 51634
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7427823543548584,
+      "learning_rate": 3.553370044759476e-05,
+      "loss": 2.9968,
+      "step": 51635
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7487204074859619,
+      "learning_rate": 3.5526432522620704e-05,
+      "loss": 2.8834,
+      "step": 51636
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.745222806930542,
+      "learning_rate": 3.551916529421906e-05,
+      "loss": 2.8098,
+      "step": 51637
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7326517105102539,
+      "learning_rate": 3.551189876240905e-05,
+      "loss": 2.9652,
+      "step": 51638
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8086389899253845,
+      "learning_rate": 3.5504632927209786e-05,
+      "loss": 3.0676,
+      "step": 51639
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7548852562904358,
+      "learning_rate": 3.5497367788640474e-05,
+      "loss": 3.0023,
+      "step": 51640
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7331960201263428,
+      "learning_rate": 3.5490103346720166e-05,
+      "loss": 2.9562,
+      "step": 51641
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7696003913879395,
+      "learning_rate": 3.548283960146795e-05,
+      "loss": 2.9089,
+      "step": 51642
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6968817710876465,
+      "learning_rate": 3.5475576552903105e-05,
+      "loss": 3.0244,
+      "step": 51643
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7616210579872131,
+      "learning_rate": 3.5468314201044624e-05,
+      "loss": 2.6087,
+      "step": 51644
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8356584310531616,
+      "learning_rate": 3.546105254591172e-05,
+      "loss": 2.9079,
+      "step": 51645
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7322165369987488,
+      "learning_rate": 3.5453791587523504e-05,
+      "loss": 2.9639,
+      "step": 51646
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7493136525154114,
+      "learning_rate": 3.544653132589911e-05,
+      "loss": 2.6329,
+      "step": 51647
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7908979654312134,
+      "learning_rate": 3.5439271761057645e-05,
+      "loss": 2.7651,
+      "step": 51648
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7793807983398438,
+      "learning_rate": 3.543201289301816e-05,
+      "loss": 2.8588,
+      "step": 51649
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7199450135231018,
+      "learning_rate": 3.542475472179981e-05,
+      "loss": 3.1165,
+      "step": 51650
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7530863285064697,
+      "learning_rate": 3.5417497247421844e-05,
+      "loss": 2.7485,
+      "step": 51651
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7242304086685181,
+      "learning_rate": 3.541024046990318e-05,
+      "loss": 3.0574,
+      "step": 51652
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7129125595092773,
+      "learning_rate": 3.540298438926308e-05,
+      "loss": 2.7127,
+      "step": 51653
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7528355717658997,
+      "learning_rate": 3.539572900552061e-05,
+      "loss": 2.8536,
+      "step": 51654
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7420986890792847,
+      "learning_rate": 3.53884743186948e-05,
+      "loss": 2.8166,
+      "step": 51655
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7624697089195251,
+      "learning_rate": 3.538122032880487e-05,
+      "loss": 3.1127,
+      "step": 51656
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7528888583183289,
+      "learning_rate": 3.537396703586983e-05,
+      "loss": 3.0215,
+      "step": 51657
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7675604224205017,
+      "learning_rate": 3.5366714439908905e-05,
+      "loss": 2.966,
+      "step": 51658
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7084840536117554,
+      "learning_rate": 3.535946254094104e-05,
+      "loss": 2.7934,
+      "step": 51659
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7773526310920715,
+      "learning_rate": 3.5352211338985425e-05,
+      "loss": 2.9984,
+      "step": 51660
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7066696882247925,
+      "learning_rate": 3.534496083406121e-05,
+      "loss": 2.8627,
+      "step": 51661
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7332533597946167,
+      "learning_rate": 3.533771102618744e-05,
+      "loss": 2.885,
+      "step": 51662
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7373495697975159,
+      "learning_rate": 3.533046191538321e-05,
+      "loss": 3.0135,
+      "step": 51663
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7333385944366455,
+      "learning_rate": 3.532321350166757e-05,
+      "loss": 2.8784,
+      "step": 51664
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.726533055305481,
+      "learning_rate": 3.531596578505966e-05,
+      "loss": 2.9567,
+      "step": 51665
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7740032076835632,
+      "learning_rate": 3.530871876557852e-05,
+      "loss": 2.9343,
+      "step": 51666
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7073380351066589,
+      "learning_rate": 3.530147244324327e-05,
+      "loss": 2.7355,
+      "step": 51667
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7208665013313293,
+      "learning_rate": 3.5294226818073056e-05,
+      "loss": 2.9038,
+      "step": 51668
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7397687435150146,
+      "learning_rate": 3.528698189008692e-05,
+      "loss": 2.8913,
+      "step": 51669
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7066610455513,
+      "learning_rate": 3.527973765930394e-05,
+      "loss": 2.8714,
+      "step": 51670
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7497684359550476,
+      "learning_rate": 3.527249412574308e-05,
+      "loss": 2.8729,
+      "step": 51671
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7150473594665527,
+      "learning_rate": 3.5265251289423544e-05,
+      "loss": 2.8148,
+      "step": 51672
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7562102675437927,
+      "learning_rate": 3.525800915036446e-05,
+      "loss": 2.991,
+      "step": 51673
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7330208420753479,
+      "learning_rate": 3.525076770858475e-05,
+      "loss": 2.917,
+      "step": 51674
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7095870971679688,
+      "learning_rate": 3.5243526964103655e-05,
+      "loss": 2.8007,
+      "step": 51675
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7798789739608765,
+      "learning_rate": 3.523628691694006e-05,
+      "loss": 2.918,
+      "step": 51676
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7631597518920898,
+      "learning_rate": 3.522904756711322e-05,
+      "loss": 2.9471,
+      "step": 51677
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8316076993942261,
+      "learning_rate": 3.522180891464209e-05,
+      "loss": 3.0965,
+      "step": 51678
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.739554762840271,
+      "learning_rate": 3.521457095954572e-05,
+      "loss": 2.9548,
+      "step": 51679
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7704421877861023,
+      "learning_rate": 3.5207333701843255e-05,
+      "loss": 2.9717,
+      "step": 51680
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7329279780387878,
+      "learning_rate": 3.520009714155365e-05,
+      "loss": 2.9602,
+      "step": 51681
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7119458913803101,
+      "learning_rate": 3.5192861278696025e-05,
+      "loss": 2.7271,
+      "step": 51682
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7648967504501343,
+      "learning_rate": 3.51856261132895e-05,
+      "loss": 2.7971,
+      "step": 51683
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7791581749916077,
+      "learning_rate": 3.5178391645353085e-05,
+      "loss": 2.8109,
+      "step": 51684
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7299320697784424,
+      "learning_rate": 3.517115787490581e-05,
+      "loss": 2.6825,
+      "step": 51685
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7683125734329224,
+      "learning_rate": 3.5163924801966684e-05,
+      "loss": 2.733,
+      "step": 51686
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7981067895889282,
+      "learning_rate": 3.515669242655482e-05,
+      "loss": 2.8742,
+      "step": 51687
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7027236819267273,
+      "learning_rate": 3.5149460748689315e-05,
+      "loss": 2.9845,
+      "step": 51688
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7060105204582214,
+      "learning_rate": 3.5142229768389085e-05,
+      "loss": 2.709,
+      "step": 51689
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7767622470855713,
+      "learning_rate": 3.513499948567331e-05,
+      "loss": 2.9245,
+      "step": 51690
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7285616993904114,
+      "learning_rate": 3.512776990056098e-05,
+      "loss": 3.0044,
+      "step": 51691
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7438006401062012,
+      "learning_rate": 3.512054101307107e-05,
+      "loss": 2.9719,
+      "step": 51692
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7528930306434631,
+      "learning_rate": 3.511331282322274e-05,
+      "loss": 3.2001,
+      "step": 51693
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7622228860855103,
+      "learning_rate": 3.5106085331034873e-05,
+      "loss": 2.8069,
+      "step": 51694
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8420156836509705,
+      "learning_rate": 3.509885853652668e-05,
+      "loss": 2.8725,
+      "step": 51695
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7685613632202148,
+      "learning_rate": 3.5091632439717065e-05,
+      "loss": 3.0688,
+      "step": 51696
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7986699342727661,
+      "learning_rate": 3.5084407040625094e-05,
+      "loss": 2.8336,
+      "step": 51697
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8661927580833435,
+      "learning_rate": 3.507718233926989e-05,
+      "loss": 2.9024,
+      "step": 51698
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8043288588523865,
+      "learning_rate": 3.506995833567038e-05,
+      "loss": 2.8473,
+      "step": 51699
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7660233974456787,
+      "learning_rate": 3.5062735029845634e-05,
+      "loss": 2.9776,
+      "step": 51700
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8327121734619141,
+      "learning_rate": 3.505551242181458e-05,
+      "loss": 2.9838,
+      "step": 51701
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.786426842212677,
+      "learning_rate": 3.5048290511596334e-05,
+      "loss": 2.7796,
+      "step": 51702
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7408648133277893,
+      "learning_rate": 3.5041069299209944e-05,
+      "loss": 3.0313,
+      "step": 51703
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7075411677360535,
+      "learning_rate": 3.50338487846743e-05,
+      "loss": 2.8537,
+      "step": 51704
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7919765114784241,
+      "learning_rate": 3.5026628968008616e-05,
+      "loss": 3.0262,
+      "step": 51705
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.726375937461853,
+      "learning_rate": 3.501940984923175e-05,
+      "loss": 2.9436,
+      "step": 51706
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7896806597709656,
+      "learning_rate": 3.501219142836275e-05,
+      "loss": 3.012,
+      "step": 51707
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.741202175617218,
+      "learning_rate": 3.500497370542066e-05,
+      "loss": 2.7639,
+      "step": 51708
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7486560940742493,
+      "learning_rate": 3.4997756680424404e-05,
+      "loss": 3.1071,
+      "step": 51709
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7122935652732849,
+      "learning_rate": 3.4990540353393136e-05,
+      "loss": 2.9733,
+      "step": 51710
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.739435076713562,
+      "learning_rate": 3.498332472434574e-05,
+      "loss": 2.9056,
+      "step": 51711
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7047844529151917,
+      "learning_rate": 3.49761097933013e-05,
+      "loss": 2.8209,
+      "step": 51712
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7015252113342285,
+      "learning_rate": 3.4968895560278796e-05,
+      "loss": 2.9192,
+      "step": 51713
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8801385760307312,
+      "learning_rate": 3.4961682025297125e-05,
+      "loss": 3.0679,
+      "step": 51714
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7651318311691284,
+      "learning_rate": 3.495446918837545e-05,
+      "loss": 2.8485,
+      "step": 51715
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7404123544692993,
+      "learning_rate": 3.494725704953265e-05,
+      "loss": 2.8387,
+      "step": 51716
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7318333387374878,
+      "learning_rate": 3.494004560878776e-05,
+      "loss": 2.6693,
+      "step": 51717
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7572652697563171,
+      "learning_rate": 3.493283486615983e-05,
+      "loss": 2.8279,
+      "step": 51718
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7449601888656616,
+      "learning_rate": 3.4925624821667755e-05,
+      "loss": 2.6999,
+      "step": 51719
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7463000416755676,
+      "learning_rate": 3.4918415475330606e-05,
+      "loss": 2.8917,
+      "step": 51720
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7627390623092651,
+      "learning_rate": 3.491120682716735e-05,
+      "loss": 3.0798,
+      "step": 51721
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.75359708070755,
+      "learning_rate": 3.490399887719689e-05,
+      "loss": 2.8438,
+      "step": 51722
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6942808628082275,
+      "learning_rate": 3.489679162543836e-05,
+      "loss": 2.7695,
+      "step": 51723
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7942733764648438,
+      "learning_rate": 3.488958507191057e-05,
+      "loss": 3.0674,
+      "step": 51724
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7533266544342041,
+      "learning_rate": 3.488237921663266e-05,
+      "loss": 3.0008,
+      "step": 51725
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.760011613368988,
+      "learning_rate": 3.48751740596235e-05,
+      "loss": 2.8886,
+      "step": 51726
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7659701704978943,
+      "learning_rate": 3.4867969600902135e-05,
+      "loss": 2.9795,
+      "step": 51727
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7460302710533142,
+      "learning_rate": 3.4860765840487545e-05,
+      "loss": 2.9011,
+      "step": 51728
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7115795016288757,
+      "learning_rate": 3.485356277839859e-05,
+      "loss": 3.0382,
+      "step": 51729
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7279272675514221,
+      "learning_rate": 3.4846360414654384e-05,
+      "loss": 3.0661,
+      "step": 51730
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7116730809211731,
+      "learning_rate": 3.483915874927378e-05,
+      "loss": 3.0457,
+      "step": 51731
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7574999928474426,
+      "learning_rate": 3.4831957782275785e-05,
+      "loss": 3.053,
+      "step": 51732
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.752925455570221,
+      "learning_rate": 3.4824757513679466e-05,
+      "loss": 2.9957,
+      "step": 51733
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7213291525840759,
+      "learning_rate": 3.481755794350367e-05,
+      "loss": 2.759,
+      "step": 51734
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7119690775871277,
+      "learning_rate": 3.481035907176741e-05,
+      "loss": 2.8161,
+      "step": 51735
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6793041229248047,
+      "learning_rate": 3.480316089848958e-05,
+      "loss": 2.7194,
+      "step": 51736
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7623587250709534,
+      "learning_rate": 3.479596342368919e-05,
+      "loss": 2.6806,
+      "step": 51737
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7153224349021912,
+      "learning_rate": 3.478876664738522e-05,
+      "loss": 2.838,
+      "step": 51738
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7242084741592407,
+      "learning_rate": 3.478157056959657e-05,
+      "loss": 2.9762,
+      "step": 51739
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7159818410873413,
+      "learning_rate": 3.477437519034225e-05,
+      "loss": 2.732,
+      "step": 51740
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7122927904129028,
+      "learning_rate": 3.4767180509641144e-05,
+      "loss": 2.8784,
+      "step": 51741
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.745762825012207,
+      "learning_rate": 3.475998652751231e-05,
+      "loss": 2.8922,
+      "step": 51742
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7402251958847046,
+      "learning_rate": 3.4752793243974584e-05,
+      "loss": 3.2264,
+      "step": 51743
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7608932256698608,
+      "learning_rate": 3.4745600659046936e-05,
+      "loss": 2.9075,
+      "step": 51744
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7457864284515381,
+      "learning_rate": 3.473840877274837e-05,
+      "loss": 2.9522,
+      "step": 51745
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7215996384620667,
+      "learning_rate": 3.4731217585097716e-05,
+      "loss": 3.1533,
+      "step": 51746
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7540760636329651,
+      "learning_rate": 3.472402709611405e-05,
+      "loss": 3.0193,
+      "step": 51747
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7310689091682434,
+      "learning_rate": 3.471683730581619e-05,
+      "loss": 2.9448,
+      "step": 51748
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7167690992355347,
+      "learning_rate": 3.4709648214223166e-05,
+      "loss": 2.8068,
+      "step": 51749
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7109416127204895,
+      "learning_rate": 3.470245982135388e-05,
+      "loss": 2.8514,
+      "step": 51750
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7370871305465698,
+      "learning_rate": 3.469527212722719e-05,
+      "loss": 2.8947,
+      "step": 51751
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7531648278236389,
+      "learning_rate": 3.468808513186219e-05,
+      "loss": 2.9873,
+      "step": 51752
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7331883311271667,
+      "learning_rate": 3.468089883527762e-05,
+      "loss": 2.7971,
+      "step": 51753
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7884783744812012,
+      "learning_rate": 3.4673713237492496e-05,
+      "loss": 2.839,
+      "step": 51754
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7267898321151733,
+      "learning_rate": 3.4666528338525846e-05,
+      "loss": 2.8982,
+      "step": 51755
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7529955506324768,
+      "learning_rate": 3.465934413839645e-05,
+      "loss": 3.0185,
+      "step": 51756
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7315231561660767,
+      "learning_rate": 3.4652160637123327e-05,
+      "loss": 2.8829,
+      "step": 51757
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7301185131072998,
+      "learning_rate": 3.464497783472523e-05,
+      "loss": 2.7218,
+      "step": 51758
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7687699198722839,
+      "learning_rate": 3.463779573122124e-05,
+      "loss": 3.0069,
+      "step": 51759
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7508442401885986,
+      "learning_rate": 3.463061432663025e-05,
+      "loss": 3.0149,
+      "step": 51760
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7293921113014221,
+      "learning_rate": 3.462343362097113e-05,
+      "loss": 2.8922,
+      "step": 51761
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7335087656974792,
+      "learning_rate": 3.4616253614262846e-05,
+      "loss": 2.7265,
+      "step": 51762
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6986225247383118,
+      "learning_rate": 3.460907430652421e-05,
+      "loss": 3.1249,
+      "step": 51763
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7187238931655884,
+      "learning_rate": 3.4601895697774274e-05,
+      "loss": 2.8012,
+      "step": 51764
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7392198443412781,
+      "learning_rate": 3.4594717788031853e-05,
+      "loss": 2.8605,
+      "step": 51765
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7785656452178955,
+      "learning_rate": 3.458754057731581e-05,
+      "loss": 2.8712,
+      "step": 51766
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7567797899246216,
+      "learning_rate": 3.458036406564518e-05,
+      "loss": 2.7214,
+      "step": 51767
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8188661932945251,
+      "learning_rate": 3.4573188253038696e-05,
+      "loss": 3.044,
+      "step": 51768
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8261837959289551,
+      "learning_rate": 3.45660131395154e-05,
+      "loss": 2.8088,
+      "step": 51769
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7429391145706177,
+      "learning_rate": 3.4558838725094154e-05,
+      "loss": 2.9782,
+      "step": 51770
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7897228002548218,
+      "learning_rate": 3.45516650097939e-05,
+      "loss": 2.9229,
+      "step": 51771
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8158841729164124,
+      "learning_rate": 3.4544491993633415e-05,
+      "loss": 3.0301,
+      "step": 51772
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7363582253456116,
+      "learning_rate": 3.45373196766316e-05,
+      "loss": 3.0389,
+      "step": 51773
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7058192491531372,
+      "learning_rate": 3.453014805880744e-05,
+      "loss": 2.7813,
+      "step": 51774
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7266924977302551,
+      "learning_rate": 3.452297714017981e-05,
+      "loss": 2.8937,
+      "step": 51775
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7514150738716125,
+      "learning_rate": 3.45158069207675e-05,
+      "loss": 3.0089,
+      "step": 51776
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7654863595962524,
+      "learning_rate": 3.4508637400589526e-05,
+      "loss": 2.95,
+      "step": 51777
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7285271286964417,
+      "learning_rate": 3.45014685796647e-05,
+      "loss": 2.9363,
+      "step": 51778
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7652084827423096,
+      "learning_rate": 3.4494300458011856e-05,
+      "loss": 2.9796,
+      "step": 51779
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7151200771331787,
+      "learning_rate": 3.448713303564997e-05,
+      "loss": 3.1283,
+      "step": 51780
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7224323749542236,
+      "learning_rate": 3.447996631259783e-05,
+      "loss": 3.054,
+      "step": 51781
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7343427538871765,
+      "learning_rate": 3.447280028887445e-05,
+      "loss": 2.9277,
+      "step": 51782
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7346922755241394,
+      "learning_rate": 3.446563496449851e-05,
+      "loss": 3.0434,
+      "step": 51783
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7594969868659973,
+      "learning_rate": 3.445847033948901e-05,
+      "loss": 2.8904,
+      "step": 51784
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7676674723625183,
+      "learning_rate": 3.445130641386482e-05,
+      "loss": 2.7869,
+      "step": 51785
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7546485066413879,
+      "learning_rate": 3.44441431876448e-05,
+      "loss": 2.9141,
+      "step": 51786
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7444491982460022,
+      "learning_rate": 3.4436980660847836e-05,
+      "loss": 2.9501,
+      "step": 51787
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7077409625053406,
+      "learning_rate": 3.442981883349264e-05,
+      "loss": 2.9668,
+      "step": 51788
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7607021331787109,
+      "learning_rate": 3.442265770559823e-05,
+      "loss": 2.7839,
+      "step": 51789
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.790345311164856,
+      "learning_rate": 3.441549727718347e-05,
+      "loss": 2.909,
+      "step": 51790
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7288457155227661,
+      "learning_rate": 3.440833754826713e-05,
+      "loss": 2.85,
+      "step": 51791
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7285025119781494,
+      "learning_rate": 3.4401178518868163e-05,
+      "loss": 3.0604,
+      "step": 51792
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7310810685157776,
+      "learning_rate": 3.4394020189005364e-05,
+      "loss": 2.8663,
+      "step": 51793
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.725584089756012,
+      "learning_rate": 3.438686255869758e-05,
+      "loss": 3.018,
+      "step": 51794
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7113367319107056,
+      "learning_rate": 3.43797056279637e-05,
+      "loss": 2.9908,
+      "step": 51795
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8494749665260315,
+      "learning_rate": 3.4372549396822526e-05,
+      "loss": 2.9683,
+      "step": 51796
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7407485246658325,
+      "learning_rate": 3.436539386529299e-05,
+      "loss": 2.7897,
+      "step": 51797
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7727249264717102,
+      "learning_rate": 3.4358239033393806e-05,
+      "loss": 2.8892,
+      "step": 51798
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8405013084411621,
+      "learning_rate": 3.43510849011439e-05,
+      "loss": 3.0491,
+      "step": 51799
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7366344332695007,
+      "learning_rate": 3.434393146856218e-05,
+      "loss": 2.8452,
+      "step": 51800
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7424528002738953,
+      "learning_rate": 3.433677873566743e-05,
+      "loss": 2.9742,
+      "step": 51801
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.721091091632843,
+      "learning_rate": 3.4329626702478455e-05,
+      "loss": 2.9626,
+      "step": 51802
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.770412027835846,
+      "learning_rate": 3.432247536901408e-05,
+      "loss": 2.8381,
+      "step": 51803
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7273968458175659,
+      "learning_rate": 3.431532473529314e-05,
+      "loss": 2.6277,
+      "step": 51804
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7612076997756958,
+      "learning_rate": 3.430817480133458e-05,
+      "loss": 2.9437,
+      "step": 51805
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7737910151481628,
+      "learning_rate": 3.4301025567157084e-05,
+      "loss": 3.02,
+      "step": 51806
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8094823956489563,
+      "learning_rate": 3.429387703277964e-05,
+      "loss": 2.9737,
+      "step": 51807
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7228150963783264,
+      "learning_rate": 3.428672919822096e-05,
+      "loss": 2.7584,
+      "step": 51808
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7346405982971191,
+      "learning_rate": 3.4279582063499845e-05,
+      "loss": 2.948,
+      "step": 51809
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7100589275360107,
+      "learning_rate": 3.427243562863523e-05,
+      "loss": 2.7957,
+      "step": 51810
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7395407557487488,
+      "learning_rate": 3.426528989364584e-05,
+      "loss": 2.9512,
+      "step": 51811
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7701348066329956,
+      "learning_rate": 3.425814485855056e-05,
+      "loss": 3.0815,
+      "step": 51812
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8105944991111755,
+      "learning_rate": 3.425100052336814e-05,
+      "loss": 2.8488,
+      "step": 51813
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7398055791854858,
+      "learning_rate": 3.424385688811751e-05,
+      "loss": 2.9291,
+      "step": 51814
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7211189866065979,
+      "learning_rate": 3.4236713952817394e-05,
+      "loss": 2.9474,
+      "step": 51815
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7564665675163269,
+      "learning_rate": 3.4229571717486606e-05,
+      "loss": 2.7423,
+      "step": 51816
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7213353514671326,
+      "learning_rate": 3.4222430182144015e-05,
+      "loss": 2.8272,
+      "step": 51817
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7244040966033936,
+      "learning_rate": 3.42152893468083e-05,
+      "loss": 2.907,
+      "step": 51818
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6921009421348572,
+      "learning_rate": 3.420814921149847e-05,
+      "loss": 2.9494,
+      "step": 51819
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7488122582435608,
+      "learning_rate": 3.420100977623312e-05,
+      "loss": 2.8671,
+      "step": 51820
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7758278846740723,
+      "learning_rate": 3.419387104103118e-05,
+      "loss": 2.7779,
+      "step": 51821
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7232535481452942,
+      "learning_rate": 3.41867330059115e-05,
+      "loss": 2.8373,
+      "step": 51822
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.716827392578125,
+      "learning_rate": 3.41795956708928e-05,
+      "loss": 2.9449,
+      "step": 51823
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7418515086174011,
+      "learning_rate": 3.417245903599389e-05,
+      "loss": 2.9679,
+      "step": 51824
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7049885392189026,
+      "learning_rate": 3.41653231012335e-05,
+      "loss": 2.8945,
+      "step": 51825
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.757987916469574,
+      "learning_rate": 3.415818786663047e-05,
+      "loss": 2.8117,
+      "step": 51826
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7701760530471802,
+      "learning_rate": 3.415105333220369e-05,
+      "loss": 2.8238,
+      "step": 51827
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.75865238904953,
+      "learning_rate": 3.414391949797182e-05,
+      "loss": 3.0257,
+      "step": 51828
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.698841392993927,
+      "learning_rate": 3.413678636395373e-05,
+      "loss": 2.8413,
+      "step": 51829
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7989115715026855,
+      "learning_rate": 3.412965393016821e-05,
+      "loss": 2.8112,
+      "step": 51830
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6878560185432434,
+      "learning_rate": 3.412252219663392e-05,
+      "loss": 2.8933,
+      "step": 51831
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7924463748931885,
+      "learning_rate": 3.411539116336983e-05,
+      "loss": 2.914,
+      "step": 51832
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7778879404067993,
+      "learning_rate": 3.410826083039454e-05,
+      "loss": 2.887,
+      "step": 51833
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7213044762611389,
+      "learning_rate": 3.4101131197726996e-05,
+      "loss": 2.6942,
+      "step": 51834
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7068144679069519,
+      "learning_rate": 3.4094002265385846e-05,
+      "loss": 2.758,
+      "step": 51835
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7102383375167847,
+      "learning_rate": 3.4086874033389974e-05,
+      "loss": 3.0547,
+      "step": 51836
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7071000337600708,
+      "learning_rate": 3.407974650175811e-05,
+      "loss": 2.9646,
+      "step": 51837
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7384864687919617,
+      "learning_rate": 3.4072619670508925e-05,
+      "loss": 2.7728,
+      "step": 51838
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.739606499671936,
+      "learning_rate": 3.406549353966135e-05,
+      "loss": 2.8419,
+      "step": 51839
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8046254515647888,
+      "learning_rate": 3.405836810923406e-05,
+      "loss": 3.0422,
+      "step": 51840
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7787489891052246,
+      "learning_rate": 3.405124337924581e-05,
+      "loss": 3.0144,
+      "step": 51841
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7587060332298279,
+      "learning_rate": 3.404411934971548e-05,
+      "loss": 3.011,
+      "step": 51842
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7384954690933228,
+      "learning_rate": 3.403699602066167e-05,
+      "loss": 2.9519,
+      "step": 51843
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7247627973556519,
+      "learning_rate": 3.402987339210331e-05,
+      "loss": 2.9142,
+      "step": 51844
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7663771510124207,
+      "learning_rate": 3.402275146405907e-05,
+      "loss": 2.8239,
+      "step": 51845
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7334901094436646,
+      "learning_rate": 3.401563023654766e-05,
+      "loss": 3.017,
+      "step": 51846
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7264374494552612,
+      "learning_rate": 3.400850970958797e-05,
+      "loss": 2.8086,
+      "step": 51847
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7451301217079163,
+      "learning_rate": 3.4001389883198584e-05,
+      "loss": 2.8351,
+      "step": 51848
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7210157513618469,
+      "learning_rate": 3.3994270757398415e-05,
+      "loss": 2.8837,
+      "step": 51849
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7271472811698914,
+      "learning_rate": 3.398715233220608e-05,
+      "loss": 2.8446,
+      "step": 51850
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7448669075965881,
+      "learning_rate": 3.398003460764044e-05,
+      "loss": 2.896,
+      "step": 51851
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7264134287834167,
+      "learning_rate": 3.397291758372021e-05,
+      "loss": 2.7349,
+      "step": 51852
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7169459462165833,
+      "learning_rate": 3.396580126046403e-05,
+      "loss": 3.0358,
+      "step": 51853
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7395771145820618,
+      "learning_rate": 3.395868563789084e-05,
+      "loss": 2.7717,
+      "step": 51854
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7105684876441956,
+      "learning_rate": 3.395157071601917e-05,
+      "loss": 2.9395,
+      "step": 51855
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7309338450431824,
+      "learning_rate": 3.394445649486785e-05,
+      "loss": 2.8171,
+      "step": 51856
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7164181470870972,
+      "learning_rate": 3.393734297445574e-05,
+      "loss": 2.8214,
+      "step": 51857
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7518510818481445,
+      "learning_rate": 3.393023015480144e-05,
+      "loss": 2.8898,
+      "step": 51858
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7415131330490112,
+      "learning_rate": 3.392311803592368e-05,
+      "loss": 2.9758,
+      "step": 51859
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.746828556060791,
+      "learning_rate": 3.3916006617841175e-05,
+      "loss": 2.955,
+      "step": 51860
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7056711316108704,
+      "learning_rate": 3.390889590057271e-05,
+      "loss": 2.8957,
+      "step": 51861
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8097109198570251,
+      "learning_rate": 3.390178588413707e-05,
+      "loss": 2.9484,
+      "step": 51862
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7409765720367432,
+      "learning_rate": 3.3894676568552834e-05,
+      "loss": 2.7206,
+      "step": 51863
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8118706345558167,
+      "learning_rate": 3.388756795383887e-05,
+      "loss": 2.8796,
+      "step": 51864
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.744475781917572,
+      "learning_rate": 3.3880460040013815e-05,
+      "loss": 2.8114,
+      "step": 51865
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7227175831794739,
+      "learning_rate": 3.387335282709642e-05,
+      "loss": 3.1192,
+      "step": 51866
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7669481635093689,
+      "learning_rate": 3.3866246315105415e-05,
+      "loss": 2.7376,
+      "step": 51867
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7482892274856567,
+      "learning_rate": 3.3859140504059446e-05,
+      "loss": 2.8004,
+      "step": 51868
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7517153024673462,
+      "learning_rate": 3.385203539397733e-05,
+      "loss": 2.7044,
+      "step": 51869
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7399962544441223,
+      "learning_rate": 3.384493098487768e-05,
+      "loss": 3.0025,
+      "step": 51870
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7390328049659729,
+      "learning_rate": 3.383782727677926e-05,
+      "loss": 2.8516,
+      "step": 51871
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7233524322509766,
+      "learning_rate": 3.383072426970085e-05,
+      "loss": 2.804,
+      "step": 51872
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7317050695419312,
+      "learning_rate": 3.382362196366106e-05,
+      "loss": 2.9562,
+      "step": 51873
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7686887383460999,
+      "learning_rate": 3.3816520358678655e-05,
+      "loss": 2.997,
+      "step": 51874
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7436739802360535,
+      "learning_rate": 3.3809419454772214e-05,
+      "loss": 2.8637,
+      "step": 51875
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.701702892780304,
+      "learning_rate": 3.3802319251960554e-05,
+      "loss": 2.7465,
+      "step": 51876
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7306755185127258,
+      "learning_rate": 3.379521975026239e-05,
+      "loss": 2.7137,
+      "step": 51877
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7624945044517517,
+      "learning_rate": 3.378812094969635e-05,
+      "loss": 3.029,
+      "step": 51878
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7094208002090454,
+      "learning_rate": 3.378102285028121e-05,
+      "loss": 3.0725,
+      "step": 51879
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7654349207878113,
+      "learning_rate": 3.377392545203567e-05,
+      "loss": 3.1119,
+      "step": 51880
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7229974865913391,
+      "learning_rate": 3.376682875497826e-05,
+      "loss": 2.8137,
+      "step": 51881
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7172889709472656,
+      "learning_rate": 3.375973275912784e-05,
+      "loss": 2.8475,
+      "step": 51882
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7357591986656189,
+      "learning_rate": 3.375263746450301e-05,
+      "loss": 2.9436,
+      "step": 51883
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7437008619308472,
+      "learning_rate": 3.374554287112256e-05,
+      "loss": 3.1229,
+      "step": 51884
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7146857976913452,
+      "learning_rate": 3.373844897900504e-05,
+      "loss": 2.7566,
+      "step": 51885
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7032164335250854,
+      "learning_rate": 3.3731355788169176e-05,
+      "loss": 2.8108,
+      "step": 51886
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.738189697265625,
+      "learning_rate": 3.372426329863377e-05,
+      "loss": 2.9045,
+      "step": 51887
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7533491253852844,
+      "learning_rate": 3.3717171510417396e-05,
+      "loss": 2.9161,
+      "step": 51888
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7233903408050537,
+      "learning_rate": 3.371008042353874e-05,
+      "loss": 2.9135,
+      "step": 51889
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7648455500602722,
+      "learning_rate": 3.370299003801643e-05,
+      "loss": 2.9126,
+      "step": 51890
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.742719829082489,
+      "learning_rate": 3.369590035386921e-05,
+      "loss": 2.9763,
+      "step": 51891
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7518740892410278,
+      "learning_rate": 3.368881137111577e-05,
+      "loss": 3.0087,
+      "step": 51892
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7219116687774658,
+      "learning_rate": 3.3681723089774714e-05,
+      "loss": 3.0071,
+      "step": 51893
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7408004403114319,
+      "learning_rate": 3.367463550986478e-05,
+      "loss": 2.9956,
+      "step": 51894
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7436817288398743,
+      "learning_rate": 3.3667548631404604e-05,
+      "loss": 2.7571,
+      "step": 51895
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6922502517700195,
+      "learning_rate": 3.366046245441282e-05,
+      "loss": 2.9126,
+      "step": 51896
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7235960364341736,
+      "learning_rate": 3.365337697890813e-05,
+      "loss": 2.8327,
+      "step": 51897
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7371718287467957,
+      "learning_rate": 3.3646292204909164e-05,
+      "loss": 2.8076,
+      "step": 51898
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7488693594932556,
+      "learning_rate": 3.3639208132434635e-05,
+      "loss": 2.727,
+      "step": 51899
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7194048762321472,
+      "learning_rate": 3.363212476150313e-05,
+      "loss": 2.8375,
+      "step": 51900
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7807648777961731,
+      "learning_rate": 3.3625042092133406e-05,
+      "loss": 2.9876,
+      "step": 51901
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.741208016872406,
+      "learning_rate": 3.3617960124344e-05,
+      "loss": 2.77,
+      "step": 51902
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7476096749305725,
+      "learning_rate": 3.361087885815368e-05,
+      "loss": 3.1433,
+      "step": 51903
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8125763535499573,
+      "learning_rate": 3.3603798293581016e-05,
+      "loss": 2.8455,
+      "step": 51904
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7425322532653809,
+      "learning_rate": 3.359671843064464e-05,
+      "loss": 3.098,
+      "step": 51905
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7436474561691284,
+      "learning_rate": 3.358963926936327e-05,
+      "loss": 2.9924,
+      "step": 51906
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.8901980519294739,
+      "learning_rate": 3.358256080975548e-05,
+      "loss": 2.8039,
+      "step": 51907
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7209775447845459,
+      "learning_rate": 3.357548305183996e-05,
+      "loss": 3.2191,
+      "step": 51908
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7874775528907776,
+      "learning_rate": 3.3568405995635393e-05,
+      "loss": 2.9348,
+      "step": 51909
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7514326572418213,
+      "learning_rate": 3.3561329641160376e-05,
+      "loss": 3.0049,
+      "step": 51910
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7351415157318115,
+      "learning_rate": 3.3554253988433513e-05,
+      "loss": 2.8308,
+      "step": 51911
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7183605432510376,
+      "learning_rate": 3.354717903747343e-05,
+      "loss": 2.9062,
+      "step": 51912
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2949225902557373,
+      "learning_rate": 3.354010478829881e-05,
+      "loss": 3.0845,
+      "step": 51913
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7274050116539001,
+      "learning_rate": 3.35330312409283e-05,
+      "loss": 3.066,
+      "step": 51914
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7558318376541138,
+      "learning_rate": 3.352595839538046e-05,
+      "loss": 2.9214,
+      "step": 51915
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7233791351318359,
+      "learning_rate": 3.3518886251674004e-05,
+      "loss": 2.8911,
+      "step": 51916
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7367327809333801,
+      "learning_rate": 3.351181480982755e-05,
+      "loss": 2.8778,
+      "step": 51917
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7649989128112793,
+      "learning_rate": 3.350474406985959e-05,
+      "loss": 2.802,
+      "step": 51918
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7352694869041443,
+      "learning_rate": 3.3497674031788935e-05,
+      "loss": 2.871,
+      "step": 51919
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7183537483215332,
+      "learning_rate": 3.3490604695634073e-05,
+      "loss": 2.9789,
+      "step": 51920
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7267533540725708,
+      "learning_rate": 3.348353606141369e-05,
+      "loss": 2.9397,
+      "step": 51921
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7344212532043457,
+      "learning_rate": 3.347646812914633e-05,
+      "loss": 2.9678,
+      "step": 51922
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7288460731506348,
+      "learning_rate": 3.346940089885065e-05,
+      "loss": 3.0565,
+      "step": 51923
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.77585369348526,
+      "learning_rate": 3.346233437054538e-05,
+      "loss": 2.9148,
+      "step": 51924
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7512287497520447,
+      "learning_rate": 3.345526854424898e-05,
+      "loss": 2.9999,
+      "step": 51925
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7456086874008179,
+      "learning_rate": 3.344820341998012e-05,
+      "loss": 3.0275,
+      "step": 51926
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7627673149108887,
+      "learning_rate": 3.344113899775732e-05,
+      "loss": 2.8022,
+      "step": 51927
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.701185405254364,
+      "learning_rate": 3.343407527759929e-05,
+      "loss": 2.6771,
+      "step": 51928
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7408918738365173,
+      "learning_rate": 3.3427012259524623e-05,
+      "loss": 2.8085,
+      "step": 51929
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7534040212631226,
+      "learning_rate": 3.341994994355186e-05,
+      "loss": 3.0257,
+      "step": 51930
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7601995468139648,
+      "learning_rate": 3.341288832969973e-05,
+      "loss": 2.8248,
+      "step": 51931
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7661672830581665,
+      "learning_rate": 3.340582741798671e-05,
+      "loss": 3.0491,
+      "step": 51932
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7597264051437378,
+      "learning_rate": 3.339876720843139e-05,
+      "loss": 2.9227,
+      "step": 51933
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.726146936416626,
+      "learning_rate": 3.3391707701052484e-05,
+      "loss": 2.941,
+      "step": 51934
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7189964652061462,
+      "learning_rate": 3.338464889586842e-05,
+      "loss": 2.6657,
+      "step": 51935
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7691967487335205,
+      "learning_rate": 3.337759079289797e-05,
+      "loss": 2.7853,
+      "step": 51936
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7658074498176575,
+      "learning_rate": 3.3370533392159557e-05,
+      "loss": 2.9422,
+      "step": 51937
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7648226618766785,
+      "learning_rate": 3.336347669367191e-05,
+      "loss": 2.8923,
+      "step": 51938
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7221632599830627,
+      "learning_rate": 3.3356420697453535e-05,
+      "loss": 2.7206,
+      "step": 51939
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7435133457183838,
+      "learning_rate": 3.3349365403522986e-05,
+      "loss": 2.8739,
+      "step": 51940
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.744724690914154,
+      "learning_rate": 3.3342310811898923e-05,
+      "loss": 2.8973,
+      "step": 51941
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7191993594169617,
+      "learning_rate": 3.3335256922599886e-05,
+      "loss": 2.7973,
+      "step": 51942
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7432810068130493,
+      "learning_rate": 3.33282037356444e-05,
+      "loss": 2.9852,
+      "step": 51943
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7368122339248657,
+      "learning_rate": 3.3321151251051216e-05,
+      "loss": 3.0071,
+      "step": 51944
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7477759122848511,
+      "learning_rate": 3.331409946883872e-05,
+      "loss": 2.988,
+      "step": 51945
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8247474431991577,
+      "learning_rate": 3.330704838902563e-05,
+      "loss": 2.8604,
+      "step": 51946
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7637084126472473,
+      "learning_rate": 3.3299998011630424e-05,
+      "loss": 2.9437,
+      "step": 51947
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7311221361160278,
+      "learning_rate": 3.3292948336671635e-05,
+      "loss": 2.9512,
+      "step": 51948
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7546210885047913,
+      "learning_rate": 3.328589936416797e-05,
+      "loss": 2.8564,
+      "step": 51949
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7457635402679443,
+      "learning_rate": 3.327885109413785e-05,
+      "loss": 2.8564,
+      "step": 51950
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7271265983581543,
+      "learning_rate": 3.327180352659996e-05,
+      "loss": 3.0531,
+      "step": 51951
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7246139049530029,
+      "learning_rate": 3.326475666157275e-05,
+      "loss": 2.6776,
+      "step": 51952
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7810348868370056,
+      "learning_rate": 3.325771049907492e-05,
+      "loss": 2.8616,
+      "step": 51953
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7812083959579468,
+      "learning_rate": 3.32506650391249e-05,
+      "loss": 3.0086,
+      "step": 51954
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7078095078468323,
+      "learning_rate": 3.3243620281741254e-05,
+      "loss": 2.9684,
+      "step": 51955
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7382186651229858,
+      "learning_rate": 3.323657622694264e-05,
+      "loss": 3.0665,
+      "step": 51956
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7355473041534424,
+      "learning_rate": 3.322953287474749e-05,
+      "loss": 2.7563,
+      "step": 51957
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7071644067764282,
+      "learning_rate": 3.32224902251744e-05,
+      "loss": 2.8823,
+      "step": 51958
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7337292432785034,
+      "learning_rate": 3.321544827824196e-05,
+      "loss": 2.8112,
+      "step": 51959
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7419911026954651,
+      "learning_rate": 3.320840703396872e-05,
+      "loss": 2.8555,
+      "step": 51960
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7350503206253052,
+      "learning_rate": 3.3201366492373195e-05,
+      "loss": 3.1303,
+      "step": 51961
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7591217160224915,
+      "learning_rate": 3.3194326653473836e-05,
+      "loss": 2.8945,
+      "step": 51962
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7995786666870117,
+      "learning_rate": 3.318728751728929e-05,
+      "loss": 2.9841,
+      "step": 51963
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7559335231781006,
+      "learning_rate": 3.318024908383816e-05,
+      "loss": 2.9551,
+      "step": 51964
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7337844371795654,
+      "learning_rate": 3.3173211353138816e-05,
+      "loss": 3.106,
+      "step": 51965
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7507258057594299,
+      "learning_rate": 3.316617432520995e-05,
+      "loss": 2.6799,
+      "step": 51966
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7264440059661865,
+      "learning_rate": 3.315913800006994e-05,
+      "loss": 2.8526,
+      "step": 51967
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7675163745880127,
+      "learning_rate": 3.315210237773751e-05,
+      "loss": 3.0471,
+      "step": 51968
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7322779893875122,
+      "learning_rate": 3.314506745823108e-05,
+      "loss": 2.9175,
+      "step": 51969
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7262869477272034,
+      "learning_rate": 3.31380332415691e-05,
+      "loss": 2.886,
+      "step": 51970
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7132663726806641,
+      "learning_rate": 3.313099972777025e-05,
+      "loss": 2.7285,
+      "step": 51971
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7315619587898254,
+      "learning_rate": 3.3123966916852915e-05,
+      "loss": 2.9182,
+      "step": 51972
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7278238534927368,
+      "learning_rate": 3.3116934808835726e-05,
+      "loss": 2.7525,
+      "step": 51973
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7661489248275757,
+      "learning_rate": 3.310990340373719e-05,
+      "loss": 2.9219,
+      "step": 51974
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7610893845558167,
+      "learning_rate": 3.3102872701575836e-05,
+      "loss": 2.8953,
+      "step": 51975
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7103796601295471,
+      "learning_rate": 3.309584270237013e-05,
+      "loss": 2.8619,
+      "step": 51976
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.733978271484375,
+      "learning_rate": 3.3088813406138545e-05,
+      "loss": 2.899,
+      "step": 51977
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7456706762313843,
+      "learning_rate": 3.308178481289976e-05,
+      "loss": 2.9013,
+      "step": 51978
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7410678863525391,
+      "learning_rate": 3.3074756922672085e-05,
+      "loss": 2.8654,
+      "step": 51979
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7431950569152832,
+      "learning_rate": 3.3067729735474115e-05,
+      "loss": 2.8395,
+      "step": 51980
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.715957522392273,
+      "learning_rate": 3.306070325132446e-05,
+      "loss": 2.9237,
+      "step": 51981
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7735357284545898,
+      "learning_rate": 3.3053677470241544e-05,
+      "loss": 2.9148,
+      "step": 51982
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7633320689201355,
+      "learning_rate": 3.304665239224385e-05,
+      "loss": 2.6345,
+      "step": 51983
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.6995530128479004,
+      "learning_rate": 3.303962801734984e-05,
+      "loss": 3.1224,
+      "step": 51984
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7412270903587341,
+      "learning_rate": 3.303260434557808e-05,
+      "loss": 2.9119,
+      "step": 51985
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7831922173500061,
+      "learning_rate": 3.302558137694711e-05,
+      "loss": 2.9452,
+      "step": 51986
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7183090448379517,
+      "learning_rate": 3.301855911147533e-05,
+      "loss": 2.9392,
+      "step": 51987
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7193189263343811,
+      "learning_rate": 3.3011537549181335e-05,
+      "loss": 3.1388,
+      "step": 51988
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7489942312240601,
+      "learning_rate": 3.3004516690083505e-05,
+      "loss": 2.946,
+      "step": 51989
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7693890333175659,
+      "learning_rate": 3.2997496534200464e-05,
+      "loss": 2.8692,
+      "step": 51990
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7755201458930969,
+      "learning_rate": 3.2990477081550624e-05,
+      "loss": 2.934,
+      "step": 51991
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7237342596054077,
+      "learning_rate": 3.2983458332152415e-05,
+      "loss": 2.7404,
+      "step": 51992
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.711622953414917,
+      "learning_rate": 3.2976440286024474e-05,
+      "loss": 3.0484,
+      "step": 51993
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7582530975341797,
+      "learning_rate": 3.2969422943185095e-05,
+      "loss": 2.8522,
+      "step": 51994
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7546455264091492,
+      "learning_rate": 3.29624063036529e-05,
+      "loss": 2.8292,
+      "step": 51995
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7362459301948547,
+      "learning_rate": 3.29553903674464e-05,
+      "loss": 2.8576,
+      "step": 51996
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7272038459777832,
+      "learning_rate": 3.2948375134583984e-05,
+      "loss": 2.9262,
+      "step": 51997
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.787972092628479,
+      "learning_rate": 3.2941360605084166e-05,
+      "loss": 2.8167,
+      "step": 51998
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7820271253585815,
+      "learning_rate": 3.293434677896534e-05,
+      "loss": 2.9891,
+      "step": 51999
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7015359997749329,
+      "learning_rate": 3.292733365624605e-05,
+      "loss": 2.8756,
+      "step": 52000
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7298908233642578,
+      "learning_rate": 3.292032123694481e-05,
+      "loss": 2.7706,
+      "step": 52001
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7448387145996094,
+      "learning_rate": 3.2913309521080026e-05,
+      "loss": 2.8888,
+      "step": 52002
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7021864056587219,
+      "learning_rate": 3.2906298508670195e-05,
+      "loss": 2.9838,
+      "step": 52003
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7519044280052185,
+      "learning_rate": 3.289928819973381e-05,
+      "loss": 2.8275,
+      "step": 52004
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7517359256744385,
+      "learning_rate": 3.289227859428919e-05,
+      "loss": 2.9263,
+      "step": 52005
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7345372438430786,
+      "learning_rate": 3.288526969235499e-05,
+      "loss": 2.9273,
+      "step": 52006
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.702253520488739,
+      "learning_rate": 3.287826149394952e-05,
+      "loss": 3.0666,
+      "step": 52007
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7254359126091003,
+      "learning_rate": 3.2871253999091377e-05,
+      "loss": 2.7956,
+      "step": 52008
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7682890295982361,
+      "learning_rate": 3.2864247207798864e-05,
+      "loss": 2.8114,
+      "step": 52009
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.718375027179718,
+      "learning_rate": 3.2857241120090515e-05,
+      "loss": 2.8538,
+      "step": 52010
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7394582629203796,
+      "learning_rate": 3.285023573598484e-05,
+      "loss": 2.9124,
+      "step": 52011
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.723473310470581,
+      "learning_rate": 3.284323105550023e-05,
+      "loss": 3.0113,
+      "step": 52012
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.6897356510162354,
+      "learning_rate": 3.283622707865513e-05,
+      "loss": 2.8036,
+      "step": 52013
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7808870077133179,
+      "learning_rate": 3.282922380546793e-05,
+      "loss": 2.8074,
+      "step": 52014
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7283383011817932,
+      "learning_rate": 3.2822221235957145e-05,
+      "loss": 2.8458,
+      "step": 52015
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7218790650367737,
+      "learning_rate": 3.2815219370141264e-05,
+      "loss": 3.0506,
+      "step": 52016
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7908827066421509,
+      "learning_rate": 3.280821820803861e-05,
+      "loss": 3.0478,
+      "step": 52017
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7488206028938293,
+      "learning_rate": 3.280121774966775e-05,
+      "loss": 2.793,
+      "step": 52018
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7366180419921875,
+      "learning_rate": 3.2794217995047054e-05,
+      "loss": 3.0707,
+      "step": 52019
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7356551885604858,
+      "learning_rate": 3.278721894419494e-05,
+      "loss": 2.9331,
+      "step": 52020
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7605820298194885,
+      "learning_rate": 3.278022059712988e-05,
+      "loss": 3.1874,
+      "step": 52021
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7392238974571228,
+      "learning_rate": 3.2773222953870236e-05,
+      "loss": 2.8032,
+      "step": 52022
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7410476803779602,
+      "learning_rate": 3.2766226014434596e-05,
+      "loss": 2.8852,
+      "step": 52023
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7377091646194458,
+      "learning_rate": 3.2759229778841166e-05,
+      "loss": 2.7027,
+      "step": 52024
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7168262004852295,
+      "learning_rate": 3.275223424710854e-05,
+      "loss": 2.813,
+      "step": 52025
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7007832527160645,
+      "learning_rate": 3.2745239419255145e-05,
+      "loss": 2.9249,
+      "step": 52026
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7348765134811401,
+      "learning_rate": 3.273824529529935e-05,
+      "loss": 2.8679,
+      "step": 52027
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7212089896202087,
+      "learning_rate": 3.273125187525958e-05,
+      "loss": 2.7796,
+      "step": 52028
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7870141863822937,
+      "learning_rate": 3.2724259159154186e-05,
+      "loss": 2.965,
+      "step": 52029
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.728225588798523,
+      "learning_rate": 3.2717267147001656e-05,
+      "loss": 2.9082,
+      "step": 52030
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7556633353233337,
+      "learning_rate": 3.271027583882048e-05,
+      "loss": 2.9904,
+      "step": 52031
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7504242658615112,
+      "learning_rate": 3.270328523462893e-05,
+      "loss": 2.9538,
+      "step": 52032
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7282172441482544,
+      "learning_rate": 3.2696295334445566e-05,
+      "loss": 2.9926,
+      "step": 52033
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7356157302856445,
+      "learning_rate": 3.2689306138288675e-05,
+      "loss": 2.938,
+      "step": 52034
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.6893876194953918,
+      "learning_rate": 3.268231764617667e-05,
+      "loss": 2.9002,
+      "step": 52035
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7842584848403931,
+      "learning_rate": 3.267532985812808e-05,
+      "loss": 2.8638,
+      "step": 52036
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7505773901939392,
+      "learning_rate": 3.2668342774161105e-05,
+      "loss": 3.0113,
+      "step": 52037
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7909178733825684,
+      "learning_rate": 3.266135639429438e-05,
+      "loss": 2.8813,
+      "step": 52038
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7430846095085144,
+      "learning_rate": 3.265437071854609e-05,
+      "loss": 3.109,
+      "step": 52039
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7298345565795898,
+      "learning_rate": 3.264738574693481e-05,
+      "loss": 3.0871,
+      "step": 52040
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7301103472709656,
+      "learning_rate": 3.264040147947889e-05,
+      "loss": 2.9725,
+      "step": 52041
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7341718077659607,
+      "learning_rate": 3.26334179161966e-05,
+      "loss": 2.6133,
+      "step": 52042
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7384917140007019,
+      "learning_rate": 3.262643505710651e-05,
+      "loss": 2.7426,
+      "step": 52043
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.754947304725647,
+      "learning_rate": 3.2619452902226885e-05,
+      "loss": 2.7363,
+      "step": 52044
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7241181135177612,
+      "learning_rate": 3.261247145157613e-05,
+      "loss": 2.871,
+      "step": 52045
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7246091365814209,
+      "learning_rate": 3.2605490705172734e-05,
+      "loss": 2.969,
+      "step": 52046
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7573584914207458,
+      "learning_rate": 3.259851066303498e-05,
+      "loss": 2.9414,
+      "step": 52047
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7820131182670593,
+      "learning_rate": 3.259153132518133e-05,
+      "loss": 3.2345,
+      "step": 52048
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7728943824768066,
+      "learning_rate": 3.2584552691630125e-05,
+      "loss": 2.8005,
+      "step": 52049
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7590402960777283,
+      "learning_rate": 3.257757476239966e-05,
+      "loss": 2.6722,
+      "step": 52050
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7413563132286072,
+      "learning_rate": 3.257059753750848e-05,
+      "loss": 2.722,
+      "step": 52051
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7250611782073975,
+      "learning_rate": 3.256362101697484e-05,
+      "loss": 2.6977,
+      "step": 52052
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7582236528396606,
+      "learning_rate": 3.255664520081719e-05,
+      "loss": 3.023,
+      "step": 52053
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7816905975341797,
+      "learning_rate": 3.254967008905378e-05,
+      "loss": 2.8675,
+      "step": 52054
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7680097818374634,
+      "learning_rate": 3.254269568170317e-05,
+      "loss": 2.8494,
+      "step": 52055
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7169177532196045,
+      "learning_rate": 3.25357219787836e-05,
+      "loss": 2.951,
+      "step": 52056
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7729882597923279,
+      "learning_rate": 3.25287489803134e-05,
+      "loss": 2.806,
+      "step": 52057
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7043116688728333,
+      "learning_rate": 3.252177668631109e-05,
+      "loss": 3.1595,
+      "step": 52058
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8605642318725586,
+      "learning_rate": 3.2514805096794835e-05,
+      "loss": 2.8124,
+      "step": 52059
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7150072455406189,
+      "learning_rate": 3.2507834211783214e-05,
+      "loss": 2.7993,
+      "step": 52060
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7066315412521362,
+      "learning_rate": 3.250086403129439e-05,
+      "loss": 2.9328,
+      "step": 52061
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7706637978553772,
+      "learning_rate": 3.2493894555346866e-05,
+      "loss": 3.0378,
+      "step": 52062
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7277361750602722,
+      "learning_rate": 3.248692578395898e-05,
+      "loss": 2.8421,
+      "step": 52063
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7525270581245422,
+      "learning_rate": 3.2479957717148966e-05,
+      "loss": 2.9334,
+      "step": 52064
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7777685523033142,
+      "learning_rate": 3.247299035493529e-05,
+      "loss": 2.6474,
+      "step": 52065
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7585991621017456,
+      "learning_rate": 3.246602369733622e-05,
+      "loss": 2.9298,
+      "step": 52066
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7407344579696655,
+      "learning_rate": 3.24590577443702e-05,
+      "loss": 2.8826,
+      "step": 52067
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7381274700164795,
+      "learning_rate": 3.245209249605555e-05,
+      "loss": 2.8998,
+      "step": 52068
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7720596790313721,
+      "learning_rate": 3.2445127952410554e-05,
+      "loss": 3.0174,
+      "step": 52069
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7466363310813904,
+      "learning_rate": 3.243816411345364e-05,
+      "loss": 2.9214,
+      "step": 52070
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7581253051757812,
+      "learning_rate": 3.243120097920311e-05,
+      "loss": 2.9456,
+      "step": 52071
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7453657388687134,
+      "learning_rate": 3.2424238549677265e-05,
+      "loss": 2.8891,
+      "step": 52072
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7070062756538391,
+      "learning_rate": 3.2417276824894514e-05,
+      "loss": 3.0115,
+      "step": 52073
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7681336402893066,
+      "learning_rate": 3.2410315804873114e-05,
+      "loss": 2.6473,
+      "step": 52074
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7477194666862488,
+      "learning_rate": 3.240335548963151e-05,
+      "loss": 2.9022,
+      "step": 52075
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7516164183616638,
+      "learning_rate": 3.23963958791879e-05,
+      "loss": 2.9424,
+      "step": 52076
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7572428584098816,
+      "learning_rate": 3.2389436973560754e-05,
+      "loss": 2.8975,
+      "step": 52077
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7532944679260254,
+      "learning_rate": 3.238247877276831e-05,
+      "loss": 2.9374,
+      "step": 52078
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7349500060081482,
+      "learning_rate": 3.2375521276828866e-05,
+      "loss": 2.7757,
+      "step": 52079
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7738845944404602,
+      "learning_rate": 3.236856448576085e-05,
+      "loss": 3.076,
+      "step": 52080
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7966225743293762,
+      "learning_rate": 3.2361608399582454e-05,
+      "loss": 2.8035,
+      "step": 52081
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.6876529455184937,
+      "learning_rate": 3.235465301831208e-05,
+      "loss": 2.9115,
+      "step": 52082
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7187408208847046,
+      "learning_rate": 3.2347698341968095e-05,
+      "loss": 2.8788,
+      "step": 52083
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7126637697219849,
+      "learning_rate": 3.2340744370568783e-05,
+      "loss": 3.1122,
+      "step": 52084
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7737832069396973,
+      "learning_rate": 3.233379110413242e-05,
+      "loss": 2.8319,
+      "step": 52085
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7817880511283875,
+      "learning_rate": 3.23268385426773e-05,
+      "loss": 2.8349,
+      "step": 52086
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7609059810638428,
+      "learning_rate": 3.231988668622174e-05,
+      "loss": 3.1345,
+      "step": 52087
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7383360266685486,
+      "learning_rate": 3.2312935534784154e-05,
+      "loss": 2.9601,
+      "step": 52088
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7559726238250732,
+      "learning_rate": 3.2305985088382736e-05,
+      "loss": 2.8189,
+      "step": 52089
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7402336001396179,
+      "learning_rate": 3.2299035347035896e-05,
+      "loss": 2.9232,
+      "step": 52090
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8144344687461853,
+      "learning_rate": 3.229208631076179e-05,
+      "loss": 2.8724,
+      "step": 52091
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7357531785964966,
+      "learning_rate": 3.22851379795789e-05,
+      "loss": 2.8621,
+      "step": 52092
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7645347714424133,
+      "learning_rate": 3.227819035350542e-05,
+      "loss": 2.7017,
+      "step": 52093
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7534157037734985,
+      "learning_rate": 3.227124343255959e-05,
+      "loss": 2.8386,
+      "step": 52094
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7897367477416992,
+      "learning_rate": 3.226429721675984e-05,
+      "loss": 2.8818,
+      "step": 52095
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7672426700592041,
+      "learning_rate": 3.225735170612435e-05,
+      "loss": 2.8568,
+      "step": 52096
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7439735531806946,
+      "learning_rate": 3.2250406900671466e-05,
+      "loss": 2.8941,
+      "step": 52097
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7530331611633301,
+      "learning_rate": 3.224346280041955e-05,
+      "loss": 2.8021,
+      "step": 52098
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7255603671073914,
+      "learning_rate": 3.223651940538682e-05,
+      "loss": 3.0742,
+      "step": 52099
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7502477765083313,
+      "learning_rate": 3.2229576715591555e-05,
+      "loss": 2.8602,
+      "step": 52100
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8245431184768677,
+      "learning_rate": 3.222263473105198e-05,
+      "loss": 2.7466,
+      "step": 52101
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7484453320503235,
+      "learning_rate": 3.2215693451786474e-05,
+      "loss": 2.9278,
+      "step": 52102
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.720234215259552,
+      "learning_rate": 3.2208752877813357e-05,
+      "loss": 3.0219,
+      "step": 52103
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7335119843482971,
+      "learning_rate": 3.220181300915079e-05,
+      "loss": 2.795,
+      "step": 52104
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7557951807975769,
+      "learning_rate": 3.219487384581715e-05,
+      "loss": 3.0479,
+      "step": 52105
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7810825705528259,
+      "learning_rate": 3.218793538783069e-05,
+      "loss": 2.9214,
+      "step": 52106
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7902493476867676,
+      "learning_rate": 3.218099763520959e-05,
+      "loss": 2.8729,
+      "step": 52107
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8319610953330994,
+      "learning_rate": 3.217406058797225e-05,
+      "loss": 2.9528,
+      "step": 52108
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7634714841842651,
+      "learning_rate": 3.2167124246136876e-05,
+      "loss": 2.9904,
+      "step": 52109
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7566140294075012,
+      "learning_rate": 3.216018860972176e-05,
+      "loss": 2.822,
+      "step": 52110
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8258447051048279,
+      "learning_rate": 3.215325367874512e-05,
+      "loss": 2.7601,
+      "step": 52111
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7029900550842285,
+      "learning_rate": 3.2146319453225266e-05,
+      "loss": 3.0402,
+      "step": 52112
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7495381832122803,
+      "learning_rate": 3.2139385933180526e-05,
+      "loss": 2.8941,
+      "step": 52113
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7451898455619812,
+      "learning_rate": 3.213245311862909e-05,
+      "loss": 2.9559,
+      "step": 52114
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7116692662239075,
+      "learning_rate": 3.212552100958919e-05,
+      "loss": 3.0828,
+      "step": 52115
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7679808735847473,
+      "learning_rate": 3.21185896060791e-05,
+      "loss": 2.9068,
+      "step": 52116
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7367801070213318,
+      "learning_rate": 3.2111658908117065e-05,
+      "loss": 2.9342,
+      "step": 52117
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7176463007926941,
+      "learning_rate": 3.210472891572143e-05,
+      "loss": 3.1503,
+      "step": 52118
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.6963474750518799,
+      "learning_rate": 3.2097799628910295e-05,
+      "loss": 2.9381,
+      "step": 52119
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7654833197593689,
+      "learning_rate": 3.209087104770209e-05,
+      "loss": 2.8578,
+      "step": 52120
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7469300627708435,
+      "learning_rate": 3.208394317211496e-05,
+      "loss": 3.1074,
+      "step": 52121
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.702953577041626,
+      "learning_rate": 3.20770160021671e-05,
+      "loss": 2.6675,
+      "step": 52122
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7122648358345032,
+      "learning_rate": 3.207008953787688e-05,
+      "loss": 2.9066,
+      "step": 52123
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7659664750099182,
+      "learning_rate": 3.206316377926244e-05,
+      "loss": 2.8832,
+      "step": 52124
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.747031033039093,
+      "learning_rate": 3.20562387263421e-05,
+      "loss": 2.965,
+      "step": 52125
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7918295860290527,
+      "learning_rate": 3.204931437913398e-05,
+      "loss": 2.8316,
+      "step": 52126
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7406182289123535,
+      "learning_rate": 3.204239073765641e-05,
+      "loss": 2.9045,
+      "step": 52127
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7530829906463623,
+      "learning_rate": 3.2035467801927696e-05,
+      "loss": 2.9757,
+      "step": 52128
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7679747343063354,
+      "learning_rate": 3.2028545571965994e-05,
+      "loss": 2.9212,
+      "step": 52129
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7454062104225159,
+      "learning_rate": 3.202162404778949e-05,
+      "loss": 2.9671,
+      "step": 52130
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7858791947364807,
+      "learning_rate": 3.2014703229416426e-05,
+      "loss": 2.8775,
+      "step": 52131
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7846336364746094,
+      "learning_rate": 3.2007783116865104e-05,
+      "loss": 2.9201,
+      "step": 52132
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.7320681810379028,
+      "learning_rate": 3.2000863710153635e-05,
+      "loss": 3.0948,
+      "step": 52133
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8238586783409119,
+      "learning_rate": 3.1993945009300336e-05,
+      "loss": 2.9597,
+      "step": 52134
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7886196970939636,
+      "learning_rate": 3.198702701432346e-05,
+      "loss": 2.9869,
+      "step": 52135
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7563168406486511,
+      "learning_rate": 3.19801097252412e-05,
+      "loss": 2.9357,
+      "step": 52136
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8196958899497986,
+      "learning_rate": 3.197319314207169e-05,
+      "loss": 2.9284,
+      "step": 52137
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7236234545707703,
+      "learning_rate": 3.196627726483321e-05,
+      "loss": 2.8384,
+      "step": 52138
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.773149847984314,
+      "learning_rate": 3.195936209354395e-05,
+      "loss": 2.8404,
+      "step": 52139
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7744463682174683,
+      "learning_rate": 3.1952447628222176e-05,
+      "loss": 2.7944,
+      "step": 52140
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7418228387832642,
+      "learning_rate": 3.194553386888604e-05,
+      "loss": 2.9338,
+      "step": 52141
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7621162533760071,
+      "learning_rate": 3.193862081555383e-05,
+      "loss": 2.9538,
+      "step": 52142
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7875955104827881,
+      "learning_rate": 3.193170846824369e-05,
+      "loss": 2.9326,
+      "step": 52143
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7359634637832642,
+      "learning_rate": 3.192479682697381e-05,
+      "loss": 2.7486,
+      "step": 52144
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7574135661125183,
+      "learning_rate": 3.191788589176244e-05,
+      "loss": 2.8255,
+      "step": 52145
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7023094296455383,
+      "learning_rate": 3.1910975662627734e-05,
+      "loss": 2.9397,
+      "step": 52146
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7390046119689941,
+      "learning_rate": 3.190406613958798e-05,
+      "loss": 3.0639,
+      "step": 52147
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7670930027961731,
+      "learning_rate": 3.189715732266126e-05,
+      "loss": 2.6807,
+      "step": 52148
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.711762011051178,
+      "learning_rate": 3.1890249211865826e-05,
+      "loss": 2.7934,
+      "step": 52149
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7005589008331299,
+      "learning_rate": 3.1883341807219965e-05,
+      "loss": 2.9111,
+      "step": 52150
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7569049596786499,
+      "learning_rate": 3.1876435108741725e-05,
+      "loss": 2.8733,
+      "step": 52151
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8212657570838928,
+      "learning_rate": 3.18695291164494e-05,
+      "loss": 2.9394,
+      "step": 52152
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7402049899101257,
+      "learning_rate": 3.186262383036107e-05,
+      "loss": 2.8861,
+      "step": 52153
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7410562038421631,
+      "learning_rate": 3.185571925049495e-05,
+      "loss": 3.0182,
+      "step": 52154
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7571350336074829,
+      "learning_rate": 3.1848815376869366e-05,
+      "loss": 2.7717,
+      "step": 52155
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7215419411659241,
+      "learning_rate": 3.1841912209502294e-05,
+      "loss": 2.86,
+      "step": 52156
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.751708984375,
+      "learning_rate": 3.18350097484121e-05,
+      "loss": 3.1578,
+      "step": 52157
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7202328443527222,
+      "learning_rate": 3.182810799361689e-05,
+      "loss": 2.8002,
+      "step": 52158
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7052213549613953,
+      "learning_rate": 3.182120694513476e-05,
+      "loss": 3.1323,
+      "step": 52159
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7351638674736023,
+      "learning_rate": 3.181430660298402e-05,
+      "loss": 2.8612,
+      "step": 52160
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7632426619529724,
+      "learning_rate": 3.180740696718269e-05,
+      "loss": 3.0659,
+      "step": 52161
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7189815640449524,
+      "learning_rate": 3.180050803774913e-05,
+      "loss": 3.0106,
+      "step": 52162
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7511869668960571,
+      "learning_rate": 3.179360981470138e-05,
+      "loss": 3.0049,
+      "step": 52163
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7162873148918152,
+      "learning_rate": 3.178671229805767e-05,
+      "loss": 2.9714,
+      "step": 52164
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8090042471885681,
+      "learning_rate": 3.177981548783615e-05,
+      "loss": 2.873,
+      "step": 52165
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7131547927856445,
+      "learning_rate": 3.1772919384054897e-05,
+      "loss": 2.6244,
+      "step": 52166
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7170288562774658,
+      "learning_rate": 3.176602398673224e-05,
+      "loss": 3.1226,
+      "step": 52167
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7279961705207825,
+      "learning_rate": 3.1759129295886184e-05,
+      "loss": 2.9755,
+      "step": 52168
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7459250688552856,
+      "learning_rate": 3.175223531153497e-05,
+      "loss": 2.7423,
+      "step": 52169
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.806098461151123,
+      "learning_rate": 3.1745342033696764e-05,
+      "loss": 2.8613,
+      "step": 52170
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7358162999153137,
+      "learning_rate": 3.1738449462389656e-05,
+      "loss": 2.8704,
+      "step": 52171
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7452125549316406,
+      "learning_rate": 3.173155759763194e-05,
+      "loss": 2.9032,
+      "step": 52172
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.765661358833313,
+      "learning_rate": 3.1724666439441626e-05,
+      "loss": 2.9674,
+      "step": 52173
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7364770770072937,
+      "learning_rate": 3.171777598783687e-05,
+      "loss": 2.7907,
+      "step": 52174
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.740664005279541,
+      "learning_rate": 3.171088624283592e-05,
+      "loss": 2.8336,
+      "step": 52175
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7158206701278687,
+      "learning_rate": 3.17039972044568e-05,
+      "loss": 2.734,
+      "step": 52176
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7403131127357483,
+      "learning_rate": 3.169710887271779e-05,
+      "loss": 2.9151,
+      "step": 52177
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7874448299407959,
+      "learning_rate": 3.169022124763688e-05,
+      "loss": 2.9587,
+      "step": 52178
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7513279318809509,
+      "learning_rate": 3.168333432923236e-05,
+      "loss": 2.9218,
+      "step": 52179
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.715566873550415,
+      "learning_rate": 3.16764481175223e-05,
+      "loss": 2.8593,
+      "step": 52180
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7262329459190369,
+      "learning_rate": 3.166956261252477e-05,
+      "loss": 2.9149,
+      "step": 52181
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7445135712623596,
+      "learning_rate": 3.166267781425804e-05,
+      "loss": 2.9653,
+      "step": 52182
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.731885552406311,
+      "learning_rate": 3.165579372274011e-05,
+      "loss": 2.8945,
+      "step": 52183
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7854867577552795,
+      "learning_rate": 3.164891033798919e-05,
+      "loss": 2.7433,
+      "step": 52184
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7866995334625244,
+      "learning_rate": 3.164202766002343e-05,
+      "loss": 2.8823,
+      "step": 52185
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7712637782096863,
+      "learning_rate": 3.1635145688860916e-05,
+      "loss": 2.944,
+      "step": 52186
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.758851170539856,
+      "learning_rate": 3.162826442451981e-05,
+      "loss": 2.9999,
+      "step": 52187
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7671577334403992,
+      "learning_rate": 3.162138386701812e-05,
+      "loss": 2.9154,
+      "step": 52188
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7265655398368835,
+      "learning_rate": 3.161450401637407e-05,
+      "loss": 2.9974,
+      "step": 52189
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7680802941322327,
+      "learning_rate": 3.160762487260582e-05,
+      "loss": 2.9132,
+      "step": 52190
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7444314360618591,
+      "learning_rate": 3.160074643573137e-05,
+      "loss": 2.8905,
+      "step": 52191
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7568455934524536,
+      "learning_rate": 3.1593868705768955e-05,
+      "loss": 2.7307,
+      "step": 52192
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7822867631912231,
+      "learning_rate": 3.1586991682736585e-05,
+      "loss": 2.93,
+      "step": 52193
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7543140053749084,
+      "learning_rate": 3.158011536665248e-05,
+      "loss": 3.1412,
+      "step": 52194
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.853643000125885,
+      "learning_rate": 3.15732397575347e-05,
+      "loss": 2.9038,
+      "step": 52195
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8402360677719116,
+      "learning_rate": 3.1566364855401285e-05,
+      "loss": 2.9202,
+      "step": 52196
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7252039313316345,
+      "learning_rate": 3.155949066027046e-05,
+      "loss": 2.977,
+      "step": 52197
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7317278385162354,
+      "learning_rate": 3.155261717216021e-05,
+      "loss": 2.7395,
+      "step": 52198
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7297314405441284,
+      "learning_rate": 3.154574439108869e-05,
+      "loss": 2.8092,
+      "step": 52199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7250798940658569,
+      "learning_rate": 3.153887231707408e-05,
+      "loss": 3.1066,
+      "step": 52200
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7466351389884949,
+      "learning_rate": 3.153200095013444e-05,
+      "loss": 3.1056,
+      "step": 52201
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7196610569953918,
+      "learning_rate": 3.1525130290287816e-05,
+      "loss": 2.9067,
+      "step": 52202
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7877478003501892,
+      "learning_rate": 3.151826033755229e-05,
+      "loss": 2.6556,
+      "step": 52203
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7279103398323059,
+      "learning_rate": 3.151139109194596e-05,
+      "loss": 2.9329,
+      "step": 52204
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7403162717819214,
+      "learning_rate": 3.1504522553487035e-05,
+      "loss": 2.9713,
+      "step": 52205
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.748703122138977,
+      "learning_rate": 3.1497654722193466e-05,
+      "loss": 2.7022,
+      "step": 52206
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7111794948577881,
+      "learning_rate": 3.149078759808348e-05,
+      "loss": 3.1579,
+      "step": 52207
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7396597266197205,
+      "learning_rate": 3.148392118117505e-05,
+      "loss": 2.6688,
+      "step": 52208
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7661385536193848,
+      "learning_rate": 3.1477055471486315e-05,
+      "loss": 3.2509,
+      "step": 52209
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7724836468696594,
+      "learning_rate": 3.1470190469035274e-05,
+      "loss": 3.0427,
+      "step": 52210
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7351310849189758,
+      "learning_rate": 3.146332617384009e-05,
+      "loss": 2.949,
+      "step": 52211
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7441316246986389,
+      "learning_rate": 3.145646258591884e-05,
+      "loss": 2.9306,
+      "step": 52212
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7312101125717163,
+      "learning_rate": 3.1449599705289554e-05,
+      "loss": 2.9497,
+      "step": 52213
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.733696699142456,
+      "learning_rate": 3.144273753197041e-05,
+      "loss": 2.8247,
+      "step": 52214
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7087571620941162,
+      "learning_rate": 3.143587606597933e-05,
+      "loss": 2.8334,
+      "step": 52215
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7312347292900085,
+      "learning_rate": 3.142901530733453e-05,
+      "loss": 2.869,
+      "step": 52216
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6981326341629028,
+      "learning_rate": 3.1422155256054024e-05,
+      "loss": 2.8639,
+      "step": 52217
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7454235553741455,
+      "learning_rate": 3.141529591215578e-05,
+      "loss": 3.1351,
+      "step": 52218
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7256028652191162,
+      "learning_rate": 3.1408437275658074e-05,
+      "loss": 2.776,
+      "step": 52219
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7506396174430847,
+      "learning_rate": 3.1401579346578753e-05,
+      "loss": 2.7559,
+      "step": 52220
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7295669913291931,
+      "learning_rate": 3.139472212493598e-05,
+      "loss": 2.8567,
+      "step": 52221
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.731708288192749,
+      "learning_rate": 3.1387865610747864e-05,
+      "loss": 2.7866,
+      "step": 52222
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.718250036239624,
+      "learning_rate": 3.138100980403243e-05,
+      "loss": 2.9393,
+      "step": 52223
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7222898006439209,
+      "learning_rate": 3.137415470480773e-05,
+      "loss": 2.9184,
+      "step": 52224
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7631383538246155,
+      "learning_rate": 3.136730031309175e-05,
+      "loss": 2.9492,
+      "step": 52225
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7158065438270569,
+      "learning_rate": 3.13604466289026e-05,
+      "loss": 3.0133,
+      "step": 52226
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7890329360961914,
+      "learning_rate": 3.135359365225838e-05,
+      "loss": 3.0817,
+      "step": 52227
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7782434225082397,
+      "learning_rate": 3.134674138317703e-05,
+      "loss": 3.0138,
+      "step": 52228
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7466015815734863,
+      "learning_rate": 3.133988982167671e-05,
+      "loss": 2.823,
+      "step": 52229
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.712049663066864,
+      "learning_rate": 3.1333038967775404e-05,
+      "loss": 2.9232,
+      "step": 52230
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7094315886497498,
+      "learning_rate": 3.1326188821491196e-05,
+      "loss": 2.7197,
+      "step": 52231
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7306052446365356,
+      "learning_rate": 3.131933938284207e-05,
+      "loss": 2.9825,
+      "step": 52232
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6810746788978577,
+      "learning_rate": 3.131249065184608e-05,
+      "loss": 2.7697,
+      "step": 52233
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.748337984085083,
+      "learning_rate": 3.130564262852131e-05,
+      "loss": 2.9846,
+      "step": 52234
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7128395438194275,
+      "learning_rate": 3.1298795312885716e-05,
+      "loss": 3.0067,
+      "step": 52235
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7342303395271301,
+      "learning_rate": 3.1291948704957416e-05,
+      "loss": 3.1391,
+      "step": 52236
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7613341808319092,
+      "learning_rate": 3.1285102804754424e-05,
+      "loss": 2.9238,
+      "step": 52237
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8025845289230347,
+      "learning_rate": 3.1278257612294765e-05,
+      "loss": 3.0679,
+      "step": 52238
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7101073265075684,
+      "learning_rate": 3.127141312759649e-05,
+      "loss": 2.7353,
+      "step": 52239
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7481630444526672,
+      "learning_rate": 3.126456935067748e-05,
+      "loss": 2.8775,
+      "step": 52240
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.747065544128418,
+      "learning_rate": 3.125772628155593e-05,
+      "loss": 2.9329,
+      "step": 52241
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7435370683670044,
+      "learning_rate": 3.125088392024985e-05,
+      "loss": 2.9153,
+      "step": 52242
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.781072199344635,
+      "learning_rate": 3.124404226677712e-05,
+      "loss": 2.9379,
+      "step": 52243
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7462438344955444,
+      "learning_rate": 3.123720132115598e-05,
+      "loss": 2.897,
+      "step": 52244
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7858874797821045,
+      "learning_rate": 3.12303610834043e-05,
+      "loss": 2.8322,
+      "step": 52245
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7304945588111877,
+      "learning_rate": 3.1223521553540064e-05,
+      "loss": 2.8908,
+      "step": 52246
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7903547286987305,
+      "learning_rate": 3.12166827315814e-05,
+      "loss": 2.8174,
+      "step": 52247
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7469950914382935,
+      "learning_rate": 3.120984461754622e-05,
+      "loss": 2.9731,
+      "step": 52248
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7844982743263245,
+      "learning_rate": 3.120300721145261e-05,
+      "loss": 2.8559,
+      "step": 52249
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7452136278152466,
+      "learning_rate": 3.119617051331853e-05,
+      "loss": 3.1268,
+      "step": 52250
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7502493262290955,
+      "learning_rate": 3.1189334523161954e-05,
+      "loss": 2.8521,
+      "step": 52251
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6964902877807617,
+      "learning_rate": 3.1182499241001045e-05,
+      "loss": 2.8879,
+      "step": 52252
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7556952834129333,
+      "learning_rate": 3.1175664666853686e-05,
+      "loss": 2.8724,
+      "step": 52253
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6863505244255066,
+      "learning_rate": 3.116883080073789e-05,
+      "loss": 2.9686,
+      "step": 52254
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7049617767333984,
+      "learning_rate": 3.1161997642671574e-05,
+      "loss": 2.9675,
+      "step": 52255
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7351288199424744,
+      "learning_rate": 3.115516519267284e-05,
+      "loss": 2.7977,
+      "step": 52256
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7706930637359619,
+      "learning_rate": 3.114833345075972e-05,
+      "loss": 2.9956,
+      "step": 52257
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.771697998046875,
+      "learning_rate": 3.114150241695008e-05,
+      "loss": 2.7364,
+      "step": 52258
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7425963878631592,
+      "learning_rate": 3.1134672091262035e-05,
+      "loss": 2.963,
+      "step": 52259
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7799915075302124,
+      "learning_rate": 3.11278424737135e-05,
+      "loss": 2.9248,
+      "step": 52260
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7096208930015564,
+      "learning_rate": 3.112101356432247e-05,
+      "loss": 2.7819,
+      "step": 52261
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.747329831123352,
+      "learning_rate": 3.111418536310699e-05,
+      "loss": 2.8839,
+      "step": 52262
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7840654253959656,
+      "learning_rate": 3.110735787008492e-05,
+      "loss": 2.8895,
+      "step": 52263
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7264786958694458,
+      "learning_rate": 3.1100531085274394e-05,
+      "loss": 2.8788,
+      "step": 52264
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7728217840194702,
+      "learning_rate": 3.1093705008693245e-05,
+      "loss": 2.9602,
+      "step": 52265
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7420971393585205,
+      "learning_rate": 3.1086879640359596e-05,
+      "loss": 2.8934,
+      "step": 52266
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.714773952960968,
+      "learning_rate": 3.108005498029136e-05,
+      "loss": 2.8666,
+      "step": 52267
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7264063954353333,
+      "learning_rate": 3.107323102850642e-05,
+      "loss": 2.8107,
+      "step": 52268
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7227268815040588,
+      "learning_rate": 3.1066407785022895e-05,
+      "loss": 2.8292,
+      "step": 52269
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7567482590675354,
+      "learning_rate": 3.1059585249858675e-05,
+      "loss": 2.8256,
+      "step": 52270
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7323058247566223,
+      "learning_rate": 3.1052763423031714e-05,
+      "loss": 2.9365,
+      "step": 52271
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7168264389038086,
+      "learning_rate": 3.104594230456009e-05,
+      "loss": 3.0581,
+      "step": 52272
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7276723980903625,
+      "learning_rate": 3.1039121894461625e-05,
+      "loss": 2.7855,
+      "step": 52273
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7067778706550598,
+      "learning_rate": 3.103230219275441e-05,
+      "loss": 3.0388,
+      "step": 52274
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7861632704734802,
+      "learning_rate": 3.1025483199456327e-05,
+      "loss": 2.954,
+      "step": 52275
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6696493625640869,
+      "learning_rate": 3.1018664914585326e-05,
+      "loss": 2.9169,
+      "step": 52276
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7829317450523376,
+      "learning_rate": 3.1011847338159454e-05,
+      "loss": 2.9342,
+      "step": 52277
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7882497310638428,
+      "learning_rate": 3.1005030470196514e-05,
+      "loss": 2.8775,
+      "step": 52278
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7552138566970825,
+      "learning_rate": 3.099821431071464e-05,
+      "loss": 2.9103,
+      "step": 52279
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7698383927345276,
+      "learning_rate": 3.099139885973166e-05,
+      "loss": 2.8783,
+      "step": 52280
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.782609760761261,
+      "learning_rate": 3.09845841172656e-05,
+      "loss": 2.975,
+      "step": 52281
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7387914061546326,
+      "learning_rate": 3.0977770083334366e-05,
+      "loss": 2.8243,
+      "step": 52282
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7554448843002319,
+      "learning_rate": 3.097095675795588e-05,
+      "loss": 2.9946,
+      "step": 52283
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7315139770507812,
+      "learning_rate": 3.096414414114813e-05,
+      "loss": 2.9418,
+      "step": 52284
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7560511827468872,
+      "learning_rate": 3.0957332232929036e-05,
+      "loss": 3.0498,
+      "step": 52285
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7678272724151611,
+      "learning_rate": 3.095052103331651e-05,
+      "loss": 2.7948,
+      "step": 52286
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7785819172859192,
+      "learning_rate": 3.0943710542328613e-05,
+      "loss": 2.7721,
+      "step": 52287
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7176530361175537,
+      "learning_rate": 3.093690075998323e-05,
+      "loss": 2.9313,
+      "step": 52288
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7264245748519897,
+      "learning_rate": 3.0930091686298206e-05,
+      "loss": 2.7374,
+      "step": 52289
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7617298364639282,
+      "learning_rate": 3.092328332129154e-05,
+      "loss": 2.8061,
+      "step": 52290
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7552651166915894,
+      "learning_rate": 3.0916475664981165e-05,
+      "loss": 2.9829,
+      "step": 52291
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7785048484802246,
+      "learning_rate": 3.0909668717384986e-05,
+      "loss": 2.8478,
+      "step": 52292
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.719987154006958,
+      "learning_rate": 3.090286247852094e-05,
+      "loss": 2.8228,
+      "step": 52293
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.735836386680603,
+      "learning_rate": 3.089605694840702e-05,
+      "loss": 2.9333,
+      "step": 52294
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.763145923614502,
+      "learning_rate": 3.0889252127061045e-05,
+      "loss": 3.0371,
+      "step": 52295
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7010157108306885,
+      "learning_rate": 3.088244801450104e-05,
+      "loss": 3.0665,
+      "step": 52296
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7367874979972839,
+      "learning_rate": 3.0875644610744873e-05,
+      "loss": 2.93,
+      "step": 52297
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7362725138664246,
+      "learning_rate": 3.0868841915810414e-05,
+      "loss": 2.8234,
+      "step": 52298
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6777426600456238,
+      "learning_rate": 3.086203992971568e-05,
+      "loss": 2.9562,
+      "step": 52299
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7489884495735168,
+      "learning_rate": 3.085523865247851e-05,
+      "loss": 2.8091,
+      "step": 52300
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7176619172096252,
+      "learning_rate": 3.084843808411687e-05,
+      "loss": 3.1135,
+      "step": 52301
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7485775351524353,
+      "learning_rate": 3.084163822464857e-05,
+      "loss": 2.8253,
+      "step": 52302
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7570214867591858,
+      "learning_rate": 3.08348390740917e-05,
+      "loss": 2.7535,
+      "step": 52303
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7418150305747986,
+      "learning_rate": 3.082804063246401e-05,
+      "loss": 3.0901,
+      "step": 52304
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6917791962623596,
+      "learning_rate": 3.0821242899783425e-05,
+      "loss": 2.8208,
+      "step": 52305
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7729413509368896,
+      "learning_rate": 3.081444587606796e-05,
+      "loss": 2.76,
+      "step": 52306
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7332080006599426,
+      "learning_rate": 3.0807649561335335e-05,
+      "loss": 3.0991,
+      "step": 52307
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7556661367416382,
+      "learning_rate": 3.08008539556036e-05,
+      "loss": 2.9618,
+      "step": 52308
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7616263031959534,
+      "learning_rate": 3.079405905889064e-05,
+      "loss": 2.7836,
+      "step": 52309
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7406206727027893,
+      "learning_rate": 3.078726487121431e-05,
+      "loss": 2.9042,
+      "step": 52310
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7701447010040283,
+      "learning_rate": 3.078047139259253e-05,
+      "loss": 3.0468,
+      "step": 52311
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7630977630615234,
+      "learning_rate": 3.077367862304312e-05,
+      "loss": 2.8137,
+      "step": 52312
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7166601419448853,
+      "learning_rate": 3.0766886562584024e-05,
+      "loss": 2.9846,
+      "step": 52313
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7290651202201843,
+      "learning_rate": 3.07600952112332e-05,
+      "loss": 2.9109,
+      "step": 52314
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7534510493278503,
+      "learning_rate": 3.075330456900841e-05,
+      "loss": 3.0675,
+      "step": 52315
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7178420424461365,
+      "learning_rate": 3.074651463592762e-05,
+      "loss": 2.9887,
+      "step": 52316
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7508209347724915,
+      "learning_rate": 3.0739725412008665e-05,
+      "loss": 2.9242,
+      "step": 52317
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7520042061805725,
+      "learning_rate": 3.073293689726952e-05,
+      "loss": 2.9403,
+      "step": 52318
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7215971946716309,
+      "learning_rate": 3.072614909172798e-05,
+      "loss": 3.038,
+      "step": 52319
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.761830747127533,
+      "learning_rate": 3.071936199540189e-05,
+      "loss": 2.9039,
+      "step": 52320
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7258598208427429,
+      "learning_rate": 3.0712575608309206e-05,
+      "loss": 2.8936,
+      "step": 52321
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7424689531326294,
+      "learning_rate": 3.070578993046775e-05,
+      "loss": 2.8634,
+      "step": 52322
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7346766591072083,
+      "learning_rate": 3.06990049618954e-05,
+      "loss": 3.0474,
+      "step": 52323
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7245206236839294,
+      "learning_rate": 3.0692220702610115e-05,
+      "loss": 2.8944,
+      "step": 52324
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7064627408981323,
+      "learning_rate": 3.0685437152629676e-05,
+      "loss": 2.9548,
+      "step": 52325
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7573798894882202,
+      "learning_rate": 3.0678654311971974e-05,
+      "loss": 2.772,
+      "step": 52326
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6997696161270142,
+      "learning_rate": 3.0671872180654825e-05,
+      "loss": 2.7829,
+      "step": 52327
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7369567155838013,
+      "learning_rate": 3.066509075869611e-05,
+      "loss": 2.819,
+      "step": 52328
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7621482014656067,
+      "learning_rate": 3.065831004611375e-05,
+      "loss": 2.9502,
+      "step": 52329
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7581804394721985,
+      "learning_rate": 3.065153004292554e-05,
+      "loss": 2.7999,
+      "step": 52330
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7452824711799622,
+      "learning_rate": 3.064475074914939e-05,
+      "loss": 3.0601,
+      "step": 52331
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7450490593910217,
+      "learning_rate": 3.063797216480316e-05,
+      "loss": 2.7313,
+      "step": 52332
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7387819290161133,
+      "learning_rate": 3.063119428990459e-05,
+      "loss": 3.134,
+      "step": 52333
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7497106790542603,
+      "learning_rate": 3.062441712447171e-05,
+      "loss": 2.9456,
+      "step": 52334
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7112284898757935,
+      "learning_rate": 3.061764066852217e-05,
+      "loss": 2.6568,
+      "step": 52335
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.753256618976593,
+      "learning_rate": 3.0610864922073985e-05,
+      "loss": 2.8336,
+      "step": 52336
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6917237639427185,
+      "learning_rate": 3.0604089885144914e-05,
+      "loss": 2.8997,
+      "step": 52337
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7741192579269409,
+      "learning_rate": 3.05973155577528e-05,
+      "loss": 2.942,
+      "step": 52338
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7585358619689941,
+      "learning_rate": 3.059054193991557e-05,
+      "loss": 2.755,
+      "step": 52339
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7429476380348206,
+      "learning_rate": 3.058376903165097e-05,
+      "loss": 3.0445,
+      "step": 52340
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7957443594932556,
+      "learning_rate": 3.057699683297693e-05,
+      "loss": 3.0225,
+      "step": 52341
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7200005650520325,
+      "learning_rate": 3.057022534391113e-05,
+      "loss": 3.0441,
+      "step": 52342
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7461026310920715,
+      "learning_rate": 3.056345456447151e-05,
+      "loss": 2.8421,
+      "step": 52343
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7428950071334839,
+      "learning_rate": 3.055668449467598e-05,
+      "loss": 2.6936,
+      "step": 52344
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7124583721160889,
+      "learning_rate": 3.05499151345422e-05,
+      "loss": 3.1548,
+      "step": 52345
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7571352124214172,
+      "learning_rate": 3.054314648408815e-05,
+      "loss": 2.9257,
+      "step": 52346
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7435238361358643,
+      "learning_rate": 3.0536378543331616e-05,
+      "loss": 2.7556,
+      "step": 52347
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.733097493648529,
+      "learning_rate": 3.052961131229032e-05,
+      "loss": 2.807,
+      "step": 52348
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7678351998329163,
+      "learning_rate": 3.052284479098225e-05,
+      "loss": 2.7884,
+      "step": 52349
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7466233372688293,
+      "learning_rate": 3.051607897942504e-05,
+      "loss": 2.9157,
+      "step": 52350
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7528746724128723,
+      "learning_rate": 3.050931387763671e-05,
+      "loss": 2.8802,
+      "step": 52351
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7836788296699524,
+      "learning_rate": 3.0502549485634908e-05,
+      "loss": 2.6929,
+      "step": 52352
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7055717706680298,
+      "learning_rate": 3.0495785803437557e-05,
+      "loss": 2.9488,
+      "step": 52353
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7400059103965759,
+      "learning_rate": 3.0489022831062472e-05,
+      "loss": 2.755,
+      "step": 52354
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7582606077194214,
+      "learning_rate": 3.048226056852744e-05,
+      "loss": 2.8012,
+      "step": 52355
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.751268208026886,
+      "learning_rate": 3.0475499015850247e-05,
+      "loss": 2.6739,
+      "step": 52356
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.741468071937561,
+      "learning_rate": 3.0468738173048678e-05,
+      "loss": 2.9144,
+      "step": 52357
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.6897956132888794,
+      "learning_rate": 3.0461978040140557e-05,
+      "loss": 3.1279,
+      "step": 52358
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7293424606323242,
+      "learning_rate": 3.0455218617143796e-05,
+      "loss": 3.0598,
+      "step": 52359
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8046936988830566,
+      "learning_rate": 3.044845990407605e-05,
+      "loss": 2.8674,
+      "step": 52360
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7003205418586731,
+      "learning_rate": 3.044170190095524e-05,
+      "loss": 2.8303,
+      "step": 52361
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7055088877677917,
+      "learning_rate": 3.0434944607799116e-05,
+      "loss": 2.8384,
+      "step": 52362
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7319667935371399,
+      "learning_rate": 3.04281880246254e-05,
+      "loss": 2.9059,
+      "step": 52363
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7617000937461853,
+      "learning_rate": 3.042143215145201e-05,
+      "loss": 2.8728,
+      "step": 52364
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7651677131652832,
+      "learning_rate": 3.041467698829666e-05,
+      "loss": 2.963,
+      "step": 52365
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7394811511039734,
+      "learning_rate": 3.0407922535177176e-05,
+      "loss": 2.7771,
+      "step": 52366
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7691500782966614,
+      "learning_rate": 3.040116879211131e-05,
+      "loss": 2.866,
+      "step": 52367
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7116948962211609,
+      "learning_rate": 3.039441575911694e-05,
+      "loss": 2.9602,
+      "step": 52368
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7746070623397827,
+      "learning_rate": 3.0387663436211796e-05,
+      "loss": 2.8734,
+      "step": 52369
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.743488073348999,
+      "learning_rate": 3.038091182341359e-05,
+      "loss": 2.7054,
+      "step": 52370
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7217757105827332,
+      "learning_rate": 3.0374160920740242e-05,
+      "loss": 2.8834,
+      "step": 52371
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7398301362991333,
+      "learning_rate": 3.0367410728209408e-05,
+      "loss": 3.0355,
+      "step": 52372
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7479260563850403,
+      "learning_rate": 3.036066124583897e-05,
+      "loss": 2.9872,
+      "step": 52373
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7478479743003845,
+      "learning_rate": 3.0353912473646615e-05,
+      "loss": 2.9438,
+      "step": 52374
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7435922026634216,
+      "learning_rate": 3.0347164411650126e-05,
+      "loss": 2.8484,
+      "step": 52375
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7409965991973877,
+      "learning_rate": 3.0340417059867396e-05,
+      "loss": 2.9877,
+      "step": 52376
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.748325526714325,
+      "learning_rate": 3.0333670418316103e-05,
+      "loss": 2.9524,
+      "step": 52377
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7367417216300964,
+      "learning_rate": 3.0326924487014005e-05,
+      "loss": 2.951,
+      "step": 52378
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.753197431564331,
+      "learning_rate": 3.032017926597885e-05,
+      "loss": 3.1165,
+      "step": 52379
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7246016263961792,
+      "learning_rate": 3.031343475522846e-05,
+      "loss": 2.8368,
+      "step": 52380
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7062520980834961,
+      "learning_rate": 3.0306690954780622e-05,
+      "loss": 2.9998,
+      "step": 52381
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.73405522108078,
+      "learning_rate": 3.0299947864652984e-05,
+      "loss": 2.8911,
+      "step": 52382
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7310712337493896,
+      "learning_rate": 3.0293205484863435e-05,
+      "loss": 2.8359,
+      "step": 52383
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.759092390537262,
+      "learning_rate": 3.0286463815429695e-05,
+      "loss": 3.0487,
+      "step": 52384
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7694963812828064,
+      "learning_rate": 3.0279722856369447e-05,
+      "loss": 2.7858,
+      "step": 52385
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7476155161857605,
+      "learning_rate": 3.0272982607700546e-05,
+      "loss": 3.0141,
+      "step": 52386
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7267019152641296,
+      "learning_rate": 3.0266243069440645e-05,
+      "loss": 2.9373,
+      "step": 52387
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7420546412467957,
+      "learning_rate": 3.0259504241607625e-05,
+      "loss": 3.0561,
+      "step": 52388
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7914209961891174,
+      "learning_rate": 3.0252766124219076e-05,
+      "loss": 3.0165,
+      "step": 52389
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7075768113136292,
+      "learning_rate": 3.024602871729288e-05,
+      "loss": 3.1116,
+      "step": 52390
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7824947237968445,
+      "learning_rate": 3.023929202084676e-05,
+      "loss": 2.9569,
+      "step": 52391
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7449918985366821,
+      "learning_rate": 3.0232556034898336e-05,
+      "loss": 2.8483,
+      "step": 52392
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7132568359375,
+      "learning_rate": 3.0225820759465525e-05,
+      "loss": 2.9501,
+      "step": 52393
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8174427151679993,
+      "learning_rate": 3.0219086194565913e-05,
+      "loss": 2.925,
+      "step": 52394
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7686358690261841,
+      "learning_rate": 3.0212352340217283e-05,
+      "loss": 2.7436,
+      "step": 52395
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7603972554206848,
+      "learning_rate": 3.020561919643749e-05,
+      "loss": 2.8535,
+      "step": 52396
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7688196301460266,
+      "learning_rate": 3.0198886763244124e-05,
+      "loss": 2.8584,
+      "step": 52397
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7407582402229309,
+      "learning_rate": 3.0192155040655e-05,
+      "loss": 2.8672,
+      "step": 52398
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7440164685249329,
+      "learning_rate": 3.01854240286878e-05,
+      "loss": 2.8893,
+      "step": 52399
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.768831193447113,
+      "learning_rate": 3.0178693727360247e-05,
+      "loss": 2.7461,
+      "step": 52400
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7303265333175659,
+      "learning_rate": 3.0171964136690098e-05,
+      "loss": 2.874,
+      "step": 52401
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7380855083465576,
+      "learning_rate": 3.016523525669503e-05,
+      "loss": 2.6449,
+      "step": 52402
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7118068933486938,
+      "learning_rate": 3.015850708739287e-05,
+      "loss": 2.978,
+      "step": 52403
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7120081782341003,
+      "learning_rate": 3.0151779628801198e-05,
+      "loss": 2.8128,
+      "step": 52404
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7590805292129517,
+      "learning_rate": 3.0145052880937903e-05,
+      "loss": 2.8198,
+      "step": 52405
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7395599484443665,
+      "learning_rate": 3.013832684382057e-05,
+      "loss": 2.7616,
+      "step": 52406
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7489631772041321,
+      "learning_rate": 3.0131601517466885e-05,
+      "loss": 2.9875,
+      "step": 52407
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7233737707138062,
+      "learning_rate": 3.012487690189467e-05,
+      "loss": 2.8304,
+      "step": 52408
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.808874785900116,
+      "learning_rate": 3.011815299712157e-05,
+      "loss": 2.8516,
+      "step": 52409
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7949105501174927,
+      "learning_rate": 3.0111429803165277e-05,
+      "loss": 2.6781,
+      "step": 52410
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7281270027160645,
+      "learning_rate": 3.0104707320043643e-05,
+      "loss": 2.8587,
+      "step": 52411
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7180661559104919,
+      "learning_rate": 3.0097985547774216e-05,
+      "loss": 2.7965,
+      "step": 52412
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7658478021621704,
+      "learning_rate": 3.0091264486374788e-05,
+      "loss": 2.6224,
+      "step": 52413
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7765020132064819,
+      "learning_rate": 3.0084544135862943e-05,
+      "loss": 3.065,
+      "step": 52414
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7255092263221741,
+      "learning_rate": 3.0077824496256496e-05,
+      "loss": 2.8975,
+      "step": 52415
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7271882891654968,
+      "learning_rate": 3.007110556757314e-05,
+      "loss": 2.7939,
+      "step": 52416
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7662795782089233,
+      "learning_rate": 3.0064387349830486e-05,
+      "loss": 2.9394,
+      "step": 52417
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7720481753349304,
+      "learning_rate": 3.005766984304636e-05,
+      "loss": 3.0693,
+      "step": 52418
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7715625762939453,
+      "learning_rate": 3.005095304723831e-05,
+      "loss": 2.8276,
+      "step": 52419
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7382410764694214,
+      "learning_rate": 3.0044236962424162e-05,
+      "loss": 2.9642,
+      "step": 52420
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7307043075561523,
+      "learning_rate": 3.0037521588621527e-05,
+      "loss": 2.7719,
+      "step": 52421
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7969446182250977,
+      "learning_rate": 3.0030806925848062e-05,
+      "loss": 2.9994,
+      "step": 52422
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7565550208091736,
+      "learning_rate": 3.0024092974121517e-05,
+      "loss": 2.9169,
+      "step": 52423
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7802337408065796,
+      "learning_rate": 3.0017379733459545e-05,
+      "loss": 2.9471,
+      "step": 52424
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7368726134300232,
+      "learning_rate": 3.00106672038798e-05,
+      "loss": 2.9638,
+      "step": 52425
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7419968247413635,
+      "learning_rate": 3.000395538540007e-05,
+      "loss": 2.9303,
+      "step": 52426
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7627372741699219,
+      "learning_rate": 2.9997244278037967e-05,
+      "loss": 3.1326,
+      "step": 52427
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7172695994377136,
+      "learning_rate": 2.999053388181115e-05,
+      "loss": 2.9077,
+      "step": 52428
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7625597715377808,
+      "learning_rate": 2.9983824196737237e-05,
+      "loss": 2.6087,
+      "step": 52429
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7408584952354431,
+      "learning_rate": 2.997711522283398e-05,
+      "loss": 3.0352,
+      "step": 52430
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7761721611022949,
+      "learning_rate": 2.9970406960119096e-05,
+      "loss": 2.7375,
+      "step": 52431
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7523345351219177,
+      "learning_rate": 2.996369940861011e-05,
+      "loss": 3.0537,
+      "step": 52432
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7538020610809326,
+      "learning_rate": 2.9956992568324833e-05,
+      "loss": 2.8131,
+      "step": 52433
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7494543790817261,
+      "learning_rate": 2.995028643928089e-05,
+      "loss": 2.9724,
+      "step": 52434
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8455016016960144,
+      "learning_rate": 2.9943581021495833e-05,
+      "loss": 2.8156,
+      "step": 52435
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7785493731498718,
+      "learning_rate": 2.9936876314987513e-05,
+      "loss": 2.874,
+      "step": 52436
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.740427553653717,
+      "learning_rate": 2.9930172319773383e-05,
+      "loss": 2.8312,
+      "step": 52437
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7532118558883667,
+      "learning_rate": 2.9923469035871294e-05,
+      "loss": 2.6936,
+      "step": 52438
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7429023385047913,
+      "learning_rate": 2.9916766463298736e-05,
+      "loss": 2.7284,
+      "step": 52439
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7911916971206665,
+      "learning_rate": 2.9910064602073426e-05,
+      "loss": 2.9946,
+      "step": 52440
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7958114743232727,
+      "learning_rate": 2.990336345221308e-05,
+      "loss": 2.8705,
+      "step": 52441
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7775363326072693,
+      "learning_rate": 2.9896663013735324e-05,
+      "loss": 2.8921,
+      "step": 52442
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7311971187591553,
+      "learning_rate": 2.9889963286657737e-05,
+      "loss": 2.9018,
+      "step": 52443
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7621795535087585,
+      "learning_rate": 2.9883264270997977e-05,
+      "loss": 3.0524,
+      "step": 52444
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7280380725860596,
+      "learning_rate": 2.987656596677376e-05,
+      "loss": 2.9503,
+      "step": 52445
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7853645086288452,
+      "learning_rate": 2.9869868374002637e-05,
+      "loss": 2.7609,
+      "step": 52446
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7341321706771851,
+      "learning_rate": 2.9863171492702263e-05,
+      "loss": 2.7723,
+      "step": 52447
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7179156541824341,
+      "learning_rate": 2.9856475322890396e-05,
+      "loss": 2.89,
+      "step": 52448
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7569622993469238,
+      "learning_rate": 2.9849779864584577e-05,
+      "loss": 2.9834,
+      "step": 52449
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7575824856758118,
+      "learning_rate": 2.9843085117802433e-05,
+      "loss": 2.9927,
+      "step": 52450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7787556052207947,
+      "learning_rate": 2.983639108256155e-05,
+      "loss": 2.9165,
+      "step": 52451
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7417843341827393,
+      "learning_rate": 2.9829697758879645e-05,
+      "loss": 3.0732,
+      "step": 52452
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.756346583366394,
+      "learning_rate": 2.9823005146774336e-05,
+      "loss": 3.1536,
+      "step": 52453
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7447629570960999,
+      "learning_rate": 2.981631324626321e-05,
+      "loss": 2.9677,
+      "step": 52454
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7455997467041016,
+      "learning_rate": 2.9809622057363992e-05,
+      "loss": 2.8583,
+      "step": 52455
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7502301335334778,
+      "learning_rate": 2.9802931580094126e-05,
+      "loss": 2.9962,
+      "step": 52456
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7398320436477661,
+      "learning_rate": 2.9796241814471432e-05,
+      "loss": 3.0755,
+      "step": 52457
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8331220149993896,
+      "learning_rate": 2.9789552760513403e-05,
+      "loss": 2.8654,
+      "step": 52458
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.711363673210144,
+      "learning_rate": 2.978286441823765e-05,
+      "loss": 3.0171,
+      "step": 52459
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7725111246109009,
+      "learning_rate": 2.97761767876619e-05,
+      "loss": 2.9687,
+      "step": 52460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7222541570663452,
+      "learning_rate": 2.9769489868803598e-05,
+      "loss": 2.7685,
+      "step": 52461
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7226037979125977,
+      "learning_rate": 2.9762803661680502e-05,
+      "loss": 2.8108,
+      "step": 52462
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7615007162094116,
+      "learning_rate": 2.9756118166310193e-05,
+      "loss": 2.9072,
+      "step": 52463
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7283841967582703,
+      "learning_rate": 2.9749433382710265e-05,
+      "loss": 2.8891,
+      "step": 52464
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7138139605522156,
+      "learning_rate": 2.9742749310898327e-05,
+      "loss": 2.6945,
+      "step": 52465
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7099862694740295,
+      "learning_rate": 2.9736065950891908e-05,
+      "loss": 2.9704,
+      "step": 52466
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.6979156136512756,
+      "learning_rate": 2.9729383302708688e-05,
+      "loss": 2.8211,
+      "step": 52467
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.734841525554657,
+      "learning_rate": 2.972270136636632e-05,
+      "loss": 3.0207,
+      "step": 52468
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7860302329063416,
+      "learning_rate": 2.971602014188229e-05,
+      "loss": 2.7577,
+      "step": 52469
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7355887293815613,
+      "learning_rate": 2.9709339629274285e-05,
+      "loss": 3.0183,
+      "step": 52470
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.737175464630127,
+      "learning_rate": 2.970265982855986e-05,
+      "loss": 2.9608,
+      "step": 52471
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7097364664077759,
+      "learning_rate": 2.9695980739756564e-05,
+      "loss": 2.9803,
+      "step": 52472
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7723646759986877,
+      "learning_rate": 2.9689302362882084e-05,
+      "loss": 2.8259,
+      "step": 52473
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7352739572525024,
+      "learning_rate": 2.9682624697953873e-05,
+      "loss": 2.9392,
+      "step": 52474
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7245145440101624,
+      "learning_rate": 2.9675947744989716e-05,
+      "loss": 2.7403,
+      "step": 52475
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7283601760864258,
+      "learning_rate": 2.9669271504006997e-05,
+      "loss": 2.7105,
+      "step": 52476
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7669225335121155,
+      "learning_rate": 2.9662595975023406e-05,
+      "loss": 2.8408,
+      "step": 52477
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7420752048492432,
+      "learning_rate": 2.965592115805656e-05,
+      "loss": 2.8891,
+      "step": 52478
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7681543827056885,
+      "learning_rate": 2.964924705312398e-05,
+      "loss": 2.959,
+      "step": 52479
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8160313367843628,
+      "learning_rate": 2.9642573660243252e-05,
+      "loss": 3.0549,
+      "step": 52480
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7001559138298035,
+      "learning_rate": 2.9635900979431927e-05,
+      "loss": 2.8459,
+      "step": 52481
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7368727326393127,
+      "learning_rate": 2.9629229010707588e-05,
+      "loss": 2.7764,
+      "step": 52482
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7512417435646057,
+      "learning_rate": 2.962255775408786e-05,
+      "loss": 2.8818,
+      "step": 52483
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7434666752815247,
+      "learning_rate": 2.9615887209590263e-05,
+      "loss": 2.8293,
+      "step": 52484
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7331339120864868,
+      "learning_rate": 2.960921737723241e-05,
+      "loss": 2.8604,
+      "step": 52485
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7475157976150513,
+      "learning_rate": 2.9602548257031854e-05,
+      "loss": 2.8809,
+      "step": 52486
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7474045753479004,
+      "learning_rate": 2.959587984900609e-05,
+      "loss": 3.0424,
+      "step": 52487
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7134113311767578,
+      "learning_rate": 2.9589212153172794e-05,
+      "loss": 2.8692,
+      "step": 52488
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7441120147705078,
+      "learning_rate": 2.958254516954939e-05,
+      "loss": 2.7825,
+      "step": 52489
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7090118527412415,
+      "learning_rate": 2.9575878898153594e-05,
+      "loss": 2.9669,
+      "step": 52490
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7831944227218628,
+      "learning_rate": 2.956921333900286e-05,
+      "loss": 2.934,
+      "step": 52491
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7588694095611572,
+      "learning_rate": 2.9562548492114814e-05,
+      "loss": 2.9626,
+      "step": 52492
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7268178462982178,
+      "learning_rate": 2.955588435750693e-05,
+      "loss": 2.8937,
+      "step": 52493
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7331435084342957,
+      "learning_rate": 2.954922093519677e-05,
+      "loss": 2.879,
+      "step": 52494
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7241976261138916,
+      "learning_rate": 2.9542558225201984e-05,
+      "loss": 2.7338,
+      "step": 52495
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7318156957626343,
+      "learning_rate": 2.953589622753999e-05,
+      "loss": 2.8821,
+      "step": 52496
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7937421202659607,
+      "learning_rate": 2.9529234942228376e-05,
+      "loss": 2.8645,
+      "step": 52497
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7748422622680664,
+      "learning_rate": 2.952257436928476e-05,
+      "loss": 2.8881,
+      "step": 52498
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7224046587944031,
+      "learning_rate": 2.951591450872659e-05,
+      "loss": 2.9246,
+      "step": 52499
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7512742877006531,
+      "learning_rate": 2.9509255360571493e-05,
+      "loss": 2.9546,
+      "step": 52500
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7295030951499939,
+      "learning_rate": 2.9502596924836952e-05,
+      "loss": 2.9228,
+      "step": 52501
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7399193644523621,
+      "learning_rate": 2.949593920154045e-05,
+      "loss": 2.7575,
+      "step": 52502
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7585479021072388,
+      "learning_rate": 2.948928219069968e-05,
+      "loss": 2.8837,
+      "step": 52503
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8097811341285706,
+      "learning_rate": 2.948262589233199e-05,
+      "loss": 2.9717,
+      "step": 52504
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7251771688461304,
+      "learning_rate": 2.9475970306455065e-05,
+      "loss": 2.9217,
+      "step": 52505
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7347126603126526,
+      "learning_rate": 2.9469315433086327e-05,
+      "loss": 2.8441,
+      "step": 52506
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7708790302276611,
+      "learning_rate": 2.9462661272243394e-05,
+      "loss": 2.7953,
+      "step": 52507
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7497082948684692,
+      "learning_rate": 2.9456007823943716e-05,
+      "loss": 2.8825,
+      "step": 52508
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7288712859153748,
+      "learning_rate": 2.9449355088204817e-05,
+      "loss": 2.8716,
+      "step": 52509
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.745967447757721,
+      "learning_rate": 2.9442703065044314e-05,
+      "loss": 2.9418,
+      "step": 52510
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7650946378707886,
+      "learning_rate": 2.9436051754479595e-05,
+      "loss": 3.0354,
+      "step": 52511
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7825707793235779,
+      "learning_rate": 2.9429401156528242e-05,
+      "loss": 2.9016,
+      "step": 52512
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.775109052658081,
+      "learning_rate": 2.9422751271207845e-05,
+      "loss": 2.8788,
+      "step": 52513
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7160129547119141,
+      "learning_rate": 2.941610209853582e-05,
+      "loss": 2.8551,
+      "step": 52514
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7564792037010193,
+      "learning_rate": 2.9409453638529755e-05,
+      "loss": 2.9526,
+      "step": 52515
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7617117166519165,
+      "learning_rate": 2.9402805891207006e-05,
+      "loss": 2.9431,
+      "step": 52516
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7508881688117981,
+      "learning_rate": 2.9396158856585216e-05,
+      "loss": 2.9826,
+      "step": 52517
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7413908839225769,
+      "learning_rate": 2.9389512534681914e-05,
+      "loss": 2.6746,
+      "step": 52518
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.749812126159668,
+      "learning_rate": 2.938286692551448e-05,
+      "loss": 2.9919,
+      "step": 52519
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7722899317741394,
+      "learning_rate": 2.937622202910057e-05,
+      "loss": 3.0812,
+      "step": 52520
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7824468016624451,
+      "learning_rate": 2.936957784545757e-05,
+      "loss": 2.9144,
+      "step": 52521
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7449400424957275,
+      "learning_rate": 2.936293437460303e-05,
+      "loss": 3.1251,
+      "step": 52522
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7934387922286987,
+      "learning_rate": 2.9356291616554473e-05,
+      "loss": 2.9064,
+      "step": 52523
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7479174733161926,
+      "learning_rate": 2.934964957132928e-05,
+      "loss": 3.1097,
+      "step": 52524
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7839178442955017,
+      "learning_rate": 2.934300823894511e-05,
+      "loss": 2.9528,
+      "step": 52525
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7495297789573669,
+      "learning_rate": 2.9336367619419276e-05,
+      "loss": 2.9153,
+      "step": 52526
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7545944452285767,
+      "learning_rate": 2.9329727712769435e-05,
+      "loss": 2.9136,
+      "step": 52527
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7327878475189209,
+      "learning_rate": 2.9323088519012938e-05,
+      "loss": 3.0669,
+      "step": 52528
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7432766556739807,
+      "learning_rate": 2.9316450038167372e-05,
+      "loss": 3.0313,
+      "step": 52529
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7075245380401611,
+      "learning_rate": 2.930981227025022e-05,
+      "loss": 2.9219,
+      "step": 52530
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.740532398223877,
+      "learning_rate": 2.930317521527884e-05,
+      "loss": 2.7855,
+      "step": 52531
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7468576431274414,
+      "learning_rate": 2.9296538873270847e-05,
+      "loss": 2.8628,
+      "step": 52532
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7329097390174866,
+      "learning_rate": 2.9289903244243662e-05,
+      "loss": 2.8902,
+      "step": 52533
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7266642451286316,
+      "learning_rate": 2.9283268328214737e-05,
+      "loss": 3.0384,
+      "step": 52534
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7914463877677917,
+      "learning_rate": 2.9276634125201627e-05,
+      "loss": 3.044,
+      "step": 52535
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7478669285774231,
+      "learning_rate": 2.927000063522178e-05,
+      "loss": 2.7363,
+      "step": 52536
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7697821855545044,
+      "learning_rate": 2.9263367858292653e-05,
+      "loss": 2.6884,
+      "step": 52537
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7177620530128479,
+      "learning_rate": 2.9256735794431662e-05,
+      "loss": 3.0663,
+      "step": 52538
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7495757937431335,
+      "learning_rate": 2.9250104443656295e-05,
+      "loss": 2.8574,
+      "step": 52539
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7898780703544617,
+      "learning_rate": 2.924347380598414e-05,
+      "loss": 3.1215,
+      "step": 52540
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7186577320098877,
+      "learning_rate": 2.9236843881432482e-05,
+      "loss": 2.8895,
+      "step": 52541
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7339335680007935,
+      "learning_rate": 2.9230214670018936e-05,
+      "loss": 3.0893,
+      "step": 52542
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7203998565673828,
+      "learning_rate": 2.9223586171760827e-05,
+      "loss": 2.7479,
+      "step": 52543
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7361055612564087,
+      "learning_rate": 2.9216958386675738e-05,
+      "loss": 2.9828,
+      "step": 52544
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7605053186416626,
+      "learning_rate": 2.9210331314781088e-05,
+      "loss": 2.8262,
+      "step": 52545
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7568164467811584,
+      "learning_rate": 2.920370495609423e-05,
+      "loss": 2.8678,
+      "step": 52546
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.703392744064331,
+      "learning_rate": 2.9197079310632787e-05,
+      "loss": 2.7076,
+      "step": 52547
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7546159625053406,
+      "learning_rate": 2.9190454378414007e-05,
+      "loss": 3.0159,
+      "step": 52548
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7149426937103271,
+      "learning_rate": 2.9183830159455513e-05,
+      "loss": 2.7881,
+      "step": 52549
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7450437545776367,
+      "learning_rate": 2.9177206653774722e-05,
+      "loss": 2.8648,
+      "step": 52550
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7561696171760559,
+      "learning_rate": 2.917058386138902e-05,
+      "loss": 2.7633,
+      "step": 52551
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7118037939071655,
+      "learning_rate": 2.9163961782315924e-05,
+      "loss": 2.9854,
+      "step": 52552
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7305014133453369,
+      "learning_rate": 2.9157340416572727e-05,
+      "loss": 2.6679,
+      "step": 52553
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7449045181274414,
+      "learning_rate": 2.915071976417701e-05,
+      "loss": 2.8047,
+      "step": 52554
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7418058514595032,
+      "learning_rate": 2.9144099825146194e-05,
+      "loss": 2.8945,
+      "step": 52555
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7265969514846802,
+      "learning_rate": 2.9137480599497664e-05,
+      "loss": 3.0225,
+      "step": 52556
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7401600480079651,
+      "learning_rate": 2.913086208724894e-05,
+      "loss": 2.9199,
+      "step": 52557
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7653848528862,
+      "learning_rate": 2.912424428841731e-05,
+      "loss": 2.897,
+      "step": 52558
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7716215252876282,
+      "learning_rate": 2.9117627203020354e-05,
+      "loss": 3.0183,
+      "step": 52559
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7675553560256958,
+      "learning_rate": 2.9111010831075467e-05,
+      "loss": 2.9278,
+      "step": 52560
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8107246160507202,
+      "learning_rate": 2.910439517259996e-05,
+      "loss": 2.7924,
+      "step": 52561
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7437394261360168,
+      "learning_rate": 2.909778022761139e-05,
+      "loss": 2.9086,
+      "step": 52562
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7718828320503235,
+      "learning_rate": 2.909116599612711e-05,
+      "loss": 2.9937,
+      "step": 52563
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7451728582382202,
+      "learning_rate": 2.9084552478164537e-05,
+      "loss": 2.7001,
+      "step": 52564
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7395138144493103,
+      "learning_rate": 2.9077939673741158e-05,
+      "loss": 3.1498,
+      "step": 52565
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7220340967178345,
+      "learning_rate": 2.907132758287436e-05,
+      "loss": 3.1024,
+      "step": 52566
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7452659010887146,
+      "learning_rate": 2.9064716205581527e-05,
+      "loss": 3.0528,
+      "step": 52567
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7208335995674133,
+      "learning_rate": 2.905810554188005e-05,
+      "loss": 2.9171,
+      "step": 52568
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7047619223594666,
+      "learning_rate": 2.9051495591787376e-05,
+      "loss": 2.9405,
+      "step": 52569
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7219128608703613,
+      "learning_rate": 2.9044886355320996e-05,
+      "loss": 2.9738,
+      "step": 52570
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7547860145568848,
+      "learning_rate": 2.9038277832498157e-05,
+      "loss": 2.9313,
+      "step": 52571
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7108772397041321,
+      "learning_rate": 2.9031670023336418e-05,
+      "loss": 3.0976,
+      "step": 52572
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7670025825500488,
+      "learning_rate": 2.9025062927853094e-05,
+      "loss": 2.9123,
+      "step": 52573
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7210864424705505,
+      "learning_rate": 2.901845654606554e-05,
+      "loss": 2.8945,
+      "step": 52574
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7419854998588562,
+      "learning_rate": 2.9011850877991306e-05,
+      "loss": 3.1037,
+      "step": 52575
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7315715551376343,
+      "learning_rate": 2.9005245923647614e-05,
+      "loss": 2.96,
+      "step": 52576
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7764679789543152,
+      "learning_rate": 2.8998641683052047e-05,
+      "loss": 2.8197,
+      "step": 52577
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7059036493301392,
+      "learning_rate": 2.899203815622183e-05,
+      "loss": 2.9622,
+      "step": 52578
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8265849947929382,
+      "learning_rate": 2.8985435343174414e-05,
+      "loss": 2.9801,
+      "step": 52579
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8718390464782715,
+      "learning_rate": 2.897883324392728e-05,
+      "loss": 2.6337,
+      "step": 52580
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7311578989028931,
+      "learning_rate": 2.897223185849772e-05,
+      "loss": 2.6949,
+      "step": 52581
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7237415909767151,
+      "learning_rate": 2.896563118690315e-05,
+      "loss": 2.9607,
+      "step": 52582
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.730317234992981,
+      "learning_rate": 2.895903122916089e-05,
+      "loss": 2.8626,
+      "step": 52583
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7898061871528625,
+      "learning_rate": 2.895243198528836e-05,
+      "loss": 2.8689,
+      "step": 52584
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7148504853248596,
+      "learning_rate": 2.8945833455303045e-05,
+      "loss": 2.8797,
+      "step": 52585
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.79593425989151,
+      "learning_rate": 2.893923563922217e-05,
+      "loss": 2.9757,
+      "step": 52586
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7755904793739319,
+      "learning_rate": 2.8932638537063245e-05,
+      "loss": 2.9006,
+      "step": 52587
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7300575375556946,
+      "learning_rate": 2.8926042148843566e-05,
+      "loss": 2.7823,
+      "step": 52588
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7470569014549255,
+      "learning_rate": 2.8919446474580477e-05,
+      "loss": 2.8235,
+      "step": 52589
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7756972908973694,
+      "learning_rate": 2.8912851514291468e-05,
+      "loss": 2.8028,
+      "step": 52590
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7556537389755249,
+      "learning_rate": 2.890625726799376e-05,
+      "loss": 2.7269,
+      "step": 52591
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7390879392623901,
+      "learning_rate": 2.889966373570487e-05,
+      "loss": 2.8715,
+      "step": 52592
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7211794853210449,
+      "learning_rate": 2.889307091744202e-05,
+      "loss": 2.8193,
+      "step": 52593
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7417266964912415,
+      "learning_rate": 2.888647881322269e-05,
+      "loss": 2.8811,
+      "step": 52594
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.738362193107605,
+      "learning_rate": 2.887988742306424e-05,
+      "loss": 2.9706,
+      "step": 52595
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7565693855285645,
+      "learning_rate": 2.8873296746983888e-05,
+      "loss": 2.7883,
+      "step": 52596
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.786318838596344,
+      "learning_rate": 2.8866706784999184e-05,
+      "loss": 2.8329,
+      "step": 52597
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7326686978340149,
+      "learning_rate": 2.8860117537127314e-05,
+      "loss": 2.9177,
+      "step": 52598
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.741766095161438,
+      "learning_rate": 2.8853529003385767e-05,
+      "loss": 2.8992,
+      "step": 52599
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7155507206916809,
+      "learning_rate": 2.8846941183791793e-05,
+      "loss": 2.9925,
+      "step": 52600
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7600530385971069,
+      "learning_rate": 2.8840354078362782e-05,
+      "loss": 2.8607,
+      "step": 52601
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7748110294342041,
+      "learning_rate": 2.8833767687116184e-05,
+      "loss": 2.8176,
+      "step": 52602
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7540781497955322,
+      "learning_rate": 2.8827182010069217e-05,
+      "loss": 2.8926,
+      "step": 52603
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7902787327766418,
+      "learning_rate": 2.882059704723927e-05,
+      "loss": 3.0199,
+      "step": 52604
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7752910852432251,
+      "learning_rate": 2.8814012798643627e-05,
+      "loss": 2.8814,
+      "step": 52605
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7527629137039185,
+      "learning_rate": 2.8807429264299708e-05,
+      "loss": 2.6384,
+      "step": 52606
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7904459834098816,
+      "learning_rate": 2.8800846444224867e-05,
+      "loss": 2.8519,
+      "step": 52607
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6968057155609131,
+      "learning_rate": 2.8794264338436325e-05,
+      "loss": 2.8082,
+      "step": 52608
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7696243524551392,
+      "learning_rate": 2.8787682946951596e-05,
+      "loss": 2.8339,
+      "step": 52609
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8164247870445251,
+      "learning_rate": 2.878110226978787e-05,
+      "loss": 2.82,
+      "step": 52610
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7112769484519958,
+      "learning_rate": 2.87745223069625e-05,
+      "loss": 2.7388,
+      "step": 52611
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7222334146499634,
+      "learning_rate": 2.876794305849287e-05,
+      "loss": 2.8311,
+      "step": 52612
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7716138362884521,
+      "learning_rate": 2.8761364524396234e-05,
+      "loss": 2.9988,
+      "step": 52613
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6985487341880798,
+      "learning_rate": 2.8754786704690012e-05,
+      "loss": 2.7424,
+      "step": 52614
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7328038811683655,
+      "learning_rate": 2.874820959939146e-05,
+      "loss": 2.8344,
+      "step": 52615
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7214050889015198,
+      "learning_rate": 2.8741633208517922e-05,
+      "loss": 2.8049,
+      "step": 52616
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7892204523086548,
+      "learning_rate": 2.873505753208676e-05,
+      "loss": 2.8323,
+      "step": 52617
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.743575930595398,
+      "learning_rate": 2.8728482570115185e-05,
+      "loss": 2.9289,
+      "step": 52618
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7116771936416626,
+      "learning_rate": 2.8721908322620625e-05,
+      "loss": 2.827,
+      "step": 52619
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7374192476272583,
+      "learning_rate": 2.871533478962029e-05,
+      "loss": 3.0038,
+      "step": 52620
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7295014262199402,
+      "learning_rate": 2.8708761971131577e-05,
+      "loss": 2.6753,
+      "step": 52621
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7758492827415466,
+      "learning_rate": 2.87021898671718e-05,
+      "loss": 3.0055,
+      "step": 52622
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7198835015296936,
+      "learning_rate": 2.869561847775821e-05,
+      "loss": 2.8754,
+      "step": 52623
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7312856316566467,
+      "learning_rate": 2.8689047802908193e-05,
+      "loss": 2.7471,
+      "step": 52624
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7765367031097412,
+      "learning_rate": 2.8682477842639007e-05,
+      "loss": 2.952,
+      "step": 52625
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7989129424095154,
+      "learning_rate": 2.86759085969679e-05,
+      "loss": 2.8156,
+      "step": 52626
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7187029719352722,
+      "learning_rate": 2.8669340065912295e-05,
+      "loss": 3.0734,
+      "step": 52627
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7583007216453552,
+      "learning_rate": 2.8662772249489373e-05,
+      "loss": 2.8671,
+      "step": 52628
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7283254265785217,
+      "learning_rate": 2.865620514771656e-05,
+      "loss": 2.874,
+      "step": 52629
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.764652669429779,
+      "learning_rate": 2.8649638760611004e-05,
+      "loss": 2.9456,
+      "step": 52630
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7280563116073608,
+      "learning_rate": 2.8643073088190126e-05,
+      "loss": 3.0906,
+      "step": 52631
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7654474973678589,
+      "learning_rate": 2.863650813047118e-05,
+      "loss": 3.0266,
+      "step": 52632
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7615987062454224,
+      "learning_rate": 2.8629943887471418e-05,
+      "loss": 2.9073,
+      "step": 52633
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7482545375823975,
+      "learning_rate": 2.8623380359208194e-05,
+      "loss": 2.8876,
+      "step": 52634
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7490363121032715,
+      "learning_rate": 2.861681754569869e-05,
+      "loss": 2.7844,
+      "step": 52635
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7606791853904724,
+      "learning_rate": 2.861025544696026e-05,
+      "loss": 2.9984,
+      "step": 52636
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7316489815711975,
+      "learning_rate": 2.860369406301026e-05,
+      "loss": 2.8326,
+      "step": 52637
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7653703093528748,
+      "learning_rate": 2.859713339386588e-05,
+      "loss": 2.8809,
+      "step": 52638
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7543324828147888,
+      "learning_rate": 2.8590573439544427e-05,
+      "loss": 2.7399,
+      "step": 52639
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.739680290222168,
+      "learning_rate": 2.8584014200063132e-05,
+      "loss": 2.6263,
+      "step": 52640
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7400057315826416,
+      "learning_rate": 2.8577455675439277e-05,
+      "loss": 2.9609,
+      "step": 52641
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7001802325248718,
+      "learning_rate": 2.8570897865690247e-05,
+      "loss": 2.8452,
+      "step": 52642
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7247951030731201,
+      "learning_rate": 2.856434077083316e-05,
+      "loss": 2.8749,
+      "step": 52643
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7461290955543518,
+      "learning_rate": 2.855778439088544e-05,
+      "loss": 2.8568,
+      "step": 52644
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7752999067306519,
+      "learning_rate": 2.8551228725864205e-05,
+      "loss": 2.9557,
+      "step": 52645
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7620311379432678,
+      "learning_rate": 2.8544673775786842e-05,
+      "loss": 2.9193,
+      "step": 52646
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7228102087974548,
+      "learning_rate": 2.853811954067057e-05,
+      "loss": 2.7064,
+      "step": 52647
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7312363386154175,
+      "learning_rate": 2.8531566020532604e-05,
+      "loss": 2.8944,
+      "step": 52648
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7253252267837524,
+      "learning_rate": 2.8525013215390268e-05,
+      "loss": 3.0867,
+      "step": 52649
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7094355225563049,
+      "learning_rate": 2.8518461125260782e-05,
+      "loss": 3.0171,
+      "step": 52650
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7288451194763184,
+      "learning_rate": 2.85119097501614e-05,
+      "loss": 2.8495,
+      "step": 52651
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7485461235046387,
+      "learning_rate": 2.850535909010947e-05,
+      "loss": 2.9792,
+      "step": 52652
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7915205359458923,
+      "learning_rate": 2.8498809145122182e-05,
+      "loss": 2.8687,
+      "step": 52653
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8087054491043091,
+      "learning_rate": 2.8492259915216754e-05,
+      "loss": 2.746,
+      "step": 52654
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7752242088317871,
+      "learning_rate": 2.848571140041044e-05,
+      "loss": 2.8696,
+      "step": 52655
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7321742177009583,
+      "learning_rate": 2.8479163600720458e-05,
+      "loss": 3.0578,
+      "step": 52656
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7605643272399902,
+      "learning_rate": 2.847261651616419e-05,
+      "loss": 2.7562,
+      "step": 52657
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7781768441200256,
+      "learning_rate": 2.8466070146758734e-05,
+      "loss": 2.7825,
+      "step": 52658
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7603253722190857,
+      "learning_rate": 2.845952449252147e-05,
+      "loss": 2.8124,
+      "step": 52659
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.750426173210144,
+      "learning_rate": 2.845297955346951e-05,
+      "loss": 3.084,
+      "step": 52660
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.765870213508606,
+      "learning_rate": 2.8446435329620122e-05,
+      "loss": 2.8798,
+      "step": 52661
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7285685539245605,
+      "learning_rate": 2.843989182099061e-05,
+      "loss": 2.8378,
+      "step": 52662
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7301684617996216,
+      "learning_rate": 2.8433349027598107e-05,
+      "loss": 2.9406,
+      "step": 52663
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7690022587776184,
+      "learning_rate": 2.8426806949459956e-05,
+      "loss": 2.8652,
+      "step": 52664
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.766257107257843,
+      "learning_rate": 2.842026558659325e-05,
+      "loss": 2.9153,
+      "step": 52665
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7352449893951416,
+      "learning_rate": 2.8413724939015336e-05,
+      "loss": 3.1021,
+      "step": 52666
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.777466893196106,
+      "learning_rate": 2.840718500674344e-05,
+      "loss": 2.8008,
+      "step": 52667
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8289501667022705,
+      "learning_rate": 2.840064578979474e-05,
+      "loss": 2.8083,
+      "step": 52668
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7426512837409973,
+      "learning_rate": 2.8394107288186496e-05,
+      "loss": 2.7525,
+      "step": 52669
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7094964981079102,
+      "learning_rate": 2.8387569501935825e-05,
+      "loss": 3.0038,
+      "step": 52670
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7811385989189148,
+      "learning_rate": 2.8381032431060013e-05,
+      "loss": 2.8479,
+      "step": 52671
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7416138052940369,
+      "learning_rate": 2.837449607557635e-05,
+      "loss": 3.1491,
+      "step": 52672
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7839840650558472,
+      "learning_rate": 2.8367960435501946e-05,
+      "loss": 2.6479,
+      "step": 52673
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.78972327709198,
+      "learning_rate": 2.8361425510854096e-05,
+      "loss": 2.8504,
+      "step": 52674
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7591714262962341,
+      "learning_rate": 2.8354891301649985e-05,
+      "loss": 3.0018,
+      "step": 52675
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7825368046760559,
+      "learning_rate": 2.8348357807906764e-05,
+      "loss": 2.7301,
+      "step": 52676
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.727197527885437,
+      "learning_rate": 2.8341825029641717e-05,
+      "loss": 2.9514,
+      "step": 52677
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7893258333206177,
+      "learning_rate": 2.8335292966872003e-05,
+      "loss": 2.8807,
+      "step": 52678
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7999505996704102,
+      "learning_rate": 2.832876161961487e-05,
+      "loss": 2.8625,
+      "step": 52679
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7148997187614441,
+      "learning_rate": 2.832223098788744e-05,
+      "loss": 2.9241,
+      "step": 52680
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7615914344787598,
+      "learning_rate": 2.831570107170703e-05,
+      "loss": 2.8288,
+      "step": 52681
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7838194370269775,
+      "learning_rate": 2.830917187109073e-05,
+      "loss": 2.7449,
+      "step": 52682
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7049852609634399,
+      "learning_rate": 2.8302643386055855e-05,
+      "loss": 2.8416,
+      "step": 52683
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7759522199630737,
+      "learning_rate": 2.8296115616619497e-05,
+      "loss": 2.8045,
+      "step": 52684
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.726335883140564,
+      "learning_rate": 2.8289588562798836e-05,
+      "loss": 2.8608,
+      "step": 52685
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7436689138412476,
+      "learning_rate": 2.8283062224611165e-05,
+      "loss": 2.8539,
+      "step": 52686
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7301530838012695,
+      "learning_rate": 2.827653660207353e-05,
+      "loss": 2.9732,
+      "step": 52687
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7957533001899719,
+      "learning_rate": 2.8270011695203222e-05,
+      "loss": 2.867,
+      "step": 52688
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6973696947097778,
+      "learning_rate": 2.826348750401749e-05,
+      "loss": 2.9357,
+      "step": 52689
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8173954486846924,
+      "learning_rate": 2.8256964028533425e-05,
+      "loss": 2.7172,
+      "step": 52690
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7530957460403442,
+      "learning_rate": 2.8250441268768174e-05,
+      "loss": 3.071,
+      "step": 52691
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7341398596763611,
+      "learning_rate": 2.8243919224738964e-05,
+      "loss": 2.9068,
+      "step": 52692
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7384122014045715,
+      "learning_rate": 2.8237397896462944e-05,
+      "loss": 2.7069,
+      "step": 52693
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7298815250396729,
+      "learning_rate": 2.8230877283957365e-05,
+      "loss": 3.116,
+      "step": 52694
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7248790860176086,
+      "learning_rate": 2.8224357387239316e-05,
+      "loss": 2.8794,
+      "step": 52695
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.739068329334259,
+      "learning_rate": 2.821783820632605e-05,
+      "loss": 3.0346,
+      "step": 52696
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7384850382804871,
+      "learning_rate": 2.8211319741234684e-05,
+      "loss": 2.8443,
+      "step": 52697
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7337626218795776,
+      "learning_rate": 2.820480199198234e-05,
+      "loss": 2.7513,
+      "step": 52698
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7873049378395081,
+      "learning_rate": 2.8198284958586302e-05,
+      "loss": 2.8253,
+      "step": 52699
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7321355938911438,
+      "learning_rate": 2.8191768641063628e-05,
+      "loss": 2.7804,
+      "step": 52700
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7380569577217102,
+      "learning_rate": 2.8185253039431566e-05,
+      "loss": 2.9944,
+      "step": 52701
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8345293402671814,
+      "learning_rate": 2.8178738153707202e-05,
+      "loss": 3.1125,
+      "step": 52702
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.779485821723938,
+      "learning_rate": 2.8172223983907692e-05,
+      "loss": 2.6968,
+      "step": 52703
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7306099534034729,
+      "learning_rate": 2.8165710530050322e-05,
+      "loss": 2.9204,
+      "step": 52704
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8526567816734314,
+      "learning_rate": 2.8159197792152112e-05,
+      "loss": 2.9036,
+      "step": 52705
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7224339246749878,
+      "learning_rate": 2.815268577023031e-05,
+      "loss": 3.0588,
+      "step": 52706
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7223872542381287,
+      "learning_rate": 2.8146174464301907e-05,
+      "loss": 3.0034,
+      "step": 52707
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7458459138870239,
+      "learning_rate": 2.813966387438419e-05,
+      "loss": 3.0477,
+      "step": 52708
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7503674626350403,
+      "learning_rate": 2.813315400049434e-05,
+      "loss": 2.8695,
+      "step": 52709
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7620118260383606,
+      "learning_rate": 2.8126644842649382e-05,
+      "loss": 2.8405,
+      "step": 52710
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7874787449836731,
+      "learning_rate": 2.8120136400866566e-05,
+      "loss": 2.8848,
+      "step": 52711
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7471845746040344,
+      "learning_rate": 2.811362867516298e-05,
+      "loss": 2.8329,
+      "step": 52712
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8011273741722107,
+      "learning_rate": 2.8107121665555743e-05,
+      "loss": 2.7753,
+      "step": 52713
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6914375424385071,
+      "learning_rate": 2.8100615372062042e-05,
+      "loss": 2.9165,
+      "step": 52714
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7664171457290649,
+      "learning_rate": 2.809410979469896e-05,
+      "loss": 2.8735,
+      "step": 52715
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7507244348526001,
+      "learning_rate": 2.808760493348372e-05,
+      "loss": 2.8179,
+      "step": 52716
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7342262864112854,
+      "learning_rate": 2.808110078843331e-05,
+      "loss": 2.9128,
+      "step": 52717
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7676442265510559,
+      "learning_rate": 2.8074597359565045e-05,
+      "loss": 2.9506,
+      "step": 52718
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7457807064056396,
+      "learning_rate": 2.8068094646895945e-05,
+      "loss": 3.0169,
+      "step": 52719
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.700971782207489,
+      "learning_rate": 2.8061592650443065e-05,
+      "loss": 2.9474,
+      "step": 52720
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7742531895637512,
+      "learning_rate": 2.8055091370223692e-05,
+      "loss": 2.8221,
+      "step": 52721
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7480669617652893,
+      "learning_rate": 2.8048590806254813e-05,
+      "loss": 2.7798,
+      "step": 52722
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7456562519073486,
+      "learning_rate": 2.804209095855361e-05,
+      "loss": 2.9058,
+      "step": 52723
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.865938127040863,
+      "learning_rate": 2.803559182713727e-05,
+      "loss": 2.8553,
+      "step": 52724
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7843738794326782,
+      "learning_rate": 2.8029093412022785e-05,
+      "loss": 2.737,
+      "step": 52725
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7294140458106995,
+      "learning_rate": 2.802259571322737e-05,
+      "loss": 2.752,
+      "step": 52726
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7814497351646423,
+      "learning_rate": 2.801609873076808e-05,
+      "loss": 3.033,
+      "step": 52727
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7317684888839722,
+      "learning_rate": 2.8009602464661996e-05,
+      "loss": 2.8929,
+      "step": 52728
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7592427134513855,
+      "learning_rate": 2.8003106914926344e-05,
+      "loss": 2.9463,
+      "step": 52729
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.053802728652954,
+      "learning_rate": 2.7996612081578074e-05,
+      "loss": 2.9645,
+      "step": 52730
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7302281856536865,
+      "learning_rate": 2.7990117964634475e-05,
+      "loss": 2.8003,
+      "step": 52731
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7234871983528137,
+      "learning_rate": 2.7983624564112494e-05,
+      "loss": 2.8953,
+      "step": 52732
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7112483382225037,
+      "learning_rate": 2.7977131880029324e-05,
+      "loss": 2.7117,
+      "step": 52733
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7166719436645508,
+      "learning_rate": 2.7970639912402047e-05,
+      "loss": 2.9582,
+      "step": 52734
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7226663827896118,
+      "learning_rate": 2.796414866124769e-05,
+      "loss": 2.8281,
+      "step": 52735
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7304128408432007,
+      "learning_rate": 2.7957658126583494e-05,
+      "loss": 3.0941,
+      "step": 52736
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7455897331237793,
+      "learning_rate": 2.7951168308426387e-05,
+      "loss": 2.9331,
+      "step": 52737
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7342404127120972,
+      "learning_rate": 2.7944679206793553e-05,
+      "loss": 2.9336,
+      "step": 52738
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7524769902229309,
+      "learning_rate": 2.7938190821702145e-05,
+      "loss": 2.8062,
+      "step": 52739
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7616720199584961,
+      "learning_rate": 2.7931703153169182e-05,
+      "loss": 2.9295,
+      "step": 52740
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7344446182250977,
+      "learning_rate": 2.792521620121172e-05,
+      "loss": 2.8753,
+      "step": 52741
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7952720522880554,
+      "learning_rate": 2.7918729965846876e-05,
+      "loss": 2.8517,
+      "step": 52742
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7229446768760681,
+      "learning_rate": 2.79122444470917e-05,
+      "loss": 2.7266,
+      "step": 52743
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.751518964767456,
+      "learning_rate": 2.7905759644963355e-05,
+      "loss": 2.9713,
+      "step": 52744
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7525866627693176,
+      "learning_rate": 2.7899275559478818e-05,
+      "loss": 2.9393,
+      "step": 52745
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.752153217792511,
+      "learning_rate": 2.7892792190655312e-05,
+      "loss": 2.991,
+      "step": 52746
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7784674167633057,
+      "learning_rate": 2.788630953850972e-05,
+      "loss": 2.9634,
+      "step": 52747
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7399433255195618,
+      "learning_rate": 2.78798276030593e-05,
+      "loss": 2.9404,
+      "step": 52748
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7343171834945679,
+      "learning_rate": 2.787334638432104e-05,
+      "loss": 3.0185,
+      "step": 52749
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7647814750671387,
+      "learning_rate": 2.7866865882311985e-05,
+      "loss": 3.0713,
+      "step": 52750
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7639553546905518,
+      "learning_rate": 2.7860386097049257e-05,
+      "loss": 2.974,
+      "step": 52751
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7180127501487732,
+      "learning_rate": 2.785390702854985e-05,
+      "loss": 2.7201,
+      "step": 52752
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7458502054214478,
+      "learning_rate": 2.7847428676830876e-05,
+      "loss": 2.8863,
+      "step": 52753
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7427824139595032,
+      "learning_rate": 2.7840951041909455e-05,
+      "loss": 2.755,
+      "step": 52754
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7325022220611572,
+      "learning_rate": 2.7834474123802574e-05,
+      "loss": 3.1051,
+      "step": 52755
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8776165246963501,
+      "learning_rate": 2.7827997922527324e-05,
+      "loss": 2.9282,
+      "step": 52756
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7489161491394043,
+      "learning_rate": 2.7821522438100685e-05,
+      "loss": 3.0858,
+      "step": 52757
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7696153521537781,
+      "learning_rate": 2.781504767053985e-05,
+      "loss": 2.8448,
+      "step": 52758
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7301738858222961,
+      "learning_rate": 2.780857361986173e-05,
+      "loss": 3.0967,
+      "step": 52759
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7664462924003601,
+      "learning_rate": 2.780210028608345e-05,
+      "loss": 2.7462,
+      "step": 52760
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7320300340652466,
+      "learning_rate": 2.77956276692221e-05,
+      "loss": 3.0839,
+      "step": 52761
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7702510356903076,
+      "learning_rate": 2.7789155769294657e-05,
+      "loss": 3.0114,
+      "step": 52762
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7481235861778259,
+      "learning_rate": 2.7782684586318216e-05,
+      "loss": 2.9768,
+      "step": 52763
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.779035747051239,
+      "learning_rate": 2.7776214120309726e-05,
+      "loss": 3.0398,
+      "step": 52764
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7233691215515137,
+      "learning_rate": 2.7769744371286308e-05,
+      "loss": 2.8899,
+      "step": 52765
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7301437854766846,
+      "learning_rate": 2.7763275339265044e-05,
+      "loss": 2.8502,
+      "step": 52766
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7371978759765625,
+      "learning_rate": 2.7756807024262863e-05,
+      "loss": 2.8388,
+      "step": 52767
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7594659328460693,
+      "learning_rate": 2.7750339426296908e-05,
+      "loss": 2.7252,
+      "step": 52768
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7644268870353699,
+      "learning_rate": 2.774387254538414e-05,
+      "loss": 2.8623,
+      "step": 52769
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7400708794593811,
+      "learning_rate": 2.7737406381541637e-05,
+      "loss": 2.8575,
+      "step": 52770
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7719500064849854,
+      "learning_rate": 2.7730940934786426e-05,
+      "loss": 2.852,
+      "step": 52771
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7356804609298706,
+      "learning_rate": 2.7724476205135425e-05,
+      "loss": 2.996,
+      "step": 52772
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7251750230789185,
+      "learning_rate": 2.7718012192605854e-05,
+      "loss": 2.897,
+      "step": 52773
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7100973129272461,
+      "learning_rate": 2.7711548897214562e-05,
+      "loss": 2.9647,
+      "step": 52774
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7529676556587219,
+      "learning_rate": 2.770508631897864e-05,
+      "loss": 3.0783,
+      "step": 52775
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7607330083847046,
+      "learning_rate": 2.7698624457915175e-05,
+      "loss": 2.8693,
+      "step": 52776
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7323965430259705,
+      "learning_rate": 2.769216331404115e-05,
+      "loss": 2.9267,
+      "step": 52777
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7149645686149597,
+      "learning_rate": 2.768570288737352e-05,
+      "loss": 2.8751,
+      "step": 52778
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8055224418640137,
+      "learning_rate": 2.7679243177929332e-05,
+      "loss": 2.9824,
+      "step": 52779
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7920107841491699,
+      "learning_rate": 2.7672784185725584e-05,
+      "loss": 3.1636,
+      "step": 52780
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6878994107246399,
+      "learning_rate": 2.766632591077935e-05,
+      "loss": 2.9685,
+      "step": 52781
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7483735084533691,
+      "learning_rate": 2.7659868353107563e-05,
+      "loss": 2.7888,
+      "step": 52782
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7039675116539001,
+      "learning_rate": 2.765341151272733e-05,
+      "loss": 2.8918,
+      "step": 52783
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.752712070941925,
+      "learning_rate": 2.7646955389655544e-05,
+      "loss": 3.069,
+      "step": 52784
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7546181082725525,
+      "learning_rate": 2.7640499983909326e-05,
+      "loss": 2.7278,
+      "step": 52785
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7798104286193848,
+      "learning_rate": 2.763404529550559e-05,
+      "loss": 2.8005,
+      "step": 52786
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6820088028907776,
+      "learning_rate": 2.7627591324461295e-05,
+      "loss": 2.9015,
+      "step": 52787
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7711160182952881,
+      "learning_rate": 2.762113807079359e-05,
+      "loss": 2.7493,
+      "step": 52788
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.77363121509552,
+      "learning_rate": 2.7614685534519332e-05,
+      "loss": 2.8504,
+      "step": 52789
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7659868597984314,
+      "learning_rate": 2.760823371565557e-05,
+      "loss": 2.6641,
+      "step": 52790
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7323107719421387,
+      "learning_rate": 2.7601782614219326e-05,
+      "loss": 2.7258,
+      "step": 52791
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7489632368087769,
+      "learning_rate": 2.7595332230227584e-05,
+      "loss": 3.0457,
+      "step": 52792
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7547075748443604,
+      "learning_rate": 2.75888825636973e-05,
+      "loss": 2.9536,
+      "step": 52793
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7597982883453369,
+      "learning_rate": 2.758243361464543e-05,
+      "loss": 2.7106,
+      "step": 52794
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7556941509246826,
+      "learning_rate": 2.7575985383089018e-05,
+      "loss": 2.908,
+      "step": 52795
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7499327659606934,
+      "learning_rate": 2.7569537869045055e-05,
+      "loss": 3.0735,
+      "step": 52796
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.780232310295105,
+      "learning_rate": 2.7563091072530465e-05,
+      "loss": 2.8993,
+      "step": 52797
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7690110206604004,
+      "learning_rate": 2.7556644993562295e-05,
+      "loss": 2.6908,
+      "step": 52798
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7913340926170349,
+      "learning_rate": 2.75501996321575e-05,
+      "loss": 2.8206,
+      "step": 52799
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7513294219970703,
+      "learning_rate": 2.7543754988333e-05,
+      "loss": 2.8581,
+      "step": 52800
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7517967820167542,
+      "learning_rate": 2.753731106210585e-05,
+      "loss": 2.8545,
+      "step": 52801
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7514926195144653,
+      "learning_rate": 2.7530867853492933e-05,
+      "loss": 3.0491,
+      "step": 52802
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7327825427055359,
+      "learning_rate": 2.7524425362511337e-05,
+      "loss": 2.9297,
+      "step": 52803
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7828764915466309,
+      "learning_rate": 2.7517983589177917e-05,
+      "loss": 2.8612,
+      "step": 52804
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7338088154792786,
+      "learning_rate": 2.751154253350969e-05,
+      "loss": 3.1315,
+      "step": 52805
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7811881303787231,
+      "learning_rate": 2.7505102195523675e-05,
+      "loss": 2.6917,
+      "step": 52806
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7544569969177246,
+      "learning_rate": 2.7498662575236762e-05,
+      "loss": 2.8917,
+      "step": 52807
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7520846128463745,
+      "learning_rate": 2.7492223672665937e-05,
+      "loss": 2.9407,
+      "step": 52808
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7595059871673584,
+      "learning_rate": 2.7485785487828083e-05,
+      "loss": 3.0334,
+      "step": 52809
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7427202463150024,
+      "learning_rate": 2.7479348020740223e-05,
+      "loss": 2.8506,
+      "step": 52810
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7459549903869629,
+      "learning_rate": 2.7472911271419374e-05,
+      "loss": 2.9055,
+      "step": 52811
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7440264225006104,
+      "learning_rate": 2.746647523988239e-05,
+      "loss": 2.9919,
+      "step": 52812
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7285496592521667,
+      "learning_rate": 2.7460039926146293e-05,
+      "loss": 3.0071,
+      "step": 52813
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7452382445335388,
+      "learning_rate": 2.7453605330228e-05,
+      "loss": 2.8223,
+      "step": 52814
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8756013512611389,
+      "learning_rate": 2.74471714521444e-05,
+      "loss": 2.8339,
+      "step": 52815
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8711315393447876,
+      "learning_rate": 2.744073829191258e-05,
+      "loss": 3.0862,
+      "step": 52816
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7606200575828552,
+      "learning_rate": 2.743430584954932e-05,
+      "loss": 2.7368,
+      "step": 52817
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7606265544891357,
+      "learning_rate": 2.7427874125071714e-05,
+      "loss": 2.8173,
+      "step": 52818
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7671118378639221,
+      "learning_rate": 2.7421443118496544e-05,
+      "loss": 3.0363,
+      "step": 52819
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7543085813522339,
+      "learning_rate": 2.7415012829840933e-05,
+      "loss": 2.8662,
+      "step": 52820
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7956048846244812,
+      "learning_rate": 2.7408583259121697e-05,
+      "loss": 2.999,
+      "step": 52821
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7940734028816223,
+      "learning_rate": 2.7402154406355726e-05,
+      "loss": 2.9409,
+      "step": 52822
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7047262191772461,
+      "learning_rate": 2.739572627156007e-05,
+      "loss": 2.9213,
+      "step": 52823
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7310606241226196,
+      "learning_rate": 2.7389298854751586e-05,
+      "loss": 2.7791,
+      "step": 52824
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7745860815048218,
+      "learning_rate": 2.738287215594719e-05,
+      "loss": 3.0216,
+      "step": 52825
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7691651582717896,
+      "learning_rate": 2.7376446175163903e-05,
+      "loss": 2.6237,
+      "step": 52826
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7258844971656799,
+      "learning_rate": 2.7370020912418543e-05,
+      "loss": 2.9344,
+      "step": 52827
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7222638130187988,
+      "learning_rate": 2.7363596367728135e-05,
+      "loss": 2.5812,
+      "step": 52828
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7643725872039795,
+      "learning_rate": 2.735717254110953e-05,
+      "loss": 3.0799,
+      "step": 52829
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7751563191413879,
+      "learning_rate": 2.735074943257961e-05,
+      "loss": 2.8096,
+      "step": 52830
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7652669548988342,
+      "learning_rate": 2.73443270421554e-05,
+      "loss": 2.9164,
+      "step": 52831
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7439308166503906,
+      "learning_rate": 2.733790536985372e-05,
+      "loss": 2.983,
+      "step": 52832
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7255779504776001,
+      "learning_rate": 2.7331484415691584e-05,
+      "loss": 2.8595,
+      "step": 52833
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7431380152702332,
+      "learning_rate": 2.7325064179685753e-05,
+      "loss": 2.9898,
+      "step": 52834
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7448784708976746,
+      "learning_rate": 2.7318644661853307e-05,
+      "loss": 2.9715,
+      "step": 52835
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7783845067024231,
+      "learning_rate": 2.7312225862211067e-05,
+      "loss": 2.908,
+      "step": 52836
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7135199308395386,
+      "learning_rate": 2.7305807780775856e-05,
+      "loss": 2.8114,
+      "step": 52837
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7589677572250366,
+      "learning_rate": 2.729939041756476e-05,
+      "loss": 2.9255,
+      "step": 52838
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7436524033546448,
+      "learning_rate": 2.7292973772594527e-05,
+      "loss": 2.9792,
+      "step": 52839
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.786425769329071,
+      "learning_rate": 2.7286557845882185e-05,
+      "loss": 2.6977,
+      "step": 52840
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7813881039619446,
+      "learning_rate": 2.728014263744448e-05,
+      "loss": 3.0427,
+      "step": 52841
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7063042521476746,
+      "learning_rate": 2.7273728147298466e-05,
+      "loss": 2.9036,
+      "step": 52842
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7354524731636047,
+      "learning_rate": 2.7267314375460968e-05,
+      "loss": 2.9452,
+      "step": 52843
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7480248212814331,
+      "learning_rate": 2.72609013219488e-05,
+      "loss": 2.9096,
+      "step": 52844
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7572676539421082,
+      "learning_rate": 2.725448898677899e-05,
+      "loss": 2.8905,
+      "step": 52845
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7566613554954529,
+      "learning_rate": 2.7248077369968312e-05,
+      "loss": 2.8314,
+      "step": 52846
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.733538806438446,
+      "learning_rate": 2.724166647153373e-05,
+      "loss": 2.9796,
+      "step": 52847
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7428073287010193,
+      "learning_rate": 2.7235256291492124e-05,
+      "loss": 2.7486,
+      "step": 52848
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7868484258651733,
+      "learning_rate": 2.7228846829860318e-05,
+      "loss": 2.7314,
+      "step": 52849
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7082564830780029,
+      "learning_rate": 2.7222438086655295e-05,
+      "loss": 3.042,
+      "step": 52850
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7359206676483154,
+      "learning_rate": 2.721603006189388e-05,
+      "loss": 3.0088,
+      "step": 52851
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7429801821708679,
+      "learning_rate": 2.7209622755592852e-05,
+      "loss": 2.9179,
+      "step": 52852
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7138932347297668,
+      "learning_rate": 2.720321616776927e-05,
+      "loss": 2.8191,
+      "step": 52853
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7627942562103271,
+      "learning_rate": 2.7196810298439852e-05,
+      "loss": 2.9606,
+      "step": 52854
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7788562178611755,
+      "learning_rate": 2.7190405147621587e-05,
+      "loss": 2.7794,
+      "step": 52855
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7393107414245605,
+      "learning_rate": 2.7184000715331222e-05,
+      "loss": 2.879,
+      "step": 52856
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7325177192687988,
+      "learning_rate": 2.7177597001585782e-05,
+      "loss": 2.9611,
+      "step": 52857
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7476690411567688,
+      "learning_rate": 2.717119400640202e-05,
+      "loss": 2.9348,
+      "step": 52858
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7651125192642212,
+      "learning_rate": 2.7164791729796753e-05,
+      "loss": 2.7011,
+      "step": 52859
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7937335968017578,
+      "learning_rate": 2.7158390171787003e-05,
+      "loss": 3.0789,
+      "step": 52860
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7635074257850647,
+      "learning_rate": 2.715198933238949e-05,
+      "loss": 2.9107,
+      "step": 52861
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7641463875770569,
+      "learning_rate": 2.7145589211621132e-05,
+      "loss": 2.8345,
+      "step": 52862
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8104754090309143,
+      "learning_rate": 2.713918980949882e-05,
+      "loss": 2.9759,
+      "step": 52863
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7333548069000244,
+      "learning_rate": 2.7132791126039367e-05,
+      "loss": 3.0388,
+      "step": 52864
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.752964437007904,
+      "learning_rate": 2.712639316125963e-05,
+      "loss": 2.9652,
+      "step": 52865
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.73786860704422,
+      "learning_rate": 2.7119995915176396e-05,
+      "loss": 3.0213,
+      "step": 52866
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7262935638427734,
+      "learning_rate": 2.7113599387806582e-05,
+      "loss": 2.8607,
+      "step": 52867
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7377870678901672,
+      "learning_rate": 2.710720357916708e-05,
+      "loss": 2.9051,
+      "step": 52868
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7554672956466675,
+      "learning_rate": 2.7100808489274638e-05,
+      "loss": 2.7477,
+      "step": 52869
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7196999192237854,
+      "learning_rate": 2.7094414118146213e-05,
+      "loss": 2.8733,
+      "step": 52870
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7686446309089661,
+      "learning_rate": 2.708802046579849e-05,
+      "loss": 3.0189,
+      "step": 52871
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7374565601348877,
+      "learning_rate": 2.7081627532248486e-05,
+      "loss": 3.0224,
+      "step": 52872
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.727088212966919,
+      "learning_rate": 2.7075235317512923e-05,
+      "loss": 2.7337,
+      "step": 52873
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7013735175132751,
+      "learning_rate": 2.706884382160862e-05,
+      "loss": 2.9789,
+      "step": 52874
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7275285124778748,
+      "learning_rate": 2.7062453044552536e-05,
+      "loss": 2.843,
+      "step": 52875
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7339413166046143,
+      "learning_rate": 2.705606298636135e-05,
+      "loss": 2.786,
+      "step": 52876
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8053582906723022,
+      "learning_rate": 2.704967364705195e-05,
+      "loss": 3.011,
+      "step": 52877
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.770816445350647,
+      "learning_rate": 2.7043285026641258e-05,
+      "loss": 3.0154,
+      "step": 52878
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7276818752288818,
+      "learning_rate": 2.7036897125145995e-05,
+      "loss": 3.0039,
+      "step": 52879
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7705271244049072,
+      "learning_rate": 2.7030509942583013e-05,
+      "loss": 2.8089,
+      "step": 52880
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.764750063419342,
+      "learning_rate": 2.7024123478969095e-05,
+      "loss": 2.9019,
+      "step": 52881
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8144281506538391,
+      "learning_rate": 2.70177377343211e-05,
+      "loss": 2.7056,
+      "step": 52882
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7661029696464539,
+      "learning_rate": 2.7011352708655908e-05,
+      "loss": 2.907,
+      "step": 52883
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7692121267318726,
+      "learning_rate": 2.700496840199021e-05,
+      "loss": 2.8668,
+      "step": 52884
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7304009199142456,
+      "learning_rate": 2.699858481434096e-05,
+      "loss": 2.9795,
+      "step": 52885
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7058482766151428,
+      "learning_rate": 2.6992201945724878e-05,
+      "loss": 2.8171,
+      "step": 52886
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8239343762397766,
+      "learning_rate": 2.6985819796158746e-05,
+      "loss": 2.7458,
+      "step": 52887
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.76352459192276,
+      "learning_rate": 2.6979438365659456e-05,
+      "loss": 2.9991,
+      "step": 52888
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7338518500328064,
+      "learning_rate": 2.6973057654243756e-05,
+      "loss": 2.8587,
+      "step": 52889
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7867611050605774,
+      "learning_rate": 2.6966677661928504e-05,
+      "loss": 2.6982,
+      "step": 52890
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7336326837539673,
+      "learning_rate": 2.696029838873045e-05,
+      "loss": 2.8797,
+      "step": 52891
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7487276196479797,
+      "learning_rate": 2.6953919834666417e-05,
+      "loss": 2.8777,
+      "step": 52892
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7201322913169861,
+      "learning_rate": 2.6947541999753253e-05,
+      "loss": 2.806,
+      "step": 52893
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.731619119644165,
+      "learning_rate": 2.6941164884007715e-05,
+      "loss": 2.782,
+      "step": 52894
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7519465684890747,
+      "learning_rate": 2.6934788487446623e-05,
+      "loss": 2.7115,
+      "step": 52895
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7223137021064758,
+      "learning_rate": 2.6928412810086665e-05,
+      "loss": 2.8679,
+      "step": 52896
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7320907711982727,
+      "learning_rate": 2.692203785194472e-05,
+      "loss": 3.1397,
+      "step": 52897
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7210008502006531,
+      "learning_rate": 2.6915663613037618e-05,
+      "loss": 2.9665,
+      "step": 52898
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6909995675086975,
+      "learning_rate": 2.690929009338204e-05,
+      "loss": 2.9763,
+      "step": 52899
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7912375926971436,
+      "learning_rate": 2.6902917292994906e-05,
+      "loss": 2.9634,
+      "step": 52900
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7833209037780762,
+      "learning_rate": 2.6896545211892938e-05,
+      "loss": 2.7593,
+      "step": 52901
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7575777769088745,
+      "learning_rate": 2.6890173850092855e-05,
+      "loss": 2.8503,
+      "step": 52902
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7618677616119385,
+      "learning_rate": 2.6883803207611543e-05,
+      "loss": 3.023,
+      "step": 52903
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7287328839302063,
+      "learning_rate": 2.6877433284465656e-05,
+      "loss": 2.6533,
+      "step": 52904
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7309133410453796,
+      "learning_rate": 2.687106408067211e-05,
+      "loss": 2.9655,
+      "step": 52905
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7655085921287537,
+      "learning_rate": 2.6864695596247566e-05,
+      "loss": 2.9034,
+      "step": 52906
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7788105010986328,
+      "learning_rate": 2.685832783120887e-05,
+      "loss": 2.8593,
+      "step": 52907
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7491944432258606,
+      "learning_rate": 2.6851960785572813e-05,
+      "loss": 2.7122,
+      "step": 52908
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7422747015953064,
+      "learning_rate": 2.684559445935611e-05,
+      "loss": 2.8433,
+      "step": 52909
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7405474781990051,
+      "learning_rate": 2.6839228852575523e-05,
+      "loss": 2.984,
+      "step": 52910
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.719196617603302,
+      "learning_rate": 2.683286396524783e-05,
+      "loss": 2.7548,
+      "step": 52911
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7382709980010986,
+      "learning_rate": 2.6826499797389822e-05,
+      "loss": 2.9205,
+      "step": 52912
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7906863689422607,
+      "learning_rate": 2.6820136349018184e-05,
+      "loss": 2.7707,
+      "step": 52913
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.753178596496582,
+      "learning_rate": 2.681377362014977e-05,
+      "loss": 2.8353,
+      "step": 52914
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7814869284629822,
+      "learning_rate": 2.68074116108013e-05,
+      "loss": 2.7692,
+      "step": 52915
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7554951310157776,
+      "learning_rate": 2.680105032098956e-05,
+      "loss": 2.6514,
+      "step": 52916
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7147135138511658,
+      "learning_rate": 2.67946897507313e-05,
+      "loss": 3.0395,
+      "step": 52917
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7095446586608887,
+      "learning_rate": 2.6788329900043147e-05,
+      "loss": 3.0561,
+      "step": 52918
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7295847535133362,
+      "learning_rate": 2.6781970768941985e-05,
+      "loss": 3.0686,
+      "step": 52919
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7395632863044739,
+      "learning_rate": 2.6775612357444564e-05,
+      "loss": 3.1226,
+      "step": 52920
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7481321096420288,
+      "learning_rate": 2.6769254665567575e-05,
+      "loss": 2.9356,
+      "step": 52921
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7256965637207031,
+      "learning_rate": 2.67628976933278e-05,
+      "loss": 2.8715,
+      "step": 52922
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7288814783096313,
+      "learning_rate": 2.6756541440741996e-05,
+      "loss": 3.017,
+      "step": 52923
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7406113743782043,
+      "learning_rate": 2.675018590782678e-05,
+      "loss": 3.1521,
+      "step": 52924
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7255039811134338,
+      "learning_rate": 2.6743831094599077e-05,
+      "loss": 2.865,
+      "step": 52925
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7836463451385498,
+      "learning_rate": 2.673747700107547e-05,
+      "loss": 2.8227,
+      "step": 52926
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7792579531669617,
+      "learning_rate": 2.673112362727281e-05,
+      "loss": 2.8882,
+      "step": 52927
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7477447986602783,
+      "learning_rate": 2.6724770973207754e-05,
+      "loss": 3.0428,
+      "step": 52928
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7685944437980652,
+      "learning_rate": 2.671841903889702e-05,
+      "loss": 2.9894,
+      "step": 52929
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7276458740234375,
+      "learning_rate": 2.671206782435743e-05,
+      "loss": 2.9397,
+      "step": 52930
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.766247034072876,
+      "learning_rate": 2.670571732960567e-05,
+      "loss": 3.0083,
+      "step": 52931
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.759260356426239,
+      "learning_rate": 2.669936755465846e-05,
+      "loss": 2.7322,
+      "step": 52932
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7431995272636414,
+      "learning_rate": 2.6693018499532483e-05,
+      "loss": 2.9465,
+      "step": 52933
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.744221568107605,
+      "learning_rate": 2.668667016424446e-05,
+      "loss": 3.0348,
+      "step": 52934
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.8303017616271973,
+      "learning_rate": 2.6680322548811216e-05,
+      "loss": 2.8454,
+      "step": 52935
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7433030605316162,
+      "learning_rate": 2.6673975653249336e-05,
+      "loss": 2.8534,
+      "step": 52936
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7234331965446472,
+      "learning_rate": 2.666762947757567e-05,
+      "loss": 2.7294,
+      "step": 52937
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7342804074287415,
+      "learning_rate": 2.666128402180684e-05,
+      "loss": 3.0371,
+      "step": 52938
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7552002668380737,
+      "learning_rate": 2.6654939285959564e-05,
+      "loss": 3.075,
+      "step": 52939
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6919820308685303,
+      "learning_rate": 2.6648595270050598e-05,
+      "loss": 2.9163,
+      "step": 52940
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7407262921333313,
+      "learning_rate": 2.6642251974096597e-05,
+      "loss": 2.9234,
+      "step": 52941
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7760119438171387,
+      "learning_rate": 2.663590939811434e-05,
+      "loss": 3.0161,
+      "step": 52942
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7739540338516235,
+      "learning_rate": 2.6629567542120422e-05,
+      "loss": 2.8185,
+      "step": 52943
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7417166829109192,
+      "learning_rate": 2.662322640613166e-05,
+      "loss": 2.893,
+      "step": 52944
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7252646088600159,
+      "learning_rate": 2.661688599016474e-05,
+      "loss": 2.8558,
+      "step": 52945
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7560940980911255,
+      "learning_rate": 2.6610546294236245e-05,
+      "loss": 2.9693,
+      "step": 52946
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7531810402870178,
+      "learning_rate": 2.6604207318363003e-05,
+      "loss": 3.0296,
+      "step": 52947
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7631690502166748,
+      "learning_rate": 2.6597869062561627e-05,
+      "loss": 3.013,
+      "step": 52948
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7429794669151306,
+      "learning_rate": 2.659153152684884e-05,
+      "loss": 2.7711,
+      "step": 52949
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7256251573562622,
+      "learning_rate": 2.6585194711241397e-05,
+      "loss": 3.0047,
+      "step": 52950
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7361646890640259,
+      "learning_rate": 2.6578858615755882e-05,
+      "loss": 3.0397,
+      "step": 52951
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7536941766738892,
+      "learning_rate": 2.6572523240409082e-05,
+      "loss": 2.8646,
+      "step": 52952
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7746854424476624,
+      "learning_rate": 2.6566188585217652e-05,
+      "loss": 2.9317,
+      "step": 52953
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7454334497451782,
+      "learning_rate": 2.6559854650198174e-05,
+      "loss": 2.8275,
+      "step": 52954
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7651907205581665,
+      "learning_rate": 2.6553521435367474e-05,
+      "loss": 2.7026,
+      "step": 52955
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7534736394882202,
+      "learning_rate": 2.6547188940742136e-05,
+      "loss": 2.9386,
+      "step": 52956
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7591540217399597,
+      "learning_rate": 2.6540857166338915e-05,
+      "loss": 2.9647,
+      "step": 52957
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8195220232009888,
+      "learning_rate": 2.6534526112174392e-05,
+      "loss": 2.8516,
+      "step": 52958
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8042005300521851,
+      "learning_rate": 2.652819577826536e-05,
+      "loss": 2.8863,
+      "step": 52959
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7308369278907776,
+      "learning_rate": 2.6521866164628437e-05,
+      "loss": 3.1317,
+      "step": 52960
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7242509126663208,
+      "learning_rate": 2.6515537271280206e-05,
+      "loss": 2.8885,
+      "step": 52961
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7605120539665222,
+      "learning_rate": 2.6509209098237494e-05,
+      "loss": 3.0218,
+      "step": 52962
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.729441225528717,
+      "learning_rate": 2.6502881645516848e-05,
+      "loss": 2.9623,
+      "step": 52963
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7368941903114319,
+      "learning_rate": 2.649655491313496e-05,
+      "loss": 3.0673,
+      "step": 52964
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7427717447280884,
+      "learning_rate": 2.6490228901108546e-05,
+      "loss": 2.8994,
+      "step": 52965
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7409499883651733,
+      "learning_rate": 2.648390360945426e-05,
+      "loss": 2.772,
+      "step": 52966
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7534723877906799,
+      "learning_rate": 2.6477579038188722e-05,
+      "loss": 3.1012,
+      "step": 52967
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7700952887535095,
+      "learning_rate": 2.647125518732852e-05,
+      "loss": 2.9597,
+      "step": 52968
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.718622088432312,
+      "learning_rate": 2.646493205689044e-05,
+      "loss": 2.9911,
+      "step": 52969
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7912169694900513,
+      "learning_rate": 2.6458609646891106e-05,
+      "loss": 3.0782,
+      "step": 52970
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.766054630279541,
+      "learning_rate": 2.6452287957347095e-05,
+      "loss": 3.1029,
+      "step": 52971
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.816627025604248,
+      "learning_rate": 2.644596698827517e-05,
+      "loss": 2.7479,
+      "step": 52972
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.779098391532898,
+      "learning_rate": 2.6439646739691876e-05,
+      "loss": 2.7438,
+      "step": 52973
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7731761932373047,
+      "learning_rate": 2.6433327211613974e-05,
+      "loss": 2.9227,
+      "step": 52974
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7547659277915955,
+      "learning_rate": 2.6427008404058014e-05,
+      "loss": 2.8468,
+      "step": 52975
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7428092956542969,
+      "learning_rate": 2.6420690317040616e-05,
+      "loss": 2.9101,
+      "step": 52976
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7234500050544739,
+      "learning_rate": 2.6414372950578533e-05,
+      "loss": 2.7825,
+      "step": 52977
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7860729098320007,
+      "learning_rate": 2.6408056304688253e-05,
+      "loss": 2.832,
+      "step": 52978
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7206407189369202,
+      "learning_rate": 2.6401740379386526e-05,
+      "loss": 2.6774,
+      "step": 52979
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.715305507183075,
+      "learning_rate": 2.639542517469001e-05,
+      "loss": 2.9311,
+      "step": 52980
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7753638625144958,
+      "learning_rate": 2.6389110690615288e-05,
+      "loss": 2.929,
+      "step": 52981
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7320675253868103,
+      "learning_rate": 2.6382796927178984e-05,
+      "loss": 2.7352,
+      "step": 52982
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7664274573326111,
+      "learning_rate": 2.6376483884397682e-05,
+      "loss": 2.772,
+      "step": 52983
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.723810076713562,
+      "learning_rate": 2.6370171562288068e-05,
+      "loss": 2.9876,
+      "step": 52984
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7204124927520752,
+      "learning_rate": 2.6363859960866828e-05,
+      "loss": 2.9197,
+      "step": 52985
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.748243510723114,
+      "learning_rate": 2.6357549080150452e-05,
+      "loss": 2.8726,
+      "step": 52986
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7590616941452026,
+      "learning_rate": 2.635123892015566e-05,
+      "loss": 2.8701,
+      "step": 52987
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7760512828826904,
+      "learning_rate": 2.6344929480899068e-05,
+      "loss": 2.7923,
+      "step": 52988
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7821771502494812,
+      "learning_rate": 2.6338620762397266e-05,
+      "loss": 2.9296,
+      "step": 52989
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7788903117179871,
+      "learning_rate": 2.6332312764666808e-05,
+      "loss": 2.7867,
+      "step": 52990
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7191170454025269,
+      "learning_rate": 2.6326005487724345e-05,
+      "loss": 3.0973,
+      "step": 52991
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7111521363258362,
+      "learning_rate": 2.6319698931586598e-05,
+      "loss": 2.9587,
+      "step": 52992
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.742542028427124,
+      "learning_rate": 2.6313393096270053e-05,
+      "loss": 2.8454,
+      "step": 52993
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7643880844116211,
+      "learning_rate": 2.6307087981791365e-05,
+      "loss": 2.9678,
+      "step": 52994
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7396053075790405,
+      "learning_rate": 2.630078358816712e-05,
+      "loss": 2.7297,
+      "step": 52995
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7279162406921387,
+      "learning_rate": 2.6294479915413968e-05,
+      "loss": 2.7479,
+      "step": 52996
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7429473996162415,
+      "learning_rate": 2.62881769635485e-05,
+      "loss": 2.9331,
+      "step": 52997
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7297216057777405,
+      "learning_rate": 2.6281874732587238e-05,
+      "loss": 2.9241,
+      "step": 52998
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7150318026542664,
+      "learning_rate": 2.6275573222546898e-05,
+      "loss": 2.9074,
+      "step": 52999
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7341070175170898,
+      "learning_rate": 2.6269272433443933e-05,
+      "loss": 2.8459,
+      "step": 53000
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.741511881351471,
+      "learning_rate": 2.6262972365295067e-05,
+      "loss": 3.1279,
+      "step": 53001
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7074028849601746,
+      "learning_rate": 2.625667301811688e-05,
+      "loss": 2.8032,
+      "step": 53002
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7410276532173157,
+      "learning_rate": 2.6250374391925933e-05,
+      "loss": 2.6586,
+      "step": 53003
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7217987775802612,
+      "learning_rate": 2.624407648673881e-05,
+      "loss": 2.6289,
+      "step": 53004
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7761947512626648,
+      "learning_rate": 2.623777930257206e-05,
+      "loss": 2.86,
+      "step": 53005
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7420976161956787,
+      "learning_rate": 2.6231482839442307e-05,
+      "loss": 3.0115,
+      "step": 53006
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7364739179611206,
+      "learning_rate": 2.6225187097366206e-05,
+      "loss": 2.8002,
+      "step": 53007
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7955524325370789,
+      "learning_rate": 2.621889207636021e-05,
+      "loss": 2.9859,
+      "step": 53008
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7260406613349915,
+      "learning_rate": 2.6212597776441034e-05,
+      "loss": 2.896,
+      "step": 53009
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7395499348640442,
+      "learning_rate": 2.6206304197625107e-05,
+      "loss": 2.9315,
+      "step": 53010
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7385746836662292,
+      "learning_rate": 2.6200011339929138e-05,
+      "loss": 2.8443,
+      "step": 53011
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7338225245475769,
+      "learning_rate": 2.6193719203369658e-05,
+      "loss": 2.9998,
+      "step": 53012
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7368943691253662,
+      "learning_rate": 2.6187427787963143e-05,
+      "loss": 2.812,
+      "step": 53013
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7941758036613464,
+      "learning_rate": 2.6181137093726358e-05,
+      "loss": 2.964,
+      "step": 53014
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8088509440422058,
+      "learning_rate": 2.6174847120675646e-05,
+      "loss": 3.0882,
+      "step": 53015
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7890652418136597,
+      "learning_rate": 2.6168557868827733e-05,
+      "loss": 2.8184,
+      "step": 53016
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7560929656028748,
+      "learning_rate": 2.6162269338199172e-05,
+      "loss": 2.9436,
+      "step": 53017
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7155969738960266,
+      "learning_rate": 2.615598152880648e-05,
+      "loss": 2.9308,
+      "step": 53018
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7594807147979736,
+      "learning_rate": 2.6149694440666246e-05,
+      "loss": 2.8756,
+      "step": 53019
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7847410440444946,
+      "learning_rate": 2.6143408073794957e-05,
+      "loss": 3.0293,
+      "step": 53020
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8248916864395142,
+      "learning_rate": 2.6137122428209234e-05,
+      "loss": 2.7244,
+      "step": 53021
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7598782777786255,
+      "learning_rate": 2.6130837503925662e-05,
+      "loss": 2.9428,
+      "step": 53022
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8169916868209839,
+      "learning_rate": 2.6124553300960725e-05,
+      "loss": 2.9181,
+      "step": 53023
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7497507333755493,
+      "learning_rate": 2.611826981933105e-05,
+      "loss": 2.7606,
+      "step": 53024
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7234208583831787,
+      "learning_rate": 2.611198705905312e-05,
+      "loss": 2.9675,
+      "step": 53025
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6970089077949524,
+      "learning_rate": 2.6105705020143487e-05,
+      "loss": 2.9544,
+      "step": 53026
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.702825129032135,
+      "learning_rate": 2.6099423702618738e-05,
+      "loss": 2.8923,
+      "step": 53027
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7164490222930908,
+      "learning_rate": 2.6093143106495362e-05,
+      "loss": 3.0565,
+      "step": 53028
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8021504878997803,
+      "learning_rate": 2.6086863231789978e-05,
+      "loss": 2.9865,
+      "step": 53029
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7775247097015381,
+      "learning_rate": 2.608058407851904e-05,
+      "loss": 2.9305,
+      "step": 53030
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.903372049331665,
+      "learning_rate": 2.60743056466991e-05,
+      "loss": 2.945,
+      "step": 53031
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7477425932884216,
+      "learning_rate": 2.6068027936346814e-05,
+      "loss": 2.6972,
+      "step": 53032
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7447633147239685,
+      "learning_rate": 2.6061750947478598e-05,
+      "loss": 3.0102,
+      "step": 53033
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7586161494255066,
+      "learning_rate": 2.605547468011101e-05,
+      "loss": 2.9697,
+      "step": 53034
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7298742532730103,
+      "learning_rate": 2.6049199134260535e-05,
+      "loss": 2.9407,
+      "step": 53035
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7363532185554504,
+      "learning_rate": 2.6042924309943724e-05,
+      "loss": 3.0353,
+      "step": 53036
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7362156510353088,
+      "learning_rate": 2.6036650207177202e-05,
+      "loss": 2.9521,
+      "step": 53037
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7948529720306396,
+      "learning_rate": 2.603037682597735e-05,
+      "loss": 2.9741,
+      "step": 53038
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7342039942741394,
+      "learning_rate": 2.6024104166360826e-05,
+      "loss": 2.9473,
+      "step": 53039
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7215148210525513,
+      "learning_rate": 2.601783222834408e-05,
+      "loss": 3.0339,
+      "step": 53040
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7647704482078552,
+      "learning_rate": 2.601156101194357e-05,
+      "loss": 2.7878,
+      "step": 53041
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7760405540466309,
+      "learning_rate": 2.6005290517175948e-05,
+      "loss": 2.7535,
+      "step": 53042
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7145057916641235,
+      "learning_rate": 2.5999020744057598e-05,
+      "loss": 2.9127,
+      "step": 53043
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7269275784492493,
+      "learning_rate": 2.599275169260514e-05,
+      "loss": 2.6882,
+      "step": 53044
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7115113735198975,
+      "learning_rate": 2.5986483362835e-05,
+      "loss": 2.9195,
+      "step": 53045
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7383373379707336,
+      "learning_rate": 2.5980215754763788e-05,
+      "loss": 2.9078,
+      "step": 53046
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6950742602348328,
+      "learning_rate": 2.5973948868407936e-05,
+      "loss": 2.8095,
+      "step": 53047
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7306373715400696,
+      "learning_rate": 2.596768270378392e-05,
+      "loss": 2.9741,
+      "step": 53048
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7336162328720093,
+      "learning_rate": 2.5961417260908335e-05,
+      "loss": 2.9961,
+      "step": 53049
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7189518213272095,
+      "learning_rate": 2.5955152539797598e-05,
+      "loss": 3.0896,
+      "step": 53050
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8078373670578003,
+      "learning_rate": 2.5948888540468226e-05,
+      "loss": 2.5814,
+      "step": 53051
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7885643839836121,
+      "learning_rate": 2.594262526293681e-05,
+      "loss": 3.1085,
+      "step": 53052
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7464584112167358,
+      "learning_rate": 2.5936362707219704e-05,
+      "loss": 2.8979,
+      "step": 53053
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7252410650253296,
+      "learning_rate": 2.593010087333356e-05,
+      "loss": 2.8299,
+      "step": 53054
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7533923387527466,
+      "learning_rate": 2.5923839761294762e-05,
+      "loss": 2.694,
+      "step": 53055
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.700678825378418,
+      "learning_rate": 2.5917579371119768e-05,
+      "loss": 2.9615,
+      "step": 53056
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7296587824821472,
+      "learning_rate": 2.591131970282516e-05,
+      "loss": 2.9133,
+      "step": 53057
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7433834671974182,
+      "learning_rate": 2.5905060756427365e-05,
+      "loss": 2.9297,
+      "step": 53058
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7818628549575806,
+      "learning_rate": 2.589880253194293e-05,
+      "loss": 2.846,
+      "step": 53059
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.775662362575531,
+      "learning_rate": 2.5892545029388245e-05,
+      "loss": 2.7622,
+      "step": 53060
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7537034153938293,
+      "learning_rate": 2.5886288248779896e-05,
+      "loss": 3.0378,
+      "step": 53061
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7831193804740906,
+      "learning_rate": 2.5880032190134303e-05,
+      "loss": 2.782,
+      "step": 53062
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7730798125267029,
+      "learning_rate": 2.587377685346792e-05,
+      "loss": 2.9905,
+      "step": 53063
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7518443465232849,
+      "learning_rate": 2.5867522238797266e-05,
+      "loss": 2.8506,
+      "step": 53064
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7517948150634766,
+      "learning_rate": 2.5861268346138796e-05,
+      "loss": 2.8146,
+      "step": 53065
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7461619973182678,
+      "learning_rate": 2.5855015175508998e-05,
+      "loss": 2.994,
+      "step": 53066
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7383305430412292,
+      "learning_rate": 2.5848762726924287e-05,
+      "loss": 2.8297,
+      "step": 53067
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7097854614257812,
+      "learning_rate": 2.584251100040122e-05,
+      "loss": 2.6958,
+      "step": 53068
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7683972120285034,
+      "learning_rate": 2.5836259995956254e-05,
+      "loss": 2.7007,
+      "step": 53069
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7571167945861816,
+      "learning_rate": 2.583000971360574e-05,
+      "loss": 2.9229,
+      "step": 53070
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7330532670021057,
+      "learning_rate": 2.5823760153366258e-05,
+      "loss": 2.9509,
+      "step": 53071
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7204066514968872,
+      "learning_rate": 2.581751131525417e-05,
+      "loss": 2.9927,
+      "step": 53072
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7530893683433533,
+      "learning_rate": 2.5811263199285992e-05,
+      "loss": 2.9785,
+      "step": 53073
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7310274839401245,
+      "learning_rate": 2.5805015805478245e-05,
+      "loss": 2.8125,
+      "step": 53074
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7056111097335815,
+      "learning_rate": 2.579876913384725e-05,
+      "loss": 2.8229,
+      "step": 53075
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7395758628845215,
+      "learning_rate": 2.5792523184409597e-05,
+      "loss": 3.0898,
+      "step": 53076
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7391893267631531,
+      "learning_rate": 2.578627795718167e-05,
+      "loss": 2.7922,
+      "step": 53077
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.719342052936554,
+      "learning_rate": 2.5780033452179815e-05,
+      "loss": 2.8224,
+      "step": 53078
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7190227508544922,
+      "learning_rate": 2.5773789669420662e-05,
+      "loss": 2.9418,
+      "step": 53079
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7570742964744568,
+      "learning_rate": 2.576754660892053e-05,
+      "loss": 2.5958,
+      "step": 53080
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7037019729614258,
+      "learning_rate": 2.5761304270695936e-05,
+      "loss": 2.9506,
+      "step": 53081
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7258500456809998,
+      "learning_rate": 2.5755062654763237e-05,
+      "loss": 2.6689,
+      "step": 53082
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7900306582450867,
+      "learning_rate": 2.5748821761138982e-05,
+      "loss": 3.0713,
+      "step": 53083
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7837575078010559,
+      "learning_rate": 2.5742581589839563e-05,
+      "loss": 2.656,
+      "step": 53084
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.715522050857544,
+      "learning_rate": 2.5736342140881326e-05,
+      "loss": 2.9993,
+      "step": 53085
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7533067464828491,
+      "learning_rate": 2.5730103414280834e-05,
+      "loss": 3.0117,
+      "step": 53086
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7553693652153015,
+      "learning_rate": 2.5723865410054434e-05,
+      "loss": 2.9827,
+      "step": 53087
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7062312364578247,
+      "learning_rate": 2.5717628128218548e-05,
+      "loss": 3.022,
+      "step": 53088
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7795845866203308,
+      "learning_rate": 2.571139156878973e-05,
+      "loss": 2.8414,
+      "step": 53089
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7023152709007263,
+      "learning_rate": 2.5705155731784333e-05,
+      "loss": 2.8888,
+      "step": 53090
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7439648509025574,
+      "learning_rate": 2.5698920617218743e-05,
+      "loss": 2.8256,
+      "step": 53091
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7769492268562317,
+      "learning_rate": 2.569268622510935e-05,
+      "loss": 2.7845,
+      "step": 53092
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7483662962913513,
+      "learning_rate": 2.5686452555472637e-05,
+      "loss": 2.8814,
+      "step": 53093
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7647069096565247,
+      "learning_rate": 2.568021960832506e-05,
+      "loss": 2.6949,
+      "step": 53094
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7527300715446472,
+      "learning_rate": 2.5673987383682936e-05,
+      "loss": 3.0339,
+      "step": 53095
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7335328459739685,
+      "learning_rate": 2.566775588156279e-05,
+      "loss": 2.869,
+      "step": 53096
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7456409335136414,
+      "learning_rate": 2.566152510198094e-05,
+      "loss": 2.9675,
+      "step": 53097
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7429218888282776,
+      "learning_rate": 2.5655295044953873e-05,
+      "loss": 2.9004,
+      "step": 53098
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7677924036979675,
+      "learning_rate": 2.5649065710497974e-05,
+      "loss": 2.9512,
+      "step": 53099
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7493651509284973,
+      "learning_rate": 2.5642837098629566e-05,
+      "loss": 2.9513,
+      "step": 53100
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7118934392929077,
+      "learning_rate": 2.5636609209365168e-05,
+      "loss": 2.6457,
+      "step": 53101
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8193805813789368,
+      "learning_rate": 2.5630382042721098e-05,
+      "loss": 2.668,
+      "step": 53102
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6974309086799622,
+      "learning_rate": 2.5624155598713815e-05,
+      "loss": 2.8641,
+      "step": 53103
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7470758557319641,
+      "learning_rate": 2.5617929877359732e-05,
+      "loss": 2.9929,
+      "step": 53104
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7814863324165344,
+      "learning_rate": 2.5611704878675244e-05,
+      "loss": 2.8942,
+      "step": 53105
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7379601001739502,
+      "learning_rate": 2.5605480602676698e-05,
+      "loss": 3.0074,
+      "step": 53106
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7781730890274048,
+      "learning_rate": 2.5599257049380452e-05,
+      "loss": 2.663,
+      "step": 53107
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7469813227653503,
+      "learning_rate": 2.5593034218802956e-05,
+      "loss": 2.8231,
+      "step": 53108
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7671642303466797,
+      "learning_rate": 2.5586812110960665e-05,
+      "loss": 2.9087,
+      "step": 53109
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7788114547729492,
+      "learning_rate": 2.558059072586983e-05,
+      "loss": 2.7514,
+      "step": 53110
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7065609693527222,
+      "learning_rate": 2.5574370063546944e-05,
+      "loss": 2.991,
+      "step": 53111
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.725106418132782,
+      "learning_rate": 2.5568150124008324e-05,
+      "loss": 2.8381,
+      "step": 53112
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7204715013504028,
+      "learning_rate": 2.5561930907270456e-05,
+      "loss": 2.9109,
+      "step": 53113
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7312618494033813,
+      "learning_rate": 2.5555712413349594e-05,
+      "loss": 3.0618,
+      "step": 53114
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7476996779441833,
+      "learning_rate": 2.554949464226216e-05,
+      "loss": 2.8415,
+      "step": 53115
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.763667106628418,
+      "learning_rate": 2.554327759402457e-05,
+      "loss": 2.8231,
+      "step": 53116
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.759936511516571,
+      "learning_rate": 2.5537061268653114e-05,
+      "loss": 2.825,
+      "step": 53117
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7484183311462402,
+      "learning_rate": 2.5530845666164212e-05,
+      "loss": 2.7023,
+      "step": 53118
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.676468014717102,
+      "learning_rate": 2.5524630786574285e-05,
+      "loss": 3.0118,
+      "step": 53119
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7524356842041016,
+      "learning_rate": 2.5518416629899687e-05,
+      "loss": 2.7162,
+      "step": 53120
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7328376173973083,
+      "learning_rate": 2.5512203196156735e-05,
+      "loss": 2.9075,
+      "step": 53121
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7236394286155701,
+      "learning_rate": 2.550599048536175e-05,
+      "loss": 2.8896,
+      "step": 53122
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7498464584350586,
+      "learning_rate": 2.5499778497531188e-05,
+      "loss": 2.8998,
+      "step": 53123
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7795760631561279,
+      "learning_rate": 2.54935672326814e-05,
+      "loss": 2.6818,
+      "step": 53124
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7325364351272583,
+      "learning_rate": 2.5487356690828707e-05,
+      "loss": 2.8785,
+      "step": 53125
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7392412424087524,
+      "learning_rate": 2.54811468719895e-05,
+      "loss": 2.9473,
+      "step": 53126
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7106963992118835,
+      "learning_rate": 2.547493777618016e-05,
+      "loss": 2.9328,
+      "step": 53127
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7459375262260437,
+      "learning_rate": 2.546872940341691e-05,
+      "loss": 2.8673,
+      "step": 53128
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7813073992729187,
+      "learning_rate": 2.546252175371627e-05,
+      "loss": 2.8378,
+      "step": 53129
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7099730372428894,
+      "learning_rate": 2.545631482709446e-05,
+      "loss": 2.8736,
+      "step": 53130
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.739006757736206,
+      "learning_rate": 2.5450108623567933e-05,
+      "loss": 2.7253,
+      "step": 53131
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7848395109176636,
+      "learning_rate": 2.5443903143152912e-05,
+      "loss": 2.7427,
+      "step": 53132
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7411284446716309,
+      "learning_rate": 2.543769838586581e-05,
+      "loss": 2.8126,
+      "step": 53133
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8115738034248352,
+      "learning_rate": 2.5431494351723058e-05,
+      "loss": 3.1662,
+      "step": 53134
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7761831879615784,
+      "learning_rate": 2.542529104074087e-05,
+      "loss": 2.9151,
+      "step": 53135
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7265602350234985,
+      "learning_rate": 2.541908845293563e-05,
+      "loss": 3.0349,
+      "step": 53136
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.721573531627655,
+      "learning_rate": 2.5412886588323634e-05,
+      "loss": 2.9599,
+      "step": 53137
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7648131251335144,
+      "learning_rate": 2.5406685446921228e-05,
+      "loss": 2.9716,
+      "step": 53138
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7431436777114868,
+      "learning_rate": 2.5400485028744833e-05,
+      "loss": 2.8647,
+      "step": 53139
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7635769844055176,
+      "learning_rate": 2.5394285333810637e-05,
+      "loss": 3.0528,
+      "step": 53140
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7431437373161316,
+      "learning_rate": 2.538808636213513e-05,
+      "loss": 3.0272,
+      "step": 53141
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.735590934753418,
+      "learning_rate": 2.5381888113734528e-05,
+      "loss": 2.8362,
+      "step": 53142
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7066057920455933,
+      "learning_rate": 2.537569058862512e-05,
+      "loss": 3.017,
+      "step": 53143
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7370970845222473,
+      "learning_rate": 2.5369493786823358e-05,
+      "loss": 2.9394,
+      "step": 53144
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.721082866191864,
+      "learning_rate": 2.5363297708345465e-05,
+      "loss": 2.8916,
+      "step": 53145
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7915112376213074,
+      "learning_rate": 2.5357102353207827e-05,
+      "loss": 2.8815,
+      "step": 53146
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7461155652999878,
+      "learning_rate": 2.5350907721426663e-05,
+      "loss": 2.838,
+      "step": 53147
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7456420660018921,
+      "learning_rate": 2.5344713813018425e-05,
+      "loss": 2.8849,
+      "step": 53148
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7861379384994507,
+      "learning_rate": 2.5338520627999337e-05,
+      "loss": 2.6795,
+      "step": 53149
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7006232738494873,
+      "learning_rate": 2.533232816638565e-05,
+      "loss": 3.0351,
+      "step": 53150
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7393845319747925,
+      "learning_rate": 2.5326136428193823e-05,
+      "loss": 2.7669,
+      "step": 53151
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7685011029243469,
+      "learning_rate": 2.531994541344007e-05,
+      "loss": 2.8029,
+      "step": 53152
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7489452958106995,
+      "learning_rate": 2.531375512214071e-05,
+      "loss": 3.1667,
+      "step": 53153
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7370985150337219,
+      "learning_rate": 2.5307565554312038e-05,
+      "loss": 2.8493,
+      "step": 53154
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.7306023240089417,
+      "learning_rate": 2.5301376709970368e-05,
+      "loss": 2.9711,
+      "step": 53155
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7589641809463501,
+      "learning_rate": 2.5295188589132054e-05,
+      "loss": 2.9249,
+      "step": 53156
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7181271314620972,
+      "learning_rate": 2.5289001191813353e-05,
+      "loss": 2.8004,
+      "step": 53157
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.765256941318512,
+      "learning_rate": 2.5282814518030547e-05,
+      "loss": 2.964,
+      "step": 53158
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6980361342430115,
+      "learning_rate": 2.5276628567799894e-05,
+      "loss": 2.819,
+      "step": 53159
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6991076469421387,
+      "learning_rate": 2.5270443341137714e-05,
+      "loss": 3.0081,
+      "step": 53160
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8100741505622864,
+      "learning_rate": 2.526425883806039e-05,
+      "loss": 2.9064,
+      "step": 53161
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7279528975486755,
+      "learning_rate": 2.5258075058584047e-05,
+      "loss": 3.0166,
+      "step": 53162
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7336760759353638,
+      "learning_rate": 2.5251892002725138e-05,
+      "loss": 2.7742,
+      "step": 53163
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7446197271347046,
+      "learning_rate": 2.524570967049988e-05,
+      "loss": 3.0359,
+      "step": 53164
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7815460562705994,
+      "learning_rate": 2.5239528061924462e-05,
+      "loss": 3.029,
+      "step": 53165
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8095081448554993,
+      "learning_rate": 2.5233347177015307e-05,
+      "loss": 2.8958,
+      "step": 53166
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7752934694290161,
+      "learning_rate": 2.5227167015788597e-05,
+      "loss": 2.8065,
+      "step": 53167
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7434821724891663,
+      "learning_rate": 2.5220987578260688e-05,
+      "loss": 2.9032,
+      "step": 53168
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7885562181472778,
+      "learning_rate": 2.5214808864447765e-05,
+      "loss": 3.0296,
+      "step": 53169
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7058457136154175,
+      "learning_rate": 2.520863087436622e-05,
+      "loss": 2.7885,
+      "step": 53170
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7025486826896667,
+      "learning_rate": 2.5202453608032236e-05,
+      "loss": 2.8992,
+      "step": 53171
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7743384838104248,
+      "learning_rate": 2.5196277065462066e-05,
+      "loss": 2.8032,
+      "step": 53172
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7513494491577148,
+      "learning_rate": 2.5190101246672068e-05,
+      "loss": 2.8837,
+      "step": 53173
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7396803498268127,
+      "learning_rate": 2.5183926151678392e-05,
+      "loss": 2.8922,
+      "step": 53174
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7196879386901855,
+      "learning_rate": 2.5177751780497358e-05,
+      "loss": 3.0125,
+      "step": 53175
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7186749577522278,
+      "learning_rate": 2.517157813314532e-05,
+      "loss": 2.7121,
+      "step": 53176
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7624224424362183,
+      "learning_rate": 2.5165405209638366e-05,
+      "loss": 2.9514,
+      "step": 53177
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7520177364349365,
+      "learning_rate": 2.5159233009992917e-05,
+      "loss": 3.0939,
+      "step": 53178
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7938098907470703,
+      "learning_rate": 2.5153061534225128e-05,
+      "loss": 2.8881,
+      "step": 53179
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7314983606338501,
+      "learning_rate": 2.514689078235125e-05,
+      "loss": 2.8824,
+      "step": 53180
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7542241811752319,
+      "learning_rate": 2.5140720754387634e-05,
+      "loss": 2.9394,
+      "step": 53181
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7400225400924683,
+      "learning_rate": 2.513455145035037e-05,
+      "loss": 2.7797,
+      "step": 53182
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.747697651386261,
+      "learning_rate": 2.5128382870255882e-05,
+      "loss": 2.9337,
+      "step": 53183
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8165611624717712,
+      "learning_rate": 2.5122215014120252e-05,
+      "loss": 2.7655,
+      "step": 53184
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7541787624359131,
+      "learning_rate": 2.5116047881959866e-05,
+      "loss": 2.9598,
+      "step": 53185
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7399340271949768,
+      "learning_rate": 2.5109881473790914e-05,
+      "loss": 2.9438,
+      "step": 53186
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7630729079246521,
+      "learning_rate": 2.5103715789629585e-05,
+      "loss": 2.9433,
+      "step": 53187
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7590308785438538,
+      "learning_rate": 2.509755082949223e-05,
+      "loss": 3.0062,
+      "step": 53188
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7435056567192078,
+      "learning_rate": 2.5091386593394935e-05,
+      "loss": 2.7753,
+      "step": 53189
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7512586712837219,
+      "learning_rate": 2.5085223081354023e-05,
+      "loss": 2.9262,
+      "step": 53190
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.723823070526123,
+      "learning_rate": 2.507906029338581e-05,
+      "loss": 2.9895,
+      "step": 53191
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.743118166923523,
+      "learning_rate": 2.5072898229506423e-05,
+      "loss": 2.8739,
+      "step": 53192
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7139184474945068,
+      "learning_rate": 2.506673688973211e-05,
+      "loss": 2.7876,
+      "step": 53193
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8011366724967957,
+      "learning_rate": 2.5060576274079024e-05,
+      "loss": 3.0433,
+      "step": 53194
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7116770148277283,
+      "learning_rate": 2.505441638256349e-05,
+      "loss": 3.1634,
+      "step": 53195
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7735740542411804,
+      "learning_rate": 2.5048257215201762e-05,
+      "loss": 2.8701,
+      "step": 53196
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7631982564926147,
+      "learning_rate": 2.5042098772009955e-05,
+      "loss": 2.9832,
+      "step": 53197
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7225585579872131,
+      "learning_rate": 2.5035941053004392e-05,
+      "loss": 3.0262,
+      "step": 53198
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.744941234588623,
+      "learning_rate": 2.5029784058201197e-05,
+      "loss": 2.9931,
+      "step": 53199
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7380294799804688,
+      "learning_rate": 2.5023627787616685e-05,
+      "loss": 2.8191,
+      "step": 53200
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6965009570121765,
+      "learning_rate": 2.501747224126701e-05,
+      "loss": 2.9667,
+      "step": 53201
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.767867922782898,
+      "learning_rate": 2.501131741916833e-05,
+      "loss": 2.8884,
+      "step": 53202
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7580838203430176,
+      "learning_rate": 2.5005163321336996e-05,
+      "loss": 2.9514,
+      "step": 53203
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7149063944816589,
+      "learning_rate": 2.499900994778906e-05,
+      "loss": 2.8192,
+      "step": 53204
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.771682858467102,
+      "learning_rate": 2.4992857298540813e-05,
+      "loss": 2.8842,
+      "step": 53205
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7764263153076172,
+      "learning_rate": 2.498670537360854e-05,
+      "loss": 2.8959,
+      "step": 53206
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7236824631690979,
+      "learning_rate": 2.4980554173008326e-05,
+      "loss": 2.844,
+      "step": 53207
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7100843191146851,
+      "learning_rate": 2.4974403696756394e-05,
+      "loss": 2.8599,
+      "step": 53208
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7423948049545288,
+      "learning_rate": 2.4968253944868933e-05,
+      "loss": 2.8558,
+      "step": 53209
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.698421061038971,
+      "learning_rate": 2.4962104917362124e-05,
+      "loss": 2.9021,
+      "step": 53210
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7575321793556213,
+      "learning_rate": 2.495595661425226e-05,
+      "loss": 3.1162,
+      "step": 53211
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.710386335849762,
+      "learning_rate": 2.4949809035555425e-05,
+      "loss": 2.816,
+      "step": 53212
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7282596826553345,
+      "learning_rate": 2.4943662181287937e-05,
+      "loss": 2.9855,
+      "step": 53213
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7500553727149963,
+      "learning_rate": 2.4937516051465888e-05,
+      "loss": 2.7933,
+      "step": 53214
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8065747022628784,
+      "learning_rate": 2.493137064610543e-05,
+      "loss": 2.8999,
+      "step": 53215
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7221754789352417,
+      "learning_rate": 2.492522596522285e-05,
+      "loss": 2.7116,
+      "step": 53216
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7535700798034668,
+      "learning_rate": 2.4919082008834236e-05,
+      "loss": 2.842,
+      "step": 53217
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7541929483413696,
+      "learning_rate": 2.4912938776955872e-05,
+      "loss": 2.9126,
+      "step": 53218
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7476422190666199,
+      "learning_rate": 2.4906796269603847e-05,
+      "loss": 2.9692,
+      "step": 53219
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7244073152542114,
+      "learning_rate": 2.4900654486794346e-05,
+      "loss": 2.8498,
+      "step": 53220
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7529792189598083,
+      "learning_rate": 2.489451342854366e-05,
+      "loss": 3.0183,
+      "step": 53221
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7540156841278076,
+      "learning_rate": 2.4888373094867876e-05,
+      "loss": 2.6817,
+      "step": 53222
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7746258974075317,
+      "learning_rate": 2.4882233485783144e-05,
+      "loss": 3.0357,
+      "step": 53223
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7871257662773132,
+      "learning_rate": 2.487609460130565e-05,
+      "loss": 2.6511,
+      "step": 53224
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7155481576919556,
+      "learning_rate": 2.4869956441451587e-05,
+      "loss": 3.1273,
+      "step": 53225
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7250311970710754,
+      "learning_rate": 2.4863819006237072e-05,
+      "loss": 2.9099,
+      "step": 53226
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7574231028556824,
+      "learning_rate": 2.4857682295678326e-05,
+      "loss": 3.0626,
+      "step": 53227
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7355062961578369,
+      "learning_rate": 2.4851546309791536e-05,
+      "loss": 2.9971,
+      "step": 53228
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7345170974731445,
+      "learning_rate": 2.484541104859282e-05,
+      "loss": 3.0486,
+      "step": 53229
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6943804621696472,
+      "learning_rate": 2.4839276512098304e-05,
+      "loss": 2.799,
+      "step": 53230
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7673830389976501,
+      "learning_rate": 2.483314270032417e-05,
+      "loss": 2.7265,
+      "step": 53231
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7196083068847656,
+      "learning_rate": 2.4827009613286575e-05,
+      "loss": 2.816,
+      "step": 53232
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.734291136264801,
+      "learning_rate": 2.4820877251001702e-05,
+      "loss": 2.8073,
+      "step": 53233
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7641286253929138,
+      "learning_rate": 2.4814745613485643e-05,
+      "loss": 3.1057,
+      "step": 53234
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7599497437477112,
+      "learning_rate": 2.4808614700754647e-05,
+      "loss": 2.8137,
+      "step": 53235
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7263058423995972,
+      "learning_rate": 2.480248451282474e-05,
+      "loss": 2.8649,
+      "step": 53236
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7398238182067871,
+      "learning_rate": 2.479635504971217e-05,
+      "loss": 2.7207,
+      "step": 53237
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7094271779060364,
+      "learning_rate": 2.4790226311433025e-05,
+      "loss": 2.9685,
+      "step": 53238
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7355355620384216,
+      "learning_rate": 2.478409829800343e-05,
+      "loss": 2.8561,
+      "step": 53239
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.747589111328125,
+      "learning_rate": 2.4777971009439602e-05,
+      "loss": 2.9834,
+      "step": 53240
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7668625116348267,
+      "learning_rate": 2.477184444575756e-05,
+      "loss": 2.7875,
+      "step": 53241
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7156922221183777,
+      "learning_rate": 2.476571860697353e-05,
+      "loss": 2.7711,
+      "step": 53242
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7356650233268738,
+      "learning_rate": 2.475959349310369e-05,
+      "loss": 2.8436,
+      "step": 53243
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7467832565307617,
+      "learning_rate": 2.47534691041641e-05,
+      "loss": 2.7338,
+      "step": 53244
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7915269732475281,
+      "learning_rate": 2.4747345440170885e-05,
+      "loss": 2.8669,
+      "step": 53245
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7523009181022644,
+      "learning_rate": 2.4741222501140155e-05,
+      "loss": 2.8385,
+      "step": 53246
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8412205576896667,
+      "learning_rate": 2.4735100287088038e-05,
+      "loss": 3.0301,
+      "step": 53247
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7007306814193726,
+      "learning_rate": 2.4728978798030787e-05,
+      "loss": 3.0305,
+      "step": 53248
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7333875298500061,
+      "learning_rate": 2.472285803398435e-05,
+      "loss": 2.9612,
+      "step": 53249
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8363428711891174,
+      "learning_rate": 2.4716737994964986e-05,
+      "loss": 2.8032,
+      "step": 53250
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7706584334373474,
+      "learning_rate": 2.4710618680988746e-05,
+      "loss": 3.0626,
+      "step": 53251
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7772002816200256,
+      "learning_rate": 2.4704500092071723e-05,
+      "loss": 2.8498,
+      "step": 53252
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7295860648155212,
+      "learning_rate": 2.4698382228230095e-05,
+      "loss": 2.9828,
+      "step": 53253
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.763123631477356,
+      "learning_rate": 2.4692265089479922e-05,
+      "loss": 2.9188,
+      "step": 53254
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8055310845375061,
+      "learning_rate": 2.468614867583736e-05,
+      "loss": 2.993,
+      "step": 53255
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7372824549674988,
+      "learning_rate": 2.4680032987318454e-05,
+      "loss": 2.5725,
+      "step": 53256
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8814607262611389,
+      "learning_rate": 2.4673918023939364e-05,
+      "loss": 2.8363,
+      "step": 53257
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7216055393218994,
+      "learning_rate": 2.4667803785716245e-05,
+      "loss": 2.7199,
+      "step": 53258
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7859689593315125,
+      "learning_rate": 2.4661690272665114e-05,
+      "loss": 2.9632,
+      "step": 53259
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7562228441238403,
+      "learning_rate": 2.4655577484802125e-05,
+      "loss": 2.9153,
+      "step": 53260
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7715989351272583,
+      "learning_rate": 2.46494654221433e-05,
+      "loss": 2.8636,
+      "step": 53261
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7629818916320801,
+      "learning_rate": 2.464335408470479e-05,
+      "loss": 3.0001,
+      "step": 53262
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8690528273582458,
+      "learning_rate": 2.4637243472502723e-05,
+      "loss": 2.9002,
+      "step": 53263
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7542612552642822,
+      "learning_rate": 2.4631133585553144e-05,
+      "loss": 2.8384,
+      "step": 53264
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7296197414398193,
+      "learning_rate": 2.462502442387221e-05,
+      "loss": 2.9057,
+      "step": 53265
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7588483095169067,
+      "learning_rate": 2.461891598747594e-05,
+      "loss": 2.5271,
+      "step": 53266
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.750971794128418,
+      "learning_rate": 2.461280827638039e-05,
+      "loss": 2.8642,
+      "step": 53267
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7146446704864502,
+      "learning_rate": 2.4606701290601783e-05,
+      "loss": 2.8347,
+      "step": 53268
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.755339503288269,
+      "learning_rate": 2.4600595030156066e-05,
+      "loss": 2.8039,
+      "step": 53269
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7227151393890381,
+      "learning_rate": 2.4594489495059434e-05,
+      "loss": 3.0266,
+      "step": 53270
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7845844626426697,
+      "learning_rate": 2.4588384685327865e-05,
+      "loss": 2.7491,
+      "step": 53271
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7608388066291809,
+      "learning_rate": 2.4582280600977522e-05,
+      "loss": 3.0839,
+      "step": 53272
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7561764121055603,
+      "learning_rate": 2.4576177242024453e-05,
+      "loss": 3.1059,
+      "step": 53273
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7092791199684143,
+      "learning_rate": 2.457007460848468e-05,
+      "loss": 2.9953,
+      "step": 53274
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7406812906265259,
+      "learning_rate": 2.4563972700374357e-05,
+      "loss": 3.1362,
+      "step": 53275
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7598647475242615,
+      "learning_rate": 2.4557871517709503e-05,
+      "loss": 2.6158,
+      "step": 53276
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7269623279571533,
+      "learning_rate": 2.4551771060506177e-05,
+      "loss": 2.9399,
+      "step": 53277
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7937368154525757,
+      "learning_rate": 2.4545671328780526e-05,
+      "loss": 2.7327,
+      "step": 53278
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7488956451416016,
+      "learning_rate": 2.453957232254854e-05,
+      "loss": 2.8383,
+      "step": 53279
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7417857646942139,
+      "learning_rate": 2.4533474041826338e-05,
+      "loss": 2.8689,
+      "step": 53280
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7532116770744324,
+      "learning_rate": 2.4527376486629945e-05,
+      "loss": 2.963,
+      "step": 53281
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7646198868751526,
+      "learning_rate": 2.452127965697538e-05,
+      "loss": 3.0262,
+      "step": 53282
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7706944346427917,
+      "learning_rate": 2.4515183552878824e-05,
+      "loss": 2.781,
+      "step": 53283
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7329809665679932,
+      "learning_rate": 2.450908817435617e-05,
+      "loss": 2.8377,
+      "step": 53284
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7508072257041931,
+      "learning_rate": 2.4502993521423643e-05,
+      "loss": 2.8186,
+      "step": 53285
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.755642831325531,
+      "learning_rate": 2.4496899594097118e-05,
+      "loss": 2.888,
+      "step": 53286
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7248871922492981,
+      "learning_rate": 2.4490806392392826e-05,
+      "loss": 2.7755,
+      "step": 53287
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7461181879043579,
+      "learning_rate": 2.4484713916326715e-05,
+      "loss": 2.8476,
+      "step": 53288
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7649577856063843,
+      "learning_rate": 2.4478622165914774e-05,
+      "loss": 2.784,
+      "step": 53289
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7114865779876709,
+      "learning_rate": 2.447253114117319e-05,
+      "loss": 2.9854,
+      "step": 53290
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7199516892433167,
+      "learning_rate": 2.4466440842117886e-05,
+      "loss": 2.9036,
+      "step": 53291
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7920954823493958,
+      "learning_rate": 2.4460351268764943e-05,
+      "loss": 2.7358,
+      "step": 53292
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.762523889541626,
+      "learning_rate": 2.4454262421130456e-05,
+      "loss": 2.8051,
+      "step": 53293
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7442542910575867,
+      "learning_rate": 2.4448174299230404e-05,
+      "loss": 2.7057,
+      "step": 53294
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7322526574134827,
+      "learning_rate": 2.4442086903080815e-05,
+      "loss": 2.954,
+      "step": 53295
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7498273849487305,
+      "learning_rate": 2.4436000232697706e-05,
+      "loss": 2.9617,
+      "step": 53296
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8006744384765625,
+      "learning_rate": 2.4429914288097162e-05,
+      "loss": 2.8717,
+      "step": 53297
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7188971638679504,
+      "learning_rate": 2.4423829069295207e-05,
+      "loss": 2.9337,
+      "step": 53298
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.728274941444397,
+      "learning_rate": 2.4417744576307796e-05,
+      "loss": 2.8373,
+      "step": 53299
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7777703404426575,
+      "learning_rate": 2.441166080915108e-05,
+      "loss": 2.7575,
+      "step": 53300
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8000141978263855,
+      "learning_rate": 2.4405577767840946e-05,
+      "loss": 3.0339,
+      "step": 53301
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.749862551689148,
+      "learning_rate": 2.4399495452393547e-05,
+      "loss": 2.854,
+      "step": 53302
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7590504884719849,
+      "learning_rate": 2.439341386282484e-05,
+      "loss": 2.7715,
+      "step": 53303
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7074437737464905,
+      "learning_rate": 2.4387332999150778e-05,
+      "loss": 2.9909,
+      "step": 53304
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7344616651535034,
+      "learning_rate": 2.4381252861387478e-05,
+      "loss": 3.0727,
+      "step": 53305
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7661343812942505,
+      "learning_rate": 2.43751734495509e-05,
+      "loss": 2.9525,
+      "step": 53306
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7796639204025269,
+      "learning_rate": 2.4369094763657093e-05,
+      "loss": 2.9349,
+      "step": 53307
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.749068021774292,
+      "learning_rate": 2.4363016803721978e-05,
+      "loss": 3.0218,
+      "step": 53308
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7495163083076477,
+      "learning_rate": 2.4356939569761713e-05,
+      "loss": 2.7993,
+      "step": 53309
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7630246877670288,
+      "learning_rate": 2.435086306179218e-05,
+      "loss": 2.8242,
+      "step": 53310
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7312656044960022,
+      "learning_rate": 2.43447872798294e-05,
+      "loss": 2.8078,
+      "step": 53311
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7347916960716248,
+      "learning_rate": 2.4338712223889467e-05,
+      "loss": 2.7272,
+      "step": 53312
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7420660853385925,
+      "learning_rate": 2.4332637893988228e-05,
+      "loss": 3.0838,
+      "step": 53313
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7539196014404297,
+      "learning_rate": 2.4326564290141802e-05,
+      "loss": 2.8889,
+      "step": 53314
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7311657071113586,
+      "learning_rate": 2.4320491412366183e-05,
+      "loss": 2.8895,
+      "step": 53315
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7519007921218872,
+      "learning_rate": 2.431441926067732e-05,
+      "loss": 2.7742,
+      "step": 53316
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7469418048858643,
+      "learning_rate": 2.4308347835091234e-05,
+      "loss": 2.7949,
+      "step": 53317
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7581943273544312,
+      "learning_rate": 2.4302277135623883e-05,
+      "loss": 2.8109,
+      "step": 53318
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7415329813957214,
+      "learning_rate": 2.4296207162291213e-05,
+      "loss": 3.1124,
+      "step": 53319
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.784935712814331,
+      "learning_rate": 2.4290137915109386e-05,
+      "loss": 2.8482,
+      "step": 53320
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6955008506774902,
+      "learning_rate": 2.4284069394094187e-05,
+      "loss": 2.8681,
+      "step": 53321
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7275387644767761,
+      "learning_rate": 2.4278001599261732e-05,
+      "loss": 2.9916,
+      "step": 53322
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7459524273872375,
+      "learning_rate": 2.4271934530627913e-05,
+      "loss": 2.751,
+      "step": 53323
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7747052311897278,
+      "learning_rate": 2.4265868188208813e-05,
+      "loss": 2.9919,
+      "step": 53324
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7346550822257996,
+      "learning_rate": 2.425980257202036e-05,
+      "loss": 2.6747,
+      "step": 53325
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7680572271347046,
+      "learning_rate": 2.4253737682078433e-05,
+      "loss": 2.7634,
+      "step": 53326
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7524115443229675,
+      "learning_rate": 2.424767351839919e-05,
+      "loss": 3.0331,
+      "step": 53327
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7559760808944702,
+      "learning_rate": 2.424161008099842e-05,
+      "loss": 2.8906,
+      "step": 53328
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7822624444961548,
+      "learning_rate": 2.4235547369892174e-05,
+      "loss": 2.6455,
+      "step": 53329
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6873143315315247,
+      "learning_rate": 2.4229485385096504e-05,
+      "loss": 2.9405,
+      "step": 53330
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7661668658256531,
+      "learning_rate": 2.4223424126627267e-05,
+      "loss": 2.9843,
+      "step": 53331
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7725754380226135,
+      "learning_rate": 2.4217363594500484e-05,
+      "loss": 2.8333,
+      "step": 53332
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7318544983863831,
+      "learning_rate": 2.421130378873204e-05,
+      "loss": 2.8713,
+      "step": 53333
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7380740642547607,
+      "learning_rate": 2.4205244709337923e-05,
+      "loss": 3.1337,
+      "step": 53334
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7293857932090759,
+      "learning_rate": 2.4199186356334156e-05,
+      "loss": 2.7604,
+      "step": 53335
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8121163845062256,
+      "learning_rate": 2.419312872973662e-05,
+      "loss": 2.935,
+      "step": 53336
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7472822070121765,
+      "learning_rate": 2.4187071829561376e-05,
+      "loss": 2.801,
+      "step": 53337
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7153192758560181,
+      "learning_rate": 2.418101565582421e-05,
+      "loss": 2.7313,
+      "step": 53338
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7651809453964233,
+      "learning_rate": 2.4174960208541238e-05,
+      "loss": 3.0331,
+      "step": 53339
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7362973093986511,
+      "learning_rate": 2.4168905487728353e-05,
+      "loss": 2.9481,
+      "step": 53340
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8571172952651978,
+      "learning_rate": 2.4162851493401403e-05,
+      "loss": 2.7162,
+      "step": 53341
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.766938328742981,
+      "learning_rate": 2.4156798225576478e-05,
+      "loss": 3.0625,
+      "step": 53342
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7803354859352112,
+      "learning_rate": 2.4150745684269436e-05,
+      "loss": 2.8998,
+      "step": 53343
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7079799771308899,
+      "learning_rate": 2.414469386949619e-05,
+      "loss": 2.8232,
+      "step": 53344
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7367957830429077,
+      "learning_rate": 2.413864278127283e-05,
+      "loss": 2.5717,
+      "step": 53345
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7332743406295776,
+      "learning_rate": 2.4132592419615184e-05,
+      "loss": 2.8145,
+      "step": 53346
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.724522590637207,
+      "learning_rate": 2.412654278453916e-05,
+      "loss": 3.0007,
+      "step": 53347
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7491048574447632,
+      "learning_rate": 2.412049387606072e-05,
+      "loss": 2.6494,
+      "step": 53348
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7689740061759949,
+      "learning_rate": 2.4114445694195783e-05,
+      "loss": 2.8103,
+      "step": 53349
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8025519251823425,
+      "learning_rate": 2.410839823896037e-05,
+      "loss": 2.7399,
+      "step": 53350
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7537023425102234,
+      "learning_rate": 2.4102351510370267e-05,
+      "loss": 2.8394,
+      "step": 53351
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7922477722167969,
+      "learning_rate": 2.409630550844156e-05,
+      "loss": 2.9793,
+      "step": 53352
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.77238529920578,
+      "learning_rate": 2.409026023319004e-05,
+      "loss": 2.83,
+      "step": 53353
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7189983129501343,
+      "learning_rate": 2.408421568463166e-05,
+      "loss": 3.0147,
+      "step": 53354
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7614080905914307,
+      "learning_rate": 2.4078171862782403e-05,
+      "loss": 2.7063,
+      "step": 53355
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.749724805355072,
+      "learning_rate": 2.4072128767658062e-05,
+      "loss": 2.812,
+      "step": 53356
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7478910684585571,
+      "learning_rate": 2.4066086399274687e-05,
+      "loss": 2.8834,
+      "step": 53357
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8119695782661438,
+      "learning_rate": 2.4060044757648102e-05,
+      "loss": 3.0316,
+      "step": 53358
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7431463003158569,
+      "learning_rate": 2.4054003842794257e-05,
+      "loss": 2.8187,
+      "step": 53359
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7376152276992798,
+      "learning_rate": 2.4047963654729075e-05,
+      "loss": 2.7177,
+      "step": 53360
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7031460404396057,
+      "learning_rate": 2.4041924193468475e-05,
+      "loss": 2.753,
+      "step": 53361
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7614917755126953,
+      "learning_rate": 2.4035885459028347e-05,
+      "loss": 2.8834,
+      "step": 53362
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7040819525718689,
+      "learning_rate": 2.4029847451424512e-05,
+      "loss": 2.9195,
+      "step": 53363
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7586194276809692,
+      "learning_rate": 2.4023810170672954e-05,
+      "loss": 3.0339,
+      "step": 53364
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7714332938194275,
+      "learning_rate": 2.4017773616789628e-05,
+      "loss": 2.9253,
+      "step": 53365
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7194476127624512,
+      "learning_rate": 2.401173778979032e-05,
+      "loss": 2.8969,
+      "step": 53366
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7926198840141296,
+      "learning_rate": 2.400570268969102e-05,
+      "loss": 3.1839,
+      "step": 53367
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7163087129592896,
+      "learning_rate": 2.3999668316507582e-05,
+      "loss": 2.7451,
+      "step": 53368
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7153766751289368,
+      "learning_rate": 2.3993634670255856e-05,
+      "loss": 2.8174,
+      "step": 53369
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7822369933128357,
+      "learning_rate": 2.3987601750951835e-05,
+      "loss": 2.7719,
+      "step": 53370
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7908663749694824,
+      "learning_rate": 2.3981569558611303e-05,
+      "loss": 2.7651,
+      "step": 53371
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7336825132369995,
+      "learning_rate": 2.3975538093250245e-05,
+      "loss": 2.8495,
+      "step": 53372
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8331197500228882,
+      "learning_rate": 2.396950735488442e-05,
+      "loss": 2.895,
+      "step": 53373
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7461316585540771,
+      "learning_rate": 2.396347734352988e-05,
+      "loss": 2.8857,
+      "step": 53374
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7184967994689941,
+      "learning_rate": 2.395744805920238e-05,
+      "loss": 2.9948,
+      "step": 53375
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8150850534439087,
+      "learning_rate": 2.39514195019178e-05,
+      "loss": 2.8987,
+      "step": 53376
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.750407338142395,
+      "learning_rate": 2.3945391671692138e-05,
+      "loss": 3.0958,
+      "step": 53377
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7473511099815369,
+      "learning_rate": 2.3939364568541075e-05,
+      "loss": 2.9058,
+      "step": 53378
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7648722529411316,
+      "learning_rate": 2.39333381924807e-05,
+      "loss": 2.9031,
+      "step": 53379
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7389246821403503,
+      "learning_rate": 2.39273125435267e-05,
+      "loss": 3.0294,
+      "step": 53380
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7998449206352234,
+      "learning_rate": 2.392128762169503e-05,
+      "loss": 2.9568,
+      "step": 53381
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7724308371543884,
+      "learning_rate": 2.3915263427001608e-05,
+      "loss": 2.7814,
+      "step": 53382
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7729584574699402,
+      "learning_rate": 2.3909239959462258e-05,
+      "loss": 2.9317,
+      "step": 53383
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7403950095176697,
+      "learning_rate": 2.390321721909283e-05,
+      "loss": 2.8143,
+      "step": 53384
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7252776026725769,
+      "learning_rate": 2.3897195205909148e-05,
+      "loss": 2.7213,
+      "step": 53385
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7848640084266663,
+      "learning_rate": 2.389117391992713e-05,
+      "loss": 2.8572,
+      "step": 53386
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7389460206031799,
+      "learning_rate": 2.3885153361162636e-05,
+      "loss": 2.7936,
+      "step": 53387
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7464485764503479,
+      "learning_rate": 2.3879133529631477e-05,
+      "loss": 2.8893,
+      "step": 53388
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7308782339096069,
+      "learning_rate": 2.3873114425349584e-05,
+      "loss": 2.8139,
+      "step": 53389
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7255467772483826,
+      "learning_rate": 2.38670960483328e-05,
+      "loss": 3.0575,
+      "step": 53390
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7490675449371338,
+      "learning_rate": 2.3861078398596857e-05,
+      "loss": 2.8725,
+      "step": 53391
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7539382576942444,
+      "learning_rate": 2.3855061476157733e-05,
+      "loss": 2.81,
+      "step": 53392
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7484925389289856,
+      "learning_rate": 2.3849045281031188e-05,
+      "loss": 2.9626,
+      "step": 53393
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7690110802650452,
+      "learning_rate": 2.3843029813233173e-05,
+      "loss": 3.0192,
+      "step": 53394
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.697595477104187,
+      "learning_rate": 2.383701507277941e-05,
+      "loss": 3.0619,
+      "step": 53395
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7498865723609924,
+      "learning_rate": 2.3831001059685884e-05,
+      "loss": 2.7544,
+      "step": 53396
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.6930327415466309,
+      "learning_rate": 2.382498777396832e-05,
+      "loss": 2.9071,
+      "step": 53397
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7832930088043213,
+      "learning_rate": 2.3818975215642532e-05,
+      "loss": 2.8844,
+      "step": 53398
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7338464856147766,
+      "learning_rate": 2.381296338472448e-05,
+      "loss": 2.8968,
+      "step": 53399
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7855767607688904,
+      "learning_rate": 2.3806952281229853e-05,
+      "loss": 2.6963,
+      "step": 53400
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7518692016601562,
+      "learning_rate": 2.3800941905174563e-05,
+      "loss": 2.9422,
+      "step": 53401
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7118070721626282,
+      "learning_rate": 2.3794932256574505e-05,
+      "loss": 2.9071,
+      "step": 53402
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8377758264541626,
+      "learning_rate": 2.3788923335445398e-05,
+      "loss": 3.1013,
+      "step": 53403
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7513577342033386,
+      "learning_rate": 2.3782915141803127e-05,
+      "loss": 3.0399,
+      "step": 53404
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.796913206577301,
+      "learning_rate": 2.3776907675663513e-05,
+      "loss": 2.7059,
+      "step": 53405
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7465709447860718,
+      "learning_rate": 2.3770900937042314e-05,
+      "loss": 2.7208,
+      "step": 53406
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7374803423881531,
+      "learning_rate": 2.376489492595548e-05,
+      "loss": 2.7431,
+      "step": 53407
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7477961182594299,
+      "learning_rate": 2.3758889642418665e-05,
+      "loss": 2.7817,
+      "step": 53408
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8048363327980042,
+      "learning_rate": 2.3752885086447827e-05,
+      "loss": 2.9601,
+      "step": 53409
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7863703370094299,
+      "learning_rate": 2.374688125805868e-05,
+      "loss": 2.7946,
+      "step": 53410
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7128931283950806,
+      "learning_rate": 2.3740878157267118e-05,
+      "loss": 2.9408,
+      "step": 53411
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7168523073196411,
+      "learning_rate": 2.3734875784088926e-05,
+      "loss": 2.8748,
+      "step": 53412
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7598782181739807,
+      "learning_rate": 2.3728874138539856e-05,
+      "loss": 2.8996,
+      "step": 53413
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7533124089241028,
+      "learning_rate": 2.3722873220635795e-05,
+      "loss": 2.7102,
+      "step": 53414
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7148170471191406,
+      "learning_rate": 2.3716873030392504e-05,
+      "loss": 2.9118,
+      "step": 53415
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7307083606719971,
+      "learning_rate": 2.371087356782576e-05,
+      "loss": 2.8594,
+      "step": 53416
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7332124710083008,
+      "learning_rate": 2.3704874832951458e-05,
+      "loss": 3.065,
+      "step": 53417
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7090801000595093,
+      "learning_rate": 2.369887682578535e-05,
+      "loss": 2.8158,
+      "step": 53418
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7521113157272339,
+      "learning_rate": 2.3692879546343224e-05,
+      "loss": 2.92,
+      "step": 53419
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7190861105918884,
+      "learning_rate": 2.368688299464083e-05,
+      "loss": 2.8477,
+      "step": 53420
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7460111975669861,
+      "learning_rate": 2.3680887170693995e-05,
+      "loss": 2.887,
+      "step": 53421
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7339314222335815,
+      "learning_rate": 2.3674892074518602e-05,
+      "loss": 2.8122,
+      "step": 53422
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7279649972915649,
+      "learning_rate": 2.366889770613031e-05,
+      "loss": 2.9055,
+      "step": 53423
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7642202973365784,
+      "learning_rate": 2.3662904065544996e-05,
+      "loss": 2.8235,
+      "step": 53424
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.726921021938324,
+      "learning_rate": 2.3656911152778357e-05,
+      "loss": 2.9702,
+      "step": 53425
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7702478170394897,
+      "learning_rate": 2.3650918967846276e-05,
+      "loss": 2.729,
+      "step": 53426
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7660560011863708,
+      "learning_rate": 2.364492751076451e-05,
+      "loss": 3.0538,
+      "step": 53427
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7796090245246887,
+      "learning_rate": 2.3638936781548745e-05,
+      "loss": 2.9503,
+      "step": 53428
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7895865440368652,
+      "learning_rate": 2.36329467802149e-05,
+      "loss": 2.9029,
+      "step": 53429
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7662897109985352,
+      "learning_rate": 2.362695750677863e-05,
+      "loss": 2.8069,
+      "step": 53430
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7575547695159912,
+      "learning_rate": 2.362096896125576e-05,
+      "loss": 3.0073,
+      "step": 53431
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.753297746181488,
+      "learning_rate": 2.3614981143662136e-05,
+      "loss": 2.9955,
+      "step": 53432
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7785336971282959,
+      "learning_rate": 2.3608994054013452e-05,
+      "loss": 2.8248,
+      "step": 53433
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7514358758926392,
+      "learning_rate": 2.3603007692325493e-05,
+      "loss": 2.9667,
+      "step": 53434
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7518543004989624,
+      "learning_rate": 2.3597022058613946e-05,
+      "loss": 3.0098,
+      "step": 53435
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8257125616073608,
+      "learning_rate": 2.3591037152894664e-05,
+      "loss": 2.9063,
+      "step": 53436
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7336806654930115,
+      "learning_rate": 2.358505297518344e-05,
+      "loss": 2.9125,
+      "step": 53437
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7743903398513794,
+      "learning_rate": 2.3579069525495918e-05,
+      "loss": 2.756,
+      "step": 53438
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7176278233528137,
+      "learning_rate": 2.3573086803848028e-05,
+      "loss": 2.8686,
+      "step": 53439
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7256504893302917,
+      "learning_rate": 2.3567104810255323e-05,
+      "loss": 2.8576,
+      "step": 53440
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7542241215705872,
+      "learning_rate": 2.356112354473375e-05,
+      "loss": 2.9628,
+      "step": 53441
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7405779957771301,
+      "learning_rate": 2.355514300729897e-05,
+      "loss": 2.891,
+      "step": 53442
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7476978898048401,
+      "learning_rate": 2.354916319796667e-05,
+      "loss": 2.7838,
+      "step": 53443
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7352883219718933,
+      "learning_rate": 2.3543184116752766e-05,
+      "loss": 2.9383,
+      "step": 53444
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.766408383846283,
+      "learning_rate": 2.3537205763672816e-05,
+      "loss": 3.1493,
+      "step": 53445
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7481442093849182,
+      "learning_rate": 2.3531228138742675e-05,
+      "loss": 2.8828,
+      "step": 53446
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7440130710601807,
+      "learning_rate": 2.352525124197813e-05,
+      "loss": 2.9028,
+      "step": 53447
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7087607383728027,
+      "learning_rate": 2.351927507339486e-05,
+      "loss": 2.7698,
+      "step": 53448
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7708470225334167,
+      "learning_rate": 2.3513299633008597e-05,
+      "loss": 2.8684,
+      "step": 53449
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7144055962562561,
+      "learning_rate": 2.3507324920835057e-05,
+      "loss": 2.8738,
+      "step": 53450
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7633524537086487,
+      "learning_rate": 2.3501350936889995e-05,
+      "loss": 2.8275,
+      "step": 53451
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.6785310506820679,
+      "learning_rate": 2.349537768118923e-05,
+      "loss": 2.8683,
+      "step": 53452
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7202582359313965,
+      "learning_rate": 2.348940515374835e-05,
+      "loss": 2.7919,
+      "step": 53453
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7880463004112244,
+      "learning_rate": 2.3483433354583214e-05,
+      "loss": 2.7937,
+      "step": 53454
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7895649075508118,
+      "learning_rate": 2.3477462283709535e-05,
+      "loss": 2.97,
+      "step": 53455
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.741170346736908,
+      "learning_rate": 2.3471491941142972e-05,
+      "loss": 2.8845,
+      "step": 53456
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7909066081047058,
+      "learning_rate": 2.3465522326899245e-05,
+      "loss": 2.7116,
+      "step": 53457
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7080055475234985,
+      "learning_rate": 2.3459553440994107e-05,
+      "loss": 2.9053,
+      "step": 53458
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7602565884590149,
+      "learning_rate": 2.345358528344331e-05,
+      "loss": 2.9799,
+      "step": 53459
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7593586444854736,
+      "learning_rate": 2.344761785426251e-05,
+      "loss": 3.0791,
+      "step": 53460
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.760124146938324,
+      "learning_rate": 2.3441651153467532e-05,
+      "loss": 2.8708,
+      "step": 53461
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.725273847579956,
+      "learning_rate": 2.343568518107396e-05,
+      "loss": 2.902,
+      "step": 53462
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7621003985404968,
+      "learning_rate": 2.3429719937097614e-05,
+      "loss": 2.8048,
+      "step": 53463
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7457227110862732,
+      "learning_rate": 2.342375542155418e-05,
+      "loss": 3.0858,
+      "step": 53464
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7379422187805176,
+      "learning_rate": 2.3417791634459248e-05,
+      "loss": 2.7287,
+      "step": 53465
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7121053338050842,
+      "learning_rate": 2.3411828575828707e-05,
+      "loss": 2.7895,
+      "step": 53466
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7786771655082703,
+      "learning_rate": 2.340586624567814e-05,
+      "loss": 3.047,
+      "step": 53467
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7441116571426392,
+      "learning_rate": 2.339990464402327e-05,
+      "loss": 2.9813,
+      "step": 53468
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7259562611579895,
+      "learning_rate": 2.339394377087985e-05,
+      "loss": 2.8765,
+      "step": 53469
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7485330700874329,
+      "learning_rate": 2.3387983626263575e-05,
+      "loss": 2.7566,
+      "step": 53470
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7584130764007568,
+      "learning_rate": 2.338202421019012e-05,
+      "loss": 3.0932,
+      "step": 53471
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7374604940414429,
+      "learning_rate": 2.3376065522675113e-05,
+      "loss": 3.0193,
+      "step": 53472
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7430775761604309,
+      "learning_rate": 2.337010756373431e-05,
+      "loss": 3.0866,
+      "step": 53473
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7617211937904358,
+      "learning_rate": 2.336415033338346e-05,
+      "loss": 2.6868,
+      "step": 53474
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7515766620635986,
+      "learning_rate": 2.335819383163815e-05,
+      "loss": 2.9058,
+      "step": 53475
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7782991528511047,
+      "learning_rate": 2.335223805851417e-05,
+      "loss": 2.9492,
+      "step": 53476
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7339718341827393,
+      "learning_rate": 2.3346283014027144e-05,
+      "loss": 3.0604,
+      "step": 53477
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7309208512306213,
+      "learning_rate": 2.334032869819272e-05,
+      "loss": 2.9403,
+      "step": 53478
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7868647575378418,
+      "learning_rate": 2.3334375111026693e-05,
+      "loss": 2.9261,
+      "step": 53479
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7272360920906067,
+      "learning_rate": 2.332842225254461e-05,
+      "loss": 2.7392,
+      "step": 53480
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.719589114189148,
+      "learning_rate": 2.332247012276226e-05,
+      "loss": 2.952,
+      "step": 53481
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7240522503852844,
+      "learning_rate": 2.331651872169523e-05,
+      "loss": 2.7877,
+      "step": 53482
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7426586747169495,
+      "learning_rate": 2.3310568049359245e-05,
+      "loss": 2.9101,
+      "step": 53483
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7182551026344299,
+      "learning_rate": 2.3304618105770024e-05,
+      "loss": 3.0309,
+      "step": 53484
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7897670269012451,
+      "learning_rate": 2.3298668890943183e-05,
+      "loss": 2.8458,
+      "step": 53485
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7790352702140808,
+      "learning_rate": 2.3292720404894416e-05,
+      "loss": 2.6728,
+      "step": 53486
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7496531009674072,
+      "learning_rate": 2.3286772647639307e-05,
+      "loss": 3.0795,
+      "step": 53487
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7485788464546204,
+      "learning_rate": 2.3280825619193577e-05,
+      "loss": 2.8998,
+      "step": 53488
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7316558361053467,
+      "learning_rate": 2.3274879319572946e-05,
+      "loss": 2.9014,
+      "step": 53489
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.766089141368866,
+      "learning_rate": 2.3268933748792972e-05,
+      "loss": 2.9425,
+      "step": 53490
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7590373754501343,
+      "learning_rate": 2.326298890686944e-05,
+      "loss": 2.9089,
+      "step": 53491
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7415793538093567,
+      "learning_rate": 2.3257044793817904e-05,
+      "loss": 2.9765,
+      "step": 53492
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7299401760101318,
+      "learning_rate": 2.3251101409654015e-05,
+      "loss": 2.9363,
+      "step": 53493
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.714733898639679,
+      "learning_rate": 2.3245158754393534e-05,
+      "loss": 2.7415,
+      "step": 53494
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7463199496269226,
+      "learning_rate": 2.3239216828051975e-05,
+      "loss": 2.841,
+      "step": 53495
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7127196788787842,
+      "learning_rate": 2.3233275630645098e-05,
+      "loss": 2.7024,
+      "step": 53496
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7690715193748474,
+      "learning_rate": 2.3227335162188486e-05,
+      "loss": 2.9423,
+      "step": 53497
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7808687090873718,
+      "learning_rate": 2.322139542269783e-05,
+      "loss": 3.0039,
+      "step": 53498
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7283824682235718,
+      "learning_rate": 2.321545641218878e-05,
+      "loss": 2.8473,
+      "step": 53499
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.762187659740448,
+      "learning_rate": 2.3209518130676863e-05,
+      "loss": 2.9331,
+      "step": 53500
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7350523471832275,
+      "learning_rate": 2.3203580578177894e-05,
+      "loss": 3.0007,
+      "step": 53501
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7983888983726501,
+      "learning_rate": 2.319764375470733e-05,
+      "loss": 2.7056,
+      "step": 53502
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.697699248790741,
+      "learning_rate": 2.319170766028092e-05,
+      "loss": 2.7097,
+      "step": 53503
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7238744497299194,
+      "learning_rate": 2.318577229491433e-05,
+      "loss": 2.9083,
+      "step": 53504
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7800061106681824,
+      "learning_rate": 2.31798376586231e-05,
+      "loss": 2.7768,
+      "step": 53505
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7480177283287048,
+      "learning_rate": 2.317390375142296e-05,
+      "loss": 2.7823,
+      "step": 53506
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7896789312362671,
+      "learning_rate": 2.3167970573329463e-05,
+      "loss": 2.7216,
+      "step": 53507
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7196285724639893,
+      "learning_rate": 2.3162038124358196e-05,
+      "loss": 2.8438,
+      "step": 53508
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8125433325767517,
+      "learning_rate": 2.315610640452491e-05,
+      "loss": 2.9548,
+      "step": 53509
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8360536694526672,
+      "learning_rate": 2.3150175413845094e-05,
+      "loss": 2.8399,
+      "step": 53510
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7761578559875488,
+      "learning_rate": 2.3144245152334506e-05,
+      "loss": 2.8066,
+      "step": 53511
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7385191917419434,
+      "learning_rate": 2.3138315620008628e-05,
+      "loss": 2.8548,
+      "step": 53512
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7106783390045166,
+      "learning_rate": 2.313238681688322e-05,
+      "loss": 2.7003,
+      "step": 53513
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7793312668800354,
+      "learning_rate": 2.3126458742973796e-05,
+      "loss": 2.7662,
+      "step": 53514
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7528009414672852,
+      "learning_rate": 2.312053139829595e-05,
+      "loss": 3.0059,
+      "step": 53515
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7911593317985535,
+      "learning_rate": 2.31146047828654e-05,
+      "loss": 2.8624,
+      "step": 53516
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7517163157463074,
+      "learning_rate": 2.310867889669763e-05,
+      "loss": 2.8081,
+      "step": 53517
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7046359777450562,
+      "learning_rate": 2.31027537398083e-05,
+      "loss": 2.8519,
+      "step": 53518
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7641608715057373,
+      "learning_rate": 2.3096829312213093e-05,
+      "loss": 2.8228,
+      "step": 53519
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7242162227630615,
+      "learning_rate": 2.3090905613927535e-05,
+      "loss": 2.9697,
+      "step": 53520
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7436241507530212,
+      "learning_rate": 2.308498264496724e-05,
+      "loss": 2.9085,
+      "step": 53521
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7740703225135803,
+      "learning_rate": 2.307906040534777e-05,
+      "loss": 2.9504,
+      "step": 53522
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7288923859596252,
+      "learning_rate": 2.3073138895084742e-05,
+      "loss": 2.8711,
+      "step": 53523
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7444291114807129,
+      "learning_rate": 2.3067218114193842e-05,
+      "loss": 3.0138,
+      "step": 53524
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7635215520858765,
+      "learning_rate": 2.3061298062690525e-05,
+      "loss": 2.8761,
+      "step": 53525
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.790448784828186,
+      "learning_rate": 2.3055378740590514e-05,
+      "loss": 2.8546,
+      "step": 53526
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7522045969963074,
+      "learning_rate": 2.3049460147909294e-05,
+      "loss": 2.9095,
+      "step": 53527
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7432355880737305,
+      "learning_rate": 2.304354228466252e-05,
+      "loss": 3.1744,
+      "step": 53528
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.759417712688446,
+      "learning_rate": 2.303762515086578e-05,
+      "loss": 2.956,
+      "step": 53529
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7882241606712341,
+      "learning_rate": 2.3031708746534562e-05,
+      "loss": 2.8042,
+      "step": 53530
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7573877573013306,
+      "learning_rate": 2.3025793071684585e-05,
+      "loss": 2.7942,
+      "step": 53531
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8343915343284607,
+      "learning_rate": 2.3019878126331304e-05,
+      "loss": 2.9512,
+      "step": 53532
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7674956917762756,
+      "learning_rate": 2.301396391049034e-05,
+      "loss": 2.8093,
+      "step": 53533
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.714229941368103,
+      "learning_rate": 2.3008050424177384e-05,
+      "loss": 2.9264,
+      "step": 53534
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.780677318572998,
+      "learning_rate": 2.3002137667407884e-05,
+      "loss": 3.0235,
+      "step": 53535
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7305729985237122,
+      "learning_rate": 2.2996225640197463e-05,
+      "loss": 2.7816,
+      "step": 53536
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.773148775100708,
+      "learning_rate": 2.2990314342561612e-05,
+      "loss": 3.1407,
+      "step": 53537
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7465836405754089,
+      "learning_rate": 2.298440377451598e-05,
+      "loss": 2.9568,
+      "step": 53538
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8020625710487366,
+      "learning_rate": 2.2978493936076125e-05,
+      "loss": 2.9573,
+      "step": 53539
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7247298955917358,
+      "learning_rate": 2.2972584827257568e-05,
+      "loss": 2.9254,
+      "step": 53540
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7181655168533325,
+      "learning_rate": 2.296667644807596e-05,
+      "loss": 2.895,
+      "step": 53541
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7766932249069214,
+      "learning_rate": 2.2960768798546824e-05,
+      "loss": 2.7412,
+      "step": 53542
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7357663512229919,
+      "learning_rate": 2.2954861878685684e-05,
+      "loss": 2.7664,
+      "step": 53543
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7322977185249329,
+      "learning_rate": 2.2948955688508085e-05,
+      "loss": 2.931,
+      "step": 53544
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7766660451889038,
+      "learning_rate": 2.294305022802959e-05,
+      "loss": 2.9752,
+      "step": 53545
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7280550003051758,
+      "learning_rate": 2.293714549726585e-05,
+      "loss": 2.6769,
+      "step": 53546
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.726729154586792,
+      "learning_rate": 2.293124149623232e-05,
+      "loss": 3.118,
+      "step": 53547
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.750844419002533,
+      "learning_rate": 2.2925338224944588e-05,
+      "loss": 2.9707,
+      "step": 53548
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7047064304351807,
+      "learning_rate": 2.2919435683418175e-05,
+      "loss": 2.7327,
+      "step": 53549
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7407240867614746,
+      "learning_rate": 2.2913533871668667e-05,
+      "loss": 2.7221,
+      "step": 53550
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7109093070030212,
+      "learning_rate": 2.2907632789711617e-05,
+      "loss": 3.0363,
+      "step": 53551
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7659242749214172,
+      "learning_rate": 2.2901732437562447e-05,
+      "loss": 2.7527,
+      "step": 53552
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7269027829170227,
+      "learning_rate": 2.2895832815236848e-05,
+      "loss": 2.7867,
+      "step": 53553
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.739622950553894,
+      "learning_rate": 2.288993392275027e-05,
+      "loss": 2.8092,
+      "step": 53554
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7424209713935852,
+      "learning_rate": 2.2884035760118235e-05,
+      "loss": 2.8659,
+      "step": 53555
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7513794302940369,
+      "learning_rate": 2.28781383273564e-05,
+      "loss": 2.9606,
+      "step": 53556
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7400728464126587,
+      "learning_rate": 2.287224162448018e-05,
+      "loss": 2.8118,
+      "step": 53557
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7726714015007019,
+      "learning_rate": 2.2866345651505168e-05,
+      "loss": 3.0265,
+      "step": 53558
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7855329513549805,
+      "learning_rate": 2.2860450408446818e-05,
+      "loss": 2.9305,
+      "step": 53559
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7705275416374207,
+      "learning_rate": 2.2854555895320713e-05,
+      "loss": 2.7834,
+      "step": 53560
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7320125699043274,
+      "learning_rate": 2.2848662112142412e-05,
+      "loss": 2.9377,
+      "step": 53561
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7487272024154663,
+      "learning_rate": 2.2842769058927336e-05,
+      "loss": 3.0346,
+      "step": 53562
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7231678366661072,
+      "learning_rate": 2.2836876735691102e-05,
+      "loss": 2.7278,
+      "step": 53563
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8296177387237549,
+      "learning_rate": 2.2830985142449164e-05,
+      "loss": 2.9324,
+      "step": 53564
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7314631342887878,
+      "learning_rate": 2.2825094279217116e-05,
+      "loss": 2.9992,
+      "step": 53565
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7270010113716125,
+      "learning_rate": 2.281920414601044e-05,
+      "loss": 2.9121,
+      "step": 53566
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7274070382118225,
+      "learning_rate": 2.2813314742844556e-05,
+      "loss": 3.0206,
+      "step": 53567
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7393171787261963,
+      "learning_rate": 2.2807426069735124e-05,
+      "loss": 2.9025,
+      "step": 53568
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7126226425170898,
+      "learning_rate": 2.2801538126697527e-05,
+      "loss": 2.7961,
+      "step": 53569
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7575780153274536,
+      "learning_rate": 2.2795650913747355e-05,
+      "loss": 3.0125,
+      "step": 53570
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7629463076591492,
+      "learning_rate": 2.2789764430900126e-05,
+      "loss": 2.834,
+      "step": 53571
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7294546961784363,
+      "learning_rate": 2.2783878678171295e-05,
+      "loss": 2.7511,
+      "step": 53572
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7730554938316345,
+      "learning_rate": 2.2777993655576388e-05,
+      "loss": 2.8679,
+      "step": 53573
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7737736105918884,
+      "learning_rate": 2.277210936313082e-05,
+      "loss": 2.808,
+      "step": 53574
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7201272249221802,
+      "learning_rate": 2.276622580085018e-05,
+      "loss": 3.0931,
+      "step": 53575
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7674169540405273,
+      "learning_rate": 2.2760342968750024e-05,
+      "loss": 2.6019,
+      "step": 53576
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.6896553039550781,
+      "learning_rate": 2.275446086684567e-05,
+      "loss": 2.7181,
+      "step": 53577
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7537925839424133,
+      "learning_rate": 2.2748579495152808e-05,
+      "loss": 3.0731,
+      "step": 53578
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7853711247444153,
+      "learning_rate": 2.274269885368679e-05,
+      "loss": 3.0916,
+      "step": 53579
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8356667160987854,
+      "learning_rate": 2.273681894246311e-05,
+      "loss": 2.9315,
+      "step": 53580
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7744783759117126,
+      "learning_rate": 2.2730939761497347e-05,
+      "loss": 2.62,
+      "step": 53581
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7801983952522278,
+      "learning_rate": 2.272506131080486e-05,
+      "loss": 2.8224,
+      "step": 53582
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.737760066986084,
+      "learning_rate": 2.2719183590401237e-05,
+      "loss": 3.0579,
+      "step": 53583
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7177762389183044,
+      "learning_rate": 2.2713306600301894e-05,
+      "loss": 2.8598,
+      "step": 53584
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7965755462646484,
+      "learning_rate": 2.2707430340522326e-05,
+      "loss": 2.925,
+      "step": 53585
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7102624177932739,
+      "learning_rate": 2.2701554811078083e-05,
+      "loss": 2.8057,
+      "step": 53586
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8156418800354004,
+      "learning_rate": 2.2695680011984583e-05,
+      "loss": 2.8195,
+      "step": 53587
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7690003514289856,
+      "learning_rate": 2.2689805943257257e-05,
+      "loss": 2.8399,
+      "step": 53588
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7723907828330994,
+      "learning_rate": 2.268393260491158e-05,
+      "loss": 3.0257,
+      "step": 53589
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7428565621376038,
+      "learning_rate": 2.267805999696308e-05,
+      "loss": 2.9596,
+      "step": 53590
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7959504127502441,
+      "learning_rate": 2.2672188119427214e-05,
+      "loss": 2.9929,
+      "step": 53591
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7445618510246277,
+      "learning_rate": 2.2666316972319397e-05,
+      "loss": 2.9063,
+      "step": 53592
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7014729976654053,
+      "learning_rate": 2.266044655565519e-05,
+      "loss": 2.7315,
+      "step": 53593
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7331639528274536,
+      "learning_rate": 2.2654576869449968e-05,
+      "loss": 3.0117,
+      "step": 53594
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7533635497093201,
+      "learning_rate": 2.2648707913719165e-05,
+      "loss": 2.9198,
+      "step": 53595
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7588867545127869,
+      "learning_rate": 2.2642839688478364e-05,
+      "loss": 2.9348,
+      "step": 53596
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.669175922870636,
+      "learning_rate": 2.2636972193742887e-05,
+      "loss": 2.9407,
+      "step": 53597
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7316837310791016,
+      "learning_rate": 2.2631105429528284e-05,
+      "loss": 2.952,
+      "step": 53598
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.767562210559845,
+      "learning_rate": 2.2625239395849916e-05,
+      "loss": 2.943,
+      "step": 53599
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7083771228790283,
+      "learning_rate": 2.261937409272333e-05,
+      "loss": 2.9877,
+      "step": 53600
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7206432223320007,
+      "learning_rate": 2.2613509520163954e-05,
+      "loss": 2.9687,
+      "step": 53601
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7313346862792969,
+      "learning_rate": 2.2607645678187138e-05,
+      "loss": 2.8454,
+      "step": 53602
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.713829517364502,
+      "learning_rate": 2.260178256680847e-05,
+      "loss": 3.1016,
+      "step": 53603
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.695851743221283,
+      "learning_rate": 2.2595920186043238e-05,
+      "loss": 2.5597,
+      "step": 53604
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7535444498062134,
+      "learning_rate": 2.2590058535906996e-05,
+      "loss": 2.924,
+      "step": 53605
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7128457427024841,
+      "learning_rate": 2.25841976164152e-05,
+      "loss": 2.8773,
+      "step": 53606
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7772616147994995,
+      "learning_rate": 2.2578337427583172e-05,
+      "loss": 3.1488,
+      "step": 53607
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.727367103099823,
+      "learning_rate": 2.2572477969426462e-05,
+      "loss": 2.798,
+      "step": 53608
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7210254073143005,
+      "learning_rate": 2.256661924196046e-05,
+      "loss": 3.0219,
+      "step": 53609
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7165561318397522,
+      "learning_rate": 2.256076124520052e-05,
+      "loss": 2.7132,
+      "step": 53610
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7372732162475586,
+      "learning_rate": 2.255490397916223e-05,
+      "loss": 2.8915,
+      "step": 53611
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7609841227531433,
+      "learning_rate": 2.2549047443860845e-05,
+      "loss": 3.088,
+      "step": 53612
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.744179904460907,
+      "learning_rate": 2.2543191639311954e-05,
+      "loss": 3.0776,
+      "step": 53613
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7504658102989197,
+      "learning_rate": 2.2537336565530807e-05,
+      "loss": 2.9734,
+      "step": 53614
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.6780826449394226,
+      "learning_rate": 2.2531482222532994e-05,
+      "loss": 2.9569,
+      "step": 53615
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7250926494598389,
+      "learning_rate": 2.2525628610333835e-05,
+      "loss": 2.8901,
+      "step": 53616
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7274686694145203,
+      "learning_rate": 2.2519775728948754e-05,
+      "loss": 2.8393,
+      "step": 53617
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7236843109130859,
+      "learning_rate": 2.25139235783932e-05,
+      "loss": 2.6181,
+      "step": 53618
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7012773752212524,
+      "learning_rate": 2.2508072158682535e-05,
+      "loss": 2.9166,
+      "step": 53619
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7341805696487427,
+      "learning_rate": 2.2502221469832273e-05,
+      "loss": 2.7775,
+      "step": 53620
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7430617809295654,
+      "learning_rate": 2.249637151185767e-05,
+      "loss": 2.943,
+      "step": 53621
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7874128222465515,
+      "learning_rate": 2.249052228477428e-05,
+      "loss": 2.9523,
+      "step": 53622
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8702890276908875,
+      "learning_rate": 2.248467378859743e-05,
+      "loss": 2.6836,
+      "step": 53623
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7678395509719849,
+      "learning_rate": 2.24788260233425e-05,
+      "loss": 2.9465,
+      "step": 53624
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7276766300201416,
+      "learning_rate": 2.247297898902498e-05,
+      "loss": 2.8315,
+      "step": 53625
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7727230191230774,
+      "learning_rate": 2.2467132685660196e-05,
+      "loss": 3.0265,
+      "step": 53626
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7668293714523315,
+      "learning_rate": 2.246128711326356e-05,
+      "loss": 2.9094,
+      "step": 53627
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7255916595458984,
+      "learning_rate": 2.2455442271850532e-05,
+      "loss": 2.8169,
+      "step": 53628
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7411938905715942,
+      "learning_rate": 2.24495981614364e-05,
+      "loss": 2.9508,
+      "step": 53629
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7330008745193481,
+      "learning_rate": 2.2443754782036648e-05,
+      "loss": 2.8352,
+      "step": 53630
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7218151092529297,
+      "learning_rate": 2.2437912133666634e-05,
+      "loss": 2.706,
+      "step": 53631
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7262304425239563,
+      "learning_rate": 2.2432070216341715e-05,
+      "loss": 2.8294,
+      "step": 53632
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7365819215774536,
+      "learning_rate": 2.2426229030077335e-05,
+      "loss": 2.846,
+      "step": 53633
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7333903312683105,
+      "learning_rate": 2.2420388574888793e-05,
+      "loss": 3.0362,
+      "step": 53634
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7654792666435242,
+      "learning_rate": 2.241454885079157e-05,
+      "loss": 2.7503,
+      "step": 53635
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7742012739181519,
+      "learning_rate": 2.2408709857800988e-05,
+      "loss": 2.7352,
+      "step": 53636
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7555630803108215,
+      "learning_rate": 2.240287159593247e-05,
+      "loss": 2.9539,
+      "step": 53637
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7577449083328247,
+      "learning_rate": 2.2397034065201336e-05,
+      "loss": 2.9425,
+      "step": 53638
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7749025821685791,
+      "learning_rate": 2.2391197265622975e-05,
+      "loss": 3.015,
+      "step": 53639
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7856140732765198,
+      "learning_rate": 2.2385361197212805e-05,
+      "loss": 2.7987,
+      "step": 53640
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7463555932044983,
+      "learning_rate": 2.2379525859986148e-05,
+      "loss": 2.8118,
+      "step": 53641
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7431473135948181,
+      "learning_rate": 2.2373691253958358e-05,
+      "loss": 2.8837,
+      "step": 53642
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7310684323310852,
+      "learning_rate": 2.2367857379144892e-05,
+      "loss": 2.8888,
+      "step": 53643
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7498796582221985,
+      "learning_rate": 2.236202423556107e-05,
+      "loss": 2.9496,
+      "step": 53644
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.740679919719696,
+      "learning_rate": 2.2356191823222214e-05,
+      "loss": 3.0156,
+      "step": 53645
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7141246199607849,
+      "learning_rate": 2.235036014214371e-05,
+      "loss": 2.8782,
+      "step": 53646
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7410897612571716,
+      "learning_rate": 2.2344529192340876e-05,
+      "loss": 2.6084,
+      "step": 53647
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7512717247009277,
+      "learning_rate": 2.2338698973829207e-05,
+      "loss": 2.8707,
+      "step": 53648
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7618011832237244,
+      "learning_rate": 2.2332869486623884e-05,
+      "loss": 2.8932,
+      "step": 53649
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8862194418907166,
+      "learning_rate": 2.2327040730740397e-05,
+      "loss": 2.8414,
+      "step": 53650
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7280387878417969,
+      "learning_rate": 2.2321212706194037e-05,
+      "loss": 2.7621,
+      "step": 53651
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.714809000492096,
+      "learning_rate": 2.2315385413000152e-05,
+      "loss": 2.623,
+      "step": 53652
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7935085892677307,
+      "learning_rate": 2.2309558851174138e-05,
+      "loss": 2.7519,
+      "step": 53653
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.736251711845398,
+      "learning_rate": 2.230373302073124e-05,
+      "loss": 2.8681,
+      "step": 53654
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7445536851882935,
+      "learning_rate": 2.2297907921686953e-05,
+      "loss": 3.1613,
+      "step": 53655
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7414441108703613,
+      "learning_rate": 2.2292083554056427e-05,
+      "loss": 2.8987,
+      "step": 53656
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7777810096740723,
+      "learning_rate": 2.2286259917855152e-05,
+      "loss": 2.6787,
+      "step": 53657
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7500730156898499,
+      "learning_rate": 2.2280437013098452e-05,
+      "loss": 2.768,
+      "step": 53658
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7526910901069641,
+      "learning_rate": 2.2274614839801607e-05,
+      "loss": 2.9552,
+      "step": 53659
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7558931112289429,
+      "learning_rate": 2.2268793397980013e-05,
+      "loss": 3.0392,
+      "step": 53660
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7288579344749451,
+      "learning_rate": 2.2262972687648916e-05,
+      "loss": 2.9854,
+      "step": 53661
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7054191827774048,
+      "learning_rate": 2.225715270882368e-05,
+      "loss": 2.9694,
+      "step": 53662
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7429760098457336,
+      "learning_rate": 2.2251333461519717e-05,
+      "loss": 2.8739,
+      "step": 53663
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8079425692558289,
+      "learning_rate": 2.224551494575222e-05,
+      "loss": 2.8049,
+      "step": 53664
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7400147318840027,
+      "learning_rate": 2.2239697161536642e-05,
+      "loss": 2.9006,
+      "step": 53665
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7143925428390503,
+      "learning_rate": 2.223388010888817e-05,
+      "loss": 2.6057,
+      "step": 53666
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7217711806297302,
+      "learning_rate": 2.2228063787822292e-05,
+      "loss": 2.896,
+      "step": 53667
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8171392679214478,
+      "learning_rate": 2.2222248198354198e-05,
+      "loss": 3.0967,
+      "step": 53668
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7449054718017578,
+      "learning_rate": 2.221643334049924e-05,
+      "loss": 2.9392,
+      "step": 53669
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7087463140487671,
+      "learning_rate": 2.221061921427274e-05,
+      "loss": 2.9487,
+      "step": 53670
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7898861169815063,
+      "learning_rate": 2.2204805819689987e-05,
+      "loss": 2.8475,
+      "step": 53671
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.762144923210144,
+      "learning_rate": 2.21989931567663e-05,
+      "loss": 2.948,
+      "step": 53672
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7354418635368347,
+      "learning_rate": 2.2193181225517065e-05,
+      "loss": 2.6357,
+      "step": 53673
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7770733833312988,
+      "learning_rate": 2.2187370025957506e-05,
+      "loss": 2.9122,
+      "step": 53674
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7444289922714233,
+      "learning_rate": 2.2181559558102947e-05,
+      "loss": 2.8764,
+      "step": 53675
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7131602764129639,
+      "learning_rate": 2.217574982196867e-05,
+      "loss": 2.8945,
+      "step": 53676
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8043651580810547,
+      "learning_rate": 2.2169940817569965e-05,
+      "loss": 2.7772,
+      "step": 53677
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8861671686172485,
+      "learning_rate": 2.216413254492222e-05,
+      "loss": 2.8719,
+      "step": 53678
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.782197892665863,
+      "learning_rate": 2.2158325004040655e-05,
+      "loss": 2.927,
+      "step": 53679
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7546136379241943,
+      "learning_rate": 2.2152518194940626e-05,
+      "loss": 2.9038,
+      "step": 53680
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7524154186248779,
+      "learning_rate": 2.2146712117637388e-05,
+      "loss": 2.8427,
+      "step": 53681
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7260668277740479,
+      "learning_rate": 2.2140906772146195e-05,
+      "loss": 2.9093,
+      "step": 53682
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7640723586082458,
+      "learning_rate": 2.2135102158482432e-05,
+      "loss": 2.7328,
+      "step": 53683
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7455698251724243,
+      "learning_rate": 2.2129298276661255e-05,
+      "loss": 2.8891,
+      "step": 53684
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7532725930213928,
+      "learning_rate": 2.212349512669812e-05,
+      "loss": 2.9091,
+      "step": 53685
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7638143301010132,
+      "learning_rate": 2.2117692708608148e-05,
+      "loss": 3.1224,
+      "step": 53686
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7833783626556396,
+      "learning_rate": 2.211189102240669e-05,
+      "loss": 2.9242,
+      "step": 53687
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7878255844116211,
+      "learning_rate": 2.2106090068109107e-05,
+      "loss": 2.9914,
+      "step": 53688
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7649726867675781,
+      "learning_rate": 2.210028984573058e-05,
+      "loss": 2.6449,
+      "step": 53689
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7419236898422241,
+      "learning_rate": 2.2094490355286397e-05,
+      "loss": 3.0778,
+      "step": 53690
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7762550115585327,
+      "learning_rate": 2.2088691596791784e-05,
+      "loss": 2.869,
+      "step": 53691
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7269838452339172,
+      "learning_rate": 2.2082893570262163e-05,
+      "loss": 2.8345,
+      "step": 53692
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7062630653381348,
+      "learning_rate": 2.2077096275712648e-05,
+      "loss": 2.7138,
+      "step": 53693
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7006743550300598,
+      "learning_rate": 2.2071299713158564e-05,
+      "loss": 2.9197,
+      "step": 53694
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.738308846950531,
+      "learning_rate": 2.206550388261523e-05,
+      "loss": 2.6824,
+      "step": 53695
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8569402694702148,
+      "learning_rate": 2.2059708784097874e-05,
+      "loss": 3.0809,
+      "step": 53696
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7651105523109436,
+      "learning_rate": 2.2053914417621776e-05,
+      "loss": 2.6283,
+      "step": 53697
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7134808897972107,
+      "learning_rate": 2.2048120783202128e-05,
+      "loss": 2.8568,
+      "step": 53698
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7216563820838928,
+      "learning_rate": 2.2042327880854215e-05,
+      "loss": 2.7479,
+      "step": 53699
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7495296597480774,
+      "learning_rate": 2.2036535710593362e-05,
+      "loss": 3.0511,
+      "step": 53700
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7457597255706787,
+      "learning_rate": 2.2030744272434754e-05,
+      "loss": 2.8722,
+      "step": 53701
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.76603764295578,
+      "learning_rate": 2.2024953566393678e-05,
+      "loss": 3.0893,
+      "step": 53702
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7615640163421631,
+      "learning_rate": 2.201916359248542e-05,
+      "loss": 3.0252,
+      "step": 53703
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.765140950679779,
+      "learning_rate": 2.201337435072511e-05,
+      "loss": 2.9289,
+      "step": 53704
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7220942974090576,
+      "learning_rate": 2.200758584112813e-05,
+      "loss": 2.8011,
+      "step": 53705
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7161285281181335,
+      "learning_rate": 2.2001798063709632e-05,
+      "loss": 2.7574,
+      "step": 53706
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7688199281692505,
+      "learning_rate": 2.1996011018484936e-05,
+      "loss": 3.0325,
+      "step": 53707
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7416340708732605,
+      "learning_rate": 2.199022470546917e-05,
+      "loss": 2.8929,
+      "step": 53708
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7346240282058716,
+      "learning_rate": 2.198443912467769e-05,
+      "loss": 2.9675,
+      "step": 53709
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7140522599220276,
+      "learning_rate": 2.1978654276125708e-05,
+      "loss": 2.9636,
+      "step": 53710
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.761841356754303,
+      "learning_rate": 2.1972870159828482e-05,
+      "loss": 2.8097,
+      "step": 53711
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7178420424461365,
+      "learning_rate": 2.196708677580117e-05,
+      "loss": 2.9565,
+      "step": 53712
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7384170889854431,
+      "learning_rate": 2.196130412405902e-05,
+      "loss": 2.9932,
+      "step": 53713
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.742639422416687,
+      "learning_rate": 2.1955522204617258e-05,
+      "loss": 2.967,
+      "step": 53714
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7487492561340332,
+      "learning_rate": 2.1949741017491206e-05,
+      "loss": 2.8679,
+      "step": 53715
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7340525984764099,
+      "learning_rate": 2.1943960562695983e-05,
+      "loss": 2.7455,
+      "step": 53716
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7590996623039246,
+      "learning_rate": 2.193818084024691e-05,
+      "loss": 2.8737,
+      "step": 53717
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7017925977706909,
+      "learning_rate": 2.1932401850159142e-05,
+      "loss": 2.7985,
+      "step": 53718
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7645330429077148,
+      "learning_rate": 2.1926623592447834e-05,
+      "loss": 2.939,
+      "step": 53719
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7296159863471985,
+      "learning_rate": 2.1920846067128372e-05,
+      "loss": 2.9681,
+      "step": 53720
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7412862181663513,
+      "learning_rate": 2.1915069274215812e-05,
+      "loss": 2.8342,
+      "step": 53721
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7657079100608826,
+      "learning_rate": 2.190929321372551e-05,
+      "loss": 3.0628,
+      "step": 53722
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7571131587028503,
+      "learning_rate": 2.1903517885672552e-05,
+      "loss": 2.7669,
+      "step": 53723
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7139904499053955,
+      "learning_rate": 2.1897743290072257e-05,
+      "loss": 2.8135,
+      "step": 53724
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7659240365028381,
+      "learning_rate": 2.1891969426939783e-05,
+      "loss": 2.8425,
+      "step": 53725
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7518433928489685,
+      "learning_rate": 2.188619629629028e-05,
+      "loss": 2.9292,
+      "step": 53726
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8315146565437317,
+      "learning_rate": 2.1880423898139078e-05,
+      "loss": 2.921,
+      "step": 53727
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7377954125404358,
+      "learning_rate": 2.1874652232501255e-05,
+      "loss": 2.7456,
+      "step": 53728
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7297799587249756,
+      "learning_rate": 2.186888129939207e-05,
+      "loss": 2.5994,
+      "step": 53729
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7472618818283081,
+      "learning_rate": 2.186311109882678e-05,
+      "loss": 2.7483,
+      "step": 53730
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7658689618110657,
+      "learning_rate": 2.1857341630820503e-05,
+      "loss": 2.6907,
+      "step": 53731
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7699265480041504,
+      "learning_rate": 2.185157289538846e-05,
+      "loss": 2.9277,
+      "step": 53732
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8159839510917664,
+      "learning_rate": 2.1845804892545872e-05,
+      "loss": 2.8331,
+      "step": 53733
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7433233857154846,
+      "learning_rate": 2.1840037622307827e-05,
+      "loss": 2.7793,
+      "step": 53734
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7243334054946899,
+      "learning_rate": 2.183427108468968e-05,
+      "loss": 2.8757,
+      "step": 53735
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7512285709381104,
+      "learning_rate": 2.1828505279706454e-05,
+      "loss": 2.91,
+      "step": 53736
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.761920690536499,
+      "learning_rate": 2.182274020737347e-05,
+      "loss": 2.9553,
+      "step": 53737
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7395731210708618,
+      "learning_rate": 2.181697586770578e-05,
+      "loss": 2.8619,
+      "step": 53738
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7716213464736938,
+      "learning_rate": 2.18112122607187e-05,
+      "loss": 2.9312,
+      "step": 53739
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7732451558113098,
+      "learning_rate": 2.1805449386427364e-05,
+      "loss": 2.8891,
+      "step": 53740
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7086454629898071,
+      "learning_rate": 2.1799687244846852e-05,
+      "loss": 3.0304,
+      "step": 53741
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7271209359169006,
+      "learning_rate": 2.1793925835992486e-05,
+      "loss": 2.7164,
+      "step": 53742
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7848542332649231,
+      "learning_rate": 2.1788165159879356e-05,
+      "loss": 2.7783,
+      "step": 53743
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7181041240692139,
+      "learning_rate": 2.1782405216522615e-05,
+      "loss": 2.8273,
+      "step": 53744
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7394196391105652,
+      "learning_rate": 2.177664600593755e-05,
+      "loss": 2.9158,
+      "step": 53745
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7684555053710938,
+      "learning_rate": 2.1770887528139245e-05,
+      "loss": 2.8852,
+      "step": 53746
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.733661413192749,
+      "learning_rate": 2.1765129783142897e-05,
+      "loss": 2.8123,
+      "step": 53747
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7537463307380676,
+      "learning_rate": 2.1759372770963558e-05,
+      "loss": 2.9477,
+      "step": 53748
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.752604067325592,
+      "learning_rate": 2.175361649161651e-05,
+      "loss": 2.8565,
+      "step": 53749
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8050605654716492,
+      "learning_rate": 2.1747860945116946e-05,
+      "loss": 2.6993,
+      "step": 53750
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7391548156738281,
+      "learning_rate": 2.1742106131479886e-05,
+      "loss": 2.7675,
+      "step": 53751
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7486521005630493,
+      "learning_rate": 2.173635205072065e-05,
+      "loss": 2.7678,
+      "step": 53752
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7553504705429077,
+      "learning_rate": 2.173059870285423e-05,
+      "loss": 2.8386,
+      "step": 53753
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8861485123634338,
+      "learning_rate": 2.1724846087895943e-05,
+      "loss": 2.651,
+      "step": 53754
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7568973898887634,
+      "learning_rate": 2.171909420586081e-05,
+      "loss": 2.9888,
+      "step": 53755
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7912908792495728,
+      "learning_rate": 2.171334305676402e-05,
+      "loss": 2.9017,
+      "step": 53756
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7801294326782227,
+      "learning_rate": 2.1707592640620762e-05,
+      "loss": 2.8552,
+      "step": 53757
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.749573290348053,
+      "learning_rate": 2.1701842957446092e-05,
+      "loss": 2.9134,
+      "step": 53758
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.778415322303772,
+      "learning_rate": 2.1696094007255194e-05,
+      "loss": 2.8841,
+      "step": 53759
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7514265179634094,
+      "learning_rate": 2.169034579006329e-05,
+      "loss": 3.0026,
+      "step": 53760
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7774766683578491,
+      "learning_rate": 2.168459830588547e-05,
+      "loss": 2.7802,
+      "step": 53761
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7230890393257141,
+      "learning_rate": 2.167885155473682e-05,
+      "loss": 2.9125,
+      "step": 53762
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7347655892372131,
+      "learning_rate": 2.1673105536632497e-05,
+      "loss": 3.0178,
+      "step": 53763
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7173894643783569,
+      "learning_rate": 2.166736025158762e-05,
+      "loss": 3.0613,
+      "step": 53764
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7474822402000427,
+      "learning_rate": 2.1661615699617407e-05,
+      "loss": 2.8945,
+      "step": 53765
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7168195247650146,
+      "learning_rate": 2.1655871880736887e-05,
+      "loss": 2.8247,
+      "step": 53766
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7810943722724915,
+      "learning_rate": 2.1650128794961276e-05,
+      "loss": 2.7536,
+      "step": 53767
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7393297553062439,
+      "learning_rate": 2.1644386442305627e-05,
+      "loss": 2.957,
+      "step": 53768
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.758391797542572,
+      "learning_rate": 2.1638644822785166e-05,
+      "loss": 2.6468,
+      "step": 53769
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.764182448387146,
+      "learning_rate": 2.1632903936414848e-05,
+      "loss": 2.9961,
+      "step": 53770
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7735395431518555,
+      "learning_rate": 2.162716378320989e-05,
+      "loss": 2.9979,
+      "step": 53771
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7840405702590942,
+      "learning_rate": 2.162142436318548e-05,
+      "loss": 2.9104,
+      "step": 53772
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7420419454574585,
+      "learning_rate": 2.1615685676356577e-05,
+      "loss": 3.1512,
+      "step": 53773
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7456751465797424,
+      "learning_rate": 2.160994772273846e-05,
+      "loss": 2.9972,
+      "step": 53774
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7378044128417969,
+      "learning_rate": 2.16042105023461e-05,
+      "loss": 2.7904,
+      "step": 53775
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.744476318359375,
+      "learning_rate": 2.1598474015194733e-05,
+      "loss": 2.807,
+      "step": 53776
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7326436638832092,
+      "learning_rate": 2.1592738261299393e-05,
+      "loss": 2.8704,
+      "step": 53777
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.752892792224884,
+      "learning_rate": 2.158700324067516e-05,
+      "loss": 2.941,
+      "step": 53778
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7676821947097778,
+      "learning_rate": 2.1581268953337227e-05,
+      "loss": 2.8962,
+      "step": 53779
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.788492739200592,
+      "learning_rate": 2.157553539930058e-05,
+      "loss": 3.0261,
+      "step": 53780
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7496923804283142,
+      "learning_rate": 2.1569802578580407e-05,
+      "loss": 3.1289,
+      "step": 53781
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7280400395393372,
+      "learning_rate": 2.156407049119183e-05,
+      "loss": 3.0553,
+      "step": 53782
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8010007739067078,
+      "learning_rate": 2.1558339137149904e-05,
+      "loss": 2.7921,
+      "step": 53783
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7570450305938721,
+      "learning_rate": 2.155260851646975e-05,
+      "loss": 3.0598,
+      "step": 53784
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7340518832206726,
+      "learning_rate": 2.1546878629166355e-05,
+      "loss": 2.8535,
+      "step": 53785
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7815223336219788,
+      "learning_rate": 2.1541149475254903e-05,
+      "loss": 3.059,
+      "step": 53786
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7372803688049316,
+      "learning_rate": 2.1535421054750558e-05,
+      "loss": 3.0539,
+      "step": 53787
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7437090277671814,
+      "learning_rate": 2.1529693367668233e-05,
+      "loss": 3.0135,
+      "step": 53788
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7286654710769653,
+      "learning_rate": 2.152396641402315e-05,
+      "loss": 2.7342,
+      "step": 53789
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7460008263587952,
+      "learning_rate": 2.1518240193830306e-05,
+      "loss": 3.0216,
+      "step": 53790
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7398113012313843,
+      "learning_rate": 2.151251470710488e-05,
+      "loss": 3.0744,
+      "step": 53791
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7812997698783875,
+      "learning_rate": 2.150678995386189e-05,
+      "loss": 2.8752,
+      "step": 53792
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7865747809410095,
+      "learning_rate": 2.1501065934116368e-05,
+      "loss": 2.8237,
+      "step": 53793
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7350339293479919,
+      "learning_rate": 2.1495342647883496e-05,
+      "loss": 2.929,
+      "step": 53794
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7692163586616516,
+      "learning_rate": 2.148962009517823e-05,
+      "loss": 2.9889,
+      "step": 53795
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7704609036445618,
+      "learning_rate": 2.1483898276015688e-05,
+      "loss": 2.9512,
+      "step": 53796
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7423082590103149,
+      "learning_rate": 2.147817719041103e-05,
+      "loss": 2.7966,
+      "step": 53797
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7098038196563721,
+      "learning_rate": 2.147245683837927e-05,
+      "loss": 3.0031,
+      "step": 53798
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7264339923858643,
+      "learning_rate": 2.1466737219935405e-05,
+      "loss": 2.8246,
+      "step": 53799
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7618200778961182,
+      "learning_rate": 2.1461018335094548e-05,
+      "loss": 3.1037,
+      "step": 53800
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.742719292640686,
+      "learning_rate": 2.1455300183871728e-05,
+      "loss": 2.9129,
+      "step": 53801
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8027572631835938,
+      "learning_rate": 2.1449582766282125e-05,
+      "loss": 3.0597,
+      "step": 53802
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7896417379379272,
+      "learning_rate": 2.1443866082340632e-05,
+      "loss": 3.015,
+      "step": 53803
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7622906565666199,
+      "learning_rate": 2.1438150132062436e-05,
+      "loss": 3.051,
+      "step": 53804
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7213104367256165,
+      "learning_rate": 2.143243491546256e-05,
+      "loss": 2.6144,
+      "step": 53805
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.754871666431427,
+      "learning_rate": 2.142672043255599e-05,
+      "loss": 2.8604,
+      "step": 53806
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7541650533676147,
+      "learning_rate": 2.1421006683357844e-05,
+      "loss": 2.7727,
+      "step": 53807
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7695738673210144,
+      "learning_rate": 2.141529366788315e-05,
+      "loss": 2.9745,
+      "step": 53808
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7393126487731934,
+      "learning_rate": 2.140958138614699e-05,
+      "loss": 2.868,
+      "step": 53809
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7608216404914856,
+      "learning_rate": 2.140386983816429e-05,
+      "loss": 2.8814,
+      "step": 53810
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7869523763656616,
+      "learning_rate": 2.1398159023950235e-05,
+      "loss": 2.9099,
+      "step": 53811
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7133737206459045,
+      "learning_rate": 2.1392448943519812e-05,
+      "loss": 2.9117,
+      "step": 53812
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7602190375328064,
+      "learning_rate": 2.1386739596888116e-05,
+      "loss": 2.8612,
+      "step": 53813
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7913804650306702,
+      "learning_rate": 2.138103098407009e-05,
+      "loss": 2.9442,
+      "step": 53814
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7600404620170593,
+      "learning_rate": 2.1375323105080768e-05,
+      "loss": 2.8422,
+      "step": 53815
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7508252263069153,
+      "learning_rate": 2.1369615959935192e-05,
+      "loss": 2.6938,
+      "step": 53816
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7605404853820801,
+      "learning_rate": 2.1363909548648527e-05,
+      "loss": 2.6986,
+      "step": 53817
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7708483934402466,
+      "learning_rate": 2.1358203871235623e-05,
+      "loss": 2.7698,
+      "step": 53818
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7216460704803467,
+      "learning_rate": 2.1352498927711603e-05,
+      "loss": 2.9233,
+      "step": 53819
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7428836226463318,
+      "learning_rate": 2.134679471809152e-05,
+      "loss": 2.8189,
+      "step": 53820
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.757232666015625,
+      "learning_rate": 2.1341091242390296e-05,
+      "loss": 2.8876,
+      "step": 53821
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7275024652481079,
+      "learning_rate": 2.133538850062305e-05,
+      "loss": 2.7804,
+      "step": 53822
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7810689210891724,
+      "learning_rate": 2.132968649280471e-05,
+      "loss": 2.8834,
+      "step": 53823
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.758514404296875,
+      "learning_rate": 2.132398521895039e-05,
+      "loss": 2.9051,
+      "step": 53824
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7726489305496216,
+      "learning_rate": 2.1318284679075016e-05,
+      "loss": 3.0325,
+      "step": 53825
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7153679728507996,
+      "learning_rate": 2.1312584873193705e-05,
+      "loss": 2.9691,
+      "step": 53826
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7290853261947632,
+      "learning_rate": 2.1306885801321384e-05,
+      "loss": 3.0509,
+      "step": 53827
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7138325572013855,
+      "learning_rate": 2.1301187463473067e-05,
+      "loss": 2.9116,
+      "step": 53828
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7441121339797974,
+      "learning_rate": 2.1295489859663817e-05,
+      "loss": 2.9767,
+      "step": 53829
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7585853934288025,
+      "learning_rate": 2.1289792989908583e-05,
+      "loss": 2.8873,
+      "step": 53830
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7859900593757629,
+      "learning_rate": 2.1284096854222388e-05,
+      "loss": 2.9058,
+      "step": 53831
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.762042224407196,
+      "learning_rate": 2.1278401452620287e-05,
+      "loss": 2.8423,
+      "step": 53832
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7395182251930237,
+      "learning_rate": 2.1272706785117167e-05,
+      "loss": 2.7781,
+      "step": 53833
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7580464482307434,
+      "learning_rate": 2.126701285172815e-05,
+      "loss": 2.6739,
+      "step": 53834
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7301293015480042,
+      "learning_rate": 2.126131965246819e-05,
+      "loss": 2.7645,
+      "step": 53835
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7253274321556091,
+      "learning_rate": 2.1255627187352244e-05,
+      "loss": 2.7732,
+      "step": 53836
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7815554738044739,
+      "learning_rate": 2.124993545639533e-05,
+      "loss": 2.714,
+      "step": 53837
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7133461236953735,
+      "learning_rate": 2.1244244459612436e-05,
+      "loss": 2.8703,
+      "step": 53838
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7465064525604248,
+      "learning_rate": 2.1238554197018587e-05,
+      "loss": 2.9164,
+      "step": 53839
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7770967483520508,
+      "learning_rate": 2.123286466862867e-05,
+      "loss": 2.7855,
+      "step": 53840
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7370714545249939,
+      "learning_rate": 2.12271758744578e-05,
+      "loss": 2.7894,
+      "step": 53841
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.747188925743103,
+      "learning_rate": 2.122148781452091e-05,
+      "loss": 3.0637,
+      "step": 53842
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.747463047504425,
+      "learning_rate": 2.1215800488832913e-05,
+      "loss": 2.6843,
+      "step": 53843
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7289397120475769,
+      "learning_rate": 2.1210113897408898e-05,
+      "loss": 2.9257,
+      "step": 53844
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7071889042854309,
+      "learning_rate": 2.120442804026372e-05,
+      "loss": 2.7178,
+      "step": 53845
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7718331813812256,
+      "learning_rate": 2.119874291741247e-05,
+      "loss": 3.0457,
+      "step": 53846
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7785387635231018,
+      "learning_rate": 2.1193058528870066e-05,
+      "loss": 2.9793,
+      "step": 53847
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7537055015563965,
+      "learning_rate": 2.11873748746515e-05,
+      "loss": 2.723,
+      "step": 53848
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7677419185638428,
+      "learning_rate": 2.118169195477175e-05,
+      "loss": 2.7947,
+      "step": 53849
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7995899319648743,
+      "learning_rate": 2.1176009769245683e-05,
+      "loss": 3.0872,
+      "step": 53850
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.688861608505249,
+      "learning_rate": 2.1170328318088447e-05,
+      "loss": 2.8195,
+      "step": 53851
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8810485005378723,
+      "learning_rate": 2.11646476013148e-05,
+      "loss": 2.9762,
+      "step": 53852
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7384539842605591,
+      "learning_rate": 2.1158967618939858e-05,
+      "loss": 2.8733,
+      "step": 53853
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7322173118591309,
+      "learning_rate": 2.115328837097855e-05,
+      "loss": 2.847,
+      "step": 53854
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8136117458343506,
+      "learning_rate": 2.114760985744576e-05,
+      "loss": 2.916,
+      "step": 53855
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7917467951774597,
+      "learning_rate": 2.114193207835657e-05,
+      "loss": 2.8955,
+      "step": 53856
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7615936994552612,
+      "learning_rate": 2.1136255033725846e-05,
+      "loss": 3.037,
+      "step": 53857
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7316914200782776,
+      "learning_rate": 2.1130578723568538e-05,
+      "loss": 3.0313,
+      "step": 53858
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.705625593662262,
+      "learning_rate": 2.112490314789963e-05,
+      "loss": 3.0341,
+      "step": 53859
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7249335646629333,
+      "learning_rate": 2.1119228306734014e-05,
+      "loss": 3.0307,
+      "step": 53860
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7347077131271362,
+      "learning_rate": 2.1113554200086748e-05,
+      "loss": 2.8507,
+      "step": 53861
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.734248161315918,
+      "learning_rate": 2.1107880827972646e-05,
+      "loss": 2.9392,
+      "step": 53862
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7667978405952454,
+      "learning_rate": 2.1102208190406767e-05,
+      "loss": 2.7851,
+      "step": 53863
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7558903098106384,
+      "learning_rate": 2.1096536287403997e-05,
+      "loss": 2.7192,
+      "step": 53864
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8619865775108337,
+      "learning_rate": 2.1090865118979194e-05,
+      "loss": 2.8567,
+      "step": 53865
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7786526083946228,
+      "learning_rate": 2.1085194685147478e-05,
+      "loss": 2.949,
+      "step": 53866
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7425805330276489,
+      "learning_rate": 2.1079524985923602e-05,
+      "loss": 2.7537,
+      "step": 53867
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6905446648597717,
+      "learning_rate": 2.1073856021322555e-05,
+      "loss": 2.6834,
+      "step": 53868
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7856373190879822,
+      "learning_rate": 2.1068187791359393e-05,
+      "loss": 2.9288,
+      "step": 53869
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7731509804725647,
+      "learning_rate": 2.1062520296048903e-05,
+      "loss": 2.9401,
+      "step": 53870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7262278199195862,
+      "learning_rate": 2.105685353540607e-05,
+      "loss": 2.8644,
+      "step": 53871
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.730444610118866,
+      "learning_rate": 2.105118750944572e-05,
+      "loss": 3.1441,
+      "step": 53872
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7428088188171387,
+      "learning_rate": 2.1045522218182908e-05,
+      "loss": 2.9372,
+      "step": 53873
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7640175819396973,
+      "learning_rate": 2.103985766163252e-05,
+      "loss": 3.0142,
+      "step": 53874
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8164355158805847,
+      "learning_rate": 2.103419383980941e-05,
+      "loss": 2.9677,
+      "step": 53875
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7192038893699646,
+      "learning_rate": 2.10285307527286e-05,
+      "loss": 2.8549,
+      "step": 53876
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7759872078895569,
+      "learning_rate": 2.1022868400404913e-05,
+      "loss": 2.9487,
+      "step": 53877
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7276328802108765,
+      "learning_rate": 2.1017206782853334e-05,
+      "loss": 2.7763,
+      "step": 53878
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7632603049278259,
+      "learning_rate": 2.1011545900088757e-05,
+      "loss": 2.974,
+      "step": 53879
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7106670141220093,
+      "learning_rate": 2.1005885752126027e-05,
+      "loss": 2.8494,
+      "step": 53880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7603490948677063,
+      "learning_rate": 2.1000226338980143e-05,
+      "loss": 2.8562,
+      "step": 53881
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7455410361289978,
+      "learning_rate": 2.0994567660665916e-05,
+      "loss": 2.8456,
+      "step": 53882
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7279428243637085,
+      "learning_rate": 2.098890971719831e-05,
+      "loss": 3.1957,
+      "step": 53883
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8473446369171143,
+      "learning_rate": 2.098325250859224e-05,
+      "loss": 2.9781,
+      "step": 53884
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7343671321868896,
+      "learning_rate": 2.0977596034862632e-05,
+      "loss": 2.9825,
+      "step": 53885
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7839120030403137,
+      "learning_rate": 2.0971940296024304e-05,
+      "loss": 3.0148,
+      "step": 53886
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7506952285766602,
+      "learning_rate": 2.0966285292092144e-05,
+      "loss": 3.0424,
+      "step": 53887
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7599936723709106,
+      "learning_rate": 2.0960631023081077e-05,
+      "loss": 2.9168,
+      "step": 53888
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7371647953987122,
+      "learning_rate": 2.0954977489006085e-05,
+      "loss": 2.5458,
+      "step": 53889
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7845785021781921,
+      "learning_rate": 2.0949324689881896e-05,
+      "loss": 2.9261,
+      "step": 53890
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7766330242156982,
+      "learning_rate": 2.094367262572356e-05,
+      "loss": 2.9509,
+      "step": 53891
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7818014025688171,
+      "learning_rate": 2.0938021296545804e-05,
+      "loss": 2.9393,
+      "step": 53892
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7388964891433716,
+      "learning_rate": 2.0932370702363678e-05,
+      "loss": 2.8543,
+      "step": 53893
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7346888780593872,
+      "learning_rate": 2.0926720843191936e-05,
+      "loss": 2.8711,
+      "step": 53894
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7887476086616516,
+      "learning_rate": 2.0921071719045503e-05,
+      "loss": 2.9683,
+      "step": 53895
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7444952130317688,
+      "learning_rate": 2.091542332993926e-05,
+      "loss": 2.6975,
+      "step": 53896
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7575118541717529,
+      "learning_rate": 2.0909775675888073e-05,
+      "loss": 2.5881,
+      "step": 53897
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7286080121994019,
+      "learning_rate": 2.090412875690679e-05,
+      "loss": 2.7892,
+      "step": 53898
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7232897281646729,
+      "learning_rate": 2.08984825730104e-05,
+      "loss": 2.9935,
+      "step": 53899
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7478100061416626,
+      "learning_rate": 2.0892837124213657e-05,
+      "loss": 2.9702,
+      "step": 53900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7384991645812988,
+      "learning_rate": 2.0887192410531483e-05,
+      "loss": 2.8308,
+      "step": 53901
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7648786902427673,
+      "learning_rate": 2.08815484319787e-05,
+      "loss": 2.9977,
+      "step": 53902
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7234745621681213,
+      "learning_rate": 2.0875905188570163e-05,
+      "loss": 2.9809,
+      "step": 53903
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7378318905830383,
+      "learning_rate": 2.087026268032086e-05,
+      "loss": 2.9467,
+      "step": 53904
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7225425243377686,
+      "learning_rate": 2.086462090724548e-05,
+      "loss": 2.7495,
+      "step": 53905
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.744601309299469,
+      "learning_rate": 2.085897986935904e-05,
+      "loss": 2.9628,
+      "step": 53906
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7627322673797607,
+      "learning_rate": 2.085333956667633e-05,
+      "loss": 2.8884,
+      "step": 53907
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7031084299087524,
+      "learning_rate": 2.084769999921214e-05,
+      "loss": 3.1328,
+      "step": 53908
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8040719032287598,
+      "learning_rate": 2.0842061166981427e-05,
+      "loss": 2.9711,
+      "step": 53909
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.733121931552887,
+      "learning_rate": 2.0836423069998974e-05,
+      "loss": 2.9283,
+      "step": 53910
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7206590175628662,
+      "learning_rate": 2.0830785708279673e-05,
+      "loss": 2.8658,
+      "step": 53911
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7238807678222656,
+      "learning_rate": 2.082514908183831e-05,
+      "loss": 2.9172,
+      "step": 53912
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7584816217422485,
+      "learning_rate": 2.081951319068981e-05,
+      "loss": 2.977,
+      "step": 53913
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8075535297393799,
+      "learning_rate": 2.081387803484902e-05,
+      "loss": 2.8249,
+      "step": 53914
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7531391382217407,
+      "learning_rate": 2.0808243614330733e-05,
+      "loss": 2.8588,
+      "step": 53915
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7435611486434937,
+      "learning_rate": 2.0802609929149804e-05,
+      "loss": 2.9846,
+      "step": 53916
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7779226303100586,
+      "learning_rate": 2.0796976979321024e-05,
+      "loss": 2.9811,
+      "step": 53917
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7348940372467041,
+      "learning_rate": 2.079134476485924e-05,
+      "loss": 2.8087,
+      "step": 53918
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7662265300750732,
+      "learning_rate": 2.0785713285779415e-05,
+      "loss": 2.6446,
+      "step": 53919
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8061109185218811,
+      "learning_rate": 2.07800825420962e-05,
+      "loss": 2.6772,
+      "step": 53920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.744640052318573,
+      "learning_rate": 2.0774452533824583e-05,
+      "loss": 3.0171,
+      "step": 53921
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7543174624443054,
+      "learning_rate": 2.0768823260979318e-05,
+      "loss": 2.7892,
+      "step": 53922
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7616002559661865,
+      "learning_rate": 2.0763194723575193e-05,
+      "loss": 2.8543,
+      "step": 53923
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.725988507270813,
+      "learning_rate": 2.07575669216271e-05,
+      "loss": 2.82,
+      "step": 53924
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7437060475349426,
+      "learning_rate": 2.075193985514979e-05,
+      "loss": 2.7369,
+      "step": 53925
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7315745949745178,
+      "learning_rate": 2.0746313524158188e-05,
+      "loss": 2.7923,
+      "step": 53926
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7858138680458069,
+      "learning_rate": 2.0740687928666978e-05,
+      "loss": 2.9287,
+      "step": 53927
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7877421975135803,
+      "learning_rate": 2.073506306869115e-05,
+      "loss": 2.9113,
+      "step": 53928
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8002856969833374,
+      "learning_rate": 2.072943894424536e-05,
+      "loss": 2.9444,
+      "step": 53929
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7211002111434937,
+      "learning_rate": 2.0723815555344457e-05,
+      "loss": 2.7723,
+      "step": 53930
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7559928894042969,
+      "learning_rate": 2.0718192902003338e-05,
+      "loss": 2.9722,
+      "step": 53931
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7671013474464417,
+      "learning_rate": 2.0712570984236686e-05,
+      "loss": 2.9676,
+      "step": 53932
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7174298167228699,
+      "learning_rate": 2.070694980205946e-05,
+      "loss": 2.8203,
+      "step": 53933
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7641894817352295,
+      "learning_rate": 2.0701329355486274e-05,
+      "loss": 2.8446,
+      "step": 53934
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6962066292762756,
+      "learning_rate": 2.0695709644532056e-05,
+      "loss": 2.7518,
+      "step": 53935
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.74459308385849,
+      "learning_rate": 2.0690090669211658e-05,
+      "loss": 2.8212,
+      "step": 53936
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7321491837501526,
+      "learning_rate": 2.0684472429539767e-05,
+      "loss": 2.8309,
+      "step": 53937
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.739220917224884,
+      "learning_rate": 2.067885492553124e-05,
+      "loss": 3.009,
+      "step": 53938
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7226603627204895,
+      "learning_rate": 2.06732381572008e-05,
+      "loss": 2.7967,
+      "step": 53939
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7862400412559509,
+      "learning_rate": 2.06676221245633e-05,
+      "loss": 2.827,
+      "step": 53940
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7336839437484741,
+      "learning_rate": 2.0662006827633558e-05,
+      "loss": 2.9242,
+      "step": 53941
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7695524096488953,
+      "learning_rate": 2.06563922664263e-05,
+      "loss": 2.8123,
+      "step": 53942
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7591333389282227,
+      "learning_rate": 2.0650778440956383e-05,
+      "loss": 2.9191,
+      "step": 53943
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7534220218658447,
+      "learning_rate": 2.0645165351238557e-05,
+      "loss": 2.9842,
+      "step": 53944
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7719517946243286,
+      "learning_rate": 2.0639552997287546e-05,
+      "loss": 2.8474,
+      "step": 53945
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7812724113464355,
+      "learning_rate": 2.0633941379118237e-05,
+      "loss": 2.7268,
+      "step": 53946
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7152418494224548,
+      "learning_rate": 2.0628330496745317e-05,
+      "loss": 2.833,
+      "step": 53947
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7359592914581299,
+      "learning_rate": 2.0622720350183674e-05,
+      "loss": 2.8466,
+      "step": 53948
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7347102761268616,
+      "learning_rate": 2.0617110939447966e-05,
+      "loss": 3.0976,
+      "step": 53949
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7281254529953003,
+      "learning_rate": 2.0611502264553048e-05,
+      "loss": 3.0755,
+      "step": 53950
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7379975914955139,
+      "learning_rate": 2.0605894325513673e-05,
+      "loss": 2.7204,
+      "step": 53951
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7058905959129333,
+      "learning_rate": 2.0600287122344562e-05,
+      "loss": 2.7165,
+      "step": 53952
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7494770884513855,
+      "learning_rate": 2.0594680655060536e-05,
+      "loss": 2.9196,
+      "step": 53953
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7461589574813843,
+      "learning_rate": 2.0589074923676352e-05,
+      "loss": 2.7942,
+      "step": 53954
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8033571839332581,
+      "learning_rate": 2.058346992820673e-05,
+      "loss": 2.6223,
+      "step": 53955
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7271087169647217,
+      "learning_rate": 2.0577865668666527e-05,
+      "loss": 3.1622,
+      "step": 53956
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7702967524528503,
+      "learning_rate": 2.0572262145070427e-05,
+      "loss": 2.827,
+      "step": 53957
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7806286811828613,
+      "learning_rate": 2.056665935743326e-05,
+      "loss": 3.1208,
+      "step": 53958
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7661564350128174,
+      "learning_rate": 2.0561057305769703e-05,
+      "loss": 2.7039,
+      "step": 53959
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7610711455345154,
+      "learning_rate": 2.0555455990094515e-05,
+      "loss": 2.873,
+      "step": 53960
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7484396696090698,
+      "learning_rate": 2.054985541042252e-05,
+      "loss": 2.7665,
+      "step": 53961
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.734775722026825,
+      "learning_rate": 2.0544255566768376e-05,
+      "loss": 3.0531,
+      "step": 53962
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.728737473487854,
+      "learning_rate": 2.0538656459146928e-05,
+      "loss": 3.0091,
+      "step": 53963
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8020430207252502,
+      "learning_rate": 2.0533058087572805e-05,
+      "loss": 3.1086,
+      "step": 53964
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7418114542961121,
+      "learning_rate": 2.0527460452060895e-05,
+      "loss": 2.8423,
+      "step": 53965
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7765441536903381,
+      "learning_rate": 2.0521863552625884e-05,
+      "loss": 2.9005,
+      "step": 53966
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7442047595977783,
+      "learning_rate": 2.0516267389282426e-05,
+      "loss": 2.8706,
+      "step": 53967
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7826739549636841,
+      "learning_rate": 2.051067196204538e-05,
+      "loss": 2.861,
+      "step": 53968
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7431984543800354,
+      "learning_rate": 2.0505077270929392e-05,
+      "loss": 2.8538,
+      "step": 53969
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7477771043777466,
+      "learning_rate": 2.0499483315949262e-05,
+      "loss": 2.9119,
+      "step": 53970
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.772100031375885,
+      "learning_rate": 2.0493890097119703e-05,
+      "loss": 2.8383,
+      "step": 53971
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7262080311775208,
+      "learning_rate": 2.0488297614455474e-05,
+      "loss": 2.9591,
+      "step": 53972
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8009024262428284,
+      "learning_rate": 2.0482705867971293e-05,
+      "loss": 3.0581,
+      "step": 53973
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7679193019866943,
+      "learning_rate": 2.0477114857681787e-05,
+      "loss": 2.7829,
+      "step": 53974
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7369647026062012,
+      "learning_rate": 2.047152458360177e-05,
+      "loss": 3.0353,
+      "step": 53975
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7575830817222595,
+      "learning_rate": 2.046593504574604e-05,
+      "loss": 2.9625,
+      "step": 53976
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7336440682411194,
+      "learning_rate": 2.0460346244129177e-05,
+      "loss": 2.8509,
+      "step": 53977
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7487339377403259,
+      "learning_rate": 2.0454758178766007e-05,
+      "loss": 2.6161,
+      "step": 53978
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7532362341880798,
+      "learning_rate": 2.044917084967115e-05,
+      "loss": 2.8774,
+      "step": 53979
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7453672885894775,
+      "learning_rate": 2.0443584256859426e-05,
+      "loss": 2.9005,
+      "step": 53980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.750306248664856,
+      "learning_rate": 2.0437998400345522e-05,
+      "loss": 3.0054,
+      "step": 53981
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7489728927612305,
+      "learning_rate": 2.0432413280144035e-05,
+      "loss": 3.005,
+      "step": 53982
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7205755710601807,
+      "learning_rate": 2.0426828896269843e-05,
+      "loss": 2.9092,
+      "step": 53983
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7374038696289062,
+      "learning_rate": 2.042124524873754e-05,
+      "loss": 2.7873,
+      "step": 53984
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7091577649116516,
+      "learning_rate": 2.041566233756188e-05,
+      "loss": 2.8018,
+      "step": 53985
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7290316820144653,
+      "learning_rate": 2.041008016275758e-05,
+      "loss": 2.9189,
+      "step": 53986
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7330735921859741,
+      "learning_rate": 2.0404498724339336e-05,
+      "loss": 2.9191,
+      "step": 53987
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7329016923904419,
+      "learning_rate": 2.0398918022321796e-05,
+      "loss": 3.0506,
+      "step": 53988
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.764037013053894,
+      "learning_rate": 2.039333805671969e-05,
+      "loss": 2.8124,
+      "step": 53989
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7291437983512878,
+      "learning_rate": 2.0387758827547696e-05,
+      "loss": 2.7463,
+      "step": 53990
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7491414546966553,
+      "learning_rate": 2.038218033482061e-05,
+      "loss": 2.7817,
+      "step": 53991
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7548601031303406,
+      "learning_rate": 2.037660257855298e-05,
+      "loss": 2.7962,
+      "step": 53992
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7716947197914124,
+      "learning_rate": 2.0371025558759602e-05,
+      "loss": 2.8948,
+      "step": 53993
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7344406247138977,
+      "learning_rate": 2.036544927545509e-05,
+      "loss": 2.9549,
+      "step": 53994
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7470876574516296,
+      "learning_rate": 2.035987372865421e-05,
+      "loss": 3.075,
+      "step": 53995
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7501749992370605,
+      "learning_rate": 2.0354298918371638e-05,
+      "loss": 2.5949,
+      "step": 53996
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7505838871002197,
+      "learning_rate": 2.0348724844621932e-05,
+      "loss": 3.097,
+      "step": 53997
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7241684198379517,
+      "learning_rate": 2.034315150741992e-05,
+      "loss": 2.8974,
+      "step": 53998
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7406726479530334,
+      "learning_rate": 2.0337578906780185e-05,
+      "loss": 2.8339,
+      "step": 53999
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7717121839523315,
+      "learning_rate": 2.0332007042717447e-05,
+      "loss": 3.0096,
+      "step": 54000
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7408150434494019,
+      "learning_rate": 2.032643591524643e-05,
+      "loss": 2.9066,
+      "step": 54001
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7498103380203247,
+      "learning_rate": 2.0320865524381758e-05,
+      "loss": 2.8442,
+      "step": 54002
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7494451999664307,
+      "learning_rate": 2.0315295870138114e-05,
+      "loss": 2.6605,
+      "step": 54003
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7431146502494812,
+      "learning_rate": 2.0309726952530092e-05,
+      "loss": 2.8177,
+      "step": 54004
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7047357559204102,
+      "learning_rate": 2.0304158771572442e-05,
+      "loss": 2.879,
+      "step": 54005
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7716295123100281,
+      "learning_rate": 2.029859132727979e-05,
+      "loss": 3.1973,
+      "step": 54006
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7560664415359497,
+      "learning_rate": 2.029302461966682e-05,
+      "loss": 3.088,
+      "step": 54007
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7658318877220154,
+      "learning_rate": 2.0287458648748257e-05,
+      "loss": 2.7865,
+      "step": 54008
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7343339323997498,
+      "learning_rate": 2.0281893414538653e-05,
+      "loss": 2.7412,
+      "step": 54009
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7271828651428223,
+      "learning_rate": 2.027632891705273e-05,
+      "loss": 2.8022,
+      "step": 54010
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7835471034049988,
+      "learning_rate": 2.027076515630508e-05,
+      "loss": 2.8315,
+      "step": 54011
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7581990361213684,
+      "learning_rate": 2.0265202132310388e-05,
+      "loss": 2.8508,
+      "step": 54012
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7217426896095276,
+      "learning_rate": 2.0259639845083374e-05,
+      "loss": 3.0623,
+      "step": 54013
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7215791344642639,
+      "learning_rate": 2.0254078294638598e-05,
+      "loss": 2.8275,
+      "step": 54014
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7520722150802612,
+      "learning_rate": 2.0248517480990777e-05,
+      "loss": 2.8071,
+      "step": 54015
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7443130016326904,
+      "learning_rate": 2.024295740415447e-05,
+      "loss": 2.849,
+      "step": 54016
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7317799925804138,
+      "learning_rate": 2.0237398064144394e-05,
+      "loss": 2.8952,
+      "step": 54017
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7501115798950195,
+      "learning_rate": 2.0231839460975206e-05,
+      "loss": 3.1212,
+      "step": 54018
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7234840393066406,
+      "learning_rate": 2.0226281594661464e-05,
+      "loss": 3.0704,
+      "step": 54019
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7871589064598083,
+      "learning_rate": 2.0220724465217884e-05,
+      "loss": 2.9676,
+      "step": 54020
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7186853885650635,
+      "learning_rate": 2.0215168072659027e-05,
+      "loss": 3.0373,
+      "step": 54021
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7855047583580017,
+      "learning_rate": 2.0209612416999578e-05,
+      "loss": 2.8772,
+      "step": 54022
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7410944104194641,
+      "learning_rate": 2.0204057498254223e-05,
+      "loss": 2.7401,
+      "step": 54023
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7117852568626404,
+      "learning_rate": 2.019850331643752e-05,
+      "loss": 2.8494,
+      "step": 54024
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7461109757423401,
+      "learning_rate": 2.019294987156409e-05,
+      "loss": 2.9391,
+      "step": 54025
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.769916296005249,
+      "learning_rate": 2.0187397163648555e-05,
+      "loss": 2.9171,
+      "step": 54026
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7795796990394592,
+      "learning_rate": 2.018184519270557e-05,
+      "loss": 2.7288,
+      "step": 54027
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7641432285308838,
+      "learning_rate": 2.017629395874979e-05,
+      "loss": 2.8498,
+      "step": 54028
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7332988977432251,
+      "learning_rate": 2.0170743461795736e-05,
+      "loss": 2.8082,
+      "step": 54029
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.758630096912384,
+      "learning_rate": 2.0165193701858163e-05,
+      "loss": 2.8566,
+      "step": 54030
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7000086307525635,
+      "learning_rate": 2.015964467895159e-05,
+      "loss": 2.8143,
+      "step": 54031
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7220397591590881,
+      "learning_rate": 2.0154096393090647e-05,
+      "loss": 2.7702,
+      "step": 54032
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7732095718383789,
+      "learning_rate": 2.0148548844289946e-05,
+      "loss": 2.875,
+      "step": 54033
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7628332376480103,
+      "learning_rate": 2.0143002032564116e-05,
+      "loss": 2.7058,
+      "step": 54034
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7517362833023071,
+      "learning_rate": 2.0137455957927773e-05,
+      "loss": 2.7692,
+      "step": 54035
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7574434280395508,
+      "learning_rate": 2.0131910620395476e-05,
+      "loss": 2.7894,
+      "step": 54036
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7853326797485352,
+      "learning_rate": 2.0126366019981878e-05,
+      "loss": 2.861,
+      "step": 54037
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7405124306678772,
+      "learning_rate": 2.01208221567016e-05,
+      "loss": 2.9171,
+      "step": 54038
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7227654457092285,
+      "learning_rate": 2.0115279030569232e-05,
+      "loss": 2.873,
+      "step": 54039
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7326869368553162,
+      "learning_rate": 2.010973664159933e-05,
+      "loss": 2.8694,
+      "step": 54040
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8230959177017212,
+      "learning_rate": 2.0104194989806476e-05,
+      "loss": 2.8616,
+      "step": 54041
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7183669209480286,
+      "learning_rate": 2.00986540752053e-05,
+      "loss": 2.9716,
+      "step": 54042
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7086861729621887,
+      "learning_rate": 2.0093113897810453e-05,
+      "loss": 2.7061,
+      "step": 54043
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7914672493934631,
+      "learning_rate": 2.0087574457636424e-05,
+      "loss": 2.8385,
+      "step": 54044
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7433429956436157,
+      "learning_rate": 2.00820357546979e-05,
+      "loss": 2.8681,
+      "step": 54045
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7644243240356445,
+      "learning_rate": 2.0076497789009437e-05,
+      "loss": 3.195,
+      "step": 54046
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.76406329870224,
+      "learning_rate": 2.007096056058556e-05,
+      "loss": 2.6896,
+      "step": 54047
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7108646035194397,
+      "learning_rate": 2.0065424069440915e-05,
+      "loss": 2.6872,
+      "step": 54048
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7588872313499451,
+      "learning_rate": 2.0059888315590068e-05,
+      "loss": 3.081,
+      "step": 54049
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7359708547592163,
+      "learning_rate": 2.0054353299047598e-05,
+      "loss": 2.7153,
+      "step": 54050
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7428367733955383,
+      "learning_rate": 2.0048819019828068e-05,
+      "loss": 2.935,
+      "step": 54051
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7662042379379272,
+      "learning_rate": 2.0043285477946124e-05,
+      "loss": 3.0226,
+      "step": 54052
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7269078493118286,
+      "learning_rate": 2.0037752673416264e-05,
+      "loss": 2.9673,
+      "step": 54053
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7469284534454346,
+      "learning_rate": 2.0032220606253037e-05,
+      "loss": 2.9737,
+      "step": 54054
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7175904512405396,
+      "learning_rate": 2.002668927647113e-05,
+      "loss": 2.8703,
+      "step": 54055
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7442259192466736,
+      "learning_rate": 2.002115868408497e-05,
+      "loss": 2.9262,
+      "step": 54056
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7313393354415894,
+      "learning_rate": 2.001562882910921e-05,
+      "loss": 2.946,
+      "step": 54057
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7050204873085022,
+      "learning_rate": 2.0010099711558468e-05,
+      "loss": 3.024,
+      "step": 54058
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7282868027687073,
+      "learning_rate": 2.000457133144714e-05,
+      "loss": 2.8238,
+      "step": 54059
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7803441882133484,
+      "learning_rate": 1.9999043688789975e-05,
+      "loss": 2.9185,
+      "step": 54060
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7263804078102112,
+      "learning_rate": 1.9993516783601426e-05,
+      "loss": 2.9197,
+      "step": 54061
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7453001141548157,
+      "learning_rate": 1.998799061589602e-05,
+      "loss": 2.7994,
+      "step": 54062
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7197867631912231,
+      "learning_rate": 1.998246518568841e-05,
+      "loss": 3.0516,
+      "step": 54063
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7825305461883545,
+      "learning_rate": 1.997694049299302e-05,
+      "loss": 2.9953,
+      "step": 54064
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7243884801864624,
+      "learning_rate": 1.997141653782457e-05,
+      "loss": 2.828,
+      "step": 54065
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7629521489143372,
+      "learning_rate": 1.9965893320197447e-05,
+      "loss": 2.8634,
+      "step": 54066
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7828942537307739,
+      "learning_rate": 1.9960370840126306e-05,
+      "loss": 3.0588,
+      "step": 54067
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7585510611534119,
+      "learning_rate": 1.9954849097625668e-05,
+      "loss": 2.7942,
+      "step": 54068
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7742347717285156,
+      "learning_rate": 1.9949328092709994e-05,
+      "loss": 3.0219,
+      "step": 54069
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.736878514289856,
+      "learning_rate": 1.9943807825393965e-05,
+      "loss": 2.9672,
+      "step": 54070
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7557036280632019,
+      "learning_rate": 1.9938288295691974e-05,
+      "loss": 3.0019,
+      "step": 54071
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7458971738815308,
+      "learning_rate": 1.993276950361864e-05,
+      "loss": 2.762,
+      "step": 54072
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7320734858512878,
+      "learning_rate": 1.9927251449188552e-05,
+      "loss": 3.0947,
+      "step": 54073
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7172092199325562,
+      "learning_rate": 1.992173413241617e-05,
+      "loss": 2.8381,
+      "step": 54074
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.707459568977356,
+      "learning_rate": 1.9916217553316037e-05,
+      "loss": 2.9832,
+      "step": 54075
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7637947797775269,
+      "learning_rate": 1.9910701711902656e-05,
+      "loss": 3.0281,
+      "step": 54076
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7895542979240417,
+      "learning_rate": 1.990518660819057e-05,
+      "loss": 3.0742,
+      "step": 54077
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.707626223564148,
+      "learning_rate": 1.9899672242194376e-05,
+      "loss": 2.8941,
+      "step": 54078
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.743711531162262,
+      "learning_rate": 1.989415861392849e-05,
+      "loss": 3.0584,
+      "step": 54079
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7679846286773682,
+      "learning_rate": 1.9888645723407504e-05,
+      "loss": 2.858,
+      "step": 54080
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7698391675949097,
+      "learning_rate": 1.9883133570645903e-05,
+      "loss": 2.6755,
+      "step": 54081
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7134325504302979,
+      "learning_rate": 1.9877622155658246e-05,
+      "loss": 2.8351,
+      "step": 54082
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7812037467956543,
+      "learning_rate": 1.9872111478459018e-05,
+      "loss": 2.9758,
+      "step": 54083
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7417611479759216,
+      "learning_rate": 1.9866601539062677e-05,
+      "loss": 2.8881,
+      "step": 54084
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8070379495620728,
+      "learning_rate": 1.9861092337483876e-05,
+      "loss": 2.8155,
+      "step": 54085
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7638986706733704,
+      "learning_rate": 1.985558387373697e-05,
+      "loss": 2.7645,
+      "step": 54086
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.6919122934341431,
+      "learning_rate": 1.9850076147836613e-05,
+      "loss": 2.8189,
+      "step": 54087
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7425715923309326,
+      "learning_rate": 1.9844569159797165e-05,
+      "loss": 2.8075,
+      "step": 54088
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7360454201698303,
+      "learning_rate": 1.9839062909633274e-05,
+      "loss": 2.898,
+      "step": 54089
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7551466226577759,
+      "learning_rate": 1.9833557397359366e-05,
+      "loss": 2.9805,
+      "step": 54090
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7435944080352783,
+      "learning_rate": 1.98280526229899e-05,
+      "loss": 2.9451,
+      "step": 54091
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7457761168479919,
+      "learning_rate": 1.9822548586539487e-05,
+      "loss": 2.7013,
+      "step": 54092
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8180214762687683,
+      "learning_rate": 1.981704528802249e-05,
+      "loss": 3.0459,
+      "step": 54093
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7294940948486328,
+      "learning_rate": 1.9811542727453467e-05,
+      "loss": 2.8207,
+      "step": 54094
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.742396354675293,
+      "learning_rate": 1.9806040904846997e-05,
+      "loss": 2.9505,
+      "step": 54095
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7712141275405884,
+      "learning_rate": 1.9800539820217474e-05,
+      "loss": 2.8266,
+      "step": 54096
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7579017877578735,
+      "learning_rate": 1.979503947357939e-05,
+      "loss": 2.7108,
+      "step": 54097
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7335031628608704,
+      "learning_rate": 1.9789539864947224e-05,
+      "loss": 2.7703,
+      "step": 54098
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7949396967887878,
+      "learning_rate": 1.9784040994335472e-05,
+      "loss": 2.902,
+      "step": 54099
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7702135443687439,
+      "learning_rate": 1.977854286175865e-05,
+      "loss": 2.8729,
+      "step": 54100
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7160282731056213,
+      "learning_rate": 1.9773045467231185e-05,
+      "loss": 2.9078,
+      "step": 54101
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7243478298187256,
+      "learning_rate": 1.9767548810767663e-05,
+      "loss": 2.9189,
+      "step": 54102
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7236966490745544,
+      "learning_rate": 1.9762052892382407e-05,
+      "loss": 3.0583,
+      "step": 54103
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7192164063453674,
+      "learning_rate": 1.9756557712090037e-05,
+      "loss": 2.7622,
+      "step": 54104
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7658694386482239,
+      "learning_rate": 1.9751063269904945e-05,
+      "loss": 2.9541,
+      "step": 54105
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7122344374656677,
+      "learning_rate": 1.9745569565841578e-05,
+      "loss": 2.9305,
+      "step": 54106
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7958580851554871,
+      "learning_rate": 1.9740076599914468e-05,
+      "loss": 2.9211,
+      "step": 54107
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7178252339363098,
+      "learning_rate": 1.973458437213803e-05,
+      "loss": 2.8776,
+      "step": 54108
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7540079951286316,
+      "learning_rate": 1.9729092882526788e-05,
+      "loss": 2.8094,
+      "step": 54109
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.766581118106842,
+      "learning_rate": 1.9723602131095195e-05,
+      "loss": 2.7469,
+      "step": 54110
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7104828953742981,
+      "learning_rate": 1.9718112117857675e-05,
+      "loss": 2.7799,
+      "step": 54111
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7695900201797485,
+      "learning_rate": 1.9712622842828718e-05,
+      "loss": 3.009,
+      "step": 54112
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7981126308441162,
+      "learning_rate": 1.9707134306022744e-05,
+      "loss": 2.7478,
+      "step": 54113
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.714474618434906,
+      "learning_rate": 1.9701646507454237e-05,
+      "loss": 2.9426,
+      "step": 54114
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7519267201423645,
+      "learning_rate": 1.9696159447137662e-05,
+      "loss": 2.8552,
+      "step": 54115
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7607603669166565,
+      "learning_rate": 1.969067312508743e-05,
+      "loss": 2.8663,
+      "step": 54116
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7680455446243286,
+      "learning_rate": 1.968518754131807e-05,
+      "loss": 2.8828,
+      "step": 54117
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7606989145278931,
+      "learning_rate": 1.9679702695843935e-05,
+      "loss": 2.7714,
+      "step": 54118
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7711759805679321,
+      "learning_rate": 1.9674218588679548e-05,
+      "loss": 2.8402,
+      "step": 54119
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7324842214584351,
+      "learning_rate": 1.9668735219839327e-05,
+      "loss": 2.7645,
+      "step": 54120
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7444130182266235,
+      "learning_rate": 1.966325258933763e-05,
+      "loss": 2.8317,
+      "step": 54121
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7366726398468018,
+      "learning_rate": 1.965777069718908e-05,
+      "loss": 2.8117,
+      "step": 54122
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7467571496963501,
+      "learning_rate": 1.9652289543407927e-05,
+      "loss": 3.1177,
+      "step": 54123
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7339689135551453,
+      "learning_rate": 1.9646809128008668e-05,
+      "loss": 2.7936,
+      "step": 54124
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.812752902507782,
+      "learning_rate": 1.964132945100585e-05,
+      "loss": 2.9288,
+      "step": 54125
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.777722954750061,
+      "learning_rate": 1.9635850512413797e-05,
+      "loss": 2.8443,
+      "step": 54126
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8094779253005981,
+      "learning_rate": 1.9630372312246933e-05,
+      "loss": 2.9391,
+      "step": 54127
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7842881083488464,
+      "learning_rate": 1.962489485051968e-05,
+      "loss": 2.8474,
+      "step": 54128
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7789468765258789,
+      "learning_rate": 1.9619418127246524e-05,
+      "loss": 2.8385,
+      "step": 54129
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8131735920906067,
+      "learning_rate": 1.961394214244185e-05,
+      "loss": 3.0392,
+      "step": 54130
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7453126907348633,
+      "learning_rate": 1.960846689612009e-05,
+      "loss": 2.8532,
+      "step": 54131
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7671250104904175,
+      "learning_rate": 1.960299238829569e-05,
+      "loss": 2.914,
+      "step": 54132
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7735947370529175,
+      "learning_rate": 1.9597518618983044e-05,
+      "loss": 2.8704,
+      "step": 54133
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.762312650680542,
+      "learning_rate": 1.9592045588196536e-05,
+      "loss": 2.7795,
+      "step": 54134
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7702953815460205,
+      "learning_rate": 1.9586573295950658e-05,
+      "loss": 2.7177,
+      "step": 54135
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7665062546730042,
+      "learning_rate": 1.958110174225973e-05,
+      "loss": 3.0213,
+      "step": 54136
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7448751926422119,
+      "learning_rate": 1.9575630927138274e-05,
+      "loss": 2.8374,
+      "step": 54137
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7393977642059326,
+      "learning_rate": 1.957016085060058e-05,
+      "loss": 2.8929,
+      "step": 54138
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7659404873847961,
+      "learning_rate": 1.9564691512661102e-05,
+      "loss": 2.7063,
+      "step": 54139
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7388503551483154,
+      "learning_rate": 1.9559222913334293e-05,
+      "loss": 2.922,
+      "step": 54140
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7335487604141235,
+      "learning_rate": 1.9553755052634545e-05,
+      "loss": 2.8808,
+      "step": 54141
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7967286705970764,
+      "learning_rate": 1.954828793057621e-05,
+      "loss": 2.7343,
+      "step": 54142
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7063218951225281,
+      "learning_rate": 1.954282154717368e-05,
+      "loss": 3.0011,
+      "step": 54143
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7302457690238953,
+      "learning_rate": 1.9537355902441408e-05,
+      "loss": 3.0176,
+      "step": 54144
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.739165723323822,
+      "learning_rate": 1.9531890996393785e-05,
+      "loss": 2.6946,
+      "step": 54145
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7565047144889832,
+      "learning_rate": 1.9526426829045126e-05,
+      "loss": 2.8469,
+      "step": 54146
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7552198171615601,
+      "learning_rate": 1.9520963400409927e-05,
+      "loss": 2.8518,
+      "step": 54147
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7470796704292297,
+      "learning_rate": 1.9515500710502542e-05,
+      "loss": 2.8107,
+      "step": 54148
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.764629602432251,
+      "learning_rate": 1.9510038759337287e-05,
+      "loss": 2.8671,
+      "step": 54149
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7533672451972961,
+      "learning_rate": 1.950457754692869e-05,
+      "loss": 2.9093,
+      "step": 54150
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7616235613822937,
+      "learning_rate": 1.949911707329097e-05,
+      "loss": 2.8767,
+      "step": 54151
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7351246476173401,
+      "learning_rate": 1.949365733843865e-05,
+      "loss": 2.6712,
+      "step": 54152
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.80738765001297,
+      "learning_rate": 1.948819834238602e-05,
+      "loss": 2.7892,
+      "step": 54153
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7314664721488953,
+      "learning_rate": 1.948274008514753e-05,
+      "loss": 2.8074,
+      "step": 54154
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7750166058540344,
+      "learning_rate": 1.9477282566737505e-05,
+      "loss": 2.7988,
+      "step": 54155
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.739399790763855,
+      "learning_rate": 1.9471825787170304e-05,
+      "loss": 2.8234,
+      "step": 54156
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7659290432929993,
+      "learning_rate": 1.946636974646034e-05,
+      "loss": 2.828,
+      "step": 54157
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7248683571815491,
+      "learning_rate": 1.9460914444621943e-05,
+      "loss": 2.9416,
+      "step": 54158
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7538133859634399,
+      "learning_rate": 1.9455459881669566e-05,
+      "loss": 3.1345,
+      "step": 54159
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7005477547645569,
+      "learning_rate": 1.945000605761743e-05,
+      "loss": 2.7204,
+      "step": 54160
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7675425410270691,
+      "learning_rate": 1.944455297248002e-05,
+      "loss": 2.7766,
+      "step": 54161
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7757595777511597,
+      "learning_rate": 1.9439100626271697e-05,
+      "loss": 2.87,
+      "step": 54162
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7428346276283264,
+      "learning_rate": 1.943364901900678e-05,
+      "loss": 2.8913,
+      "step": 54163
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.74378901720047,
+      "learning_rate": 1.942819815069966e-05,
+      "loss": 2.7817,
+      "step": 54164
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7840437889099121,
+      "learning_rate": 1.942274802136462e-05,
+      "loss": 2.8597,
+      "step": 54165
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.735499382019043,
+      "learning_rate": 1.941729863101602e-05,
+      "loss": 2.672,
+      "step": 54166
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7412136793136597,
+      "learning_rate": 1.941184997966835e-05,
+      "loss": 2.9249,
+      "step": 54167
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7294532656669617,
+      "learning_rate": 1.9406402067335792e-05,
+      "loss": 2.917,
+      "step": 54168
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7440246343612671,
+      "learning_rate": 1.9400954894032837e-05,
+      "loss": 2.9341,
+      "step": 54169
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7838345766067505,
+      "learning_rate": 1.9395508459773745e-05,
+      "loss": 2.9521,
+      "step": 54170
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7340986728668213,
+      "learning_rate": 1.9390062764572865e-05,
+      "loss": 3.0219,
+      "step": 54171
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7282890677452087,
+      "learning_rate": 1.938461780844456e-05,
+      "loss": 2.8559,
+      "step": 54172
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7802963256835938,
+      "learning_rate": 1.9379173591403142e-05,
+      "loss": 3.0019,
+      "step": 54173
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7121721506118774,
+      "learning_rate": 1.9373730113463037e-05,
+      "loss": 3.0312,
+      "step": 54174
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7408917546272278,
+      "learning_rate": 1.9368287374638437e-05,
+      "loss": 3.026,
+      "step": 54175
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7899402976036072,
+      "learning_rate": 1.9362845374943823e-05,
+      "loss": 2.8091,
+      "step": 54176
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7442259192466736,
+      "learning_rate": 1.935740411439346e-05,
+      "loss": 2.894,
+      "step": 54177
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8089560270309448,
+      "learning_rate": 1.9351963593001664e-05,
+      "loss": 2.7595,
+      "step": 54178
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7835502624511719,
+      "learning_rate": 1.9346523810782788e-05,
+      "loss": 2.8599,
+      "step": 54179
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7364153265953064,
+      "learning_rate": 1.9341084767751158e-05,
+      "loss": 2.6572,
+      "step": 54180
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.0555264949798584,
+      "learning_rate": 1.933564646392106e-05,
+      "loss": 2.8671,
+      "step": 54181
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7606764435768127,
+      "learning_rate": 1.933020889930692e-05,
+      "loss": 2.974,
+      "step": 54182
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 1.9324772073922922e-05,
+      "loss": 2.8173,
+      "step": 54183
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7525346875190735,
+      "learning_rate": 1.9319335987783523e-05,
+      "loss": 2.8566,
+      "step": 54184
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7471001744270325,
+      "learning_rate": 1.9313900640902978e-05,
+      "loss": 2.8165,
+      "step": 54185
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7534440159797668,
+      "learning_rate": 1.9308466033295544e-05,
+      "loss": 2.922,
+      "step": 54186
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7355985045433044,
+      "learning_rate": 1.930303216497564e-05,
+      "loss": 2.7781,
+      "step": 54187
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7623375058174133,
+      "learning_rate": 1.9297599035957457e-05,
+      "loss": 2.872,
+      "step": 54188
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7745537161827087,
+      "learning_rate": 1.9292166646255447e-05,
+      "loss": 2.624,
+      "step": 54189
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7441405653953552,
+      "learning_rate": 1.92867349958838e-05,
+      "loss": 2.7594,
+      "step": 54190
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7160558104515076,
+      "learning_rate": 1.9281304084856942e-05,
+      "loss": 2.8772,
+      "step": 54191
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7814651727676392,
+      "learning_rate": 1.9275873913189055e-05,
+      "loss": 2.7706,
+      "step": 54192
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7084113359451294,
+      "learning_rate": 1.9270444480894463e-05,
+      "loss": 3.0227,
+      "step": 54193
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8035101890563965,
+      "learning_rate": 1.9265015787987558e-05,
+      "loss": 2.9222,
+      "step": 54194
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7509192824363708,
+      "learning_rate": 1.9259587834482493e-05,
+      "loss": 3.0245,
+      "step": 54195
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7465216517448425,
+      "learning_rate": 1.9254160620393656e-05,
+      "loss": 2.9304,
+      "step": 54196
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7854889035224915,
+      "learning_rate": 1.92487341457354e-05,
+      "loss": 2.6707,
+      "step": 54197
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7187995314598083,
+      "learning_rate": 1.924330841052192e-05,
+      "loss": 2.899,
+      "step": 54198
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7536890506744385,
+      "learning_rate": 1.923788341476753e-05,
+      "loss": 2.8897,
+      "step": 54199
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7636274695396423,
+      "learning_rate": 1.9232459158486492e-05,
+      "loss": 2.8353,
+      "step": 54200
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7261160612106323,
+      "learning_rate": 1.9227035641693123e-05,
+      "loss": 2.7868,
+      "step": 54201
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7559409141540527,
+      "learning_rate": 1.9221612864401716e-05,
+      "loss": 3.1095,
+      "step": 54202
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8080114722251892,
+      "learning_rate": 1.9216190826626555e-05,
+      "loss": 3.0153,
+      "step": 54203
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7144877314567566,
+      "learning_rate": 1.9210769528381898e-05,
+      "loss": 3.012,
+      "step": 54204
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.700718343257904,
+      "learning_rate": 1.9205348969682032e-05,
+      "loss": 2.8758,
+      "step": 54205
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7655426263809204,
+      "learning_rate": 1.919992915054125e-05,
+      "loss": 3.0361,
+      "step": 54206
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7685981392860413,
+      "learning_rate": 1.9194510070973834e-05,
+      "loss": 2.9705,
+      "step": 54207
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7193219661712646,
+      "learning_rate": 1.9189091730994012e-05,
+      "loss": 2.799,
+      "step": 54208
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7027367353439331,
+      "learning_rate": 1.91836741306161e-05,
+      "loss": 2.9511,
+      "step": 54209
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.744439423084259,
+      "learning_rate": 1.917825726985429e-05,
+      "loss": 2.788,
+      "step": 54210
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7260327339172363,
+      "learning_rate": 1.9172841148722905e-05,
+      "loss": 2.7213,
+      "step": 54211
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7319767475128174,
+      "learning_rate": 1.9167425767236268e-05,
+      "loss": 3.0538,
+      "step": 54212
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7671077251434326,
+      "learning_rate": 1.9162011125408595e-05,
+      "loss": 2.8832,
+      "step": 54213
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8228315711021423,
+      "learning_rate": 1.915659722325411e-05,
+      "loss": 2.8539,
+      "step": 54214
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7222673296928406,
+      "learning_rate": 1.9151184060787072e-05,
+      "loss": 2.9277,
+      "step": 54215
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7548918724060059,
+      "learning_rate": 1.9145771638021767e-05,
+      "loss": 2.8345,
+      "step": 54216
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7558291554450989,
+      "learning_rate": 1.9140359954972485e-05,
+      "loss": 2.799,
+      "step": 54217
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7385692000389099,
+      "learning_rate": 1.913494901165341e-05,
+      "loss": 2.9751,
+      "step": 54218
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7456986904144287,
+      "learning_rate": 1.9129538808078838e-05,
+      "loss": 3.0317,
+      "step": 54219
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7515402436256409,
+      "learning_rate": 1.9124129344262984e-05,
+      "loss": 3.0887,
+      "step": 54220
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7937130928039551,
+      "learning_rate": 1.9118720620220172e-05,
+      "loss": 2.9111,
+      "step": 54221
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8028191924095154,
+      "learning_rate": 1.9113312635964593e-05,
+      "loss": 2.9427,
+      "step": 54222
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7735779881477356,
+      "learning_rate": 1.9107905391510437e-05,
+      "loss": 2.6901,
+      "step": 54223
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7522846460342407,
+      "learning_rate": 1.910249888687202e-05,
+      "loss": 2.7894,
+      "step": 54224
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7548792362213135,
+      "learning_rate": 1.9097093122063534e-05,
+      "loss": 2.8319,
+      "step": 54225
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7677998542785645,
+      "learning_rate": 1.9091688097099266e-05,
+      "loss": 2.9,
+      "step": 54226
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.752396821975708,
+      "learning_rate": 1.908628381199344e-05,
+      "loss": 3.1708,
+      "step": 54227
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7591312527656555,
+      "learning_rate": 1.9080880266760312e-05,
+      "loss": 2.8663,
+      "step": 54228
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7363387942314148,
+      "learning_rate": 1.9075477461414067e-05,
+      "loss": 2.9872,
+      "step": 54229
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7834699153900146,
+      "learning_rate": 1.90700753959689e-05,
+      "loss": 2.6909,
+      "step": 54230
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7395800352096558,
+      "learning_rate": 1.906467407043909e-05,
+      "loss": 2.9373,
+      "step": 54231
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7479794025421143,
+      "learning_rate": 1.90592734848389e-05,
+      "loss": 2.7438,
+      "step": 54232
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7837538719177246,
+      "learning_rate": 1.9053873639182484e-05,
+      "loss": 2.7192,
+      "step": 54233
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8077936768531799,
+      "learning_rate": 1.9048474533484126e-05,
+      "loss": 2.7285,
+      "step": 54234
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7308048605918884,
+      "learning_rate": 1.9043076167758055e-05,
+      "loss": 3.0173,
+      "step": 54235
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7778046131134033,
+      "learning_rate": 1.9037678542018418e-05,
+      "loss": 2.8399,
+      "step": 54236
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7745692133903503,
+      "learning_rate": 1.9032281656279412e-05,
+      "loss": 2.9058,
+      "step": 54237
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7720127105712891,
+      "learning_rate": 1.9026885510555324e-05,
+      "loss": 2.9293,
+      "step": 54238
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7536343932151794,
+      "learning_rate": 1.902149010486037e-05,
+      "loss": 2.8993,
+      "step": 54239
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8127491474151611,
+      "learning_rate": 1.901609543920868e-05,
+      "loss": 2.9735,
+      "step": 54240
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7343817949295044,
+      "learning_rate": 1.9010701513614603e-05,
+      "loss": 2.9745,
+      "step": 54241
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7082345485687256,
+      "learning_rate": 1.9005308328092196e-05,
+      "loss": 3.0584,
+      "step": 54242
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7429682612419128,
+      "learning_rate": 1.8999915882655746e-05,
+      "loss": 2.7716,
+      "step": 54243
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7653400897979736,
+      "learning_rate": 1.899452417731948e-05,
+      "loss": 2.8173,
+      "step": 54244
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7767915725708008,
+      "learning_rate": 1.8989133212097485e-05,
+      "loss": 2.7113,
+      "step": 54245
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7101877331733704,
+      "learning_rate": 1.8983742987004113e-05,
+      "loss": 2.7319,
+      "step": 54246
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7132937908172607,
+      "learning_rate": 1.8978353502053388e-05,
+      "loss": 2.6421,
+      "step": 54247
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7609696984291077,
+      "learning_rate": 1.8972964757259633e-05,
+      "loss": 3.0005,
+      "step": 54248
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7206118106842041,
+      "learning_rate": 1.8967576752637036e-05,
+      "loss": 3.1412,
+      "step": 54249
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7709072828292847,
+      "learning_rate": 1.8962189488199754e-05,
+      "loss": 2.8475,
+      "step": 54250
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7561743259429932,
+      "learning_rate": 1.895680296396197e-05,
+      "loss": 2.7725,
+      "step": 54251
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7318551540374756,
+      "learning_rate": 1.8951417179937844e-05,
+      "loss": 2.8117,
+      "step": 54252
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7357078790664673,
+      "learning_rate": 1.89460321361416e-05,
+      "loss": 2.6899,
+      "step": 54253
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7411123514175415,
+      "learning_rate": 1.8940647832587452e-05,
+      "loss": 2.9539,
+      "step": 54254
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7445243000984192,
+      "learning_rate": 1.8935264269289497e-05,
+      "loss": 2.8118,
+      "step": 54255
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7355711460113525,
+      "learning_rate": 1.892988144626202e-05,
+      "loss": 2.8894,
+      "step": 54256
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7821559906005859,
+      "learning_rate": 1.892449936351914e-05,
+      "loss": 2.8614,
+      "step": 54257
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7362728714942932,
+      "learning_rate": 1.8919118021074985e-05,
+      "loss": 3.1394,
+      "step": 54258
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7443636655807495,
+      "learning_rate": 1.891373741894384e-05,
+      "loss": 2.877,
+      "step": 54259
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.745068371295929,
+      "learning_rate": 1.8908357557139732e-05,
+      "loss": 2.8646,
+      "step": 54260
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7140920162200928,
+      "learning_rate": 1.890297843567701e-05,
+      "loss": 2.768,
+      "step": 54261
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7033805847167969,
+      "learning_rate": 1.889760005456966e-05,
+      "loss": 2.8209,
+      "step": 54262
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7535343766212463,
+      "learning_rate": 1.8892222413831948e-05,
+      "loss": 2.92,
+      "step": 54263
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7832685112953186,
+      "learning_rate": 1.8886845513478054e-05,
+      "loss": 2.8499,
+      "step": 54264
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7332696914672852,
+      "learning_rate": 1.8881469353522107e-05,
+      "loss": 3.1087,
+      "step": 54265
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.727424144744873,
+      "learning_rate": 1.8876093933978285e-05,
+      "loss": 3.0913,
+      "step": 54266
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7173473834991455,
+      "learning_rate": 1.8870719254860688e-05,
+      "loss": 3.0397,
+      "step": 54267
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7304912805557251,
+      "learning_rate": 1.88653453161835e-05,
+      "loss": 2.8918,
+      "step": 54268
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8100420832633972,
+      "learning_rate": 1.885997211796094e-05,
+      "loss": 2.7616,
+      "step": 54269
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8588573336601257,
+      "learning_rate": 1.885459966020704e-05,
+      "loss": 2.7817,
+      "step": 54270
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7639005780220032,
+      "learning_rate": 1.8849227942936074e-05,
+      "loss": 3.0441,
+      "step": 54271
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7512505054473877,
+      "learning_rate": 1.884385696616214e-05,
+      "loss": 2.9208,
+      "step": 54272
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7321940660476685,
+      "learning_rate": 1.8838486729899327e-05,
+      "loss": 2.8442,
+      "step": 54273
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7625482678413391,
+      "learning_rate": 1.8833117234161886e-05,
+      "loss": 2.9836,
+      "step": 54274
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7443647384643555,
+      "learning_rate": 1.8827748478963844e-05,
+      "loss": 2.8452,
+      "step": 54275
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7935992479324341,
+      "learning_rate": 1.882238046431942e-05,
+      "loss": 2.7888,
+      "step": 54276
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7821605205535889,
+      "learning_rate": 1.8817013190242734e-05,
+      "loss": 2.7943,
+      "step": 54277
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7425271272659302,
+      "learning_rate": 1.8811646656747948e-05,
+      "loss": 3.0639,
+      "step": 54278
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.702383816242218,
+      "learning_rate": 1.8806280863849145e-05,
+      "loss": 2.6989,
+      "step": 54279
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7176997661590576,
+      "learning_rate": 1.880091581156048e-05,
+      "loss": 2.9766,
+      "step": 54280
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7591765522956848,
+      "learning_rate": 1.879555149989611e-05,
+      "loss": 2.663,
+      "step": 54281
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7460857033729553,
+      "learning_rate": 1.8790187928870092e-05,
+      "loss": 2.9084,
+      "step": 54282
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7594284415245056,
+      "learning_rate": 1.878482509849658e-05,
+      "loss": 2.7448,
+      "step": 54283
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7766778469085693,
+      "learning_rate": 1.8779463008789797e-05,
+      "loss": 2.9685,
+      "step": 54284
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7691056132316589,
+      "learning_rate": 1.877410165976373e-05,
+      "loss": 3.0542,
+      "step": 54285
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7344174385070801,
+      "learning_rate": 1.8768741051432567e-05,
+      "loss": 3.0607,
+      "step": 54286
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7203677892684937,
+      "learning_rate": 1.8763381183810468e-05,
+      "loss": 2.8125,
+      "step": 54287
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7336849570274353,
+      "learning_rate": 1.8758022056911414e-05,
+      "loss": 2.8562,
+      "step": 54288
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7745854258537292,
+      "learning_rate": 1.875266367074967e-05,
+      "loss": 2.9174,
+      "step": 54289
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7709341049194336,
+      "learning_rate": 1.8747306025339247e-05,
+      "loss": 2.6841,
+      "step": 54290
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8152937293052673,
+      "learning_rate": 1.874194912069431e-05,
+      "loss": 2.9381,
+      "step": 54291
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7402138710021973,
+      "learning_rate": 1.873659295682891e-05,
+      "loss": 2.6456,
+      "step": 54292
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7518551349639893,
+      "learning_rate": 1.8731237533757272e-05,
+      "loss": 2.8282,
+      "step": 54293
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7548336386680603,
+      "learning_rate": 1.872588285149338e-05,
+      "loss": 2.7088,
+      "step": 54294
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7288030982017517,
+      "learning_rate": 1.8720528910051356e-05,
+      "loss": 2.9081,
+      "step": 54295
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7580604553222656,
+      "learning_rate": 1.8715175709445362e-05,
+      "loss": 2.9925,
+      "step": 54296
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7717550992965698,
+      "learning_rate": 1.8709823249689414e-05,
+      "loss": 2.8237,
+      "step": 54297
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.706786572933197,
+      "learning_rate": 1.870447153079767e-05,
+      "loss": 2.9454,
+      "step": 54298
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7198885083198547,
+      "learning_rate": 1.8699120552784252e-05,
+      "loss": 2.728,
+      "step": 54299
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7398534417152405,
+      "learning_rate": 1.8693770315663215e-05,
+      "loss": 3.0013,
+      "step": 54300
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7463058829307556,
+      "learning_rate": 1.8688420819448645e-05,
+      "loss": 3.0623,
+      "step": 54301
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7530163526535034,
+      "learning_rate": 1.86830720641546e-05,
+      "loss": 2.8843,
+      "step": 54302
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7228906750679016,
+      "learning_rate": 1.8677724049795174e-05,
+      "loss": 3.0058,
+      "step": 54303
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.759463906288147,
+      "learning_rate": 1.8672376776384547e-05,
+      "loss": 2.649,
+      "step": 54304
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7761176824569702,
+      "learning_rate": 1.8667030243936708e-05,
+      "loss": 2.94,
+      "step": 54305
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7421549558639526,
+      "learning_rate": 1.8661684452465787e-05,
+      "loss": 2.8615,
+      "step": 54306
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7677056193351746,
+      "learning_rate": 1.8656339401985797e-05,
+      "loss": 2.7058,
+      "step": 54307
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7426798939704895,
+      "learning_rate": 1.8650995092510934e-05,
+      "loss": 2.8782,
+      "step": 54308
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7160890698432922,
+      "learning_rate": 1.8645651524055184e-05,
+      "loss": 2.8999,
+      "step": 54309
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7567172646522522,
+      "learning_rate": 1.86403086966326e-05,
+      "loss": 2.9189,
+      "step": 54310
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7990744113922119,
+      "learning_rate": 1.863496661025734e-05,
+      "loss": 2.9437,
+      "step": 54311
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7247997522354126,
+      "learning_rate": 1.862962526494336e-05,
+      "loss": 2.8656,
+      "step": 54312
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7634393572807312,
+      "learning_rate": 1.8624284660704815e-05,
+      "loss": 3.0048,
+      "step": 54313
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7473614811897278,
+      "learning_rate": 1.8618944797555824e-05,
+      "loss": 3.1732,
+      "step": 54314
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7102734446525574,
+      "learning_rate": 1.8613605675510345e-05,
+      "loss": 2.7127,
+      "step": 54315
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7932694554328918,
+      "learning_rate": 1.860826729458247e-05,
+      "loss": 2.9487,
+      "step": 54316
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7079724669456482,
+      "learning_rate": 1.8602929654786247e-05,
+      "loss": 2.798,
+      "step": 54317
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7377166748046875,
+      "learning_rate": 1.859759275613577e-05,
+      "loss": 2.9455,
+      "step": 54318
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8409765958786011,
+      "learning_rate": 1.859225659864506e-05,
+      "loss": 2.8922,
+      "step": 54319
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7548701167106628,
+      "learning_rate": 1.8586921182328173e-05,
+      "loss": 2.875,
+      "step": 54320
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7892856001853943,
+      "learning_rate": 1.858158650719923e-05,
+      "loss": 3.0358,
+      "step": 54321
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7447632551193237,
+      "learning_rate": 1.8576252573272155e-05,
+      "loss": 3.0135,
+      "step": 54322
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7343568205833435,
+      "learning_rate": 1.85709193805612e-05,
+      "loss": 2.9174,
+      "step": 54323
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7546588778495789,
+      "learning_rate": 1.8565586929080156e-05,
+      "loss": 3.1323,
+      "step": 54324
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7940450310707092,
+      "learning_rate": 1.8560255218843213e-05,
+      "loss": 2.7931,
+      "step": 54325
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7970552444458008,
+      "learning_rate": 1.8554924249864423e-05,
+      "loss": 3.0955,
+      "step": 54326
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7288427352905273,
+      "learning_rate": 1.8549594022157775e-05,
+      "loss": 2.8776,
+      "step": 54327
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8218358159065247,
+      "learning_rate": 1.8544264535737364e-05,
+      "loss": 2.8839,
+      "step": 54328
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7489563226699829,
+      "learning_rate": 1.8538935790617136e-05,
+      "loss": 2.6361,
+      "step": 54329
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7439214587211609,
+      "learning_rate": 1.8533607786811257e-05,
+      "loss": 2.7068,
+      "step": 54330
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7576811909675598,
+      "learning_rate": 1.8528280524333637e-05,
+      "loss": 2.7559,
+      "step": 54331
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7353628873825073,
+      "learning_rate": 1.8522954003198342e-05,
+      "loss": 2.9782,
+      "step": 54332
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7459601163864136,
+      "learning_rate": 1.851762822341949e-05,
+      "loss": 2.9133,
+      "step": 54333
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7878472208976746,
+      "learning_rate": 1.8512303185010936e-05,
+      "loss": 2.8641,
+      "step": 54334
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7898440957069397,
+      "learning_rate": 1.8506978887986835e-05,
+      "loss": 2.9113,
+      "step": 54335
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7459374666213989,
+      "learning_rate": 1.850165533236121e-05,
+      "loss": 3.0018,
+      "step": 54336
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7519286870956421,
+      "learning_rate": 1.849633251814805e-05,
+      "loss": 2.7782,
+      "step": 54337
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7953622341156006,
+      "learning_rate": 1.8491010445361376e-05,
+      "loss": 2.8511,
+      "step": 54338
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7520739436149597,
+      "learning_rate": 1.8485689114015146e-05,
+      "loss": 2.993,
+      "step": 54339
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7880058884620667,
+      "learning_rate": 1.8480368524123445e-05,
+      "loss": 2.6661,
+      "step": 54340
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7255048751831055,
+      "learning_rate": 1.84750486757003e-05,
+      "loss": 2.8142,
+      "step": 54341
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7529579997062683,
+      "learning_rate": 1.8469729568759662e-05,
+      "loss": 2.6264,
+      "step": 54342
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7077673077583313,
+      "learning_rate": 1.8464411203315655e-05,
+      "loss": 3.071,
+      "step": 54343
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7397609949111938,
+      "learning_rate": 1.84590935793821e-05,
+      "loss": 2.9625,
+      "step": 54344
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7642784714698792,
+      "learning_rate": 1.8453776696973187e-05,
+      "loss": 2.8608,
+      "step": 54345
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7922868728637695,
+      "learning_rate": 1.844846055610284e-05,
+      "loss": 2.6326,
+      "step": 54346
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7658751606941223,
+      "learning_rate": 1.8443145156785012e-05,
+      "loss": 2.8669,
+      "step": 54347
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7865391969680786,
+      "learning_rate": 1.8437830499033822e-05,
+      "loss": 2.7268,
+      "step": 54348
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7619051337242126,
+      "learning_rate": 1.843251658286313e-05,
+      "loss": 2.6806,
+      "step": 54349
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7652508616447449,
+      "learning_rate": 1.8427203408286994e-05,
+      "loss": 2.8923,
+      "step": 54350
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.718989372253418,
+      "learning_rate": 1.8421890975319465e-05,
+      "loss": 3.2317,
+      "step": 54351
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.757768988609314,
+      "learning_rate": 1.84165792839745e-05,
+      "loss": 2.825,
+      "step": 54352
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7357819676399231,
+      "learning_rate": 1.841126833426605e-05,
+      "loss": 3.0808,
+      "step": 54353
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7870209217071533,
+      "learning_rate": 1.840595812620811e-05,
+      "loss": 2.8403,
+      "step": 54354
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7332582473754883,
+      "learning_rate": 1.8400648659814664e-05,
+      "loss": 2.9084,
+      "step": 54355
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7562378644943237,
+      "learning_rate": 1.8395339935099772e-05,
+      "loss": 2.9404,
+      "step": 54356
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7957108020782471,
+      "learning_rate": 1.8390031952077287e-05,
+      "loss": 2.9112,
+      "step": 54357
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7108379006385803,
+      "learning_rate": 1.8384724710761334e-05,
+      "loss": 2.8825,
+      "step": 54358
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7623836398124695,
+      "learning_rate": 1.8379418211165796e-05,
+      "loss": 3.0071,
+      "step": 54359
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7817762494087219,
+      "learning_rate": 1.8374112453304634e-05,
+      "loss": 2.6759,
+      "step": 54360
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8142674565315247,
+      "learning_rate": 1.83688074371919e-05,
+      "loss": 2.9088,
+      "step": 54361
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7520138621330261,
+      "learning_rate": 1.836350316284152e-05,
+      "loss": 2.876,
+      "step": 54362
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7290118336677551,
+      "learning_rate": 1.835819963026748e-05,
+      "loss": 2.8468,
+      "step": 54363
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7808724641799927,
+      "learning_rate": 1.83528968394837e-05,
+      "loss": 2.9203,
+      "step": 54364
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7678559422492981,
+      "learning_rate": 1.834759479050417e-05,
+      "loss": 2.9192,
+      "step": 54365
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7148266434669495,
+      "learning_rate": 1.834229348334292e-05,
+      "loss": 2.7605,
+      "step": 54366
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.714754045009613,
+      "learning_rate": 1.8336992918013894e-05,
+      "loss": 2.8237,
+      "step": 54367
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7212704420089722,
+      "learning_rate": 1.8331693094530985e-05,
+      "loss": 2.8664,
+      "step": 54368
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7763723731040955,
+      "learning_rate": 1.8326394012908153e-05,
+      "loss": 2.7022,
+      "step": 54369
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7871028780937195,
+      "learning_rate": 1.8321095673159414e-05,
+      "loss": 2.9555,
+      "step": 54370
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7615554928779602,
+      "learning_rate": 1.8315798075298728e-05,
+      "loss": 2.8821,
+      "step": 54371
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7369310259819031,
+      "learning_rate": 1.8310501219339947e-05,
+      "loss": 2.7315,
+      "step": 54372
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7455233931541443,
+      "learning_rate": 1.8305205105297162e-05,
+      "loss": 3.1036,
+      "step": 54373
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7491878867149353,
+      "learning_rate": 1.829990973318426e-05,
+      "loss": 2.8756,
+      "step": 54374
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8159949779510498,
+      "learning_rate": 1.829461510301513e-05,
+      "loss": 2.8182,
+      "step": 54375
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7366239428520203,
+      "learning_rate": 1.828932121480383e-05,
+      "loss": 2.8917,
+      "step": 54376
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7913699150085449,
+      "learning_rate": 1.8284028068564183e-05,
+      "loss": 3.1585,
+      "step": 54377
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7340931296348572,
+      "learning_rate": 1.8278735664310208e-05,
+      "loss": 2.9116,
+      "step": 54378
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7172591686248779,
+      "learning_rate": 1.8273444002055828e-05,
+      "loss": 2.9317,
+      "step": 54379
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7502052783966064,
+      "learning_rate": 1.8268153081814994e-05,
+      "loss": 2.9903,
+      "step": 54380
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7538665533065796,
+      "learning_rate": 1.8262862903601606e-05,
+      "loss": 3.0005,
+      "step": 54381
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7647122740745544,
+      "learning_rate": 1.8257573467429608e-05,
+      "loss": 2.9253,
+      "step": 54382
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7346746921539307,
+      "learning_rate": 1.8252284773312964e-05,
+      "loss": 2.8669,
+      "step": 54383
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7537041306495667,
+      "learning_rate": 1.8246996821265557e-05,
+      "loss": 2.7086,
+      "step": 54384
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.74997878074646,
+      "learning_rate": 1.8241709611301313e-05,
+      "loss": 2.7979,
+      "step": 54385
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7360642552375793,
+      "learning_rate": 1.823642314343422e-05,
+      "loss": 2.7866,
+      "step": 54386
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7375506162643433,
+      "learning_rate": 1.8231137417678133e-05,
+      "loss": 2.8911,
+      "step": 54387
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7365680932998657,
+      "learning_rate": 1.8225852434047038e-05,
+      "loss": 2.8581,
+      "step": 54388
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7538241147994995,
+      "learning_rate": 1.822056819255483e-05,
+      "loss": 2.9333,
+      "step": 54389
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7558910846710205,
+      "learning_rate": 1.821528469321536e-05,
+      "loss": 2.7096,
+      "step": 54390
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7916548252105713,
+      "learning_rate": 1.8210001936042652e-05,
+      "loss": 2.9807,
+      "step": 54391
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.1039891242980957,
+      "learning_rate": 1.8204719921050492e-05,
+      "loss": 3.1819,
+      "step": 54392
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.734824538230896,
+      "learning_rate": 1.819943864825294e-05,
+      "loss": 2.6976,
+      "step": 54393
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7892003655433655,
+      "learning_rate": 1.819415811766378e-05,
+      "loss": 3.0551,
+      "step": 54394
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7752202749252319,
+      "learning_rate": 1.818887832929704e-05,
+      "loss": 2.8143,
+      "step": 54395
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7521406412124634,
+      "learning_rate": 1.818359928316654e-05,
+      "loss": 2.8164,
+      "step": 54396
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7605895400047302,
+      "learning_rate": 1.8178320979286164e-05,
+      "loss": 2.945,
+      "step": 54397
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8226302862167358,
+      "learning_rate": 1.8173043417669908e-05,
+      "loss": 2.8654,
+      "step": 54398
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7271564602851868,
+      "learning_rate": 1.8167766598331558e-05,
+      "loss": 2.9806,
+      "step": 54399
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7447865009307861,
+      "learning_rate": 1.81624905212851e-05,
+      "loss": 2.9206,
+      "step": 54400
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7446723580360413,
+      "learning_rate": 1.815721518654436e-05,
+      "loss": 3.0564,
+      "step": 54401
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7353099584579468,
+      "learning_rate": 1.815194059412336e-05,
+      "loss": 3.0469,
+      "step": 54402
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7726219296455383,
+      "learning_rate": 1.8146666744035852e-05,
+      "loss": 2.9872,
+      "step": 54403
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7748692035675049,
+      "learning_rate": 1.8141393636295764e-05,
+      "loss": 2.8217,
+      "step": 54404
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7315913438796997,
+      "learning_rate": 1.813612127091705e-05,
+      "loss": 2.871,
+      "step": 54405
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7552880644798279,
+      "learning_rate": 1.8130849647913492e-05,
+      "loss": 2.8242,
+      "step": 54406
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.726218581199646,
+      "learning_rate": 1.812557876729902e-05,
+      "loss": 2.8563,
+      "step": 54407
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7617534399032593,
+      "learning_rate": 1.8120308629087587e-05,
+      "loss": 2.8983,
+      "step": 54408
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7186073064804077,
+      "learning_rate": 1.811503923329295e-05,
+      "loss": 2.9901,
+      "step": 54409
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7599871754646301,
+      "learning_rate": 1.810977057992913e-05,
+      "loss": 2.8198,
+      "step": 54410
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7000196576118469,
+      "learning_rate": 1.810450266900988e-05,
+      "loss": 2.902,
+      "step": 54411
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7205071449279785,
+      "learning_rate": 1.8099235500549092e-05,
+      "loss": 2.6609,
+      "step": 54412
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7560635209083557,
+      "learning_rate": 1.809396907456072e-05,
+      "loss": 2.936,
+      "step": 54413
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7410910129547119,
+      "learning_rate": 1.8088703391058523e-05,
+      "loss": 2.9418,
+      "step": 54414
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7488538026809692,
+      "learning_rate": 1.8083438450056487e-05,
+      "loss": 2.6645,
+      "step": 54415
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7629244923591614,
+      "learning_rate": 1.8078174251568333e-05,
+      "loss": 2.7683,
+      "step": 54416
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7645202279090881,
+      "learning_rate": 1.8072910795608088e-05,
+      "loss": 2.9679,
+      "step": 54417
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7587352395057678,
+      "learning_rate": 1.8067648082189536e-05,
+      "loss": 2.8207,
+      "step": 54418
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7425807118415833,
+      "learning_rate": 1.80623861113265e-05,
+      "loss": 3.1311,
+      "step": 54419
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7521542906761169,
+      "learning_rate": 1.8057124883032904e-05,
+      "loss": 3.1478,
+      "step": 54420
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7295096516609192,
+      "learning_rate": 1.805186439732257e-05,
+      "loss": 3.0066,
+      "step": 54421
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7017486095428467,
+      "learning_rate": 1.8046604654209317e-05,
+      "loss": 2.8773,
+      "step": 54422
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7578108310699463,
+      "learning_rate": 1.8041345653707106e-05,
+      "loss": 2.8673,
+      "step": 54423
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.771789014339447,
+      "learning_rate": 1.8036087395829758e-05,
+      "loss": 2.9348,
+      "step": 54424
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7727000117301941,
+      "learning_rate": 1.8030829880591057e-05,
+      "loss": 2.6452,
+      "step": 54425
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7393551468849182,
+      "learning_rate": 1.802557310800483e-05,
+      "loss": 3.0339,
+      "step": 54426
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7382111549377441,
+      "learning_rate": 1.8020317078085e-05,
+      "loss": 2.8841,
+      "step": 54427
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7548832297325134,
+      "learning_rate": 1.801506179084542e-05,
+      "loss": 3.0882,
+      "step": 54428
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.747711718082428,
+      "learning_rate": 1.800980724629988e-05,
+      "loss": 2.8361,
+      "step": 54429
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7639087438583374,
+      "learning_rate": 1.8004553444462265e-05,
+      "loss": 2.8952,
+      "step": 54430
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7807071208953857,
+      "learning_rate": 1.799930038534634e-05,
+      "loss": 2.8251,
+      "step": 54431
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7129517197608948,
+      "learning_rate": 1.7994048068966018e-05,
+      "loss": 2.9591,
+      "step": 54432
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7340091466903687,
+      "learning_rate": 1.7988796495335088e-05,
+      "loss": 2.899,
+      "step": 54433
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7638095617294312,
+      "learning_rate": 1.798354566446738e-05,
+      "loss": 2.9115,
+      "step": 54434
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7314512729644775,
+      "learning_rate": 1.797829557637678e-05,
+      "loss": 3.0898,
+      "step": 54435
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7057316899299622,
+      "learning_rate": 1.7973046231077037e-05,
+      "loss": 2.7938,
+      "step": 54436
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7399982213973999,
+      "learning_rate": 1.796779762858198e-05,
+      "loss": 2.8381,
+      "step": 54437
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7212401032447815,
+      "learning_rate": 1.7962549768905532e-05,
+      "loss": 2.5719,
+      "step": 54438
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7532590627670288,
+      "learning_rate": 1.7957302652061444e-05,
+      "loss": 2.8661,
+      "step": 54439
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7178873419761658,
+      "learning_rate": 1.7952056278063542e-05,
+      "loss": 2.9399,
+      "step": 54440
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7503253221511841,
+      "learning_rate": 1.794681064692558e-05,
+      "loss": 3.0511,
+      "step": 54441
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7387188673019409,
+      "learning_rate": 1.794156575866148e-05,
+      "loss": 2.8939,
+      "step": 54442
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7133545279502869,
+      "learning_rate": 1.793632161328503e-05,
+      "loss": 2.8469,
+      "step": 54443
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7400594353675842,
+      "learning_rate": 1.7931078210809958e-05,
+      "loss": 2.8812,
+      "step": 54444
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7527070641517639,
+      "learning_rate": 1.792583555125021e-05,
+      "loss": 2.89,
+      "step": 54445
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7184710502624512,
+      "learning_rate": 1.7920593634619483e-05,
+      "loss": 3.0791,
+      "step": 54446
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7595903873443604,
+      "learning_rate": 1.7915352460931664e-05,
+      "loss": 2.9072,
+      "step": 54447
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7046235799789429,
+      "learning_rate": 1.7910112030200542e-05,
+      "loss": 2.66,
+      "step": 54448
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.728029727935791,
+      "learning_rate": 1.7904872342439803e-05,
+      "loss": 2.9842,
+      "step": 54449
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7607099413871765,
+      "learning_rate": 1.7899633397663406e-05,
+      "loss": 2.8981,
+      "step": 54450
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7469753623008728,
+      "learning_rate": 1.789439519588507e-05,
+      "loss": 2.8449,
+      "step": 54451
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7470847368240356,
+      "learning_rate": 1.7889157737118554e-05,
+      "loss": 3.0769,
+      "step": 54452
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7539829611778259,
+      "learning_rate": 1.7883921021377777e-05,
+      "loss": 2.7504,
+      "step": 54453
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7649708986282349,
+      "learning_rate": 1.7878685048676467e-05,
+      "loss": 2.929,
+      "step": 54454
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7718356847763062,
+      "learning_rate": 1.7873449819028372e-05,
+      "loss": 3.0847,
+      "step": 54455
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7208101153373718,
+      "learning_rate": 1.786821533244729e-05,
+      "loss": 3.0591,
+      "step": 54456
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7297426462173462,
+      "learning_rate": 1.7862981588947032e-05,
+      "loss": 2.8899,
+      "step": 54457
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7583280801773071,
+      "learning_rate": 1.785774858854143e-05,
+      "loss": 2.958,
+      "step": 54458
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7504193186759949,
+      "learning_rate": 1.785251633124417e-05,
+      "loss": 2.8244,
+      "step": 54459
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.78138267993927,
+      "learning_rate": 1.7847284817069107e-05,
+      "loss": 2.8839,
+      "step": 54460
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7497425675392151,
+      "learning_rate": 1.784205404603003e-05,
+      "loss": 2.8106,
+      "step": 54461
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8312497735023499,
+      "learning_rate": 1.7836824018140628e-05,
+      "loss": 2.8887,
+      "step": 54462
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7753313779830933,
+      "learning_rate": 1.7831594733414757e-05,
+      "loss": 2.8713,
+      "step": 54463
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7584692239761353,
+      "learning_rate": 1.7826366191866105e-05,
+      "loss": 2.8697,
+      "step": 54464
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.799094021320343,
+      "learning_rate": 1.782113839350856e-05,
+      "loss": 2.9074,
+      "step": 54465
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7895060777664185,
+      "learning_rate": 1.781591133835578e-05,
+      "loss": 2.7601,
+      "step": 54466
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7699868083000183,
+      "learning_rate": 1.7810685026421588e-05,
+      "loss": 2.8981,
+      "step": 54467
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7355355620384216,
+      "learning_rate": 1.7805459457719805e-05,
+      "loss": 2.941,
+      "step": 54468
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7582590579986572,
+      "learning_rate": 1.780023463226412e-05,
+      "loss": 2.6234,
+      "step": 54469
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7793235182762146,
+      "learning_rate": 1.779501055006829e-05,
+      "loss": 2.7773,
+      "step": 54470
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7583927512168884,
+      "learning_rate": 1.7789787211146067e-05,
+      "loss": 2.8734,
+      "step": 54471
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7218800187110901,
+      "learning_rate": 1.7784564615511244e-05,
+      "loss": 2.9254,
+      "step": 54472
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7587677240371704,
+      "learning_rate": 1.7779342763177572e-05,
+      "loss": 2.8399,
+      "step": 54473
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7498154640197754,
+      "learning_rate": 1.7774121654158745e-05,
+      "loss": 2.8976,
+      "step": 54474
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7345309257507324,
+      "learning_rate": 1.7768901288468652e-05,
+      "loss": 2.9342,
+      "step": 54475
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.758578360080719,
+      "learning_rate": 1.7763681666120942e-05,
+      "loss": 2.8475,
+      "step": 54476
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7470417618751526,
+      "learning_rate": 1.7758462787129346e-05,
+      "loss": 2.9,
+      "step": 54477
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.750644326210022,
+      "learning_rate": 1.7753244651507648e-05,
+      "loss": 2.934,
+      "step": 54478
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7721165418624878,
+      "learning_rate": 1.774802725926954e-05,
+      "loss": 2.9427,
+      "step": 54479
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7746807336807251,
+      "learning_rate": 1.774281061042887e-05,
+      "loss": 3.035,
+      "step": 54480
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7496808767318726,
+      "learning_rate": 1.773759470499927e-05,
+      "loss": 3.0492,
+      "step": 54481
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7302466630935669,
+      "learning_rate": 1.773237954299459e-05,
+      "loss": 2.8497,
+      "step": 54482
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7932860851287842,
+      "learning_rate": 1.7727165124428455e-05,
+      "loss": 2.9827,
+      "step": 54483
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7311192154884338,
+      "learning_rate": 1.7721951449314654e-05,
+      "loss": 3.0511,
+      "step": 54484
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7317575812339783,
+      "learning_rate": 1.7716738517666906e-05,
+      "loss": 2.745,
+      "step": 54485
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7372452020645142,
+      "learning_rate": 1.7711526329498906e-05,
+      "loss": 2.8255,
+      "step": 54486
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7596213221549988,
+      "learning_rate": 1.770631488482447e-05,
+      "loss": 2.8587,
+      "step": 54487
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7744680643081665,
+      "learning_rate": 1.770110418365722e-05,
+      "loss": 2.7446,
+      "step": 54488
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7649564743041992,
+      "learning_rate": 1.7695894226010953e-05,
+      "loss": 2.7356,
+      "step": 54489
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7747321724891663,
+      "learning_rate": 1.7690685011899385e-05,
+      "loss": 2.8946,
+      "step": 54490
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7615075707435608,
+      "learning_rate": 1.7685476541336242e-05,
+      "loss": 2.9659,
+      "step": 54491
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7637103796005249,
+      "learning_rate": 1.768026881433521e-05,
+      "loss": 2.7237,
+      "step": 54492
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7580251693725586,
+      "learning_rate": 1.7675061830909976e-05,
+      "loss": 3.0487,
+      "step": 54493
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7343547940254211,
+      "learning_rate": 1.76698555910743e-05,
+      "loss": 2.7701,
+      "step": 54494
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7214182615280151,
+      "learning_rate": 1.7664650094841937e-05,
+      "loss": 2.8685,
+      "step": 54495
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7279322147369385,
+      "learning_rate": 1.7659445342226473e-05,
+      "loss": 2.869,
+      "step": 54496
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7043575048446655,
+      "learning_rate": 1.765424133324177e-05,
+      "loss": 3.072,
+      "step": 54497
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7192384600639343,
+      "learning_rate": 1.7649038067901445e-05,
+      "loss": 2.9503,
+      "step": 54498
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7618083357810974,
+      "learning_rate": 1.764383554621919e-05,
+      "loss": 2.9244,
+      "step": 54499
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7713229656219482,
+      "learning_rate": 1.7638633768208753e-05,
+      "loss": 2.8477,
+      "step": 54500
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.761009931564331,
+      "learning_rate": 1.763343273388377e-05,
+      "loss": 2.7269,
+      "step": 54501
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7590358257293701,
+      "learning_rate": 1.7628232443258016e-05,
+      "loss": 2.6942,
+      "step": 54502
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7583818435668945,
+      "learning_rate": 1.7623032896345123e-05,
+      "loss": 2.8209,
+      "step": 54503
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7613120675086975,
+      "learning_rate": 1.7617834093158878e-05,
+      "loss": 2.9961,
+      "step": 54504
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7189532518386841,
+      "learning_rate": 1.76126360337129e-05,
+      "loss": 3.0106,
+      "step": 54505
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7215641736984253,
+      "learning_rate": 1.7607438718020848e-05,
+      "loss": 2.9489,
+      "step": 54506
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7565929889678955,
+      "learning_rate": 1.760224214609648e-05,
+      "loss": 2.9464,
+      "step": 54507
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8237677216529846,
+      "learning_rate": 1.7597046317953444e-05,
+      "loss": 3.1464,
+      "step": 54508
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7600015997886658,
+      "learning_rate": 1.7591851233605405e-05,
+      "loss": 2.7613,
+      "step": 54509
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7842651009559631,
+      "learning_rate": 1.7586656893066144e-05,
+      "loss": 2.8621,
+      "step": 54510
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7692767977714539,
+      "learning_rate": 1.7581463296349252e-05,
+      "loss": 2.752,
+      "step": 54511
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7494772672653198,
+      "learning_rate": 1.7576270443468454e-05,
+      "loss": 2.8582,
+      "step": 54512
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.742486298084259,
+      "learning_rate": 1.7571078334437405e-05,
+      "loss": 2.8898,
+      "step": 54513
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7653214335441589,
+      "learning_rate": 1.7565886969269728e-05,
+      "loss": 3.1248,
+      "step": 54514
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7274088859558105,
+      "learning_rate": 1.756069634797921e-05,
+      "loss": 2.7404,
+      "step": 54515
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7633993625640869,
+      "learning_rate": 1.7555506470579404e-05,
+      "loss": 2.8211,
+      "step": 54516
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7414184808731079,
+      "learning_rate": 1.7550317337084107e-05,
+      "loss": 2.8421,
+      "step": 54517
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7275792956352234,
+      "learning_rate": 1.7545128947506836e-05,
+      "loss": 2.9229,
+      "step": 54518
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7488593459129333,
+      "learning_rate": 1.7539941301861414e-05,
+      "loss": 2.8447,
+      "step": 54519
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8635866641998291,
+      "learning_rate": 1.7534754400161398e-05,
+      "loss": 2.927,
+      "step": 54520
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7082469463348389,
+      "learning_rate": 1.7529568242420445e-05,
+      "loss": 2.8341,
+      "step": 54521
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7854346632957458,
+      "learning_rate": 1.7524382828652307e-05,
+      "loss": 3.1416,
+      "step": 54522
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7766318321228027,
+      "learning_rate": 1.7519198158870507e-05,
+      "loss": 3.0129,
+      "step": 54523
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7670121788978577,
+      "learning_rate": 1.7514014233088804e-05,
+      "loss": 2.9533,
+      "step": 54524
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7497897148132324,
+      "learning_rate": 1.7508831051320848e-05,
+      "loss": 2.9596,
+      "step": 54525
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7375215888023376,
+      "learning_rate": 1.7503648613580266e-05,
+      "loss": 3.1485,
+      "step": 54526
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7439850568771362,
+      "learning_rate": 1.7498466919880717e-05,
+      "loss": 2.9282,
+      "step": 54527
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7792314887046814,
+      "learning_rate": 1.7493285970235783e-05,
+      "loss": 3.0288,
+      "step": 54528
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7756353616714478,
+      "learning_rate": 1.7488105764659156e-05,
+      "loss": 2.7858,
+      "step": 54529
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7475228905677795,
+      "learning_rate": 1.7482926303164557e-05,
+      "loss": 2.8076,
+      "step": 54530
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7505745887756348,
+      "learning_rate": 1.7477747585765478e-05,
+      "loss": 2.9298,
+      "step": 54531
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7236295342445374,
+      "learning_rate": 1.7472569612475705e-05,
+      "loss": 2.9612,
+      "step": 54532
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7206342816352844,
+      "learning_rate": 1.7467392383308765e-05,
+      "loss": 2.8051,
+      "step": 54533
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7881414294242859,
+      "learning_rate": 1.7462215898278408e-05,
+      "loss": 3.0171,
+      "step": 54534
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7353194952011108,
+      "learning_rate": 1.745704015739816e-05,
+      "loss": 2.9517,
+      "step": 54535
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7909258008003235,
+      "learning_rate": 1.7451865160681678e-05,
+      "loss": 2.8948,
+      "step": 54536
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7583587765693665,
+      "learning_rate": 1.7446690908142613e-05,
+      "loss": 2.7776,
+      "step": 54537
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7082347273826599,
+      "learning_rate": 1.744151739979456e-05,
+      "loss": 2.8779,
+      "step": 54538
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7351507544517517,
+      "learning_rate": 1.7436344635651167e-05,
+      "loss": 3.0366,
+      "step": 54539
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7974845170974731,
+      "learning_rate": 1.743117261572613e-05,
+      "loss": 2.8264,
+      "step": 54540
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.736272394657135,
+      "learning_rate": 1.742600134003297e-05,
+      "loss": 3.0235,
+      "step": 54541
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8139263391494751,
+      "learning_rate": 1.7420830808585373e-05,
+      "loss": 2.8582,
+      "step": 54542
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7663759589195251,
+      "learning_rate": 1.7415661021396865e-05,
+      "loss": 2.965,
+      "step": 54543
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7662950754165649,
+      "learning_rate": 1.74104919784811e-05,
+      "loss": 2.7665,
+      "step": 54544
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7392333149909973,
+      "learning_rate": 1.7405323679851768e-05,
+      "loss": 2.9475,
+      "step": 54545
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7623760104179382,
+      "learning_rate": 1.740015612552239e-05,
+      "loss": 2.9041,
+      "step": 54546
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7544460296630859,
+      "learning_rate": 1.739498931550669e-05,
+      "loss": 2.6681,
+      "step": 54547
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7664691209793091,
+      "learning_rate": 1.7389823249818124e-05,
+      "loss": 2.902,
+      "step": 54548
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7272012829780579,
+      "learning_rate": 1.7384657928470448e-05,
+      "loss": 2.7784,
+      "step": 54549
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.747775137424469,
+      "learning_rate": 1.7379493351477114e-05,
+      "loss": 2.6218,
+      "step": 54550
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7551100850105286,
+      "learning_rate": 1.7374329518851814e-05,
+      "loss": 2.5569,
+      "step": 54551
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7571572661399841,
+      "learning_rate": 1.7369166430608172e-05,
+      "loss": 2.973,
+      "step": 54552
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7804607152938843,
+      "learning_rate": 1.7364004086759707e-05,
+      "loss": 2.9752,
+      "step": 54553
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.739106297492981,
+      "learning_rate": 1.7358842487320113e-05,
+      "loss": 2.9756,
+      "step": 54554
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7254763841629028,
+      "learning_rate": 1.7353681632302873e-05,
+      "loss": 3.0763,
+      "step": 54555
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7539757490158081,
+      "learning_rate": 1.7348521521721713e-05,
+      "loss": 2.7454,
+      "step": 54556
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7677503228187561,
+      "learning_rate": 1.7343362155590123e-05,
+      "loss": 2.9946,
+      "step": 54557
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7571077346801758,
+      "learning_rate": 1.733820353392169e-05,
+      "loss": 3.0087,
+      "step": 54558
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7669923901557922,
+      "learning_rate": 1.733304565673007e-05,
+      "loss": 2.9143,
+      "step": 54559
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.73554927110672,
+      "learning_rate": 1.7327888524028755e-05,
+      "loss": 2.9558,
+      "step": 54560
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7662096619606018,
+      "learning_rate": 1.7322732135831396e-05,
+      "loss": 2.7881,
+      "step": 54561
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7233344912528992,
+      "learning_rate": 1.7317576492151587e-05,
+      "loss": 2.831,
+      "step": 54562
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7593091726303101,
+      "learning_rate": 1.731242159300288e-05,
+      "loss": 2.9025,
+      "step": 54563
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7726497650146484,
+      "learning_rate": 1.7307267438398865e-05,
+      "loss": 2.8003,
+      "step": 54564
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.765876829624176,
+      "learning_rate": 1.7302114028353032e-05,
+      "loss": 2.9848,
+      "step": 54565
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7233142852783203,
+      "learning_rate": 1.7296961362879036e-05,
+      "loss": 2.9807,
+      "step": 54566
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7348734140396118,
+      "learning_rate": 1.72918094419905e-05,
+      "loss": 2.7916,
+      "step": 54567
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.769679069519043,
+      "learning_rate": 1.728665826570088e-05,
+      "loss": 2.8728,
+      "step": 54568
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7657221555709839,
+      "learning_rate": 1.7281507834023832e-05,
+      "loss": 2.8739,
+      "step": 54569
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7933992743492126,
+      "learning_rate": 1.727635814697281e-05,
+      "loss": 2.812,
+      "step": 54570
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7689788937568665,
+      "learning_rate": 1.727120920456151e-05,
+      "loss": 2.8414,
+      "step": 54571
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7207183837890625,
+      "learning_rate": 1.7266061006803444e-05,
+      "loss": 2.7137,
+      "step": 54572
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7436949610710144,
+      "learning_rate": 1.7260913553712107e-05,
+      "loss": 2.9456,
+      "step": 54573
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7527655363082886,
+      "learning_rate": 1.7255766845301157e-05,
+      "loss": 2.8905,
+      "step": 54574
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7638602256774902,
+      "learning_rate": 1.7250620881584077e-05,
+      "loss": 2.7258,
+      "step": 54575
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7570220232009888,
+      "learning_rate": 1.7245475662574428e-05,
+      "loss": 2.6924,
+      "step": 54576
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7669765949249268,
+      "learning_rate": 1.724033118828583e-05,
+      "loss": 2.797,
+      "step": 54577
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7875416278839111,
+      "learning_rate": 1.723518745873177e-05,
+      "loss": 2.7294,
+      "step": 54578
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7687140703201294,
+      "learning_rate": 1.723004447392581e-05,
+      "loss": 2.9238,
+      "step": 54579
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7571902275085449,
+      "learning_rate": 1.7224902233881433e-05,
+      "loss": 2.9232,
+      "step": 54580
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7242570519447327,
+      "learning_rate": 1.7219760738612264e-05,
+      "loss": 2.9942,
+      "step": 54581
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7527267336845398,
+      "learning_rate": 1.721461998813186e-05,
+      "loss": 2.8456,
+      "step": 54582
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7353515028953552,
+      "learning_rate": 1.720947998245368e-05,
+      "loss": 2.8948,
+      "step": 54583
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7328376770019531,
+      "learning_rate": 1.7204340721591335e-05,
+      "loss": 2.8345,
+      "step": 54584
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7656261920928955,
+      "learning_rate": 1.719920220555836e-05,
+      "loss": 2.8635,
+      "step": 54585
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7430219650268555,
+      "learning_rate": 1.7194064434368206e-05,
+      "loss": 2.8203,
+      "step": 54586
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7474393844604492,
+      "learning_rate": 1.718892740803449e-05,
+      "loss": 2.9497,
+      "step": 54587
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.75533527135849,
+      "learning_rate": 1.7183791126570678e-05,
+      "loss": 2.9873,
+      "step": 54588
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.776290237903595,
+      "learning_rate": 1.7178655589990387e-05,
+      "loss": 2.8499,
+      "step": 54589
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.743800699710846,
+      "learning_rate": 1.717352079830704e-05,
+      "loss": 2.7576,
+      "step": 54590
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7225151658058167,
+      "learning_rate": 1.7168386751534192e-05,
+      "loss": 2.8802,
+      "step": 54591
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7379876971244812,
+      "learning_rate": 1.716325344968543e-05,
+      "loss": 2.721,
+      "step": 54592
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7471773028373718,
+      "learning_rate": 1.715812089277422e-05,
+      "loss": 2.7912,
+      "step": 54593
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7949302792549133,
+      "learning_rate": 1.7152989080814106e-05,
+      "loss": 2.771,
+      "step": 54594
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7417853474617004,
+      "learning_rate": 1.7147858013818517e-05,
+      "loss": 2.923,
+      "step": 54595
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7578975558280945,
+      "learning_rate": 1.714272769180104e-05,
+      "loss": 2.8479,
+      "step": 54596
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.77840256690979,
+      "learning_rate": 1.7137598114775232e-05,
+      "loss": 2.7247,
+      "step": 54597
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7270179986953735,
+      "learning_rate": 1.7132469282754513e-05,
+      "loss": 2.7254,
+      "step": 54598
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7670302391052246,
+      "learning_rate": 1.7127341195752475e-05,
+      "loss": 2.8518,
+      "step": 54599
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7435368895530701,
+      "learning_rate": 1.7122213853782572e-05,
+      "loss": 2.9362,
+      "step": 54600
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7616455554962158,
+      "learning_rate": 1.7117087256858262e-05,
+      "loss": 2.7466,
+      "step": 54601
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7222978472709656,
+      "learning_rate": 1.711196140499317e-05,
+      "loss": 2.826,
+      "step": 54602
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7720111012458801,
+      "learning_rate": 1.7106836298200675e-05,
+      "loss": 2.9449,
+      "step": 54603
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7425702214241028,
+      "learning_rate": 1.7101711936494378e-05,
+      "loss": 3.0421,
+      "step": 54604
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7751484513282776,
+      "learning_rate": 1.709658831988766e-05,
+      "loss": 3.0768,
+      "step": 54605
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.731267511844635,
+      "learning_rate": 1.7091465448394148e-05,
+      "loss": 2.7284,
+      "step": 54606
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7741591930389404,
+      "learning_rate": 1.7086343322027263e-05,
+      "loss": 2.9372,
+      "step": 54607
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7876058220863342,
+      "learning_rate": 1.708122194080046e-05,
+      "loss": 2.9196,
+      "step": 54608
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7788753509521484,
+      "learning_rate": 1.7076101304727296e-05,
+      "loss": 2.8865,
+      "step": 54609
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7326127290725708,
+      "learning_rate": 1.7070981413821227e-05,
+      "loss": 2.9553,
+      "step": 54610
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7484868764877319,
+      "learning_rate": 1.706586226809571e-05,
+      "loss": 2.9617,
+      "step": 54611
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7754936814308167,
+      "learning_rate": 1.706074386756433e-05,
+      "loss": 2.8784,
+      "step": 54612
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8197011947631836,
+      "learning_rate": 1.7055626212240413e-05,
+      "loss": 2.9881,
+      "step": 54613
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7415262460708618,
+      "learning_rate": 1.7050509302137584e-05,
+      "loss": 2.9983,
+      "step": 54614
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7331694960594177,
+      "learning_rate": 1.704539313726929e-05,
+      "loss": 2.9561,
+      "step": 54615
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7985134720802307,
+      "learning_rate": 1.70402777176489e-05,
+      "loss": 2.9792,
+      "step": 54616
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7267589569091797,
+      "learning_rate": 1.7035163043290022e-05,
+      "loss": 2.8766,
+      "step": 54617
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7310107946395874,
+      "learning_rate": 1.703004911420599e-05,
+      "loss": 2.987,
+      "step": 54618
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7549310326576233,
+      "learning_rate": 1.702493593041042e-05,
+      "loss": 2.8447,
+      "step": 54619
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7123493552207947,
+      "learning_rate": 1.7019823491916673e-05,
+      "loss": 3.1187,
+      "step": 54620
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7951881885528564,
+      "learning_rate": 1.7014711798738268e-05,
+      "loss": 2.8816,
+      "step": 54621
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.725439190864563,
+      "learning_rate": 1.7009600850888693e-05,
+      "loss": 2.9719,
+      "step": 54622
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7655175924301147,
+      "learning_rate": 1.7004490648381274e-05,
+      "loss": 2.8131,
+      "step": 54623
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7330092191696167,
+      "learning_rate": 1.6999381191229634e-05,
+      "loss": 2.9509,
+      "step": 54624
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7745311856269836,
+      "learning_rate": 1.6994272479447124e-05,
+      "loss": 2.7781,
+      "step": 54625
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7374331951141357,
+      "learning_rate": 1.6989164513047238e-05,
+      "loss": 3.0763,
+      "step": 54626
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7071889042854309,
+      "learning_rate": 1.698405729204343e-05,
+      "loss": 2.7063,
+      "step": 54627
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7430211901664734,
+      "learning_rate": 1.6978950816449153e-05,
+      "loss": 3.0131,
+      "step": 54628
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.746117889881134,
+      "learning_rate": 1.697384508627787e-05,
+      "loss": 2.7566,
+      "step": 54629
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7581526041030884,
+      "learning_rate": 1.6968740101542965e-05,
+      "loss": 3.0198,
+      "step": 54630
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7774915099143982,
+      "learning_rate": 1.6963635862257996e-05,
+      "loss": 3.039,
+      "step": 54631
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7348177433013916,
+      "learning_rate": 1.695853236843625e-05,
+      "loss": 3.1106,
+      "step": 54632
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7624771595001221,
+      "learning_rate": 1.6953429620091284e-05,
+      "loss": 3.0076,
+      "step": 54633
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7285212874412537,
+      "learning_rate": 1.6948327617236557e-05,
+      "loss": 2.7591,
+      "step": 54634
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8262482285499573,
+      "learning_rate": 1.6943226359885385e-05,
+      "loss": 3.0077,
+      "step": 54635
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7384718656539917,
+      "learning_rate": 1.6938125848051363e-05,
+      "loss": 2.8514,
+      "step": 54636
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7086389064788818,
+      "learning_rate": 1.6933026081747814e-05,
+      "loss": 2.9698,
+      "step": 54637
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7307590246200562,
+      "learning_rate": 1.6927927060988156e-05,
+      "loss": 2.8567,
+      "step": 54638
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7535133957862854,
+      "learning_rate": 1.6922828785785912e-05,
+      "loss": 2.821,
+      "step": 54639
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7423950433731079,
+      "learning_rate": 1.6917731256154443e-05,
+      "loss": 2.8587,
+      "step": 54640
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7522428035736084,
+      "learning_rate": 1.69126344721072e-05,
+      "loss": 2.9274,
+      "step": 54641
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7286722660064697,
+      "learning_rate": 1.6907538433657543e-05,
+      "loss": 2.7226,
+      "step": 54642
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7772533297538757,
+      "learning_rate": 1.690244314081902e-05,
+      "loss": 2.9504,
+      "step": 54643
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7469460964202881,
+      "learning_rate": 1.6897348593604965e-05,
+      "loss": 2.8693,
+      "step": 54644
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7319382429122925,
+      "learning_rate": 1.6892254792028757e-05,
+      "loss": 2.8529,
+      "step": 54645
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7492722272872925,
+      "learning_rate": 1.6887161736103893e-05,
+      "loss": 2.8104,
+      "step": 54646
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7667667865753174,
+      "learning_rate": 1.688206942584376e-05,
+      "loss": 2.8156,
+      "step": 54647
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.6894829869270325,
+      "learning_rate": 1.6876977861261744e-05,
+      "loss": 2.8207,
+      "step": 54648
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7800548672676086,
+      "learning_rate": 1.6871887042371334e-05,
+      "loss": 2.9097,
+      "step": 54649
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7656291127204895,
+      "learning_rate": 1.6866796969185827e-05,
+      "loss": 2.7117,
+      "step": 54650
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7687351107597351,
+      "learning_rate": 1.686170764171877e-05,
+      "loss": 2.8556,
+      "step": 54651
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7354574203491211,
+      "learning_rate": 1.6856619059983423e-05,
+      "loss": 2.8432,
+      "step": 54652
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7067791819572449,
+      "learning_rate": 1.685153122399321e-05,
+      "loss": 2.976,
+      "step": 54653
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7321593165397644,
+      "learning_rate": 1.684644413376165e-05,
+      "loss": 2.8863,
+      "step": 54654
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7532573342323303,
+      "learning_rate": 1.6841357789302e-05,
+      "loss": 2.8874,
+      "step": 54655
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7408707141876221,
+      "learning_rate": 1.6836272190627787e-05,
+      "loss": 3.2351,
+      "step": 54656
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7293885946273804,
+      "learning_rate": 1.683118733775226e-05,
+      "loss": 2.8992,
+      "step": 54657
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7179583311080933,
+      "learning_rate": 1.682610323068895e-05,
+      "loss": 2.8923,
+      "step": 54658
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7808348536491394,
+      "learning_rate": 1.6821019869451167e-05,
+      "loss": 3.0609,
+      "step": 54659
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7644177675247192,
+      "learning_rate": 1.6815937254052314e-05,
+      "loss": 2.9049,
+      "step": 54660
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7662442326545715,
+      "learning_rate": 1.6810855384505804e-05,
+      "loss": 2.8631,
+      "step": 54661
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7450445890426636,
+      "learning_rate": 1.6805774260824968e-05,
+      "loss": 2.835,
+      "step": 54662
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7103903889656067,
+      "learning_rate": 1.6800693883023185e-05,
+      "loss": 2.7432,
+      "step": 54663
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7254341244697571,
+      "learning_rate": 1.6795614251113953e-05,
+      "loss": 3.0087,
+      "step": 54664
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7645226716995239,
+      "learning_rate": 1.6790535365110557e-05,
+      "loss": 2.6737,
+      "step": 54665
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7188624143600464,
+      "learning_rate": 1.6785457225026387e-05,
+      "loss": 2.6871,
+      "step": 54666
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7680506110191345,
+      "learning_rate": 1.678037983087477e-05,
+      "loss": 2.8942,
+      "step": 54667
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7478795051574707,
+      "learning_rate": 1.677530318266912e-05,
+      "loss": 2.8025,
+      "step": 54668
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7576417922973633,
+      "learning_rate": 1.6770227280422866e-05,
+      "loss": 2.8098,
+      "step": 54669
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7802990078926086,
+      "learning_rate": 1.6765152124149262e-05,
+      "loss": 2.8422,
+      "step": 54670
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7456069588661194,
+      "learning_rate": 1.6760077713861797e-05,
+      "loss": 2.8779,
+      "step": 54671
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7156590819358826,
+      "learning_rate": 1.675500404957376e-05,
+      "loss": 2.8347,
+      "step": 54672
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7358556389808655,
+      "learning_rate": 1.6749931131298545e-05,
+      "loss": 3.0632,
+      "step": 54673
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7578690648078918,
+      "learning_rate": 1.6744858959049502e-05,
+      "loss": 2.8503,
+      "step": 54674
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7345317602157593,
+      "learning_rate": 1.6739787532839922e-05,
+      "loss": 2.9099,
+      "step": 54675
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7033917903900146,
+      "learning_rate": 1.6734716852683295e-05,
+      "loss": 2.906,
+      "step": 54676
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.6990805268287659,
+      "learning_rate": 1.6729646918592875e-05,
+      "loss": 2.9649,
+      "step": 54677
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7403311729431152,
+      "learning_rate": 1.6724577730582024e-05,
+      "loss": 2.6907,
+      "step": 54678
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.770260214805603,
+      "learning_rate": 1.671950928866419e-05,
+      "loss": 2.8517,
+      "step": 54679
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7441731691360474,
+      "learning_rate": 1.6714441592852633e-05,
+      "loss": 3.0706,
+      "step": 54680
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7643983960151672,
+      "learning_rate": 1.670937464316071e-05,
+      "loss": 2.9851,
+      "step": 54681
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7585827708244324,
+      "learning_rate": 1.6704308439601745e-05,
+      "loss": 2.8638,
+      "step": 54682
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7482290863990784,
+      "learning_rate": 1.6699242982189088e-05,
+      "loss": 2.8541,
+      "step": 54683
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7718914151191711,
+      "learning_rate": 1.6694178270936167e-05,
+      "loss": 2.9846,
+      "step": 54684
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8131400346755981,
+      "learning_rate": 1.66891143058562e-05,
+      "loss": 2.7676,
+      "step": 54685
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7344086170196533,
+      "learning_rate": 1.6684051086962648e-05,
+      "loss": 2.9111,
+      "step": 54686
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7355520725250244,
+      "learning_rate": 1.6678988614268762e-05,
+      "loss": 3.0156,
+      "step": 54687
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7679204940795898,
+      "learning_rate": 1.667392688778787e-05,
+      "loss": 2.7147,
+      "step": 54688
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7418351173400879,
+      "learning_rate": 1.6668865907533356e-05,
+      "loss": 2.9869,
+      "step": 54689
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7173027992248535,
+      "learning_rate": 1.666380567351845e-05,
+      "loss": 2.8766,
+      "step": 54690
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7610350251197815,
+      "learning_rate": 1.6658746185756632e-05,
+      "loss": 3.1118,
+      "step": 54691
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7524661421775818,
+      "learning_rate": 1.66536874442611e-05,
+      "loss": 2.8634,
+      "step": 54692
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7599107623100281,
+      "learning_rate": 1.6648629449045204e-05,
+      "loss": 2.6746,
+      "step": 54693
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7229357361793518,
+      "learning_rate": 1.6643572200122336e-05,
+      "loss": 2.9954,
+      "step": 54694
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8016541600227356,
+      "learning_rate": 1.663851569750578e-05,
+      "loss": 2.8605,
+      "step": 54695
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7787773609161377,
+      "learning_rate": 1.663345994120884e-05,
+      "loss": 3.0242,
+      "step": 54696
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.768181562423706,
+      "learning_rate": 1.6628404931244755e-05,
+      "loss": 2.7763,
+      "step": 54697
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7640612721443176,
+      "learning_rate": 1.6623350667626955e-05,
+      "loss": 2.9869,
+      "step": 54698
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7562863230705261,
+      "learning_rate": 1.6618297150368733e-05,
+      "loss": 2.7484,
+      "step": 54699
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.755027711391449,
+      "learning_rate": 1.6613244379483336e-05,
+      "loss": 2.9133,
+      "step": 54700
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7754169702529907,
+      "learning_rate": 1.6608192354984162e-05,
+      "loss": 3.0882,
+      "step": 54701
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7277271151542664,
+      "learning_rate": 1.6603141076884495e-05,
+      "loss": 2.8785,
+      "step": 54702
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7051646113395691,
+      "learning_rate": 1.659809054519753e-05,
+      "loss": 2.8702,
+      "step": 54703
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7471274137496948,
+      "learning_rate": 1.6593040759936715e-05,
+      "loss": 2.8466,
+      "step": 54704
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7694349884986877,
+      "learning_rate": 1.6587991721115245e-05,
+      "loss": 2.6614,
+      "step": 54705
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7603945732116699,
+      "learning_rate": 1.658294342874651e-05,
+      "loss": 2.8695,
+      "step": 54706
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7334092259407043,
+      "learning_rate": 1.6577895882843693e-05,
+      "loss": 2.7742,
+      "step": 54707
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7664968371391296,
+      "learning_rate": 1.6572849083420225e-05,
+      "loss": 2.977,
+      "step": 54708
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7536723613739014,
+      "learning_rate": 1.6567803030489292e-05,
+      "loss": 2.8594,
+      "step": 54709
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7267215251922607,
+      "learning_rate": 1.656275772406418e-05,
+      "loss": 2.8231,
+      "step": 54710
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7520238161087036,
+      "learning_rate": 1.6557713164158283e-05,
+      "loss": 2.8955,
+      "step": 54711
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7576550245285034,
+      "learning_rate": 1.6552669350784753e-05,
+      "loss": 2.8592,
+      "step": 54712
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7513378262519836,
+      "learning_rate": 1.6547626283957016e-05,
+      "loss": 2.9951,
+      "step": 54713
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7380017638206482,
+      "learning_rate": 1.6542583963688195e-05,
+      "loss": 2.9589,
+      "step": 54714
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7709711194038391,
+      "learning_rate": 1.6537542389991676e-05,
+      "loss": 2.7643,
+      "step": 54715
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7406794428825378,
+      "learning_rate": 1.6532501562880784e-05,
+      "loss": 3.0594,
+      "step": 54716
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7626711130142212,
+      "learning_rate": 1.652746148236871e-05,
+      "loss": 2.7846,
+      "step": 54717
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7286007404327393,
+      "learning_rate": 1.6522422148468738e-05,
+      "loss": 3.0798,
+      "step": 54718
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.740604817867279,
+      "learning_rate": 1.65173835611941e-05,
+      "loss": 2.8224,
+      "step": 54719
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8206152319908142,
+      "learning_rate": 1.651234572055814e-05,
+      "loss": 2.9789,
+      "step": 54720
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7075120210647583,
+      "learning_rate": 1.6507308626574122e-05,
+      "loss": 2.9661,
+      "step": 54721
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.775321364402771,
+      "learning_rate": 1.650227227925527e-05,
+      "loss": 2.9403,
+      "step": 54722
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7323336005210876,
+      "learning_rate": 1.64972366786149e-05,
+      "loss": 2.9916,
+      "step": 54723
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7589771151542664,
+      "learning_rate": 1.6492201824666274e-05,
+      "loss": 2.7981,
+      "step": 54724
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7782710194587708,
+      "learning_rate": 1.6487167717422544e-05,
+      "loss": 2.9767,
+      "step": 54725
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7076034545898438,
+      "learning_rate": 1.6482134356897138e-05,
+      "loss": 2.6416,
+      "step": 54726
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7385815382003784,
+      "learning_rate": 1.6477101743103138e-05,
+      "loss": 2.7569,
+      "step": 54727
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7414993047714233,
+      "learning_rate": 1.6472069876053973e-05,
+      "loss": 2.775,
+      "step": 54728
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7520781755447388,
+      "learning_rate": 1.646703875576273e-05,
+      "loss": 2.914,
+      "step": 54729
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7415902614593506,
+      "learning_rate": 1.64620083822428e-05,
+      "loss": 2.8865,
+      "step": 54730
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7842666506767273,
+      "learning_rate": 1.6456978755507366e-05,
+      "loss": 3.0898,
+      "step": 54731
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7630308866500854,
+      "learning_rate": 1.6451949875569658e-05,
+      "loss": 2.8664,
+      "step": 54732
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7267240285873413,
+      "learning_rate": 1.6446921742442963e-05,
+      "loss": 2.9599,
+      "step": 54733
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7458666563034058,
+      "learning_rate": 1.6441894356140474e-05,
+      "loss": 2.6751,
+      "step": 54734
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7139987349510193,
+      "learning_rate": 1.6436867716675473e-05,
+      "loss": 2.7982,
+      "step": 54735
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7688502073287964,
+      "learning_rate": 1.643184182406122e-05,
+      "loss": 2.8371,
+      "step": 54736
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7603595852851868,
+      "learning_rate": 1.6426816678310873e-05,
+      "loss": 3.0501,
+      "step": 54737
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7536769509315491,
+      "learning_rate": 1.6421792279437752e-05,
+      "loss": 2.9334,
+      "step": 54738
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7776528000831604,
+      "learning_rate": 1.641676862745508e-05,
+      "loss": 2.8698,
+      "step": 54739
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7579948902130127,
+      "learning_rate": 1.6411745722376014e-05,
+      "loss": 2.7414,
+      "step": 54740
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7683404684066772,
+      "learning_rate": 1.6406723564213843e-05,
+      "loss": 2.9881,
+      "step": 54741
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7273842692375183,
+      "learning_rate": 1.640170215298179e-05,
+      "loss": 2.7239,
+      "step": 54742
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7698961496353149,
+      "learning_rate": 1.6396681488693077e-05,
+      "loss": 2.672,
+      "step": 54743
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7344798445701599,
+      "learning_rate": 1.639166157136089e-05,
+      "loss": 2.9347,
+      "step": 54744
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7438050508499146,
+      "learning_rate": 1.638664240099856e-05,
+      "loss": 2.9031,
+      "step": 54745
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7391680479049683,
+      "learning_rate": 1.6381623977619206e-05,
+      "loss": 3.0264,
+      "step": 54746
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7523117661476135,
+      "learning_rate": 1.6376606301236017e-05,
+      "loss": 2.9136,
+      "step": 54747
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7371009588241577,
+      "learning_rate": 1.6371589371862315e-05,
+      "loss": 3.0831,
+      "step": 54748
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.718420147895813,
+      "learning_rate": 1.6366573189511224e-05,
+      "loss": 2.6528,
+      "step": 54749
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7344409227371216,
+      "learning_rate": 1.6361557754195997e-05,
+      "loss": 2.9695,
+      "step": 54750
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7529169321060181,
+      "learning_rate": 1.6356543065929862e-05,
+      "loss": 2.8281,
+      "step": 54751
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7365326881408691,
+      "learning_rate": 1.6351529124726005e-05,
+      "loss": 2.85,
+      "step": 54752
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7705813050270081,
+      "learning_rate": 1.6346515930597648e-05,
+      "loss": 3.0134,
+      "step": 54753
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7423902750015259,
+      "learning_rate": 1.634150348355795e-05,
+      "loss": 2.9457,
+      "step": 54754
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7638031244277954,
+      "learning_rate": 1.6336491783620097e-05,
+      "loss": 2.8813,
+      "step": 54755
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7596058249473572,
+      "learning_rate": 1.6331480830797416e-05,
+      "loss": 2.9061,
+      "step": 54756
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7403856515884399,
+      "learning_rate": 1.632647062510296e-05,
+      "loss": 3.0641,
+      "step": 54757
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7983343005180359,
+      "learning_rate": 1.632146116655002e-05,
+      "loss": 2.8901,
+      "step": 54758
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7619073987007141,
+      "learning_rate": 1.631645245515172e-05,
+      "loss": 2.8923,
+      "step": 54759
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7451263666152954,
+      "learning_rate": 1.6311444490921344e-05,
+      "loss": 2.9146,
+      "step": 54760
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7346713542938232,
+      "learning_rate": 1.6306437273872018e-05,
+      "loss": 2.8883,
+      "step": 54761
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7482584118843079,
+      "learning_rate": 1.6301430804016902e-05,
+      "loss": 2.8736,
+      "step": 54762
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7432484030723572,
+      "learning_rate": 1.6296425081369248e-05,
+      "loss": 2.9137,
+      "step": 54763
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7410925030708313,
+      "learning_rate": 1.6291420105942178e-05,
+      "loss": 2.7335,
+      "step": 54764
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7755419015884399,
+      "learning_rate": 1.6286415877748915e-05,
+      "loss": 2.6624,
+      "step": 54765
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7352432608604431,
+      "learning_rate": 1.628141239680265e-05,
+      "loss": 2.8728,
+      "step": 54766
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7407886385917664,
+      "learning_rate": 1.627640966311654e-05,
+      "loss": 2.9961,
+      "step": 54767
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7732663750648499,
+      "learning_rate": 1.6271407676703773e-05,
+      "loss": 2.774,
+      "step": 54768
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.8005004525184631,
+      "learning_rate": 1.6266406437577506e-05,
+      "loss": 3.0023,
+      "step": 54769
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7332467436790466,
+      "learning_rate": 1.626140594575086e-05,
+      "loss": 3.1547,
+      "step": 54770
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7251038551330566,
+      "learning_rate": 1.625640620123716e-05,
+      "loss": 2.958,
+      "step": 54771
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7420737743377686,
+      "learning_rate": 1.6251407204049394e-05,
+      "loss": 2.9342,
+      "step": 54772
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7631248831748962,
+      "learning_rate": 1.6246408954200883e-05,
+      "loss": 2.882,
+      "step": 54773
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7106769680976868,
+      "learning_rate": 1.6241411451704656e-05,
+      "loss": 2.754,
+      "step": 54774
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.77427077293396,
+      "learning_rate": 1.6236414696573994e-05,
+      "loss": 3.0064,
+      "step": 54775
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7603651285171509,
+      "learning_rate": 1.6231418688821997e-05,
+      "loss": 2.6856,
+      "step": 54776
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7737959623336792,
+      "learning_rate": 1.622642342846181e-05,
+      "loss": 2.7616,
+      "step": 54777
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7463186979293823,
+      "learning_rate": 1.622142891550663e-05,
+      "loss": 2.7421,
+      "step": 54778
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7298126816749573,
+      "learning_rate": 1.6216435149969574e-05,
+      "loss": 2.9434,
+      "step": 54779
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7219801545143127,
+      "learning_rate": 1.6211442131863772e-05,
+      "loss": 2.9801,
+      "step": 54780
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7264379858970642,
+      "learning_rate": 1.620644986120251e-05,
+      "loss": 2.9004,
+      "step": 54781
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7708029747009277,
+      "learning_rate": 1.6201458337998807e-05,
+      "loss": 2.9669,
+      "step": 54782
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7728287577629089,
+      "learning_rate": 1.6196467562265857e-05,
+      "loss": 2.7538,
+      "step": 54783
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7423531413078308,
+      "learning_rate": 1.619147753401675e-05,
+      "loss": 2.9775,
+      "step": 54784
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7586064338684082,
+      "learning_rate": 1.6186488253264707e-05,
+      "loss": 3.0552,
+      "step": 54785
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7348181009292603,
+      "learning_rate": 1.6181499720022784e-05,
+      "loss": 2.7265,
+      "step": 54786
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7246251702308655,
+      "learning_rate": 1.617651193430417e-05,
+      "loss": 2.7787,
+      "step": 54787
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7123945355415344,
+      "learning_rate": 1.6171524896122056e-05,
+      "loss": 2.9102,
+      "step": 54788
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7603493332862854,
+      "learning_rate": 1.6166538605489497e-05,
+      "loss": 2.6584,
+      "step": 54789
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7190308570861816,
+      "learning_rate": 1.6161553062419687e-05,
+      "loss": 2.8735,
+      "step": 54790
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7607088088989258,
+      "learning_rate": 1.615656826692564e-05,
+      "loss": 2.9559,
+      "step": 54791
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7394618391990662,
+      "learning_rate": 1.6151584219020585e-05,
+      "loss": 2.9037,
+      "step": 54792
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7496686577796936,
+      "learning_rate": 1.6146600918717676e-05,
+      "loss": 2.8412,
+      "step": 54793
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7344440817832947,
+      "learning_rate": 1.6141618366029974e-05,
+      "loss": 2.8324,
+      "step": 54794
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7307662963867188,
+      "learning_rate": 1.6136636560970628e-05,
+      "loss": 2.7831,
+      "step": 54795
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7606189250946045,
+      "learning_rate": 1.61316555035527e-05,
+      "loss": 2.9232,
+      "step": 54796
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7663336992263794,
+      "learning_rate": 1.6126675193789406e-05,
+      "loss": 2.7469,
+      "step": 54797
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7470273375511169,
+      "learning_rate": 1.6121695631693843e-05,
+      "loss": 2.819,
+      "step": 54798
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7757609486579895,
+      "learning_rate": 1.611671681727903e-05,
+      "loss": 2.9269,
+      "step": 54799
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.6888236999511719,
+      "learning_rate": 1.6111738750558223e-05,
+      "loss": 2.8834,
+      "step": 54800
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7634980082511902,
+      "learning_rate": 1.610676143154438e-05,
+      "loss": 3.0092,
+      "step": 54801
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7457658648490906,
+      "learning_rate": 1.6101784860250722e-05,
+      "loss": 2.8261,
+      "step": 54802
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7501567602157593,
+      "learning_rate": 1.6096809036690373e-05,
+      "loss": 2.6977,
+      "step": 54803
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.771567165851593,
+      "learning_rate": 1.6091833960876355e-05,
+      "loss": 3.0102,
+      "step": 54804
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7606413960456848,
+      "learning_rate": 1.6086859632821826e-05,
+      "loss": 3.0169,
+      "step": 54805
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7576195001602173,
+      "learning_rate": 1.608188605253984e-05,
+      "loss": 2.7713,
+      "step": 54806
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7587432861328125,
+      "learning_rate": 1.6076913220043487e-05,
+      "loss": 2.7597,
+      "step": 54807
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7228325605392456,
+      "learning_rate": 1.607194113534599e-05,
+      "loss": 3.0178,
+      "step": 54808
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7396401166915894,
+      "learning_rate": 1.6066969798460306e-05,
+      "loss": 2.9906,
+      "step": 54809
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.744340181350708,
+      "learning_rate": 1.606199920939959e-05,
+      "loss": 2.8594,
+      "step": 54810
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7660406827926636,
+      "learning_rate": 1.6057029368176933e-05,
+      "loss": 2.9305,
+      "step": 54811
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7232922911643982,
+      "learning_rate": 1.6052060274805388e-05,
+      "loss": 2.6945,
+      "step": 54812
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7410985827445984,
+      "learning_rate": 1.6047091929298118e-05,
+      "loss": 2.826,
+      "step": 54813
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7533641457557678,
+      "learning_rate": 1.6042124331668106e-05,
+      "loss": 2.9354,
+      "step": 54814
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8169004321098328,
+      "learning_rate": 1.6037157481928542e-05,
+      "loss": 2.8995,
+      "step": 54815
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7725902199745178,
+      "learning_rate": 1.6032191380092385e-05,
+      "loss": 2.8986,
+      "step": 54816
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7658239603042603,
+      "learning_rate": 1.6027226026172822e-05,
+      "loss": 2.7877,
+      "step": 54817
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7260845899581909,
+      "learning_rate": 1.6022261420182915e-05,
+      "loss": 2.7825,
+      "step": 54818
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7706580758094788,
+      "learning_rate": 1.601729756213571e-05,
+      "loss": 2.6989,
+      "step": 54819
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7238854169845581,
+      "learning_rate": 1.6012334452044305e-05,
+      "loss": 2.9405,
+      "step": 54820
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8242838382720947,
+      "learning_rate": 1.600737208992172e-05,
+      "loss": 2.927,
+      "step": 54821
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7421150803565979,
+      "learning_rate": 1.6002410475781048e-05,
+      "loss": 2.8967,
+      "step": 54822
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7626473903656006,
+      "learning_rate": 1.5997449609635404e-05,
+      "loss": 2.8926,
+      "step": 54823
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7503937482833862,
+      "learning_rate": 1.5992489491497783e-05,
+      "loss": 2.6612,
+      "step": 54824
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8117401599884033,
+      "learning_rate": 1.5987530121381344e-05,
+      "loss": 2.6266,
+      "step": 54825
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7335777282714844,
+      "learning_rate": 1.59825714992991e-05,
+      "loss": 2.9657,
+      "step": 54826
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8089480400085449,
+      "learning_rate": 1.5977613625264018e-05,
+      "loss": 2.5981,
+      "step": 54827
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7682913541793823,
+      "learning_rate": 1.5972656499289317e-05,
+      "loss": 2.8463,
+      "step": 54828
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7823068499565125,
+      "learning_rate": 1.596770012138795e-05,
+      "loss": 3.0189,
+      "step": 54829
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7197554707527161,
+      "learning_rate": 1.596274449157301e-05,
+      "loss": 2.952,
+      "step": 54830
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8423383831977844,
+      "learning_rate": 1.5957789609857486e-05,
+      "loss": 2.9469,
+      "step": 54831
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.742664635181427,
+      "learning_rate": 1.5952835476254534e-05,
+      "loss": 2.9605,
+      "step": 54832
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7450295090675354,
+      "learning_rate": 1.5947882090777142e-05,
+      "loss": 3.0918,
+      "step": 54833
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7666531801223755,
+      "learning_rate": 1.5942929453438335e-05,
+      "loss": 2.7747,
+      "step": 54834
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7433580756187439,
+      "learning_rate": 1.5937977564251238e-05,
+      "loss": 2.8439,
+      "step": 54835
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7616721391677856,
+      "learning_rate": 1.5933026423228766e-05,
+      "loss": 2.9655,
+      "step": 54836
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7969316840171814,
+      "learning_rate": 1.5928076030384053e-05,
+      "loss": 2.7916,
+      "step": 54837
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7451574206352234,
+      "learning_rate": 1.5923126385730145e-05,
+      "loss": 2.7925,
+      "step": 54838
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7240291833877563,
+      "learning_rate": 1.5918177489280003e-05,
+      "loss": 3.0301,
+      "step": 54839
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7478523254394531,
+      "learning_rate": 1.5913229341046752e-05,
+      "loss": 2.9261,
+      "step": 54840
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.715190589427948,
+      "learning_rate": 1.590828194104341e-05,
+      "loss": 2.8159,
+      "step": 54841
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7498929500579834,
+      "learning_rate": 1.5903335289282903e-05,
+      "loss": 2.826,
+      "step": 54842
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7427136301994324,
+      "learning_rate": 1.5898389385778388e-05,
+      "loss": 3.1507,
+      "step": 54843
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7392913699150085,
+      "learning_rate": 1.589344423054282e-05,
+      "loss": 2.8675,
+      "step": 54844
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7416170835494995,
+      "learning_rate": 1.588849982358925e-05,
+      "loss": 2.9029,
+      "step": 54845
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.767384946346283,
+      "learning_rate": 1.588355616493068e-05,
+      "loss": 2.8468,
+      "step": 54846
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8128290176391602,
+      "learning_rate": 1.5878613254580153e-05,
+      "loss": 2.8221,
+      "step": 54847
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7421417832374573,
+      "learning_rate": 1.58736710925507e-05,
+      "loss": 2.8124,
+      "step": 54848
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7256699800491333,
+      "learning_rate": 1.5868729678855276e-05,
+      "loss": 2.8736,
+      "step": 54849
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7774531841278076,
+      "learning_rate": 1.5863789013506966e-05,
+      "loss": 3.0403,
+      "step": 54850
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7691476941108704,
+      "learning_rate": 1.5858849096518732e-05,
+      "loss": 2.8938,
+      "step": 54851
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8459373712539673,
+      "learning_rate": 1.5853909927903596e-05,
+      "loss": 2.6245,
+      "step": 54852
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7290185689926147,
+      "learning_rate": 1.5848971507674612e-05,
+      "loss": 2.8709,
+      "step": 54853
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7624875903129578,
+      "learning_rate": 1.584403383584474e-05,
+      "loss": 2.9471,
+      "step": 54854
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7604098320007324,
+      "learning_rate": 1.583909691242703e-05,
+      "loss": 2.7269,
+      "step": 54855
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7130759954452515,
+      "learning_rate": 1.5834160737434377e-05,
+      "loss": 2.8143,
+      "step": 54856
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7458640933036804,
+      "learning_rate": 1.5829225310879868e-05,
+      "loss": 2.9933,
+      "step": 54857
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7971216440200806,
+      "learning_rate": 1.5824290632776525e-05,
+      "loss": 2.9175,
+      "step": 54858
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.743493378162384,
+      "learning_rate": 1.5819356703137275e-05,
+      "loss": 2.7029,
+      "step": 54859
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7640153765678406,
+      "learning_rate": 1.5814423521975172e-05,
+      "loss": 2.829,
+      "step": 54860
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7117670774459839,
+      "learning_rate": 1.5809491089303172e-05,
+      "loss": 2.9363,
+      "step": 54861
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7635127305984497,
+      "learning_rate": 1.58045594051343e-05,
+      "loss": 2.9184,
+      "step": 54862
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7638167142868042,
+      "learning_rate": 1.579962846948154e-05,
+      "loss": 2.9043,
+      "step": 54863
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7500227093696594,
+      "learning_rate": 1.579469828235782e-05,
+      "loss": 2.999,
+      "step": 54864
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.75905442237854,
+      "learning_rate": 1.5789768843776195e-05,
+      "loss": 2.7478,
+      "step": 54865
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7460845112800598,
+      "learning_rate": 1.578484015374959e-05,
+      "loss": 2.9451,
+      "step": 54866
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7751177549362183,
+      "learning_rate": 1.577991221229106e-05,
+      "loss": 2.8939,
+      "step": 54867
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7179179191589355,
+      "learning_rate": 1.577498501941349e-05,
+      "loss": 2.9286,
+      "step": 54868
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7725952863693237,
+      "learning_rate": 1.5770058575129972e-05,
+      "loss": 2.89,
+      "step": 54869
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7968480587005615,
+      "learning_rate": 1.57651328794534e-05,
+      "loss": 2.9152,
+      "step": 54870
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7715826630592346,
+      "learning_rate": 1.576020793239673e-05,
+      "loss": 2.9582,
+      "step": 54871
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7265507578849792,
+      "learning_rate": 1.575528373397301e-05,
+      "loss": 2.8741,
+      "step": 54872
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7969720363616943,
+      "learning_rate": 1.5750360284195107e-05,
+      "loss": 3.063,
+      "step": 54873
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7510287761688232,
+      "learning_rate": 1.574543758307607e-05,
+      "loss": 2.8564,
+      "step": 54874
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7693957090377808,
+      "learning_rate": 1.574051563062889e-05,
+      "loss": 2.9187,
+      "step": 54875
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8029640316963196,
+      "learning_rate": 1.573559442686646e-05,
+      "loss": 2.7687,
+      "step": 54876
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7065871953964233,
+      "learning_rate": 1.5730673971801833e-05,
+      "loss": 2.8054,
+      "step": 54877
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7508147358894348,
+      "learning_rate": 1.5725754265447798e-05,
+      "loss": 3.1097,
+      "step": 54878
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7067500352859497,
+      "learning_rate": 1.5720835307817448e-05,
+      "loss": 3.004,
+      "step": 54879
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7348147034645081,
+      "learning_rate": 1.5715917098923737e-05,
+      "loss": 3.0635,
+      "step": 54880
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7430323362350464,
+      "learning_rate": 1.571099963877952e-05,
+      "loss": 3.0832,
+      "step": 54881
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7505277395248413,
+      "learning_rate": 1.570608292739789e-05,
+      "loss": 3.0346,
+      "step": 54882
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7366390824317932,
+      "learning_rate": 1.5701166964791667e-05,
+      "loss": 3.0246,
+      "step": 54883
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7447764277458191,
+      "learning_rate": 1.569625175097391e-05,
+      "loss": 2.9516,
+      "step": 54884
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7516607046127319,
+      "learning_rate": 1.5691337285957504e-05,
+      "loss": 2.9391,
+      "step": 54885
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7200809121131897,
+      "learning_rate": 1.568642356975538e-05,
+      "loss": 3.0,
+      "step": 54886
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7547498941421509,
+      "learning_rate": 1.568151060238052e-05,
+      "loss": 2.9319,
+      "step": 54887
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7287806868553162,
+      "learning_rate": 1.5676598383845817e-05,
+      "loss": 2.8652,
+      "step": 54888
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.756786048412323,
+      "learning_rate": 1.5671686914164194e-05,
+      "loss": 3.0127,
+      "step": 54889
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7761844992637634,
+      "learning_rate": 1.5666776193348706e-05,
+      "loss": 2.817,
+      "step": 54890
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7763392925262451,
+      "learning_rate": 1.566186622141221e-05,
+      "loss": 2.9292,
+      "step": 54891
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7687212824821472,
+      "learning_rate": 1.5656956998367596e-05,
+      "loss": 3.0882,
+      "step": 54892
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7391537427902222,
+      "learning_rate": 1.5652048524227857e-05,
+      "loss": 3.0195,
+      "step": 54893
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7586456537246704,
+      "learning_rate": 1.564714079900584e-05,
+      "loss": 2.8656,
+      "step": 54894
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7237671613693237,
+      "learning_rate": 1.5642233822714612e-05,
+      "loss": 3.0369,
+      "step": 54895
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7231841683387756,
+      "learning_rate": 1.5637327595366954e-05,
+      "loss": 2.9005,
+      "step": 54896
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7463599443435669,
+      "learning_rate": 1.5632422116975863e-05,
+      "loss": 2.8008,
+      "step": 54897
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7573348879814148,
+      "learning_rate": 1.5627517387554256e-05,
+      "loss": 2.8954,
+      "step": 54898
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7449231147766113,
+      "learning_rate": 1.562261340711506e-05,
+      "loss": 2.8526,
+      "step": 54899
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7895654439926147,
+      "learning_rate": 1.5617710175671162e-05,
+      "loss": 2.8813,
+      "step": 54900
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7453767657279968,
+      "learning_rate": 1.561280769323545e-05,
+      "loss": 2.8936,
+      "step": 54901
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7513246536254883,
+      "learning_rate": 1.5607905959820887e-05,
+      "loss": 2.9529,
+      "step": 54902
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.752334475517273,
+      "learning_rate": 1.5603004975440357e-05,
+      "loss": 2.9262,
+      "step": 54903
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7545332312583923,
+      "learning_rate": 1.5598104740106788e-05,
+      "loss": 2.6894,
+      "step": 54904
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7315773963928223,
+      "learning_rate": 1.5593205253833096e-05,
+      "loss": 3.0838,
+      "step": 54905
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7586761713027954,
+      "learning_rate": 1.5588306516632175e-05,
+      "loss": 2.8601,
+      "step": 54906
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7025976181030273,
+      "learning_rate": 1.5583408528516916e-05,
+      "loss": 2.8994,
+      "step": 54907
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7706955075263977,
+      "learning_rate": 1.5578511289500174e-05,
+      "loss": 2.8067,
+      "step": 54908
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7359408736228943,
+      "learning_rate": 1.5573614799594902e-05,
+      "loss": 2.7985,
+      "step": 54909
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7297497391700745,
+      "learning_rate": 1.556871905881403e-05,
+      "loss": 2.9224,
+      "step": 54910
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7188913822174072,
+      "learning_rate": 1.556382406717037e-05,
+      "loss": 2.9255,
+      "step": 54911
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7384754419326782,
+      "learning_rate": 1.5558929824676892e-05,
+      "loss": 2.882,
+      "step": 54912
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7281101942062378,
+      "learning_rate": 1.5554036331346444e-05,
+      "loss": 3.1627,
+      "step": 54913
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7926753163337708,
+      "learning_rate": 1.5549143587191914e-05,
+      "loss": 3.0256,
+      "step": 54914
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7407711148262024,
+      "learning_rate": 1.5544251592226198e-05,
+      "loss": 3.0445,
+      "step": 54915
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7497155666351318,
+      "learning_rate": 1.553936034646218e-05,
+      "loss": 3.0041,
+      "step": 54916
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7246167063713074,
+      "learning_rate": 1.5534469849912755e-05,
+      "loss": 2.8765,
+      "step": 54917
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7616339325904846,
+      "learning_rate": 1.5529580102590743e-05,
+      "loss": 2.975,
+      "step": 54918
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7942472696304321,
+      "learning_rate": 1.552469110450907e-05,
+      "loss": 2.9639,
+      "step": 54919
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7623060345649719,
+      "learning_rate": 1.5519802855680685e-05,
+      "loss": 2.798,
+      "step": 54920
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7489006519317627,
+      "learning_rate": 1.551491535611835e-05,
+      "loss": 2.7875,
+      "step": 54921
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7411266565322876,
+      "learning_rate": 1.5510028605834988e-05,
+      "loss": 2.9106,
+      "step": 54922
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7480968832969666,
+      "learning_rate": 1.5505142604843422e-05,
+      "loss": 2.9138,
+      "step": 54923
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7965890765190125,
+      "learning_rate": 1.5500257353156574e-05,
+      "loss": 2.8865,
+      "step": 54924
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7516406774520874,
+      "learning_rate": 1.5495372850787337e-05,
+      "loss": 2.8225,
+      "step": 54925
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8011462092399597,
+      "learning_rate": 1.549048909774846e-05,
+      "loss": 2.7632,
+      "step": 54926
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7487853169441223,
+      "learning_rate": 1.548560609405294e-05,
+      "loss": 2.9072,
+      "step": 54927
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7621021866798401,
+      "learning_rate": 1.5480723839713592e-05,
+      "loss": 2.98,
+      "step": 54928
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7990534901618958,
+      "learning_rate": 1.5475842334743216e-05,
+      "loss": 2.9085,
+      "step": 54929
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7323390245437622,
+      "learning_rate": 1.5470961579154727e-05,
+      "loss": 2.9067,
+      "step": 54930
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7541958093643188,
+      "learning_rate": 1.5466081572960952e-05,
+      "loss": 2.6722,
+      "step": 54931
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7375277876853943,
+      "learning_rate": 1.5461202316174783e-05,
+      "loss": 3.0888,
+      "step": 54932
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7689357399940491,
+      "learning_rate": 1.5456323808809033e-05,
+      "loss": 2.816,
+      "step": 54933
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7837892174720764,
+      "learning_rate": 1.545144605087657e-05,
+      "loss": 2.7925,
+      "step": 54934
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7377240061759949,
+      "learning_rate": 1.5446569042390246e-05,
+      "loss": 2.8422,
+      "step": 54935
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.716027021408081,
+      "learning_rate": 1.544169278336288e-05,
+      "loss": 2.9464,
+      "step": 54936
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7693297266960144,
+      "learning_rate": 1.5436817273807335e-05,
+      "loss": 2.8894,
+      "step": 54937
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.766839861869812,
+      "learning_rate": 1.5431942513736428e-05,
+      "loss": 2.981,
+      "step": 54938
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7598301768302917,
+      "learning_rate": 1.5427068503163052e-05,
+      "loss": 3.0225,
+      "step": 54939
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7385877966880798,
+      "learning_rate": 1.5422195242099966e-05,
+      "loss": 2.9501,
+      "step": 54940
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7691483497619629,
+      "learning_rate": 1.5417322730560055e-05,
+      "loss": 2.7536,
+      "step": 54941
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7527814507484436,
+      "learning_rate": 1.541245096855621e-05,
+      "loss": 2.7586,
+      "step": 54942
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7725242376327515,
+      "learning_rate": 1.5407579956101155e-05,
+      "loss": 2.7225,
+      "step": 54943
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7793623208999634,
+      "learning_rate": 1.5402709693207816e-05,
+      "loss": 2.852,
+      "step": 54944
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7739753127098083,
+      "learning_rate": 1.5397840179888876e-05,
+      "loss": 2.8733,
+      "step": 54945
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7654193043708801,
+      "learning_rate": 1.5392971416157296e-05,
+      "loss": 2.7769,
+      "step": 54946
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7343190908432007,
+      "learning_rate": 1.5388103402025863e-05,
+      "loss": 3.0919,
+      "step": 54947
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7587972283363342,
+      "learning_rate": 1.538323613750737e-05,
+      "loss": 3.0448,
+      "step": 54948
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7687961459159851,
+      "learning_rate": 1.537836962261467e-05,
+      "loss": 2.8649,
+      "step": 54949
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7500821948051453,
+      "learning_rate": 1.5373503857360592e-05,
+      "loss": 2.7358,
+      "step": 54950
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7614321708679199,
+      "learning_rate": 1.5368638841757885e-05,
+      "loss": 3.0014,
+      "step": 54951
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7479397058486938,
+      "learning_rate": 1.5363774575819444e-05,
+      "loss": 2.8803,
+      "step": 54952
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7300918102264404,
+      "learning_rate": 1.5358911059557986e-05,
+      "loss": 2.969,
+      "step": 54953
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7309346199035645,
+      "learning_rate": 1.5354048292986444e-05,
+      "loss": 2.9086,
+      "step": 54954
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7472579479217529,
+      "learning_rate": 1.5349186276117498e-05,
+      "loss": 2.7324,
+      "step": 54955
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7331668734550476,
+      "learning_rate": 1.5344325008964075e-05,
+      "loss": 2.9536,
+      "step": 54956
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7437717318534851,
+      "learning_rate": 1.53394644915389e-05,
+      "loss": 2.7823,
+      "step": 54957
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7608617544174194,
+      "learning_rate": 1.5334604723854727e-05,
+      "loss": 2.9898,
+      "step": 54958
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8812674880027771,
+      "learning_rate": 1.5329745705924512e-05,
+      "loss": 2.9538,
+      "step": 54959
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7427992224693298,
+      "learning_rate": 1.532488743776088e-05,
+      "loss": 2.6872,
+      "step": 54960
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7801377773284912,
+      "learning_rate": 1.532002991937672e-05,
+      "loss": 2.7277,
+      "step": 54961
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7416569590568542,
+      "learning_rate": 1.5315173150784853e-05,
+      "loss": 2.9374,
+      "step": 54962
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7589424252510071,
+      "learning_rate": 1.5310317131998007e-05,
+      "loss": 2.9005,
+      "step": 54963
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7642767429351807,
+      "learning_rate": 1.5305461863029e-05,
+      "loss": 2.8343,
+      "step": 54964
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7222141027450562,
+      "learning_rate": 1.530060734389066e-05,
+      "loss": 2.8669,
+      "step": 54965
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8169352412223816,
+      "learning_rate": 1.529575357459567e-05,
+      "loss": 2.7694,
+      "step": 54966
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7054728269577026,
+      "learning_rate": 1.5290900555156893e-05,
+      "loss": 2.8031,
+      "step": 54967
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7217891812324524,
+      "learning_rate": 1.5286048285587082e-05,
+      "loss": 2.7711,
+      "step": 54968
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7295551300048828,
+      "learning_rate": 1.528119676589903e-05,
+      "loss": 2.9962,
+      "step": 54969
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7057387232780457,
+      "learning_rate": 1.527634599610549e-05,
+      "loss": 3.0983,
+      "step": 54970
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.779369056224823,
+      "learning_rate": 1.5271495976219316e-05,
+      "loss": 2.9026,
+      "step": 54971
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7789977788925171,
+      "learning_rate": 1.5266646706253205e-05,
+      "loss": 2.9408,
+      "step": 54972
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7889942526817322,
+      "learning_rate": 1.526179818621991e-05,
+      "loss": 2.7538,
+      "step": 54973
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7462673187255859,
+      "learning_rate": 1.5256950416132286e-05,
+      "loss": 2.8916,
+      "step": 54974
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7497276067733765,
+      "learning_rate": 1.5252103396003023e-05,
+      "loss": 2.956,
+      "step": 54975
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7257047891616821,
+      "learning_rate": 1.524725712584488e-05,
+      "loss": 2.831,
+      "step": 54976
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7067046761512756,
+      "learning_rate": 1.5242411605670746e-05,
+      "loss": 2.7964,
+      "step": 54977
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7375363111495972,
+      "learning_rate": 1.5237566835493275e-05,
+      "loss": 2.8754,
+      "step": 54978
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7760704159736633,
+      "learning_rate": 1.523272281532526e-05,
+      "loss": 2.8453,
+      "step": 54979
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.772099494934082,
+      "learning_rate": 1.5227879545179388e-05,
+      "loss": 2.8503,
+      "step": 54980
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7476798295974731,
+      "learning_rate": 1.5223037025068485e-05,
+      "loss": 2.9263,
+      "step": 54981
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7684528827667236,
+      "learning_rate": 1.521819525500534e-05,
+      "loss": 2.8822,
+      "step": 54982
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7638944387435913,
+      "learning_rate": 1.5213354235002606e-05,
+      "loss": 2.865,
+      "step": 54983
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7224180102348328,
+      "learning_rate": 1.5208513965073144e-05,
+      "loss": 2.8717,
+      "step": 54984
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8036991953849792,
+      "learning_rate": 1.5203674445229608e-05,
+      "loss": 2.8207,
+      "step": 54985
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.701972246170044,
+      "learning_rate": 1.5198835675484821e-05,
+      "loss": 2.6852,
+      "step": 54986
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7487140893936157,
+      "learning_rate": 1.5193997655851508e-05,
+      "loss": 2.9361,
+      "step": 54987
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7961825132369995,
+      "learning_rate": 1.5189160386342324e-05,
+      "loss": 2.8853,
+      "step": 54988
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7591695785522461,
+      "learning_rate": 1.5184323866970127e-05,
+      "loss": 2.8568,
+      "step": 54989
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8069265484809875,
+      "learning_rate": 1.5179488097747571e-05,
+      "loss": 2.7135,
+      "step": 54990
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.729277491569519,
+      "learning_rate": 1.5174653078687449e-05,
+      "loss": 2.9049,
+      "step": 54991
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7785135507583618,
+      "learning_rate": 1.516981880980248e-05,
+      "loss": 2.9743,
+      "step": 54992
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7566962242126465,
+      "learning_rate": 1.5164985291105392e-05,
+      "loss": 2.7836,
+      "step": 54993
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7411288022994995,
+      "learning_rate": 1.5160152522608937e-05,
+      "loss": 2.9298,
+      "step": 54994
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7546302676200867,
+      "learning_rate": 1.5155320504325773e-05,
+      "loss": 3.0254,
+      "step": 54995
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7514198422431946,
+      "learning_rate": 1.5150489236268692e-05,
+      "loss": 2.8508,
+      "step": 54996
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7835859060287476,
+      "learning_rate": 1.5145658718450415e-05,
+      "loss": 2.929,
+      "step": 54997
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7201144695281982,
+      "learning_rate": 1.5140828950883633e-05,
+      "loss": 2.9676,
+      "step": 54998
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7485184073448181,
+      "learning_rate": 1.51359999335811e-05,
+      "loss": 3.0914,
+      "step": 54999
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7460274696350098,
+      "learning_rate": 1.5131171666555507e-05,
+      "loss": 2.8109,
+      "step": 55000
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7447788715362549,
+      "learning_rate": 1.5126344149819613e-05,
+      "loss": 2.808,
+      "step": 55001
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7432894706726074,
+      "learning_rate": 1.5121517383386073e-05,
+      "loss": 2.8884,
+      "step": 55002
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7191660404205322,
+      "learning_rate": 1.5116691367267642e-05,
+      "loss": 2.7407,
+      "step": 55003
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7479361295700073,
+      "learning_rate": 1.511186610147701e-05,
+      "loss": 2.7467,
+      "step": 55004
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7603451609611511,
+      "learning_rate": 1.5107041586026902e-05,
+      "loss": 3.1482,
+      "step": 55005
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7904460430145264,
+      "learning_rate": 1.5102217820929974e-05,
+      "loss": 2.9543,
+      "step": 55006
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7052236795425415,
+      "learning_rate": 1.5097394806199048e-05,
+      "loss": 3.0005,
+      "step": 55007
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.824068546295166,
+      "learning_rate": 1.5092572541846748e-05,
+      "loss": 2.8321,
+      "step": 55008
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7605258822441101,
+      "learning_rate": 1.5087751027885764e-05,
+      "loss": 2.7439,
+      "step": 55009
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7721036076545715,
+      "learning_rate": 1.5082930264328786e-05,
+      "loss": 2.9184,
+      "step": 55010
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7376681566238403,
+      "learning_rate": 1.5078110251188535e-05,
+      "loss": 2.7824,
+      "step": 55011
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7896484732627869,
+      "learning_rate": 1.5073290988477737e-05,
+      "loss": 2.9285,
+      "step": 55012
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7744725346565247,
+      "learning_rate": 1.5068472476209048e-05,
+      "loss": 2.919,
+      "step": 55013
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7636879086494446,
+      "learning_rate": 1.5063654714395157e-05,
+      "loss": 2.7188,
+      "step": 55014
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7519050240516663,
+      "learning_rate": 1.5058837703048788e-05,
+      "loss": 2.9375,
+      "step": 55015
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7617154121398926,
+      "learning_rate": 1.505402144218263e-05,
+      "loss": 2.8453,
+      "step": 55016
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7616671919822693,
+      "learning_rate": 1.5049205931809238e-05,
+      "loss": 2.9701,
+      "step": 55017
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7466253042221069,
+      "learning_rate": 1.5044391171941439e-05,
+      "loss": 2.8328,
+      "step": 55018
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7612420320510864,
+      "learning_rate": 1.5039577162591921e-05,
+      "loss": 2.83,
+      "step": 55019
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7458781003952026,
+      "learning_rate": 1.5034763903773239e-05,
+      "loss": 2.8543,
+      "step": 55020
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7177784442901611,
+      "learning_rate": 1.5029951395498219e-05,
+      "loss": 2.9089,
+      "step": 55021
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7244158983230591,
+      "learning_rate": 1.5025139637779415e-05,
+      "loss": 2.7326,
+      "step": 55022
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7530938386917114,
+      "learning_rate": 1.5020328630629586e-05,
+      "loss": 2.9726,
+      "step": 55023
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7853320240974426,
+      "learning_rate": 1.5015518374061352e-05,
+      "loss": 2.8278,
+      "step": 55024
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7918479442596436,
+      "learning_rate": 1.5010708868087373e-05,
+      "loss": 2.889,
+      "step": 55025
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.748648464679718,
+      "learning_rate": 1.5005900112720371e-05,
+      "loss": 3.0377,
+      "step": 55026
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.755605161190033,
+      "learning_rate": 1.5001092107972968e-05,
+      "loss": 2.9121,
+      "step": 55027
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7698187828063965,
+      "learning_rate": 1.4996284853857787e-05,
+      "loss": 2.7703,
+      "step": 55028
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.683066189289093,
+      "learning_rate": 1.499147835038762e-05,
+      "loss": 2.8249,
+      "step": 55029
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7477896809577942,
+      "learning_rate": 1.4986672597575056e-05,
+      "loss": 2.9249,
+      "step": 55030
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7523871064186096,
+      "learning_rate": 1.4981867595432718e-05,
+      "loss": 2.9331,
+      "step": 55031
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7780340909957886,
+      "learning_rate": 1.4977063343973261e-05,
+      "loss": 2.859,
+      "step": 55032
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7800566554069519,
+      "learning_rate": 1.4972259843209344e-05,
+      "loss": 2.6747,
+      "step": 55033
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7878885269165039,
+      "learning_rate": 1.496745709315369e-05,
+      "loss": 3.0291,
+      "step": 55034
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7875646352767944,
+      "learning_rate": 1.4962655093818887e-05,
+      "loss": 2.8538,
+      "step": 55035
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7212032079696655,
+      "learning_rate": 1.4957853845217594e-05,
+      "loss": 2.7787,
+      "step": 55036
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7758206725120544,
+      "learning_rate": 1.4953053347362498e-05,
+      "loss": 2.8677,
+      "step": 55037
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7322551012039185,
+      "learning_rate": 1.4948253600266124e-05,
+      "loss": 3.0312,
+      "step": 55038
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7753440141677856,
+      "learning_rate": 1.494345460394123e-05,
+      "loss": 2.9871,
+      "step": 55039
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7763801217079163,
+      "learning_rate": 1.4938656358400402e-05,
+      "loss": 2.8968,
+      "step": 55040
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7767189741134644,
+      "learning_rate": 1.4933858863656334e-05,
+      "loss": 2.7693,
+      "step": 55041
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.749442458152771,
+      "learning_rate": 1.4929062119721546e-05,
+      "loss": 2.8442,
+      "step": 55042
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7398672103881836,
+      "learning_rate": 1.4924266126608763e-05,
+      "loss": 2.7577,
+      "step": 55043
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7189925909042358,
+      "learning_rate": 1.4919470884330642e-05,
+      "loss": 3.0464,
+      "step": 55044
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.758366584777832,
+      "learning_rate": 1.4914676392899772e-05,
+      "loss": 3.0093,
+      "step": 55045
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7701453566551208,
+      "learning_rate": 1.4909882652328775e-05,
+      "loss": 3.0331,
+      "step": 55046
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7704036831855774,
+      "learning_rate": 1.4905089662630243e-05,
+      "loss": 2.8136,
+      "step": 55047
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.781955897808075,
+      "learning_rate": 1.4900297423816832e-05,
+      "loss": 3.0821,
+      "step": 55048
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7298367619514465,
+      "learning_rate": 1.489550593590123e-05,
+      "loss": 3.0272,
+      "step": 55049
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7215245366096497,
+      "learning_rate": 1.489071519889593e-05,
+      "loss": 2.8109,
+      "step": 55050
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7225742340087891,
+      "learning_rate": 1.4885925212813688e-05,
+      "loss": 2.7193,
+      "step": 55051
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7542920708656311,
+      "learning_rate": 1.4881135977667025e-05,
+      "loss": 2.8853,
+      "step": 55052
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7337054014205933,
+      "learning_rate": 1.4876347493468532e-05,
+      "loss": 2.9647,
+      "step": 55053
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7353600859642029,
+      "learning_rate": 1.48715597602309e-05,
+      "loss": 2.9208,
+      "step": 55054
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7551136016845703,
+      "learning_rate": 1.4866772777966685e-05,
+      "loss": 2.9143,
+      "step": 55055
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7284271717071533,
+      "learning_rate": 1.4861986546688543e-05,
+      "loss": 2.7553,
+      "step": 55056
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.769647479057312,
+      "learning_rate": 1.485720106640903e-05,
+      "loss": 2.7599,
+      "step": 55057
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.793817937374115,
+      "learning_rate": 1.485241633714077e-05,
+      "loss": 2.8884,
+      "step": 55058
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7645768523216248,
+      "learning_rate": 1.4847632358896422e-05,
+      "loss": 3.0118,
+      "step": 55059
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7738844156265259,
+      "learning_rate": 1.4842849131688472e-05,
+      "loss": 3.0922,
+      "step": 55060
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7235080003738403,
+      "learning_rate": 1.4838066655529613e-05,
+      "loss": 2.8157,
+      "step": 55061
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7338597774505615,
+      "learning_rate": 1.4833284930432366e-05,
+      "loss": 3.099,
+      "step": 55062
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7195146679878235,
+      "learning_rate": 1.4828503956409355e-05,
+      "loss": 2.7882,
+      "step": 55063
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7406694889068604,
+      "learning_rate": 1.4823723733473237e-05,
+      "loss": 2.6486,
+      "step": 55064
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7771344184875488,
+      "learning_rate": 1.48189442616365e-05,
+      "loss": 3.0643,
+      "step": 55065
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7858314514160156,
+      "learning_rate": 1.4814165540911805e-05,
+      "loss": 2.9493,
+      "step": 55066
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8330277800559998,
+      "learning_rate": 1.4809387571311738e-05,
+      "loss": 2.9517,
+      "step": 55067
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7209185361862183,
+      "learning_rate": 1.4804610352848823e-05,
+      "loss": 2.8796,
+      "step": 55068
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7240145802497864,
+      "learning_rate": 1.4799833885535684e-05,
+      "loss": 2.7944,
+      "step": 55069
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7134544253349304,
+      "learning_rate": 1.4795058169384877e-05,
+      "loss": 2.7025,
+      "step": 55070
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7314766049385071,
+      "learning_rate": 1.4790283204409059e-05,
+      "loss": 2.7722,
+      "step": 55071
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7572324275970459,
+      "learning_rate": 1.4785508990620654e-05,
+      "loss": 3.037,
+      "step": 55072
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7572822570800781,
+      "learning_rate": 1.4780735528032418e-05,
+      "loss": 2.9088,
+      "step": 55073
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7391690611839294,
+      "learning_rate": 1.4775962816656806e-05,
+      "loss": 2.8931,
+      "step": 55074
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7284706830978394,
+      "learning_rate": 1.4771190856506377e-05,
+      "loss": 3.114,
+      "step": 55075
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7849894165992737,
+      "learning_rate": 1.4766419647593785e-05,
+      "loss": 2.9319,
+      "step": 55076
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7333627939224243,
+      "learning_rate": 1.476164918993149e-05,
+      "loss": 3.0572,
+      "step": 55077
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7046851515769958,
+      "learning_rate": 1.4756879483532146e-05,
+      "loss": 2.9404,
+      "step": 55078
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7099255919456482,
+      "learning_rate": 1.4752110528408312e-05,
+      "loss": 2.7765,
+      "step": 55079
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7698432803153992,
+      "learning_rate": 1.4747342324572509e-05,
+      "loss": 2.8692,
+      "step": 55080
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7886145114898682,
+      "learning_rate": 1.4742574872037327e-05,
+      "loss": 2.9755,
+      "step": 55081
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7380675077438354,
+      "learning_rate": 1.4737808170815258e-05,
+      "loss": 2.6219,
+      "step": 55082
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.80084228515625,
+      "learning_rate": 1.4733042220918923e-05,
+      "loss": 2.9648,
+      "step": 55083
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7491824626922607,
+      "learning_rate": 1.4728277022360846e-05,
+      "loss": 2.7953,
+      "step": 55084
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7680736184120178,
+      "learning_rate": 1.4723512575153584e-05,
+      "loss": 2.9031,
+      "step": 55085
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.76555997133255,
+      "learning_rate": 1.4718748879309727e-05,
+      "loss": 2.863,
+      "step": 55086
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7460500597953796,
+      "learning_rate": 1.4713985934841731e-05,
+      "loss": 2.8605,
+      "step": 55087
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7447842359542847,
+      "learning_rate": 1.4709223741762221e-05,
+      "loss": 2.8689,
+      "step": 55088
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7283835411071777,
+      "learning_rate": 1.470446230008372e-05,
+      "loss": 2.7695,
+      "step": 55089
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.775287926197052,
+      "learning_rate": 1.4699701609818714e-05,
+      "loss": 2.8405,
+      "step": 55090
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7444217801094055,
+      "learning_rate": 1.4694941670979832e-05,
+      "loss": 2.9038,
+      "step": 55091
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7645972371101379,
+      "learning_rate": 1.4690182483579527e-05,
+      "loss": 2.7941,
+      "step": 55092
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7570917010307312,
+      "learning_rate": 1.468542404763039e-05,
+      "loss": 2.9379,
+      "step": 55093
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7789246439933777,
+      "learning_rate": 1.4680666363144944e-05,
+      "loss": 3.1158,
+      "step": 55094
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7554366588592529,
+      "learning_rate": 1.4675909430135713e-05,
+      "loss": 2.782,
+      "step": 55095
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7414656281471252,
+      "learning_rate": 1.4671153248615253e-05,
+      "loss": 2.6158,
+      "step": 55096
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7360018491744995,
+      "learning_rate": 1.4666397818596021e-05,
+      "loss": 2.9317,
+      "step": 55097
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.740975022315979,
+      "learning_rate": 1.4661643140090606e-05,
+      "loss": 2.8232,
+      "step": 55098
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7482433319091797,
+      "learning_rate": 1.4656889213111467e-05,
+      "loss": 2.8816,
+      "step": 55099
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7271791696548462,
+      "learning_rate": 1.4652136037671158e-05,
+      "loss": 2.9859,
+      "step": 55100
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7431482076644897,
+      "learning_rate": 1.464738361378227e-05,
+      "loss": 2.9572,
+      "step": 55101
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7354775071144104,
+      "learning_rate": 1.4642631941457194e-05,
+      "loss": 3.016,
+      "step": 55102
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7678640484809875,
+      "learning_rate": 1.463788102070862e-05,
+      "loss": 2.709,
+      "step": 55103
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7526733875274658,
+      "learning_rate": 1.4633130851548835e-05,
+      "loss": 2.9976,
+      "step": 55104
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.740717887878418,
+      "learning_rate": 1.4628381433990466e-05,
+      "loss": 2.8201,
+      "step": 55105
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7562867999076843,
+      "learning_rate": 1.4623632768046101e-05,
+      "loss": 2.9115,
+      "step": 55106
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7273349761962891,
+      "learning_rate": 1.4618884853728063e-05,
+      "loss": 2.9555,
+      "step": 55107
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7348599433898926,
+      "learning_rate": 1.4614137691049044e-05,
+      "loss": 2.7613,
+      "step": 55108
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7380170226097107,
+      "learning_rate": 1.46093912800214e-05,
+      "loss": 2.8046,
+      "step": 55109
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7288892269134521,
+      "learning_rate": 1.4604645620657751e-05,
+      "loss": 2.7613,
+      "step": 55110
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7518560290336609,
+      "learning_rate": 1.4599900712970524e-05,
+      "loss": 2.834,
+      "step": 55111
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.792517900466919,
+      "learning_rate": 1.459515655697221e-05,
+      "loss": 3.1014,
+      "step": 55112
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7188156247138977,
+      "learning_rate": 1.4590413152675362e-05,
+      "loss": 3.0312,
+      "step": 55113
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7061635255813599,
+      "learning_rate": 1.4585670500092405e-05,
+      "loss": 2.9461,
+      "step": 55114
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7597672939300537,
+      "learning_rate": 1.458092859923583e-05,
+      "loss": 2.8916,
+      "step": 55115
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7571054697036743,
+      "learning_rate": 1.4576187450118227e-05,
+      "loss": 3.2875,
+      "step": 55116
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7508683204650879,
+      "learning_rate": 1.4571447052752017e-05,
+      "loss": 3.0029,
+      "step": 55117
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7968552708625793,
+      "learning_rate": 1.4566707407149691e-05,
+      "loss": 2.9155,
+      "step": 55118
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7436490654945374,
+      "learning_rate": 1.4561968513323641e-05,
+      "loss": 2.799,
+      "step": 55119
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.6988527178764343,
+      "learning_rate": 1.4557230371286488e-05,
+      "loss": 2.8147,
+      "step": 55120
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7517215609550476,
+      "learning_rate": 1.4552492981050656e-05,
+      "loss": 2.9274,
+      "step": 55121
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7683263421058655,
+      "learning_rate": 1.4547756342628603e-05,
+      "loss": 2.8543,
+      "step": 55122
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7408250570297241,
+      "learning_rate": 1.454302045603285e-05,
+      "loss": 2.9393,
+      "step": 55123
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.737571656703949,
+      "learning_rate": 1.4538285321275823e-05,
+      "loss": 2.8324,
+      "step": 55124
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7951434850692749,
+      "learning_rate": 1.4533550938370008e-05,
+      "loss": 2.7856,
+      "step": 55125
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7539615631103516,
+      "learning_rate": 1.4528817307327933e-05,
+      "loss": 2.8728,
+      "step": 55126
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7781716585159302,
+      "learning_rate": 1.4524084428161953e-05,
+      "loss": 2.78,
+      "step": 55127
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7636751532554626,
+      "learning_rate": 1.4519352300884623e-05,
+      "loss": 2.9116,
+      "step": 55128
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7897639274597168,
+      "learning_rate": 1.451462092550837e-05,
+      "loss": 2.8244,
+      "step": 55129
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7303542494773865,
+      "learning_rate": 1.4509890302045612e-05,
+      "loss": 2.9801,
+      "step": 55130
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7917243838310242,
+      "learning_rate": 1.4505160430508944e-05,
+      "loss": 2.9172,
+      "step": 55131
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7893748879432678,
+      "learning_rate": 1.450043131091072e-05,
+      "loss": 3.0641,
+      "step": 55132
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7223455309867859,
+      "learning_rate": 1.4495702943263432e-05,
+      "loss": 2.8755,
+      "step": 55133
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7736308574676514,
+      "learning_rate": 1.4490975327579436e-05,
+      "loss": 3.098,
+      "step": 55134
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7550448179244995,
+      "learning_rate": 1.4486248463871287e-05,
+      "loss": 2.8793,
+      "step": 55135
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7876956462860107,
+      "learning_rate": 1.4481522352151475e-05,
+      "loss": 2.8375,
+      "step": 55136
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7860339283943176,
+      "learning_rate": 1.4476796992432327e-05,
+      "loss": 3.0902,
+      "step": 55137
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7922260761260986,
+      "learning_rate": 1.4472072384726363e-05,
+      "loss": 2.7761,
+      "step": 55138
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7500098347663879,
+      "learning_rate": 1.4467348529046041e-05,
+      "loss": 2.9636,
+      "step": 55139
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7248625159263611,
+      "learning_rate": 1.4462625425403718e-05,
+      "loss": 2.9241,
+      "step": 55140
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7339884638786316,
+      "learning_rate": 1.4457903073811916e-05,
+      "loss": 2.9563,
+      "step": 55141
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7312856316566467,
+      "learning_rate": 1.4453181474283026e-05,
+      "loss": 2.8138,
+      "step": 55142
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7703109979629517,
+      "learning_rate": 1.4448460626829505e-05,
+      "loss": 2.9011,
+      "step": 55143
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7444011569023132,
+      "learning_rate": 1.4443740531463777e-05,
+      "loss": 2.8922,
+      "step": 55144
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7480001449584961,
+      "learning_rate": 1.4439021188198264e-05,
+      "loss": 2.7758,
+      "step": 55145
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7840633988380432,
+      "learning_rate": 1.4434302597045454e-05,
+      "loss": 3.0501,
+      "step": 55146
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8444885015487671,
+      "learning_rate": 1.4429584758017743e-05,
+      "loss": 2.9258,
+      "step": 55147
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7418254017829895,
+      "learning_rate": 1.442486767112755e-05,
+      "loss": 2.9051,
+      "step": 55148
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.75757896900177,
+      "learning_rate": 1.4420151336387231e-05,
+      "loss": 2.8969,
+      "step": 55149
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7526350021362305,
+      "learning_rate": 1.4415435753809313e-05,
+      "loss": 2.8936,
+      "step": 55150
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7957975268363953,
+      "learning_rate": 1.4410720923406183e-05,
+      "loss": 2.763,
+      "step": 55151
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7414007186889648,
+      "learning_rate": 1.4406006845190233e-05,
+      "loss": 2.7035,
+      "step": 55152
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7431672215461731,
+      "learning_rate": 1.4401293519173918e-05,
+      "loss": 2.8199,
+      "step": 55153
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7612506151199341,
+      "learning_rate": 1.439658094536963e-05,
+      "loss": 2.7203,
+      "step": 55154
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7056882977485657,
+      "learning_rate": 1.4391869123789756e-05,
+      "loss": 2.8934,
+      "step": 55155
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7619946002960205,
+      "learning_rate": 1.4387158054446756e-05,
+      "loss": 2.7752,
+      "step": 55156
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7698616981506348,
+      "learning_rate": 1.4382447737352987e-05,
+      "loss": 2.8134,
+      "step": 55157
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8463789224624634,
+      "learning_rate": 1.4377738172520903e-05,
+      "loss": 2.8462,
+      "step": 55158
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7743161916732788,
+      "learning_rate": 1.4373029359962862e-05,
+      "loss": 2.8833,
+      "step": 55159
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7336245775222778,
+      "learning_rate": 1.4368321299691355e-05,
+      "loss": 2.9437,
+      "step": 55160
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7166826128959656,
+      "learning_rate": 1.436361399171867e-05,
+      "loss": 3.0139,
+      "step": 55161
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7472900748252869,
+      "learning_rate": 1.4358907436057266e-05,
+      "loss": 2.816,
+      "step": 55162
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7503640055656433,
+      "learning_rate": 1.435420163271953e-05,
+      "loss": 2.651,
+      "step": 55163
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7015929222106934,
+      "learning_rate": 1.4349496581717823e-05,
+      "loss": 2.8712,
+      "step": 55164
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7873293161392212,
+      "learning_rate": 1.4344792283064565e-05,
+      "loss": 2.7243,
+      "step": 55165
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7600153684616089,
+      "learning_rate": 1.4340088736772214e-05,
+      "loss": 3.0151,
+      "step": 55166
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8037029504776001,
+      "learning_rate": 1.4335385942853028e-05,
+      "loss": 2.9545,
+      "step": 55167
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7690410017967224,
+      "learning_rate": 1.4330683901319496e-05,
+      "loss": 2.8564,
+      "step": 55168
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7575297355651855,
+      "learning_rate": 1.4325982612183973e-05,
+      "loss": 2.8627,
+      "step": 55169
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7149128913879395,
+      "learning_rate": 1.4321282075458784e-05,
+      "loss": 2.8766,
+      "step": 55170
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7330735325813293,
+      "learning_rate": 1.431658229115642e-05,
+      "loss": 2.8966,
+      "step": 55171
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.725845456123352,
+      "learning_rate": 1.4311883259289136e-05,
+      "loss": 2.7962,
+      "step": 55172
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.712296724319458,
+      "learning_rate": 1.4307184979869391e-05,
+      "loss": 2.8348,
+      "step": 55173
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8258801102638245,
+      "learning_rate": 1.4302487452909539e-05,
+      "loss": 2.9634,
+      "step": 55174
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7133929133415222,
+      "learning_rate": 1.4297790678421972e-05,
+      "loss": 2.7334,
+      "step": 55175
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7767179012298584,
+      "learning_rate": 1.4293094656419046e-05,
+      "loss": 2.8227,
+      "step": 55176
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7630220651626587,
+      "learning_rate": 1.4288399386913086e-05,
+      "loss": 2.9548,
+      "step": 55177
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.756858766078949,
+      "learning_rate": 1.4283704869916512e-05,
+      "loss": 3.0724,
+      "step": 55178
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.734491765499115,
+      "learning_rate": 1.4279011105441651e-05,
+      "loss": 2.8499,
+      "step": 55179
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7555728554725647,
+      "learning_rate": 1.4274318093500925e-05,
+      "loss": 3.0742,
+      "step": 55180
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.779068648815155,
+      "learning_rate": 1.4269625834106625e-05,
+      "loss": 3.0113,
+      "step": 55181
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7292513847351074,
+      "learning_rate": 1.4264934327271172e-05,
+      "loss": 2.9021,
+      "step": 55182
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7384148240089417,
+      "learning_rate": 1.4260243573006891e-05,
+      "loss": 2.6843,
+      "step": 55183
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.72471022605896,
+      "learning_rate": 1.4255553571326106e-05,
+      "loss": 3.1414,
+      "step": 55184
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.761789083480835,
+      "learning_rate": 1.425086432224124e-05,
+      "loss": 3.0294,
+      "step": 55185
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7743604779243469,
+      "learning_rate": 1.424617582576455e-05,
+      "loss": 2.8922,
+      "step": 55186
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7332843542098999,
+      "learning_rate": 1.4241488081908426e-05,
+      "loss": 2.8659,
+      "step": 55187
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.739586353302002,
+      "learning_rate": 1.423680109068529e-05,
+      "loss": 2.7717,
+      "step": 55188
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7435079216957092,
+      "learning_rate": 1.4232114852107368e-05,
+      "loss": 2.8082,
+      "step": 55189
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7537873983383179,
+      "learning_rate": 1.4227429366187115e-05,
+      "loss": 2.7587,
+      "step": 55190
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7387575507164001,
+      "learning_rate": 1.4222744632936822e-05,
+      "loss": 2.9726,
+      "step": 55191
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.762256920337677,
+      "learning_rate": 1.421806065236878e-05,
+      "loss": 2.7777,
+      "step": 55192
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7154507637023926,
+      "learning_rate": 1.4213377424495375e-05,
+      "loss": 2.9309,
+      "step": 55193
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7694316506385803,
+      "learning_rate": 1.4208694949328903e-05,
+      "loss": 2.4981,
+      "step": 55194
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.7170578241348267,
+      "learning_rate": 1.4204013226881784e-05,
+      "loss": 3.0198,
+      "step": 55195
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7894667387008667,
+      "learning_rate": 1.419933225716624e-05,
+      "loss": 2.8454,
+      "step": 55196
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7934728860855103,
+      "learning_rate": 1.4194652040194698e-05,
+      "loss": 3.0578,
+      "step": 55197
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7518163919448853,
+      "learning_rate": 1.4189972575979448e-05,
+      "loss": 2.7094,
+      "step": 55198
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7911694645881653,
+      "learning_rate": 1.4185293864532743e-05,
+      "loss": 3.0337,
+      "step": 55199
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7032492756843567,
+      "learning_rate": 1.418061590586701e-05,
+      "loss": 2.9355,
+      "step": 55200
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7293274402618408,
+      "learning_rate": 1.4175938699994504e-05,
+      "loss": 2.9049,
+      "step": 55201
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7434591054916382,
+      "learning_rate": 1.417126224692755e-05,
+      "loss": 2.8035,
+      "step": 55202
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7389995455741882,
+      "learning_rate": 1.4166586546678538e-05,
+      "loss": 2.8651,
+      "step": 55203
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7393417358398438,
+      "learning_rate": 1.416191159925969e-05,
+      "loss": 2.8286,
+      "step": 55204
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7257364988327026,
+      "learning_rate": 1.415723740468343e-05,
+      "loss": 2.9278,
+      "step": 55205
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8316163420677185,
+      "learning_rate": 1.4152563962961916e-05,
+      "loss": 2.9318,
+      "step": 55206
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.718009889125824,
+      "learning_rate": 1.4147891274107503e-05,
+      "loss": 3.0043,
+      "step": 55207
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7439994812011719,
+      "learning_rate": 1.4143219338132616e-05,
+      "loss": 2.8376,
+      "step": 55208
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7357783317565918,
+      "learning_rate": 1.4138548155049411e-05,
+      "loss": 2.8897,
+      "step": 55209
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7854210734367371,
+      "learning_rate": 1.4133877724870312e-05,
+      "loss": 2.7374,
+      "step": 55210
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7327690124511719,
+      "learning_rate": 1.412920804760751e-05,
+      "loss": 2.7055,
+      "step": 55211
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.784988284111023,
+      "learning_rate": 1.4124539123273392e-05,
+      "loss": 2.7917,
+      "step": 55212
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7121521830558777,
+      "learning_rate": 1.4119870951880252e-05,
+      "loss": 2.8236,
+      "step": 55213
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7981328368186951,
+      "learning_rate": 1.4115203533440278e-05,
+      "loss": 3.0959,
+      "step": 55214
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7848892211914062,
+      "learning_rate": 1.4110536867965894e-05,
+      "loss": 2.7955,
+      "step": 55215
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7660860419273376,
+      "learning_rate": 1.410587095546929e-05,
+      "loss": 3.0071,
+      "step": 55216
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7169570326805115,
+      "learning_rate": 1.4101205795962788e-05,
+      "loss": 2.6379,
+      "step": 55217
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7584700584411621,
+      "learning_rate": 1.409654138945875e-05,
+      "loss": 2.9686,
+      "step": 55218
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7803013920783997,
+      "learning_rate": 1.4091877735969393e-05,
+      "loss": 2.8314,
+      "step": 55219
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7412531971931458,
+      "learning_rate": 1.408721483550701e-05,
+      "loss": 2.9973,
+      "step": 55220
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8721296191215515,
+      "learning_rate": 1.4082552688083827e-05,
+      "loss": 2.917,
+      "step": 55221
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7726236581802368,
+      "learning_rate": 1.4077891293712162e-05,
+      "loss": 2.9424,
+      "step": 55222
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.761952817440033,
+      "learning_rate": 1.4073230652404378e-05,
+      "loss": 2.6168,
+      "step": 55223
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7393391132354736,
+      "learning_rate": 1.4068570764172626e-05,
+      "loss": 2.9725,
+      "step": 55224
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7465383410453796,
+      "learning_rate": 1.4063911629029268e-05,
+      "loss": 2.8849,
+      "step": 55225
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7391934394836426,
+      "learning_rate": 1.405925324698649e-05,
+      "loss": 2.9616,
+      "step": 55226
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7626303434371948,
+      "learning_rate": 1.4054595618056652e-05,
+      "loss": 2.9186,
+      "step": 55227
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7824798226356506,
+      "learning_rate": 1.4049938742251976e-05,
+      "loss": 3.2266,
+      "step": 55228
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7267372012138367,
+      "learning_rate": 1.4045282619584685e-05,
+      "loss": 2.7209,
+      "step": 55229
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7535328269004822,
+      "learning_rate": 1.4040627250067138e-05,
+      "loss": 2.9377,
+      "step": 55230
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7468425035476685,
+      "learning_rate": 1.403597263371149e-05,
+      "loss": 2.8851,
+      "step": 55231
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7248378396034241,
+      "learning_rate": 1.4031318770530065e-05,
+      "loss": 2.6145,
+      "step": 55232
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7613323926925659,
+      "learning_rate": 1.4026665660535152e-05,
+      "loss": 2.9701,
+      "step": 55233
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7856761813163757,
+      "learning_rate": 1.4022013303738977e-05,
+      "loss": 2.8771,
+      "step": 55234
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7455305457115173,
+      "learning_rate": 1.4017361700153762e-05,
+      "loss": 2.9148,
+      "step": 55235
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7511041760444641,
+      "learning_rate": 1.4012710849791731e-05,
+      "loss": 2.6375,
+      "step": 55236
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7273489832878113,
+      "learning_rate": 1.4008060752665206e-05,
+      "loss": 2.7796,
+      "step": 55237
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.744505763053894,
+      "learning_rate": 1.4003411408786414e-05,
+      "loss": 3.0129,
+      "step": 55238
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7265143394470215,
+      "learning_rate": 1.3998762818167575e-05,
+      "loss": 2.7688,
+      "step": 55239
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8545629978179932,
+      "learning_rate": 1.3994114980821014e-05,
+      "loss": 2.9477,
+      "step": 55240
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7982408404350281,
+      "learning_rate": 1.3989467896758887e-05,
+      "loss": 2.9287,
+      "step": 55241
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7544611692428589,
+      "learning_rate": 1.398482156599342e-05,
+      "loss": 2.8994,
+      "step": 55242
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8148059248924255,
+      "learning_rate": 1.3980175988536935e-05,
+      "loss": 2.81,
+      "step": 55243
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7339653372764587,
+      "learning_rate": 1.3975531164401555e-05,
+      "loss": 2.9793,
+      "step": 55244
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7554219961166382,
+      "learning_rate": 1.397088709359967e-05,
+      "loss": 3.0484,
+      "step": 55245
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8348968625068665,
+      "learning_rate": 1.3966243776143338e-05,
+      "loss": 3.0106,
+      "step": 55246
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7876124382019043,
+      "learning_rate": 1.3961601212044881e-05,
+      "loss": 2.8469,
+      "step": 55247
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7836396098136902,
+      "learning_rate": 1.3956959401316558e-05,
+      "loss": 2.7991,
+      "step": 55248
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7819403409957886,
+      "learning_rate": 1.3952318343970559e-05,
+      "loss": 3.033,
+      "step": 55249
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7380668520927429,
+      "learning_rate": 1.3947678040019105e-05,
+      "loss": 2.8926,
+      "step": 55250
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7316756248474121,
+      "learning_rate": 1.3943038489474356e-05,
+      "loss": 2.9032,
+      "step": 55251
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7308788299560547,
+      "learning_rate": 1.3938399692348667e-05,
+      "loss": 2.911,
+      "step": 55252
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7654592990875244,
+      "learning_rate": 1.393376164865413e-05,
+      "loss": 2.6837,
+      "step": 55253
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7962247133255005,
+      "learning_rate": 1.3929124358402999e-05,
+      "loss": 2.9718,
+      "step": 55254
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7708144187927246,
+      "learning_rate": 1.3924487821607534e-05,
+      "loss": 2.8453,
+      "step": 55255
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7123581767082214,
+      "learning_rate": 1.3919852038279888e-05,
+      "loss": 2.7315,
+      "step": 55256
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7534207105636597,
+      "learning_rate": 1.3915217008432355e-05,
+      "loss": 3.0169,
+      "step": 55257
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7725705504417419,
+      "learning_rate": 1.3910582732076991e-05,
+      "loss": 2.8719,
+      "step": 55258
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7859570384025574,
+      "learning_rate": 1.3905949209226119e-05,
+      "loss": 2.9936,
+      "step": 55259
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7947921752929688,
+      "learning_rate": 1.3901316439891962e-05,
+      "loss": 2.7604,
+      "step": 55260
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7809910774230957,
+      "learning_rate": 1.3896684424086613e-05,
+      "loss": 2.9163,
+      "step": 55261
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7197073698043823,
+      "learning_rate": 1.3892053161822392e-05,
+      "loss": 2.9279,
+      "step": 55262
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7367950677871704,
+      "learning_rate": 1.3887422653111424e-05,
+      "loss": 2.7356,
+      "step": 55263
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7174615263938904,
+      "learning_rate": 1.3882792897965899e-05,
+      "loss": 3.0581,
+      "step": 55264
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7707169055938721,
+      "learning_rate": 1.3878163896398076e-05,
+      "loss": 2.6856,
+      "step": 55265
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7517051100730896,
+      "learning_rate": 1.3873535648420043e-05,
+      "loss": 2.8619,
+      "step": 55266
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7613734602928162,
+      "learning_rate": 1.3868908154044089e-05,
+      "loss": 2.7424,
+      "step": 55267
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7365747690200806,
+      "learning_rate": 1.386428141328234e-05,
+      "loss": 3.0878,
+      "step": 55268
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7309760451316833,
+      "learning_rate": 1.385965542614702e-05,
+      "loss": 2.7943,
+      "step": 55269
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7792554497718811,
+      "learning_rate": 1.3855030192650318e-05,
+      "loss": 2.9993,
+      "step": 55270
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7756322622299194,
+      "learning_rate": 1.385040571280439e-05,
+      "loss": 2.5423,
+      "step": 55271
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7910360097885132,
+      "learning_rate": 1.3845781986621463e-05,
+      "loss": 3.0719,
+      "step": 55272
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7808008790016174,
+      "learning_rate": 1.384115901411359e-05,
+      "loss": 2.9894,
+      "step": 55273
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7794164419174194,
+      "learning_rate": 1.3836536795293063e-05,
+      "loss": 2.6242,
+      "step": 55274
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7962504625320435,
+      "learning_rate": 1.3831915330172038e-05,
+      "loss": 2.9159,
+      "step": 55275
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7393701076507568,
+      "learning_rate": 1.382729461876264e-05,
+      "loss": 2.8513,
+      "step": 55276
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7692285776138306,
+      "learning_rate": 1.3822674661077126e-05,
+      "loss": 2.7925,
+      "step": 55277
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.721187949180603,
+      "learning_rate": 1.3818055457127586e-05,
+      "loss": 3.0527,
+      "step": 55278
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7725158929824829,
+      "learning_rate": 1.3813437006926209e-05,
+      "loss": 2.9564,
+      "step": 55279
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7646429538726807,
+      "learning_rate": 1.3808819310485186e-05,
+      "loss": 2.8219,
+      "step": 55280
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7827332019805908,
+      "learning_rate": 1.3804202367816608e-05,
+      "loss": 2.9961,
+      "step": 55281
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7357837557792664,
+      "learning_rate": 1.3799586178932697e-05,
+      "loss": 2.9484,
+      "step": 55282
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7550036311149597,
+      "learning_rate": 1.3794970743845612e-05,
+      "loss": 2.9146,
+      "step": 55283
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7835947275161743,
+      "learning_rate": 1.3790356062567475e-05,
+      "loss": 2.8771,
+      "step": 55284
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7576488852500916,
+      "learning_rate": 1.378574213511051e-05,
+      "loss": 2.6269,
+      "step": 55285
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7808263301849365,
+      "learning_rate": 1.3781128961486742e-05,
+      "loss": 2.813,
+      "step": 55286
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7787500619888306,
+      "learning_rate": 1.377651654170846e-05,
+      "loss": 2.8559,
+      "step": 55287
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8117019534111023,
+      "learning_rate": 1.3771904875787687e-05,
+      "loss": 2.8457,
+      "step": 55288
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.712166965007782,
+      "learning_rate": 1.3767293963736647e-05,
+      "loss": 2.8873,
+      "step": 55289
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7200855016708374,
+      "learning_rate": 1.3762683805567498e-05,
+      "loss": 2.8848,
+      "step": 55290
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7672809958457947,
+      "learning_rate": 1.3758074401292296e-05,
+      "loss": 2.8106,
+      "step": 55291
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7226176261901855,
+      "learning_rate": 1.3753465750923297e-05,
+      "loss": 2.8904,
+      "step": 55292
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.740872323513031,
+      "learning_rate": 1.3748857854472594e-05,
+      "loss": 2.9025,
+      "step": 55293
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7510790824890137,
+      "learning_rate": 1.3744250711952242e-05,
+      "loss": 2.9526,
+      "step": 55294
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7161646485328674,
+      "learning_rate": 1.37396443233745e-05,
+      "loss": 3.074,
+      "step": 55295
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7294967770576477,
+      "learning_rate": 1.3735038688751387e-05,
+      "loss": 2.8639,
+      "step": 55296
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7264935970306396,
+      "learning_rate": 1.3730433808095165e-05,
+      "loss": 2.8598,
+      "step": 55297
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8006035685539246,
+      "learning_rate": 1.3725829681417822e-05,
+      "loss": 2.9085,
+      "step": 55298
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7237944602966309,
+      "learning_rate": 1.372122630873158e-05,
+      "loss": 2.7831,
+      "step": 55299
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7685567140579224,
+      "learning_rate": 1.3716623690048534e-05,
+      "loss": 3.0923,
+      "step": 55300
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8011992573738098,
+      "learning_rate": 1.371202182538077e-05,
+      "loss": 2.9499,
+      "step": 55301
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7467533349990845,
+      "learning_rate": 1.370742071474048e-05,
+      "loss": 3.0305,
+      "step": 55302
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7417289614677429,
+      "learning_rate": 1.3702820358139721e-05,
+      "loss": 2.7992,
+      "step": 55303
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7307183742523193,
+      "learning_rate": 1.3698220755590616e-05,
+      "loss": 2.9265,
+      "step": 55304
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7761890888214111,
+      "learning_rate": 1.3693621907105357e-05,
+      "loss": 2.7917,
+      "step": 55305
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7544965744018555,
+      "learning_rate": 1.3689023812695999e-05,
+      "loss": 2.9144,
+      "step": 55306
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8068240284919739,
+      "learning_rate": 1.3684426472374632e-05,
+      "loss": 2.7519,
+      "step": 55307
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7504112124443054,
+      "learning_rate": 1.3679829886153349e-05,
+      "loss": 2.7891,
+      "step": 55308
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7085170149803162,
+      "learning_rate": 1.367523405404427e-05,
+      "loss": 2.6752,
+      "step": 55309
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7529444694519043,
+      "learning_rate": 1.3670638976059589e-05,
+      "loss": 3.0175,
+      "step": 55310
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7292396426200867,
+      "learning_rate": 1.3666044652211295e-05,
+      "loss": 2.9474,
+      "step": 55311
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.787841260433197,
+      "learning_rate": 1.3661451082511577e-05,
+      "loss": 2.6969,
+      "step": 55312
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7327471375465393,
+      "learning_rate": 1.3656858266972427e-05,
+      "loss": 2.9287,
+      "step": 55313
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7643541693687439,
+      "learning_rate": 1.3652266205606065e-05,
+      "loss": 2.8279,
+      "step": 55314
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.6762557625770569,
+      "learning_rate": 1.364767489842452e-05,
+      "loss": 2.8347,
+      "step": 55315
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.726814866065979,
+      "learning_rate": 1.3643084345439847e-05,
+      "loss": 2.7628,
+      "step": 55316
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.770945131778717,
+      "learning_rate": 1.3638494546664236e-05,
+      "loss": 2.8228,
+      "step": 55317
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7129945158958435,
+      "learning_rate": 1.3633905502109676e-05,
+      "loss": 2.8855,
+      "step": 55318
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7419125437736511,
+      "learning_rate": 1.362931721178826e-05,
+      "loss": 2.8758,
+      "step": 55319
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7454127669334412,
+      "learning_rate": 1.3624729675712176e-05,
+      "loss": 2.9655,
+      "step": 55320
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7969364523887634,
+      "learning_rate": 1.362014289389345e-05,
+      "loss": 3.003,
+      "step": 55321
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7323818206787109,
+      "learning_rate": 1.3615556866344135e-05,
+      "loss": 2.8169,
+      "step": 55322
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7280770540237427,
+      "learning_rate": 1.3610971593076293e-05,
+      "loss": 2.7861,
+      "step": 55323
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.808602511882782,
+      "learning_rate": 1.3606387074102042e-05,
+      "loss": 2.9433,
+      "step": 55324
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.717870831489563,
+      "learning_rate": 1.360180330943348e-05,
+      "loss": 2.9154,
+      "step": 55325
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7248071432113647,
+      "learning_rate": 1.3597220299082623e-05,
+      "loss": 2.9233,
+      "step": 55326
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7266684174537659,
+      "learning_rate": 1.3592638043061598e-05,
+      "loss": 3.1069,
+      "step": 55327
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7501939535140991,
+      "learning_rate": 1.3588056541382397e-05,
+      "loss": 2.9581,
+      "step": 55328
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7233149409294128,
+      "learning_rate": 1.3583475794057207e-05,
+      "loss": 2.8132,
+      "step": 55329
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8018295168876648,
+      "learning_rate": 1.3578895801097989e-05,
+      "loss": 3.065,
+      "step": 55330
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7703627347946167,
+      "learning_rate": 1.3574316562516796e-05,
+      "loss": 2.7958,
+      "step": 55331
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7846996188163757,
+      "learning_rate": 1.3569738078325788e-05,
+      "loss": 2.72,
+      "step": 55332
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8439915180206299,
+      "learning_rate": 1.3565160348536918e-05,
+      "loss": 2.673,
+      "step": 55333
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7853070497512817,
+      "learning_rate": 1.3560583373162315e-05,
+      "loss": 2.8157,
+      "step": 55334
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7770476937294006,
+      "learning_rate": 1.3556007152213966e-05,
+      "loss": 2.9171,
+      "step": 55335
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7779586315155029,
+      "learning_rate": 1.3551431685704028e-05,
+      "loss": 2.9688,
+      "step": 55336
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7522245049476624,
+      "learning_rate": 1.3546856973644493e-05,
+      "loss": 2.7268,
+      "step": 55337
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7922734022140503,
+      "learning_rate": 1.354228301604735e-05,
+      "loss": 2.7946,
+      "step": 55338
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7396979331970215,
+      "learning_rate": 1.3537709812924757e-05,
+      "loss": 2.9228,
+      "step": 55339
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7690479755401611,
+      "learning_rate": 1.3533137364288671e-05,
+      "loss": 2.8392,
+      "step": 55340
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8230964541435242,
+      "learning_rate": 1.3528565670151149e-05,
+      "loss": 2.7559,
+      "step": 55341
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7428261637687683,
+      "learning_rate": 1.3523994730524312e-05,
+      "loss": 2.8789,
+      "step": 55342
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7513474822044373,
+      "learning_rate": 1.351942454542012e-05,
+      "loss": 2.711,
+      "step": 55343
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7638099193572998,
+      "learning_rate": 1.351485511485063e-05,
+      "loss": 2.6492,
+      "step": 55344
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7367132306098938,
+      "learning_rate": 1.3510286438827866e-05,
+      "loss": 2.936,
+      "step": 55345
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7802484035491943,
+      "learning_rate": 1.3505718517363851e-05,
+      "loss": 2.9882,
+      "step": 55346
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7542453408241272,
+      "learning_rate": 1.3501151350470673e-05,
+      "loss": 2.8119,
+      "step": 55347
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.79583340883255,
+      "learning_rate": 1.3496584938160292e-05,
+      "loss": 2.8344,
+      "step": 55348
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7419581413269043,
+      "learning_rate": 1.3492019280444798e-05,
+      "loss": 2.6465,
+      "step": 55349
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7605627775192261,
+      "learning_rate": 1.348745437733615e-05,
+      "loss": 2.7418,
+      "step": 55350
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7467504143714905,
+      "learning_rate": 1.3482890228846466e-05,
+      "loss": 2.8421,
+      "step": 55351
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7766313552856445,
+      "learning_rate": 1.3478326834987673e-05,
+      "loss": 2.8587,
+      "step": 55352
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7753242254257202,
+      "learning_rate": 1.3473764195771797e-05,
+      "loss": 2.7824,
+      "step": 55353
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7845016121864319,
+      "learning_rate": 1.3469202311210925e-05,
+      "loss": 2.9468,
+      "step": 55354
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7222416400909424,
+      "learning_rate": 1.3464641181317015e-05,
+      "loss": 3.0865,
+      "step": 55355
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7512432932853699,
+      "learning_rate": 1.3460080806102057e-05,
+      "loss": 2.6878,
+      "step": 55356
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7436035871505737,
+      "learning_rate": 1.3455521185578177e-05,
+      "loss": 2.7787,
+      "step": 55357
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7184932827949524,
+      "learning_rate": 1.3450962319757263e-05,
+      "loss": 2.8446,
+      "step": 55358
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.748602569103241,
+      "learning_rate": 1.3446404208651406e-05,
+      "loss": 2.948,
+      "step": 55359
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8889462947845459,
+      "learning_rate": 1.344184685227253e-05,
+      "loss": 2.9219,
+      "step": 55360
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7768132090568542,
+      "learning_rate": 1.343729025063266e-05,
+      "loss": 2.8009,
+      "step": 55361
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.727043628692627,
+      "learning_rate": 1.343273440374385e-05,
+      "loss": 2.7721,
+      "step": 55362
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7611899375915527,
+      "learning_rate": 1.3428179311618058e-05,
+      "loss": 2.8881,
+      "step": 55363
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7601465582847595,
+      "learning_rate": 1.3423624974267312e-05,
+      "loss": 2.8482,
+      "step": 55364
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7595193386077881,
+      "learning_rate": 1.3419071391703596e-05,
+      "loss": 2.8653,
+      "step": 55365
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8344078660011292,
+      "learning_rate": 1.3414518563938837e-05,
+      "loss": 2.969,
+      "step": 55366
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7677531242370605,
+      "learning_rate": 1.3409966490985125e-05,
+      "loss": 2.7756,
+      "step": 55367
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7744829654693604,
+      "learning_rate": 1.3405415172854384e-05,
+      "loss": 2.9537,
+      "step": 55368
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7642647624015808,
+      "learning_rate": 1.3400864609558637e-05,
+      "loss": 2.6661,
+      "step": 55369
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7297582626342773,
+      "learning_rate": 1.3396314801109808e-05,
+      "loss": 2.977,
+      "step": 55370
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7670851945877075,
+      "learning_rate": 1.3391765747519956e-05,
+      "loss": 2.9117,
+      "step": 55371
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7472463250160217,
+      "learning_rate": 1.3387217448801069e-05,
+      "loss": 2.8681,
+      "step": 55372
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7477490305900574,
+      "learning_rate": 1.3382669904965105e-05,
+      "loss": 2.8268,
+      "step": 55373
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7499014735221863,
+      "learning_rate": 1.3378123116023987e-05,
+      "loss": 2.9298,
+      "step": 55374
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.748065710067749,
+      "learning_rate": 1.3373577081989707e-05,
+      "loss": 2.9597,
+      "step": 55375
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7136443257331848,
+      "learning_rate": 1.3369031802874286e-05,
+      "loss": 2.8244,
+      "step": 55376
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7487014532089233,
+      "learning_rate": 1.3364487278689684e-05,
+      "loss": 2.7234,
+      "step": 55377
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7109510898590088,
+      "learning_rate": 1.3359943509447858e-05,
+      "loss": 3.0295,
+      "step": 55378
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7548745274543762,
+      "learning_rate": 1.3355400495160762e-05,
+      "loss": 2.9223,
+      "step": 55379
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7589823007583618,
+      "learning_rate": 1.3350858235840423e-05,
+      "loss": 2.8757,
+      "step": 55380
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8094698190689087,
+      "learning_rate": 1.3346316731498697e-05,
+      "loss": 2.9904,
+      "step": 55381
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7765593528747559,
+      "learning_rate": 1.334177598214764e-05,
+      "loss": 2.9588,
+      "step": 55382
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7499661445617676,
+      "learning_rate": 1.3337235987799144e-05,
+      "loss": 2.8294,
+      "step": 55383
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7597896456718445,
+      "learning_rate": 1.3332696748465232e-05,
+      "loss": 2.8076,
+      "step": 55384
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7389174103736877,
+      "learning_rate": 1.3328158264157762e-05,
+      "loss": 3.1259,
+      "step": 55385
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7625584006309509,
+      "learning_rate": 1.3323620534888823e-05,
+      "loss": 2.8586,
+      "step": 55386
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7467266917228699,
+      "learning_rate": 1.3319083560670308e-05,
+      "loss": 3.0767,
+      "step": 55387
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7760413885116577,
+      "learning_rate": 1.3314547341514104e-05,
+      "loss": 3.0268,
+      "step": 55388
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7491160035133362,
+      "learning_rate": 1.3310011877432236e-05,
+      "loss": 2.9444,
+      "step": 55389
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8000679612159729,
+      "learning_rate": 1.3305477168436596e-05,
+      "loss": 2.9765,
+      "step": 55390
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.0552611351013184,
+      "learning_rate": 1.3300943214539138e-05,
+      "loss": 2.9166,
+      "step": 55391
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7722744345664978,
+      "learning_rate": 1.3296410015751858e-05,
+      "loss": 2.8614,
+      "step": 55392
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7588552832603455,
+      "learning_rate": 1.329187757208664e-05,
+      "loss": 2.7327,
+      "step": 55393
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7379563450813293,
+      "learning_rate": 1.3287345883555445e-05,
+      "loss": 2.8572,
+      "step": 55394
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7401456236839294,
+      "learning_rate": 1.328281495017023e-05,
+      "loss": 2.8642,
+      "step": 55395
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7446950078010559,
+      "learning_rate": 1.3278284771942849e-05,
+      "loss": 2.8721,
+      "step": 55396
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.788352370262146,
+      "learning_rate": 1.3273755348885329e-05,
+      "loss": 2.9431,
+      "step": 55397
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8451722264289856,
+      "learning_rate": 1.3269226681009525e-05,
+      "loss": 2.822,
+      "step": 55398
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7008662223815918,
+      "learning_rate": 1.3264698768327432e-05,
+      "loss": 3.1527,
+      "step": 55399
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7523859143257141,
+      "learning_rate": 1.3260171610850867e-05,
+      "loss": 2.9673,
+      "step": 55400
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7550097703933716,
+      "learning_rate": 1.3255645208591893e-05,
+      "loss": 2.9408,
+      "step": 55401
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7374807596206665,
+      "learning_rate": 1.3251119561562362e-05,
+      "loss": 2.815,
+      "step": 55402
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7401943802833557,
+      "learning_rate": 1.3246594669774169e-05,
+      "loss": 2.8424,
+      "step": 55403
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7593181729316711,
+      "learning_rate": 1.3242070533239302e-05,
+      "loss": 2.9994,
+      "step": 55404
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7926674485206604,
+      "learning_rate": 1.3237547151969552e-05,
+      "loss": 2.9654,
+      "step": 55405
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7539723515510559,
+      "learning_rate": 1.323302452597701e-05,
+      "loss": 2.8782,
+      "step": 55406
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.776358962059021,
+      "learning_rate": 1.3228502655273398e-05,
+      "loss": 3.0188,
+      "step": 55407
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7997563481330872,
+      "learning_rate": 1.322398153987081e-05,
+      "loss": 2.9002,
+      "step": 55408
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7761156558990479,
+      "learning_rate": 1.3219461179781033e-05,
+      "loss": 2.8597,
+      "step": 55409
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7830085158348083,
+      "learning_rate": 1.3214941575015958e-05,
+      "loss": 2.8994,
+      "step": 55410
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7592552304267883,
+      "learning_rate": 1.3210422725587577e-05,
+      "loss": 2.755,
+      "step": 55411
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.716784656047821,
+      "learning_rate": 1.3205904631507714e-05,
+      "loss": 2.8458,
+      "step": 55412
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7735759615898132,
+      "learning_rate": 1.3201387292788324e-05,
+      "loss": 2.9543,
+      "step": 55413
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7621521949768066,
+      "learning_rate": 1.3196870709441298e-05,
+      "loss": 3.0033,
+      "step": 55414
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7094641923904419,
+      "learning_rate": 1.3192354881478496e-05,
+      "loss": 3.1288,
+      "step": 55415
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7761197090148926,
+      "learning_rate": 1.3187839808911871e-05,
+      "loss": 3.1512,
+      "step": 55416
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8005975484848022,
+      "learning_rate": 1.318332549175325e-05,
+      "loss": 2.9146,
+      "step": 55417
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7150023579597473,
+      "learning_rate": 1.3178811930014554e-05,
+      "loss": 2.6611,
+      "step": 55418
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7599121332168579,
+      "learning_rate": 1.3174299123707677e-05,
+      "loss": 2.8705,
+      "step": 55419
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8056515455245972,
+      "learning_rate": 1.3169787072844473e-05,
+      "loss": 2.8577,
+      "step": 55420
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7697814106941223,
+      "learning_rate": 1.3165275777436867e-05,
+      "loss": 3.1584,
+      "step": 55421
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7207685112953186,
+      "learning_rate": 1.3160765237496685e-05,
+      "loss": 3.1243,
+      "step": 55422
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7401747107505798,
+      "learning_rate": 1.3156255453035914e-05,
+      "loss": 3.0593,
+      "step": 55423
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8038659691810608,
+      "learning_rate": 1.3151746424066312e-05,
+      "loss": 2.9738,
+      "step": 55424
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7309437990188599,
+      "learning_rate": 1.3147238150599804e-05,
+      "loss": 2.9488,
+      "step": 55425
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.752806544303894,
+      "learning_rate": 1.3142730632648313e-05,
+      "loss": 2.7383,
+      "step": 55426
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.72652268409729,
+      "learning_rate": 1.3138223870223597e-05,
+      "loss": 2.7769,
+      "step": 55427
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.75226891040802,
+      "learning_rate": 1.3133717863337612e-05,
+      "loss": 3.0681,
+      "step": 55428
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7673428654670715,
+      "learning_rate": 1.3129212612002215e-05,
+      "loss": 2.7021,
+      "step": 55429
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7407193779945374,
+      "learning_rate": 1.3124708116229233e-05,
+      "loss": 2.9799,
+      "step": 55430
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7485948801040649,
+      "learning_rate": 1.3120204376030652e-05,
+      "loss": 3.0959,
+      "step": 55431
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7666661739349365,
+      "learning_rate": 1.3115701391418165e-05,
+      "loss": 2.7309,
+      "step": 55432
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7030153870582581,
+      "learning_rate": 1.3111199162403697e-05,
+      "loss": 2.716,
+      "step": 55433
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6960538625717163,
+      "learning_rate": 1.3106697688999169e-05,
+      "loss": 2.7599,
+      "step": 55434
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6722820997238159,
+      "learning_rate": 1.310219697121634e-05,
+      "loss": 2.941,
+      "step": 55435
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7488369941711426,
+      "learning_rate": 1.3097697009067132e-05,
+      "loss": 2.8376,
+      "step": 55436
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7796913981437683,
+      "learning_rate": 1.3093197802563338e-05,
+      "loss": 2.9606,
+      "step": 55437
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6959371566772461,
+      "learning_rate": 1.3088699351716914e-05,
+      "loss": 2.9329,
+      "step": 55438
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7195696234703064,
+      "learning_rate": 1.3084201656539617e-05,
+      "loss": 2.8452,
+      "step": 55439
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7293813824653625,
+      "learning_rate": 1.3079704717043272e-05,
+      "loss": 2.7457,
+      "step": 55440
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7376397252082825,
+      "learning_rate": 1.3075208533239834e-05,
+      "loss": 2.6492,
+      "step": 55441
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7778947353363037,
+      "learning_rate": 1.3070713105140996e-05,
+      "loss": 2.8703,
+      "step": 55442
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7461794018745422,
+      "learning_rate": 1.3066218432758712e-05,
+      "loss": 2.8688,
+      "step": 55443
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7236195802688599,
+      "learning_rate": 1.306172451610481e-05,
+      "loss": 2.9004,
+      "step": 55444
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8100160360336304,
+      "learning_rate": 1.305723135519111e-05,
+      "loss": 2.7097,
+      "step": 55445
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7482362389564514,
+      "learning_rate": 1.305273895002944e-05,
+      "loss": 2.6586,
+      "step": 55446
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7682929635047913,
+      "learning_rate": 1.3048247300631587e-05,
+      "loss": 2.9453,
+      "step": 55447
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7298302054405212,
+      "learning_rate": 1.304375640700941e-05,
+      "loss": 2.8216,
+      "step": 55448
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7340006828308105,
+      "learning_rate": 1.303926626917483e-05,
+      "loss": 2.8007,
+      "step": 55449
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7444607615470886,
+      "learning_rate": 1.3034776887139509e-05,
+      "loss": 2.9086,
+      "step": 55450
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7466793060302734,
+      "learning_rate": 1.3030288260915433e-05,
+      "loss": 3.0851,
+      "step": 55451
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7335530519485474,
+      "learning_rate": 1.3025800390514296e-05,
+      "loss": 2.8182,
+      "step": 55452
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7638733983039856,
+      "learning_rate": 1.302131327594802e-05,
+      "loss": 2.7318,
+      "step": 55453
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8187944293022156,
+      "learning_rate": 1.3016826917228362e-05,
+      "loss": 3.0425,
+      "step": 55454
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7817355990409851,
+      "learning_rate": 1.3012341314367147e-05,
+      "loss": 2.9438,
+      "step": 55455
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7346907258033752,
+      "learning_rate": 1.3007856467376198e-05,
+      "loss": 3.1035,
+      "step": 55456
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7391301989555359,
+      "learning_rate": 1.3003372376267273e-05,
+      "loss": 3.0594,
+      "step": 55457
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7475512027740479,
+      "learning_rate": 1.2998889041052263e-05,
+      "loss": 3.0051,
+      "step": 55458
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7690040469169617,
+      "learning_rate": 1.2994406461742957e-05,
+      "loss": 2.7683,
+      "step": 55459
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7919315099716187,
+      "learning_rate": 1.298992463835118e-05,
+      "loss": 2.8209,
+      "step": 55460
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7413168549537659,
+      "learning_rate": 1.2985443570888687e-05,
+      "loss": 2.9064,
+      "step": 55461
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7446701526641846,
+      "learning_rate": 1.2980963259367238e-05,
+      "loss": 2.9471,
+      "step": 55462
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7528586983680725,
+      "learning_rate": 1.2976483703798724e-05,
+      "loss": 2.7709,
+      "step": 55463
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7157964706420898,
+      "learning_rate": 1.2972004904194965e-05,
+      "loss": 3.0465,
+      "step": 55464
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7323057651519775,
+      "learning_rate": 1.2967526860567623e-05,
+      "loss": 2.8482,
+      "step": 55465
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7675830721855164,
+      "learning_rate": 1.2963049572928652e-05,
+      "loss": 2.8903,
+      "step": 55466
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7248068451881409,
+      "learning_rate": 1.2958573041289744e-05,
+      "loss": 2.8582,
+      "step": 55467
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8281282186508179,
+      "learning_rate": 1.2954097265662689e-05,
+      "loss": 3.0083,
+      "step": 55468
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7674946784973145,
+      "learning_rate": 1.2949622246059311e-05,
+      "loss": 2.9369,
+      "step": 55469
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7439644932746887,
+      "learning_rate": 1.2945147982491367e-05,
+      "loss": 2.9113,
+      "step": 55470
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7129039764404297,
+      "learning_rate": 1.2940674474970713e-05,
+      "loss": 2.8134,
+      "step": 55471
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8135375380516052,
+      "learning_rate": 1.293620172350901e-05,
+      "loss": 2.8027,
+      "step": 55472
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.735965371131897,
+      "learning_rate": 1.2931729728118112e-05,
+      "loss": 2.8844,
+      "step": 55473
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7808622717857361,
+      "learning_rate": 1.2927258488809845e-05,
+      "loss": 3.1059,
+      "step": 55474
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6862757205963135,
+      "learning_rate": 1.2922788005595896e-05,
+      "loss": 2.6411,
+      "step": 55475
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7670363187789917,
+      "learning_rate": 1.2918318278488093e-05,
+      "loss": 2.93,
+      "step": 55476
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8030156493186951,
+      "learning_rate": 1.2913849307498158e-05,
+      "loss": 2.9131,
+      "step": 55477
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7516763806343079,
+      "learning_rate": 1.2909381092637883e-05,
+      "loss": 2.9154,
+      "step": 55478
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7685103416442871,
+      "learning_rate": 1.290491363391909e-05,
+      "loss": 2.8171,
+      "step": 55479
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7892995476722717,
+      "learning_rate": 1.290044693135347e-05,
+      "loss": 2.8467,
+      "step": 55480
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7564141154289246,
+      "learning_rate": 1.2895980984952847e-05,
+      "loss": 2.9793,
+      "step": 55481
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7547937631607056,
+      "learning_rate": 1.2891515794728979e-05,
+      "loss": 2.723,
+      "step": 55482
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7510333061218262,
+      "learning_rate": 1.2887051360693524e-05,
+      "loss": 2.8364,
+      "step": 55483
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.738362193107605,
+      "learning_rate": 1.2882587682858403e-05,
+      "loss": 3.0928,
+      "step": 55484
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7294523119926453,
+      "learning_rate": 1.2878124761235208e-05,
+      "loss": 2.8948,
+      "step": 55485
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7492228746414185,
+      "learning_rate": 1.2873662595835833e-05,
+      "loss": 2.9309,
+      "step": 55486
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7402193546295166,
+      "learning_rate": 1.286920118667193e-05,
+      "loss": 2.9529,
+      "step": 55487
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.708809494972229,
+      "learning_rate": 1.2864740533755357e-05,
+      "loss": 2.8814,
+      "step": 55488
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.749814510345459,
+      "learning_rate": 1.2860280637097776e-05,
+      "loss": 3.1164,
+      "step": 55489
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7216050624847412,
+      "learning_rate": 1.2855821496710905e-05,
+      "loss": 3.0579,
+      "step": 55490
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7654860019683838,
+      "learning_rate": 1.2851363112606572e-05,
+      "loss": 2.9416,
+      "step": 55491
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7191634178161621,
+      "learning_rate": 1.2846905484796466e-05,
+      "loss": 2.8358,
+      "step": 55492
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7747024893760681,
+      "learning_rate": 1.2842448613292378e-05,
+      "loss": 2.883,
+      "step": 55493
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7576825022697449,
+      "learning_rate": 1.2837992498105998e-05,
+      "loss": 2.858,
+      "step": 55494
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7463220953941345,
+      "learning_rate": 1.283353713924905e-05,
+      "loss": 2.8142,
+      "step": 55495
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7193440198898315,
+      "learning_rate": 1.282908253673336e-05,
+      "loss": 2.7727,
+      "step": 55496
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7219351530075073,
+      "learning_rate": 1.2824628690570582e-05,
+      "loss": 2.8686,
+      "step": 55497
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7690739035606384,
+      "learning_rate": 1.2820175600772475e-05,
+      "loss": 3.065,
+      "step": 55498
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7162508964538574,
+      "learning_rate": 1.281572326735073e-05,
+      "loss": 2.8442,
+      "step": 55499
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7276342511177063,
+      "learning_rate": 1.281127169031707e-05,
+      "loss": 3.0282,
+      "step": 55500
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7825616002082825,
+      "learning_rate": 1.2806820869683288e-05,
+      "loss": 2.9913,
+      "step": 55501
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.742638111114502,
+      "learning_rate": 1.280237080546107e-05,
+      "loss": 2.76,
+      "step": 55502
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.750740647315979,
+      "learning_rate": 1.2797921497662145e-05,
+      "loss": 3.0382,
+      "step": 55503
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7699446082115173,
+      "learning_rate": 1.2793472946298233e-05,
+      "loss": 2.8473,
+      "step": 55504
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7575497627258301,
+      "learning_rate": 1.2789025151380994e-05,
+      "loss": 2.777,
+      "step": 55505
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.9555964469909668,
+      "learning_rate": 1.2784578112922217e-05,
+      "loss": 2.9411,
+      "step": 55506
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.785099983215332,
+      "learning_rate": 1.278013183093356e-05,
+      "loss": 2.6507,
+      "step": 55507
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7826633453369141,
+      "learning_rate": 1.277568630542678e-05,
+      "loss": 2.8064,
+      "step": 55508
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.745834469795227,
+      "learning_rate": 1.2771241536413534e-05,
+      "loss": 2.7561,
+      "step": 55509
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7851356267929077,
+      "learning_rate": 1.2766797523905614e-05,
+      "loss": 2.7275,
+      "step": 55510
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.747675895690918,
+      "learning_rate": 1.2762354267914643e-05,
+      "loss": 2.7524,
+      "step": 55511
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6990408897399902,
+      "learning_rate": 1.2757911768452344e-05,
+      "loss": 2.7738,
+      "step": 55512
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7718357443809509,
+      "learning_rate": 1.2753470025530444e-05,
+      "loss": 2.9192,
+      "step": 55513
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7508569359779358,
+      "learning_rate": 1.2749029039160597e-05,
+      "loss": 2.9562,
+      "step": 55514
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.777606189250946,
+      "learning_rate": 1.2744588809354495e-05,
+      "loss": 3.0925,
+      "step": 55515
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7705445885658264,
+      "learning_rate": 1.2740149336123928e-05,
+      "loss": 2.9461,
+      "step": 55516
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7426809072494507,
+      "learning_rate": 1.2735710619480488e-05,
+      "loss": 2.9199,
+      "step": 55517
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7444571852684021,
+      "learning_rate": 1.2731272659435932e-05,
+      "loss": 2.8854,
+      "step": 55518
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7581663131713867,
+      "learning_rate": 1.2726835456001916e-05,
+      "loss": 2.9637,
+      "step": 55519
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.77213054895401,
+      "learning_rate": 1.2722399009190098e-05,
+      "loss": 2.7211,
+      "step": 55520
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7387886643409729,
+      "learning_rate": 1.2717963319012203e-05,
+      "loss": 2.9042,
+      "step": 55521
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7408025860786438,
+      "learning_rate": 1.2713528385479888e-05,
+      "loss": 2.9561,
+      "step": 55522
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.756806492805481,
+      "learning_rate": 1.2709094208604908e-05,
+      "loss": 2.6417,
+      "step": 55523
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.750133752822876,
+      "learning_rate": 1.2704660788398824e-05,
+      "loss": 3.0107,
+      "step": 55524
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7915104627609253,
+      "learning_rate": 1.270022812487339e-05,
+      "loss": 2.6029,
+      "step": 55525
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7304241061210632,
+      "learning_rate": 1.26957962180403e-05,
+      "loss": 2.8826,
+      "step": 55526
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8078506588935852,
+      "learning_rate": 1.269136506791114e-05,
+      "loss": 2.917,
+      "step": 55527
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7720636129379272,
+      "learning_rate": 1.268693467449764e-05,
+      "loss": 2.9377,
+      "step": 55528
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7359655499458313,
+      "learning_rate": 1.2682505037811452e-05,
+      "loss": 2.8316,
+      "step": 55529
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7408030033111572,
+      "learning_rate": 1.2678076157864237e-05,
+      "loss": 3.0292,
+      "step": 55530
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7604274749755859,
+      "learning_rate": 1.2673648034667749e-05,
+      "loss": 2.8554,
+      "step": 55531
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7211507558822632,
+      "learning_rate": 1.2669220668233481e-05,
+      "loss": 2.9626,
+      "step": 55532
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7372564077377319,
+      "learning_rate": 1.266479405857329e-05,
+      "loss": 3.081,
+      "step": 55533
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7603108882904053,
+      "learning_rate": 1.2660368205698667e-05,
+      "loss": 2.7713,
+      "step": 55534
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7562865018844604,
+      "learning_rate": 1.2655943109621337e-05,
+      "loss": 2.7846,
+      "step": 55535
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7241320013999939,
+      "learning_rate": 1.2651518770352954e-05,
+      "loss": 2.8448,
+      "step": 55536
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7706020474433899,
+      "learning_rate": 1.2647095187905177e-05,
+      "loss": 3.0305,
+      "step": 55537
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7894367575645447,
+      "learning_rate": 1.2642672362289663e-05,
+      "loss": 2.6124,
+      "step": 55538
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7493867874145508,
+      "learning_rate": 1.2638250293518004e-05,
+      "loss": 2.8271,
+      "step": 55539
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.781067967414856,
+      "learning_rate": 1.2633828981601923e-05,
+      "loss": 3.0305,
+      "step": 55540
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7739070653915405,
+      "learning_rate": 1.2629408426553044e-05,
+      "loss": 2.9046,
+      "step": 55541
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7560089230537415,
+      "learning_rate": 1.262498862838296e-05,
+      "loss": 2.9307,
+      "step": 55542
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7880851030349731,
+      "learning_rate": 1.262056958710339e-05,
+      "loss": 2.7827,
+      "step": 55543
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7638664841651917,
+      "learning_rate": 1.2616151302725896e-05,
+      "loss": 2.9314,
+      "step": 55544
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7615971565246582,
+      "learning_rate": 1.2611733775262133e-05,
+      "loss": 2.8241,
+      "step": 55545
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7791383266448975,
+      "learning_rate": 1.2607317004723827e-05,
+      "loss": 2.9721,
+      "step": 55546
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7734250426292419,
+      "learning_rate": 1.26029009911225e-05,
+      "loss": 3.0511,
+      "step": 55547
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7525005340576172,
+      "learning_rate": 1.2598485734469843e-05,
+      "loss": 3.016,
+      "step": 55548
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7330466508865356,
+      "learning_rate": 1.2594071234777413e-05,
+      "loss": 2.9917,
+      "step": 55549
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7011294364929199,
+      "learning_rate": 1.2589657492056904e-05,
+      "loss": 2.9246,
+      "step": 55550
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7659233212471008,
+      "learning_rate": 1.2585244506319936e-05,
+      "loss": 2.9226,
+      "step": 55551
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7099422812461853,
+      "learning_rate": 1.2580832277578102e-05,
+      "loss": 2.7514,
+      "step": 55552
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7204059958457947,
+      "learning_rate": 1.257642080584309e-05,
+      "loss": 2.8493,
+      "step": 55553
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.719089150428772,
+      "learning_rate": 1.2572010091126395e-05,
+      "loss": 2.8074,
+      "step": 55554
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7707274556159973,
+      "learning_rate": 1.256760013343977e-05,
+      "loss": 2.8689,
+      "step": 55555
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7827194333076477,
+      "learning_rate": 1.2563190932794775e-05,
+      "loss": 2.8556,
+      "step": 55556
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.724220335483551,
+      "learning_rate": 1.2558782489202967e-05,
+      "loss": 2.8183,
+      "step": 55557
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7621506452560425,
+      "learning_rate": 1.255437480267607e-05,
+      "loss": 2.8624,
+      "step": 55558
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7910413146018982,
+      "learning_rate": 1.2549967873225542e-05,
+      "loss": 2.9232,
+      "step": 55559
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7383802533149719,
+      "learning_rate": 1.2545561700863105e-05,
+      "loss": 2.8597,
+      "step": 55560
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7192094326019287,
+      "learning_rate": 1.2541156285600385e-05,
+      "loss": 2.9147,
+      "step": 55561
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7950126528739929,
+      "learning_rate": 1.2536751627448938e-05,
+      "loss": 2.9614,
+      "step": 55562
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7479918599128723,
+      "learning_rate": 1.2532347726420355e-05,
+      "loss": 3.1158,
+      "step": 55563
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7508048415184021,
+      "learning_rate": 1.2527944582526196e-05,
+      "loss": 3.1672,
+      "step": 55564
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7449748516082764,
+      "learning_rate": 1.2523542195778147e-05,
+      "loss": 2.9723,
+      "step": 55565
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.731986939907074,
+      "learning_rate": 1.2519140566187735e-05,
+      "loss": 3.0262,
+      "step": 55566
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7509217262268066,
+      "learning_rate": 1.2514739693766585e-05,
+      "loss": 3.002,
+      "step": 55567
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8176628947257996,
+      "learning_rate": 1.2510339578526286e-05,
+      "loss": 2.9815,
+      "step": 55568
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7532127499580383,
+      "learning_rate": 1.250594022047846e-05,
+      "loss": 2.8316,
+      "step": 55569
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7925772070884705,
+      "learning_rate": 1.2501541619634636e-05,
+      "loss": 2.817,
+      "step": 55570
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7416542768478394,
+      "learning_rate": 1.2497143776006402e-05,
+      "loss": 3.0308,
+      "step": 55571
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7497258186340332,
+      "learning_rate": 1.2492746689605349e-05,
+      "loss": 2.9368,
+      "step": 55572
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7755299210548401,
+      "learning_rate": 1.24883503604431e-05,
+      "loss": 2.9622,
+      "step": 55573
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.772011399269104,
+      "learning_rate": 1.248395478853118e-05,
+      "loss": 2.8229,
+      "step": 55574
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7620413899421692,
+      "learning_rate": 1.2479559973881248e-05,
+      "loss": 2.9222,
+      "step": 55575
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7677061557769775,
+      "learning_rate": 1.2475165916504726e-05,
+      "loss": 3.0159,
+      "step": 55576
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.72245854139328,
+      "learning_rate": 1.2470772616413371e-05,
+      "loss": 2.8573,
+      "step": 55577
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7207547426223755,
+      "learning_rate": 1.2466380073618642e-05,
+      "loss": 2.7114,
+      "step": 55578
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7339671850204468,
+      "learning_rate": 1.2461988288132097e-05,
+      "loss": 2.9585,
+      "step": 55579
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7937641143798828,
+      "learning_rate": 1.2457597259965358e-05,
+      "loss": 2.8606,
+      "step": 55580
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7468560338020325,
+      "learning_rate": 1.2453206989129949e-05,
+      "loss": 2.8659,
+      "step": 55581
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7348960638046265,
+      "learning_rate": 1.2448817475637462e-05,
+      "loss": 2.9035,
+      "step": 55582
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7143893241882324,
+      "learning_rate": 1.2444428719499488e-05,
+      "loss": 2.892,
+      "step": 55583
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.787746012210846,
+      "learning_rate": 1.2440040720727518e-05,
+      "loss": 2.8582,
+      "step": 55584
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7754120826721191,
+      "learning_rate": 1.2435653479333173e-05,
+      "loss": 2.696,
+      "step": 55585
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7287518382072449,
+      "learning_rate": 1.2431266995327916e-05,
+      "loss": 2.9811,
+      "step": 55586
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7545494437217712,
+      "learning_rate": 1.2426881268723365e-05,
+      "loss": 2.7484,
+      "step": 55587
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7642377614974976,
+      "learning_rate": 1.2422496299531081e-05,
+      "loss": 2.8666,
+      "step": 55588
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7176851034164429,
+      "learning_rate": 1.2418112087762588e-05,
+      "loss": 2.8306,
+      "step": 55589
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7618530988693237,
+      "learning_rate": 1.2413728633429476e-05,
+      "loss": 2.8173,
+      "step": 55590
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7074477076530457,
+      "learning_rate": 1.2409345936543235e-05,
+      "loss": 2.847,
+      "step": 55591
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7396611571311951,
+      "learning_rate": 1.2404963997115391e-05,
+      "loss": 2.8979,
+      "step": 55592
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7340872287750244,
+      "learning_rate": 1.24005828151576e-05,
+      "loss": 2.9145,
+      "step": 55593
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7579055428504944,
+      "learning_rate": 1.2396202390681254e-05,
+      "loss": 2.8307,
+      "step": 55594
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7400532364845276,
+      "learning_rate": 1.2391822723698009e-05,
+      "loss": 2.8784,
+      "step": 55595
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7686823606491089,
+      "learning_rate": 1.238744381421929e-05,
+      "loss": 2.8043,
+      "step": 55596
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7867351770401001,
+      "learning_rate": 1.2383065662256719e-05,
+      "loss": 2.8821,
+      "step": 55597
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7557711005210876,
+      "learning_rate": 1.2378688267821824e-05,
+      "loss": 2.6193,
+      "step": 55598
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7336000204086304,
+      "learning_rate": 1.2374311630926126e-05,
+      "loss": 2.7775,
+      "step": 55599
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8032825589179993,
+      "learning_rate": 1.2369935751581117e-05,
+      "loss": 3.0363,
+      "step": 55600
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7513951063156128,
+      "learning_rate": 1.2365560629798322e-05,
+      "loss": 2.9107,
+      "step": 55601
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7409695386886597,
+      "learning_rate": 1.2361186265589262e-05,
+      "loss": 2.9644,
+      "step": 55602
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.7681583166122437,
+      "learning_rate": 1.2356812658965532e-05,
+      "loss": 2.9018,
+      "step": 55603
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.745311439037323,
+      "learning_rate": 1.2352439809938585e-05,
+      "loss": 2.8465,
+      "step": 55604
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7664110660552979,
+      "learning_rate": 1.2348067718519949e-05,
+      "loss": 3.0799,
+      "step": 55605
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7889949679374695,
+      "learning_rate": 1.234369638472118e-05,
+      "loss": 2.9645,
+      "step": 55606
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7036034464836121,
+      "learning_rate": 1.233932580855367e-05,
+      "loss": 2.9552,
+      "step": 55607
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.726105272769928,
+      "learning_rate": 1.2334955990029105e-05,
+      "loss": 3.079,
+      "step": 55608
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.741583526134491,
+      "learning_rate": 1.2330586929158814e-05,
+      "loss": 2.8413,
+      "step": 55609
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7955318689346313,
+      "learning_rate": 1.2326218625954454e-05,
+      "loss": 2.9633,
+      "step": 55610
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.874595046043396,
+      "learning_rate": 1.2321851080427448e-05,
+      "loss": 2.8424,
+      "step": 55611
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7419061064720154,
+      "learning_rate": 1.2317484292589352e-05,
+      "loss": 2.7988,
+      "step": 55612
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8153803944587708,
+      "learning_rate": 1.2313118262451627e-05,
+      "loss": 2.7867,
+      "step": 55613
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7266865968704224,
+      "learning_rate": 1.2308752990025761e-05,
+      "loss": 2.5696,
+      "step": 55614
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7737489938735962,
+      "learning_rate": 1.230438847532328e-05,
+      "loss": 3.1006,
+      "step": 55615
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.762378990650177,
+      "learning_rate": 1.2300024718355673e-05,
+      "loss": 3.0235,
+      "step": 55616
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.710425078868866,
+      "learning_rate": 1.2295661719134397e-05,
+      "loss": 2.9223,
+      "step": 55617
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7368878722190857,
+      "learning_rate": 1.2291299477671047e-05,
+      "loss": 2.8503,
+      "step": 55618
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7075875997543335,
+      "learning_rate": 1.2286937993976976e-05,
+      "loss": 2.8905,
+      "step": 55619
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7213077545166016,
+      "learning_rate": 1.2282577268063777e-05,
+      "loss": 2.8651,
+      "step": 55620
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7512006163597107,
+      "learning_rate": 1.2278217299942939e-05,
+      "loss": 2.6974,
+      "step": 55621
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7457561492919922,
+      "learning_rate": 1.2273858089625821e-05,
+      "loss": 3.0884,
+      "step": 55622
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8021249771118164,
+      "learning_rate": 1.2269499637124047e-05,
+      "loss": 2.9176,
+      "step": 55623
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7707574963569641,
+      "learning_rate": 1.2265141942448975e-05,
+      "loss": 2.8524,
+      "step": 55624
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7147628664970398,
+      "learning_rate": 1.2260785005612195e-05,
+      "loss": 2.8479,
+      "step": 55625
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7359575629234314,
+      "learning_rate": 1.2256428826625098e-05,
+      "loss": 2.7975,
+      "step": 55626
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7357754707336426,
+      "learning_rate": 1.2252073405499241e-05,
+      "loss": 2.9908,
+      "step": 55627
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8136209845542908,
+      "learning_rate": 1.2247718742246016e-05,
+      "loss": 2.988,
+      "step": 55628
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7402750253677368,
+      "learning_rate": 1.2243364836876912e-05,
+      "loss": 2.8614,
+      "step": 55629
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7957865595817566,
+      "learning_rate": 1.2239011689403422e-05,
+      "loss": 2.7521,
+      "step": 55630
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7921202182769775,
+      "learning_rate": 1.2234659299836969e-05,
+      "loss": 2.8699,
+      "step": 55631
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7857522964477539,
+      "learning_rate": 1.223030766818901e-05,
+      "loss": 2.8791,
+      "step": 55632
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7415589094161987,
+      "learning_rate": 1.2225956794471103e-05,
+      "loss": 2.8202,
+      "step": 55633
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7329033017158508,
+      "learning_rate": 1.222160667869464e-05,
+      "loss": 2.8956,
+      "step": 55634
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7467180490493774,
+      "learning_rate": 1.2217257320871076e-05,
+      "loss": 2.7925,
+      "step": 55635
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.841463029384613,
+      "learning_rate": 1.2212908721011837e-05,
+      "loss": 2.9325,
+      "step": 55636
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7936882376670837,
+      "learning_rate": 1.2208560879128416e-05,
+      "loss": 2.9749,
+      "step": 55637
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7346864938735962,
+      "learning_rate": 1.2204213795232298e-05,
+      "loss": 2.9747,
+      "step": 55638
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7802562713623047,
+      "learning_rate": 1.2199867469334845e-05,
+      "loss": 2.9534,
+      "step": 55639
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7328370213508606,
+      "learning_rate": 1.2195521901447614e-05,
+      "loss": 2.9276,
+      "step": 55640
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7416670918464661,
+      "learning_rate": 1.2191177091581927e-05,
+      "loss": 2.9684,
+      "step": 55641
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7478400468826294,
+      "learning_rate": 1.2186833039749344e-05,
+      "loss": 3.1439,
+      "step": 55642
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7259265780448914,
+      "learning_rate": 1.218248974596122e-05,
+      "loss": 2.9068,
+      "step": 55643
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7286504507064819,
+      "learning_rate": 1.2178147210229017e-05,
+      "loss": 2.7912,
+      "step": 55644
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.804137647151947,
+      "learning_rate": 1.2173805432564221e-05,
+      "loss": 3.0854,
+      "step": 55645
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7600886821746826,
+      "learning_rate": 1.2169464412978192e-05,
+      "loss": 3.1028,
+      "step": 55646
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7712217569351196,
+      "learning_rate": 1.2165124151482419e-05,
+      "loss": 3.0731,
+      "step": 55647
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7367176413536072,
+      "learning_rate": 1.2160784648088295e-05,
+      "loss": 2.9482,
+      "step": 55648
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7509863376617432,
+      "learning_rate": 1.2156445902807278e-05,
+      "loss": 2.8918,
+      "step": 55649
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7302809357643127,
+      "learning_rate": 1.2152107915650821e-05,
+      "loss": 3.0575,
+      "step": 55650
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7326139211654663,
+      "learning_rate": 1.2147770686630254e-05,
+      "loss": 2.9015,
+      "step": 55651
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.722706139087677,
+      "learning_rate": 1.2143434215757097e-05,
+      "loss": 2.8909,
+      "step": 55652
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7505283355712891,
+      "learning_rate": 1.213909850304271e-05,
+      "loss": 2.7945,
+      "step": 55653
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7195035815238953,
+      "learning_rate": 1.213476354849855e-05,
+      "loss": 2.7821,
+      "step": 55654
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7279053330421448,
+      "learning_rate": 1.2130429352136039e-05,
+      "loss": 2.8279,
+      "step": 55655
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7183561325073242,
+      "learning_rate": 1.2126095913966538e-05,
+      "loss": 2.8743,
+      "step": 55656
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7786692976951599,
+      "learning_rate": 1.2121763234001536e-05,
+      "loss": 2.9192,
+      "step": 55657
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7574261426925659,
+      "learning_rate": 1.211743131225239e-05,
+      "loss": 2.8047,
+      "step": 55658
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7224125862121582,
+      "learning_rate": 1.2113100148730526e-05,
+      "loss": 3.0287,
+      "step": 55659
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7393490076065063,
+      "learning_rate": 1.2108769743447366e-05,
+      "loss": 2.8423,
+      "step": 55660
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.746552586555481,
+      "learning_rate": 1.2104440096414269e-05,
+      "loss": 2.7179,
+      "step": 55661
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7337487936019897,
+      "learning_rate": 1.2100111207642693e-05,
+      "loss": 2.864,
+      "step": 55662
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7381230592727661,
+      "learning_rate": 1.2095783077143994e-05,
+      "loss": 2.9342,
+      "step": 55663
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7539899945259094,
+      "learning_rate": 1.2091455704929631e-05,
+      "loss": 2.8978,
+      "step": 55664
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8014904856681824,
+      "learning_rate": 1.208712909101096e-05,
+      "loss": 2.7139,
+      "step": 55665
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7007734775543213,
+      "learning_rate": 1.2082803235399374e-05,
+      "loss": 2.9909,
+      "step": 55666
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7687050700187683,
+      "learning_rate": 1.2078478138106262e-05,
+      "loss": 2.9276,
+      "step": 55667
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7429559230804443,
+      "learning_rate": 1.2074153799143016e-05,
+      "loss": 2.8966,
+      "step": 55668
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7226108312606812,
+      "learning_rate": 1.2069830218521025e-05,
+      "loss": 2.9716,
+      "step": 55669
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7323675751686096,
+      "learning_rate": 1.206550739625175e-05,
+      "loss": 2.9101,
+      "step": 55670
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7690631151199341,
+      "learning_rate": 1.2061185332346478e-05,
+      "loss": 2.8661,
+      "step": 55671
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8044900894165039,
+      "learning_rate": 1.2056864026816671e-05,
+      "loss": 2.8463,
+      "step": 55672
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7190569043159485,
+      "learning_rate": 1.2052543479673582e-05,
+      "loss": 3.0036,
+      "step": 55673
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7784371376037598,
+      "learning_rate": 1.2048223690928704e-05,
+      "loss": 2.9049,
+      "step": 55674
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7170578241348267,
+      "learning_rate": 1.2043904660593429e-05,
+      "loss": 2.8486,
+      "step": 55675
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7676547765731812,
+      "learning_rate": 1.2039586388679046e-05,
+      "loss": 2.9257,
+      "step": 55676
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7858740091323853,
+      "learning_rate": 1.2035268875197014e-05,
+      "loss": 2.7855,
+      "step": 55677
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7509922981262207,
+      "learning_rate": 1.2030952120158621e-05,
+      "loss": 3.0946,
+      "step": 55678
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7644749283790588,
+      "learning_rate": 1.2026636123575329e-05,
+      "loss": 2.892,
+      "step": 55679
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7936258912086487,
+      "learning_rate": 1.2022320885458458e-05,
+      "loss": 2.7818,
+      "step": 55680
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7399991750717163,
+      "learning_rate": 1.2018006405819335e-05,
+      "loss": 3.1416,
+      "step": 55681
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7325205206871033,
+      "learning_rate": 1.2013692684669386e-05,
+      "loss": 2.7667,
+      "step": 55682
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7739435434341431,
+      "learning_rate": 1.2009379722019896e-05,
+      "loss": 2.9389,
+      "step": 55683
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.769698441028595,
+      "learning_rate": 1.2005067517882294e-05,
+      "loss": 2.8377,
+      "step": 55684
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.752373456954956,
+      "learning_rate": 1.2000756072267936e-05,
+      "loss": 2.7622,
+      "step": 55685
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7731847763061523,
+      "learning_rate": 1.1996445385188181e-05,
+      "loss": 3.0762,
+      "step": 55686
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7705298662185669,
+      "learning_rate": 1.199213545665435e-05,
+      "loss": 2.8453,
+      "step": 55687
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7608867287635803,
+      "learning_rate": 1.1987826286677771e-05,
+      "loss": 3.0201,
+      "step": 55688
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7582928538322449,
+      "learning_rate": 1.1983517875269832e-05,
+      "loss": 2.9205,
+      "step": 55689
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7286206483840942,
+      "learning_rate": 1.1979210222441925e-05,
+      "loss": 2.9747,
+      "step": 55690
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7775378227233887,
+      "learning_rate": 1.1974903328205276e-05,
+      "loss": 2.792,
+      "step": 55691
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7497450709342957,
+      "learning_rate": 1.1970597192571374e-05,
+      "loss": 2.9682,
+      "step": 55692
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7265482544898987,
+      "learning_rate": 1.1966291815551477e-05,
+      "loss": 2.9189,
+      "step": 55693
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7717911601066589,
+      "learning_rate": 1.1961987197156876e-05,
+      "loss": 3.1288,
+      "step": 55694
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.803691565990448,
+      "learning_rate": 1.1957683337399026e-05,
+      "loss": 2.978,
+      "step": 55695
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7337557673454285,
+      "learning_rate": 1.1953380236289156e-05,
+      "loss": 2.7589,
+      "step": 55696
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7882424592971802,
+      "learning_rate": 1.1949077893838688e-05,
+      "loss": 2.9169,
+      "step": 55697
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7700529098510742,
+      "learning_rate": 1.194477631005888e-05,
+      "loss": 2.88,
+      "step": 55698
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7207964658737183,
+      "learning_rate": 1.1940475484961087e-05,
+      "loss": 2.7995,
+      "step": 55699
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7262753844261169,
+      "learning_rate": 1.193617541855667e-05,
+      "loss": 2.9706,
+      "step": 55700
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7570127844810486,
+      "learning_rate": 1.1931876110856953e-05,
+      "loss": 2.9817,
+      "step": 55701
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7883241176605225,
+      "learning_rate": 1.1927577561873225e-05,
+      "loss": 2.7953,
+      "step": 55702
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.773369550704956,
+      "learning_rate": 1.1923279771616778e-05,
+      "loss": 2.8493,
+      "step": 55703
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7068461179733276,
+      "learning_rate": 1.1918982740098971e-05,
+      "loss": 3.0174,
+      "step": 55704
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7062078714370728,
+      "learning_rate": 1.1914686467331158e-05,
+      "loss": 2.6588,
+      "step": 55705
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7688685059547424,
+      "learning_rate": 1.1910390953324567e-05,
+      "loss": 2.9203,
+      "step": 55706
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7777459621429443,
+      "learning_rate": 1.190609619809062e-05,
+      "loss": 2.9701,
+      "step": 55707
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7545894384384155,
+      "learning_rate": 1.1901802201640542e-05,
+      "loss": 2.8712,
+      "step": 55708
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7288576364517212,
+      "learning_rate": 1.1897508963985659e-05,
+      "loss": 2.7988,
+      "step": 55709
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.806057333946228,
+      "learning_rate": 1.1893216485137325e-05,
+      "loss": 2.8109,
+      "step": 55710
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7674176692962646,
+      "learning_rate": 1.18889247651068e-05,
+      "loss": 2.8256,
+      "step": 55711
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7585336565971375,
+      "learning_rate": 1.1884633803905408e-05,
+      "loss": 2.8867,
+      "step": 55712
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7617801427841187,
+      "learning_rate": 1.1880343601544406e-05,
+      "loss": 2.9724,
+      "step": 55713
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.814563512802124,
+      "learning_rate": 1.1876054158035152e-05,
+      "loss": 2.8553,
+      "step": 55714
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7293022871017456,
+      "learning_rate": 1.187176547338894e-05,
+      "loss": 2.9574,
+      "step": 55715
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7652244567871094,
+      "learning_rate": 1.1867477547616988e-05,
+      "loss": 2.9344,
+      "step": 55716
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7348045706748962,
+      "learning_rate": 1.1863190380730725e-05,
+      "loss": 3.0233,
+      "step": 55717
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.6973322033882141,
+      "learning_rate": 1.1858903972741308e-05,
+      "loss": 2.9189,
+      "step": 55718
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7177212238311768,
+      "learning_rate": 1.1854618323660092e-05,
+      "loss": 2.5284,
+      "step": 55719
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7871966361999512,
+      "learning_rate": 1.1850333433498372e-05,
+      "loss": 3.0959,
+      "step": 55720
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7772104144096375,
+      "learning_rate": 1.1846049302267369e-05,
+      "loss": 2.8529,
+      "step": 55721
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7406902313232422,
+      "learning_rate": 1.1841765929978476e-05,
+      "loss": 2.917,
+      "step": 55722
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7036511301994324,
+      "learning_rate": 1.1837483316642915e-05,
+      "loss": 3.015,
+      "step": 55723
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7832501530647278,
+      "learning_rate": 1.183320146227198e-05,
+      "loss": 2.8413,
+      "step": 55724
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7509165406227112,
+      "learning_rate": 1.182892036687686e-05,
+      "loss": 2.9039,
+      "step": 55725
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7857611179351807,
+      "learning_rate": 1.1824640030468945e-05,
+      "loss": 3.0673,
+      "step": 55726
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7503923177719116,
+      "learning_rate": 1.1820360453059463e-05,
+      "loss": 2.8432,
+      "step": 55727
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7694743275642395,
+      "learning_rate": 1.1816081634659668e-05,
+      "loss": 2.9747,
+      "step": 55728
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7916954755783081,
+      "learning_rate": 1.1811803575280921e-05,
+      "loss": 2.7779,
+      "step": 55729
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8194757699966431,
+      "learning_rate": 1.1807526274934375e-05,
+      "loss": 2.7108,
+      "step": 55730
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7391253113746643,
+      "learning_rate": 1.1803249733631326e-05,
+      "loss": 3.1041,
+      "step": 55731
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7366408705711365,
+      "learning_rate": 1.1798973951383094e-05,
+      "loss": 2.9705,
+      "step": 55732
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7290082573890686,
+      "learning_rate": 1.179469892820084e-05,
+      "loss": 2.6135,
+      "step": 55733
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7617344856262207,
+      "learning_rate": 1.1790424664095954e-05,
+      "loss": 3.099,
+      "step": 55734
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7532030940055847,
+      "learning_rate": 1.178615115907956e-05,
+      "loss": 2.7876,
+      "step": 55735
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7590699195861816,
+      "learning_rate": 1.1781878413163015e-05,
+      "loss": 2.9633,
+      "step": 55736
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7126039862632751,
+      "learning_rate": 1.1777606426357544e-05,
+      "loss": 2.9342,
+      "step": 55737
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7936747074127197,
+      "learning_rate": 1.1773335198674338e-05,
+      "loss": 2.7536,
+      "step": 55738
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7752701044082642,
+      "learning_rate": 1.1769064730124756e-05,
+      "loss": 3.0024,
+      "step": 55739
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7142334580421448,
+      "learning_rate": 1.176479502071992e-05,
+      "loss": 3.15,
+      "step": 55740
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7375084757804871,
+      "learning_rate": 1.1760526070471154e-05,
+      "loss": 2.8894,
+      "step": 55741
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7870582342147827,
+      "learning_rate": 1.1756257879389752e-05,
+      "loss": 2.8652,
+      "step": 55742
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7322790026664734,
+      "learning_rate": 1.1751990447486836e-05,
+      "loss": 3.0524,
+      "step": 55743
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7427972555160522,
+      "learning_rate": 1.1747723774773733e-05,
+      "loss": 2.6524,
+      "step": 55744
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.766864538192749,
+      "learning_rate": 1.1743457861261662e-05,
+      "loss": 2.7924,
+      "step": 55745
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.729448139667511,
+      "learning_rate": 1.1739192706961786e-05,
+      "loss": 2.9511,
+      "step": 55746
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7730950117111206,
+      "learning_rate": 1.1734928311885461e-05,
+      "loss": 2.9649,
+      "step": 55747
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7573014497756958,
+      "learning_rate": 1.173066467604381e-05,
+      "loss": 2.9444,
+      "step": 55748
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.768448531627655,
+      "learning_rate": 1.172640179944816e-05,
+      "loss": 2.8058,
+      "step": 55749
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7829905152320862,
+      "learning_rate": 1.1722139682109633e-05,
+      "loss": 2.6185,
+      "step": 55750
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7292187809944153,
+      "learning_rate": 1.1717878324039554e-05,
+      "loss": 2.919,
+      "step": 55751
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7217176556587219,
+      "learning_rate": 1.171361772524908e-05,
+      "loss": 2.7037,
+      "step": 55752
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7316499948501587,
+      "learning_rate": 1.1709357885749437e-05,
+      "loss": 2.9193,
+      "step": 55753
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.748083233833313,
+      "learning_rate": 1.1705098805551916e-05,
+      "loss": 2.8312,
+      "step": 55754
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7519057393074036,
+      "learning_rate": 1.1700840484667639e-05,
+      "loss": 2.8975,
+      "step": 55755
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7586154937744141,
+      "learning_rate": 1.1696582923107866e-05,
+      "loss": 2.8254,
+      "step": 55756
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7953873872756958,
+      "learning_rate": 1.1692326120883822e-05,
+      "loss": 2.7867,
+      "step": 55757
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.733325183391571,
+      "learning_rate": 1.1688070078006695e-05,
+      "loss": 2.8756,
+      "step": 55758
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7922911047935486,
+      "learning_rate": 1.1683814794487744e-05,
+      "loss": 3.1538,
+      "step": 55759
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7304345369338989,
+      "learning_rate": 1.1679560270338095e-05,
+      "loss": 3.0398,
+      "step": 55760
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8140565752983093,
+      "learning_rate": 1.1675306505568972e-05,
+      "loss": 2.7469,
+      "step": 55761
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7204803824424744,
+      "learning_rate": 1.1671053500191663e-05,
+      "loss": 2.8597,
+      "step": 55762
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7560591101646423,
+      "learning_rate": 1.1666801254217262e-05,
+      "loss": 2.7704,
+      "step": 55763
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7345255017280579,
+      "learning_rate": 1.1662549767657059e-05,
+      "loss": 2.8837,
+      "step": 55764
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7856339812278748,
+      "learning_rate": 1.1658299040522145e-05,
+      "loss": 2.748,
+      "step": 55765
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7450258135795593,
+      "learning_rate": 1.1654049072823845e-05,
+      "loss": 2.8668,
+      "step": 55766
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.751708984375,
+      "learning_rate": 1.1649799864573284e-05,
+      "loss": 2.9728,
+      "step": 55767
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7992194890975952,
+      "learning_rate": 1.1645551415781618e-05,
+      "loss": 2.9086,
+      "step": 55768
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7067645192146301,
+      "learning_rate": 1.1641303726460106e-05,
+      "loss": 2.8046,
+      "step": 55769
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7081477046012878,
+      "learning_rate": 1.1637056796619871e-05,
+      "loss": 3.1195,
+      "step": 55770
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.733273983001709,
+      "learning_rate": 1.1632810626272104e-05,
+      "loss": 2.7449,
+      "step": 55771
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7776265740394592,
+      "learning_rate": 1.1628565215428099e-05,
+      "loss": 2.8875,
+      "step": 55772
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7776868939399719,
+      "learning_rate": 1.1624320564098911e-05,
+      "loss": 2.5992,
+      "step": 55773
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.783747136592865,
+      "learning_rate": 1.1620076672295797e-05,
+      "loss": 2.9476,
+      "step": 55774
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7857709527015686,
+      "learning_rate": 1.161583354002985e-05,
+      "loss": 3.107,
+      "step": 55775
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.729192316532135,
+      "learning_rate": 1.1611591167312295e-05,
+      "loss": 2.7004,
+      "step": 55776
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7840185165405273,
+      "learning_rate": 1.1607349554154355e-05,
+      "loss": 3.0569,
+      "step": 55777
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7330710887908936,
+      "learning_rate": 1.1603108700567121e-05,
+      "loss": 2.8325,
+      "step": 55778
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7933141589164734,
+      "learning_rate": 1.1598868606561817e-05,
+      "loss": 2.984,
+      "step": 55779
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7475676536560059,
+      "learning_rate": 1.1594629272149568e-05,
+      "loss": 2.7584,
+      "step": 55780
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7124994397163391,
+      "learning_rate": 1.1590390697341567e-05,
+      "loss": 2.9443,
+      "step": 55781
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7357338666915894,
+      "learning_rate": 1.1586152882149002e-05,
+      "loss": 3.0484,
+      "step": 55782
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7515398859977722,
+      "learning_rate": 1.1581915826582966e-05,
+      "loss": 3.0142,
+      "step": 55783
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8320146203041077,
+      "learning_rate": 1.1577679530654681e-05,
+      "loss": 2.9811,
+      "step": 55784
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7434693574905396,
+      "learning_rate": 1.1573443994375243e-05,
+      "loss": 2.9766,
+      "step": 55785
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7687200307846069,
+      "learning_rate": 1.1569209217755871e-05,
+      "loss": 2.8711,
+      "step": 55786
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7312271595001221,
+      "learning_rate": 1.1564975200807692e-05,
+      "loss": 2.8606,
+      "step": 55787
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.77076256275177,
+      "learning_rate": 1.1560741943541862e-05,
+      "loss": 2.7525,
+      "step": 55788
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7475319504737854,
+      "learning_rate": 1.1556509445969542e-05,
+      "loss": 2.8676,
+      "step": 55789
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7481474876403809,
+      "learning_rate": 1.155227770810182e-05,
+      "loss": 2.707,
+      "step": 55790
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7318161725997925,
+      "learning_rate": 1.154804672994989e-05,
+      "loss": 2.7252,
+      "step": 55791
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7487803101539612,
+      "learning_rate": 1.154381651152494e-05,
+      "loss": 2.6743,
+      "step": 55792
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7442032694816589,
+      "learning_rate": 1.1539587052837995e-05,
+      "loss": 2.8245,
+      "step": 55793
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7299758195877075,
+      "learning_rate": 1.1535358353900314e-05,
+      "loss": 2.7304,
+      "step": 55794
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7423626780509949,
+      "learning_rate": 1.1531130414722988e-05,
+      "loss": 2.8714,
+      "step": 55795
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7554141879081726,
+      "learning_rate": 1.1526903235317142e-05,
+      "loss": 2.9592,
+      "step": 55796
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7774835824966431,
+      "learning_rate": 1.1522676815693899e-05,
+      "loss": 2.9311,
+      "step": 55797
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7268118262290955,
+      "learning_rate": 1.1518451155864383e-05,
+      "loss": 2.9979,
+      "step": 55798
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7733696103096008,
+      "learning_rate": 1.1514226255839786e-05,
+      "loss": 2.7952,
+      "step": 55799
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7510420680046082,
+      "learning_rate": 1.1510002115631167e-05,
+      "loss": 2.9407,
+      "step": 55800
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7454401254653931,
+      "learning_rate": 1.1505778735249715e-05,
+      "loss": 2.8337,
+      "step": 55801
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7559973001480103,
+      "learning_rate": 1.1501556114706489e-05,
+      "loss": 2.7546,
+      "step": 55802
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7908011674880981,
+      "learning_rate": 1.149733425401268e-05,
+      "loss": 2.935,
+      "step": 55803
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7167772650718689,
+      "learning_rate": 1.1493113153179345e-05,
+      "loss": 2.8562,
+      "step": 55804
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.8508129119873047,
+      "learning_rate": 1.1488892812217609e-05,
+      "loss": 2.9246,
+      "step": 55805
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7644029855728149,
+      "learning_rate": 1.148467323113863e-05,
+      "loss": 2.9711,
+      "step": 55806
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7415115237236023,
+      "learning_rate": 1.1480454409953466e-05,
+      "loss": 2.8969,
+      "step": 55807
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7756478190422058,
+      "learning_rate": 1.1476236348673273e-05,
+      "loss": 3.0643,
+      "step": 55808
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7457125782966614,
+      "learning_rate": 1.1472019047309177e-05,
+      "loss": 2.823,
+      "step": 55809
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.763870358467102,
+      "learning_rate": 1.1467802505872236e-05,
+      "loss": 3.125,
+      "step": 55810
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8004308342933655,
+      "learning_rate": 1.1463586724373574e-05,
+      "loss": 2.7832,
+      "step": 55811
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7365848422050476,
+      "learning_rate": 1.1459371702824283e-05,
+      "loss": 2.9564,
+      "step": 55812
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.746489942073822,
+      "learning_rate": 1.1455157441235452e-05,
+      "loss": 2.8939,
+      "step": 55813
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7295230031013489,
+      "learning_rate": 1.1450943939618273e-05,
+      "loss": 2.8244,
+      "step": 55814
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7783901691436768,
+      "learning_rate": 1.1446731197983706e-05,
+      "loss": 3.006,
+      "step": 55815
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7580480575561523,
+      "learning_rate": 1.1442519216342972e-05,
+      "loss": 2.9875,
+      "step": 55816
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7256256341934204,
+      "learning_rate": 1.1438307994707097e-05,
+      "loss": 3.119,
+      "step": 55817
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7361882925033569,
+      "learning_rate": 1.1434097533087139e-05,
+      "loss": 2.7991,
+      "step": 55818
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7718648910522461,
+      "learning_rate": 1.1429887831494289e-05,
+      "loss": 3.0094,
+      "step": 55819
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7573488354682922,
+      "learning_rate": 1.1425678889939538e-05,
+      "loss": 2.8855,
+      "step": 55820
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7509820461273193,
+      "learning_rate": 1.1421470708434044e-05,
+      "loss": 3.0757,
+      "step": 55821
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7302854657173157,
+      "learning_rate": 1.1417263286988798e-05,
+      "loss": 3.0187,
+      "step": 55822
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7011269927024841,
+      "learning_rate": 1.1413056625614991e-05,
+      "loss": 2.9904,
+      "step": 55823
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8346005082130432,
+      "learning_rate": 1.1408850724323648e-05,
+      "loss": 3.1022,
+      "step": 55824
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7266502380371094,
+      "learning_rate": 1.140464558312586e-05,
+      "loss": 2.9047,
+      "step": 55825
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7775142788887024,
+      "learning_rate": 1.1400441202032717e-05,
+      "loss": 2.9262,
+      "step": 55826
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7800171971321106,
+      "learning_rate": 1.1396237581055212e-05,
+      "loss": 2.7697,
+      "step": 55827
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7560826539993286,
+      "learning_rate": 1.1392034720204468e-05,
+      "loss": 2.7127,
+      "step": 55828
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7661461234092712,
+      "learning_rate": 1.138783261949161e-05,
+      "loss": 2.9436,
+      "step": 55829
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7515476942062378,
+      "learning_rate": 1.138363127892763e-05,
+      "loss": 2.8215,
+      "step": 55830
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7410834431648254,
+      "learning_rate": 1.137943069852365e-05,
+      "loss": 2.9039,
+      "step": 55831
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7249680161476135,
+      "learning_rate": 1.1375230878290697e-05,
+      "loss": 2.8317,
+      "step": 55832
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7653201818466187,
+      "learning_rate": 1.1371031818239795e-05,
+      "loss": 2.7605,
+      "step": 55833
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7415288090705872,
+      "learning_rate": 1.1366833518382102e-05,
+      "loss": 3.0566,
+      "step": 55834
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7455683946609497,
+      "learning_rate": 1.1362635978728574e-05,
+      "loss": 2.8293,
+      "step": 55835
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7558901906013489,
+      "learning_rate": 1.1358439199290337e-05,
+      "loss": 2.8531,
+      "step": 55836
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6817432641983032,
+      "learning_rate": 1.1354243180078415e-05,
+      "loss": 3.1373,
+      "step": 55837
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7174025774002075,
+      "learning_rate": 1.1350047921103866e-05,
+      "loss": 2.8915,
+      "step": 55838
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.759171187877655,
+      "learning_rate": 1.1345853422377749e-05,
+      "loss": 2.965,
+      "step": 55839
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7580920457839966,
+      "learning_rate": 1.1341659683911086e-05,
+      "loss": 3.0421,
+      "step": 55840
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7607132792472839,
+      "learning_rate": 1.133746670571497e-05,
+      "loss": 2.8414,
+      "step": 55841
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7223249077796936,
+      "learning_rate": 1.133327448780036e-05,
+      "loss": 3.2551,
+      "step": 55842
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7324064373970032,
+      "learning_rate": 1.1329083030178376e-05,
+      "loss": 2.8549,
+      "step": 55843
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8532474040985107,
+      "learning_rate": 1.1324892332860047e-05,
+      "loss": 2.9252,
+      "step": 55844
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7820866703987122,
+      "learning_rate": 1.1320702395856362e-05,
+      "loss": 2.8606,
+      "step": 55845
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7099783420562744,
+      "learning_rate": 1.1316513219178413e-05,
+      "loss": 2.785,
+      "step": 55846
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7845707535743713,
+      "learning_rate": 1.1312324802837225e-05,
+      "loss": 2.8446,
+      "step": 55847
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7959213852882385,
+      "learning_rate": 1.1308137146843787e-05,
+      "loss": 2.9116,
+      "step": 55848
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.780271589756012,
+      "learning_rate": 1.130395025120916e-05,
+      "loss": 2.7504,
+      "step": 55849
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7800261378288269,
+      "learning_rate": 1.1299764115944365e-05,
+      "loss": 3.0029,
+      "step": 55850
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7617143392562866,
+      "learning_rate": 1.1295578741060463e-05,
+      "loss": 2.8185,
+      "step": 55851
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7690078616142273,
+      "learning_rate": 1.1291394126568376e-05,
+      "loss": 2.8685,
+      "step": 55852
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7068555355072021,
+      "learning_rate": 1.1287210272479264e-05,
+      "loss": 2.9051,
+      "step": 55853
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7279587388038635,
+      "learning_rate": 1.1283027178804082e-05,
+      "loss": 2.832,
+      "step": 55854
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7110167145729065,
+      "learning_rate": 1.1278844845553792e-05,
+      "loss": 2.6555,
+      "step": 55855
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7727231383323669,
+      "learning_rate": 1.1274663272739482e-05,
+      "loss": 2.8709,
+      "step": 55856
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7116361856460571,
+      "learning_rate": 1.127048246037211e-05,
+      "loss": 3.0165,
+      "step": 55857
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7671318054199219,
+      "learning_rate": 1.1266302408462734e-05,
+      "loss": 2.8943,
+      "step": 55858
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7194930911064148,
+      "learning_rate": 1.1262123117022382e-05,
+      "loss": 2.8822,
+      "step": 55859
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7566820979118347,
+      "learning_rate": 1.1257944586061974e-05,
+      "loss": 2.7942,
+      "step": 55860
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.717607855796814,
+      "learning_rate": 1.1253766815592669e-05,
+      "loss": 2.8937,
+      "step": 55861
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7694738507270813,
+      "learning_rate": 1.1249589805625292e-05,
+      "loss": 2.8219,
+      "step": 55862
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.729606032371521,
+      "learning_rate": 1.1245413556170901e-05,
+      "loss": 2.5149,
+      "step": 55863
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7487224340438843,
+      "learning_rate": 1.1241238067240588e-05,
+      "loss": 2.6698,
+      "step": 55864
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7842611074447632,
+      "learning_rate": 1.1237063338845209e-05,
+      "loss": 2.8821,
+      "step": 55865
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7518981695175171,
+      "learning_rate": 1.123288937099589e-05,
+      "loss": 2.8636,
+      "step": 55866
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7385653257369995,
+      "learning_rate": 1.1228716163703556e-05,
+      "loss": 2.8735,
+      "step": 55867
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7072750329971313,
+      "learning_rate": 1.1224543716979195e-05,
+      "loss": 2.8221,
+      "step": 55868
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7266203165054321,
+      "learning_rate": 1.1220372030833836e-05,
+      "loss": 3.007,
+      "step": 55869
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7844318151473999,
+      "learning_rate": 1.12162011052784e-05,
+      "loss": 2.8807,
+      "step": 55870
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.866727352142334,
+      "learning_rate": 1.121203094032398e-05,
+      "loss": 2.8678,
+      "step": 55871
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7634037137031555,
+      "learning_rate": 1.12078615359814e-05,
+      "loss": 2.876,
+      "step": 55872
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.72465580701828,
+      "learning_rate": 1.1203692892261785e-05,
+      "loss": 2.9704,
+      "step": 55873
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7288641929626465,
+      "learning_rate": 1.1199525009176059e-05,
+      "loss": 2.9168,
+      "step": 55874
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7715079188346863,
+      "learning_rate": 1.1195357886735213e-05,
+      "loss": 2.7081,
+      "step": 55875
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7418809533119202,
+      "learning_rate": 1.1191191524950239e-05,
+      "loss": 2.8808,
+      "step": 55876
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7965562343597412,
+      "learning_rate": 1.1187025923832027e-05,
+      "loss": 3.0647,
+      "step": 55877
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6859378218650818,
+      "learning_rate": 1.1182861083391636e-05,
+      "loss": 2.8672,
+      "step": 55878
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7590106129646301,
+      "learning_rate": 1.1178697003639958e-05,
+      "loss": 2.8654,
+      "step": 55879
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7052371501922607,
+      "learning_rate": 1.117453368458805e-05,
+      "loss": 2.7231,
+      "step": 55880
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7554574608802795,
+      "learning_rate": 1.1170371126246835e-05,
+      "loss": 2.8292,
+      "step": 55881
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7286297678947449,
+      "learning_rate": 1.116620932862724e-05,
+      "loss": 3.073,
+      "step": 55882
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7669945359230042,
+      "learning_rate": 1.1162048291740321e-05,
+      "loss": 3.0026,
+      "step": 55883
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7861215472221375,
+      "learning_rate": 1.1157888015596972e-05,
+      "loss": 2.9076,
+      "step": 55884
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7036914825439453,
+      "learning_rate": 1.1153728500208115e-05,
+      "loss": 2.9243,
+      "step": 55885
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7453054785728455,
+      "learning_rate": 1.1149569745584774e-05,
+      "loss": 2.9301,
+      "step": 55886
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7406743764877319,
+      "learning_rate": 1.1145411751737843e-05,
+      "loss": 3.0202,
+      "step": 55887
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7263259291648865,
+      "learning_rate": 1.1141254518678344e-05,
+      "loss": 3.0923,
+      "step": 55888
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7786001563072205,
+      "learning_rate": 1.113709804641717e-05,
+      "loss": 3.0409,
+      "step": 55889
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.757474958896637,
+      "learning_rate": 1.1132942334965311e-05,
+      "loss": 2.7983,
+      "step": 55890
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7858365774154663,
+      "learning_rate": 1.1128787384333692e-05,
+      "loss": 3.0745,
+      "step": 55891
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7730826735496521,
+      "learning_rate": 1.1124633194533206e-05,
+      "loss": 2.7281,
+      "step": 55892
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7334781885147095,
+      "learning_rate": 1.1120479765574875e-05,
+      "loss": 2.8114,
+      "step": 55893
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7341852784156799,
+      "learning_rate": 1.111632709746959e-05,
+      "loss": 2.957,
+      "step": 55894
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7085254788398743,
+      "learning_rate": 1.1112175190228279e-05,
+      "loss": 2.9096,
+      "step": 55895
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7142402529716492,
+      "learning_rate": 1.1108024043861963e-05,
+      "loss": 3.1765,
+      "step": 55896
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7631176710128784,
+      "learning_rate": 1.1103873658381502e-05,
+      "loss": 2.923,
+      "step": 55897
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8920885920524597,
+      "learning_rate": 1.109972403379782e-05,
+      "loss": 3.0669,
+      "step": 55898
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7416764497756958,
+      "learning_rate": 1.109557517012184e-05,
+      "loss": 2.7871,
+      "step": 55899
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8150529265403748,
+      "learning_rate": 1.1091427067364523e-05,
+      "loss": 2.9976,
+      "step": 55900
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7230525612831116,
+      "learning_rate": 1.1087279725536824e-05,
+      "loss": 2.9518,
+      "step": 55901
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.762174665927887,
+      "learning_rate": 1.1083133144649569e-05,
+      "loss": 3.2105,
+      "step": 55902
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7683262228965759,
+      "learning_rate": 1.1078987324713817e-05,
+      "loss": 2.7405,
+      "step": 55903
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7210031151771545,
+      "learning_rate": 1.1074842265740325e-05,
+      "loss": 3.0056,
+      "step": 55904
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7431241273880005,
+      "learning_rate": 1.1070697967740149e-05,
+      "loss": 2.7492,
+      "step": 55905
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7543492317199707,
+      "learning_rate": 1.106655443072415e-05,
+      "loss": 2.8373,
+      "step": 55906
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7546671628952026,
+      "learning_rate": 1.1062411654703218e-05,
+      "loss": 2.8681,
+      "step": 55907
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7264155745506287,
+      "learning_rate": 1.105826963968831e-05,
+      "loss": 2.7632,
+      "step": 55908
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7544650435447693,
+      "learning_rate": 1.1054128385690254e-05,
+      "loss": 2.8906,
+      "step": 55909
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7051803469657898,
+      "learning_rate": 1.1049987892720036e-05,
+      "loss": 2.8865,
+      "step": 55910
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7877517938613892,
+      "learning_rate": 1.1045848160788585e-05,
+      "loss": 2.9269,
+      "step": 55911
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7090002298355103,
+      "learning_rate": 1.1041709189906756e-05,
+      "loss": 3.0367,
+      "step": 55912
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7225410342216492,
+      "learning_rate": 1.1037570980085442e-05,
+      "loss": 2.7446,
+      "step": 55913
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7321907877922058,
+      "learning_rate": 1.1033433531335533e-05,
+      "loss": 2.8573,
+      "step": 55914
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7515150904655457,
+      "learning_rate": 1.1029296843667923e-05,
+      "loss": 2.9509,
+      "step": 55915
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7628140449523926,
+      "learning_rate": 1.10251609170936e-05,
+      "loss": 3.1318,
+      "step": 55916
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7710598111152649,
+      "learning_rate": 1.1021025751623324e-05,
+      "loss": 2.8181,
+      "step": 55917
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7168538570404053,
+      "learning_rate": 1.1016891347268087e-05,
+      "loss": 2.9143,
+      "step": 55918
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7750139236450195,
+      "learning_rate": 1.1012757704038778e-05,
+      "loss": 2.9098,
+      "step": 55919
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8607103824615479,
+      "learning_rate": 1.1008624821946188e-05,
+      "loss": 2.8608,
+      "step": 55920
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7440026998519897,
+      "learning_rate": 1.1004492701001278e-05,
+      "loss": 2.8417,
+      "step": 55921
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7108123898506165,
+      "learning_rate": 1.1000361341214902e-05,
+      "loss": 2.9667,
+      "step": 55922
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7313125729560852,
+      "learning_rate": 1.099623074259799e-05,
+      "loss": 2.6806,
+      "step": 55923
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8081473708152771,
+      "learning_rate": 1.099210090516136e-05,
+      "loss": 2.704,
+      "step": 55924
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7507513761520386,
+      "learning_rate": 1.098797182891591e-05,
+      "loss": 2.9803,
+      "step": 55925
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.770849347114563,
+      "learning_rate": 1.0983843513872559e-05,
+      "loss": 3.0487,
+      "step": 55926
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7244208455085754,
+      "learning_rate": 1.0979715960042135e-05,
+      "loss": 2.879,
+      "step": 55927
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7546032667160034,
+      "learning_rate": 1.0975589167435527e-05,
+      "loss": 3.0663,
+      "step": 55928
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8063235878944397,
+      "learning_rate": 1.097146313606353e-05,
+      "loss": 2.9658,
+      "step": 55929
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.714979887008667,
+      "learning_rate": 1.0967337865937098e-05,
+      "loss": 2.9623,
+      "step": 55930
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7428481578826904,
+      "learning_rate": 1.096321335706709e-05,
+      "loss": 2.9092,
+      "step": 55931
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7758767008781433,
+      "learning_rate": 1.0959089609464334e-05,
+      "loss": 2.7544,
+      "step": 55932
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7635376453399658,
+      "learning_rate": 1.0954966623139748e-05,
+      "loss": 2.9114,
+      "step": 55933
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7332720160484314,
+      "learning_rate": 1.095084439810413e-05,
+      "loss": 2.836,
+      "step": 55934
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7916608452796936,
+      "learning_rate": 1.0946722934368334e-05,
+      "loss": 2.814,
+      "step": 55935
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7521407008171082,
+      "learning_rate": 1.0942602231943287e-05,
+      "loss": 2.8053,
+      "step": 55936
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8015977144241333,
+      "learning_rate": 1.0938482290839745e-05,
+      "loss": 2.9304,
+      "step": 55937
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7351771593093872,
+      "learning_rate": 1.0934363111068668e-05,
+      "loss": 2.9438,
+      "step": 55938
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7315554022789001,
+      "learning_rate": 1.0930244692640777e-05,
+      "loss": 2.8405,
+      "step": 55939
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7486281394958496,
+      "learning_rate": 1.0926127035567067e-05,
+      "loss": 2.9316,
+      "step": 55940
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.721809446811676,
+      "learning_rate": 1.0922010139858262e-05,
+      "loss": 2.9926,
+      "step": 55941
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7924941778182983,
+      "learning_rate": 1.0917894005525251e-05,
+      "loss": 2.8259,
+      "step": 55942
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7480098009109497,
+      "learning_rate": 1.0913778632578862e-05,
+      "loss": 2.9488,
+      "step": 55943
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7476303577423096,
+      "learning_rate": 1.090966402102995e-05,
+      "loss": 2.8346,
+      "step": 55944
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7705007195472717,
+      "learning_rate": 1.0905550170889344e-05,
+      "loss": 2.9332,
+      "step": 55945
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.745185375213623,
+      "learning_rate": 1.0901437082167896e-05,
+      "loss": 2.8909,
+      "step": 55946
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7382268905639648,
+      "learning_rate": 1.0897324754876402e-05,
+      "loss": 2.8413,
+      "step": 55947
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8054601550102234,
+      "learning_rate": 1.0893213189025751e-05,
+      "loss": 2.7561,
+      "step": 55948
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7425912022590637,
+      "learning_rate": 1.0889102384626702e-05,
+      "loss": 2.9528,
+      "step": 55949
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7311367988586426,
+      "learning_rate": 1.0884992341690113e-05,
+      "loss": 2.9087,
+      "step": 55950
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.752752959728241,
+      "learning_rate": 1.0880883060226875e-05,
+      "loss": 3.0692,
+      "step": 55951
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.74880051612854,
+      "learning_rate": 1.087677454024768e-05,
+      "loss": 2.8871,
+      "step": 55952
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.776472806930542,
+      "learning_rate": 1.087266678176345e-05,
+      "loss": 2.823,
+      "step": 55953
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7268320322036743,
+      "learning_rate": 1.0868559784784947e-05,
+      "loss": 2.6937,
+      "step": 55954
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7472983598709106,
+      "learning_rate": 1.086445354932306e-05,
+      "loss": 3.2036,
+      "step": 55955
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7563039064407349,
+      "learning_rate": 1.0860348075388514e-05,
+      "loss": 2.9643,
+      "step": 55956
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.72945237159729,
+      "learning_rate": 1.0856243362992167e-05,
+      "loss": 2.9227,
+      "step": 55957
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7762035727500916,
+      "learning_rate": 1.0852139412144877e-05,
+      "loss": 2.5892,
+      "step": 55958
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7469337582588196,
+      "learning_rate": 1.0848036222857337e-05,
+      "loss": 3.0108,
+      "step": 55959
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6988431215286255,
+      "learning_rate": 1.084393379514047e-05,
+      "loss": 3.0516,
+      "step": 55960
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7568329572677612,
+      "learning_rate": 1.0839832129005e-05,
+      "loss": 2.9372,
+      "step": 55961
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7815048098564148,
+      "learning_rate": 1.0835731224461786e-05,
+      "loss": 2.7753,
+      "step": 55962
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7493932843208313,
+      "learning_rate": 1.0831631081521618e-05,
+      "loss": 2.9575,
+      "step": 55963
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7615856528282166,
+      "learning_rate": 1.0827531700195224e-05,
+      "loss": 2.6857,
+      "step": 55964
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.889343798160553,
+      "learning_rate": 1.0823433080493526e-05,
+      "loss": 2.6804,
+      "step": 55965
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7530947923660278,
+      "learning_rate": 1.0819335222427183e-05,
+      "loss": 2.8449,
+      "step": 55966
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7794963121414185,
+      "learning_rate": 1.0815238126007086e-05,
+      "loss": 2.8913,
+      "step": 55967
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7236289381980896,
+      "learning_rate": 1.0811141791244026e-05,
+      "loss": 2.8246,
+      "step": 55968
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7335292100906372,
+      "learning_rate": 1.0807046218148729e-05,
+      "loss": 2.9881,
+      "step": 55969
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7387674450874329,
+      "learning_rate": 1.0802951406732018e-05,
+      "loss": 2.9562,
+      "step": 55970
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.754050076007843,
+      "learning_rate": 1.0798857357004719e-05,
+      "loss": 2.7057,
+      "step": 55971
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7682558298110962,
+      "learning_rate": 1.0794764068977524e-05,
+      "loss": 2.7377,
+      "step": 55972
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7760759592056274,
+      "learning_rate": 1.0790671542661289e-05,
+      "loss": 2.8588,
+      "step": 55973
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.74440997838974,
+      "learning_rate": 1.0786579778066739e-05,
+      "loss": 2.9655,
+      "step": 55974
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7802349925041199,
+      "learning_rate": 1.07824887752047e-05,
+      "loss": 2.9066,
+      "step": 55975
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7338876724243164,
+      "learning_rate": 1.0778398534085898e-05,
+      "loss": 3.0789,
+      "step": 55976
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7808542251586914,
+      "learning_rate": 1.0774309054721187e-05,
+      "loss": 2.7526,
+      "step": 55977
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7372686862945557,
+      "learning_rate": 1.077022033712126e-05,
+      "loss": 2.9052,
+      "step": 55978
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8141536116600037,
+      "learning_rate": 1.0766132381296876e-05,
+      "loss": 2.8219,
+      "step": 55979
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7406730651855469,
+      "learning_rate": 1.076204518725886e-05,
+      "loss": 2.9798,
+      "step": 55980
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7653099298477173,
+      "learning_rate": 1.0757958755017937e-05,
+      "loss": 2.756,
+      "step": 55981
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7098361849784851,
+      "learning_rate": 1.0753873084584896e-05,
+      "loss": 2.767,
+      "step": 55982
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7831395268440247,
+      "learning_rate": 1.0749788175970497e-05,
+      "loss": 2.6796,
+      "step": 55983
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.79066401720047,
+      "learning_rate": 1.0745704029185464e-05,
+      "loss": 2.7994,
+      "step": 55984
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7196535468101501,
+      "learning_rate": 1.0741620644240622e-05,
+      "loss": 2.7152,
+      "step": 55985
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8042973875999451,
+      "learning_rate": 1.0737538021146663e-05,
+      "loss": 2.9082,
+      "step": 55986
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8539407849311829,
+      "learning_rate": 1.073345615991431e-05,
+      "loss": 3.0031,
+      "step": 55987
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7349897027015686,
+      "learning_rate": 1.0729375060554424e-05,
+      "loss": 2.8356,
+      "step": 55988
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7442134022712708,
+      "learning_rate": 1.072529472307766e-05,
+      "loss": 2.9911,
+      "step": 55989
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.742074191570282,
+      "learning_rate": 1.072121514749481e-05,
+      "loss": 2.9064,
+      "step": 55990
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7211217880249023,
+      "learning_rate": 1.07171363338166e-05,
+      "loss": 2.8363,
+      "step": 55991
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7274338006973267,
+      "learning_rate": 1.0713058282053788e-05,
+      "loss": 2.9167,
+      "step": 55992
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7637966871261597,
+      "learning_rate": 1.0708980992217131e-05,
+      "loss": 3.002,
+      "step": 55993
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7385900616645813,
+      "learning_rate": 1.0704904464317287e-05,
+      "loss": 2.7543,
+      "step": 55994
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7418367862701416,
+      "learning_rate": 1.0700828698365117e-05,
+      "loss": 2.9536,
+      "step": 55995
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7280617356300354,
+      "learning_rate": 1.0696753694371207e-05,
+      "loss": 2.6508,
+      "step": 55996
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7581601738929749,
+      "learning_rate": 1.069267945234642e-05,
+      "loss": 3.0828,
+      "step": 55997
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7517957091331482,
+      "learning_rate": 1.0688605972301444e-05,
+      "loss": 2.8601,
+      "step": 55998
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7825131416320801,
+      "learning_rate": 1.0684533254246974e-05,
+      "loss": 2.9613,
+      "step": 55999
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7765049934387207,
+      "learning_rate": 1.0680461298193799e-05,
+      "loss": 2.908,
+      "step": 56000
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7475263476371765,
+      "learning_rate": 1.0676390104152577e-05,
+      "loss": 2.9421,
+      "step": 56001
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7581568360328674,
+      "learning_rate": 1.0672319672134033e-05,
+      "loss": 2.7795,
+      "step": 56002
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7449358105659485,
+      "learning_rate": 1.0668250002148993e-05,
+      "loss": 2.8471,
+      "step": 56003
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7291288375854492,
+      "learning_rate": 1.0664181094208012e-05,
+      "loss": 3.0978,
+      "step": 56004
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7702457308769226,
+      "learning_rate": 1.0660112948321953e-05,
+      "loss": 2.9775,
+      "step": 56005
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.738068699836731,
+      "learning_rate": 1.065604556450147e-05,
+      "loss": 2.9591,
+      "step": 56006
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7106137871742249,
+      "learning_rate": 1.0651978942757256e-05,
+      "loss": 2.7704,
+      "step": 56007
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7228193879127502,
+      "learning_rate": 1.0647913083100101e-05,
+      "loss": 2.8967,
+      "step": 56008
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.728057324886322,
+      "learning_rate": 1.0643847985540566e-05,
+      "loss": 2.8884,
+      "step": 56009
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8260751366615295,
+      "learning_rate": 1.0639783650089506e-05,
+      "loss": 3.0548,
+      "step": 56010
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7846874594688416,
+      "learning_rate": 1.0635720076757548e-05,
+      "loss": 2.8825,
+      "step": 56011
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7633529901504517,
+      "learning_rate": 1.0631657265555382e-05,
+      "loss": 2.8856,
+      "step": 56012
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7845501899719238,
+      "learning_rate": 1.0627595216493801e-05,
+      "loss": 2.9424,
+      "step": 56013
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7749620676040649,
+      "learning_rate": 1.0623533929583461e-05,
+      "loss": 2.994,
+      "step": 56014
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7692545056343079,
+      "learning_rate": 1.0619473404834988e-05,
+      "loss": 3.0687,
+      "step": 56015
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7629958391189575,
+      "learning_rate": 1.0615413642259141e-05,
+      "loss": 3.0601,
+      "step": 56016
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7727113962173462,
+      "learning_rate": 1.0611354641866576e-05,
+      "loss": 2.9892,
+      "step": 56017
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7810448408126831,
+      "learning_rate": 1.0607296403668052e-05,
+      "loss": 3.0302,
+      "step": 56018
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7546478509902954,
+      "learning_rate": 1.0603238927674197e-05,
+      "loss": 3.0782,
+      "step": 56019
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7549473643302917,
+      "learning_rate": 1.0599182213895729e-05,
+      "loss": 2.8536,
+      "step": 56020
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.712246298789978,
+      "learning_rate": 1.0595126262343313e-05,
+      "loss": 2.8927,
+      "step": 56021
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7353459000587463,
+      "learning_rate": 1.0591071073027601e-05,
+      "loss": 2.7289,
+      "step": 56022
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7388055920600891,
+      "learning_rate": 1.0587016645959356e-05,
+      "loss": 2.7817,
+      "step": 56023
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7375566959381104,
+      "learning_rate": 1.0582962981149167e-05,
+      "loss": 2.7332,
+      "step": 56024
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7719786167144775,
+      "learning_rate": 1.0578910078607794e-05,
+      "loss": 2.8114,
+      "step": 56025
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7156137824058533,
+      "learning_rate": 1.057485793834586e-05,
+      "loss": 2.8983,
+      "step": 56026
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7468981742858887,
+      "learning_rate": 1.0570806560374023e-05,
+      "loss": 2.8096,
+      "step": 56027
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7423037886619568,
+      "learning_rate": 1.0566755944703009e-05,
+      "loss": 2.9614,
+      "step": 56028
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7885465621948242,
+      "learning_rate": 1.0562706091343475e-05,
+      "loss": 2.8188,
+      "step": 56029
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7959953546524048,
+      "learning_rate": 1.055865700030608e-05,
+      "loss": 2.7377,
+      "step": 56030
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7130370736122131,
+      "learning_rate": 1.0554608671601417e-05,
+      "loss": 2.8816,
+      "step": 56031
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.724517822265625,
+      "learning_rate": 1.055056110524024e-05,
+      "loss": 2.898,
+      "step": 56032
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7943875789642334,
+      "learning_rate": 1.0546514301233178e-05,
+      "loss": 2.956,
+      "step": 56033
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7979520559310913,
+      "learning_rate": 1.0542468259590853e-05,
+      "loss": 2.7335,
+      "step": 56034
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7505634427070618,
+      "learning_rate": 1.0538422980324025e-05,
+      "loss": 2.8819,
+      "step": 56035
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.761986255645752,
+      "learning_rate": 1.0534378463443283e-05,
+      "loss": 2.7005,
+      "step": 56036
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7043732404708862,
+      "learning_rate": 1.0530334708959255e-05,
+      "loss": 2.8908,
+      "step": 56037
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7133161425590515,
+      "learning_rate": 1.0526291716882596e-05,
+      "loss": 2.9596,
+      "step": 56038
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7343462109565735,
+      "learning_rate": 1.0522249487223966e-05,
+      "loss": 2.8832,
+      "step": 56039
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7378050684928894,
+      "learning_rate": 1.0518208019994056e-05,
+      "loss": 2.9357,
+      "step": 56040
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7832443714141846,
+      "learning_rate": 1.0514167315203458e-05,
+      "loss": 3.0828,
+      "step": 56041
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.715339183807373,
+      "learning_rate": 1.0510127372862831e-05,
+      "loss": 2.8073,
+      "step": 56042
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7633505463600159,
+      "learning_rate": 1.0506088192982831e-05,
+      "loss": 2.9054,
+      "step": 56043
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7171920537948608,
+      "learning_rate": 1.050204977557405e-05,
+      "loss": 2.8434,
+      "step": 56044
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7434439063072205,
+      "learning_rate": 1.0498012120647215e-05,
+      "loss": 2.8091,
+      "step": 56045
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7520685791969299,
+      "learning_rate": 1.0493975228212814e-05,
+      "loss": 2.7975,
+      "step": 56046
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7988944053649902,
+      "learning_rate": 1.048993909828164e-05,
+      "loss": 2.6339,
+      "step": 56047
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7427051663398743,
+      "learning_rate": 1.048590373086422e-05,
+      "loss": 2.7419,
+      "step": 56048
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7187119126319885,
+      "learning_rate": 1.0481869125971176e-05,
+      "loss": 2.964,
+      "step": 56049
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7609080672264099,
+      "learning_rate": 1.0477835283613233e-05,
+      "loss": 2.6344,
+      "step": 56050
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7719628810882568,
+      "learning_rate": 1.0473802203800918e-05,
+      "loss": 2.8794,
+      "step": 56051
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.807428240776062,
+      "learning_rate": 1.0469769886544921e-05,
+      "loss": 2.9451,
+      "step": 56052
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7161225080490112,
+      "learning_rate": 1.0465738331855767e-05,
+      "loss": 2.7363,
+      "step": 56053
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7533316612243652,
+      "learning_rate": 1.0461707539744179e-05,
+      "loss": 2.9044,
+      "step": 56054
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7733919620513916,
+      "learning_rate": 1.045767751022072e-05,
+      "loss": 3.0109,
+      "step": 56055
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7934540510177612,
+      "learning_rate": 1.0453648243296008e-05,
+      "loss": 2.6885,
+      "step": 56056
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.75040203332901,
+      "learning_rate": 1.0449619738980675e-05,
+      "loss": 3.0411,
+      "step": 56057
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7220397591590881,
+      "learning_rate": 1.0445591997285341e-05,
+      "loss": 2.8238,
+      "step": 56058
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7255467772483826,
+      "learning_rate": 1.0441565018220532e-05,
+      "loss": 3.0506,
+      "step": 56059
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7547920346260071,
+      "learning_rate": 1.0437538801796973e-05,
+      "loss": 3.0475,
+      "step": 56060
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7408540844917297,
+      "learning_rate": 1.0433513348025158e-05,
+      "loss": 2.8226,
+      "step": 56061
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7349470853805542,
+      "learning_rate": 1.0429488656915807e-05,
+      "loss": 2.9271,
+      "step": 56062
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7243402600288391,
+      "learning_rate": 1.0425464728479382e-05,
+      "loss": 2.8937,
+      "step": 56063
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7938183546066284,
+      "learning_rate": 1.0421441562726607e-05,
+      "loss": 2.8097,
+      "step": 56064
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7476469278335571,
+      "learning_rate": 1.0417419159668006e-05,
+      "loss": 2.5328,
+      "step": 56065
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7038164734840393,
+      "learning_rate": 1.041339751931417e-05,
+      "loss": 2.7658,
+      "step": 56066
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6968178749084473,
+      "learning_rate": 1.0409376641675759e-05,
+      "loss": 2.946,
+      "step": 56067
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7435596585273743,
+      "learning_rate": 1.0405356526763264e-05,
+      "loss": 2.8976,
+      "step": 56068
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7599115371704102,
+      "learning_rate": 1.0401337174587344e-05,
+      "loss": 2.6582,
+      "step": 56069
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7440100312232971,
+      "learning_rate": 1.0397318585158587e-05,
+      "loss": 2.7613,
+      "step": 56070
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7963297963142395,
+      "learning_rate": 1.0393300758487522e-05,
+      "loss": 2.7282,
+      "step": 56071
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.746436357498169,
+      "learning_rate": 1.0389283694584805e-05,
+      "loss": 2.777,
+      "step": 56072
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7619957327842712,
+      "learning_rate": 1.0385267393460994e-05,
+      "loss": 2.6955,
+      "step": 56073
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7717376947402954,
+      "learning_rate": 1.0381251855126582e-05,
+      "loss": 3.0108,
+      "step": 56074
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7113671898841858,
+      "learning_rate": 1.037723707959226e-05,
+      "loss": 2.7965,
+      "step": 56075
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7118061184883118,
+      "learning_rate": 1.0373223066868552e-05,
+      "loss": 2.8349,
+      "step": 56076
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7409781217575073,
+      "learning_rate": 1.0369209816966051e-05,
+      "loss": 2.7353,
+      "step": 56077
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7771452069282532,
+      "learning_rate": 1.0365197329895281e-05,
+      "loss": 2.8024,
+      "step": 56078
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.787254810333252,
+      "learning_rate": 1.0361185605666866e-05,
+      "loss": 2.937,
+      "step": 56079
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7171097993850708,
+      "learning_rate": 1.0357174644291332e-05,
+      "loss": 2.9785,
+      "step": 56080
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7196240425109863,
+      "learning_rate": 1.0353164445779238e-05,
+      "loss": 2.988,
+      "step": 56081
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.718910813331604,
+      "learning_rate": 1.0349155010141208e-05,
+      "loss": 2.6988,
+      "step": 56082
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7512689828872681,
+      "learning_rate": 1.03451463373877e-05,
+      "loss": 3.0488,
+      "step": 56083
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7203948497772217,
+      "learning_rate": 1.0341138427529339e-05,
+      "loss": 2.8411,
+      "step": 56084
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7465165257453918,
+      "learning_rate": 1.0337131280576716e-05,
+      "loss": 2.9288,
+      "step": 56085
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7512971758842468,
+      "learning_rate": 1.0333124896540324e-05,
+      "loss": 2.981,
+      "step": 56086
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7287713885307312,
+      "learning_rate": 1.0329119275430785e-05,
+      "loss": 2.9989,
+      "step": 56087
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7980208992958069,
+      "learning_rate": 1.0325114417258529e-05,
+      "loss": 2.9048,
+      "step": 56088
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6989551782608032,
+      "learning_rate": 1.0321110322034175e-05,
+      "loss": 2.8285,
+      "step": 56089
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7511205077171326,
+      "learning_rate": 1.0317106989768287e-05,
+      "loss": 2.8994,
+      "step": 56090
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7218790650367737,
+      "learning_rate": 1.0313104420471386e-05,
+      "loss": 2.7294,
+      "step": 56091
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7532414197921753,
+      "learning_rate": 1.030910261415403e-05,
+      "loss": 2.9913,
+      "step": 56092
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7547132968902588,
+      "learning_rate": 1.0305101570826713e-05,
+      "loss": 2.8585,
+      "step": 56093
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7735399603843689,
+      "learning_rate": 1.0301101290500025e-05,
+      "loss": 3.0481,
+      "step": 56094
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.728217363357544,
+      "learning_rate": 1.0297101773184523e-05,
+      "loss": 2.887,
+      "step": 56095
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7460715174674988,
+      "learning_rate": 1.0293103018890637e-05,
+      "loss": 2.7163,
+      "step": 56096
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7449331283569336,
+      "learning_rate": 1.0289105027628985e-05,
+      "loss": 2.903,
+      "step": 56097
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.738858699798584,
+      "learning_rate": 1.0285107799410063e-05,
+      "loss": 2.7658,
+      "step": 56098
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7564826011657715,
+      "learning_rate": 1.0281111334244396e-05,
+      "loss": 2.8346,
+      "step": 56099
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7498881816864014,
+      "learning_rate": 1.0277115632142574e-05,
+      "loss": 2.8828,
+      "step": 56100
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7890697717666626,
+      "learning_rate": 1.0273120693115056e-05,
+      "loss": 2.8409,
+      "step": 56101
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7382162809371948,
+      "learning_rate": 1.02691265171724e-05,
+      "loss": 2.915,
+      "step": 56102
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7483049035072327,
+      "learning_rate": 1.0265133104325063e-05,
+      "loss": 2.6881,
+      "step": 56103
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7414224147796631,
+      "learning_rate": 1.0261140454583572e-05,
+      "loss": 2.8794,
+      "step": 56104
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7294985055923462,
+      "learning_rate": 1.0257148567958551e-05,
+      "loss": 2.8673,
+      "step": 56105
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7478199005126953,
+      "learning_rate": 1.0253157444460391e-05,
+      "loss": 3.0047,
+      "step": 56106
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7180767059326172,
+      "learning_rate": 1.0249167084099685e-05,
+      "loss": 2.7526,
+      "step": 56107
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7234374284744263,
+      "learning_rate": 1.0245177486886858e-05,
+      "loss": 3.1474,
+      "step": 56108
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7626802325248718,
+      "learning_rate": 1.0241188652832533e-05,
+      "loss": 2.8545,
+      "step": 56109
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7502772212028503,
+      "learning_rate": 1.0237200581947136e-05,
+      "loss": 2.9798,
+      "step": 56110
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7730661034584045,
+      "learning_rate": 1.0233213274241158e-05,
+      "loss": 3.0912,
+      "step": 56111
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7321888208389282,
+      "learning_rate": 1.022922672972516e-05,
+      "loss": 2.8724,
+      "step": 56112
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7915605902671814,
+      "learning_rate": 1.0225240948409563e-05,
+      "loss": 2.7356,
+      "step": 56113
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7326378226280212,
+      "learning_rate": 1.0221255930304961e-05,
+      "loss": 2.7905,
+      "step": 56114
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7629287838935852,
+      "learning_rate": 1.0217271675421779e-05,
+      "loss": 2.915,
+      "step": 56115
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.721251904964447,
+      "learning_rate": 1.0213288183770574e-05,
+      "loss": 2.9442,
+      "step": 56116
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8092251420021057,
+      "learning_rate": 1.0209305455361771e-05,
+      "loss": 2.9196,
+      "step": 56117
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7478073835372925,
+      "learning_rate": 1.0205323490205863e-05,
+      "loss": 2.8916,
+      "step": 56118
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7610535621643066,
+      "learning_rate": 1.0201342288313374e-05,
+      "loss": 2.9998,
+      "step": 56119
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7558162212371826,
+      "learning_rate": 1.0197361849694763e-05,
+      "loss": 2.7903,
+      "step": 56120
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7537033557891846,
+      "learning_rate": 1.0193382174360554e-05,
+      "loss": 2.9119,
+      "step": 56121
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7669283747673035,
+      "learning_rate": 1.0189403262321205e-05,
+      "loss": 2.9549,
+      "step": 56122
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7572535872459412,
+      "learning_rate": 1.0185425113587175e-05,
+      "loss": 3.1026,
+      "step": 56123
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7835591435432434,
+      "learning_rate": 1.018144772816899e-05,
+      "loss": 2.9209,
+      "step": 56124
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8195697665214539,
+      "learning_rate": 1.017747110607704e-05,
+      "loss": 2.9498,
+      "step": 56125
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7439544796943665,
+      "learning_rate": 1.0173495247321884e-05,
+      "loss": 2.8571,
+      "step": 56126
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7486630082130432,
+      "learning_rate": 1.016952015191398e-05,
+      "loss": 2.938,
+      "step": 56127
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7225660085678101,
+      "learning_rate": 1.0165545819863751e-05,
+      "loss": 2.8997,
+      "step": 56128
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7393267750740051,
+      "learning_rate": 1.0161572251181692e-05,
+      "loss": 2.9516,
+      "step": 56129
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7707771062850952,
+      "learning_rate": 1.0157599445878296e-05,
+      "loss": 2.9941,
+      "step": 56130
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8051019906997681,
+      "learning_rate": 1.0153627403963983e-05,
+      "loss": 2.863,
+      "step": 56131
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7474145293235779,
+      "learning_rate": 1.014965612544928e-05,
+      "loss": 2.6743,
+      "step": 56132
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7893514633178711,
+      "learning_rate": 1.0145685610344544e-05,
+      "loss": 2.8199,
+      "step": 56133
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7323321104049683,
+      "learning_rate": 1.0141715858660305e-05,
+      "loss": 2.9556,
+      "step": 56134
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7719254493713379,
+      "learning_rate": 1.0137746870406983e-05,
+      "loss": 2.7592,
+      "step": 56135
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7276996970176697,
+      "learning_rate": 1.013377864559507e-05,
+      "loss": 2.6894,
+      "step": 56136
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.750486433506012,
+      "learning_rate": 1.0129811184234993e-05,
+      "loss": 2.9752,
+      "step": 56137
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7667336463928223,
+      "learning_rate": 1.0125844486337242e-05,
+      "loss": 2.9073,
+      "step": 56138
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.74593186378479,
+      "learning_rate": 1.012187855191221e-05,
+      "loss": 2.8232,
+      "step": 56139
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7445417046546936,
+      "learning_rate": 1.0117913380970355e-05,
+      "loss": 2.8888,
+      "step": 56140
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7170975208282471,
+      "learning_rate": 1.0113948973522101e-05,
+      "loss": 3.0145,
+      "step": 56141
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7595486640930176,
+      "learning_rate": 1.010998532957794e-05,
+      "loss": 2.8864,
+      "step": 56142
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7461397051811218,
+      "learning_rate": 1.0106022449148266e-05,
+      "loss": 2.9398,
+      "step": 56143
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7885101437568665,
+      "learning_rate": 1.01020603322436e-05,
+      "loss": 2.6727,
+      "step": 56144
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7303288578987122,
+      "learning_rate": 1.009809897887427e-05,
+      "loss": 2.9738,
+      "step": 56145
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7623375654220581,
+      "learning_rate": 1.0094138389050733e-05,
+      "loss": 3.0115,
+      "step": 56146
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.9273803234100342,
+      "learning_rate": 1.009017856278348e-05,
+      "loss": 2.7699,
+      "step": 56147
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7486945390701294,
+      "learning_rate": 1.008621950008287e-05,
+      "loss": 3.0087,
+      "step": 56148
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7255769968032837,
+      "learning_rate": 1.0082261200959397e-05,
+      "loss": 2.9543,
+      "step": 56149
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7860700488090515,
+      "learning_rate": 1.0078303665423448e-05,
+      "loss": 2.8486,
+      "step": 56150
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7196409106254578,
+      "learning_rate": 1.0074346893485418e-05,
+      "loss": 2.7253,
+      "step": 56151
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7964147329330444,
+      "learning_rate": 1.00703908851558e-05,
+      "loss": 2.9744,
+      "step": 56152
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7458089590072632,
+      "learning_rate": 1.0066435640444947e-05,
+      "loss": 2.7989,
+      "step": 56153
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7688996195793152,
+      "learning_rate": 1.0062481159363356e-05,
+      "loss": 3.0695,
+      "step": 56154
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7817563414573669,
+      "learning_rate": 1.0058527441921316e-05,
+      "loss": 2.6904,
+      "step": 56155
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6953161358833313,
+      "learning_rate": 1.0054574488129318e-05,
+      "loss": 2.7915,
+      "step": 56156
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7510791420936584,
+      "learning_rate": 1.0050622297997824e-05,
+      "loss": 2.8286,
+      "step": 56157
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7296958565711975,
+      "learning_rate": 1.0046670871537154e-05,
+      "loss": 3.0983,
+      "step": 56158
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7343621850013733,
+      "learning_rate": 1.0042720208757771e-05,
+      "loss": 2.971,
+      "step": 56159
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7544857263565063,
+      "learning_rate": 1.0038770309670064e-05,
+      "loss": 2.8618,
+      "step": 56160
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7312347292900085,
+      "learning_rate": 1.0034821174284391e-05,
+      "loss": 2.8683,
+      "step": 56161
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.723470151424408,
+      "learning_rate": 1.0030872802611245e-05,
+      "loss": 3.1332,
+      "step": 56162
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7472372055053711,
+      "learning_rate": 1.0026925194660918e-05,
+      "loss": 2.6761,
+      "step": 56163
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7491852641105652,
+      "learning_rate": 1.00229783504439e-05,
+      "loss": 2.791,
+      "step": 56164
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7424899339675903,
+      "learning_rate": 1.001903226997055e-05,
+      "loss": 2.9423,
+      "step": 56165
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7745535373687744,
+      "learning_rate": 1.0015086953251262e-05,
+      "loss": 2.9281,
+      "step": 56166
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8208742141723633,
+      "learning_rate": 1.0011142400296424e-05,
+      "loss": 2.8546,
+      "step": 56167
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8146986365318298,
+      "learning_rate": 1.0007198611116397e-05,
+      "loss": 2.9061,
+      "step": 56168
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7721244096755981,
+      "learning_rate": 1.0003255585721636e-05,
+      "loss": 3.1029,
+      "step": 56169
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.709169328212738,
+      "learning_rate": 9.999313324122471e-06,
+      "loss": 2.8192,
+      "step": 56170
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7202749848365784,
+      "learning_rate": 9.99537182632929e-06,
+      "loss": 2.6902,
+      "step": 56171
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7772741913795471,
+      "learning_rate": 9.99143109235252e-06,
+      "loss": 3.0664,
+      "step": 56172
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7416417002677917,
+      "learning_rate": 9.987491122202484e-06,
+      "loss": 3.0866,
+      "step": 56173
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7932614088058472,
+      "learning_rate": 9.983551915889609e-06,
+      "loss": 2.6837,
+      "step": 56174
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7134155035018921,
+      "learning_rate": 9.97961347342422e-06,
+      "loss": 2.6047,
+      "step": 56175
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.736059308052063,
+      "learning_rate": 9.975675794816707e-06,
+      "loss": 2.8903,
+      "step": 56176
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7584584951400757,
+      "learning_rate": 9.971738880077495e-06,
+      "loss": 2.9494,
+      "step": 56177
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6984370946884155,
+      "learning_rate": 9.967802729216844e-06,
+      "loss": 2.7811,
+      "step": 56178
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7537134885787964,
+      "learning_rate": 9.963867342245213e-06,
+      "loss": 2.9652,
+      "step": 56179
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7283973693847656,
+      "learning_rate": 9.959932719172924e-06,
+      "loss": 2.8149,
+      "step": 56180
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7228909730911255,
+      "learning_rate": 9.955998860010372e-06,
+      "loss": 3.0026,
+      "step": 56181
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8229911923408508,
+      "learning_rate": 9.952065764767913e-06,
+      "loss": 2.7358,
+      "step": 56182
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.768810510635376,
+      "learning_rate": 9.94813343345584e-06,
+      "loss": 2.9345,
+      "step": 56183
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7733370065689087,
+      "learning_rate": 9.944201866084612e-06,
+      "loss": 2.9988,
+      "step": 56184
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7900514602661133,
+      "learning_rate": 9.940271062664484e-06,
+      "loss": 2.9167,
+      "step": 56185
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.797058641910553,
+      "learning_rate": 9.936341023205918e-06,
+      "loss": 2.7364,
+      "step": 56186
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.766906201839447,
+      "learning_rate": 9.932411747719171e-06,
+      "loss": 2.899,
+      "step": 56187
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7349963784217834,
+      "learning_rate": 9.928483236214634e-06,
+      "loss": 2.7932,
+      "step": 56188
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.808515191078186,
+      "learning_rate": 9.924555488702667e-06,
+      "loss": 2.8549,
+      "step": 56189
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7727739810943604,
+      "learning_rate": 9.920628505193529e-06,
+      "loss": 2.8072,
+      "step": 56190
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.761241614818573,
+      "learning_rate": 9.916702285697708e-06,
+      "loss": 2.6368,
+      "step": 56191
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7611913681030273,
+      "learning_rate": 9.9127768302254e-06,
+      "loss": 2.9305,
+      "step": 56192
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7446513772010803,
+      "learning_rate": 9.908852138787027e-06,
+      "loss": 2.8921,
+      "step": 56193
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7649831771850586,
+      "learning_rate": 9.904928211392915e-06,
+      "loss": 2.7833,
+      "step": 56194
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7408809065818787,
+      "learning_rate": 9.90100504805339e-06,
+      "loss": 2.9076,
+      "step": 56195
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7453528642654419,
+      "learning_rate": 9.897082648778809e-06,
+      "loss": 2.7996,
+      "step": 56196
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7997094392776489,
+      "learning_rate": 9.893161013579499e-06,
+      "loss": 2.9272,
+      "step": 56197
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7552264928817749,
+      "learning_rate": 9.889240142465715e-06,
+      "loss": 2.9482,
+      "step": 56198
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7569772005081177,
+      "learning_rate": 9.885320035447853e-06,
+      "loss": 2.9521,
+      "step": 56199
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7403649091720581,
+      "learning_rate": 9.881400692536235e-06,
+      "loss": 2.9042,
+      "step": 56200
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7069177627563477,
+      "learning_rate": 9.877482113741186e-06,
+      "loss": 2.686,
+      "step": 56201
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7120211124420166,
+      "learning_rate": 9.873564299072968e-06,
+      "loss": 2.6873,
+      "step": 56202
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7966209650039673,
+      "learning_rate": 9.869647248542001e-06,
+      "loss": 2.9283,
+      "step": 56203
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6995888352394104,
+      "learning_rate": 9.865730962158546e-06,
+      "loss": 3.067,
+      "step": 56204
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7135257124900818,
+      "learning_rate": 9.861815439932864e-06,
+      "loss": 2.9348,
+      "step": 56205
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8033514022827148,
+      "learning_rate": 9.857900681875374e-06,
+      "loss": 2.8451,
+      "step": 56206
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8119677901268005,
+      "learning_rate": 9.853986687996273e-06,
+      "loss": 2.8265,
+      "step": 56207
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7657623291015625,
+      "learning_rate": 9.850073458305952e-06,
+      "loss": 2.9276,
+      "step": 56208
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7248706817626953,
+      "learning_rate": 9.846160992814734e-06,
+      "loss": 2.9206,
+      "step": 56209
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7554598450660706,
+      "learning_rate": 9.842249291532878e-06,
+      "loss": 2.9856,
+      "step": 56210
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.789419949054718,
+      "learning_rate": 9.838338354470675e-06,
+      "loss": 3.0536,
+      "step": 56211
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7719467282295227,
+      "learning_rate": 9.834428181638487e-06,
+      "loss": 2.8967,
+      "step": 56212
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7833728194236755,
+      "learning_rate": 9.830518773046536e-06,
+      "loss": 2.6544,
+      "step": 56213
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7524414658546448,
+      "learning_rate": 9.826610128705148e-06,
+      "loss": 2.7656,
+      "step": 56214
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7353940606117249,
+      "learning_rate": 9.822702248624614e-06,
+      "loss": 2.8325,
+      "step": 56215
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7165583372116089,
+      "learning_rate": 9.818795132815294e-06,
+      "loss": 2.8988,
+      "step": 56216
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.6943876147270203,
+      "learning_rate": 9.814888781287378e-06,
+      "loss": 2.8794,
+      "step": 56217
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7502477765083313,
+      "learning_rate": 9.810983194051193e-06,
+      "loss": 2.8726,
+      "step": 56218
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7636768817901611,
+      "learning_rate": 9.80707837111706e-06,
+      "loss": 2.9083,
+      "step": 56219
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7751126885414124,
+      "learning_rate": 9.80317431249521e-06,
+      "loss": 2.9759,
+      "step": 56220
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7540944218635559,
+      "learning_rate": 9.799271018195964e-06,
+      "loss": 2.8304,
+      "step": 56221
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7522205114364624,
+      "learning_rate": 9.795368488229583e-06,
+      "loss": 2.9295,
+      "step": 56222
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7293267846107483,
+      "learning_rate": 9.791466722606323e-06,
+      "loss": 3.1562,
+      "step": 56223
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7595233917236328,
+      "learning_rate": 9.78756572133651e-06,
+      "loss": 2.8134,
+      "step": 56224
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7647051215171814,
+      "learning_rate": 9.783665484430436e-06,
+      "loss": 2.8663,
+      "step": 56225
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7567137479782104,
+      "learning_rate": 9.779766011898293e-06,
+      "loss": 2.9385,
+      "step": 56226
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7509698271751404,
+      "learning_rate": 9.775867303750372e-06,
+      "loss": 3.0103,
+      "step": 56227
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7402850389480591,
+      "learning_rate": 9.771969359996967e-06,
+      "loss": 2.8077,
+      "step": 56228
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7876865267753601,
+      "learning_rate": 9.768072180648401e-06,
+      "loss": 2.8664,
+      "step": 56229
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7342726588249207,
+      "learning_rate": 9.7641757657148e-06,
+      "loss": 3.1046,
+      "step": 56230
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8054348826408386,
+      "learning_rate": 9.760280115206553e-06,
+      "loss": 3.0441,
+      "step": 56231
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7675466537475586,
+      "learning_rate": 9.756385229133823e-06,
+      "loss": 2.9284,
+      "step": 56232
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7613694071769714,
+      "learning_rate": 9.752491107506966e-06,
+      "loss": 2.923,
+      "step": 56233
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7603533267974854,
+      "learning_rate": 9.748597750336207e-06,
+      "loss": 2.7409,
+      "step": 56234
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.783909261226654,
+      "learning_rate": 9.744705157631705e-06,
+      "loss": 2.7609,
+      "step": 56235
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7282220721244812,
+      "learning_rate": 9.740813329403851e-06,
+      "loss": 2.8911,
+      "step": 56236
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7373564839363098,
+      "learning_rate": 9.736922265662805e-06,
+      "loss": 2.8638,
+      "step": 56237
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7997520565986633,
+      "learning_rate": 9.733031966418826e-06,
+      "loss": 2.9606,
+      "step": 56238
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7475751638412476,
+      "learning_rate": 9.729142431682201e-06,
+      "loss": 2.7857,
+      "step": 56239
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7757993936538696,
+      "learning_rate": 9.725253661463161e-06,
+      "loss": 2.6651,
+      "step": 56240
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7897785902023315,
+      "learning_rate": 9.721365655771962e-06,
+      "loss": 2.7296,
+      "step": 56241
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7303987145423889,
+      "learning_rate": 9.717478414618763e-06,
+      "loss": 2.9488,
+      "step": 56242
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7232341170310974,
+      "learning_rate": 9.713591938013854e-06,
+      "loss": 2.8354,
+      "step": 56243
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7750096321105957,
+      "learning_rate": 9.709706225967529e-06,
+      "loss": 3.0524,
+      "step": 56244
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7706706523895264,
+      "learning_rate": 9.705821278489912e-06,
+      "loss": 2.8424,
+      "step": 56245
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8001003265380859,
+      "learning_rate": 9.701937095591362e-06,
+      "loss": 2.8951,
+      "step": 56246
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7747387290000916,
+      "learning_rate": 9.698053677282037e-06,
+      "loss": 2.9802,
+      "step": 56247
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7447709441184998,
+      "learning_rate": 9.694171023572095e-06,
+      "loss": 2.7805,
+      "step": 56248
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7555181384086609,
+      "learning_rate": 9.690289134471896e-06,
+      "loss": 2.9242,
+      "step": 56249
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7425608038902283,
+      "learning_rate": 9.686408009991565e-06,
+      "loss": 2.7431,
+      "step": 56250
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7470229864120483,
+      "learning_rate": 9.682527650141426e-06,
+      "loss": 2.8323,
+      "step": 56251
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7438477873802185,
+      "learning_rate": 9.678648054931571e-06,
+      "loss": 2.9314,
+      "step": 56252
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7268569469451904,
+      "learning_rate": 9.674769224372292e-06,
+      "loss": 2.9639,
+      "step": 56253
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7507839798927307,
+      "learning_rate": 9.670891158473814e-06,
+      "loss": 2.7267,
+      "step": 56254
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7674872875213623,
+      "learning_rate": 9.667013857246363e-06,
+      "loss": 2.801,
+      "step": 56255
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7318063974380493,
+      "learning_rate": 9.663137320700098e-06,
+      "loss": 2.9449,
+      "step": 56256
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7555484771728516,
+      "learning_rate": 9.659261548845243e-06,
+      "loss": 2.8206,
+      "step": 56257
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8074846267700195,
+      "learning_rate": 9.655386541691989e-06,
+      "loss": 2.9654,
+      "step": 56258
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7630916833877563,
+      "learning_rate": 9.65151229925063e-06,
+      "loss": 2.931,
+      "step": 56259
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7579756379127502,
+      "learning_rate": 9.647638821531256e-06,
+      "loss": 3.0187,
+      "step": 56260
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7585728764533997,
+      "learning_rate": 9.64376610854416e-06,
+      "loss": 2.6944,
+      "step": 56261
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7346169352531433,
+      "learning_rate": 9.639894160299499e-06,
+      "loss": 2.5751,
+      "step": 56262
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7276310920715332,
+      "learning_rate": 9.636022976807467e-06,
+      "loss": 2.8546,
+      "step": 56263
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7806369066238403,
+      "learning_rate": 9.632152558078287e-06,
+      "loss": 2.9523,
+      "step": 56264
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7958837747573853,
+      "learning_rate": 9.628282904122087e-06,
+      "loss": 2.7377,
+      "step": 56265
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7492504715919495,
+      "learning_rate": 9.624414014949156e-06,
+      "loss": 3.0474,
+      "step": 56266
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7660962343215942,
+      "learning_rate": 9.620545890569587e-06,
+      "loss": 3.0414,
+      "step": 56267
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.775435745716095,
+      "learning_rate": 9.616678530993671e-06,
+      "loss": 3.2265,
+      "step": 56268
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7225391268730164,
+      "learning_rate": 9.6128119362315e-06,
+      "loss": 2.8305,
+      "step": 56269
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7277866005897522,
+      "learning_rate": 9.608946106293302e-06,
+      "loss": 2.715,
+      "step": 56270
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7221185564994812,
+      "learning_rate": 9.605081041189266e-06,
+      "loss": 2.7211,
+      "step": 56271
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7964680194854736,
+      "learning_rate": 9.60121674092955e-06,
+      "loss": 2.9876,
+      "step": 56272
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7531980276107788,
+      "learning_rate": 9.59735320552435e-06,
+      "loss": 2.7597,
+      "step": 56273
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.74334716796875,
+      "learning_rate": 9.593490434983787e-06,
+      "loss": 2.8664,
+      "step": 56274
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7619105577468872,
+      "learning_rate": 9.58962842931812e-06,
+      "loss": 2.6577,
+      "step": 56275
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8216807842254639,
+      "learning_rate": 9.585767188537474e-06,
+      "loss": 2.9393,
+      "step": 56276
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7856898903846741,
+      "learning_rate": 9.581906712652043e-06,
+      "loss": 2.8374,
+      "step": 56277
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7424861788749695,
+      "learning_rate": 9.578047001671984e-06,
+      "loss": 2.8822,
+      "step": 56278
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.758293092250824,
+      "learning_rate": 9.574188055607423e-06,
+      "loss": 2.972,
+      "step": 56279
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7339947819709778,
+      "learning_rate": 9.570329874468553e-06,
+      "loss": 3.113,
+      "step": 56280
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7537075281143188,
+      "learning_rate": 9.566472458265594e-06,
+      "loss": 2.7329,
+      "step": 56281
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7673822045326233,
+      "learning_rate": 9.562615807008577e-06,
+      "loss": 3.0186,
+      "step": 56282
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7142574787139893,
+      "learning_rate": 9.558759920707792e-06,
+      "loss": 2.6684,
+      "step": 56283
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7520612478256226,
+      "learning_rate": 9.55490479937333e-06,
+      "loss": 2.6922,
+      "step": 56284
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7886697053909302,
+      "learning_rate": 9.551050443015318e-06,
+      "loss": 2.95,
+      "step": 56285
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7925397157669067,
+      "learning_rate": 9.54719685164398e-06,
+      "loss": 2.953,
+      "step": 56286
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7573472261428833,
+      "learning_rate": 9.543344025269406e-06,
+      "loss": 2.975,
+      "step": 56287
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7683876752853394,
+      "learning_rate": 9.539491963901757e-06,
+      "loss": 2.8721,
+      "step": 56288
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7631601095199585,
+      "learning_rate": 9.535640667551192e-06,
+      "loss": 3.1299,
+      "step": 56289
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7261856198310852,
+      "learning_rate": 9.531790136227869e-06,
+      "loss": 2.7541,
+      "step": 56290
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7447924017906189,
+      "learning_rate": 9.527940369941911e-06,
+      "loss": 2.9447,
+      "step": 56291
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7883365154266357,
+      "learning_rate": 9.524091368703413e-06,
+      "loss": 2.8184,
+      "step": 56292
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7671542763710022,
+      "learning_rate": 9.520243132522597e-06,
+      "loss": 2.9793,
+      "step": 56293
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7295643091201782,
+      "learning_rate": 9.516395661409526e-06,
+      "loss": 3.0381,
+      "step": 56294
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7493358850479126,
+      "learning_rate": 9.512548955374355e-06,
+      "loss": 2.9604,
+      "step": 56295
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7209811806678772,
+      "learning_rate": 9.508703014427278e-06,
+      "loss": 3.0627,
+      "step": 56296
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7741036415100098,
+      "learning_rate": 9.504857838578317e-06,
+      "loss": 2.9489,
+      "step": 56297
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7175572514533997,
+      "learning_rate": 9.501013427837666e-06,
+      "loss": 2.8748,
+      "step": 56298
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7129703760147095,
+      "learning_rate": 9.497169782215485e-06,
+      "loss": 2.9452,
+      "step": 56299
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7331249117851257,
+      "learning_rate": 9.493326901721798e-06,
+      "loss": 2.7358,
+      "step": 56300
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.738678514957428,
+      "learning_rate": 9.489484786366797e-06,
+      "loss": 2.7248,
+      "step": 56301
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7397850751876831,
+      "learning_rate": 9.485643436160572e-06,
+      "loss": 2.9859,
+      "step": 56302
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7723121643066406,
+      "learning_rate": 9.48180285111325e-06,
+      "loss": 3.1972,
+      "step": 56303
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7921167612075806,
+      "learning_rate": 9.477963031234958e-06,
+      "loss": 2.9502,
+      "step": 56304
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7509664297103882,
+      "learning_rate": 9.474123976535819e-06,
+      "loss": 2.9147,
+      "step": 56305
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8044818043708801,
+      "learning_rate": 9.470285687025925e-06,
+      "loss": 3.1336,
+      "step": 56306
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7625141143798828,
+      "learning_rate": 9.466448162715335e-06,
+      "loss": 2.882,
+      "step": 56307
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7402327656745911,
+      "learning_rate": 9.46261140361424e-06,
+      "loss": 2.8615,
+      "step": 56308
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7257512807846069,
+      "learning_rate": 9.4587754097327e-06,
+      "loss": 3.0802,
+      "step": 56309
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7531664371490479,
+      "learning_rate": 9.454940181080839e-06,
+      "loss": 2.9292,
+      "step": 56310
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7305158972740173,
+      "learning_rate": 9.45110571766875e-06,
+      "loss": 2.9042,
+      "step": 56311
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7849349975585938,
+      "learning_rate": 9.447272019506524e-06,
+      "loss": 2.812,
+      "step": 56312
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7790784239768982,
+      "learning_rate": 9.443439086604287e-06,
+      "loss": 2.886,
+      "step": 56313
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7468773126602173,
+      "learning_rate": 9.439606918972097e-06,
+      "loss": 2.9739,
+      "step": 56314
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7422334551811218,
+      "learning_rate": 9.435775516620049e-06,
+      "loss": 2.983,
+      "step": 56315
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7719339728355408,
+      "learning_rate": 9.431944879558263e-06,
+      "loss": 2.9528,
+      "step": 56316
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7582297325134277,
+      "learning_rate": 9.428115007796799e-06,
+      "loss": 2.8616,
+      "step": 56317
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7723546624183655,
+      "learning_rate": 9.424285901345785e-06,
+      "loss": 2.8218,
+      "step": 56318
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7500159740447998,
+      "learning_rate": 9.420457560215245e-06,
+      "loss": 2.9789,
+      "step": 56319
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7486183047294617,
+      "learning_rate": 9.416629984415335e-06,
+      "loss": 2.8109,
+      "step": 56320
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7632740139961243,
+      "learning_rate": 9.412803173956085e-06,
+      "loss": 2.9913,
+      "step": 56321
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7619692087173462,
+      "learning_rate": 9.408977128847584e-06,
+      "loss": 3.0314,
+      "step": 56322
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7509559392929077,
+      "learning_rate": 9.405151849099923e-06,
+      "loss": 2.7083,
+      "step": 56323
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7356145977973938,
+      "learning_rate": 9.40132733472313e-06,
+      "loss": 2.7204,
+      "step": 56324
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7453984618186951,
+      "learning_rate": 9.397503585727328e-06,
+      "loss": 2.845,
+      "step": 56325
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7576645612716675,
+      "learning_rate": 9.393680602122578e-06,
+      "loss": 2.8996,
+      "step": 56326
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7218955755233765,
+      "learning_rate": 9.389858383918969e-06,
+      "loss": 2.8574,
+      "step": 56327
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7386839389801025,
+      "learning_rate": 9.386036931126529e-06,
+      "loss": 2.9469,
+      "step": 56328
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7222349047660828,
+      "learning_rate": 9.382216243755313e-06,
+      "loss": 2.7775,
+      "step": 56329
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7363526225090027,
+      "learning_rate": 9.378396321815419e-06,
+      "loss": 2.9486,
+      "step": 56330
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7235442996025085,
+      "learning_rate": 9.374577165316932e-06,
+      "loss": 3.037,
+      "step": 56331
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7777265310287476,
+      "learning_rate": 9.37075877426985e-06,
+      "loss": 2.9327,
+      "step": 56332
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7499440908432007,
+      "learning_rate": 9.36694114868426e-06,
+      "loss": 3.0098,
+      "step": 56333
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7149835824966431,
+      "learning_rate": 9.363124288570189e-06,
+      "loss": 2.913,
+      "step": 56334
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7982455492019653,
+      "learning_rate": 9.359308193937765e-06,
+      "loss": 3.024,
+      "step": 56335
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7221076488494873,
+      "learning_rate": 9.355492864796977e-06,
+      "loss": 2.7995,
+      "step": 56336
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.6967481970787048,
+      "learning_rate": 9.35167830115785e-06,
+      "loss": 3.0547,
+      "step": 56337
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7800359129905701,
+      "learning_rate": 9.347864503030511e-06,
+      "loss": 2.9932,
+      "step": 56338
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7424789071083069,
+      "learning_rate": 9.344051470424884e-06,
+      "loss": 2.8447,
+      "step": 56339
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7663801908493042,
+      "learning_rate": 9.340239203351129e-06,
+      "loss": 2.9393,
+      "step": 56340
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7594396471977234,
+      "learning_rate": 9.336427701819272e-06,
+      "loss": 2.8301,
+      "step": 56341
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7261060476303101,
+      "learning_rate": 9.332616965839335e-06,
+      "loss": 2.8903,
+      "step": 56342
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7838093042373657,
+      "learning_rate": 9.328806995421311e-06,
+      "loss": 2.7609,
+      "step": 56343
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7690815925598145,
+      "learning_rate": 9.324997790575262e-06,
+      "loss": 2.8137,
+      "step": 56344
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7286287546157837,
+      "learning_rate": 9.321189351311275e-06,
+      "loss": 2.7991,
+      "step": 56345
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7787858843803406,
+      "learning_rate": 9.31738167763928e-06,
+      "loss": 3.0519,
+      "step": 56346
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7564954161643982,
+      "learning_rate": 9.313574769569366e-06,
+      "loss": 2.8195,
+      "step": 56347
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7487925887107849,
+      "learning_rate": 9.309768627111558e-06,
+      "loss": 2.8282,
+      "step": 56348
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7704711556434631,
+      "learning_rate": 9.305963250275916e-06,
+      "loss": 2.811,
+      "step": 56349
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7544106841087341,
+      "learning_rate": 9.302158639072399e-06,
+      "loss": 3.0789,
+      "step": 56350
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7106571793556213,
+      "learning_rate": 9.298354793510998e-06,
+      "loss": 2.7505,
+      "step": 56351
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7845272421836853,
+      "learning_rate": 9.294551713601805e-06,
+      "loss": 2.916,
+      "step": 56352
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7282042503356934,
+      "learning_rate": 9.290749399354847e-06,
+      "loss": 3.0183,
+      "step": 56353
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7017480134963989,
+      "learning_rate": 9.28694785078008e-06,
+      "loss": 2.8869,
+      "step": 56354
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7535594701766968,
+      "learning_rate": 9.283147067887564e-06,
+      "loss": 2.8366,
+      "step": 56355
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7693657279014587,
+      "learning_rate": 9.279347050687291e-06,
+      "loss": 2.8678,
+      "step": 56356
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7219848036766052,
+      "learning_rate": 9.275547799189253e-06,
+      "loss": 2.8987,
+      "step": 56357
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7361370325088501,
+      "learning_rate": 9.271749313403509e-06,
+      "loss": 2.8727,
+      "step": 56358
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7325665950775146,
+      "learning_rate": 9.26795159333995e-06,
+      "loss": 2.8354,
+      "step": 56359
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7923938035964966,
+      "learning_rate": 9.264154639008736e-06,
+      "loss": 2.6835,
+      "step": 56360
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7122077345848083,
+      "learning_rate": 9.260358450419724e-06,
+      "loss": 2.8346,
+      "step": 56361
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7342128753662109,
+      "learning_rate": 9.256563027582975e-06,
+      "loss": 2.8962,
+      "step": 56362
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7379400730133057,
+      "learning_rate": 9.252768370508513e-06,
+      "loss": 2.8967,
+      "step": 56363
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7097499966621399,
+      "learning_rate": 9.248974479206328e-06,
+      "loss": 2.8588,
+      "step": 56364
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7544580101966858,
+      "learning_rate": 9.245181353686348e-06,
+      "loss": 2.9067,
+      "step": 56365
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7791709899902344,
+      "learning_rate": 9.241388993958598e-06,
+      "loss": 2.976,
+      "step": 56366
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8830820918083191,
+      "learning_rate": 9.23759740003307e-06,
+      "loss": 3.0506,
+      "step": 56367
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7556170225143433,
+      "learning_rate": 9.233806571919788e-06,
+      "loss": 2.7625,
+      "step": 56368
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7081811428070068,
+      "learning_rate": 9.230016509628679e-06,
+      "loss": 2.8684,
+      "step": 56369
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7261504530906677,
+      "learning_rate": 9.226227213169735e-06,
+      "loss": 2.8581,
+      "step": 56370
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.768895149230957,
+      "learning_rate": 9.22243868255298e-06,
+      "loss": 2.9572,
+      "step": 56371
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7371829152107239,
+      "learning_rate": 9.218650917788339e-06,
+      "loss": 2.7464,
+      "step": 56372
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7369073033332825,
+      "learning_rate": 9.214863918885806e-06,
+      "loss": 2.6433,
+      "step": 56373
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7547151446342468,
+      "learning_rate": 9.211077685855372e-06,
+      "loss": 2.8662,
+      "step": 56374
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7417933344841003,
+      "learning_rate": 9.207292218706996e-06,
+      "loss": 3.0114,
+      "step": 56375
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7300156354904175,
+      "learning_rate": 9.203507517450603e-06,
+      "loss": 3.0027,
+      "step": 56376
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7735124230384827,
+      "learning_rate": 9.199723582096253e-06,
+      "loss": 2.8641,
+      "step": 56377
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7115468382835388,
+      "learning_rate": 9.195940412653868e-06,
+      "loss": 2.8722,
+      "step": 56378
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7530733942985535,
+      "learning_rate": 9.19215800913341e-06,
+      "loss": 2.9117,
+      "step": 56379
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.74192875623703,
+      "learning_rate": 9.188376371544837e-06,
+      "loss": 2.8185,
+      "step": 56380
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7729009389877319,
+      "learning_rate": 9.184595499898073e-06,
+      "loss": 2.9038,
+      "step": 56381
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7162991166114807,
+      "learning_rate": 9.180815394203145e-06,
+      "loss": 3.1287,
+      "step": 56382
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7535189390182495,
+      "learning_rate": 9.177036054470011e-06,
+      "loss": 2.8736,
+      "step": 56383
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7650796175003052,
+      "learning_rate": 9.17325748070853e-06,
+      "loss": 3.0441,
+      "step": 56384
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7328662872314453,
+      "learning_rate": 9.169479672928759e-06,
+      "loss": 2.8645,
+      "step": 56385
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7376196384429932,
+      "learning_rate": 9.165702631140625e-06,
+      "loss": 2.7676,
+      "step": 56386
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7664865255355835,
+      "learning_rate": 9.161926355354022e-06,
+      "loss": 3.0326,
+      "step": 56387
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7669900059700012,
+      "learning_rate": 9.158150845578971e-06,
+      "loss": 2.8375,
+      "step": 56388
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7463096976280212,
+      "learning_rate": 9.154376101825333e-06,
+      "loss": 2.9926,
+      "step": 56389
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7658016085624695,
+      "learning_rate": 9.1506021241031e-06,
+      "loss": 2.7995,
+      "step": 56390
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7736062407493591,
+      "learning_rate": 9.146828912422199e-06,
+      "loss": 2.858,
+      "step": 56391
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7192699313163757,
+      "learning_rate": 9.14305646679262e-06,
+      "loss": 2.8394,
+      "step": 56392
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7871600985527039,
+      "learning_rate": 9.139284787224254e-06,
+      "loss": 2.9598,
+      "step": 56393
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7646672129631042,
+      "learning_rate": 9.135513873726996e-06,
+      "loss": 2.9829,
+      "step": 56394
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7649540305137634,
+      "learning_rate": 9.131743726310835e-06,
+      "loss": 3.1376,
+      "step": 56395
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7846724987030029,
+      "learning_rate": 9.127974344985634e-06,
+      "loss": 2.955,
+      "step": 56396
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7587399482727051,
+      "learning_rate": 9.124205729761414e-06,
+      "loss": 2.7941,
+      "step": 56397
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7872162461280823,
+      "learning_rate": 9.120437880648068e-06,
+      "loss": 2.9246,
+      "step": 56398
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7765682339668274,
+      "learning_rate": 9.116670797655456e-06,
+      "loss": 2.6701,
+      "step": 56399
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7208048701286316,
+      "learning_rate": 9.112904480793604e-06,
+      "loss": 2.915,
+      "step": 56400
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7376406192779541,
+      "learning_rate": 9.109138930072401e-06,
+      "loss": 3.0416,
+      "step": 56401
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7587045431137085,
+      "learning_rate": 9.105374145501676e-06,
+      "loss": 3.0432,
+      "step": 56402
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7954713106155396,
+      "learning_rate": 9.101610127091453e-06,
+      "loss": 2.8843,
+      "step": 56403
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.783104419708252,
+      "learning_rate": 9.097846874851588e-06,
+      "loss": 3.0071,
+      "step": 56404
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7557005286216736,
+      "learning_rate": 9.094084388792012e-06,
+      "loss": 2.8479,
+      "step": 56405
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.738347053527832,
+      "learning_rate": 9.090322668922612e-06,
+      "loss": 3.0269,
+      "step": 56406
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7137772440910339,
+      "learning_rate": 9.086561715253349e-06,
+      "loss": 2.7575,
+      "step": 56407
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7108136415481567,
+      "learning_rate": 9.082801527794115e-06,
+      "loss": 3.1663,
+      "step": 56408
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.773616373538971,
+      "learning_rate": 9.079042106554734e-06,
+      "loss": 2.8878,
+      "step": 56409
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7863094806671143,
+      "learning_rate": 9.075283451545202e-06,
+      "loss": 3.0319,
+      "step": 56410
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7484992146492004,
+      "learning_rate": 9.071525562775373e-06,
+      "loss": 2.9823,
+      "step": 56411
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7688180804252625,
+      "learning_rate": 9.067768440255141e-06,
+      "loss": 2.727,
+      "step": 56412
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7612563967704773,
+      "learning_rate": 9.064012083994433e-06,
+      "loss": 2.9456,
+      "step": 56413
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7126185297966003,
+      "learning_rate": 9.060256494003105e-06,
+      "loss": 2.7921,
+      "step": 56414
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7291039824485779,
+      "learning_rate": 9.056501670291149e-06,
+      "loss": 2.8641,
+      "step": 56415
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7190093398094177,
+      "learning_rate": 9.052747612868294e-06,
+      "loss": 3.0766,
+      "step": 56416
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7553203701972961,
+      "learning_rate": 9.04899432174453e-06,
+      "loss": 2.7905,
+      "step": 56417
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7836436629295349,
+      "learning_rate": 9.045241796929713e-06,
+      "loss": 2.6367,
+      "step": 56418
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7655666470527649,
+      "learning_rate": 9.041490038433741e-06,
+      "loss": 2.9534,
+      "step": 56419
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7657229900360107,
+      "learning_rate": 9.037739046266501e-06,
+      "loss": 3.0535,
+      "step": 56420
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7775158882141113,
+      "learning_rate": 9.033988820437821e-06,
+      "loss": 2.8456,
+      "step": 56421
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.773569643497467,
+      "learning_rate": 9.030239360957658e-06,
+      "loss": 2.8731,
+      "step": 56422
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7964552640914917,
+      "learning_rate": 9.02649066783584e-06,
+      "loss": 2.8055,
+      "step": 56423
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7493492960929871,
+      "learning_rate": 9.022742741082223e-06,
+      "loss": 2.7148,
+      "step": 56424
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7704905867576599,
+      "learning_rate": 9.018995580706733e-06,
+      "loss": 2.834,
+      "step": 56425
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.773236095905304,
+      "learning_rate": 9.015249186719164e-06,
+      "loss": 2.7859,
+      "step": 56426
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7138835787773132,
+      "learning_rate": 9.011503559129474e-06,
+      "loss": 3.0637,
+      "step": 56427
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7717339396476746,
+      "learning_rate": 9.007758697947453e-06,
+      "loss": 3.0914,
+      "step": 56428
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7018489837646484,
+      "learning_rate": 9.004014603182996e-06,
+      "loss": 2.8645,
+      "step": 56429
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7404462099075317,
+      "learning_rate": 9.000271274845994e-06,
+      "loss": 2.8071,
+      "step": 56430
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.74892258644104,
+      "learning_rate": 8.996528712946238e-06,
+      "loss": 3.0029,
+      "step": 56431
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8299066424369812,
+      "learning_rate": 8.992786917493622e-06,
+      "loss": 2.8981,
+      "step": 56432
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7275087237358093,
+      "learning_rate": 8.989045888498003e-06,
+      "loss": 2.8402,
+      "step": 56433
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.73467618227005,
+      "learning_rate": 8.985305625969209e-06,
+      "loss": 2.5677,
+      "step": 56434
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7989902496337891,
+      "learning_rate": 8.981566129917162e-06,
+      "loss": 2.9348,
+      "step": 56435
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7334526777267456,
+      "learning_rate": 8.977827400351623e-06,
+      "loss": 2.8775,
+      "step": 56436
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7814991474151611,
+      "learning_rate": 8.974089437282517e-06,
+      "loss": 3.0359,
+      "step": 56437
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7895364165306091,
+      "learning_rate": 8.970352240719637e-06,
+      "loss": 3.0237,
+      "step": 56438
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7760238647460938,
+      "learning_rate": 8.966615810672805e-06,
+      "loss": 2.7975,
+      "step": 56439
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.795097827911377,
+      "learning_rate": 8.962880147151951e-06,
+      "loss": 2.7911,
+      "step": 56440
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7412228584289551,
+      "learning_rate": 8.959145250166799e-06,
+      "loss": 2.7482,
+      "step": 56441
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7412148118019104,
+      "learning_rate": 8.955411119727307e-06,
+      "loss": 2.8579,
+      "step": 56442
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.740685224533081,
+      "learning_rate": 8.951677755843201e-06,
+      "loss": 2.8466,
+      "step": 56443
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7448912262916565,
+      "learning_rate": 8.947945158524372e-06,
+      "loss": 2.7933,
+      "step": 56444
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7429138422012329,
+      "learning_rate": 8.94421332778068e-06,
+      "loss": 2.8862,
+      "step": 56445
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7522229552268982,
+      "learning_rate": 8.940482263621884e-06,
+      "loss": 2.8816,
+      "step": 56446
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7623513340950012,
+      "learning_rate": 8.936751966057842e-06,
+      "loss": 2.8082,
+      "step": 56447
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7645907402038574,
+      "learning_rate": 8.93302243509838e-06,
+      "loss": 2.9118,
+      "step": 56448
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7380853891372681,
+      "learning_rate": 8.92929367075329e-06,
+      "loss": 2.8021,
+      "step": 56449
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.714244544506073,
+      "learning_rate": 8.925565673032465e-06,
+      "loss": 3.1272,
+      "step": 56450
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7148142457008362,
+      "learning_rate": 8.921838441945695e-06,
+      "loss": 2.8177,
+      "step": 56451
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7204702496528625,
+      "learning_rate": 8.918111977502772e-06,
+      "loss": 2.8349,
+      "step": 56452
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7750880718231201,
+      "learning_rate": 8.914386279713492e-06,
+      "loss": 3.0569,
+      "step": 56453
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7485977411270142,
+      "learning_rate": 8.910661348587711e-06,
+      "loss": 2.8497,
+      "step": 56454
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7468576431274414,
+      "learning_rate": 8.90693718413522e-06,
+      "loss": 2.9756,
+      "step": 56455
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7802996635437012,
+      "learning_rate": 8.903213786365847e-06,
+      "loss": 2.9238,
+      "step": 56456
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7729799747467041,
+      "learning_rate": 8.899491155289385e-06,
+      "loss": 2.8883,
+      "step": 56457
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7460412383079529,
+      "learning_rate": 8.895769290915621e-06,
+      "loss": 2.7785,
+      "step": 56458
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8159653544425964,
+      "learning_rate": 8.89204819325442e-06,
+      "loss": 2.9526,
+      "step": 56459
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7661879658699036,
+      "learning_rate": 8.888327862315536e-06,
+      "loss": 2.9368,
+      "step": 56460
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.769864022731781,
+      "learning_rate": 8.884608298108731e-06,
+      "loss": 2.9032,
+      "step": 56461
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7339568734169006,
+      "learning_rate": 8.880889500643861e-06,
+      "loss": 2.8907,
+      "step": 56462
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7468345761299133,
+      "learning_rate": 8.877171469930689e-06,
+      "loss": 2.7929,
+      "step": 56463
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7140334844589233,
+      "learning_rate": 8.873454205979002e-06,
+      "loss": 2.695,
+      "step": 56464
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7177031636238098,
+      "learning_rate": 8.869737708798663e-06,
+      "loss": 2.9442,
+      "step": 56465
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7148962616920471,
+      "learning_rate": 8.866021978399396e-06,
+      "loss": 2.8238,
+      "step": 56466
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.771285891532898,
+      "learning_rate": 8.862307014790992e-06,
+      "loss": 2.9251,
+      "step": 56467
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7293102741241455,
+      "learning_rate": 8.858592817983212e-06,
+      "loss": 3.0754,
+      "step": 56468
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.776573657989502,
+      "learning_rate": 8.85487938798588e-06,
+      "loss": 2.9431,
+      "step": 56469
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7331776022911072,
+      "learning_rate": 8.85116672480879e-06,
+      "loss": 3.0407,
+      "step": 56470
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7505592703819275,
+      "learning_rate": 8.847454828461632e-06,
+      "loss": 2.9197,
+      "step": 56471
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7655132412910461,
+      "learning_rate": 8.843743698954297e-06,
+      "loss": 3.0714,
+      "step": 56472
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7211343050003052,
+      "learning_rate": 8.840033336296516e-06,
+      "loss": 2.8666,
+      "step": 56473
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7068425416946411,
+      "learning_rate": 8.836323740498009e-06,
+      "loss": 2.8191,
+      "step": 56474
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.6717082262039185,
+      "learning_rate": 8.832614911568636e-06,
+      "loss": 2.8159,
+      "step": 56475
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7441225051879883,
+      "learning_rate": 8.828906849518092e-06,
+      "loss": 2.8976,
+      "step": 56476
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7472232580184937,
+      "learning_rate": 8.825199554356166e-06,
+      "loss": 2.7381,
+      "step": 56477
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7638667225837708,
+      "learning_rate": 8.821493026092619e-06,
+      "loss": 2.8284,
+      "step": 56478
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7200614213943481,
+      "learning_rate": 8.817787264737242e-06,
+      "loss": 2.9348,
+      "step": 56479
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7430014610290527,
+      "learning_rate": 8.81408227029976e-06,
+      "loss": 2.9065,
+      "step": 56480
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7274750471115112,
+      "learning_rate": 8.810378042789968e-06,
+      "loss": 2.7577,
+      "step": 56481
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7740485072135925,
+      "learning_rate": 8.80667458221762e-06,
+      "loss": 2.9349,
+      "step": 56482
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7242769598960876,
+      "learning_rate": 8.802971888592381e-06,
+      "loss": 2.9105,
+      "step": 56483
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7862998843193054,
+      "learning_rate": 8.799269961924104e-06,
+      "loss": 2.8458,
+      "step": 56484
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7596186399459839,
+      "learning_rate": 8.79556880222252e-06,
+      "loss": 2.8756,
+      "step": 56485
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7429347634315491,
+      "learning_rate": 8.791868409497316e-06,
+      "loss": 3.0002,
+      "step": 56486
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7703331708908081,
+      "learning_rate": 8.788168783758354e-06,
+      "loss": 2.9832,
+      "step": 56487
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7370572090148926,
+      "learning_rate": 8.784469925015292e-06,
+      "loss": 2.7631,
+      "step": 56488
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7434906959533691,
+      "learning_rate": 8.780771833277855e-06,
+      "loss": 2.8704,
+      "step": 56489
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7217363119125366,
+      "learning_rate": 8.777074508555871e-06,
+      "loss": 2.8909,
+      "step": 56490
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7547829747200012,
+      "learning_rate": 8.773377950858995e-06,
+      "loss": 2.8094,
+      "step": 56491
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7625556588172913,
+      "learning_rate": 8.769682160196989e-06,
+      "loss": 2.9916,
+      "step": 56492
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7367002964019775,
+      "learning_rate": 8.765987136579577e-06,
+      "loss": 2.7937,
+      "step": 56493
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7374213337898254,
+      "learning_rate": 8.762292880016553e-06,
+      "loss": 2.9097,
+      "step": 56494
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8057516813278198,
+      "learning_rate": 8.758599390517607e-06,
+      "loss": 2.7651,
+      "step": 56495
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7465009093284607,
+      "learning_rate": 8.754906668092398e-06,
+      "loss": 2.9281,
+      "step": 56496
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7300065755844116,
+      "learning_rate": 8.751214712750754e-06,
+      "loss": 2.9257,
+      "step": 56497
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7591533064842224,
+      "learning_rate": 8.747523524502332e-06,
+      "loss": 3.0157,
+      "step": 56498
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7155272960662842,
+      "learning_rate": 8.743833103356923e-06,
+      "loss": 2.881,
+      "step": 56499
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7301709651947021,
+      "learning_rate": 8.740143449324155e-06,
+      "loss": 3.1471,
+      "step": 56500
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7165641784667969,
+      "learning_rate": 8.736454562413786e-06,
+      "loss": 2.7683,
+      "step": 56501
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7362425923347473,
+      "learning_rate": 8.732766442635607e-06,
+      "loss": 2.9066,
+      "step": 56502
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7725816965103149,
+      "learning_rate": 8.729079089999246e-06,
+      "loss": 2.9066,
+      "step": 56503
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7778238654136658,
+      "learning_rate": 8.725392504514429e-06,
+      "loss": 2.8673,
+      "step": 56504
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7198028564453125,
+      "learning_rate": 8.721706686190843e-06,
+      "loss": 2.9211,
+      "step": 56505
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7817999720573425,
+      "learning_rate": 8.71802163503822e-06,
+      "loss": 2.8808,
+      "step": 56506
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7423765063285828,
+      "learning_rate": 8.714337351066314e-06,
+      "loss": 2.7144,
+      "step": 56507
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7737480998039246,
+      "learning_rate": 8.710653834284753e-06,
+      "loss": 2.9229,
+      "step": 56508
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7494272589683533,
+      "learning_rate": 8.706971084703295e-06,
+      "loss": 2.9935,
+      "step": 56509
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7207717895507812,
+      "learning_rate": 8.7032891023316e-06,
+      "loss": 3.0677,
+      "step": 56510
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7812270522117615,
+      "learning_rate": 8.699607887179361e-06,
+      "loss": 2.7982,
+      "step": 56511
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8005214929580688,
+      "learning_rate": 8.695927439256334e-06,
+      "loss": 2.6689,
+      "step": 56512
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7496790885925293,
+      "learning_rate": 8.692247758572146e-06,
+      "loss": 2.8616,
+      "step": 56513
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7499210238456726,
+      "learning_rate": 8.688568845136523e-06,
+      "loss": 2.803,
+      "step": 56514
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7113456130027771,
+      "learning_rate": 8.684890698959125e-06,
+      "loss": 2.8449,
+      "step": 56515
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7013689279556274,
+      "learning_rate": 8.681213320049674e-06,
+      "loss": 2.8038,
+      "step": 56516
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7134240865707397,
+      "learning_rate": 8.677536708417865e-06,
+      "loss": 2.8203,
+      "step": 56517
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7750184535980225,
+      "learning_rate": 8.673860864073323e-06,
+      "loss": 2.7887,
+      "step": 56518
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7498697638511658,
+      "learning_rate": 8.670185787025807e-06,
+      "loss": 3.0101,
+      "step": 56519
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7713249921798706,
+      "learning_rate": 8.66651147728491e-06,
+      "loss": 2.7569,
+      "step": 56520
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7590934038162231,
+      "learning_rate": 8.662837934860322e-06,
+      "loss": 2.8114,
+      "step": 56521
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.729499101638794,
+      "learning_rate": 8.659165159761838e-06,
+      "loss": 2.8747,
+      "step": 56522
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7570987939834595,
+      "learning_rate": 8.655493151998983e-06,
+      "loss": 3.1679,
+      "step": 56523
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7177196145057678,
+      "learning_rate": 8.651821911581513e-06,
+      "loss": 2.9119,
+      "step": 56524
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7343611121177673,
+      "learning_rate": 8.648151438519091e-06,
+      "loss": 3.0303,
+      "step": 56525
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7677000164985657,
+      "learning_rate": 8.644481732821306e-06,
+      "loss": 2.9359,
+      "step": 56526
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7537042498588562,
+      "learning_rate": 8.640812794497953e-06,
+      "loss": 2.7783,
+      "step": 56527
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7486345171928406,
+      "learning_rate": 8.637144623558557e-06,
+      "loss": 2.7331,
+      "step": 56528
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7860642671585083,
+      "learning_rate": 8.633477220012908e-06,
+      "loss": 2.9114,
+      "step": 56529
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7259573340415955,
+      "learning_rate": 8.629810583870567e-06,
+      "loss": 2.9674,
+      "step": 56530
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7208335995674133,
+      "learning_rate": 8.62614471514126e-06,
+      "loss": 2.7895,
+      "step": 56531
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7486009001731873,
+      "learning_rate": 8.62247961383461e-06,
+      "loss": 2.948,
+      "step": 56532
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7948107719421387,
+      "learning_rate": 8.618815279960244e-06,
+      "loss": 2.8621,
+      "step": 56533
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7149216532707214,
+      "learning_rate": 8.615151713527857e-06,
+      "loss": 2.8832,
+      "step": 56534
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7751069068908691,
+      "learning_rate": 8.611488914547071e-06,
+      "loss": 2.9773,
+      "step": 56535
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7153540849685669,
+      "learning_rate": 8.607826883027547e-06,
+      "loss": 2.687,
+      "step": 56536
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7932222485542297,
+      "learning_rate": 8.604165618978942e-06,
+      "loss": 3.1045,
+      "step": 56537
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7211452126502991,
+      "learning_rate": 8.60050512241085e-06,
+      "loss": 2.8193,
+      "step": 56538
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7320334911346436,
+      "learning_rate": 8.596845393332997e-06,
+      "loss": 2.969,
+      "step": 56539
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7381606101989746,
+      "learning_rate": 8.593186431754973e-06,
+      "loss": 2.72,
+      "step": 56540
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7109737992286682,
+      "learning_rate": 8.589528237686372e-06,
+      "loss": 2.9596,
+      "step": 56541
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7556430101394653,
+      "learning_rate": 8.585870811136885e-06,
+      "loss": 3.0741,
+      "step": 56542
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7720571160316467,
+      "learning_rate": 8.58221415211614e-06,
+      "loss": 3.1015,
+      "step": 56543
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7691468000411987,
+      "learning_rate": 8.57855826063376e-06,
+      "loss": 2.952,
+      "step": 56544
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8022409677505493,
+      "learning_rate": 8.57490313669934e-06,
+      "loss": 2.8529,
+      "step": 56545
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7315953969955444,
+      "learning_rate": 8.57124878032257e-06,
+      "loss": 2.7401,
+      "step": 56546
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7592272162437439,
+      "learning_rate": 8.567595191513044e-06,
+      "loss": 2.9078,
+      "step": 56547
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7328471541404724,
+      "learning_rate": 8.563942370280386e-06,
+      "loss": 2.691,
+      "step": 56548
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7198687195777893,
+      "learning_rate": 8.560290316634222e-06,
+      "loss": 2.7965,
+      "step": 56549
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7822607159614563,
+      "learning_rate": 8.556639030584112e-06,
+      "loss": 2.9445,
+      "step": 56550
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.789264976978302,
+      "learning_rate": 8.552988512139747e-06,
+      "loss": 2.85,
+      "step": 56551
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7752166390419006,
+      "learning_rate": 8.549338761310753e-06,
+      "loss": 2.541,
+      "step": 56552
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7067755460739136,
+      "learning_rate": 8.545689778106724e-06,
+      "loss": 2.9171,
+      "step": 56553
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7491674423217773,
+      "learning_rate": 8.54204156253725e-06,
+      "loss": 2.7476,
+      "step": 56554
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8012225031852722,
+      "learning_rate": 8.53839411461189e-06,
+      "loss": 2.9535,
+      "step": 56555
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7741601467132568,
+      "learning_rate": 8.534747434340306e-06,
+      "loss": 2.8989,
+      "step": 56556
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.735129714012146,
+      "learning_rate": 8.531101521732153e-06,
+      "loss": 2.8754,
+      "step": 56557
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.730719268321991,
+      "learning_rate": 8.527456376796959e-06,
+      "loss": 2.8079,
+      "step": 56558
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8632533550262451,
+      "learning_rate": 8.523811999544383e-06,
+      "loss": 2.8242,
+      "step": 56559
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7945080399513245,
+      "learning_rate": 8.52016838998395e-06,
+      "loss": 2.7674,
+      "step": 56560
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7971088290214539,
+      "learning_rate": 8.516525548125319e-06,
+      "loss": 2.7071,
+      "step": 56561
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7828199863433838,
+      "learning_rate": 8.512883473978083e-06,
+      "loss": 2.7697,
+      "step": 56562
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7509476542472839,
+      "learning_rate": 8.509242167551767e-06,
+      "loss": 2.7219,
+      "step": 56563
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7151806950569153,
+      "learning_rate": 8.505601628856062e-06,
+      "loss": 3.0501,
+      "step": 56564
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7569729089736938,
+      "learning_rate": 8.50196185790043e-06,
+      "loss": 2.9743,
+      "step": 56565
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7372108101844788,
+      "learning_rate": 8.498322854694562e-06,
+      "loss": 2.8322,
+      "step": 56566
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7561090588569641,
+      "learning_rate": 8.494684619248049e-06,
+      "loss": 2.9245,
+      "step": 56567
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7554269433021545,
+      "learning_rate": 8.491047151570418e-06,
+      "loss": 2.9634,
+      "step": 56568
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7970679402351379,
+      "learning_rate": 8.487410451671261e-06,
+      "loss": 2.8352,
+      "step": 56569
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7189784049987793,
+      "learning_rate": 8.483774519560137e-06,
+      "loss": 2.9002,
+      "step": 56570
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.768104612827301,
+      "learning_rate": 8.480139355246674e-06,
+      "loss": 3.1014,
+      "step": 56571
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7429171800613403,
+      "learning_rate": 8.476504958740427e-06,
+      "loss": 2.9454,
+      "step": 56572
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7118484973907471,
+      "learning_rate": 8.472871330050923e-06,
+      "loss": 3.0559,
+      "step": 56573
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7167723178863525,
+      "learning_rate": 8.46923846918779e-06,
+      "loss": 2.8801,
+      "step": 56574
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.749420702457428,
+      "learning_rate": 8.465606376160617e-06,
+      "loss": 2.8208,
+      "step": 56575
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8014590740203857,
+      "learning_rate": 8.461975050978898e-06,
+      "loss": 2.9306,
+      "step": 56576
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.855155348777771,
+      "learning_rate": 8.458344493652225e-06,
+      "loss": 2.9504,
+      "step": 56577
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7464328408241272,
+      "learning_rate": 8.454714704190125e-06,
+      "loss": 2.8316,
+      "step": 56578
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7837406992912292,
+      "learning_rate": 8.451085682602255e-06,
+      "loss": 2.8898,
+      "step": 56579
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7430059313774109,
+      "learning_rate": 8.447457428898108e-06,
+      "loss": 2.8781,
+      "step": 56580
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7241303324699402,
+      "learning_rate": 8.443829943087244e-06,
+      "loss": 3.1459,
+      "step": 56581
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7481164932250977,
+      "learning_rate": 8.440203225179187e-06,
+      "loss": 2.721,
+      "step": 56582
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.766409158706665,
+      "learning_rate": 8.436577275183564e-06,
+      "loss": 3.0709,
+      "step": 56583
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7170672416687012,
+      "learning_rate": 8.432952093109901e-06,
+      "loss": 3.1805,
+      "step": 56584
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7190765142440796,
+      "learning_rate": 8.429327678967656e-06,
+      "loss": 2.8354,
+      "step": 56585
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.783243715763092,
+      "learning_rate": 8.425704032766523e-06,
+      "loss": 2.836,
+      "step": 56586
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.716397762298584,
+      "learning_rate": 8.422081154515925e-06,
+      "loss": 2.8644,
+      "step": 56587
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.766613781452179,
+      "learning_rate": 8.418459044225456e-06,
+      "loss": 2.8583,
+      "step": 56588
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7538684606552124,
+      "learning_rate": 8.414837701904642e-06,
+      "loss": 2.8837,
+      "step": 56589
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7688279151916504,
+      "learning_rate": 8.411217127563075e-06,
+      "loss": 2.9859,
+      "step": 56590
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8095577359199524,
+      "learning_rate": 8.407597321210214e-06,
+      "loss": 2.7778,
+      "step": 56591
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7337608337402344,
+      "learning_rate": 8.403978282855617e-06,
+      "loss": 2.8053,
+      "step": 56592
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7341989278793335,
+      "learning_rate": 8.40036001250881e-06,
+      "loss": 2.9954,
+      "step": 56593
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7581641674041748,
+      "learning_rate": 8.396742510179388e-06,
+      "loss": 3.1145,
+      "step": 56594
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7279971241950989,
+      "learning_rate": 8.393125775876775e-06,
+      "loss": 2.8688,
+      "step": 56595
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7769571542739868,
+      "learning_rate": 8.389509809610562e-06,
+      "loss": 2.6323,
+      "step": 56596
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7689619660377502,
+      "learning_rate": 8.385894611390276e-06,
+      "loss": 2.814,
+      "step": 56597
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7330824732780457,
+      "learning_rate": 8.38228018122541e-06,
+      "loss": 2.7816,
+      "step": 56598
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7108086943626404,
+      "learning_rate": 8.378666519125487e-06,
+      "loss": 2.8276,
+      "step": 56599
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7666386365890503,
+      "learning_rate": 8.375053625100037e-06,
+      "loss": 2.9543,
+      "step": 56600
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7531747817993164,
+      "learning_rate": 8.371441499158582e-06,
+      "loss": 3.0132,
+      "step": 56601
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7575362920761108,
+      "learning_rate": 8.367830141310583e-06,
+      "loss": 3.0232,
+      "step": 56602
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.741112232208252,
+      "learning_rate": 8.364219551565631e-06,
+      "loss": 2.7388,
+      "step": 56603
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7264926433563232,
+      "learning_rate": 8.360609729933187e-06,
+      "loss": 2.987,
+      "step": 56604
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8191945552825928,
+      "learning_rate": 8.357000676422808e-06,
+      "loss": 2.8198,
+      "step": 56605
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8269351720809937,
+      "learning_rate": 8.353392391043922e-06,
+      "loss": 2.8031,
+      "step": 56606
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7800099849700928,
+      "learning_rate": 8.349784873806087e-06,
+      "loss": 2.8467,
+      "step": 56607
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7215859293937683,
+      "learning_rate": 8.34617812471876e-06,
+      "loss": 2.7812,
+      "step": 56608
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8046889901161194,
+      "learning_rate": 8.342572143791504e-06,
+      "loss": 2.8322,
+      "step": 56609
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.753553569316864,
+      "learning_rate": 8.338966931033741e-06,
+      "loss": 2.8108,
+      "step": 56610
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7166416049003601,
+      "learning_rate": 8.335362486455066e-06,
+      "loss": 2.8806,
+      "step": 56611
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7240424156188965,
+      "learning_rate": 8.331758810064903e-06,
+      "loss": 3.0349,
+      "step": 56612
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7516683340072632,
+      "learning_rate": 8.328155901872713e-06,
+      "loss": 2.9413,
+      "step": 56613
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7496482729911804,
+      "learning_rate": 8.324553761888053e-06,
+      "loss": 2.8857,
+      "step": 56614
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7219851016998291,
+      "learning_rate": 8.320952390120383e-06,
+      "loss": 2.9247,
+      "step": 56615
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7659447193145752,
+      "learning_rate": 8.317351786579196e-06,
+      "loss": 2.9269,
+      "step": 56616
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7846766114234924,
+      "learning_rate": 8.31375195127395e-06,
+      "loss": 2.7706,
+      "step": 56617
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8673431277275085,
+      "learning_rate": 8.310152884214173e-06,
+      "loss": 3.0868,
+      "step": 56618
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8626456260681152,
+      "learning_rate": 8.30655458540932e-06,
+      "loss": 2.8726,
+      "step": 56619
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7804566621780396,
+      "learning_rate": 8.30295705486882e-06,
+      "loss": 2.7799,
+      "step": 56620
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8007548451423645,
+      "learning_rate": 8.29936029260223e-06,
+      "loss": 2.9062,
+      "step": 56621
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7293591499328613,
+      "learning_rate": 8.295764298618946e-06,
+      "loss": 2.8577,
+      "step": 56622
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8285486698150635,
+      "learning_rate": 8.29216907292849e-06,
+      "loss": 2.9366,
+      "step": 56623
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7310425043106079,
+      "learning_rate": 8.288574615540355e-06,
+      "loss": 2.931,
+      "step": 56624
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7126893401145935,
+      "learning_rate": 8.284980926463936e-06,
+      "loss": 2.839,
+      "step": 56625
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7340129017829895,
+      "learning_rate": 8.28138800570879e-06,
+      "loss": 2.925,
+      "step": 56626
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7461667060852051,
+      "learning_rate": 8.277795853284308e-06,
+      "loss": 2.7116,
+      "step": 56627
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7524672150611877,
+      "learning_rate": 8.274204469199919e-06,
+      "loss": 3.0097,
+      "step": 56628
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7391806244850159,
+      "learning_rate": 8.27061385346518e-06,
+      "loss": 2.9233,
+      "step": 56629
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7427461743354797,
+      "learning_rate": 8.267024006089484e-06,
+      "loss": 2.8764,
+      "step": 56630
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7367878556251526,
+      "learning_rate": 8.263434927082291e-06,
+      "loss": 2.9975,
+      "step": 56631
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7743518948554993,
+      "learning_rate": 8.259846616453058e-06,
+      "loss": 2.8299,
+      "step": 56632
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.9217862486839294,
+      "learning_rate": 8.25625907421128e-06,
+      "loss": 2.9932,
+      "step": 56633
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7349005937576294,
+      "learning_rate": 8.25267230036638e-06,
+      "loss": 2.9098,
+      "step": 56634
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7806947827339172,
+      "learning_rate": 8.24908629492772e-06,
+      "loss": 2.6354,
+      "step": 56635
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7729756832122803,
+      "learning_rate": 8.24550105790489e-06,
+      "loss": 2.8451,
+      "step": 56636
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7258002758026123,
+      "learning_rate": 8.241916589307219e-06,
+      "loss": 2.9029,
+      "step": 56637
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7425938248634338,
+      "learning_rate": 8.238332889144194e-06,
+      "loss": 3.0206,
+      "step": 56638
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7938366532325745,
+      "learning_rate": 8.234749957425247e-06,
+      "loss": 3.0195,
+      "step": 56639
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7334127426147461,
+      "learning_rate": 8.231167794159832e-06,
+      "loss": 2.958,
+      "step": 56640
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7349879145622253,
+      "learning_rate": 8.22758639935741e-06,
+      "loss": 2.8197,
+      "step": 56641
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7187401056289673,
+      "learning_rate": 8.22400577302731e-06,
+      "loss": 2.9082,
+      "step": 56642
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7203710079193115,
+      "learning_rate": 8.220425915179018e-06,
+      "loss": 2.7315,
+      "step": 56643
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7575718760490417,
+      "learning_rate": 8.216846825821999e-06,
+      "loss": 2.829,
+      "step": 56644
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7450301647186279,
+      "learning_rate": 8.213268504965643e-06,
+      "loss": 2.7181,
+      "step": 56645
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8067429065704346,
+      "learning_rate": 8.209690952619407e-06,
+      "loss": 2.8633,
+      "step": 56646
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.795412540435791,
+      "learning_rate": 8.206114168792622e-06,
+      "loss": 2.8272,
+      "step": 56647
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8101593852043152,
+      "learning_rate": 8.202538153494842e-06,
+      "loss": 3.12,
+      "step": 56648
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7850880026817322,
+      "learning_rate": 8.19896290673543e-06,
+      "loss": 3.0746,
+      "step": 56649
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7382371425628662,
+      "learning_rate": 8.195388428523741e-06,
+      "loss": 2.7615,
+      "step": 56650
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7291812896728516,
+      "learning_rate": 8.191814718869271e-06,
+      "loss": 2.6933,
+      "step": 56651
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7332150936126709,
+      "learning_rate": 8.18824177778138e-06,
+      "loss": 2.9268,
+      "step": 56652
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7187190055847168,
+      "learning_rate": 8.18466960526949e-06,
+      "loss": 2.8866,
+      "step": 56653
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7524594664573669,
+      "learning_rate": 8.181098201343028e-06,
+      "loss": 3.1315,
+      "step": 56654
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7417961955070496,
+      "learning_rate": 8.177527566011422e-06,
+      "loss": 2.8357,
+      "step": 56655
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7433927059173584,
+      "learning_rate": 8.173957699284029e-06,
+      "loss": 2.8197,
+      "step": 56656
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7373107671737671,
+      "learning_rate": 8.170388601170275e-06,
+      "loss": 2.847,
+      "step": 56657
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7926795482635498,
+      "learning_rate": 8.166820271679552e-06,
+      "loss": 2.6819,
+      "step": 56658
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7191329598426819,
+      "learning_rate": 8.163252710821222e-06,
+      "loss": 2.9689,
+      "step": 56659
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.820548415184021,
+      "learning_rate": 8.15968591860474e-06,
+      "loss": 2.8692,
+      "step": 56660
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7487797141075134,
+      "learning_rate": 8.156119895039503e-06,
+      "loss": 3.0475,
+      "step": 56661
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.78996741771698,
+      "learning_rate": 8.152554640134834e-06,
+      "loss": 3.1179,
+      "step": 56662
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7686530351638794,
+      "learning_rate": 8.148990153900225e-06,
+      "loss": 2.9168,
+      "step": 56663
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8370349407196045,
+      "learning_rate": 8.145426436345004e-06,
+      "loss": 2.891,
+      "step": 56664
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.756450355052948,
+      "learning_rate": 8.141863487478528e-06,
+      "loss": 2.8712,
+      "step": 56665
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7915993332862854,
+      "learning_rate": 8.138301307310224e-06,
+      "loss": 2.7854,
+      "step": 56666
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7291852235794067,
+      "learning_rate": 8.134739895849452e-06,
+      "loss": 2.91,
+      "step": 56667
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7370338439941406,
+      "learning_rate": 8.131179253105635e-06,
+      "loss": 2.8508,
+      "step": 56668
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7495514750480652,
+      "learning_rate": 8.127619379088102e-06,
+      "loss": 3.1002,
+      "step": 56669
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.6904610395431519,
+      "learning_rate": 8.124060273806242e-06,
+      "loss": 2.8653,
+      "step": 56670
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7376260757446289,
+      "learning_rate": 8.120501937269453e-06,
+      "loss": 2.8539,
+      "step": 56671
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7773568034172058,
+      "learning_rate": 8.116944369487088e-06,
+      "loss": 2.6758,
+      "step": 56672
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8228160738945007,
+      "learning_rate": 8.11338757046851e-06,
+      "loss": 2.7239,
+      "step": 56673
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7620078921318054,
+      "learning_rate": 8.109831540223077e-06,
+      "loss": 2.8486,
+      "step": 56674
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7939218282699585,
+      "learning_rate": 8.106276278760183e-06,
+      "loss": 2.8879,
+      "step": 56675
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7919220924377441,
+      "learning_rate": 8.102721786089183e-06,
+      "loss": 2.9234,
+      "step": 56676
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7480902671813965,
+      "learning_rate": 8.099168062219474e-06,
+      "loss": 2.7817,
+      "step": 56677
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7824358344078064,
+      "learning_rate": 8.095615107160347e-06,
+      "loss": 2.8962,
+      "step": 56678
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7388315200805664,
+      "learning_rate": 8.09206292092116e-06,
+      "loss": 2.9825,
+      "step": 56679
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.749704122543335,
+      "learning_rate": 8.088511503511342e-06,
+      "loss": 2.8423,
+      "step": 56680
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7328793406486511,
+      "learning_rate": 8.084960854940181e-06,
+      "loss": 2.9085,
+      "step": 56681
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7136931419372559,
+      "learning_rate": 8.081410975217073e-06,
+      "loss": 2.9693,
+      "step": 56682
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7360411286354065,
+      "learning_rate": 8.077861864351343e-06,
+      "loss": 2.7465,
+      "step": 56683
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7628071308135986,
+      "learning_rate": 8.074313522352315e-06,
+      "loss": 2.7308,
+      "step": 56684
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.72130286693573,
+      "learning_rate": 8.070765949229419e-06,
+      "loss": 2.6445,
+      "step": 56685
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7512112259864807,
+      "learning_rate": 8.06721914499191e-06,
+      "loss": 2.8764,
+      "step": 56686
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7311689853668213,
+      "learning_rate": 8.063673109649148e-06,
+      "loss": 3.1255,
+      "step": 56687
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.806206226348877,
+      "learning_rate": 8.060127843210528e-06,
+      "loss": 2.7643,
+      "step": 56688
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7583916187286377,
+      "learning_rate": 8.056583345685308e-06,
+      "loss": 3.0265,
+      "step": 56689
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7623338103294373,
+      "learning_rate": 8.053039617082846e-06,
+      "loss": 2.8745,
+      "step": 56690
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7402061820030212,
+      "learning_rate": 8.049496657412535e-06,
+      "loss": 2.9607,
+      "step": 56691
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7634629607200623,
+      "learning_rate": 8.045954466683668e-06,
+      "loss": 3.1609,
+      "step": 56692
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7831898927688599,
+      "learning_rate": 8.042413044905572e-06,
+      "loss": 2.6583,
+      "step": 56693
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7754256129264832,
+      "learning_rate": 8.038872392087537e-06,
+      "loss": 2.5636,
+      "step": 56694
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7307144999504089,
+      "learning_rate": 8.035332508238923e-06,
+      "loss": 2.8217,
+      "step": 56695
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7803258895874023,
+      "learning_rate": 8.031793393369057e-06,
+      "loss": 2.9664,
+      "step": 56696
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7296281456947327,
+      "learning_rate": 8.028255047487265e-06,
+      "loss": 2.8883,
+      "step": 56697
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7694997191429138,
+      "learning_rate": 8.02471747060287e-06,
+      "loss": 2.7386,
+      "step": 56698
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7115613222122192,
+      "learning_rate": 8.021180662725169e-06,
+      "loss": 2.8441,
+      "step": 56699
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7649816274642944,
+      "learning_rate": 8.017644623863484e-06,
+      "loss": 2.9336,
+      "step": 56700
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7806012630462646,
+      "learning_rate": 8.014109354027142e-06,
+      "loss": 2.9071,
+      "step": 56701
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7786357402801514,
+      "learning_rate": 8.010574853225405e-06,
+      "loss": 2.8436,
+      "step": 56702
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7318533658981323,
+      "learning_rate": 8.00704112146766e-06,
+      "loss": 2.9391,
+      "step": 56703
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7530311942100525,
+      "learning_rate": 8.003508158763138e-06,
+      "loss": 3.1211,
+      "step": 56704
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7505061626434326,
+      "learning_rate": 7.999975965121164e-06,
+      "loss": 3.1264,
+      "step": 56705
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7486022114753723,
+      "learning_rate": 7.996444540551128e-06,
+      "loss": 2.7554,
+      "step": 56706
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7808429002761841,
+      "learning_rate": 7.992913885062224e-06,
+      "loss": 2.8135,
+      "step": 56707
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7680838704109192,
+      "learning_rate": 7.989383998663812e-06,
+      "loss": 2.9188,
+      "step": 56708
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7538081407546997,
+      "learning_rate": 7.985854881365117e-06,
+      "loss": 2.7569,
+      "step": 56709
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7408961057662964,
+      "learning_rate": 7.982326533175466e-06,
+      "loss": 2.9055,
+      "step": 56710
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7339892983436584,
+      "learning_rate": 7.978798954104216e-06,
+      "loss": 2.8456,
+      "step": 56711
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7620311975479126,
+      "learning_rate": 7.975272144160594e-06,
+      "loss": 2.9099,
+      "step": 56712
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7628583908081055,
+      "learning_rate": 7.971746103353926e-06,
+      "loss": 2.7379,
+      "step": 56713
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7465094327926636,
+      "learning_rate": 7.968220831693473e-06,
+      "loss": 2.7516,
+      "step": 56714
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7672311067581177,
+      "learning_rate": 7.964696329188492e-06,
+      "loss": 2.8941,
+      "step": 56715
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7001157402992249,
+      "learning_rate": 7.96117259584831e-06,
+      "loss": 2.779,
+      "step": 56716
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7452120780944824,
+      "learning_rate": 7.957649631682183e-06,
+      "loss": 3.0356,
+      "step": 56717
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7730103731155396,
+      "learning_rate": 7.954127436699441e-06,
+      "loss": 2.8371,
+      "step": 56718
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7829596996307373,
+      "learning_rate": 7.950606010909277e-06,
+      "loss": 2.8766,
+      "step": 56719
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.733018696308136,
+      "learning_rate": 7.947085354321047e-06,
+      "loss": 2.9807,
+      "step": 56720
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7372322678565979,
+      "learning_rate": 7.94356546694398e-06,
+      "loss": 2.8052,
+      "step": 56721
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7484538555145264,
+      "learning_rate": 7.940046348787332e-06,
+      "loss": 2.9547,
+      "step": 56722
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7925270795822144,
+      "learning_rate": 7.936527999860432e-06,
+      "loss": 2.9496,
+      "step": 56723
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7163591980934143,
+      "learning_rate": 7.93301042017247e-06,
+      "loss": 2.7463,
+      "step": 56724
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.777164101600647,
+      "learning_rate": 7.92949360973274e-06,
+      "loss": 2.9165,
+      "step": 56725
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7506938576698303,
+      "learning_rate": 7.925977568550568e-06,
+      "loss": 2.9781,
+      "step": 56726
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7980271577835083,
+      "learning_rate": 7.922462296635112e-06,
+      "loss": 2.6755,
+      "step": 56727
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7719228863716125,
+      "learning_rate": 7.9189477939957e-06,
+      "loss": 2.7433,
+      "step": 56728
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7837870121002197,
+      "learning_rate": 7.91543406064159e-06,
+      "loss": 2.8218,
+      "step": 56729
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7866212725639343,
+      "learning_rate": 7.911921096581975e-06,
+      "loss": 2.895,
+      "step": 56730
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7204878926277161,
+      "learning_rate": 7.908408901826147e-06,
+      "loss": 2.9522,
+      "step": 56731
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7222285270690918,
+      "learning_rate": 7.904897476383365e-06,
+      "loss": 2.9197,
+      "step": 56732
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7325133085250854,
+      "learning_rate": 7.90138682026289e-06,
+      "loss": 2.6879,
+      "step": 56733
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7699395418167114,
+      "learning_rate": 7.897876933473912e-06,
+      "loss": 2.8361,
+      "step": 56734
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7754160165786743,
+      "learning_rate": 7.894367816025726e-06,
+      "loss": 2.81,
+      "step": 56735
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7943594455718994,
+      "learning_rate": 7.89085946792759e-06,
+      "loss": 2.8848,
+      "step": 56736
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7334774136543274,
+      "learning_rate": 7.887351889188665e-06,
+      "loss": 2.7652,
+      "step": 56737
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7373555898666382,
+      "learning_rate": 7.883845079818273e-06,
+      "loss": 3.0028,
+      "step": 56738
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7737788558006287,
+      "learning_rate": 7.880339039825578e-06,
+      "loss": 2.7939,
+      "step": 56739
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7455878257751465,
+      "learning_rate": 7.87683376921987e-06,
+      "loss": 2.8291,
+      "step": 56740
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7712538242340088,
+      "learning_rate": 7.873329268010343e-06,
+      "loss": 3.0044,
+      "step": 56741
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7128819227218628,
+      "learning_rate": 7.869825536206254e-06,
+      "loss": 2.7565,
+      "step": 56742
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.72657710313797,
+      "learning_rate": 7.866322573816864e-06,
+      "loss": 2.8434,
+      "step": 56743
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7716096639633179,
+      "learning_rate": 7.862820380851299e-06,
+      "loss": 2.961,
+      "step": 56744
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7939549684524536,
+      "learning_rate": 7.859318957318884e-06,
+      "loss": 2.9316,
+      "step": 56745
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7576471567153931,
+      "learning_rate": 7.855818303228779e-06,
+      "loss": 2.9361,
+      "step": 56746
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7495378255844116,
+      "learning_rate": 7.852318418590208e-06,
+      "loss": 2.9346,
+      "step": 56747
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7718560099601746,
+      "learning_rate": 7.848819303412435e-06,
+      "loss": 2.8327,
+      "step": 56748
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7532568573951721,
+      "learning_rate": 7.845320957704614e-06,
+      "loss": 2.7433,
+      "step": 56749
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7588929533958435,
+      "learning_rate": 7.841823381476043e-06,
+      "loss": 2.962,
+      "step": 56750
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7554545998573303,
+      "learning_rate": 7.838326574735875e-06,
+      "loss": 3.0306,
+      "step": 56751
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7670568227767944,
+      "learning_rate": 7.834830537493275e-06,
+      "loss": 2.9727,
+      "step": 56752
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7550936341285706,
+      "learning_rate": 7.831335269757532e-06,
+      "loss": 2.9098,
+      "step": 56753
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7789241075515747,
+      "learning_rate": 7.827840771537808e-06,
+      "loss": 2.7227,
+      "step": 56754
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7854425311088562,
+      "learning_rate": 7.82434704284336e-06,
+      "loss": 2.8727,
+      "step": 56755
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7839749455451965,
+      "learning_rate": 7.820854083683314e-06,
+      "loss": 2.9157,
+      "step": 56756
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7156407833099365,
+      "learning_rate": 7.81736189406693e-06,
+      "loss": 3.0155,
+      "step": 56757
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7155882716178894,
+      "learning_rate": 7.813870474003403e-06,
+      "loss": 2.8871,
+      "step": 56758
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7472594380378723,
+      "learning_rate": 7.810379823501855e-06,
+      "loss": 3.0024,
+      "step": 56759
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7841302156448364,
+      "learning_rate": 7.806889942571582e-06,
+      "loss": 2.7603,
+      "step": 56760
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7454701662063599,
+      "learning_rate": 7.803400831221707e-06,
+      "loss": 2.7723,
+      "step": 56761
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7371299862861633,
+      "learning_rate": 7.799912489461423e-06,
+      "loss": 2.9579,
+      "step": 56762
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7872686386108398,
+      "learning_rate": 7.796424917299959e-06,
+      "loss": 2.9897,
+      "step": 56763
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7827180624008179,
+      "learning_rate": 7.792938114746439e-06,
+      "loss": 3.0709,
+      "step": 56764
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7246116399765015,
+      "learning_rate": 7.78945208181012e-06,
+      "loss": 2.8398,
+      "step": 56765
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7627786993980408,
+      "learning_rate": 7.785966818500166e-06,
+      "loss": 3.0078,
+      "step": 56766
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7191957831382751,
+      "learning_rate": 7.7824823248257e-06,
+      "loss": 3.119,
+      "step": 56767
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7956440448760986,
+      "learning_rate": 7.778998600795949e-06,
+      "loss": 3.0264,
+      "step": 56768
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7572888135910034,
+      "learning_rate": 7.775515646420071e-06,
+      "loss": 2.9061,
+      "step": 56769
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7518506646156311,
+      "learning_rate": 7.772033461707261e-06,
+      "loss": 2.9363,
+      "step": 56770
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7562054991722107,
+      "learning_rate": 7.768552046666642e-06,
+      "loss": 2.9044,
+      "step": 56771
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7331022024154663,
+      "learning_rate": 7.765071401307443e-06,
+      "loss": 2.8371,
+      "step": 56772
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7433456778526306,
+      "learning_rate": 7.761591525638822e-06,
+      "loss": 2.8736,
+      "step": 56773
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7637978196144104,
+      "learning_rate": 7.75811241966987e-06,
+      "loss": 2.9288,
+      "step": 56774
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7655687928199768,
+      "learning_rate": 7.75463408340985e-06,
+      "loss": 2.8614,
+      "step": 56775
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7311526536941528,
+      "learning_rate": 7.751156516867884e-06,
+      "loss": 2.7973,
+      "step": 56776
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7516604661941528,
+      "learning_rate": 7.747679720053069e-06,
+      "loss": 2.8696,
+      "step": 56777
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7627346515655518,
+      "learning_rate": 7.744203692974692e-06,
+      "loss": 2.7741,
+      "step": 56778
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8112263679504395,
+      "learning_rate": 7.740728435641818e-06,
+      "loss": 2.8013,
+      "step": 56779
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7233673930168152,
+      "learning_rate": 7.737253948063638e-06,
+      "loss": 3.0573,
+      "step": 56780
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7762109637260437,
+      "learning_rate": 7.733780230249243e-06,
+      "loss": 2.8076,
+      "step": 56781
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7639904618263245,
+      "learning_rate": 7.730307282207794e-06,
+      "loss": 2.7387,
+      "step": 56782
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7313291430473328,
+      "learning_rate": 7.726835103948548e-06,
+      "loss": 2.7993,
+      "step": 56783
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7252292633056641,
+      "learning_rate": 7.7233636954805e-06,
+      "loss": 2.8332,
+      "step": 56784
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7580043077468872,
+      "learning_rate": 7.719893056812908e-06,
+      "loss": 2.9502,
+      "step": 56785
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7498751878738403,
+      "learning_rate": 7.716423187954835e-06,
+      "loss": 2.8644,
+      "step": 56786
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7402743697166443,
+      "learning_rate": 7.712954088915469e-06,
+      "loss": 2.7173,
+      "step": 56787
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7868022322654724,
+      "learning_rate": 7.709485759703971e-06,
+      "loss": 2.769,
+      "step": 56788
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.762657880783081,
+      "learning_rate": 7.706018200329367e-06,
+      "loss": 2.8656,
+      "step": 56789
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8196261525154114,
+      "learning_rate": 7.702551410800884e-06,
+      "loss": 2.8877,
+      "step": 56790
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7575928568840027,
+      "learning_rate": 7.699085391127614e-06,
+      "loss": 2.9055,
+      "step": 56791
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7336806654930115,
+      "learning_rate": 7.695620141318715e-06,
+      "loss": 2.9307,
+      "step": 56792
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7909324765205383,
+      "learning_rate": 7.692155661383314e-06,
+      "loss": 2.8882,
+      "step": 56793
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.778021514415741,
+      "learning_rate": 7.688691951330506e-06,
+      "loss": 2.8321,
+      "step": 56794
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7695522904396057,
+      "learning_rate": 7.685229011169447e-06,
+      "loss": 2.7739,
+      "step": 56795
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7543647289276123,
+      "learning_rate": 7.681766840909197e-06,
+      "loss": 2.9972,
+      "step": 56796
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7531213164329529,
+      "learning_rate": 7.678305440558884e-06,
+      "loss": 2.674,
+      "step": 56797
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7810092568397522,
+      "learning_rate": 7.674844810127734e-06,
+      "loss": 2.8329,
+      "step": 56798
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7242538332939148,
+      "learning_rate": 7.671384949624736e-06,
+      "loss": 2.9462,
+      "step": 56799
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7586749792098999,
+      "learning_rate": 7.667925859059087e-06,
+      "loss": 2.7034,
+      "step": 56800
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.757546067237854,
+      "learning_rate": 7.664467538439845e-06,
+      "loss": 2.7994,
+      "step": 56801
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7679392695426941,
+      "learning_rate": 7.6610099877761e-06,
+      "loss": 2.8346,
+      "step": 56802
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7457258701324463,
+      "learning_rate": 7.657553207077049e-06,
+      "loss": 2.8294,
+      "step": 56803
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.734887957572937,
+      "learning_rate": 7.654097196351716e-06,
+      "loss": 2.8035,
+      "step": 56804
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7640114426612854,
+      "learning_rate": 7.650641955609227e-06,
+      "loss": 2.8279,
+      "step": 56805
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7584049105644226,
+      "learning_rate": 7.647187484858674e-06,
+      "loss": 3.168,
+      "step": 56806
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7257455587387085,
+      "learning_rate": 7.643733784109151e-06,
+      "loss": 2.6585,
+      "step": 56807
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7480292916297913,
+      "learning_rate": 7.640280853369784e-06,
+      "loss": 2.7445,
+      "step": 56808
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7356035709381104,
+      "learning_rate": 7.6368286926497e-06,
+      "loss": 2.9441,
+      "step": 56809
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.719250500202179,
+      "learning_rate": 7.633377301957889e-06,
+      "loss": 2.8578,
+      "step": 56810
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7773850560188293,
+      "learning_rate": 7.629926681303512e-06,
+      "loss": 2.7369,
+      "step": 56811
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7469416856765747,
+      "learning_rate": 7.626476830695627e-06,
+      "loss": 2.9075,
+      "step": 56812
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.768189549446106,
+      "learning_rate": 7.623027750143329e-06,
+      "loss": 2.776,
+      "step": 56813
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7666859030723572,
+      "learning_rate": 7.61957943965571e-06,
+      "loss": 2.8001,
+      "step": 56814
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.734056293964386,
+      "learning_rate": 7.616131899241862e-06,
+      "loss": 2.8107,
+      "step": 56815
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7309668064117432,
+      "learning_rate": 7.612685128910878e-06,
+      "loss": 3.0441,
+      "step": 56816
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7235714793205261,
+      "learning_rate": 7.609239128671785e-06,
+      "loss": 2.8479,
+      "step": 56817
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8136651515960693,
+      "learning_rate": 7.6057938985336746e-06,
+      "loss": 2.8428,
+      "step": 56818
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7454987168312073,
+      "learning_rate": 7.602349438505606e-06,
+      "loss": 2.7822,
+      "step": 56819
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.775428831577301,
+      "learning_rate": 7.59890574859674e-06,
+      "loss": 2.9205,
+      "step": 56820
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.82244873046875,
+      "learning_rate": 7.595462828816035e-06,
+      "loss": 2.9713,
+      "step": 56821
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7343226075172424,
+      "learning_rate": 7.5920206791726165e-06,
+      "loss": 2.7359,
+      "step": 56822
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7130812406539917,
+      "learning_rate": 7.588579299675579e-06,
+      "loss": 3.0358,
+      "step": 56823
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.752113401889801,
+      "learning_rate": 7.585138690333881e-06,
+      "loss": 2.9996,
+      "step": 56824
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7490484118461609,
+      "learning_rate": 7.581698851156715e-06,
+      "loss": 2.9933,
+      "step": 56825
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7558857202529907,
+      "learning_rate": 7.578259782153006e-06,
+      "loss": 2.846,
+      "step": 56826
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7821259498596191,
+      "learning_rate": 7.57482148333195e-06,
+      "loss": 2.5186,
+      "step": 56827
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7697935700416565,
+      "learning_rate": 7.571383954702504e-06,
+      "loss": 2.9788,
+      "step": 56828
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7339929342269897,
+      "learning_rate": 7.567947196273727e-06,
+      "loss": 3.0044,
+      "step": 56829
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.724521815776825,
+      "learning_rate": 7.5645112080547464e-06,
+      "loss": 2.8909,
+      "step": 56830
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7704073786735535,
+      "learning_rate": 7.561075990054555e-06,
+      "loss": 3.2447,
+      "step": 56831
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7758163213729858,
+      "learning_rate": 7.5576415422822115e-06,
+      "loss": 2.9757,
+      "step": 56832
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7906506657600403,
+      "learning_rate": 7.554207864746742e-06,
+      "loss": 2.8299,
+      "step": 56833
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7277704477310181,
+      "learning_rate": 7.550774957457173e-06,
+      "loss": 2.7329,
+      "step": 56834
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7903933525085449,
+      "learning_rate": 7.54734282042263e-06,
+      "loss": 2.7854,
+      "step": 56835
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7526589632034302,
+      "learning_rate": 7.543911453652074e-06,
+      "loss": 2.937,
+      "step": 56836
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.790673017501831,
+      "learning_rate": 7.540480857154596e-06,
+      "loss": 3.0866,
+      "step": 56837
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7465153336524963,
+      "learning_rate": 7.537051030939223e-06,
+      "loss": 2.8482,
+      "step": 56838
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7059990167617798,
+      "learning_rate": 7.5336219750149475e-06,
+      "loss": 2.9167,
+      "step": 56839
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7449349164962769,
+      "learning_rate": 7.530193689390828e-06,
+      "loss": 2.6752,
+      "step": 56840
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7419530153274536,
+      "learning_rate": 7.52676617407586e-06,
+      "loss": 2.9918,
+      "step": 56841
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7866668105125427,
+      "learning_rate": 7.5233394290791334e-06,
+      "loss": 2.8204,
+      "step": 56842
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8241013288497925,
+      "learning_rate": 7.519913454409643e-06,
+      "loss": 2.8006,
+      "step": 56843
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.793013334274292,
+      "learning_rate": 7.516488250076414e-06,
+      "loss": 2.8997,
+      "step": 56844
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7368309497833252,
+      "learning_rate": 7.513063816088472e-06,
+      "loss": 2.9748,
+      "step": 56845
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7466829419136047,
+      "learning_rate": 7.50964015245481e-06,
+      "loss": 2.785,
+      "step": 56846
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7598251104354858,
+      "learning_rate": 7.506217259184488e-06,
+      "loss": 2.7236,
+      "step": 56847
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7561330199241638,
+      "learning_rate": 7.502795136286465e-06,
+      "loss": 2.999,
+      "step": 56848
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7700143456459045,
+      "learning_rate": 7.499373783769768e-06,
+      "loss": 2.8617,
+      "step": 56849
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7483510971069336,
+      "learning_rate": 7.49595320164349e-06,
+      "loss": 2.8559,
+      "step": 56850
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7109506726264954,
+      "learning_rate": 7.4925333899165216e-06,
+      "loss": 2.8932,
+      "step": 56851
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7281533479690552,
+      "learning_rate": 7.489114348597991e-06,
+      "loss": 3.0285,
+      "step": 56852
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7932998538017273,
+      "learning_rate": 7.48569607769679e-06,
+      "loss": 2.87,
+      "step": 56853
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7219394445419312,
+      "learning_rate": 7.4822785772219786e-06,
+      "loss": 2.8928,
+      "step": 56854
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7718573808670044,
+      "learning_rate": 7.478861847182582e-06,
+      "loss": 2.8527,
+      "step": 56855
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7460842728614807,
+      "learning_rate": 7.475445887587528e-06,
+      "loss": 2.8612,
+      "step": 56856
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7437963485717773,
+      "learning_rate": 7.472030698445875e-06,
+      "loss": 2.8628,
+      "step": 56857
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7210949659347534,
+      "learning_rate": 7.468616279766548e-06,
+      "loss": 2.7437,
+      "step": 56858
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.6923820376396179,
+      "learning_rate": 7.465202631558642e-06,
+      "loss": 2.7774,
+      "step": 56859
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7626507878303528,
+      "learning_rate": 7.461789753831082e-06,
+      "loss": 2.8715,
+      "step": 56860
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7916741371154785,
+      "learning_rate": 7.458377646592828e-06,
+      "loss": 2.9131,
+      "step": 56861
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7709150314331055,
+      "learning_rate": 7.454966309852972e-06,
+      "loss": 2.7865,
+      "step": 56862
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7656967639923096,
+      "learning_rate": 7.4515557436203746e-06,
+      "loss": 2.9982,
+      "step": 56863
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7165960669517517,
+      "learning_rate": 7.448145947904094e-06,
+      "loss": 2.8934,
+      "step": 56864
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7551760673522949,
+      "learning_rate": 7.444736922713091e-06,
+      "loss": 2.9164,
+      "step": 56865
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7682600021362305,
+      "learning_rate": 7.441328668056357e-06,
+      "loss": 2.9947,
+      "step": 56866
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7713602781295776,
+      "learning_rate": 7.4379211839428856e-06,
+      "loss": 2.8302,
+      "step": 56867
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7858055830001831,
+      "learning_rate": 7.434514470381603e-06,
+      "loss": 2.838,
+      "step": 56868
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7591269016265869,
+      "learning_rate": 7.431108527381469e-06,
+      "loss": 2.764,
+      "step": 56869
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7732019424438477,
+      "learning_rate": 7.427703354951542e-06,
+      "loss": 3.1269,
+      "step": 56870
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7430728077888489,
+      "learning_rate": 7.424298953100716e-06,
+      "loss": 2.8704,
+      "step": 56871
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7464427947998047,
+      "learning_rate": 7.420895321837983e-06,
+      "loss": 2.9217,
+      "step": 56872
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7523966431617737,
+      "learning_rate": 7.41749246117227e-06,
+      "loss": 2.8089,
+      "step": 56873
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7557439804077148,
+      "learning_rate": 7.414090371112636e-06,
+      "loss": 2.7274,
+      "step": 56874
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7846986651420593,
+      "learning_rate": 7.41068905166794e-06,
+      "loss": 2.8855,
+      "step": 56875
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7404799461364746,
+      "learning_rate": 7.4072885028471755e-06,
+      "loss": 2.7038,
+      "step": 56876
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7172025442123413,
+      "learning_rate": 7.403888724659334e-06,
+      "loss": 2.9229,
+      "step": 56877
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7421782612800598,
+      "learning_rate": 7.400489717113278e-06,
+      "loss": 2.8372,
+      "step": 56878
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7487596869468689,
+      "learning_rate": 7.397091480218065e-06,
+      "loss": 2.6678,
+      "step": 56879
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.753854513168335,
+      "learning_rate": 7.393694013982587e-06,
+      "loss": 2.9184,
+      "step": 56880
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7490181922912598,
+      "learning_rate": 7.390297318415839e-06,
+      "loss": 2.9672,
+      "step": 56881
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7417166233062744,
+      "learning_rate": 7.386901393526712e-06,
+      "loss": 2.9964,
+      "step": 56882
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7824398875236511,
+      "learning_rate": 7.383506239324166e-06,
+      "loss": 2.9129,
+      "step": 56883
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7355906367301941,
+      "learning_rate": 7.380111855817128e-06,
+      "loss": 2.8434,
+      "step": 56884
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7218039035797119,
+      "learning_rate": 7.3767182430145895e-06,
+      "loss": 2.8935,
+      "step": 56885
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7684163451194763,
+      "learning_rate": 7.373325400925445e-06,
+      "loss": 2.976,
+      "step": 56886
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7165760397911072,
+      "learning_rate": 7.3699333295586526e-06,
+      "loss": 2.6787,
+      "step": 56887
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7151129841804504,
+      "learning_rate": 7.366542028923139e-06,
+      "loss": 2.7651,
+      "step": 56888
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7509085536003113,
+      "learning_rate": 7.3631514990278305e-06,
+      "loss": 3.0883,
+      "step": 56889
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7478203773498535,
+      "learning_rate": 7.359761739881686e-06,
+      "loss": 3.0038,
+      "step": 56890
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.77152019739151,
+      "learning_rate": 7.3563727514935665e-06,
+      "loss": 2.804,
+      "step": 56891
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7167760729789734,
+      "learning_rate": 7.352984533872464e-06,
+      "loss": 3.1426,
+      "step": 56892
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7409500479698181,
+      "learning_rate": 7.34959708702727e-06,
+      "loss": 2.9238,
+      "step": 56893
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8209675550460815,
+      "learning_rate": 7.346210410966946e-06,
+      "loss": 2.8103,
+      "step": 56894
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7210500836372375,
+      "learning_rate": 7.342824505700351e-06,
+      "loss": 2.6596,
+      "step": 56895
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7498226761817932,
+      "learning_rate": 7.339439371236444e-06,
+      "loss": 2.8775,
+      "step": 56896
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7758424282073975,
+      "learning_rate": 7.336055007584152e-06,
+      "loss": 2.7742,
+      "step": 56897
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7447173595428467,
+      "learning_rate": 7.3326714147523e-06,
+      "loss": 2.8494,
+      "step": 56898
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.837697446346283,
+      "learning_rate": 7.329288592749916e-06,
+      "loss": 2.8712,
+      "step": 56899
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7390157580375671,
+      "learning_rate": 7.3259065415858575e-06,
+      "loss": 2.8616,
+      "step": 56900
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.735798180103302,
+      "learning_rate": 7.322525261269019e-06,
+      "loss": 2.9517,
+      "step": 56901
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7630640268325806,
+      "learning_rate": 7.319144751808325e-06,
+      "loss": 2.9783,
+      "step": 56902
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7421290278434753,
+      "learning_rate": 7.315765013212705e-06,
+      "loss": 2.8817,
+      "step": 56903
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7206147313117981,
+      "learning_rate": 7.312386045491015e-06,
+      "loss": 2.7915,
+      "step": 56904
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7598541975021362,
+      "learning_rate": 7.309007848652149e-06,
+      "loss": 2.647,
+      "step": 56905
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7555434703826904,
+      "learning_rate": 7.305630422705033e-06,
+      "loss": 2.7925,
+      "step": 56906
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7854859232902527,
+      "learning_rate": 7.3022537676585944e-06,
+      "loss": 2.8326,
+      "step": 56907
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7405829429626465,
+      "learning_rate": 7.298877883521626e-06,
+      "loss": 2.8358,
+      "step": 56908
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7246938347816467,
+      "learning_rate": 7.295502770303152e-06,
+      "loss": 2.6278,
+      "step": 56909
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7304655313491821,
+      "learning_rate": 7.292128428011934e-06,
+      "loss": 2.824,
+      "step": 56910
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7219448685646057,
+      "learning_rate": 7.288754856656964e-06,
+      "loss": 3.1004,
+      "step": 56911
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7531955242156982,
+      "learning_rate": 7.285382056247069e-06,
+      "loss": 2.8265,
+      "step": 56912
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7347313761711121,
+      "learning_rate": 7.282010026791107e-06,
+      "loss": 2.758,
+      "step": 56913
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.740077555179596,
+      "learning_rate": 7.27863876829804e-06,
+      "loss": 2.9192,
+      "step": 56914
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7309461832046509,
+      "learning_rate": 7.275268280776659e-06,
+      "loss": 2.8396,
+      "step": 56915
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7398518919944763,
+      "learning_rate": 7.2718985642358894e-06,
+      "loss": 2.9457,
+      "step": 56916
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7491543292999268,
+      "learning_rate": 7.268529618684627e-06,
+      "loss": 2.9195,
+      "step": 56917
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7731919884681702,
+      "learning_rate": 7.265161444131729e-06,
+      "loss": 3.1226,
+      "step": 56918
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.802014946937561,
+      "learning_rate": 7.261794040586055e-06,
+      "loss": 2.9917,
+      "step": 56919
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7600812315940857,
+      "learning_rate": 7.258427408056467e-06,
+      "loss": 2.9308,
+      "step": 56920
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7489023804664612,
+      "learning_rate": 7.255061546551821e-06,
+      "loss": 3.0007,
+      "step": 56921
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7434468269348145,
+      "learning_rate": 7.251696456081046e-06,
+      "loss": 2.9466,
+      "step": 56922
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7413731813430786,
+      "learning_rate": 7.248332136652901e-06,
+      "loss": 3.0299,
+      "step": 56923
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7570599913597107,
+      "learning_rate": 7.2449685882763784e-06,
+      "loss": 2.9096,
+      "step": 56924
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7353522777557373,
+      "learning_rate": 7.241605810960238e-06,
+      "loss": 2.7581,
+      "step": 56925
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7793547511100769,
+      "learning_rate": 7.2382438047133395e-06,
+      "loss": 3.1442,
+      "step": 56926
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7294006943702698,
+      "learning_rate": 7.234882569544609e-06,
+      "loss": 3.0892,
+      "step": 56927
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7296475172042847,
+      "learning_rate": 7.231522105462806e-06,
+      "loss": 2.8816,
+      "step": 56928
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8307911157608032,
+      "learning_rate": 7.22816241247689e-06,
+      "loss": 2.9092,
+      "step": 56929
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.744428277015686,
+      "learning_rate": 7.224803490595587e-06,
+      "loss": 2.8779,
+      "step": 56930
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.758772075176239,
+      "learning_rate": 7.221445339827825e-06,
+      "loss": 2.9967,
+      "step": 56931
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7177093029022217,
+      "learning_rate": 7.218087960182428e-06,
+      "loss": 2.8032,
+      "step": 56932
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7423251867294312,
+      "learning_rate": 7.21473135166829e-06,
+      "loss": 2.9131,
+      "step": 56933
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7697576880455017,
+      "learning_rate": 7.211375514294171e-06,
+      "loss": 3.128,
+      "step": 56934
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7272599935531616,
+      "learning_rate": 7.208020448068897e-06,
+      "loss": 2.6826,
+      "step": 56935
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7582986354827881,
+      "learning_rate": 7.204666153001393e-06,
+      "loss": 2.9562,
+      "step": 56936
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.757316529750824,
+      "learning_rate": 7.201312629100453e-06,
+      "loss": 2.9547,
+      "step": 56937
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7660689949989319,
+      "learning_rate": 7.197959876374871e-06,
+      "loss": 2.8556,
+      "step": 56938
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7317332625389099,
+      "learning_rate": 7.194607894833537e-06,
+      "loss": 2.8184,
+      "step": 56939
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7949168682098389,
+      "learning_rate": 7.1912566844852475e-06,
+      "loss": 3.0257,
+      "step": 56940
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7534281611442566,
+      "learning_rate": 7.187906245338826e-06,
+      "loss": 2.8538,
+      "step": 56941
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7638459205627441,
+      "learning_rate": 7.184556577403134e-06,
+      "loss": 2.8792,
+      "step": 56942
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.761776864528656,
+      "learning_rate": 7.181207680686929e-06,
+      "loss": 2.9767,
+      "step": 56943
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7768969535827637,
+      "learning_rate": 7.1778595551991055e-06,
+      "loss": 2.87,
+      "step": 56944
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7668014168739319,
+      "learning_rate": 7.1745122009484235e-06,
+      "loss": 2.8332,
+      "step": 56945
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7426735162734985,
+      "learning_rate": 7.171165617943741e-06,
+      "loss": 2.8732,
+      "step": 56946
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7256327867507935,
+      "learning_rate": 7.167819806193853e-06,
+      "loss": 2.9715,
+      "step": 56947
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8112422227859497,
+      "learning_rate": 7.164474765707518e-06,
+      "loss": 2.9562,
+      "step": 56948
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7298985719680786,
+      "learning_rate": 7.161130496493661e-06,
+      "loss": 2.8423,
+      "step": 56949
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7860491275787354,
+      "learning_rate": 7.157786998560977e-06,
+      "loss": 2.8787,
+      "step": 56950
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.815058708190918,
+      "learning_rate": 7.154444271918358e-06,
+      "loss": 2.8567,
+      "step": 56951
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7936457395553589,
+      "learning_rate": 7.151102316574564e-06,
+      "loss": 2.978,
+      "step": 56952
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7552315592765808,
+      "learning_rate": 7.147761132538387e-06,
+      "loss": 2.8227,
+      "step": 56953
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7764397263526917,
+      "learning_rate": 7.144420719818689e-06,
+      "loss": 3.0444,
+      "step": 56954
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7704004049301147,
+      "learning_rate": 7.141081078424227e-06,
+      "loss": 2.8896,
+      "step": 56955
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.749974250793457,
+      "learning_rate": 7.137742208363762e-06,
+      "loss": 2.7538,
+      "step": 56956
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7289097309112549,
+      "learning_rate": 7.1344041096461525e-06,
+      "loss": 2.9296,
+      "step": 56957
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7585612535476685,
+      "learning_rate": 7.131066782280126e-06,
+      "loss": 3.0644,
+      "step": 56958
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7690691947937012,
+      "learning_rate": 7.127730226274509e-06,
+      "loss": 3.0722,
+      "step": 56959
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7478421330451965,
+      "learning_rate": 7.124394441638092e-06,
+      "loss": 2.7564,
+      "step": 56960
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7582880258560181,
+      "learning_rate": 7.121059428379672e-06,
+      "loss": 2.7356,
+      "step": 56961
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7731198668479919,
+      "learning_rate": 7.117725186508006e-06,
+      "loss": 2.8882,
+      "step": 56962
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7167331576347351,
+      "learning_rate": 7.114391716031887e-06,
+      "loss": 2.9934,
+      "step": 56963
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7596493363380432,
+      "learning_rate": 7.111059016960108e-06,
+      "loss": 3.0592,
+      "step": 56964
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7319817543029785,
+      "learning_rate": 7.107727089301396e-06,
+      "loss": 3.0243,
+      "step": 56965
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7279293537139893,
+      "learning_rate": 7.1043959330645774e-06,
+      "loss": 2.96,
+      "step": 56966
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7954239249229431,
+      "learning_rate": 7.101065548258411e-06,
+      "loss": 2.9018,
+      "step": 56967
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7724497318267822,
+      "learning_rate": 7.097735934891657e-06,
+      "loss": 3.1738,
+      "step": 56968
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.753320574760437,
+      "learning_rate": 7.094407092973142e-06,
+      "loss": 2.981,
+      "step": 56969
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8049317002296448,
+      "learning_rate": 7.091079022511559e-06,
+      "loss": 2.9705,
+      "step": 56970
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7446392774581909,
+      "learning_rate": 7.0877517235157e-06,
+      "loss": 2.6321,
+      "step": 56971
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7473111152648926,
+      "learning_rate": 7.0844251959943255e-06,
+      "loss": 2.8549,
+      "step": 56972
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7095595598220825,
+      "learning_rate": 7.081099439956195e-06,
+      "loss": 2.7745,
+      "step": 56973
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7316811084747314,
+      "learning_rate": 7.077774455410102e-06,
+      "loss": 2.6908,
+      "step": 56974
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7425003051757812,
+      "learning_rate": 7.074450242364738e-06,
+      "loss": 2.8079,
+      "step": 56975
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7391384840011597,
+      "learning_rate": 7.07112680082893e-06,
+      "loss": 2.8774,
+      "step": 56976
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.763640820980072,
+      "learning_rate": 7.067804130811405e-06,
+      "loss": 2.8268,
+      "step": 56977
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7273032069206238,
+      "learning_rate": 7.06448223232089e-06,
+      "loss": 2.7182,
+      "step": 56978
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7752225399017334,
+      "learning_rate": 7.061161105366176e-06,
+      "loss": 2.8235,
+      "step": 56979
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.755729615688324,
+      "learning_rate": 7.057840749955956e-06,
+      "loss": 2.8752,
+      "step": 56980
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7277241349220276,
+      "learning_rate": 7.054521166099026e-06,
+      "loss": 2.7167,
+      "step": 56981
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7544986605644226,
+      "learning_rate": 7.0512023538040756e-06,
+      "loss": 2.9328,
+      "step": 56982
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7631686925888062,
+      "learning_rate": 7.047884313079899e-06,
+      "loss": 2.9111,
+      "step": 56983
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7494498491287231,
+      "learning_rate": 7.044567043935257e-06,
+      "loss": 2.8915,
+      "step": 56984
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7393820285797119,
+      "learning_rate": 7.0412505463787734e-06,
+      "loss": 3.0258,
+      "step": 56985
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7384428381919861,
+      "learning_rate": 7.037934820419311e-06,
+      "loss": 2.8008,
+      "step": 56986
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7311258316040039,
+      "learning_rate": 7.034619866065494e-06,
+      "loss": 2.9212,
+      "step": 56987
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7373533248901367,
+      "learning_rate": 7.0313056833261495e-06,
+      "loss": 2.8112,
+      "step": 56988
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7468560338020325,
+      "learning_rate": 7.0279922722099365e-06,
+      "loss": 2.9788,
+      "step": 56989
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7494136691093445,
+      "learning_rate": 7.024679632725616e-06,
+      "loss": 2.9388,
+      "step": 56990
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.731553852558136,
+      "learning_rate": 7.0213677648819135e-06,
+      "loss": 2.7017,
+      "step": 56991
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8054585456848145,
+      "learning_rate": 7.018056668687555e-06,
+      "loss": 2.8854,
+      "step": 56992
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7437418103218079,
+      "learning_rate": 7.014746344151234e-06,
+      "loss": 2.8213,
+      "step": 56993
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7302021980285645,
+      "learning_rate": 7.01143679128171e-06,
+      "loss": 2.8911,
+      "step": 56994
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7534018158912659,
+      "learning_rate": 7.00812801008761e-06,
+      "loss": 2.8335,
+      "step": 56995
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7726359367370605,
+      "learning_rate": 7.004820000577793e-06,
+      "loss": 2.9413,
+      "step": 56996
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.748771071434021,
+      "learning_rate": 7.0015127627608195e-06,
+      "loss": 2.7926,
+      "step": 56997
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7633996605873108,
+      "learning_rate": 6.998206296645515e-06,
+      "loss": 2.8268,
+      "step": 56998
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7452352643013,
+      "learning_rate": 6.994900602240572e-06,
+      "loss": 2.7504,
+      "step": 56999
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7363735437393188,
+      "learning_rate": 6.991595679554618e-06,
+      "loss": 3.0628,
+      "step": 57000
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7469814419746399,
+      "learning_rate": 6.988291528596446e-06,
+      "loss": 3.0343,
+      "step": 57001
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7548006176948547,
+      "learning_rate": 6.984988149374682e-06,
+      "loss": 3.1684,
+      "step": 57002
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7495371699333191,
+      "learning_rate": 6.981685541898086e-06,
+      "loss": 2.9532,
+      "step": 57003
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7482824325561523,
+      "learning_rate": 6.97838370617535e-06,
+      "loss": 3.0097,
+      "step": 57004
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7665938138961792,
+      "learning_rate": 6.975082642215168e-06,
+      "loss": 2.7349,
+      "step": 57005
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7205333113670349,
+      "learning_rate": 6.9717823500262e-06,
+      "loss": 2.8713,
+      "step": 57006
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7368008494377136,
+      "learning_rate": 6.9684828296171725e-06,
+      "loss": 2.7659,
+      "step": 57007
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7547926306724548,
+      "learning_rate": 6.965184080996744e-06,
+      "loss": 2.9229,
+      "step": 57008
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7362197041511536,
+      "learning_rate": 6.9618861041736415e-06,
+      "loss": 2.6904,
+      "step": 57009
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7418313026428223,
+      "learning_rate": 6.958588899156525e-06,
+      "loss": 2.8835,
+      "step": 57010
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7631124258041382,
+      "learning_rate": 6.955292465954121e-06,
+      "loss": 2.7645,
+      "step": 57011
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7663556337356567,
+      "learning_rate": 6.951996804575055e-06,
+      "loss": 2.9622,
+      "step": 57012
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7453634738922119,
+      "learning_rate": 6.948701915028054e-06,
+      "loss": 2.8942,
+      "step": 57013
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7672542333602905,
+      "learning_rate": 6.9454077973217444e-06,
+      "loss": 2.9699,
+      "step": 57014
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7669967412948608,
+      "learning_rate": 6.9421144514648195e-06,
+      "loss": 2.8731,
+      "step": 57015
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7587398886680603,
+      "learning_rate": 6.938821877466005e-06,
+      "loss": 2.8788,
+      "step": 57016
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7381514310836792,
+      "learning_rate": 6.935530075333895e-06,
+      "loss": 2.9416,
+      "step": 57017
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7473018765449524,
+      "learning_rate": 6.932239045077215e-06,
+      "loss": 3.012,
+      "step": 57018
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7571165561676025,
+      "learning_rate": 6.928948786704624e-06,
+      "loss": 2.8817,
+      "step": 57019
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7279015183448792,
+      "learning_rate": 6.925659300224784e-06,
+      "loss": 3.0053,
+      "step": 57020
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7720736861228943,
+      "learning_rate": 6.9223705856463525e-06,
+      "loss": 2.7265,
+      "step": 57021
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7333340048789978,
+      "learning_rate": 6.919082642977991e-06,
+      "loss": 2.9004,
+      "step": 57022
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7707780599594116,
+      "learning_rate": 6.915795472228325e-06,
+      "loss": 3.0141,
+      "step": 57023
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7591186165809631,
+      "learning_rate": 6.912509073406114e-06,
+      "loss": 2.719,
+      "step": 57024
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7694956064224243,
+      "learning_rate": 6.909223446519885e-06,
+      "loss": 2.7744,
+      "step": 57025
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7224785089492798,
+      "learning_rate": 6.9059385915783974e-06,
+      "loss": 2.9739,
+      "step": 57026
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8058514595031738,
+      "learning_rate": 6.9026545085902445e-06,
+      "loss": 3.1169,
+      "step": 57027
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7608110308647156,
+      "learning_rate": 6.899371197564085e-06,
+      "loss": 2.9059,
+      "step": 57028
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7489415407180786,
+      "learning_rate": 6.896088658508581e-06,
+      "loss": 2.8504,
+      "step": 57029
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.802652895450592,
+      "learning_rate": 6.892806891432356e-06,
+      "loss": 2.8233,
+      "step": 57030
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7284513711929321,
+      "learning_rate": 6.889525896344106e-06,
+      "loss": 2.678,
+      "step": 57031
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7275854349136353,
+      "learning_rate": 6.886245673252389e-06,
+      "loss": 3.0578,
+      "step": 57032
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7676070332527161,
+      "learning_rate": 6.882966222165898e-06,
+      "loss": 2.9871,
+      "step": 57033
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7430452108383179,
+      "learning_rate": 6.879687543093293e-06,
+      "loss": 2.8993,
+      "step": 57034
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.740665853023529,
+      "learning_rate": 6.876409636043168e-06,
+      "loss": 2.8878,
+      "step": 57035
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.744522750377655,
+      "learning_rate": 6.8731325010241815e-06,
+      "loss": 3.1277,
+      "step": 57036
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7291746735572815,
+      "learning_rate": 6.869856138044927e-06,
+      "loss": 3.0474,
+      "step": 57037
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7282450795173645,
+      "learning_rate": 6.866580547114031e-06,
+      "loss": 2.6989,
+      "step": 57038
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7871297597885132,
+      "learning_rate": 6.863305728240187e-06,
+      "loss": 2.9547,
+      "step": 57039
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7655256390571594,
+      "learning_rate": 6.860031681431954e-06,
+      "loss": 2.9997,
+      "step": 57040
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7627614736557007,
+      "learning_rate": 6.8567584066980265e-06,
+      "loss": 2.9494,
+      "step": 57041
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7453312873840332,
+      "learning_rate": 6.853485904046962e-06,
+      "loss": 3.21,
+      "step": 57042
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7098878622055054,
+      "learning_rate": 6.850214173487389e-06,
+      "loss": 2.8033,
+      "step": 57043
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.79835045337677,
+      "learning_rate": 6.846943215027934e-06,
+      "loss": 2.8388,
+      "step": 57044
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7601944804191589,
+      "learning_rate": 6.843673028677188e-06,
+      "loss": 3.1851,
+      "step": 57045
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.826819896697998,
+      "learning_rate": 6.840403614443846e-06,
+      "loss": 2.7549,
+      "step": 57046
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7625347971916199,
+      "learning_rate": 6.8371349723364e-06,
+      "loss": 2.9151,
+      "step": 57047
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7577785849571228,
+      "learning_rate": 6.833867102363544e-06,
+      "loss": 2.7609,
+      "step": 57048
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7831384539604187,
+      "learning_rate": 6.830600004533903e-06,
+      "loss": 2.9857,
+      "step": 57049
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7404026985168457,
+      "learning_rate": 6.827333678855973e-06,
+      "loss": 2.9262,
+      "step": 57050
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7409828305244446,
+      "learning_rate": 6.824068125338478e-06,
+      "loss": 2.8237,
+      "step": 57051
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7415507435798645,
+      "learning_rate": 6.820803343989911e-06,
+      "loss": 2.8341,
+      "step": 57052
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7397423982620239,
+      "learning_rate": 6.817539334818967e-06,
+      "loss": 3.0505,
+      "step": 57053
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7370979189872742,
+      "learning_rate": 6.81427609783417e-06,
+      "loss": 2.8316,
+      "step": 57054
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7267847657203674,
+      "learning_rate": 6.8110136330441485e-06,
+      "loss": 2.8226,
+      "step": 57055
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7403401136398315,
+      "learning_rate": 6.807751940457529e-06,
+      "loss": 3.0404,
+      "step": 57056
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.770000696182251,
+      "learning_rate": 6.804491020082836e-06,
+      "loss": 3.0793,
+      "step": 57057
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7584140300750732,
+      "learning_rate": 6.801230871928698e-06,
+      "loss": 2.8472,
+      "step": 57058
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7541932463645935,
+      "learning_rate": 6.7979714960036735e-06,
+      "loss": 2.9838,
+      "step": 57059
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.749034583568573,
+      "learning_rate": 6.794712892316356e-06,
+      "loss": 2.8473,
+      "step": 57060
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7517021298408508,
+      "learning_rate": 6.791455060875372e-06,
+      "loss": 2.8163,
+      "step": 57061
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.746054470539093,
+      "learning_rate": 6.788198001689249e-06,
+      "loss": 3.0738,
+      "step": 57062
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7522591352462769,
+      "learning_rate": 6.784941714766579e-06,
+      "loss": 2.8944,
+      "step": 57063
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7277589440345764,
+      "learning_rate": 6.781686200115987e-06,
+      "loss": 2.8983,
+      "step": 57064
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7314507961273193,
+      "learning_rate": 6.7784314577459364e-06,
+      "loss": 2.8121,
+      "step": 57065
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7789071202278137,
+      "learning_rate": 6.7751774876650845e-06,
+      "loss": 3.0653,
+      "step": 57066
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8103259205818176,
+      "learning_rate": 6.7719242898819915e-06,
+      "loss": 2.6541,
+      "step": 57067
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7770789265632629,
+      "learning_rate": 6.768671864405218e-06,
+      "loss": 3.0381,
+      "step": 57068
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7559372186660767,
+      "learning_rate": 6.765420211243322e-06,
+      "loss": 2.9393,
+      "step": 57069
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.722319483757019,
+      "learning_rate": 6.7621693304048984e-06,
+      "loss": 3.1307,
+      "step": 57070
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7697762846946716,
+      "learning_rate": 6.758919221898473e-06,
+      "loss": 2.8636,
+      "step": 57071
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7274612188339233,
+      "learning_rate": 6.755669885732607e-06,
+      "loss": 2.9749,
+      "step": 57072
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7847182750701904,
+      "learning_rate": 6.7524213219158905e-06,
+      "loss": 2.8848,
+      "step": 57073
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7678075432777405,
+      "learning_rate": 6.749173530456819e-06,
+      "loss": 2.7446,
+      "step": 57074
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8545045852661133,
+      "learning_rate": 6.745926511363986e-06,
+      "loss": 2.9995,
+      "step": 57075
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7134187817573547,
+      "learning_rate": 6.742680264645983e-06,
+      "loss": 2.8353,
+      "step": 57076
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8113738298416138,
+      "learning_rate": 6.7394347903113045e-06,
+      "loss": 3.1161,
+      "step": 57077
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7588936686515808,
+      "learning_rate": 6.736190088368543e-06,
+      "loss": 2.9629,
+      "step": 57078
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7868394255638123,
+      "learning_rate": 6.732946158826224e-06,
+      "loss": 2.9781,
+      "step": 57079
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7183898091316223,
+      "learning_rate": 6.72970300169281e-06,
+      "loss": 2.8361,
+      "step": 57080
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7224221229553223,
+      "learning_rate": 6.726460616976992e-06,
+      "loss": 2.7046,
+      "step": 57081
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.724663257598877,
+      "learning_rate": 6.723219004687197e-06,
+      "loss": 2.8677,
+      "step": 57082
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7192972302436829,
+      "learning_rate": 6.7199781648320185e-06,
+      "loss": 2.7882,
+      "step": 57083
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7668144106864929,
+      "learning_rate": 6.716738097419949e-06,
+      "loss": 2.9125,
+      "step": 57084
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7557151317596436,
+      "learning_rate": 6.713498802459583e-06,
+      "loss": 2.6406,
+      "step": 57085
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7683967351913452,
+      "learning_rate": 6.710260279959412e-06,
+      "loss": 2.9452,
+      "step": 57086
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7353421449661255,
+      "learning_rate": 6.707022529927964e-06,
+      "loss": 2.8917,
+      "step": 57087
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7499251961708069,
+      "learning_rate": 6.703785552373764e-06,
+      "loss": 2.8837,
+      "step": 57088
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7639588713645935,
+      "learning_rate": 6.700549347305339e-06,
+      "loss": 2.915,
+      "step": 57089
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.776925265789032,
+      "learning_rate": 6.697313914731217e-06,
+      "loss": 2.9083,
+      "step": 57090
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.755670428276062,
+      "learning_rate": 6.694079254659957e-06,
+      "loss": 2.8506,
+      "step": 57091
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.789338231086731,
+      "learning_rate": 6.690845367100017e-06,
+      "loss": 2.8634,
+      "step": 57092
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7923612594604492,
+      "learning_rate": 6.68761225205996e-06,
+      "loss": 2.8226,
+      "step": 57093
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.731760561466217,
+      "learning_rate": 6.6843799095482766e-06,
+      "loss": 2.9201,
+      "step": 57094
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7637943029403687,
+      "learning_rate": 6.681148339573461e-06,
+      "loss": 2.7399,
+      "step": 57095
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.759672224521637,
+      "learning_rate": 6.677917542144107e-06,
+      "loss": 3.0211,
+      "step": 57096
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7676330208778381,
+      "learning_rate": 6.674687517268606e-06,
+      "loss": 2.6788,
+      "step": 57097
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7366164326667786,
+      "learning_rate": 6.671458264955554e-06,
+      "loss": 2.8141,
+      "step": 57098
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7516197562217712,
+      "learning_rate": 6.668229785213441e-06,
+      "loss": 3.1885,
+      "step": 57099
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7431179285049438,
+      "learning_rate": 6.6650020780507295e-06,
+      "loss": 2.7607,
+      "step": 57100
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7496405243873596,
+      "learning_rate": 6.661775143475978e-06,
+      "loss": 3.0291,
+      "step": 57101
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7139589786529541,
+      "learning_rate": 6.658548981497647e-06,
+      "loss": 2.7426,
+      "step": 57102
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7403964400291443,
+      "learning_rate": 6.655323592124262e-06,
+      "loss": 2.9492,
+      "step": 57103
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7503599524497986,
+      "learning_rate": 6.652098975364251e-06,
+      "loss": 2.7976,
+      "step": 57104
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7594204545021057,
+      "learning_rate": 6.648875131226173e-06,
+      "loss": 2.8069,
+      "step": 57105
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7432661652565002,
+      "learning_rate": 6.645652059718521e-06,
+      "loss": 2.7111,
+      "step": 57106
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7444175481796265,
+      "learning_rate": 6.642429760849754e-06,
+      "loss": 2.7686,
+      "step": 57107
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7241867780685425,
+      "learning_rate": 6.639208234628402e-06,
+      "loss": 2.8685,
+      "step": 57108
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8182801008224487,
+      "learning_rate": 6.635987481062854e-06,
+      "loss": 2.8802,
+      "step": 57109
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7803454995155334,
+      "learning_rate": 6.632767500161673e-06,
+      "loss": 2.9425,
+      "step": 57110
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7847126126289368,
+      "learning_rate": 6.629548291933351e-06,
+      "loss": 2.8971,
+      "step": 57111
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7423437833786011,
+      "learning_rate": 6.626329856386314e-06,
+      "loss": 2.8664,
+      "step": 57112
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7208489775657654,
+      "learning_rate": 6.62311219352909e-06,
+      "loss": 2.8964,
+      "step": 57113
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7785851955413818,
+      "learning_rate": 6.619895303370071e-06,
+      "loss": 2.9177,
+      "step": 57114
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7598791122436523,
+      "learning_rate": 6.61667918591785e-06,
+      "loss": 2.933,
+      "step": 57115
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7468469142913818,
+      "learning_rate": 6.613463841180822e-06,
+      "loss": 3.0654,
+      "step": 57116
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7566969394683838,
+      "learning_rate": 6.610249269167445e-06,
+      "loss": 2.978,
+      "step": 57117
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7173053622245789,
+      "learning_rate": 6.607035469886213e-06,
+      "loss": 2.6706,
+      "step": 57118
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7708240747451782,
+      "learning_rate": 6.603822443345586e-06,
+      "loss": 2.9867,
+      "step": 57119
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7301391959190369,
+      "learning_rate": 6.60061018955399e-06,
+      "loss": 2.9992,
+      "step": 57120
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7776297330856323,
+      "learning_rate": 6.597398708519986e-06,
+      "loss": 2.88,
+      "step": 57121
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7679793834686279,
+      "learning_rate": 6.594188000251932e-06,
+      "loss": 2.8798,
+      "step": 57122
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.740397036075592,
+      "learning_rate": 6.590978064758356e-06,
+      "loss": 2.9748,
+      "step": 57123
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7653552293777466,
+      "learning_rate": 6.587768902047618e-06,
+      "loss": 2.9602,
+      "step": 57124
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.70607590675354,
+      "learning_rate": 6.584560512128278e-06,
+      "loss": 3.0756,
+      "step": 57125
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7225512266159058,
+      "learning_rate": 6.581352895008696e-06,
+      "loss": 2.8843,
+      "step": 57126
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7174662351608276,
+      "learning_rate": 6.578146050697364e-06,
+      "loss": 2.8845,
+      "step": 57127
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7353945374488831,
+      "learning_rate": 6.574939979202776e-06,
+      "loss": 2.7851,
+      "step": 57128
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7632526159286499,
+      "learning_rate": 6.571734680533292e-06,
+      "loss": 2.8001,
+      "step": 57129
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7371880412101746,
+      "learning_rate": 6.568530154697404e-06,
+      "loss": 2.8652,
+      "step": 57130
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7216587066650391,
+      "learning_rate": 6.5653264017035075e-06,
+      "loss": 2.9515,
+      "step": 57131
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7436114549636841,
+      "learning_rate": 6.562123421560062e-06,
+      "loss": 2.7013,
+      "step": 57132
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7178574800491333,
+      "learning_rate": 6.5589212142755255e-06,
+      "loss": 2.8224,
+      "step": 57133
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.723756730556488,
+      "learning_rate": 6.555719779858293e-06,
+      "loss": 2.9817,
+      "step": 57134
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7632118463516235,
+      "learning_rate": 6.552519118316857e-06,
+      "loss": 2.7139,
+      "step": 57135
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7216606140136719,
+      "learning_rate": 6.549319229659611e-06,
+      "loss": 2.9235,
+      "step": 57136
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7410150766372681,
+      "learning_rate": 6.546120113894981e-06,
+      "loss": 2.6225,
+      "step": 57137
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7452429533004761,
+      "learning_rate": 6.542921771031395e-06,
+      "loss": 2.729,
+      "step": 57138
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7587102651596069,
+      "learning_rate": 6.5397242010772454e-06,
+      "loss": 2.7456,
+      "step": 57139
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7959299087524414,
+      "learning_rate": 6.536527404041025e-06,
+      "loss": 3.0694,
+      "step": 57140
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7172296643257141,
+      "learning_rate": 6.533331379931061e-06,
+      "loss": 2.7504,
+      "step": 57141
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.773472785949707,
+      "learning_rate": 6.5301361287558475e-06,
+      "loss": 2.8911,
+      "step": 57142
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7361921072006226,
+      "learning_rate": 6.526941650523776e-06,
+      "loss": 3.1896,
+      "step": 57143
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7535603046417236,
+      "learning_rate": 6.523747945243307e-06,
+      "loss": 2.7142,
+      "step": 57144
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7661078572273254,
+      "learning_rate": 6.520555012922768e-06,
+      "loss": 2.8727,
+      "step": 57145
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8038497567176819,
+      "learning_rate": 6.5173628535705846e-06,
+      "loss": 2.9623,
+      "step": 57146
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7339990139007568,
+      "learning_rate": 6.514171467195184e-06,
+      "loss": 2.9381,
+      "step": 57147
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7659273743629456,
+      "learning_rate": 6.510980853805026e-06,
+      "loss": 2.6611,
+      "step": 57148
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7171333432197571,
+      "learning_rate": 6.507791013408403e-06,
+      "loss": 2.8451,
+      "step": 57149
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7424218058586121,
+      "learning_rate": 6.50460194601381e-06,
+      "loss": 2.8667,
+      "step": 57150
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7660609483718872,
+      "learning_rate": 6.501413651629605e-06,
+      "loss": 2.9709,
+      "step": 57151
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7458734512329102,
+      "learning_rate": 6.498226130264183e-06,
+      "loss": 2.931,
+      "step": 57152
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7954019904136658,
+      "learning_rate": 6.495039381925971e-06,
+      "loss": 2.6553,
+      "step": 57153
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.760521650314331,
+      "learning_rate": 6.491853406623293e-06,
+      "loss": 2.7921,
+      "step": 57154
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7642186284065247,
+      "learning_rate": 6.488668204364644e-06,
+      "loss": 2.8834,
+      "step": 57155
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7433639764785767,
+      "learning_rate": 6.485483775158317e-06,
+      "loss": 3.1153,
+      "step": 57156
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7498500943183899,
+      "learning_rate": 6.482300119012706e-06,
+      "loss": 2.9402,
+      "step": 57157
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7769145965576172,
+      "learning_rate": 6.479117235936304e-06,
+      "loss": 2.9213,
+      "step": 57158
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7752602100372314,
+      "learning_rate": 6.475935125937404e-06,
+      "loss": 3.0248,
+      "step": 57159
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7823769450187683,
+      "learning_rate": 6.472753789024365e-06,
+      "loss": 3.0033,
+      "step": 57160
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7235212922096252,
+      "learning_rate": 6.469573225205615e-06,
+      "loss": 2.8306,
+      "step": 57161
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7436525225639343,
+      "learning_rate": 6.466393434489514e-06,
+      "loss": 2.8062,
+      "step": 57162
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.716906726360321,
+      "learning_rate": 6.463214416884455e-06,
+      "loss": 2.84,
+      "step": 57163
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7609844207763672,
+      "learning_rate": 6.460036172398764e-06,
+      "loss": 2.8809,
+      "step": 57164
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7516579031944275,
+      "learning_rate": 6.4568587010409015e-06,
+      "loss": 2.942,
+      "step": 57165
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7477497458457947,
+      "learning_rate": 6.453682002819161e-06,
+      "loss": 2.815,
+      "step": 57166
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7754057049751282,
+      "learning_rate": 6.450506077741901e-06,
+      "loss": 2.8088,
+      "step": 57167
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7623452544212341,
+      "learning_rate": 6.447330925817518e-06,
+      "loss": 2.8484,
+      "step": 57168
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7394402623176575,
+      "learning_rate": 6.444156547054369e-06,
+      "loss": 2.9405,
+      "step": 57169
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.779207706451416,
+      "learning_rate": 6.440982941460848e-06,
+      "loss": 3.1458,
+      "step": 57170
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7456086874008179,
+      "learning_rate": 6.437810109045249e-06,
+      "loss": 3.0439,
+      "step": 57171
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7450244426727295,
+      "learning_rate": 6.434638049815966e-06,
+      "loss": 2.8157,
+      "step": 57172
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7540341019630432,
+      "learning_rate": 6.4314667637813565e-06,
+      "loss": 3.0942,
+      "step": 57173
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7208600044250488,
+      "learning_rate": 6.428296250949749e-06,
+      "loss": 2.9982,
+      "step": 57174
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7499627470970154,
+      "learning_rate": 6.425126511329537e-06,
+      "loss": 2.8532,
+      "step": 57175
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7473828196525574,
+      "learning_rate": 6.421957544929013e-06,
+      "loss": 2.9155,
+      "step": 57176
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7623142004013062,
+      "learning_rate": 6.41878935175657e-06,
+      "loss": 2.984,
+      "step": 57177
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7398785352706909,
+      "learning_rate": 6.415621931820536e-06,
+      "loss": 2.7673,
+      "step": 57178
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7367839813232422,
+      "learning_rate": 6.412455285129236e-06,
+      "loss": 3.085,
+      "step": 57179
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7085956335067749,
+      "learning_rate": 6.409289411691065e-06,
+      "loss": 2.8524,
+      "step": 57180
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.741699755191803,
+      "learning_rate": 6.406124311514316e-06,
+      "loss": 2.7711,
+      "step": 57181
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7420092225074768,
+      "learning_rate": 6.402959984607281e-06,
+      "loss": 3.0438,
+      "step": 57182
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7586853504180908,
+      "learning_rate": 6.3997964309783875e-06,
+      "loss": 2.9436,
+      "step": 57183
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7290069460868835,
+      "learning_rate": 6.396633650635929e-06,
+      "loss": 2.8053,
+      "step": 57184
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7516114115715027,
+      "learning_rate": 6.393471643588233e-06,
+      "loss": 2.7453,
+      "step": 57185
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7613343596458435,
+      "learning_rate": 6.390310409843591e-06,
+      "loss": 2.723,
+      "step": 57186
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.6871271729469299,
+      "learning_rate": 6.387149949410398e-06,
+      "loss": 2.9468,
+      "step": 57187
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7292281985282898,
+      "learning_rate": 6.383990262296945e-06,
+      "loss": 3.0544,
+      "step": 57188
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7564751505851746,
+      "learning_rate": 6.380831348511528e-06,
+      "loss": 2.873,
+      "step": 57189
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7569683194160461,
+      "learning_rate": 6.377673208062539e-06,
+      "loss": 2.8391,
+      "step": 57190
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7515416741371155,
+      "learning_rate": 6.374515840958206e-06,
+      "loss": 2.9058,
+      "step": 57191
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7558345794677734,
+      "learning_rate": 6.371359247206887e-06,
+      "loss": 2.8219,
+      "step": 57192
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7959738969802856,
+      "learning_rate": 6.368203426816909e-06,
+      "loss": 2.8152,
+      "step": 57193
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7814188599586487,
+      "learning_rate": 6.365048379796567e-06,
+      "loss": 3.0779,
+      "step": 57194
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7512837648391724,
+      "learning_rate": 6.361894106154253e-06,
+      "loss": 2.8264,
+      "step": 57195
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7334195375442505,
+      "learning_rate": 6.358740605898094e-06,
+      "loss": 2.921,
+      "step": 57196
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7542886137962341,
+      "learning_rate": 6.35558787903655e-06,
+      "loss": 2.8447,
+      "step": 57197
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7210304737091064,
+      "learning_rate": 6.352435925577848e-06,
+      "loss": 2.9945,
+      "step": 57198
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.767892062664032,
+      "learning_rate": 6.349284745530347e-06,
+      "loss": 2.9185,
+      "step": 57199
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7845554351806641,
+      "learning_rate": 6.3461343389023086e-06,
+      "loss": 3.0824,
+      "step": 57200
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7469670176506042,
+      "learning_rate": 6.342984705701992e-06,
+      "loss": 2.7764,
+      "step": 57201
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7826364636421204,
+      "learning_rate": 6.339835845937791e-06,
+      "loss": 2.9049,
+      "step": 57202
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7540563344955444,
+      "learning_rate": 6.336687759617965e-06,
+      "loss": 2.9289,
+      "step": 57203
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7304716110229492,
+      "learning_rate": 6.33354044675074e-06,
+      "loss": 2.9662,
+      "step": 57204
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7389193773269653,
+      "learning_rate": 6.330393907344478e-06,
+      "loss": 2.9829,
+      "step": 57205
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8157293200492859,
+      "learning_rate": 6.3272481414074036e-06,
+      "loss": 3.1955,
+      "step": 57206
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.775015652179718,
+      "learning_rate": 6.324103148947879e-06,
+      "loss": 2.8361,
+      "step": 57207
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7417725920677185,
+      "learning_rate": 6.320958929974096e-06,
+      "loss": 3.1331,
+      "step": 57208
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7489861845970154,
+      "learning_rate": 6.3178154844944484e-06,
+      "loss": 2.947,
+      "step": 57209
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7145342826843262,
+      "learning_rate": 6.31467281251713e-06,
+      "loss": 3.0041,
+      "step": 57210
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7301230430603027,
+      "learning_rate": 6.311530914050433e-06,
+      "loss": 2.8497,
+      "step": 57211
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8062174916267395,
+      "learning_rate": 6.3083897891026855e-06,
+      "loss": 2.843,
+      "step": 57212
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7669421434402466,
+      "learning_rate": 6.305249437682047e-06,
+      "loss": 2.9091,
+      "step": 57213
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7261912822723389,
+      "learning_rate": 6.302109859796878e-06,
+      "loss": 2.7274,
+      "step": 57214
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7557966709136963,
+      "learning_rate": 6.298971055455471e-06,
+      "loss": 2.7214,
+      "step": 57215
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.776716947555542,
+      "learning_rate": 6.29583302466602e-06,
+      "loss": 3.1463,
+      "step": 57216
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7491167783737183,
+      "learning_rate": 6.292695767436817e-06,
+      "loss": 2.8682,
+      "step": 57217
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7415738701820374,
+      "learning_rate": 6.289559283776158e-06,
+      "loss": 2.9238,
+      "step": 57218
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.731604278087616,
+      "learning_rate": 6.286423573692234e-06,
+      "loss": 2.946,
+      "step": 57219
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7397477626800537,
+      "learning_rate": 6.283288637193373e-06,
+      "loss": 2.7806,
+      "step": 57220
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.761749804019928,
+      "learning_rate": 6.280154474287769e-06,
+      "loss": 2.791,
+      "step": 57221
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7503309845924377,
+      "learning_rate": 6.277021084983747e-06,
+      "loss": 2.8343,
+      "step": 57222
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7354109883308411,
+      "learning_rate": 6.273888469289501e-06,
+      "loss": 2.7524,
+      "step": 57223
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7324250936508179,
+      "learning_rate": 6.270756627213325e-06,
+      "loss": 2.9752,
+      "step": 57224
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7345625162124634,
+      "learning_rate": 6.267625558763445e-06,
+      "loss": 2.7768,
+      "step": 57225
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7356343269348145,
+      "learning_rate": 6.264495263948088e-06,
+      "loss": 2.7994,
+      "step": 57226
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.72329181432724,
+      "learning_rate": 6.261365742775548e-06,
+      "loss": 2.7858,
+      "step": 57227
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7713742852210999,
+      "learning_rate": 6.258236995254018e-06,
+      "loss": 2.6734,
+      "step": 57228
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7680825591087341,
+      "learning_rate": 6.255109021391724e-06,
+      "loss": 3.1817,
+      "step": 57229
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7381764650344849,
+      "learning_rate": 6.251981821196994e-06,
+      "loss": 2.9703,
+      "step": 57230
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7881075739860535,
+      "learning_rate": 6.24885539467802e-06,
+      "loss": 2.8687,
+      "step": 57231
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7882162928581238,
+      "learning_rate": 6.245729741843031e-06,
+      "loss": 2.9799,
+      "step": 57232
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8076366782188416,
+      "learning_rate": 6.242604862700218e-06,
+      "loss": 2.7061,
+      "step": 57233
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7945789694786072,
+      "learning_rate": 6.239480757257842e-06,
+      "loss": 2.9682,
+      "step": 57234
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7510719895362854,
+      "learning_rate": 6.236357425524164e-06,
+      "loss": 3.0232,
+      "step": 57235
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7661550045013428,
+      "learning_rate": 6.233234867507375e-06,
+      "loss": 2.9934,
+      "step": 57236
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7666568160057068,
+      "learning_rate": 6.2301130832157046e-06,
+      "loss": 2.8113,
+      "step": 57237
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7529629468917847,
+      "learning_rate": 6.226992072657377e-06,
+      "loss": 2.7529,
+      "step": 57238
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7184775471687317,
+      "learning_rate": 6.223871835840655e-06,
+      "loss": 2.7139,
+      "step": 57239
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7889968752861023,
+      "learning_rate": 6.220752372773696e-06,
+      "loss": 3.0028,
+      "step": 57240
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7780910134315491,
+      "learning_rate": 6.217633683464696e-06,
+      "loss": 2.9095,
+      "step": 57241
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7787454724311829,
+      "learning_rate": 6.214515767921946e-06,
+      "loss": 2.9341,
+      "step": 57242
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7292056083679199,
+      "learning_rate": 6.211398626153607e-06,
+      "loss": 2.8772,
+      "step": 57243
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7210398316383362,
+      "learning_rate": 6.208282258167907e-06,
+      "loss": 2.7536,
+      "step": 57244
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7634195685386658,
+      "learning_rate": 6.205166663973038e-06,
+      "loss": 2.9861,
+      "step": 57245
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7329500317573547,
+      "learning_rate": 6.202051843577261e-06,
+      "loss": 3.0233,
+      "step": 57246
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7299970388412476,
+      "learning_rate": 6.198937796988734e-06,
+      "loss": 2.5595,
+      "step": 57247
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7431080937385559,
+      "learning_rate": 6.19582452421562e-06,
+      "loss": 2.9949,
+      "step": 57248
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7681297063827515,
+      "learning_rate": 6.1927120252661444e-06,
+      "loss": 3.1091,
+      "step": 57249
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7236172556877136,
+      "learning_rate": 6.189600300148567e-06,
+      "loss": 2.9795,
+      "step": 57250
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7395146489143372,
+      "learning_rate": 6.186489348871016e-06,
+      "loss": 3.0194,
+      "step": 57251
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7280433773994446,
+      "learning_rate": 6.183379171441716e-06,
+      "loss": 2.8707,
+      "step": 57252
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7451938390731812,
+      "learning_rate": 6.180269767868862e-06,
+      "loss": 2.9683,
+      "step": 57253
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7515122890472412,
+      "learning_rate": 6.177161138160614e-06,
+      "loss": 2.8751,
+      "step": 57254
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7310147881507874,
+      "learning_rate": 6.174053282325198e-06,
+      "loss": 2.7883,
+      "step": 57255
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7602556943893433,
+      "learning_rate": 6.170946200370741e-06,
+      "loss": 3.0935,
+      "step": 57256
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7617813348770142,
+      "learning_rate": 6.1678398923055035e-06,
+      "loss": 3.003,
+      "step": 57257
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7423365116119385,
+      "learning_rate": 6.164734358137579e-06,
+      "loss": 2.8405,
+      "step": 57258
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.730692446231842,
+      "learning_rate": 6.161629597875195e-06,
+      "loss": 2.8546,
+      "step": 57259
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7807807922363281,
+      "learning_rate": 6.158525611526577e-06,
+      "loss": 2.837,
+      "step": 57260
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7491143345832825,
+      "learning_rate": 6.155422399099852e-06,
+      "loss": 3.056,
+      "step": 57261
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7525192499160767,
+      "learning_rate": 6.1523199606031806e-06,
+      "loss": 2.7682,
+      "step": 57262
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8082860708236694,
+      "learning_rate": 6.149218296044722e-06,
+      "loss": 2.7799,
+      "step": 57263
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7616897821426392,
+      "learning_rate": 6.146117405432638e-06,
+      "loss": 2.6663,
+      "step": 57264
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.784067690372467,
+      "learning_rate": 6.143017288775187e-06,
+      "loss": 3.032,
+      "step": 57265
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.773425817489624,
+      "learning_rate": 6.139917946080464e-06,
+      "loss": 2.8674,
+      "step": 57266
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7454070448875427,
+      "learning_rate": 6.136819377356628e-06,
+      "loss": 2.6727,
+      "step": 57267
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7858648896217346,
+      "learning_rate": 6.133721582611906e-06,
+      "loss": 2.9446,
+      "step": 57268
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7687196135520935,
+      "learning_rate": 6.130624561854325e-06,
+      "loss": 2.6409,
+      "step": 57269
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7145859599113464,
+      "learning_rate": 6.127528315092178e-06,
+      "loss": 3.2014,
+      "step": 57270
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7361599802970886,
+      "learning_rate": 6.1244328423335265e-06,
+      "loss": 3.015,
+      "step": 57271
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7314938902854919,
+      "learning_rate": 6.121338143586596e-06,
+      "loss": 3.0502,
+      "step": 57272
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7757865786552429,
+      "learning_rate": 6.11824421885948e-06,
+      "loss": 2.767,
+      "step": 57273
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7310301661491394,
+      "learning_rate": 6.115151068160373e-06,
+      "loss": 2.7884,
+      "step": 57274
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.748052179813385,
+      "learning_rate": 6.112058691497401e-06,
+      "loss": 2.7668,
+      "step": 57275
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7008647918701172,
+      "learning_rate": 6.108967088878691e-06,
+      "loss": 2.8638,
+      "step": 57276
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7348527312278748,
+      "learning_rate": 6.105876260312404e-06,
+      "loss": 2.8012,
+      "step": 57277
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7351037859916687,
+      "learning_rate": 6.102786205806664e-06,
+      "loss": 2.9923,
+      "step": 57278
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7870645523071289,
+      "learning_rate": 6.099696925369635e-06,
+      "loss": 2.9804,
+      "step": 57279
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7273508310317993,
+      "learning_rate": 6.096608419009441e-06,
+      "loss": 2.9093,
+      "step": 57280
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7449030876159668,
+      "learning_rate": 6.09352068673421e-06,
+      "loss": 3.0118,
+      "step": 57281
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7714568972587585,
+      "learning_rate": 6.090433728552102e-06,
+      "loss": 3.0912,
+      "step": 57282
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7424775958061218,
+      "learning_rate": 6.08734754447121e-06,
+      "loss": 2.9796,
+      "step": 57283
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7018720507621765,
+      "learning_rate": 6.084262134499696e-06,
+      "loss": 2.7236,
+      "step": 57284
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8280545473098755,
+      "learning_rate": 6.081177498645651e-06,
+      "loss": 2.8014,
+      "step": 57285
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7586017847061157,
+      "learning_rate": 6.078093636917236e-06,
+      "loss": 2.9029,
+      "step": 57286
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7319121956825256,
+      "learning_rate": 6.075010549322545e-06,
+      "loss": 2.954,
+      "step": 57287
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7442706823348999,
+      "learning_rate": 6.071928235869705e-06,
+      "loss": 2.993,
+      "step": 57288
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7516229152679443,
+      "learning_rate": 6.068846696566842e-06,
+      "loss": 2.8055,
+      "step": 57289
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7649211883544922,
+      "learning_rate": 6.065765931422084e-06,
+      "loss": 2.8407,
+      "step": 57290
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7471636533737183,
+      "learning_rate": 6.062685940443523e-06,
+      "loss": 2.7969,
+      "step": 57291
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7413542866706848,
+      "learning_rate": 6.0596067236392874e-06,
+      "loss": 2.9349,
+      "step": 57292
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7625101804733276,
+      "learning_rate": 6.056528281017436e-06,
+      "loss": 3.0211,
+      "step": 57293
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7629085779190063,
+      "learning_rate": 6.0534506125861635e-06,
+      "loss": 2.8523,
+      "step": 57294
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7582331895828247,
+      "learning_rate": 6.050373718353496e-06,
+      "loss": 2.83,
+      "step": 57295
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7564645409584045,
+      "learning_rate": 6.047297598327561e-06,
+      "loss": 2.7993,
+      "step": 57296
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7481827735900879,
+      "learning_rate": 6.044222252516551e-06,
+      "loss": 2.9593,
+      "step": 57297
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7227786183357239,
+      "learning_rate": 6.041147680928427e-06,
+      "loss": 3.0073,
+      "step": 57298
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7628014087677002,
+      "learning_rate": 6.038073883571348e-06,
+      "loss": 3.0438,
+      "step": 57299
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7520178556442261,
+      "learning_rate": 6.035000860453409e-06,
+      "loss": 2.8691,
+      "step": 57300
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7767812013626099,
+      "learning_rate": 6.031928611582704e-06,
+      "loss": 2.8563,
+      "step": 57301
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.738016664981842,
+      "learning_rate": 6.028857136967325e-06,
+      "loss": 2.9511,
+      "step": 57302
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7443080544471741,
+      "learning_rate": 6.0257864366153655e-06,
+      "loss": 2.7384,
+      "step": 57303
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7690123915672302,
+      "learning_rate": 6.02271651053492e-06,
+      "loss": 2.7267,
+      "step": 57304
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.757183313369751,
+      "learning_rate": 6.019647358734048e-06,
+      "loss": 2.8482,
+      "step": 57305
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7601922750473022,
+      "learning_rate": 6.0165789812208435e-06,
+      "loss": 2.9199,
+      "step": 57306
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7741249799728394,
+      "learning_rate": 6.0135113780034e-06,
+      "loss": 3.0289,
+      "step": 57307
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7413105368614197,
+      "learning_rate": 6.010444549089777e-06,
+      "loss": 2.9368,
+      "step": 57308
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.763995349407196,
+      "learning_rate": 6.007378494488102e-06,
+      "loss": 2.8091,
+      "step": 57309
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.742984414100647,
+      "learning_rate": 6.004313214206369e-06,
+      "loss": 2.9815,
+      "step": 57310
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7137545347213745,
+      "learning_rate": 6.001248708252704e-06,
+      "loss": 2.8202,
+      "step": 57311
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.74273282289505,
+      "learning_rate": 5.998184976635201e-06,
+      "loss": 2.8075,
+      "step": 57312
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7143562436103821,
+      "learning_rate": 5.995122019361853e-06,
+      "loss": 2.8815,
+      "step": 57313
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.713367760181427,
+      "learning_rate": 5.992059836440788e-06,
+      "loss": 2.8194,
+      "step": 57314
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.6914703845977783,
+      "learning_rate": 5.988998427880032e-06,
+      "loss": 2.8552,
+      "step": 57315
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7266353368759155,
+      "learning_rate": 5.985937793687679e-06,
+      "loss": 2.7779,
+      "step": 57316
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7659033536911011,
+      "learning_rate": 5.982877933871822e-06,
+      "loss": 3.0615,
+      "step": 57317
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.740502119064331,
+      "learning_rate": 5.979818848440421e-06,
+      "loss": 2.8445,
+      "step": 57318
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7166577577590942,
+      "learning_rate": 5.976760537401637e-06,
+      "loss": 2.8114,
+      "step": 57319
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7707402110099792,
+      "learning_rate": 5.973703000763497e-06,
+      "loss": 2.8491,
+      "step": 57320
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7468615174293518,
+      "learning_rate": 5.9706462385339936e-06,
+      "loss": 2.8557,
+      "step": 57321
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7503215670585632,
+      "learning_rate": 5.967590250721222e-06,
+      "loss": 2.9284,
+      "step": 57322
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7697805166244507,
+      "learning_rate": 5.96453503733324e-06,
+      "loss": 2.8139,
+      "step": 57323
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7666358351707458,
+      "learning_rate": 5.96148059837811e-06,
+      "loss": 3.2273,
+      "step": 57324
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.1106570959091187,
+      "learning_rate": 5.958426933863791e-06,
+      "loss": 2.9179,
+      "step": 57325
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7379875183105469,
+      "learning_rate": 5.955374043798444e-06,
+      "loss": 2.8579,
+      "step": 57326
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7157080173492432,
+      "learning_rate": 5.952321928190029e-06,
+      "loss": 2.6449,
+      "step": 57327
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7337937355041504,
+      "learning_rate": 5.949270587046606e-06,
+      "loss": 2.6561,
+      "step": 57328
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7767958641052246,
+      "learning_rate": 5.946220020376235e-06,
+      "loss": 2.7252,
+      "step": 57329
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.699264645576477,
+      "learning_rate": 5.943170228186878e-06,
+      "loss": 2.7687,
+      "step": 57330
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.769777238368988,
+      "learning_rate": 5.940121210486659e-06,
+      "loss": 3.0783,
+      "step": 57331
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7534322738647461,
+      "learning_rate": 5.93707296728354e-06,
+      "loss": 2.7708,
+      "step": 57332
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7632072567939758,
+      "learning_rate": 5.934025498585615e-06,
+      "loss": 2.759,
+      "step": 57333
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7482007741928101,
+      "learning_rate": 5.930978804400876e-06,
+      "loss": 3.1279,
+      "step": 57334
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7381764054298401,
+      "learning_rate": 5.927932884737318e-06,
+      "loss": 2.8895,
+      "step": 57335
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7939985394477844,
+      "learning_rate": 5.924887739602968e-06,
+      "loss": 2.8252,
+      "step": 57336
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7845645546913147,
+      "learning_rate": 5.921843369005885e-06,
+      "loss": 3.0684,
+      "step": 57337
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7730729579925537,
+      "learning_rate": 5.918799772954064e-06,
+      "loss": 3.0756,
+      "step": 57338
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7449852824211121,
+      "learning_rate": 5.9157569514555635e-06,
+      "loss": 3.0701,
+      "step": 57339
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7435621023178101,
+      "learning_rate": 5.912714904518312e-06,
+      "loss": 2.9434,
+      "step": 57340
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7603535652160645,
+      "learning_rate": 5.909673632150436e-06,
+      "loss": 3.0084,
+      "step": 57341
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7573125958442688,
+      "learning_rate": 5.9066331343598285e-06,
+      "loss": 2.795,
+      "step": 57342
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7497143745422363,
+      "learning_rate": 5.90359341115455e-06,
+      "loss": 2.8122,
+      "step": 57343
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7319371104240417,
+      "learning_rate": 5.900554462542662e-06,
+      "loss": 2.9278,
+      "step": 57344
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8178033828735352,
+      "learning_rate": 5.897516288532056e-06,
+      "loss": 2.8397,
+      "step": 57345
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8565611243247986,
+      "learning_rate": 5.894478889130794e-06,
+      "loss": 2.8193,
+      "step": 57346
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.744626522064209,
+      "learning_rate": 5.8914422643468685e-06,
+      "loss": 2.9301,
+      "step": 57347
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7861790060997009,
+      "learning_rate": 5.88840641418834e-06,
+      "loss": 2.8523,
+      "step": 57348
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7643438577651978,
+      "learning_rate": 5.885371338663103e-06,
+      "loss": 2.8766,
+      "step": 57349
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8093913197517395,
+      "learning_rate": 5.882337037779183e-06,
+      "loss": 3.0652,
+      "step": 57350
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7430794835090637,
+      "learning_rate": 5.879303511544576e-06,
+      "loss": 2.8223,
+      "step": 57351
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8071761131286621,
+      "learning_rate": 5.876270759967305e-06,
+      "loss": 2.9618,
+      "step": 57352
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7300369143486023,
+      "learning_rate": 5.873238783055334e-06,
+      "loss": 2.9846,
+      "step": 57353
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7400018572807312,
+      "learning_rate": 5.870207580816655e-06,
+      "loss": 2.6373,
+      "step": 57354
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7521073818206787,
+      "learning_rate": 5.8671771532592284e-06,
+      "loss": 2.9761,
+      "step": 57355
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7645672559738159,
+      "learning_rate": 5.864147500391047e-06,
+      "loss": 2.6966,
+      "step": 57356
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7233006358146667,
+      "learning_rate": 5.861118622220073e-06,
+      "loss": 2.8905,
+      "step": 57357
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7672364115715027,
+      "learning_rate": 5.858090518754332e-06,
+      "loss": 2.6379,
+      "step": 57358
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7448053359985352,
+      "learning_rate": 5.855063190001785e-06,
+      "loss": 2.89,
+      "step": 57359
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7966521978378296,
+      "learning_rate": 5.8520366359703585e-06,
+      "loss": 2.9453,
+      "step": 57360
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.851170003414154,
+      "learning_rate": 5.849010856668079e-06,
+      "loss": 2.8829,
+      "step": 57361
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8057568073272705,
+      "learning_rate": 5.845985852102874e-06,
+      "loss": 2.6775,
+      "step": 57362
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7434202432632446,
+      "learning_rate": 5.842961622282738e-06,
+      "loss": 2.8152,
+      "step": 57363
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.846804678440094,
+      "learning_rate": 5.839938167215663e-06,
+      "loss": 3.005,
+      "step": 57364
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7839913368225098,
+      "learning_rate": 5.836915486909544e-06,
+      "loss": 2.8143,
+      "step": 57365
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7269675731658936,
+      "learning_rate": 5.833893581372373e-06,
+      "loss": 2.9256,
+      "step": 57366
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7645124793052673,
+      "learning_rate": 5.830872450612112e-06,
+      "loss": 2.9203,
+      "step": 57367
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7519145011901855,
+      "learning_rate": 5.82785209463672e-06,
+      "loss": 2.7965,
+      "step": 57368
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7483285665512085,
+      "learning_rate": 5.824832513454192e-06,
+      "loss": 2.703,
+      "step": 57369
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.767500638961792,
+      "learning_rate": 5.82181370707242e-06,
+      "loss": 2.6382,
+      "step": 57370
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7527577877044678,
+      "learning_rate": 5.8187956754993655e-06,
+      "loss": 2.8592,
+      "step": 57371
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7957414984703064,
+      "learning_rate": 5.815778418742989e-06,
+      "loss": 2.7785,
+      "step": 57372
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7185649871826172,
+      "learning_rate": 5.812761936811217e-06,
+      "loss": 2.8294,
+      "step": 57373
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7688578367233276,
+      "learning_rate": 5.8097462297120425e-06,
+      "loss": 2.8467,
+      "step": 57374
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7571620941162109,
+      "learning_rate": 5.806731297453326e-06,
+      "loss": 2.8394,
+      "step": 57375
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7824880480766296,
+      "learning_rate": 5.803717140043129e-06,
+      "loss": 2.8248,
+      "step": 57376
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7680121064186096,
+      "learning_rate": 5.800703757489311e-06,
+      "loss": 3.0829,
+      "step": 57377
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7361317276954651,
+      "learning_rate": 5.7976911497997655e-06,
+      "loss": 2.8745,
+      "step": 57378
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7746545076370239,
+      "learning_rate": 5.79467931698252e-06,
+      "loss": 2.9367,
+      "step": 57379
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7424500584602356,
+      "learning_rate": 5.791668259045468e-06,
+      "loss": 2.7415,
+      "step": 57380
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.9100118279457092,
+      "learning_rate": 5.788657975996536e-06,
+      "loss": 3.0618,
+      "step": 57381
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.768410325050354,
+      "learning_rate": 5.785648467843651e-06,
+      "loss": 2.8411,
+      "step": 57382
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7742539644241333,
+      "learning_rate": 5.782639734594741e-06,
+      "loss": 2.8681,
+      "step": 57383
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7879564762115479,
+      "learning_rate": 5.779631776257765e-06,
+      "loss": 2.8448,
+      "step": 57384
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7167067527770996,
+      "learning_rate": 5.776624592840584e-06,
+      "loss": 2.8608,
+      "step": 57385
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7402490377426147,
+      "learning_rate": 5.773618184351192e-06,
+      "loss": 3.1005,
+      "step": 57386
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7294678092002869,
+      "learning_rate": 5.770612550797416e-06,
+      "loss": 2.8117,
+      "step": 57387
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7386787533760071,
+      "learning_rate": 5.767607692187215e-06,
+      "loss": 2.7278,
+      "step": 57388
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7458301782608032,
+      "learning_rate": 5.7646036085285505e-06,
+      "loss": 2.8675,
+      "step": 57389
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7428733706474304,
+      "learning_rate": 5.76160029982925e-06,
+      "loss": 2.8125,
+      "step": 57390
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7372967004776001,
+      "learning_rate": 5.758597766097306e-06,
+      "loss": 3.0284,
+      "step": 57391
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7986836433410645,
+      "learning_rate": 5.755596007340579e-06,
+      "loss": 2.8643,
+      "step": 57392
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7363787293434143,
+      "learning_rate": 5.752595023566964e-06,
+      "loss": 3.0305,
+      "step": 57393
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7744364738464355,
+      "learning_rate": 5.7495948147844194e-06,
+      "loss": 2.8855,
+      "step": 57394
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7604561448097229,
+      "learning_rate": 5.746595381000774e-06,
+      "loss": 2.8601,
+      "step": 57395
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7391270399093628,
+      "learning_rate": 5.74359672222402e-06,
+      "loss": 2.9664,
+      "step": 57396
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7320667505264282,
+      "learning_rate": 5.740598838461951e-06,
+      "loss": 3.0343,
+      "step": 57397
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7917554378509521,
+      "learning_rate": 5.73760172972253e-06,
+      "loss": 2.9968,
+      "step": 57398
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.772669792175293,
+      "learning_rate": 5.734605396013647e-06,
+      "loss": 2.9576,
+      "step": 57399
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7252935767173767,
+      "learning_rate": 5.731609837343132e-06,
+      "loss": 3.0033,
+      "step": 57400
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7601112127304077,
+      "learning_rate": 5.728615053718977e-06,
+      "loss": 3.03,
+      "step": 57401
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7780544757843018,
+      "learning_rate": 5.725621045148976e-06,
+      "loss": 2.9666,
+      "step": 57402
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.760009229183197,
+      "learning_rate": 5.72262781164109e-06,
+      "loss": 2.9728,
+      "step": 57403
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7839109301567078,
+      "learning_rate": 5.719635353203145e-06,
+      "loss": 3.0657,
+      "step": 57404
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7375196218490601,
+      "learning_rate": 5.716643669843035e-06,
+      "loss": 2.8473,
+      "step": 57405
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7924519777297974,
+      "learning_rate": 5.713652761568688e-06,
+      "loss": 2.8607,
+      "step": 57406
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7020806074142456,
+      "learning_rate": 5.71066262838793e-06,
+      "loss": 2.9978,
+      "step": 57407
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.744614839553833,
+      "learning_rate": 5.707673270308655e-06,
+      "loss": 3.1015,
+      "step": 57408
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7562986612319946,
+      "learning_rate": 5.704684687338723e-06,
+      "loss": 2.7133,
+      "step": 57409
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7199798226356506,
+      "learning_rate": 5.7016968794859955e-06,
+      "loss": 2.8839,
+      "step": 57410
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7654950022697449,
+      "learning_rate": 5.698709846758364e-06,
+      "loss": 2.9377,
+      "step": 57411
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.737543523311615,
+      "learning_rate": 5.695723589163692e-06,
+      "loss": 2.7927,
+      "step": 57412
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7589178681373596,
+      "learning_rate": 5.692738106709871e-06,
+      "loss": 2.9463,
+      "step": 57413
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7573418021202087,
+      "learning_rate": 5.689753399404729e-06,
+      "loss": 2.8789,
+      "step": 57414
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7789357900619507,
+      "learning_rate": 5.686769467256125e-06,
+      "loss": 2.8112,
+      "step": 57415
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7354355454444885,
+      "learning_rate": 5.683786310271921e-06,
+      "loss": 2.6794,
+      "step": 57416
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7149415016174316,
+      "learning_rate": 5.680803928459976e-06,
+      "loss": 2.7977,
+      "step": 57417
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7695300579071045,
+      "learning_rate": 5.677822321828152e-06,
+      "loss": 2.9736,
+      "step": 57418
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7486740350723267,
+      "learning_rate": 5.674841490384341e-06,
+      "loss": 2.9578,
+      "step": 57419
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8076779842376709,
+      "learning_rate": 5.671861434136304e-06,
+      "loss": 2.9316,
+      "step": 57420
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.766984224319458,
+      "learning_rate": 5.668882153091969e-06,
+      "loss": 2.9472,
+      "step": 57421
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7838339805603027,
+      "learning_rate": 5.665903647259162e-06,
+      "loss": 2.7034,
+      "step": 57422
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.764181911945343,
+      "learning_rate": 5.662925916645677e-06,
+      "loss": 2.7349,
+      "step": 57423
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7412382960319519,
+      "learning_rate": 5.659948961259442e-06,
+      "loss": 3.072,
+      "step": 57424
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7331834435462952,
+      "learning_rate": 5.656972781108216e-06,
+      "loss": 2.9661,
+      "step": 57425
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7448732852935791,
+      "learning_rate": 5.653997376199925e-06,
+      "loss": 2.9843,
+      "step": 57426
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.739306628704071,
+      "learning_rate": 5.6510227465423e-06,
+      "loss": 2.8614,
+      "step": 57427
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8040658235549927,
+      "learning_rate": 5.6480488921432645e-06,
+      "loss": 2.9276,
+      "step": 57428
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7358850240707397,
+      "learning_rate": 5.645075813010647e-06,
+      "loss": 2.9018,
+      "step": 57429
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7447613477706909,
+      "learning_rate": 5.642103509152207e-06,
+      "loss": 2.9166,
+      "step": 57430
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.8018415570259094,
+      "learning_rate": 5.6391319805758396e-06,
+      "loss": 2.8888,
+      "step": 57431
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7493090033531189,
+      "learning_rate": 5.6361612272893045e-06,
+      "loss": 2.8736,
+      "step": 57432
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.770797073841095,
+      "learning_rate": 5.633191249300529e-06,
+      "loss": 2.8206,
+      "step": 57433
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7289892435073853,
+      "learning_rate": 5.630222046617205e-06,
+      "loss": 2.8424,
+      "step": 57434
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7826147079467773,
+      "learning_rate": 5.627253619247296e-06,
+      "loss": 2.9898,
+      "step": 57435
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7384172081947327,
+      "learning_rate": 5.624285967198528e-06,
+      "loss": 2.7443,
+      "step": 57436
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7605634927749634,
+      "learning_rate": 5.621319090478693e-06,
+      "loss": 2.8757,
+      "step": 57437
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7587273120880127,
+      "learning_rate": 5.6183529890957205e-06,
+      "loss": 2.8125,
+      "step": 57438
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.6983641982078552,
+      "learning_rate": 5.615387663057302e-06,
+      "loss": 2.8265,
+      "step": 57439
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7505119442939758,
+      "learning_rate": 5.612423112371267e-06,
+      "loss": 3.1734,
+      "step": 57440
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7802063822746277,
+      "learning_rate": 5.609459337045507e-06,
+      "loss": 3.1378,
+      "step": 57441
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7453617453575134,
+      "learning_rate": 5.606496337087751e-06,
+      "loss": 2.9801,
+      "step": 57442
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7429621815681458,
+      "learning_rate": 5.603534112505859e-06,
+      "loss": 3.0071,
+      "step": 57443
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7750768661499023,
+      "learning_rate": 5.60057266330759e-06,
+      "loss": 2.9811,
+      "step": 57444
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.767142117023468,
+      "learning_rate": 5.5976119895007385e-06,
+      "loss": 2.6831,
+      "step": 57445
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8097765445709229,
+      "learning_rate": 5.594652091093133e-06,
+      "loss": 3.0158,
+      "step": 57446
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.735511064529419,
+      "learning_rate": 5.591692968092532e-06,
+      "loss": 2.8753,
+      "step": 57447
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.778668224811554,
+      "learning_rate": 5.588734620506763e-06,
+      "loss": 2.624,
+      "step": 57448
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7237783670425415,
+      "learning_rate": 5.585777048343587e-06,
+      "loss": 2.9627,
+      "step": 57449
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7732894420623779,
+      "learning_rate": 5.582820251610864e-06,
+      "loss": 2.9355,
+      "step": 57450
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7555185556411743,
+      "learning_rate": 5.5798642303163225e-06,
+      "loss": 2.781,
+      "step": 57451
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7874925136566162,
+      "learning_rate": 5.576908984467721e-06,
+      "loss": 2.9256,
+      "step": 57452
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7143713235855103,
+      "learning_rate": 5.573954514072887e-06,
+      "loss": 2.8541,
+      "step": 57453
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7951681613922119,
+      "learning_rate": 5.571000819139582e-06,
+      "loss": 2.9457,
+      "step": 57454
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7409695982933044,
+      "learning_rate": 5.568047899675599e-06,
+      "loss": 3.0258,
+      "step": 57455
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7536093592643738,
+      "learning_rate": 5.565095755688731e-06,
+      "loss": 2.9839,
+      "step": 57456
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7845450639724731,
+      "learning_rate": 5.56214438718674e-06,
+      "loss": 2.8603,
+      "step": 57457
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7714955806732178,
+      "learning_rate": 5.559193794177419e-06,
+      "loss": 3.0423,
+      "step": 57458
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7862733602523804,
+      "learning_rate": 5.556243976668462e-06,
+      "loss": 2.7072,
+      "step": 57459
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7352171540260315,
+      "learning_rate": 5.553294934667696e-06,
+      "loss": 3.0502,
+      "step": 57460
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7902106642723083,
+      "learning_rate": 5.5503466681829146e-06,
+      "loss": 2.8126,
+      "step": 57461
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7630064487457275,
+      "learning_rate": 5.547399177221812e-06,
+      "loss": 2.8373,
+      "step": 57462
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7489845156669617,
+      "learning_rate": 5.544452461792248e-06,
+      "loss": 2.9392,
+      "step": 57463
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7420606017112732,
+      "learning_rate": 5.541506521901884e-06,
+      "loss": 2.9887,
+      "step": 57464
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7396260499954224,
+      "learning_rate": 5.538561357558546e-06,
+      "loss": 2.5488,
+      "step": 57465
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7430559992790222,
+      "learning_rate": 5.535616968769963e-06,
+      "loss": 2.9337,
+      "step": 57466
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7690236568450928,
+      "learning_rate": 5.53267335554386e-06,
+      "loss": 2.9588,
+      "step": 57467
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.733445405960083,
+      "learning_rate": 5.529730517888064e-06,
+      "loss": 2.8503,
+      "step": 57468
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7059511542320251,
+      "learning_rate": 5.526788455810271e-06,
+      "loss": 2.9426,
+      "step": 57469
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7520898580551147,
+      "learning_rate": 5.52384716931824e-06,
+      "loss": 2.9003,
+      "step": 57470
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8237969875335693,
+      "learning_rate": 5.520906658419733e-06,
+      "loss": 2.8435,
+      "step": 57471
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7647461891174316,
+      "learning_rate": 5.517966923122508e-06,
+      "loss": 2.9031,
+      "step": 57472
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7231850624084473,
+      "learning_rate": 5.51502796343426e-06,
+      "loss": 2.7187,
+      "step": 57473
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.6980883479118347,
+      "learning_rate": 5.512089779362716e-06,
+      "loss": 2.8957,
+      "step": 57474
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7455374598503113,
+      "learning_rate": 5.50915237091567e-06,
+      "loss": 2.9012,
+      "step": 57475
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7731330990791321,
+      "learning_rate": 5.506215738100883e-06,
+      "loss": 3.0828,
+      "step": 57476
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7092301249504089,
+      "learning_rate": 5.503279880926015e-06,
+      "loss": 2.9492,
+      "step": 57477
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7573080062866211,
+      "learning_rate": 5.5003447993988255e-06,
+      "loss": 2.8745,
+      "step": 57478
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7752591371536255,
+      "learning_rate": 5.497410493527077e-06,
+      "loss": 2.9229,
+      "step": 57479
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7782385945320129,
+      "learning_rate": 5.4944769633184616e-06,
+      "loss": 3.0516,
+      "step": 57480
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7580782175064087,
+      "learning_rate": 5.491544208780707e-06,
+      "loss": 2.9722,
+      "step": 57481
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.749438464641571,
+      "learning_rate": 5.488612229921541e-06,
+      "loss": 2.9657,
+      "step": 57482
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7620990872383118,
+      "learning_rate": 5.485681026748723e-06,
+      "loss": 2.8918,
+      "step": 57483
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7435450553894043,
+      "learning_rate": 5.48275059926988e-06,
+      "loss": 2.8467,
+      "step": 57484
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7541936635971069,
+      "learning_rate": 5.4798209474928075e-06,
+      "loss": 2.6854,
+      "step": 57485
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7550050020217896,
+      "learning_rate": 5.4768920714252315e-06,
+      "loss": 2.6602,
+      "step": 57486
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7599036693572998,
+      "learning_rate": 5.473963971074846e-06,
+      "loss": 2.8434,
+      "step": 57487
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.714177668094635,
+      "learning_rate": 5.471036646449378e-06,
+      "loss": 2.844,
+      "step": 57488
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7540480494499207,
+      "learning_rate": 5.468110097556455e-06,
+      "loss": 2.8204,
+      "step": 57489
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7340836524963379,
+      "learning_rate": 5.4651843244038705e-06,
+      "loss": 2.8181,
+      "step": 57490
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7581061124801636,
+      "learning_rate": 5.462259326999319e-06,
+      "loss": 2.8814,
+      "step": 57491
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.834175169467926,
+      "learning_rate": 5.45933510535046e-06,
+      "loss": 2.9842,
+      "step": 57492
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7358565330505371,
+      "learning_rate": 5.456411659465054e-06,
+      "loss": 2.8919,
+      "step": 57493
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7109126448631287,
+      "learning_rate": 5.453488989350762e-06,
+      "loss": 2.7427,
+      "step": 57494
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.724243700504303,
+      "learning_rate": 5.450567095015312e-06,
+      "loss": 2.5712,
+      "step": 57495
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7592142224311829,
+      "learning_rate": 5.447645976466364e-06,
+      "loss": 2.9613,
+      "step": 57496
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7279868125915527,
+      "learning_rate": 5.444725633711611e-06,
+      "loss": 2.8802,
+      "step": 57497
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7003878355026245,
+      "learning_rate": 5.4418060667588135e-06,
+      "loss": 3.1777,
+      "step": 57498
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7382853031158447,
+      "learning_rate": 5.438887275615566e-06,
+      "loss": 2.8485,
+      "step": 57499
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7092037200927734,
+      "learning_rate": 5.43596926028963e-06,
+      "loss": 2.8656,
+      "step": 57500
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.756615936756134,
+      "learning_rate": 5.433052020788631e-06,
+      "loss": 3.0391,
+      "step": 57501
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7301920056343079,
+      "learning_rate": 5.4301355571202965e-06,
+      "loss": 2.8625,
+      "step": 57502
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.753453254699707,
+      "learning_rate": 5.427219869292288e-06,
+      "loss": 2.8667,
+      "step": 57503
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8032575845718384,
+      "learning_rate": 5.424304957312298e-06,
+      "loss": 2.8938,
+      "step": 57504
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7729468941688538,
+      "learning_rate": 5.421390821187988e-06,
+      "loss": 3.0574,
+      "step": 57505
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7835939526557922,
+      "learning_rate": 5.418477460927051e-06,
+      "loss": 2.7424,
+      "step": 57506
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7130961418151855,
+      "learning_rate": 5.415564876537149e-06,
+      "loss": 2.6947,
+      "step": 57507
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7707082033157349,
+      "learning_rate": 5.412653068025941e-06,
+      "loss": 2.8117,
+      "step": 57508
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7709498405456543,
+      "learning_rate": 5.409742035401154e-06,
+      "loss": 2.7995,
+      "step": 57509
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7126640677452087,
+      "learning_rate": 5.406831778670351e-06,
+      "loss": 2.8213,
+      "step": 57510
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7614896893501282,
+      "learning_rate": 5.403922297841323e-06,
+      "loss": 2.8555,
+      "step": 57511
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7267091870307922,
+      "learning_rate": 5.4010135929216e-06,
+      "loss": 2.9359,
+      "step": 57512
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7336955070495605,
+      "learning_rate": 5.398105663918972e-06,
+      "loss": 2.9989,
+      "step": 57513
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7180055975914001,
+      "learning_rate": 5.39519851084097e-06,
+      "loss": 3.036,
+      "step": 57514
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8044130206108093,
+      "learning_rate": 5.392292133695386e-06,
+      "loss": 2.8144,
+      "step": 57515
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7591328024864197,
+      "learning_rate": 5.389386532489781e-06,
+      "loss": 2.801,
+      "step": 57516
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7301998138427734,
+      "learning_rate": 5.386481707231816e-06,
+      "loss": 2.7639,
+      "step": 57517
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7751278281211853,
+      "learning_rate": 5.3835776579291835e-06,
+      "loss": 2.7864,
+      "step": 57518
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7999823689460754,
+      "learning_rate": 5.3806743845894784e-06,
+      "loss": 2.9398,
+      "step": 57519
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7624409794807434,
+      "learning_rate": 5.377771887220394e-06,
+      "loss": 2.5767,
+      "step": 57520
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8037823438644409,
+      "learning_rate": 5.374870165829559e-06,
+      "loss": 2.9523,
+      "step": 57521
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7656051516532898,
+      "learning_rate": 5.371969220424599e-06,
+      "loss": 2.8153,
+      "step": 57522
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7563663125038147,
+      "learning_rate": 5.369069051013208e-06,
+      "loss": 2.8219,
+      "step": 57523
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7026633620262146,
+      "learning_rate": 5.366169657602948e-06,
+      "loss": 2.8431,
+      "step": 57524
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7382919192314148,
+      "learning_rate": 5.363271040201511e-06,
+      "loss": 2.9779,
+      "step": 57525
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.0166808366775513,
+      "learning_rate": 5.360373198816492e-06,
+      "loss": 2.9759,
+      "step": 57526
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7511675357818604,
+      "learning_rate": 5.357476133455552e-06,
+      "loss": 2.8831,
+      "step": 57527
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7667033076286316,
+      "learning_rate": 5.354579844126317e-06,
+      "loss": 2.8983,
+      "step": 57528
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7955288290977478,
+      "learning_rate": 5.351684330836414e-06,
+      "loss": 2.9765,
+      "step": 57529
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7648731470108032,
+      "learning_rate": 5.348789593593505e-06,
+      "loss": 2.9163,
+      "step": 57530
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8392326235771179,
+      "learning_rate": 5.3458956324051504e-06,
+      "loss": 2.8243,
+      "step": 57531
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7558670043945312,
+      "learning_rate": 5.343002447278977e-06,
+      "loss": 3.025,
+      "step": 57532
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7797101736068726,
+      "learning_rate": 5.340110038222644e-06,
+      "loss": 3.0272,
+      "step": 57533
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7675113081932068,
+      "learning_rate": 5.337218405243748e-06,
+      "loss": 3.2485,
+      "step": 57534
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7596341371536255,
+      "learning_rate": 5.334327548349948e-06,
+      "loss": 3.0808,
+      "step": 57535
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7368146777153015,
+      "learning_rate": 5.3314374675487714e-06,
+      "loss": 2.724,
+      "step": 57536
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7727293968200684,
+      "learning_rate": 5.328548162847879e-06,
+      "loss": 2.7899,
+      "step": 57537
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.71891850233078,
+      "learning_rate": 5.325659634254931e-06,
+      "loss": 2.8625,
+      "step": 57538
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7327159643173218,
+      "learning_rate": 5.322771881777421e-06,
+      "loss": 2.9382,
+      "step": 57539
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7162185311317444,
+      "learning_rate": 5.319884905423077e-06,
+      "loss": 2.7911,
+      "step": 57540
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.740092396736145,
+      "learning_rate": 5.316998705199393e-06,
+      "loss": 2.9277,
+      "step": 57541
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.771077036857605,
+      "learning_rate": 5.31411328111403e-06,
+      "loss": 2.7363,
+      "step": 57542
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.728895902633667,
+      "learning_rate": 5.31122863317458e-06,
+      "loss": 2.8725,
+      "step": 57543
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7603982090950012,
+      "learning_rate": 5.308344761388639e-06,
+      "loss": 2.8479,
+      "step": 57544
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7482584118843079,
+      "learning_rate": 5.305461665763833e-06,
+      "loss": 2.876,
+      "step": 57545
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7513059377670288,
+      "learning_rate": 5.302579346307723e-06,
+      "loss": 2.9574,
+      "step": 57546
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7721797227859497,
+      "learning_rate": 5.299697803027869e-06,
+      "loss": 3.2145,
+      "step": 57547
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7654960751533508,
+      "learning_rate": 5.296817035931933e-06,
+      "loss": 3.0258,
+      "step": 57548
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.793851912021637,
+      "learning_rate": 5.293937045027441e-06,
+      "loss": 2.9649,
+      "step": 57549
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7595252394676208,
+      "learning_rate": 5.2910578303220205e-06,
+      "loss": 2.8733,
+      "step": 57550
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7790116667747498,
+      "learning_rate": 5.2881793918232e-06,
+      "loss": 2.889,
+      "step": 57551
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.74222332239151,
+      "learning_rate": 5.285301729538638e-06,
+      "loss": 2.9103,
+      "step": 57552
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.793062150478363,
+      "learning_rate": 5.282424843475863e-06,
+      "loss": 2.8367,
+      "step": 57553
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7489209175109863,
+      "learning_rate": 5.279548733642436e-06,
+      "loss": 2.8782,
+      "step": 57554
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7790075540542603,
+      "learning_rate": 5.276673400045983e-06,
+      "loss": 2.9005,
+      "step": 57555
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7113046646118164,
+      "learning_rate": 5.273798842694033e-06,
+      "loss": 2.7339,
+      "step": 57556
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7641512751579285,
+      "learning_rate": 5.2709250615941445e-06,
+      "loss": 2.8315,
+      "step": 57557
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7478774189949036,
+      "learning_rate": 5.26805205675398e-06,
+      "loss": 2.7826,
+      "step": 57558
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7339596748352051,
+      "learning_rate": 5.265179828181032e-06,
+      "loss": 2.9666,
+      "step": 57559
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7537974715232849,
+      "learning_rate": 5.2623083758828624e-06,
+      "loss": 2.9265,
+      "step": 57560
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7144533395767212,
+      "learning_rate": 5.259437699867064e-06,
+      "loss": 2.8843,
+      "step": 57561
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7479895949363708,
+      "learning_rate": 5.256567800141165e-06,
+      "loss": 2.9927,
+      "step": 57562
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7054181098937988,
+      "learning_rate": 5.253698676712759e-06,
+      "loss": 2.8793,
+      "step": 57563
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7824179530143738,
+      "learning_rate": 5.250830329589339e-06,
+      "loss": 2.8978,
+      "step": 57564
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7174307703971863,
+      "learning_rate": 5.2479627587785675e-06,
+      "loss": 2.8663,
+      "step": 57565
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7782198786735535,
+      "learning_rate": 5.245095964287904e-06,
+      "loss": 2.8694,
+      "step": 57566
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7361534833908081,
+      "learning_rate": 5.242229946124943e-06,
+      "loss": 2.8386,
+      "step": 57567
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7740105986595154,
+      "learning_rate": 5.239364704297211e-06,
+      "loss": 2.7616,
+      "step": 57568
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7601125240325928,
+      "learning_rate": 5.236500238812269e-06,
+      "loss": 3.121,
+      "step": 57569
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7358449697494507,
+      "learning_rate": 5.233636549677678e-06,
+      "loss": 2.883,
+      "step": 57570
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7319931387901306,
+      "learning_rate": 5.230773636900898e-06,
+      "loss": 2.9506,
+      "step": 57571
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7816628217697144,
+      "learning_rate": 5.22791150048959e-06,
+      "loss": 2.8939,
+      "step": 57572
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7447727918624878,
+      "learning_rate": 5.225050140451215e-06,
+      "loss": 2.9117,
+      "step": 57573
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7741653919219971,
+      "learning_rate": 5.222189556793332e-06,
+      "loss": 2.7785,
+      "step": 57574
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7777344584465027,
+      "learning_rate": 5.219329749523471e-06,
+      "loss": 2.7844,
+      "step": 57575
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7560518980026245,
+      "learning_rate": 5.216470718649157e-06,
+      "loss": 3.047,
+      "step": 57576
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7396130561828613,
+      "learning_rate": 5.213612464177918e-06,
+      "loss": 2.8337,
+      "step": 57577
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7314813137054443,
+      "learning_rate": 5.210754986117316e-06,
+      "loss": 2.9577,
+      "step": 57578
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7825134992599487,
+      "learning_rate": 5.207898284474809e-06,
+      "loss": 2.8162,
+      "step": 57579
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.787757933139801,
+      "learning_rate": 5.205042359257994e-06,
+      "loss": 3.0029,
+      "step": 57580
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7091009616851807,
+      "learning_rate": 5.202187210474396e-06,
+      "loss": 2.8445,
+      "step": 57581
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7412608861923218,
+      "learning_rate": 5.199332838131443e-06,
+      "loss": 2.9387,
+      "step": 57582
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7515140771865845,
+      "learning_rate": 5.196479242236762e-06,
+      "loss": 2.7867,
+      "step": 57583
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7683895826339722,
+      "learning_rate": 5.1936264227977806e-06,
+      "loss": 3.0592,
+      "step": 57584
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7576361298561096,
+      "learning_rate": 5.1907743798220606e-06,
+      "loss": 2.7596,
+      "step": 57585
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7240979075431824,
+      "learning_rate": 5.187923113317094e-06,
+      "loss": 3.0416,
+      "step": 57586
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7744309306144714,
+      "learning_rate": 5.1850726232904095e-06,
+      "loss": 2.9025,
+      "step": 57587
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7236972451210022,
+      "learning_rate": 5.182222909749501e-06,
+      "loss": 2.9219,
+      "step": 57588
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7626051306724548,
+      "learning_rate": 5.179373972701895e-06,
+      "loss": 3.0846,
+      "step": 57589
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7822689414024353,
+      "learning_rate": 5.176525812155085e-06,
+      "loss": 2.8351,
+      "step": 57590
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7402531504631042,
+      "learning_rate": 5.173678428116534e-06,
+      "loss": 2.8016,
+      "step": 57591
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8098210692405701,
+      "learning_rate": 5.170831820593802e-06,
+      "loss": 3.0032,
+      "step": 57592
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7769423723220825,
+      "learning_rate": 5.167985989594348e-06,
+      "loss": 2.9816,
+      "step": 57593
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7293099761009216,
+      "learning_rate": 5.1651409351256665e-06,
+      "loss": 2.828,
+      "step": 57594
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7770448923110962,
+      "learning_rate": 5.162296657195253e-06,
+      "loss": 2.9258,
+      "step": 57595
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7682694792747498,
+      "learning_rate": 5.159453155810633e-06,
+      "loss": 2.7589,
+      "step": 57596
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7171539068222046,
+      "learning_rate": 5.156610430979302e-06,
+      "loss": 2.7792,
+      "step": 57597
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7691710591316223,
+      "learning_rate": 5.1537684827086535e-06,
+      "loss": 2.9293,
+      "step": 57598
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7626630663871765,
+      "learning_rate": 5.1509273110062475e-06,
+      "loss": 2.9575,
+      "step": 57599
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7831852436065674,
+      "learning_rate": 5.148086915879579e-06,
+      "loss": 2.7166,
+      "step": 57600
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.734420895576477,
+      "learning_rate": 5.145247297336075e-06,
+      "loss": 2.883,
+      "step": 57601
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7334008812904358,
+      "learning_rate": 5.142408455383262e-06,
+      "loss": 3.034,
+      "step": 57602
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7297717928886414,
+      "learning_rate": 5.139570390028602e-06,
+      "loss": 2.6785,
+      "step": 57603
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7485255002975464,
+      "learning_rate": 5.136733101279556e-06,
+      "loss": 3.0351,
+      "step": 57604
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7516413927078247,
+      "learning_rate": 5.133896589143616e-06,
+      "loss": 2.8879,
+      "step": 57605
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8280989527702332,
+      "learning_rate": 5.1310608536282105e-06,
+      "loss": 2.9283,
+      "step": 57606
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7576739192008972,
+      "learning_rate": 5.128225894740867e-06,
+      "loss": 2.8524,
+      "step": 57607
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7196022868156433,
+      "learning_rate": 5.125391712489047e-06,
+      "loss": 2.8059,
+      "step": 57608
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7515164017677307,
+      "learning_rate": 5.122558306880143e-06,
+      "loss": 2.6472,
+      "step": 57609
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7482687830924988,
+      "learning_rate": 5.119725677921715e-06,
+      "loss": 3.0543,
+      "step": 57610
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7849657535552979,
+      "learning_rate": 5.116893825621194e-06,
+      "loss": 2.604,
+      "step": 57611
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.737377941608429,
+      "learning_rate": 5.114062749986003e-06,
+      "loss": 2.8361,
+      "step": 57612
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7955058217048645,
+      "learning_rate": 5.1112324510236055e-06,
+      "loss": 2.9009,
+      "step": 57613
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7713934183120728,
+      "learning_rate": 5.108402928741462e-06,
+      "loss": 2.5417,
+      "step": 57614
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7756964564323425,
+      "learning_rate": 5.105574183147065e-06,
+      "loss": 2.8647,
+      "step": 57615
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7626116871833801,
+      "learning_rate": 5.10274621424781e-06,
+      "loss": 2.8613,
+      "step": 57616
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7126087546348572,
+      "learning_rate": 5.099919022051157e-06,
+      "loss": 2.7191,
+      "step": 57617
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7495712637901306,
+      "learning_rate": 5.097092606564601e-06,
+      "loss": 2.7508,
+      "step": 57618
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.743903398513794,
+      "learning_rate": 5.094266967795502e-06,
+      "loss": 2.9745,
+      "step": 57619
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7743878960609436,
+      "learning_rate": 5.091442105751387e-06,
+      "loss": 2.6665,
+      "step": 57620
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7580164074897766,
+      "learning_rate": 5.0886180204396165e-06,
+      "loss": 2.9658,
+      "step": 57621
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7179275155067444,
+      "learning_rate": 5.085794711867719e-06,
+      "loss": 2.9423,
+      "step": 57622
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7473773956298828,
+      "learning_rate": 5.0829721800430544e-06,
+      "loss": 2.9537,
+      "step": 57623
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7446900606155396,
+      "learning_rate": 5.080150424973051e-06,
+      "loss": 2.8643,
+      "step": 57624
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7130290865898132,
+      "learning_rate": 5.077329446665235e-06,
+      "loss": 2.8381,
+      "step": 57625
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7973248362541199,
+      "learning_rate": 5.0745092451269345e-06,
+      "loss": 2.7452,
+      "step": 57626
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7995738387107849,
+      "learning_rate": 5.071689820365643e-06,
+      "loss": 2.9036,
+      "step": 57627
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8131319880485535,
+      "learning_rate": 5.068871172388755e-06,
+      "loss": 2.9542,
+      "step": 57628
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7483770251274109,
+      "learning_rate": 5.066053301203665e-06,
+      "loss": 2.8858,
+      "step": 57629
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7863949537277222,
+      "learning_rate": 5.0632362068178665e-06,
+      "loss": 2.9802,
+      "step": 57630
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7554341554641724,
+      "learning_rate": 5.0604198892387205e-06,
+      "loss": 3.0817,
+      "step": 57631
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7444338798522949,
+      "learning_rate": 5.057604348473687e-06,
+      "loss": 3.0471,
+      "step": 57632
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8026116490364075,
+      "learning_rate": 5.054789584530161e-06,
+      "loss": 2.6147,
+      "step": 57633
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7201833128929138,
+      "learning_rate": 5.051975597415536e-06,
+      "loss": 2.7799,
+      "step": 57634
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7888780236244202,
+      "learning_rate": 5.0491623871372396e-06,
+      "loss": 3.0041,
+      "step": 57635
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7324066162109375,
+      "learning_rate": 5.0463499537027e-06,
+      "loss": 2.7168,
+      "step": 57636
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7467460632324219,
+      "learning_rate": 5.043538297119309e-06,
+      "loss": 2.7614,
+      "step": 57637
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8163952827453613,
+      "learning_rate": 5.040727417394464e-06,
+      "loss": 3.0423,
+      "step": 57638
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.788611888885498,
+      "learning_rate": 5.037917314535589e-06,
+      "loss": 3.0102,
+      "step": 57639
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7810007929801941,
+      "learning_rate": 5.03510798855008e-06,
+      "loss": 2.9274,
+      "step": 57640
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.8209431171417236,
+      "learning_rate": 5.032299439445297e-06,
+      "loss": 2.7132,
+      "step": 57641
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7978459596633911,
+      "learning_rate": 5.0294916672287025e-06,
+      "loss": 2.7835,
+      "step": 57642
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7896865010261536,
+      "learning_rate": 5.026684671907655e-06,
+      "loss": 2.7713,
+      "step": 57643
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7331417202949524,
+      "learning_rate": 5.023878453489516e-06,
+      "loss": 2.8776,
+      "step": 57644
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7624796628952026,
+      "learning_rate": 5.0210730119817465e-06,
+      "loss": 2.9129,
+      "step": 57645
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7527821660041809,
+      "learning_rate": 5.018268347391674e-06,
+      "loss": 2.9413,
+      "step": 57646
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7554188966751099,
+      "learning_rate": 5.015464459726759e-06,
+      "loss": 3.1033,
+      "step": 57647
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7641181349754333,
+      "learning_rate": 5.012661348994329e-06,
+      "loss": 2.8095,
+      "step": 57648
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7661232352256775,
+      "learning_rate": 5.009859015201745e-06,
+      "loss": 2.9625,
+      "step": 57649
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7631051540374756,
+      "learning_rate": 5.0070574583564335e-06,
+      "loss": 2.8323,
+      "step": 57650
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8133715391159058,
+      "learning_rate": 5.004256678465757e-06,
+      "loss": 2.9095,
+      "step": 57651
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7536367774009705,
+      "learning_rate": 5.001456675537108e-06,
+      "loss": 2.8894,
+      "step": 57652
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7896614074707031,
+      "learning_rate": 4.998657449577848e-06,
+      "loss": 3.0761,
+      "step": 57653
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 4.995859000595337e-06,
+      "loss": 2.9384,
+      "step": 57654
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7240200638771057,
+      "learning_rate": 4.99306132859697e-06,
+      "loss": 2.6961,
+      "step": 57655
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7582331895828247,
+      "learning_rate": 4.990264433590108e-06,
+      "loss": 2.7058,
+      "step": 57656
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7294660210609436,
+      "learning_rate": 4.98746831558211e-06,
+      "loss": 2.9779,
+      "step": 57657
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.752255916595459,
+      "learning_rate": 4.984672974580306e-06,
+      "loss": 2.8786,
+      "step": 57658
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.751939594745636,
+      "learning_rate": 4.981878410592122e-06,
+      "loss": 2.8781,
+      "step": 57659
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7334133982658386,
+      "learning_rate": 4.979084623624918e-06,
+      "loss": 2.9869,
+      "step": 57660
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7954682111740112,
+      "learning_rate": 4.976291613686023e-06,
+      "loss": 3.1021,
+      "step": 57661
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7579163312911987,
+      "learning_rate": 4.973499380782797e-06,
+      "loss": 2.9179,
+      "step": 57662
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.743008553981781,
+      "learning_rate": 4.970707924922568e-06,
+      "loss": 2.8985,
+      "step": 57663
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7410622835159302,
+      "learning_rate": 4.9679172461127295e-06,
+      "loss": 3.0076,
+      "step": 57664
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.725355327129364,
+      "learning_rate": 4.965127344360608e-06,
+      "loss": 2.8972,
+      "step": 57665
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7321346998214722,
+      "learning_rate": 4.962338219673567e-06,
+      "loss": 3.1501,
+      "step": 57666
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7218225002288818,
+      "learning_rate": 4.9595498720589655e-06,
+      "loss": 2.7632,
+      "step": 57667
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7458818554878235,
+      "learning_rate": 4.9567623015240975e-06,
+      "loss": 3.0234,
+      "step": 57668
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7871835231781006,
+      "learning_rate": 4.9539755080763575e-06,
+      "loss": 2.8553,
+      "step": 57669
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.9065406322479248,
+      "learning_rate": 4.95118949172304e-06,
+      "loss": 2.8418,
+      "step": 57670
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7578248977661133,
+      "learning_rate": 4.948404252471539e-06,
+      "loss": 2.8822,
+      "step": 57671
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.734098494052887,
+      "learning_rate": 4.945619790329147e-06,
+      "loss": 2.7124,
+      "step": 57672
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7335187196731567,
+      "learning_rate": 4.942836105303161e-06,
+      "loss": 2.9535,
+      "step": 57673
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7521882057189941,
+      "learning_rate": 4.940053197401006e-06,
+      "loss": 2.7288,
+      "step": 57674
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8029778003692627,
+      "learning_rate": 4.937271066629944e-06,
+      "loss": 2.9228,
+      "step": 57675
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8154861330986023,
+      "learning_rate": 4.934489712997369e-06,
+      "loss": 3.0219,
+      "step": 57676
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7279578447341919,
+      "learning_rate": 4.931709136510542e-06,
+      "loss": 2.9194,
+      "step": 57677
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7490201592445374,
+      "learning_rate": 4.9289293371767905e-06,
+      "loss": 2.8508,
+      "step": 57678
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8128371834754944,
+      "learning_rate": 4.926150315003441e-06,
+      "loss": 2.6265,
+      "step": 57679
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7674548029899597,
+      "learning_rate": 4.923372069997822e-06,
+      "loss": 2.8969,
+      "step": 57680
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7656154036521912,
+      "learning_rate": 4.920594602167261e-06,
+      "loss": 2.8899,
+      "step": 57681
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.733749270439148,
+      "learning_rate": 4.917817911519084e-06,
+      "loss": 2.8042,
+      "step": 57682
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7814816832542419,
+      "learning_rate": 4.915041998060587e-06,
+      "loss": 2.9326,
+      "step": 57683
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7335500121116638,
+      "learning_rate": 4.912266861799097e-06,
+      "loss": 2.7703,
+      "step": 57684
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7337460517883301,
+      "learning_rate": 4.909492502741841e-06,
+      "loss": 3.0456,
+      "step": 57685
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7310239672660828,
+      "learning_rate": 4.906718920896213e-06,
+      "loss": 2.8439,
+      "step": 57686
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7263157963752747,
+      "learning_rate": 4.903946116269508e-06,
+      "loss": 3.001,
+      "step": 57687
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7620536088943481,
+      "learning_rate": 4.90117408886902e-06,
+      "loss": 3.0321,
+      "step": 57688
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7526038289070129,
+      "learning_rate": 4.898402838702043e-06,
+      "loss": 2.9153,
+      "step": 57689
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7336723804473877,
+      "learning_rate": 4.89563236577587e-06,
+      "loss": 3.0666,
+      "step": 57690
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7981674671173096,
+      "learning_rate": 4.89286267009783e-06,
+      "loss": 2.7095,
+      "step": 57691
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7469791173934937,
+      "learning_rate": 4.890093751675183e-06,
+      "loss": 2.9837,
+      "step": 57692
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7180920243263245,
+      "learning_rate": 4.887325610515225e-06,
+      "loss": 2.9881,
+      "step": 57693
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7620313763618469,
+      "learning_rate": 4.884558246625248e-06,
+      "loss": 2.7828,
+      "step": 57694
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7135055661201477,
+      "learning_rate": 4.8817916600125465e-06,
+      "loss": 3.048,
+      "step": 57695
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7366616129875183,
+      "learning_rate": 4.879025850684415e-06,
+      "loss": 2.9772,
+      "step": 57696
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.745126485824585,
+      "learning_rate": 4.876260818648181e-06,
+      "loss": 2.8637,
+      "step": 57697
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7213049530982971,
+      "learning_rate": 4.873496563911039e-06,
+      "loss": 2.9629,
+      "step": 57698
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7770732641220093,
+      "learning_rate": 4.8707330864803166e-06,
+      "loss": 2.9155,
+      "step": 57699
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.773853063583374,
+      "learning_rate": 4.8679703863632735e-06,
+      "loss": 2.8713,
+      "step": 57700
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7769290804862976,
+      "learning_rate": 4.865208463567172e-06,
+      "loss": 2.8884,
+      "step": 57701
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7928144335746765,
+      "learning_rate": 4.8624473180993385e-06,
+      "loss": 2.7892,
+      "step": 57702
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7701162695884705,
+      "learning_rate": 4.859686949967034e-06,
+      "loss": 2.9104,
+      "step": 57703
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7658120393753052,
+      "learning_rate": 4.856927359177487e-06,
+      "loss": 2.825,
+      "step": 57704
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7138972282409668,
+      "learning_rate": 4.854168545738024e-06,
+      "loss": 3.0144,
+      "step": 57705
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7636241316795349,
+      "learning_rate": 4.8514105096558396e-06,
+      "loss": 2.7451,
+      "step": 57706
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7346954941749573,
+      "learning_rate": 4.848653250938261e-06,
+      "loss": 2.9506,
+      "step": 57707
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7780959010124207,
+      "learning_rate": 4.8458967695925166e-06,
+      "loss": 2.8082,
+      "step": 57708
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7654991149902344,
+      "learning_rate": 4.843141065625899e-06,
+      "loss": 2.7958,
+      "step": 57709
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7330126166343689,
+      "learning_rate": 4.840386139045604e-06,
+      "loss": 2.9405,
+      "step": 57710
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7495510578155518,
+      "learning_rate": 4.837631989858959e-06,
+      "loss": 2.8339,
+      "step": 57711
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7610169053077698,
+      "learning_rate": 4.83487861807319e-06,
+      "loss": 2.6812,
+      "step": 57712
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.711486279964447,
+      "learning_rate": 4.832126023695526e-06,
+      "loss": 2.9422,
+      "step": 57713
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7498459815979004,
+      "learning_rate": 4.829374206733261e-06,
+      "loss": 2.8501,
+      "step": 57714
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.755424439907074,
+      "learning_rate": 4.826623167193588e-06,
+      "loss": 2.9558,
+      "step": 57715
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.744842529296875,
+      "learning_rate": 4.82387290508377e-06,
+      "loss": 2.8703,
+      "step": 57716
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7416000962257385,
+      "learning_rate": 4.821123420411099e-06,
+      "loss": 2.9342,
+      "step": 57717
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.782110869884491,
+      "learning_rate": 4.818374713182771e-06,
+      "loss": 2.8408,
+      "step": 57718
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7619156241416931,
+      "learning_rate": 4.815626783406046e-06,
+      "loss": 2.8473,
+      "step": 57719
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7477062940597534,
+      "learning_rate": 4.812879631088151e-06,
+      "loss": 3.0257,
+      "step": 57720
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7299256920814514,
+      "learning_rate": 4.810133256236282e-06,
+      "loss": 3.0029,
+      "step": 57721
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7986459136009216,
+      "learning_rate": 4.807387658857764e-06,
+      "loss": 2.4808,
+      "step": 57722
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7746559381484985,
+      "learning_rate": 4.804642838959727e-06,
+      "loss": 3.0949,
+      "step": 57723
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8096112012863159,
+      "learning_rate": 4.8018987965494635e-06,
+      "loss": 2.8608,
+      "step": 57724
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7485320568084717,
+      "learning_rate": 4.799155531634203e-06,
+      "loss": 2.904,
+      "step": 57725
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7414878010749817,
+      "learning_rate": 4.796413044221136e-06,
+      "loss": 2.7812,
+      "step": 57726
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7604761123657227,
+      "learning_rate": 4.793671334317528e-06,
+      "loss": 2.8507,
+      "step": 57727
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7408043146133423,
+      "learning_rate": 4.7909304019305705e-06,
+      "loss": 3.0301,
+      "step": 57728
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8125025033950806,
+      "learning_rate": 4.788190247067492e-06,
+      "loss": 3.0556,
+      "step": 57729
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7148528099060059,
+      "learning_rate": 4.785450869735452e-06,
+      "loss": 2.9549,
+      "step": 57730
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7859530448913574,
+      "learning_rate": 4.782712269941747e-06,
+      "loss": 2.904,
+      "step": 57731
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.814154863357544,
+      "learning_rate": 4.7799744476936015e-06,
+      "loss": 2.845,
+      "step": 57732
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7224166393280029,
+      "learning_rate": 4.777237402998147e-06,
+      "loss": 2.8495,
+      "step": 57733
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7980074286460876,
+      "learning_rate": 4.7745011358626406e-06,
+      "loss": 2.9546,
+      "step": 57734
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7275083661079407,
+      "learning_rate": 4.7717656462943125e-06,
+      "loss": 2.9702,
+      "step": 57735
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7441866993904114,
+      "learning_rate": 4.769030934300289e-06,
+      "loss": 2.918,
+      "step": 57736
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8218494653701782,
+      "learning_rate": 4.766296999887864e-06,
+      "loss": 3.025,
+      "step": 57737
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7137600779533386,
+      "learning_rate": 4.763563843064167e-06,
+      "loss": 2.8481,
+      "step": 57738
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8052471876144409,
+      "learning_rate": 4.760831463836423e-06,
+      "loss": 2.8723,
+      "step": 57739
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7791904807090759,
+      "learning_rate": 4.7580998622118285e-06,
+      "loss": 2.9187,
+      "step": 57740
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7228248119354248,
+      "learning_rate": 4.755369038197576e-06,
+      "loss": 3.0401,
+      "step": 57741
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.796574056148529,
+      "learning_rate": 4.752638991800895e-06,
+      "loss": 2.9358,
+      "step": 57742
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7155166268348694,
+      "learning_rate": 4.749909723028911e-06,
+      "loss": 2.9773,
+      "step": 57743
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.731598973274231,
+      "learning_rate": 4.747181231888852e-06,
+      "loss": 2.8988,
+      "step": 57744
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7063519954681396,
+      "learning_rate": 4.74445351838788e-06,
+      "loss": 2.8026,
+      "step": 57745
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7575142979621887,
+      "learning_rate": 4.741726582533189e-06,
+      "loss": 2.9443,
+      "step": 57746
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7412130832672119,
+      "learning_rate": 4.739000424331973e-06,
+      "loss": 3.0455,
+      "step": 57747
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.700270414352417,
+      "learning_rate": 4.736275043791393e-06,
+      "loss": 2.9013,
+      "step": 57748
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8509933352470398,
+      "learning_rate": 4.733550440918676e-06,
+      "loss": 2.6319,
+      "step": 57749
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7422335147857666,
+      "learning_rate": 4.730826615720951e-06,
+      "loss": 2.8367,
+      "step": 57750
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7345046401023865,
+      "learning_rate": 4.7281035682054105e-06,
+      "loss": 3.0125,
+      "step": 57751
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7160439491271973,
+      "learning_rate": 4.725381298379216e-06,
+      "loss": 2.8065,
+      "step": 57752
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7227357625961304,
+      "learning_rate": 4.722659806249496e-06,
+      "loss": 2.978,
+      "step": 57753
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7689091563224792,
+      "learning_rate": 4.71993909182351e-06,
+      "loss": 2.9034,
+      "step": 57754
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7610191106796265,
+      "learning_rate": 4.717219155108354e-06,
+      "loss": 2.8094,
+      "step": 57755
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.777751088142395,
+      "learning_rate": 4.714499996111254e-06,
+      "loss": 2.7979,
+      "step": 57756
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8778126239776611,
+      "learning_rate": 4.711781614839305e-06,
+      "loss": 2.9273,
+      "step": 57757
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7235648036003113,
+      "learning_rate": 4.709064011299668e-06,
+      "loss": 2.98,
+      "step": 57758
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7848798036575317,
+      "learning_rate": 4.70634718549957e-06,
+      "loss": 2.9884,
+      "step": 57759
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7237803936004639,
+      "learning_rate": 4.703631137446107e-06,
+      "loss": 2.9722,
+      "step": 57760
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7664092183113098,
+      "learning_rate": 4.700915867146438e-06,
+      "loss": 2.7523,
+      "step": 57761
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7664266228675842,
+      "learning_rate": 4.698201374607724e-06,
+      "loss": 2.7736,
+      "step": 57762
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7319455742835999,
+      "learning_rate": 4.695487659837127e-06,
+      "loss": 2.956,
+      "step": 57763
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7283288836479187,
+      "learning_rate": 4.692774722841807e-06,
+      "loss": 2.8592,
+      "step": 57764
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7501567602157593,
+      "learning_rate": 4.69006256362886e-06,
+      "loss": 2.7725,
+      "step": 57765
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8033075332641602,
+      "learning_rate": 4.687351182205446e-06,
+      "loss": 2.627,
+      "step": 57766
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7505348324775696,
+      "learning_rate": 4.684640578578725e-06,
+      "loss": 2.9283,
+      "step": 57767
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7713310718536377,
+      "learning_rate": 4.681930752755791e-06,
+      "loss": 2.8058,
+      "step": 57768
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7473810315132141,
+      "learning_rate": 4.679221704743874e-06,
+      "loss": 3.0927,
+      "step": 57769
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7242414951324463,
+      "learning_rate": 4.67651343455e-06,
+      "loss": 2.848,
+      "step": 57770
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7761023640632629,
+      "learning_rate": 4.673805942181397e-06,
+      "loss": 3.0277,
+      "step": 57771
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7542916536331177,
+      "learning_rate": 4.671099227645159e-06,
+      "loss": 2.9206,
+      "step": 57772
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7616918087005615,
+      "learning_rate": 4.668393290948347e-06,
+      "loss": 3.0023,
+      "step": 57773
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7441499829292297,
+      "learning_rate": 4.665688132098222e-06,
+      "loss": 2.8658,
+      "step": 57774
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.736674964427948,
+      "learning_rate": 4.662983751101779e-06,
+      "loss": 2.8954,
+      "step": 57775
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7363643050193787,
+      "learning_rate": 4.660280147966211e-06,
+      "loss": 3.0603,
+      "step": 57776
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7268117070198059,
+      "learning_rate": 4.657577322698647e-06,
+      "loss": 2.9473,
+      "step": 57777
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7369086742401123,
+      "learning_rate": 4.65487527530618e-06,
+      "loss": 2.6568,
+      "step": 57778
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7631237506866455,
+      "learning_rate": 4.652174005795939e-06,
+      "loss": 3.0302,
+      "step": 57779
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7441269755363464,
+      "learning_rate": 4.649473514174984e-06,
+      "loss": 2.884,
+      "step": 57780
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7539127469062805,
+      "learning_rate": 4.646773800450543e-06,
+      "loss": 2.6154,
+      "step": 57781
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7788946628570557,
+      "learning_rate": 4.644074864629576e-06,
+      "loss": 2.783,
+      "step": 57782
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7208992838859558,
+      "learning_rate": 4.641376706719313e-06,
+      "loss": 2.7661,
+      "step": 57783
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7645790576934814,
+      "learning_rate": 4.6386793267268465e-06,
+      "loss": 2.8266,
+      "step": 57784
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7291238307952881,
+      "learning_rate": 4.635982724659237e-06,
+      "loss": 2.8439,
+      "step": 57785
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7457494139671326,
+      "learning_rate": 4.633286900523614e-06,
+      "loss": 2.883,
+      "step": 57786
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7361904382705688,
+      "learning_rate": 4.630591854327037e-06,
+      "loss": 3.0172,
+      "step": 57787
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7620532512664795,
+      "learning_rate": 4.627897586076667e-06,
+      "loss": 2.8486,
+      "step": 57788
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7753808498382568,
+      "learning_rate": 4.625204095779566e-06,
+      "loss": 2.9601,
+      "step": 57789
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7386515736579895,
+      "learning_rate": 4.6225113834427955e-06,
+      "loss": 2.8754,
+      "step": 57790
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7125532627105713,
+      "learning_rate": 4.619819449073548e-06,
+      "loss": 2.8305,
+      "step": 57791
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7690471410751343,
+      "learning_rate": 4.617128292678784e-06,
+      "loss": 2.8357,
+      "step": 57792
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.717765212059021,
+      "learning_rate": 4.614437914265701e-06,
+      "loss": 2.9072,
+      "step": 57793
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7456265091896057,
+      "learning_rate": 4.611748313841324e-06,
+      "loss": 2.7729,
+      "step": 57794
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7556934952735901,
+      "learning_rate": 4.6090594914127475e-06,
+      "loss": 3.028,
+      "step": 57795
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.76350337266922,
+      "learning_rate": 4.6063714469871005e-06,
+      "loss": 2.9979,
+      "step": 57796
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7897090911865234,
+      "learning_rate": 4.6036841805713765e-06,
+      "loss": 2.8443,
+      "step": 57797
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7721547484397888,
+      "learning_rate": 4.6009976921726695e-06,
+      "loss": 3.1241,
+      "step": 57798
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7226487398147583,
+      "learning_rate": 4.598311981798142e-06,
+      "loss": 2.8602,
+      "step": 57799
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7703262567520142,
+      "learning_rate": 4.59562704945482e-06,
+      "loss": 2.8151,
+      "step": 57800
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7636512517929077,
+      "learning_rate": 4.592942895149732e-06,
+      "loss": 2.9814,
+      "step": 57801
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7318775057792664,
+      "learning_rate": 4.590259518889971e-06,
+      "loss": 2.7938,
+      "step": 57802
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7615649700164795,
+      "learning_rate": 4.587576920682601e-06,
+      "loss": 2.8404,
+      "step": 57803
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7781710624694824,
+      "learning_rate": 4.584895100534747e-06,
+      "loss": 3.0723,
+      "step": 57804
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7660056948661804,
+      "learning_rate": 4.582214058453404e-06,
+      "loss": 2.945,
+      "step": 57805
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7571398019790649,
+      "learning_rate": 4.579533794445667e-06,
+      "loss": 2.8892,
+      "step": 57806
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.793786346912384,
+      "learning_rate": 4.576854308518563e-06,
+      "loss": 3.0665,
+      "step": 57807
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7984482645988464,
+      "learning_rate": 4.574175600679186e-06,
+      "loss": 2.8812,
+      "step": 57808
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7446393966674805,
+      "learning_rate": 4.571497670934565e-06,
+      "loss": 2.9753,
+      "step": 57809
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7223838567733765,
+      "learning_rate": 4.56882051929176e-06,
+      "loss": 2.9849,
+      "step": 57810
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7122632265090942,
+      "learning_rate": 4.566144145757833e-06,
+      "loss": 3.0396,
+      "step": 57811
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7451797723770142,
+      "learning_rate": 4.56346855033981e-06,
+      "loss": 2.9919,
+      "step": 57812
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7426049113273621,
+      "learning_rate": 4.560793733044754e-06,
+      "loss": 2.8361,
+      "step": 57813
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7814031839370728,
+      "learning_rate": 4.558119693879725e-06,
+      "loss": 2.8659,
+      "step": 57814
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7301146388053894,
+      "learning_rate": 4.55544643285175e-06,
+      "loss": 2.7683,
+      "step": 57815
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7373456954956055,
+      "learning_rate": 4.552773949967892e-06,
+      "loss": 3.0187,
+      "step": 57816
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7450728416442871,
+      "learning_rate": 4.5501022452351095e-06,
+      "loss": 2.842,
+      "step": 57817
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7419819831848145,
+      "learning_rate": 4.5474313186605326e-06,
+      "loss": 2.803,
+      "step": 57818
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7292543649673462,
+      "learning_rate": 4.544761170251154e-06,
+      "loss": 3.1129,
+      "step": 57819
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7770516276359558,
+      "learning_rate": 4.542091800014003e-06,
+      "loss": 2.9207,
+      "step": 57820
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7825976610183716,
+      "learning_rate": 4.5394232079561385e-06,
+      "loss": 3.0597,
+      "step": 57821
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7760852575302124,
+      "learning_rate": 4.53675539408459e-06,
+      "loss": 2.6327,
+      "step": 57822
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7495847344398499,
+      "learning_rate": 4.534088358406318e-06,
+      "loss": 2.6761,
+      "step": 57823
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7523453235626221,
+      "learning_rate": 4.531422100928417e-06,
+      "loss": 2.7543,
+      "step": 57824
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7610305547714233,
+      "learning_rate": 4.528756621657881e-06,
+      "loss": 2.9113,
+      "step": 57825
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7986166477203369,
+      "learning_rate": 4.526091920601737e-06,
+      "loss": 2.74,
+      "step": 57826
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7516844272613525,
+      "learning_rate": 4.523427997766982e-06,
+      "loss": 2.9315,
+      "step": 57827
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8079379796981812,
+      "learning_rate": 4.520764853160674e-06,
+      "loss": 3.0084,
+      "step": 57828
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8045254349708557,
+      "learning_rate": 4.518102486789843e-06,
+      "loss": 2.6265,
+      "step": 57829
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7560080289840698,
+      "learning_rate": 4.515440898661382e-06,
+      "loss": 2.8943,
+      "step": 57830
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7986272573471069,
+      "learning_rate": 4.512780088782453e-06,
+      "loss": 2.8917,
+      "step": 57831
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7806861400604248,
+      "learning_rate": 4.510120057159949e-06,
+      "loss": 2.9683,
+      "step": 57832
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7481813430786133,
+      "learning_rate": 4.507460803800933e-06,
+      "loss": 2.7959,
+      "step": 57833
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8359472155570984,
+      "learning_rate": 4.504802328712398e-06,
+      "loss": 2.8192,
+      "step": 57834
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7175796031951904,
+      "learning_rate": 4.502144631901339e-06,
+      "loss": 3.0029,
+      "step": 57835
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7417463660240173,
+      "learning_rate": 4.499487713374783e-06,
+      "loss": 3.0287,
+      "step": 57836
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7876279950141907,
+      "learning_rate": 4.496831573139692e-06,
+      "loss": 2.8647,
+      "step": 57837
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7579513788223267,
+      "learning_rate": 4.494176211203093e-06,
+      "loss": 3.0034,
+      "step": 57838
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7461996078491211,
+      "learning_rate": 4.491521627571948e-06,
+      "loss": 2.7366,
+      "step": 57839
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.753563404083252,
+      "learning_rate": 4.48886782225325e-06,
+      "loss": 3.0483,
+      "step": 57840
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7457968592643738,
+      "learning_rate": 4.4862147952539954e-06,
+      "loss": 2.7279,
+      "step": 57841
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7886890769004822,
+      "learning_rate": 4.4835625465812095e-06,
+      "loss": 3.0078,
+      "step": 57842
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7757450342178345,
+      "learning_rate": 4.4809110762418225e-06,
+      "loss": 2.752,
+      "step": 57843
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.731127917766571,
+      "learning_rate": 4.47826038424286e-06,
+      "loss": 2.929,
+      "step": 57844
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7428131103515625,
+      "learning_rate": 4.475610470591284e-06,
+      "loss": 2.9674,
+      "step": 57845
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8183236718177795,
+      "learning_rate": 4.472961335294056e-06,
+      "loss": 2.9806,
+      "step": 57846
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7529081702232361,
+      "learning_rate": 4.47031297835817e-06,
+      "loss": 2.944,
+      "step": 57847
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7067267894744873,
+      "learning_rate": 4.467665399790654e-06,
+      "loss": 2.9451,
+      "step": 57848
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.793649435043335,
+      "learning_rate": 4.465018599598369e-06,
+      "loss": 2.9698,
+      "step": 57849
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7690135836601257,
+      "learning_rate": 4.462372577788376e-06,
+      "loss": 2.6836,
+      "step": 57850
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7430137991905212,
+      "learning_rate": 4.459727334367635e-06,
+      "loss": 2.9179,
+      "step": 57851
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7567710876464844,
+      "learning_rate": 4.457082869343076e-06,
+      "loss": 3.0389,
+      "step": 57852
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7652338743209839,
+      "learning_rate": 4.454439182721692e-06,
+      "loss": 2.9754,
+      "step": 57853
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7574499845504761,
+      "learning_rate": 4.451796274510411e-06,
+      "loss": 2.7802,
+      "step": 57854
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7804326415061951,
+      "learning_rate": 4.449154144716227e-06,
+      "loss": 2.8233,
+      "step": 57855
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.777418315410614,
+      "learning_rate": 4.446512793346102e-06,
+      "loss": 2.8151,
+      "step": 57856
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7421988248825073,
+      "learning_rate": 4.443872220406963e-06,
+      "loss": 2.8111,
+      "step": 57857
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.758773922920227,
+      "learning_rate": 4.441232425905805e-06,
+      "loss": 3.0517,
+      "step": 57858
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8324280381202698,
+      "learning_rate": 4.438593409849556e-06,
+      "loss": 3.0362,
+      "step": 57859
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7193976044654846,
+      "learning_rate": 4.435955172245176e-06,
+      "loss": 2.7782,
+      "step": 57860
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7793514132499695,
+      "learning_rate": 4.433317713099593e-06,
+      "loss": 2.9797,
+      "step": 57861
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7374630570411682,
+      "learning_rate": 4.43068103241977e-06,
+      "loss": 2.6941,
+      "step": 57862
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7534642219543457,
+      "learning_rate": 4.428045130212665e-06,
+      "loss": 2.9008,
+      "step": 57863
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7688164114952087,
+      "learning_rate": 4.425410006485175e-06,
+      "loss": 3.0181,
+      "step": 57864
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7269670963287354,
+      "learning_rate": 4.4227756612442936e-06,
+      "loss": 2.9657,
+      "step": 57865
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7430902123451233,
+      "learning_rate": 4.420142094496948e-06,
+      "loss": 2.9997,
+      "step": 57866
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7281354069709778,
+      "learning_rate": 4.417509306250033e-06,
+      "loss": 2.9365,
+      "step": 57867
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7742909789085388,
+      "learning_rate": 4.414877296510544e-06,
+      "loss": 2.9117,
+      "step": 57868
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7314521670341492,
+      "learning_rate": 4.412246065285341e-06,
+      "loss": 2.8858,
+      "step": 57869
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7391765713691711,
+      "learning_rate": 4.409615612581418e-06,
+      "loss": 3.1135,
+      "step": 57870
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7244545221328735,
+      "learning_rate": 4.406985938405672e-06,
+      "loss": 2.8085,
+      "step": 57871
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7341313362121582,
+      "learning_rate": 4.404357042765028e-06,
+      "loss": 2.8818,
+      "step": 57872
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7704923152923584,
+      "learning_rate": 4.4017289256664476e-06,
+      "loss": 2.8435,
+      "step": 57873
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8078938722610474,
+      "learning_rate": 4.399101587116827e-06,
+      "loss": 3.1185,
+      "step": 57874
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7866788506507874,
+      "learning_rate": 4.396475027123059e-06,
+      "loss": 2.8965,
+      "step": 57875
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7199814915657043,
+      "learning_rate": 4.393849245692105e-06,
+      "loss": 2.9652,
+      "step": 57876
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7180720567703247,
+      "learning_rate": 4.391224242830827e-06,
+      "loss": 2.8224,
+      "step": 57877
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.731706976890564,
+      "learning_rate": 4.388600018546218e-06,
+      "loss": 2.8166,
+      "step": 57878
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7510423064231873,
+      "learning_rate": 4.385976572845107e-06,
+      "loss": 2.8502,
+      "step": 57879
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7061409950256348,
+      "learning_rate": 4.383353905734488e-06,
+      "loss": 2.8161,
+      "step": 57880
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.756230354309082,
+      "learning_rate": 4.380732017221189e-06,
+      "loss": 2.8982,
+      "step": 57881
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7549283504486084,
+      "learning_rate": 4.378110907312138e-06,
+      "loss": 2.9815,
+      "step": 57882
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7772846817970276,
+      "learning_rate": 4.375490576014296e-06,
+      "loss": 2.7412,
+      "step": 57883
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7680914998054504,
+      "learning_rate": 4.372871023334457e-06,
+      "loss": 2.6235,
+      "step": 57884
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7407665848731995,
+      "learning_rate": 4.370252249279615e-06,
+      "loss": 2.8474,
+      "step": 57885
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7720359563827515,
+      "learning_rate": 4.367634253856633e-06,
+      "loss": 2.807,
+      "step": 57886
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7055439352989197,
+      "learning_rate": 4.365017037072405e-06,
+      "loss": 2.6587,
+      "step": 57887
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8091383576393127,
+      "learning_rate": 4.362400598933824e-06,
+      "loss": 2.8048,
+      "step": 57888
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7102215886116028,
+      "learning_rate": 4.3597849394477855e-06,
+      "loss": 2.9037,
+      "step": 57889
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7323922514915466,
+      "learning_rate": 4.357170058621151e-06,
+      "loss": 2.7498,
+      "step": 57890
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7890856266021729,
+      "learning_rate": 4.354555956460881e-06,
+      "loss": 2.8841,
+      "step": 57891
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.782171905040741,
+      "learning_rate": 4.3519426329737705e-06,
+      "loss": 2.9299,
+      "step": 57892
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7346777319908142,
+      "learning_rate": 4.349330088166781e-06,
+      "loss": 2.9818,
+      "step": 57893
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7390797138214111,
+      "learning_rate": 4.346718322046705e-06,
+      "loss": 2.9289,
+      "step": 57894
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.710119366645813,
+      "learning_rate": 4.344107334620539e-06,
+      "loss": 2.9441,
+      "step": 57895
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8123124241828918,
+      "learning_rate": 4.341497125895044e-06,
+      "loss": 3.1647,
+      "step": 57896
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7103381156921387,
+      "learning_rate": 4.33888769587718e-06,
+      "loss": 2.8699,
+      "step": 57897
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.740925669670105,
+      "learning_rate": 4.336279044573776e-06,
+      "loss": 2.9577,
+      "step": 57898
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7418505549430847,
+      "learning_rate": 4.33367117199166e-06,
+      "loss": 2.951,
+      "step": 57899
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7685325145721436,
+      "learning_rate": 4.331064078137791e-06,
+      "loss": 2.8862,
+      "step": 57900
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.787038266658783,
+      "learning_rate": 4.328457763019033e-06,
+      "loss": 2.7146,
+      "step": 57901
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7468493580818176,
+      "learning_rate": 4.325852226642179e-06,
+      "loss": 2.8801,
+      "step": 57902
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7677332162857056,
+      "learning_rate": 4.323247469014124e-06,
+      "loss": 2.8442,
+      "step": 57903
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7528624534606934,
+      "learning_rate": 4.320643490141729e-06,
+      "loss": 2.6777,
+      "step": 57904
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7097703218460083,
+      "learning_rate": 4.318040290031888e-06,
+      "loss": 2.7669,
+      "step": 57905
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.6983256936073303,
+      "learning_rate": 4.315437868691363e-06,
+      "loss": 2.9228,
+      "step": 57906
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8027907013893127,
+      "learning_rate": 4.312836226127115e-06,
+      "loss": 2.9804,
+      "step": 57907
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7508549690246582,
+      "learning_rate": 4.3102353623459376e-06,
+      "loss": 2.9204,
+      "step": 57908
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7843279242515564,
+      "learning_rate": 4.3076352773546925e-06,
+      "loss": 2.8192,
+      "step": 57909
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7532942295074463,
+      "learning_rate": 4.305035971160242e-06,
+      "loss": 2.9188,
+      "step": 57910
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7582915425300598,
+      "learning_rate": 4.3024374437694135e-06,
+      "loss": 2.9539,
+      "step": 57911
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7695998549461365,
+      "learning_rate": 4.299839695189034e-06,
+      "loss": 2.9009,
+      "step": 57912
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.739365816116333,
+      "learning_rate": 4.2972427254259974e-06,
+      "loss": 2.8835,
+      "step": 57913
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.745867133140564,
+      "learning_rate": 4.294646534487068e-06,
+      "loss": 2.8355,
+      "step": 57914
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7075350880622864,
+      "learning_rate": 4.2920511223791695e-06,
+      "loss": 2.8435,
+      "step": 57915
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8138481974601746,
+      "learning_rate": 4.2894564891091e-06,
+      "loss": 3.016,
+      "step": 57916
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9259201288223267,
+      "learning_rate": 4.286862634683685e-06,
+      "loss": 2.7868,
+      "step": 57917
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7305136322975159,
+      "learning_rate": 4.284269559109787e-06,
+      "loss": 2.81,
+      "step": 57918
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7330555319786072,
+      "learning_rate": 4.2816772623941675e-06,
+      "loss": 2.6985,
+      "step": 57919
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7787506580352783,
+      "learning_rate": 4.27908574454372e-06,
+      "loss": 2.8231,
+      "step": 57920
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7437611222267151,
+      "learning_rate": 4.2764950055652395e-06,
+      "loss": 2.8805,
+      "step": 57921
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7861391305923462,
+      "learning_rate": 4.273905045465553e-06,
+      "loss": 2.9361,
+      "step": 57922
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7396217584609985,
+      "learning_rate": 4.27131586425149e-06,
+      "loss": 2.7683,
+      "step": 57923
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7936074137687683,
+      "learning_rate": 4.268727461929877e-06,
+      "loss": 2.794,
+      "step": 57924
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7759957313537598,
+      "learning_rate": 4.2661398385075426e-06,
+      "loss": 2.7761,
+      "step": 57925
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7496296167373657,
+      "learning_rate": 4.263552993991248e-06,
+      "loss": 2.6912,
+      "step": 57926
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7550476789474487,
+      "learning_rate": 4.2609669283878215e-06,
+      "loss": 2.8715,
+      "step": 57927
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7111081480979919,
+      "learning_rate": 4.258381641704123e-06,
+      "loss": 2.7266,
+      "step": 57928
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7496201992034912,
+      "learning_rate": 4.255797133946915e-06,
+      "loss": 2.8366,
+      "step": 57929
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.802704393863678,
+      "learning_rate": 4.253213405123057e-06,
+      "loss": 2.9926,
+      "step": 57930
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7297253608703613,
+      "learning_rate": 4.250630455239312e-06,
+      "loss": 3.123,
+      "step": 57931
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.71900874376297,
+      "learning_rate": 4.248048284302441e-06,
+      "loss": 2.8253,
+      "step": 57932
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7305431962013245,
+      "learning_rate": 4.245466892319338e-06,
+      "loss": 2.8313,
+      "step": 57933
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7327029705047607,
+      "learning_rate": 4.242886279296731e-06,
+      "loss": 2.9302,
+      "step": 57934
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7928759455680847,
+      "learning_rate": 4.240306445241448e-06,
+      "loss": 2.8044,
+      "step": 57935
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.731713056564331,
+      "learning_rate": 4.237727390160251e-06,
+      "loss": 3.0344,
+      "step": 57936
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.736788809299469,
+      "learning_rate": 4.235149114059966e-06,
+      "loss": 2.8996,
+      "step": 57937
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7987184524536133,
+      "learning_rate": 4.232571616947422e-06,
+      "loss": 2.8778,
+      "step": 57938
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7894659638404846,
+      "learning_rate": 4.229994898829348e-06,
+      "loss": 2.9336,
+      "step": 57939
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7339268922805786,
+      "learning_rate": 4.227418959712536e-06,
+      "loss": 2.855,
+      "step": 57940
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7853479385375977,
+      "learning_rate": 4.224843799603783e-06,
+      "loss": 2.9414,
+      "step": 57941
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8146184086799622,
+      "learning_rate": 4.2222694185098495e-06,
+      "loss": 2.8615,
+      "step": 57942
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.825524091720581,
+      "learning_rate": 4.2196958164375625e-06,
+      "loss": 2.9887,
+      "step": 57943
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7087740898132324,
+      "learning_rate": 4.2171229933936515e-06,
+      "loss": 2.742,
+      "step": 57944
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7632535099983215,
+      "learning_rate": 4.21455094938491e-06,
+      "loss": 2.7973,
+      "step": 57945
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7354146838188171,
+      "learning_rate": 4.211979684418165e-06,
+      "loss": 3.1476,
+      "step": 57946
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7568386793136597,
+      "learning_rate": 4.20940919850008e-06,
+      "loss": 2.8029,
+      "step": 57947
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7767660617828369,
+      "learning_rate": 4.206839491637515e-06,
+      "loss": 2.9156,
+      "step": 57948
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7283722162246704,
+      "learning_rate": 4.204270563837198e-06,
+      "loss": 2.8352,
+      "step": 57949
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7674680948257446,
+      "learning_rate": 4.201702415105923e-06,
+      "loss": 2.6455,
+      "step": 57950
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.751236081123352,
+      "learning_rate": 4.199135045450386e-06,
+      "loss": 2.8124,
+      "step": 57951
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7168905735015869,
+      "learning_rate": 4.196568454877447e-06,
+      "loss": 2.817,
+      "step": 57952
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7038781046867371,
+      "learning_rate": 4.194002643393834e-06,
+      "loss": 2.5922,
+      "step": 57953
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.790746808052063,
+      "learning_rate": 4.191437611006243e-06,
+      "loss": 2.7898,
+      "step": 57954
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7640130519866943,
+      "learning_rate": 4.188873357721501e-06,
+      "loss": 2.9048,
+      "step": 57955
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7775148749351501,
+      "learning_rate": 4.186309883546335e-06,
+      "loss": 2.9071,
+      "step": 57956
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7022102475166321,
+      "learning_rate": 4.183747188487474e-06,
+      "loss": 2.9208,
+      "step": 57957
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8010618686676025,
+      "learning_rate": 4.181185272551712e-06,
+      "loss": 2.9989,
+      "step": 57958
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7519500851631165,
+      "learning_rate": 4.178624135745779e-06,
+      "loss": 2.8834,
+      "step": 57959
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.745065450668335,
+      "learning_rate": 4.176063778076433e-06,
+      "loss": 2.8246,
+      "step": 57960
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7399493455886841,
+      "learning_rate": 4.17350419955037e-06,
+      "loss": 2.997,
+      "step": 57961
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7513877153396606,
+      "learning_rate": 4.170945400174386e-06,
+      "loss": 2.8798,
+      "step": 57962
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7985098958015442,
+      "learning_rate": 4.168387379955207e-06,
+      "loss": 2.7076,
+      "step": 57963
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7976999878883362,
+      "learning_rate": 4.1658301388995285e-06,
+      "loss": 2.992,
+      "step": 57964
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7618657946586609,
+      "learning_rate": 4.163273677014145e-06,
+      "loss": 2.8952,
+      "step": 57965
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9525282979011536,
+      "learning_rate": 4.1607179943057176e-06,
+      "loss": 2.7617,
+      "step": 57966
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7083631753921509,
+      "learning_rate": 4.1581630907810745e-06,
+      "loss": 2.9582,
+      "step": 57967
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7903333902359009,
+      "learning_rate": 4.155608966446911e-06,
+      "loss": 2.9069,
+      "step": 57968
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7692322134971619,
+      "learning_rate": 4.153055621309886e-06,
+      "loss": 2.8151,
+      "step": 57969
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8629139065742493,
+      "learning_rate": 4.150503055376797e-06,
+      "loss": 2.9081,
+      "step": 57970
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7383148074150085,
+      "learning_rate": 4.147951268654337e-06,
+      "loss": 2.7084,
+      "step": 57971
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7752810716629028,
+      "learning_rate": 4.145400261149235e-06,
+      "loss": 2.7297,
+      "step": 57972
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7584665417671204,
+      "learning_rate": 4.142850032868217e-06,
+      "loss": 2.7295,
+      "step": 57973
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7689148783683777,
+      "learning_rate": 4.140300583817979e-06,
+      "loss": 2.9014,
+      "step": 57974
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7338578104972839,
+      "learning_rate": 4.137751914005283e-06,
+      "loss": 2.6482,
+      "step": 57975
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7252892255783081,
+      "learning_rate": 4.135204023436822e-06,
+      "loss": 3.0192,
+      "step": 57976
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7320716977119446,
+      "learning_rate": 4.132656912119226e-06,
+      "loss": 2.8298,
+      "step": 57977
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7503402829170227,
+      "learning_rate": 4.130110580059321e-06,
+      "loss": 2.7203,
+      "step": 57978
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.756325364112854,
+      "learning_rate": 4.127565027263735e-06,
+      "loss": 2.9428,
+      "step": 57979
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7448549270629883,
+      "learning_rate": 4.125020253739231e-06,
+      "loss": 2.8511,
+      "step": 57980
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7655499577522278,
+      "learning_rate": 4.122476259492469e-06,
+      "loss": 2.8666,
+      "step": 57981
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7265483736991882,
+      "learning_rate": 4.119933044530177e-06,
+      "loss": 2.8987,
+      "step": 57982
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7448196411132812,
+      "learning_rate": 4.117390608859017e-06,
+      "loss": 2.792,
+      "step": 57983
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7171865105628967,
+      "learning_rate": 4.114848952485683e-06,
+      "loss": 3.0098,
+      "step": 57984
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8641087412834167,
+      "learning_rate": 4.112308075416937e-06,
+      "loss": 2.7329,
+      "step": 57985
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7868609428405762,
+      "learning_rate": 4.109767977659406e-06,
+      "loss": 2.8346,
+      "step": 57986
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7726017236709595,
+      "learning_rate": 4.107228659219819e-06,
+      "loss": 2.8714,
+      "step": 57987
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7346259951591492,
+      "learning_rate": 4.104690120104803e-06,
+      "loss": 2.9884,
+      "step": 57988
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7412421107292175,
+      "learning_rate": 4.102152360321121e-06,
+      "loss": 2.9378,
+      "step": 57989
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7275198101997375,
+      "learning_rate": 4.099615379875432e-06,
+      "loss": 2.8435,
+      "step": 57990
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7367424368858337,
+      "learning_rate": 4.097079178774399e-06,
+      "loss": 2.8953,
+      "step": 57991
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7552502751350403,
+      "learning_rate": 4.094543757024715e-06,
+      "loss": 2.7933,
+      "step": 57992
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7371522188186646,
+      "learning_rate": 4.09200911463301e-06,
+      "loss": 3.0965,
+      "step": 57993
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.785754919052124,
+      "learning_rate": 4.089475251606045e-06,
+      "loss": 2.8669,
+      "step": 57994
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.793599009513855,
+      "learning_rate": 4.086942167950446e-06,
+      "loss": 2.6429,
+      "step": 57995
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7560648918151855,
+      "learning_rate": 4.084409863672877e-06,
+      "loss": 3.0258,
+      "step": 57996
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7576151490211487,
+      "learning_rate": 4.081878338780065e-06,
+      "loss": 2.9662,
+      "step": 57997
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7509180903434753,
+      "learning_rate": 4.079347593278604e-06,
+      "loss": 2.7924,
+      "step": 57998
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7320362329483032,
+      "learning_rate": 4.076817627175189e-06,
+      "loss": 3.1727,
+      "step": 57999
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7936714887619019,
+      "learning_rate": 4.074288440476514e-06,
+      "loss": 2.9039,
+      "step": 58000
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7300601005554199,
+      "learning_rate": 4.071760033189175e-06,
+      "loss": 2.7854,
+      "step": 58001
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8125167489051819,
+      "learning_rate": 4.0692324053199e-06,
+      "loss": 2.8679,
+      "step": 58002
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.768974244594574,
+      "learning_rate": 4.066705556875283e-06,
+      "loss": 2.8779,
+      "step": 58003
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8136953711509705,
+      "learning_rate": 4.0641794878620184e-06,
+      "loss": 2.9216,
+      "step": 58004
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7406480312347412,
+      "learning_rate": 4.061654198286768e-06,
+      "loss": 2.853,
+      "step": 58005
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7325662970542908,
+      "learning_rate": 4.059129688156127e-06,
+      "loss": 2.8341,
+      "step": 58006
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7513902187347412,
+      "learning_rate": 4.056605957476822e-06,
+      "loss": 3.0636,
+      "step": 58007
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7709435224533081,
+      "learning_rate": 4.0540830062554485e-06,
+      "loss": 2.8569,
+      "step": 58008
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7840178608894348,
+      "learning_rate": 4.051560834498668e-06,
+      "loss": 2.9059,
+      "step": 58009
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7339916825294495,
+      "learning_rate": 4.049039442213109e-06,
+      "loss": 3.0242,
+      "step": 58010
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7548692226409912,
+      "learning_rate": 4.046518829405432e-06,
+      "loss": 2.9171,
+      "step": 58011
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7953229546546936,
+      "learning_rate": 4.0439989960822985e-06,
+      "loss": 3.0365,
+      "step": 58012
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7593258023262024,
+      "learning_rate": 4.04147994225027e-06,
+      "loss": 2.9056,
+      "step": 58013
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7055078744888306,
+      "learning_rate": 4.0389616679160074e-06,
+      "loss": 2.7517,
+      "step": 58014
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7557649612426758,
+      "learning_rate": 4.036444173086206e-06,
+      "loss": 2.7834,
+      "step": 58015
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7741184234619141,
+      "learning_rate": 4.0339274577674275e-06,
+      "loss": 2.812,
+      "step": 58016
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7468675374984741,
+      "learning_rate": 4.031411521966332e-06,
+      "loss": 3.0378,
+      "step": 58017
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7847301959991455,
+      "learning_rate": 4.028896365689516e-06,
+      "loss": 2.9056,
+      "step": 58018
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7446919083595276,
+      "learning_rate": 4.026381988943672e-06,
+      "loss": 2.8941,
+      "step": 58019
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7204757332801819,
+      "learning_rate": 4.0238683917353634e-06,
+      "loss": 2.852,
+      "step": 58020
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7593426704406738,
+      "learning_rate": 4.021355574071217e-06,
+      "loss": 2.8235,
+      "step": 58021
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7240512371063232,
+      "learning_rate": 4.018843535957861e-06,
+      "loss": 2.7313,
+      "step": 58022
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7632853388786316,
+      "learning_rate": 4.01633227740189e-06,
+      "loss": 2.9329,
+      "step": 58023
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.753861665725708,
+      "learning_rate": 4.013821798409933e-06,
+      "loss": 2.913,
+      "step": 58024
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8168571591377258,
+      "learning_rate": 4.0113120989886176e-06,
+      "loss": 2.8373,
+      "step": 58025
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7445526719093323,
+      "learning_rate": 4.008803179144571e-06,
+      "loss": 2.951,
+      "step": 58026
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7250993251800537,
+      "learning_rate": 4.006295038884355e-06,
+      "loss": 3.1706,
+      "step": 58027
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7294281721115112,
+      "learning_rate": 4.003787678214565e-06,
+      "loss": 2.9983,
+      "step": 58028
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8010278344154358,
+      "learning_rate": 4.0012810971418285e-06,
+      "loss": 2.7596,
+      "step": 58029
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7343051433563232,
+      "learning_rate": 3.998775295672773e-06,
+      "loss": 2.9007,
+      "step": 58030
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7521501779556274,
+      "learning_rate": 3.9962702738139955e-06,
+      "loss": 2.9278,
+      "step": 58031
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7345162630081177,
+      "learning_rate": 3.993766031572055e-06,
+      "loss": 3.1938,
+      "step": 58032
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7414243817329407,
+      "learning_rate": 3.9912625689535795e-06,
+      "loss": 2.9453,
+      "step": 58033
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7303183674812317,
+      "learning_rate": 3.988759885965131e-06,
+      "loss": 2.8327,
+      "step": 58034
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7680356502532959,
+      "learning_rate": 3.986257982613339e-06,
+      "loss": 2.9224,
+      "step": 58035
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7452159523963928,
+      "learning_rate": 3.983756858904763e-06,
+      "loss": 2.9883,
+      "step": 58036
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7528615593910217,
+      "learning_rate": 3.981256514845999e-06,
+      "loss": 2.7967,
+      "step": 58037
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7940719723701477,
+      "learning_rate": 3.97875695044364e-06,
+      "loss": 2.8682,
+      "step": 58038
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7146586179733276,
+      "learning_rate": 3.9762581657042496e-06,
+      "loss": 3.025,
+      "step": 58039
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7448782920837402,
+      "learning_rate": 3.973760160634453e-06,
+      "loss": 2.828,
+      "step": 58040
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7157915830612183,
+      "learning_rate": 3.971262935240782e-06,
+      "loss": 2.9029,
+      "step": 58041
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.824224591255188,
+      "learning_rate": 3.968766489529862e-06,
+      "loss": 2.7699,
+      "step": 58042
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7506048679351807,
+      "learning_rate": 3.966270823508189e-06,
+      "loss": 2.8328,
+      "step": 58043
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7083612084388733,
+      "learning_rate": 3.963775937182423e-06,
+      "loss": 2.8861,
+      "step": 58044
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7859985828399658,
+      "learning_rate": 3.9612818305590934e-06,
+      "loss": 2.6817,
+      "step": 58045
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7196922898292542,
+      "learning_rate": 3.958788503644761e-06,
+      "loss": 3.0316,
+      "step": 58046
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7509509325027466,
+      "learning_rate": 3.95629595644602e-06,
+      "loss": 2.9547,
+      "step": 58047
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7594537138938904,
+      "learning_rate": 3.953804188969401e-06,
+      "loss": 3.013,
+      "step": 58048
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7385445237159729,
+      "learning_rate": 3.951313201221495e-06,
+      "loss": 2.6765,
+      "step": 58049
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7691807746887207,
+      "learning_rate": 3.9488229932088665e-06,
+      "loss": 2.778,
+      "step": 58050
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8221585154533386,
+      "learning_rate": 3.946333564938042e-06,
+      "loss": 2.7925,
+      "step": 58051
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.7647941708564758,
+      "learning_rate": 3.943844916415617e-06,
+      "loss": 2.9116,
+      "step": 58052
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7364143133163452,
+      "learning_rate": 3.941357047648119e-06,
+      "loss": 2.872,
+      "step": 58053
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7513567805290222,
+      "learning_rate": 3.9388699586421104e-06,
+      "loss": 2.9102,
+      "step": 58054
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8263747692108154,
+      "learning_rate": 3.936383649404151e-06,
+      "loss": 3.1531,
+      "step": 58055
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7924360632896423,
+      "learning_rate": 3.93389811994077e-06,
+      "loss": 3.0245,
+      "step": 58056
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7682927250862122,
+      "learning_rate": 3.931413370258529e-06,
+      "loss": 2.8791,
+      "step": 58057
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7378009557723999,
+      "learning_rate": 3.928929400363922e-06,
+      "loss": 2.8736,
+      "step": 58058
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.742211103439331,
+      "learning_rate": 3.926446210263579e-06,
+      "loss": 2.6486,
+      "step": 58059
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7150681614875793,
+      "learning_rate": 3.923963799963992e-06,
+      "loss": 3.034,
+      "step": 58060
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.726213276386261,
+      "learning_rate": 3.92148216947169e-06,
+      "loss": 2.9354,
+      "step": 58061
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7555464506149292,
+      "learning_rate": 3.919001318793235e-06,
+      "loss": 2.6879,
+      "step": 58062
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7286418080329895,
+      "learning_rate": 3.916521247935156e-06,
+      "loss": 2.8126,
+      "step": 58063
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7420178651809692,
+      "learning_rate": 3.914041956903979e-06,
+      "loss": 2.7675,
+      "step": 58064
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7674248814582825,
+      "learning_rate": 3.9115634457062e-06,
+      "loss": 3.0303,
+      "step": 58065
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7217745184898376,
+      "learning_rate": 3.909085714348381e-06,
+      "loss": 2.9029,
+      "step": 58066
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8204011917114258,
+      "learning_rate": 3.906608762837082e-06,
+      "loss": 2.9336,
+      "step": 58067
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.744408369064331,
+      "learning_rate": 3.904132591178766e-06,
+      "loss": 2.9035,
+      "step": 58068
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7998400926589966,
+      "learning_rate": 3.901657199379993e-06,
+      "loss": 2.8445,
+      "step": 58069
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7387979030609131,
+      "learning_rate": 3.899182587447258e-06,
+      "loss": 2.7588,
+      "step": 58070
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7533370852470398,
+      "learning_rate": 3.89670875538709e-06,
+      "loss": 2.8385,
+      "step": 58071
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7749383449554443,
+      "learning_rate": 3.8942357032060166e-06,
+      "loss": 2.8858,
+      "step": 58072
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7244678735733032,
+      "learning_rate": 3.891763430910533e-06,
+      "loss": 2.9102,
+      "step": 58073
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7526236176490784,
+      "learning_rate": 3.8892919385071664e-06,
+      "loss": 2.7091,
+      "step": 58074
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7498911619186401,
+      "learning_rate": 3.886821226002379e-06,
+      "loss": 2.7809,
+      "step": 58075
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8182533383369446,
+      "learning_rate": 3.884351293402766e-06,
+      "loss": 2.8096,
+      "step": 58076
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7636764645576477,
+      "learning_rate": 3.8818821407147875e-06,
+      "loss": 2.8166,
+      "step": 58077
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7577216625213623,
+      "learning_rate": 3.879413767944905e-06,
+      "loss": 3.0238,
+      "step": 58078
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7836326360702515,
+      "learning_rate": 3.876946175099682e-06,
+      "loss": 2.8152,
+      "step": 58079
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7404009699821472,
+      "learning_rate": 3.874479362185545e-06,
+      "loss": 2.831,
+      "step": 58080
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7890130877494812,
+      "learning_rate": 3.872013329209089e-06,
+      "loss": 2.7029,
+      "step": 58081
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.821405827999115,
+      "learning_rate": 3.869548076176743e-06,
+      "loss": 2.731,
+      "step": 58082
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7491986155509949,
+      "learning_rate": 3.867083603095e-06,
+      "loss": 2.6854,
+      "step": 58083
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7356759309768677,
+      "learning_rate": 3.86461990997039e-06,
+      "loss": 3.0514,
+      "step": 58084
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7587080597877502,
+      "learning_rate": 3.862156996809407e-06,
+      "loss": 3.1083,
+      "step": 58085
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7740174531936646,
+      "learning_rate": 3.859694863618445e-06,
+      "loss": 2.9711,
+      "step": 58086
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.761070191860199,
+      "learning_rate": 3.857233510404101e-06,
+      "loss": 3.0597,
+      "step": 58087
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7623142004013062,
+      "learning_rate": 3.854772937172768e-06,
+      "loss": 2.7424,
+      "step": 58088
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7792325615882874,
+      "learning_rate": 3.852313143931007e-06,
+      "loss": 2.8965,
+      "step": 58089
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7377564311027527,
+      "learning_rate": 3.849854130685248e-06,
+      "loss": 2.9561,
+      "step": 58090
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7853538393974304,
+      "learning_rate": 3.847395897441985e-06,
+      "loss": 3.1712,
+      "step": 58091
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7288347482681274,
+      "learning_rate": 3.844938444207679e-06,
+      "loss": 2.7964,
+      "step": 58092
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7623826265335083,
+      "learning_rate": 3.842481770988792e-06,
+      "loss": 2.8275,
+      "step": 58093
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7762326598167419,
+      "learning_rate": 3.840025877791852e-06,
+      "loss": 2.961,
+      "step": 58094
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7291229963302612,
+      "learning_rate": 3.837570764623254e-06,
+      "loss": 2.9422,
+      "step": 58095
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7547798752784729,
+      "learning_rate": 3.8351164314894935e-06,
+      "loss": 2.7757,
+      "step": 58096
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7603956460952759,
+      "learning_rate": 3.8326628783970306e-06,
+      "loss": 3.2117,
+      "step": 58097
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7306252121925354,
+      "learning_rate": 3.8302101053523605e-06,
+      "loss": 2.7616,
+      "step": 58098
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7425011992454529,
+      "learning_rate": 3.8277581123619114e-06,
+      "loss": 2.9913,
+      "step": 58099
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8286489248275757,
+      "learning_rate": 3.825306899432145e-06,
+      "loss": 2.7551,
+      "step": 58100
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7279012799263,
+      "learning_rate": 3.8228564665695235e-06,
+      "loss": 2.9012,
+      "step": 58101
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.713568925857544,
+      "learning_rate": 3.820406813780508e-06,
+      "loss": 2.9402,
+      "step": 58102
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7257249355316162,
+      "learning_rate": 3.8179579410715255e-06,
+      "loss": 2.6947,
+      "step": 58103
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7878603935241699,
+      "learning_rate": 3.815509848449039e-06,
+      "loss": 2.8945,
+      "step": 58104
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7164974808692932,
+      "learning_rate": 3.8130625359194756e-06,
+      "loss": 3.0317,
+      "step": 58105
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.731782853603363,
+      "learning_rate": 3.810616003489364e-06,
+      "loss": 2.9602,
+      "step": 58106
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7645469903945923,
+      "learning_rate": 3.808170251165066e-06,
+      "loss": 2.8414,
+      "step": 58107
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7678006887435913,
+      "learning_rate": 3.805725278953009e-06,
+      "loss": 2.8792,
+      "step": 58108
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.727119505405426,
+      "learning_rate": 3.8032810868596888e-06,
+      "loss": 3.0004,
+      "step": 58109
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8259202837944031,
+      "learning_rate": 3.8008376748915324e-06,
+      "loss": 2.7632,
+      "step": 58110
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7696822285652161,
+      "learning_rate": 3.798395043054936e-06,
+      "loss": 2.7003,
+      "step": 58111
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.753521203994751,
+      "learning_rate": 3.7959531913563932e-06,
+      "loss": 2.9795,
+      "step": 58112
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7405487298965454,
+      "learning_rate": 3.7935121198023e-06,
+      "loss": 2.8325,
+      "step": 58113
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7650240063667297,
+      "learning_rate": 3.7910718283990837e-06,
+      "loss": 3.0367,
+      "step": 58114
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7302567958831787,
+      "learning_rate": 3.788632317153173e-06,
+      "loss": 2.7721,
+      "step": 58115
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7484518885612488,
+      "learning_rate": 3.786193586070996e-06,
+      "loss": 2.791,
+      "step": 58116
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7105563282966614,
+      "learning_rate": 3.7837556351589802e-06,
+      "loss": 2.9578,
+      "step": 58117
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7681797742843628,
+      "learning_rate": 3.7813184644235217e-06,
+      "loss": 3.0148,
+      "step": 58118
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7496532797813416,
+      "learning_rate": 3.7788820738711145e-06,
+      "loss": 3.0347,
+      "step": 58119
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7610386610031128,
+      "learning_rate": 3.7764464635080873e-06,
+      "loss": 2.8305,
+      "step": 58120
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7656939625740051,
+      "learning_rate": 3.774011633340901e-06,
+      "loss": 2.881,
+      "step": 58121
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7909092307090759,
+      "learning_rate": 3.771577583375951e-06,
+      "loss": 2.9376,
+      "step": 58122
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7618745565414429,
+      "learning_rate": 3.7691443136196653e-06,
+      "loss": 2.592,
+      "step": 58123
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7755120992660522,
+      "learning_rate": 3.766711824078439e-06,
+      "loss": 2.8602,
+      "step": 58124
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7302433252334595,
+      "learning_rate": 3.7642801147586664e-06,
+      "loss": 2.7635,
+      "step": 58125
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7398534417152405,
+      "learning_rate": 3.7618491856667764e-06,
+      "loss": 2.9394,
+      "step": 58126
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7428982853889465,
+      "learning_rate": 3.7594190368091638e-06,
+      "loss": 2.8178,
+      "step": 58127
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7572886347770691,
+      "learning_rate": 3.756989668192256e-06,
+      "loss": 2.7583,
+      "step": 58128
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7500163912773132,
+      "learning_rate": 3.7545610798224156e-06,
+      "loss": 2.7213,
+      "step": 58129
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7317116260528564,
+      "learning_rate": 3.7521332717060372e-06,
+      "loss": 2.727,
+      "step": 58130
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7358356714248657,
+      "learning_rate": 3.749706243849482e-06,
+      "loss": 2.8328,
+      "step": 58131
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7284991145133972,
+      "learning_rate": 3.747279996259245e-06,
+      "loss": 2.8797,
+      "step": 58132
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7599244713783264,
+      "learning_rate": 3.744854528941621e-06,
+      "loss": 2.8693,
+      "step": 58133
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.791732907295227,
+      "learning_rate": 3.742429841903072e-06,
+      "loss": 2.6926,
+      "step": 58134
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7750145792961121,
+      "learning_rate": 3.740005935149959e-06,
+      "loss": 3.0105,
+      "step": 58135
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7046689391136169,
+      "learning_rate": 3.7375828086886106e-06,
+      "loss": 2.9164,
+      "step": 58136
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.729168176651001,
+      "learning_rate": 3.735160462525455e-06,
+      "loss": 3.0215,
+      "step": 58137
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7688641548156738,
+      "learning_rate": 3.732738896666887e-06,
+      "loss": 2.8869,
+      "step": 58138
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7332656383514404,
+      "learning_rate": 3.730318111119268e-06,
+      "loss": 2.721,
+      "step": 58139
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7618536353111267,
+      "learning_rate": 3.72789810588896e-06,
+      "loss": 2.9654,
+      "step": 58140
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7432697415351868,
+      "learning_rate": 3.7254788809823577e-06,
+      "loss": 2.8195,
+      "step": 58141
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7419174313545227,
+      "learning_rate": 3.7230604364057894e-06,
+      "loss": 2.9438,
+      "step": 58142
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7732837200164795,
+      "learning_rate": 3.7206427721657162e-06,
+      "loss": 2.9231,
+      "step": 58143
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7398931980133057,
+      "learning_rate": 3.7182258882684333e-06,
+      "loss": 2.7001,
+      "step": 58144
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7343015670776367,
+      "learning_rate": 3.715809784720303e-06,
+      "loss": 2.7543,
+      "step": 58145
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7587280869483948,
+      "learning_rate": 3.713394461527719e-06,
+      "loss": 2.8208,
+      "step": 58146
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.74312824010849,
+      "learning_rate": 3.7109799186970437e-06,
+      "loss": 2.7435,
+      "step": 58147
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7362629175186157,
+      "learning_rate": 3.708566156234605e-06,
+      "loss": 2.8137,
+      "step": 58148
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7702608108520508,
+      "learning_rate": 3.7061531741467975e-06,
+      "loss": 2.9806,
+      "step": 58149
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7515386343002319,
+      "learning_rate": 3.7037409724399504e-06,
+      "loss": 2.8289,
+      "step": 58150
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.762421190738678,
+      "learning_rate": 3.7013295511204576e-06,
+      "loss": 2.8378,
+      "step": 58151
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7281259298324585,
+      "learning_rate": 3.6989189101945817e-06,
+      "loss": 2.8135,
+      "step": 58152
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.782008171081543,
+      "learning_rate": 3.69650904966875e-06,
+      "loss": 3.1768,
+      "step": 58153
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7807068824768066,
+      "learning_rate": 3.6940999695493247e-06,
+      "loss": 2.9561,
+      "step": 58154
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8265916705131531,
+      "learning_rate": 3.691691669842567e-06,
+      "loss": 2.907,
+      "step": 58155
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7538347244262695,
+      "learning_rate": 3.689284150554905e-06,
+      "loss": 2.9961,
+      "step": 58156
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7269682288169861,
+      "learning_rate": 3.6868774116926346e-06,
+      "loss": 2.894,
+      "step": 58157
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7205626964569092,
+      "learning_rate": 3.6844714532620836e-06,
+      "loss": 2.9546,
+      "step": 58158
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7215340733528137,
+      "learning_rate": 3.6820662752696127e-06,
+      "loss": 2.8961,
+      "step": 58159
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7290651202201843,
+      "learning_rate": 3.6796618777215516e-06,
+      "loss": 2.9625,
+      "step": 58160
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7407297492027283,
+      "learning_rate": 3.6772582606242605e-06,
+      "loss": 2.9215,
+      "step": 58161
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8198636770248413,
+      "learning_rate": 3.6748554239840023e-06,
+      "loss": 2.954,
+      "step": 58162
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7506884932518005,
+      "learning_rate": 3.6724533678071376e-06,
+      "loss": 2.9087,
+      "step": 58163
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7431238889694214,
+      "learning_rate": 3.670052092100062e-06,
+      "loss": 2.8626,
+      "step": 58164
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7417303323745728,
+      "learning_rate": 3.6676515968690035e-06,
+      "loss": 3.2512,
+      "step": 58165
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7299365997314453,
+      "learning_rate": 3.6652518821203236e-06,
+      "loss": 2.8178,
+      "step": 58166
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7602360248565674,
+      "learning_rate": 3.6628529478603176e-06,
+      "loss": 2.8408,
+      "step": 58167
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7500886917114258,
+      "learning_rate": 3.6604547940953468e-06,
+      "loss": 2.9354,
+      "step": 58168
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.740761399269104,
+      "learning_rate": 3.6580574208317058e-06,
+      "loss": 3.0434,
+      "step": 58169
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7828353643417358,
+      "learning_rate": 3.6556608280757237e-06,
+      "loss": 2.9932,
+      "step": 58170
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8029820322990417,
+      "learning_rate": 3.6532650158336952e-06,
+      "loss": 2.9728,
+      "step": 58171
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8066024780273438,
+      "learning_rate": 3.650869984111915e-06,
+      "loss": 2.7727,
+      "step": 58172
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7479804158210754,
+      "learning_rate": 3.6484757329167115e-06,
+      "loss": 2.7737,
+      "step": 58173
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7437139749526978,
+      "learning_rate": 3.6460822622544126e-06,
+      "loss": 2.8592,
+      "step": 58174
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7401143312454224,
+      "learning_rate": 3.6436895721312476e-06,
+      "loss": 2.7352,
+      "step": 58175
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7213879823684692,
+      "learning_rate": 3.6412976625536105e-06,
+      "loss": 3.0583,
+      "step": 58176
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7483210563659668,
+      "learning_rate": 3.6389065335277632e-06,
+      "loss": 2.8794,
+      "step": 58177
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7725558280944824,
+      "learning_rate": 3.636516185059968e-06,
+      "loss": 2.8978,
+      "step": 58178
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7521719336509705,
+      "learning_rate": 3.6341266171566185e-06,
+      "loss": 2.8083,
+      "step": 58179
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7107489705085754,
+      "learning_rate": 3.6317378298238774e-06,
+      "loss": 2.9908,
+      "step": 58180
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7555409669876099,
+      "learning_rate": 3.6293498230681397e-06,
+      "loss": 2.6925,
+      "step": 58181
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7039787769317627,
+      "learning_rate": 3.6269625968955997e-06,
+      "loss": 2.9181,
+      "step": 58182
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7815123200416565,
+      "learning_rate": 3.6245761513126525e-06,
+      "loss": 2.9034,
+      "step": 58183
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8096544146537781,
+      "learning_rate": 3.622190486325527e-06,
+      "loss": 2.9037,
+      "step": 58184
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7660767436027527,
+      "learning_rate": 3.6198056019404843e-06,
+      "loss": 2.7394,
+      "step": 58185
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7837504744529724,
+      "learning_rate": 3.617421498163886e-06,
+      "loss": 2.9246,
+      "step": 58186
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7592591643333435,
+      "learning_rate": 3.6150381750019276e-06,
+      "loss": 2.9072,
+      "step": 58187
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7563009858131409,
+      "learning_rate": 3.6126556324609034e-06,
+      "loss": 2.9115,
+      "step": 58188
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7264691591262817,
+      "learning_rate": 3.6102738705471425e-06,
+      "loss": 2.9466,
+      "step": 58189
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.6998595595359802,
+      "learning_rate": 3.6078928892668392e-06,
+      "loss": 3.0228,
+      "step": 58190
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7344692945480347,
+      "learning_rate": 3.6055126886263554e-06,
+      "loss": 2.8275,
+      "step": 58191
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7686775326728821,
+      "learning_rate": 3.6031332686318525e-06,
+      "loss": 2.7886,
+      "step": 58192
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8161377906799316,
+      "learning_rate": 3.600754629289693e-06,
+      "loss": 2.9067,
+      "step": 58193
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7631113529205322,
+      "learning_rate": 3.598376770606104e-06,
+      "loss": 2.8216,
+      "step": 58194
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7431334853172302,
+      "learning_rate": 3.5959996925873145e-06,
+      "loss": 2.821,
+      "step": 58195
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7493770718574524,
+      "learning_rate": 3.593623395239653e-06,
+      "loss": 2.7067,
+      "step": 58196
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7191861271858215,
+      "learning_rate": 3.5912478785693144e-06,
+      "loss": 2.8691,
+      "step": 58197
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8084812760353088,
+      "learning_rate": 3.5888731425825603e-06,
+      "loss": 3.0337,
+      "step": 58198
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7551262974739075,
+      "learning_rate": 3.5864991872857186e-06,
+      "loss": 2.9638,
+      "step": 58199
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7352034449577332,
+      "learning_rate": 3.5841260126849515e-06,
+      "loss": 2.9625,
+      "step": 58200
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7375224232673645,
+      "learning_rate": 3.581753618786587e-06,
+      "loss": 2.8832,
+      "step": 58201
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7833859920501709,
+      "learning_rate": 3.5793820055968535e-06,
+      "loss": 2.7742,
+      "step": 58202
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7902634143829346,
+      "learning_rate": 3.577011173121913e-06,
+      "loss": 3.0064,
+      "step": 58203
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7929283380508423,
+      "learning_rate": 3.574641121368127e-06,
+      "loss": 3.0257,
+      "step": 58204
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7383139729499817,
+      "learning_rate": 3.5722718503416904e-06,
+      "loss": 2.9243,
+      "step": 58205
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7614343762397766,
+      "learning_rate": 3.5699033600488313e-06,
+      "loss": 3.0933,
+      "step": 58206
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7375363707542419,
+      "learning_rate": 3.5675356504957786e-06,
+      "loss": 2.8334,
+      "step": 58207
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7111465930938721,
+      "learning_rate": 3.5651687216887938e-06,
+      "loss": 2.6562,
+      "step": 58208
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7754095196723938,
+      "learning_rate": 3.5628025736341383e-06,
+      "loss": 2.7742,
+      "step": 58209
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7329739332199097,
+      "learning_rate": 3.5604372063379737e-06,
+      "loss": 2.8782,
+      "step": 58210
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7588919401168823,
+      "learning_rate": 3.5580726198065958e-06,
+      "loss": 2.7934,
+      "step": 58211
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7436473369598389,
+      "learning_rate": 3.5557088140461655e-06,
+      "loss": 2.8764,
+      "step": 58212
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7899497151374817,
+      "learning_rate": 3.553345789062978e-06,
+      "loss": 2.7362,
+      "step": 58213
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7157718539237976,
+      "learning_rate": 3.550983544863195e-06,
+      "loss": 3.0378,
+      "step": 58214
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7522872686386108,
+      "learning_rate": 3.548622081453112e-06,
+      "loss": 2.7286,
+      "step": 58215
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7334341406822205,
+      "learning_rate": 3.5462613988388898e-06,
+      "loss": 2.8558,
+      "step": 58216
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7479165196418762,
+      "learning_rate": 3.543901497026724e-06,
+      "loss": 2.782,
+      "step": 58217
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.744875967502594,
+      "learning_rate": 3.54154237602291e-06,
+      "loss": 2.9049,
+      "step": 58218
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7260187864303589,
+      "learning_rate": 3.539184035833609e-06,
+      "loss": 2.8262,
+      "step": 58219
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7581130266189575,
+      "learning_rate": 3.5368264764650157e-06,
+      "loss": 2.9337,
+      "step": 58220
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.745539128780365,
+      "learning_rate": 3.534469697923392e-06,
+      "loss": 3.1823,
+      "step": 58221
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7698248028755188,
+      "learning_rate": 3.5321137002149e-06,
+      "loss": 2.9272,
+      "step": 58222
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7561793923377991,
+      "learning_rate": 3.5297584833458015e-06,
+      "loss": 2.6776,
+      "step": 58223
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7512466311454773,
+      "learning_rate": 3.5274040473222578e-06,
+      "loss": 2.94,
+      "step": 58224
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7217196226119995,
+      "learning_rate": 3.5250503921504305e-06,
+      "loss": 2.9191,
+      "step": 58225
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7432135343551636,
+      "learning_rate": 3.5226975178366144e-06,
+      "loss": 3.0536,
+      "step": 58226
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7715436220169067,
+      "learning_rate": 3.5203454243869056e-06,
+      "loss": 2.866,
+      "step": 58227
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.765325129032135,
+      "learning_rate": 3.517994111807565e-06,
+      "loss": 2.8093,
+      "step": 58228
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7218096256256104,
+      "learning_rate": 3.5156435801047877e-06,
+      "loss": 2.7881,
+      "step": 58229
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7781936526298523,
+      "learning_rate": 3.5132938292847355e-06,
+      "loss": 2.9954,
+      "step": 58230
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.759574294090271,
+      "learning_rate": 3.510944859353604e-06,
+      "loss": 2.7673,
+      "step": 58231
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.723047137260437,
+      "learning_rate": 3.508596670317587e-06,
+      "loss": 2.9904,
+      "step": 58232
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7782467007637024,
+      "learning_rate": 3.506249262182881e-06,
+      "loss": 2.993,
+      "step": 58233
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7288681864738464,
+      "learning_rate": 3.5039026349556133e-06,
+      "loss": 2.8057,
+      "step": 58234
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7624087333679199,
+      "learning_rate": 3.5015567886420125e-06,
+      "loss": 2.8476,
+      "step": 58235
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7774996757507324,
+      "learning_rate": 3.499211723248274e-06,
+      "loss": 2.8622,
+      "step": 58236
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7889575958251953,
+      "learning_rate": 3.4968674387805597e-06,
+      "loss": 2.9525,
+      "step": 58237
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8073905110359192,
+      "learning_rate": 3.4945239352449974e-06,
+      "loss": 2.9227,
+      "step": 58238
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.781495988368988,
+      "learning_rate": 3.4921812126478153e-06,
+      "loss": 3.019,
+      "step": 58239
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7901045680046082,
+      "learning_rate": 3.489839270995143e-06,
+      "loss": 2.8305,
+      "step": 58240
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8189294934272766,
+      "learning_rate": 3.4874981102932074e-06,
+      "loss": 2.8926,
+      "step": 58241
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7915399670600891,
+      "learning_rate": 3.4851577305481047e-06,
+      "loss": 2.8993,
+      "step": 58242
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7558654546737671,
+      "learning_rate": 3.4828181317660296e-06,
+      "loss": 2.8288,
+      "step": 58243
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7701000571250916,
+      "learning_rate": 3.480479313953144e-06,
+      "loss": 2.6183,
+      "step": 58244
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7547457814216614,
+      "learning_rate": 3.4781412771156424e-06,
+      "loss": 2.9023,
+      "step": 58245
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7245228290557861,
+      "learning_rate": 3.4758040212596206e-06,
+      "loss": 2.8595,
+      "step": 58246
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7754863500595093,
+      "learning_rate": 3.473467546391273e-06,
+      "loss": 2.9257,
+      "step": 58247
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7499856948852539,
+      "learning_rate": 3.471131852516762e-06,
+      "loss": 2.9872,
+      "step": 58248
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7747548222541809,
+      "learning_rate": 3.468796939642182e-06,
+      "loss": 2.9552,
+      "step": 58249
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8232293128967285,
+      "learning_rate": 3.466462807773729e-06,
+      "loss": 3.0993,
+      "step": 58250
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7535151839256287,
+      "learning_rate": 3.464129456917564e-06,
+      "loss": 3.1768,
+      "step": 58251
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7690832018852234,
+      "learning_rate": 3.4617968870798154e-06,
+      "loss": 2.5916,
+      "step": 58252
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7519869208335876,
+      "learning_rate": 3.459465098266645e-06,
+      "loss": 2.8841,
+      "step": 58253
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7166727185249329,
+      "learning_rate": 3.457134090484115e-06,
+      "loss": 2.9734,
+      "step": 58254
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.742997944355011,
+      "learning_rate": 3.4548038637384536e-06,
+      "loss": 2.9862,
+      "step": 58255
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7113354802131653,
+      "learning_rate": 3.452474418035789e-06,
+      "loss": 2.8711,
+      "step": 58256
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.785720944404602,
+      "learning_rate": 3.450145753382183e-06,
+      "loss": 2.9458,
+      "step": 58257
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7967292666435242,
+      "learning_rate": 3.4478178697838644e-06,
+      "loss": 2.5159,
+      "step": 58258
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7993360757827759,
+      "learning_rate": 3.445490767246928e-06,
+      "loss": 2.8102,
+      "step": 58259
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7960979342460632,
+      "learning_rate": 3.4431644457774356e-06,
+      "loss": 3.063,
+      "step": 58260
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7577957510948181,
+      "learning_rate": 3.4408389053816486e-06,
+      "loss": 2.733,
+      "step": 58261
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8428319692611694,
+      "learning_rate": 3.438514146065563e-06,
+      "loss": 2.8959,
+      "step": 58262
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.740744411945343,
+      "learning_rate": 3.4361901678353734e-06,
+      "loss": 3.0609,
+      "step": 58263
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7100527882575989,
+      "learning_rate": 3.4338669706971743e-06,
+      "loss": 2.7642,
+      "step": 58264
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8007273077964783,
+      "learning_rate": 3.4315445546570953e-06,
+      "loss": 2.9573,
+      "step": 58265
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7537129521369934,
+      "learning_rate": 3.429222919721297e-06,
+      "loss": 2.8178,
+      "step": 58266
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8160079717636108,
+      "learning_rate": 3.4269020658958093e-06,
+      "loss": 2.7717,
+      "step": 58267
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7378734350204468,
+      "learning_rate": 3.4245819931867924e-06,
+      "loss": 2.7693,
+      "step": 58268
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7598668932914734,
+      "learning_rate": 3.422262701600342e-06,
+      "loss": 3.0711,
+      "step": 58269
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7291812300682068,
+      "learning_rate": 3.4199441911425874e-06,
+      "loss": 2.9735,
+      "step": 58270
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8131197094917297,
+      "learning_rate": 3.4176264618196224e-06,
+      "loss": 2.922,
+      "step": 58271
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7432801127433777,
+      "learning_rate": 3.415309513637543e-06,
+      "loss": 2.7121,
+      "step": 58272
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7480484843254089,
+      "learning_rate": 3.412993346602444e-06,
+      "loss": 3.0394,
+      "step": 58273
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7310391664505005,
+      "learning_rate": 3.410677960720487e-06,
+      "loss": 3.0235,
+      "step": 58274
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7628144025802612,
+      "learning_rate": 3.4083633559976676e-06,
+      "loss": 2.85,
+      "step": 58275
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7469602227210999,
+      "learning_rate": 3.4060495324401803e-06,
+      "loss": 2.8982,
+      "step": 58276
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8014663457870483,
+      "learning_rate": 3.4037364900540544e-06,
+      "loss": 2.9782,
+      "step": 58277
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7652974724769592,
+      "learning_rate": 3.4014242288454176e-06,
+      "loss": 2.901,
+      "step": 58278
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7459307909011841,
+      "learning_rate": 3.3991127488203318e-06,
+      "loss": 2.9058,
+      "step": 58279
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8366508483886719,
+      "learning_rate": 3.3968020499849256e-06,
+      "loss": 2.9857,
+      "step": 58280
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7131357192993164,
+      "learning_rate": 3.3944921323452613e-06,
+      "loss": 2.8848,
+      "step": 58281
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7664152383804321,
+      "learning_rate": 3.3921829959074e-06,
+      "loss": 2.9115,
+      "step": 58282
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7806337475776672,
+      "learning_rate": 3.3898746406774703e-06,
+      "loss": 2.8592,
+      "step": 58283
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7475855350494385,
+      "learning_rate": 3.3875670666615007e-06,
+      "loss": 2.9584,
+      "step": 58284
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.754424512386322,
+      "learning_rate": 3.3852602738655866e-06,
+      "loss": 2.8993,
+      "step": 58285
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7555055022239685,
+      "learning_rate": 3.3829542622958563e-06,
+      "loss": 3.0222,
+      "step": 58286
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7311819791793823,
+      "learning_rate": 3.3806490319583047e-06,
+      "loss": 2.8419,
+      "step": 58287
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7578787803649902,
+      "learning_rate": 3.3783445828590605e-06,
+      "loss": 2.871,
+      "step": 58288
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.9047726392745972,
+      "learning_rate": 3.3760409150041524e-06,
+      "loss": 2.8261,
+      "step": 58289
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7717176079750061,
+      "learning_rate": 3.373738028399675e-06,
+      "loss": 2.8182,
+      "step": 58290
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7483989000320435,
+      "learning_rate": 3.3714359230516906e-06,
+      "loss": 2.8682,
+      "step": 58291
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8349034786224365,
+      "learning_rate": 3.369134598966228e-06,
+      "loss": 2.9128,
+      "step": 58292
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7524193525314331,
+      "learning_rate": 3.3668340561494145e-06,
+      "loss": 2.8352,
+      "step": 58293
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7450447082519531,
+      "learning_rate": 3.364534294607246e-06,
+      "loss": 3.0256,
+      "step": 58294
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7853724360466003,
+      "learning_rate": 3.362235314345818e-06,
+      "loss": 2.8471,
+      "step": 58295
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8045984506607056,
+      "learning_rate": 3.3599371153711917e-06,
+      "loss": 2.9097,
+      "step": 58296
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7121866941452026,
+      "learning_rate": 3.3576396976893626e-06,
+      "loss": 2.7478,
+      "step": 58297
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7629473805427551,
+      "learning_rate": 3.355343061306459e-06,
+      "loss": 2.8422,
+      "step": 58298
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7671418786048889,
+      "learning_rate": 3.353047206228443e-06,
+      "loss": 3.1245,
+      "step": 58299
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7228366136550903,
+      "learning_rate": 3.350752132461443e-06,
+      "loss": 3.0361,
+      "step": 58300
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7080792188644409,
+      "learning_rate": 3.3484578400114537e-06,
+      "loss": 3.03,
+      "step": 58301
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7466286420822144,
+      "learning_rate": 3.3461643288845375e-06,
+      "loss": 2.765,
+      "step": 58302
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7471815943717957,
+      "learning_rate": 3.343871599086756e-06,
+      "loss": 2.9213,
+      "step": 58303
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7371976375579834,
+      "learning_rate": 3.3415796506241045e-06,
+      "loss": 2.9685,
+      "step": 58304
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6997452974319458,
+      "learning_rate": 3.3392884835026444e-06,
+      "loss": 2.9548,
+      "step": 58305
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.801490068435669,
+      "learning_rate": 3.336998097728405e-06,
+      "loss": 3.1318,
+      "step": 58306
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7446818947792053,
+      "learning_rate": 3.3347084933074143e-06,
+      "loss": 2.9129,
+      "step": 58307
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7398616075515747,
+      "learning_rate": 3.3324196702457006e-06,
+      "loss": 2.9168,
+      "step": 58308
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7449392676353455,
+      "learning_rate": 3.3301316285493263e-06,
+      "loss": 3.0701,
+      "step": 58309
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7428542375564575,
+      "learning_rate": 3.327844368224253e-06,
+      "loss": 2.6963,
+      "step": 58310
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7714420557022095,
+      "learning_rate": 3.3255578892765753e-06,
+      "loss": 2.8949,
+      "step": 58311
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7456600069999695,
+      "learning_rate": 3.323272191712256e-06,
+      "loss": 2.841,
+      "step": 58312
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7867556810379028,
+      "learning_rate": 3.3209872755373566e-06,
+      "loss": 2.8416,
+      "step": 58313
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7435100674629211,
+      "learning_rate": 3.318703140757872e-06,
+      "loss": 2.958,
+      "step": 58314
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.752841591835022,
+      "learning_rate": 3.3164197873798316e-06,
+      "loss": 3.1411,
+      "step": 58315
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7200860381126404,
+      "learning_rate": 3.3141372154092293e-06,
+      "loss": 2.8908,
+      "step": 58316
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6982949376106262,
+      "learning_rate": 3.311855424852128e-06,
+      "loss": 2.5464,
+      "step": 58317
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7470500469207764,
+      "learning_rate": 3.309574415714489e-06,
+      "loss": 2.7845,
+      "step": 58318
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7575530409812927,
+      "learning_rate": 3.3072941880023075e-06,
+      "loss": 2.935,
+      "step": 58319
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7366733551025391,
+      "learning_rate": 3.305014741721612e-06,
+      "loss": 3.0839,
+      "step": 58320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7596519589424133,
+      "learning_rate": 3.3027360768784318e-06,
+      "loss": 2.7866,
+      "step": 58321
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7350767254829407,
+      "learning_rate": 3.3004581934787276e-06,
+      "loss": 2.8478,
+      "step": 58322
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.730754554271698,
+      "learning_rate": 3.298181091528529e-06,
+      "loss": 2.8066,
+      "step": 58323
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7694946527481079,
+      "learning_rate": 3.295904771033797e-06,
+      "loss": 2.6976,
+      "step": 58324
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7357961535453796,
+      "learning_rate": 3.293629232000561e-06,
+      "loss": 2.8541,
+      "step": 58325
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7392043471336365,
+      "learning_rate": 3.2913544744348152e-06,
+      "loss": 2.7927,
+      "step": 58326
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7261336445808411,
+      "learning_rate": 3.289080498342522e-06,
+      "loss": 2.9183,
+      "step": 58327
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7417194843292236,
+      "learning_rate": 3.2868073037297104e-06,
+      "loss": 2.9671,
+      "step": 58328
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7913023829460144,
+      "learning_rate": 3.2845348906023084e-06,
+      "loss": 3.0921,
+      "step": 58329
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7364912033081055,
+      "learning_rate": 3.2822632589663776e-06,
+      "loss": 2.9238,
+      "step": 58330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7688376903533936,
+      "learning_rate": 3.279992408827847e-06,
+      "loss": 2.9304,
+      "step": 58331
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7457802295684814,
+      "learning_rate": 3.277722340192712e-06,
+      "loss": 3.1047,
+      "step": 58332
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.74704509973526,
+      "learning_rate": 3.2754530530669343e-06,
+      "loss": 2.9558,
+      "step": 58333
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7239862680435181,
+      "learning_rate": 3.2731845474565086e-06,
+      "loss": 2.8178,
+      "step": 58334
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7056124806404114,
+      "learning_rate": 3.270916823367431e-06,
+      "loss": 3.0095,
+      "step": 58335
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7435981035232544,
+      "learning_rate": 3.2686498808056296e-06,
+      "loss": 2.7761,
+      "step": 58336
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7090030908584595,
+      "learning_rate": 3.266383719777099e-06,
+      "loss": 2.9752,
+      "step": 58337
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8070139288902283,
+      "learning_rate": 3.2641183402878356e-06,
+      "loss": 2.9494,
+      "step": 58338
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7761786580085754,
+      "learning_rate": 3.2618537423437672e-06,
+      "loss": 2.7368,
+      "step": 58339
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7421362400054932,
+      "learning_rate": 3.259589925950856e-06,
+      "loss": 2.8487,
+      "step": 58340
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8992583155632019,
+      "learning_rate": 3.257326891115064e-06,
+      "loss": 2.9035,
+      "step": 58341
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7621567249298096,
+      "learning_rate": 3.255064637842386e-06,
+      "loss": 2.8351,
+      "step": 58342
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7418533563613892,
+      "learning_rate": 3.252803166138751e-06,
+      "loss": 3.0166,
+      "step": 58343
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7442709803581238,
+      "learning_rate": 3.2505424760101207e-06,
+      "loss": 2.6756,
+      "step": 58344
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7631102800369263,
+      "learning_rate": 3.248282567462457e-06,
+      "loss": 3.0295,
+      "step": 58345
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8174268007278442,
+      "learning_rate": 3.246023440501688e-06,
+      "loss": 2.9048,
+      "step": 58346
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.736268162727356,
+      "learning_rate": 3.243765095133777e-06,
+      "loss": 2.951,
+      "step": 58347
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7341980934143066,
+      "learning_rate": 3.2415075313647177e-06,
+      "loss": 2.8567,
+      "step": 58348
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7812278866767883,
+      "learning_rate": 3.239250749200373e-06,
+      "loss": 2.882,
+      "step": 58349
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7396385073661804,
+      "learning_rate": 3.2369947486467707e-06,
+      "loss": 3.0547,
+      "step": 58350
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7556043267250061,
+      "learning_rate": 3.2347395297097732e-06,
+      "loss": 2.7976,
+      "step": 58351
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7392325401306152,
+      "learning_rate": 3.2324850923953426e-06,
+      "loss": 2.9214,
+      "step": 58352
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7273942232131958,
+      "learning_rate": 3.230231436709474e-06,
+      "loss": 2.8526,
+      "step": 58353
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7350383996963501,
+      "learning_rate": 3.2279785626580622e-06,
+      "loss": 3.0958,
+      "step": 58354
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7365357875823975,
+      "learning_rate": 3.2257264702470363e-06,
+      "loss": 2.8979,
+      "step": 58355
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.742205023765564,
+      "learning_rate": 3.223475159482325e-06,
+      "loss": 2.8678,
+      "step": 58356
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7722266912460327,
+      "learning_rate": 3.2212246303698565e-06,
+      "loss": 2.7893,
+      "step": 58357
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7623163461685181,
+      "learning_rate": 3.2189748829155593e-06,
+      "loss": 2.8126,
+      "step": 58358
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.794621467590332,
+      "learning_rate": 3.2167259171253625e-06,
+      "loss": 2.8671,
+      "step": 58359
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.737347424030304,
+      "learning_rate": 3.2144777330052274e-06,
+      "loss": 2.8197,
+      "step": 58360
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7127636075019836,
+      "learning_rate": 3.2122303305610162e-06,
+      "loss": 2.803,
+      "step": 58361
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7416470646858215,
+      "learning_rate": 3.2099837097986914e-06,
+      "loss": 2.7396,
+      "step": 58362
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7706217765808105,
+      "learning_rate": 3.2077378707241142e-06,
+      "loss": 2.7705,
+      "step": 58363
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7568953037261963,
+      "learning_rate": 3.2054928133432467e-06,
+      "loss": 2.9229,
+      "step": 58364
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7913509607315063,
+      "learning_rate": 3.2032485376620175e-06,
+      "loss": 2.9663,
+      "step": 58365
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.777898371219635,
+      "learning_rate": 3.2010050436862887e-06,
+      "loss": 2.7852,
+      "step": 58366
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7000056505203247,
+      "learning_rate": 3.198762331421989e-06,
+      "loss": 2.7677,
+      "step": 58367
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7845345139503479,
+      "learning_rate": 3.19652040087508e-06,
+      "loss": 2.7989,
+      "step": 58368
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7394914627075195,
+      "learning_rate": 3.194279252051357e-06,
+      "loss": 2.9253,
+      "step": 58369
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.741482675075531,
+      "learning_rate": 3.1920388849568157e-06,
+      "loss": 3.07,
+      "step": 58370
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7681283950805664,
+      "learning_rate": 3.1897992995973175e-06,
+      "loss": 2.9483,
+      "step": 58371
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7793736457824707,
+      "learning_rate": 3.187560495978758e-06,
+      "loss": 2.5742,
+      "step": 58372
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7607495784759521,
+      "learning_rate": 3.1853224741070326e-06,
+      "loss": 3.0333,
+      "step": 58373
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7457079887390137,
+      "learning_rate": 3.1830852339880366e-06,
+      "loss": 2.9459,
+      "step": 58374
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7789652347564697,
+      "learning_rate": 3.180848775627698e-06,
+      "loss": 2.9568,
+      "step": 58375
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7540808916091919,
+      "learning_rate": 3.178613099031879e-06,
+      "loss": 2.8091,
+      "step": 58376
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7654391527175903,
+      "learning_rate": 3.176378204206442e-06,
+      "loss": 2.8447,
+      "step": 58377
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7442813515663147,
+      "learning_rate": 3.174144091157316e-06,
+      "loss": 2.8316,
+      "step": 58378
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7185924053192139,
+      "learning_rate": 3.1719107598903615e-06,
+      "loss": 2.8791,
+      "step": 58379
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7041293978691101,
+      "learning_rate": 3.169678210411475e-06,
+      "loss": 2.8301,
+      "step": 58380
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8128548860549927,
+      "learning_rate": 3.167446442726518e-06,
+      "loss": 2.9627,
+      "step": 58381
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7193200588226318,
+      "learning_rate": 3.165215456841419e-06,
+      "loss": 2.8726,
+      "step": 58382
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7626433372497559,
+      "learning_rate": 3.162985252761974e-06,
+      "loss": 2.8444,
+      "step": 58383
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7526370286941528,
+      "learning_rate": 3.1607558304941107e-06,
+      "loss": 2.6661,
+      "step": 58384
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7516125440597534,
+      "learning_rate": 3.1585271900436913e-06,
+      "loss": 2.9895,
+      "step": 58385
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7663487792015076,
+      "learning_rate": 3.1562993314165784e-06,
+      "loss": 2.832,
+      "step": 58386
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7314614653587341,
+      "learning_rate": 3.154072254618667e-06,
+      "loss": 3.0411,
+      "step": 58387
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7240493893623352,
+      "learning_rate": 3.151845959655752e-06,
+      "loss": 2.9518,
+      "step": 58388
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7454181909561157,
+      "learning_rate": 3.1496204465337626e-06,
+      "loss": 2.7987,
+      "step": 58389
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6941863894462585,
+      "learning_rate": 3.1473957152585936e-06,
+      "loss": 3.1225,
+      "step": 58390
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7873305678367615,
+      "learning_rate": 3.1451717658360075e-06,
+      "loss": 2.7987,
+      "step": 58391
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7635713219642639,
+      "learning_rate": 3.1429485982719326e-06,
+      "loss": 2.6695,
+      "step": 58392
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.74675452709198,
+      "learning_rate": 3.1407262125721645e-06,
+      "loss": 2.7484,
+      "step": 58393
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7379273176193237,
+      "learning_rate": 3.138504608742598e-06,
+      "loss": 2.9954,
+      "step": 58394
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7270514965057373,
+      "learning_rate": 3.1362837867890957e-06,
+      "loss": 3.2038,
+      "step": 58395
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7507408261299133,
+      "learning_rate": 3.134063746717486e-06,
+      "loss": 2.8354,
+      "step": 58396
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7894320487976074,
+      "learning_rate": 3.13184448853363e-06,
+      "loss": 2.9427,
+      "step": 58397
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7546813488006592,
+      "learning_rate": 3.129626012243358e-06,
+      "loss": 2.8778,
+      "step": 58398
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7203192114830017,
+      "learning_rate": 3.127408317852498e-06,
+      "loss": 2.9069,
+      "step": 58399
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7543355226516724,
+      "learning_rate": 3.125191405366911e-06,
+      "loss": 2.9173,
+      "step": 58400
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7346734404563904,
+      "learning_rate": 3.1229752747924606e-06,
+      "loss": 2.8988,
+      "step": 58401
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8272573351860046,
+      "learning_rate": 3.1207599261349416e-06,
+      "loss": 2.8884,
+      "step": 58402
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7785396575927734,
+      "learning_rate": 3.1185453594001818e-06,
+      "loss": 2.7603,
+      "step": 58403
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7275567650794983,
+      "learning_rate": 3.1163315745940777e-06,
+      "loss": 2.8534,
+      "step": 58404
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7752771377563477,
+      "learning_rate": 3.114118571722424e-06,
+      "loss": 2.741,
+      "step": 58405
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7439063191413879,
+      "learning_rate": 3.111906350791016e-06,
+      "loss": 3.0018,
+      "step": 58406
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7584403157234192,
+      "learning_rate": 3.109694911805716e-06,
+      "loss": 3.0572,
+      "step": 58407
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7162322998046875,
+      "learning_rate": 3.1074842547723523e-06,
+      "loss": 2.9417,
+      "step": 58408
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7319763898849487,
+      "learning_rate": 3.1052743796967207e-06,
+      "loss": 3.1582,
+      "step": 58409
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.825232744216919,
+      "learning_rate": 3.1030652865846828e-06,
+      "loss": 2.8801,
+      "step": 58410
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7098069787025452,
+      "learning_rate": 3.100856975442001e-06,
+      "loss": 2.9849,
+      "step": 58411
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.754648745059967,
+      "learning_rate": 3.0986494462745372e-06,
+      "loss": 3.0793,
+      "step": 58412
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7526900768280029,
+      "learning_rate": 3.0964426990880865e-06,
+      "loss": 2.9493,
+      "step": 58413
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7582096457481384,
+      "learning_rate": 3.0942367338884775e-06,
+      "loss": 2.8612,
+      "step": 58414
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7545999884605408,
+      "learning_rate": 3.0920315506815063e-06,
+      "loss": 2.8995,
+      "step": 58415
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7414794564247131,
+      "learning_rate": 3.0898271494729677e-06,
+      "loss": 2.8369,
+      "step": 58416
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8262287378311157,
+      "learning_rate": 3.08762353026869e-06,
+      "loss": 2.9537,
+      "step": 58417
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7526206374168396,
+      "learning_rate": 3.0854206930744696e-06,
+      "loss": 2.8687,
+      "step": 58418
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7402799725532532,
+      "learning_rate": 3.0832186378961343e-06,
+      "loss": 2.8745,
+      "step": 58419
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.784649133682251,
+      "learning_rate": 3.0810173647394463e-06,
+      "loss": 2.6984,
+      "step": 58420
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7722074389457703,
+      "learning_rate": 3.0788168736102014e-06,
+      "loss": 2.8043,
+      "step": 58421
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8019753694534302,
+      "learning_rate": 3.0766171645141946e-06,
+      "loss": 2.8064,
+      "step": 58422
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7800315618515015,
+      "learning_rate": 3.0744182374572548e-06,
+      "loss": 2.8841,
+      "step": 58423
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7406907081604004,
+      "learning_rate": 3.0722200924451434e-06,
+      "loss": 2.9577,
+      "step": 58424
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7482556104660034,
+      "learning_rate": 3.0700227294836897e-06,
+      "loss": 2.8487,
+      "step": 58425
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7338652610778809,
+      "learning_rate": 3.0678261485785892e-06,
+      "loss": 3.0761,
+      "step": 58426
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7629140019416809,
+      "learning_rate": 3.0656303497357373e-06,
+      "loss": 2.8937,
+      "step": 58427
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7479938864707947,
+      "learning_rate": 3.0634353329608954e-06,
+      "loss": 2.6075,
+      "step": 58428
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.793538510799408,
+      "learning_rate": 3.061241098259759e-06,
+      "loss": 3.0803,
+      "step": 58429
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7822150588035583,
+      "learning_rate": 3.059047645638191e-06,
+      "loss": 2.6345,
+      "step": 58430
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7450809478759766,
+      "learning_rate": 3.056854975101952e-06,
+      "loss": 2.9367,
+      "step": 58431
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7621826529502869,
+      "learning_rate": 3.0546630866568057e-06,
+      "loss": 3.0372,
+      "step": 58432
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7780464887619019,
+      "learning_rate": 3.052471980308513e-06,
+      "loss": 2.9461,
+      "step": 58433
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7118364572525024,
+      "learning_rate": 3.0502816560628695e-06,
+      "loss": 2.8477,
+      "step": 58434
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7661958336830139,
+      "learning_rate": 3.0480921139256378e-06,
+      "loss": 2.9218,
+      "step": 58435
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7564506530761719,
+      "learning_rate": 3.045903353902579e-06,
+      "loss": 2.9434,
+      "step": 58436
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7816385626792908,
+      "learning_rate": 3.043715375999456e-06,
+      "loss": 2.8841,
+      "step": 58437
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7606683373451233,
+      "learning_rate": 3.0415281802220304e-06,
+      "loss": 2.7296,
+      "step": 58438
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7658647298812866,
+      "learning_rate": 3.0393417665760976e-06,
+      "loss": 3.0555,
+      "step": 58439
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7768723964691162,
+      "learning_rate": 3.0371561350673537e-06,
+      "loss": 2.8682,
+      "step": 58440
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7276012301445007,
+      "learning_rate": 3.0349712857016263e-06,
+      "loss": 2.7861,
+      "step": 58441
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7483054995536804,
+      "learning_rate": 3.0327872184846113e-06,
+      "loss": 2.8149,
+      "step": 58442
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7850237488746643,
+      "learning_rate": 3.030603933422071e-06,
+      "loss": 2.8655,
+      "step": 58443
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.764310359954834,
+      "learning_rate": 3.0284214305197675e-06,
+      "loss": 3.0662,
+      "step": 58444
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7962012887001038,
+      "learning_rate": 3.0262397097834956e-06,
+      "loss": 2.8438,
+      "step": 58445
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.789114773273468,
+      "learning_rate": 3.0240587712188848e-06,
+      "loss": 3.0808,
+      "step": 58446
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7645512819290161,
+      "learning_rate": 3.0218786148317965e-06,
+      "loss": 2.8729,
+      "step": 58447
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7275797724723816,
+      "learning_rate": 3.019699240627893e-06,
+      "loss": 3.0125,
+      "step": 58448
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7716579437255859,
+      "learning_rate": 3.01752064861297e-06,
+      "loss": 2.9713,
+      "step": 58449
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.781461775302887,
+      "learning_rate": 3.015342838792756e-06,
+      "loss": 2.8235,
+      "step": 58450
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7181127667427063,
+      "learning_rate": 3.0131658111729463e-06,
+      "loss": 2.9488,
+      "step": 58451
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.8236247897148132,
+      "learning_rate": 3.0109895657593365e-06,
+      "loss": 2.9258,
+      "step": 58452
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7421223521232605,
+      "learning_rate": 3.0088141025575887e-06,
+      "loss": 2.8039,
+      "step": 58453
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7448740601539612,
+      "learning_rate": 3.0066394215734645e-06,
+      "loss": 2.7816,
+      "step": 58454
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.751390814781189,
+      "learning_rate": 3.0044655228126937e-06,
+      "loss": 2.8791,
+      "step": 58455
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7451471090316772,
+      "learning_rate": 3.0022924062810373e-06,
+      "loss": 2.8559,
+      "step": 58456
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7556320428848267,
+      "learning_rate": 3.0001200719841578e-06,
+      "loss": 2.661,
+      "step": 58457
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7909460067749023,
+      "learning_rate": 2.9979485199278175e-06,
+      "loss": 3.1855,
+      "step": 58458
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7238598465919495,
+      "learning_rate": 2.9957777501177115e-06,
+      "loss": 2.8134,
+      "step": 58459
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7427256107330322,
+      "learning_rate": 2.9936077625595356e-06,
+      "loss": 3.0505,
+      "step": 58460
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7333883047103882,
+      "learning_rate": 2.9914385572590848e-06,
+      "loss": 2.9026,
+      "step": 58461
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8171818256378174,
+      "learning_rate": 2.9892701342220215e-06,
+      "loss": 2.8903,
+      "step": 58462
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7717416286468506,
+      "learning_rate": 2.987102493454041e-06,
+      "loss": 2.8525,
+      "step": 58463
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7565333247184753,
+      "learning_rate": 2.9849356349608724e-06,
+      "loss": 2.8967,
+      "step": 58464
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7647886872291565,
+      "learning_rate": 2.9827695587482434e-06,
+      "loss": 2.9653,
+      "step": 58465
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7452661395072937,
+      "learning_rate": 2.9806042648217843e-06,
+      "loss": 2.8538,
+      "step": 58466
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7374056577682495,
+      "learning_rate": 2.9784397531872894e-06,
+      "loss": 2.9242,
+      "step": 58467
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7226657271385193,
+      "learning_rate": 2.9762760238504212e-06,
+      "loss": 2.7999,
+      "step": 58468
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7068235874176025,
+      "learning_rate": 2.9741130768168752e-06,
+      "loss": 2.7401,
+      "step": 58469
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7041826248168945,
+      "learning_rate": 2.971950912092347e-06,
+      "loss": 2.8399,
+      "step": 58470
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7928381562232971,
+      "learning_rate": 2.9697895296825645e-06,
+      "loss": 2.8014,
+      "step": 58471
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7238459587097168,
+      "learning_rate": 2.9676289295931577e-06,
+      "loss": 2.9021,
+      "step": 58472
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7136267423629761,
+      "learning_rate": 2.9654691118298544e-06,
+      "loss": 2.937,
+      "step": 58473
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7531981468200684,
+      "learning_rate": 2.963310076398351e-06,
+      "loss": 2.7484,
+      "step": 58474
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7254064083099365,
+      "learning_rate": 2.9611518233043084e-06,
+      "loss": 2.8743,
+      "step": 58475
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.74024498462677,
+      "learning_rate": 2.9589943525534566e-06,
+      "loss": 2.8673,
+      "step": 58476
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7722803354263306,
+      "learning_rate": 2.9568376641514237e-06,
+      "loss": 3.0154,
+      "step": 58477
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7294988036155701,
+      "learning_rate": 2.9546817581039385e-06,
+      "loss": 2.8068,
+      "step": 58478
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7921958565711975,
+      "learning_rate": 2.95252663441663e-06,
+      "loss": 3.0001,
+      "step": 58479
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7250039577484131,
+      "learning_rate": 2.9503722930951933e-06,
+      "loss": 2.9304,
+      "step": 58480
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8294291496276855,
+      "learning_rate": 2.948218734145291e-06,
+      "loss": 2.7402,
+      "step": 58481
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7494039535522461,
+      "learning_rate": 2.946065957572652e-06,
+      "loss": 2.7678,
+      "step": 58482
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8283053040504456,
+      "learning_rate": 2.9439139633828714e-06,
+      "loss": 2.863,
+      "step": 58483
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.772268533706665,
+      "learning_rate": 2.941762751581678e-06,
+      "loss": 2.7249,
+      "step": 58484
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7302533388137817,
+      "learning_rate": 2.939612322174734e-06,
+      "loss": 2.8957,
+      "step": 58485
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.787757933139801,
+      "learning_rate": 2.9374626751676347e-06,
+      "loss": 2.9369,
+      "step": 58486
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7752677798271179,
+      "learning_rate": 2.935313810566109e-06,
+      "loss": 2.8947,
+      "step": 58487
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7386978268623352,
+      "learning_rate": 2.933165728375786e-06,
+      "loss": 2.7775,
+      "step": 58488
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7692047953605652,
+      "learning_rate": 2.9310184286023276e-06,
+      "loss": 3.0412,
+      "step": 58489
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7471635937690735,
+      "learning_rate": 2.928871911251396e-06,
+      "loss": 2.7332,
+      "step": 58490
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7705933451652527,
+      "learning_rate": 2.926726176328653e-06,
+      "loss": 3.0526,
+      "step": 58491
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7803830504417419,
+      "learning_rate": 2.9245812238397282e-06,
+      "loss": 2.9832,
+      "step": 58492
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7428114414215088,
+      "learning_rate": 2.9224370537903163e-06,
+      "loss": 2.7619,
+      "step": 58493
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7303468585014343,
+      "learning_rate": 2.9202936661860134e-06,
+      "loss": 2.7892,
+      "step": 58494
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7733170390129089,
+      "learning_rate": 2.918151061032481e-06,
+      "loss": 2.8862,
+      "step": 58495
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7337782979011536,
+      "learning_rate": 2.9160092383353483e-06,
+      "loss": 2.9276,
+      "step": 58496
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7292686700820923,
+      "learning_rate": 2.9138681981002775e-06,
+      "loss": 3.0101,
+      "step": 58497
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7879104018211365,
+      "learning_rate": 2.9117279403329307e-06,
+      "loss": 2.9083,
+      "step": 58498
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7869991660118103,
+      "learning_rate": 2.9095884650389034e-06,
+      "loss": 2.912,
+      "step": 58499
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7370448112487793,
+      "learning_rate": 2.9074497722238242e-06,
+      "loss": 3.0133,
+      "step": 58500
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7162608504295349,
+      "learning_rate": 2.9053118618933556e-06,
+      "loss": 3.022,
+      "step": 58501
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7638165950775146,
+      "learning_rate": 2.9031747340531264e-06,
+      "loss": 2.965,
+      "step": 58502
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.731220543384552,
+      "learning_rate": 2.9010383887087317e-06,
+      "loss": 3.0144,
+      "step": 58503
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.778514564037323,
+      "learning_rate": 2.8989028258658674e-06,
+      "loss": 2.801,
+      "step": 58504
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8099600076675415,
+      "learning_rate": 2.8967680455300956e-06,
+      "loss": 3.0025,
+      "step": 58505
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7909402847290039,
+      "learning_rate": 2.8946340477070452e-06,
+      "loss": 2.8844,
+      "step": 58506
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7233417630195618,
+      "learning_rate": 2.892500832402378e-06,
+      "loss": 2.7969,
+      "step": 58507
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7570949792861938,
+      "learning_rate": 2.8903683996216562e-06,
+      "loss": 2.9808,
+      "step": 58508
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7265278100967407,
+      "learning_rate": 2.8882367493705425e-06,
+      "loss": 2.9376,
+      "step": 58509
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7586358189582825,
+      "learning_rate": 2.886105881654599e-06,
+      "loss": 2.8498,
+      "step": 58510
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7601178884506226,
+      "learning_rate": 2.8839757964794873e-06,
+      "loss": 2.9384,
+      "step": 58511
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7448052167892456,
+      "learning_rate": 2.8818464938508366e-06,
+      "loss": 2.997,
+      "step": 58512
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7501569986343384,
+      "learning_rate": 2.879717973774176e-06,
+      "loss": 2.7712,
+      "step": 58513
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7544119954109192,
+      "learning_rate": 2.8775902362552004e-06,
+      "loss": 2.9557,
+      "step": 58514
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7905291318893433,
+      "learning_rate": 2.875463281299439e-06,
+      "loss": 3.0345,
+      "step": 58515
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7742028832435608,
+      "learning_rate": 2.8733371089125213e-06,
+      "loss": 2.854,
+      "step": 58516
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.748437762260437,
+      "learning_rate": 2.871211719100075e-06,
+      "loss": 2.7507,
+      "step": 58517
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7571814060211182,
+      "learning_rate": 2.8690871118676295e-06,
+      "loss": 2.6589,
+      "step": 58518
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7534952759742737,
+      "learning_rate": 2.8669632872208805e-06,
+      "loss": 2.8444,
+      "step": 58519
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.710351824760437,
+      "learning_rate": 2.864840245165323e-06,
+      "loss": 2.7474,
+      "step": 58520
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7488248944282532,
+      "learning_rate": 2.8627179857066195e-06,
+      "loss": 2.7427,
+      "step": 58521
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7320175766944885,
+      "learning_rate": 2.8605965088503324e-06,
+      "loss": 2.8392,
+      "step": 58522
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.738500714302063,
+      "learning_rate": 2.8584758146020236e-06,
+      "loss": 2.6562,
+      "step": 58523
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7765610218048096,
+      "learning_rate": 2.856355902967322e-06,
+      "loss": 3.1511,
+      "step": 58524
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7138228416442871,
+      "learning_rate": 2.85423677395179e-06,
+      "loss": 2.9996,
+      "step": 58525
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7560122013092041,
+      "learning_rate": 2.8521184275609897e-06,
+      "loss": 2.9072,
+      "step": 58526
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7315943241119385,
+      "learning_rate": 2.85000086380055e-06,
+      "loss": 2.8984,
+      "step": 58527
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7746604084968567,
+      "learning_rate": 2.8478840826759996e-06,
+      "loss": 2.7893,
+      "step": 58528
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7341941595077515,
+      "learning_rate": 2.8457680841929344e-06,
+      "loss": 2.7045,
+      "step": 58529
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7307849526405334,
+      "learning_rate": 2.843652868356949e-06,
+      "loss": 2.7645,
+      "step": 58530
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7367571592330933,
+      "learning_rate": 2.841538435173607e-06,
+      "loss": 2.8463,
+      "step": 58531
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7715376615524292,
+      "learning_rate": 2.8394247846484363e-06,
+      "loss": 2.8366,
+      "step": 58532
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7914366722106934,
+      "learning_rate": 2.8373119167869994e-06,
+      "loss": 2.777,
+      "step": 58533
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7575352191925049,
+      "learning_rate": 2.835199831594959e-06,
+      "loss": 3.0308,
+      "step": 58534
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7674602270126343,
+      "learning_rate": 2.8330885290777426e-06,
+      "loss": 2.7934,
+      "step": 58535
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7100291848182678,
+      "learning_rate": 2.830978009241047e-06,
+      "loss": 2.8665,
+      "step": 58536
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7356333136558533,
+      "learning_rate": 2.828868272090334e-06,
+      "loss": 2.9221,
+      "step": 58537
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7471137642860413,
+      "learning_rate": 2.8267593176311665e-06,
+      "loss": 3.0157,
+      "step": 58538
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7663685083389282,
+      "learning_rate": 2.8246511458691722e-06,
+      "loss": 2.8722,
+      "step": 58539
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7581278085708618,
+      "learning_rate": 2.8225437568098143e-06,
+      "loss": 2.776,
+      "step": 58540
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7169347405433655,
+      "learning_rate": 2.820437150458721e-06,
+      "loss": 2.5599,
+      "step": 58541
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7184317111968994,
+      "learning_rate": 2.818331326821355e-06,
+      "loss": 2.8356,
+      "step": 58542
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7458834648132324,
+      "learning_rate": 2.8162262859033447e-06,
+      "loss": 2.8992,
+      "step": 58543
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7352158427238464,
+      "learning_rate": 2.8141220277102196e-06,
+      "loss": 2.8397,
+      "step": 58544
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.716495156288147,
+      "learning_rate": 2.8120185522474415e-06,
+      "loss": 2.9937,
+      "step": 58545
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7433305978775024,
+      "learning_rate": 2.809915859520673e-06,
+      "loss": 3.0178,
+      "step": 58546
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7520804405212402,
+      "learning_rate": 2.8078139495353423e-06,
+      "loss": 2.777,
+      "step": 58547
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.728420078754425,
+      "learning_rate": 2.8057128222970458e-06,
+      "loss": 3.1996,
+      "step": 58548
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.816613495349884,
+      "learning_rate": 2.8036124778113123e-06,
+      "loss": 3.1029,
+      "step": 58549
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7988955974578857,
+      "learning_rate": 2.8015129160836704e-06,
+      "loss": 2.818,
+      "step": 58550
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7273043394088745,
+      "learning_rate": 2.7994141371196487e-06,
+      "loss": 2.9679,
+      "step": 58551
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7395277619361877,
+      "learning_rate": 2.797316140924777e-06,
+      "loss": 2.86,
+      "step": 58552
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7469648122787476,
+      "learning_rate": 2.79521892750455e-06,
+      "loss": 2.8988,
+      "step": 58553
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.789919376373291,
+      "learning_rate": 2.7931224968645304e-06,
+      "loss": 2.7315,
+      "step": 58554
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7301388382911682,
+      "learning_rate": 2.7910268490102473e-06,
+      "loss": 3.0406,
+      "step": 58555
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7471961975097656,
+      "learning_rate": 2.7889319839471956e-06,
+      "loss": 2.7793,
+      "step": 58556
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8272915482521057,
+      "learning_rate": 2.7868379016808717e-06,
+      "loss": 2.7463,
+      "step": 58557
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7283830046653748,
+      "learning_rate": 2.7847446022168375e-06,
+      "loss": 2.9614,
+      "step": 58558
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.697257399559021,
+      "learning_rate": 2.782652085560588e-06,
+      "loss": 2.6746,
+      "step": 58559
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7829978466033936,
+      "learning_rate": 2.7805603517176203e-06,
+      "loss": 3.048,
+      "step": 58560
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7517426609992981,
+      "learning_rate": 2.7784694006934618e-06,
+      "loss": 2.9577,
+      "step": 58561
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8072444796562195,
+      "learning_rate": 2.7763792324936083e-06,
+      "loss": 2.8453,
+      "step": 58562
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7768639326095581,
+      "learning_rate": 2.7742898471235562e-06,
+      "loss": 2.779,
+      "step": 58563
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7537323832511902,
+      "learning_rate": 2.7722012445888676e-06,
+      "loss": 2.6229,
+      "step": 58564
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7503150701522827,
+      "learning_rate": 2.7701134248949707e-06,
+      "loss": 2.7897,
+      "step": 58565
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7244771122932434,
+      "learning_rate": 2.7680263880473953e-06,
+      "loss": 2.7409,
+      "step": 58566
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7575398087501526,
+      "learning_rate": 2.765940134051636e-06,
+      "loss": 3.0002,
+      "step": 58567
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8123334646224976,
+      "learning_rate": 2.7638546629131566e-06,
+      "loss": 3.0256,
+      "step": 58568
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8360390663146973,
+      "learning_rate": 2.7617699746375178e-06,
+      "loss": 3.0575,
+      "step": 58569
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7276893854141235,
+      "learning_rate": 2.7596860692301492e-06,
+      "loss": 2.9831,
+      "step": 58570
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7142115831375122,
+      "learning_rate": 2.75760294669658e-06,
+      "loss": 3.0064,
+      "step": 58571
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7129017114639282,
+      "learning_rate": 2.755520607042272e-06,
+      "loss": 2.8834,
+      "step": 58572
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7552473545074463,
+      "learning_rate": 2.753439050272721e-06,
+      "loss": 2.7786,
+      "step": 58573
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7475464940071106,
+      "learning_rate": 2.751358276393423e-06,
+      "loss": 2.9663,
+      "step": 58574
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7509301900863647,
+      "learning_rate": 2.749278285409806e-06,
+      "loss": 2.6761,
+      "step": 58575
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7471511960029602,
+      "learning_rate": 2.7471990773274e-06,
+      "loss": 2.9407,
+      "step": 58576
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7476446628570557,
+      "learning_rate": 2.7451206521516334e-06,
+      "loss": 2.8716,
+      "step": 58577
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7277437448501587,
+      "learning_rate": 2.743043009888035e-06,
+      "loss": 3.0029,
+      "step": 58578
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7496187686920166,
+      "learning_rate": 2.7409661505420675e-06,
+      "loss": 2.9366,
+      "step": 58579
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7967445254325867,
+      "learning_rate": 2.738890074119193e-06,
+      "loss": 2.8992,
+      "step": 58580
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7687461376190186,
+      "learning_rate": 2.736814780624874e-06,
+      "loss": 2.6873,
+      "step": 58581
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7759029269218445,
+      "learning_rate": 2.7347402700645393e-06,
+      "loss": 3.1086,
+      "step": 58582
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8095799684524536,
+      "learning_rate": 2.7326665424436844e-06,
+      "loss": 2.8247,
+      "step": 58583
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7569171786308289,
+      "learning_rate": 2.730593597767805e-06,
+      "loss": 2.8072,
+      "step": 58584
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7547518014907837,
+      "learning_rate": 2.72852143604233e-06,
+      "loss": 2.8641,
+      "step": 58585
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7172838449478149,
+      "learning_rate": 2.7264500572727222e-06,
+      "loss": 2.9092,
+      "step": 58586
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7331339120864868,
+      "learning_rate": 2.724379461464443e-06,
+      "loss": 3.0988,
+      "step": 58587
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7577721476554871,
+      "learning_rate": 2.722309648622889e-06,
+      "loss": 2.5406,
+      "step": 58588
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7274854183197021,
+      "learning_rate": 2.7202406187535887e-06,
+      "loss": 2.9189,
+      "step": 58589
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7488541007041931,
+      "learning_rate": 2.7181723718619707e-06,
+      "loss": 2.788,
+      "step": 58590
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7572417855262756,
+      "learning_rate": 2.7161049079534646e-06,
+      "loss": 2.7814,
+      "step": 58591
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7595935463905334,
+      "learning_rate": 2.7140382270334994e-06,
+      "loss": 2.8215,
+      "step": 58592
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7744547724723816,
+      "learning_rate": 2.71197232910757e-06,
+      "loss": 2.7036,
+      "step": 58593
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7563602924346924,
+      "learning_rate": 2.709907214181106e-06,
+      "loss": 2.7836,
+      "step": 58594
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7460562586784363,
+      "learning_rate": 2.7078428822595034e-06,
+      "loss": 2.9322,
+      "step": 58595
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7538842558860779,
+      "learning_rate": 2.705779333348257e-06,
+      "loss": 2.8706,
+      "step": 58596
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7640366554260254,
+      "learning_rate": 2.703716567452763e-06,
+      "loss": 2.8659,
+      "step": 58597
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7012739777565002,
+      "learning_rate": 2.7016545845784497e-06,
+      "loss": 2.9261,
+      "step": 58598
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7518980503082275,
+      "learning_rate": 2.6995933847307473e-06,
+      "loss": 2.8051,
+      "step": 58599
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7138656973838806,
+      "learning_rate": 2.697532967915117e-06,
+      "loss": 2.7991,
+      "step": 58600
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7334188222885132,
+      "learning_rate": 2.6954733341369884e-06,
+      "loss": 2.888,
+      "step": 58601
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7425467371940613,
+      "learning_rate": 2.6934144834017567e-06,
+      "loss": 2.5813,
+      "step": 58602
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7227640151977539,
+      "learning_rate": 2.6913564157148515e-06,
+      "loss": 2.7079,
+      "step": 58603
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7601994276046753,
+      "learning_rate": 2.689299131081668e-06,
+      "loss": 2.9336,
+      "step": 58604
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7707348465919495,
+      "learning_rate": 2.6872426295076686e-06,
+      "loss": 3.0102,
+      "step": 58605
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7580685019493103,
+      "learning_rate": 2.685186910998283e-06,
+      "loss": 3.0894,
+      "step": 58606
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7323053479194641,
+      "learning_rate": 2.6831319755588387e-06,
+      "loss": 3.0489,
+      "step": 58607
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7686645984649658,
+      "learning_rate": 2.681077823194866e-06,
+      "loss": 2.8456,
+      "step": 58608
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7553392052650452,
+      "learning_rate": 2.6790244539116935e-06,
+      "loss": 2.83,
+      "step": 58609
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7549324631690979,
+      "learning_rate": 2.6769718677147167e-06,
+      "loss": 3.0199,
+      "step": 58610
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.758166491985321,
+      "learning_rate": 2.6749200646093982e-06,
+      "loss": 2.9844,
+      "step": 58611
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7355962991714478,
+      "learning_rate": 2.6728690446011336e-06,
+      "loss": 3.0423,
+      "step": 58612
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7758812308311462,
+      "learning_rate": 2.6708188076952854e-06,
+      "loss": 2.8497,
+      "step": 58613
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7655168175697327,
+      "learning_rate": 2.668769353897282e-06,
+      "loss": 2.911,
+      "step": 58614
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.731803834438324,
+      "learning_rate": 2.6667206832125197e-06,
+      "loss": 2.5736,
+      "step": 58615
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.727342426776886,
+      "learning_rate": 2.6646727956463943e-06,
+      "loss": 3.0545,
+      "step": 58616
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7434524297714233,
+      "learning_rate": 2.6626256912043343e-06,
+      "loss": 3.0437,
+      "step": 58617
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.805879533290863,
+      "learning_rate": 2.660579369891669e-06,
+      "loss": 2.9437,
+      "step": 58618
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.76210618019104,
+      "learning_rate": 2.658533831713794e-06,
+      "loss": 2.8384,
+      "step": 58619
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7613382935523987,
+      "learning_rate": 2.6564890766761047e-06,
+      "loss": 2.7502,
+      "step": 58620
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7467081546783447,
+      "learning_rate": 2.654445104784031e-06,
+      "loss": 2.6637,
+      "step": 58621
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7348057627677917,
+      "learning_rate": 2.652401916042901e-06,
+      "loss": 2.9316,
+      "step": 58622
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8229193687438965,
+      "learning_rate": 2.6503595104581444e-06,
+      "loss": 3.0189,
+      "step": 58623
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.741585910320282,
+      "learning_rate": 2.648317888035123e-06,
+      "loss": 2.9296,
+      "step": 58624
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7843425869941711,
+      "learning_rate": 2.6462770487791663e-06,
+      "loss": 2.562,
+      "step": 58625
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7439048290252686,
+      "learning_rate": 2.644236992695703e-06,
+      "loss": 3.0224,
+      "step": 58626
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7619640827178955,
+      "learning_rate": 2.6421977197900956e-06,
+      "loss": 2.9144,
+      "step": 58627
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7499315142631531,
+      "learning_rate": 2.64015923006774e-06,
+      "loss": 2.5829,
+      "step": 58628
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7854250073432922,
+      "learning_rate": 2.6381215235339315e-06,
+      "loss": 2.7211,
+      "step": 58629
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7892012000083923,
+      "learning_rate": 2.6360846001940994e-06,
+      "loss": 2.8919,
+      "step": 58630
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7646967768669128,
+      "learning_rate": 2.634048460053573e-06,
+      "loss": 3.0164,
+      "step": 58631
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7222882509231567,
+      "learning_rate": 2.6320131031177805e-06,
+      "loss": 2.8806,
+      "step": 58632
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7930301427841187,
+      "learning_rate": 2.6299785293920184e-06,
+      "loss": 2.9666,
+      "step": 58633
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.775130569934845,
+      "learning_rate": 2.627944738881649e-06,
+      "loss": 2.9753,
+      "step": 58634
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7208459377288818,
+      "learning_rate": 2.625911731592034e-06,
+      "loss": 3.2084,
+      "step": 58635
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7527031898498535,
+      "learning_rate": 2.62387950752857e-06,
+      "loss": 2.8473,
+      "step": 58636
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7546898722648621,
+      "learning_rate": 2.621848066696519e-06,
+      "loss": 2.799,
+      "step": 58637
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7190507054328918,
+      "learning_rate": 2.6198174091013435e-06,
+      "loss": 2.8375,
+      "step": 58638
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7207900285720825,
+      "learning_rate": 2.617787534748339e-06,
+      "loss": 2.6189,
+      "step": 58639
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7409490346908569,
+      "learning_rate": 2.6157584436428016e-06,
+      "loss": 2.9473,
+      "step": 58640
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7580714225769043,
+      "learning_rate": 2.6137301357901596e-06,
+      "loss": 2.8106,
+      "step": 58641
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.792218804359436,
+      "learning_rate": 2.6117026111957097e-06,
+      "loss": 2.783,
+      "step": 58642
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7418773770332336,
+      "learning_rate": 2.6096758698648137e-06,
+      "loss": 3.0417,
+      "step": 58643
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7960341572761536,
+      "learning_rate": 2.607649911802767e-06,
+      "loss": 2.8398,
+      "step": 58644
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7470495700836182,
+      "learning_rate": 2.605624737014966e-06,
+      "loss": 3.0263,
+      "step": 58645
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7760054469108582,
+      "learning_rate": 2.60360034550674e-06,
+      "loss": 2.8589,
+      "step": 58646
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7305024266242981,
+      "learning_rate": 2.6015767372833507e-06,
+      "loss": 2.8561,
+      "step": 58647
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.753297746181488,
+      "learning_rate": 2.599553912350194e-06,
+      "loss": 2.9018,
+      "step": 58648
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7611088156700134,
+      "learning_rate": 2.5975318707125325e-06,
+      "loss": 2.9396,
+      "step": 58649
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.768486499786377,
+      "learning_rate": 2.595510612375762e-06,
+      "loss": 2.6987,
+      "step": 58650
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8263199925422668,
+      "learning_rate": 2.593490137345211e-06,
+      "loss": 2.8593,
+      "step": 58651
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7508795857429504,
+      "learning_rate": 2.5914704456261428e-06,
+      "loss": 2.9293,
+      "step": 58652
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.790963888168335,
+      "learning_rate": 2.5894515372238854e-06,
+      "loss": 2.7323,
+      "step": 58653
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7731552124023438,
+      "learning_rate": 2.5874334121438356e-06,
+      "loss": 2.994,
+      "step": 58654
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7483537197113037,
+      "learning_rate": 2.5854160703911885e-06,
+      "loss": 2.8272,
+      "step": 58655
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.749880850315094,
+      "learning_rate": 2.5833995119713403e-06,
+      "loss": 2.9543,
+      "step": 58656
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7379136681556702,
+      "learning_rate": 2.581383736889553e-06,
+      "loss": 2.9406,
+      "step": 58657
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.753685474395752,
+      "learning_rate": 2.5793687451512223e-06,
+      "loss": 2.8851,
+      "step": 58658
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7053732872009277,
+      "learning_rate": 2.5773545367615444e-06,
+      "loss": 2.9854,
+      "step": 58659
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7595216035842896,
+      "learning_rate": 2.5753411117258817e-06,
+      "loss": 2.8745,
+      "step": 58660
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7526916861534119,
+      "learning_rate": 2.5733284700495295e-06,
+      "loss": 2.9997,
+      "step": 58661
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7947744131088257,
+      "learning_rate": 2.5713166117377836e-06,
+      "loss": 2.8262,
+      "step": 58662
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7724664211273193,
+      "learning_rate": 2.5693055367959737e-06,
+      "loss": 2.7696,
+      "step": 58663
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7890022397041321,
+      "learning_rate": 2.5672952452293617e-06,
+      "loss": 2.8872,
+      "step": 58664
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7602983713150024,
+      "learning_rate": 2.5652857370432432e-06,
+      "loss": 2.9562,
+      "step": 58665
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7163755297660828,
+      "learning_rate": 2.5632770122429148e-06,
+      "loss": 2.8898,
+      "step": 58666
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7645668983459473,
+      "learning_rate": 2.5612690708337046e-06,
+      "loss": 2.8835,
+      "step": 58667
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7817904949188232,
+      "learning_rate": 2.5592619128208424e-06,
+      "loss": 3.0378,
+      "step": 58668
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7333517670631409,
+      "learning_rate": 2.5572555382096236e-06,
+      "loss": 2.9638,
+      "step": 58669
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7220652103424072,
+      "learning_rate": 2.555249947005378e-06,
+      "loss": 2.7169,
+      "step": 58670
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7610824108123779,
+      "learning_rate": 2.5532451392133667e-06,
+      "loss": 2.8406,
+      "step": 58671
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8026726245880127,
+      "learning_rate": 2.55124111483882e-06,
+      "loss": 3.1763,
+      "step": 58672
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7766146659851074,
+      "learning_rate": 2.5492378738871e-06,
+      "loss": 2.7902,
+      "step": 58673
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7230070233345032,
+      "learning_rate": 2.5472354163634357e-06,
+      "loss": 2.7006,
+      "step": 58674
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7731808423995972,
+      "learning_rate": 2.5452337422730894e-06,
+      "loss": 3.0314,
+      "step": 58675
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7285119891166687,
+      "learning_rate": 2.5432328516213574e-06,
+      "loss": 2.8947,
+      "step": 58676
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.746861457824707,
+      "learning_rate": 2.5412327444135017e-06,
+      "loss": 2.9094,
+      "step": 58677
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7949832677841187,
+      "learning_rate": 2.539233420654818e-06,
+      "loss": 2.8981,
+      "step": 58678
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7782156467437744,
+      "learning_rate": 2.5372348803505026e-06,
+      "loss": 2.8062,
+      "step": 58679
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7405202984809875,
+      "learning_rate": 2.5352371235058845e-06,
+      "loss": 2.943,
+      "step": 58680
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7959614396095276,
+      "learning_rate": 2.5332401501262256e-06,
+      "loss": 2.7178,
+      "step": 58681
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7623839974403381,
+      "learning_rate": 2.5312439602167555e-06,
+      "loss": 2.9395,
+      "step": 58682
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7368699312210083,
+      "learning_rate": 2.529248553782737e-06,
+      "loss": 2.8478,
+      "step": 58683
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7328764796257019,
+      "learning_rate": 2.527253930829398e-06,
+      "loss": 2.7829,
+      "step": 58684
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7400158643722534,
+      "learning_rate": 2.525260091362036e-06,
+      "loss": 2.8827,
+      "step": 58685
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7456002831459045,
+      "learning_rate": 2.523267035385912e-06,
+      "loss": 2.8261,
+      "step": 58686
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.726157009601593,
+      "learning_rate": 2.5212747629062224e-06,
+      "loss": 2.7118,
+      "step": 58687
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7584595680236816,
+      "learning_rate": 2.519283273928263e-06,
+      "loss": 2.9325,
+      "step": 58688
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8131120204925537,
+      "learning_rate": 2.5172925684572633e-06,
+      "loss": 3.0036,
+      "step": 58689
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.760067880153656,
+      "learning_rate": 2.5153026464984517e-06,
+      "loss": 2.9375,
+      "step": 58690
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7592241764068604,
+      "learning_rate": 2.513313508057091e-06,
+      "loss": 2.9265,
+      "step": 58691
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.800504744052887,
+      "learning_rate": 2.5113251531383773e-06,
+      "loss": 2.8072,
+      "step": 58692
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7303928732872009,
+      "learning_rate": 2.509337581747606e-06,
+      "loss": 2.8293,
+      "step": 58693
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8083999156951904,
+      "learning_rate": 2.5073507938900064e-06,
+      "loss": 2.6546,
+      "step": 58694
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7412149310112,
+      "learning_rate": 2.505364789570774e-06,
+      "loss": 2.9867,
+      "step": 58695
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7698215246200562,
+      "learning_rate": 2.5033795687951387e-06,
+      "loss": 3.118,
+      "step": 58696
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.77435702085495,
+      "learning_rate": 2.501395131568362e-06,
+      "loss": 3.0678,
+      "step": 58697
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7428318858146667,
+      "learning_rate": 2.499411477895674e-06,
+      "loss": 2.9809,
+      "step": 58698
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7360471487045288,
+      "learning_rate": 2.497428607782237e-06,
+      "loss": 2.9411,
+      "step": 58699
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7438157796859741,
+      "learning_rate": 2.4954465212333465e-06,
+      "loss": 2.7105,
+      "step": 58700
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6961084604263306,
+      "learning_rate": 2.493465218254198e-06,
+      "loss": 2.7633,
+      "step": 58701
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7811095118522644,
+      "learning_rate": 2.491484698849988e-06,
+      "loss": 2.8801,
+      "step": 58702
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7979738116264343,
+      "learning_rate": 2.489504963025979e-06,
+      "loss": 2.9504,
+      "step": 58703
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8025102615356445,
+      "learning_rate": 2.487526010787333e-06,
+      "loss": 2.6882,
+      "step": 58704
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.771865725517273,
+      "learning_rate": 2.4855478421393126e-06,
+      "loss": 2.8791,
+      "step": 58705
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7456363439559937,
+      "learning_rate": 2.4835704570870806e-06,
+      "loss": 2.6352,
+      "step": 58706
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7263805866241455,
+      "learning_rate": 2.481593855635833e-06,
+      "loss": 2.8166,
+      "step": 58707
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7499218583106995,
+      "learning_rate": 2.479618037790865e-06,
+      "loss": 2.9985,
+      "step": 58708
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.724136471748352,
+      "learning_rate": 2.4776430035572723e-06,
+      "loss": 2.9608,
+      "step": 58709
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7608489990234375,
+      "learning_rate": 2.475668752940352e-06,
+      "loss": 2.6573,
+      "step": 58710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7402869462966919,
+      "learning_rate": 2.4736952859452653e-06,
+      "loss": 3.0642,
+      "step": 58711
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7280595302581787,
+      "learning_rate": 2.471722602577175e-06,
+      "loss": 2.778,
+      "step": 58712
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7287282943725586,
+      "learning_rate": 2.469750702841311e-06,
+      "loss": 2.8341,
+      "step": 58713
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7806359529495239,
+      "learning_rate": 2.4677795867428684e-06,
+      "loss": 3.1066,
+      "step": 58714
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7894913554191589,
+      "learning_rate": 2.4658092542870434e-06,
+      "loss": 3.0778,
+      "step": 58715
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7530649304389954,
+      "learning_rate": 2.4638397054789984e-06,
+      "loss": 2.9428,
+      "step": 58716
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7815943360328674,
+      "learning_rate": 2.461870940323929e-06,
+      "loss": 3.0238,
+      "step": 58717
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7556338310241699,
+      "learning_rate": 2.459902958827065e-06,
+      "loss": 2.8127,
+      "step": 58718
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7679249048233032,
+      "learning_rate": 2.4579357609935344e-06,
+      "loss": 2.9174,
+      "step": 58719
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8016083836555481,
+      "learning_rate": 2.455969346828568e-06,
+      "loss": 2.8592,
+      "step": 58720
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7322315573692322,
+      "learning_rate": 2.4540037163372606e-06,
+      "loss": 3.0443,
+      "step": 58721
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7427347898483276,
+      "learning_rate": 2.452038869524875e-06,
+      "loss": 2.7349,
+      "step": 58722
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7253776788711548,
+      "learning_rate": 2.45007480639654e-06,
+      "loss": 2.9283,
+      "step": 58723
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7481330037117004,
+      "learning_rate": 2.4481115269574526e-06,
+      "loss": 2.8383,
+      "step": 58724
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7602149248123169,
+      "learning_rate": 2.446149031212774e-06,
+      "loss": 2.7756,
+      "step": 58725
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7619256377220154,
+      "learning_rate": 2.4441873191676677e-06,
+      "loss": 3.0721,
+      "step": 58726
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7644228935241699,
+      "learning_rate": 2.4422263908272953e-06,
+      "loss": 2.8749,
+      "step": 58727
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7094855904579163,
+      "learning_rate": 2.4402662461968538e-06,
+      "loss": 3.0339,
+      "step": 58728
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7785953283309937,
+      "learning_rate": 2.4383068852814715e-06,
+      "loss": 2.9188,
+      "step": 58729
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7782620191574097,
+      "learning_rate": 2.4363483080863446e-06,
+      "loss": 2.7007,
+      "step": 58730
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7648835778236389,
+      "learning_rate": 2.434390514616602e-06,
+      "loss": 3.0109,
+      "step": 58731
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7509472966194153,
+      "learning_rate": 2.4324335048773734e-06,
+      "loss": 2.8961,
+      "step": 58732
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7752648591995239,
+      "learning_rate": 2.430477278873888e-06,
+      "loss": 2.8416,
+      "step": 58733
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.775607168674469,
+      "learning_rate": 2.428521836611241e-06,
+      "loss": 2.7793,
+      "step": 58734
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7678676247596741,
+      "learning_rate": 2.426567178094596e-06,
+      "loss": 2.8429,
+      "step": 58735
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7633538246154785,
+      "learning_rate": 2.4246133033291148e-06,
+      "loss": 2.7533,
+      "step": 58736
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7499951124191284,
+      "learning_rate": 2.4226602123199267e-06,
+      "loss": 3.0187,
+      "step": 58737
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8302484750747681,
+      "learning_rate": 2.420707905072161e-06,
+      "loss": 2.9427,
+      "step": 58738
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7577258348464966,
+      "learning_rate": 2.418756381591014e-06,
+      "loss": 2.967,
+      "step": 58739
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7333148121833801,
+      "learning_rate": 2.416805641881581e-06,
+      "loss": 2.869,
+      "step": 58740
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8242916464805603,
+      "learning_rate": 2.4148556859490242e-06,
+      "loss": 2.9112,
+      "step": 58741
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7202953696250916,
+      "learning_rate": 2.4129065137984404e-06,
+      "loss": 2.6746,
+      "step": 58742
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7955062389373779,
+      "learning_rate": 2.410958125434992e-06,
+      "loss": 2.7812,
+      "step": 58743
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8054195642471313,
+      "learning_rate": 2.4090105208638078e-06,
+      "loss": 3.0059,
+      "step": 58744
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7250577807426453,
+      "learning_rate": 2.4070637000900174e-06,
+      "loss": 2.8463,
+      "step": 58745
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6978896260261536,
+      "learning_rate": 2.4051176631187497e-06,
+      "loss": 2.8534,
+      "step": 58746
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.794734001159668,
+      "learning_rate": 2.403172409955134e-06,
+      "loss": 2.6841,
+      "step": 58747
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7843784093856812,
+      "learning_rate": 2.4012279406043e-06,
+      "loss": 2.7228,
+      "step": 58748
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7873679399490356,
+      "learning_rate": 2.3992842550713433e-06,
+      "loss": 2.8963,
+      "step": 58749
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7977191209793091,
+      "learning_rate": 2.397341353361393e-06,
+      "loss": 2.8242,
+      "step": 58750
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7681131958961487,
+      "learning_rate": 2.395399235479578e-06,
+      "loss": 2.8686,
+      "step": 58751
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7318379282951355,
+      "learning_rate": 2.393457901430995e-06,
+      "loss": 2.9615,
+      "step": 58752
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7517814040184021,
+      "learning_rate": 2.391517351220773e-06,
+      "loss": 2.8989,
+      "step": 58753
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7598768472671509,
+      "learning_rate": 2.3895775848540413e-06,
+      "loss": 2.9079,
+      "step": 58754
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7105078101158142,
+      "learning_rate": 2.387638602335862e-06,
+      "loss": 2.8294,
+      "step": 58755
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7269749641418457,
+      "learning_rate": 2.385700403671398e-06,
+      "loss": 2.7675,
+      "step": 58756
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7759285569190979,
+      "learning_rate": 2.383762988865678e-06,
+      "loss": 2.9059,
+      "step": 58757
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.74515700340271,
+      "learning_rate": 2.381826357923866e-06,
+      "loss": 2.7272,
+      "step": 58758
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7521591186523438,
+      "learning_rate": 2.3798905108510567e-06,
+      "loss": 3.0477,
+      "step": 58759
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7557122707366943,
+      "learning_rate": 2.3779554476523134e-06,
+      "loss": 2.7349,
+      "step": 58760
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7690261602401733,
+      "learning_rate": 2.3760211683327644e-06,
+      "loss": 2.8652,
+      "step": 58761
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.780361533164978,
+      "learning_rate": 2.374087672897507e-06,
+      "loss": 2.8442,
+      "step": 58762
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7434288263320923,
+      "learning_rate": 2.3721549613516355e-06,
+      "loss": 2.88,
+      "step": 58763
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8056708574295044,
+      "learning_rate": 2.3702230337002135e-06,
+      "loss": 2.9113,
+      "step": 58764
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8034791946411133,
+      "learning_rate": 2.368291889948337e-06,
+      "loss": 2.7607,
+      "step": 58765
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7488557696342468,
+      "learning_rate": 2.3663615301011018e-06,
+      "loss": 2.7314,
+      "step": 58766
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7275140881538391,
+      "learning_rate": 2.364431954163604e-06,
+      "loss": 2.7287,
+      "step": 58767
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7067748308181763,
+      "learning_rate": 2.3625031621409053e-06,
+      "loss": 2.8275,
+      "step": 58768
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7348287105560303,
+      "learning_rate": 2.3605751540380692e-06,
+      "loss": 2.9979,
+      "step": 58769
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7248157262802124,
+      "learning_rate": 2.3586479298602245e-06,
+      "loss": 2.8282,
+      "step": 58770
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7620055079460144,
+      "learning_rate": 2.3567214896124007e-06,
+      "loss": 2.7483,
+      "step": 58771
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7315212488174438,
+      "learning_rate": 2.3547958332997276e-06,
+      "loss": 2.7802,
+      "step": 58772
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7276974320411682,
+      "learning_rate": 2.3528709609272e-06,
+      "loss": 2.9266,
+      "step": 58773
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7391351461410522,
+      "learning_rate": 2.3509468724999146e-06,
+      "loss": 2.9255,
+      "step": 58774
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8038753271102905,
+      "learning_rate": 2.3490235680230006e-06,
+      "loss": 2.572,
+      "step": 58775
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7655202150344849,
+      "learning_rate": 2.3471010475014207e-06,
+      "loss": 2.8412,
+      "step": 58776
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7686148881912231,
+      "learning_rate": 2.345179310940337e-06,
+      "loss": 3.075,
+      "step": 58777
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7740659117698669,
+      "learning_rate": 2.3432583583447462e-06,
+      "loss": 2.96,
+      "step": 58778
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6870248913764954,
+      "learning_rate": 2.34133818971971e-06,
+      "loss": 2.7546,
+      "step": 58779
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7673691511154175,
+      "learning_rate": 2.3394188050703587e-06,
+      "loss": 2.7017,
+      "step": 58780
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7266324162483215,
+      "learning_rate": 2.337500204401621e-06,
+      "loss": 2.9339,
+      "step": 58781
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7386431694030762,
+      "learning_rate": 2.335582387718693e-06,
+      "loss": 2.8721,
+      "step": 58782
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7468572854995728,
+      "learning_rate": 2.3336653550265037e-06,
+      "loss": 2.9022,
+      "step": 58783
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7643420100212097,
+      "learning_rate": 2.331749106330183e-06,
+      "loss": 3.2001,
+      "step": 58784
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7512932419776917,
+      "learning_rate": 2.3298336416347263e-06,
+      "loss": 2.9548,
+      "step": 58785
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7336491942405701,
+      "learning_rate": 2.3279189609451964e-06,
+      "loss": 2.8733,
+      "step": 58786
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7166445851325989,
+      "learning_rate": 2.326005064266656e-06,
+      "loss": 2.8924,
+      "step": 58787
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8269556760787964,
+      "learning_rate": 2.3240919516041014e-06,
+      "loss": 2.9191,
+      "step": 58788
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7374542355537415,
+      "learning_rate": 2.322179622962628e-06,
+      "loss": 2.9622,
+      "step": 58789
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7649866342544556,
+      "learning_rate": 2.3202680783472315e-06,
+      "loss": 2.8605,
+      "step": 58790
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7545380592346191,
+      "learning_rate": 2.3183573177629757e-06,
+      "loss": 2.9829,
+      "step": 58791
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7572214007377625,
+      "learning_rate": 2.316447341214889e-06,
+      "loss": 2.9821,
+      "step": 58792
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7295137047767639,
+      "learning_rate": 2.3145381487079674e-06,
+      "loss": 2.9972,
+      "step": 58793
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7803434133529663,
+      "learning_rate": 2.3126297402472404e-06,
+      "loss": 2.8749,
+      "step": 58794
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7383474707603455,
+      "learning_rate": 2.310722115837771e-06,
+      "loss": 2.8434,
+      "step": 58795
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8103850483894348,
+      "learning_rate": 2.308815275484588e-06,
+      "loss": 2.4542,
+      "step": 58796
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7786707282066345,
+      "learning_rate": 2.3069092191926873e-06,
+      "loss": 3.1745,
+      "step": 58797
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7713927626609802,
+      "learning_rate": 2.3050039469670656e-06,
+      "loss": 2.824,
+      "step": 58798
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7706761956214905,
+      "learning_rate": 2.3030994588128183e-06,
+      "loss": 2.8247,
+      "step": 58799
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7205535173416138,
+      "learning_rate": 2.301195754734875e-06,
+      "loss": 2.8544,
+      "step": 58800
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7274318933486938,
+      "learning_rate": 2.299292834738298e-06,
+      "loss": 2.9493,
+      "step": 58801
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7716490030288696,
+      "learning_rate": 2.2973906988281167e-06,
+      "loss": 2.9651,
+      "step": 58802
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7678560614585876,
+      "learning_rate": 2.295489347009294e-06,
+      "loss": 2.9299,
+      "step": 58803
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7848706245422363,
+      "learning_rate": 2.2935887792868257e-06,
+      "loss": 2.8571,
+      "step": 58804
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7634941339492798,
+      "learning_rate": 2.291688995665808e-06,
+      "loss": 2.9207,
+      "step": 58805
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7758709788322449,
+      "learning_rate": 2.2897899961511367e-06,
+      "loss": 2.8365,
+      "step": 58806
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7400304675102234,
+      "learning_rate": 2.287891780747908e-06,
+      "loss": 2.952,
+      "step": 58807
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7655829787254333,
+      "learning_rate": 2.285994349461051e-06,
+      "loss": 2.6617,
+      "step": 58808
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.738251268863678,
+      "learning_rate": 2.284097702295562e-06,
+      "loss": 2.9595,
+      "step": 58809
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7782675623893738,
+      "learning_rate": 2.2822018392565035e-06,
+      "loss": 2.7474,
+      "step": 58810
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7073926329612732,
+      "learning_rate": 2.2803067603488378e-06,
+      "loss": 2.8823,
+      "step": 58811
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8041476607322693,
+      "learning_rate": 2.278412465577528e-06,
+      "loss": 2.7922,
+      "step": 58812
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7299349308013916,
+      "learning_rate": 2.2765189549475703e-06,
+      "loss": 3.0036,
+      "step": 58813
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7464435696601868,
+      "learning_rate": 2.2746262284639607e-06,
+      "loss": 2.8143,
+      "step": 58814
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7710967659950256,
+      "learning_rate": 2.2727342861317275e-06,
+      "loss": 2.8977,
+      "step": 58815
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7730777263641357,
+      "learning_rate": 2.270843127955768e-06,
+      "loss": 2.773,
+      "step": 58816
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7543210983276367,
+      "learning_rate": 2.2689527539411445e-06,
+      "loss": 2.9257,
+      "step": 58817
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7473542094230652,
+      "learning_rate": 2.2670631640927527e-06,
+      "loss": 2.7792,
+      "step": 58818
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7523601055145264,
+      "learning_rate": 2.2651743584156555e-06,
+      "loss": 2.9197,
+      "step": 58819
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.750109851360321,
+      "learning_rate": 2.2632863369147824e-06,
+      "loss": 3.0966,
+      "step": 58820
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7598704099655151,
+      "learning_rate": 2.261399099595129e-06,
+      "loss": 2.9957,
+      "step": 58821
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7527625560760498,
+      "learning_rate": 2.2595126464616254e-06,
+      "loss": 2.9311,
+      "step": 58822
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7135601043701172,
+      "learning_rate": 2.257626977519267e-06,
+      "loss": 2.9232,
+      "step": 58823
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7807445526123047,
+      "learning_rate": 2.2557420927730163e-06,
+      "loss": 2.9763,
+      "step": 58824
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7743086814880371,
+      "learning_rate": 2.2538579922278364e-06,
+      "loss": 3.0103,
+      "step": 58825
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7185844779014587,
+      "learning_rate": 2.2519746758886904e-06,
+      "loss": 2.9725,
+      "step": 58826
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7576574683189392,
+      "learning_rate": 2.2500921437605736e-06,
+      "loss": 2.8631,
+      "step": 58827
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7718368172645569,
+      "learning_rate": 2.2482103958483823e-06,
+      "loss": 3.0295,
+      "step": 58828
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7165724635124207,
+      "learning_rate": 2.246329432157079e-06,
+      "loss": 2.7588,
+      "step": 58829
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7765059471130371,
+      "learning_rate": 2.2444492526916935e-06,
+      "loss": 2.832,
+      "step": 58830
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7383131980895996,
+      "learning_rate": 2.2425698574570884e-06,
+      "loss": 2.8534,
+      "step": 58831
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7530897259712219,
+      "learning_rate": 2.2406912464582594e-06,
+      "loss": 2.5967,
+      "step": 58832
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7495601773262024,
+      "learning_rate": 2.238813419700136e-06,
+      "loss": 2.8738,
+      "step": 58833
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7114120125770569,
+      "learning_rate": 2.236936377187681e-06,
+      "loss": 2.8921,
+      "step": 58834
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8221262693405151,
+      "learning_rate": 2.2350601189258243e-06,
+      "loss": 2.7301,
+      "step": 58835
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7930930852890015,
+      "learning_rate": 2.233184644919528e-06,
+      "loss": 2.8669,
+      "step": 58836
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7594324350357056,
+      "learning_rate": 2.2313099551737213e-06,
+      "loss": 2.8567,
+      "step": 58837
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.733824610710144,
+      "learning_rate": 2.2294360496933007e-06,
+      "loss": 2.9,
+      "step": 58838
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7296682000160217,
+      "learning_rate": 2.2275629284832618e-06,
+      "loss": 2.8983,
+      "step": 58839
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7384452819824219,
+      "learning_rate": 2.2256905915485013e-06,
+      "loss": 2.8469,
+      "step": 58840
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7113006114959717,
+      "learning_rate": 2.2238190388939815e-06,
+      "loss": 2.9981,
+      "step": 58841
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7444003224372864,
+      "learning_rate": 2.2219482705246317e-06,
+      "loss": 2.7707,
+      "step": 58842
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7594712972640991,
+      "learning_rate": 2.220078286445315e-06,
+      "loss": 2.9082,
+      "step": 58843
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7738532423973083,
+      "learning_rate": 2.2182090866610604e-06,
+      "loss": 2.8424,
+      "step": 58844
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7955849170684814,
+      "learning_rate": 2.2163406711766975e-06,
+      "loss": 2.9444,
+      "step": 58845
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7212498188018799,
+      "learning_rate": 2.2144730399971554e-06,
+      "loss": 2.9951,
+      "step": 58846
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7770746350288391,
+      "learning_rate": 2.2126061931274307e-06,
+      "loss": 2.9731,
+      "step": 58847
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7356753945350647,
+      "learning_rate": 2.210740130572353e-06,
+      "loss": 2.8022,
+      "step": 58848
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7566166520118713,
+      "learning_rate": 2.208874852336917e-06,
+      "loss": 2.7831,
+      "step": 58849
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7645593881607056,
+      "learning_rate": 2.2070103584259867e-06,
+      "loss": 2.8821,
+      "step": 58850
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.729720413684845,
+      "learning_rate": 2.2051466488444246e-06,
+      "loss": 3.0565,
+      "step": 58851
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7721166610717773,
+      "learning_rate": 2.20328372359726e-06,
+      "loss": 2.9014,
+      "step": 58852
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7474196553230286,
+      "learning_rate": 2.201421582689289e-06,
+      "loss": 2.9263,
+      "step": 58853
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7505802512168884,
+      "learning_rate": 2.1995602261255074e-06,
+      "loss": 2.9056,
+      "step": 58854
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7443969249725342,
+      "learning_rate": 2.1976996539107117e-06,
+      "loss": 3.0748,
+      "step": 58855
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7555038332939148,
+      "learning_rate": 2.1958398660498976e-06,
+      "loss": 2.9408,
+      "step": 58856
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7873604893684387,
+      "learning_rate": 2.193980862547928e-06,
+      "loss": 2.9474,
+      "step": 58857
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7833068370819092,
+      "learning_rate": 2.1921226434096995e-06,
+      "loss": 2.9086,
+      "step": 58858
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7554247379302979,
+      "learning_rate": 2.1902652086401075e-06,
+      "loss": 2.8782,
+      "step": 58859
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7960243821144104,
+      "learning_rate": 2.188408558244048e-06,
+      "loss": 2.8816,
+      "step": 58860
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8111169338226318,
+      "learning_rate": 2.1865526922263842e-06,
+      "loss": 2.7859,
+      "step": 58861
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.800476610660553,
+      "learning_rate": 2.1846976105920456e-06,
+      "loss": 2.8443,
+      "step": 58862
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7481632828712463,
+      "learning_rate": 2.1828433133458613e-06,
+      "loss": 2.7816,
+      "step": 58863
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7516404986381531,
+      "learning_rate": 2.180989800492794e-06,
+      "loss": 2.7836,
+      "step": 58864
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7697651982307434,
+      "learning_rate": 2.1791370720376733e-06,
+      "loss": 2.772,
+      "step": 58865
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7722952365875244,
+      "learning_rate": 2.177285127985362e-06,
+      "loss": 2.7993,
+      "step": 58866
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7531412839889526,
+      "learning_rate": 2.1754339683407897e-06,
+      "loss": 2.6921,
+      "step": 58867
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7838771939277649,
+      "learning_rate": 2.1735835931088187e-06,
+      "loss": 3.0806,
+      "step": 58868
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7424795031547546,
+      "learning_rate": 2.1717340022942785e-06,
+      "loss": 2.8725,
+      "step": 58869
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.6980968713760376,
+      "learning_rate": 2.169885195902099e-06,
+      "loss": 2.9858,
+      "step": 58870
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7613954544067383,
+      "learning_rate": 2.1680371739371093e-06,
+      "loss": 3.0112,
+      "step": 58871
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7577490210533142,
+      "learning_rate": 2.166189936404206e-06,
+      "loss": 3.0586,
+      "step": 58872
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7090437412261963,
+      "learning_rate": 2.1643434833082176e-06,
+      "loss": 2.9531,
+      "step": 58873
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7910329699516296,
+      "learning_rate": 2.1624978146540407e-06,
+      "loss": 2.6705,
+      "step": 58874
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7528069019317627,
+      "learning_rate": 2.1606529304465046e-06,
+      "loss": 3.0284,
+      "step": 58875
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.759083092212677,
+      "learning_rate": 2.158808830690506e-06,
+      "loss": 2.9894,
+      "step": 58876
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.764187753200531,
+      "learning_rate": 2.156965515390907e-06,
+      "loss": 2.8765,
+      "step": 58877
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7952934503555298,
+      "learning_rate": 2.155122984552504e-06,
+      "loss": 2.8857,
+      "step": 58878
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.751152753829956,
+      "learning_rate": 2.1532812381801932e-06,
+      "loss": 2.8625,
+      "step": 58879
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7216384410858154,
+      "learning_rate": 2.1514402762788376e-06,
+      "loss": 3.1293,
+      "step": 58880
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7260209321975708,
+      "learning_rate": 2.1496000988532327e-06,
+      "loss": 2.8408,
+      "step": 58881
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.756483793258667,
+      "learning_rate": 2.147760705908308e-06,
+      "loss": 2.9624,
+      "step": 58882
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8064687848091125,
+      "learning_rate": 2.1459220974487933e-06,
+      "loss": 2.9032,
+      "step": 58883
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8111887574195862,
+      "learning_rate": 2.1440842734796514e-06,
+      "loss": 2.8218,
+      "step": 58884
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7237271070480347,
+      "learning_rate": 2.142247234005645e-06,
+      "loss": 2.879,
+      "step": 58885
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.699209451675415,
+      "learning_rate": 2.140410979031637e-06,
+      "loss": 2.7129,
+      "step": 58886
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7380729913711548,
+      "learning_rate": 2.13857550856249e-06,
+      "loss": 2.9703,
+      "step": 58887
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7376232147216797,
+      "learning_rate": 2.136740822602967e-06,
+      "loss": 2.9186,
+      "step": 58888
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7537232637405396,
+      "learning_rate": 2.1349069211579637e-06,
+      "loss": 3.0193,
+      "step": 58889
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7629387378692627,
+      "learning_rate": 2.133073804232277e-06,
+      "loss": 2.8616,
+      "step": 58890
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7387884259223938,
+      "learning_rate": 2.1312414718307693e-06,
+      "loss": 2.8796,
+      "step": 58891
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.701157808303833,
+      "learning_rate": 2.129409923958236e-06,
+      "loss": 2.829,
+      "step": 58892
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8402901291847229,
+      "learning_rate": 2.127579160619508e-06,
+      "loss": 2.6893,
+      "step": 58893
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.732913076877594,
+      "learning_rate": 2.125749181819414e-06,
+      "loss": 2.7791,
+      "step": 58894
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7669530510902405,
+      "learning_rate": 2.1239199875627834e-06,
+      "loss": 2.8068,
+      "step": 58895
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7334358096122742,
+      "learning_rate": 2.122091577854379e-06,
+      "loss": 2.9818,
+      "step": 58896
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.785017192363739,
+      "learning_rate": 2.1202639526990973e-06,
+      "loss": 2.8513,
+      "step": 58897
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7877793312072754,
+      "learning_rate": 2.1184371121017006e-06,
+      "loss": 2.8978,
+      "step": 58898
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7523209452629089,
+      "learning_rate": 2.116611056067019e-06,
+      "loss": 2.6516,
+      "step": 58899
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7709547877311707,
+      "learning_rate": 2.114785784599815e-06,
+      "loss": 2.7821,
+      "step": 58900
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7097203731536865,
+      "learning_rate": 2.1129612977049847e-06,
+      "loss": 2.9419,
+      "step": 58901
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7327682375907898,
+      "learning_rate": 2.1111375953872576e-06,
+      "loss": 2.8539,
+      "step": 58902
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7576892375946045,
+      "learning_rate": 2.1093146776514637e-06,
+      "loss": 2.9477,
+      "step": 58903
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7579607367515564,
+      "learning_rate": 2.1074925445024314e-06,
+      "loss": 2.7468,
+      "step": 58904
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7820842862129211,
+      "learning_rate": 2.1056711959449248e-06,
+      "loss": 2.9472,
+      "step": 58905
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7186954021453857,
+      "learning_rate": 2.1038506319837056e-06,
+      "loss": 2.9416,
+      "step": 58906
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.740797758102417,
+      "learning_rate": 2.102030852623671e-06,
+      "loss": 3.0116,
+      "step": 58907
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7477077841758728,
+      "learning_rate": 2.100211857869549e-06,
+      "loss": 3.0106,
+      "step": 58908
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7214314341545105,
+      "learning_rate": 2.098393647726104e-06,
+      "loss": 2.8677,
+      "step": 58909
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7203336358070374,
+      "learning_rate": 2.0965762221981654e-06,
+      "loss": 3.0648,
+      "step": 58910
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7489813566207886,
+      "learning_rate": 2.094759581290528e-06,
+      "loss": 2.7633,
+      "step": 58911
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7301415801048279,
+      "learning_rate": 2.092943725007956e-06,
+      "loss": 2.837,
+      "step": 58912
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8440554141998291,
+      "learning_rate": 2.091128653355245e-06,
+      "loss": 2.796,
+      "step": 58913
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7419830560684204,
+      "learning_rate": 2.089314366337158e-06,
+      "loss": 2.8407,
+      "step": 58914
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7549934387207031,
+      "learning_rate": 2.0875008639584912e-06,
+      "loss": 3.0267,
+      "step": 58915
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7184531092643738,
+      "learning_rate": 2.085688146224007e-06,
+      "loss": 2.6725,
+      "step": 58916
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7466804385185242,
+      "learning_rate": 2.0838762131384688e-06,
+      "loss": 2.9272,
+      "step": 58917
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7794800400733948,
+      "learning_rate": 2.0820650647066726e-06,
+      "loss": 2.7772,
+      "step": 58918
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7532921433448792,
+      "learning_rate": 2.0802547009333814e-06,
+      "loss": 3.0,
+      "step": 58919
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7411693334579468,
+      "learning_rate": 2.078445121823358e-06,
+      "loss": 3.0414,
+      "step": 58920
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7391944527626038,
+      "learning_rate": 2.076636327381398e-06,
+      "loss": 2.8976,
+      "step": 58921
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7387174963951111,
+      "learning_rate": 2.0748283176122317e-06,
+      "loss": 3.035,
+      "step": 58922
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7118051052093506,
+      "learning_rate": 2.0730210925206213e-06,
+      "loss": 2.8181,
+      "step": 58923
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7901133894920349,
+      "learning_rate": 2.0712146521113636e-06,
+      "loss": 3.0142,
+      "step": 58924
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7497255206108093,
+      "learning_rate": 2.0694089963891545e-06,
+      "loss": 2.7844,
+      "step": 58925
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7415169477462769,
+      "learning_rate": 2.067604125358824e-06,
+      "loss": 2.911,
+      "step": 58926
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.749427855014801,
+      "learning_rate": 2.065800039025034e-06,
+      "loss": 2.9933,
+      "step": 58927
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7294812202453613,
+      "learning_rate": 2.0639967373926146e-06,
+      "loss": 2.869,
+      "step": 58928
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7828423380851746,
+      "learning_rate": 2.0621942204662956e-06,
+      "loss": 2.7197,
+      "step": 58929
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7263489365577698,
+      "learning_rate": 2.060392488250806e-06,
+      "loss": 2.85,
+      "step": 58930
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7711845636367798,
+      "learning_rate": 2.0585915407509424e-06,
+      "loss": 3.1033,
+      "step": 58931
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7376527190208435,
+      "learning_rate": 2.0567913779713675e-06,
+      "loss": 2.7261,
+      "step": 58932
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7345303893089294,
+      "learning_rate": 2.0549919999168442e-06,
+      "loss": 2.7508,
+      "step": 58933
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7842707633972168,
+      "learning_rate": 2.0531934065921685e-06,
+      "loss": 2.9854,
+      "step": 58934
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7627545595169067,
+      "learning_rate": 2.051395598002037e-06,
+      "loss": 2.6343,
+      "step": 58935
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7491629719734192,
+      "learning_rate": 2.049598574151179e-06,
+      "loss": 2.8958,
+      "step": 58936
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7489625811576843,
+      "learning_rate": 2.047802335044324e-06,
+      "loss": 2.8476,
+      "step": 58937
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7320138812065125,
+      "learning_rate": 2.046006880686235e-06,
+      "loss": 2.9481,
+      "step": 58938
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7232726812362671,
+      "learning_rate": 2.0442122110816084e-06,
+      "loss": 2.924,
+      "step": 58939
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7281123399734497,
+      "learning_rate": 2.0424183262352066e-06,
+      "loss": 2.9472,
+      "step": 58940
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7523177266120911,
+      "learning_rate": 2.0406252261517263e-06,
+      "loss": 2.8302,
+      "step": 58941
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.738097608089447,
+      "learning_rate": 2.0388329108358636e-06,
+      "loss": 2.8233,
+      "step": 58942
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.727954089641571,
+      "learning_rate": 2.0370413802923813e-06,
+      "loss": 2.9305,
+      "step": 58943
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7692335844039917,
+      "learning_rate": 2.035250634526009e-06,
+      "loss": 2.9345,
+      "step": 58944
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7667322754859924,
+      "learning_rate": 2.0334606735414426e-06,
+      "loss": 2.9262,
+      "step": 58945
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8380236625671387,
+      "learning_rate": 2.0316714973434124e-06,
+      "loss": 3.0058,
+      "step": 58946
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.768031120300293,
+      "learning_rate": 2.0298831059365807e-06,
+      "loss": 3.2082,
+      "step": 58947
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.773243248462677,
+      "learning_rate": 2.028095499325677e-06,
+      "loss": 2.9019,
+      "step": 58948
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7496214509010315,
+      "learning_rate": 2.026308677515465e-06,
+      "loss": 2.9999,
+      "step": 58949
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7690075635910034,
+      "learning_rate": 2.024522640510573e-06,
+      "loss": 2.6612,
+      "step": 58950
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8003986477851868,
+      "learning_rate": 2.022737388315765e-06,
+      "loss": 2.8753,
+      "step": 58951
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7260475158691406,
+      "learning_rate": 2.020952920935737e-06,
+      "loss": 3.0065,
+      "step": 58952
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7463489770889282,
+      "learning_rate": 2.0191692383751514e-06,
+      "loss": 3.0693,
+      "step": 58953
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7311463356018066,
+      "learning_rate": 2.0173863406387047e-06,
+      "loss": 2.9125,
+      "step": 58954
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8149186372756958,
+      "learning_rate": 2.0156042277311267e-06,
+      "loss": 3.1956,
+      "step": 58955
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7813419103622437,
+      "learning_rate": 2.0138228996571136e-06,
+      "loss": 2.8011,
+      "step": 58956
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7746787071228027,
+      "learning_rate": 2.0120423564213285e-06,
+      "loss": 2.9801,
+      "step": 58957
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8212281465530396,
+      "learning_rate": 2.0102625980285002e-06,
+      "loss": 2.8556,
+      "step": 58958
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7449041604995728,
+      "learning_rate": 2.008483624483259e-06,
+      "loss": 3.0733,
+      "step": 58959
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7603235840797424,
+      "learning_rate": 2.006705435790368e-06,
+      "loss": 2.9102,
+      "step": 58960
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7589277625083923,
+      "learning_rate": 2.0049280319544225e-06,
+      "loss": 2.7227,
+      "step": 58961
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.752943754196167,
+      "learning_rate": 2.0031514129801527e-06,
+      "loss": 2.7721,
+      "step": 58962
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.768645703792572,
+      "learning_rate": 2.0013755788722217e-06,
+      "loss": 2.7773,
+      "step": 58963
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7890374660491943,
+      "learning_rate": 1.999600529635359e-06,
+      "loss": 2.7951,
+      "step": 58964
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7113425135612488,
+      "learning_rate": 1.9978262652741605e-06,
+      "loss": 2.9277,
+      "step": 58965
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.797177255153656,
+      "learning_rate": 1.996052785793323e-06,
+      "loss": 2.67,
+      "step": 58966
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7570541501045227,
+      "learning_rate": 1.9942800911975755e-06,
+      "loss": 2.9775,
+      "step": 58967
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8025187849998474,
+      "learning_rate": 1.9925081814915145e-06,
+      "loss": 2.9072,
+      "step": 58968
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7739835381507874,
+      "learning_rate": 1.9907370566798363e-06,
+      "loss": 2.955,
+      "step": 58969
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7543403506278992,
+      "learning_rate": 1.9889667167672042e-06,
+      "loss": 2.8965,
+      "step": 58970
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8121760487556458,
+      "learning_rate": 1.9871971617582804e-06,
+      "loss": 2.8277,
+      "step": 58971
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8004959225654602,
+      "learning_rate": 1.9854283916577285e-06,
+      "loss": 2.7487,
+      "step": 58972
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7454361915588379,
+      "learning_rate": 1.983660406470211e-06,
+      "loss": 2.7616,
+      "step": 58973
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7641414999961853,
+      "learning_rate": 1.9818932062003578e-06,
+      "loss": 2.9533,
+      "step": 58974
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7749798893928528,
+      "learning_rate": 1.980126790852865e-06,
+      "loss": 2.5778,
+      "step": 58975
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7524572610855103,
+      "learning_rate": 1.978361160432362e-06,
+      "loss": 2.9732,
+      "step": 58976
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7702212929725647,
+      "learning_rate": 1.9765963149434793e-06,
+      "loss": 2.9104,
+      "step": 58977
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.802010715007782,
+      "learning_rate": 1.974832254390879e-06,
+      "loss": 2.927,
+      "step": 58978
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7283293604850769,
+      "learning_rate": 1.9730689787792573e-06,
+      "loss": 3.0435,
+      "step": 58979
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7269130349159241,
+      "learning_rate": 1.971306488113178e-06,
+      "loss": 2.8377,
+      "step": 58980
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7806401252746582,
+      "learning_rate": 1.9695447823973364e-06,
+      "loss": 2.7224,
+      "step": 58981
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7446566224098206,
+      "learning_rate": 1.967783861636363e-06,
+      "loss": 2.9339,
+      "step": 58982
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8244256377220154,
+      "learning_rate": 1.9660237258348533e-06,
+      "loss": 2.8912,
+      "step": 58983
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7497897148132324,
+      "learning_rate": 1.9642643749975373e-06,
+      "loss": 2.8148,
+      "step": 58984
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8421869874000549,
+      "learning_rate": 1.9625058091289447e-06,
+      "loss": 3.0151,
+      "step": 58985
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7372134327888489,
+      "learning_rate": 1.9607480282337716e-06,
+      "loss": 2.8176,
+      "step": 58986
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7009263634681702,
+      "learning_rate": 1.958991032316615e-06,
+      "loss": 2.819,
+      "step": 58987
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7520594000816345,
+      "learning_rate": 1.9572348213821365e-06,
+      "loss": 2.8624,
+      "step": 58988
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7330684661865234,
+      "learning_rate": 1.9554793954349333e-06,
+      "loss": 2.7898,
+      "step": 58989
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7308124303817749,
+      "learning_rate": 1.953724754479635e-06,
+      "loss": 2.8689,
+      "step": 58990
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7374723553657532,
+      "learning_rate": 1.9519708985208717e-06,
+      "loss": 2.7598,
+      "step": 58991
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7027939558029175,
+      "learning_rate": 1.9502178275632718e-06,
+      "loss": 2.9557,
+      "step": 58992
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7598909139633179,
+      "learning_rate": 1.9484655416113993e-06,
+      "loss": 2.9631,
+      "step": 58993
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7432723641395569,
+      "learning_rate": 1.94671404066995e-06,
+      "loss": 3.0194,
+      "step": 58994
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7426236867904663,
+      "learning_rate": 1.944963324743454e-06,
+      "loss": 2.9044,
+      "step": 58995
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7190657258033752,
+      "learning_rate": 1.943213393836607e-06,
+      "loss": 2.4641,
+      "step": 58996
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7499857544898987,
+      "learning_rate": 1.9414642479539386e-06,
+      "loss": 3.0042,
+      "step": 58997
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7538277506828308,
+      "learning_rate": 1.939715887100113e-06,
+      "loss": 3.1177,
+      "step": 58998
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7844749093055725,
+      "learning_rate": 1.937968311279692e-06,
+      "loss": 2.7512,
+      "step": 58999
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7818543314933777,
+      "learning_rate": 1.9362215204972718e-06,
+      "loss": 2.9125,
+      "step": 59000
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7285245656967163,
+      "learning_rate": 1.9344755147575496e-06,
+      "loss": 2.6752,
+      "step": 59001
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.721545934677124,
+      "learning_rate": 1.932730294064988e-06,
+      "loss": 2.9226,
+      "step": 59002
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7440168857574463,
+      "learning_rate": 1.9309858584242834e-06,
+      "loss": 2.9274,
+      "step": 59003
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7558407783508301,
+      "learning_rate": 1.929242207840032e-06,
+      "loss": 2.8321,
+      "step": 59004
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7752707600593567,
+      "learning_rate": 1.9274993423167295e-06,
+      "loss": 2.9114,
+      "step": 59005
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7502283453941345,
+      "learning_rate": 1.9257572618590732e-06,
+      "loss": 2.8542,
+      "step": 59006
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7198836207389832,
+      "learning_rate": 1.9240159664715926e-06,
+      "loss": 2.6527,
+      "step": 59007
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7290161848068237,
+      "learning_rate": 1.9222754561588836e-06,
+      "loss": 2.8487,
+      "step": 59008
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7256250381469727,
+      "learning_rate": 1.920535730925543e-06,
+      "loss": 2.6648,
+      "step": 59009
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7327529191970825,
+      "learning_rate": 1.9187967907761335e-06,
+      "loss": 2.9445,
+      "step": 59010
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7599136233329773,
+      "learning_rate": 1.9170586357152518e-06,
+      "loss": 2.9671,
+      "step": 59011
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7350993752479553,
+      "learning_rate": 1.9153212657474605e-06,
+      "loss": 2.881,
+      "step": 59012
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7257763147354126,
+      "learning_rate": 1.913584680877356e-06,
+      "loss": 2.8071,
+      "step": 59013
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7590259909629822,
+      "learning_rate": 1.9118488811095013e-06,
+      "loss": 2.8561,
+      "step": 59014
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7977140545845032,
+      "learning_rate": 1.9101138664484593e-06,
+      "loss": 2.8963,
+      "step": 59015
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.734058141708374,
+      "learning_rate": 1.908379636898827e-06,
+      "loss": 2.9871,
+      "step": 59016
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7733717560768127,
+      "learning_rate": 1.9066461924651665e-06,
+      "loss": 2.7593,
+      "step": 59017
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7138202786445618,
+      "learning_rate": 1.9049135331520081e-06,
+      "loss": 2.9411,
+      "step": 59018
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7705768346786499,
+      "learning_rate": 1.9031816589639148e-06,
+      "loss": 2.9631,
+      "step": 59019
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7369682192802429,
+      "learning_rate": 1.9014505699054827e-06,
+      "loss": 2.8449,
+      "step": 59020
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7815206050872803,
+      "learning_rate": 1.8997202659812749e-06,
+      "loss": 2.7891,
+      "step": 59021
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7516481280326843,
+      "learning_rate": 1.8979907471958212e-06,
+      "loss": 2.9268,
+      "step": 59022
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8083335161209106,
+      "learning_rate": 1.8962620135537176e-06,
+      "loss": 3.0219,
+      "step": 59023
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7490785717964172,
+      "learning_rate": 1.894534065059461e-06,
+      "loss": 2.9312,
+      "step": 59024
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.6919969916343689,
+      "learning_rate": 1.8928069017176472e-06,
+      "loss": 2.8647,
+      "step": 59025
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7737188339233398,
+      "learning_rate": 1.8910805235328063e-06,
+      "loss": 2.7855,
+      "step": 59026
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7691394686698914,
+      "learning_rate": 1.889354930509468e-06,
+      "loss": 3.0189,
+      "step": 59027
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7691125273704529,
+      "learning_rate": 1.8876301226522284e-06,
+      "loss": 2.7889,
+      "step": 59028
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7242297530174255,
+      "learning_rate": 1.8859060999655507e-06,
+      "loss": 2.8643,
+      "step": 59029
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7567639946937561,
+      "learning_rate": 1.8841828624540645e-06,
+      "loss": 2.7662,
+      "step": 59030
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7616976499557495,
+      "learning_rate": 1.882460410122233e-06,
+      "loss": 2.828,
+      "step": 59031
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7366723418235779,
+      "learning_rate": 1.8807387429746524e-06,
+      "loss": 2.8366,
+      "step": 59032
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7073797583580017,
+      "learning_rate": 1.8790178610158525e-06,
+      "loss": 2.9282,
+      "step": 59033
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7266165614128113,
+      "learning_rate": 1.8772977642502963e-06,
+      "loss": 3.0461,
+      "step": 59034
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7223436832427979,
+      "learning_rate": 1.87557845268258e-06,
+      "loss": 2.7378,
+      "step": 59035
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8283471465110779,
+      "learning_rate": 1.8738599263172338e-06,
+      "loss": 2.844,
+      "step": 59036
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7933211326599121,
+      "learning_rate": 1.8721421851587537e-06,
+      "loss": 2.9484,
+      "step": 59037
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7600369453430176,
+      "learning_rate": 1.8704252292116696e-06,
+      "loss": 2.9175,
+      "step": 59038
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.788422703742981,
+      "learning_rate": 1.8687090584805108e-06,
+      "loss": 2.7721,
+      "step": 59039
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7650947570800781,
+      "learning_rate": 1.8669936729698076e-06,
+      "loss": 2.7601,
+      "step": 59040
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.748820960521698,
+      "learning_rate": 1.865279072684056e-06,
+      "loss": 2.8145,
+      "step": 59041
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7527621388435364,
+      "learning_rate": 1.8635652576277527e-06,
+      "loss": 2.9823,
+      "step": 59042
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7002708315849304,
+      "learning_rate": 1.8618522278054938e-06,
+      "loss": 2.8998,
+      "step": 59043
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7159339785575867,
+      "learning_rate": 1.860139983221709e-06,
+      "loss": 2.6089,
+      "step": 59044
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.737927258014679,
+      "learning_rate": 1.8584285238809281e-06,
+      "loss": 2.9601,
+      "step": 59045
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7862527370452881,
+      "learning_rate": 1.8567178497876811e-06,
+      "loss": 2.7792,
+      "step": 59046
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7486521005630493,
+      "learning_rate": 1.8550079609464974e-06,
+      "loss": 2.7664,
+      "step": 59047
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7823449969291687,
+      "learning_rate": 1.8532988573618068e-06,
+      "loss": 2.8471,
+      "step": 59048
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7296319603919983,
+      "learning_rate": 1.8515905390381392e-06,
+      "loss": 2.7969,
+      "step": 59049
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8135117888450623,
+      "learning_rate": 1.8498830059799908e-06,
+      "loss": 2.8586,
+      "step": 59050
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.6989290714263916,
+      "learning_rate": 1.8481762581918913e-06,
+      "loss": 2.6872,
+      "step": 59051
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7613128423690796,
+      "learning_rate": 1.8464702956783372e-06,
+      "loss": 2.995,
+      "step": 59052
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.721194863319397,
+      "learning_rate": 1.8447651184437584e-06,
+      "loss": 2.7919,
+      "step": 59053
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.726525068283081,
+      "learning_rate": 1.8430607264927178e-06,
+      "loss": 2.972,
+      "step": 59054
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7626053690910339,
+      "learning_rate": 1.8413571198296784e-06,
+      "loss": 2.8219,
+      "step": 59055
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7047833204269409,
+      "learning_rate": 1.8396542984591035e-06,
+      "loss": 2.7473,
+      "step": 59056
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7405434846878052,
+      "learning_rate": 1.8379522623854892e-06,
+      "loss": 2.7802,
+      "step": 59057
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7458202242851257,
+      "learning_rate": 1.8362510116133323e-06,
+      "loss": 3.0058,
+      "step": 59058
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7674630880355835,
+      "learning_rate": 1.8345505461471287e-06,
+      "loss": 2.9578,
+      "step": 59059
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7461906671524048,
+      "learning_rate": 1.8328508659913087e-06,
+      "loss": 2.8953,
+      "step": 59060
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.745392382144928,
+      "learning_rate": 1.831151971150402e-06,
+      "loss": 2.8253,
+      "step": 59061
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7358865141868591,
+      "learning_rate": 1.829453861628838e-06,
+      "loss": 2.9244,
+      "step": 59062
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7252581715583801,
+      "learning_rate": 1.8277565374311132e-06,
+      "loss": 3.0891,
+      "step": 59063
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7202037572860718,
+      "learning_rate": 1.8260599985616908e-06,
+      "loss": 2.5241,
+      "step": 59064
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7681000232696533,
+      "learning_rate": 1.824364245025034e-06,
+      "loss": 2.9659,
+      "step": 59065
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7250438928604126,
+      "learning_rate": 1.822669276825639e-06,
+      "loss": 2.7523,
+      "step": 59066
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7999386787414551,
+      "learning_rate": 1.8209750939679357e-06,
+      "loss": 2.7248,
+      "step": 59067
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8048508167266846,
+      "learning_rate": 1.8192816964563873e-06,
+      "loss": 2.6697,
+      "step": 59068
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7457453012466431,
+      "learning_rate": 1.81758908429549e-06,
+      "loss": 2.8159,
+      "step": 59069
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7637097239494324,
+      "learning_rate": 1.8158972574896403e-06,
+      "loss": 2.748,
+      "step": 59070
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7343689799308777,
+      "learning_rate": 1.814206216043368e-06,
+      "loss": 2.8942,
+      "step": 59071
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7599355578422546,
+      "learning_rate": 1.8125159599610694e-06,
+      "loss": 2.9799,
+      "step": 59072
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7494969964027405,
+      "learning_rate": 1.8108264892472412e-06,
+      "loss": 2.8893,
+      "step": 59073
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7399177551269531,
+      "learning_rate": 1.80913780390628e-06,
+      "loss": 2.8731,
+      "step": 59074
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7720394134521484,
+      "learning_rate": 1.8074499039426815e-06,
+      "loss": 2.7274,
+      "step": 59075
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7690413594245911,
+      "learning_rate": 1.8057627893608428e-06,
+      "loss": 2.7685,
+      "step": 59076
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7422636151313782,
+      "learning_rate": 1.8040764601652601e-06,
+      "loss": 2.884,
+      "step": 59077
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7094586491584778,
+      "learning_rate": 1.8023909163603634e-06,
+      "loss": 2.8035,
+      "step": 59078
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7637078166007996,
+      "learning_rate": 1.8007061579505488e-06,
+      "loss": 2.9295,
+      "step": 59079
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7474511861801147,
+      "learning_rate": 1.7990221849402797e-06,
+      "loss": 2.9187,
+      "step": 59080
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7161756157875061,
+      "learning_rate": 1.797338997334019e-06,
+      "loss": 2.7149,
+      "step": 59081
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7648656368255615,
+      "learning_rate": 1.7956565951361635e-06,
+      "loss": 2.979,
+      "step": 59082
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8284847736358643,
+      "learning_rate": 1.7939749783511758e-06,
+      "loss": 3.0267,
+      "step": 59083
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7770487070083618,
+      "learning_rate": 1.7922941469834528e-06,
+      "loss": 2.8505,
+      "step": 59084
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7366689443588257,
+      "learning_rate": 1.7906141010374576e-06,
+      "loss": 2.9935,
+      "step": 59085
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7799773812294006,
+      "learning_rate": 1.788934840517553e-06,
+      "loss": 2.8051,
+      "step": 59086
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7727737426757812,
+      "learning_rate": 1.7872563654282357e-06,
+      "loss": 3.0461,
+      "step": 59087
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7719646692276001,
+      "learning_rate": 1.7855786757738687e-06,
+      "loss": 2.8494,
+      "step": 59088
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7487027049064636,
+      "learning_rate": 1.7839017715589155e-06,
+      "loss": 2.9247,
+      "step": 59089
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7525886297225952,
+      "learning_rate": 1.7822256527877721e-06,
+      "loss": 2.9167,
+      "step": 59090
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7553564310073853,
+      "learning_rate": 1.7805503194648684e-06,
+      "loss": 2.7595,
+      "step": 59091
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8537721633911133,
+      "learning_rate": 1.778875771594568e-06,
+      "loss": 2.9898,
+      "step": 59092
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7500604391098022,
+      "learning_rate": 1.7772020091813333e-06,
+      "loss": 2.9763,
+      "step": 59093
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7806376814842224,
+      "learning_rate": 1.775529032229528e-06,
+      "loss": 2.7487,
+      "step": 59094
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7601209282875061,
+      "learning_rate": 1.7738568407436149e-06,
+      "loss": 2.9419,
+      "step": 59095
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7778134346008301,
+      "learning_rate": 1.7721854347279573e-06,
+      "loss": 2.723,
+      "step": 59096
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.759236216545105,
+      "learning_rate": 1.7705148141869853e-06,
+      "loss": 3.0486,
+      "step": 59097
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8189762234687805,
+      "learning_rate": 1.7688449791250614e-06,
+      "loss": 2.8238,
+      "step": 59098
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7347763776779175,
+      "learning_rate": 1.7671759295465826e-06,
+      "loss": 2.8377,
+      "step": 59099
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7606347799301147,
+      "learning_rate": 1.7655076654560119e-06,
+      "loss": 2.7916,
+      "step": 59100
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7474910020828247,
+      "learning_rate": 1.7638401868576791e-06,
+      "loss": 2.9414,
+      "step": 59101
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7264478206634521,
+      "learning_rate": 1.7621734937559806e-06,
+      "loss": 2.8281,
+      "step": 59102
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7830918431282043,
+      "learning_rate": 1.7605075861553464e-06,
+      "loss": 2.9221,
+      "step": 59103
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.78126060962677,
+      "learning_rate": 1.7588424640601062e-06,
+      "loss": 2.9048,
+      "step": 59104
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7438468933105469,
+      "learning_rate": 1.7571781274746898e-06,
+      "loss": 2.7915,
+      "step": 59105
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8499040007591248,
+      "learning_rate": 1.755514576403494e-06,
+      "loss": 2.8397,
+      "step": 59106
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7694287300109863,
+      "learning_rate": 1.7538518108508481e-06,
+      "loss": 2.8528,
+      "step": 59107
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7242769598960876,
+      "learning_rate": 1.752189830821149e-06,
+      "loss": 2.9493,
+      "step": 59108
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7706876397132874,
+      "learning_rate": 1.750528636318793e-06,
+      "loss": 2.7071,
+      "step": 59109
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7799972295761108,
+      "learning_rate": 1.7488682273481769e-06,
+      "loss": 2.9786,
+      "step": 59110
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7658340930938721,
+      "learning_rate": 1.7472086039135968e-06,
+      "loss": 2.991,
+      "step": 59111
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7521490454673767,
+      "learning_rate": 1.7455497660194828e-06,
+      "loss": 2.8154,
+      "step": 59112
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7272855639457703,
+      "learning_rate": 1.7438917136702312e-06,
+      "loss": 2.84,
+      "step": 59113
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7962722182273865,
+      "learning_rate": 1.7422344468701055e-06,
+      "loss": 3.0468,
+      "step": 59114
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.770363986492157,
+      "learning_rate": 1.7405779656235686e-06,
+      "loss": 2.8827,
+      "step": 59115
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7188537120819092,
+      "learning_rate": 1.7389222699349503e-06,
+      "loss": 2.8618,
+      "step": 59116
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7084076404571533,
+      "learning_rate": 1.7372673598085808e-06,
+      "loss": 2.8814,
+      "step": 59117
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7412644028663635,
+      "learning_rate": 1.7356132352488562e-06,
+      "loss": 2.8781,
+      "step": 59118
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7647395730018616,
+      "learning_rate": 1.733959896260173e-06,
+      "loss": 2.8254,
+      "step": 59119
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7721149921417236,
+      "learning_rate": 1.7323073428467948e-06,
+      "loss": 2.8524,
+      "step": 59120
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7480499148368835,
+      "learning_rate": 1.730655575013118e-06,
+      "loss": 2.8018,
+      "step": 59121
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7373523116111755,
+      "learning_rate": 1.7290045927634721e-06,
+      "loss": 2.8485,
+      "step": 59122
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7831674218177795,
+      "learning_rate": 1.727354396102254e-06,
+      "loss": 3.0131,
+      "step": 59123
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7761872410774231,
+      "learning_rate": 1.72570498503376e-06,
+      "loss": 2.8869,
+      "step": 59124
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7596009969711304,
+      "learning_rate": 1.7240563595623868e-06,
+      "loss": 2.6972,
+      "step": 59125
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.734153151512146,
+      "learning_rate": 1.7224085196923977e-06,
+      "loss": 2.7864,
+      "step": 59126
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7314983606338501,
+      "learning_rate": 1.720761465428222e-06,
+      "loss": 2.9053,
+      "step": 59127
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7744563817977905,
+      "learning_rate": 1.7191151967741568e-06,
+      "loss": 2.905,
+      "step": 59128
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7810060381889343,
+      "learning_rate": 1.7174697137344983e-06,
+      "loss": 2.9127,
+      "step": 59129
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7507235407829285,
+      "learning_rate": 1.7158250163136432e-06,
+      "loss": 2.8359,
+      "step": 59130
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7610126733779907,
+      "learning_rate": 1.7141811045158883e-06,
+      "loss": 2.6502,
+      "step": 59131
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7299541234970093,
+      "learning_rate": 1.712537978345596e-06,
+      "loss": 2.7764,
+      "step": 59132
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.805625319480896,
+      "learning_rate": 1.7108956378070637e-06,
+      "loss": 2.8168,
+      "step": 59133
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7253496646881104,
+      "learning_rate": 1.7092540829046208e-06,
+      "loss": 2.8982,
+      "step": 59134
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8161607384681702,
+      "learning_rate": 1.7076133136425974e-06,
+      "loss": 2.846,
+      "step": 59135
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7335495948791504,
+      "learning_rate": 1.7059733300253232e-06,
+      "loss": 2.9378,
+      "step": 59136
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7281621098518372,
+      "learning_rate": 1.7043341320570613e-06,
+      "loss": 2.8579,
+      "step": 59137
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7644967436790466,
+      "learning_rate": 1.702695719742242e-06,
+      "loss": 2.8762,
+      "step": 59138
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7520266771316528,
+      "learning_rate": 1.7010580930850614e-06,
+      "loss": 2.8353,
+      "step": 59139
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7870677709579468,
+      "learning_rate": 1.6994212520899164e-06,
+      "loss": 2.9208,
+      "step": 59140
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7785607576370239,
+      "learning_rate": 1.69778519676107e-06,
+      "loss": 3.0277,
+      "step": 59141
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7652883529663086,
+      "learning_rate": 1.6961499271028521e-06,
+      "loss": 2.9603,
+      "step": 59142
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7231030464172363,
+      "learning_rate": 1.6945154431195596e-06,
+      "loss": 2.5437,
+      "step": 59143
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7375117540359497,
+      "learning_rate": 1.6928817448155218e-06,
+      "loss": 2.6275,
+      "step": 59144
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7867879867553711,
+      "learning_rate": 1.6912488321950023e-06,
+      "loss": 2.8344,
+      "step": 59145
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8134239912033081,
+      "learning_rate": 1.689616705262331e-06,
+      "loss": 2.8047,
+      "step": 59146
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7559427618980408,
+      "learning_rate": 1.6879853640218044e-06,
+      "loss": 2.6349,
+      "step": 59147
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7431520223617554,
+      "learning_rate": 1.686354808477719e-06,
+      "loss": 2.9635,
+      "step": 59148
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7744369506835938,
+      "learning_rate": 1.684725038634338e-06,
+      "loss": 2.972,
+      "step": 59149
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.769906222820282,
+      "learning_rate": 1.6830960544959915e-06,
+      "loss": 2.7403,
+      "step": 59150
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.772413432598114,
+      "learning_rate": 1.681467856066976e-06,
+      "loss": 2.9986,
+      "step": 59151
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7344859838485718,
+      "learning_rate": 1.6798404433515545e-06,
+      "loss": 2.8915,
+      "step": 59152
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7600609660148621,
+      "learning_rate": 1.6782138163539903e-06,
+      "loss": 2.6796,
+      "step": 59153
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7428922653198242,
+      "learning_rate": 1.6765879750786138e-06,
+      "loss": 3.1002,
+      "step": 59154
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7847473621368408,
+      "learning_rate": 1.6749629195296877e-06,
+      "loss": 2.7684,
+      "step": 59155
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7154970169067383,
+      "learning_rate": 1.6733386497115086e-06,
+      "loss": 2.9555,
+      "step": 59156
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8147268295288086,
+      "learning_rate": 1.6717151656283734e-06,
+      "loss": 2.8499,
+      "step": 59157
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7564356923103333,
+      "learning_rate": 1.6700924672844783e-06,
+      "loss": 2.8353,
+      "step": 59158
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7662482261657715,
+      "learning_rate": 1.6684705546841537e-06,
+      "loss": 2.9374,
+      "step": 59159
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7386744022369385,
+      "learning_rate": 1.6668494278316624e-06,
+      "loss": 2.749,
+      "step": 59160
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7575583457946777,
+      "learning_rate": 1.665229086731268e-06,
+      "loss": 3.0314,
+      "step": 59161
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7565872073173523,
+      "learning_rate": 1.6636095313872666e-06,
+      "loss": 2.854,
+      "step": 59162
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7207484245300293,
+      "learning_rate": 1.6619907618038887e-06,
+      "loss": 2.8848,
+      "step": 59163
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7588822245597839,
+      "learning_rate": 1.660372777985397e-06,
+      "loss": 2.7857,
+      "step": 59164
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7523000836372375,
+      "learning_rate": 1.6587555799360885e-06,
+      "loss": 2.9389,
+      "step": 59165
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7837392687797546,
+      "learning_rate": 1.657139167660193e-06,
+      "loss": 2.7966,
+      "step": 59166
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7007728815078735,
+      "learning_rate": 1.655523541161974e-06,
+      "loss": 2.8785,
+      "step": 59167
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7474599480628967,
+      "learning_rate": 1.653908700445694e-06,
+      "loss": 2.8779,
+      "step": 59168
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8092578649520874,
+      "learning_rate": 1.6522946455155839e-06,
+      "loss": 2.7932,
+      "step": 59169
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7371110916137695,
+      "learning_rate": 1.6506813763759064e-06,
+      "loss": 2.8497,
+      "step": 59170
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7597226500511169,
+      "learning_rate": 1.6490688930309581e-06,
+      "loss": 2.9094,
+      "step": 59171
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7393887639045715,
+      "learning_rate": 1.6474571954849358e-06,
+      "loss": 2.8139,
+      "step": 59172
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7919648885726929,
+      "learning_rate": 1.6458462837420693e-06,
+      "loss": 2.8186,
+      "step": 59173
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7349705696105957,
+      "learning_rate": 1.6442361578066221e-06,
+      "loss": 2.7144,
+      "step": 59174
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7724583745002747,
+      "learning_rate": 1.6426268176828239e-06,
+      "loss": 2.7673,
+      "step": 59175
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7804393172264099,
+      "learning_rate": 1.641018263374938e-06,
+      "loss": 2.8634,
+      "step": 59176
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8103668689727783,
+      "learning_rate": 1.6394104948872277e-06,
+      "loss": 2.7951,
+      "step": 59177
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7489713430404663,
+      "learning_rate": 1.6378035122238565e-06,
+      "loss": 3.0172,
+      "step": 59178
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7497350573539734,
+      "learning_rate": 1.6361973153890873e-06,
+      "loss": 2.7531,
+      "step": 59179
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7449610829353333,
+      "learning_rate": 1.6345919043871835e-06,
+      "loss": 2.9164,
+      "step": 59180
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7574251294136047,
+      "learning_rate": 1.6329872792223087e-06,
+      "loss": 2.7664,
+      "step": 59181
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7449432611465454,
+      "learning_rate": 1.6313834398987258e-06,
+      "loss": 2.9572,
+      "step": 59182
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7493433356285095,
+      "learning_rate": 1.629780386420665e-06,
+      "loss": 2.8492,
+      "step": 59183
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.721378743648529,
+      "learning_rate": 1.6281781187923226e-06,
+      "loss": 2.8584,
+      "step": 59184
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.705966055393219,
+      "learning_rate": 1.6265766370179623e-06,
+      "loss": 2.7012,
+      "step": 59185
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7675456404685974,
+      "learning_rate": 1.6249759411017805e-06,
+      "loss": 2.7796,
+      "step": 59186
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7658663988113403,
+      "learning_rate": 1.623376031048007e-06,
+      "loss": 2.6216,
+      "step": 59187
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7195037603378296,
+      "learning_rate": 1.6217769068607722e-06,
+      "loss": 2.9578,
+      "step": 59188
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7412412762641907,
+      "learning_rate": 1.6201785685444058e-06,
+      "loss": 2.7394,
+      "step": 59189
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7388929724693298,
+      "learning_rate": 1.6185810161030378e-06,
+      "loss": 2.9017,
+      "step": 59190
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8015137910842896,
+      "learning_rate": 1.616984249540898e-06,
+      "loss": 2.7832,
+      "step": 59191
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.733881413936615,
+      "learning_rate": 1.6153882688622167e-06,
+      "loss": 2.8523,
+      "step": 59192
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7632375359535217,
+      "learning_rate": 1.6137930740711901e-06,
+      "loss": 2.7231,
+      "step": 59193
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7348890900611877,
+      "learning_rate": 1.6121986651719821e-06,
+      "loss": 2.8821,
+      "step": 59194
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.729511559009552,
+      "learning_rate": 1.610605042168822e-06,
+      "loss": 2.9414,
+      "step": 59195
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7740534543991089,
+      "learning_rate": 1.609012205065907e-06,
+      "loss": 2.7154,
+      "step": 59196
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7699583768844604,
+      "learning_rate": 1.6074201538674335e-06,
+      "loss": 2.8821,
+      "step": 59197
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7296841144561768,
+      "learning_rate": 1.605828888577565e-06,
+      "loss": 2.6342,
+      "step": 59198
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7484259009361267,
+      "learning_rate": 1.6042384092005312e-06,
+      "loss": 2.9587,
+      "step": 59199
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7556095719337463,
+      "learning_rate": 1.6026487157404954e-06,
+      "loss": 2.8836,
+      "step": 59200
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7266550064086914,
+      "learning_rate": 1.6010598082016545e-06,
+      "loss": 2.7881,
+      "step": 59201
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.745608925819397,
+      "learning_rate": 1.5994716865882384e-06,
+      "loss": 3.12,
+      "step": 59202
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7120503783226013,
+      "learning_rate": 1.5978843509043103e-06,
+      "loss": 3.1001,
+      "step": 59203
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7117788195610046,
+      "learning_rate": 1.5962978011541672e-06,
+      "loss": 3.027,
+      "step": 59204
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7518680691719055,
+      "learning_rate": 1.5947120373419719e-06,
+      "loss": 2.94,
+      "step": 59205
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.731158435344696,
+      "learning_rate": 1.5931270594718215e-06,
+      "loss": 2.7666,
+      "step": 59206
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7914332151412964,
+      "learning_rate": 1.5915428675479792e-06,
+      "loss": 2.8187,
+      "step": 59207
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7586977481842041,
+      "learning_rate": 1.5899594615746081e-06,
+      "loss": 2.8927,
+      "step": 59208
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7375707030296326,
+      "learning_rate": 1.588376841555805e-06,
+      "loss": 2.7874,
+      "step": 59209
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7598499655723572,
+      "learning_rate": 1.5867950074958001e-06,
+      "loss": 2.8557,
+      "step": 59210
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.759164571762085,
+      "learning_rate": 1.5852139593987566e-06,
+      "loss": 2.8819,
+      "step": 59211
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7925102710723877,
+      "learning_rate": 1.583633697268838e-06,
+      "loss": 2.7934,
+      "step": 59212
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7286711931228638,
+      "learning_rate": 1.582054221110174e-06,
+      "loss": 2.9716,
+      "step": 59213
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7402360439300537,
+      "learning_rate": 1.5804755309269612e-06,
+      "loss": 2.8322,
+      "step": 59214
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7518147230148315,
+      "learning_rate": 1.57889762672333e-06,
+      "loss": 2.8124,
+      "step": 59215
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7689666748046875,
+      "learning_rate": 1.5773205085034435e-06,
+      "loss": 2.7527,
+      "step": 59216
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.746090829372406,
+      "learning_rate": 1.5757441762714984e-06,
+      "loss": 2.9052,
+      "step": 59217
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7587335109710693,
+      "learning_rate": 1.5741686300315581e-06,
+      "loss": 2.7332,
+      "step": 59218
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7321277260780334,
+      "learning_rate": 1.5725938697878525e-06,
+      "loss": 3.0402,
+      "step": 59219
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.774876058101654,
+      "learning_rate": 1.5710198955445119e-06,
+      "loss": 2.9392,
+      "step": 59220
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7118234634399414,
+      "learning_rate": 1.5694467073056661e-06,
+      "loss": 2.7393,
+      "step": 59221
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8065067529678345,
+      "learning_rate": 1.5678743050754784e-06,
+      "loss": 2.8755,
+      "step": 59222
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.815037190914154,
+      "learning_rate": 1.5663026888580454e-06,
+      "loss": 2.7749,
+      "step": 59223
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7758423089981079,
+      "learning_rate": 1.564731858657531e-06,
+      "loss": 2.9076,
+      "step": 59224
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7859812378883362,
+      "learning_rate": 1.5631618144780977e-06,
+      "loss": 3.0618,
+      "step": 59225
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7478824257850647,
+      "learning_rate": 1.5615925563238763e-06,
+      "loss": 2.906,
+      "step": 59226
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7796872854232788,
+      "learning_rate": 1.5600240841989632e-06,
+      "loss": 2.8319,
+      "step": 59227
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7634370923042297,
+      "learning_rate": 1.5584563981075215e-06,
+      "loss": 2.7921,
+      "step": 59228
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7156879901885986,
+      "learning_rate": 1.5568894980536484e-06,
+      "loss": 2.8655,
+      "step": 59229
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7343173027038574,
+      "learning_rate": 1.5553233840415068e-06,
+      "loss": 3.1111,
+      "step": 59230
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7309796810150146,
+      "learning_rate": 1.5537580560751938e-06,
+      "loss": 2.6718,
+      "step": 59231
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8148305416107178,
+      "learning_rate": 1.5521935141588726e-06,
+      "loss": 2.8729,
+      "step": 59232
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7874312400817871,
+      "learning_rate": 1.5506297582966066e-06,
+      "loss": 2.889,
+      "step": 59233
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7957163453102112,
+      "learning_rate": 1.549066788492559e-06,
+      "loss": 2.9847,
+      "step": 59234
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7534181475639343,
+      "learning_rate": 1.5475046047508265e-06,
+      "loss": 2.8653,
+      "step": 59235
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7766826748847961,
+      "learning_rate": 1.5459432070755395e-06,
+      "loss": 2.8438,
+      "step": 59236
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7522812485694885,
+      "learning_rate": 1.5443825954707945e-06,
+      "loss": 2.7315,
+      "step": 59237
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7763208746910095,
+      "learning_rate": 1.5428227699407213e-06,
+      "loss": 2.9692,
+      "step": 59238
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.773130476474762,
+      "learning_rate": 1.5412637304893838e-06,
+      "loss": 2.821,
+      "step": 59239
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7237739562988281,
+      "learning_rate": 1.539705477120945e-06,
+      "loss": 2.8297,
+      "step": 59240
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7151420712471008,
+      "learning_rate": 1.538148009839435e-06,
+      "loss": 2.758,
+      "step": 59241
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7747194170951843,
+      "learning_rate": 1.5365913286490506e-06,
+      "loss": 2.7573,
+      "step": 59242
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7364190220832825,
+      "learning_rate": 1.5350354335538217e-06,
+      "loss": 3.0683,
+      "step": 59243
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.6902226209640503,
+      "learning_rate": 1.533480324557912e-06,
+      "loss": 2.7821,
+      "step": 59244
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7629439234733582,
+      "learning_rate": 1.5319260016653178e-06,
+      "loss": 2.8934,
+      "step": 59245
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7819312214851379,
+      "learning_rate": 1.530372464880203e-06,
+      "loss": 2.9913,
+      "step": 59246
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7529870867729187,
+      "learning_rate": 1.5288197142066638e-06,
+      "loss": 2.7037,
+      "step": 59247
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7887903451919556,
+      "learning_rate": 1.5272677496487307e-06,
+      "loss": 2.9547,
+      "step": 59248
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7321487665176392,
+      "learning_rate": 1.5257165712106e-06,
+      "loss": 2.9963,
+      "step": 59249
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7392261028289795,
+      "learning_rate": 1.5241661788962357e-06,
+      "loss": 2.9604,
+      "step": 59250
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7557159662246704,
+      "learning_rate": 1.5226165727098005e-06,
+      "loss": 2.964,
+      "step": 59251
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7541767358779907,
+      "learning_rate": 1.521067752655325e-06,
+      "loss": 3.0179,
+      "step": 59252
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7616634368896484,
+      "learning_rate": 1.519519718736939e-06,
+      "loss": 2.7647,
+      "step": 59253
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7685297727584839,
+      "learning_rate": 1.5179724709586727e-06,
+      "loss": 2.9063,
+      "step": 59254
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7325605154037476,
+      "learning_rate": 1.5164260093246228e-06,
+      "loss": 3.0109,
+      "step": 59255
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7943716645240784,
+      "learning_rate": 1.514880333838886e-06,
+      "loss": 2.7804,
+      "step": 59256
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7822527885437012,
+      "learning_rate": 1.5133354445054923e-06,
+      "loss": 3.128,
+      "step": 59257
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7548781037330627,
+      "learning_rate": 1.5117913413285388e-06,
+      "loss": 2.8991,
+      "step": 59258
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.769331157207489,
+      "learning_rate": 1.5102480243120885e-06,
+      "loss": 3.0113,
+      "step": 59259
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.772138237953186,
+      "learning_rate": 1.5087054934601716e-06,
+      "loss": 2.8884,
+      "step": 59260
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7810102105140686,
+      "learning_rate": 1.507163748776885e-06,
+      "loss": 2.8167,
+      "step": 59261
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7660080194473267,
+      "learning_rate": 1.505622790266292e-06,
+      "loss": 2.8325,
+      "step": 59262
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7257230281829834,
+      "learning_rate": 1.504082617932456e-06,
+      "loss": 3.0289,
+      "step": 59263
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7466867566108704,
+      "learning_rate": 1.5025432317794073e-06,
+      "loss": 2.8704,
+      "step": 59264
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7224690914154053,
+      "learning_rate": 1.501004631811209e-06,
+      "loss": 2.7124,
+      "step": 59265
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7552944421768188,
+      "learning_rate": 1.4994668180319248e-06,
+      "loss": 3.1567,
+      "step": 59266
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8041907548904419,
+      "learning_rate": 1.4979297904455844e-06,
+      "loss": 2.9041,
+      "step": 59267
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7628178596496582,
+      "learning_rate": 1.4963935490562516e-06,
+      "loss": 2.8904,
+      "step": 59268
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8382160663604736,
+      "learning_rate": 1.49485809386799e-06,
+      "loss": 2.7538,
+      "step": 59269
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7363545298576355,
+      "learning_rate": 1.4933234248847958e-06,
+      "loss": 2.8377,
+      "step": 59270
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7144303321838379,
+      "learning_rate": 1.491789542110766e-06,
+      "loss": 2.7339,
+      "step": 59271
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8086472749710083,
+      "learning_rate": 1.4902564455498978e-06,
+      "loss": 2.9529,
+      "step": 59272
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7652781009674072,
+      "learning_rate": 1.488724135206254e-06,
+      "loss": 2.8901,
+      "step": 59273
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7442313432693481,
+      "learning_rate": 1.487192611083865e-06,
+      "loss": 3.0013,
+      "step": 59274
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7640436887741089,
+      "learning_rate": 1.4856618731867277e-06,
+      "loss": 2.8635,
+      "step": 59275
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7747520208358765,
+      "learning_rate": 1.4841319215189385e-06,
+      "loss": 2.9427,
+      "step": 59276
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7614535689353943,
+      "learning_rate": 1.482602756084461e-06,
+      "loss": 2.713,
+      "step": 59277
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7723492980003357,
+      "learning_rate": 1.4810743768873923e-06,
+      "loss": 2.9073,
+      "step": 59278
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7300555109977722,
+      "learning_rate": 1.4795467839317288e-06,
+      "loss": 2.9138,
+      "step": 59279
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7472530603408813,
+      "learning_rate": 1.4780199772214674e-06,
+      "loss": 2.831,
+      "step": 59280
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7535228729248047,
+      "learning_rate": 1.4764939567606382e-06,
+      "loss": 2.9556,
+      "step": 59281
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7726250886917114,
+      "learning_rate": 1.4749687225533047e-06,
+      "loss": 2.7113,
+      "step": 59282
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7266640663146973,
+      "learning_rate": 1.4734442746034303e-06,
+      "loss": 2.8155,
+      "step": 59283
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7508488297462463,
+      "learning_rate": 1.471920612915045e-06,
+      "loss": 2.8604,
+      "step": 59284
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.777144193649292,
+      "learning_rate": 1.4703977374921794e-06,
+      "loss": 2.6361,
+      "step": 59285
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7734659910202026,
+      "learning_rate": 1.4688756483388297e-06,
+      "loss": 2.8084,
+      "step": 59286
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7497792840003967,
+      "learning_rate": 1.4673543454590264e-06,
+      "loss": 2.8345,
+      "step": 59287
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.753085732460022,
+      "learning_rate": 1.4658338288567329e-06,
+      "loss": 2.9196,
+      "step": 59288
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7959982752799988,
+      "learning_rate": 1.4643140985359792e-06,
+      "loss": 2.8826,
+      "step": 59289
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7402199506759644,
+      "learning_rate": 1.462795154500762e-06,
+      "loss": 2.8387,
+      "step": 59290
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7226002216339111,
+      "learning_rate": 1.461276996755112e-06,
+      "loss": 3.0343,
+      "step": 59291
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7589437365531921,
+      "learning_rate": 1.459759625302992e-06,
+      "loss": 2.9562,
+      "step": 59292
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7250106930732727,
+      "learning_rate": 1.458243040148399e-06,
+      "loss": 2.8674,
+      "step": 59293
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7297521233558655,
+      "learning_rate": 1.4567272412953633e-06,
+      "loss": 3.1333,
+      "step": 59294
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.729756236076355,
+      "learning_rate": 1.4552122287478485e-06,
+      "loss": 2.7591,
+      "step": 59295
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7715573906898499,
+      "learning_rate": 1.4536980025098178e-06,
+      "loss": 3.0325,
+      "step": 59296
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7817413806915283,
+      "learning_rate": 1.4521845625853347e-06,
+      "loss": 3.0629,
+      "step": 59297
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7538647651672363,
+      "learning_rate": 1.4506719089782958e-06,
+      "loss": 2.9156,
+      "step": 59298
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7621957063674927,
+      "learning_rate": 1.4491600416927652e-06,
+      "loss": 3.0193,
+      "step": 59299
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7765817642211914,
+      "learning_rate": 1.4476489607326725e-06,
+      "loss": 3.0129,
+      "step": 59300
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7483084797859192,
+      "learning_rate": 1.4461386661020146e-06,
+      "loss": 2.8993,
+      "step": 59301
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7720745801925659,
+      "learning_rate": 1.4446291578047886e-06,
+      "loss": 3.0589,
+      "step": 59302
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7810670137405396,
+      "learning_rate": 1.4431204358449244e-06,
+      "loss": 3.0455,
+      "step": 59303
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7453815937042236,
+      "learning_rate": 1.441612500226419e-06,
+      "loss": 2.8832,
+      "step": 59304
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7267022132873535,
+      "learning_rate": 1.440105350953269e-06,
+      "loss": 2.8214,
+      "step": 59305
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7542246580123901,
+      "learning_rate": 1.438598988029438e-06,
+      "loss": 2.7414,
+      "step": 59306
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7183325290679932,
+      "learning_rate": 1.4370934114588228e-06,
+      "loss": 2.7841,
+      "step": 59307
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.9062812924385071,
+      "learning_rate": 1.435588621245487e-06,
+      "loss": 2.7707,
+      "step": 59308
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7738885283470154,
+      "learning_rate": 1.4340846173933606e-06,
+      "loss": 2.9159,
+      "step": 59309
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7519917488098145,
+      "learning_rate": 1.432581399906374e-06,
+      "loss": 3.0687,
+      "step": 59310
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7735452055931091,
+      "learning_rate": 1.4310789687885238e-06,
+      "loss": 2.6822,
+      "step": 59311
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.767966091632843,
+      "learning_rate": 1.4295773240437068e-06,
+      "loss": 2.9608,
+      "step": 59312
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7551448941230774,
+      "learning_rate": 1.4280764656759535e-06,
+      "loss": 2.9766,
+      "step": 59313
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.785728394985199,
+      "learning_rate": 1.426576393689194e-06,
+      "loss": 3.004,
+      "step": 59314
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7843324542045593,
+      "learning_rate": 1.425077108087358e-06,
+      "loss": 2.6908,
+      "step": 59315
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.717927098274231,
+      "learning_rate": 1.4235786088744094e-06,
+      "loss": 3.0576,
+      "step": 59316
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7239921689033508,
+      "learning_rate": 1.4220808960542784e-06,
+      "loss": 3.0828,
+      "step": 59317
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7748315334320068,
+      "learning_rate": 1.4205839696309618e-06,
+      "loss": 2.8963,
+      "step": 59318
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7714270353317261,
+      "learning_rate": 1.419087829608323e-06,
+      "loss": 2.9912,
+      "step": 59319
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7787078619003296,
+      "learning_rate": 1.4175924759903589e-06,
+      "loss": 2.7653,
+      "step": 59320
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7504279613494873,
+      "learning_rate": 1.4160979087809999e-06,
+      "loss": 2.7674,
+      "step": 59321
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7364410161972046,
+      "learning_rate": 1.4146041279841759e-06,
+      "loss": 3.0725,
+      "step": 59322
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.757007896900177,
+      "learning_rate": 1.413111133603817e-06,
+      "loss": 2.8924,
+      "step": 59323
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7655906677246094,
+      "learning_rate": 1.4116189256438538e-06,
+      "loss": 2.8821,
+      "step": 59324
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7602041363716125,
+      "learning_rate": 1.410127504108216e-06,
+      "loss": 2.6149,
+      "step": 59325
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7541115283966064,
+      "learning_rate": 1.4086368690008675e-06,
+      "loss": 2.7603,
+      "step": 59326
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7562124133110046,
+      "learning_rate": 1.4071470203256718e-06,
+      "loss": 2.9612,
+      "step": 59327
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7242172956466675,
+      "learning_rate": 1.4056579580865923e-06,
+      "loss": 2.708,
+      "step": 59328
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7529643774032593,
+      "learning_rate": 1.4041696822875592e-06,
+      "loss": 2.7851,
+      "step": 59329
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6918560266494751,
+      "learning_rate": 1.4026821929324695e-06,
+      "loss": 2.9248,
+      "step": 59330
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7533132433891296,
+      "learning_rate": 1.4011954900252532e-06,
+      "loss": 2.9044,
+      "step": 59331
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7325230240821838,
+      "learning_rate": 1.3997095735698071e-06,
+      "loss": 2.7677,
+      "step": 59332
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6999819278717041,
+      "learning_rate": 1.3982244435700617e-06,
+      "loss": 2.7991,
+      "step": 59333
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7240120768547058,
+      "learning_rate": 1.3967401000299472e-06,
+      "loss": 2.9905,
+      "step": 59334
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7363443374633789,
+      "learning_rate": 1.3952565429533268e-06,
+      "loss": 2.7389,
+      "step": 59335
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7331762909889221,
+      "learning_rate": 1.3937737723441644e-06,
+      "loss": 2.6583,
+      "step": 59336
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7030410766601562,
+      "learning_rate": 1.39229178820629e-06,
+      "loss": 2.8128,
+      "step": 59337
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7303512692451477,
+      "learning_rate": 1.3908105905436672e-06,
+      "loss": 2.8716,
+      "step": 59338
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7628925442695618,
+      "learning_rate": 1.389330179360193e-06,
+      "loss": 2.9677,
+      "step": 59339
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7441055178642273,
+      "learning_rate": 1.387850554659764e-06,
+      "loss": 2.8874,
+      "step": 59340
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7332181334495544,
+      "learning_rate": 1.386371716446244e-06,
+      "loss": 2.8901,
+      "step": 59341
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7477689385414124,
+      "learning_rate": 1.384893664723563e-06,
+      "loss": 2.9624,
+      "step": 59342
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7767438888549805,
+      "learning_rate": 1.3834163994955848e-06,
+      "loss": 2.825,
+      "step": 59343
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7393341660499573,
+      "learning_rate": 1.3819399207662396e-06,
+      "loss": 2.8816,
+      "step": 59344
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7812148928642273,
+      "learning_rate": 1.3804642285393908e-06,
+      "loss": 2.8379,
+      "step": 59345
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7248178124427795,
+      "learning_rate": 1.3789893228189352e-06,
+      "loss": 3.0399,
+      "step": 59346
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7927574515342712,
+      "learning_rate": 1.3775152036087367e-06,
+      "loss": 3.02,
+      "step": 59347
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7457157969474792,
+      "learning_rate": 1.3760418709126919e-06,
+      "loss": 3.0966,
+      "step": 59348
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6921155452728271,
+      "learning_rate": 1.3745693247346978e-06,
+      "loss": 2.9358,
+      "step": 59349
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7026863098144531,
+      "learning_rate": 1.3730975650785847e-06,
+      "loss": 2.6633,
+      "step": 59350
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7563304901123047,
+      "learning_rate": 1.3716265919482827e-06,
+      "loss": 2.8672,
+      "step": 59351
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7299355268478394,
+      "learning_rate": 1.370156405347622e-06,
+      "loss": 2.9305,
+      "step": 59352
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7180936336517334,
+      "learning_rate": 1.3686870052805332e-06,
+      "loss": 3.0106,
+      "step": 59353
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8197305798530579,
+      "learning_rate": 1.3672183917508461e-06,
+      "loss": 2.9598,
+      "step": 59354
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7417368292808533,
+      "learning_rate": 1.3657505647624245e-06,
+      "loss": 2.879,
+      "step": 59355
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7476125955581665,
+      "learning_rate": 1.364283524319132e-06,
+      "loss": 2.804,
+      "step": 59356
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7566918134689331,
+      "learning_rate": 1.362817270424832e-06,
+      "loss": 2.9898,
+      "step": 59357
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7430882453918457,
+      "learning_rate": 1.3613518030834215e-06,
+      "loss": 2.7312,
+      "step": 59358
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7406376004219055,
+      "learning_rate": 1.3598871222987307e-06,
+      "loss": 2.9305,
+      "step": 59359
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7154626846313477,
+      "learning_rate": 1.3584232280746231e-06,
+      "loss": 2.8426,
+      "step": 59360
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7587353587150574,
+      "learning_rate": 1.3569601204149628e-06,
+      "loss": 2.8959,
+      "step": 59361
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7413969039916992,
+      "learning_rate": 1.3554977993235794e-06,
+      "loss": 2.8985,
+      "step": 59362
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6970677971839905,
+      "learning_rate": 1.354036264804337e-06,
+      "loss": 2.7983,
+      "step": 59363
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7779316902160645,
+      "learning_rate": 1.3525755168610985e-06,
+      "loss": 2.8532,
+      "step": 59364
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7617775201797485,
+      "learning_rate": 1.3511155554976948e-06,
+      "loss": 2.9153,
+      "step": 59365
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7746046781539917,
+      "learning_rate": 1.3496563807179895e-06,
+      "loss": 3.0756,
+      "step": 59366
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7900286316871643,
+      "learning_rate": 1.3481979925257792e-06,
+      "loss": 2.6014,
+      "step": 59367
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7511681914329529,
+      "learning_rate": 1.3467403909249608e-06,
+      "loss": 2.8521,
+      "step": 59368
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7791048884391785,
+      "learning_rate": 1.3452835759193648e-06,
+      "loss": 2.9309,
+      "step": 59369
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7215959429740906,
+      "learning_rate": 1.343827547512788e-06,
+      "loss": 3.0088,
+      "step": 59370
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7659755349159241,
+      "learning_rate": 1.3423723057091274e-06,
+      "loss": 2.7044,
+      "step": 59371
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7261947393417358,
+      "learning_rate": 1.3409178505121465e-06,
+      "loss": 2.7411,
+      "step": 59372
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7190525531768799,
+      "learning_rate": 1.339464181925709e-06,
+      "loss": 2.938,
+      "step": 59373
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7865150570869446,
+      "learning_rate": 1.3380112999536786e-06,
+      "loss": 3.0685,
+      "step": 59374
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7611143589019775,
+      "learning_rate": 1.3365592045998186e-06,
+      "loss": 2.7982,
+      "step": 59375
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.758155107498169,
+      "learning_rate": 1.3351078958679929e-06,
+      "loss": 2.8102,
+      "step": 59376
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7684991955757141,
+      "learning_rate": 1.3336573737619981e-06,
+      "loss": 2.8202,
+      "step": 59377
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.72589510679245,
+      "learning_rate": 1.3322076382856984e-06,
+      "loss": 2.9137,
+      "step": 59378
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7442227005958557,
+      "learning_rate": 1.3307586894428568e-06,
+      "loss": 2.8947,
+      "step": 59379
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8001748919487,
+      "learning_rate": 1.3293105272373372e-06,
+      "loss": 2.8781,
+      "step": 59380
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.77564537525177,
+      "learning_rate": 1.3278631516729366e-06,
+      "loss": 2.9313,
+      "step": 59381
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7333300709724426,
+      "learning_rate": 1.3264165627534518e-06,
+      "loss": 2.889,
+      "step": 59382
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7711548805236816,
+      "learning_rate": 1.3249707604827131e-06,
+      "loss": 2.8388,
+      "step": 59383
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7647919058799744,
+      "learning_rate": 1.3235257448644843e-06,
+      "loss": 2.7657,
+      "step": 59384
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7608219981193542,
+      "learning_rate": 1.3220815159026288e-06,
+      "loss": 2.745,
+      "step": 59385
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7038655877113342,
+      "learning_rate": 1.3206380736009437e-06,
+      "loss": 2.6167,
+      "step": 59386
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7791205048561096,
+      "learning_rate": 1.3191954179632258e-06,
+      "loss": 3.025,
+      "step": 59387
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7197181582450867,
+      "learning_rate": 1.3177535489932388e-06,
+      "loss": 3.1071,
+      "step": 59388
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7548245191574097,
+      "learning_rate": 1.3163124666948132e-06,
+      "loss": 2.7108,
+      "step": 59389
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7533859014511108,
+      "learning_rate": 1.3148721710717459e-06,
+      "loss": 3.086,
+      "step": 59390
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7900393009185791,
+      "learning_rate": 1.3134326621278002e-06,
+      "loss": 2.8453,
+      "step": 59391
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7366978526115417,
+      "learning_rate": 1.3119939398668068e-06,
+      "loss": 3.0731,
+      "step": 59392
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7741217613220215,
+      "learning_rate": 1.3105560042925623e-06,
+      "loss": 2.9944,
+      "step": 59393
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7335111498832703,
+      "learning_rate": 1.3091188554087971e-06,
+      "loss": 2.7327,
+      "step": 59394
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7100944519042969,
+      "learning_rate": 1.3076824932193086e-06,
+      "loss": 2.8545,
+      "step": 59395
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7313215136528015,
+      "learning_rate": 1.3062469177279267e-06,
+      "loss": 2.8571,
+      "step": 59396
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7661868929862976,
+      "learning_rate": 1.304812128938415e-06,
+      "loss": 3.0734,
+      "step": 59397
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7975308895111084,
+      "learning_rate": 1.3033781268545374e-06,
+      "loss": 2.8728,
+      "step": 59398
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7336994409561157,
+      "learning_rate": 1.3019449114800573e-06,
+      "loss": 3.1197,
+      "step": 59399
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7279320359230042,
+      "learning_rate": 1.3005124828188052e-06,
+      "loss": 2.7903,
+      "step": 59400
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7265657186508179,
+      "learning_rate": 1.299080840874478e-06,
+      "loss": 3.0084,
+      "step": 59401
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7682440876960754,
+      "learning_rate": 1.2976499856509059e-06,
+      "loss": 2.8712,
+      "step": 59402
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7393474578857422,
+      "learning_rate": 1.2962199171518528e-06,
+      "loss": 2.8719,
+      "step": 59403
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7249324917793274,
+      "learning_rate": 1.2947906353810489e-06,
+      "loss": 2.9401,
+      "step": 59404
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7434907555580139,
+      "learning_rate": 1.293362140342258e-06,
+      "loss": 2.9189,
+      "step": 59405
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.766069233417511,
+      "learning_rate": 1.29193443203931e-06,
+      "loss": 3.0494,
+      "step": 59406
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7799139618873596,
+      "learning_rate": 1.2905075104759022e-06,
+      "loss": 2.7501,
+      "step": 59407
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7259254455566406,
+      "learning_rate": 1.2890813756557983e-06,
+      "loss": 2.9274,
+      "step": 59408
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8382498621940613,
+      "learning_rate": 1.2876560275827619e-06,
+      "loss": 3.0196,
+      "step": 59409
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7565951943397522,
+      "learning_rate": 1.2862314662605565e-06,
+      "loss": 3.0605,
+      "step": 59410
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7980879545211792,
+      "learning_rate": 1.2848076916929462e-06,
+      "loss": 2.7056,
+      "step": 59411
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8130079507827759,
+      "learning_rate": 1.283384703883661e-06,
+      "loss": 2.6706,
+      "step": 59412
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7338296175003052,
+      "learning_rate": 1.2819625028364312e-06,
+      "loss": 2.8004,
+      "step": 59413
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7697415351867676,
+      "learning_rate": 1.2805410885550205e-06,
+      "loss": 2.8974,
+      "step": 59414
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7453495860099792,
+      "learning_rate": 1.2791204610431927e-06,
+      "loss": 3.0127,
+      "step": 59415
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7462559342384338,
+      "learning_rate": 1.277700620304678e-06,
+      "loss": 2.7122,
+      "step": 59416
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7518516778945923,
+      "learning_rate": 1.2762815663431735e-06,
+      "loss": 2.7392,
+      "step": 59417
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7591348886489868,
+      "learning_rate": 1.2748632991625096e-06,
+      "loss": 3.0127,
+      "step": 59418
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7799931764602661,
+      "learning_rate": 1.2734458187663166e-06,
+      "loss": 2.912,
+      "step": 59419
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7421641945838928,
+      "learning_rate": 1.2720291251583913e-06,
+      "loss": 2.6909,
+      "step": 59420
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7436515092849731,
+      "learning_rate": 1.2706132183424644e-06,
+      "loss": 2.7775,
+      "step": 59421
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7163355350494385,
+      "learning_rate": 1.2691980983222327e-06,
+      "loss": 2.8094,
+      "step": 59422
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7329570651054382,
+      "learning_rate": 1.26778376510146e-06,
+      "loss": 2.9062,
+      "step": 59423
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8021432161331177,
+      "learning_rate": 1.26637021868381e-06,
+      "loss": 3.0601,
+      "step": 59424
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7619763016700745,
+      "learning_rate": 1.2649574590730794e-06,
+      "loss": 2.9556,
+      "step": 59425
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7866322994232178,
+      "learning_rate": 1.2635454862729656e-06,
+      "loss": 2.7857,
+      "step": 59426
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7534366250038147,
+      "learning_rate": 1.2621343002871654e-06,
+      "loss": 2.8646,
+      "step": 59427
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7509728074073792,
+      "learning_rate": 1.2607239011194092e-06,
+      "loss": 2.8734,
+      "step": 59428
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7655785083770752,
+      "learning_rate": 1.2593142887734276e-06,
+      "loss": 2.8618,
+      "step": 59429
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.759252667427063,
+      "learning_rate": 1.2579054632529173e-06,
+      "loss": 3.0101,
+      "step": 59430
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7218188047409058,
+      "learning_rate": 1.2564974245615756e-06,
+      "loss": 2.8201,
+      "step": 59431
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7249210476875305,
+      "learning_rate": 1.2550901727031326e-06,
+      "loss": 2.7672,
+      "step": 59432
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7824299335479736,
+      "learning_rate": 1.2536837076812856e-06,
+      "loss": 2.8039,
+      "step": 59433
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.731982946395874,
+      "learning_rate": 1.2522780294997315e-06,
+      "loss": 2.997,
+      "step": 59434
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7653811573982239,
+      "learning_rate": 1.2508731381622006e-06,
+      "loss": 3.0187,
+      "step": 59435
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7498577833175659,
+      "learning_rate": 1.2494690336723568e-06,
+      "loss": 2.7194,
+      "step": 59436
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7105596661567688,
+      "learning_rate": 1.2480657160339303e-06,
+      "loss": 2.8219,
+      "step": 59437
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7923392653465271,
+      "learning_rate": 1.2466631852506182e-06,
+      "loss": 2.8113,
+      "step": 59438
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.725375235080719,
+      "learning_rate": 1.2452614413260509e-06,
+      "loss": 2.9244,
+      "step": 59439
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7582253217697144,
+      "learning_rate": 1.2438604842640253e-06,
+      "loss": 2.7547,
+      "step": 59440
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7434356808662415,
+      "learning_rate": 1.242460314068139e-06,
+      "loss": 2.8055,
+      "step": 59441
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7622381448745728,
+      "learning_rate": 1.2410609307421216e-06,
+      "loss": 2.8985,
+      "step": 59442
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8195493817329407,
+      "learning_rate": 1.2396623342896373e-06,
+      "loss": 2.8076,
+      "step": 59443
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7499621510505676,
+      "learning_rate": 1.2382645247144163e-06,
+      "loss": 2.7527,
+      "step": 59444
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7483765482902527,
+      "learning_rate": 1.2368675020200891e-06,
+      "loss": 2.9304,
+      "step": 59445
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7556241154670715,
+      "learning_rate": 1.2354712662103528e-06,
+      "loss": 2.9852,
+      "step": 59446
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7922104597091675,
+      "learning_rate": 1.2340758172889043e-06,
+      "loss": 2.9386,
+      "step": 59447
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7229467034339905,
+      "learning_rate": 1.232681155259374e-06,
+      "loss": 2.897,
+      "step": 59448
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7952060699462891,
+      "learning_rate": 1.2312872801254926e-06,
+      "loss": 3.0477,
+      "step": 59449
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8594965934753418,
+      "learning_rate": 1.2298941918908566e-06,
+      "loss": 2.9014,
+      "step": 59450
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7167303562164307,
+      "learning_rate": 1.2285018905592303e-06,
+      "loss": 2.8262,
+      "step": 59451
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7530221343040466,
+      "learning_rate": 1.2271103761342105e-06,
+      "loss": 2.9817,
+      "step": 59452
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7742487192153931,
+      "learning_rate": 1.225719648619461e-06,
+      "loss": 2.8349,
+      "step": 59453
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7406141757965088,
+      "learning_rate": 1.2243297080186786e-06,
+      "loss": 2.9665,
+      "step": 59454
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7695170044898987,
+      "learning_rate": 1.2229405543355274e-06,
+      "loss": 2.9037,
+      "step": 59455
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7280238270759583,
+      "learning_rate": 1.221552187573671e-06,
+      "loss": 2.9269,
+      "step": 59456
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.730404257774353,
+      "learning_rate": 1.2201646077367067e-06,
+      "loss": 2.8521,
+      "step": 59457
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7469308376312256,
+      "learning_rate": 1.2187778148283312e-06,
+      "loss": 2.7984,
+      "step": 59458
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7723640203475952,
+      "learning_rate": 1.217391808852175e-06,
+      "loss": 2.9174,
+      "step": 59459
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8354815244674683,
+      "learning_rate": 1.2160065898119352e-06,
+      "loss": 2.7031,
+      "step": 59460
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7310277819633484,
+      "learning_rate": 1.2146221577112426e-06,
+      "loss": 2.8573,
+      "step": 59461
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.761574923992157,
+      "learning_rate": 1.2132385125537269e-06,
+      "loss": 2.947,
+      "step": 59462
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8336168527603149,
+      "learning_rate": 1.2118556543430192e-06,
+      "loss": 2.9162,
+      "step": 59463
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7921255230903625,
+      "learning_rate": 1.2104735830827827e-06,
+      "loss": 2.8752,
+      "step": 59464
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7641667723655701,
+      "learning_rate": 1.2090922987766816e-06,
+      "loss": 2.8504,
+      "step": 59465
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7452090382575989,
+      "learning_rate": 1.2077118014282794e-06,
+      "loss": 2.8647,
+      "step": 59466
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.751675546169281,
+      "learning_rate": 1.2063320910413067e-06,
+      "loss": 2.8153,
+      "step": 59467
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7424731254577637,
+      "learning_rate": 1.204953167619327e-06,
+      "loss": 3.0888,
+      "step": 59468
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8056619167327881,
+      "learning_rate": 1.2035750311660042e-06,
+      "loss": 2.851,
+      "step": 59469
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8469576239585876,
+      "learning_rate": 1.2021976816849687e-06,
+      "loss": 2.8366,
+      "step": 59470
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7873582243919373,
+      "learning_rate": 1.2008211191798177e-06,
+      "loss": 2.9648,
+      "step": 59471
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7305603623390198,
+      "learning_rate": 1.1994453436542152e-06,
+      "loss": 2.845,
+      "step": 59472
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7844927906990051,
+      "learning_rate": 1.198070355111791e-06,
+      "loss": 2.9788,
+      "step": 59473
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7984238862991333,
+      "learning_rate": 1.1966961535561094e-06,
+      "loss": 2.9127,
+      "step": 59474
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7224608659744263,
+      "learning_rate": 1.1953227389908339e-06,
+      "loss": 2.9176,
+      "step": 59475
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7413891553878784,
+      "learning_rate": 1.1939501114195616e-06,
+      "loss": 2.6806,
+      "step": 59476
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8167316317558289,
+      "learning_rate": 1.192578270845923e-06,
+      "loss": 3.0055,
+      "step": 59477
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7244284749031067,
+      "learning_rate": 1.1912072172735488e-06,
+      "loss": 2.8573,
+      "step": 59478
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7515481114387512,
+      "learning_rate": 1.1898369507060024e-06,
+      "loss": 2.9796,
+      "step": 59479
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7766403555870056,
+      "learning_rate": 1.1884674711469477e-06,
+      "loss": 2.9782,
+      "step": 59480
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7261124849319458,
+      "learning_rate": 1.1870987785999154e-06,
+      "loss": 2.7586,
+      "step": 59481
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7415921092033386,
+      "learning_rate": 1.1857308730685688e-06,
+      "loss": 3.0122,
+      "step": 59482
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7948103547096252,
+      "learning_rate": 1.1843637545565388e-06,
+      "loss": 3.0193,
+      "step": 59483
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7435404062271118,
+      "learning_rate": 1.1829974230673557e-06,
+      "loss": 2.734,
+      "step": 59484
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7635899186134338,
+      "learning_rate": 1.1816318786046497e-06,
+      "loss": 2.918,
+      "step": 59485
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7679666876792908,
+      "learning_rate": 1.1802671211720182e-06,
+      "loss": 2.9842,
+      "step": 59486
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7481985092163086,
+      "learning_rate": 1.1789031507730584e-06,
+      "loss": 2.8345,
+      "step": 59487
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7324191927909851,
+      "learning_rate": 1.1775399674113672e-06,
+      "loss": 2.9773,
+      "step": 59488
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7569427490234375,
+      "learning_rate": 1.1761775710905086e-06,
+      "loss": 2.8726,
+      "step": 59489
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7332199811935425,
+      "learning_rate": 1.174815961814113e-06,
+      "loss": 2.9145,
+      "step": 59490
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7667021155357361,
+      "learning_rate": 1.1734551395857107e-06,
+      "loss": 2.8075,
+      "step": 59491
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7241703271865845,
+      "learning_rate": 1.1720951044089323e-06,
+      "loss": 2.8421,
+      "step": 59492
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.807370662689209,
+      "learning_rate": 1.1707358562873414e-06,
+      "loss": 2.9485,
+      "step": 59493
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7350105047225952,
+      "learning_rate": 1.1693773952245355e-06,
+      "loss": 3.1253,
+      "step": 59494
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8460718989372253,
+      "learning_rate": 1.1680197212240783e-06,
+      "loss": 2.8301,
+      "step": 59495
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7515153884887695,
+      "learning_rate": 1.1666628342895e-06,
+      "loss": 2.8304,
+      "step": 59496
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.718813955783844,
+      "learning_rate": 1.1653067344244649e-06,
+      "loss": 2.9743,
+      "step": 59497
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7459069490432739,
+      "learning_rate": 1.163951421632503e-06,
+      "loss": 2.8279,
+      "step": 59498
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7909800410270691,
+      "learning_rate": 1.1625968959171782e-06,
+      "loss": 3.0892,
+      "step": 59499
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7388293743133545,
+      "learning_rate": 1.1612431572820547e-06,
+      "loss": 2.8286,
+      "step": 59500
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7680931091308594,
+      "learning_rate": 1.1598902057306958e-06,
+      "loss": 2.8482,
+      "step": 59501
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7312542200088501,
+      "learning_rate": 1.1585380412666657e-06,
+      "loss": 2.8631,
+      "step": 59502
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7777981758117676,
+      "learning_rate": 1.157186663893528e-06,
+      "loss": 2.8585,
+      "step": 59503
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7600824236869812,
+      "learning_rate": 1.1558360736148797e-06,
+      "loss": 2.6733,
+      "step": 59504
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7767745852470398,
+      "learning_rate": 1.1544862704342184e-06,
+      "loss": 2.8698,
+      "step": 59505
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7815971970558167,
+      "learning_rate": 1.1531372543551409e-06,
+      "loss": 2.8926,
+      "step": 59506
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7735201716423035,
+      "learning_rate": 1.151789025381178e-06,
+      "loss": 2.8795,
+      "step": 59507
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7491256594657898,
+      "learning_rate": 1.150441583515893e-06,
+      "loss": 2.9923,
+      "step": 59508
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7839617133140564,
+      "learning_rate": 1.1490949287628503e-06,
+      "loss": 3.0807,
+      "step": 59509
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7311417460441589,
+      "learning_rate": 1.1477490611255468e-06,
+      "loss": 3.0293,
+      "step": 59510
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7277370691299438,
+      "learning_rate": 1.1464039806075464e-06,
+      "loss": 2.8279,
+      "step": 59511
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7706161141395569,
+      "learning_rate": 1.1450596872124463e-06,
+      "loss": 2.918,
+      "step": 59512
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7380478978157043,
+      "learning_rate": 1.14371618094371e-06,
+      "loss": 2.7686,
+      "step": 59513
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7525132894515991,
+      "learning_rate": 1.1423734618049352e-06,
+      "loss": 2.6612,
+      "step": 59514
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7478514909744263,
+      "learning_rate": 1.1410315297996187e-06,
+      "loss": 3.1156,
+      "step": 59515
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7614080309867859,
+      "learning_rate": 1.139690384931291e-06,
+      "loss": 2.9505,
+      "step": 59516
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7383436560630798,
+      "learning_rate": 1.1383500272035163e-06,
+      "loss": 2.9673,
+      "step": 59517
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7430320382118225,
+      "learning_rate": 1.1370104566197913e-06,
+      "loss": 2.9268,
+      "step": 59518
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7891849279403687,
+      "learning_rate": 1.1356716731836802e-06,
+      "loss": 2.8843,
+      "step": 59519
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7453582286834717,
+      "learning_rate": 1.1343336768987132e-06,
+      "loss": 2.9038,
+      "step": 59520
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7106955051422119,
+      "learning_rate": 1.1329964677683878e-06,
+      "loss": 2.5896,
+      "step": 59521
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.807589054107666,
+      "learning_rate": 1.1316600457962011e-06,
+      "loss": 2.8919,
+      "step": 59522
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7693127989768982,
+      "learning_rate": 1.1303244109857168e-06,
+      "loss": 3.0337,
+      "step": 59523
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.747882068157196,
+      "learning_rate": 1.1289895633404656e-06,
+      "loss": 3.0611,
+      "step": 59524
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7988489270210266,
+      "learning_rate": 1.127655502863911e-06,
+      "loss": 2.8766,
+      "step": 59525
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7369797229766846,
+      "learning_rate": 1.1263222295595842e-06,
+      "loss": 2.8861,
+      "step": 59526
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7736074924468994,
+      "learning_rate": 1.124989743431015e-06,
+      "loss": 2.7278,
+      "step": 59527
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7230579257011414,
+      "learning_rate": 1.1236580444817345e-06,
+      "loss": 2.8485,
+      "step": 59528
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7344254851341248,
+      "learning_rate": 1.1223271327151728e-06,
+      "loss": 3.0521,
+      "step": 59529
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7863304615020752,
+      "learning_rate": 1.1209970081348941e-06,
+      "loss": 2.9299,
+      "step": 59530
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7944401502609253,
+      "learning_rate": 1.1196676707443953e-06,
+      "loss": 2.8285,
+      "step": 59531
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7780875563621521,
+      "learning_rate": 1.1183391205471737e-06,
+      "loss": 2.8872,
+      "step": 59532
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7192845344543457,
+      "learning_rate": 1.1170113575467266e-06,
+      "loss": 2.7766,
+      "step": 59533
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7729279398918152,
+      "learning_rate": 1.1156843817465178e-06,
+      "loss": 2.7897,
+      "step": 59534
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7623705267906189,
+      "learning_rate": 1.1143581931501111e-06,
+      "loss": 2.882,
+      "step": 59535
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7770814895629883,
+      "learning_rate": 1.1130327917609704e-06,
+      "loss": 3.0556,
+      "step": 59536
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7293316721916199,
+      "learning_rate": 1.1117081775825598e-06,
+      "loss": 2.7412,
+      "step": 59537
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7572553157806396,
+      "learning_rate": 1.1103843506184096e-06,
+      "loss": 2.7117,
+      "step": 59538
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7205670475959778,
+      "learning_rate": 1.1090613108719505e-06,
+      "loss": 2.7834,
+      "step": 59539
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7953411936759949,
+      "learning_rate": 1.1077390583467126e-06,
+      "loss": 3.0259,
+      "step": 59540
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7684890031814575,
+      "learning_rate": 1.1064175930461604e-06,
+      "loss": 2.7954,
+      "step": 59541
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7398133873939514,
+      "learning_rate": 1.1050969149738243e-06,
+      "loss": 3.0277,
+      "step": 59542
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7606955170631409,
+      "learning_rate": 1.1037770241331013e-06,
+      "loss": 2.8503,
+      "step": 59543
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7371569871902466,
+      "learning_rate": 1.1024579205275218e-06,
+      "loss": 2.629,
+      "step": 59544
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7682278156280518,
+      "learning_rate": 1.101139604160517e-06,
+      "loss": 3.0296,
+      "step": 59545
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8034104704856873,
+      "learning_rate": 1.0998220750356169e-06,
+      "loss": 2.9309,
+      "step": 59546
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7584922909736633,
+      "learning_rate": 1.098505333156252e-06,
+      "loss": 2.854,
+      "step": 59547
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7490361928939819,
+      "learning_rate": 1.0971893785258868e-06,
+      "loss": 3.0433,
+      "step": 59548
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7814920544624329,
+      "learning_rate": 1.0958742111480179e-06,
+      "loss": 2.7952,
+      "step": 59549
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7534535527229309,
+      "learning_rate": 1.094559831026076e-06,
+      "loss": 2.8577,
+      "step": 59550
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.722969114780426,
+      "learning_rate": 1.0932462381635254e-06,
+      "loss": 2.7215,
+      "step": 59551
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7845919728279114,
+      "learning_rate": 1.0919334325638629e-06,
+      "loss": 2.7127,
+      "step": 59552
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8240389227867126,
+      "learning_rate": 1.0906214142305191e-06,
+      "loss": 2.735,
+      "step": 59553
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7913576364517212,
+      "learning_rate": 1.0893101831669249e-06,
+      "loss": 2.9265,
+      "step": 59554
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.716317892074585,
+      "learning_rate": 1.0879997393766105e-06,
+      "loss": 2.9504,
+      "step": 59555
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7530320882797241,
+      "learning_rate": 1.08669008286294e-06,
+      "loss": 2.8871,
+      "step": 59556
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7626606822013855,
+      "learning_rate": 1.0853812136294104e-06,
+      "loss": 3.0875,
+      "step": 59557
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7616795301437378,
+      "learning_rate": 1.084073131679486e-06,
+      "loss": 2.6925,
+      "step": 59558
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7134222388267517,
+      "learning_rate": 1.0827658370165304e-06,
+      "loss": 2.921,
+      "step": 59559
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7397261261940002,
+      "learning_rate": 1.0814593296441077e-06,
+      "loss": 2.9264,
+      "step": 59560
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7546602487564087,
+      "learning_rate": 1.0801536095655482e-06,
+      "loss": 3.096,
+      "step": 59561
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7439787983894348,
+      "learning_rate": 1.0788486767843497e-06,
+      "loss": 2.8096,
+      "step": 59562
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7837499380111694,
+      "learning_rate": 1.0775445313039088e-06,
+      "loss": 2.8704,
+      "step": 59563
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7200571894645691,
+      "learning_rate": 1.0762411731277232e-06,
+      "loss": 2.9517,
+      "step": 59564
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7449080348014832,
+      "learning_rate": 1.0749386022591899e-06,
+      "loss": 2.9333,
+      "step": 59565
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7231045365333557,
+      "learning_rate": 1.0736368187017064e-06,
+      "loss": 2.7141,
+      "step": 59566
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8130992650985718,
+      "learning_rate": 1.0723358224587698e-06,
+      "loss": 2.8928,
+      "step": 59567
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7753105759620667,
+      "learning_rate": 1.071035613533744e-06,
+      "loss": 2.8285,
+      "step": 59568
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7972942590713501,
+      "learning_rate": 1.0697361919300928e-06,
+      "loss": 2.8247,
+      "step": 59569
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.710925817489624,
+      "learning_rate": 1.068437557651214e-06,
+      "loss": 2.8817,
+      "step": 59570
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7548450827598572,
+      "learning_rate": 1.0671397107005708e-06,
+      "loss": 2.7388,
+      "step": 59571
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7482631802558899,
+      "learning_rate": 1.0658426510815276e-06,
+      "loss": 2.7826,
+      "step": 59572
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7267739772796631,
+      "learning_rate": 1.064546378797515e-06,
+      "loss": 2.8136,
+      "step": 59573
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7370942831039429,
+      "learning_rate": 1.0632508938519634e-06,
+      "loss": 2.8792,
+      "step": 59574
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7242228984832764,
+      "learning_rate": 1.0619561962482703e-06,
+      "loss": 2.8447,
+      "step": 59575
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7722229361534119,
+      "learning_rate": 1.0606622859898662e-06,
+      "loss": 3.0327,
+      "step": 59576
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7654498219490051,
+      "learning_rate": 1.0593691630801482e-06,
+      "loss": 2.9213,
+      "step": 59577
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7603773474693298,
+      "learning_rate": 1.0580768275225139e-06,
+      "loss": 3.1623,
+      "step": 59578
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7189480662345886,
+      "learning_rate": 1.0567852793203602e-06,
+      "loss": 2.812,
+      "step": 59579
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7826395034790039,
+      "learning_rate": 1.055494518477118e-06,
+      "loss": 3.1349,
+      "step": 59580
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7658017873764038,
+      "learning_rate": 1.0542045449961844e-06,
+      "loss": 2.6386,
+      "step": 59581
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7522681951522827,
+      "learning_rate": 1.0529153588809235e-06,
+      "loss": 2.8757,
+      "step": 59582
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7336506843566895,
+      "learning_rate": 1.0516269601347326e-06,
+      "loss": 2.8125,
+      "step": 59583
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7140693068504333,
+      "learning_rate": 1.0503393487610423e-06,
+      "loss": 2.9674,
+      "step": 59584
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7803701758384705,
+      "learning_rate": 1.0490525247632165e-06,
+      "loss": 3.1645,
+      "step": 59585
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7349176406860352,
+      "learning_rate": 1.0477664881446857e-06,
+      "loss": 2.7669,
+      "step": 59586
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7476712465286255,
+      "learning_rate": 1.0464812389087806e-06,
+      "loss": 2.765,
+      "step": 59587
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.764438271522522,
+      "learning_rate": 1.0451967770588987e-06,
+      "loss": 2.971,
+      "step": 59588
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7536563873291016,
+      "learning_rate": 1.043913102598437e-06,
+      "loss": 2.9274,
+      "step": 59589
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.728675365447998,
+      "learning_rate": 1.0426302155307597e-06,
+      "loss": 2.7336,
+      "step": 59590
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7635043859481812,
+      "learning_rate": 1.041348115859264e-06,
+      "loss": 2.8814,
+      "step": 59591
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7409845590591431,
+      "learning_rate": 1.0400668035873472e-06,
+      "loss": 2.8674,
+      "step": 59592
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.707777202129364,
+      "learning_rate": 1.03878627871834e-06,
+      "loss": 2.8546,
+      "step": 59593
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7888078689575195,
+      "learning_rate": 1.0375065412556393e-06,
+      "loss": 3.0258,
+      "step": 59594
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7485946416854858,
+      "learning_rate": 1.0362275912026096e-06,
+      "loss": 2.9028,
+      "step": 59595
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7295870780944824,
+      "learning_rate": 1.0349494285626147e-06,
+      "loss": 3.0208,
+      "step": 59596
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7414456009864807,
+      "learning_rate": 1.0336720533390187e-06,
+      "loss": 2.789,
+      "step": 59597
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7335713505744934,
+      "learning_rate": 1.0323954655352185e-06,
+      "loss": 3.0562,
+      "step": 59598
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.759948194026947,
+      "learning_rate": 1.0311196651545118e-06,
+      "loss": 2.9113,
+      "step": 59599
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7210058569908142,
+      "learning_rate": 1.029844652200329e-06,
+      "loss": 3.0582,
+      "step": 59600
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.736574113368988,
+      "learning_rate": 1.028570426676001e-06,
+      "loss": 2.7906,
+      "step": 59601
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7816761136054993,
+      "learning_rate": 1.0272969885848915e-06,
+      "loss": 2.9032,
+      "step": 59602
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8219104409217834,
+      "learning_rate": 1.0260243379302979e-06,
+      "loss": 3.044,
+      "step": 59603
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7174280285835266,
+      "learning_rate": 1.024752474715651e-06,
+      "loss": 2.9151,
+      "step": 59604
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7537269592285156,
+      "learning_rate": 1.023481398944248e-06,
+      "loss": 2.9196,
+      "step": 59605
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7246042490005493,
+      "learning_rate": 1.022211110619453e-06,
+      "loss": 2.7453,
+      "step": 59606
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7886763215065002,
+      "learning_rate": 1.0209416097446632e-06,
+      "loss": 2.8192,
+      "step": 59607
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7710185050964355,
+      "learning_rate": 1.0196728963231427e-06,
+      "loss": 3.0203,
+      "step": 59608
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.704352080821991,
+      "learning_rate": 1.018404970358222e-06,
+      "loss": 2.8758,
+      "step": 59609
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7111163139343262,
+      "learning_rate": 1.0171378318533318e-06,
+      "loss": 2.7523,
+      "step": 59610
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7470084428787231,
+      "learning_rate": 1.0158714808117362e-06,
+      "loss": 2.7751,
+      "step": 59611
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7764518857002258,
+      "learning_rate": 1.0146059172368327e-06,
+      "loss": 2.9253,
+      "step": 59612
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7668911218643188,
+      "learning_rate": 1.0133411411318849e-06,
+      "loss": 3.1101,
+      "step": 59613
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7347301840782166,
+      "learning_rate": 1.012077152500257e-06,
+      "loss": 2.9405,
+      "step": 59614
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8694640398025513,
+      "learning_rate": 1.0108139513452795e-06,
+      "loss": 2.8178,
+      "step": 59615
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7998950481414795,
+      "learning_rate": 1.0095515376702833e-06,
+      "loss": 2.9485,
+      "step": 59616
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7624552249908447,
+      "learning_rate": 1.0082899114785659e-06,
+      "loss": 3.0671,
+      "step": 59617
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7087655067443848,
+      "learning_rate": 1.0070290727734909e-06,
+      "loss": 2.8172,
+      "step": 59618
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7436306476593018,
+      "learning_rate": 1.0057690215583558e-06,
+      "loss": 2.9391,
+      "step": 59619
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7129530906677246,
+      "learning_rate": 1.004509757836458e-06,
+      "loss": 2.7657,
+      "step": 59620
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7246747016906738,
+      "learning_rate": 1.0032512816111616e-06,
+      "loss": 3.0554,
+      "step": 59621
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7818371057510376,
+      "learning_rate": 1.0019935928857637e-06,
+      "loss": 2.7496,
+      "step": 59622
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.754097580909729,
+      "learning_rate": 1.0007366916635618e-06,
+      "loss": 2.7842,
+      "step": 59623
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7770971655845642,
+      "learning_rate": 9.994805779478865e-07,
+      "loss": 2.9256,
+      "step": 59624
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7818499803543091,
+      "learning_rate": 9.982252517420352e-07,
+      "loss": 3.0954,
+      "step": 59625
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7363375425338745,
+      "learning_rate": 9.969707130493055e-07,
+      "loss": 3.0082,
+      "step": 59626
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7672472596168518,
+      "learning_rate": 9.957169618729944e-07,
+      "loss": 2.9003,
+      "step": 59627
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7299450039863586,
+      "learning_rate": 9.944639982164326e-07,
+      "loss": 3.1018,
+      "step": 59628
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.788252592086792,
+      "learning_rate": 9.932118220829177e-07,
+      "loss": 2.7675,
+      "step": 59629
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.760606050491333,
+      "learning_rate": 9.91960433475747e-07,
+      "loss": 2.8997,
+      "step": 59630
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7685892581939697,
+      "learning_rate": 9.907098323981845e-07,
+      "loss": 2.9651,
+      "step": 59631
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7383074760437012,
+      "learning_rate": 9.894600188535606e-07,
+      "loss": 2.9564,
+      "step": 59632
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7410311698913574,
+      "learning_rate": 9.882109928451398e-07,
+      "loss": 2.6135,
+      "step": 59633
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7539855241775513,
+      "learning_rate": 9.869627543762527e-07,
+      "loss": 3.037,
+      "step": 59634
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7680968046188354,
+      "learning_rate": 9.857153034501297e-07,
+      "loss": 2.7394,
+      "step": 59635
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.779704213142395,
+      "learning_rate": 9.84468640070102e-07,
+      "loss": 2.7469,
+      "step": 59636
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7694498896598816,
+      "learning_rate": 9.83222764239433e-07,
+      "loss": 3.0253,
+      "step": 59637
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8191694617271423,
+      "learning_rate": 9.819776759614206e-07,
+      "loss": 3.0428,
+      "step": 59638
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.722665548324585,
+      "learning_rate": 9.807333752392953e-07,
+      "loss": 2.944,
+      "step": 59639
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7409653067588806,
+      "learning_rate": 9.79489862076388e-07,
+      "loss": 2.8522,
+      "step": 59640
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7272565364837646,
+      "learning_rate": 9.782471364759625e-07,
+      "loss": 2.7308,
+      "step": 59641
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.766656756401062,
+      "learning_rate": 9.770051984413164e-07,
+      "loss": 2.7798,
+      "step": 59642
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7929888367652893,
+      "learning_rate": 9.757640479756468e-07,
+      "loss": 2.9093,
+      "step": 59643
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7375426888465881,
+      "learning_rate": 9.745236850822846e-07,
+      "loss": 2.6784,
+      "step": 59644
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7349624037742615,
+      "learning_rate": 9.732841097644605e-07,
+      "loss": 2.7962,
+      "step": 59645
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7367823719978333,
+      "learning_rate": 9.72045322025472e-07,
+      "loss": 2.975,
+      "step": 59646
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7537251114845276,
+      "learning_rate": 9.70807321868583e-07,
+      "loss": 2.8237,
+      "step": 59647
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7877868413925171,
+      "learning_rate": 9.695701092970243e-07,
+      "loss": 2.7918,
+      "step": 59648
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7532494068145752,
+      "learning_rate": 9.683336843140598e-07,
+      "loss": 2.8354,
+      "step": 59649
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7658734321594238,
+      "learning_rate": 9.67098046922954e-07,
+      "loss": 2.9192,
+      "step": 59650
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7385966777801514,
+      "learning_rate": 9.658631971269704e-07,
+      "loss": 2.8873,
+      "step": 59651
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7445787787437439,
+      "learning_rate": 9.6462913492934e-07,
+      "loss": 2.6811,
+      "step": 59652
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7418826818466187,
+      "learning_rate": 9.63395860333327e-07,
+      "loss": 2.8437,
+      "step": 59653
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7458414435386658,
+      "learning_rate": 9.621633733421953e-07,
+      "loss": 2.7761,
+      "step": 59654
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7673033475875854,
+      "learning_rate": 9.609316739591754e-07,
+      "loss": 2.9863,
+      "step": 59655
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7251089811325073,
+      "learning_rate": 9.597007621874985e-07,
+      "loss": 3.0085,
+      "step": 59656
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7382851243019104,
+      "learning_rate": 9.584706380304286e-07,
+      "loss": 2.6704,
+      "step": 59657
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7357835173606873,
+      "learning_rate": 9.572413014911628e-07,
+      "loss": 2.9542,
+      "step": 59658
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7621243596076965,
+      "learning_rate": 9.560127525730321e-07,
+      "loss": 2.9173,
+      "step": 59659
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8021424412727356,
+      "learning_rate": 9.547849912791673e-07,
+      "loss": 2.9653,
+      "step": 59660
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7782572507858276,
+      "learning_rate": 9.535580176128654e-07,
+      "loss": 2.9197,
+      "step": 59661
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7518317103385925,
+      "learning_rate": 9.523318315773242e-07,
+      "loss": 2.8542,
+      "step": 59662
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7433437705039978,
+      "learning_rate": 9.511064331757745e-07,
+      "loss": 2.9361,
+      "step": 59663
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7323741912841797,
+      "learning_rate": 9.498818224114802e-07,
+      "loss": 2.8354,
+      "step": 59664
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7862437963485718,
+      "learning_rate": 9.486579992876386e-07,
+      "loss": 2.9296,
+      "step": 59665
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.729877769947052,
+      "learning_rate": 9.474349638074808e-07,
+      "loss": 2.6364,
+      "step": 59666
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7364502549171448,
+      "learning_rate": 9.462127159742372e-07,
+      "loss": 2.9278,
+      "step": 59667
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.748285174369812,
+      "learning_rate": 9.449912557911055e-07,
+      "loss": 2.8261,
+      "step": 59668
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7488347887992859,
+      "learning_rate": 9.437705832613163e-07,
+      "loss": 3.0538,
+      "step": 59669
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7590417265892029,
+      "learning_rate": 9.425506983880671e-07,
+      "loss": 2.8349,
+      "step": 59670
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7655706405639648,
+      "learning_rate": 9.413316011745886e-07,
+      "loss": 2.8614,
+      "step": 59671
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7407287955284119,
+      "learning_rate": 9.401132916240784e-07,
+      "loss": 3.0396,
+      "step": 59672
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7478367686271667,
+      "learning_rate": 9.388957697397669e-07,
+      "loss": 2.9103,
+      "step": 59673
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8015658259391785,
+      "learning_rate": 9.37679035524852e-07,
+      "loss": 2.7994,
+      "step": 59674
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.748598039150238,
+      "learning_rate": 9.364630889825309e-07,
+      "loss": 2.7964,
+      "step": 59675
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7474225759506226,
+      "learning_rate": 9.352479301160009e-07,
+      "loss": 2.935,
+      "step": 59676
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7985930442810059,
+      "learning_rate": 9.340335589284931e-07,
+      "loss": 2.9996,
+      "step": 59677
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.739578127861023,
+      "learning_rate": 9.328199754231713e-07,
+      "loss": 2.8553,
+      "step": 59678
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.745047926902771,
+      "learning_rate": 9.316071796032331e-07,
+      "loss": 2.7658,
+      "step": 59679
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7639679312705994,
+      "learning_rate": 9.303951714719093e-07,
+      "loss": 2.7573,
+      "step": 59680
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8055102229118347,
+      "learning_rate": 9.29183951032364e-07,
+      "loss": 2.7784,
+      "step": 59681
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.717705488204956,
+      "learning_rate": 9.279735182877613e-07,
+      "loss": 2.8754,
+      "step": 59682
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7145783305168152,
+      "learning_rate": 9.267638732413651e-07,
+      "loss": 2.9277,
+      "step": 59683
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7795868515968323,
+      "learning_rate": 9.255550158962733e-07,
+      "loss": 2.8612,
+      "step": 59684
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7477320432662964,
+      "learning_rate": 9.243469462557163e-07,
+      "loss": 2.8429,
+      "step": 59685
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7685227990150452,
+      "learning_rate": 9.231396643228917e-07,
+      "loss": 2.7086,
+      "step": 59686
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.742152214050293,
+      "learning_rate": 9.219331701009635e-07,
+      "loss": 2.9458,
+      "step": 59687
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7278732657432556,
+      "learning_rate": 9.207274635930961e-07,
+      "loss": 2.8947,
+      "step": 59688
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7534875273704529,
+      "learning_rate": 9.195225448024535e-07,
+      "loss": 2.974,
+      "step": 59689
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.726291835308075,
+      "learning_rate": 9.183184137322664e-07,
+      "loss": 2.9126,
+      "step": 59690
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.746126651763916,
+      "learning_rate": 9.171150703856323e-07,
+      "loss": 2.9024,
+      "step": 59691
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7456505298614502,
+      "learning_rate": 9.15912514765782e-07,
+      "loss": 3.0167,
+      "step": 59692
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7747290134429932,
+      "learning_rate": 9.147107468758464e-07,
+      "loss": 2.956,
+      "step": 59693
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7368700504302979,
+      "learning_rate": 9.135097667189895e-07,
+      "loss": 2.9151,
+      "step": 59694
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7123827934265137,
+      "learning_rate": 9.123095742984088e-07,
+      "loss": 3.1211,
+      "step": 59695
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7461838722229004,
+      "learning_rate": 9.111101696172351e-07,
+      "loss": 2.9004,
+      "step": 59696
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7076005935668945,
+      "learning_rate": 9.099115526786327e-07,
+      "loss": 2.8209,
+      "step": 59697
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7806337475776672,
+      "learning_rate": 9.087137234857656e-07,
+      "loss": 2.6983,
+      "step": 59698
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7522020936012268,
+      "learning_rate": 9.075166820417978e-07,
+      "loss": 2.8614,
+      "step": 59699
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7688117623329163,
+      "learning_rate": 9.063204283498604e-07,
+      "loss": 2.8533,
+      "step": 59700
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7308453917503357,
+      "learning_rate": 9.051249624131174e-07,
+      "loss": 2.8086,
+      "step": 59701
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7483466267585754,
+      "learning_rate": 9.039302842346996e-07,
+      "loss": 2.9024,
+      "step": 59702
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7657617926597595,
+      "learning_rate": 9.027363938178045e-07,
+      "loss": 2.7308,
+      "step": 59703
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7511147856712341,
+      "learning_rate": 9.015432911654962e-07,
+      "loss": 2.9577,
+      "step": 59704
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7356677651405334,
+      "learning_rate": 9.003509762809724e-07,
+      "loss": 2.991,
+      "step": 59705
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7413914203643799,
+      "learning_rate": 8.991594491673637e-07,
+      "loss": 3.2518,
+      "step": 59706
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7317723631858826,
+      "learning_rate": 8.97968709827801e-07,
+      "loss": 2.9233,
+      "step": 59707
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7266475558280945,
+      "learning_rate": 8.967787582654484e-07,
+      "loss": 2.9267,
+      "step": 59708
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.725741446018219,
+      "learning_rate": 8.955895944834035e-07,
+      "loss": 2.8691,
+      "step": 59709
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7397601008415222,
+      "learning_rate": 8.94401218484797e-07,
+      "loss": 2.9947,
+      "step": 59710
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.735898494720459,
+      "learning_rate": 8.932136302727933e-07,
+      "loss": 2.7562,
+      "step": 59711
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7711181044578552,
+      "learning_rate": 8.920268298504896e-07,
+      "loss": 2.9557,
+      "step": 59712
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8324527740478516,
+      "learning_rate": 8.908408172210169e-07,
+      "loss": 2.8636,
+      "step": 59713
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7208682894706726,
+      "learning_rate": 8.896555923875393e-07,
+      "loss": 2.7489,
+      "step": 59714
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7793273329734802,
+      "learning_rate": 8.884711553530876e-07,
+      "loss": 2.862,
+      "step": 59715
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7881961464881897,
+      "learning_rate": 8.872875061208928e-07,
+      "loss": 2.8456,
+      "step": 59716
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7748516798019409,
+      "learning_rate": 8.861046446939857e-07,
+      "loss": 3.0182,
+      "step": 59717
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7874419093132019,
+      "learning_rate": 8.849225710755304e-07,
+      "loss": 2.9961,
+      "step": 59718
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7328853011131287,
+      "learning_rate": 8.83741285268591e-07,
+      "loss": 3.0178,
+      "step": 59719
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7477407455444336,
+      "learning_rate": 8.825607872763319e-07,
+      "loss": 2.914,
+      "step": 59720
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7703258991241455,
+      "learning_rate": 8.813810771018503e-07,
+      "loss": 3.1145,
+      "step": 59721
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7548531889915466,
+      "learning_rate": 8.802021547482441e-07,
+      "loss": 3.0983,
+      "step": 59722
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7501577734947205,
+      "learning_rate": 8.790240202186105e-07,
+      "loss": 2.8297,
+      "step": 59723
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8373538255691528,
+      "learning_rate": 8.778466735160472e-07,
+      "loss": 2.8036,
+      "step": 59724
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7723232507705688,
+      "learning_rate": 8.76670114643685e-07,
+      "loss": 2.98,
+      "step": 59725
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7283118963241577,
+      "learning_rate": 8.754943436046214e-07,
+      "loss": 2.9201,
+      "step": 59726
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7659836411476135,
+      "learning_rate": 8.743193604019205e-07,
+      "loss": 2.7259,
+      "step": 59727
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7766382098197937,
+      "learning_rate": 8.731451650386801e-07,
+      "loss": 2.8198,
+      "step": 59728
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7227720618247986,
+      "learning_rate": 8.719717575180308e-07,
+      "loss": 3.1198,
+      "step": 59729
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7672495245933533,
+      "learning_rate": 8.707991378430034e-07,
+      "loss": 2.8466,
+      "step": 59730
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.754241406917572,
+      "learning_rate": 8.696273060167624e-07,
+      "loss": 2.6828,
+      "step": 59731
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7628318667411804,
+      "learning_rate": 8.684562620423385e-07,
+      "loss": 3.0211,
+      "step": 59732
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.788650393486023,
+      "learning_rate": 8.672860059228293e-07,
+      "loss": 2.8669,
+      "step": 59733
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7480198740959167,
+      "learning_rate": 8.661165376613321e-07,
+      "loss": 2.9814,
+      "step": 59734
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7865045666694641,
+      "learning_rate": 8.649478572609114e-07,
+      "loss": 2.9745,
+      "step": 59735
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7339010238647461,
+      "learning_rate": 8.637799647246646e-07,
+      "loss": 2.803,
+      "step": 59736
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.6831942200660706,
+      "learning_rate": 8.626128600556226e-07,
+      "loss": 3.0673,
+      "step": 59737
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7489328980445862,
+      "learning_rate": 8.614465432568829e-07,
+      "loss": 2.8345,
+      "step": 59738
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8042427897453308,
+      "learning_rate": 8.602810143315431e-07,
+      "loss": 2.8714,
+      "step": 59739
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8107337355613708,
+      "learning_rate": 8.591162732826339e-07,
+      "loss": 2.9028,
+      "step": 59740
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7713976502418518,
+      "learning_rate": 8.579523201132866e-07,
+      "loss": 2.9202,
+      "step": 59741
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8133545517921448,
+      "learning_rate": 8.567891548264649e-07,
+      "loss": 2.8562,
+      "step": 59742
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8308982253074646,
+      "learning_rate": 8.556267774253001e-07,
+      "loss": 2.7826,
+      "step": 59743
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.756340503692627,
+      "learning_rate": 8.544651879128894e-07,
+      "loss": 2.8316,
+      "step": 59744
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7788317203521729,
+      "learning_rate": 8.533043862921973e-07,
+      "loss": 2.8322,
+      "step": 59745
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.735393226146698,
+      "learning_rate": 8.521443725663546e-07,
+      "loss": 2.8576,
+      "step": 59746
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7246791124343872,
+      "learning_rate": 8.509851467383588e-07,
+      "loss": 3.0481,
+      "step": 59747
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7521356344223022,
+      "learning_rate": 8.498267088113409e-07,
+      "loss": 2.7354,
+      "step": 59748
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.803340494632721,
+      "learning_rate": 8.48669058788265e-07,
+      "loss": 2.8304,
+      "step": 59749
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7315862774848938,
+      "learning_rate": 8.47512196672262e-07,
+      "loss": 2.8125,
+      "step": 59750
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7708979845046997,
+      "learning_rate": 8.463561224662962e-07,
+      "loss": 2.8332,
+      "step": 59751
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7568145990371704,
+      "learning_rate": 8.452008361734985e-07,
+      "loss": 2.7078,
+      "step": 59752
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7380116581916809,
+      "learning_rate": 8.440463377968332e-07,
+      "loss": 2.9881,
+      "step": 59753
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7458478212356567,
+      "learning_rate": 8.428926273393977e-07,
+      "loss": 2.9401,
+      "step": 59754
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7849690914154053,
+      "learning_rate": 8.417397048041897e-07,
+      "loss": 2.8784,
+      "step": 59755
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8106399178504944,
+      "learning_rate": 8.405875701942732e-07,
+      "loss": 2.7235,
+      "step": 59756
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7472584247589111,
+      "learning_rate": 8.394362235126795e-07,
+      "loss": 3.1152,
+      "step": 59757
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7843793630599976,
+      "learning_rate": 8.382856647624392e-07,
+      "loss": 2.928,
+      "step": 59758
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7767054438591003,
+      "learning_rate": 8.371358939465833e-07,
+      "loss": 2.9102,
+      "step": 59759
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7358002066612244,
+      "learning_rate": 8.359869110681095e-07,
+      "loss": 2.8749,
+      "step": 59760
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.768272340297699,
+      "learning_rate": 8.348387161301151e-07,
+      "loss": 2.8866,
+      "step": 59761
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7254506945610046,
+      "learning_rate": 8.336913091355313e-07,
+      "loss": 2.7688,
+      "step": 59762
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7313871383666992,
+      "learning_rate": 8.325446900874555e-07,
+      "loss": 2.6363,
+      "step": 59763
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.793598473072052,
+      "learning_rate": 8.313988589888853e-07,
+      "loss": 2.9504,
+      "step": 59764
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.73553466796875,
+      "learning_rate": 8.302538158428185e-07,
+      "loss": 2.9293,
+      "step": 59765
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.765264093875885,
+      "learning_rate": 8.29109560652319e-07,
+      "loss": 3.1213,
+      "step": 59766
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7166268229484558,
+      "learning_rate": 8.279660934203181e-07,
+      "loss": 2.8594,
+      "step": 59767
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.772975742816925,
+      "learning_rate": 8.268234141499131e-07,
+      "loss": 2.7602,
+      "step": 59768
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.73553466796875,
+      "learning_rate": 8.256815228440683e-07,
+      "loss": 2.8789,
+      "step": 59769
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7122424244880676,
+      "learning_rate": 8.245404195058147e-07,
+      "loss": 2.7645,
+      "step": 59770
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.6854143738746643,
+      "learning_rate": 8.234001041381166e-07,
+      "loss": 2.9399,
+      "step": 59771
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7698032855987549,
+      "learning_rate": 8.222605767440383e-07,
+      "loss": 3.0868,
+      "step": 59772
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7176466584205627,
+      "learning_rate": 8.211218373265105e-07,
+      "loss": 2.824,
+      "step": 59773
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7282741665840149,
+      "learning_rate": 8.199838858885977e-07,
+      "loss": 2.9953,
+      "step": 59774
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8401013612747192,
+      "learning_rate": 8.18846722433264e-07,
+      "loss": 2.7525,
+      "step": 59775
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7844291925430298,
+      "learning_rate": 8.177103469635071e-07,
+      "loss": 2.899,
+      "step": 59776
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7597390413284302,
+      "learning_rate": 8.165747594823247e-07,
+      "loss": 2.7735,
+      "step": 59777
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7756280899047852,
+      "learning_rate": 8.154399599926809e-07,
+      "loss": 2.8091,
+      "step": 59778
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7339857816696167,
+      "learning_rate": 8.143059484976066e-07,
+      "loss": 2.9528,
+      "step": 59779
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.764804482460022,
+      "learning_rate": 8.131727250000663e-07,
+      "loss": 2.8768,
+      "step": 59780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8054780960083008,
+      "learning_rate": 8.120402895030908e-07,
+      "loss": 2.5517,
+      "step": 59781
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7895975708961487,
+      "learning_rate": 8.109086420095779e-07,
+      "loss": 2.8077,
+      "step": 59782
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7449101805686951,
+      "learning_rate": 8.097777825225582e-07,
+      "loss": 2.8612,
+      "step": 59783
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7465765476226807,
+      "learning_rate": 8.086477110449962e-07,
+      "loss": 2.7231,
+      "step": 59784
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7886302471160889,
+      "learning_rate": 8.075184275798896e-07,
+      "loss": 2.8589,
+      "step": 59785
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7831330895423889,
+      "learning_rate": 8.063899321302025e-07,
+      "loss": 2.7825,
+      "step": 59786
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7341232895851135,
+      "learning_rate": 8.052622246988993e-07,
+      "loss": 2.8403,
+      "step": 59787
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7725659608840942,
+      "learning_rate": 8.041353052889443e-07,
+      "loss": 3.0116,
+      "step": 59788
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7446717023849487,
+      "learning_rate": 8.030091739033018e-07,
+      "loss": 2.9361,
+      "step": 59789
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7172529697418213,
+      "learning_rate": 8.018838305450026e-07,
+      "loss": 2.827,
+      "step": 59790
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7514923810958862,
+      "learning_rate": 8.007592752169112e-07,
+      "loss": 2.8141,
+      "step": 59791
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7799364924430847,
+      "learning_rate": 7.996355079220585e-07,
+      "loss": 2.9459,
+      "step": 59792
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.801102876663208,
+      "learning_rate": 7.985125286633753e-07,
+      "loss": 2.6798,
+      "step": 59793
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7336550354957581,
+      "learning_rate": 7.973903374438262e-07,
+      "loss": 3.0554,
+      "step": 59794
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7697311043739319,
+      "learning_rate": 7.962689342663753e-07,
+      "loss": 2.8884,
+      "step": 59795
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7298890948295593,
+      "learning_rate": 7.951483191339869e-07,
+      "loss": 2.6709,
+      "step": 59796
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7298159003257751,
+      "learning_rate": 7.940284920495588e-07,
+      "loss": 2.8218,
+      "step": 59797
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7589902281761169,
+      "learning_rate": 7.929094530161217e-07,
+      "loss": 2.8853,
+      "step": 59798
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7164919972419739,
+      "learning_rate": 7.917912020365402e-07,
+      "loss": 2.8939,
+      "step": 59799
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7424035668373108,
+      "learning_rate": 7.906737391137785e-07,
+      "loss": 2.9849,
+      "step": 59800
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7874431014060974,
+      "learning_rate": 7.895570642508342e-07,
+      "loss": 2.6638,
+      "step": 59801
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7526593804359436,
+      "learning_rate": 7.88441177450605e-07,
+      "loss": 2.9343,
+      "step": 59802
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7278794050216675,
+      "learning_rate": 7.873260787160551e-07,
+      "loss": 2.8957,
+      "step": 59803
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7460549473762512,
+      "learning_rate": 7.862117680500823e-07,
+      "loss": 2.8864,
+      "step": 59804
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7620648145675659,
+      "learning_rate": 7.850982454556509e-07,
+      "loss": 2.8314,
+      "step": 59805
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.727797269821167,
+      "learning_rate": 7.839855109356585e-07,
+      "loss": 2.9278,
+      "step": 59806
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7091959714889526,
+      "learning_rate": 7.828735644931028e-07,
+      "loss": 2.9797,
+      "step": 59807
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7799173593521118,
+      "learning_rate": 7.81762406130848e-07,
+      "loss": 2.9589,
+      "step": 59808
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7307581901550293,
+      "learning_rate": 7.806520358518587e-07,
+      "loss": 2.7646,
+      "step": 59809
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8110495805740356,
+      "learning_rate": 7.795424536590322e-07,
+      "loss": 2.9862,
+      "step": 59810
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7516154646873474,
+      "learning_rate": 7.784336595552998e-07,
+      "loss": 3.0308,
+      "step": 59811
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7222862839698792,
+      "learning_rate": 7.773256535436257e-07,
+      "loss": 2.7501,
+      "step": 59812
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7540051937103271,
+      "learning_rate": 7.762184356268408e-07,
+      "loss": 2.8516,
+      "step": 59813
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7463213205337524,
+      "learning_rate": 7.751120058079429e-07,
+      "loss": 2.9547,
+      "step": 59814
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7319959402084351,
+      "learning_rate": 7.740063640897964e-07,
+      "loss": 2.8742,
+      "step": 59815
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7268669605255127,
+      "learning_rate": 7.729015104753322e-07,
+      "loss": 2.9301,
+      "step": 59816
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7571326494216919,
+      "learning_rate": 7.71797444967448e-07,
+      "loss": 2.7898,
+      "step": 59817
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7747343182563782,
+      "learning_rate": 7.706941675690747e-07,
+      "loss": 3.0083,
+      "step": 59818
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7545844912528992,
+      "learning_rate": 7.695916782831102e-07,
+      "loss": 2.9409,
+      "step": 59819
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7999106049537659,
+      "learning_rate": 7.68489977112452e-07,
+      "loss": 2.7857,
+      "step": 59820
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7516934275627136,
+      "learning_rate": 7.673890640599978e-07,
+      "loss": 2.9531,
+      "step": 59821
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.73682701587677,
+      "learning_rate": 7.662889391286453e-07,
+      "loss": 2.9481,
+      "step": 59822
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7407304048538208,
+      "learning_rate": 7.651896023213255e-07,
+      "loss": 2.8385,
+      "step": 59823
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8045844435691833,
+      "learning_rate": 7.640910536408695e-07,
+      "loss": 2.9524,
+      "step": 59824
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7355616092681885,
+      "learning_rate": 7.629932930902416e-07,
+      "loss": 2.7532,
+      "step": 59825
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8066126704216003,
+      "learning_rate": 7.618963206722728e-07,
+      "loss": 2.9467,
+      "step": 59826
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8547759652137756,
+      "learning_rate": 7.608001363899274e-07,
+      "loss": 2.8454,
+      "step": 59827
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.77346271276474,
+      "learning_rate": 7.597047402460033e-07,
+      "loss": 2.6539,
+      "step": 59828
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7750498652458191,
+      "learning_rate": 7.586101322434313e-07,
+      "loss": 2.8954,
+      "step": 59829
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7115117907524109,
+      "learning_rate": 7.575163123850758e-07,
+      "loss": 2.933,
+      "step": 59830
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7660773992538452,
+      "learning_rate": 7.564232806738679e-07,
+      "loss": 3.1174,
+      "step": 59831
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7857538461685181,
+      "learning_rate": 7.553310371126053e-07,
+      "loss": 2.6361,
+      "step": 59832
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.789605438709259,
+      "learning_rate": 7.542395817042523e-07,
+      "loss": 2.6682,
+      "step": 59833
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7484418153762817,
+      "learning_rate": 7.5314891445164e-07,
+      "loss": 2.6815,
+      "step": 59834
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7630303502082825,
+      "learning_rate": 7.520590353575994e-07,
+      "loss": 2.6285,
+      "step": 59835
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8174473643302917,
+      "learning_rate": 7.509699444250949e-07,
+      "loss": 2.7668,
+      "step": 59836
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.791915774345398,
+      "learning_rate": 7.498816416569242e-07,
+      "loss": 2.9532,
+      "step": 59837
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7464105486869812,
+      "learning_rate": 7.487941270559516e-07,
+      "loss": 3.0629,
+      "step": 59838
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7361699938774109,
+      "learning_rate": 7.47707400625075e-07,
+      "loss": 2.7977,
+      "step": 59839
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7796388268470764,
+      "learning_rate": 7.466214623671585e-07,
+      "loss": 2.9098,
+      "step": 59840
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7234344482421875,
+      "learning_rate": 7.455363122850333e-07,
+      "loss": 2.918,
+      "step": 59841
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7764577269554138,
+      "learning_rate": 7.444519503815971e-07,
+      "loss": 2.8202,
+      "step": 59842
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7591822147369385,
+      "learning_rate": 7.43368376659681e-07,
+      "loss": 2.964,
+      "step": 59843
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.6883577704429626,
+      "learning_rate": 7.422855911221159e-07,
+      "loss": 2.7551,
+      "step": 59844
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7411786913871765,
+      "learning_rate": 7.412035937717665e-07,
+      "loss": 2.8426,
+      "step": 59845
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7492934465408325,
+      "learning_rate": 7.4012238461153e-07,
+      "loss": 2.8528,
+      "step": 59846
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7484312057495117,
+      "learning_rate": 7.390419636442047e-07,
+      "loss": 2.831,
+      "step": 59847
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7504763007164001,
+      "learning_rate": 7.379623308726213e-07,
+      "loss": 2.8901,
+      "step": 59848
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.2700088024139404,
+      "learning_rate": 7.368834862997109e-07,
+      "loss": 2.9359,
+      "step": 59849
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7634099125862122,
+      "learning_rate": 7.358054299282045e-07,
+      "loss": 2.8977,
+      "step": 59850
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7338115572929382,
+      "learning_rate": 7.347281617610001e-07,
+      "loss": 2.7591,
+      "step": 59851
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7483226656913757,
+      "learning_rate": 7.336516818009286e-07,
+      "loss": 2.9123,
+      "step": 59852
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7367458343505859,
+      "learning_rate": 7.32575990050821e-07,
+      "loss": 2.9619,
+      "step": 59853
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7470386624336243,
+      "learning_rate": 7.315010865135418e-07,
+      "loss": 2.7604,
+      "step": 59854
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7699258923530579,
+      "learning_rate": 7.304269711918553e-07,
+      "loss": 3.1449,
+      "step": 59855
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7648332118988037,
+      "learning_rate": 7.293536440886594e-07,
+      "loss": 2.8605,
+      "step": 59856
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7431793212890625,
+      "learning_rate": 7.282811052067183e-07,
+      "loss": 3.0201,
+      "step": 59857
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7206470966339111,
+      "learning_rate": 7.272093545489299e-07,
+      "loss": 2.9679,
+      "step": 59858
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7647475600242615,
+      "learning_rate": 7.261383921180253e-07,
+      "loss": 3.2728,
+      "step": 59859
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7028852701187134,
+      "learning_rate": 7.250682179169021e-07,
+      "loss": 2.8486,
+      "step": 59860
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7771363854408264,
+      "learning_rate": 7.239988319483581e-07,
+      "loss": 2.6647,
+      "step": 59861
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.802079975605011,
+      "learning_rate": 7.22930234215191e-07,
+      "loss": 2.7629,
+      "step": 59862
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7468277812004089,
+      "learning_rate": 7.21862424720232e-07,
+      "loss": 2.5919,
+      "step": 59863
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7966805696487427,
+      "learning_rate": 7.207954034663122e-07,
+      "loss": 2.8663,
+      "step": 59864
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7449951767921448,
+      "learning_rate": 7.197291704561959e-07,
+      "loss": 2.8764,
+      "step": 59865
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7162653803825378,
+      "learning_rate": 7.186637256927141e-07,
+      "loss": 2.9887,
+      "step": 59866
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.762417733669281,
+      "learning_rate": 7.175990691786981e-07,
+      "loss": 2.6655,
+      "step": 59867
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7331001162528992,
+      "learning_rate": 7.165352009169123e-07,
+      "loss": 2.7909,
+      "step": 59868
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8558750152587891,
+      "learning_rate": 7.154721209101544e-07,
+      "loss": 2.9628,
+      "step": 59869
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7514426112174988,
+      "learning_rate": 7.144098291612888e-07,
+      "loss": 2.8311,
+      "step": 59870
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7667511701583862,
+      "learning_rate": 7.133483256730465e-07,
+      "loss": 2.8383,
+      "step": 59871
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7563773393630981,
+      "learning_rate": 7.122876104482256e-07,
+      "loss": 2.8637,
+      "step": 59872
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7660837173461914,
+      "learning_rate": 7.112276834896568e-07,
+      "loss": 2.8503,
+      "step": 59873
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7688391804695129,
+      "learning_rate": 7.101685448001382e-07,
+      "loss": 2.9298,
+      "step": 59874
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7841348052024841,
+      "learning_rate": 7.091101943824007e-07,
+      "loss": 3.0863,
+      "step": 59875
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.748285174369812,
+      "learning_rate": 7.080526322392754e-07,
+      "loss": 2.9875,
+      "step": 59876
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7872505784034729,
+      "learning_rate": 7.069958583735602e-07,
+      "loss": 3.2147,
+      "step": 59877
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7570214867591858,
+      "learning_rate": 7.059398727880194e-07,
+      "loss": 3.0622,
+      "step": 59878
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.9395710825920105,
+      "learning_rate": 7.048846754853843e-07,
+      "loss": 3.0173,
+      "step": 59879
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7911810874938965,
+      "learning_rate": 7.038302664685192e-07,
+      "loss": 2.8858,
+      "step": 59880
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7433484792709351,
+      "learning_rate": 7.027766457401551e-07,
+      "loss": 2.9579,
+      "step": 59881
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7412406206130981,
+      "learning_rate": 7.017238133030901e-07,
+      "loss": 2.8247,
+      "step": 59882
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.742946207523346,
+      "learning_rate": 7.006717691600883e-07,
+      "loss": 2.6349,
+      "step": 59883
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7795896530151367,
+      "learning_rate": 6.99620513313881e-07,
+      "loss": 2.8827,
+      "step": 59884
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7590773701667786,
+      "learning_rate": 6.985700457672993e-07,
+      "loss": 2.8935,
+      "step": 59885
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7604755163192749,
+      "learning_rate": 6.975203665230744e-07,
+      "loss": 3.0704,
+      "step": 59886
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7391711473464966,
+      "learning_rate": 6.964714755839707e-07,
+      "loss": 2.9104,
+      "step": 59887
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7480840682983398,
+      "learning_rate": 6.954233729527859e-07,
+      "loss": 2.9565,
+      "step": 59888
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7480906844139099,
+      "learning_rate": 6.943760586322178e-07,
+      "loss": 2.8197,
+      "step": 59889
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7429905533790588,
+      "learning_rate": 6.933295326250976e-07,
+      "loss": 2.9929,
+      "step": 59890
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8021376132965088,
+      "learning_rate": 6.922837949341231e-07,
+      "loss": 2.9372,
+      "step": 59891
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8098527193069458,
+      "learning_rate": 6.912388455620588e-07,
+      "loss": 3.0109,
+      "step": 59892
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7434927225112915,
+      "learning_rate": 6.901946845117023e-07,
+      "loss": 2.8893,
+      "step": 59893
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7502673864364624,
+      "learning_rate": 6.891513117857517e-07,
+      "loss": 2.8706,
+      "step": 59894
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7409077882766724,
+      "learning_rate": 6.881087273869712e-07,
+      "loss": 2.8049,
+      "step": 59895
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7963376045227051,
+      "learning_rate": 6.870669313180921e-07,
+      "loss": 2.9638,
+      "step": 59896
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7674747705459595,
+      "learning_rate": 6.860259235818788e-07,
+      "loss": 2.8135,
+      "step": 59897
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8309979438781738,
+      "learning_rate": 6.849857041810958e-07,
+      "loss": 2.8022,
+      "step": 59898
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7403390407562256,
+      "learning_rate": 6.839462731184409e-07,
+      "loss": 2.8496,
+      "step": 59899
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7429600954055786,
+      "learning_rate": 6.829076303966785e-07,
+      "loss": 3.2261,
+      "step": 59900
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7548747658729553,
+      "learning_rate": 6.818697760185065e-07,
+      "loss": 2.8569,
+      "step": 59901
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7258870005607605,
+      "learning_rate": 6.808327099867228e-07,
+      "loss": 2.9454,
+      "step": 59902
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7824847102165222,
+      "learning_rate": 6.797964323039918e-07,
+      "loss": 3.0032,
+      "step": 59903
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.769087016582489,
+      "learning_rate": 6.787609429730779e-07,
+      "loss": 2.9478,
+      "step": 59904
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7909684777259827,
+      "learning_rate": 6.777262419967122e-07,
+      "loss": 2.8437,
+      "step": 59905
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7738947868347168,
+      "learning_rate": 6.766923293775928e-07,
+      "loss": 2.8174,
+      "step": 59906
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7569680213928223,
+      "learning_rate": 6.756592051184507e-07,
+      "loss": 2.7229,
+      "step": 59907
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7944665551185608,
+      "learning_rate": 6.746268692220502e-07,
+      "loss": 2.8974,
+      "step": 59908
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7685111165046692,
+      "learning_rate": 6.735953216910561e-07,
+      "loss": 2.7728,
+      "step": 59909
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7883465886116028,
+      "learning_rate": 6.725645625281995e-07,
+      "loss": 2.8538,
+      "step": 59910
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7175097465515137,
+      "learning_rate": 6.715345917362113e-07,
+      "loss": 2.7597,
+      "step": 59911
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8032987713813782,
+      "learning_rate": 6.705054093177897e-07,
+      "loss": 2.6333,
+      "step": 59912
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8348165154457092,
+      "learning_rate": 6.694770152756657e-07,
+      "loss": 2.7465,
+      "step": 59913
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.78910893201828,
+      "learning_rate": 6.684494096125037e-07,
+      "loss": 2.8535,
+      "step": 59914
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7422067523002625,
+      "learning_rate": 6.674225923310683e-07,
+      "loss": 2.9404,
+      "step": 59915
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8071993589401245,
+      "learning_rate": 6.66396563434024e-07,
+      "loss": 3.0353,
+      "step": 59916
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7702639698982239,
+      "learning_rate": 6.653713229240686e-07,
+      "loss": 3.0289,
+      "step": 59917
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7507279515266418,
+      "learning_rate": 6.643468708039335e-07,
+      "loss": 2.84,
+      "step": 59918
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7871813178062439,
+      "learning_rate": 6.633232070762828e-07,
+      "loss": 2.923,
+      "step": 59919
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7500841617584229,
+      "learning_rate": 6.62300331743848e-07,
+      "loss": 2.9329,
+      "step": 59920
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7141773104667664,
+      "learning_rate": 6.612782448092935e-07,
+      "loss": 3.0867,
+      "step": 59921
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7696955800056458,
+      "learning_rate": 6.602569462753171e-07,
+      "loss": 2.8553,
+      "step": 59922
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7573153376579285,
+      "learning_rate": 6.5923643614465e-07,
+      "loss": 2.8073,
+      "step": 59923
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7639854550361633,
+      "learning_rate": 6.582167144198902e-07,
+      "loss": 2.7228,
+      "step": 59924
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7519087195396423,
+      "learning_rate": 6.571977811038354e-07,
+      "loss": 2.9471,
+      "step": 59925
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7760326266288757,
+      "learning_rate": 6.561796361990834e-07,
+      "loss": 2.9599,
+      "step": 59926
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.791580855846405,
+      "learning_rate": 6.551622797083322e-07,
+      "loss": 2.843,
+      "step": 59927
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7381553053855896,
+      "learning_rate": 6.541457116342797e-07,
+      "loss": 2.8464,
+      "step": 59928
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7461539506912231,
+      "learning_rate": 6.531299319796234e-07,
+      "loss": 2.7977,
+      "step": 59929
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7161452174186707,
+      "learning_rate": 6.52114940746995e-07,
+      "loss": 3.0583,
+      "step": 59930
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7718291878700256,
+      "learning_rate": 6.51100737939092e-07,
+      "loss": 2.8127,
+      "step": 59931
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8032710552215576,
+      "learning_rate": 6.500873235585458e-07,
+      "loss": 2.8657,
+      "step": 59932
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8120545744895935,
+      "learning_rate": 6.490746976080874e-07,
+      "loss": 2.8896,
+      "step": 59933
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7528083920478821,
+      "learning_rate": 6.480628600903482e-07,
+      "loss": 2.6816,
+      "step": 59934
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7332311868667603,
+      "learning_rate": 6.470518110080259e-07,
+      "loss": 2.7482,
+      "step": 59935
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7682661414146423,
+      "learning_rate": 6.460415503637184e-07,
+      "loss": 2.9757,
+      "step": 59936
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7852498292922974,
+      "learning_rate": 6.45032078160157e-07,
+      "loss": 2.8943,
+      "step": 59937
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7649948596954346,
+      "learning_rate": 6.440233943999396e-07,
+      "loss": 3.1078,
+      "step": 59938
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7823342084884644,
+      "learning_rate": 6.430154990857639e-07,
+      "loss": 2.9817,
+      "step": 59939
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.755197286605835,
+      "learning_rate": 6.420083922202945e-07,
+      "loss": 2.9656,
+      "step": 59940
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7865545153617859,
+      "learning_rate": 6.410020738061294e-07,
+      "loss": 2.8751,
+      "step": 59941
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7866171598434448,
+      "learning_rate": 6.399965438459664e-07,
+      "loss": 2.8668,
+      "step": 59942
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7472968697547913,
+      "learning_rate": 6.389918023424367e-07,
+      "loss": 2.7944,
+      "step": 59943
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8500422835350037,
+      "learning_rate": 6.379878492982049e-07,
+      "loss": 2.9968,
+      "step": 59944
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7453992962837219,
+      "learning_rate": 6.369846847158689e-07,
+      "loss": 2.8663,
+      "step": 59945
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7211142182350159,
+      "learning_rate": 6.359823085981264e-07,
+      "loss": 2.7665,
+      "step": 59946
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7596474885940552,
+      "learning_rate": 6.349807209476088e-07,
+      "loss": 2.7561,
+      "step": 59947
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7471470832824707,
+      "learning_rate": 6.339799217668806e-07,
+      "loss": 2.7884,
+      "step": 59948
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7682774662971497,
+      "learning_rate": 6.329799110586731e-07,
+      "loss": 2.7054,
+      "step": 59949
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7464492321014404,
+      "learning_rate": 6.31980688825584e-07,
+      "loss": 2.7138,
+      "step": 59950
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6935986876487732,
+      "learning_rate": 6.309822550702448e-07,
+      "loss": 2.8231,
+      "step": 59951
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7591943740844727,
+      "learning_rate": 6.299846097952865e-07,
+      "loss": 2.8893,
+      "step": 59952
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7476579546928406,
+      "learning_rate": 6.289877530033072e-07,
+      "loss": 2.9504,
+      "step": 59953
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8381856083869934,
+      "learning_rate": 6.279916846969713e-07,
+      "loss": 2.689,
+      "step": 59954
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.779679536819458,
+      "learning_rate": 6.2699640487891e-07,
+      "loss": 2.6562,
+      "step": 59955
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7561830282211304,
+      "learning_rate": 6.260019135517213e-07,
+      "loss": 2.7303,
+      "step": 59956
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.764441728591919,
+      "learning_rate": 6.250082107180032e-07,
+      "loss": 2.9123,
+      "step": 59957
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7792513370513916,
+      "learning_rate": 6.240152963804202e-07,
+      "loss": 2.6153,
+      "step": 59958
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7713596820831299,
+      "learning_rate": 6.230231705415367e-07,
+      "loss": 3.1653,
+      "step": 59959
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7923070192337036,
+      "learning_rate": 6.220318332040175e-07,
+      "loss": 2.753,
+      "step": 59960
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7460241317749023,
+      "learning_rate": 6.210412843704605e-07,
+      "loss": 2.9271,
+      "step": 59961
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.768636167049408,
+      "learning_rate": 6.2005152404343e-07,
+      "loss": 2.8513,
+      "step": 59962
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7382229566574097,
+      "learning_rate": 6.19062552225591e-07,
+      "loss": 2.8161,
+      "step": 59963
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7440184354782104,
+      "learning_rate": 6.180743689195078e-07,
+      "loss": 2.9522,
+      "step": 59964
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7206867337226868,
+      "learning_rate": 6.170869741278117e-07,
+      "loss": 2.7925,
+      "step": 59965
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7849396467208862,
+      "learning_rate": 6.161003678531006e-07,
+      "loss": 3.0001,
+      "step": 59966
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7822026610374451,
+      "learning_rate": 6.151145500979726e-07,
+      "loss": 3.0031,
+      "step": 59967
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7590682506561279,
+      "learning_rate": 6.141295208649921e-07,
+      "loss": 2.9123,
+      "step": 59968
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8161299824714661,
+      "learning_rate": 6.131452801567571e-07,
+      "loss": 2.7404,
+      "step": 59969
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7368728518486023,
+      "learning_rate": 6.121618279758989e-07,
+      "loss": 2.7845,
+      "step": 59970
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7626436352729797,
+      "learning_rate": 6.111791643250152e-07,
+      "loss": 2.8567,
+      "step": 59971
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7329202890396118,
+      "learning_rate": 6.101972892066376e-07,
+      "loss": 2.7161,
+      "step": 59972
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7684486508369446,
+      "learning_rate": 6.092162026233971e-07,
+      "loss": 2.9794,
+      "step": 59973
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.769387423992157,
+      "learning_rate": 6.082359045778251e-07,
+      "loss": 2.7875,
+      "step": 59974
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7294189929962158,
+      "learning_rate": 6.072563950725862e-07,
+      "loss": 2.8859,
+      "step": 59975
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7302929759025574,
+      "learning_rate": 6.062776741101783e-07,
+      "loss": 2.7874,
+      "step": 59976
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7487363815307617,
+      "learning_rate": 6.052997416932326e-07,
+      "loss": 2.9571,
+      "step": 59977
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7794578075408936,
+      "learning_rate": 6.043225978243138e-07,
+      "loss": 2.9774,
+      "step": 59978
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7639981508255005,
+      "learning_rate": 6.033462425059865e-07,
+      "loss": 2.8534,
+      "step": 59979
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7297922968864441,
+      "learning_rate": 6.023706757408487e-07,
+      "loss": 2.799,
+      "step": 59980
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7623937726020813,
+      "learning_rate": 6.013958975313982e-07,
+      "loss": 2.9303,
+      "step": 59981
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7628301382064819,
+      "learning_rate": 6.004219078802663e-07,
+      "loss": 2.7251,
+      "step": 59982
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7557390928268433,
+      "learning_rate": 5.994487067900178e-07,
+      "loss": 2.8481,
+      "step": 59983
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7709507346153259,
+      "learning_rate": 5.984762942631838e-07,
+      "loss": 2.871,
+      "step": 59984
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.718964159488678,
+      "learning_rate": 5.975046703023623e-07,
+      "loss": 2.7692,
+      "step": 59985
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7481675744056702,
+      "learning_rate": 5.965338349100845e-07,
+      "loss": 2.899,
+      "step": 59986
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7227299809455872,
+      "learning_rate": 5.955637880889153e-07,
+      "loss": 2.7444,
+      "step": 59987
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7137525677680969,
+      "learning_rate": 5.945945298413856e-07,
+      "loss": 2.8117,
+      "step": 59988
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7731427550315857,
+      "learning_rate": 5.936260601700937e-07,
+      "loss": 2.776,
+      "step": 59989
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7331191897392273,
+      "learning_rate": 5.926583790775707e-07,
+      "loss": 3.0075,
+      "step": 59990
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7784517407417297,
+      "learning_rate": 5.916914865663481e-07,
+      "loss": 2.8882,
+      "step": 59991
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.847660481929779,
+      "learning_rate": 5.907253826390234e-07,
+      "loss": 2.9467,
+      "step": 59992
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7578979730606079,
+      "learning_rate": 5.897600672980618e-07,
+      "loss": 2.915,
+      "step": 59993
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.754513144493103,
+      "learning_rate": 5.88795540546061e-07,
+      "loss": 2.7695,
+      "step": 59994
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7934916615486145,
+      "learning_rate": 5.878318023855855e-07,
+      "loss": 2.6653,
+      "step": 59995
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7949754595756531,
+      "learning_rate": 5.868688528191002e-07,
+      "loss": 2.8544,
+      "step": 59996
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7312231659889221,
+      "learning_rate": 5.859066918492028e-07,
+      "loss": 2.908,
+      "step": 59997
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7361305952072144,
+      "learning_rate": 5.849453194783915e-07,
+      "loss": 2.8526,
+      "step": 59998
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7572147250175476,
+      "learning_rate": 5.839847357091975e-07,
+      "loss": 2.8984,
+      "step": 59999
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.793131411075592,
+      "learning_rate": 5.830249405441855e-07,
+      "loss": 2.9666,
+      "step": 60000
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7692931294441223,
+      "learning_rate": 5.820659339858536e-07,
+      "loss": 3.1132,
+      "step": 60001
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7415686249732971,
+      "learning_rate": 5.811077160367328e-07,
+      "loss": 2.9689,
+      "step": 60002
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.704154372215271,
+      "learning_rate": 5.801502866993546e-07,
+      "loss": 2.8167,
+      "step": 60003
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7278680205345154,
+      "learning_rate": 5.791936459762169e-07,
+      "loss": 2.8319,
+      "step": 60004
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7455959320068359,
+      "learning_rate": 5.782377938698845e-07,
+      "loss": 3.024,
+      "step": 60005
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7692423462867737,
+      "learning_rate": 5.77282730382822e-07,
+      "loss": 2.9902,
+      "step": 60006
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7454026937484741,
+      "learning_rate": 5.76328455517594e-07,
+      "loss": 2.8497,
+      "step": 60007
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7537476420402527,
+      "learning_rate": 5.753749692766984e-07,
+      "loss": 2.8039,
+      "step": 60008
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7972829937934875,
+      "learning_rate": 5.744222716626334e-07,
+      "loss": 2.9203,
+      "step": 60009
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8087637424468994,
+      "learning_rate": 5.734703626778969e-07,
+      "loss": 2.9301,
+      "step": 60010
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7381844520568848,
+      "learning_rate": 5.725192423250203e-07,
+      "loss": 2.8868,
+      "step": 60011
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7417834997177124,
+      "learning_rate": 5.715689106065346e-07,
+      "loss": 2.7835,
+      "step": 60012
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7412571310997009,
+      "learning_rate": 5.706193675248716e-07,
+      "loss": 3.0335,
+      "step": 60013
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7355944514274597,
+      "learning_rate": 5.696706130825624e-07,
+      "loss": 2.7241,
+      "step": 60014
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.738745927810669,
+      "learning_rate": 5.687226472821383e-07,
+      "loss": 2.7394,
+      "step": 60015
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6923105716705322,
+      "learning_rate": 5.67775470126064e-07,
+      "loss": 2.9196,
+      "step": 60016
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7854421734809875,
+      "learning_rate": 5.668290816168708e-07,
+      "loss": 2.734,
+      "step": 60017
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8272637128829956,
+      "learning_rate": 5.658834817569901e-07,
+      "loss": 2.9645,
+      "step": 60018
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.72818922996521,
+      "learning_rate": 5.649386705489533e-07,
+      "loss": 2.9149,
+      "step": 60019
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7304298877716064,
+      "learning_rate": 5.63994647995225e-07,
+      "loss": 2.7626,
+      "step": 60020
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8027445673942566,
+      "learning_rate": 5.630514140983366e-07,
+      "loss": 2.9628,
+      "step": 60021
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.719323456287384,
+      "learning_rate": 5.621089688607194e-07,
+      "loss": 2.6681,
+      "step": 60022
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.806006669998169,
+      "learning_rate": 5.611673122849048e-07,
+      "loss": 2.7381,
+      "step": 60023
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7545458078384399,
+      "learning_rate": 5.60226444373324e-07,
+      "loss": 2.8362,
+      "step": 60024
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8001763224601746,
+      "learning_rate": 5.592863651285084e-07,
+      "loss": 3.0595,
+      "step": 60025
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7519499659538269,
+      "learning_rate": 5.583470745528562e-07,
+      "loss": 2.7645,
+      "step": 60026
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7754395604133606,
+      "learning_rate": 5.574085726489319e-07,
+      "loss": 2.7446,
+      "step": 60027
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7979375123977661,
+      "learning_rate": 5.564708594191337e-07,
+      "loss": 3.0058,
+      "step": 60028
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7920042872428894,
+      "learning_rate": 5.555339348659926e-07,
+      "loss": 2.7863,
+      "step": 60029
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7542409300804138,
+      "learning_rate": 5.54597798991907e-07,
+      "loss": 2.8707,
+      "step": 60030
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7596069574356079,
+      "learning_rate": 5.536624517994081e-07,
+      "loss": 2.9896,
+      "step": 60031
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7304089665412903,
+      "learning_rate": 5.527278932909274e-07,
+      "loss": 2.6327,
+      "step": 60032
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7943688631057739,
+      "learning_rate": 5.517941234689294e-07,
+      "loss": 2.9162,
+      "step": 60033
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7383855581283569,
+      "learning_rate": 5.508611423358789e-07,
+      "loss": 2.8657,
+      "step": 60034
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.717839777469635,
+      "learning_rate": 5.499289498942405e-07,
+      "loss": 3.0219,
+      "step": 60035
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8198238015174866,
+      "learning_rate": 5.489975461464458e-07,
+      "loss": 2.9683,
+      "step": 60036
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7249385118484497,
+      "learning_rate": 5.480669310949593e-07,
+      "loss": 2.7602,
+      "step": 60037
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.771135687828064,
+      "learning_rate": 5.471371047422457e-07,
+      "loss": 3.0002,
+      "step": 60038
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.759382426738739,
+      "learning_rate": 5.462080670907365e-07,
+      "loss": 2.7144,
+      "step": 60039
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8611246943473816,
+      "learning_rate": 5.452798181428631e-07,
+      "loss": 2.7942,
+      "step": 60040
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7444459795951843,
+      "learning_rate": 5.443523579011233e-07,
+      "loss": 3.0667,
+      "step": 60041
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7149555683135986,
+      "learning_rate": 5.434256863679154e-07,
+      "loss": 2.9428,
+      "step": 60042
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7191066741943359,
+      "learning_rate": 5.424998035456707e-07,
+      "loss": 2.6826,
+      "step": 60043
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7511018514633179,
+      "learning_rate": 5.415747094368872e-07,
+      "loss": 2.7892,
+      "step": 60044
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7677654027938843,
+      "learning_rate": 5.40650404043963e-07,
+      "loss": 2.8827,
+      "step": 60045
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7885034084320068,
+      "learning_rate": 5.397268873692961e-07,
+      "loss": 2.6797,
+      "step": 60046
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7618846893310547,
+      "learning_rate": 5.388041594153847e-07,
+      "loss": 3.108,
+      "step": 60047
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7315735816955566,
+      "learning_rate": 5.378822201846267e-07,
+      "loss": 2.8573,
+      "step": 60048
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7194553017616272,
+      "learning_rate": 5.369610696794535e-07,
+      "loss": 2.9314,
+      "step": 60049
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.712523341178894,
+      "learning_rate": 5.360407079022966e-07,
+      "loss": 2.717,
+      "step": 60050
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.725117027759552,
+      "learning_rate": 5.35121134855554e-07,
+      "loss": 2.6336,
+      "step": 60051
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7298504114151001,
+      "learning_rate": 5.342023505416903e-07,
+      "loss": 2.9929,
+      "step": 60052
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.2413877248764038,
+      "learning_rate": 5.332843549631038e-07,
+      "loss": 2.911,
+      "step": 60053
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7253249883651733,
+      "learning_rate": 5.323671481222258e-07,
+      "loss": 2.8556,
+      "step": 60054
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7611720561981201,
+      "learning_rate": 5.31450730021421e-07,
+      "loss": 2.8577,
+      "step": 60055
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7636274695396423,
+      "learning_rate": 5.305351006631542e-07,
+      "loss": 2.7728,
+      "step": 60056
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7457626461982727,
+      "learning_rate": 5.296202600498567e-07,
+      "loss": 2.8102,
+      "step": 60057
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7568612098693848,
+      "learning_rate": 5.2870620818386e-07,
+      "loss": 2.9234,
+      "step": 60058
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7343557476997375,
+      "learning_rate": 5.277929450676288e-07,
+      "loss": 2.979,
+      "step": 60059
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7438076734542847,
+      "learning_rate": 5.268804707035946e-07,
+      "loss": 3.08,
+      "step": 60060
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7437184453010559,
+      "learning_rate": 5.259687850940886e-07,
+      "loss": 2.7556,
+      "step": 60061
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7388575077056885,
+      "learning_rate": 5.250578882415423e-07,
+      "loss": 2.9538,
+      "step": 60062
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7527971267700195,
+      "learning_rate": 5.24147780148354e-07,
+      "loss": 2.9749,
+      "step": 60063
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7302408814430237,
+      "learning_rate": 5.232384608169549e-07,
+      "loss": 2.6886,
+      "step": 60064
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7744280695915222,
+      "learning_rate": 5.223299302496764e-07,
+      "loss": 2.6481,
+      "step": 60065
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.72216796875,
+      "learning_rate": 5.2142218844895e-07,
+      "loss": 2.9079,
+      "step": 60066
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7193196415901184,
+      "learning_rate": 5.205152354171738e-07,
+      "loss": 2.98,
+      "step": 60067
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7564364075660706,
+      "learning_rate": 5.196090711567457e-07,
+      "loss": 2.8694,
+      "step": 60068
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7751148343086243,
+      "learning_rate": 5.187036956699975e-07,
+      "loss": 2.8992,
+      "step": 60069
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7742836475372314,
+      "learning_rate": 5.177991089593603e-07,
+      "loss": 2.8661,
+      "step": 60070
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7468364834785461,
+      "learning_rate": 5.16895311027199e-07,
+      "loss": 2.8607,
+      "step": 60071
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7487417459487915,
+      "learning_rate": 5.159923018759115e-07,
+      "loss": 2.8908,
+      "step": 60072
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7274312376976013,
+      "learning_rate": 5.150900815078296e-07,
+      "loss": 2.9371,
+      "step": 60073
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7443316578865051,
+      "learning_rate": 5.141886499253844e-07,
+      "loss": 2.9809,
+      "step": 60074
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7913233041763306,
+      "learning_rate": 5.132880071309408e-07,
+      "loss": 2.9794,
+      "step": 60075
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.763105034828186,
+      "learning_rate": 5.123881531268637e-07,
+      "loss": 2.7934,
+      "step": 60076
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7337465882301331,
+      "learning_rate": 5.114890879155176e-07,
+      "loss": 2.7711,
+      "step": 60077
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.785071849822998,
+      "learning_rate": 5.105908114992341e-07,
+      "loss": 2.9069,
+      "step": 60078
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7351061701774597,
+      "learning_rate": 5.096933238804446e-07,
+      "loss": 2.5839,
+      "step": 60079
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7270272970199585,
+      "learning_rate": 5.087966250614806e-07,
+      "loss": 2.8432,
+      "step": 60080
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8075821399688721,
+      "learning_rate": 5.079007150447067e-07,
+      "loss": 2.7532,
+      "step": 60081
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7147870659828186,
+      "learning_rate": 5.07005593832488e-07,
+      "loss": 3.1726,
+      "step": 60082
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.772592306137085,
+      "learning_rate": 5.06111261427189e-07,
+      "loss": 3.0678,
+      "step": 60083
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7498898506164551,
+      "learning_rate": 5.052177178311412e-07,
+      "loss": 2.8506,
+      "step": 60084
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7638654708862305,
+      "learning_rate": 5.043249630467094e-07,
+      "loss": 3.0761,
+      "step": 60085
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7409616112709045,
+      "learning_rate": 5.034329970762586e-07,
+      "loss": 2.9633,
+      "step": 60086
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7529875040054321,
+      "learning_rate": 5.025418199220866e-07,
+      "loss": 2.938,
+      "step": 60087
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7796822190284729,
+      "learning_rate": 5.01651431586625e-07,
+      "loss": 2.8388,
+      "step": 60088
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7700410485267639,
+      "learning_rate": 5.007618320721718e-07,
+      "loss": 2.8562,
+      "step": 60089
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7736303210258484,
+      "learning_rate": 4.998730213810587e-07,
+      "loss": 2.8441,
+      "step": 60090
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7461416721343994,
+      "learning_rate": 4.989849995156503e-07,
+      "loss": 3.0117,
+      "step": 60091
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7124243974685669,
+      "learning_rate": 4.980977664782448e-07,
+      "loss": 3.0132,
+      "step": 60092
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7641018629074097,
+      "learning_rate": 4.972113222712404e-07,
+      "loss": 3.0398,
+      "step": 60093
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.763166069984436,
+      "learning_rate": 4.963256668969351e-07,
+      "loss": 2.9907,
+      "step": 60094
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7739465236663818,
+      "learning_rate": 4.954408003576271e-07,
+      "loss": 2.6657,
+      "step": 60095
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7320348024368286,
+      "learning_rate": 4.945567226557478e-07,
+      "loss": 2.9288,
+      "step": 60096
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8021144270896912,
+      "learning_rate": 4.936734337935289e-07,
+      "loss": 2.7831,
+      "step": 60097
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.6665416359901428,
+      "learning_rate": 4.927909337733349e-07,
+      "loss": 2.8185,
+      "step": 60098
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.737612783908844,
+      "learning_rate": 4.919092225975307e-07,
+      "loss": 2.7675,
+      "step": 60099
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7620866894721985,
+      "learning_rate": 4.91028300268348e-07,
+      "loss": 2.9269,
+      "step": 60100
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7583200335502625,
+      "learning_rate": 4.901481667881846e-07,
+      "loss": 2.9917,
+      "step": 60101
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7368113398551941,
+      "learning_rate": 4.892688221593055e-07,
+      "loss": 2.9468,
+      "step": 60102
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.754863440990448,
+      "learning_rate": 4.883902663840422e-07,
+      "loss": 2.7378,
+      "step": 60103
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.793886125087738,
+      "learning_rate": 4.875124994647595e-07,
+      "loss": 3.0714,
+      "step": 60104
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7754970192909241,
+      "learning_rate": 4.866355214036887e-07,
+      "loss": 2.759,
+      "step": 60105
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7628046870231628,
+      "learning_rate": 4.85759332203195e-07,
+      "loss": 2.9111,
+      "step": 60106
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7230570912361145,
+      "learning_rate": 4.848839318655429e-07,
+      "loss": 2.9432,
+      "step": 60107
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7568901777267456,
+      "learning_rate": 4.840093203930973e-07,
+      "loss": 2.9762,
+      "step": 60108
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.772777259349823,
+      "learning_rate": 4.83135497788123e-07,
+      "loss": 3.0028,
+      "step": 60109
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7773007750511169,
+      "learning_rate": 4.822624640529182e-07,
+      "loss": 2.7886,
+      "step": 60110
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7324758768081665,
+      "learning_rate": 4.813902191897812e-07,
+      "loss": 2.8291,
+      "step": 60111
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7719749808311462,
+      "learning_rate": 4.805187632010432e-07,
+      "loss": 2.7614,
+      "step": 60112
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7728280425071716,
+      "learning_rate": 4.796480960889359e-07,
+      "loss": 2.9761,
+      "step": 60113
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7225713133811951,
+      "learning_rate": 4.787782178558241e-07,
+      "loss": 2.757,
+      "step": 60114
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7266071438789368,
+      "learning_rate": 4.779091285039726e-07,
+      "loss": 3.0808,
+      "step": 60115
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7436891198158264,
+      "learning_rate": 4.770408280356464e-07,
+      "loss": 2.7847,
+      "step": 60116
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7610815167427063,
+      "learning_rate": 4.761733164531434e-07,
+      "loss": 3.0614,
+      "step": 60117
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7271731495857239,
+      "learning_rate": 4.753065937587619e-07,
+      "loss": 2.7171,
+      "step": 60118
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7084574103355408,
+      "learning_rate": 4.7444065995480006e-07,
+      "loss": 2.8051,
+      "step": 60119
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7272463440895081,
+      "learning_rate": 4.7357551504348944e-07,
+      "loss": 2.9871,
+      "step": 60120
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7564078569412231,
+      "learning_rate": 4.727111590271615e-07,
+      "loss": 3.0835,
+      "step": 60121
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7226554155349731,
+      "learning_rate": 4.718475919080478e-07,
+      "loss": 3.0259,
+      "step": 60122
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.747903048992157,
+      "learning_rate": 4.7098481368844645e-07,
+      "loss": 2.9517,
+      "step": 60123
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7668612599372864,
+      "learning_rate": 4.7012282437062234e-07,
+      "loss": 2.9348,
+      "step": 60124
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7929732203483582,
+      "learning_rate": 4.692616239568403e-07,
+      "loss": 3.0129,
+      "step": 60125
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7587389945983887,
+      "learning_rate": 4.6840121244939856e-07,
+      "loss": 2.7101,
+      "step": 60126
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7391680479049683,
+      "learning_rate": 4.675415898505286e-07,
+      "loss": 2.7574,
+      "step": 60127
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.827909529209137,
+      "learning_rate": 4.6668275616249526e-07,
+      "loss": 2.8657,
+      "step": 60128
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7175806164741516,
+      "learning_rate": 4.658247113875968e-07,
+      "loss": 2.9622,
+      "step": 60129
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7418898940086365,
+      "learning_rate": 4.649674555280647e-07,
+      "loss": 3.021,
+      "step": 60130
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7666599154472351,
+      "learning_rate": 4.641109885861638e-07,
+      "loss": 2.723,
+      "step": 60131
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.775105357170105,
+      "learning_rate": 4.632553105641257e-07,
+      "loss": 2.871,
+      "step": 60132
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7589561939239502,
+      "learning_rate": 4.624004214642485e-07,
+      "loss": 2.9596,
+      "step": 60133
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7122417092323303,
+      "learning_rate": 4.615463212887638e-07,
+      "loss": 2.7842,
+      "step": 60134
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7434190511703491,
+      "learning_rate": 4.6069301003990313e-07,
+      "loss": 2.9867,
+      "step": 60135
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8169651627540588,
+      "learning_rate": 4.598404877199646e-07,
+      "loss": 2.8957,
+      "step": 60136
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7982842922210693,
+      "learning_rate": 4.5898875433111327e-07,
+      "loss": 2.9529,
+      "step": 60137
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7457339763641357,
+      "learning_rate": 4.581378098756472e-07,
+      "loss": 2.8845,
+      "step": 60138
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7173131704330444,
+      "learning_rate": 4.5728765435583124e-07,
+      "loss": 2.9729,
+      "step": 60139
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.791340708732605,
+      "learning_rate": 4.564382877738304e-07,
+      "loss": 2.6271,
+      "step": 60140
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7958931922912598,
+      "learning_rate": 4.5558971013194277e-07,
+      "loss": 2.7756,
+      "step": 60141
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7099641561508179,
+      "learning_rate": 4.547419214324e-07,
+      "loss": 2.8837,
+      "step": 60142
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7637771368026733,
+      "learning_rate": 4.5389492167736685e-07,
+      "loss": 2.9816,
+      "step": 60143
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7536028027534485,
+      "learning_rate": 4.5304871086917495e-07,
+      "loss": 2.8259,
+      "step": 60144
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8726342916488647,
+      "learning_rate": 4.5220328900998917e-07,
+      "loss": 3.1647,
+      "step": 60145
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7555391192436218,
+      "learning_rate": 4.5135865610204103e-07,
+      "loss": 2.7219,
+      "step": 60146
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7423574328422546,
+      "learning_rate": 4.5051481214759545e-07,
+      "loss": 2.8423,
+      "step": 60147
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7736955285072327,
+      "learning_rate": 4.4967175714881733e-07,
+      "loss": 2.8912,
+      "step": 60148
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7960270643234253,
+      "learning_rate": 4.488294911079382e-07,
+      "loss": 3.2692,
+      "step": 60149
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7632052898406982,
+      "learning_rate": 4.47988014027223e-07,
+      "loss": 2.9191,
+      "step": 60150
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7297966480255127,
+      "learning_rate": 4.471473259088365e-07,
+      "loss": 2.8145,
+      "step": 60151
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.707733154296875,
+      "learning_rate": 4.463074267550104e-07,
+      "loss": 2.9402,
+      "step": 60152
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.741661012172699,
+      "learning_rate": 4.454683165679762e-07,
+      "loss": 2.7414,
+      "step": 60153
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7709798812866211,
+      "learning_rate": 4.4462999534993214e-07,
+      "loss": 2.7895,
+      "step": 60154
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7329643964767456,
+      "learning_rate": 4.4379246310307646e-07,
+      "loss": 2.8292,
+      "step": 60155
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7478441596031189,
+      "learning_rate": 4.4295571982964074e-07,
+      "loss": 3.0081,
+      "step": 60156
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7324329614639282,
+      "learning_rate": 4.421197655317565e-07,
+      "loss": 2.9113,
+      "step": 60157
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7825899720191956,
+      "learning_rate": 4.412846002117221e-07,
+      "loss": 2.8504,
+      "step": 60158
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7741754055023193,
+      "learning_rate": 4.4045022387166897e-07,
+      "loss": 2.8648,
+      "step": 60159
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7594071626663208,
+      "learning_rate": 4.396166365138287e-07,
+      "loss": 3.0879,
+      "step": 60160
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7644873261451721,
+      "learning_rate": 4.3878383814039964e-07,
+      "loss": 2.9545,
+      "step": 60161
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7412201762199402,
+      "learning_rate": 4.379518287535466e-07,
+      "loss": 2.967,
+      "step": 60162
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7475193738937378,
+      "learning_rate": 4.3712060835550123e-07,
+      "loss": 2.8789,
+      "step": 60163
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7200668454170227,
+      "learning_rate": 4.3629017694839505e-07,
+      "loss": 2.7571,
+      "step": 60164
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7569831013679504,
+      "learning_rate": 4.354605345344597e-07,
+      "loss": 2.8523,
+      "step": 60165
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7711689472198486,
+      "learning_rate": 4.346316811158601e-07,
+      "loss": 2.9118,
+      "step": 60166
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7586641311645508,
+      "learning_rate": 4.3380361669479447e-07,
+      "loss": 2.9045,
+      "step": 60167
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7284442782402039,
+      "learning_rate": 4.3297634127346106e-07,
+      "loss": 2.7352,
+      "step": 60168
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8342050909996033,
+      "learning_rate": 4.3214985485399144e-07,
+      "loss": 2.7207,
+      "step": 60169
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.749860405921936,
+      "learning_rate": 4.3132415743858394e-07,
+      "loss": 2.7941,
+      "step": 60170
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7597050070762634,
+      "learning_rate": 4.3049924902940345e-07,
+      "loss": 2.8716,
+      "step": 60171
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7795093059539795,
+      "learning_rate": 4.2967512962864825e-07,
+      "loss": 2.7667,
+      "step": 60172
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7680733799934387,
+      "learning_rate": 4.2885179923848323e-07,
+      "loss": 2.8817,
+      "step": 60173
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.780337929725647,
+      "learning_rate": 4.280292578610733e-07,
+      "loss": 2.7562,
+      "step": 60174
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7037748694419861,
+      "learning_rate": 4.272075054985835e-07,
+      "loss": 2.8465,
+      "step": 60175
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7473371028900146,
+      "learning_rate": 4.263865421531454e-07,
+      "loss": 3.058,
+      "step": 60176
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.766514778137207,
+      "learning_rate": 4.255663678269905e-07,
+      "loss": 3.0962,
+      "step": 60177
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.705461859703064,
+      "learning_rate": 4.2474698252221715e-07,
+      "loss": 2.9497,
+      "step": 60178
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7880750298500061,
+      "learning_rate": 4.2392838624099033e-07,
+      "loss": 2.8343,
+      "step": 60179
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8290342092514038,
+      "learning_rate": 4.231105789855083e-07,
+      "loss": 2.8913,
+      "step": 60180
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8501545190811157,
+      "learning_rate": 4.2229356075786923e-07,
+      "loss": 2.9309,
+      "step": 60181
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7908932566642761,
+      "learning_rate": 4.214773315602715e-07,
+      "loss": 2.9298,
+      "step": 60182
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7290691137313843,
+      "learning_rate": 4.206618913948467e-07,
+      "loss": 2.7574,
+      "step": 60183
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7517605423927307,
+      "learning_rate": 4.198472402637598e-07,
+      "loss": 2.9407,
+      "step": 60184
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7302461862564087,
+      "learning_rate": 4.190333781691424e-07,
+      "loss": 2.9164,
+      "step": 60185
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7860396504402161,
+      "learning_rate": 4.182203051130928e-07,
+      "loss": 2.8992,
+      "step": 60186
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7854883670806885,
+      "learning_rate": 4.174080210978426e-07,
+      "loss": 3.107,
+      "step": 60187
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7467091083526611,
+      "learning_rate": 4.165965261254567e-07,
+      "loss": 2.832,
+      "step": 60188
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7421848773956299,
+      "learning_rate": 4.157858201981001e-07,
+      "loss": 2.874,
+      "step": 60189
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7487695217132568,
+      "learning_rate": 4.1497590331790453e-07,
+      "loss": 2.6692,
+      "step": 60190
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7340686321258545,
+      "learning_rate": 4.141667754870348e-07,
+      "loss": 2.8941,
+      "step": 60191
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7664749622344971,
+      "learning_rate": 4.13358436707556e-07,
+      "loss": 2.8168,
+      "step": 60192
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7620031833648682,
+      "learning_rate": 4.1255088698166625e-07,
+      "loss": 3.0271,
+      "step": 60193
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7645248174667358,
+      "learning_rate": 4.117441263114307e-07,
+      "loss": 2.8387,
+      "step": 60194
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7493310570716858,
+      "learning_rate": 4.1093815469901423e-07,
+      "loss": 2.8177,
+      "step": 60195
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7595289349555969,
+      "learning_rate": 4.101329721465485e-07,
+      "loss": 3.058,
+      "step": 60196
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7337210774421692,
+      "learning_rate": 4.093285786560985e-07,
+      "loss": 2.8429,
+      "step": 60197
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7361249327659607,
+      "learning_rate": 4.085249742298624e-07,
+      "loss": 2.9273,
+      "step": 60198
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7934818863868713,
+      "learning_rate": 4.077221588699053e-07,
+      "loss": 2.8338,
+      "step": 60199
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8189910650253296,
+      "learning_rate": 4.0692013257832556e-07,
+      "loss": 2.6777,
+      "step": 60200
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.777704119682312,
+      "learning_rate": 4.0611889535728803e-07,
+      "loss": 2.9845,
+      "step": 60201
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7534101605415344,
+      "learning_rate": 4.0531844720885774e-07,
+      "loss": 3.0629,
+      "step": 60202
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7779284119606018,
+      "learning_rate": 4.0451878813516636e-07,
+      "loss": 2.8526,
+      "step": 60203
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.72135990858078,
+      "learning_rate": 4.0371991813831215e-07,
+      "loss": 2.9209,
+      "step": 60204
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7906358242034912,
+      "learning_rate": 4.0292183722039353e-07,
+      "loss": 3.1225,
+      "step": 60205
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.6974494457244873,
+      "learning_rate": 4.02124545383542e-07,
+      "loss": 2.844,
+      "step": 60206
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7315931916236877,
+      "learning_rate": 4.013280426297893e-07,
+      "loss": 2.9354,
+      "step": 60207
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.759009838104248,
+      "learning_rate": 4.005323289613338e-07,
+      "loss": 2.8429,
+      "step": 60208
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7365983128547668,
+      "learning_rate": 3.9973740438017376e-07,
+      "loss": 2.7836,
+      "step": 60209
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7780154943466187,
+      "learning_rate": 3.989432688884742e-07,
+      "loss": 2.786,
+      "step": 60210
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7027831673622131,
+      "learning_rate": 3.981499224883e-07,
+      "loss": 2.8762,
+      "step": 60211
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7347671389579773,
+      "learning_rate": 3.9735736518171634e-07,
+      "loss": 2.6765,
+      "step": 60212
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7240535616874695,
+      "learning_rate": 3.965655969708548e-07,
+      "loss": 2.9078,
+      "step": 60213
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7486701011657715,
+      "learning_rate": 3.957746178577803e-07,
+      "loss": 2.8639,
+      "step": 60214
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7389553785324097,
+      "learning_rate": 3.94984427844558e-07,
+      "loss": 2.7024,
+      "step": 60215
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.737718403339386,
+      "learning_rate": 3.941950269333194e-07,
+      "loss": 3.0266,
+      "step": 60216
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7455750703811646,
+      "learning_rate": 3.9340641512606295e-07,
+      "loss": 2.8594,
+      "step": 60217
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7492016553878784,
+      "learning_rate": 3.9261859242495363e-07,
+      "loss": 2.9803,
+      "step": 60218
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7766376733779907,
+      "learning_rate": 3.918315588319898e-07,
+      "loss": 2.68,
+      "step": 60219
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7816788554191589,
+      "learning_rate": 3.910453143493364e-07,
+      "loss": 2.9721,
+      "step": 60220
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7525882720947266,
+      "learning_rate": 3.902598589789585e-07,
+      "loss": 2.6368,
+      "step": 60221
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.746664822101593,
+      "learning_rate": 3.8947519272298777e-07,
+      "loss": 2.7792,
+      "step": 60222
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7470570802688599,
+      "learning_rate": 3.886913155834892e-07,
+      "loss": 3.0567,
+      "step": 60223
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7868690490722656,
+      "learning_rate": 3.8790822756249453e-07,
+      "loss": 2.7597,
+      "step": 60224
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8167101144790649,
+      "learning_rate": 3.87125928662102e-07,
+      "loss": 2.9416,
+      "step": 60225
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7606449127197266,
+      "learning_rate": 3.863444188843767e-07,
+      "loss": 2.9537,
+      "step": 60226
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7369158864021301,
+      "learning_rate": 3.8556369823135036e-07,
+      "loss": 2.8176,
+      "step": 60227
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7974393963813782,
+      "learning_rate": 3.847837667050546e-07,
+      "loss": 2.9026,
+      "step": 60228
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7653765082359314,
+      "learning_rate": 3.8400462430762113e-07,
+      "loss": 2.6229,
+      "step": 60229
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7237482070922852,
+      "learning_rate": 3.8322627104101497e-07,
+      "loss": 2.9362,
+      "step": 60230
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7616258263587952,
+      "learning_rate": 3.8244870690736785e-07,
+      "loss": 2.702,
+      "step": 60231
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.750545859336853,
+      "learning_rate": 3.8167193190864477e-07,
+      "loss": 2.8297,
+      "step": 60232
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7552297711372375,
+      "learning_rate": 3.808959460469774e-07,
+      "loss": 2.7552,
+      "step": 60233
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7343993782997131,
+      "learning_rate": 3.801207493243308e-07,
+      "loss": 2.9365,
+      "step": 60234
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8061990737915039,
+      "learning_rate": 3.793463417428033e-07,
+      "loss": 2.784,
+      "step": 60235
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7554232478141785,
+      "learning_rate": 3.7857272330442667e-07,
+      "loss": 2.8843,
+      "step": 60236
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8099880814552307,
+      "learning_rate": 3.7779989401119925e-07,
+      "loss": 2.8103,
+      "step": 60237
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7447828054428101,
+      "learning_rate": 3.770278538652194e-07,
+      "loss": 2.6748,
+      "step": 60238
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7810490727424622,
+      "learning_rate": 3.7625660286845215e-07,
+      "loss": 2.7598,
+      "step": 60239
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8725793361663818,
+      "learning_rate": 3.754861410229626e-07,
+      "loss": 2.9231,
+      "step": 60240
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7612780332565308,
+      "learning_rate": 3.747164683307824e-07,
+      "loss": 2.8793,
+      "step": 60241
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7492809295654297,
+      "learning_rate": 3.7394758479394326e-07,
+      "loss": 2.8316,
+      "step": 60242
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7240146398544312,
+      "learning_rate": 3.731794904144769e-07,
+      "loss": 3.118,
+      "step": 60243
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7847809791564941,
+      "learning_rate": 3.7241218519434845e-07,
+      "loss": 2.8566,
+      "step": 60244
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7559931874275208,
+      "learning_rate": 3.7164566913565617e-07,
+      "loss": 2.8552,
+      "step": 60245
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7063673734664917,
+      "learning_rate": 3.7087994224036524e-07,
+      "loss": 2.8538,
+      "step": 60246
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7377515435218811,
+      "learning_rate": 3.7011500451054055e-07,
+      "loss": 2.9346,
+      "step": 60247
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7750132083892822,
+      "learning_rate": 3.693508559481473e-07,
+      "loss": 2.8041,
+      "step": 60248
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7176759243011475,
+      "learning_rate": 3.6858749655521714e-07,
+      "loss": 2.7881,
+      "step": 60249
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.834689199924469,
+      "learning_rate": 3.678249263337818e-07,
+      "loss": 2.8782,
+      "step": 60250
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7550405263900757,
+      "learning_rate": 3.6706314528580637e-07,
+      "loss": 3.157,
+      "step": 60251
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7085062861442566,
+      "learning_rate": 3.663021534133226e-07,
+      "loss": 2.9406,
+      "step": 60252
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7482970952987671,
+      "learning_rate": 3.655419507183288e-07,
+      "loss": 3.069,
+      "step": 60253
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7461667656898499,
+      "learning_rate": 3.647825372028568e-07,
+      "loss": 2.9194,
+      "step": 60254
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7867854237556458,
+      "learning_rate": 3.640239128688716e-07,
+      "loss": 2.7486,
+      "step": 60255
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7498422861099243,
+      "learning_rate": 3.6326607771837156e-07,
+      "loss": 2.6522,
+      "step": 60256
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.75998854637146,
+      "learning_rate": 3.625090317533552e-07,
+      "loss": 2.9697,
+      "step": 60257
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7807403802871704,
+      "learning_rate": 3.617527749758542e-07,
+      "loss": 2.9362,
+      "step": 60258
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.773429811000824,
+      "learning_rate": 3.609973073878003e-07,
+      "loss": 2.9117,
+      "step": 60259
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.76775062084198,
+      "learning_rate": 3.602426289912252e-07,
+      "loss": 3.1239,
+      "step": 60260
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7371132969856262,
+      "learning_rate": 3.594887397880941e-07,
+      "loss": 2.8399,
+      "step": 60261
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.747292697429657,
+      "learning_rate": 3.5873563978040534e-07,
+      "loss": 2.7784,
+      "step": 60262
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8072085976600647,
+      "learning_rate": 3.5798332897015727e-07,
+      "loss": 2.8365,
+      "step": 60263
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7212876677513123,
+      "learning_rate": 3.57231807359315e-07,
+      "loss": 2.7879,
+      "step": 60264
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7406131029129028,
+      "learning_rate": 3.564810749498437e-07,
+      "loss": 2.7748,
+      "step": 60265
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7893876433372498,
+      "learning_rate": 3.5573113174374167e-07,
+      "loss": 2.9208,
+      "step": 60266
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7580415606498718,
+      "learning_rate": 3.549819777429408e-07,
+      "loss": 2.708,
+      "step": 60267
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7362975478172302,
+      "learning_rate": 3.5423361294950603e-07,
+      "loss": 2.9144,
+      "step": 60268
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7409021854400635,
+      "learning_rate": 3.5348603736530254e-07,
+      "loss": 2.9298,
+      "step": 60269
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7252013683319092,
+      "learning_rate": 3.527392509923621e-07,
+      "loss": 3.0379,
+      "step": 60270
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8025727272033691,
+      "learning_rate": 3.5199325383264976e-07,
+      "loss": 2.7628,
+      "step": 60271
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.775205135345459,
+      "learning_rate": 3.512480458881306e-07,
+      "loss": 2.9703,
+      "step": 60272
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7527696490287781,
+      "learning_rate": 3.505036271607364e-07,
+      "loss": 2.7959,
+      "step": 60273
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.732417106628418,
+      "learning_rate": 3.4975999765243233e-07,
+      "loss": 2.8526,
+      "step": 60274
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7460353374481201,
+      "learning_rate": 3.490171573652167e-07,
+      "loss": 2.827,
+      "step": 60275
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.785345733165741,
+      "learning_rate": 3.482751063010214e-07,
+      "loss": 2.8966,
+      "step": 60276
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7670798301696777,
+      "learning_rate": 3.4753384446177813e-07,
+      "loss": 3.0136,
+      "step": 60277
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7101062536239624,
+      "learning_rate": 3.467933718494853e-07,
+      "loss": 2.7194,
+      "step": 60278
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7125613689422607,
+      "learning_rate": 3.4605368846607473e-07,
+      "loss": 2.9547,
+      "step": 60279
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7298769950866699,
+      "learning_rate": 3.453147943134782e-07,
+      "loss": 2.9709,
+      "step": 60280
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7311604619026184,
+      "learning_rate": 3.4457668939362747e-07,
+      "loss": 2.7934,
+      "step": 60281
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7600909471511841,
+      "learning_rate": 3.43839373708521e-07,
+      "loss": 3.0138,
+      "step": 60282
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.9641423225402832,
+      "learning_rate": 3.431028472600572e-07,
+      "loss": 2.7382,
+      "step": 60283
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7784249186515808,
+      "learning_rate": 3.423671100502012e-07,
+      "loss": 2.7052,
+      "step": 60284
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7559043765068054,
+      "learning_rate": 3.4163216208088485e-07,
+      "loss": 2.7005,
+      "step": 60285
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7473515272140503,
+      "learning_rate": 3.408980033540398e-07,
+      "loss": 2.9085,
+      "step": 60286
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7658305168151855,
+      "learning_rate": 3.401646338715647e-07,
+      "loss": 2.7508,
+      "step": 60287
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7382382154464722,
+      "learning_rate": 3.394320536354578e-07,
+      "loss": 2.912,
+      "step": 60288
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.762175440788269,
+      "learning_rate": 3.387002626476176e-07,
+      "loss": 3.04,
+      "step": 60289
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7681100964546204,
+      "learning_rate": 3.379692609099427e-07,
+      "loss": 2.9685,
+      "step": 60290
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7338895797729492,
+      "learning_rate": 3.372390484244314e-07,
+      "loss": 2.8639,
+      "step": 60291
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7698550224304199,
+      "learning_rate": 3.3650962519291557e-07,
+      "loss": 2.8677,
+      "step": 60292
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7156473398208618,
+      "learning_rate": 3.3578099121739365e-07,
+      "loss": 2.9148,
+      "step": 60293
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7647261023521423,
+      "learning_rate": 3.350531464997308e-07,
+      "loss": 2.7391,
+      "step": 60294
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7407509088516235,
+      "learning_rate": 3.3432609104189214e-07,
+      "loss": 2.776,
+      "step": 60295
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.713916540145874,
+      "learning_rate": 3.335998248457428e-07,
+      "loss": 2.6472,
+      "step": 60296
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7673091888427734,
+      "learning_rate": 3.328743479132145e-07,
+      "loss": 2.8011,
+      "step": 60297
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7312276363372803,
+      "learning_rate": 3.3214966024623926e-07,
+      "loss": 2.7984,
+      "step": 60298
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7208895087242126,
+      "learning_rate": 3.3142576184668204e-07,
+      "loss": 2.8252,
+      "step": 60299
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7697820663452148,
+      "learning_rate": 3.30702652716508e-07,
+      "loss": 2.7858,
+      "step": 60300
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7742111682891846,
+      "learning_rate": 3.299803328575823e-07,
+      "loss": 2.994,
+      "step": 60301
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7125436663627625,
+      "learning_rate": 3.2925880227183677e-07,
+      "loss": 2.9656,
+      "step": 60302
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7392163872718811,
+      "learning_rate": 3.2853806096113655e-07,
+      "loss": 2.8482,
+      "step": 60303
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.77128005027771,
+      "learning_rate": 3.2781810892738016e-07,
+      "loss": 3.0908,
+      "step": 60304
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7209950685501099,
+      "learning_rate": 3.27098946172466e-07,
+      "loss": 2.9793,
+      "step": 60305
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7680009007453918,
+      "learning_rate": 3.2638057269832594e-07,
+      "loss": 2.9299,
+      "step": 60306
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7300169467926025,
+      "learning_rate": 3.2566298850682513e-07,
+      "loss": 2.9231,
+      "step": 60307
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7227309942245483,
+      "learning_rate": 3.249461935998621e-07,
+      "loss": 2.7252,
+      "step": 60308
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7580020427703857,
+      "learning_rate": 3.242301879793019e-07,
+      "loss": 2.8591,
+      "step": 60309
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7477824687957764,
+      "learning_rate": 3.235149716470431e-07,
+      "loss": 2.9945,
+      "step": 60310
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7373795509338379,
+      "learning_rate": 3.2280054460498414e-07,
+      "loss": 2.882,
+      "step": 60311
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.766094982624054,
+      "learning_rate": 3.22086906854957e-07,
+      "loss": 2.8065,
+      "step": 60312
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7640054225921631,
+      "learning_rate": 3.213740583989266e-07,
+      "loss": 2.9561,
+      "step": 60313
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7449299097061157,
+      "learning_rate": 3.206619992386916e-07,
+      "loss": 3.0813,
+      "step": 60314
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7240998148918152,
+      "learning_rate": 3.1995072937618385e-07,
+      "loss": 3.1171,
+      "step": 60315
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.756550133228302,
+      "learning_rate": 3.1924024881323505e-07,
+      "loss": 2.882,
+      "step": 60316
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7245644330978394,
+      "learning_rate": 3.1853055755174386e-07,
+      "loss": 3.0473,
+      "step": 60317
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7418863773345947,
+      "learning_rate": 3.17821655593542e-07,
+      "loss": 2.9368,
+      "step": 60318
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7780527472496033,
+      "learning_rate": 3.1711354294056137e-07,
+      "loss": 2.8981,
+      "step": 60319
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.6818873286247253,
+      "learning_rate": 3.1640621959460046e-07,
+      "loss": 2.7333,
+      "step": 60320
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7510486245155334,
+      "learning_rate": 3.1569968555755777e-07,
+      "loss": 2.8713,
+      "step": 60321
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7911639213562012,
+      "learning_rate": 3.1499394083129845e-07,
+      "loss": 2.9481,
+      "step": 60322
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.779403030872345,
+      "learning_rate": 3.1428898541765444e-07,
+      "loss": 2.9089,
+      "step": 60323
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7606043815612793,
+      "learning_rate": 3.135848193184909e-07,
+      "loss": 2.7561,
+      "step": 60324
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7843817472457886,
+      "learning_rate": 3.128814425357062e-07,
+      "loss": 2.6959,
+      "step": 60325
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7207468152046204,
+      "learning_rate": 3.1217885507106576e-07,
+      "loss": 2.6606,
+      "step": 60326
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.753822386264801,
+      "learning_rate": 3.1147705692650126e-07,
+      "loss": 2.8586,
+      "step": 60327
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7353950142860413,
+      "learning_rate": 3.107760481038113e-07,
+      "loss": 2.9383,
+      "step": 60328
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7305023074150085,
+      "learning_rate": 3.1007582860489433e-07,
+      "loss": 2.8052,
+      "step": 60329
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.716469943523407,
+      "learning_rate": 3.0937639843154893e-07,
+      "loss": 2.8077,
+      "step": 60330
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.78750079870224,
+      "learning_rate": 3.0867775758560697e-07,
+      "loss": 2.8025,
+      "step": 60331
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7497711777687073,
+      "learning_rate": 3.079799060689336e-07,
+      "loss": 3.0581,
+      "step": 60332
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7364689111709595,
+      "learning_rate": 3.072828438833941e-07,
+      "loss": 2.6442,
+      "step": 60333
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8130603432655334,
+      "learning_rate": 3.065865710307536e-07,
+      "loss": 3.0189,
+      "step": 60334
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7645628452301025,
+      "learning_rate": 3.058910875129106e-07,
+      "loss": 2.8934,
+      "step": 60335
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.759429395198822,
+      "learning_rate": 3.051963933316637e-07,
+      "loss": 2.8775,
+      "step": 60336
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7373213171958923,
+      "learning_rate": 3.04502488488878e-07,
+      "loss": 2.9541,
+      "step": 60337
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7441393136978149,
+      "learning_rate": 3.038093729863522e-07,
+      "loss": 2.9521,
+      "step": 60338
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7805241346359253,
+      "learning_rate": 3.0311704682588477e-07,
+      "loss": 2.8231,
+      "step": 60339
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7624837160110474,
+      "learning_rate": 3.024255100093409e-07,
+      "loss": 2.7851,
+      "step": 60340
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.800739049911499,
+      "learning_rate": 3.0173476253851915e-07,
+      "loss": 2.8968,
+      "step": 60341
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7413939833641052,
+      "learning_rate": 3.0104480441528465e-07,
+      "loss": 2.7343,
+      "step": 60342
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7429508566856384,
+      "learning_rate": 3.0035563564140275e-07,
+      "loss": 3.1652,
+      "step": 60343
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7063744068145752,
+      "learning_rate": 2.9966725621870527e-07,
+      "loss": 2.7918,
+      "step": 60344
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8287339806556702,
+      "learning_rate": 2.989796661489907e-07,
+      "loss": 2.8146,
+      "step": 60345
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7384853959083557,
+      "learning_rate": 2.98292865434091e-07,
+      "loss": 3.0369,
+      "step": 60346
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7586947679519653,
+      "learning_rate": 2.976068540758381e-07,
+      "loss": 3.0617,
+      "step": 60347
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7649457454681396,
+      "learning_rate": 2.9692163207596375e-07,
+      "loss": 2.9689,
+      "step": 60348
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8336549401283264,
+      "learning_rate": 2.9623719943633326e-07,
+      "loss": 3.0257,
+      "step": 60349
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.767746090888977,
+      "learning_rate": 2.955535561587452e-07,
+      "loss": 2.9512,
+      "step": 60350
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.737592339515686,
+      "learning_rate": 2.948707022449981e-07,
+      "loss": 3.0337,
+      "step": 60351
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7662484049797058,
+      "learning_rate": 2.941886376968905e-07,
+      "loss": 3.0496,
+      "step": 60352
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7627364993095398,
+      "learning_rate": 2.9350736251618765e-07,
+      "loss": 2.8868,
+      "step": 60353
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7466115951538086,
+      "learning_rate": 2.9282687670468817e-07,
+      "loss": 3.1975,
+      "step": 60354
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7700613141059875,
+      "learning_rate": 2.921471802642572e-07,
+      "loss": 2.9977,
+      "step": 60355
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7370407581329346,
+      "learning_rate": 2.9146827319659337e-07,
+      "loss": 2.7743,
+      "step": 60356
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7551794648170471,
+      "learning_rate": 2.907901555035286e-07,
+      "loss": 2.8775,
+      "step": 60357
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7633122801780701,
+      "learning_rate": 2.90112827186828e-07,
+      "loss": 3.004,
+      "step": 60358
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7845262289047241,
+      "learning_rate": 2.8943628824832365e-07,
+      "loss": 2.8964,
+      "step": 60359
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7491711378097534,
+      "learning_rate": 2.8876053868974734e-07,
+      "loss": 2.9076,
+      "step": 60360
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7596331238746643,
+      "learning_rate": 2.880855785128644e-07,
+      "loss": 2.9886,
+      "step": 60361
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7640942931175232,
+      "learning_rate": 2.8741140771953996e-07,
+      "loss": 2.879,
+      "step": 60362
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7326937317848206,
+      "learning_rate": 2.867380263114394e-07,
+      "loss": 3.1048,
+      "step": 60363
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7499207854270935,
+      "learning_rate": 2.860654342904278e-07,
+      "loss": 2.9645,
+      "step": 60364
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8236726522445679,
+      "learning_rate": 2.853936316582039e-07,
+      "loss": 2.878,
+      "step": 60365
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7456938624382019,
+      "learning_rate": 2.847226184165996e-07,
+      "loss": 2.9094,
+      "step": 60366
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7293276786804199,
+      "learning_rate": 2.8405239456734675e-07,
+      "loss": 2.7994,
+      "step": 60367
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7146532535552979,
+      "learning_rate": 2.833829601122439e-07,
+      "loss": 2.8819,
+      "step": 60368
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.784432053565979,
+      "learning_rate": 2.827143150529898e-07,
+      "loss": 2.6586,
+      "step": 60369
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7396759986877441,
+      "learning_rate": 2.8204645939141623e-07,
+      "loss": 2.786,
+      "step": 60370
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7426009774208069,
+      "learning_rate": 2.813793931292219e-07,
+      "loss": 2.9405,
+      "step": 60371
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7698811292648315,
+      "learning_rate": 2.8071311626820526e-07,
+      "loss": 2.6749,
+      "step": 60372
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8511273860931396,
+      "learning_rate": 2.8004762881009833e-07,
+      "loss": 2.877,
+      "step": 60373
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7467965483665466,
+      "learning_rate": 2.793829307566664e-07,
+      "loss": 2.8784,
+      "step": 60374
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7436198592185974,
+      "learning_rate": 2.7871902210967467e-07,
+      "loss": 3.0372,
+      "step": 60375
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7805924415588379,
+      "learning_rate": 2.780559028708551e-07,
+      "loss": 3.0216,
+      "step": 60376
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7895100712776184,
+      "learning_rate": 2.7739357304193963e-07,
+      "loss": 2.9302,
+      "step": 60377
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7379295229911804,
+      "learning_rate": 2.7673203262469355e-07,
+      "loss": 2.8006,
+      "step": 60378
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7495266199111938,
+      "learning_rate": 2.760712816208155e-07,
+      "loss": 2.8561,
+      "step": 60379
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7267303466796875,
+      "learning_rate": 2.754113200321373e-07,
+      "loss": 2.9655,
+      "step": 60380
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.743676483631134,
+      "learning_rate": 2.7475214786032427e-07,
+      "loss": 2.9309,
+      "step": 60381
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7842069864273071,
+      "learning_rate": 2.7409376510710847e-07,
+      "loss": 2.9816,
+      "step": 60382
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7070533037185669,
+      "learning_rate": 2.7343617177428835e-07,
+      "loss": 2.7144,
+      "step": 60383
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7402557134628296,
+      "learning_rate": 2.7277936786352925e-07,
+      "loss": 3.0865,
+      "step": 60384
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7583931684494019,
+      "learning_rate": 2.7212335337656323e-07,
+      "loss": 2.7086,
+      "step": 60385
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7335100173950195,
+      "learning_rate": 2.714681283151554e-07,
+      "loss": 2.8412,
+      "step": 60386
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7490924596786499,
+      "learning_rate": 2.708136926810378e-07,
+      "loss": 2.7074,
+      "step": 60387
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7380550503730774,
+      "learning_rate": 2.7016004647590905e-07,
+      "loss": 2.6201,
+      "step": 60388
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7886814475059509,
+      "learning_rate": 2.695071897014678e-07,
+      "loss": 3.0093,
+      "step": 60389
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7457057237625122,
+      "learning_rate": 2.688551223594793e-07,
+      "loss": 2.8093,
+      "step": 60390
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7019795179367065,
+      "learning_rate": 2.682038444516421e-07,
+      "loss": 2.7947,
+      "step": 60391
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7453329563140869,
+      "learning_rate": 2.675533559796883e-07,
+      "loss": 2.8498,
+      "step": 60392
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7653692364692688,
+      "learning_rate": 2.6690365694528317e-07,
+      "loss": 2.88,
+      "step": 60393
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7203850746154785,
+      "learning_rate": 2.662547473501919e-07,
+      "loss": 2.7761,
+      "step": 60394
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7470075488090515,
+      "learning_rate": 2.656066271960799e-07,
+      "loss": 2.8994,
+      "step": 60395
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.765425980091095,
+      "learning_rate": 2.649592964846792e-07,
+      "loss": 2.7976,
+      "step": 60396
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7347993850708008,
+      "learning_rate": 2.6431275521772156e-07,
+      "loss": 2.7715,
+      "step": 60397
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7307003736495972,
+      "learning_rate": 2.6366700339683913e-07,
+      "loss": 3.1203,
+      "step": 60398
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7308559417724609,
+      "learning_rate": 2.630220410237971e-07,
+      "loss": 2.9197,
+      "step": 60399
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7948404550552368,
+      "learning_rate": 2.6237786810026085e-07,
+      "loss": 2.9589,
+      "step": 60400
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7453910112380981,
+      "learning_rate": 2.61734484627929e-07,
+      "loss": 2.9592,
+      "step": 60401
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7211994528770447,
+      "learning_rate": 2.610918906085002e-07,
+      "loss": 3.1642,
+      "step": 60402
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7621414065361023,
+      "learning_rate": 2.604500860436731e-07,
+      "loss": 2.7984,
+      "step": 60403
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.722687840461731,
+      "learning_rate": 2.5980907093514635e-07,
+      "loss": 3.0943,
+      "step": 60404
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7349399924278259,
+      "learning_rate": 2.5916884528458523e-07,
+      "loss": 2.7707,
+      "step": 60405
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7808841466903687,
+      "learning_rate": 2.5852940909368846e-07,
+      "loss": 2.7073,
+      "step": 60406
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7772266268730164,
+      "learning_rate": 2.578907623641546e-07,
+      "loss": 2.9222,
+      "step": 60407
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7723431587219238,
+      "learning_rate": 2.572529050976491e-07,
+      "loss": 2.8101,
+      "step": 60408
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7670159935951233,
+      "learning_rate": 2.5661583729583713e-07,
+      "loss": 2.8798,
+      "step": 60409
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.780825138092041,
+      "learning_rate": 2.5597955896041744e-07,
+      "loss": 2.6376,
+      "step": 60410
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7525875568389893,
+      "learning_rate": 2.5534407009305536e-07,
+      "loss": 2.8498,
+      "step": 60411
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7023805975914001,
+      "learning_rate": 2.547093706954162e-07,
+      "loss": 2.9365,
+      "step": 60412
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7754830121994019,
+      "learning_rate": 2.540754607691986e-07,
+      "loss": 2.9444,
+      "step": 60413
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7659097909927368,
+      "learning_rate": 2.534423403160679e-07,
+      "loss": 2.9425,
+      "step": 60414
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8276565074920654,
+      "learning_rate": 2.5281000933768946e-07,
+      "loss": 2.9218,
+      "step": 60415
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.778061032295227,
+      "learning_rate": 2.521784678356953e-07,
+      "loss": 3.0158,
+      "step": 60416
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7523320317268372,
+      "learning_rate": 2.515477158118173e-07,
+      "loss": 2.7615,
+      "step": 60417
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8072141408920288,
+      "learning_rate": 2.50917753267621e-07,
+      "loss": 2.6866,
+      "step": 60418
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7441281676292419,
+      "learning_rate": 2.502885802048715e-07,
+      "loss": 2.6342,
+      "step": 60419
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7458040714263916,
+      "learning_rate": 2.4966019662513435e-07,
+      "loss": 2.9695,
+      "step": 60420
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7723721265792847,
+      "learning_rate": 2.490326025301082e-07,
+      "loss": 2.8109,
+      "step": 60421
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7250966429710388,
+      "learning_rate": 2.484057979214582e-07,
+      "loss": 3.1312,
+      "step": 60422
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7306153178215027,
+      "learning_rate": 2.4777978280081657e-07,
+      "loss": 2.9188,
+      "step": 60423
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7456044554710388,
+      "learning_rate": 2.4715455716984853e-07,
+      "loss": 2.8441,
+      "step": 60424
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7753781080245972,
+      "learning_rate": 2.4653012103015293e-07,
+      "loss": 2.8846,
+      "step": 60425
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.6862159371376038,
+      "learning_rate": 2.4590647438342823e-07,
+      "loss": 2.8749,
+      "step": 60426
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7013908624649048,
+      "learning_rate": 2.452836172313066e-07,
+      "loss": 2.9398,
+      "step": 60427
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7356927990913391,
+      "learning_rate": 2.4466154957538676e-07,
+      "loss": 2.9671,
+      "step": 60428
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7475108504295349,
+      "learning_rate": 2.4404027141736726e-07,
+      "loss": 2.9092,
+      "step": 60429
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7366794943809509,
+      "learning_rate": 2.4341978275884686e-07,
+      "loss": 2.8098,
+      "step": 60430
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7213911414146423,
+      "learning_rate": 2.428000836014576e-07,
+      "loss": 3.0121,
+      "step": 60431
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7313321232795715,
+      "learning_rate": 2.4218117394686487e-07,
+      "loss": 2.7566,
+      "step": 60432
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7520000338554382,
+      "learning_rate": 2.415630537966673e-07,
+      "loss": 2.8234,
+      "step": 60433
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7432150840759277,
+      "learning_rate": 2.4094572315253025e-07,
+      "loss": 3.0233,
+      "step": 60434
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7823925018310547,
+      "learning_rate": 2.403291820160191e-07,
+      "loss": 2.9009,
+      "step": 60435
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7720388770103455,
+      "learning_rate": 2.397134303887993e-07,
+      "loss": 2.8322,
+      "step": 60436
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7479499578475952,
+      "learning_rate": 2.390984682724695e-07,
+      "loss": 2.8643,
+      "step": 60437
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7411769032478333,
+      "learning_rate": 2.384842956686617e-07,
+      "loss": 2.8443,
+      "step": 60438
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7253438830375671,
+      "learning_rate": 2.3787091257904122e-07,
+      "loss": 3.2558,
+      "step": 60439
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7320099472999573,
+      "learning_rate": 2.3725831900514026e-07,
+      "loss": 2.7542,
+      "step": 60440
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7726000547409058,
+      "learning_rate": 2.3664651494862407e-07,
+      "loss": 3.1502,
+      "step": 60441
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7863835096359253,
+      "learning_rate": 2.3603550041109142e-07,
+      "loss": 2.9089,
+      "step": 60442
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7607067823410034,
+      "learning_rate": 2.3542527539414102e-07,
+      "loss": 2.8335,
+      "step": 60443
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7399455308914185,
+      "learning_rate": 2.348158398994049e-07,
+      "loss": 2.9403,
+      "step": 60444
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7538642883300781,
+      "learning_rate": 2.3420719392844844e-07,
+      "loss": 2.9633,
+      "step": 60445
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7989528775215149,
+      "learning_rate": 2.335993374829037e-07,
+      "loss": 2.8171,
+      "step": 60446
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7250673770904541,
+      "learning_rate": 2.3299227056436942e-07,
+      "loss": 2.9535,
+      "step": 60447
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7597386240959167,
+      "learning_rate": 2.323859931744776e-07,
+      "loss": 2.8269,
+      "step": 60448
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7532151937484741,
+      "learning_rate": 2.3178050531476034e-07,
+      "loss": 2.8806,
+      "step": 60449
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7555036544799805,
+      "learning_rate": 2.3117580698681638e-07,
+      "loss": 3.0434,
+      "step": 60450
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7302855253219604,
+      "learning_rate": 2.3057189819231103e-07,
+      "loss": 2.9344,
+      "step": 60451
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.715503454208374,
+      "learning_rate": 2.2996877893274313e-07,
+      "loss": 2.769,
+      "step": 60452
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.770855188369751,
+      "learning_rate": 2.2936644920977798e-07,
+      "loss": 2.8592,
+      "step": 60453
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7369658350944519,
+      "learning_rate": 2.2876490902498102e-07,
+      "loss": 2.8822,
+      "step": 60454
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7520059943199158,
+      "learning_rate": 2.2816415837988432e-07,
+      "loss": 2.9528,
+      "step": 60455
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7750919461250305,
+      "learning_rate": 2.2756419727611997e-07,
+      "loss": 2.8362,
+      "step": 60456
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7011710405349731,
+      "learning_rate": 2.2696502571528664e-07,
+      "loss": 2.9554,
+      "step": 60457
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7666123509407043,
+      "learning_rate": 2.2636664369891643e-07,
+      "loss": 2.6842,
+      "step": 60458
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7649023532867432,
+      "learning_rate": 2.2576905122860812e-07,
+      "loss": 2.9592,
+      "step": 60459
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7662659287452698,
+      "learning_rate": 2.2517224830592707e-07,
+      "loss": 3.0518,
+      "step": 60460
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7530821561813354,
+      "learning_rate": 2.245762349324387e-07,
+      "loss": 2.9532,
+      "step": 60461
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7605764865875244,
+      "learning_rate": 2.2398101110974176e-07,
+      "loss": 2.7051,
+      "step": 60462
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.760724663734436,
+      "learning_rate": 2.2338657683936834e-07,
+      "loss": 2.9107,
+      "step": 60463
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.744367241859436,
+      "learning_rate": 2.2279293212291715e-07,
+      "loss": 2.7573,
+      "step": 60464
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7295060157775879,
+      "learning_rate": 2.2220007696195364e-07,
+      "loss": 2.9798,
+      "step": 60465
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7500439286231995,
+      "learning_rate": 2.2160801135797658e-07,
+      "loss": 2.9385,
+      "step": 60466
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7618116140365601,
+      "learning_rate": 2.21016735312618e-07,
+      "loss": 3.0254,
+      "step": 60467
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7346891760826111,
+      "learning_rate": 2.2042624882741e-07,
+      "loss": 2.9205,
+      "step": 60468
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7819466590881348,
+      "learning_rate": 2.1983655190391804e-07,
+      "loss": 2.892,
+      "step": 60469
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7500108480453491,
+      "learning_rate": 2.1924764454364084e-07,
+      "loss": 2.7599,
+      "step": 60470
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7795835137367249,
+      "learning_rate": 2.1865952674821051e-07,
+      "loss": 2.8849,
+      "step": 60471
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7483556270599365,
+      "learning_rate": 2.1807219851912582e-07,
+      "loss": 2.7881,
+      "step": 60472
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7419742345809937,
+      "learning_rate": 2.1748565985791888e-07,
+      "loss": 3.0436,
+      "step": 60473
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7780374884605408,
+      "learning_rate": 2.168999107661884e-07,
+      "loss": 2.9787,
+      "step": 60474
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7674023509025574,
+      "learning_rate": 2.163149512454665e-07,
+      "loss": 3.093,
+      "step": 60475
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7375624775886536,
+      "learning_rate": 2.1573078129725196e-07,
+      "loss": 2.7765,
+      "step": 60476
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7869064211845398,
+      "learning_rate": 2.1514740092311023e-07,
+      "loss": 2.8789,
+      "step": 60477
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7125802040100098,
+      "learning_rate": 2.1456481012454006e-07,
+      "loss": 2.7467,
+      "step": 60478
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7873156666755676,
+      "learning_rate": 2.1398300890317354e-07,
+      "loss": 2.6907,
+      "step": 60479
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7791264653205872,
+      "learning_rate": 2.1340199726044282e-07,
+      "loss": 2.898,
+      "step": 60480
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7683501839637756,
+      "learning_rate": 2.1282177519791333e-07,
+      "loss": 2.8212,
+      "step": 60481
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7381917834281921,
+      "learning_rate": 2.1224234271715045e-07,
+      "loss": 2.8748,
+      "step": 60482
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7684051394462585,
+      "learning_rate": 2.1166369981961973e-07,
+      "loss": 2.9751,
+      "step": 60483
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7232508659362793,
+      "learning_rate": 2.1108584650688652e-07,
+      "loss": 2.8718,
+      "step": 60484
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7385919094085693,
+      "learning_rate": 2.1050878278041637e-07,
+      "loss": 3.0691,
+      "step": 60485
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7586545944213867,
+      "learning_rate": 2.09932508641808e-07,
+      "loss": 2.9754,
+      "step": 60486
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7664059996604919,
+      "learning_rate": 2.0935702409252686e-07,
+      "loss": 2.9511,
+      "step": 60487
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7628438472747803,
+      "learning_rate": 2.0878232913413839e-07,
+      "loss": 2.8217,
+      "step": 60488
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.765557587146759,
+      "learning_rate": 2.082084237680748e-07,
+      "loss": 2.8077,
+      "step": 60489
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7436291575431824,
+      "learning_rate": 2.0763530799593474e-07,
+      "loss": 2.8652,
+      "step": 60490
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7807187438011169,
+      "learning_rate": 2.070629818191505e-07,
+      "loss": 2.835,
+      "step": 60491
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7364583611488342,
+      "learning_rate": 2.064914452392874e-07,
+      "loss": 2.8279,
+      "step": 60492
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8078753352165222,
+      "learning_rate": 2.05920698257811e-07,
+      "loss": 2.6383,
+      "step": 60493
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7190173268318176,
+      "learning_rate": 2.0535074087625337e-07,
+      "loss": 2.8113,
+      "step": 60494
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7492725253105164,
+      "learning_rate": 2.047815730961133e-07,
+      "loss": 2.8811,
+      "step": 60495
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7870933413505554,
+      "learning_rate": 2.0421319491888966e-07,
+      "loss": 3.0474,
+      "step": 60496
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7799807786941528,
+      "learning_rate": 2.0364560634604787e-07,
+      "loss": 2.8235,
+      "step": 60497
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7293304800987244,
+      "learning_rate": 2.030788073791201e-07,
+      "loss": 2.9151,
+      "step": 60498
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7670882344245911,
+      "learning_rate": 2.0251279801957177e-07,
+      "loss": 2.7792,
+      "step": 60499
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7522982358932495,
+      "learning_rate": 2.0194757826893503e-07,
+      "loss": 2.9497,
+      "step": 60500
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.811562716960907,
+      "learning_rate": 2.013831481286421e-07,
+      "loss": 3.0968,
+      "step": 60501
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7589870095252991,
+      "learning_rate": 2.0081950760022502e-07,
+      "loss": 2.9288,
+      "step": 60502
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7747796177864075,
+      "learning_rate": 2.0025665668514935e-07,
+      "loss": 2.8051,
+      "step": 60503
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7064770460128784,
+      "learning_rate": 1.9969459538488052e-07,
+      "loss": 3.077,
+      "step": 60504
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.784367024898529,
+      "learning_rate": 1.991333237009507e-07,
+      "loss": 2.9097,
+      "step": 60505
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7294931411743164,
+      "learning_rate": 1.9857284163479203e-07,
+      "loss": 2.8172,
+      "step": 60506
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7817466259002686,
+      "learning_rate": 1.9801314918787003e-07,
+      "loss": 2.8352,
+      "step": 60507
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.790314257144928,
+      "learning_rate": 1.974542463617168e-07,
+      "loss": 2.8238,
+      "step": 60508
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.833737313747406,
+      "learning_rate": 1.9689613315776454e-07,
+      "loss": 2.7384,
+      "step": 60509
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7663082480430603,
+      "learning_rate": 1.963388095774787e-07,
+      "loss": 2.7838,
+      "step": 60510
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7980212569236755,
+      "learning_rate": 1.9578227562235816e-07,
+      "loss": 3.0168,
+      "step": 60511
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7827617526054382,
+      "learning_rate": 1.9522653129383504e-07,
+      "loss": 2.9978,
+      "step": 60512
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7780979871749878,
+      "learning_rate": 1.9467157659337486e-07,
+      "loss": 2.8574,
+      "step": 60513
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.716793954372406,
+      "learning_rate": 1.941174115224764e-07,
+      "loss": 2.8676,
+      "step": 60514
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7576918601989746,
+      "learning_rate": 1.9356403608257187e-07,
+      "loss": 2.9088,
+      "step": 60515
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.725763201713562,
+      "learning_rate": 1.9301145027509346e-07,
+      "loss": 2.8202,
+      "step": 60516
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7284558415412903,
+      "learning_rate": 1.9245965410153996e-07,
+      "loss": 2.9063,
+      "step": 60517
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.729720950126648,
+      "learning_rate": 1.9190864756334355e-07,
+      "loss": 2.6978,
+      "step": 60518
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7677045464515686,
+      "learning_rate": 1.9135843066196976e-07,
+      "loss": 2.9292,
+      "step": 60519
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7284916639328003,
+      "learning_rate": 1.9080900339885075e-07,
+      "loss": 3.099,
+      "step": 60520
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7096062898635864,
+      "learning_rate": 1.90260365775452e-07,
+      "loss": 2.7886,
+      "step": 60521
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7152775526046753,
+      "learning_rate": 1.8971251779320574e-07,
+      "loss": 2.8109,
+      "step": 60522
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7443001866340637,
+      "learning_rate": 1.8916545945354412e-07,
+      "loss": 2.9791,
+      "step": 60523
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7410553693771362,
+      "learning_rate": 1.8861919075793263e-07,
+      "loss": 2.8849,
+      "step": 60524
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6965992450714111,
+      "learning_rate": 1.880737117078035e-07,
+      "loss": 2.9143,
+      "step": 60525
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7935270667076111,
+      "learning_rate": 1.8752902230458888e-07,
+      "loss": 2.9024,
+      "step": 60526
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.725361168384552,
+      "learning_rate": 1.8698512254968766e-07,
+      "loss": 2.9605,
+      "step": 60527
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.745184063911438,
+      "learning_rate": 1.8644201244459866e-07,
+      "loss": 2.8531,
+      "step": 60528
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7021890878677368,
+      "learning_rate": 1.8589969199072074e-07,
+      "loss": 3.0086,
+      "step": 60529
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7811173796653748,
+      "learning_rate": 1.853581611894861e-07,
+      "loss": 2.9315,
+      "step": 60530
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7329429984092712,
+      "learning_rate": 1.848174200422936e-07,
+      "loss": 2.9401,
+      "step": 60531
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7532490491867065,
+      "learning_rate": 1.8427746855060877e-07,
+      "loss": 2.6845,
+      "step": 60532
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7217630743980408,
+      "learning_rate": 1.8373830671583045e-07,
+      "loss": 2.6904,
+      "step": 60533
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.772171676158905,
+      "learning_rate": 1.8319993453939087e-07,
+      "loss": 2.8626,
+      "step": 60534
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6981392502784729,
+      "learning_rate": 1.8266235202272217e-07,
+      "loss": 3.0572,
+      "step": 60535
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7536487579345703,
+      "learning_rate": 1.8212555916718997e-07,
+      "loss": 3.0188,
+      "step": 60536
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7514350414276123,
+      "learning_rate": 1.815895559742264e-07,
+      "loss": 2.8476,
+      "step": 60537
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7271486520767212,
+      "learning_rate": 1.8105434244529705e-07,
+      "loss": 2.8713,
+      "step": 60538
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8349723815917969,
+      "learning_rate": 1.805199185817341e-07,
+      "loss": 2.9083,
+      "step": 60539
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7469309568405151,
+      "learning_rate": 1.7998628438500306e-07,
+      "loss": 2.9315,
+      "step": 60540
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7527533769607544,
+      "learning_rate": 1.7945343985646955e-07,
+      "loss": 3.0218,
+      "step": 60541
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.713559091091156,
+      "learning_rate": 1.789213849975657e-07,
+      "loss": 2.7901,
+      "step": 60542
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7423522472381592,
+      "learning_rate": 1.7839011980965712e-07,
+      "loss": 2.95,
+      "step": 60543
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7502104640007019,
+      "learning_rate": 1.7785964429420928e-07,
+      "loss": 2.8922,
+      "step": 60544
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7708683609962463,
+      "learning_rate": 1.7732995845255448e-07,
+      "loss": 2.8545,
+      "step": 60545
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7669790983200073,
+      "learning_rate": 1.7680106228609158e-07,
+      "loss": 3.0175,
+      "step": 60546
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7533451318740845,
+      "learning_rate": 1.7627295579625277e-07,
+      "loss": 2.8942,
+      "step": 60547
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7395743131637573,
+      "learning_rate": 1.7574563898440364e-07,
+      "loss": 2.8993,
+      "step": 60548
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.777221143245697,
+      "learning_rate": 1.7521911185194305e-07,
+      "loss": 2.8378,
+      "step": 60549
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7704327702522278,
+      "learning_rate": 1.7469337440026986e-07,
+      "loss": 2.7663,
+      "step": 60550
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7657343149185181,
+      "learning_rate": 1.741684266307164e-07,
+      "loss": 2.8244,
+      "step": 60551
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7424664497375488,
+      "learning_rate": 1.736442685447148e-07,
+      "loss": 2.9918,
+      "step": 60552
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7621636390686035,
+      "learning_rate": 1.7312090014363068e-07,
+      "loss": 2.7005,
+      "step": 60553
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8045515418052673,
+      "learning_rate": 1.7259832142886286e-07,
+      "loss": 2.9842,
+      "step": 60554
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7256165742874146,
+      "learning_rate": 1.7207653240174369e-07,
+      "loss": 2.92,
+      "step": 60555
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7653000354766846,
+      "learning_rate": 1.7155553306367197e-07,
+      "loss": 2.6685,
+      "step": 60556
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7511091828346252,
+      "learning_rate": 1.7103532341604665e-07,
+      "loss": 2.8257,
+      "step": 60557
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7537457346916199,
+      "learning_rate": 1.7051590346019994e-07,
+      "loss": 2.954,
+      "step": 60558
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7920700311660767,
+      "learning_rate": 1.6999727319749745e-07,
+      "loss": 2.7778,
+      "step": 60559
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7353650331497192,
+      "learning_rate": 1.6947943262930475e-07,
+      "loss": 2.7668,
+      "step": 60560
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.68696129322052,
+      "learning_rate": 1.689623817570207e-07,
+      "loss": 2.8671,
+      "step": 60561
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7720268368721008,
+      "learning_rate": 1.684461205819776e-07,
+      "loss": 2.7461,
+      "step": 60562
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.787878692150116,
+      "learning_rate": 1.679306491055743e-07,
+      "loss": 2.6584,
+      "step": 60563
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7417284846305847,
+      "learning_rate": 1.674159673291098e-07,
+      "loss": 2.8816,
+      "step": 60564
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7512297034263611,
+      "learning_rate": 1.669020752539829e-07,
+      "loss": 2.8187,
+      "step": 60565
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7542157769203186,
+      "learning_rate": 1.6638897288152598e-07,
+      "loss": 2.9312,
+      "step": 60566
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7483489513397217,
+      "learning_rate": 1.6587666021307123e-07,
+      "loss": 2.934,
+      "step": 60567
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7624621987342834,
+      "learning_rate": 1.6536513725001754e-07,
+      "loss": 2.9115,
+      "step": 60568
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7325168251991272,
+      "learning_rate": 1.648544039936972e-07,
+      "loss": 2.9444,
+      "step": 60569
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.693103015422821,
+      "learning_rate": 1.6434446044540916e-07,
+      "loss": 2.6785,
+      "step": 60570
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7448264360427856,
+      "learning_rate": 1.638353066065856e-07,
+      "loss": 2.8162,
+      "step": 60571
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7847685217857361,
+      "learning_rate": 1.633269424784589e-07,
+      "loss": 2.9311,
+      "step": 60572
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7694564461708069,
+      "learning_rate": 1.628193680624612e-07,
+      "loss": 2.8487,
+      "step": 60573
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7881042957305908,
+      "learning_rate": 1.623125833598915e-07,
+      "loss": 2.7117,
+      "step": 60574
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7773247957229614,
+      "learning_rate": 1.6180658837208204e-07,
+      "loss": 3.0141,
+      "step": 60575
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7404429912567139,
+      "learning_rate": 1.613013831003651e-07,
+      "loss": 2.8635,
+      "step": 60576
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7321801781654358,
+      "learning_rate": 1.6079696754607298e-07,
+      "loss": 2.7727,
+      "step": 60577
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7246294617652893,
+      "learning_rate": 1.6029334171053786e-07,
+      "loss": 3.004,
+      "step": 60578
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7736529111862183,
+      "learning_rate": 1.597905055950921e-07,
+      "loss": 2.9635,
+      "step": 60579
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.746509850025177,
+      "learning_rate": 1.592884592010346e-07,
+      "loss": 2.8717,
+      "step": 60580
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7547891736030579,
+      "learning_rate": 1.587872025296977e-07,
+      "loss": 3.016,
+      "step": 60581
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7664916515350342,
+      "learning_rate": 1.5828673558244687e-07,
+      "loss": 2.9747,
+      "step": 60582
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7931352853775024,
+      "learning_rate": 1.5778705836054784e-07,
+      "loss": 2.8982,
+      "step": 60583
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7977129817008972,
+      "learning_rate": 1.5728817086533286e-07,
+      "loss": 3.0114,
+      "step": 60584
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7142460346221924,
+      "learning_rate": 1.5679007309810088e-07,
+      "loss": 2.9114,
+      "step": 60585
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7245585918426514,
+      "learning_rate": 1.5629276506018418e-07,
+      "loss": 2.7651,
+      "step": 60586
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.742992639541626,
+      "learning_rate": 1.557962467528817e-07,
+      "loss": 2.851,
+      "step": 60587
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7510650753974915,
+      "learning_rate": 1.5530051817752576e-07,
+      "loss": 2.9322,
+      "step": 60588
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7841531038284302,
+      "learning_rate": 1.5480557933538195e-07,
+      "loss": 3.0033,
+      "step": 60589
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7130355834960938,
+      "learning_rate": 1.5431143022778257e-07,
+      "loss": 2.8831,
+      "step": 60590
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7567395567893982,
+      "learning_rate": 1.538180708560266e-07,
+      "loss": 3.1319,
+      "step": 60591
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7231013178825378,
+      "learning_rate": 1.5332550122141295e-07,
+      "loss": 2.9103,
+      "step": 60592
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7480908036231995,
+      "learning_rate": 1.528337213252073e-07,
+      "loss": 2.7543,
+      "step": 60593
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7661453485488892,
+      "learning_rate": 1.5234273116874197e-07,
+      "loss": 2.8553,
+      "step": 60594
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7607757449150085,
+      "learning_rate": 1.5185253075331584e-07,
+      "loss": 3.1001,
+      "step": 60595
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7835237383842468,
+      "learning_rate": 1.5136312008019458e-07,
+      "loss": 2.6377,
+      "step": 60596
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7366989254951477,
+      "learning_rate": 1.5087449915064387e-07,
+      "loss": 2.9224,
+      "step": 60597
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7715787887573242,
+      "learning_rate": 1.5038666796602926e-07,
+      "loss": 2.8592,
+      "step": 60598
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7971885800361633,
+      "learning_rate": 1.4989962652754983e-07,
+      "loss": 2.7836,
+      "step": 60599
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.770579993724823,
+      "learning_rate": 1.4941337483657113e-07,
+      "loss": 2.8284,
+      "step": 60600
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7559013962745667,
+      "learning_rate": 1.489279128942922e-07,
+      "loss": 2.7782,
+      "step": 60601
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.737792432308197,
+      "learning_rate": 1.4844324070204528e-07,
+      "loss": 2.8605,
+      "step": 60602
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.726133406162262,
+      "learning_rate": 1.4795935826109606e-07,
+      "loss": 3.0003,
+      "step": 60603
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7365215420722961,
+      "learning_rate": 1.4747626557271018e-07,
+      "loss": 2.8262,
+      "step": 60604
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7618759274482727,
+      "learning_rate": 1.4699396263815332e-07,
+      "loss": 2.8041,
+      "step": 60605
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7196887731552124,
+      "learning_rate": 1.4651244945869113e-07,
+      "loss": 2.7637,
+      "step": 60606
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7778249382972717,
+      "learning_rate": 1.4603172603565583e-07,
+      "loss": 2.9692,
+      "step": 60607
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8094426989555359,
+      "learning_rate": 1.455517923702465e-07,
+      "loss": 2.815,
+      "step": 60608
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.706639289855957,
+      "learning_rate": 1.450726484637288e-07,
+      "loss": 2.6698,
+      "step": 60609
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6968713998794556,
+      "learning_rate": 1.4459429431736835e-07,
+      "loss": 3.1378,
+      "step": 60610
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7427403926849365,
+      "learning_rate": 1.4411672993246415e-07,
+      "loss": 2.8479,
+      "step": 60611
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7301240563392639,
+      "learning_rate": 1.4363995531024853e-07,
+      "loss": 2.8367,
+      "step": 60612
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7708475589752197,
+      "learning_rate": 1.4316397045198712e-07,
+      "loss": 2.9659,
+      "step": 60613
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7885755300521851,
+      "learning_rate": 1.426887753589123e-07,
+      "loss": 2.8787,
+      "step": 60614
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.73357093334198,
+      "learning_rate": 1.4221437003228974e-07,
+      "loss": 2.7029,
+      "step": 60615
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7780753374099731,
+      "learning_rate": 1.4174075447335177e-07,
+      "loss": 2.9124,
+      "step": 60616
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7349255084991455,
+      "learning_rate": 1.41267928683364e-07,
+      "loss": 2.872,
+      "step": 60617
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7401142120361328,
+      "learning_rate": 1.4079589266355885e-07,
+      "loss": 2.8839,
+      "step": 60618
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7998162508010864,
+      "learning_rate": 1.403246464152019e-07,
+      "loss": 2.874,
+      "step": 60619
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7657843232154846,
+      "learning_rate": 1.3985418993952558e-07,
+      "loss": 2.8038,
+      "step": 60620
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7320613861083984,
+      "learning_rate": 1.3938452323776217e-07,
+      "loss": 2.9106,
+      "step": 60621
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7517797350883484,
+      "learning_rate": 1.3891564631114404e-07,
+      "loss": 3.0539,
+      "step": 60622
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7234874963760376,
+      "learning_rate": 1.3844755916090355e-07,
+      "loss": 2.8527,
+      "step": 60623
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7384620904922485,
+      "learning_rate": 1.3798026178830634e-07,
+      "loss": 2.853,
+      "step": 60624
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7372459173202515,
+      "learning_rate": 1.3751375419455145e-07,
+      "loss": 2.9442,
+      "step": 60625
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7490679025650024,
+      "learning_rate": 1.3704803638087125e-07,
+      "loss": 2.808,
+      "step": 60626
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7829398512840271,
+      "learning_rate": 1.3658310834849807e-07,
+      "loss": 2.9018,
+      "step": 60627
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7897228002548218,
+      "learning_rate": 1.3611897009866425e-07,
+      "loss": 2.5938,
+      "step": 60628
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7101447582244873,
+      "learning_rate": 1.3565562163256882e-07,
+      "loss": 2.9465,
+      "step": 60629
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7695879340171814,
+      "learning_rate": 1.3519306295144416e-07,
+      "loss": 2.8979,
+      "step": 60630
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.752558708190918,
+      "learning_rate": 1.347312940565226e-07,
+      "loss": 3.0605,
+      "step": 60631
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7636677026748657,
+      "learning_rate": 1.342703149490032e-07,
+      "loss": 2.7709,
+      "step": 60632
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7717976570129395,
+      "learning_rate": 1.3381012563008497e-07,
+      "loss": 2.8797,
+      "step": 60633
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7932553291320801,
+      "learning_rate": 1.3335072610103359e-07,
+      "loss": 2.944,
+      "step": 60634
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7431058883666992,
+      "learning_rate": 1.3289211636301478e-07,
+      "loss": 2.8666,
+      "step": 60635
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7574412226676941,
+      "learning_rate": 1.324342964172276e-07,
+      "loss": 3.0864,
+      "step": 60636
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7627542018890381,
+      "learning_rate": 1.3197726626493764e-07,
+      "loss": 2.9098,
+      "step": 60637
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7745662331581116,
+      "learning_rate": 1.315210259072774e-07,
+      "loss": 2.9873,
+      "step": 60638
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7328659296035767,
+      "learning_rate": 1.3106557534547923e-07,
+      "loss": 2.9549,
+      "step": 60639
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.778195321559906,
+      "learning_rate": 1.3061091458077543e-07,
+      "loss": 2.9793,
+      "step": 60640
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7515281438827515,
+      "learning_rate": 1.3015704361429846e-07,
+      "loss": 2.9819,
+      "step": 60641
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7318151593208313,
+      "learning_rate": 1.2970396244728065e-07,
+      "loss": 2.7252,
+      "step": 60642
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7786916494369507,
+      "learning_rate": 1.2925167108092105e-07,
+      "loss": 2.9772,
+      "step": 60643
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7459295392036438,
+      "learning_rate": 1.288001695164187e-07,
+      "loss": 2.9008,
+      "step": 60644
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7117395401000977,
+      "learning_rate": 1.2834945775490601e-07,
+      "loss": 3.1014,
+      "step": 60645
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7417547702789307,
+      "learning_rate": 1.2789953579764866e-07,
+      "loss": 3.09,
+      "step": 60646
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8106539249420166,
+      "learning_rate": 1.2745040364574576e-07,
+      "loss": 3.0429,
+      "step": 60647
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7398137450218201,
+      "learning_rate": 1.2700206130046297e-07,
+      "loss": 3.0541,
+      "step": 60648
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7454041242599487,
+      "learning_rate": 1.265545087629327e-07,
+      "loss": 2.9233,
+      "step": 60649
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7723972201347351,
+      "learning_rate": 1.2610774603435403e-07,
+      "loss": 3.01,
+      "step": 60650
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7595714926719666,
+      "learning_rate": 1.2566177311589264e-07,
+      "loss": 3.0519,
+      "step": 60651
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7491887807846069,
+      "learning_rate": 1.2521659000871433e-07,
+      "loss": 2.7545,
+      "step": 60652
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7535812258720398,
+      "learning_rate": 1.2477219671401807e-07,
+      "loss": 2.9838,
+      "step": 60653
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7194299101829529,
+      "learning_rate": 1.2432859323296962e-07,
+      "loss": 2.8807,
+      "step": 60654
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7006655335426331,
+      "learning_rate": 1.2388577956670143e-07,
+      "loss": 2.7205,
+      "step": 60655
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7317060232162476,
+      "learning_rate": 1.2344375571641251e-07,
+      "loss": 2.8463,
+      "step": 60656
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.761465311050415,
+      "learning_rate": 1.2300252168326862e-07,
+      "loss": 2.9015,
+      "step": 60657
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7406482696533203,
+      "learning_rate": 1.225620774684355e-07,
+      "loss": 2.8137,
+      "step": 60658
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7317981123924255,
+      "learning_rate": 1.2212242307304554e-07,
+      "loss": 2.7945,
+      "step": 60659
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7996775507926941,
+      "learning_rate": 1.216835584982645e-07,
+      "loss": 2.9023,
+      "step": 60660
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7539355158805847,
+      "learning_rate": 1.212454837452581e-07,
+      "loss": 3.0568,
+      "step": 60661
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7302529215812683,
+      "learning_rate": 1.2080819881519211e-07,
+      "loss": 2.985,
+      "step": 60662
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6857694387435913,
+      "learning_rate": 1.203717037091989e-07,
+      "loss": 2.8932,
+      "step": 60663
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7334001660346985,
+      "learning_rate": 1.1993599842844426e-07,
+      "loss": 2.8093,
+      "step": 60664
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7142249345779419,
+      "learning_rate": 1.1950108297406058e-07,
+      "loss": 2.8508,
+      "step": 60665
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7907318472862244,
+      "learning_rate": 1.1906695734718031e-07,
+      "loss": 2.9828,
+      "step": 60666
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8050836324691772,
+      "learning_rate": 1.1863362154896916e-07,
+      "loss": 2.9504,
+      "step": 60667
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7420530915260315,
+      "learning_rate": 1.1820107558055958e-07,
+      "loss": 2.9651,
+      "step": 60668
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7614012360572815,
+      "learning_rate": 1.177693194431173e-07,
+      "loss": 2.9209,
+      "step": 60669
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7007195949554443,
+      "learning_rate": 1.1733835313774143e-07,
+      "loss": 2.9823,
+      "step": 60670
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7247536778450012,
+      "learning_rate": 1.169081766655977e-07,
+      "loss": 2.8296,
+      "step": 60671
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7323319315910339,
+      "learning_rate": 1.1647879002778526e-07,
+      "loss": 2.6532,
+      "step": 60672
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7220818996429443,
+      "learning_rate": 1.1605019322546982e-07,
+      "loss": 2.8848,
+      "step": 60673
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7621148824691772,
+      "learning_rate": 1.1562238625975051e-07,
+      "loss": 2.9724,
+      "step": 60674
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7443565130233765,
+      "learning_rate": 1.1519536913175976e-07,
+      "loss": 3.0197,
+      "step": 60675
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7332092523574829,
+      "learning_rate": 1.1476914184262997e-07,
+      "loss": 2.8808,
+      "step": 60676
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7590422630310059,
+      "learning_rate": 1.1434370439352691e-07,
+      "loss": 2.9004,
+      "step": 60677
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7833712697029114,
+      "learning_rate": 1.1391905678548308e-07,
+      "loss": 2.8703,
+      "step": 60678
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7459626197814941,
+      "learning_rate": 1.134951990196975e-07,
+      "loss": 2.9135,
+      "step": 60679
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7762752175331116,
+      "learning_rate": 1.130721310972027e-07,
+      "loss": 2.8371,
+      "step": 60680
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7452636361122131,
+      "learning_rate": 1.1264985301919771e-07,
+      "loss": 2.8966,
+      "step": 60681
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.766267716884613,
+      "learning_rate": 1.1222836478674834e-07,
+      "loss": 2.8336,
+      "step": 60682
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7369273900985718,
+      "learning_rate": 1.1180766640098704e-07,
+      "loss": 2.8865,
+      "step": 60683
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7170395255088806,
+      "learning_rate": 1.1138775786301291e-07,
+      "loss": 2.7637,
+      "step": 60684
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7494436502456665,
+      "learning_rate": 1.1096863917389175e-07,
+      "loss": 3.0208,
+      "step": 60685
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7833617329597473,
+      "learning_rate": 1.1055031033478934e-07,
+      "loss": 2.8597,
+      "step": 60686
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7761126160621643,
+      "learning_rate": 1.1013277134680476e-07,
+      "loss": 2.6651,
+      "step": 60687
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7350229024887085,
+      "learning_rate": 1.0971602221097054e-07,
+      "loss": 2.9107,
+      "step": 60688
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7333272695541382,
+      "learning_rate": 1.093000629284524e-07,
+      "loss": 2.8596,
+      "step": 60689
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8076415061950684,
+      "learning_rate": 1.0888489350034946e-07,
+      "loss": 2.7711,
+      "step": 60690
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7704684734344482,
+      "learning_rate": 1.0847051392769424e-07,
+      "loss": 2.7579,
+      "step": 60691
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7633429169654846,
+      "learning_rate": 1.0805692421161915e-07,
+      "loss": 2.75,
+      "step": 60692
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7828161716461182,
+      "learning_rate": 1.0764412435319004e-07,
+      "loss": 2.8519,
+      "step": 60693
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8346174955368042,
+      "learning_rate": 1.0723211435353929e-07,
+      "loss": 2.9347,
+      "step": 60694
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.783769965171814,
+      "learning_rate": 1.0682089421369944e-07,
+      "loss": 2.8745,
+      "step": 60695
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7348896861076355,
+      "learning_rate": 1.0641046393480291e-07,
+      "loss": 2.9594,
+      "step": 60696
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7758384346961975,
+      "learning_rate": 1.060008235178822e-07,
+      "loss": 2.971,
+      "step": 60697
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7444601058959961,
+      "learning_rate": 1.0559197296406974e-07,
+      "loss": 2.8435,
+      "step": 60698
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7654844522476196,
+      "learning_rate": 1.0518391227439804e-07,
+      "loss": 2.7562,
+      "step": 60699
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7547435760498047,
+      "learning_rate": 1.0477664144993292e-07,
+      "loss": 2.9196,
+      "step": 60700
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7515742778778076,
+      "learning_rate": 1.043701604918068e-07,
+      "loss": 2.9464,
+      "step": 60701
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7705299854278564,
+      "learning_rate": 1.0396446940101888e-07,
+      "loss": 2.8524,
+      "step": 60702
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7585739493370056,
+      "learning_rate": 1.0355956817870159e-07,
+      "loss": 3.0153,
+      "step": 60703
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7425493001937866,
+      "learning_rate": 1.0315545682585413e-07,
+      "loss": 2.9251,
+      "step": 60704
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7635918855667114,
+      "learning_rate": 1.0275213534360893e-07,
+      "loss": 2.8483,
+      "step": 60705
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8116132616996765,
+      "learning_rate": 1.0234960373299849e-07,
+      "loss": 2.6353,
+      "step": 60706
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7795093059539795,
+      "learning_rate": 1.0194786199508865e-07,
+      "loss": 2.9259,
+      "step": 60707
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7156280279159546,
+      "learning_rate": 1.0154691013091187e-07,
+      "loss": 2.8244,
+      "step": 60708
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7367256879806519,
+      "learning_rate": 1.0114674814156731e-07,
+      "loss": 2.8971,
+      "step": 60709
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.762520968914032,
+      "learning_rate": 1.0074737602805416e-07,
+      "loss": 2.7293,
+      "step": 60710
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7359768748283386,
+      "learning_rate": 1.0034879379147154e-07,
+      "loss": 2.8846,
+      "step": 60711
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7226070761680603,
+      "learning_rate": 9.995100143285195e-08,
+      "loss": 3.0858,
+      "step": 60712
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7261221408843994,
+      "learning_rate": 9.955399895326121e-08,
+      "loss": 2.9844,
+      "step": 60713
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7427242398262024,
+      "learning_rate": 9.915778635369854e-08,
+      "loss": 2.7948,
+      "step": 60714
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.771078884601593,
+      "learning_rate": 9.876236363526303e-08,
+      "loss": 2.7511,
+      "step": 60715
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7642545104026794,
+      "learning_rate": 9.836773079895389e-08,
+      "loss": 2.8683,
+      "step": 60716
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7525956034660339,
+      "learning_rate": 9.797388784583692e-08,
+      "loss": 2.9183,
+      "step": 60717
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7317299246788025,
+      "learning_rate": 9.758083477694468e-08,
+      "loss": 2.7959,
+      "step": 60718
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7813636064529419,
+      "learning_rate": 9.71885715932763e-08,
+      "loss": 2.9581,
+      "step": 60719
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7140172719955444,
+      "learning_rate": 9.679709829593096e-08,
+      "loss": 2.9738,
+      "step": 60720
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7551311254501343,
+      "learning_rate": 9.640641488587453e-08,
+      "loss": 2.7313,
+      "step": 60721
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7514950037002563,
+      "learning_rate": 9.601652136420612e-08,
+      "loss": 2.9285,
+      "step": 60722
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7382096648216248,
+      "learning_rate": 9.562741773185833e-08,
+      "loss": 3.1155,
+      "step": 60723
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7395986318588257,
+      "learning_rate": 9.523910398996359e-08,
+      "loss": 2.9011,
+      "step": 60724
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.742736279964447,
+      "learning_rate": 9.485158013945449e-08,
+      "loss": 2.9065,
+      "step": 60725
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7521002888679504,
+      "learning_rate": 9.446484618139682e-08,
+      "loss": 2.7803,
+      "step": 60726
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7703086733818054,
+      "learning_rate": 9.407890211682312e-08,
+      "loss": 2.8006,
+      "step": 60727
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7542996406555176,
+      "learning_rate": 9.369374794669926e-08,
+      "loss": 2.8211,
+      "step": 60728
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7727019190788269,
+      "learning_rate": 9.330938367209107e-08,
+      "loss": 2.8717,
+      "step": 60729
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7284082174301147,
+      "learning_rate": 9.292580929396443e-08,
+      "loss": 2.7599,
+      "step": 60730
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7392890453338623,
+      "learning_rate": 9.254302481338516e-08,
+      "loss": 2.9226,
+      "step": 60731
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7863142490386963,
+      "learning_rate": 9.216103023128584e-08,
+      "loss": 2.753,
+      "step": 60732
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7285134196281433,
+      "learning_rate": 9.177982554873231e-08,
+      "loss": 3.173,
+      "step": 60733
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7787915468215942,
+      "learning_rate": 9.139941076672374e-08,
+      "loss": 2.8017,
+      "step": 60734
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.77415531873703,
+      "learning_rate": 9.101978588625935e-08,
+      "loss": 2.6794,
+      "step": 60735
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7192097902297974,
+      "learning_rate": 9.064095090830504e-08,
+      "loss": 2.8903,
+      "step": 60736
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.755970299243927,
+      "learning_rate": 9.026290583389329e-08,
+      "loss": 3.218,
+      "step": 60737
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7275537848472595,
+      "learning_rate": 8.988565066402332e-08,
+      "loss": 2.9964,
+      "step": 60738
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8885321617126465,
+      "learning_rate": 8.950918539966102e-08,
+      "loss": 2.9317,
+      "step": 60739
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7572287321090698,
+      "learning_rate": 8.913351004180558e-08,
+      "loss": 3.1145,
+      "step": 60740
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7967362999916077,
+      "learning_rate": 8.875862459145622e-08,
+      "loss": 2.6688,
+      "step": 60741
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7556520700454712,
+      "learning_rate": 8.838452904961213e-08,
+      "loss": 2.9846,
+      "step": 60742
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.6918498873710632,
+      "learning_rate": 8.801122341723921e-08,
+      "loss": 2.8424,
+      "step": 60743
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7346342206001282,
+      "learning_rate": 8.763870769533666e-08,
+      "loss": 2.6878,
+      "step": 60744
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7678686380386353,
+      "learning_rate": 8.726698188483705e-08,
+      "loss": 2.9052,
+      "step": 60745
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7246375679969788,
+      "learning_rate": 8.689604598680621e-08,
+      "loss": 2.956,
+      "step": 60746
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7658019065856934,
+      "learning_rate": 8.652590000214343e-08,
+      "loss": 2.8864,
+      "step": 60747
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7121071219444275,
+      "learning_rate": 8.61565439318479e-08,
+      "loss": 2.8179,
+      "step": 60748
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.767242431640625,
+      "learning_rate": 8.578797777688551e-08,
+      "loss": 2.8497,
+      "step": 60749
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7547940611839294,
+      "learning_rate": 8.542020153828877e-08,
+      "loss": 2.9006,
+      "step": 60750
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7395439743995667,
+      "learning_rate": 8.505321521692366e-08,
+      "loss": 3.0071,
+      "step": 60751
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7863125801086426,
+      "learning_rate": 8.468701881382267e-08,
+      "loss": 3.1082,
+      "step": 60752
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7661166191101074,
+      "learning_rate": 8.432161232995172e-08,
+      "loss": 2.8509,
+      "step": 60753
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7914922833442688,
+      "learning_rate": 8.395699576624338e-08,
+      "loss": 2.7245,
+      "step": 60754
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7542526125907898,
+      "learning_rate": 8.359316912366355e-08,
+      "loss": 2.7699,
+      "step": 60755
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7157665491104126,
+      "learning_rate": 8.323013240321141e-08,
+      "loss": 2.9214,
+      "step": 60756
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7522284388542175,
+      "learning_rate": 8.286788560578628e-08,
+      "loss": 2.7608,
+      "step": 60757
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7462372779846191,
+      "learning_rate": 8.250642873235402e-08,
+      "loss": 2.8285,
+      "step": 60758
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7638521790504456,
+      "learning_rate": 8.214576178391386e-08,
+      "loss": 3.0214,
+      "step": 60759
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7532297968864441,
+      "learning_rate": 8.178588476136505e-08,
+      "loss": 2.8109,
+      "step": 60760
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7298749089241028,
+      "learning_rate": 8.14267976656735e-08,
+      "loss": 2.7234,
+      "step": 60761
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7774704694747925,
+      "learning_rate": 8.10685004977718e-08,
+      "loss": 2.7798,
+      "step": 60762
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.756490170955658,
+      "learning_rate": 8.071099325862584e-08,
+      "loss": 2.8804,
+      "step": 60763
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7454630732536316,
+      "learning_rate": 8.035427594916822e-08,
+      "loss": 2.839,
+      "step": 60764
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7813576459884644,
+      "learning_rate": 7.99983485703315e-08,
+      "loss": 2.8881,
+      "step": 60765
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7779887318611145,
+      "learning_rate": 7.964321112304828e-08,
+      "loss": 2.9005,
+      "step": 60766
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7656338214874268,
+      "learning_rate": 7.928886360828446e-08,
+      "loss": 3.1341,
+      "step": 60767
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7824438810348511,
+      "learning_rate": 7.893530602693931e-08,
+      "loss": 2.8625,
+      "step": 60768
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.779429018497467,
+      "learning_rate": 7.858253837997874e-08,
+      "loss": 2.8873,
+      "step": 60769
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7629287242889404,
+      "learning_rate": 7.823056066830202e-08,
+      "loss": 2.9589,
+      "step": 60770
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7588117718696594,
+      "learning_rate": 7.787937289284174e-08,
+      "loss": 2.9203,
+      "step": 60771
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7689938545227051,
+      "learning_rate": 7.752897505449718e-08,
+      "loss": 2.9673,
+      "step": 60772
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7638649344444275,
+      "learning_rate": 7.717936715426753e-08,
+      "loss": 2.893,
+      "step": 60773
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7803759574890137,
+      "learning_rate": 7.683054919298549e-08,
+      "loss": 2.8572,
+      "step": 60774
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7284832000732422,
+      "learning_rate": 7.648252117165021e-08,
+      "loss": 2.7997,
+      "step": 60775
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7143325209617615,
+      "learning_rate": 7.61352830910944e-08,
+      "loss": 2.8185,
+      "step": 60776
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8019963502883911,
+      "learning_rate": 7.578883495231725e-08,
+      "loss": 2.6413,
+      "step": 60777
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7633552551269531,
+      "learning_rate": 7.544317675615142e-08,
+      "loss": 2.8431,
+      "step": 60778
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7556493878364563,
+      "learning_rate": 7.50983085035961e-08,
+      "loss": 2.885,
+      "step": 60779
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7214013338088989,
+      "learning_rate": 7.475423019548399e-08,
+      "loss": 2.959,
+      "step": 60780
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7588320970535278,
+      "learning_rate": 7.441094183271435e-08,
+      "loss": 2.9404,
+      "step": 60781
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7851195335388184,
+      "learning_rate": 7.406844341628638e-08,
+      "loss": 2.9118,
+      "step": 60782
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7186158299446106,
+      "learning_rate": 7.372673494699943e-08,
+      "loss": 2.8767,
+      "step": 60783
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7348068952560425,
+      "learning_rate": 7.338581642578611e-08,
+      "loss": 2.8697,
+      "step": 60784
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7313429713249207,
+      "learning_rate": 7.304568785357901e-08,
+      "loss": 2.8737,
+      "step": 60785
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7902020812034607,
+      "learning_rate": 7.270634923124408e-08,
+      "loss": 2.9459,
+      "step": 60786
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7413740158081055,
+      "learning_rate": 7.236780055968061e-08,
+      "loss": 2.8579,
+      "step": 60787
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7744763493537903,
+      "learning_rate": 7.203004183975459e-08,
+      "loss": 2.8673,
+      "step": 60788
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7383904457092285,
+      "learning_rate": 7.169307307239857e-08,
+      "loss": 2.976,
+      "step": 60789
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7870457768440247,
+      "learning_rate": 7.135689425847857e-08,
+      "loss": 2.8683,
+      "step": 60790
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7270851135253906,
+      "learning_rate": 7.102150539889384e-08,
+      "loss": 2.9019,
+      "step": 60791
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7581509947776794,
+      "learning_rate": 7.068690649451036e-08,
+      "loss": 2.8886,
+      "step": 60792
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7382477521896362,
+      "learning_rate": 7.035309754619411e-08,
+      "loss": 2.9169,
+      "step": 60793
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7643795013427734,
+      "learning_rate": 7.002007855487768e-08,
+      "loss": 2.9914,
+      "step": 60794
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7508678436279297,
+      "learning_rate": 6.968784952136042e-08,
+      "loss": 2.7102,
+      "step": 60795
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7405849099159241,
+      "learning_rate": 6.935641044660823e-08,
+      "loss": 2.7978,
+      "step": 60796
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7330188155174255,
+      "learning_rate": 6.902576133142047e-08,
+      "loss": 2.8835,
+      "step": 60797
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7529171109199524,
+      "learning_rate": 6.86959021766964e-08,
+      "loss": 2.7461,
+      "step": 60798
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7461307048797607,
+      "learning_rate": 6.836683298333534e-08,
+      "loss": 3.0741,
+      "step": 60799
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.755508542060852,
+      "learning_rate": 6.803855375213663e-08,
+      "loss": 2.6922,
+      "step": 60800
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8063923120498657,
+      "learning_rate": 6.771106448399955e-08,
+      "loss": 2.7744,
+      "step": 60801
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7900208234786987,
+      "learning_rate": 6.738436517982338e-08,
+      "loss": 2.8385,
+      "step": 60802
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7695721387863159,
+      "learning_rate": 6.705845584040748e-08,
+      "loss": 2.9264,
+      "step": 60803
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7548896670341492,
+      "learning_rate": 6.673333646665113e-08,
+      "loss": 2.9377,
+      "step": 60804
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7459438443183899,
+      "learning_rate": 6.640900705938701e-08,
+      "loss": 2.9739,
+      "step": 60805
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7769594788551331,
+      "learning_rate": 6.608546761948108e-08,
+      "loss": 2.9764,
+      "step": 60806
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7430137395858765,
+      "learning_rate": 6.576271814776601e-08,
+      "loss": 2.8683,
+      "step": 60807
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7870651483535767,
+      "learning_rate": 6.544075864514109e-08,
+      "loss": 2.7308,
+      "step": 60808
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7871001958847046,
+      "learning_rate": 6.511958911240567e-08,
+      "loss": 2.9413,
+      "step": 60809
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7236528992652893,
+      "learning_rate": 6.479920955042573e-08,
+      "loss": 2.7938,
+      "step": 60810
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8005480170249939,
+      "learning_rate": 6.447961996003392e-08,
+      "loss": 2.7432,
+      "step": 60811
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7919958233833313,
+      "learning_rate": 6.416082034206294e-08,
+      "loss": 2.8455,
+      "step": 60812
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8858550190925598,
+      "learning_rate": 6.384281069737873e-08,
+      "loss": 2.9674,
+      "step": 60813
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7702856659889221,
+      "learning_rate": 6.352559102681399e-08,
+      "loss": 2.6873,
+      "step": 60814
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7305554747581482,
+      "learning_rate": 6.320916133120135e-08,
+      "loss": 2.8038,
+      "step": 60815
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.818423330783844,
+      "learning_rate": 6.289352161137351e-08,
+      "loss": 2.7824,
+      "step": 60816
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7536776661872864,
+      "learning_rate": 6.257867186816312e-08,
+      "loss": 2.9083,
+      "step": 60817
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7545386552810669,
+      "learning_rate": 6.226461210240286e-08,
+      "loss": 3.1492,
+      "step": 60818
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7835391163825989,
+      "learning_rate": 6.195134231489207e-08,
+      "loss": 3.1372,
+      "step": 60819
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7253636717796326,
+      "learning_rate": 6.163886250649675e-08,
+      "loss": 3.0669,
+      "step": 60820
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7546964287757874,
+      "learning_rate": 6.132717267801624e-08,
+      "loss": 2.9428,
+      "step": 60821
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7309494614601135,
+      "learning_rate": 6.10162728302499e-08,
+      "loss": 2.8188,
+      "step": 60822
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7510556578636169,
+      "learning_rate": 6.070616296406372e-08,
+      "loss": 2.7979,
+      "step": 60823
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7674130201339722,
+      "learning_rate": 6.039684308025705e-08,
+      "loss": 2.7607,
+      "step": 60824
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7727053761482239,
+      "learning_rate": 6.008831317962926e-08,
+      "loss": 2.949,
+      "step": 60825
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7896957993507385,
+      "learning_rate": 5.9780573263013e-08,
+      "loss": 2.7591,
+      "step": 60826
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7202666997909546,
+      "learning_rate": 5.947362333120764e-08,
+      "loss": 2.9216,
+      "step": 60827
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7396377921104431,
+      "learning_rate": 5.916746338501255e-08,
+      "loss": 2.9348,
+      "step": 60828
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8150199055671692,
+      "learning_rate": 5.886209342526038e-08,
+      "loss": 2.7997,
+      "step": 60829
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7385109066963196,
+      "learning_rate": 5.8557513452750506e-08,
+      "loss": 2.8084,
+      "step": 60830
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.739365816116333,
+      "learning_rate": 5.8253723468248966e-08,
+      "loss": 2.8535,
+      "step": 60831
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7645565271377563,
+      "learning_rate": 5.795072347258844e-08,
+      "loss": 2.9378,
+      "step": 60832
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7741032838821411,
+      "learning_rate": 5.764851346656829e-08,
+      "loss": 2.9119,
+      "step": 60833
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7111135125160217,
+      "learning_rate": 5.734709345098787e-08,
+      "loss": 2.8008,
+      "step": 60834
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7414789199829102,
+      "learning_rate": 5.704646342661323e-08,
+      "loss": 2.8277,
+      "step": 60835
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7525594830513,
+      "learning_rate": 5.674662339427705e-08,
+      "loss": 2.9346,
+      "step": 60836
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.704296886920929,
+      "learning_rate": 5.644757335471206e-08,
+      "loss": 2.8076,
+      "step": 60837
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.793901264667511,
+      "learning_rate": 5.614931330875094e-08,
+      "loss": 2.8382,
+      "step": 60838
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7604386210441589,
+      "learning_rate": 5.585184325715975e-08,
+      "loss": 2.8852,
+      "step": 60839
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7613590955734253,
+      "learning_rate": 5.555516320073783e-08,
+      "loss": 2.7332,
+      "step": 60840
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7492443323135376,
+      "learning_rate": 5.5259273140284554e-08,
+      "loss": 2.9632,
+      "step": 60841
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7389005422592163,
+      "learning_rate": 5.4964173076499364e-08,
+      "loss": 2.99,
+      "step": 60842
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7318647503852844,
+      "learning_rate": 5.466986301024823e-08,
+      "loss": 2.8089,
+      "step": 60843
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7469201683998108,
+      "learning_rate": 5.4376342942263895e-08,
+      "loss": 2.7664,
+      "step": 60844
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7827994227409363,
+      "learning_rate": 5.4083612873345726e-08,
+      "loss": 2.7729,
+      "step": 60845
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7702500224113464,
+      "learning_rate": 5.3791672804226474e-08,
+      "loss": 2.8159,
+      "step": 60846
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7574726343154907,
+      "learning_rate": 5.350052273567218e-08,
+      "loss": 2.8351,
+      "step": 60847
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7728380560874939,
+      "learning_rate": 5.321016266851552e-08,
+      "loss": 3.0487,
+      "step": 60848
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7415326833724976,
+      "learning_rate": 5.2920592603455934e-08,
+      "loss": 2.5731,
+      "step": 60849
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7318070530891418,
+      "learning_rate": 5.263181254125948e-08,
+      "loss": 2.7188,
+      "step": 60850
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8076432943344116,
+      "learning_rate": 5.2343822482725504e-08,
+      "loss": 2.7616,
+      "step": 60851
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7778842449188232,
+      "learning_rate": 5.205662242858677e-08,
+      "loss": 2.9999,
+      "step": 60852
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7371053695678711,
+      "learning_rate": 5.1770212379609324e-08,
+      "loss": 2.8389,
+      "step": 60853
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.732855498790741,
+      "learning_rate": 5.148459233652591e-08,
+      "loss": 2.8063,
+      "step": 60854
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7521831393241882,
+      "learning_rate": 5.119976230013589e-08,
+      "loss": 2.8495,
+      "step": 60855
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7384352087974548,
+      "learning_rate": 5.0915722271138715e-08,
+      "loss": 2.8919,
+      "step": 60856
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.762461245059967,
+      "learning_rate": 5.0632472250300424e-08,
+      "loss": 2.8156,
+      "step": 60857
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.774097740650177,
+      "learning_rate": 5.035001223838708e-08,
+      "loss": 2.8582,
+      "step": 60858
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.758370578289032,
+      "learning_rate": 5.0068342236098124e-08,
+      "loss": 2.7938,
+      "step": 60859
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7590804100036621,
+      "learning_rate": 4.9787462244232914e-08,
+      "loss": 2.8904,
+      "step": 60860
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7966086268424988,
+      "learning_rate": 4.950737226349089e-08,
+      "loss": 3.0966,
+      "step": 60861
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7693399786949158,
+      "learning_rate": 4.9228072294604795e-08,
+      "loss": 2.9406,
+      "step": 60862
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7839014530181885,
+      "learning_rate": 4.8949562338340684e-08,
+      "loss": 2.771,
+      "step": 60863
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7283673882484436,
+      "learning_rate": 4.867184239543131e-08,
+      "loss": 2.9914,
+      "step": 60864
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7859243750572205,
+      "learning_rate": 4.839491246657612e-08,
+      "loss": 3.0732,
+      "step": 60865
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7411079406738281,
+      "learning_rate": 4.811877255250785e-08,
+      "loss": 3.0077,
+      "step": 60866
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7194960117340088,
+      "learning_rate": 4.7843422653992546e-08,
+      "loss": 2.9794,
+      "step": 60867
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7908714413642883,
+      "learning_rate": 4.756886277169636e-08,
+      "loss": 2.8156,
+      "step": 60868
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7900540828704834,
+      "learning_rate": 4.729509290641864e-08,
+      "loss": 2.8043,
+      "step": 60869
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8169203400611877,
+      "learning_rate": 4.702211305882553e-08,
+      "loss": 2.8463,
+      "step": 60870
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7235054969787598,
+      "learning_rate": 4.674992322961646e-08,
+      "loss": 2.7804,
+      "step": 60871
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7315838932991028,
+      "learning_rate": 4.64785234195908e-08,
+      "loss": 2.9179,
+      "step": 60872
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7634678483009338,
+      "learning_rate": 4.620791362938137e-08,
+      "loss": 2.7995,
+      "step": 60873
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7515854835510254,
+      "learning_rate": 4.593809385975422e-08,
+      "loss": 2.7075,
+      "step": 60874
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.746206521987915,
+      "learning_rate": 4.56690641113755e-08,
+      "loss": 2.7542,
+      "step": 60875
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7026474475860596,
+      "learning_rate": 4.540082438497794e-08,
+      "loss": 2.9067,
+      "step": 60876
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7854546308517456,
+      "learning_rate": 4.51333746812943e-08,
+      "loss": 2.8615,
+      "step": 60877
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8138797879219055,
+      "learning_rate": 4.4866715000990703e-08,
+      "loss": 2.6686,
+      "step": 60878
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7269437313079834,
+      "learning_rate": 4.46008453447666e-08,
+      "loss": 2.9791,
+      "step": 60879
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7314205765724182,
+      "learning_rate": 4.4335765713321427e-08,
+      "loss": 3.0126,
+      "step": 60880
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7562000155448914,
+      "learning_rate": 4.407147610742123e-08,
+      "loss": 2.915,
+      "step": 60881
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7194030284881592,
+      "learning_rate": 4.3807976527665544e-08,
+      "loss": 2.7785,
+      "step": 60882
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7603053450584412,
+      "learning_rate": 4.3545266974820417e-08,
+      "loss": 2.7567,
+      "step": 60883
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7493316531181335,
+      "learning_rate": 4.328334744951867e-08,
+      "loss": 3.0762,
+      "step": 60884
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7748812437057495,
+      "learning_rate": 4.3022217952493055e-08,
+      "loss": 2.9854,
+      "step": 60885
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7610214352607727,
+      "learning_rate": 4.2761878484443014e-08,
+      "loss": 2.941,
+      "step": 60886
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7458277940750122,
+      "learning_rate": 4.250232904600137e-08,
+      "loss": 2.921,
+      "step": 60887
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7483231425285339,
+      "learning_rate": 4.224356963786757e-08,
+      "loss": 2.8299,
+      "step": 60888
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.738000214099884,
+      "learning_rate": 4.1985600260774354e-08,
+      "loss": 3.1012,
+      "step": 60889
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7348437309265137,
+      "learning_rate": 4.1728420915354555e-08,
+      "loss": 2.8678,
+      "step": 60890
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7560616135597229,
+      "learning_rate": 4.14720316022743e-08,
+      "loss": 3.0925,
+      "step": 60891
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7360788583755493,
+      "learning_rate": 4.121643232223304e-08,
+      "loss": 2.85,
+      "step": 60892
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7723987698554993,
+      "learning_rate": 4.0961623075930206e-08,
+      "loss": 2.9986,
+      "step": 60893
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7306909561157227,
+      "learning_rate": 4.070760386396532e-08,
+      "loss": 2.8415,
+      "step": 60894
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7794232368469238,
+      "learning_rate": 4.045437468707113e-08,
+      "loss": 2.8476,
+      "step": 60895
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7389131784439087,
+      "learning_rate": 4.020193554588047e-08,
+      "loss": 2.9256,
+      "step": 60896
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7388023734092712,
+      "learning_rate": 3.9950286441092773e-08,
+      "loss": 3.1202,
+      "step": 60897
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7390792369842529,
+      "learning_rate": 3.9699427373307554e-08,
+      "loss": 2.8996,
+      "step": 60898
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7606891989707947,
+      "learning_rate": 3.9449358343257574e-08,
+      "loss": 3.14,
+      "step": 60899
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7230342626571655,
+      "learning_rate": 3.920007935157565e-08,
+      "loss": 2.8227,
+      "step": 60900
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7700236439704895,
+      "learning_rate": 3.8951590398894614e-08,
+      "loss": 2.8222,
+      "step": 60901
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7606983780860901,
+      "learning_rate": 3.87038914858806e-08,
+      "loss": 2.849,
+      "step": 60902
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7816833853721619,
+      "learning_rate": 3.845698261319974e-08,
+      "loss": 2.6993,
+      "step": 60903
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.768175482749939,
+      "learning_rate": 3.8210863781484856e-08,
+      "loss": 2.8003,
+      "step": 60904
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7766599655151367,
+      "learning_rate": 3.796553499140209e-08,
+      "loss": 3.0448,
+      "step": 60905
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7720254063606262,
+      "learning_rate": 3.7720996243584265e-08,
+      "loss": 2.8401,
+      "step": 60906
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7150791883468628,
+      "learning_rate": 3.747724753869752e-08,
+      "loss": 2.7285,
+      "step": 60907
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7499209046363831,
+      "learning_rate": 3.723428887734137e-08,
+      "loss": 2.8719,
+      "step": 60908
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7711845636367798,
+      "learning_rate": 3.6992120260215254e-08,
+      "loss": 2.8742,
+      "step": 60909
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7651515007019043,
+      "learning_rate": 3.6750741687885385e-08,
+      "loss": 2.8158,
+      "step": 60910
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7973775267601013,
+      "learning_rate": 3.651015316105121e-08,
+      "loss": 3.1713,
+      "step": 60911
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8048893213272095,
+      "learning_rate": 3.627035468031225e-08,
+      "loss": 3.031,
+      "step": 60912
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7640774250030518,
+      "learning_rate": 3.603134624630133e-08,
+      "loss": 2.8489,
+      "step": 60913
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7189072370529175,
+      "learning_rate": 3.579312785965127e-08,
+      "loss": 2.9895,
+      "step": 60914
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7827427387237549,
+      "learning_rate": 3.5555699521028214e-08,
+      "loss": 3.075,
+      "step": 60915
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7645692229270935,
+      "learning_rate": 3.531906123099837e-08,
+      "loss": 2.8066,
+      "step": 60916
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7669183611869812,
+      "learning_rate": 3.5083212990194564e-08,
+      "loss": 2.787,
+      "step": 60917
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7498992085456848,
+      "learning_rate": 3.4848154799249627e-08,
+      "loss": 2.9104,
+      "step": 60918
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.761599600315094,
+      "learning_rate": 3.4613886658796383e-08,
+      "loss": 3.0023,
+      "step": 60919
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8176621794700623,
+      "learning_rate": 3.438040856946767e-08,
+      "loss": 2.9239,
+      "step": 60920
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7003955245018005,
+      "learning_rate": 3.414772053182968e-08,
+      "loss": 2.9555,
+      "step": 60921
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7190863490104675,
+      "learning_rate": 3.391582254651526e-08,
+      "loss": 2.682,
+      "step": 60922
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7530226707458496,
+      "learning_rate": 3.368471461412392e-08,
+      "loss": 2.7699,
+      "step": 60923
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7347545027732849,
+      "learning_rate": 3.345439673528849e-08,
+      "loss": 2.9471,
+      "step": 60924
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7659326791763306,
+      "learning_rate": 3.322486891060849e-08,
+      "loss": 2.7471,
+      "step": 60925
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7832285165786743,
+      "learning_rate": 3.299613114068345e-08,
+      "loss": 2.9932,
+      "step": 60926
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7939525842666626,
+      "learning_rate": 3.276818342611287e-08,
+      "loss": 2.9293,
+      "step": 60927
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7634779810905457,
+      "learning_rate": 3.254102576749629e-08,
+      "loss": 2.8537,
+      "step": 60928
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7426297664642334,
+      "learning_rate": 3.231465816546652e-08,
+      "loss": 2.8445,
+      "step": 60929
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7596210837364197,
+      "learning_rate": 3.2089080620556483e-08,
+      "loss": 2.7078,
+      "step": 60930
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.743432879447937,
+      "learning_rate": 3.1864293133399e-08,
+      "loss": 2.8809,
+      "step": 60931
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7715290784835815,
+      "learning_rate": 3.1640295704593586e-08,
+      "loss": 3.0125,
+      "step": 60932
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7802637815475464,
+      "learning_rate": 3.141708833470646e-08,
+      "loss": 2.9165,
+      "step": 60933
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8331025242805481,
+      "learning_rate": 3.119467102433715e-08,
+      "loss": 2.8806,
+      "step": 60934
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7259206771850586,
+      "learning_rate": 3.097304377408516e-08,
+      "loss": 2.8832,
+      "step": 60935
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7143881916999817,
+      "learning_rate": 3.0752206584516714e-08,
+      "loss": 2.9913,
+      "step": 60936
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7747646570205688,
+      "learning_rate": 3.053215945619802e-08,
+      "loss": 2.8695,
+      "step": 60937
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7718039155006409,
+      "learning_rate": 3.031290238972861e-08,
+      "loss": 3.0108,
+      "step": 60938
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7405833601951599,
+      "learning_rate": 3.0094435385707994e-08,
+      "loss": 2.7529,
+      "step": 60939
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7925773859024048,
+      "learning_rate": 2.987675844466908e-08,
+      "loss": 2.6726,
+      "step": 60940
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7442820072174072,
+      "learning_rate": 2.9659871567211392e-08,
+      "loss": 2.8788,
+      "step": 60941
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7715669870376587,
+      "learning_rate": 2.9443774753901138e-08,
+      "loss": 2.6613,
+      "step": 60942
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7934419512748718,
+      "learning_rate": 2.922846800530454e-08,
+      "loss": 2.6129,
+      "step": 60943
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7691748738288879,
+      "learning_rate": 2.9013951321987804e-08,
+      "loss": 2.9847,
+      "step": 60944
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8082806468009949,
+      "learning_rate": 2.880022470451715e-08,
+      "loss": 2.9681,
+      "step": 60945
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7494584918022156,
+      "learning_rate": 2.8587288153458788e-08,
+      "loss": 2.8827,
+      "step": 60946
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7465475797653198,
+      "learning_rate": 2.8375141669345624e-08,
+      "loss": 2.9779,
+      "step": 60947
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7119285464286804,
+      "learning_rate": 2.816378525281049e-08,
+      "loss": 2.8423,
+      "step": 60948
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7398651838302612,
+      "learning_rate": 2.7953218904319674e-08,
+      "loss": 2.6514,
+      "step": 60949
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7157944440841675,
+      "learning_rate": 2.7743442624506008e-08,
+      "loss": 2.8982,
+      "step": 60950
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7688831686973572,
+      "learning_rate": 2.753445641383578e-08,
+      "loss": 3.0429,
+      "step": 60951
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7500354647636414,
+      "learning_rate": 2.7326260272941825e-08,
+      "loss": 3.1836,
+      "step": 60952
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7020663619041443,
+      "learning_rate": 2.7118854202357044e-08,
+      "loss": 2.8623,
+      "step": 60953
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7152649164199829,
+      "learning_rate": 2.6912238202581038e-08,
+      "loss": 2.8368,
+      "step": 60954
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7284510731697083,
+      "learning_rate": 2.6706412274180022e-08,
+      "loss": 3.0018,
+      "step": 60955
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7411639094352722,
+      "learning_rate": 2.650137641772021e-08,
+      "loss": 2.9187,
+      "step": 60956
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.77473384141922,
+      "learning_rate": 2.629713063373451e-08,
+      "loss": 3.0106,
+      "step": 60957
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.729745626449585,
+      "learning_rate": 2.6093674922722518e-08,
+      "loss": 2.9179,
+      "step": 60958
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7914096117019653,
+      "learning_rate": 2.589100928528376e-08,
+      "loss": 2.9409,
+      "step": 60959
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7265556454658508,
+      "learning_rate": 2.5689133721884526e-08,
+      "loss": 2.8163,
+      "step": 60960
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7446621656417847,
+      "learning_rate": 2.5488048233124337e-08,
+      "loss": 2.7302,
+      "step": 60961
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7492318153381348,
+      "learning_rate": 2.5287752819469488e-08,
+      "loss": 2.8235,
+      "step": 60962
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8737034797668457,
+      "learning_rate": 2.508824748148619e-08,
+      "loss": 2.7881,
+      "step": 60963
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7302938103675842,
+      "learning_rate": 2.4889532219674047e-08,
+      "loss": 2.8977,
+      "step": 60964
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7328683137893677,
+      "learning_rate": 2.4691607034565963e-08,
+      "loss": 3.0039,
+      "step": 60965
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8015636801719666,
+      "learning_rate": 2.4494471926728155e-08,
+      "loss": 2.8645,
+      "step": 60966
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7779310941696167,
+      "learning_rate": 2.4298126896593607e-08,
+      "loss": 2.8827,
+      "step": 60967
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7854167222976685,
+      "learning_rate": 2.4102571944761838e-08,
+      "loss": 3.0293,
+      "step": 60968
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7807804346084595,
+      "learning_rate": 2.3907807071699148e-08,
+      "loss": 2.6994,
+      "step": 60969
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7262849807739258,
+      "learning_rate": 2.371383227793844e-08,
+      "loss": 3.0185,
+      "step": 60970
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7822308540344238,
+      "learning_rate": 2.352064756397931e-08,
+      "loss": 2.7,
+      "step": 60971
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7708417773246765,
+      "learning_rate": 2.3328252930321366e-08,
+      "loss": 3.0354,
+      "step": 60972
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7239246964454651,
+      "learning_rate": 2.3136648377497512e-08,
+      "loss": 2.9021,
+      "step": 60973
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7216629981994629,
+      "learning_rate": 2.294583390600735e-08,
+      "loss": 2.932,
+      "step": 60974
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7417978048324585,
+      "learning_rate": 2.2755809516350475e-08,
+      "loss": 3.0051,
+      "step": 60975
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7490457892417908,
+      "learning_rate": 2.2566575208993187e-08,
+      "loss": 2.5854,
+      "step": 60976
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7109230756759644,
+      "learning_rate": 2.2378130984501695e-08,
+      "loss": 2.6226,
+      "step": 60977
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7185120582580566,
+      "learning_rate": 2.2190476843308992e-08,
+      "loss": 2.9134,
+      "step": 60978
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7175775170326233,
+      "learning_rate": 2.2003612785947977e-08,
+      "loss": 2.7592,
+      "step": 60979
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7656946182250977,
+      "learning_rate": 2.181753881288495e-08,
+      "loss": 3.0205,
+      "step": 60980
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7306361198425293,
+      "learning_rate": 2.1632254924619507e-08,
+      "loss": 2.7981,
+      "step": 60981
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8056076169013977,
+      "learning_rate": 2.1447761121651254e-08,
+      "loss": 3.096,
+      "step": 60982
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7194265127182007,
+      "learning_rate": 2.1264057404446476e-08,
+      "loss": 2.9004,
+      "step": 60983
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7339944243431091,
+      "learning_rate": 2.108114377353809e-08,
+      "loss": 2.9615,
+      "step": 60984
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7202855348587036,
+      "learning_rate": 2.089902022932577e-08,
+      "loss": 3.0398,
+      "step": 60985
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7589322328567505,
+      "learning_rate": 2.0717686772342422e-08,
+      "loss": 3.0222,
+      "step": 60986
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7830872535705566,
+      "learning_rate": 2.0537143403087653e-08,
+      "loss": 2.7202,
+      "step": 60987
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7454860210418701,
+      "learning_rate": 2.0357390121961138e-08,
+      "loss": 2.89,
+      "step": 60988
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7104257941246033,
+      "learning_rate": 2.017842692952909e-08,
+      "loss": 2.9121,
+      "step": 60989
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7989444136619568,
+      "learning_rate": 2.0000253826157885e-08,
+      "loss": 2.8338,
+      "step": 60990
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7443826794624329,
+      "learning_rate": 1.9822870812413738e-08,
+      "loss": 2.7031,
+      "step": 60991
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7373210191726685,
+      "learning_rate": 1.9646277888729635e-08,
+      "loss": 2.858,
+      "step": 60992
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7563875317573547,
+      "learning_rate": 1.9470475055538558e-08,
+      "loss": 2.8906,
+      "step": 60993
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7365525960922241,
+      "learning_rate": 1.9295462313340116e-08,
+      "loss": 2.7371,
+      "step": 60994
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8209661841392517,
+      "learning_rate": 1.9121239662600595e-08,
+      "loss": 2.8336,
+      "step": 60995
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7196007370948792,
+      "learning_rate": 1.894780710371968e-08,
+      "loss": 3.0186,
+      "step": 60996
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7250151634216309,
+      "learning_rate": 1.8775164637230278e-08,
+      "loss": 2.6314,
+      "step": 60997
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7289470434188843,
+      "learning_rate": 1.8603312263565374e-08,
+      "loss": 3.0658,
+      "step": 60998
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7411735653877258,
+      "learning_rate": 1.8432249983157954e-08,
+      "loss": 2.9024,
+      "step": 60999
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7378327250480652,
+      "learning_rate": 1.8261977796441007e-08,
+      "loss": 2.8826,
+      "step": 61000
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8322823643684387,
+      "learning_rate": 1.8092495703914134e-08,
+      "loss": 2.8086,
+      "step": 61001
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7469660043716431,
+      "learning_rate": 1.792380370601032e-08,
+      "loss": 2.8937,
+      "step": 61002
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.6969466209411621,
+      "learning_rate": 1.7755901803129246e-08,
+      "loss": 2.7511,
+      "step": 61003
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7883122563362122,
+      "learning_rate": 1.7588789995770515e-08,
+      "loss": 2.7053,
+      "step": 61004
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7031104564666748,
+      "learning_rate": 1.7422468284333802e-08,
+      "loss": 2.7567,
+      "step": 61005
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7445712089538574,
+      "learning_rate": 1.7256936669285402e-08,
+      "loss": 2.9512,
+      "step": 61006
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7441263794898987,
+      "learning_rate": 1.709219515105831e-08,
+      "loss": 3.0192,
+      "step": 61007
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7398879528045654,
+      "learning_rate": 1.6928243730052194e-08,
+      "loss": 2.6434,
+      "step": 61008
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8084988594055176,
+      "learning_rate": 1.676508240676666e-08,
+      "loss": 2.6561,
+      "step": 61009
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7381051778793335,
+      "learning_rate": 1.6602711181534778e-08,
+      "loss": 3.1026,
+      "step": 61010
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.77137690782547,
+      "learning_rate": 1.6441130054889452e-08,
+      "loss": 2.9732,
+      "step": 61011
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7231003046035767,
+      "learning_rate": 1.6280339027163747e-08,
+      "loss": 2.6597,
+      "step": 61012
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.740614116191864,
+      "learning_rate": 1.612033809882396e-08,
+      "loss": 2.9157,
+      "step": 61013
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7222123146057129,
+      "learning_rate": 1.5961127270336382e-08,
+      "loss": 2.9297,
+      "step": 61014
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7570497989654541,
+      "learning_rate": 1.5802706542034083e-08,
+      "loss": 2.7988,
+      "step": 61015
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7413751482963562,
+      "learning_rate": 1.5645075914383352e-08,
+      "loss": 2.7553,
+      "step": 61016
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.715862512588501,
+      "learning_rate": 1.5488235387783877e-08,
+      "loss": 3.0391,
+      "step": 61017
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.782159686088562,
+      "learning_rate": 1.5332184962668638e-08,
+      "loss": 2.856,
+      "step": 61018
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7552375197410583,
+      "learning_rate": 1.517692463940401e-08,
+      "loss": 2.7686,
+      "step": 61019
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7790356874465942,
+      "learning_rate": 1.5022454418456285e-08,
+      "loss": 3.0508,
+      "step": 61020
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7251024842262268,
+      "learning_rate": 1.486877430022515e-08,
+      "loss": 2.9912,
+      "step": 61021
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7824087142944336,
+      "learning_rate": 1.4715884285076974e-08,
+      "loss": 2.7077,
+      "step": 61022
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7169434428215027,
+      "learning_rate": 1.4563784373411435e-08,
+      "loss": 2.813,
+      "step": 61023
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7796780467033386,
+      "learning_rate": 1.441247456569483e-08,
+      "loss": 2.8747,
+      "step": 61024
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.756627082824707,
+      "learning_rate": 1.4261954862260228e-08,
+      "loss": 2.8956,
+      "step": 61025
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7208350300788879,
+      "learning_rate": 1.4112225263507304e-08,
+      "loss": 2.7472,
+      "step": 61026
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7269884347915649,
+      "learning_rate": 1.3963285769869049e-08,
+      "loss": 2.8458,
+      "step": 61027
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7311272621154785,
+      "learning_rate": 1.3815136381711833e-08,
+      "loss": 2.8329,
+      "step": 61028
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7485470175743103,
+      "learning_rate": 1.366777709943534e-08,
+      "loss": 3.0685,
+      "step": 61029
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7784717679023743,
+      "learning_rate": 1.3521207923439247e-08,
+      "loss": 2.9271,
+      "step": 61030
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7870509624481201,
+      "learning_rate": 1.3375428854089931e-08,
+      "loss": 2.8426,
+      "step": 61031
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7719137072563171,
+      "learning_rate": 1.3230439891753763e-08,
+      "loss": 2.9635,
+      "step": 61032
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7624124884605408,
+      "learning_rate": 1.3086241036863732e-08,
+      "loss": 3.0412,
+      "step": 61033
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7434390187263489,
+      "learning_rate": 1.2942832289752903e-08,
+      "loss": 2.7478,
+      "step": 61034
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7709497213363647,
+      "learning_rate": 1.2800213650820957e-08,
+      "loss": 2.8803,
+      "step": 61035
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7721742987632751,
+      "learning_rate": 1.2658385120434266e-08,
+      "loss": 2.8181,
+      "step": 61036
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7293633222579956,
+      "learning_rate": 1.2517346698959207e-08,
+      "loss": 2.9901,
+      "step": 61037
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7672916650772095,
+      "learning_rate": 1.2377098386828766e-08,
+      "loss": 2.7232,
+      "step": 61038
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7433645725250244,
+      "learning_rate": 1.2237640184309394e-08,
+      "loss": 2.8997,
+      "step": 61039
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7337106466293335,
+      "learning_rate": 1.209897209183408e-08,
+      "loss": 2.7652,
+      "step": 61040
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.74492347240448,
+      "learning_rate": 1.1961094109769199e-08,
+      "loss": 3.0415,
+      "step": 61041
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.751065731048584,
+      "learning_rate": 1.1824006238481121e-08,
+      "loss": 2.6755,
+      "step": 61042
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7474420666694641,
+      "learning_rate": 1.1687708478302915e-08,
+      "loss": 2.7026,
+      "step": 61043
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7618042826652527,
+      "learning_rate": 1.1552200829600955e-08,
+      "loss": 2.8862,
+      "step": 61044
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7556697130203247,
+      "learning_rate": 1.1417483292708308e-08,
+      "loss": 2.9045,
+      "step": 61045
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7072124481201172,
+      "learning_rate": 1.1283555868057958e-08,
+      "loss": 2.8852,
+      "step": 61046
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7252053618431091,
+      "learning_rate": 1.1150418555916362e-08,
+      "loss": 2.6886,
+      "step": 61047
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7440884709358215,
+      "learning_rate": 1.1018071356683201e-08,
+      "loss": 2.9452,
+      "step": 61048
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7469719052314758,
+      "learning_rate": 1.0886514270691538e-08,
+      "loss": 3.1659,
+      "step": 61049
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.771772563457489,
+      "learning_rate": 1.0755747298307748e-08,
+      "loss": 2.9415,
+      "step": 61050
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7458013892173767,
+      "learning_rate": 1.0625770439831594e-08,
+      "loss": 2.8189,
+      "step": 61051
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.749146044254303,
+      "learning_rate": 1.0496583695629447e-08,
+      "loss": 2.8342,
+      "step": 61052
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7614862322807312,
+      "learning_rate": 1.0368187066067679e-08,
+      "loss": 2.9507,
+      "step": 61053
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7297611832618713,
+      "learning_rate": 1.0240580551479361e-08,
+      "loss": 2.8405,
+      "step": 61054
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.764214277267456,
+      "learning_rate": 1.0113764152130944e-08,
+      "loss": 2.7654,
+      "step": 61055
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7660698890686035,
+      "learning_rate": 9.987737868455415e-09,
+      "loss": 2.8231,
+      "step": 61056
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7836101055145264,
+      "learning_rate": 9.862501700719227e-09,
+      "loss": 2.6716,
+      "step": 61057
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7371546626091003,
+      "learning_rate": 9.73805564925545e-09,
+      "loss": 2.7436,
+      "step": 61058
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7891651391983032,
+      "learning_rate": 9.614399714430455e-09,
+      "loss": 2.951,
+      "step": 61059
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7586222290992737,
+      "learning_rate": 9.491533896544002e-09,
+      "loss": 2.8016,
+      "step": 61060
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7199023365974426,
+      "learning_rate": 9.369458195929158e-09,
+      "loss": 2.7525,
+      "step": 61061
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7390625476837158,
+      "learning_rate": 9.248172612885686e-09,
+      "loss": 3.1196,
+      "step": 61062
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7458911538124084,
+      "learning_rate": 9.127677147746649e-09,
+      "loss": 2.8515,
+      "step": 61063
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7433958053588867,
+      "learning_rate": 9.007971800845115e-09,
+      "loss": 2.6842,
+      "step": 61064
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7755656242370605,
+      "learning_rate": 8.889056572480847e-09,
+      "loss": 2.8729,
+      "step": 61065
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.730906069278717,
+      "learning_rate": 8.7709314629536e-09,
+      "loss": 2.811,
+      "step": 61066
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7480701804161072,
+      "learning_rate": 8.653596472596447e-09,
+      "loss": 2.8485,
+      "step": 61067
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7432823777198792,
+      "learning_rate": 8.537051601709144e-09,
+      "loss": 2.8094,
+      "step": 61068
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.796856701374054,
+      "learning_rate": 8.421296850591452e-09,
+      "loss": 2.7903,
+      "step": 61069
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7544580698013306,
+      "learning_rate": 8.306332219576439e-09,
+      "loss": 2.7556,
+      "step": 61070
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7667417526245117,
+      "learning_rate": 8.192157708930558e-09,
+      "loss": 2.8409,
+      "step": 61071
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7658982872962952,
+      "learning_rate": 8.078773318986875e-09,
+      "loss": 2.8028,
+      "step": 61072
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7310402989387512,
+      "learning_rate": 7.966179050045152e-09,
+      "loss": 2.904,
+      "step": 61073
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7373026609420776,
+      "learning_rate": 7.854374902338533e-09,
+      "loss": 3.0008,
+      "step": 61074
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.732184886932373,
+      "learning_rate": 7.743360876233395e-09,
+      "loss": 2.732,
+      "step": 61075
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7509750723838806,
+      "learning_rate": 7.633136972029497e-09,
+      "loss": 2.9609,
+      "step": 61076
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8037346601486206,
+      "learning_rate": 7.523703189926677e-09,
+      "loss": 2.8011,
+      "step": 61077
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7447194457054138,
+      "learning_rate": 7.415059530324619e-09,
+      "loss": 3.1165,
+      "step": 61078
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7197108268737793,
+      "learning_rate": 7.30720599342316e-09,
+      "loss": 2.8209,
+      "step": 61079
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7522173523902893,
+      "learning_rate": 7.200142579555368e-09,
+      "loss": 3.2279,
+      "step": 61080
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7585819363594055,
+      "learning_rate": 7.093869288987697e-09,
+      "loss": 2.8398,
+      "step": 61081
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.741621732711792,
+      "learning_rate": 6.9883861219866e-09,
+      "loss": 2.9234,
+      "step": 61082
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.758293092250824,
+      "learning_rate": 6.883693078851838e-09,
+      "loss": 2.9133,
+      "step": 61083
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7666619420051575,
+      "learning_rate": 6.779790159849863e-09,
+      "loss": 2.7775,
+      "step": 61084
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7205373644828796,
+      "learning_rate": 6.67667736524713e-09,
+      "loss": 3.0212,
+      "step": 61085
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7747965455055237,
+      "learning_rate": 6.574354695343398e-09,
+      "loss": 2.7993,
+      "step": 61086
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7606662511825562,
+      "learning_rate": 6.472822150371815e-09,
+      "loss": 2.7932,
+      "step": 61087
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7349268198013306,
+      "learning_rate": 6.372079730598833e-09,
+      "loss": 2.9177,
+      "step": 61088
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7886388897895813,
+      "learning_rate": 6.272127436324215e-09,
+      "loss": 2.8516,
+      "step": 61089
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.733219563961029,
+      "learning_rate": 6.172965267814411e-09,
+      "loss": 2.7334,
+      "step": 61090
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8257836699485779,
+      "learning_rate": 6.074593225302571e-09,
+      "loss": 2.939,
+      "step": 61091
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7202447056770325,
+      "learning_rate": 5.9770113090218396e-09,
+      "loss": 2.8128,
+      "step": 61092
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7605451941490173,
+      "learning_rate": 5.880219519305285e-09,
+      "loss": 2.9223,
+      "step": 61093
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7240636348724365,
+      "learning_rate": 5.784217856352746e-09,
+      "loss": 3.0476,
+      "step": 61094
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7533585429191589,
+      "learning_rate": 5.6890063203973715e-09,
+      "loss": 2.95,
+      "step": 61095
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.724474310874939,
+      "learning_rate": 5.594584911772227e-09,
+      "loss": 3.0003,
+      "step": 61096
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7521463632583618,
+      "learning_rate": 5.5009536306438455e-09,
+      "loss": 2.5749,
+      "step": 61097
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7722815275192261,
+      "learning_rate": 5.408112477278681e-09,
+      "loss": 2.865,
+      "step": 61098
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.776710033416748,
+      "learning_rate": 5.316061451976494e-09,
+      "loss": 2.9713,
+      "step": 61099
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7619010806083679,
+      "learning_rate": 5.2248005549038184e-09,
+      "loss": 3.0012,
+      "step": 61100
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7714282870292664,
+      "learning_rate": 5.134329786327107e-09,
+      "loss": 2.6743,
+      "step": 61101
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.707321286201477,
+      "learning_rate": 5.044649146512814e-09,
+      "loss": 2.9208,
+      "step": 61102
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7693123817443848,
+      "learning_rate": 4.955758635694085e-09,
+      "loss": 2.8076,
+      "step": 61103
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7465909123420715,
+      "learning_rate": 4.867658254037454e-09,
+      "loss": 2.8324,
+      "step": 61104
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.732196033000946,
+      "learning_rate": 4.780348001842682e-09,
+      "loss": 2.9903,
+      "step": 61105
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7602418065071106,
+      "learning_rate": 4.693827879342915e-09,
+      "loss": 2.9082,
+      "step": 61106
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7443466782569885,
+      "learning_rate": 4.608097886737994e-09,
+      "loss": 2.9654,
+      "step": 61107
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7329009175300598,
+      "learning_rate": 4.523158024227758e-09,
+      "loss": 2.9078,
+      "step": 61108
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8062663078308105,
+      "learning_rate": 4.439008292111967e-09,
+      "loss": 2.8382,
+      "step": 61109
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7804901003837585,
+      "learning_rate": 4.355648690557157e-09,
+      "loss": 2.9814,
+      "step": 61110
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7507225871086121,
+      "learning_rate": 4.273079219763165e-09,
+      "loss": 2.7954,
+      "step": 61111
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7602078914642334,
+      "learning_rate": 4.191299879996446e-09,
+      "loss": 2.7327,
+      "step": 61112
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7694182991981506,
+      "learning_rate": 4.110310671490147e-09,
+      "loss": 2.9445,
+      "step": 61113
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7346636652946472,
+      "learning_rate": 4.030111594377494e-09,
+      "loss": 2.5756,
+      "step": 61114
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7528190612792969,
+      "learning_rate": 3.950702648924942e-09,
+      "loss": 2.7316,
+      "step": 61115
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7616682052612305,
+      "learning_rate": 3.872083835332329e-09,
+      "loss": 2.8147,
+      "step": 61116
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7209632396697998,
+      "learning_rate": 3.794255153799497e-09,
+      "loss": 2.9357,
+      "step": 61117
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7706297039985657,
+      "learning_rate": 3.717216604559592e-09,
+      "loss": 2.9647,
+      "step": 61118
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7737950682640076,
+      "learning_rate": 3.6409681877791475e-09,
+      "loss": 2.6065,
+      "step": 61119
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7609835863113403,
+      "learning_rate": 3.5655099036580036e-09,
+      "loss": 2.9106,
+      "step": 61120
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.774637758731842,
+      "learning_rate": 3.4908417524293075e-09,
+      "loss": 2.862,
+      "step": 61121
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8165029883384705,
+      "learning_rate": 3.4169637342595922e-09,
+      "loss": 2.9376,
+      "step": 61122
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.817268967628479,
+      "learning_rate": 3.3438758493486984e-09,
+      "loss": 3.0128,
+      "step": 61123
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7474488615989685,
+      "learning_rate": 3.2715780979297723e-09,
+      "loss": 2.8821,
+      "step": 61124
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.72201007604599,
+      "learning_rate": 3.200070480136041e-09,
+      "loss": 3.1782,
+      "step": 61125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.762576699256897,
+      "learning_rate": 3.1293529961673445e-09,
+      "loss": 2.8246,
+      "step": 61126
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8136159777641296,
+      "learning_rate": 3.0594256462235234e-09,
+      "loss": 2.9168,
+      "step": 61127
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7495198249816895,
+      "learning_rate": 2.9902884304711106e-09,
+      "loss": 2.7297,
+      "step": 61128
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7615594267845154,
+      "learning_rate": 2.921941349143253e-09,
+      "loss": 2.8838,
+      "step": 61129
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7518624067306519,
+      "learning_rate": 2.854384402373178e-09,
+      "loss": 3.0472,
+      "step": 61130
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7348370552062988,
+      "learning_rate": 2.7876175903274177e-09,
+      "loss": 2.8619,
+      "step": 61131
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7087485790252686,
+      "learning_rate": 2.7216409132058136e-09,
+      "loss": 3.0459,
+      "step": 61132
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8339840769767761,
+      "learning_rate": 2.6564543712082053e-09,
+      "loss": 2.7792,
+      "step": 61133
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7480267882347107,
+      "learning_rate": 2.5920579644678195e-09,
+      "loss": 2.9,
+      "step": 61134
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7389785647392273,
+      "learning_rate": 2.52845169315119e-09,
+      "loss": 2.9488,
+      "step": 61135
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7411351799964905,
+      "learning_rate": 2.4656355574581565e-09,
+      "loss": 2.8339,
+      "step": 61136
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8139830231666565,
+      "learning_rate": 2.4036095575219462e-09,
+      "loss": 2.8145,
+      "step": 61137
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7364872097969055,
+      "learning_rate": 2.342373693509092e-09,
+      "loss": 2.7267,
+      "step": 61138
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7832111120223999,
+      "learning_rate": 2.281927965586128e-09,
+      "loss": 2.8519,
+      "step": 61139
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7762227058410645,
+      "learning_rate": 2.2222723739528935e-09,
+      "loss": 2.7967,
+      "step": 61140
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7120994329452515,
+      "learning_rate": 2.1634069187093096e-09,
+      "loss": 2.878,
+      "step": 61141
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7865113019943237,
+      "learning_rate": 2.105331600021909e-09,
+      "loss": 2.6527,
+      "step": 61142
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7547078728675842,
+      "learning_rate": 2.0480464180572252e-09,
+      "loss": 3.0927,
+      "step": 61143
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7572701573371887,
+      "learning_rate": 1.991551372981792e-09,
+      "loss": 2.9788,
+      "step": 61144
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7272855639457703,
+      "learning_rate": 1.9358464648955295e-09,
+      "loss": 2.9688,
+      "step": 61145
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8515251278877258,
+      "learning_rate": 1.8809316939982775e-09,
+      "loss": 2.8112,
+      "step": 61146
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7926790714263916,
+      "learning_rate": 1.8268070603899565e-09,
+      "loss": 2.7903,
+      "step": 61147
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.795349657535553,
+      "learning_rate": 1.7734725642370994e-09,
+      "loss": 2.7626,
+      "step": 61148
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7533791661262512,
+      "learning_rate": 1.7209282057062401e-09,
+      "loss": 2.8706,
+      "step": 61149
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7342902421951294,
+      "learning_rate": 1.6691739848972985e-09,
+      "loss": 2.7107,
+      "step": 61150
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7694773077964783,
+      "learning_rate": 1.6182099019435013e-09,
+      "loss": 2.9749,
+      "step": 61151
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7363452315330505,
+      "learning_rate": 1.5680359569780754e-09,
+      "loss": 2.7131,
+      "step": 61152
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6961015462875366,
+      "learning_rate": 1.5186521501675541e-09,
+      "loss": 2.7004,
+      "step": 61153
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7018082141876221,
+      "learning_rate": 1.4700584816118576e-09,
+      "loss": 2.9433,
+      "step": 61154
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6965725421905518,
+      "learning_rate": 1.4222549514442127e-09,
+      "loss": 3.0143,
+      "step": 61155
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7362546324729919,
+      "learning_rate": 1.3752415597978462e-09,
+      "loss": 2.652,
+      "step": 61156
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7468457818031311,
+      "learning_rate": 1.3290183068059845e-09,
+      "loss": 2.9477,
+      "step": 61157
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.764543890953064,
+      "learning_rate": 1.283585192568548e-09,
+      "loss": 2.9168,
+      "step": 61158
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7274712920188904,
+      "learning_rate": 1.2389422172187636e-09,
+      "loss": 2.7423,
+      "step": 61159
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8610571026802063,
+      "learning_rate": 1.1950893808565509e-09,
+      "loss": 2.8983,
+      "step": 61160
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7543455362319946,
+      "learning_rate": 1.1520266836151371e-09,
+      "loss": 3.0349,
+      "step": 61161
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7443333864212036,
+      "learning_rate": 1.1097541256277487e-09,
+      "loss": 2.9159,
+      "step": 61162
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7243948578834534,
+      "learning_rate": 1.0682717069609993e-09,
+      "loss": 2.8332,
+      "step": 61163
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7481058239936829,
+      "learning_rate": 1.0275794277481154e-09,
+      "loss": 3.0435,
+      "step": 61164
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7074810862541199,
+      "learning_rate": 9.876772880890171e-10,
+      "loss": 2.8396,
+      "step": 61165
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.743200421333313,
+      "learning_rate": 9.485652880836248e-10,
+      "loss": 2.8234,
+      "step": 61166
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7826438546180725,
+      "learning_rate": 9.102434278318582e-10,
+      "loss": 2.8359,
+      "step": 61167
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7499393820762634,
+      "learning_rate": 8.72711707500251e-10,
+      "loss": 2.7631,
+      "step": 61168
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7657136917114258,
+      "learning_rate": 8.359701270888031e-10,
+      "loss": 2.7475,
+      "step": 61169
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.72947758436203,
+      "learning_rate": 8.000186867307413e-10,
+      "loss": 2.9276,
+      "step": 61170
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7392745018005371,
+      "learning_rate": 7.648573865592921e-10,
+      "loss": 2.9282,
+      "step": 61171
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7172539830207825,
+      "learning_rate": 7.304862266410693e-10,
+      "loss": 2.8242,
+      "step": 61172
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7763723731040955,
+      "learning_rate": 6.969052070426862e-10,
+      "loss": 3.1106,
+      "step": 61173
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7424915432929993,
+      "learning_rate": 6.641143278973693e-10,
+      "loss": 3.0323,
+      "step": 61174
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7958371043205261,
+      "learning_rate": 6.321135892384254e-10,
+      "loss": 2.972,
+      "step": 61175
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7552845478057861,
+      "learning_rate": 6.009029911990815e-10,
+      "loss": 2.764,
+      "step": 61176
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7244768142700195,
+      "learning_rate": 5.70482533812644e-10,
+      "loss": 3.1299,
+      "step": 61177
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7536823153495789,
+      "learning_rate": 5.408522172123397e-10,
+      "loss": 2.8117,
+      "step": 61178
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7438461780548096,
+      "learning_rate": 5.120120414314754e-10,
+      "loss": 2.8251,
+      "step": 61179
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.767284095287323,
+      "learning_rate": 4.839620066032778e-10,
+      "loss": 2.6252,
+      "step": 61180
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.755458652973175,
+      "learning_rate": 4.5670211272774705e-10,
+      "loss": 2.9643,
+      "step": 61181
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7305696606636047,
+      "learning_rate": 4.3023235993810966e-10,
+      "loss": 3.0228,
+      "step": 61182
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7119986414909363,
+      "learning_rate": 4.0455274826767246e-10,
+      "loss": 2.8719,
+      "step": 61183
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7614222764968872,
+      "learning_rate": 3.7966327778304883e-10,
+      "loss": 2.9814,
+      "step": 61184
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8122992515563965,
+      "learning_rate": 3.555639485841588e-10,
+      "loss": 2.9432,
+      "step": 61185
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7403243780136108,
+      "learning_rate": 3.322547607043091e-10,
+      "loss": 2.8608,
+      "step": 61186
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7525769472122192,
+      "learning_rate": 3.0973571421011313e-10,
+      "loss": 2.9349,
+      "step": 61187
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7958530187606812,
+      "learning_rate": 2.880068091681842e-10,
+      "loss": 3.0496,
+      "step": 61188
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7971166968345642,
+      "learning_rate": 2.67068045611829e-10,
+      "loss": 3.0409,
+      "step": 61189
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7553383708000183,
+      "learning_rate": 2.469194236409677e-10,
+      "loss": 2.8486,
+      "step": 61190
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7261549830436707,
+      "learning_rate": 2.2756094325560025e-10,
+      "loss": 2.9742,
+      "step": 61191
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7534186840057373,
+      "learning_rate": 2.089926045556467e-10,
+      "loss": 2.5935,
+      "step": 61192
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7727741003036499,
+      "learning_rate": 1.9121440757441376e-10,
+      "loss": 2.9467,
+      "step": 61193
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7001724243164062,
+      "learning_rate": 1.742263523452081e-10,
+      "loss": 2.7758,
+      "step": 61194
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7367728352546692,
+      "learning_rate": 1.5802843893464312e-10,
+      "loss": 2.7828,
+      "step": 61195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7262899875640869,
+      "learning_rate": 1.4262066734271881e-10,
+      "loss": 2.8854,
+      "step": 61196
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7722108364105225,
+      "learning_rate": 1.2800303766935526e-10,
+      "loss": 2.9377,
+      "step": 61197
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.774972677230835,
+      "learning_rate": 1.1417554991455247e-10,
+      "loss": 2.8263,
+      "step": 61198
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7445690035820007,
+      "learning_rate": 1.011382041116171e-10,
+      "loss": 2.9051,
+      "step": 61199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7341079115867615,
+      "learning_rate": 8.889100032716257e-11,
+      "loss": 2.925,
+      "step": 61200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7602269649505615,
+      "learning_rate": 7.743393852788216e-11,
+      "loss": 2.9037,
+      "step": 61201
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7879922986030579,
+      "learning_rate": 6.676701884700264e-11,
+      "loss": 2.7574,
+      "step": 61202
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7422546148300171,
+      "learning_rate": 5.689024121791064e-11,
+      "loss": 2.8568,
+      "step": 61203
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7315881252288818,
+      "learning_rate": 4.7803605707219525e-11,
+      "loss": 2.8086,
+      "step": 61204
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7382916212081909,
+      "learning_rate": 3.9507112348236e-11,
+      "loss": 2.7754,
+      "step": 61205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8105321526527405,
+      "learning_rate": 3.2000761107653375e-11,
+      "loss": 2.7572,
+      "step": 61206
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7308658361434937,
+      "learning_rate": 2.5284552085391706e-11,
+      "loss": 2.9413,
+      "step": 61207
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8163118362426758,
+      "learning_rate": 1.9358485248144318e-11,
+      "loss": 2.8004,
+      "step": 61208
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7237694263458252,
+      "learning_rate": 1.4222560629217894e-11,
+      "loss": 2.8694,
+      "step": 61209
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7595729231834412,
+      "learning_rate": 9.876778261919127e-12,
+      "loss": 2.8303,
+      "step": 61210
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7996221780776978,
+      "learning_rate": 6.3211380796346356e-12,
+      "loss": 3.0761,
+      "step": 61211
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.793536901473999,
+      "learning_rate": 3.5556401822844914e-12,
+      "loss": 2.7739,
+      "step": 61212
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7907678484916687,
+      "learning_rate": 1.5802845365620042e-12,
+      "loss": 2.7948,
+      "step": 61213
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7259610891342163,
+      "learning_rate": 3.9507114246717374e-13,
+      "loss": 2.8744,
+      "step": 61214
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8341996669769287,
+      "learning_rate": 0.0,
+      "loss": 3.0732,
+      "step": 61215
+    },
+    {
+      "epoch": 3.0,
+      "step": 61215,
+      "total_flos": 1.1279892023790797e+17,
+      "train_loss": 3.1076881288265,
+      "train_runtime": 9220.7966,
+      "train_samples_per_second": 424.877,
+      "train_steps_per_second": 6.639
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 61215,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 1.1279892023790797e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}